Mitochondrial RNA Modifications in Pancreatic β-Cells: A Novel Axis in Early Diabetes Pathogenesis

Abstract

Mitochondrial RNA (mtRNA) modifications have emerged as critical regulators of pancreatic β-cell bioenergetics, influencing glucose-stimulated insulin secretion (GSIS) and the early pathogenesis of diabetes mellitus (DM). This review synthesizes current evidence on the diversity, mechanisms, and functional implications of mtRNA modifications—such as N6-methyladenosine (m6A), 5-methylcytosine (m5C), pseudouridine (Ψ), and 5-formylcytosine (f5C)—within β-cell mitochondria. These chemical marks, installed and recognized by specific writer, eraser, and reader proteins, regulate mitochondrial translation, oxidative phosphorylation (OXPHOS) complex assembly, and redox balance. Defects in mtRNA modification machinery, exemplified by β-cell-specific knockout of TFB1M, MRM2, or PUS1, impair ribosome biogenesis, disrupt ATP production, and precipitate insulin secretory failure, as demonstrated in human islets, rodent models, and monogenic diabetes syndromes. Advances in epitranscriptomic mapping technologies—including nanopore direct RNA sequencing, RNA immunoprecipitation (RIP)-seq, and mass spectrometry—have enabled high-resolution profiling of mtRNA modification landscapes under physiological and diabetic conditions, revealing their dynamic regulation in response to metabolic stress. Furthermore, mtRNA modifications interact with environmental stressors, such as oxidative damage and toxic metals, modulating β-cell vulnerability via pathways like the mitochondrial unfolded protein response (UPRmt). Therapeutically, modulation of RNA-modifying enzymes or restoration of specific chemical marks holds promise for preserving β-cell function, with potential applications in early diagnosis, risk stratification, and precision medicine approaches for DM. Despite substantial progress, critical gaps remain in understanding the interplay between mtRNA modifications, mitochondrial-nuclear crosstalk, and β-cell plasticity. Addressing these gaps will be pivotal for translating mtRNA biology into novel biomarkers and targeted interventions for early-stage diabetes.

1. Introduction

Diabetes mellitus is a major global health burden, with prevalence projected to exceed 600 million individuals by 2040, predominantly driven by type 2 diabetes (T2DM), which is characterized by insulin resistance and progressive β-cell dysfunction [1,2]. Pancreatic β-cells are uniquely dependent on mitochondrial function to sustain insulin secretion, and early mitochondrial impairment is a key feature of diabetes pathogenesis. Emerging evidence suggests that, beyond classical genetic and epigenetic mechanisms, post-transcriptional regulation through mitochondrial RNA (mtRNA) modifications, the so-called mito-epitranscriptome, plays a critical role in β-cell mitochondrial function and metabolic homeostasis. Mitochondrial oxidative phosphorylation (OXPHOS) is central to glucose-stimulated insulin secretion (GSIS), as ATP production links glucose metabolism to insulin granule exocytosis. Consequently, β-cells are highly sensitive to disruptions in mitochondrial gene expression and bioenergetics, making mitochondrial regulatory mechanisms particularly relevant in diabetes [3,4].

Recent advances have identified diverse chemical modifications on mitochondrial RNAs, including N6-methyladenosine (m6A), 5-methylcytosine (m5C), pseudouridine (Ψ), and 5-formylcytosine (f5C), which occur across mitochondrial tRNAs, rRNAs, and mRNAs [5,6]. These modifications are installed by specific enzymes (“writers”) and interpreted by “reader” proteins to regulate RNA stability, structure, and translation efficiency. Through these mechanisms, mtRNA modifications directly influence mitochondrial translation, respiratory chain assembly, and cellular responses to metabolic stress. Technological developments such as nanopore sequencing and high-resolution RNA mapping have further enabled site-specific characterization of these modifications, revealing their dynamic regulation in both physiological and disease states [7,8].

Importantly, mitochondrial RNA modifications have been implicated as both drivers and amplifiers of β-cell dysfunction. Genetic disruption of mtRNA-modifying enzymes (including TFB1M, MRM2, and PUS1) impairs mitochondrial translation, reduces OXPHOS efficiency, and compromises insulin secretion, supporting a causal role in β-cell failure [9,10,11,12,13]. Conversely, metabolic stressors associated with diabetes, such as hyperglycemia, lipotoxicity, and oxidative stress, can alter mtRNA modification patterns, suggesting that these epitranscriptomic changes may also arise secondary to cellular dysfunction [14]. Thus, mtRNA modifications likely participate in bidirectional and feed-forward mechanisms that exacerbate mitochondrial impairment during disease progression [15].

At the molecular level, mitochondrial transcripts harbor complex and highly coordinated modification landscapes. Each mitochondrial tRNA contains multiple chemical marks, forming distinct epitranscriptomic signatures that regulate RNA folding, stability, and translational fidelity [16,17,18,19]. Moreover, multiple modifications may coexist on the same RNA molecule, potentially generating combinatorial regulatory effects that fine-tune mitochondrial gene expression. However, current methodological limitations restrict the ability to resolve such multi-site interactions, and the functional interplay between different RNA modifications remains poorly understood [20].

Disruption of mtRNA modifications can compromise mitochondrial proteostasis, leading to defective synthesis of respiratory chain components, reduced ATP production, and increased reactive oxygen species (ROS) generation. These alterations are particularly detrimental in β-cells, which possess limited antioxidant defenses and rely on tightly regulated mitochondrial activity for insulin secretion. Consequently, impaired mtRNA modification has been linked to bioenergetic failure, oxidative stress, and β-cell apoptosis—hallmarks of both T1D and T2D pathophysiology [16,21,22].

Clinical and genetic evidence further supports the importance of mtRNA modifications in metabolic disease. Mutations in genes encoding RNA-modifying enzymes are associated with a spectrum of disorders ranging from insulinopenic diabetes to multisystem mitochondrial diseases, underscoring their essential and non-redundant roles in cellular homeostasis [16,17]. These findings position mtRNA modifications as potential biomarkers and therapeutic targets in diabetes, particularly in strategies aimed at restoring mitochondrial function.

Despite these advances, key questions remain unresolved, including whether mtRNA modification defects represent initiating events in β-cell dysfunction or secondary adaptations to metabolic stress, and how combinatorial modification patterns regulate mitochondrial gene expression. Addressing these challenges will require integrative approaches combining high-resolution epitranscriptomic mapping, functional genomics, and temporally controlled disease models. This review aims to synthesize current knowledge on mitochondrial RNA modifications in pancreatic β-cells, with a focus on their roles in regulating mitochondrial translation, bioenergetics, and stress responses, and to evaluate their contribution to the onset and progression of diabetes.

2. Overview of Mitochondrial RNA Biology in Pancreatic β-Cells

2.1. Mitochondrial Transcription and RNA Processing

Mitochondrial transcription and RNA processing are of primary importance for optimal mitochondrial function in pancreatic β-cells with respect to energy production and secretion of insulin (Figure 1) [23]. This section aligns with the overall study of mitochondrial bioenergetics and epitranscriptomic control.

2.1.1. Polycistronic Transcription of Human mtDNA by POLRMT

Polycistronic transcription of the mitochondrial genome by mitochondrial RNA polymerase (POLRMT) represents an evolutionary strategy to transcribe all mitochondrial-encoded RNAs (mRNAs, tRNAs, rRNAs) as a single long transcript to achieve streamlined expression of mitochondrial proteins [6]. In contrast to the large size of the nuclear genome, the human mitochondrial genome is very small, only comprising ~16.5 kilobases, encoding only 13 mRNAs for subunits of OXPHOS, 22 tRNAs, and 2 rRNAs. Because pancreatic β-cells are very metabolically active, they rely heavily on oxidative phosphorylation to support insulin secretion [24].

Additionally, the adaptability of the transcriptional process may allow pancreatic β-cells to rapidly adjust to variable cellular metabolism during glucose-stimulated insulin secretion, when demands on β-cell metabolism are particularly high. Synchronized transcription of all mitochondrial genes encoding OXPHOS subunits achieves sufficient ATP production for insulin granule exocytosis [19,25,26]. The efficiency of the mitochondrial transcription system is tied to its integrity. The activity of POLRMT is under tight regulation by interactions with nuclear-encoded transcription factors that may allow it to respond to metabolic conditions [13]. As an example, glucose signaling has been found to regulate POLRMT activity, which suggests that pancreatic β-cells have specific mechanisms to couple POLRMT activity with fluctuating energy status.

Recent evidence suggests that transcription may be regulated through chemical modifications of mitochondrial RNAs. Chemical marks such as m5C, m6A, and Ψ have been implicated in the regulation of RNA [18,27,28]. If mitochondrial RNA modifications represent inputs to control the rate or fidelity of POLRMT-driven transcription, this indicates that transcription is intricately controlled by a complex interplay of genetic, epitranscriptomic, and metabolic signaling in pancreatic β-cells. However, the interplay between these inputs remains unexplored. Following transcription by POLRMT, mitochondrial RNAs are processed in a process known as the “tRNA punctuation model”. The tRNAs within the polycistronic mitochondrial RNA transcript represent recognition points to trigger endonuclease cleavage. This liberates tRNAs, rRNAs, and mRNAs from the transcript into mature form, which are necessary for translation [19,29]. Therefore, in pancreatic β-cells, which are heavily reliant on mitochondrial respiration, efficient processing is of utmost importance for the translation of proteins involved in oxidative phosphorylation and ATP production for glucose-stimulated insulin secretion. Mitochondrial tRNAs are shorter in length and thus are less stable and more vulnerable to stress conditions than those encoded by the nuclear genome [19]. Thus, inefficient tRNA processing leads to a loss of functional mRNA templates to be translated, negatively impacting ATP synthesis and increasing the likelihood of β-cell failure, leading to impaired glucose homeostasis.

Processing activity seems to be dependent on chemical modification marks. Modifications such as m5C and m6A have been demonstrated to alter RNA secondary structure and alter binding affinities for RNA-binding proteins, which suggests that processing of mitochondrial RNAs may be directly affected by epitranscriptomic changes [16]. Consistent with the idea that the processes involved in mitochondrial RNA processing are vulnerable in pancreatic β-cells, the RNA-binding proteins, ribosomal assembly factors, and processing enzymes required for RNA integrity exhibit islet-specific expression. This expression pattern suggests that these proteins have tissue-specific roles and are critical for function. Thus, it may be that pancreatic β-cell mitochondria are sensitive to disruption of any one of these processes.

An intriguing possibility is that prediabetic conditions may disrupt the epitranscriptomic landscape in such a way as to dysregulate transcription or processing before the onset of diabetes. Perhaps misprocessed RNAs in the cell instigate an induction of the transcription/processing machinery, as compensation. This hypothesis suggests that in response to a specific dysregulation of expression, pancreatic β-cells attempt to compensate through mitochondrial or cellular quality control mechanisms, and these mechanisms may fail over time, leading to impaired function. These experiments have not yet been executed in pancreatic β-cells. Functional investigations into transcription and processing can be combined, for example, in metabolomic assays or cell respiration tests, such as Seahorse [16,19].

The combined results of transcription and processing, coordinated with chemical modifications, all lend support to the fact that pancreatic β-cell mitochondrial function is exquisitely regulated by multiple facets. However, the relationship between all of these inputs is poorly understood. In addition, if perturbations to this system result in cellular dysfunction, strategies to re-establish proper function have not been studied and require further research. The interplay between them has the potential to be more complex than simple cause and effect. Rather, perhaps each system influences the others, providing feedback mechanisms to finely control the availability of essential proteins for mitochondrial activity. It may be that the proper activity of all three systems ensures the proper abundance of each product, so that if even one system is disrupted by mutation or in response to external signals, this triggers compensation or feedback control. If the feedback loop is disrupted over a long period of time, this disruption may result in β-cell impairment, prediabetes, and ultimately type 2 diabetes.

2.1.2. RNA Cleavage into rRNAs, tRNAs, mRNAs

Mitochondrial RNA cleavage is a crucial process for the production of mature rRNAs, tRNAs, and mRNAs in pancreatic β-cells, which are required for proper mitochondrial translation and insulin secretion [30]. Mitochondrial tRNA cleavage is critical for mitochondrial protein synthesis and is extremely susceptible to errors. Improper mitochondrial tRNA cleavage is common in β-cells that contain mutations in the tRNA sequences or have impaired cleavage enzymes, which leads to improperly processed RNAs that are not incorporated into the ribosomes. tRNA-mediated cleavage is essential for mitochondrial gene expression and is highly sensitive to even minor mutations, as improper cleavage can lead to the impairment of mitochondrial translation, ATP production, and insulin secretion [31].

Mutations in tRNA genes or abnormal mitochondrial processing enzymes can interrupt proper mitochondrial RNA processing and cause incomplete maturation of mitochondrial RNA precursors. Accumulation of unprocessed transcripts and a reduction in mature tRNAs have been reported with mutations in mitochondrial tRNA genes, which disrupt tRNA processing [18]. It has been demonstrated that point mutations in tRNAs are responsible for this effect, as they disrupt tRNA cleavage [18]. Since mutations in tRNA genes directly disrupt tRNA processing, they have been reported to impair mitochondrial translation, thereby reducing ATP production and insulin secretion [18]. In addition, improper rRNA maturation has been reported as a result of a deficiency in MRM2 [21]. Correction of mutations in tRNA genes restores impaired mitochondrial function and prevents diabetes, further confirming that tRNA processing, performed by mitochondrial processing enzymes, is essential for maintaining proper function of mitochondrial RNA and the proper production of proteins from the ribosome [18].

Mitochondrial tRNAs range from 59 to 75 nucleotides in length, whereas their nuclear-encoded homologs usually exceed 75 nucleotides in length, indicating that mitochondrial tRNAs are significantly shorter. In addition to being significantly shorter, the mitochondrial tRNAs contain several structural features that do not resemble the structure of nuclear-encoded tRNAs. The major structural variations in mitochondrial tRNAs are associated with their stability in the mitochondrial environment. These modifications are critical for stability, proper tRNA processing, proper protein synthesis by the ribosome, and the assembly of mature mitochondrial tRNAs and rRNAs [19,32]. If the modifications in tRNAs are not made, processing efficiency is impaired, reducing mitochondrial translational output, thereby preventing adequate levels of ATP in the cell and eventually damaging β-cells. Evidence has demonstrated that even mild reductions in mitochondrial ATP production will impair insulin secretion in β-cells [18]. Therefore, the process of tRNA cleavage, which is critical for mRNA translation in the mitochondria, must be maintained properly to ensure efficient production of energy to sustain the health of β-cells. Metabolic disturbances that negatively impact tRNA processing will reduce protein output and eventually negatively impact energy production in the mitochondria, damaging β-cells and leading to the onset of diabetes [33].

Inactivation of TFB1M in murine β-cells results in the degradation of 12S rRNA, thereby making ribosomal subunits unstable and nonfunctional. Reduced mitochondrial translation due to destabilized ribosomal subunits has been reported to impair insulin secretion [21,34]. With both the basal and glucose-stimulated mitochondrial responses impaired in β-cells with reduced expression of MRM2, these results indicated that the enzymes involved in mitochondrial RNA processing are responsible for proper mitochondrial function [10].

It has been demonstrated that post-transcriptional modifications influence the structure of mRNA by causing physical changes to mRNA molecules [16]. Many of the chemical marks on RNA, such as m6A, m5C, and Ψ, are critical for the functional activity of mature RNAs, but these modifications can influence the structure and stability of the mRNA in the mitochondrion. These chemical marks may affect mRNA cleavage. Accumulation of unprocessed RNA precursors was identified in β-cells deficient in many of these RNA-modifying enzymes, further indicating the possibility that a defect in RNA modifications leads to improper RNA processing in the mitochondria, thereby contributing to dysregulation of cellular energy production [18,21]. Alterations in RNA modifications caused by oxidative stress or nutrient overload can potentially affect RNA-processing dynamics, leading to early β-cell dysfunction [16]. The role of RNA modification and processing is an expanding area of investigation in diabetes research.

2.1.3. Key RNA-Binding Factors and Islet-Specific Expression

RNA-binding proteins are found to have distinct expression patterns in pancreatic β-cells and are of utmost importance in influencing mitochondrial RNA stability and post-transcriptional processes required for adequate insulin secretion [35]. The specialized RNA-binding proteins in β-cells tend to have high levels or unique patterns of expression compared to the RNA-binding proteins present in other cells. RNA-binding proteins are primarily responsible for the stabilization of RNAs, and in the absence of such proteins, there are elevated risks of destabilization, thus impairing mitochondrial translation and bioenergetics. Proteins that bind to mitochondrial tRNA and rRNA play roles in their folding and maturation, crucial steps for the synthesis of OXPHOS proteins [19,36]. Furthermore, due to the essential role of OXPHOS in the generation of ATP that ultimately leads to glucose-stimulated insulin secretion (GSIS), disruption of these RNA-binding protein interactions would result in a compromised mitochondrial function and therefore β-cell failure in the presence of metabolic stress. Characterization of the mechanisms responsible for regulating the selective expression of certain RNA-binding proteins in β-cells is key for understanding the factors that make these proteins prone to dysfunction in diabetes [35].

The β-cell-specific expression of RNA-binding proteins may have a selective advantage by allowing for rapid control of the activity of mitochondria depending on the levels of glucose within the cells. Even so, this characteristic may also pose a liability because even small changes in the expression and/or activity of such RNA-binding proteins can disrupt β-cell mitochondrial function to a greater extent than in other cell types. Recently, data from studies that combine islet transcriptomics and proteomics indicate that this may be a valuable approach for identifying vital RNA-binding proteins that are responsible for various β-cell mitochondrial activities and that these proteins could become a novel target for diabetes [16].

RNA-binding enzymes are also involved in RNA modification, a vital process for translation accuracy and proper mitochondrial bioenergetics. PUS1 stabilizes rRNA and tRNA by catalyzing the formation of Ψ in mitochondrial RNAs, ultimately enhancing the proper function of the ribosome, translation, and ATP production. Ablation of this enzyme causes instability of mitochondrial rRNA and tRNA, as well as a reduction in ETC function, leading to lower ATP production and oxidative stress [37]. Another enzyme involved in mitochondrial RNA modification is TRMT61B, a methyltransferase that catalyzes N¹-methyladenosine (m¹A) formation in mitochondrial RNAs, including both mitochondrial rRNA and certain mitochondrial tRNAs. These findings further highlight the coordinated roles of RNA-binding and RNA-modifying proteins in maintaining mitochondrial RNA integrity and translational capacity [18,19,38].

RNA modifications may also be responsible for alterations in RNA-binding affinity for proteins within β-cell mitochondria. RNA modifications such as m6A, m5C, and Ψ affect the structure and electrostatic interactions of mitochondrial RNA. When modifications affect the structure of RNA, it causes either an enhanced or decreased binding affinity to RNA-binding proteins. The effects can depend on the position of the modified nucleotide(s), their chemical features, or the expression of RNA-binding proteins [39]. This process plays an important part in determining the fate of the RNA as well as the ability of β-cells to regulate the function of mitochondrial output dependent on the concentration of glucose. In this process, RNA-binding proteins may also preferentially bind with RNA that is in a modified state, and therefore, the sites in modified mitochondrial RNAs act in a context-dependent manner to influence translation rates [40]. Considering the interaction of modifications on the regulation of translation, this could mean that RNA modifications can exert effects directly in ribosome engagement with RNA substrates. Disruptions in the RNA modification process caused by certain environmental cues or by genetic mutations, together with the imbalance in RNA-binding protein expression, are capable of indirectly altering the binding preferences, which can result in different and unpredictable effects on the process of translation [16].

RNA editing also acts as a post-transcriptional regulation to affect the function of mitochondrial RNA in β-cells. IFNα and IL-1β, inflammatory cytokines involved in the pathogenesis of T1D, induce in human β-cells the expression of ADAR1, the enzyme that catalyzes A-to-I editing in RNAs, suggesting that a stress-induced transcriptome remodeling may result as a consequence of ADAR1 activation [41]. The ectopic overexpression or silencing of ADAR1 expression in β-cells was able to significantly alter the splicing and expression of some of the most common transcripts in these cells. The loss or induction of ADAR1 also had effects on RNA stability and mitochondrial function. While some of the edits introduced by this RNA-modifying process are used for β-cells to deal with acute challenges, other edits are not suitable and impair cellular health. The accumulation of unfavorable RNA editing in β-cells can be induced in conditions of chronic inflammation, such as that observed in T1D or T2D, and this could compromise β-cell fitness, which leads to cellular death. RNA editing, and enzymes involved in this process, may represent new targets for personalized interventions in β-cell health in response to both metabolic and immune stressors [41,42].

2.2. The Mitochondrial Epitranscriptome

Mitochondrial RNA modifications are an important regulatory mechanism governing the bioenergetic capacity and function of pancreatic β-cells (Figure 2). The exploration of chemical marks, enzymes, and advanced mapping uncovers how these epitranscriptomic dynamics influence mitochondrial translation, redox homeostasis, and, in turn, insulin secretion. This knowledge connects to our overall picture of the mitochondrial contribution to the development of diabetes, suggesting new possibilities for early diagnosis and therapies.

2.2.1. Chemical Marks in Mitochondrial RNA

Mitochondrial RNA in pancreatic β-cells is widely modified by a vast number of chemical marks, with more than 100 distinct RNA modifications present in eukaryotic systems [21]. Among the more than 100 RNA modifications identified in cellular RNAs, m6A, m5C, and pseudouridine (Ψ) have been detected in mitochondrial RNAs and contribute to the regulation of mitochondrial gene expression and RNA stability [43,44]. These modifications occur in different mitochondrial RNA species, including mitochondrial tRNAs, rRNAs, and, in some cases, mRNAs, where they influence RNA structure, ribosome function, and translational efficiency. In contrast, 5-formylcytidine (f5C) has been identified at a specific site in mitochondrial tRNA^Met, where it occupies the wobble position of the anticodon and enables proper decoding of AUA codons as methionine. The distribution and functional roles of these modifications collectively contribute to the regulation of mitochondrial translation and energy metabolism. The pancreas, especially the pancreatic β-cell, depends on well-tuned mitochondrial function for optimal energy production. Therefore, these RNA modifications are vital for mitochondrial functions and subsequent regulation of the glucose-stimulated insulin secretion in pancreatic β-cells [16,19,39]. However, the mechanisms governing their strategic arrangement in mitochondrial RNAs, as well as the role of each modification in mitochondrial function, are yet to be discovered.

The arsenal of mitochondrial RNA modifications in β-cells appears highly adaptable to allow dynamic adjustments to adapt to the metabolic demands placed upon β-cells. m6A and m5C are located in regulatory regions in mitochondrial RNAs, critical for the RNA stability and translational accuracy, which directly influence the mitochondrial protein synthesis and indirectly regulate insulin secretion [16,45]. These findings suggest that the mitochondrial epitranscriptome is refined in β-cells and can be reprogrammed to optimize β-cell functions, especially under stress conditions like acute glucose exposure [16]. The question of whether the metabolic flexibility of β-cells is highly associated with dynamic RNA modifications in mitochondria still needs further investigation. The importance of these modifications is revealed in their extensive clustering at key sites such as the anticodon loop in tRNA and the catalytic centers of rRNAs. Their positioning ensures translational accuracy as well as maximal energy production, which are required for β-cell survival and optimal insulin secretion [19]. However, whether the clustering of RNA modifications indicates a universal regulatory mechanism in mitochondria or if unique features of β-cells have led to these phenomena remains elusive.

The functional impact of these modifications also reveals the complexity in the regulation of the mitochondrial epitranscriptome and its impact on β-cell functions, as they often cooperate or interact. The presence of multiple modifications such as m6A, Ψ, and m5C within the same transcript will have synergistic or counteracting effects to maintain β-cell functions. In β-cells, the complex modification landscape in mitochondrial tRNAs and mRNAs enhances the ability to respond quickly to fluctuating extracellular glucose levels, as well as to stress conditions [16,46]. However, whether the presence of m6A may recruit or enhance the activity of PUSs or DNMTs, which promote the addition of Ψ or m5C, is unknown. Although the complex modification landscape in mitochondrial RNAs allows β-cells to cope with drastic changes in nutrient status, the loss of these modifications can alter mitochondrial biogenesis, mitochondrial metabolism, mitochondrial dynamics, and ultimately impact overall cell functions and survival [16,19]. Loss of chemical marks may cause site-specific defects, altering mitochondrial gene expression and leading to impaired ATP production, an imbalance in redox status, and subsequent disruption of β-cell functions.

Characterizing the mitochondrial epitranscriptome in β-cells opens opportunities to identify novel therapeutic and diagnostic strategies for type 2 diabetes. Indeed, recent publications have suggested that certain RNA modifications in β-cell mitochondria are associated with early β-cell dysfunction in diabetes [16,47]. Therefore, establishing tools to properly map and quantify the mitochondrial RNA modification sites in both normal and diabetic β-cells will facilitate early diagnosis and intervention. However, as these are early-stage findings, further investigation is required to properly validate their clinical relevance. Nine methylated residues and one pseudouridine site have already been identified in mammalian mitochondrial rRNAs. The methylation sites of mitochondrial rRNAs, consisting of m6A, m5C, and Ψ, are indispensable for mitochondrial ribosome assembly and translational fidelity [21]. Both processes are crucial for β-cell energy metabolism as well as for glucose-stimulated insulin secretion. However, how each modification plays a role in ribosome stability, activity, and quality control remains unexplored.

The stability of ribosomes is dependent on several rRNA modifications; thus, the lack of these modifications compromises ribosome stability and ribosome assembly. Subsequently, it reduces mitochondrial protein synthesis and compromises β-cell bioenergetic function [21,48]. Moreover, the electron transport chain activity of mitochondria is significantly compromised, thereby leading to decreased insulin secretion in β-cells. Thus, this would eventually lead to diabetes development [21]. Therapeutic stabilization of rRNA modifications can be used to reverse β-cell failure in type 2 diabetes.

Mammalian mitochondrial rRNAs contain a limited number of RNA modifications, including nine methylated residues and a single pseudouridine site (Ψ1397) in the 16S rRNA, which is catalyzed by the pseudouridine synthase RPUSD4. Proper installation of this modification is important for mitochondrial ribosome function and translation. The cooperative or competitive interactions between the different modification networks could ensure the overall high efficiency of mitochondrial translation and the proper function of β-cells [19].

The relatively short size and distinctive architecture of mitochondrial tRNAs, as compared with that of nuclear-encoded tRNAs, make them highly susceptible to changes in folding and interaction with ribosomes [19]. Efficient translation ensures that OXPHOS components are synthesized for β-cell respiration, eventually stimulating the glucose-induced ATP production. For example, depletion of m5C at the specific site on tRNA^Leu(UUR) leads to drastic reductions in aminoacylation and mitochondrial translation [14]. Interestingly, the modification pattern of tRNAs is greatly modulated and impacted by environmental and metabolic stress. Therefore, how these modifications change and what factors regulate their presence or absence could have significant consequences in β-cell function [19].

RNA modifications such as m6A and Ψ are indispensable for the stability and proper folding of RNA molecules, the association with proteins, the translational efficiency, and the mRNA turnover rate [16]. They act as dynamic hubs for coordinating mitochondrial and cellular functions in response to altered nutrient conditions in β-cells. For example, specific m6A modifications in mitochondrial mRNAs strongly influence mitochondrial protein synthesis and thereby affect the functional output of the OXPHOS complexes [16]. Meanwhile, Ψ enhances the thermal stability of RNA, facilitating the incorporation of ribosomal proteins in the vicinity. Furthermore, m6A and Ψ on RNAs affect ribonucleoprotein complex recruitment. These RNA modifications are the main hub integrating mitochondrial and cellular metabolism. Given that β-cells are heavily exposed to oxidative and nutrient stress, it will be of great value to study how stress regulates these processes. Since they impact mitochondrial protein synthesis as well as ATP production, therapeutic manipulation of these modifications could be an effective and useful strategy to rescue these processes in diabetic β-cells [16].

In addition to environmental cues, toxic metals such as cadmium, mercury, and lead are emerging as novel chemical stressors of mitochondrial RNA and subsequent translation. Cells exposed to these toxic metals will undergo translational errors and increased oxidative stress, eventually promoting β-cell failure [49]. These toxic metals also may alter molecular epitranscriptomic pathways, causing translational errors, increased oxidative stress, and β-cell failure. The interaction between toxic chemicals and epitranscriptomic elements will open a novel opportunity to further dissect the causal link between pollution and diabetes and identify novel targets for early diagnosis and treatment [50].

2.2.2. Enzymatic Machinery of RNA Modifications

The enzymes responsible for mitochondrial RNA modifications in pancreatic β-cells consist of writer, eraser, and reader proteins that are specialized to perform the functions of installing, removing, and recognizing chemical modifications. RNA writer enzymes, including methyltransferases and pseudouridine synthases, install chemical modifications, such as m6A, m5C, and Ψ, into mitochondrial RNA. TFB1M and MRM2 catalyze methylation events on mitochondrial RNA, while PUS1 installs Ψ into rRNAs and tRNAs. Experimental depletion of RNA modification writer enzymes, through approaches such as gene knockdown or knockout, leads to destabilized mitochondrial ribosomes, defective assembly of respiratory chain complexes, reduced ATP production, and impaired insulin secretion [18,21]. A cellular model that lacks MRM2 has reduced growth on a galactose-based medium, implying reduced mitochondrial metabolism. The results of MRM2 inactivation showed reduced mitochondrial translation and respiratory chain activity [51].

Mitochondrial readers are thought to regulate mitochondrial RNA stability, translation, and protein synthesis. Even though only a little information is available about the function of reader proteins, there could potentially be undiscovered mitochondrial reader proteins. Therefore, reader proteins can selectively bind modified RNA residues, influencing stability and translation, and thus they are also regulators of mitochondrial gene expression [16,19]. Therefore, the reader proteins may bind to specific sites where m6A occurs and then alter the conformation of RNA to enhance its stability or impact translation. Readers can also lead to conformational changes within the RNA molecule so that other factors can or cannot bind to that RNA. This action might affect the protein synthesis of the cell [16]. Epitranscriptomic modifications of mRNA are installed by RNA methyltransferases (“writers”) and can be reversed by RNA demethylases, often referred to as RNA “erasers.” These proteins can be categorized into FTO and ALKBH family proteins [52]. Unfortunately, to date, there is no evidence showing that these erasers are present in the mitochondria. Therefore, RNA methylations might remain in the mitochondria, and the changes cannot be reversed [16].

In this context, metabolic and oxidative stress associated with diabetes may promote the accumulation of dysfunctional RNA modifications, contributing to impaired mitochondrial function [52]. If mitochondrial RNA modifications are indeed stable or only slowly reversible due to the absence of dedicated demethylase activities, this raises important implications for mitochondrial function and β-cell physiology. Persistent RNA modifications could serve as a form of molecular “memory,” encoding prior exposure to metabolic or oxidative stress and thereby influencing mitochondrial translation efficiency and bioenergetic output long after the initiating insult has resolved. In β-cells, which rely on tightly regulated mitochondrial activity to sustain insulin secretion, such epitranscriptomic memory may contribute to prolonged functional impairment and increased vulnerability to subsequent stress. This concept aligns with emerging models of metabolic memory in diabetes, in which transient hyperglycemia or oxidative stress induces durable molecular changes that perpetuate cellular dysfunction. Thus, the apparent irreversibility of mitochondrial RNA modifications may represent a critical mechanism linking early metabolic stress to long-term β-cell failure. Also, an association has been made through clinical analysis. Specifically, humans bearing mutations in the genes that encode for the enzymes responsible for RNA modification can have serious conditions that may range from insulinopenic diabetes to multisystemic mitochondrial disease [19,52]. Thus, the specific and non-redundant functions of the enzymes responsible for modifying RNA have been supported in maintaining cell health.

Additionally, the activation of mitochondrial RNA modification enzymes may constitute an adaptive response during metabolic stress that can be manipulated to preserve β-cell integrity in diabetes [16]. Interestingly, RNA metabolism has been shown to have a huge impact on cell adaptation to stress. Therefore, the therapeutic implications can possibly benefit the pancreatic β-cell.

2.2.3. Mapping Approaches for Mitochondrial RNA Modifications

Mapping mitochondrial RNA modifications in pancreatic β-cells is important for understanding their role in β-cell bioenergetics and function (Figure 3). Nanopore direct RNA sequencing allows direct single-molecule analysis of modifications such as m6A and Ψ at base-level resolution without reverse transcription or amplification. The small size of the mitochondrial transcriptome of β-cells allows for more robust characterization of RNA modifications [16]. Nanopore direct RNA sequencing can be utilized to study the distribution and dynamic regulation of modifications in response to glucose stimulus and metabolic stress. Major limitations of nanopore direct RNA sequencing include low specificity to closely related modified nucleotides as well as the high error rate of the technique, and nanopore direct RNA sequencing often requires complementary experiments to validate the results and examine the impact of modifications on RNA or cell function [16,53].

RNA immunoprecipitation (RIP) can also be used to enrich mitochondrial RNA species associated with RNA-binding proteins or RNA-modifying enzymes in β-cells. By using antibodies targeting specific RNA-binding proteins or modification “reader” proteins, RIP-seq enables the identification of RNA–protein interaction sites and provides insight into how RNA modifications influence transcript stability and translation under stress or disease conditions [19]. However, direct mapping of RNA modifications such as m6A or m5C generally requires specialized approaches, including antibody-based methods (e.g., MeRIP-seq for m6A) or chemical and bisulfite sequencing–based techniques for m5C detection. While RIP-seq has contributed to understanding RNA–protein interactions involved in mitochondrial RNA regulation, the method has several limitations, including antibody specificity, limited spatial resolution, and the inability to quantify modification levels or map changes at single-nucleotide resolution [54].

Several methodological caveats exist when mapping mitochondrial RNA modifications in β-cells. Due to the size of the mitochondrial RNA transcriptome as well as the proximity of nuclear-encoded RNAs, very high-purity preparations are crucial for mapping. To avoid contamination from mRNA or nuclear-encoded RNA, isolation from highly pure mitochondria or total mitochondrial RNA is critical. Some modifications, such as 5-hydroxycytosine (5-hC), are very rare in the transcriptome and therefore require a highly sensitive technique to ensure reliability [55]. Another caveat to RNA modification mapping is the lack of single-nucleotide resolution of some mapping techniques, such as RIP and mass spectrometry. Knowing with base-level resolution the location of RNA modifications will be necessary to elucidate their function in mitochondrial biology of β-cells [56].

3. Mechanisms Linking RNA Modifications to β-Cell Function

Understanding how RNA modifications modulate mitochondrial translation and respiration is key to deciphering the mechanisms underlying β-cell dysfunction (Figure 4). The roles of specific epitranscriptomic marks in maintaining mitochondrial integrity, energy production, and insulin secretion are explored in this section, connecting molecular changes to the pathophysiology of diabetes. These mechanisms contribute to the broader understanding of mitochondrial bioenergetics and the role of epitranscriptomic regulation, thus representing potential targets for therapeutic intervention.

3.1. Impact on Mitochondrial Translation and Respiration

Mitochondrial RNA modifications influence the efficiency and fidelity of mitochondrial translation that ultimately affects respiratory and ATP-producing function in β-cells (

Table 1. Characteristics and functional consequences of mitochondrial RNA modifications in pancreatic β-cells.

RNA Mark	Enzyme Class	Known Reader	Effect on Translation/Respiration	Phenotype in β-Cell	References
m6A (N6-methyladenosine)	Writer: METTL3/METTL14 Mitochondrial-associated: METTL17 Eraser: FTO, ALKBH5	YTHDF1/2, IGF2BP2 (β-cell relevance: inferred)	↑ Translation efficiency ↑ OXPHOS protein expression	↑ ATP production ↑ Insulin secretion (likely)	[58,59,60]
m5C (5-methylcytosine)	Writer: NSUN4 Eraser: Not established	ALYREF (primarily nuclear; mitochondrial role inferred)	↑ Ribosome biogenesis ↑ RNA stability	↑ Mitochondrial protein synthesis (likely)	[61,62,63]
Ψ (Pseudouridine)	Writer: PUS1, TRUB2 Eraser: Not identified	Not identified	↑ tRNA stability ↑ Translation accuracy	↑ OXPHOS assembly ↓ Oxidative stress (inferred)	[64,65]
m1A (1-methyladenosine)	Writer: TRMT61A/B Eraser: ALKBH3	No established dedicated readers	↑ tRNA function Variable effects on translation	Context-dependent effects on respiration (inferred)	[66,67,68,69]
m7G (7-methylguanosine)	Writer: METTL1 Eraser: Not established	eIF4E (cap-binding protein, not a modification “reader”)	↑ Translation initiation (cap-dependent)	↑ ATP synthesis (indirect)	[70,71,72,73]
ac4C (N4-acetylcytidine)	Writer: NAT10 Eraser: Not established	Not identified	↑ rRNA processing ↑ Ribosome function	↑ Mitochondrial biogenesis (emerging)	[74,75,76]

) [57]. Epitranscriptomic marks play a crucial role in governing ribosomal integrity, complex assembly, and bioenergetic balance. Understanding the mechanisms underlying these effects provides novel insights into their involvement in insulin secretion. Integrating such molecular mechanisms into mitochondrial biology provides crucial insights into the early pathogenesis of diabetes and enhances the overall resilience of β-cells.

3.1.1. m6A in mRNAs and Effects on Ribosomal Fidelity and Complex Assembly

N6-methyladenosine (m6A) is a well-characterized epitranscriptomic modification primarily found on nuclear-encoded mRNAs, where it regulates RNA stability, translation, and cellular stress responses. While direct m6A modification of mitochondrial DNA (mtDNA)-encoded transcripts in mammals remains unestablished, m6A-dependent regulation of nuclear-encoded mitochondrial proteins plays an important indirect role in mitochondrial function, including ribosomal activity and respiratory chain assembly in β-cells. Through modulation of nuclear transcripts encoding mitochondrial ribosomal proteins, assembly factors, and electron transport chain (ETC) components, m6A influences mitochondrial translation efficiency and proteome integrity. Disruption of m6A “writer” activity has been associated with impaired translational fidelity, amino acid misincorporation, and defective OXPHOS subunit assembly. These alterations ultimately reduce ATP production and compromise the bioenergetic capacity required for glucose-stimulated insulin secretion (GSIS) [16,21,77].

In this context, m6A contributes to the regulation of mitochondrial function by controlling the expression and translation of nuclear-encoded genes essential for ETC complex I–V assembly. Under m6A-deficient conditions, reduced translation of these key components leads to incomplete OXPHOS assembly and diminished respiratory efficiency, thereby impairing ATP-dependent insulin granule exocytosis. These findings support a role for RNA epitranscriptomic regulation as an upstream modulator of mitochondrial bioenergetics in β-cells [16,21]. Alterations in m6A reader proteins, such as YTHDF1 and YTHDF2, have been reported under diabetic or hyperglycemic conditions, potentially reflecting adaptive responses to cellular stress. However, their direct involvement in mitochondrial regulation in β-cells remains incompletely defined, and current evidence is largely derived from non-mitochondrial or non-β-cell systems [16,78]. Therefore, their role in mitochondrial dysfunction should be interpreted with caution.

Overall, while m6A does not appear to directly modify mitochondrial transcripts in mammals, it exerts significant indirect control over mitochondrial function through regulation of nuclear-encoded genes. Disruption of this regulatory axis contributes to impaired mitochondrial proteostasis, increased oxidative stress, and β-cell dysfunction, highlighting its relevance in the early pathogenesis of diabetes.

3.1.2. m5C in tRNAs and Effects on Aminoacylation and ATP Production

Mitochondrial RNA modifications such as m5C in mitochondrial transfer RNAs (tRNAs) play a critical role in ensuring accurate mitochondrial translation, tRNA aminoacylation, and ATP production in pancreatic β-cells. The loss of m5C methylation is a direct inducer of poor aminoacylation, resulting in incorrect mitochondrial protein synthesis, OXPHOS complex impairment, and thus, low ATP production. m5C maintains the accurate mitochondrial protein synthesis in pancreatic β-cells [44]. Aminoacylation is a process whereby tRNAs are “charged” by binding to their cognate amino acids. tRNAs such as tRNA^Leu(UUR) have m5C, which plays a key role in the tRNA structure for recognition by aminoacyl-tRNA synthetases. If the m5C is removed, the tRNAs are no longer correctly or efficiently aminoacylated [79]. In this regard, in the absence of adequate m5C levels, the tRNAs are not readily charged, causing erroneous synthesis of protein products. The m5C modification, thus, enhances the speed and accuracy of translation.

Preclinical studies in diabetic animal models and human islets reveal that m5C levels are negatively correlated with ATP production and GSIS. In mitochondria-specific deficient m5C methylation, particularly at sites such as mitochondrial tRNA^Leu(UUR), islet cells fail to adequately stimulate ATP production. Islets from individuals with impaired fasting glucose (IFG) exhibit reduced m5C levels, defective aminoacylation, and impaired mitochondrial translation. However, emerging evidence indicates that the functional impact of m5C deficiency in mitochondrial tRNAs may be more nuanced than previously assumed. Notably, complete loss of m5C at positions 48–50 in multiple mitochondrial tRNAs via NSUN2 knockout resulted in minimal effects on mitochondrial translation and respiratory chain complex abundance [80]. These findings suggest that not all m5C sites are equally critical for mitochondrial function and that compensatory mechanisms or site-specific dependencies may exist within the mitochondrial translation system. NSUN2, a key m5C methyltransferase predominantly localized in the nucleus but also detected within mitochondria, plays a broader role in cellular metabolism beyond mitochondrial RNA modification. Recent evidence further highlights its systemic metabolic relevance. For instance, NSUN2 is upregulated in T2DM models and clinical samples, and that its knockdown improves glucose tolerance, reduces insulin resistance, and alleviates lipid accumulation through m5C-dependent regulation of metabolic genes such as ACSL6 [81]. These findings indicate that NSUN2-mediated m5C modifications may exert tissue-specific and context-dependent effects, influencing not only mitochondrial RNA biology but also broader metabolic pathways.

m5C deficiency predisposes cells to increased vulnerability to metabolic stress. Loss-of-function enzymes responsible for the m5C modification contribute to metabolic stress by limiting the adaptive response of mitochondria to high glucose levels or oxidative stress. These deficiencies can promote instability and abnormal protein synthesis by inhibiting the correct modification of tRNAs through m5C. Therefore, deficient m5C modification of tRNAs exacerbates mitochondrial dysfunction and accelerates pancreatic β-cell failure, leading to the onset and progression of diabetes [82]. This strengthens the assumption that m5C dysfunction drives the etiology and progression of the disease by directly dysregulating energy expenditure in many cell types.

3.2. Redox Homeostasis, ROS Signaling, and Mitochondrial Unfolded Protein Response

Redox state and stress response pathways in pancreatic β-cells modulate protein quality and oxidative damage control in mitochondria. Perturbations of these pathways promote mitochondrial unfolded protein response, oxidative stress, and cell apoptosis to eventually cause β-cell failure [83]. Elucidating these connected mechanisms is key to determining the role of mitochondrial RNA modifications on cellular resilience and β-cell function in diabetes development.

3.2.1. Misfolded Proteins from Aberrant Marks and UPRmt Activation

Aberrant mitochondrial RNA (mtRNA) modifications in pancreatic β-cells, particularly defects in m6A, m5C, and pseudouridine (Ψ), disrupt translational fidelity and promote protein misfolding within the mitochondrial matrix. Deficiencies in RNA modification enzymes destabilize mitochondrial ribosomal subunits, accelerate rRNA degradation, and impair the assembly of functional 55S ribosomes. Collectively, these defects impair mitochondrial translation, particularly of OXPHOS components, leading to proteostatic imbalance and mitochondrial dysfunction [17,18,21,84]. Disruption of mitochondrial proteostasis triggers activation of the mitochondrial unfolded protein response (UPRmt), an adaptive signaling pathway aimed at restoring protein homeostasis. UPRmt induces the expression of mitochondrial chaperones, proteases, and quality control systems to refold or degrade misfolded proteins and maintain bioenergetic function. However, sustained RNA modification defects prolong UPRmt activation, diverting cellular resources away from mitochondrial translation and biogenesis, ultimately compromising metabolic efficiency in β-cells [16,21]. To provide a consolidated overview, the key mitochondrial RNA-modifying enzymes and their functional consequences in β-cells are summarized in

Table 2. Mechanistic pathways linking mitochondrial RNA modifications to β-cell dysfunction and diabetes-related phenotypes.

Enzyme	Modification	RNA Target	β-cell Phenotype on KO/Depletion	Human Disease Link	Reference
TFB1M	m6A	12S rRNA	Impaired mitoribosome assembly, reduced mitochondrial translation, decreased ATP production, defective GSIS	Diabetes-like phenotype, mitochondrial dysfunction	[45,46]
MRM2	2′-O-methylation (Um)	16S rRNA	Destabilized large ribosomal subunit, impaired OXPHOS translation, reduced ATP and insulin secretion	Mitochondrial disease phenotypes	[59,85]
PUS1	Pseudouridine (Ψ)	mt-tRNA and 16S rRNA	Reduced ribosome integrity, decreased ATP production, impaired GSIS, increased ROS	Insulinopenic diabetes, mitochondrial disorders	[52,57]
TRMT61B (mt-mRNA)	m6A	mt-mRNA	Altered mitochondrial translation efficiency, dysregulated protein synthesis, impaired bioenergetics	Associated with mitochondrial dysfunction	[44,86]
TRMT61B (mt-tRNA)	m1A	mt-tRNA	Disrupted tRNA stability and translation fidelity, impaired OXPHOS function	Mitochondrial disease phenotypes	[44,86]

.

Given their high dependence on oxidative phosphorylation for glucose-stimulated insulin secretion (GSIS), β-cells are particularly sensitive to proteostatic imbalance. While transient UPRmt activation is protective (supporting recovery of mitochondrial function), chronic activation reduces ATP availability, suppresses protein synthesis, and impairs OXPHOS capacity. This maladaptive phase is characterized by persistent mitochondrial dysfunction, reduced biogenesis, and progressive loss of cellular homeostasis, culminating in apoptosis [16,21,87]. Prolonged UPRmt activation is also associated with elevated reactive oxygen species (ROS) production, driven by defective electron transport and protein degradation. Due to inherently low antioxidant defenses, β-cells are especially vulnerable to oxidative damage affecting mtDNA, rRNA, and mitochondrial proteins. This oxidative stress further amplifies mitochondrial dysfunction and promotes apoptotic signaling, contributing to β-cell loss in both T1D and T2D [88].

Overall, this enhances cellular vulnerability to oxidative and metabolic stress, consequently lowering the threshold for cell death, leading to the initiation of the diabetic state. Site-specific RNA modifications are master regulators of cellular resilience, and their proper placement can be applied in novel β-cell regenerative therapies for T1D and T2D [18,21,89].

3.2.2. Chronic UPRmt, ROS, DNA Damage, and Apoptosis

Persistent activation of UPRmt in β-cells causes elevated levels of ROS, leading to damage of mtDNA/RNA. This also triggers the loss of antioxidant enzymes. Oxidative DNA/RNA damages (8-oxo-guanine) are increased under diabetic conditions. Accumulation of oxidatively damaged mtRNA precedes mtDNA accumulation in diabetic nephropathy, suggesting that mtRNA might serve as an earlier indicator for diagnosis [90].

Chronic UPRmt impairs cellular antioxidant function by depleting glutathione stores, leading to oxidative damage and apoptotic events. Defects in mtRNA modifications such as m6A, m5C, and Ψ are also shown to induce UPRmt, increase ROS, deplete glutathione, and trigger apoptosis [91]. ER stress and NADPH oxidase-induced ROS also aggravate β-cell mitochondrial function to initiate β-cell dysfunction in T1D and T2D [92].

The chronic UPRmt becomes maladaptive in the β-cells due to elevated biosynthesis for the proper function of insulin. It can lead to apoptotic responses of the β-cells. The threshold of these cells for UPRmt is lower, making them more prone to damage [91,93]. Accumulation of damaged or incomplete transcripts and ROS in mitochondria can cause impaired mitochondrial ribosome function, which leads to a vicious cycle to further accumulate misfolded proteins and damage of mtDNA/RNA. Strategies to improve mtRNA modifications can lead to better β-cell function.

3.2.3. PUS1 Depletion Effects in INS-1 Cells

Since PUS1 and the Ψ residue have been demonstrated to impact folding of rRNAs and tRNAs for proper ribosome activity, it is probable that the loss of this protein could lead to misfolded rRNAs and tRNAs that would further destabilize the mitochondrial ribosomes. The result of dysfunctional or incomplete mitoribosome activity inhibits the translation of mitochondrial proteins and severely impairs the mitochondrial function. This establishes that mitochondrial epitranscriptomic regulation is important for β-cell mitochondrial bioenergetics [21].

Thus, the loss of PUS1 also negatively impacts mitoribosome translation of mitochondrial-encoded OXPHOS proteins, resulting in reduced complex activity, decreased ATP production, mitochondrial membrane potential, and dysfunctional β-cell bioenergetics and secretion of insulin. These combined mechanisms of deficient mitochondrial translation demonstrate that PUS1 is critical to β-cell energy regulation [18,21]. Independent evidence supports the assertion that deficiencies in tRNA and rRNA modifications can cause impaired synthesis of OXPHOS proteins, destabilizing respiratory complexes and releasing ROS in response to metabolic stress [18].

Improper modification of nucleic acids can also destabilize mitochondrial function and overall β-cell function [18,94]. Furthermore, the increased oxidative stress response in the β-cell lines as a result of PUS1 knockdown can disrupt nucleic acid and protein function. The β-cell is particularly susceptible to increased ROS, as this leads to heightened apoptosis rates. These factors combined lead to the impaired stability and activity of the electron transport chain, increased oxidative stress, and mitoribosome vulnerability to cellular apoptosis [18,21].

Further, depletion of PUS1 and the resultant modification defects can have a similar effect to specific gene knockdown on the cellular mitoribosomes and overall cell function. Targeted knockdown of mitochondrial RNAs using antisense oligonucleotides (ASOs) can result in analogous effects to PUS1 depletion in β-cells. ASOs can be mitochondrial-targeted, permitting the specific knockdown of mitochondrial RNA transcripts and examination of overall mitochondrial protein synthesis defects [95]. The use of these probes revealed a general decrease in respiratory protein expression and a reduction in mitochondrial membrane potential and bioenergetic capacity when the indicated mRNA targets were knocked down by the use of ASOs in HeLa cells. In addition, fluorescence staining techniques can monitor the effect of RNA knockdown on mitochondrial membrane potential. These methods indicate that knockdown via ASO exposure to β-cells results in reduced, diffuse staining of mitochondria within the cytoplasm of the cells, compared to control cell lines, which stain at a much higher level with clearly defined mitochondrial organelles [95]. These experimental data demonstrate that mitochondrial RNA modifications, and more specifically pseudouridine modifications, are crucial contributors to β-cell function through proper assembly and function of the mitoribosome and downstream translation of OXPHOS complex subunits within the mitochondria.

4. Dysregulated Mitochondrial RNA Modifications in Diabetes Models

Mitochondrial RNA modifications are central to normal β-cell function and energy metabolism and are altered in diabetes. The following section explores site-specific changes from various preclinical models and humans, including how these epitranscriptomic aberrations result in defective mitochondrial translation and bioenergetics. Within the broader themes of mitochondrial bioenergetics and disease, this offers an improved insight into the early molecular events causing β-cell failure and subsequent diabetes development.

4.1. Preclinical Rodent and Human Islet Studies

In this chapter, we focus on the alteration of mitochondrial RNA modifications observed in different preclinical models of the disease, highlighting how this process can lead to impairments in β-cell function and energy metabolism (Figure 5). This provides an understanding of site-specific alterations of the mRNA modifications in rodent and human islets, representing the onset of diabetes. Within the framework of mitochondrial bioenergetics in pathological processes, we discuss potential therapeutic strategies.

4.1.1. STZ-Induced and db/db Mouse Models: Site-Specific Changes

In studies of STZ-induced diabetic models and db/db mice, defects in site-specific mitochondrial tRNA modifications, including m5C and other epitranscriptomic marks, have been linked to β-cell dysfunction and diminished insulin secretion in diabetic states. The loss of certain RNA modifications has been directly correlated to declines in β-cell function, suggesting early impairment of mitochondrial RNA regulation [18,19]. Moreover, closer inspection of these models has identified significant decreases of m5C at site-specific functional regions, most notably in the tRNA^Leu(UUR) locus, leading to defective aminoacylation efficiency. This impairment in site-specific RNA modifications further decreases mitochondrial translation and bioenergetic function.

Additional studies, including those in human monogenic diabetes and mitochondrial mutation models (e.g., m.3243A>G and TRMT10A deficiency), have shown that mutation or deficiency of mitochondrial RNA-modifying enzymes can lead to incomplete tRNA maturation and impaired mitochondrial translation. In db/db mice, these bioenergetic defects in β-cells are associated with inadequate insulin secretion and glucose intolerance, whereas in human mtDNA mutation or monogenic models, similar mitochondrial dysfunction arises from genetically defined defects in RNA modification pathways [19]. For instance, levels of fragmented mitochondrial tRNA fragments, or tRFs, were higher in db/db pancreatic islets and serum of db/db mice as compared to their control counterparts, implying increased mitochondrial and β-cell stress. These altered levels of circulating mitochondrial tRFs in diabetic mice point to their potential as molecular biomarkers for mitochondrial and β-cell dysfunction [19,96]. Further in-depth analyses of these mechanisms found destabilization of certain mitochondrial rRNA, which consequently disrupted ribosome formation, hindering the translation of essential proteins involved in oxidative phosphorylation complex assembly.

Indeed, dysregulated tRNA fragmentation and modifications have been shown to directly impair glucose tolerance and lead to insufficient insulin secretion [19,96]. ATP, a key molecule for energy generation in the glucose-stimulated insulin secretion pathway, is often reduced under conditions of defective tRNA modification, reflecting impaired mitochondrial bioenergetics and altered ATP/ADP ratios in diabetic models. Similarly, alterations in expression and levels of mitochondrial tRFs, which were also found in STZ-induced diabetic models, have been linked to increased oxidative damage and redox imbalance, common mechanisms that are understood to promote β-cell apoptosis.

Finally, the accumulation of mtDNA and RNA mutations, mitochondrial dysregulation, and loss of RNA modification control, observed in both human genetic models and experimental diabetic mouse models, correlate with progressive metabolic stress and the development of diabetes [19,93]. Thus, defective tRNA fragmentation and modification appear to contribute to the development of β-cell dysfunction. These site-specific losses of modification, such as loss of m5C in the wobble stem of tRNA, have been directly detected prior to marked insulin resistance and during a stage in which β-cell insulin secretion is normal. Moreover, these losses in m5C are still easily observable during early insulin resistance with β-cell hypersecretion, and furthermore are detectable later on, but at lower levels, during frank diabetic states [19].

These findings challenge the common notion that mitochondrial and RNA dysregulation are merely a secondary response, but rather serve as reversible factors in the development of β-cell stress and diabetes progression.

4.1.2. Human Islets from IFG Donors: Reduced m5C in tRNA^Leu(UUR)

The reduction in m5C modification on mitochondrial tRNA^Leu(UUR) has been observed in pancreatic islets from impaired fasting glucose (IFG) individuals, thus being the early epitranscriptomic disruption in prediabetes. Quantitative measurement also confirmed that a deficit of m5C on IFG samples was significantly lower compared to normoglycemic controls [17,19]. Thus, the potential of this modification as a molecular indicator for β-cell dysfunction is feasible.

m5C of tRNA^Leu(UUR) plays an important role in the correct structure and codon-anticodon pairing of the particular tRNA molecule and is essential for the proper mitochondrial translation. The loss of m5C from the mentioned tRNA species is responsible for an altered mitochondrial protein synthesis that leads to deficient production of OXPHOS subunits [18]. This altered OXPHOS complex impairs the energy necessary for glucose-stimulated insulin secretion.

The m5C reduction in IFG islets is significantly related to the lower aminoacylation of tRNA^Leu(UUR). This finding supports the hypothesis that aminoacylation deficits contribute to decreased expression of OXPHOS complex subunits during translation. Loss of aminoacylation impairs oxidative phosphorylation-dependent ATP production for energy, coupling glucose to insulin secretion [18,97]. This hypothesis also confirms several findings in diabetic animal models, highlighting the conserved feature in early mitochondrial dysfunction in prediabetes.

The diminished levels of m5C on tRNA^Leu(UUR) were a characteristic of mitochondrial dysfunction found only on this tRNA species and not throughout mitochondrial tRNA. As mentioned before, the peculiar sensitivity of the specific tRNA^Leu(UUR) to metabolic stress indicates that tRNA^Leu(UUR) acts as a sentinel for β-cell homeostasis. This cell-specific defect also indicates an intrinsic vulnerability of β-cells to stress-induced mitochondrial RNA modifications [19,98]. Therefore, targeting this RNA modification may be a worthwhile therapeutic approach.

The downstream consequence of m5C-deficient tRNA^Leu(UUR) in mitochondria leads to defective translation and, therefore, diminished oxidative capacity to produce ATP and an increase in oxidative stress. The consequence is the increased production of ROS that inhibits the electron transfer process in the oxidative phosphorylation cycle. When ROS overwhelms, the cellular redox equilibrium will be affected, damaging the mitochondrial macromolecules. Simultaneously, the decreased production of ATP impairs ATP-dependent antioxidant and defense mechanisms in scavenging free radicals, thus exacerbating the condition. Increased oxidative damage caused by excess ROS also leads to a chronic inflammatory response through the activation of pro-apoptotic pathways. This leads to β-cell damage and death by a variety of biochemical changes in this stage [18,99].

As the loss of m5C from tRNA^Leu(UUR) appears at the IFG stage, which is before a drop in β-cell mass and overt hyperglycemia, this epitranscriptomic change is postulated to be upstream of this condition. This further implicates its possibility to be an early signal of β-cell dysfunction that leads to prediabetes [19]. As well as being a biomarker candidate, m5C on tRNA^Leu(UUR) is a good target for therapeutics since it can be reversed by adding or removing the methyl group with small molecules. The causal effect of loss of m5C from tRNA^Leu(UUR) has been further demonstrated with a number of independent experimental models, where reduced or abolished expression of certain tRNA-modifying enzymes results in a phenotype very similar to what is observed in IFG, that is, defective insulin secretion and mitochondrial dysfunction [18].

Lastly, the specific and marked susceptibility of tRNA^Leu(UUR) to m5C depletion suggests the potential utilization of this site-specific RNA modification as a biomarker for mitochondrial dysfunction and β-cell stress. Additionally, incorporating m5C detection of tRNA^Leu(UUR) into existing circulating biomarker panels for the diagnosis of T2D and the estimation of pancreatic β-cell mitochondrial health may be a step forward towards personalized medicine [82]. The role of mitochondrial RNA modifications like m5C, therefore, includes signaling and integration of various metabolic, genetic, and environmental inputs on β-cell stress status [18]. In consequence, mitochondrial RNA modifications seem to have emerged as major players of various cellular functions in the β-cells.

4.2. Translational and Therapeutic Implications of Mitochondrial RNA Modifications

Mitochondrial RNA (mtRNA) modifications represent an emerging regulatory layer with significant implications for β-cell function and diabetes pathophysiology. Given their central role in mitochondrial translation and oxidative phosphorylation (OXPHOS), dysregulation of these modifications contributes to impaired ATP production, increased reactive oxygen species (ROS), and defective glucose-stimulated insulin secretion (GSIS). These features position mtRNA modifications as both mechanistic drivers of β-cell dysfunction and potential targets for therapeutic intervention [43,44,51].

A particularly relevant example is TRMT61B, a mitochondrial RNA methyltransferase with dual-substrate specificity that modifies both mitochondrial tRNAs and mRNAs. TRMT61B catalyzes N1-methyladenosine (m1A) formation in mt-tRNAs, which is essential for proper tRNA folding, stability, and decoding efficiency during mitochondrial translation. In parallel, it has been shown to introduce N6-methyladenosine (m6A) modifications in mitochondrial mRNAs, thereby influencing translational dynamics and protein synthesis within the organelle [100]. This dual functionality positions TRMT61B as a key integrator of mitochondrial gene expression at multiple regulatory levels. Dysregulation of TRMT61B activity may simultaneously impair tRNA maturation and mRNA translation, resulting in compounded defects in OXPHOS and ATP production [101]. In pancreatic β-cells, which rely heavily on coordinated mitochondrial translation for insulin secretion, such disruptions may exacerbate bioenergetic stress and functional decline, highlighting TRMT61B as a potential node of vulnerability and therapeutic interest.

Targeting mtRNA modification pathways may therefore offer novel therapeutic opportunities. Strategies aimed at restoring or modulating the activity of RNA-modifying enzymes could improve mitochondrial translational fidelity and respiratory chain function. In addition, interventions that mitigate downstream consequences, such as oxidative stress and proteostatic imbalance, may further enhance β-cell resilience under diabetic conditions. Beyond therapeutic applications, mtRNA modifications may also serve as biomarkers of mitochondrial health and β-cell function. Altered RNA modification patterns could reflect early mitochondrial dysfunction prior to overt metabolic impairment, providing opportunities for early detection and risk stratification in diabetes. Integrating epitranscriptomic profiling with functional and clinical datasets may therefore improve diagnostic precision and enable more targeted intervention strategies. The summary of preclinical evidence is presented in

Table 3. Preclinical models and human genetic syndromes implicating mitochondrial RNA epitranscriptome in diabetes.

Model	Site-Specific RNA Modification	Genetic Syndrome/Gene	Clinical Phenotype	Reference
Human diabetes nephropathy	Increased Ψ and m6A modifications	-	Diabetic nephropathy, ESRD risk	[46]
STZ-induced diabetic mice	Altered expression of metabolic enzyme genes	-	Diabetic retinopathy phenotype	[102]
db/db diabetic mice	Differential gene expression from RNA-seq	-	Obesity-associated diabetes model	[103]
Human mitochondrial RNA	1-methyladenosine (m1A) at 16S rRNA position 947	TRMT61B	Mitochondrial translation dysfunction	[104]
Human mitochondrial RNA	Pseudouridine modification defects	PUS1	Mitochondrial myopathy, lactic acidosis, and sideroblastic anemia (MLASA)	[104]
Human β-cells, INS-1 cells	Loss of Ψ incorporation in mitochondrial tRNAs and rRNAs	PUS1	Early-onset insulinopenic diabetes, mitochondrial myopathy, impaired OXPHOS, UPRᵐᵗ activation	[105]
Mitochondrial translation system	Pseudouridine deficiency affecting tRNA stability	PUS1	Reduced mitoribosome assembly, decreased OXPHOS protein synthesis	[21]
Human patients with PUS1 mutations	Defective pseudouridylation of mitochondrial tRNAs	PUS1	Early-onset insulinopenic diabetes, mitochondrial dysfunction	[106]
Mitochondrial stress models	ROS-induced mitochondrial dysfunction	PUS1-related	Oxidative stress, proteostatic stress, mitochondrial damage cycle	[107,108]
Human T2DM patients	Defective methylation of tRNA^Trp and tRNA^Ser(AGY)	TRMT61B	Impaired glucose-stimulated insulin secretion, early insulin dependence	[82]
β-cell apoptosis models	Mitochondrial protein homeostasis disruption	Mimitin deficiency	β-cell apoptosis, reduced proliferation, and caspase activation	[109]

Notes: STZ: Streptozotocin-induced diabetes model; IFG: Impaired fasting glucose; T2DM: Type 2 diabetes mellitus; ESRD: End-stage renal disease; m⁶A: N⁶-methyladenosine; m¹A: 1-methyladenosine; m⁵C: 5-methylcytosine; Ψ: Pseudouridine; OXPHOS: Oxidative phosphorylation; MLASA: Mitochondrial myopathy, lactic acidosis, and sideroblastic anemia; UPRᵐᵗ: Mitochondrial unfolded protein response; ROS: Reactive oxygen species.

.

TRMT61B is responsible for tRNA methylation in mitochondria. TRMT61B deficiency has been linked to defective methylation of the anticodon loop of tRNA^Trp and tRNA^Ser(AGY) in mitochondria. Deficient methylation on tRNA molecules destabilizes mitochondrial ribosomes and impairs protein synthesis in mitochondria. Lack of the TRMT61B protein thus reduces ATP levels and contributes to defective coupling between glucose metabolism and insulin secretion in β-cells. Importantly, these recent findings indicate that diabetes may develop even in the absence of mutations in mitochondrial DNA, highlighting a key role for mitochondrial RNA modifications as independent contributors to the etiology of diabetes [19,82].

Typically, patients with mitochondrial RNA modification mutations present with a range of metabolic and multisystemic features, often displaying sensorineural hearing loss, optic atrophy, and cardiomyopathy, reflecting the high-energy burden in these tissues, as well as many of the other metabolic complications. Thus, β-cells of diabetic patients carrying these mutations undergo apoptosis due to impaired ATP production, with neurological and sensorineural complications caused by the fact that these tissues rely heavily on ATP-dependent processes. Given that β-cells are the most highly metabolic cells of the body, it is likely that they are the first to undergo cell death as a consequence of defective mitochondrial RNA modification mutations [19,82]. Furthermore, incomplete penetrance has also been reported, meaning that some carriers of these mutations have impaired glucose tolerance [19,82]. This result indicates the importance of mitochondrial RNA modification in β-cell function even without a complete heteroplasmic background.

The high allelic heterogeneity in genes encoding mitochondrial RNA-modifying enzymes raises many questions about the potential mechanism of development of this type of diabetes, with a need for the implementation of molecular diagnostics for these rarer diseases. The recent advent of high-resolution molecular diagnostics has enabled accurate measurement of these modifications (e.g., mass spectrometry and direct RNA sequencing) and facilitated the detection and treatment of a great variety of human diseases. As such, patients could be diagnosed with subclinical β-cell failure and perhaps delay the onset of diabetes [19,110].

Currently, no mechanism is known to remedy this defect in patients with mitochondrial RNA modification mutations, and therefore, care is simply supportive. However, in the future, methods to reestablish Ψ and m5C in tRNAs by treating patients with a specific small-molecule analog or a CRISPR delivery system in the form of gene therapy may reverse the effect of disease progression. Altogether, defective mitochondrial RNA modification, characterized in different rare human mitochondrial diabetes syndromes, shares a common mechanism, suggesting that this type of disturbance may link both rare and more common forms of the disease [111]. This possibility breaks the common paradigm that diabetes is classified solely between genetic and the disease, but more that there is a continuum between both. Overall, the evidence was summarized and categorized as follows in

Table 4. Levels of evidence linking mitochondrial RNA modifications to β-cell dysfunction and diabetes.

Category of Evidence	Type of Study/Model	Key Findings	Interpretation for Causality
Direct Mechanistic (Causal)	β-cell-specific genetic knockouts (e.g., TFB1M, MRM2)	Loss of mitochondrial RNA-modifying enzymes impairs mitochondrial translation, disrupts OXPHOS, reduces ATP production, and leads to defective insulin secretion and diabetes phenotypes	Strong evidence that disruption of mitochondrial RNA modification machinery is sufficient to drive β-cell dysfunction
	Rare human monogenic disorders affecting mtRNA modification enzymes	Mutations in mitochondrial RNA-modifying enzymes are associated with insulinopenic diabetes and β-cell failure	Supports a direct, causal role of mtRNA modification defects in β-cell dysfunction in humans
Associative (Correlative)	Stress-induced models (hyperglycemia, oxidative stress, lipotoxicity)	Altered mitochondrial RNA modification patterns observed following metabolic or oxidative stress exposure	Indicates that mtRNA modifications respond dynamically to cellular stress; does not establish whether changes are initiating or secondary
	Human islets from IFG or T2DM donors	Differences in mitochondrial RNA modification levels compared with non-diabetic controls	Suggests clinical relevance, but causality cannot be inferred due to confounding metabolic changes.
Hypothetical/Speculative	Epitranscriptomic “memory” models	Persistent mtRNA modifications are proposed to encode prior metabolic or oxidative stress exposure	Conceptually plausible model; currently lacks direct experimental validation
	Therapeutic reversal of mtRNA modifications	Restoration of specific RNA modifications proposed to improve β-cell function	Forward-looking hypothesis; feasibility and efficacy remain to be demonstrated.

.

5. Emerging Biomarker and Therapeutic Opportunities

Advancements in mitochondrial RNA biology offer novel opportunities for early diagnosis and therapy of diabetes (Figure 6). The authors report on circulating RNA fragments as potential biomarkers for early diagnosis as well as on enzyme modulators for the therapy of β-cell dysfunction caused by mitochondrial epitranscriptomic disruption. With the goal of improving diabetes management and β-cell survival, their work is relevant in the broader context of mitochondrial mechanisms involved in diabetes pathogenesis.

5.1. Circulating Mitochondrial RNA Fragments as Biomarkers

5.1.1. Detection of Mitochondrial tRFs in Plasma

Circulating mitochondrial-derived tRNA fragments (mt-tRFs) could reflect real-time mitochondrial stress and β-cell injury in diabetes, as these fragments are found in the plasma and urine of humans with diabetes and in diabetic animal models. For example, higher levels of mitochondrial nucleic acids are detectable in the urine of both diabetic mice and diabetic human beings [112]. The abundance and diversity of mt-tRFs found in plasma are related to mitochondrial RNA processing disturbances, presence of tRNA mutations or modification defects in diabetes, as tRNA fragment patterns undergo substantial alterations in response to a high-fat diet, an established metabolic stressor [19]. This suggests that production of mt-tRFs is associated with mitotranscriptomic dysregulation and begs the question of whether there are selective regulatory pathways involved in the synthesis of these fragments.

Levels of mitochondrial DNA (mtDNA) and RNA were found to be higher in the urine than in plasma during the early stages of diabetic nephropathy. Furthermore, it has been proposed that the urinary mtDNA/creatinine ratio could serve as an early predictive marker for the disease. In contrast, lower levels of plasma mtDNA were observed in diabetic individuals as opposed to normal controls [112]. This could be attributed to several factors: the compartment in which the mtDNA fragments are measured, tissue-specific levels of mtDNA release during diabetes, and/or varying clearance kinetics.

Recent innovations in next-generation sequencing and mass spectrometry enable the mapping of mt-tRFs carrying the chemical modifications m6A and m5C and linking the differential modification status of RNA fragments to disease stages, insulin resistance, and organ dysfunction [113]. Furthermore, studies on defects in RNA modification enzymes affecting tRNA methylation demonstrate that altering mt-tRF output can have direct effects on the production and secretion of insulin [18]. Moreover, it remains possible that plasma mt-tRFs derive not only from β-cells but also from other cells.

It has been postulated that the selective profile and modification status of mt-tRFs found in plasma could constitute a form of “liquid biopsy” to assess mitochondrial defects in β-cells for risk stratification and personalized medicine early in the diabetes trajectory [19]. It will be important to develop new analytical techniques that specifically reflect mitochondrial activity and the release of mt-tRFs from β-cells, as well as sensitive methods for mt-tRF analysis that can detect lower levels of expression in plasma. The observation that increased levels of mt-tRFs are seen in the urine during times of mitochondrial damage and impaired organ function, while exciting, can be impacted by factors not directly related to the pancreas [112]. The ability to define reference ranges for mt-tRF signatures has been greatly impacted by developments in RNA sequencing and mass spectrometry technologies that yield reproducible, clinically relevant quantification assays.

5.1.2. Correlation of m6A-Modified CYTB Fragments with HOMA-IR

The relationship between m6A-modified CYTB RNA fragments in the mitochondria and the HOMA-IR level represents an emerging role of epitranscriptomic modifications in diabetes. HOMA-IR, a value calculated from fasting insulin and glucose levels, is used as an indicator of insulin resistance [114]. Clinical studies reveal a strong positive correlation between the plasma concentration of m6A-modified CYTB fragments and the HOMA-IR value, linking mitochondrial dysfunction to insulin resistance [115]. These circulating m6A-modified CYTB fragments have potential as non-invasive biomarkers for the metabolic status of individuals. However, additional mechanistic studies are needed to ascertain their causative role.

Additionally, clinical studies show that people with type 2 diabetes have significantly increased m6A-modified CYTB RNA fragments in their plasma compared to healthy individuals [115]. Levels of these modified CYTB RNA fragments increase when β-cell function declines, indicating that mitochondrial RNA modifications play a crucial role in metabolic homeostasis. With future validation and longitudinal studies, these fragments could have prognostic value to track the progression of diabetes. This suggests that they reflect changes in the metabolic function and have the potential to be used to categorize individuals by risk.

When m6A writers and readers in mitochondria are disrupted by knockout and pharmacological inhibition, the half-life of CYTB RNA decreases, and its fragmentation increases. This effect also leads to decreased activity of complex III in the electron transport chain and increased CYTB RNA debris in circulation, demonstrating that m6A is critical for maintaining the functional integrity of the mitochondrial respiratory chain [16]. However, the roles and regulation of other epitranscriptomic marks still need to be investigated. The higher expression level of m6A writers and readers in β-cells in the presence of glucotoxicity and oxidative stress indicates that the levels of this modification are highly regulated by metabolic demands. This may explain the altered levels of these fragments in disease states. Moreover, the plasma m6A-modified CYTB RNA fragments can also be used to indicate the status of mitochondria, especially under metabolic stress.

5.2. Targeting RNA-Modifying Enzymes in Islets

Targeting mitochondrial RNA-modifying enzymes to restore mitochondrial function and improve β-cell health can be achieved via small-molecule modulators and gene editing [116]. These options would address the underlying molecular root cause of β-cell failure by altering epitranscriptomic signatures, thus paving a therapeutic avenue in the pathogenesis of mitochondrial bioenergetics and diabetes.

5.2.1. Small-Molecule Modulators of RNA Modifications

Small-molecule agents targeted at RNA modification enzymes (methyltransferases (“writers”) and demethylases (“erasers”) could potentially be used to influence the epitranscriptomic landscape of mitochondrial RNAs in pancreatic β-cells, therefore impacting mitochondrial gene expression and energy metabolism, and indirectly regulating insulin secretion. To date, several studies indicate that small-molecule modulation of RNA modification enzymes can influence mitochondrial translation, OXPHOS complex assembly, and ATP production in pancreatic β-cells; however, a more complete understanding of how this affects mitochondrial function is required for future clinical trials [16,46].

In an attempt to restore m6A/m5C and improve mitochondrial RNA translation, specific small-molecule inhibitors and activators to control RNA-modifying enzymes have been explored. The experimental data potentially implies that restoring mitochondrial m6A/m5C RNA modification patterns restores mitochondrial translation accuracy and reverses energy imbalances in β-cells under metabolic stress [117]. These findings support the idea that restoring mitochondrial RNA modification patterns pharmacologically may represent an effective therapy to protect against β-cell dysfunction associated with diabetes [16].

Small-molecule regulators of RNA modification enzymes have been shown to influence insulin secretion in in vitro and in vivo models of diabetes. By directly affecting the metabolic coupling of β-cells, small-molecule modulators of RNA modifications have been shown to boost glucose-stimulated insulin release (GSIS) [118]. Further studies need to be carried out to fully understand the mechanism through which small-molecule regulation of RNA modification impacts β-cell metabolism and the links to improved insulin secretion.

It has been demonstrated that small-molecule modulators can significantly decrease oxidative stress in vivo and in vitro, thus restoring electron transport chain function and limiting ROS levels in diabetic conditions. By relieving β-cell oxidative stress induced by chronic hyperglycemia, small molecules may prolong the lifespan of pancreatic β-cells; however, the molecular mechanisms of oxidative damage in β-cells and how mitochondrial membrane stress contributes to mitotranscriptomic imbalance remain unknown [16,119].

A major hurdle in developing small-molecule-based therapies is balancing off-target effects while ensuring beneficial effects on mitochondrial RNA modification. Improving small-molecule bioavailability with specific targeting and delivery methods could prove beneficial for future clinical applications of these therapies. A number of advanced delivery strategies, such as mitochondria-targeted small-molecule formulations and delivery vehicles, are being actively explored to overcome this concern [16].

Preliminary studies in cell and animal models using RNA modification modulators such as small molecules have provided a proof-of-concept for their therapeutic efficacy, indicating the impact of pharmacologic m6A/m5C in mitochondria to enhance the OXPHOS efficiency, ATP production, and decrease the amount of oxidative stress. Further preclinical investigations of pharmacologic modulation for mitochondrial RNA modifications are required to demonstrate more broadly the efficacy of this mechanism for regulating human insulin secretion [77,120]. A recent study using a novel small molecule inhibitor and m6A small molecule activator in rat INS-1 β-cells cultured in high glucose and FFA showed improvements in both energy metabolism and mitochondrial protein synthesis. Future investigations are required to establish how these small-molecule modulators translate to improved glucose tolerance in vivo [77]. Research has suggested that altering mitochondrial RNA modification with pharmacological interventions can reduce oxidative stress in pancreatic β-cells. An increased level of mROS leads to oxidative stress, DNA damage, and ultimately the dysfunction of pancreatic β-cells in diabetes, indicating the significance of the regulation of ROS [77,121].

Strategies for early intervention of a small-molecule therapy that targets a particular pathway or gene represent an avenue of investigation, but require further optimization in terms of drug design and delivery. The delivery of exogenous small-molecule modulators to mitochondria represents one of the biggest challenges in the field of mitopharmaceuticals. This is primarily due to the dual-membrane structure of mitochondria. MITO-Porter is a newly developed system that delivers therapeutic antisense oligonucleotides (ODNs) into mitochondria by encapsulating them in a liposome. The efficacy and targeting specificity of the MITO-Porter system to mitochondria were demonstrated in β-cells. Experimental data reveal that it can significantly reduce mitochondrial-encoded genes and alter cell energy levels [95,122]. The results obtained by the MITO-Porter system suggest that the success of MITO-Porter is governed by the appropriate choice of particle size, surface charge, and a MITO-Porter liposome, which is capable of being adsorbed to the surface of mitochondrial outer membranes to ensure targeting specificity and avoid off-target effects [95]. MITO-Porter systems may be improved and further developed to allow clinical application in humans, in particular to prevent and/or treat various pathological conditions. Challenges regarding stability, bioavailability, immune system stimulation, and tissue-specific targeting will need to be overcome before MITO-Porter systems can be translated into human applications for the clinical delivery of oligonucleotides or small-molecule drugs to treat β-cell dysfunction in diabetes [95,123].

The potential off-target effect of RNA modification modulators is another major challenge in future translational work, as in long-term intervention for β-cell regeneration in diabetes, unintended effects on nuclear/cytoplasmic RNA modification pathways could result in unforeseen toxicity or immune response. The specificity and targeted activity of small-molecule inhibitors can be maximized by employing islet-targeted delivery strategies [16,85]. The rational design of small-molecule agents capable of selectively modulating mitochondrial RNA-modifying enzymes is a major focus of this field. Structural knowledge of RNA-modifying enzymes and structural and functional insight of binding domains offer great advantages for designing target-specific modulators. A high-throughput screening approach of chemical libraries for small-molecule inhibitors in combination with structural and functional analysis has been employed to discover candidate modulators with minimal off-target effect and good oral bioavailability for mitochondrial delivery to protect the viability of β-cells [16,124].

5.2.2. CRISPR-dCas13b Approaches for Epitranscriptomic Editing

CRISPR-dCas13b provides a programmable technology to edit mitochondrial RNA in pancreatic β-cells by fusing catalytically inactive Cas13b to RNA-modifying enzymatic domains and using sequence-specific RNAs to add or remove a chemical mark to mitochondrial transcripts. Caution must be taken when considering the efficacy and specificity of this technology, and the scalability needs to be examined. The ability of CRISPR-dCas13b to perform both gain- and loss-of-function experiments helps understand the function of understudied chemical marks. The ability to restore specific marks also creates the possibility of identifying novel drug targets in the mitotranscriptome, but whether this is also feasible in vivo through direct RNA modification remains unclear [18]. CRISPR-dCas13b allows us to study RNA modification as a dynamic cell response to metabolic stress, due to its reversibility and transcript-specificity as compared with gene-editing technology [125]. This is particularly useful for modeling β-cell adaptations and to test whether ATP deficit, chronic ROS production, or translation infidelity may be reversible by restoring the missing modifications in mitomRNAs. Furthermore, we can determine the long-term or acute changes in mRNA chemical marks during the pathogenesis of diabetes. CRISPR-dCas13b for the induction of RNA chemical modification is not applicable as a single and discrete tool. Instead, it should be combined with the ability of mapping or detecting these modifications at single-nucleotide resolution, such as the newly developed Nanopore and mass spectrometry strategies, to confirm that RNA was edited and modified. For instance, CRISPR-dCas13 b-induced RNA modification could be accompanied by oxidative or endoplasmic reticulum stress in order to investigate the RNA modification landscape of diabetes in β-cells to determine their involvement in adaptive changes. Thus, CRISPR-dCas13b enables us to integrate RNA chemical modification with other functional omics to identify novel therapeutic targets against mitochondrial dysfunction [18,86].

Mitochondrial import is a highly controlled process. This process can be artificially circumvented by a variety of methods for delivering molecules or chemicals of any charge and size directly to the mitochondria of any cell, ranging from liposomal methods to targeting to an internal matrix using charged groups. For instance, the MITO-Porter system utilizes liposomal technology with optimized size and charge to ensure sufficient encapsulation to avoid leakage, specific cell targeting, and ultimately efficient mitochondrial entry. The stability of MITO-Porters, as well as immunogenicity, are open question. Future research to reduce the size of Cas13b, for instance, will be a step closer to efficient mitochondrial-targeted delivery in vivo [95]. The use of the CRISPR-dCas13b system can involve several potential concerns that require thoughtful consideration. First, in a complex organism with highly metabolically active cells, it is difficult to completely control the risk of off-target chemical marking. Single-molecule sequencing and high-throughput screening are required to confirm specificity. As RNA chemical marking technology is very new, the long-term safety and clinical consequences in vivo need to be extensively studied. Other important considerations for this technique are its limited efficiency of transfection and ethical considerations, such as the risk of bias and disparities related to gender, age, and ethnicity. Ultimately, it is imperative that investigators work together to balance the potential benefits of this technology with the need for equity, informed consent, and transparent accountability [126,127].

An interesting use of RNA chemical marking techniques, in combination with the high-resolution mapping and monitoring in real time in β-cells, would be the development of individual-based or personalized treatments based on a patient’s mitochondrial RNA-modification fingerprint. The ability to perform on-the-fly, direct nanopore sequencing coupled to CRISPR-dCas13b-mediated modification could allow for individualized diagnosis and treatment within a medical visit, tailored specifically to each patient’s condition [16,18]. While these approaches are conceptually attractive, their translation into routine clinical practice remains highly challenging. At present, the isolation of β-cell-specific mitochondrial RNA from living patients is not feasible, as pancreatic β-cells are not readily accessible and mitochondrial enrichment typically requires invasive tissue sampling or ex vivo cell isolation. In addition, absolute quantification and high-resolution mapping of mitochondrial RNA modifications rely on technically demanding, time-intensive, and costly workflows that are incompatible with rapid clinical decision-making.

6. Conclusions

We conclude that mitochondrial RNA modifications are crucial players that ensure proper maintenance of mitochondrial functions. These modifications regulate mRNA stability and translational fidelity, which directly contribute to the appropriate assembly of oxidative phosphorylation protein complexes and redox balance in β-cells. Inherited and acquired defects in RNA-modifying enzymes result in deficient mitochondrial protein translation, deficient ATP production, and the induction of mitochondrial unfolded protein response and oxidative stress, causing β-cell dysfunction. Our findings reveal that a few cases of monogenic diabetes involve impaired mitochondrial RNA modification activities because of mutations in mitochondrial RNA-modifying enzymes. Several studies have reported the presence of fragmented mitochondrial RNA in the bloodstream, which may serve as a potential diagnostic and prognostic biomarker in patients with early stages of pre-diabetes. Current efforts are also focused on developing small-molecule modulators and CRISPR-dCas13b-based modulators of these processes in β-cells to correct the effects of RNA modification deficiencies.

The results of our review expand on the significance of epitranscriptomic studies, positioning the mito-epitranscriptome as a new control layer to the emerging paradigm in metabolic disease pathogenesis, in which the mitochondrion represents not only the executor but also the sensor. It contributes and complements conceptual frameworks based on genome- and proteome-wide analyses, demonstrating that RNA modification defects at defined modification sites can initiate or accelerate the development of metabolic disease. In addition, defining functional and molecular consequences of mitochondrial RNA modifications in β-cells helps to bridge our understanding of rare genetic (monogenic) diseases of mitochondrial energy metabolism with common multifactorial forms of metabolic disease and diabetes.

Although our review makes a step forward, further investigation of many aspects of mitochondrial RNA modification biology in β-cells is necessary. Limitations that have been identified in the current available evidence include difficulties in distinguishing mitochondrial RNA modifications and their modulators from their nuclear counterparts and difficulties in establishing long-term or causative links in vivo in humans due to incomplete or low-purity mitochondrial preparations, lack of human data on the identity, stability, and activity of ‘erasing’ enzymes that remove these modifications, challenges in integration with other omics analyses and other regulators of protein expression, and inability to identify and understand the reversibility or adaptability of these RNA modification changes under normal physiological or diabetic conditions.

Therefore, future investigations should address many aspects of the mito-epitranscriptome in β-cells. For example, the impact of genetic variation and environmental/nutritional factors should be further explored. The research can be extended by defining more diagnostic/prognostic mitochondrial RNA modifications through studies of people with diabetes. In vivo therapeutic strategies should be tested through the development of specific small-molecule or RNA modification modulators, CRISPR-dCas13b-based modulators, or improved islet- and mitochondria-targeted delivery systems. A high priority in the field must also be to improve methods for purifying and characterizing β-cell mitochondria. Finally, integrating the mito-epitranscriptome with other omics data will provide more in-depth systems-level insights into the role of RNA modification and metabolic dysregulation on β-cell functions.

In conclusion, this review aims to show the central role of mitochondrial RNA modifications on β-cell function and in the pathogenesis of diabetes. Our results show how the defective control of these RNA modifications could result in mitochondrial failure, thus generating oxidative stress that leads to β-cell death. This review also emphasizes several exciting translational opportunities to identify circulating mitochondrial RNA fragments that serve as a useful prognostic/diagnostic biomarker to better stratify pre-diabetic patients. In addition, a variety of therapeutic approaches currently in preclinical development hold great promise for curing these conditions in humans. However, further studies are necessary to identify a complete atlas of dynamic and reversible mitochondrial RNA modifications in response to diverse signals. Future research should prioritize clarifying the causal versus consequential roles of mitochondrial RNA modifications in β-cell dysfunction using temporally resolved and cell-specific models. Advances in high-resolution and single-molecule epitranscriptomic profiling will be essential to define combinatorial RNA modification patterns and their functional interactions in mitochondrial gene regulation. Additionally, integrating epitranscriptomic data with mitochondrial functional assays may facilitate the identification of clinically relevant biomarkers and therapeutic targets for diabetes.