MicroRNA chemical modifications in post-transcriptional gene silencing and human diseases.

Recent findings indicate that small RNAs undergo various modifications, which profoundly impact their roles in diverse biological processes. 38 Here, we outline the effects of miRNA modifications in their biogenesis and activities, emphasizing different types of modifications, including methylation, pseudouridylation, and other modifications, such as phosphorylation, RNA editing (A to I), acetylation, and oxidation. The chemical structure of these modifications is also illustrated in Figure 2 . Figure 2 Different types of chemical modifications in miRNA Several types of chemical modifications in miRNA have been discovered. These chemical modifications include (A) 2′-O-methylation (Nm), (B) N6-methyladenosine (m6A), (C) 5-methylcytidine (m5C), (D) N7-methylguanosine (m7G), (E) pseudouridylation (Ψ), (F) phosphorylation, (G) adenosine (A) to inosine (I), (H) N4-acetylcytidine (ac4C), and (I) 8-oxoguanine (o8G). Different types of chemical modifications in miRNA Several types of chemical modifications in miRNA have been discovered. These chemical modifications include (A) 2′-O-methylation (Nm), (B) N6-methyladenosine (m6A), (C) 5-methylcytidine (m5C), (D) N7-methylguanosine (m7G), (E) pseudouridylation (Ψ), (F) phosphorylation, (G) adenosine (A) to inosine (I), (H) N4-acetylcytidine (ac4C), and (I) 8-oxoguanine (o8G). A necessary miRNA modification is 2′-O-methylation (Nm) ( Figure 3 A), a strongly conserved and abundant modification in which a hydrogen atom (-H) on the 2′-hydroxyl group (-OH) of the ribose is replaced with a methyl group (-CH 3 ). 39 Nm modifications occur in multiple RNA types, including small nuclear RNAs (snRNAs), ribosomal RNAs (rRNAs), transfer RNAs (tRNAs), and mRNAs. 40 Additionally, Nm is found at the 3′ ends of small RNAs, namely miRNAs and small interfering RNAs (siRNAs), in plants, 41 as well as in AGO2-loaded siRNAs and miRNAs in flies and PIWI-interacting RNAs (piRNAs) in animals. 42 In Drosophila , piRNAs are 2′-O-methylated at their 3′ ends by Pimet, the fly homolog of Arabidopsis HUA ENHANCER 1 (HEN1) methyltransferase. This modification stabilizes piRNAs and is crucial for their function in transposon silencing. Pimet specifically methylates single-stranded piRNAs associated with PIWI proteins, distinguishing it from plant HEN1, which methylates miRNA duplexes. 43 Throughout small RNA maturation in humans, Nm modifications occur at the 3′ end nucleotides of these RNAs after being processed by Dicer or PIWI proteins, 44 and this modification is crucial for forming well-established structures that resist 3′-5′ truncation and 3′-uridine-triggered degradation. 45 Moreover, Nm is involved in enhancing miRNA stability by affecting thermodynamic properties, such as base stacking and structural rigidity. 46 Figure 3 Stabilization of miRNAs through 2′-O-methylation and N6-methyladenosine in miRNA maturation and stability (A) The figure illustrates the role of 2′-O-methylation (Nm) in their stability and function and the enhancement of miRNA binding to AGO through Nm, catalyzed by HEN1. 227 This modification, which modifies pre-miRNA, protects the miRNA from 3′ to 5′ exonuclease activity and uridine-triggered degradation, thereby increasing its stability and functional efficacy. (B) (Left) The figure highlights the role of N6-methyladenosine (m6A) in regulating miRNA maturation and stability. 38 METTL3 and METTL14 catalyze m6A. This modification that occurs on the flanking sequence of pri-miRNA, before the attachment of the microprocessor complex, enhances miRNA processing by promoting DGCR8 recognition, accelerating pri-miRNA cleavage into pre-miRNA. (Right) It also occurs on the terminal loop of pre-miRNA after the cleavage of pri-miRNA into pre-miRNA by the microprocessing complex. The presence of m6A on pri-miRNA or pre-miRNA enhances miRNA maturation and stabilizes miRNA expression. In this process, HNRNPA2/B1 binds to m6A-modified pri-miRNA or pre-miRNA to enhance processing. ALKBH5 and FTO are demethylase enzymes that can remove m6A modifications. YTHDF1/2/3 and YTHDC1/2 are readers that recognize m6A modifications, influencing miRNA stability and function. IGF2BP1/2/3 are proteins that stabilize m6A-modified miRNAs. HEN1, HUA ENHANCER 1; Me, methyl functional group; m6A, N6-methyladenosine; DGCR8, DiGeorge syndrome critical region 8; ALKBH5, AlkB homolog 5; FTO, fat mass and obesity-associated protein; HNRNPA2/B1, heterogeneous nuclear ribonucleoprotein A2/B1; YTHDF1/2/3, YTH domain family proteins 1/2/3; YTHDC1/2, YTH domain-containing proteins 1/2; IGF2BP1/2/3, insulin-like growth factor 2 mRNA-binding proteins 1/2/3. Stabilization of miRNAs through 2′-O-methylation and N6-methyladenosine in miRNA maturation and stability (A) The figure illustrates the role of 2′-O-methylation (Nm) in their stability and function and the enhancement of miRNA binding to AGO through Nm, catalyzed by HEN1. 227 This modification, which modifies pre-miRNA, protects the miRNA from 3′ to 5′ exonuclease activity and uridine-triggered degradation, thereby increasing its stability and functional efficacy. (B) (Left) The figure highlights the role of N6-methyladenosine (m6A) in regulating miRNA maturation and stability. 38 METTL3 and METTL14 catalyze m6A. This modification that occurs on the flanking sequence of pri-miRNA, before the attachment of the microprocessor complex, enhances miRNA processing by promoting DGCR8 recognition, accelerating pri-miRNA cleavage into pre-miRNA. (Right) It also occurs on the terminal loop of pre-miRNA after the cleavage of pri-miRNA into pre-miRNA by the microprocessing complex. The presence of m6A on pri-miRNA or pre-miRNA enhances miRNA maturation and stabilizes miRNA expression. In this process, HNRNPA2/B1 binds to m6A-modified pri-miRNA or pre-miRNA to enhance processing. ALKBH5 and FTO are demethylase enzymes that can remove m6A modifications. YTHDF1/2/3 and YTHDC1/2 are readers that recognize m6A modifications, influencing miRNA stability and function. IGF2BP1/2/3 are proteins that stabilize m6A-modified miRNAs. HEN1, HUA ENHANCER 1; Me, methyl functional group; m6A, N6-methyladenosine; DGCR8, DiGeorge syndrome critical region 8; ALKBH5, AlkB homolog 5; FTO, fat mass and obesity-associated protein; HNRNPA2/B1, heterogeneous nuclear ribonucleoprotein A2/B1; YTHDF1/2/3, YTH domain family proteins 1/2/3; YTHDC1/2, YTH domain-containing proteins 1/2; IGF2BP1/2/3, insulin-like growth factor 2 mRNA-binding proteins 1/2/3. The initially discovered 2′-O-methyltransferase responsible for Nm modification of small RNAs was HEN1, found in Arabidopsis as a methylase for miRNAs and siRNAs. 38 HEN1 homologs were later identified in other plant species and animals, where they methylate piRNAs, and in Drosophila , where they modify AGO2-associated small RNAs, including miRNA. 47 Research has shown that the knockout or mutation of HEN1 in Arabidopsis leads to increased heterogeneous 3′ ends and poly-U sequences, which destabilize RNAs, causing abnormal miRNA lengths and reduced levels. 48 This effect is linked to enzymes such as HEN1 SUPPRESSOR1 (HESO1) 49 and UTP:RNA uridylyltransferase 1 (URT1), 50 which are involved in the 3′ uridylation of small RNAs. Protein HEN1 contains a double-stranded RNA (dsRNA)-binding domain at its N terminus and a catalytic domain at its C terminus, 51 responsible for transferring methyl groups from S-adenosylmethionine (SAM) to miRNA/miRNA∗ duplexes. 52 In Arabidopsis HEN1, five sequential residues (Asp-Phe-Gly-Cys-Gly) within the catalytic domain form a SAM-binding motif essential for methylation. 43 In HEN1 mutants, the loss of HEN1-mediated methylation results in the 3′ end uridylation of miRNAs and siRNAs, leading to their destabilization and reduced abundance. 53 Thus, HEN1 methylates miRNAs and siRNAs at the 2′-OH of their 3′-terminal nucleotide, which protects them from 3′ end uridylation and degradation, thereby ensuring their stability in vivo . 54 HEN1 plays pleiotropic roles in plant development, including regulating organ size, leaf morphology, and floral transition. Mutations in HEN1 cause reduced organ size, curled leaves, delayed flowering, and fertility defects. 55 Furthermore, mutations in the HEN1 gene are associated with accelerated neuronal degeneration and a shorter life expectancy, suggesting that Nm modification of miRNAs may influence aging-associated cell signaling pathways. 56 These results highlight the significance of HEN1 and its orthologs in stabilizing germline miRNAs across diverse plant and animal species, with species-specific effects. Interestingly, Nm modifications have similarly been discovered in mature miRNAs in mammals. For example, distinct 3′-terminal Nm patterns have been observed in miRNAs such as miR-21-5p from non-small cell lung cancer (NSCLC) cells and their paired normal tissues. This methylation appears to protect miR-21-5p from deterioration by the enzyme PNPase 1 (PNPT1), leading to its prolonged loading onto AGO2, which boosts the inhibitory effect on programmed cell death 4 (PDCD4) expression. This suggests that Nm modification enhances the stability of miRNA and shields them from breakdown by enzymes such as PNPT1. 57 Mutations in Hen1 and Ago2, which lead to the absence of Nm on miRNAs, have been shown to accelerate neurodegeneration and reduce lifespan. These findings suggest that the age-related increase in Nm of miRNA may impact age-related biological processes. It is investigated that the role of Nm at the 3′ end of miRNAs in Drosophila is mediated by the differential distribution of miRNAs between Ago1 and Ago2 during aging. Furthermore, it is highlighted that this regulatory mechanism, alongside other Ago2-mediated functions, may contribute to the modulation of age-associated events. 56 SNORD11B facilitates Nm at the G225 site of pri-let-7a through its canonical motif. This modification promotes the degradation of pri-let-7a, disrupts its interaction with DGCR8, and reduces the expression level of the mature tumor-repressor miRNA let-7a-5p. Consequently, the downregulation of let-7a-5p leads to the overexpression of oncogene translation, contributing to tumor progression. 58 m6A ( Figure 3 B), initially identified in the 1970s, is a methylation modification that occurs at the sixth nitrogen (N) atom of adenine (A). 59 The m6A modification is catalyzed by the SAM-binding proteins METTL3 and METTL14. 60 Importantly, additional cofactors such as METTL16, 61 Wilms tumor-associated protein (WTAP), 62 RNA-binding motif protein 15 (RBM15), 63 KIAA1429 (also known as VIRMA), and zinc finger CCCH domain-containing protein 13 (ZC3H13) 64 are recognized as crucial for the catalytic activity of m6A methyltransferases. Furthermore, the proteins fat mass and obesity-associated protein (FTO) and ALKB homolog 5 (ALKBH5) have been recognized as m6A demethylases. 65 Nonetheless, members of the heterogeneous nuclear ribonucleoprotein protein families (including HNRNPC and HNRNPA2/B1), members of YT521-B homology domain family 1/2/3 (YTHDF1/2/3), 66 eIF3, 63 YT521-B homology domain-containing proteins 1/2 (YTHDC1/2), 66 and insulin-like growth factor-2 mRNA-binding proteins 1/2/3 (IGF2BP1/2/3) 67 have all been identified as reader proteins that detect m6A methylation. In 2015, it was demonstrated that changes in the levels of the methyltransferase METTL3 influenced both mature miRNA expression and unprocessed pri-miRNA, in addition to its established role in mRNA regulation. 38 This finding suggests a link between m6A modification and miRNA biosynthesis. 68 As previously outlined, the initial step of miRNA biosynthesis involves the binding and recognition of pri-miRNA by the dsRNA-binding protein DGCR8 at the junction between the pri-miRNA hairpin stem and the adjacent single-stranded RNA within the nucleus. This interaction recruits the RNase III endonuclease Drosha, forming a microprocessor complex that cleaves pri-miRNA to generate a pre-miRNA. The pre-miRNA is then bound by XPO5 and transported to the cytoplasm, where Dicer processes it into mature miRNA. Notably, it is recognized that this mechanism relies on the m6A modification of RNA. 38 METTL3 can methylate pri-miRNAs, enabling recognition by HNRNPA2/B1, which then recruits and interacts with DGCR8 to bind pri-miRNA, thereby accelerating miRNA production. This highlights the significance of m6A as an essential post-transcriptional modification in the efficient biosynthesis of miRNAs within cells. 38 This discovery sheds light on the broader participation of m6A in molecular processes and its involvement in the pathogenesis of human disorders. For instance, in bladder cancer, a molecular pathway involving an elevated expression level of METTL3 facilitates DGCR8’s identification of m6A-modified pri-miR221/222, promoting the maturation of miR221/222. This leads to the suppression of phosphatase and tensin homolog (PTEN), a recognized target of miR-221/222, which contributes to abnormal cell proliferation. 69 The involvement of METTL14 in spinal tissue degeneration has been documented to regulate m6A modification of pri-miR-34a-5p, enhancing DGCR8 identification and increasing miR-34a-5p levels. This miRNA then targets silent information regulator sirtuin 1 (SIRT1), promoting TNF-α-induced cell senescence in the nucleus pulposus of intervertebral disc tissue. 70 Also, METTL14 associates with the microprocessor protein DGCR8 and enhances the processing of pri-miR-126 in an m6A-dependent way, leading to miR-126 maturation that counteracts the suppressive impact of METTL14 on tumor metastasis. 71 m7G modification of pri-let-7e by METTL1 results in mature miRNA let-7e, leading to repression of the migration of lung cancer cells. 72 In intestinal inflammation and tumor development caused by enterotoxigenic Bacteroides fragilis, the METTL14-dependent m6A modification facilitated the splicing and production of miR-149-3p, which in turn regulates T helper type 17 (Th17) differentiation. 73 The increased levels of METTL14 in colorectal cancer cells enhance the differentiation of CD4+ T cells into Th17 cells through exosomes containing miR-149-3p. 74 Besides the role of methyltransferases, m6A demethylases and reader proteins also influence miRNA-related biological mechanisms. For instance, in lung cancer, the m6A reader HNRNPA2B1 collaborates with LINC01234 to recruit DGCR8, thereby enhancing the accumulation of miR-106b-5p. This miRNA downregulates cryptochrome circadian regulator 2 (CRY2), thereby increasing the levels of c-Myc and promoting the development of lung cancer. 75 In the activation of lung fibroblasts and silica-induced lung fibrosis, the demethylase ALKBH5 demethylates pri-miR-320a-3p, preventing its engagement with DGCR8 and blocking the maturation of miR-320a-3p. This dysregulation affects target genes, such as forkhead box M1 (FOXM1), which ultimately leads to the deterioration of lung tissue. 76 The functions of small RNA modifications are involved in numerous cellular processes, both in maintaining homeostasis and in the context of human diseases. For example, the suppression of m6A modification has been shown to result in the downregulation of miR-374c-5p by reducing the stability of its RNA transcript. The findings collectively illustrate that m6A modification on pri-miRNA-374c inhibits the maturation of miR-374c-5p, consequently leading to the activation of GRM3 expression. This regulatory cascade has been implicated in promoting breast cancer (BC) cell metastasis in response to cadmium exposure. 77 HNRNPA2B1 recognizes the m6A modification site on pri-miR-106b, thereby facilitating the maturation of miR-106b-5p. The mature miR-106b-5p subsequently targets and suppresses the expression of secreted frizzled-related protein 2 (SFRP2), thereby inducing Wnt/β-catenin signaling. Collectively, these findings suggest that HNRNPA2B1 facilitates SFRP2 inhibition and Wnt/β-catenin pathway activation through m6A-regulated miR-106b-5p maturation, thereby enhancing stemness properties and promoting the progression of lung adenocarcinoma (LUAD). 78 METTL3 promotes the expression level of miR-222-3p by catalyzing the m6A modification of pri-miR-222-3p, thereby facilitating its maturation. The resulting miR-222-3p targets and negatively regulates serine/threonine stress kinase 4 (STK4). Silencing of METTL3 has been shown to upregulate the expression of STK4 by suppressing miR-222-3p, thereby inhibiting the malignant phenotypes of thyroid cancer (TC) cells and reducing tumor development and lung metastasis in nude mouse models. 79 Beyond its roles as a methylation reader, writer, and eraser, m6A modification may also aid in miRNA maturation by enhancing Dicer-mediated splicing of pre-miRNAs. In NSCLC, METTL3 has been found to promote the splicing of pre-miR-143-3p in an m6A-dependent manner, thereby facilitating the biogenesis of miR-143-3p. This, in turn, drives lung cancer metastatic invasion and angiogenic processes through the expression malfunction of vasohibin 1 (VASH1). 80 This discovery opens a promising direction for research on developing new treatments targeting NSCLC and brain metastasis. Remarkably, despite no changes in the primary transcript, several mature miRNAs have been demonstrated to decrease in cells following m6A demethylase FTO downregulation, suggesting that m6A may negatively impact miRNA biogenesis. 81 Supporting this, the methyltransferase NSUN2 hinders the maturation of pri-miR-125b into miR-125b, thereby reducing the expression levels of miR-125b. 82 This NSUN2-driven downregulation of miR-125b is potentially influenced by protease-activated receptor 2 (PAR2) and can enhance rectal cancer metastasis by disrupting the expression of the GRB2-associated binding protein 2 (Gab2) gene. 83 In contrast, in endocrine-resistant BC cells, HNRNPA2/B1 plays a more intricate role in miRNA biogenesis. Overexpression of HNRNPA2/B1 in these cells results in the upregulation of miR-671-3p, miR-1266-5p, and miR-1268a while reducing the levels of miR-222, miR-29a-3p, and miR-29b-3p, collectively contributing to decreased sensitivity to cancer drugs, such as 4-hydroxytamoxifen and fulvestrant. 84 m6A methylation plays a crucial role in RNA metabolism, including the decay of mRNAs and circular RNAs (circRNAs) through pathways such as deadenylation and endoribonucleolytic cleavage. Chen and colleagues imply that m6A modifications are enriched at miRNA target sites and can influence mRNA stability via miRNA pairing, although the mechanisms linking m6A to miRNA homeostasis remain unclear. Further research is needed to determine whether m6A directly regulates pri-miRNAs or the expression of miRNA-processing proteins. 85 miRNA biogenesis is subject to precise regulation at various stages, including transcription, processing, incorporation into the RISC, and subsequent decay. 1 Fluctuations can influence each of these phases in the levels of regulatory molecules and modifications to the miRNA’s sequence or structure. Key alterations, such as miRNA tailing, 86 RNA editing, 87 and RNA O-methylation, 41 have been extensively investigated, with significant attention paid to their roles in miRNA function. However, the role of N6-adenosine methylation in miRNA biology remains less explored. Recent studies suggest that the tRNA cytosine-methyltransferase NSun2 can methylate miR-125b at adenosine residues, 88 further highlighting the importance of RNA modifications in miRNA regulation. Moreover, m6A marks have been identified on mature miRNAs, 89 but the source of these modifications and their functional impact remain unclear. Overall, while m6A is essential for miRNA processing, its role and effects on miRNAs are still not fully elucidated. One other RNA modification is 5-methylcytidine (m5C) ( Figure 4 A), where a methyl group is inserted into the fifth carbon atom (C5) of cytosine in RNA molecules. 90 m5C modification is found in miRNAs, as well. 91 This modification has been reported to disrupt miRNA/mRNA pairing, leading to a loss of gene-silencing function by the miRNAs. For example, m5C modification disrupts the tumor-suppressive activity of miRNA-181a-5p and is associated with unfavorable prognosis in glioblastoma cases. 92 Furthermore, m5C modification can lead to structural alterations to the RISC. For instance, m5C at position 9 of miR-200c-3p, near its interaction with the RISC complex, disrupts the hydrogen bond between the miRNA and AGO Ser220, altering the collaboration between guanine at position 8 of the miRNA and Arg761 in AGO during translocation. 89 Figure 4 5-methylcytidine and N7-methylguanosine in miRNA biogenesis and function (A) This schematic representation illustrates the impact of 5-methylcytidine (m5C) modification on miRNA processing, 38 stability, and gene-silencing activity. NSUN2, an RNA methyltransferase, catalyzes the methylation of cytidine residues, introducing the m5C modification in mature miRNAs. This modification enhances miRNA stability and facilitates its loading into the RISC, where it interacts with the AGO protein. NSUN2-mediated methylation promotes the recruitment of YBX1, which in turn inhibits mature miRNA gene-silencing function. This modification disrupts the pairing of miRNA and mRNA, leading to impaired gene silencing. (B) The figure underscores the regulatory role of N7-methylguanosine (m7G), a post-transcriptional modification that influences miRNA processing, stability, and functionality. 72 The pathway initiates in the nucleus, where RNA Pol II transcribes the miRNA gene, generating pri-miRNA. m7G modification occurs at G-rich sites within pri-miRNAs, catalyzed by METTL1 and WDR4. This modification destabilizes G-quadruplex structures, enhancing miRNA processing and maturation. The m7G modification, catalyzed by the METTL1/WDR4 complex, plays a critical role in miRNA regulation through multiple mechanisms, including destabilization of G-quadruplex structures, facilitating pri-miRNA recognition and efficient processing by Drosha and DGCR8, and enhancement of miRNA stability and functionality, ensuring effective incorporation into the RISC and robust gene-silencing activity, which highlight the functional significance of m7G modifications in modulating miRNA biogenesis and post-transcriptional gene regulation. m5C, 5-methylcytidine; NSUN2, NOP2/Sun RNA methyltransferase family member 2; YBX1, Y-box binding protein 1; m7G, N7-methylguanosine; RNA Pol II, RNA polymerase II; DGCR8, DiGeorge syndrome critical region 8; XPO5, exportin 5; TRBP, TAR RNA-binding protein; AGO, Argonaute; RISC, RNA-induced silencing complex; METTL1, methyltransferase-like 1; WDR4, WD repeat domain 4. 5-methylcytidine and N7-methylguanosine in miRNA biogenesis and function (A) This schematic representation illustrates the impact of 5-methylcytidine (m5C) modification on miRNA processing, 38 stability, and gene-silencing activity. NSUN2, an RNA methyltransferase, catalyzes the methylation of cytidine residues, introducing the m5C modification in mature miRNAs. This modification enhances miRNA stability and facilitates its loading into the RISC, where it interacts with the AGO protein. NSUN2-mediated methylation promotes the recruitment of YBX1, which in turn inhibits mature miRNA gene-silencing function. This modification disrupts the pairing of miRNA and mRNA, leading to impaired gene silencing. (B) The figure underscores the regulatory role of N7-methylguanosine (m7G), a post-transcriptional modification that influences miRNA processing, stability, and functionality. 72 The pathway initiates in the nucleus, where RNA Pol II transcribes the miRNA gene, generating pri-miRNA. m7G modification occurs at G-rich sites within pri-miRNAs, catalyzed by METTL1 and WDR4. This modification destabilizes G-quadruplex structures, enhancing miRNA processing and maturation. The m7G modification, catalyzed by the METTL1/WDR4 complex, plays a critical role in miRNA regulation through multiple mechanisms, including destabilization of G-quadruplex structures, facilitating pri-miRNA recognition and efficient processing by Drosha and DGCR8, and enhancement of miRNA stability and functionality, ensuring effective incorporation into the RISC and robust gene-silencing activity, which highlight the functional significance of m7G modifications in modulating miRNA biogenesis and post-transcriptional gene regulation. m5C, 5-methylcytidine; NSUN2, NOP2/Sun RNA methyltransferase family member 2; YBX1, Y-box binding protein 1; m7G, N7-methylguanosine; RNA Pol II, RNA polymerase II; DGCR8, DiGeorge syndrome critical region 8; XPO5, exportin 5; TRBP, TAR RNA-binding protein; AGO, Argonaute; RISC, RNA-induced silencing complex; METTL1, methyltransferase-like 1; WDR4, WD repeat domain 4. m7G ( Figure 4 B) is a modification where the seventh nitrogen atom of guanine in RNA is methylated, and it has been observed in rRNAs, tRNAs, mRNAs, and miRNAs. 72 This m7G modification is catalyzed intracellularly by the proteins METTL1 and WD repeat domain 4 (WDR4). 93 The m7G modification occurs at G-rich sites within miRNA, facilitated by the RNA methyltransferase complex METTL1/WDR4. This modification influences non-canonical base pairing in pri-miRNAs, altering the resilience of secondary structures, particularly G-quadruplexes. By destabilizing these G-quadruplex structures, m7G modification enhances the processing efficiency of pri-miRNAs, facilitating their maturation into pre-miRNAs and, ultimately, into functional miRNAs. 72 For instance, METTL1 binds directly to miRNA precursors, such as pri-let-7e, and mediates m7G modification, which promotes their maturation into pre-miRNAs and mature miRNAs. The m7G modification, mapped to G11 in mature let-7e-5p, enhances processing by destabilizing inhibitory G-quadruplex structures in precursors. In lung cancer cells, METTL1 depletion increases high-mobility group protein 2 (HMGA2) protein levels and enhances cell migration, phenotypes rescued by wild-type METTL1 but not its catalytically inactive mutant. Transfection of mature let-7e-5p reduces HMGA2 levels, confirming METTL1’s role in regulating let-7e-dependent migration suppression. The METTL1/WDR4 complex methylates pri-let-7e in vitro , and m7G-modified pri-miRNAs are processed more efficiently by Drosha. 72 m7G methylation of miRNAs functions as a tumor-suppressor mechanism. This modification facilitates the conversion of pri-miRNA to pre-miRNA, thereby accelerating miRNA maturation and enhancing its processing efficiency. As a result, m7G methylation downregulates target gene expression in lung cancer, contributing to tumor suppression and suppressing oncogene expression, thereby restraining the proliferation and differentiation of tumor cells. Pseudouridine (Ψ) ( Figure 5 A), a structural isomer of uridine, commonly referred to as 5-ribosyluracil, 94 was first reported in the 1950s and is recognized as the most common form of post-transcriptional RNA modification observed throughout all kingdoms of life. Pseudouridine can be observed in tRNAs, rRNAs, snRNAs, and mRNAs. 95 It has been demonstrated that the pseudouridine synthase PUS10 influences miRNA biogenesis, as its depletion results in reduced mature miRNA levels and the accumulation of pri-miRNAs. Notably, this process occurs independently of PUS10’s catalytic activity. 96 In a separate instance, TruB1, a key mammalian pseudouridine synthase, can bind to the stem-loop of pri-let-7, thereby enhancing its interaction with the miRNA processor protein DGCR8, which promotes the maturation of the let-7 miRNA family, implicated in inhibiting cell proliferation. 97 These results suggest that the ability of cells to synthesize pseudouridine and its presence on various small RNAs are critical for modulating important biological processes. Figure 5 Pseudouridylation and phosphorylation in miRNA processing and stability (A) The figure provides a detailed representation of the miRNA biogenesis pathway, with a specific focus on the role of pseudouridylation, a critical post-transcriptional modification, in modulating miRNA processing and stability. In the nucleus, RNA Pol II transcribes the miRNA gene, yielding pri-miRNA, which undergoes pseudouridylation mediated by PUS10 and TruB1. 96 , 97 This modification stabilizes secondary structures and enhances the processing of miRNA. Pseudouridylation influences miRNA function by facilitating miRNA maturation, enhancing the efficiency of pri-miRNA processing into pre-miRNA, stabilizing miRNA secondary structures, preserving their integrity, and ensuring proper functionality in gene regulatory networks. (B) The figure highlights the regulatory influence of phosphorylation, a key post-transcriptional modification, on miRNA processing and stability. 228 Phosphorylation of miRNAs occurs at the 5′ end, where a phosphate group is added, enhancing their stability and interaction with the Ago protein. This modification is crucial for seed-based targeting, as the phosphorylated 5′ end improves miRNA binding to the 3′ UTR of target mRNAs, leading to efficient suppression of mRNA and protein expression. Additionally, phosphorylation enhances the potency of single-stranded miRNAs, increasing their effectiveness in mRNA knockdown. Substituting or removing the phosphate group at key positions, such as position 2, significantly reduces miRNA activity, emphasizing the importance of precise phosphorylation in miRNA-mediated gene regulation. RNA Pol II, RNA polymerase II; PUS10, pseudouridine synthases 10; TruB1, TruB pseudouridine synthase family member 1; UTR, untranslated region; P, phosphate. Pseudouridylation and phosphorylation in miRNA processing and stability (A) The figure provides a detailed representation of the miRNA biogenesis pathway, with a specific focus on the role of pseudouridylation, a critical post-transcriptional modification, in modulating miRNA processing and stability. In the nucleus, RNA Pol II transcribes the miRNA gene, yielding pri-miRNA, which undergoes pseudouridylation mediated by PUS10 and TruB1. 96 , 97 This modification stabilizes secondary structures and enhances the processing of miRNA. Pseudouridylation influences miRNA function by facilitating miRNA maturation, enhancing the efficiency of pri-miRNA processing into pre-miRNA, stabilizing miRNA secondary structures, preserving their integrity, and ensuring proper functionality in gene regulatory networks. (B) The figure highlights the regulatory influence of phosphorylation, a key post-transcriptional modification, on miRNA processing and stability. 228 Phosphorylation of miRNAs occurs at the 5′ end, where a phosphate group is added, enhancing their stability and interaction with the Ago protein. This modification is crucial for seed-based targeting, as the phosphorylated 5′ end improves miRNA binding to the 3′ UTR of target mRNAs, leading to efficient suppression of mRNA and protein expression. Additionally, phosphorylation enhances the potency of single-stranded miRNAs, increasing their effectiveness in mRNA knockdown. Substituting or removing the phosphate group at key positions, such as position 2, significantly reduces miRNA activity, emphasizing the importance of precise phosphorylation in miRNA-mediated gene regulation. RNA Pol II, RNA polymerase II; PUS10, pseudouridine synthases 10; TruB1, TruB pseudouridine synthase family member 1; UTR, untranslated region; P, phosphate. After the passenger strand is removed through cleavage or helicase activity, the guide strand pairs with complementary mRNAs. Research reveals that the primary mode of target identification involves the miRNA seed site (positions 2–8 at the 5′ end of the guide strand) binding to the 3′ UTR of the target mRNA. This connection typically leads to the suppression of both mRNA and protein levels. 98 Single-stranded miRNAs that are 5′ phosphorylated ( Figure 5 B) and contain 2′-fluoro modifications exhibit significant activity in cells, as demonstrated by mRNA knockdown assays. The study indicates that 5′ phosphorylation is crucial for enhancing Ago-based activity, with maximum activity observed when the single strands are phosphorylated and heavily modified with 2′-fluoro-ribose. Seed-based targeting, a hallmark of miRNA activity, was confirmed through microarray analysis, where downregulated genes exhibited complementary matches to the miRNA seed region. Notably, the substitution of a 2′-fluoro group with a 2′-O-methyl (2′-Ome) group at position 2 of the guide strand resulted in a dramatic loss of potency, underscoring the importance of specific modifications for functionality. These findings suggest that the combination of 5′ phosphorylation and 2′-fluoro modifications stabilizes the miRNA’s interaction with AGO proteins, thereby enhancing its regulatory effects. 99 A-to-I RNA editing ( Figure 6 ) introduces diversity in RNA and protein in higher eukaryotes by specifically modifying both coding and non-coding regions within nuclear transcripts. 100 The enzymes that catalyze A-to-I editing, known as ADARs, are widely expressed in mammals and particularly target partially dsRNA configurations, modifying individual adenosines based on the surrounding structure and sequence context. 101 While extensive editing occurs in long dsRNAs, 102 RNA duplexes containing bulges and loops undergo site-specific editing, as seen in numerous nuclear-encoded pre-mRNAs. 101 The predicted stem-loop configuration of all miRNA precursors resembles the partially double-stranded fold-back configuration characteristic of recognized ADAR substrate molecules, 103 suggesting that ADARs in vivo could target miRNA precursors. Since ADARs are mainly nuclear enzymes, the most probable candidates for miRNA editing are pri-miRNA and pre-miRNA precursor molecules before they are exported from the nucleus. ADAR-mediated editing of miRNA precursors could considerably influence the analysis, biogenesis, and activity of miRNAs. A-to-I modification of miRNA precursors may alter the processing speed and outcome or potentially block further maturation altogether. 104 Figure 6 RNA editing in miRNAs: Influence on biogenesis and target specificity The figure depicts the biogenesis pathway of miRNAs, emphasizing the role of RNA editing, specifically adenosine-to-inosine (A-to-I) editing, in regulating miRNA processing, stability, and target specificity. 229 The process begins in the nucleus, where RNA Pol II transcribes the miRNA gene to generate pri-miRNA. A-to-I editing can occur on pri-miRNA or pre-miRNA, modifying their structure and function and impacting miRNA processing. Edited miRNAs may be degraded or undergo altered processing, which can affect their maturation. (1) A-to-I editing by ADAR on pri-miRNA inhibits the pre-miRNA generation and leads to the degradation of pri-miRNA by TSN. (2) A-to-I editing on pri-miRNA might lead to retargeting. After complete processing of pri-miRNA to mature miRNA, this modification can alter the target specificity of miRNAs and lead to the regulation of different sets of genes. (3) A-to-I editing on pri- or pre-miRNA eventually impedes the RISC and Ago assembly and generation of mature miRNA. RNA Pol II, RNA polymerase II; DGCR8, DiGeorge syndrome critical region 8; AGO, Argonaute; RISC, RNA-induced silencing complex; ADAR, adenosine deaminases acting on RNA; TSN, Tudor-SN. RNA editing in miRNAs: Influence on biogenesis and target specificity The figure depicts the biogenesis pathway of miRNAs, emphasizing the role of RNA editing, specifically adenosine-to-inosine (A-to-I) editing, in regulating miRNA processing, stability, and target specificity. 229 The process begins in the nucleus, where RNA Pol II transcribes the miRNA gene to generate pri-miRNA. A-to-I editing can occur on pri-miRNA or pre-miRNA, modifying their structure and function and impacting miRNA processing. Edited miRNAs may be degraded or undergo altered processing, which can affect their maturation. (1) A-to-I editing by ADAR on pri-miRNA inhibits the pre-miRNA generation and leads to the degradation of pri-miRNA by TSN. (2) A-to-I editing on pri-miRNA might lead to retargeting. After complete processing of pri-miRNA to mature miRNA, this modification can alter the target specificity of miRNAs and lead to the regulation of different sets of genes. (3) A-to-I editing on pri- or pre-miRNA eventually impedes the RISC and Ago assembly and generation of mature miRNA. RNA Pol II, RNA polymerase II; DGCR8, DiGeorge syndrome critical region 8; AGO, Argonaute; RISC, RNA-induced silencing complex; ADAR, adenosine deaminases acting on RNA; TSN, Tudor-SN. A-to-I RNA editing also has a significant impact on various disorders. For instance, a decrease in A-to-I RNA editing at position 5 of miR-411-5p has been observed in multiple malignancies, such as NSCLC. The edited form of miR-411-5p targets hepatocyte growth factor receptor (MET) directly, thereby negatively regulating the mitogen-activated protein kinase (MAPK) signaling cascade. This regulation suppresses cellular proliferation, induces apoptosis, and enhances sensitivity to epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) in NSCLC-resistant cells. 105 A further study demonstrated that the edited form of miR-379-5p has been shown to inhibit cell proliferation and promote apoptosis across various tumor contexts in vitro , an effect attributed to its unique capability to target CD97, a function not observed with the wild-type miR-379-5p. Notably, the delivery of edited miR-379-5p mimics via nanoliposomal systems remarkably suppressed tumor growth and prolonged survival in murine models, highlighting its potential as a therapeutic agent. 106 Furthermore, the hypoediting of adenosine to inosine in miR-455-5p enhances the progression and spread of melanoma. 107 By acting on the precursor of miR-21, ADAR1 decreases the production of mature miR-21, which in turn promotes the polarization of macrophages toward the M2 phenotype by modulating the Foxo1-interleukin (IL)-10 pathway. 108 While some pri-miRNAs have been observed to undergo A-to-I RNA editing in vivo , 109 the biological importance of this editing remains unclear. The latest research has shown that A-to-II RNA editing influences both the biogenesis and activity of miRNAs. Multiple adenosine residues in pri-miR-142 are extensively edited by ADAR1 and/or ADAR2. In vitro processing experiments with Drosha/DGCR8 and Dicer/TAR RNA-binding protein (TRBP) complexes demonstrated that editing hinders Drosha cleavage of pri-miR-142, with the +4 and +5 sites, positioned in the dsRNA stem near the Drosha cleavage site, being the most effective at inhibiting processing from pri- to pre-miR-142. Consequently, miR-142-3p and miR-142-5p expression levels are considerably raised in the spleens of B cell-lineage-specific ADAR1 −/− and ADAR2 −/− mice compared to wild-type controls. 110 Given the inhibitory impacts of A-to-I editing on pri-miR-142 processing and the expression of mature miR-142, a considerable degree of extensively edited pri-miR-142 RNA would be expected in the spleens of wild-type mice. However, only low expression levels of edited pri-miR-142 RNA are detected. This suggests that highly edited pri-miR-142 RNAs, which Drosha cannot cleave, are rapidly processed for degradation by Tudor-SN (TSN), a ribonuclease that targets I-dsRNAs. 111 This evidence indicates that the steady-state levels of edited pri-miR-142 RNAs are regulated by the incidence of editing and the TSN function. 110 Nevertheless, several edited pri-miRNAs remain stable, and it appears that only certain edited pri-miRNAs with multiple I·U pairs are selectively degraded by TSN following A-to-I RNA editing. A primary editing site (+3) and a minor site (−1) of pri-miR-151 were discovered on its antisense strand, close to the terminal loop. Studies on ADAR1 −/− embryos showed that ADAR1 is responsible for editing both the −1 and +3 sites. Although no edited mature miR-151-3p RNAs were found, all identified pre-miR-151 molecules were fully edited at the +3 site. In vitro , assays indicated that A-to-I substitutions at the −1 or +3 sites had no impact on Drosha cleavage, but they restricted cleavage of the edited pre-miR-151 by the Dicer/TRBP complex. These findings indicate that Dicer cleavage is blocked by editing at the major +3 and minor −1 sites, resulting in the accumulation of edited pre-miR-151 RNAs in the cytoplasm. 112 Notably, in vitro investigations demonstrated that editing at the +3 site is much more efficient in pre-miR-151 than in pri-miR-151. This suggests that some pri-miRNAs not edited in the nucleus may only be edited after being processed into pre-miRNAs in the cytoplasm, likely by the cytoplasmic ADAR ADAR1p150. 112 Recent research indicates that specific miRNA precursors are modified by ADAR enzymes. A comprehensive analysis of human pri-miRNA sequences revealed A-to-I RNA-editing sites in approximately 6% of the pri-miRNAs studied. 109 Nonetheless, this may be an underestimate, as in vitro studies involving randomly chosen pri-miRNAs suggest that up to 50% (4 out of 8) of pri-miRNAs could harbor targeted A-to-I RNA-editing sites. 110 This editing process may significantly impact the processing, expression, and activities of mature miRNAs. A-to-I RNA editing modifies the dsRNA configuration of miRNA precursors, potentially influencing their processing and transport stages. Recent findings indicate that editing at two exclusive locations within pri-miRNA-142 (+4 and +5 sites) fully inhibits its cleavage by the Drosha-DGCR8 complex. Additionally, TSN facilitates the decay of pri-miRNA-142 when it undergoes extensive editing. 110 Although it has not yet been demonstrated, A-to-I RNA editing of specific sites in definite pri-miRNAs is anticipated to inhibit the export of pre-miRNA from the nucleus by XPO5 and RanGTP, as well as the processing of pre-miRNA to mature miRNA by the Dicer-TRBP complex. In studies on pri-miRNA-142, editing at specific sites, including the +40 site, did not interfere with cleavage by either Drosha or Dicer. 110 Conformational alterations in certain miRNA precursors due to editing at specific sites may be tolerated, suggesting that editing in particular pri-miRNAs could produce edited mature miRNAs, depending on the site(s) of editing. For instance, a Kaposi’s sarcoma-associated virus miRNA (miRNA-K12-10b) has been shown to undergo editing at position +2. 113 Edited miRNAs can regulate a distinct set of target genes compared to their unedited counterparts, mainly if editing occurs within the “seed sequence” (the 5′ region from +2 to +8), which is crucial for target mRNA binding. 112 Additionally, editing might influence which miRNA strand is chosen as the “effective” strand for loading onto the RISC to guide it to target mRNA, as this selection relies on the stability of the sense-antisense miRNA duplex, which can be altered by A-to-I RNA editing. 114 Recent investigations have concentrated on how edited miRNAs affect target selection. For instance, it was recently demonstrated that miR-376a-5p, anticipated to target the PRPS1 transcript, undergoes editing by ADAR2. Notably, in ADAR2-deficient mice, PRPS1 protein levels are increased. 112 Nuclear ADARs may influence this step by creating a steric hindrance that prevents Drosha/DGCR8 from binding or inhibiting Drosha’s activity via modifications at the edited base within the pri-miRNA transcript. Since ADARs can bind to pre-miRNA hairpins, it is also possible that this binding could interfere with their nuclear export, a process facilitated by XPO5. The subsequent step in miRNA processing involves the cleavage of the pre-miRNA hairpin by the Dicer complex in the cytoplasm, producing a 21–23 nt RNA duplex. Cytoplasmic ADARs may also impact this process by either binding to or editing the pre-miRNA. Our findings suggest that, similar to murine mir-376, editing can redirect the targeting of human miR-376a2. 115 Notably, our results also reveal that ADARs can impact miRNA processing at the Drosha cleavage stage even without editing, thus modifying miRNA functional activity. This implies that ADARs may affect a broader range of miRNAs than previously anticipated based solely on known edited miRNAs. 115 ADARs can interfere with miRNA processing by attaching to miRNA precursors without relying on their catalytic function. For instance, it has been shown that a deaminase-inactive form of ADAR2 (E319A mutant) can still bind to pri-miR-376a2 and inhibit its processing by Drosha, resulting in a decrease in mature miRNA levels. This demonstrates that ADARs can obstruct miRNA biogenesis solely through dsRNA-binding domains (dsRBDs) even in the absence of editing activity. Importantly, the research also indicated that the dsRBDs of ADAR2, rather than its deaminase domain, are sufficient to impede Drosha cleavage physically, reinforcing the idea that the truncated version of ADARB1 (which lacks the deaminase domain but retains the dsRBDs) may also disrupt processing. 115 A-to-I editing of miR-376 cluster transcripts can result in the suppression of exclusive gene sets, enabling precise, tissue-specific modulation of selected gene products. Many miR-376 cluster members undergo substantial A-to-I editing at two specific sites (+4 and +44) in selected tissues, including certain areas of the brain, in both Homo sapiens and mice. 112 Unlike pri-miR-142 and pri-miR-151, the editing of pri-miR-376 RNAs at these two sites does not interfere with Drosha or Dicer cleavage. Both editing sites in pri-miR-376 (+4 and +44) are positioned within the crucial 5′ seed sequences of the miR-376-5p and -3p strands. In silico predictions identified two distinct sets of target genes (around 80 each) for the unedited and edited versions of miR-376a-5p, with only minimal overlap between them. Reporter assays validated that a single A-to-I base alteration is capable of redirecting the silencing activity of miRNAs to a new set of target genes. It has also been investigated that the edited miR-376a-5p specifically represses PRPS1 in a tissue-specific manner. 112 Editing at specific sites (+1 and +2) is expected to affect miRNA strand selection by weakening the 5′ end of the pre-miRNA22 Dicer product. Additionally, a single A-to-I change within a 20–22 nt sequence could redirect a miRNA to a different target. Alternatively, miRNA editing may fine-tune their activity by adjusting interactions with target sequences. 104 Although not yet confirmed, A-to-I editing of specific pri-miRNAs at certain positions may inhibit their export from the nucleus via the RanGTP/XPO5 pathway. A-to-I RNA editing is supposed to influence the local stability of the miRNA duplex. The selection of the “effective” miRNA strand, which is loaded onto the miRISC and directs the miRNA to its target mRNA, is dependent on the local stability of the sense-antisense miRNA duplex. 114 Therefore, editing may also impact the selection of the functional miRNA strand. It has been suggested that editing miRNA target positions could influence miRNA-mediated gene silencing. Since A-to-I editing often happens within 3′ UTRs, which are frequent miRNA-binding sites, such editing could potentially create or remove miRNA target sites. However, computational analyses examining the prevalence of editing at potential miRNA target sites have found that RNA editing tends to avoid these regions. 116 Post-transcriptional processes, including miRNA editing by ADAR enzymes and alternative cleavage by Drosha, regulate miRNA levels and generate isoforms of miRNAs (isomiRs) that vary in length, sequence, or both. These modifications can interfere with miRNA processing, such as Drosha cleavage, and alter miRNA stability or target specificity, thereby enabling diverse cellular functions. 117 Current research highlights that RNA editing of pri-miRNAs can inhibit their processing by Drosha, leading to the decay of the unprocessed, edited pri-miRNAs. 110 ADAR1 and ADAR2 edit specific adenosine residues in miRNA precursors such as pri-miR-142, resulting in the modification of adenosine to inosine. This editing alters the dsRNA structure by replacing stable Watson-Crick A-U pairs with less stable I-U wobble pairs, leading to significant structural changes. Editing of pri-miR-142 at specific positions inhibits its processing by the Drosha-DGCR8 complex, the first step in miRNA biogenesis. Consequently, the production of mature miR-142-3p and miR-142-5p is substantially reduced. Highly edited pri-miR-142 is targeted for decay by TSN, a ribonuclease specific to inosine-containing RNA. This degradation prevents the accumulation of unprocessed, edited precursors, thereby reducing the steady-state phases of edited miRNA precursors. Editing at certain sites, such as positions +4 and +5 near the Drosha cleavage site, completely blocks processing. In contrast, editing at sites like +40 near the terminal loop has a minimal impact. Edited precursors are degraded more efficiently when multiple editing sites are involved. 110 Additionally, the presence of inosines in miRNAs and their precursors can trigger their degradation by TSN. 110 In rare cases, editing can even enhance miRNA-processing effectiveness, as demonstrated by the increased Drosha-mediated processing of pri-miR-203 due to A-to-I editing. 118 When editing happens in the 5′ seed sequence (nt 2–8), it can result in retargeting. A notable instance is miR-376, where editing in the 5′ seed site alters its target selectivity, causing it to regulate a different set of genes. 112 However, the editing of miRNAs does not correspond to their downregulation or upregulation. Moreover, the low levels of editing are unlikely to cause significant changes in the control of an emerging target mRNA, allowing the edited miRNAs to be redirected toward their intended targets. 119 Editing events in pri-miRNAs can result in the production of edited mature miRNAs. If these modifications occur in the identification site or “seed” area, it is likely to cause a shift in the target genes. A notable instance is mouse miR-376, where editing at the identification site alters the miRNA’s target selectivity, significantly impacting cellular processes. 112 It is suggested that miRNA editing may still be biologically significant by reducing the abundance of wild-type miRNA or generating edited variants with altered targets. For instance, the editing of pri-miRNAs, including pri-miR-142 and pri-miR-151, can inhibit Drosha and Dicer cleavage, thereby suppressing the biogenesis of the wild-type miRNA. Conversely, editing of seed sequences, such as miR-376, can redirect silencing to new targets, even if the edited form is less abundant. Additionally, edited pri-miRNAs may be degraded by TSN or retained in the nucleus, further modulating miRNA output. Thus, editing can influence miRNA function through both quantitative and qualitative changes. 120 An investigation has shown that glucose deprivation causes a time-dependent buildup of reactive oxygen species (ROS), which in turn reduces histone deacetylase (HDAC) activity, notably lowering HDAC2 levels. This inhibition of HDAC activity leads to raised acetylation of the miR-466h-5p promoter region and upregulation of this miRNA. Our findings provide a new perspective on the connection between miRNA expression regulation and molecular events under specific physiological conditions. The transcriptional stimulation of this miRNA occurred following a decrease in HDAC2 function and an enhanced acetylation of the miR-466h-5p promoter region, both triggered by metabolic oxidative stress due to glucose deprivation. 121 N4-acetylcytidine (ac4C) ( Figure 7 A) is predominantly found in rRNA, tRNA, and mRNA. The acetyltransferase NAT10 catalyzes its formation in conjunction with the adaptor protein THUMPD1 122 or accessory box C/D small nucleolar RNAs (snoRNAs). 123 It is reported that the stem-loop hairpin configuration of mRNA transcripts is critical for ac4C modification mediated by the NAT10/THUMPD1 complex. 124 Notably, the stem-loop hairpin structure is also a defining characteristic of pri-miRNA transcripts. This structural similarity suggests that pri-miRNAs could undergo a similar modification by NAT10. ac4C modifications on pri-miRNAs are catalyzed by NAT10. These modifications significantly enhanced the collaboration between pri-miRNAs and the microprocessor complex, consisting of DGCR8 and DROSHA, thereby facilitating the biogenesis of mature miRNAs and influencing tumor growth. 124 However, the precise mechanisms by which ac4C modification influences pri-miRNA secondary structure and its broader functional implications in non-cancer contexts remain unresolved. Figure 7 N4-acetylcytidine and 8-oxoguanine in miRNA stability, biogenesis, and function (A) The figure depicts the regulatory role of N4-acetylcytidine (ac4C), a chemical modification that influences miRNA stability and processing efficiency. 124 ac4C modification occurs on pri-miRNAs and is catalyzed by NAT10, assisted by THUMPD1. This modification enhances the stability of pri-miRNA and its recognition by the microprocessor complex, thereby strengthening the interaction between pri-miRNAs and the Drosha-DGCR8 microprocessor complex and facilitating efficient miRNA processing and stability. By promoting structural integrity, ac4C modification facilitates miRNA maturation and ensures efficient gene silencing activity. (B) The figure provides a schematic representation of the potential impact of the oxidative chemical modification 8-oxoguanine (o8G) on miRNA function. 152 o8G chemical modification occurs on mature miRNA due to ROS-mediated oxidative stress. This modification affects miRNA function by altering RNA structure, impairing translational efficiency, and disrupting miRNA-mRNA interactions, which can potentially lead to dysregulated gene expression. Typically, when miRNAs are oxidized, they misrecognize target mRNAs and silence non-specific genes. ac4C, N4-acetylcytidine; RNA Pol II, RNA polymerase II; DGCR8, DiGeorge syndrome critical region 8; THUMPD1, THUMP domain-containing protein 1; NAT10, N-acetyltransferase 10; o8G, 8-oxoguanine; ROS, reactive oxygen species. N4-acetylcytidine and 8-oxoguanine in miRNA stability, biogenesis, and function (A) The figure depicts the regulatory role of N4-acetylcytidine (ac4C), a chemical modification that influences miRNA stability and processing efficiency. 124 ac4C modification occurs on pri-miRNAs and is catalyzed by NAT10, assisted by THUMPD1. This modification enhances the stability of pri-miRNA and its recognition by the microprocessor complex, thereby strengthening the interaction between pri-miRNAs and the Drosha-DGCR8 microprocessor complex and facilitating efficient miRNA processing and stability. By promoting structural integrity, ac4C modification facilitates miRNA maturation and ensures efficient gene silencing activity. (B) The figure provides a schematic representation of the potential impact of the oxidative chemical modification 8-oxoguanine (o8G) on miRNA function. 152 o8G chemical modification occurs on mature miRNA due to ROS-mediated oxidative stress. This modification affects miRNA function by altering RNA structure, impairing translational efficiency, and disrupting miRNA-mRNA interactions, which can potentially lead to dysregulated gene expression. Typically, when miRNAs are oxidized, they misrecognize target mRNAs and silence non-specific genes. ac4C, N4-acetylcytidine; RNA Pol II, RNA polymerase II; DGCR8, DiGeorge syndrome critical region 8; THUMPD1, THUMP domain-containing protein 1; NAT10, N-acetyltransferase 10; o8G, 8-oxoguanine; ROS, reactive oxygen species. Oxidative modifications of nucleotides can result in the transformation of guanosine (G) to 8-oxoguanine (O 8 G) ( Figure 7 B), a lesion known to pair aberrantly with adenosine (A) and thereby promote guanosine-to-thymidine (G>T) transversion mutations in DNA. Notably, analogous oxidative alterations have also been observed in RNA, particularly in miRNAs, where guanosine residues are similarly converted to O 8 G, potentially impacting RNA structure and function. 125 ROSs are highly reactive compounds distinguished by the presence of at least one oxygen atom and greater reactivity compared to molecular oxygen. 126 These include a diverse array of species, such as the superoxide anion (O 2 − ), hydrogen peroxide (H 2 O 2 ), hydroxyl radical (OH − ), peroxynitrite (ONOO − ), and hypochlorous acid (HOCl), among others. 127 ROSs influence the expression and modification of miRNAs within cells. 128 Post-transcriptional RNA modifications introduce dynamic changes to the transcriptome, thereby regulating gene expression and orchestrating various cellular processes. 129 The 8-oxo-G modification in RNA was first identified in 1989 and has since been recognized for its impact on various cellular processes. This modification has been shown to impair translational efficiency and hinder RNA elongation. 130 Beyond its role in translation, oxidized RNA has been implicated in modulating the inflammatory response. For instance, cells transfected with oxidized mitochondrial RNA (mtRNA) extracted from H 2 O 2 -treated HA1 hamster fibroblasts exhibit reduced generation of proinflammatory cytokines, including MCP-1 and IL-6. 131 Additionally, the 8-oxo-G modification in miRNAs has been reported to influence cardiac cell apoptosis by disrupting the pairing with mRNAs encoding anti-apoptotic proteins, including Bcl-xL and Bcl-w. 132 For instance, the introduction of 7o8G-miR-1, or its variant 7U-miR-1 (where guanine at position 7 is replaced with uracil), has been demonstrated to stimulate cardiac hypertrophy in murine models. Furthermore, the mRNA targets regulated by o8G-miR-1 serve a critical function in mediating the associated pathological phenotypes. Notably, particular blockade of 7o8G-miR-1 in mouse cardiomyocytes effectively alleviated the progression of cardiac hypertrophy, underscoring its functional significance. Additionally, o8G-miR-1 has been implicated in cardiomyopathy among human patients, further linking this oxidized miRNA to cardiac pathologies. These findings highlight the oxidative modification at specific miRNA positions as a potential epitranscriptional process regulating redox-sensitive gene expression in pathophysiological contexts. 133

Writing – original draft, A.A. and P.K.K.; writing – review & editing, A.A. and P.K.K.; conceptualization, P.K.K.; funding acquisition, P.K.K.; resources, P.K.K.; supervision, P.K.K.

Recent studies underscore the extensive array of chemical modifications that miRNAs undergo, which play a pivotal role in modulating their biogenesis, stability, and functionality within post-transcriptional gene regulation. Among these modifications, methylation and RNA editing have emerged as critical mechanisms influencing miRNA-mediated regulatory pathways. 134 Methylation, the most prevalent RNA modification in eukaryotic systems, manifests in various forms, including m6A, m5C, and m7G, each exerting a profound influence on miRNA expression and regulatory functions. 135 The epitranscriptome, which encompasses the full spectrum of post-transcriptional RNA modifications, significantly impacts miRNA biogenesis, target specificity, and regulatory dynamics, with recent evidence highlighting intricate crosstalk among these modifications. 134 These small RNA modifications are integral to diverse biological processes, including stress response, metabolic regulation, immune function, and the epigenetic inheritance of environmentally induced traits. 136 The m6A modification plays a crucial role in miRNA processing and maturation, exerting a profound influence on RNA metabolism and post-transcriptional gene regulation. 137 This dynamic modification is orchestrated by m6A writers, erasers, and readers, which collectively regulate miRNA biogenesis and functional activity. 138 Notably, m6A modification can drive the aberrant maturation of oncogenic miRNAs, contributing to disease pathogenesis. For example, in pancreatic cancer cells exposed to cigarette smoke condensate, excessive miR-25-3p maturation is induced by METTL3 overexpression, leading to increased m6A modification and the activation of oncogenic signaling cascades. 139 Similarly, WTAP-mediated m6A modification facilitates the accelerated maturation of pri-miR-29b-3p in an m6A-dependent manner, thereby influencing the differentiation of bone marrow mesenchymal stem cells. 140 m6A modification emerges as a crucial post-transcriptional modulator of miRNA biogenesis and function, with significant implications for cancer development, stem cell differentiation, and disease progression. Deciphering the complex crosstalk between m6A and miRNAs may offer novel opportunities for biomarker discovery and the advancement of therapeutic strategies targeting disorders, including cancer and osteoporosis. 138 , 140 2′Ome at the 3′ end of miRNAs is indispensable for miRNA function in plants and Drosophila , highlighting its evolutionary significance. Recent investigations have demonstrated that 2′Ome also occurs in mammalian miRNAs, with a predominant fraction of miR-21-5p isolated from human NSCLC tissue exhibiting this modification. Interestingly, the 3′-terminal 2′Ome pattern differs between malignant and non-malignant lung tissues, suggesting a potential role in tumorigenesis. The methyltransferase HENMT1 has been recognized as the enzyme responsible for catalyzing the 3′-terminal Nm modification in mammalian miRNAs. This chemical modification enhances miRNA stability by increasing resistance to 3′→5′ XRN degradation and strengthening the affinity of AGO2 binding, thereby facilitating more effective silencing of target genes. 57 Nm at the 3′ end of miRNAs represents a crucial regulatory mechanism that enhances the stability and functionality of miRNAs. Although initially characterized in plants, this modification has now been identified in mammalian miRNAs, with potential implications for cancer biology. The discovery of HENMT1 as the key methyltransferase in this process offers novel insights into miRNA regulation and presents promising avenues for developing novel therapeutic strategies targeting miRNA modifications. RNA editing, particularly A-to-I editing, has emerged as a critical post-transcriptional process that modulates miRNA function and gene regulation, with significant implications for cancer biology. ADAR enzymes catalyze this modification, which can occur in both miRNA precursors and mature miRNAs, influencing their biogenesis, target specificity, and functional activity. 141 Comprehensive analyses have revealed thousands of miRNA-editing events across distinct cancer categories, with 19 recurrent A-to-I RNA-editing hotspots detected in a study encompassing 8,595 samples from 20 cancer categories. 142 The prevalence and frequency of miRNA-editing events exhibit species- and tissue-specific variations, with high editing levels predominantly observed in neuronal tissues in mice and Drosophila. In contrast, editing occurs more broadly across multiple tissue types in humans. 143 Notably, miRNA editing plays a dual role in cancer progression, depending on the specific context. For instance, editing of miR-200b, a known tumor metastasis repressor, paradoxically enhances cell invasion and migration by impairing its capability to repress ZEB1/ZEB2 while simultaneously obtaining the capacity to target leukemia inhibitory factor receptor (LIFR), a metastasis repressor. 141 Furthermore, edited miRNAs in neuronal and non-neuronal tissues acquire distinct target repertoires linked to cognitive and developmental processes. 144 A-to-I editing illustrates a crucial post-transcriptional regulatory mechanism that influences miRNA-mediated gene silencing and demonstrates potential as a biomarker for cancer prognosis and treatment intervention. 142 However, challenges remain in accurately discovering and investigating miRNA-editing alterations in tumor samples, necessitating further research and the development of advanced detection and validation methodologies. 145 m7G modification has become recognized as a crucial post-transcriptional modification influencing miRNA biogenesis, stability, and function. Recent studies have highlighted its regulatory role across various RNA species, including miRNAs, with the METTL1 and WDR4 complex playing a pivotal role in catalyzing this modification. Functionally, m7G modification has been associated with modulating miRNA metabolism and processing, potentially contributing to tumorigenesis and cancer progression. The m7G modification represents a key epitranscriptomic mark with significant implications for miRNA-mediated gene regulation, given its potential role in disease pathogenesis, particularly in cancer. 146 Future research should aim to characterize the functional consequences of this finding and explore its potential as a biomarker or therapeutic target. Recognizing the molecular mechanisms of m7G-modified miRNAs may provide novel insights into gene regulatory networks and disease progression. PUS10, a pseudouridine synthase, has been identified as an essential factor in miRNA processing. It is co-expressed with the microprocessor complex and directly interacts with pri-miRNAs, thereby facilitating miRNA maturation. Notably, this regulatory function is independent of its catalytic activity, suggesting a previously unrecognized role for pseudouridine synthases beyond their enzymatic capacity. 96 This finding broadens the functional repertoire of pseudouridine synthases and their involvement in post-transcriptional modulation. Moreover, the incorporation of pseudouridine and other nucleotide modifications in mRNA therapeutics has been indicated to modulate miRNA-regulated gene silencing. Specifically, pseudouridine and related modifications attenuate the regulatory efficiency of miRNA switches in a manner that is dependent on the individual miRNA rather than the overall proportion of modified nucleotides within the target site. 147 These observations hold significant implications for the rational design of mRNA-based therapeutic strategies and the fine-tuning of miRNA-regulated gene expression systems. Pseudouridylation emerges as a critical modulator of RNA function, influencing both miRNA processing and gene regulation. The dual role of pseudouridine synthases in enzymatic and non-enzymatic mechanisms, the impact of pseudouridine incorporation on miRNA-mediated silencing, and the potential of pseudouridylation patterns as cancer biomarkers collectively underscore the significance of this modification in RNA biology. These insights lead to the advancement of innovative therapeutic applications leveraging pseudouridylation-mediated regulatory mechanisms. m5C represents a pivotal epitranscriptomic modification implicated in the modulation of diverse RNA metabolic processes and functions. Recent investigations have provided significant insights into the presence and functional consequences of m5C modifications in miRNAs, particularly in cancer studies. 148 m5C modifications have been identified in miRNAs, with notable prevalence in cancer cells. Utilizing single-molecule quantum sequencing, researchers have successfully mapped m5C sites in hsa-miR-200c-5p isolated from colorectal cancer cells. 149 This high-resolution sequencing approach enables the precise determination of methylation sites and modification rates, yielding results comparable to those obtained from mass spectrometry (MS)-based analyses. Interestingly, findings suggest that the presence of adjacent methylation sites may influence the frequencies of cytidine and adenosine methylation, indicating potential cooperative regulatory mechanisms. 149 These observations underscore the significance of m5C modifications in miRNA biology and their potential implications in cancer pathogenesis. The development of advanced sequencing technologies and chemoproteomic methodologies has facilitated a more comprehensive and precise characterization of m5C sites in miRNAs. 91 These advancements facilitate further exploration into the function of m5C in post-transcriptional gene regulation, with potential applications in disease biomarker discovery and therapeutic intervention. o8G modifications in miRNAs have appeared as essential epitranscriptional modulators, influencing tumorigenesis and gene expression. Recent studies have identified a widespread, position-specific distribution of o8G modifications in tumor-associated miRNAs, with predominant localization in the seed site (positions 2–8) and distinct sequence clustering patterns. 150 These modifications have been shown to modulate miRNA functionality by altering target transcript selection, thereby contributing to either oncogenic or tumor-repressive pathways in relation to the particular miRNA and cancer type. Notably, guanine oxidation leading to o8G occurs more readily in RNA than in DNA under conditions of oxidative stress. 144 This differential susceptibility suggests a unique role for o8G modifications in miRNA-mediated redox-sensitive gene regulation. Additionally, o8G’s ability to form non-canonical base pairing with adenine introduces alterations in RNA-RNA interactions, potentially reprogramming post-transcriptional regulatory networks. 151 These structural and functional consequences underscore the importance of oxidative modifications in shaping the dynamics of miRNA-regulated gene expression. o8G modifications in miRNAs represent a critical mechanism through which oxidative stress can influence gene regulation and cancer progression. 152 The dynamic nature of these modifications underscores their potential as targets for therapeutic interventions aimed at modulating redox-responsive gene expression. Further investigations are necessary to clarify the exact molecular mechanisms governing o8G-modulated miRNA reprogramming and to explore their broader implications across diverse cellular and pathological contexts. ac4C modification has been identified on pri-miRNAs, with NAT10/THUMPD1 serving as the acetyltransferase responsible for catalyzing this modification. Functionally, ac4C serves a vital role in miRNA biogenesis by facilitating the processing of pri-miRNAs into pre-miRNAs through enhanced interaction with DGCR8, thereby enhancing the maturation and accumulation of functional miRNAs. The identification of ac4C modifications in pri-miRNAs represents a significant advancement in the study of RNA epigenetics. This finding deepens our insight into miRNA biogenesis and suggests a potential role for ac4C in oncogenic processes and disease progression. Moreover, targeting NAT10-mediated ac4C modifications presents a promising therapeutic avenue for advancing RNA-based interventions in cancer and other pathological conditions. 124

Chemical modifications of miRNAs have emerged as crucial modulators of their biogenesis, stability, and function. Despite substantial progress in this field, several technical, biological, and translational challenges remain unresolved. To provide a structured overview of the current landscape and future directions, this section has been organized into four key thematic sections: (1) technical challenges in detecting and mapping miRNA modifications, (2) model system limitations and context-dependent effects, (3) functional characterization and target rewiring, and (4) translational implications and therapeutic potential. This framework aims to highlight both the complexity of miRNA modification biology and the therapeutic opportunities it presents. RNA modifications have a pivotal role in regulating miRNA biogenesis, stability, and function. Although notable progress has been made in this field, numerous challenges and future research directions persist. One crucial technical limitation arises from the short length of mature miRNAs (∼22 nt) when it comes to identifying chemical modifications on mature miRNAs, which restricts the resolution of conventional sequencing techniques and makes it challenging to accurately map changes at the single-nucleotide level. For instance, the first-generation m6A mapping technique methyl-RNA immunoprecipitation and sequencing (MeRIP-seq or m6A-seq) typically identifies broad regions of 100–200 nt enriched for m6A 153 but lacks the resolution to determine the precise modified site, which limits its applicability for high-resolution mapping in small RNAs, such as mature miRNAs. Furthermore, the presence of isomiRs, which are miRNA variants with subtle sequence or length differences, introduces additional complexity to data interpretation, as some modifications may be misinterpreted as sequence variants or sequencing artifacts. For example, advances in next-generation sequencing (NGS) have indicated that miRNAs, in addition to their 21–22 nt structure, commonly appear in the form of various isoforms, referred to as isomiRs. Therefore, differentiating these variants from sequencing artifacts and errors related to the detection method remains a significant challenge. Reliable identification is vital, as mistakes in classification can result in a misunderstanding of their target interactions, clinical significance, and potential for therapy. 154 The intricate interplay among various miRNA modifications necessitates further elucidation. For instance, methylation has been shown to confer protection against uridylation-induced degradation. 45 Investigating the coordinated regulation of these modifications and their collective impact on miRNA homeostasis represents a critical avenue for future research. Advancing analytical methodologies for detecting and quantifying miRNA modifications is imperative. Although techniques such as capillary electrophoresis-MS (CE-MS) have demonstrated potential in identifying post-transcriptional modifications, 155 substantial improvements are required to enhance sensitivity, enable multiplex detection, and facilitate the analysis of low-abundance miRNAs. A key challenge remains in identifying the enzymes responsible for specific miRNA modifications and deciphering their substrate specificities. The functional implications of miRNA modifications in various physiological and pathological contexts remain incompletely understood. For instance, while pseudouridylation is known to influence ribosomal function, 156 the broader impact of miRNA modifications in immune-related diseases 74 warrants further investigation. Future investigations should aim to unravel the intricate crosstalk between various chemical modifications and their functional implications. Notably, while the influence of m6A on miRNA biogenesis has been extensively explored in animal systems, its regulatory impact in plants remains insufficiently characterized and warrants further examination. 157 Unraveling the function of these modifications in cellular differentiation, developmental processes, and disorder pathogenesis will provide valuable insights into miRNA-mediated regulatory mechanisms. Exploring the evolutionary trajectory of miRNA-modifying enzymes may yield further insights into the functional significance of these modifications across different species. For example, the evolutionary transition of uridylyltransferases from adenylyltransferases 158 highlights the adaptive mechanisms underlying RNA modification processes. In conclusion, addressing these challenges requires an interdisciplinary approach that integrates biochemistry, molecular genetics, and advanced analytical technologies. Future investigations in this domain hold the potential to unveil novel regulatory paradigms and therapeutic strategies targeting miRNA-related diseases. miRNAs have gained recognition as valuable biomarkers for assessing drug-induced toxicity and chemical exposure, owing to their stability, evolutionary conservation, and early responsiveness to toxic insults. 159 These characteristics render miRNAs promising tools in drug safety evaluation and toxicological pathology. However, the therapeutic application of chemically modified miRNAs remains challenging due to concerns regarding potential toxicity and off-target effects. Chemical modifications of oligonucleotides, including miRNAs, have been implemented to promote their stability, resistance to nuclease degradation, and therapeutic efficacy. 160 Nevertheless, these modifications can inadvertently introduce toxic effects. For instance, second-generation AMOs (anti-miRNA oligonucleotides) designed with high-binding-affinity compounds may inadvertently suppress the function of closely related sequences, leading to unintended gene silencing and off-target activity. Additionally, the incorporation of non-natural nucleic acid residues has been associated with cytotoxic effects in in vitro models. 161 To address these limitations, future research should focus on developing innovative chemical modifications that enhance miRNA stability and therapeutic potency while minimizing toxicity. The identification of non-nucleotide chemical modifiers, such as the ZEN (N,N-diethyl-4-(4-nitronaphthalen-1-ylazo)-phenylamine) compound, has demonstrated potential in enhancing binding affinity and protecting miRNAs from exonuclease degradation without compromising specificity or inducing significant cytotoxicity. 161 Moreover, the integration of chemical modifications with advanced nanoparticle-based delivery systems may enhance the pharmacokinetics and gene-silencing efficacy of miRNA therapeutics while minimizing adverse effects. 162 As miRNA-based biomarker research continues to evolve, a comprehensive evaluation of the toxicological profiles of chemically modified miRNAs will be essential to ensuring their safety and efficacy in both diagnostic and therapeutic applications. While some studies have demonstrated the role of modifications, such as 2′-F and 2′-Ome, in improving miRNA stability and potency, 163 further research is needed to assess off-target effects, immunogenicity, and long-term biocompatibility. Integrating such toxicological assessments with existing design frameworks will be critical for clinical translation. Off-target effects represent a significant challenge in RNAi (RNA interference), including siRNA investigations, where unintended gene silencing has been documented. 164 Similar effects may arise in miRNA applications due to interactions in the seed region. The propensity for off-target interactions is a crucial consideration in the advancement of miRNA-based therapies, necessitating the implementation of strategies to enhance specificity and minimize unintended effects. A significant challenge in the chemical modification of miRNAs as a promising strategy to promote their stability and therapeutic efficacy is the potential immunogenicity of modified miRNAs, which can elicit unintended immune responses, leading to adverse effects or diminished therapeutic activity. 162 The immunogenicity of chemically modified miRNAs must therefore be carefully evaluated to ensure both safety and efficacy in clinical applications. Interestingly, while immunogenicity is often regarded as an obstacle, it can also be exploited for therapeutic purposes. In the context of cancer immunotherapy, for instance, inducing immunogenic cell death (ICD) can enhance tumor immunogenicity, thereby stimulating antitumor immune responses. 165 This dual role of immunogenicity underscores the complexity of miRNA modifications and highlights the necessity of tailoring their immunomodulatory properties based on therapeutic objectives. Nanoparticle-based delivery systems offer a promising approach to regulate systemic biodistribution and achieve targeted delivery of modified miRNAs, thereby mitigating immunogenicity and improving stability. 166 Further investigation into the molecular mechanisms underlying miRNA-mediated immune regulation could provide critical insights for designing immunologically optimized miRNA therapeutics. 167 The chemical modification of miRNAs and their associated delivery systems constitutes a critical area of research aimed at addressing the inherent challenges related to stability and efficient delivery in miRNA-based therapeutics. A major limitation in the clinical application of miRNAs is their vulnerability to rapid degradation by nucleases in biological environments. 57 To enhance their stability, various chemical modifications have been employed, including 2′-Ome and phosphorothioate (PS) modifications, which confer resistance to nuclease-mediated decay and improve the pharmacokinetic profile of AMOs. 168 Additionally, these chemical modifications enhance binding affinity to target miRNAs and promote cellular uptake, thereby improving in vivo delivery efficiency. 169 However, the relatively short target sequences of miRNAs introduce challenges in the rational design of effective chemical modifications, necessitating further optimization strategies. 170 Efficient delivery of miRNAs remains another critical obstacle due to poor cellular uptake and rapid enzymatic degradation. Various delivery platforms have been explored, including viral vectors, lipid-based nanoparticles, polymeric nanoparticles, and extracellular vesicles (EVs). Among non-viral carriers, polymer-based delivery systems, including poly (lactic-co-glycolic acid) (PLGA), chitosan, polyethyleneimine (PEI), and polyamidoamine (PAMAM) dendrimers, have demonstrated advantages in biocompatibility, structural versatility, and the protection of miRNAs from degradation. 171 Furthermore, advanced nanocarriers, such as pH-sensitive multifunctional envelope-type nanodevices (MENDs), have exhibited promising potential in promoting targeted delivery of AMOs to specific tissues, including the liver. 168 Future research should prioritize the design of simplified, cost-effective nucleic acid carriers and chemical modification strategies that ensure high therapeutic efficacy, minimal toxicity, and precise tissue targeting, ultimately advancing the field of miRNA-based treatments. Despite advancements in the development of chemical modifications aimed at mitigating off-target effects, several obstacles persist. These include the need for improved subcellular localization, enhanced endosomal escape, and efficient delivery to target cells. 172 Upcoming studies should prioritize refining chemical modification strategies, optimizing delivery systems, and implementing real-time monitoring techniques to assess the biological role of delivered miRNA mimics. 173 In silico models are increasingly recognized as a powerful tool in miRNA chemical modification research, providing valuable insights and cost-effective alternatives to traditional experimental methodologies. These computational approaches have been successfully applied to various aspects of miRNA biology, including miRNA discovery, target prediction, and functional analysis, thereby accelerating the identification of therapeutic targets and regulatory pathways. 174 In silico modeling can aid in optimizing experimental design, reducing the need for extensive laboratory-based investigations and facilitating the advancement of novel miRNA-targeting therapies. 175 Recent studies suggest that synergistic therapeutic strategies, including the use of complementary miRNAs, may yield substantial clinical benefits, particularly in oncology. 176 This approach aligns with the broader trend of combination therapies in miRNA research, wherein multiple therapeutic modalities are integrated to enhance treatment efficacy. Additionally, EVs enriched with therapeutic miRNAs represent a novel delivery platform that, when combined with chemical modifications, may further enhance miRNA stability and functional activity. 177 Moving forward, the future of miRNA chemical modification investigations is likely to emphasize the development of combination therapies. This may involve integrating modified miRNAs with targeted therapeutic agents, such as EGFR-TKIs for NSCLC, 178 or exploring miRNA editing as an innovative approach in cancer therapeutics. 179 Furthermore, the convergence of nanotechnology-based delivery systems with chemically modified miRNAs could present a promising avenue for enhancing therapeutic precision and overcoming challenges such as multidrug resistance in cancer. While current research highlights miRNA regulation, such as miR-508-5p and ABC transporters, as key multidrug resistance (MDR) mechanisms, 180 incorporating nanotechnology may help tackle delivery issues like low bioavailability and unintended effects on other targets. Future research should investigate the interactions between nanoparticle carriers and miRNA therapeutics to overcome efflux pumps and specifically target cancer stem cells, thereby enhancing treatment effectiveness. As research advances, these innovative approaches might support the advancement of more efficient, personalized, and durable miRNA-based therapies. Targeting ligands has developed as a promising avenue in miRNA modification research, proposing a precise strategy for modulating miRNA biogenesis and function. The rational design of sequence-specific ligands enables the selective targeting of conserved structural elements within miRNA-processing sites, thereby facilitating the simultaneous inhibition of multiple miRNAs. This strategy has been exemplified by a dimeric small molecule that binds both the Dicer processing site and an adjacent bulge within the miR-17-92 cluster, achieving a 100-fold increase in potency compared to monomeric ligands. Beyond direct inhibition, these ligands can be further engineered to mediate targeted RNA cleavage, either through direct chemical conjugation or by leveraging endogenous cellular mechanisms. For instance, conjugation with bleomycin A5 has been shown to induce RNA-selective cleavage of the entire pri-miR-17-92 cluster, effectively silencing all six miRNAs encoded within this locus. Alternatively, ligand-based approaches can be designed to recruit endogenous nucleases or function as ribonuclease-targeting chimeras (RIBOTACs), thereby expanding their potential modes of action depending on target localization and cellular context. That said, the development of targeting ligands represents a versatile and innovative tool for miRNA modulation, with significant implications for the selective inhibition or degradation of specific miRNAs and miRNA clusters. These strategies hold particular promise for therapeutic applications in diseases such as cancer and polycystic kidney disease. 181 As this field advances, the refinement of ligand-based approaches is expected to lead to more sophisticated and tailored interventions, ultimately contributing to the development of novel therapeutic modalities for miRNA-associated diseases. New and innovative technologies are significantly enhancing our ability to analyze miRNA chemical modifications with exceptional accuracy and depth. Tools such as single-cell RNA sequencing (scRNA-seq) 182 and spatial transcriptomics 183 are particularly promising, enabling the mapping of cell-type-specific and spatially distinct miRNA modification patterns across tissues. In addition, advanced platforms, such as Nanopore direct miRNA sequencing, a type of Nanopore direct RNA sequencing (DRS), enable full-length miRNA sequencing and single-base-resolution detection of miRNA modifications, including m6A, without the need for reverse transcription, thereby preserving the native chemical landscape of miRNAs. 184 Methods, such as miCLIP-seq, m5C-RIP-seq, and RNA bisulfite sequencing (RNA-BisSeq), 86 provide nucleotide-level insights into the modification types, including methylation of miRNAs. Additionally, MS-based approaches offer a powerful orthogonal strategy for the direct detection and quantification of RNA modifications. 185 When applied to exosomal miRNAs, technologies such as high-performance liquid chromatography-MS (HPLC-MS) further enhance our understanding of miRNA modifications as circulating biomarkers in disease contexts. 186

Chemical modifications of miRNAs represent an additional layer of post-transcriptional regulation, influencing their stability, processing, and function. Methylation, pseudouridylation, phosphorylation, and RNA editing collectively modulate miRNA biogenesis and interaction with target mRNAs. These modifications can either enhance or inhibit miRNA function, demonstrating their potential to serve as biomarkers and targeted therapeutic agents. Among the various modifications, m6A has been widely investigated for its role in facilitating the recognition and processing of pri-miRNA. Its dysregulation has been implicated in several cancers where altered METTL3 activity affects miRNA expression. Similarly, m5C influences miRNA stability and target recognition, with evidence linking its dysregulation to poor prognosis in glioblastoma. m7G modification enhances miRNA-processing efficiency by destabilizing secondary structures such as G-quadruplexes, further emphasizing the intricate interplay between chemical modifications and miRNA maturation. Likewise, phosphorylation enhances miRNA-AGO interactions, impacting their stability and function in gene silencing. Pseudouridylation contributes to miRNA maturation by stabilizing secondary structures and influencing processing efficiency. RNA editing, particularly A-to-I editing, modifies the targeting specificity of miRNA, allowing for the dynamic modulation of gene expression in response to physiological changes. Despite advances in understanding the chemical modifications of miRNA, significant knowledge gaps remain. The precise mechanisms through which these modifications affect miRNA function in different biological contexts require further investigation. High-throughput sequencing, single-molecule imaging, and chemical biology approaches will be crucial in elucidating the full spectrum of miRNA modifications. Furthermore, integrating multi-omics strategies will enable the identification of novel regulatory pathways and therapeutic targets. Prospective studies should focus on developing chemical modification-based miRNA therapies, enhancing delivery strategies, and minimizing off-target effects. By overcoming these obstacles, miRNA modifications could be harnessed for precision medicine, providing novel avenues for disease diagnosis and treatment.

During the preparation of this work, the authors utilized OpenAI’s ChatGPT and Grammarly to refine, paraphrase, and verify grammar, thereby enhancing clarity and fluency. However, the core ideas, structure, and content were developed by A.A. in consultation with P.K.K. After employing these tools, the authors reviewed and edited the content as necessary and take full responsibility for the publication’s content.

MicroRNAs (miRNAs) are short, non-coding RNA molecules, typically approximately 22 nucleotides (nt) in length, mainly transcribed as primary miRNAs (pri-miRNAs) from DNA sequences. They are subsequently developed into precursor miRNAs (pre-miRNAs) and, ultimately, into mature miRNAs. While miRNAs typically bind to the 3′ untranslated region (UTR) of target messenger RNAs (mRNAs) to repress gene expression, 1 they have also been indicated to communicate with other regions, such as the 5′ UTR, gene promoters, and coding sequences. 2 Identifying the first miRNA, lin-4, in 1993 in Caenorhabditis elegans 3 marked a groundbreaking advancement in molecular biology. It was discovered that lin-14 was diminished post-transcriptionally through its 3′ UTR, with lin-4 containing a complementary sequence to this region. 4 Consequently, it was demonstrated that lin-4 modulates lin-14 post-transcriptionally. 5 The synthesis of miRNAs is categorized into canonical and non-canonical pathways, which begin with the formation of transcripts generated by RNA polymerase II or III, either during or after transcription. 1 In the canonical route ( Figure 1 A), which is the primary pathway, 2 pri-miRNAs are modified into pre-miRNAs by the microprocessor complex, which includes the RNA-binding protein (RBP) DiGeorge syndrome critical region 8 (DGCR8) and the ribonuclease III enzyme Drosha. 2 , 6 DGCR8 identifies N6-methyladenylated GGAC sequences and other motifs within the pri-miRNA 7 ; meanwhile, Drosha cleaves the pri-miRNA duplex at the base region of its distinctive hairpin configuration. This cleavage synthesizes a pre-miRNA with a 2-nt 3′ overhang. 8 After pre-miRNAs are formed, they are transported to the cytoplasm via the exportin 5 (XPO5)/RanGTP complex and further cleaved by the RNase III endonuclease Dicer. 9 During this cleavage, the terminal loop is removed, forming a mature miRNA duplex. 2 The naming of the mature miRNA configuration is based on the strand’s directionality: the 5p strand comes from the 5′ end of the pre-miRNA hairpin, while the 3p strand is generated from the 3′ end. 2 Both strands of the mature miRNA duplex can be incorporated into the Argonaute (AGO) protein family (AGO1–4 in humans) in an ATP-dependent process. 10 The selection of the 5p or 3p strand is partly determined by the thermodynamic asymmetry at the 5′ ends of the miRNA duplex, where reduced stability of the 5′ antisense terminus facilitates its preferential incorporation into the RNA-induced silencing complex (RISC). 11 Typically, the strand with lower 5′ stability or a 5′ uracil is more likely to be loaded into AGO as the guide strand. The strand that is not loaded, known as the passenger strand, is separated from the guide strand through different molecular mechanisms depending on their complementarity. If the passenger strand is perfectly matched, it is processed by AGO2 and subsequently digested by cellular degradation systems, creating a high strand selection specificity bias. In cases where there are central mismatches or the miRNA is not loaded into AGO2, the miRNA duplex is unwound passively and degraded. 1 Figure 1 Steps of miRNA biogenesis and function The figure illustrates several steps involved in miRNA biogenesis. The process begins in the nucleus, where RNA polymerase II transcribes the miRNA gene to produce pri-miRNA. (A) In the canonical pathway, the pri-miRNA is then processed by the Drosha-DGCR8 complex into pre-miRNA, which is exported to the cytoplasm by XPO5. (B) In the non-canonical pathway, the pri-miRNA is then processed to pre-miRNA through the spliceosome and exported to the cytoplasm via XPO1. 222 (C) In the cytoplasm, the pre-miRNA is further cleaved by Dicer, in association with Trans-activation response (TAR) RNA-binding protein (TRBP), to form the miRNA duplex. The miRNA duplex is then loaded onto the AGO protein, where one strand (the guide strand) is retained, and the other is degraded, resulting in miRISC construction. In general, the miRISC attaches to target mRNAs to promote translational repression, likely by disrupting the eIF4F complex. Subsequently, GW182 family proteins associated with AGO attract the poly(A)-deadenylases PAN2/3 and CCR4-NOT. PAN2/3 begins the deadenylation process, while the CCR4-NOT complex finalizes it, resulting in the removal of the m7G cap from the target mRNA by the decapping machinery. The decapped mRNA can then be subjected to 5′-3′ degradation by the exoribonuclease XRN1. 223 (D) AGO2, along with another miRNA-associated protein known as FXR1, is associated with AREs located in the 3′ UTR to facilitate translation. This process depends on AGO2 and FXR1 instead of GW182. 224 (E) cRISC, connected to some other factor, including TNRC6A, TARBP2, and DICER1, can enter the nucleus through XPO1 or IPO8. TRNC6A serves as a guiding protein in this mechanism. The nRISC has a composition that resembles the cRISC, but it may also exist solely as an AGO2-miRNA complex (data not shown). The nRISC complex might associate with additional nuclear factors, or the Argonaute protein could form a separate multi-protein complex (e.g., RNA-induced transcriptional silencing (RITS); data not shown). The nRISC interacts with corresponding nuclear transcripts, resulting in the deterioration of transcripts, or (F) AGO is capable of recruiting transcription factors and epigenetic enzymes to a gene’s promoter region, including CMPs, histone demethylases, and histone methyltransferases, to begin the process of histone modification and enhance gene transcription. On the other hand, after the transport of miRNA into the nuclei, AGO proteins assemble inhibitory complexes at the promoter region targeted by the miRNA, which mainly consists of RISC (including AGO and DICER1 proteins), polycomb group (PcG) components (YY1, EZH2, and SUZ12; data not shown), CMPs, and histone deacetylase. This interaction enables the protein inhibitor complex to approach the targeted promoter region more closely, leading to an increase in H3K27me3 modifications and a reduction in H3K4me3 modifications, thereby altering the chromatin structure and creating a non-permissive transcriptional environment. 225 , 226 XPO5, exportin 5; XPO1, exportin 1; TRBP, transactivation response element RNA-binding protein; AGO, Argonaut; eIF4F, eukaryotic initiation factor 4F; CCR4-NOT, carbon catabolite repression 4-negative on TATA-less; XRN1, 5′-3′ exoribonuclease 1; FXR1, fragile X mental retardation-related protein 1; AREs, AU-rich elements; cRISC, cytoplasmic RISC; TRNC6A, trinucleotide repeat-containing adaptor 6A; nRISC, nuclear RISC; RIST, RNA-induced initiation of transcriptional silencing; CMP, chromatin-modifying proteins. Steps of miRNA biogenesis and function The figure illustrates several steps involved in miRNA biogenesis. The process begins in the nucleus, where RNA polymerase II transcribes the miRNA gene to produce pri-miRNA. (A) In the canonical pathway, the pri-miRNA is then processed by the Drosha-DGCR8 complex into pre-miRNA, which is exported to the cytoplasm by XPO5. (B) In the non-canonical pathway, the pri-miRNA is then processed to pre-miRNA through the spliceosome and exported to the cytoplasm via XPO1. 222 (C) In the cytoplasm, the pre-miRNA is further cleaved by Dicer, in association with Trans-activation response (TAR) RNA-binding protein (TRBP), to form the miRNA duplex. The miRNA duplex is then loaded onto the AGO protein, where one strand (the guide strand) is retained, and the other is degraded, resulting in miRISC construction. In general, the miRISC attaches to target mRNAs to promote translational repression, likely by disrupting the eIF4F complex. Subsequently, GW182 family proteins associated with AGO attract the poly(A)-deadenylases PAN2/3 and CCR4-NOT. PAN2/3 begins the deadenylation process, while the CCR4-NOT complex finalizes it, resulting in the removal of the m7G cap from the target mRNA by the decapping machinery. The decapped mRNA can then be subjected to 5′-3′ degradation by the exoribonuclease XRN1. 223 (D) AGO2, along with another miRNA-associated protein known as FXR1, is associated with AREs located in the 3′ UTR to facilitate translation. This process depends on AGO2 and FXR1 instead of GW182. 224 (E) cRISC, connected to some other factor, including TNRC6A, TARBP2, and DICER1, can enter the nucleus through XPO1 or IPO8. TRNC6A serves as a guiding protein in this mechanism. The nRISC has a composition that resembles the cRISC, but it may also exist solely as an AGO2-miRNA complex (data not shown). The nRISC complex might associate with additional nuclear factors, or the Argonaute protein could form a separate multi-protein complex (e.g., RNA-induced transcriptional silencing (RITS); data not shown). The nRISC interacts with corresponding nuclear transcripts, resulting in the deterioration of transcripts, or (F) AGO is capable of recruiting transcription factors and epigenetic enzymes to a gene’s promoter region, including CMPs, histone demethylases, and histone methyltransferases, to begin the process of histone modification and enhance gene transcription. On the other hand, after the transport of miRNA into the nuclei, AGO proteins assemble inhibitory complexes at the promoter region targeted by the miRNA, which mainly consists of RISC (including AGO and DICER1 proteins), polycomb group (PcG) components (YY1, EZH2, and SUZ12; data not shown), CMPs, and histone deacetylase. This interaction enables the protein inhibitor complex to approach the targeted promoter region more closely, leading to an increase in H3K27me3 modifications and a reduction in H3K4me3 modifications, thereby altering the chromatin structure and creating a non-permissive transcriptional environment. 225 , 226 XPO5, exportin 5; XPO1, exportin 1; TRBP, transactivation response element RNA-binding protein; AGO, Argonaut; eIF4F, eukaryotic initiation factor 4F; CCR4-NOT, carbon catabolite repression 4-negative on TATA-less; XRN1, 5′-3′ exoribonuclease 1; FXR1, fragile X mental retardation-related protein 1; AREs, AU-rich elements; cRISC, cytoplasmic RISC; TRNC6A, trinucleotide repeat-containing adaptor 6A; nRISC, nuclear RISC; RIST, RNA-induced initiation of transcriptional silencing; CMP, chromatin-modifying proteins. Non-canonical ( Figure 1 B) pathways utilize various combinations of proteins from the canonical miRNA biogenesis pathway, including Dicer and AGO2. 12 Broadly, non-canonical miRNA biogenesis pathways might be categorized into those independent of Drosha/DGCR8 and those bypassing Dicer. 2 Pre-miRNAs generated by Drosha/DGCR8-independent pathways resemble substrates for Dicer processing. A notable instance of these pre-miRNAs is mirtrons, such as mir-702 and mir-877, which originated from mRNA introns during the splicing process. Another class includes endogenous short hairpin RNAs (shRNAs), which are transcribed directly as hairpins and processed by Dicer without the involvement of Drosha/DGCR8. 12 An additional instance is the 7-methylguanosine (m7G)-capped pre-miR-320. These newly formed RNAs are exported to the cytoplasm directly via XPO1 or CRM1 (XPO1) in a phosphorylated adaptor for RNA export (PHAX)-dependent manner, bypassing the demand for Drosha cleavage. There is a significant 3p strand bias, likely because the m7G cap blocks the loading of the 5p strand into the AGO protein. 13 Conversely, Dicer-independent miRNAs are developed by Drosha from endogenous shRNA transcripts. These pre-miRNAs, for instance, pre-miR-451, rely on AGO2 for maturation in the cytoplasm, as they lack the necessary length to act as Dicer substrates. 14 This process facilitates the loading of the entire pre-miRNA into AGO2, followed by AGO2-mediated slicing of the 3p strand. The maturation is completed by poly(A)-specific ribonuclease (PARN) via 3′-5′ trimming of the 5p strand. 2 , 15 Most research has demonstrated that miRNAs adhere to specific sequences within the 3′ UTR of target mRNAs, suppressing translational activity, mRNA deadenylation, and decapping. 2 However, other regions of mRNAs have also been shown to contain miRNA-binding sites, such as the 5′ UTR, coding sequences, and promoter regions. 16 The binding of miRNAs to the 5′ UTR and coding sites typically results in gene silencing, as shown in the cancer-promoting function of miR-532-5p in colorectal cancer, which operates through its interaction with the 5′ UTR of RUNX3. 17 In contrast, interactions with promoter regions have been reported to promote transcription. In the PC12 cell line, it is revealed that miR-324-3p enhances RelA gene expression through promoter activation. 18 The core of the miRISC is composed of the guide strand and AGO. The specificity of the miRISC arises from its binding to complementary sequences on target mRNA, which are recognized as miRNA response elements (MREs). The extent of complementarity between the miRNA and MRE determines whether AGO2-dependent mRNA cleavage happens or if miRISC mediates translational suppression and mRNA decay. A complementary interaction between the miRNA and MRE stimulates AGO2’s endonuclease function, resulting in mRNA cleavage. 2 Nevertheless, this complete complementarity weakens the binding between AGO and the 3′ end of the miRNA, enhancing miRNA decay. 19 The majority of miRNA:MRE interactions in animal cells are not entirely complementary, and they contain a minimum of central mismatches with their guide miRNA, which prevents AGO2’s endonuclease function. 20 In many cases, the interaction between miRNA and MRE is facilitated by the 5′ seed site (nt 2–8), as supported by computational and experimental evidence. This seed region is critical for target recognition, with canonical Watson-Crick pairing often centered on nt 2–7, and non-canonical matches, such as G-U wobble base pairs (GU), also play significant roles in miRNA-mediated regulation. 16 Nonetheless, further pairing at the 3′ end enhances the robustness and selectivity of the miRNA-target binding. 21 The assembly of a miRISC commences with the assembly of the GW182 protein family by miRNA-bound AGO. GW182 serves as a scaffold, bringing in other mediator proteins, 22 including the poly(A)-deadenylase complexes PAN2-PAN3 and CCR4-NOT, after the miRNA binds to its target mRNA. 20 PAN2/3 triggers the poly(A) tail deadenylation, which is then accomplished by the CCR4-NOT complex. The collaboration between the tryptophan (W) repeats of GW182 and the poly(A)-binding protein C (PABPC) enhances the efficiency of deadenylation. 20 Following this, decapping occurs with the help of decapping protein 2 (DCP2) and related proteins 22 and is subsequently mediated by the 5′-3′ decay of the miRNA, facilitated by exoribonuclease 1 (XRN1) ( Figure 1 C). 23 While most research focuses on how miRNAs suppress gene expression, the findings of several studies have indicated that miRNAs can upregulate gene expression, as well ( Figure 1 D). In serum-starved HEK293 cells and THP-1 monocytes, AGO2 and another miRNA-associated protein, fragile X mental retardation-related protein 1 (FXR1), interact with AU-rich elements (AREs) in the 3′ UTR of the tumor necrosis factor alpha (TNF-α) mRNA to enhance translation. miRNAs, such as let-7, were discovered to work with AGO2 and FXR1 to promote TNF-α mRNA translation when the cell cycle is halted but to interfere with translation during active cell division. 24 Additionally, elevated levels of gene expression mediated by FXR1a-related microRNP (miRNA-protein complex) have been associated with quiescent cells, such as immature Xenopus laevis oocytes. 25 This activation involves AGO2 and FXR1 rather than GW182. 26 Another example includes miRNAs, such as miR-10a, that bind to the 5′ UTR of mRNAs encoding ribosomal proteins Rps16, Rps6, and Rpl9 during amino acid starvation, 27 indicating that gene activation facilitated by miRNAs is observed under specific conditions. Human AGO2 shuttles between the nucleus and cytoplasm via Importin-8 or XPO1, interacting with TNRC6A (a GW182 family protein) that has nuclear localization and export signals ( Figure 1 E). 28 The miRISC localized in the nucleus has been implicated in gene regulatory processes, with studies showing that nuclear retention of miRNAs is dependent on Ago proteins and seed-matched targets. 29 While the exact mechanisms remain unresolved, nuclear miRNAs may participate in RNA-dependent processes such as splicing or transcriptional regulation. The miRISC also associates with euchromatin at active transcription gene loci. 30 Nevertheless, the current perspective on the timing and mechanisms of miRNA function within the nucleus remains limited. Reports indicate that the low-molecular-weight miRISC interacts with mRNAs in the nucleus, leading to nuclear mRNA decay; however, the exact molecular mechanism is not well understood. 29 The presence of miRNA within transcriptionally active genes suggests that miRISC might engage with target mRNAs during or after transcription. The roles of AGO and Drosha in mRNA splicing support the idea of co-transcriptional interactions between the miRISC and mRNA. Additionally, the miRISC may directly regulate transcription. 2 Havens et al. demonstrated that Drosha, a core component of the microprocessor complex, not only processes pri-miRNAs but also promotes the splicing of a pre-miRNA-like exon in the eukaryotic translation initiation factor 4H (eIF4H) gene. Their findings reveal that Drosha binds co-transcriptionally to structured exons, enhancing their inclusion in spliced mRNAs independently of its cleavage activity. This supports a model where the microprocessor and spliceosome coordinately process shared RNA substrates during transcription. 31 For instance, AGO2 was concentrated in the nucleus of senescent fibroblasts, where it interacted with the miRISC and retinoblastoma (Rb) to inhibit the transcription of Rb/E2F-regulated genes that enhance proliferation. Specifically, let-7f was observed to be binding to MREs in the promoters of E2F target genes CDCA8 and CDC2 in an AGO2-dependent manner. Moreover, some AGO-bound E2F target genes, including ACOX1, AKAP12, and ARNTL2, were upregulated after senescence, and AGO2 was detected by co-immunoprecipitating with euchromatin. 32 A current investigation discovered that nuclear miR-552 interacts with a DNA cruciform configuration (a stem-loop on both sense and antisense DNA strands) within the promoter of CYP2E1, thereby silencing its transcription. 33 Additionally, miRNAs such as miR-24-1 have been found to engage with genomic regions actively transcribing enhancer-derived RNA (eRNA), increasing the mRNA expression levels of nearby genes, such as FBP1 and FANCC, by facilitating a transcriptionally active chromatin state 34 and influencing alternative splicing patterns ( Figure 1 F). 35 While the full contribution of the miRISC to chromatin remodeling and transcriptional regulation remains unclear, current evidence suggests that it may act in a transcription-factor-like capacity. Genome-wide analyses have revealed the regulatory role of RNA modifications in miRNA interactions. For instance, Das Mandal and Ray demonstrated that N6-methyladenosine (m6A) modifications are enriched near miRNA target sites in 3′ UTRs and show a strong positive correlation with miRNA and RBP binding density. Their transcriptome-wide computational analysis indicated that m6A-modified mRNAs are more frequently targeted by miRNAs and RBPs, suggesting a spatial and functional interplay between m6A, miRNAs, and RBPs in post-transcriptional regulation. 36 In addition to their roles in facilitating target mRNA binding, recent studies have explored how chemical modifications affect miRNA biogenesis and function. Adenosine-to-inosine (A-to-I) editing, a prominent post-transcriptional modification of miRNAs, significantly alters miRNA function and target specificity. Large-scale analyses have demonstrated that A-to-I editing can reprogram miRNA target interactions in vivo , with edited miRNAs exhibiting differential silencing efficiencies compared to their unedited counterparts. This modification contributes to transcriptomic complexity and may be crucial in regulating gene expression across diverse biological contexts and disease states. 37 In this review, we define “chemical modifications” as covalent alterations of miRNA nucleotides, including methylation, acetylation, and base conversion, such as pseudouridylation, but excluding non-templated nucleotide extensions (tailing), which are mediated by terminal nucleotide transferases. The foundational investigations published between 2010 and 2017 were instrumental in uncovering the landscape of miRNA chemical modifications, including m6A, m5C, A-to-I editing, and other changes. These earliest studies identified major enzymatic players, such as methyltransferase-like 3 (METTL3), METTL14, NSUN2, adenosine deaminases acting on RNA (ADARs), and other related enzymes, which elucidate their roles in modulating miRNA processing, stability, and silencing efficiency. Over the past few years, the field has experienced significant growth. The most recent studies conducted between 2018 and 2025 have revealed that miRNA chemical modifications are dynamic, context dependent, and functionally relevant in various human disorders, particularly cancer, cardiovascular dysfunction, immune disorders, and neuroinflammation. Advances in epitranscriptomic technologies, such as m6A individual-nucleotide resolution cross linking and immunoprecipitation (miCLIP), m6A sequencing (m6A-seq), and nanopore-based detection, along with CRISPR-mediated editing and miRNA nanodelivery systems, have enabled high-resolution mapping and targeted modulation of modified miRNAs. This review critically integrates these early foundational discoveries with contemporary insights to provide a comprehensive overview of the mechanisms, biological significance, and therapeutic implications of miRNA chemical modifications in human disorders. To ensure a comprehensive and concentrated review, we performed a structured literature search using the PubMed, Scopus, and Web of Science databases for studies published between 1993 and 2025, with the oldest references between 1993 and 2004, amounting to ∼4%, and the highest number of references falls within the years 2017–2021, accounting for ∼39% of the total. Search terms included “miRNA chemical modification,” “miRNA methylation,” “miRNA editing,” “miRNA oxidation,” “miRNA acetylation,” “miRNA pseudouridylation,” and combinations of “modification + miRNA + function/biogenesis/stability.” We included peer-reviewed original research articles, reviews, and experimental studies written in English that addressed chemical modifications of miRNAs with mechanistic insights into biogenesis, stability, post-transcriptional regulation, or their relevance to disease. Exclusion criteria were non-English texts, non-peer-reviewed materials, conference abstracts, and papers lacking specific information on chemical modifications. This strategy resulted in the inclusion of 231 publications, which were subsequently categorized and analyzed based on the type of modification, biological context, and methodological rigor. This review presents a thorough examination of chemical modifications in miRNAs and their regulatory roles in gene expression and human diseases. It examines the effects of these modifications on miRNA biogenesis, stability, and function, highlighting their impact on post-transcriptional gene silencing. The review systematically categorizes and discusses key chemical modifications, including methylation, acetylation, oxidation, pseudouridylation, and RNA editing, along with their associated molecular mechanisms. Additionally, it presents recent findings from various model systems, such as C. elegans , Drosophila , zebrafish, and mammalian cells, to contextualize the evolutionary and functional significance of miRNA modifications. The implications of these modifications across multiple health disorders, including neurological disorders, cardiovascular diseases, and cancer, are critically analyzed ( Table 1 ). Finally, the review identifies areas lacking clarity and offers prospects for future research, emphasizing the need for advanced detection methods and the development of therapeutic strategies targeting chemically modified miRNAs. Table 1 Overview of miRNA chemical modifications in the regulation of specific genes and human diseases MicroRNA Type of chemical modification Position of the nucleotide Modification on pre- or pri- or mature miRNA Target mRNA miRNA upregulation/downregulation Promotion or inhibition the development and/or progression of a specific disease Cell type Reference miR-1 8-oxoguanine (o8G) 7 mature Gata4, Gsk3b, Atp2a2, Gja1, Dram1, Pon2, Pcgf6, Bicd2, HSPB7 upregulation induces cardiac hypertrophy H9c2, rCMC, AC16 Seok et al. 133 miR-30c o8G 4, 5 mature CDKN2C upregulation promotes cardiac fibroblast proliferation NRCFs Chang et al. 125 miR-184 8-oxo-7,8-dihydroguanosine (8OHG) 8 mature Bcl-xL, Bcl-w upregulation promotes apoptosis and myocardial Infarction H9c2 Wang et al. 132 miR-375 N6-methyladenosine 6 pri-miRNA SOX12 downregulation facilitates the proliferation and invasion of BC cells while inhibiting apoptosis T47D, MDA-MB-231, MCF-7 Zhao et al. 187 miR-181b-3p N6-methyladenosine 6 mature ARL5B downregulation facilitates cell migration and invasion in BC SKBR3, MDA-MB453 Pan et al. 188 miR-221-3p N6-methyladenosine 6 pri-miRNA HIPK2 upregulation facilitates adriamycin resistance in BC MCF-7 Pan et al. 188 miR-374c-5p N6-methyladenosine 6 pri-miRNA GRM3 downregulation enhances BC cell proliferation, migration, and invasion by upregulating GRM3 expression in response to Cd exposure T-47D, MCF-7 Yue et al. 77 miR-1246 N6-methyladenosine 6 pri-miRNA SPRED2 upregulation promotes cell migration, invasion, and metastasis in CRC sDLD-1, HCT116, LoVo Peng et al. 189 miR-1246 N6-methyladenosine 6 pri-miRNA CCNG2 upregulation facilitates the initiation and metastasis of ovarian cancer ES2, OVCAR3, A2780, SKOV3 Bi et al. 190 miR-25-3p N6-methyladenosine 6 pri-miRNA PHLPP2 upregulation cigarette smoke facilitates the progression of PC HPDE6-C7, PANC-1, BXPC-3 Zhang et al. 139 Let-7b-5p N6-methyladenosine 6 pri-Let-7b Notch signaling upregulation suppresses Notch signaling and restores sensitivity to treatment H1975, HCC827 Li et al. 191 miR-106b N6-methyladenosine 6 pri-miRNA SFRP2 upregulation strengthens stemness and facilitates cell proliferation and migration and tumor growth of LUAD A549, H1975 Rong et al. 78 miR-663 N6-methyladenosine 6 pri-miRNA SOCS6 upregulation facilitates cell proliferation, migration, invasion, and tumor growth and inhibits apoptosis A549, PC9 Li et al. 191 miR-486 N6-methyladenosine 6 pri-miRNA RAP1 upregulation propofol enhances DDP sensitivity A549, SKMES Ling et al. 192 miR-576 N6-methyladenosine 6 pri-miRNA CDK6 downregulation enhances bladder cancer cell proliferation, migration, and invasion T24, UM-UC-3 Zhou et al. 193 miR-30d N6-methyladenosine 6 pri-miRNA RUNX1 upregulation impedes aerobic glycolysis, represses tumor growth and metastasis, and diminishes Warburg effect Panc-1, MiaPaCa-2 Zhou et al. 193 miR221/222 N6-methyladenosine 6 pri-miRNA PTEN upregulation promotes cell proliferation and bladder tumorigenesis EJ, T24 Rong et al. 78 miR-222-3p N6-methyladenosine 6 pri-miRNA STK4 upregulation accelerates malignant phenotypes of thyroid carcinoma cells TPC-1, SW579 Klinge et al. 84 miR-143-3p N6-methyladenosine 6 pri-miRNA DDX6 downregulation ALKBH5 represses HASMC proliferation, enhances HAEC apoptosis, and facilitates AD progression KIAA1429 enhances HASMC proliferation, inhibits HAEC apoptosis, and inhibits AD progression HASMCs, HAECs Wang et al. 194 miR-19a N6-methyladenosine 6 pri-miRNA promotes cardiovascular endothelial cell proliferation and invasion upregulation facilitates the proliferation and invasion of ASVEC ASVEC Zhang et al. 195 miR-29b-3p N6-methyladenosine 6 pri-miRNA HDAC4 downregulation (throughout the process of osteogenic differentiation) upregulation (throughout the process of adipogenic differentiation) inhibits osteogenesis and promotes adipogenesis in patient with OP and OVX mice BMMSCs, RAW264.7 Liu et al. 140 miR-25-3p N6-methyladenosine 6 mature PTEN downregulation high glucose inhibits RPE cell proliferation and promotes cell apoptosis and pyroptosis ARPE-19 Zha et al. 196 miR-93 N6-methyladenosine 6 pri-miRNA DUSP2 upregulation CS causes inflammation and rises emphysema HBE, CHBE, THP-M Xia et al. 197 miR-126 N6-methyladenosine 6 pri-miRNA facilitates the migration and invasion of human endometrial stromal cells downregulation enhances the migration and invasion of endometrial stromal cells and endometriosis progress HESCs Li et al. 198 miR-365-3p N6-methyladenosine 6 pri-miRNA produces pain-related behaviors and neuronal sensitization in naive mice upregulation facilitates CFA-induced behavioral hypersensitivity and spinal neuron sensitization spinal cells Zhang et al. 199 miR-380-3p N6-methyladenosine 6 mature PTEN upregulation facilitates cell proliferation, migration, and EMT in PC PANC1, SW1990, Capan-2, AsPC-1, BXPC-3 Jiang et al. 200 miR-143/145 N6-methyladenosine 6 pri-miRNA KLF4 (by miR-145-5p), FSCN1 (by miR-143-3p) upregulation prevents the phenotypic transition of PASMCs PASMCs Fang et al. 201 miR-34-5p N6-methyladenosine 6 pri-miRNA SIRT1 upregulation promotes cell-cycle arrest and senescence HNPCs, NP Zhu et al. 70 miR-221/222 N6-methyladenosine 6 pri-miRNA DKK2 upregulation enhances Ang-II-induced myocardial hypertrophy NRCMs Zhang et al. 199 miR-143 N6-methyladenosine 6 pri-miRNA Yap, Ctnnd1 downregulation promotes neonatal CM proliferation, heart regeneration, and its cardiac function after MI AC-16, primary neonatal ventricular myocytes Gong et al. 202 miR-17-5p N6-methyladenosine 6 pri-miRNA MFN2 upregulation results in reduced apoptosis and decreased drug sensitivity to 5-FU and induces chemoresistance in CRC HCT116, SW480, SW620 Sun et al. 203 miR-150 N6-methyladenosine 6 pri-miRNA BDNF downregulation enhances neuropathic pain progression RN-sc Zhang et al. 204 miR-21 N6-methyladenosine 6 pri-miRNA PTEN upregulation stimulates the differentiation of lung-resident mesenchymal stem cells into myofibroblasts LR-MSCs Lu et al. 205 miR-21 N6-methyladenosine 6 pri-miRNA activates the SPRY1/ERK/NF-κB signaling transduction upregulation facilitates obstructive renal fibrosis development HK-2 Liu et al. 206 miR-320a-3p N6-methyladenosine 6 pri-miRNA FOXM1 downregulation promotes silica-induced lung fibrosis MRC-5, NIH/3T3 Sun et al. 76 miR-335 N6-methyladenosine 6 pri-miRNA Erf1 downregulation inhibits SG formation while promoting apoptosis in injured neurons and cells primary cortical neurons, PC12 Si et al. 207 miR-92b-5p N6-methyladenosine 6 pri-miRNA TIMP4 upregulation accelerates osteoarthritis progression chondrocyte, SW1353 Lin et al. 208 miR-146 N6-methyladenosine 6 pri-miRNA NUMB, NOTCH2 upregulation counteracts melittin-induced tumor-suppression bladder cancer T24, EJ, SV-HUC-1, BIU87 Yan et al. 209 miR-125b2 N6-methyladenosine 6 Pre-miRNA Gab2 downregulation promotes cancer cell migration A549, SW620, HT-29 Yang et al. 83 miR-126 N6-methyladenosine 6 pri-miRNA metastasis enhancement downregulation promotes the metastatic potential of HCC SMMC-7721, HCC 1664, Hep3B, HepG2 Zhang et al. 71 miR-146a-5p N6-methyladenosine 6 pri-miRNA promotes migration and invasion upregulation promotes cell invasion and migration in BC MCF-7, MDA-MB-231 Yi et al. 210 miR-320a N6-methyladenosine 6 pri-miRNA PIK3CA upregulation promotes osteomyelitis progression hBMSCs Gao et al. 211 miR-589-5p N6-methyladenosine 6 pri-miRNA promotes cell viability, migration, and invasion upregulation facilitates the viability, migration, and invasion of liver cancer cells Hep3B, SK-Hep1 Liu and Jiang 212 miR-99a N6-methyladenosine 6 pri-miRNA TRIB2 downregulation promotes CSC persistence and induces radioresistance in ESCC cells Eca109, TE-1, EC9706, Kyse150, Kyse410 Liu et al. 213 miR-194-2 N6-methyladenosine 6 pri-miRNA RAI1 upregulation promotes the progression of esophageal cancer esophageal cancer with ALKBH5 downregulation is more sensitive to verteporfin therapeutic Kyse150, Eca109, Kyse140, Kyse510 Chen et al. 214 miR-92b N6-methyladenosine 6 pri-miRNA PTEN downregulation DCA suppresses GBC tumor growth NOZ, GBC-SD Lin et al. 215 miR-126-5p N6-methyladenosine 6 pri-miRNA PTEN upregulation facilitates cell proliferation, migration, and invasion and inhibits apoptosis in ovarian cancer A278, SKOV3, COV504, ES2 Bi et al. 216 miR-19a N6-methyladenosine 6 pri-miRNA BAMBI upregulation enhances the growth and invasiveness of nasopharyngeal carcinoma cells C666-1, HNE3, NPC53 Gong et al. 217 miR-9-5p N4- acetylcytidine 4 pri-miRNA SH3BP4, NCOR2, LMNA, EPAS1, TES upregulation promotes cancer initiation and progression A549, H1299, 293T, DU145 Zhang et al. 124 Let-7a 2′-O-methylation 2 pri-miRNA stimulate cell growth and invasion while inhibiting the rise in apoptosis downregulation facilitates cell proliferation and invasion and impedes apoptosis in CRC, linked to a more advanced TNM stage and lymph node metastases, suggesting a worse prognosis HCT116, HCT8 Bian et al. 58 miR-455-5p adenosine to inosine 2, 17 mature CPEB1 (WT), RHO-C, MDM4, and integrin α2 (ED) downregulation promotes melanoma growth and metastasis SB2, C8161, MeWo Shoshan et al. 107 miR-487b adenosine to inosine 2 mature DNAJC9, B3GALNT2, and MAP2K4 (WT), RPS6KB1 and BMP1 (ED) upregulation enhances neovascularization after ischemia through the modification of target gene selection HUASMC, HUVEC, HUAF van der Kwast et al. 218 miR-200 adenosine to inosine Not mentioend mature trigger EMT in tumor cells downregulation facilitates EMT and liver metastasis in CRC SW480, SW620, RKO, HCT116 Shelton et al. 219 miR-411–5p adenosine to inosine 5 mature MET downregulation enhances TKI sensitivity in NSCLC-resistant cells H1299, H520, A549, HCC827, PC9 Romano et al. 105 miR-222 adenosine to inosine −4, 53 pre-miRNA p27 upregulation promotes glioblastoma proliferation and migration U118, A172 Tomaselli et al. 220 miR-221 adenosine to inosine −1, 1, 34, 64, 187 pre-miRNA p27 upregulation promotes glioblastoma proliferation and migration U118, A172 Tomaselli et al. 220 miR-21 adenosine to inosine 16, 46, 51 pre-miRNA PDCD4 upregulation promotes glioblastoma proliferation and migration U118, A172 Tomaselli et al. 220 miR-379-5p adenosine to inosine 10 mature PTK2 (WT), CD97 (ED) upregulation inhibits cancer cell proliferation, tumor growth, extended survival, and promotes apoptosis MDA-MB-231, OVCAR-8, 786-O, A549, HeyA8, PC-3 Xu et al. 106 ; Kawahara et al. 118 miR-200b adenosine to inosine 5 mature ZEB1 and ZEB2 (WT), LIFR (ED) upregulation promotes cell invasion and migration MCF10A, SLR25, OVCAR8, MDA-MB-231, 786-O, HeyA8, Hs578T Wang et al. 142 let-7 adenosine to inosine 3 pri-miRNA ACVR2B, HRAS, FGFR3 downregulation enhances LSC self-renewal K562, CD34 + cord blood cells Zipeto et al. 221 BC, breast cancer; Cd, cadmium; CRC, colorectal cancer; LUAD, lung adenocarcinoma; DDP, displatin; ALKBH5, alkB homolog 5 RNA demethylase; GBC, gallbladder carcinoma; AD, aortic dissection; OP, osteoporosis; OVX, ovariectomized rodent; CS, cigarette smoke; CFA, complete Freund’s adjuvant; EMT, epithelial-mesenchymal transition; PC, pancreatic cancer; Ang-II, angiotensin II; CMs, cardiomyocytes; MI, myocardial infarction; 5-FU, 5-fluorouracil; SG, stress granule; HCC, hepatocellular carcinoma; CSC, cancer stem-like cells; ESCC, esophageal squamous cell carcinoma; DCA, deoxycholic acid; TKI, tyrosine kinase inhibitor; NSCLC, non-small cell lung cancer; LSC, leukemia stem cell; NRCF, neonatal rat cardiac fibroblasts; HASMC, primary human aortic smooth muscle cells; HAEC, primary human aortic endothelial cells; ASVEC, aortic smooth vascular endothelial cells; RPE, retinal pigment epithelium; HBE, human bronchial epithelial cells; CHBE, chronic obstructive pulmonary disease (COPD) human bronchial epithelia; THP-M, THP-1 cells differentiated into macrophages; HESC, human endometrial stromal cells; PASMC, primary pulmonary artery smooth muscle cells; HNPC, human nucleus pulposus cells; NP, nucleus pulposus; NRCM, neonatal rat cardiomyocytes; RN-sc, rat neurons-spinal cord; LR-MSC, lung-resident mesenchymal stem and stromal cells; NOZ, human gall bladder carcinoma cell line; TNM, tumor, node, metastasis; HUASMC, human umbilical artery smooth muscle cells; HUVEC, human umbilical vein endothelial cells; HUAF, human umbilical arterial fibroblasts; WT, wild-type; ED, extracellular domain. Overview of miRNA chemical modifications in the regulation of specific genes and human diseases BC, breast cancer; Cd, cadmium; CRC, colorectal cancer; LUAD, lung adenocarcinoma; DDP, displatin; ALKBH5, alkB homolog 5 RNA demethylase; GBC, gallbladder carcinoma; AD, aortic dissection; OP, osteoporosis; OVX, ovariectomized rodent; CS, cigarette smoke; CFA, complete Freund’s adjuvant; EMT, epithelial-mesenchymal transition; PC, pancreatic cancer; Ang-II, angiotensin II; CMs, cardiomyocytes; MI, myocardial infarction; 5-FU, 5-fluorouracil; SG, stress granule; HCC, hepatocellular carcinoma; CSC, cancer stem-like cells; ESCC, esophageal squamous cell carcinoma; DCA, deoxycholic acid; TKI, tyrosine kinase inhibitor; NSCLC, non-small cell lung cancer; LSC, leukemia stem cell; NRCF, neonatal rat cardiac fibroblasts; HASMC, primary human aortic smooth muscle cells; HAEC, primary human aortic endothelial cells; ASVEC, aortic smooth vascular endothelial cells; RPE, retinal pigment epithelium; HBE, human bronchial epithelial cells; CHBE, chronic obstructive pulmonary disease (COPD) human bronchial epithelia; THP-M, THP-1 cells differentiated into macrophages; HESC, human endometrial stromal cells; PASMC, primary pulmonary artery smooth muscle cells; HNPC, human nucleus pulposus cells; NP, nucleus pulposus; NRCM, neonatal rat cardiomyocytes; RN-sc, rat neurons-spinal cord; LR-MSC, lung-resident mesenchymal stem and stromal cells; NOZ, human gall bladder carcinoma cell line; TNM, tumor, node, metastasis; HUASMC, human umbilical artery smooth muscle cells; HUVEC, human umbilical vein endothelial cells; HUAF, human umbilical arterial fibroblasts; WT, wild-type; ED, extracellular domain.

The authors declare that no conflicts of interest exist.

We thank all the researchers who have advanced the field and apologize to those whose work was not cited due to space constraints. We also thank the members of our laboratory for their helpful comments and are grateful for the financial support from the 10.13039/100009326 Cancer Research Society , Canada (operating grant IDs: 1052403 and 1278222 ).