Small non-coding RNAs (ncRNAs), which range in length from 20 to 22 nucleotides, including microRNAs (miRNAs), are among the most conserved ncRNAs. Several studies have confirmed that miRNAs are critical modulators of gene expression (1–3) (namely, proliferation, differentiation, apoptosis, and stress reactivity) and numerous diseases (4–7) (namely, neurological diseases, cardiovascular diseases, cancer, and aging). The nucleus and cytoplasm are involved in the multi-step process of miRNA synthesis. The most primitive miRNAs have long primary miRNAs (pri-miRNAs) with caps and polymeric A tails. A protein complex and microprocessor subsequently process these to form precursor miRNAs (pre-miRNAs). The pre-miRNAs are then released into the cytoplasm using Exportin5 (XPO5), where they are cleaved by RNaseIII type protein (Dicer) to produce double-stranded miRNAs. One of the two miRNA strands matures in the cytoplasm and forms the RNA-induced silencing complex (RISC), while the other degenerates instantly (8, 9). Since miRNAs only mature in the cytoplasm, it was previously determined that miRNA action was cytoplasm-specific. However, identifying miRNAs in the subcellular compartments of the nucleus reveals various novel miRNA-based functions in cellular homeostasis, which has been the subject of an increasing number of studies, some of which used data from small RNA deep sequencing (10). In this article, we described how nuclear miRNAs affected gene transcription and elaborated on the associated signal transduction in the nucleus.

Pre-miRNAs are released into the cytoplasm by a specific nuclear transporter (i.e., XPO5) in the classical miRNA processing pathway, where they mature. Therefore, scientists initially speculated that mature miRNAs were restricted to the cytoplasm. The modulatory role of miRNA in the nucleus has been verified by emerging evidence from recent reports (11). The majority of nuclear miRNAs articles that have been published are summarized in Table 1.

Meister et al. showed in 2004 using Hela cell that mature miR-21 can be found in the cytoplasm and the nucleus but that 70-nt pre-miR-21 is only found in the nucleus. This unique discovery of mature nuclear miRNA sparked the hypothesis that some mature nuclear miRNA can re-enter the nucleus (12). In a subsequent investigation, Politz et al. presented visual evidence of mature miRNA expression in the nucleus (13). They used nucleic acid probe based in situ hybridization analysis, in particular, to reveal that miR-206 aggregated in the nucleoli. However, a probe complementary to the precursor miR-206 showed no nucleolar signal. Moreover, they found that the granular component was the precise location of nuclear miRNA expression. In 2009, Papp et al. used femtosecond laser microscopy to investigate multiple single living cells exploiting the association between super quencher molecular beacon probes and human single-stranded cellular miR-122 targets. For the first time, they demonstrated that mature miR-122 penetrates the nucleus of human hepatocytes after being assembled in the cytoplasm (14). Massive advancements in high-throughput analyses over the past ten years have facilitated the discovery that mature nuclear miRNAs are widespread in most mammalian cells (15, 16). In 2019, Turunen et al. used a mouse endothelial cell line to perform nuclear-cytoplasmic fractionation and small RNA sequencing analysis. They identified 196 miRNAs (56% of all miRNAs), which were exclusively distinct between the nucleus and the cytoplasm. This suggested that miRNAs had a compartmentalized nature. Here, both compartments demonstrated a ≥2-fold enrichment of 105 miRNAs (specifically, nucleus: 46 and cytoplasm: 59). According to the M versus A plot analysis, the differential enrichment was apparent throughout a broad range of absolute expressions (i.e., counting). Nuclear miR-3535 (27.9-fold enrichment) and cytoplasmic miR-27a-5p (48-fold enrichment) were the top compartment-specific miRNAs. The content of miR-210-3p elevates in the nucleus after hypoxia (17). Given this extensive body of evidence, there is good reason to speculate that miRNAs have a functional role in the nucleus.

It is well known that miRNAs co-localize with argonaute (AGO) proteins in the nucleus, which is strongly suggestive of a role for miRNAs in modulating gene expression. In recent years, many nuclear miRNAs functions have been reported, including the direct regulation of miRNAs biogenesis and association with target gene promoters or enhancers. Figures 1, 2 present a summary of the reported functions of nuclear miRNAs.

According to recent studies, pre- and mature miRNAs were found in significant amounts (27). Determining the potential roles of the nucleolar miRNAs has thus given rise to multiple hypotheses.

In 2014, Reyes-Gutierrez et al. used bioinformatic analysis to estimate canonically structured and thermodynamically stable associations between the IGF2 transcript and all five nucleolus-based miRNAs. Based on this, the nucleolus is a primary location for targeted mRNA-miRNA associations prior to cytoplasm transport (28). Before miRNAs are redistributed to the cytoplasm under genotoxic stress to initiate assembling the RISC complex as part of the genomic defense system, they may be stored in the nucleoli.

In 2016, Atwood et al. found that the smallest RISC (AGO2 and miRNA) may impact the ribosomal RNA (rRNA) in cells (50). AGO2-rRNA association is highly dependent on miRNA and sustained PoII activity, as shown by the minimal RISC to the sensitivity of 45s rRNA interaction to both Dicer suppression and actinomycin D exposure. It is possible to hypothesize the particular properties of mature ribosomes. Additionally, it has been shown that pri- and pre-miRNA potential undergo nucleolar A-to-I editing because the nucleoli contain many RNA editing enzymes, namely, adenosine deaminase. This alteration suppresses miRNA maturation, which minimizes the cellular availability of mature miRNAs (51).

In conclusion, the model mentioned above-associated miRNA with nucleoli, indicating a strong necessity for the miRNA’s nucleolar location for the cytoplasmic post-transcriptional regulation of target transcripts.

It is still unknown how nuclear miRNAs-regulate gene expression in this manner. Recent research, however, has supported numerous potential processes thought to be involved in this process.

AGO2-related genes were significantly enriched with MAZ docking locations and neural function, according to a 2017 study by Goldie et al. The AGO1 mRNA, in contrast, associates with SC35 spliceosomes and is connected to the general metabolic processes (52). This suggests that the MAZ TF is associated with nuclear miRNA and affects the control of neuronal development through the AGO2-related miRISCs. The MAZ TF may be crucial to organize higher-level correlations between transcriptional and post-transcriptional controls in primate neurons.

In 2021, Guo et al. revealed 2021 that the TFEB-induced transactivation is greatly enhanced by miR-30b-5p knockdown using CRISPR/Cas9, thereby upregulating genes involved in autophagy and lysosomal biogenesis (29). They discovered that the miR-30b-5p interacts with the nuclear CLEAR elements to modify lysosomal biogenesis and autophagy, presumably suppressing TFEB-based transactivation.

In the mouse endothelial cell line C166, Laitinen et al. discovered in 2022 that the nuclear microRNA miR-466c regulates the production of vascular endothelial growth factor a (VEGFA) when the environment is hypoxic (30). Additionally, the miR-466c deletion in the CRISPR-Cas9 genome greatly hinders the hypoxia-induced increase in VEGFA content. They also discovered long non-coding RNAs linked to the murine VEGFA promoter and hypothesized that miR-466c physically interacts with the transcript to modify VEGFA expression.

However, because nuclear miRNA is a novel gene modulator, more research is necessary to understand the networks in miRNA-mediated control of target gene expression.

The function and underlying mechanism of nuclear miRNAs are currently being studied. Previous research showed that the human miR-29b is largely found in the nucleus compared to other animal miRNAs. The unique hexanucleotide terminal motif of miR-29b functions as a transferable nuclear localization element that regulates either its own or linked siRNAs nuclear enrichment.

It also explains the highly redundant common miRNAs 5′ sequences, which may play independent roles in response to cis-acting modulatory motifs (53). The IPO8 was also found to have a significant role in transporting mature miRNA from the cytoplasm to the nucleus. Without changing the total number of miRNAs present in the cell, IPO8 knockdown, in particular, significantly reduces the nuclear translocation of certain miRNAs that are generally enriched in the nucleus. The IPO8 and AGO2 complex interaction is also necessary for the nuclear translocation of mature miRNAs mediated by IPO8. AGO2 nuclear transfer is likewise reduced by IPO8 knockdown without changing the overall levels of AGO2 cells. The effectiveness of IPO8-mediated miRNA nuclear transport was significantly reduced when the combination of miRNAs and AGO2 was dissociated (54). A single molecule fluorescence-based evaluation of the subcellular translocation, integrity, and activity of miRNAs was developed by Sethuramasundaram et al. They discovered that where the selection, melting, and nuclear retention of the miRNA chain is dependent on AGO identity, the stability and activity of the miRNA are strongly influenced by the amount of AGO protein. These findings suggest that miRNA degradation competes with AGO loading and target association to modify the subcellular miRNA content for monitoring gene silencing (55).

Furthermore, Santovito et al. showed that miR-126-5p interacts with Mex3a on the surface of autophagic vesicles to create a miR-126-5p/AGO2 complex, enabling it to reach the nucleus. Meanwhile, it was discovered that Mex3a connects with the central U- and G-type regions of miR-126-5p with nanomolar affinity utilizing mutational and biophysical analyses. The miR-126-5p/AGO2 complex dissociates upon entry into the nucleus, and miR-126-5p then binds with caspase-3 in the same manner as its seed sequence. This limits the ability of caspase to regulate cell death and prevents caspase dimerization (56). Notably, the mechanism driving the nuclear enrichment of miRNAs is tightly linked to several widely used signaling networks. Normally, miR-133a translocates from the canonical Wnt pathway to the cardiomyocyte nucleus. The nuclear miR-133a/AGO2 complex interacts with the DNA methyltransferase 3B (DNMT3B) promoter’s complementary miR-133a target site, causing DNMT3B to repress the transcription of itself (25).

Separate nuclear and cytoplasmic compartments are used for the production of miRNAs. As reported earlier, miRNA maturation depends on the nuclear transfer of precursors. Pre-miRNAs frequently use XPO-5 as a carrier. the family of importin-β/karyopherin includes XPO-558 (57). The mechanisms behind the pre-miRNA specificity of XPO-5 have already been elucidated by structure-based research (58). Similar to other importin-β/karyopherin families of proteins, XPO-5 is composed of a Huntingtin-elongation-A subunit-TOR repeat sequence consisting of two antiparallel α-sheets linked by a turn and appears as a spiral structure overall. The importin-β/karyopherin family of proteins can adapt to various vectors due to the inherent flexibility of this configuration (59). XPO-5 recognizes pre-miRNAs with similar structural profiles as a nuclear transfer signal (60). Therefore, the biosynthesis of miRNAs includes nuclear transportation as an integral part. Pre-miRNAs require a particular and essential transport receptor, and XPO-5 is present in most investigated organisms (61). However, emerging evidence revealed that embargoes (EMB; an immediate relative of XPO-1), another export receptor for leucine-rich NES proteins, influence the nuclear-cytoplasmic delivery of pri-miRNAs in the nematode Caenorhabditis elegans, which lacks an immediate relative of XPO-5 (62),. Although the role of EMB in this process is unknown, growing evidence suggests that the different organisms may use various miRNA-mediated nuclear translocation pathways. More recently, it was discovered that decreased RANBP1 significantly reduces the expression of hsa-miR-18a, hsa-miR-183, and hsa-miR-106 (miRNAs) in colorectal cancer by preventing the nuclear transfers of the corresponding pre-miRNAs. This facilitates the pre-miRNAs to accumulate in the nucleus, which lowers the amount of mature miRNA present. Evidence suggests that RANBP1 modulates pre-miRNA nuclear transportation to affect colorectal cancer cell proliferation, invasion, and apoptosis (63). Fungal virulence in plants depends on the secretory protein VdSSR1 (secreted silencing blocker 1), which is produced from Flavobacterium flavum. Zhu et al. demonstrated that VdSSR1 chelates the ALY family of proteins, potent modulators of the TREX complex, to modulate the nuclear transfer of the AGO1-miRNA complex. This increases fungal toxicity in plants by lowering the cytoplasmic AGO1 protein and sRNA contents (64).

According to a previous study, AGO proteins interacted with a GW protein family member known as TNRC6A-C in mammals, which coordinates downstream gene-silencing processes (65). The cytoplasmic functions of TNRC6 and AGO proteins are rather well known. The nucleus contains both protein families as well. Contrary, the epigenetic modifiers were recruited by RISC and guided by miRNAs to particular loci in the genome (66). These findings improved the role that proteins play in nuclear miRNA functions.

Scientists are gradually learning new roles and functions of these small RNAs due to an extensive investigation into the function of nuclear miRNAs in recent years. According to Indrabahadur et al., the nuclear miR-let-7d joins with ncRNA to create the MiCEE complex, which causes the epigenetic silencing of bidirectional gene expression in tissues with a large genome. It is important as a key regulator of bidirectional gene transcription is thus confirmed by this (67). Furthermore, extracted chromatin RNA and large parallel sequencing revealed that let-7i miRNA substantially inhibits norepinephrine transporter via association with methyl-CpG-binding protein 2 (68).

Ohno et al. discovered that the nuclear miRNAs separate stalled Pol II from the DDX21-CDK9 complex in 2022. The aforementioned researcher’s ZMYND10 activation caused by miR-34a as a system of RNA activation (RNAa) assessment to find a significant regulator of RNAa (31). Numerous investigations have also shown that nuclear miRNA regulates disorders related to glycolipid metabolism. By altering LXRα and FXR transcriptional activities in the nucleus, miR-552-3p worsens hepatic glycolipid metabolic disorders. This offers proof that miR-552-3p is an effective therapeutic target against metabolic diseases (69).

Although the accepted thinking was that miRNAs govern mRNA stability and translation in the cytoplasm, subsequent studies have shown that they also have a regulatory role in the nucleus for transcription that involves enhancers and promoters.

An independent effect approach is the biggest drawback of the current study on nuclear miRNAs. It is essential to develop a more efficient experimental system to examine the importance and purpose of nuclear miRNAs. Additionally, the location, movement, and purpose of nuclear miRNAs could be clearer. Furthermore, it is still unclear which possible signaling pathway nuclear miRNA regulates. Therefore, additional study into these areas will undoubtedly result in a breakthrough, opening up new possibilities for creating nuclear miRNA-based therapies for treating human diseases.

XH (first author): performed the data analyses and wrote the manuscript; GY: contributed to the conception of the study; YZ: data curation; LZ: contributed significantly to analysis and manuscript preparation; HH: helped perform the analysis with constructive discussions. All authors contributed to the article and approved the submitted version.