2.1 Advancements and applications of RNA sequencing technology (RNA-Seq) in miRNA research

RNA-Seq is a high-throughput method for comprehensive analysis of all RNA molecules in biological samples.⁸ miRNA-seq, a specialized form of RNA-Seq for miRNAs, provides detailed expression profiles, highlights differences, and discovers new miRNAs, enhancing our understanding of gene regulatory networks (Figure 1A).⁹ Single-cell RNA-seq (scRNA-seq) studies cell subpopulation heterogeneity, offering insights into cell diversity, fate decisions, and disease mechanisms.¹⁰

Bulk RNA-seq data is widely used as a prognostic marker in cancer studies. In bladder cancer (BLCA), researchers developed a prognostic model based on three gene expression levels to predict patient survival.¹¹ In pancreatic ductal adenocarcinoma (PDAC), RNA-seq analysis of 24 tissues revealed over 57,000 single-cell transcriptomes, uncovering links between ductal cell subpopulations and the inactivation state of tumor-infiltrating T lymphocyte (T cell). This study provided valuable insights into PDAC heterogeneity and tumor-immune interactions, aiding in the development of new immunotherapy strategies.¹⁰

Technological advancements and cost reductions¹² have transitioned RNA-Seq from tissue-level to single-cell analysis, broadening its applications. High-throughput technologies like PANDORA-seq overcome RNA modification challenges, offering comprehensive maps of highly modified miRNAs.¹³ Nanopore MinION RNA-Seq, using Oxford Nanopore Technology, provides long-read sequences and single-molecule real-time sequencing, further advancing RNA-Seq.¹⁴

RNA-Seq technology provides rich data and analysis methods for miRNA research, deepening our understanding of miRNA interactions and regulatory networks. Its application in cancer research is particularly significant, offering new methods for selecting diagnostic and therapeutic biomarkers, and demonstrating substantial clinical potential.

2.2 CRISPR/Cas9 revolutionizes miRNA research in cancer

CRISPR/Cas9 is a revolutionary gene editing tool that enables precise regulation of miRNA expression, offering specificity and long-term stability.¹⁵ By knocking out or knocking in specific miRNAs, CRISPR/Cas9 has advanced our understanding of miRNA functions and interactions (Figure 1B).¹⁶ However, the effectiveness and adaptability of CRISPR/Cas9 for treatment are limited by its toxicity and potential mutagenicity.¹⁷ The unique RNase activity of short CRISPR RNA (crRNA) (40 nt) processed into mature Cas12a pre-crRNA helps overcome the size challenge of delivery via viral vectors.¹⁸

In cancer research, CRISPR screening identifies key genes related to cancer development and treatment response, improving our grasp of drug resistance mechanisms and guiding new treatment strategies.¹⁹ The third generation of CRISPR technology has been continuously improved, achieving direct targeted DNA modification without generating DNA double-strand breaks, precise insertion of large DNA sequences, and gene regulation without any genome editing through epigenome engineering, thereby improving safety.²⁰ The development of CRISPR-Cas9 and high-throughput sequencing technologies enhances our understanding of miRNA regulation in cancer, promoting targeted therapies.²¹

In colorectal cancer (CRC), scientists used CRISPR/Cas9 to inactivate miR-34a and miR-34b/c in the HCT116 cell line. The simultaneous inactivation enhanced cell migration and invasion, activated epithelial-mesenchymal transition (EMT), reduced chemotherapy sensitivity, and increased stress-induced autophagic flux, highlighting the complementary roles of miR-34a and miR-34b/c in regulating EMT and autophagy, providing new treatment insights.²²

2.3 Tools and algorithms for predicting miRNA targets and interactions

In recent years, miRNA target prediction and interaction identification have advanced significantly. Tools and algorithms have evolved, integrating machine learning to align with disease development trends.²³ These tools utilize miRNA-mRNA complementarity, pairing stability, and conservation to efficiently predict target genes. Popular tools for predicting miRNA–mRNA interactions include miRanda,²⁴ miRBase,²⁵ TargetScan,²⁶ PicTar,²⁷ RNAhybrid,²⁸ TarBase,²⁹ RNA22,³⁰ GenMiR++,³¹ PolymiRTS,³² miRDB,³³ miRGator,³⁴ mirecords,³⁵ mirSOM,³⁶ mirDIP,³⁷ miRTarBase,³⁸ psRNATarget,³⁹ MiRTDL,⁴⁰ microT,⁴¹ miSTAR,⁴² miRGen,⁴³ PolymiRTS,³² miRPath,⁴⁴ and StarBase.⁴⁵ Additional tools include miRWalk⁴⁶ and databases based on immunoprecipitation methods (IB), such as miRTarCLIP,⁴⁷ miRBShunter,⁴⁸ and miRTar2GO.^{23, 49} Tools specific to lncRNA–miRNA interactions include LncMirNet28⁵⁰ and miRcode,⁵¹ while circRNA–miRNA interaction tools include CircNet 2.0,⁵² CSCD2,⁵³ Circbank,⁵⁴ CircInteractome,⁵⁵ and BCMCMI⁵⁶ (Figure 1D).

The interplay between miRNAs and transcription factors (TFs) is pivotal. miRNAs target transcription factor mRNAs, while transcription factors reciprocally regulate miRNA gene expression, exerting regulatory roles pre- and post-transcriptionally.⁵⁷ Tools for predicting TF–miRNA interactions are primarily based on gene expression profiles, sequence-based target predictions, experimentally validated target databases, and phylogenetic footprinting methods^{58, 59} Among these, De novo predictions utilize sequence-based methods, while experimentally validated databases often employ CHIP-chip technology to identify TF binding sites in DNA. Emerging experimental techniques, such as AgoHIST-CLIP, offer innovative approaches to miRNA target gene prediction.⁶⁰ Additionally, some studies are developing a tool called mirConnX, which integrates regulatory networks of miRNA, TFs, and mRNA.⁶¹

However, improving the accuracy of miRNA target prediction remains a challenge. Researchers are integrating multiple biological features and leveraging single-cell sequencing data along with deep learning methods to enhance prediction accuracy.⁶² Analyzing extensive scRNA-seq data has unveiled critical insights into cancer biology, disease pathogenesis, and clinical outcomes, facilitating the development of prognostic models.^{11, 63}

2.4 Integrating multi-omics data to map miRNA networks in cancer

The regulatory role of miRNA in cells is intricate and nuanced, capable of integrating multi-omics data to construct complex biological networks (Figure 1C). This approach offers researchers a comprehensive view to unravel its mechanisms and regulatory pathways.⁶⁴

miRNA plays a crucial role in cancer research across multiple omics fields, including genomics, transcriptomics, proteomics, and metabolomics. By integrating and analyzing data from various omics, we gain a more comprehensive understanding of biological processes and disease mechanisms.⁶⁵ At the genomic level, miRNA genetic variation can influence disease susceptibility and phenotypic diversity.⁶⁶ At the transcriptomic level, bioinformatics, statistical methods, network inference, and experimental validation of mRNA networks systematically reveal miRNA's deep regulatory mechanisms in cell biology.⁶⁷ Recent advances in spatial transcriptomics offer a novel way to study the distribution of miRNAs inside and outside cells, providing an intuitive understanding of their spatial dynamics.⁶⁸ At the proteomic and metabolomic levels, miRNAs indirectly regulate protein expression by targeting specific genes, which in turn affects metabolite transfer and modulates metabolic pathways. Notably, miRNAs are also packaged into extracellular vesicles (EV-miRNA), playing a vital regulatory role.⁶⁹

miRNA network analysis identifies pivotal regulatory hubs in cancer, highlighting potential biomarker Hub miRNAs. This insight not only enhances cancer diagnosis and treatment strategies but also sheds light on personalized medicine approaches.⁷⁰ Integrating functional genomics data, such as miRNA and mRNA expression profiles, reconstructs regulatory networks for survival analysis,⁶⁴ elucidating the critical role of miRNA–mRNA interactions in cancer prognosis, identifying miRNAs involved in the disease process.⁷¹

Furthermore, analyzing miRNA cooperation networks across different cancers elucidates shared and distinct mechanisms, offering comprehensive insights for cancer treatment and prevention.⁷² Computational simulations like miRNA-disease association prediction predict miRNA-disease associations, facilitating the discovery of novel therapeutic pathways and a deeper understanding of miRNA's role in disease pathogenesis.⁶⁵ These studies not only advance our understanding of miRNA's regulatory roles in cells but also innovate new strategies for disease treatment and prevention.

3.1 miRNA interactions in cancer

miRNA interacts with various molecules to exert its biological functions, categorized into upstream and downstream interactions. Downstream targets include mRNA, RBP, miRNA, and DNA, while upstream interactions involve competing endogenous RNAs (ceRNAs). Furthermore, miRNA can be secreted extracellularly, facilitating intercellular communication (Figure 2A).

miRNAs predominantly bind to specific sequences within the 3′ UTR of target mRNAs, leading to translational inhibition, mRNA deadenylation, or decapping.³ Alterations in the 3′ UTR length can affect RNA structure, impacting miRNA binding site accessibility.⁷⁸ These binding sequences, known as MREs, determine the miRNA's effect—AGO-dependent degradation or translational inhibition of mRNA.³ A single miRNA can target hundreds of mRNAs, thereby influencing protein networks, while a single mRNA may be regulated by multiple miRNAs.⁷⁶ miRNA binding sites are also present in the 5′ UTR, coding sequences (CDS), and promoter regions. miRNAs binding to the 5′ UTR and coding regions suppress gene expression, while promoter region interactions can activate transcription, mediating translational activation.⁷⁹ Moreover, RNA editing can introduce changes in miRNA seed regions, altering target mRNA binding.⁷⁸ Depending on the mRNA's location, miRNA-induced silencing complex (miRISC) can bind to ribosome- or ER-bound mRNA to inhibit translation.³ In tumorigenesis, synergistic miRNA effects on target mRNA, particularly involving multiple MREs, present viable therapeutic strategies.⁸⁰ Additionally, feedback loops are formed when downstream mRNAs act as TFs, which can also regulate miRNA expression.²¹

The biogenesis of miRNA involves a series of processing steps facilitated by proteins such as Drosha, DGCR8, Exportin5, and Dicer, ultimately forming the miRISC.⁸¹ RBPs, which regulate gene expression posttranscriptionally, are critical in miRNA-mediated processes such as splicing, localization, and stability. Disruptions in RBP expression often impair miRNA processing, influencing target mRNAs related to cancer progression.⁸² Depending on the miRISC's location, it can promote mRNP degradation in the nucleus or maintain the mRISC:mRNP complex during cytoplasmic transport. miRNAs can also form complexes with proteins like FXR1, modulating adenosine/guanosine-rich elements (AREs) and activating translation.³ Additionally, miRNA expression levels are regulated by RBPs such as AUF1, DDX17, and ILF3.⁷⁸ Recent research has highlighted the key role of interactions between miRNAs, RBPs, and exoribonucleases in tumorigenesis.⁸³

miRNAs can interact with each other at different stages of biogenesis—directly at the pri- or mature-miRNA stages or indirectly by regulating TFs or miRNA regulators.⁸⁴ Direct interactions involve complementary pairing between miRNAs, influencing their activity and expression, while indirect interactions involve targeting TFs, which then regulate downstream miRNAs. Global interactions allow a single miRNA to regulate entire families of miRNAs, providing a flexible mechanism for gene regulation.⁷⁷

Nuclear-activated miRNAs (NamiRNAs), a newly discovered class, function as enhancers by binding to promoter and enhancer regions, promoting gene expression.⁸⁵ NamiRNAs are enriched in active enhancers and alter chromatin states, activating homologous gene transcription. In cancer, NamiRNA networks have been shown to activate tumor suppressor genes, revealing novel mechanisms in cancer development and potential therapeutic strategies.⁸⁶ Additionally, some miRNAs target DNA methyltransferases (DNMTs), influencing gene expression at both transcriptional and posttranscriptional levels. DNA methylation can regulate miRNA synthesis, but a reciprocal relationship exists when miRNAs modulate gene expression via methylation mechanisms. For instance, miR-211 silencing via DNA methylation reduces melanoma cell sensitivity to cisplatin, suggesting a potential role for epigenetic regulation in enhancing drug efficacy.⁸⁷

miRNAs also interact with other ncRNAs, including lncRNAs, circRNAs, and pseudogenes, regulating various biological processes. ceRNAs modulate mRNA expression by binding to MRE sequences, counteracting miRNA-mediated inhibition of mRNA translation.⁸⁸ miRNA–ceRNA interactions often form feedback loops, either amplifying or suppressing gene expression.^89-91 ceRNAs can also recruit TFs to influence miRNA gene expression.⁵⁷

miRNAs are secreted into extracellular fluid and transported to target cells via vesicles (such as exosomes) or bound to proteins like AGO.³ These extracellular miRNAs facilitate intercellular communication, remaining stable even under extreme conditions.⁷⁸ They may bind to proteins such as high-density lipoprotein (HDL) and NPM1, further enhancing stability. miRNA transfer between cells can occur through exosomes, microvesicles, apoptotic bodies, or direct cell–cell contact.⁴ Extracellular miRNAs act as autocrine, paracrine, or endocrine regulators, modulating cellular activities.⁹² Notably, treating cells with extracellular miRNAs has consistently led to target mRNA repression, supporting the hypothesis that they play a significant role in intercellular communication.⁹³

3.2 miRNA in tumorigenesis

The complex interaction of miRNA target molecules underscores the intricate role miRNAs play in regulating tumor suppressor genes and oncogenes (Figure 2B). Abnormal miRNA expression can result in dysregulation of these genes, leading to tumorigenesis.⁹⁴ miRNA dysregulation impacts key aspects of tumor development, including cell cycle progression, proliferation, metastasis, invasion, EMT, apoptosis, DNA repair, angiogenesis, metabolic reprogramming, drug resistance, and immune modulation.^{95, 96} In cancer cells, miRNA genome rearrangements (such as deletions or amplifications), mutations in UTRs or miRNA sequences, epigenetic methylation, transcription factor dysregulation, altered miRNA biogenesis, and ceRNA activity can all lead to aberrant miRNA expression and function, creating a vicious cycle that exacerbates cancer progression.^{21, 97} Several common oncomiRs, including miR-21, let-7, miR-17-92, miR-155, and others. miR-21, notably overexpressed in many cancers, acts as a proto-oncogene and its elevated levels often correlate with poor prognosis.⁹⁸ Conversely, miRNAs like Let-7, miR-34a/b/c, and miR-15a/miR-16-1 function as tumor suppressors. Let-7, the first discovered miRNA, suppresses tumors by targeting oncogenes such as Ras, Myc, HMGA2 (in lung cancer), and MYCN (in neuroblastoma).⁹⁹

miRNAs influence various biological functions in cancer cells by targeting key mRNAs. For instance, critical mRNAs regulating cell cycle progression and proliferation include cyclin D1 (CCND1), E2F, and cyclin-dependent kinases (CDKs).¹⁰⁰ At the EMT level, miRNAs can directly target EMT markers such as E-cadherin and Vimentin, or regulate EMT-associated TFs like Snail, ZEB, and Twist, influencing downstream signaling pathways such as Wnt, PTEN, TGF-β, and Ras, thereby promoting cancer progression.¹⁰¹ In apoptosis regulation, miRNAs target mRNAs such as APAF-1, BCL-2, and caspases.¹⁰² For DNA repair, key targets include ataxia telangiectasia mutated (ATM), essential for repairing DNA double-strand breaks.¹⁰³ miRNA regulation of angiogenesis is primarily mediated through the PI3K/Akt pathway, affecting vascular endothelial growth factor (VEGF) expression.¹⁰⁴ Metabolic reprogramming is influenced by miRNAs targeting oncogenes like c-Myc, HIF-1, and p53, which control processes such as the Warburg effect.¹⁰⁵ Specifically, c-Myc regulates glucose transporters and pyruvate dehydrogenase, influencing tumor cell metabolism.¹⁰⁶ miRNAs also modulate drug resistance by targeting multidrug resistance protein 1 (MDR1).¹⁰⁷ At the immune modulation level, miRNAs regulate pathways like mTOR and NF-kB, which are involved in inflammatory and immune responses.¹⁰⁸

Moreover, miRNAs packaged in exosomes facilitate intercellular communication, influencing carcinogenesis in recipient cells. Their regulation is intricately linked to epigenetic modifications, transcriptional and posttranscriptional processes. Target genes of miRNAs involved in these regulatory mechanisms can form feedback loops, achieving precise spatiotemporal control of gene expression crucial for cancer initiation and progression.¹⁰⁹

3.3 Epigenetic regulation of miRNA in cancer

Epigenetics encompasses inheritable changes in gene expression during cell division without altering DNA sequence.¹¹⁰ miRNAs play a crucial role in this process by directly modulating gene expression or targeting epigenetic enzymes to influence DNA methylation, histone modification, genomic imprinting, and chromatin remodeling (Figure 2C).¹⁰⁹

Conversely, epigenetic modifications profoundly impact miRNA expression,²¹ creating a complex miRNA-epigenetic feedback loop crucial for gene regulation. Dysregulation of this loop is closely linked to cancer development.¹⁰⁹ A specific subset known as “epi-miRNAs” regulates epigenetic elements, containing DNMTs, histone-lysine N-methyltransferase (EZH2), histone deacetylases (HDACs), which forms several regulatory circuits.¹⁰⁸ For example, miRNA biogenesis and function are regulated by DNA and mRNA methylation, while miRNAs can, in turn, target DNMTs, influencing genome-wide methylation patterns.⁶⁶ Dysregulated epi-miRNAs contribute significantly to cancer initiation and progression.¹¹¹ Methylation of miRNAs represents a key epigenetic modification, adding complexity to the posttranscriptional regulation of gene expression.¹¹²

miRNAs influence various epigenetic modifications. For instance, the miR-29 family targets DNMT3A and 3B to modulate DNA methylation, influencing lung tumor development.^{113, 114} miR-29b targets histone deacetylase 4 (HDAC4) to enhance histone and tubulin acetylation, inhibiting multiple myeloma progression.¹¹⁵ In genomic imprinting, miRNAs in imprinted regions influence DNA methylation processes, impacting gene expression and function. Notably, imprinted miRNAs from C2MC and C19MC clusters regulate PTEN, crucial for growth and metabolism.¹¹⁶ Additionally, miR-223 indirectly regulates chromatin structure by targeting chromatin remodeling complexes like SWI/SNF.^{110, 117}

Overall, miRNAs orchestrate a broad spectrum of molecular activities through epigenetic control, forming intricate regulatory networks with significant potential in cancer therapy. Further exploration of miRNA–epigenetic interactions is essential for advancing clinical applications, potentially reversing abnormal gene expressions and enhancing immune responses.¹⁰⁹

3.4 miRNA and Tumor Microenvironment (TME)

The TME is crucial for tumor growth and metastasis, being closely linked to processes like immune evasion, extracellular matrix (ECM) remodeling, hypoxia, abnormal angiogenesis, and metabolic reprogramming, ultimately reshaping the TME itself.¹¹⁸ In cancer, miRNA levels undergo significant changes influenced by complex interactions within the TME, involving a dynamic network of factors.¹¹⁹ As depicted in Figure 2D, miRNAs not only target critical regulatory molecules in cancer cells but also contribute to intricate signaling networks between cancer cells and their microenvironment.¹¹⁹ Various cytokines within the TME can alter miRNA expression, activating pathways that help cancer cells evade immune surveillance while promoting proliferation, metastasis, and invasion.¹²⁰ Thus, miRNAs play a significant role in modulating cell-cell interactions within the TME, influencing these critical processes.

Exosomal miRNAs are pivotal in tumor progression, modulating immune responses, remodeling the TME, promoting angiogenesis, and facilitating tumor cell metastasis through intercellular communication.^{121, 122} Moreover, miRNAs are secreted via microvesicles or exosomes, further influencing cancer cell growth and metastasis.¹¹⁹ Notably, miRNAs regulate immune cell function, thereby shaping the tumor-immune microenvironment, which profoundly impacts tumor progression.¹²³

The TME transforms miRNA into a bridge connecting diverse cell types, allowing tumor cells to evade immune surveillance, facilitating infiltration, and promoting metastasis. miRNAs primarily impact tumor development by modulating various immune cells such as macrophages, myeloid-derived suppressor cells (MDSCs), dendritic cells (DCs), and natural killer cells (NK cells). Recent studies have highlighted miRNA's role in regulating Treg cells and granulocytes, further emphasizing its broad impact on cancer.¹²³ Epstein-Barr virus (EBV) infections can release miRNAs targeting FOXP1 and PBRM1, which bind to enhancer regions of programmed cell death 1 ligand 1 (PD-L1), thereby inhibiting PD-L1 expression and promoting immune escape.¹²⁴ Beyond immune cells, cancer stem cells (CSCs) are also pivotal in the TME. In BLCA (UBC), cancer-associated fibroblast (CAF)-derived miR-146a-5p mediates tumor-stroma crosstalk, promoting CSC proliferation and chemotherapy resistance.¹²⁵ miRNAs also regulate hematopoietic stem cells (HSCs) in the blood. For instance, EV miR-181a-5p from CRC activates HSCs by modulating the IL6/STAT3 pathway. Activated HSCs secrete chemokine CCL20, which further amplifies miR-181a-5p activity, creating a positive feedback loop.¹²⁶

Besides modulating immune cells, miRNAs play a role in other specific TME processes. From an ECM perspective, tumor cells can secrete exosomal miRNAs that transform normal fibroblasts (NFs) into CAFs. These CAFs secrete growth factors and cytokines that activate adjacent ECM components.^127-129 Collagen synthesis and degradation are vital for ECM remodeling. Exosomal miRNA-29a inhibits the TGF-β2/Smad3 pathway, reducing type I and type III collagen expression.¹²⁹ Hypoxia, a hallmark of the TME, promotes the Warburg effect, which enhances glycolysis and metabolite production, leading to an acidic environment that fosters tumor invasion and metastasis. miRNAs can inhibit hypoxia-inducible factor 1α (HIF-1α) translation, inducing its degradation.¹³⁰ Additionally, EV-miR-21-5p secreted by hypoxic papillary thyroid cancer cells enhances angiogenesis in the TME by targeting TGFBI and COL4A1.¹³¹ miRNAs also target proteins like TSGA10, KLF2, and PTEN, activating molecules such as vascular endothelial growth factor receptor 2 (VEGFR2), p-AKT, and p-ERK, to promote angiogenesis by regulating the quantity and function of local blood vessels.¹²²

Metabolic reprogramming is another key cancer hallmark regulated by miRNAs. miRNA-mediated regulation of COL1A1 activates MYC signaling, inducing metabolic reprogramming in CAFs, which adapt to different metabolic environments and support tumor growth.¹³² Mitochondrial reprogramming is often associated with metabolic adaptation, crucial for energy production and biosynthesis.¹³³ miRNAs can also affect adipocytes and CAFs. For instance, Kaposi's sarcoma-associated herpes virus-derived EVs transfer miRNAs to uninfected cells, reducing mitochondrial biogenesis and respiration while inducing aerobic glycolysis.¹³³

Overall, EV miRNAs act as communication molecules within the TME, regulating various critical processes through cell–cell interactions. miRNA often compromises effective antitumor immune responses, potentially promoting tumor development. Exosome-mediated delivery of miRNAs presents a promising strategy to counteract tumor-immune evasion and metastasis.¹³⁴ miRNA serves as a crucial molecular mediator of communication between tumor cells and immune cells, underscoring its significance in advancing tumor immunotherapy through deeper investigation.¹³⁵

4.1 Characteristics and techniques for miRNA detection

miRNAs are a class of small ncRNAs that play crucial regulatory roles in organisms, exhibiting several notable characteristics that hold significant implications for biological research and medical applications. First, miRNAs demonstrate a high degree of conservation, suggesting that they may perform similar regulatory functions across diverse biological developmental processes.¹³⁶ Second, miRNA expression shows tissue-specific patterns, with varying levels of expression in specific cell types or tissues.¹³⁷ Additionally, miRNA expression exhibits tissue specificity, with significant variations in expression levels across different developmental stages and tissues, indicating dynamic regulation.¹³⁸ These characteristics collectively reveal the complexity and precision of miRNAs in gene regulation, with profound implications for understanding their biological functions and their potential applications in disease diagnosis and treatment.

miRNAs exist in various forms, including pre-miRNA, pri-miRNA, mature miRNA, and intermediate products derived during miRNA processing within tumor cells. Liquid biopsy is a noninvasive diagnostic technique that provides disease information by analyzing biomarkers (cfDNA, mRNA, circulating tumor cells, EVs, metabolites, miRNA) present in body fluids, such as blood, urine, and saliva.^{139, 140} Currently, circulating free miRNAs (cf-miRNAs) have been detected in the blood through liquid biopsy. Once these miRNAs leave the intracellular environment, they exist in association with carriers, such as Ago proteins, nucleophosmin 1 (NPM1), and HDL. However, the most common form is their presence within EVs (exosomes, microvesicles, apoptotic bodies).¹⁴¹ cf-miRNAs are stable in body fluids and show specific abnormal expression in various cancers, making them promising markers for cancer diagnosis with higher sensitivity than traditional protein markers¹⁴² (Figure 3A and Figure 3B). Compared to total circulating RNAs, EV-miRNA often targets specific cells, so detection of EV-miRNA is more likely to help physicians predict cancer prognosis.^{129, 143} Specifically, cell-free immune-related miRNAs (cf-IRmiRNAs) are recognized for their stability and abundance, making them suitable for liquid biopsies to detect cancer early.^{144, 145} These findings underscore miRNAs' value in early cancer diagnosis, potentially improving early detection rates and patient survival.

With the advancement of scientific research and technology, miRNA-based cancer detection methods are poised to make significant strides in clinical practice. Various miRNA quantification technologies are emerging, each with distinct advantages. These include quantitative reverse transcription polymerase chain reaction (qRT-PCR), in situ hybridization,¹⁴⁶ Northern blot,¹⁴⁷ microarray chips,¹⁴⁸ NGS, and nanotechnology¹⁴⁹ (Figure 3C).

qRT-PCR, which combines reverse transcription and real-time quantitative PCR, is a precise technique for quantifying miRNA and analyzing gene expression. It is known for its high specificity and sensitivity and can be classified based on primer types into Linear, T4 RNA ligase, stem-loop, and poly(A)-tailing-based methods.¹⁵⁰ Detection methods for PCR products include synergy brands and TaqMan methods. Digital PCR, an emerging absolute quantitative technique, divides single analyte miRNA molecules and PCR reagents into individual water–oil emulsion droplets, offering high sensitivity for low-abundance miRNA samples.¹⁵¹

In situ hybridization, a hybridization-based detection method, displays the spatial distribution and transport of miRNAs within cells, providing a visual representation of miRNA expression. However, due to the short sequence of miRNAs, creating specific probes is challenging.¹⁵² To overcome this, specially chemically modified LNA probes, which offer higher thermodynamic stability and reduced signal loss, are often used. Some studies utilize LNA probes with 20-O-methyl RNAs to enhance specificity and sensitivity. DIG-labeled LNA probes can be detected quantitatively using alkaline phosphatase (AP)-conjugated antibodies, while fluorescence in situ hybridization employs horseradish peroxidase-conjugated secondary antibodies to produce fluorescence.¹⁴⁹ Northern blot, a molecular biology technique, detects specific RNA sequences by transferring RNA from a gel to a membrane and hybridizing with a labeled probe. Although it confirms new miRNAs, it is time-consuming and does not quantify miRNA.¹⁴⁷

Northern blot is based on the hybridization between target miRNAs and complementary probes. It involves the separation of miRNA and pre-miRNA through electrophoresis, followed by transferring the miRNA onto a solid membrane. The membrane is then hybridized overnight with complementary oligonucleotide probes. The greatest advantage of Northern blot is its high specificity, allowing for the determination of the target miRNA's sequence and length, and it is relatively straightforward to perform in the laboratory. However, miRNAs are prone to degradation by external factors such as RNase during the experiment, and Northern blot requires a large amount of miRNA sample for detection.¹⁴⁷

Microarray chips, which detect miRNAs at high throughput, are not suitable for discovering new miRNAs or providing absolute quantification. Hydrogel microfluidic chips can capture Ago2-miRNA protein complexes from serum, enabling PCR and quantitative miRNA detection. Microfluidic surfaces can also integrate with surface-enhanced Raman spectroscopy (SERS) to enhance detection sensitivity.¹⁵³

Next-generation sequencing (NGS) offers high throughput and sensitivity for detecting and identifying new miRNAs but involves complex library preparation and data processing, with challenges related to cost and absolute quantification.¹⁵⁴ NGS can also analyze miRNA expression levels and genetic mutations, which aids in selecting specific miRNA biomarkers.¹⁵⁵

Nanotechnology plays a crucial role in quantitative miRNA detection, with methods including nanoparticle solution methods, nanostructure surface-based methods, and methods combining nanoparticle materials with solid carriers. Optical signal-based strategies using nanoparticles such as gold nanoparticles (AuNPs), quantum dots (QDs), and silver nanoclusters are common. Magnetic nanoparticles can also be used in nuclear magnetic resonance measurements.¹⁵⁶

In summary, RT-qPCR remains the most widely used method for exosomal miRNA detection due to its sensitivity and specificity. Microfluidics and deep sequencing have improved sample analysis efficiency with high throughput. Microarray chips are useful for comparing miRNA expression levels, while NGS offers advantages in discovering new miRNAs and characterizing miRNAomes. Despite its benefits, each detection technology has limitations, and developing more economical, rapid, and high-throughput methods, especially for circulating miRNAs, remains a key research direction.¹⁵⁰

4.2 Improvements and innovations in miRNA detection technology

Specificity and sensitivity are the two most critical factors that require enhancement in miRNA detection technologies. Currently, most advancements focus on optimizing these two performance metrics. Established methods for miRNA quantification, such as microarray chips, qRT-PCR, and NGS, rely heavily on various probe designs and labeling technologies to improve detection sensitivity and specificity.¹⁵⁰ To further enhance these aspects, miRNA detection is increasingly adopting complex sensor systems, with CRISPR, enzymes, and amplification technologies playing crucial roles in constructing miRNA biosensors. Portable miRNA detection sensors also offer promising solutions for point-of-care testing (POCT).

For improving specificity, clustered regularly interspaced short palindromic repeats (CRISPR) technology—a gene-editing system originally found in bacteria—has shown promise. It uses RNA molecules to precisely recognize and cut specific DNA sequences, thus targeting and modifying particular locations in the genome. The CRISPR/Cas12a system has recently been reported for direct miRNA detection,¹⁵⁷ where the CRISPR sequence provides targeting information and the enzyme executes the cutting action, significantly enhancing specificity.¹⁵⁸

Sensitivity improvements largely focus on nucleic acid and signal amplification techniques, with nanoparticles being commonly employed for signal amplification. Current amplification strategies include rolling circle amplification (RCA), exponential amplification reaction (EXPAR), catalytic hairpin assembly (CHA), hybridization chain reaction (HCR), duplex-specific nuclease signal amplification (DSNSA), and loop-mediated isothermal amplification (LAMP).¹⁵⁹ Unlike traditional qRT-PCR, these methods maintain a constant temperature during analysis and quantification,¹⁶⁰ which can be further enhanced by combining different amplification techniques. For instance, EXTRA-CRISPR can be combined with RCA,¹⁶¹ while CHA can be integrated with DSNSA.¹⁶²

Nanotechnology-based miRNA detection techniques are relatively sophisticated, involving nanoparticles, nanostructured surfaces, and solid carriers. Nanoparticles are particularly useful for amplifying signal carrier material functions. For instance, nano-morphological DNA cage-based thermophoretic assays can selectively analyze miRNA in exosomes in situ, offering high sensitivity, selectivity, and in situ detection.¹⁶³ Additionally, the combination of SERS tags and magnetic separation with nanotechnology enhances sensitivity and response speed, making it particularly effective for detecting trace amounts of miRNA.¹⁶⁴ Surface polymerization technology, when combined with surface plasmon resonance imaging (SPRI), leverages changes in the refractive index between light and electrons on the nanoparticle surface to reflect miRNA-probe binding.¹⁶⁵ Moreover, electrochemical nanobiosensor technology is gaining traction, with photoelectrochemical (PEC) biosensors based on MoS2@Ti3C2 nanohybrids enabling ultra-sensitive miRNA detection.¹⁶⁶

Hydrogels, particularly DNA hydrogels, are also emerging as vital tools for miRNA detection. DNA hydrogels, formed through chemical cross-linking or physical entanglement of DNA chains into a three-dimensional network structure, respond to specific biochemical or physical signals by undergoing structural changes. These changes allow for the specific release of miRNA, which significantly enhances detection specificity and sensitivity.¹⁶⁷ DNA hydrogels can be engineered into various forms, such as microcapsules, particles, membranes, or microneedle patches, and are integral to the development of portable biosensors tailored to different detection needs and application scenarios.¹⁶⁸ The responsiveness of DNA hydrogels can also be integrated with CRISPR technology.¹⁶⁹

Additionally, biosensing, microfluidics, and machine learning represent a novel strategy for miRNA detection.¹⁷⁰ Ramshani et al. developed an integrated system combining a lysing chip with a concentration/sensing chip. This system eliminates the need for pretreatment or analyte transfer, significantly reducing measurement time.¹⁷¹ Such microchip systems, constructed using high-performance materials to create nanobiosensing platforms, amplify multiple signals for rapid, sensitive, and specific exosomal miRNA detection, showing immense potential for future applications.

4.3 Application of miRNA as biomarker

miRNAs have emerged as valuable biomarkers for cancer diagnosis. Specific miRNAs are integral to cancer progression, leading to the development and application of miRNA-based diagnostic tools for cancer screening and early detection.^{172, 173} Furthermore, miRNAs can assist in clinical staging of cancer.¹⁵⁴ Exosomal miRNAs, which facilitate intercellular communication, play a crucial role in cancer development.¹⁷⁴ Since blood is often used as a sample for exosomal miRNA detection, this method offers the advantage of being minimally invasive.¹⁷⁵ With advancements in portable POCT devices, cancer diagnosis can now be expedited through miRNA detection (Figure 3D).^{176, 177}

The role of miRNAs in cancer prognosis assessment is increasingly prominent. Abnormal miRNA expression is a potential biomarker for overall survival and disease-free survival in cancer post-neoadjuvant chemotherapy.¹⁷⁸ Changes in miRNA levels can predict cancer recurrence risk and provide crucial insights into cancer prognosis. By constructing mRNA–miRNA–lncRNA networks and performing functional enrichment analysis, hub gene identification, and prognostic correlation analysis, we can accurately identify miRNAs linked to cancer prognosis.¹⁷⁹ Additionally, analysis of the TCGA database and the Cancer Gene Atlas database, combined with risk survival curves, ROC curves, and survival status diagrams, has successfully identified miRNAs closely related to cancer prognosis, establishing corresponding prognostic risk models ^{180, 181} certain miRNA expression levels correlate with clinical stage, lymph node metastasis, and patient survival.¹⁸² A novel computational approach, ProModule, aims to systematically identify miRNA-mediated modules with prognostic value using clustering methods.¹⁸³ Additionally, deep learning shows promise in integrating multi-omics data to predict hepatocellular carcinoma (HCC) survival rates, offering new insights for clinical cancer treatment.⁶⁴

miRNAs can also serve as biomarkers for predicting and dynamically monitoring the therapeutic efficacy of cancer treatments. In predicting efficacy, AI-driven pan-cancer analysis has highlighted miRNA's unique role in cancer staging prediction.¹⁸⁴ In breast cancer, miRNAs like miR-30c, miR-187, and miR-339-5p predict treatment outcomes.¹⁸⁵ Specifically, miR-21's expression level correlates with breast cancer invasiveness and chemotherapy resistance, providing a reference for predicting chemotherapy response and guiding personalized treatment.¹⁸⁶ In immunotherapy, miRNAs influence PD-L1 expression, a key factor in tumor-immune escape. Hsa-miR-200 is associated with high PD-L1 expression, while hsa-miR-197 is negatively correlated with PD-L1 expression.¹⁸⁷ In the dynamic monitoring of treatment effect, miRNAs are crucial in optimizing treatment strategies by regulating drug resistance and affecting metabolic pathways and gene expression related to cancer metabolism. For example, distinct mutational signatures of 13 miRNAs have been associated with disease progression in chronic lymphocytic leukemia.⁹⁷ In addition, tools like the miRNA signature classifier (MSC) and MIR-Test have demonstrated high specificity and sensitivity in reducing the false-positive rate of low-dose computed tomography (LDCT).¹⁸⁸ MSC has shown promising results in detecting disease recurrence in postoperative plasma samples and is currently undergoing prospective validation.¹⁸⁹

Research indicates that genetic and epigenetic modifications of miRNAs can serve as both diagnostic and prognostic markers in cancer. Single nucleotide polymorphisms (SNPs) in miRNAs are linked to cancer susceptibility, while ADAR proteins, which mediate A-to-I editing, influence the processing, expression, and activity of miRNAs in cancer, making them important diagnostic and prognostic markers.^{190, 191} Additionally, N6-methyladenosine (m6A)-modified miRNAs represent another key epigenetic modification associated with cancer diagnosis and prognosis.¹⁹² With new technologies such as MeRIP-Seq (m6A-seq), it is now possible to identify m6A-modified miRNAs. As detection technologies continue to advance, newly discovered miRNAs and their modifications are being evaluated as potential diagnostic and prognostic markers in clinical trials.¹⁹³ Increasingly, studies are identifying or testing miRNA biomarkers in cancer efficacy monitoring, with many clinical trials completed or underway (Table 1).

5.1 Development and use of miRNA mimics and inhibitors in cancer treatment

Dysregulated miRNA expression, acting as tumor suppressors or oncogenes (oncomiRs), plays a crucial role in many cancers. Targeted therapies against miRNAs, including miRNA mimics and inhibitors, have shown great potential in preclinical studies, as shown in Figure 4A.^{96, 203, 204} These therapeutic strategies involve using anti-miRNA molecules to inhibit upregulated miRNAs causing pathological conditions and employing miRNA mimics to replace and restore the activity of downregulated miRNAs.²⁰⁵

miRNA mimics are synthetic RNA duplexes that replicate endogenous miRNA sequences. They integrate into the RISC and restore miRNA expression in disease states.²⁰⁶ This approach has demonstrated therapeutic value in clinical trials, paving new avenues for cancer treatment.²⁰⁷ Additionally, using small miRNA inhibitors/mimics as adjuvants in vaccination strategies is an innovative method to enhance antibody responses.²⁰⁸ Although initial progress has been made in clinical trials, challenges such as improving stability, achieving specific delivery, and ensuring safety remain.²⁰⁹ To counter endonuclease degradation in the blood, strategies like chemical modification or conjugation enhance miRNA mimics' stability.²⁰³ Rigorous preclinical studies and long-term follow-up experiments are essential to evaluate safety and manage potential immune and toxic reactions.²¹⁰

miRNA inhibitors can alleviate the overexpression of target mRNA and restore cell phenotypes, making them a promising therapeutic strategy. These include natural miRNA inhibitors (such as lncRNA, circRNA, and other ceRNA) and artificial miRNA inhibitors (such as Anti-miRNA oligonucleotides (AMOs) and microRNA sponges).²¹¹ MicroRNA sponges, once transfected and stably expressed in cells, can produce long-term or permanent inhibition of the target miRNA.²¹² Other small molecule inhibitors can directly target oncomiRs and Dicer (an enzyme important in miRNA generation) to achieve tumor suppressor effects. Additionally, CRISPR can be used to directly edit oncomiRs for tumor suppression.¹⁷

However, the therapeutic potential of natural miRNAs is often limited due to degradation by RNAse or activation of toll-like receptors, which stimulate the innate immune system.¹⁵⁴ To improve miRNA stability, chemically modified nucleic acids, such as locked nucleic acids (LNAs), are employed. LNAs contain a methylene bridge between the 2’-O and 4’-C of the sugar, creating a stable, “locked” ring conformation with enhanced stability. For the issue of miRNA reduction after transfection, viral vectors can provide targeted delivery at specific times, and caged miRNAs, connected by photocleavable linkers, can be specifically released via light.^{213, 214} Several miRNA-based antitumor drugs have entered clinical trials, as shown in Table 2, offering promising new directions for cancer treatment.^{220, 221}

5.2 Nanoparticle-based miRNA therapy delivery system

In addition to improving drug stability and specificity by changing the state of the nucleic acid itself, miRNA mimics and inhibitors can also be delivered into the body through various methods, including vector delivery²²² and microinjection.²²³ Delivery vectors include viral vectors (adeno-associated virus and Lentivirus) and nonviral vectors (EVs, LIPID NANOPARTICLE, POLYMERIC DELIVERY SYSTEM, INORGANIC COMPOUND-Based nanoparticle)²²⁴ (Figure 4B).

Nanoparticles, as nonviral vectors, show significant advantages in delivering miRNA to specific locations.^{210, 225} Compared to naked miRNA delivery, nanoparticles reduce adverse off-target effects and immune responses while improving cellular uptake rates and half-life.²²⁶ To optimize nanoparticles, researchers focus on material selection, design rationality, and particle modification to achieve high stability, targeting specificity, and controlled release.²²⁷

Recent developments include lipid nanoparticles, SEND RNA delivery platforms, and Fusogenix protein-lipid carriers. These particles improve stability, facilitate storage and transportation, and enhance cellular uptake efficiency.²²⁷ LNPs are a common nanoparticle carrier. In addition to their raw materials and surface modifications, barcoded nanoparticles, introduced by Dr. Dahlman, significantly enhance high-throughput screening of LNPs, accelerating the development of targeted drug delivery systems.²²⁸ A poly-l-lysine complex, a positively charged polymerized amino acid, can also be used as a specific carrier for miRNA delivery. It forms a complex with negatively charged miRNA molecules, enhancing the stability and cellular uptake of miRNAs.²²⁹ miRNA-peptide nanoparticles can be combined with materials like hydrogels to improve stability and facilitate local drug delivery.²³⁰ Additionally, dual-responsive polypeptide nanovectors can be engineered to achieve pH-specific miRNA release in the TME.²³¹ Poly(l-lysine)-modified polyethyleneimine (PEI-PLL) copolymers, which combine the high transfection efficiency of PEI with the biodegradability of poly-l-lysine, have been shown to effectively deliver miR-21 with reduced toxicity.²³²

Exosomes, natural EVs secreted by cells, also show promise due to their superior biocompatibility and targeting capabilities.²³³ Chemical modification and optimized drug delivery systems can offer advantages like rapid synthesis, controlled release, and targeted administration through.²²⁴ For instance, MRX34, a synthetic double-stranded miR-34a mimic encapsulated in liposome nanoparticles, is undergoing phase 1 anticancer clinical trials and has shown apoptosis-inducing effects on tumor cells.²⁰⁵ Mesomir, a miR-16 mimic targeting malignant pleural mesothelioma or non-small cell lung cancer, has also entered phase I clinical trials.^234-236

However, challenges remain in applying miRNA in cancer delivery therapy. Researchers use anti-fouling molecules to improve in vivo stability, reducing protein corona formation and aggregation and enhancing colloidal stability.²³⁷ Controlled and sustained miRNA release is achieved by combining controlled triggering and stimulus-responsive materials, extending the half-life.²³⁸ Additionally, proton-accepting materials can prevent miRNA complexation and degradation after cytoplasmic release.²³⁹ Functionalizing nanoparticles with specific cell-targeting ligands improves targeting specificity.²⁴⁰ Collaboration across molecular biology, bioengineering, and clinical oncology is crucial to improve miRNA delivery's stability and specificity. Most studies are still in the preclinical stage, necessitating further toxicity research on nanoparticles.²²⁶

5.3 Combination therapy of miRNA and traditional therapy

As shown in Figure 4C, miRNA therapy has unique advantages in cancer treatment and can synergize with various traditional therapies to significantly enhance therapeutic effects. Specifically, miRNA therapy can increase sensitivity to traditional therapies, overcome drug resistance, and inhibit tumor angiogenesis, improving therapeutic outcomes.²⁴¹

In chemotherapy, miRNA can be combined with drugs to target specific molecules or signaling pathways accurately, reducing damage to normal cells and enhancing tumor cell killing.²⁴¹ miRNA therapy can also be combined with radiosensitization. Before and after surgery or radiotherapy, miRNA therapy can reduce tumor recurrence risk or increase radiotherapy sensitivity.²⁴¹ For instance, miR-34a (MIRX34) has shown a radiosensitization effect in NSCLC by binding to the DNA damage response gene RAD51.²⁴²

In immunotherapy, miRNAs play diverse roles. They can directly target cancer cells for immunotherapy, regulate immune checkpoint molecules, enhance the response to immunotherapy, and fine-tune immune responses by affecting the immunosuppressive microenvironment and antigen presentation function of DCs.^{208, 209, 243} In cytokine therapy, miRNAs participate in interferon treatment responses as adjuvants, verified in clinical trials.²⁴⁴ miRNA can enhance T cell receptor sensitivity, T cell fitness, and effector function.²⁴⁵ Additionally, combining miRNA therapy with adoptive cell immunotherapy is a research hotspot. Overexpression of miR-143 supports memory T cell differentiation and tumor metabolic reprogramming regulation, enhancing HER2-CAR T cells' specific killing effect.²⁴⁶ A novel miRNA cluster and epidermal growth factor receptor (EGFR) variant III lentiviral vector CAR-T cells significantly improved persistence and efficacy in treating glioblastoma.²⁴⁷ miR-1258 targets programmed death ligand 1 (PDL1) and affects myeloma treatment by inhibiting PDL1.²⁴⁸ Despite promising clinical trials linking T cell persistence and tumor treatment, miRNA adoptive immunotherapy is still mainly in preclinical trials.²⁴⁵ Future breakthroughs in clinical trials are expected to provide more effective cancer treatment strategies.

6.1 Mechanisms by which miRNA promotes drug resistance in cancer cells

Chemotherapy, as a systemic treatment, is primarily targeted at tumors prone to spreading and metastatic advanced tumors, showcasing its indispensable therapeutic value.²⁴⁹ However, the development of drug resistance has become the primary factor in the failure of chemotherapy for advanced cancers.²⁵⁰ Acquired drug resistance refers to resistance emerging in previously drug-sensitive tumor populations, leading to tumor recurrence. This resistance often results from genetic, epigenetic, transcriptomic, and proteomic alterations in cancer cells, driving tumor heterogeneity. Many tumors exhibit resistance through a combination of intrinsic and acquired factors.²⁵¹ Moreover, drug-sensitive tumors containing resistant clones can acquire resistance through clonal expansion.²⁵²

In this context, the regulatory role of miRNA is particularly significant. miRNA can precisely target and regulate drug-related mRNA, and its mutation, abnormal expression, or processing may cause aberrant expression levels of target gene proteins.²⁵³ When these target genes are closely related to the sensitivity of tumor cells to drugs, abnormal miRNA changes will directly affect cancer's response to chemotherapy drugs.² miRNAs are closely linked to drug resistance in hematologic malignancies. EV-miRNAs can be transferred to sensitive cells, promoting the formation of drug-resistant tumor cells. Identifying miRNAs that regulate cancer cell resistance has become a focal point in hematologic tumor research.²⁵²

Key target genes involved include anticancer drug target-related genes, pharmacokinetics-related genes, key genes in drug action signaling pathways, and DNA damage repair-related genes. As shown in Figure 5B, they collectively constitute a regulatory network for chemotherapy sensitivity.²⁵⁰ Specifically, miR-192 and miR-215 target thymine synthase (TS), the main target of fluoropyrimidine drugs used to treat colorectal cancer. By downregulating TS expression, miR-192 and miR-215 significantly increase the sensitivity of colon cancer cells to 5-FU.²⁵⁴ Conversely, the multidrug resistance-associated protein 1 (MRP1) is a key efflux transporter, and its overexpression often leads to chemotherapy resistance in breast cancer. miR-145 can target MRP1 and enhance the sensitivity of breast cancer cells to doxorubicin by inhibiting its expression.²⁵⁵ Key genes involved in DNA repair, such as excision repair cross-complementation group 1 (ERCC1) and XPC, are regulated by miR-149 and miR-488, respectively.²⁵⁶

The TME plays a critical role in mediating cancer drug resistance.²⁵¹ Factors such as tumor-TME communication, stress-induced selection pressure, and the acidic TME environment contribute to drug resistance.²⁵² EV-miRNAs reflect the traits of parental tumor miRNAs²⁵⁷ and can transfer drug resistance traits to sensitive cells, enhancing drug resistance by promoting tumor progression and increasing drug efflux.^{251, 258} In the TME, EV-miRNAs play a regulatory role in chemotherapy resistance by modulating CAFs, macrophages, CSCs, EMT, and autophagy (Figure 5C).²⁵¹ Autophagy, a metabolic process that eliminates damaged or redundant molecules and organelles, can also promote tumor resistance, particularly in hypoxic TMEs, where it delays cell apoptosis.²⁵⁹

CSCs, a subset of tumor cells with self-renewal capacity and multidirectional differentiation potential, often exhibit complex drug resistance mechanisms. Specific miRNAs (e.g., miR-34a, miR-21, and miR-155) regulate CSC properties and contribute to cancer drug resistance. The close relationship between EMT and cancer drug resistance has been established, with evidence showing that EMT enhances CSC chemotherapy resistance. EV-miRNAs play a key role in the interaction between CSCs and EMT.¹⁰¹ Additionally, CSCs communicate with the TME, further driving chemotherapy resistance.²⁶⁰ CAF-derived EV-miR-146a-5p promotes CSC proliferation in UBC and enhances drug resistance through tumor-stroma interactions.¹²⁵ MDSCs also utilize miRNAs to increase the stemness of ovarian carcinoma cells by triggering miR-101 and regulating the co-repressor gene CtBP2, inducing stem cell traits.²⁶¹

Paclitaxel (PTX) is a highly effective, low-toxic, broad-spectrum natural anticancer drug widely used in treating breast cancer, ovarian cancer, head and neck cancer, and lung cancer. PTX induces cell cycle arrest, inhibits mitosis, phosphorylates antiapoptotic proteins to inactivate B-cell lymphoma-2 (Bcl-2), activates immune responses to eliminate tumors, and regulates miRNAs associated with tumor progression. PTX's regulation of calcium signals leads to mitochondrial calcium ion release and programmed cell death.²⁶² miRNAs are also involved in PTX resistance in various cancers. For instance, in breast cancer, miR-451 mediates PTX resistance by regulating YWHAZ, and PTX-induced exosomal miR-378a-3p activates the EZH2/STAT3 signaling pathway, promoting resistance.²⁶³ In lung cancer, miR-17-5p downregulation increases Beclin1 expression, a key autophagy regulator, enhancing PTX resistance. In ovarian cancer, circCELSR1 regulates PTX resistance through FOXR2 expression, while miR-106a and miR-591 also contribute to resistance. In cervical cancer, upregulation of miR-375 correlates with increased PTX resistance. Across various cancers, miR-21 overexpression inhibits PTEN, activates the PI3K/Akt pathway, and fosters PTX resistance.²⁶⁴

6.2 Strategies targeting miRNAs associated with drug resistance

The targeted regulation of therapeutic-related mRNAs by miRNAs constitutes the core mechanism of cancer cell resistance.²⁵⁴ Specifically, miRNAs can inhibit the expression of specific efflux transporters and enzymes, significantly enhancing drug accumulation in cells and improving therapeutic efficacy. Additionally, some miRNAs can directly target pharmacological targets, such as proto-oncogenes, to achieve therapeutic effects by inhibiting their protein production. miRNAs themselves can also serve as therapeutic targets. Their expression may be used as biomarkers to predict drug resistance.²⁶⁵ This unique mechanism provides a theoretical basis for developing miRNA therapeutic strategies to overcome drug resistance, including using miRNA mimics or antagonists, regulating key proteins that interact with anticancer drugs (such as MRP), and the synergistic effects of anticancer drugs.^{252, 254}

Studying miRNA not only helps predict individual tumors' resistance to anticancer drugs but also guides oncologists in tailoring reasonable treatment plans for patients, providing a crucial basis for future drug resistance avoidance strategies.²⁵⁴ Experimental research has shown that applying ds-miR-634 mimics in vitro esophageal squamous cell carcinoma (ESCC) cells and xenograft mouse models significantly enhanced cisplatin's antitumor effect in ESCC, demonstrating the great potential of miRNA mimic drugs.⁹⁶ Among tumor suppressor genes, PTEN is notable for its frequent inactivation in cancer, with its loss contributing to treatment resistance. Recent studies suggest that the ncRNAs/PTEN axis is pivotal in conferring resistance to various cancer therapies, presenting significant therapeutic potential in overcoming resistance.²⁶⁶

By targeting resistance-related miRNAs, relevant mRNAs can be regulated to reverse PTX resistance. miRNA-200c can reverse EMT-induced PTX resistance through cathepsin L (CTSL).²⁶⁷ Ectopic expression of miR‑335 or depletion of its target gene SETD8 increases cell sensitivity to PTX.²⁶⁸ Upregulation of miR-497-5p improves breast cancer patients' response to PTX by inhibiting MALAT1 and SHOC2.²⁶⁹ miR-497-5p binds to TRPM2-AS and reverses its promotion of PTX resistance.²⁷⁰ miR-522-3p reduces PTX resistance in paclitaxel-resistant ovarian cancer in vitro by downregulating E2F2.²⁷¹

Despite the potential of miRNA therapy, miRNA-based therapies are still relatively rare in clinical trials, particularly in achieving anticancer efficacy by reversing drug resistance, which remains in the preclinical research stage. Future advancements in research and technology are expected to enhance miRNA therapy's role in cancer treatment.

6.3 Emerging technologies for studying miRNA-Related drug resistance

miRNA holds great potential and offers new strategies for treating cancer resistance by regulating genes related to drug targets, drug pharmacokinetics, DNA damage repair, and key cell signaling pathways.² However, traditional methods for synthesizing miRNA drugs often require complex artificial modification processes, increasing production costs and potentially introducing different physical, chemical, and biological properties, and even potential toxicity.²⁷²

The miRNome profile is closely related to tumor-specific histopathology and molecular characteristics, such as stromal cell infiltration and tumor driver mutations.²⁷³ miRNA mutations play an important role in human cancers' occurrence and development, as shown in Figure 5A. Therefore, the miRNA expression profile in human cancers is closely related to cancer detection, staging, progression, and response to treatment.⁹⁷ More importantly, as a class of epigenetic drugs, miRNA provides a new direction for cancer treatment at the gene regulation level. The epigenetic and genomic properties of miRNA drugs reveal differences between individuals, offering new insights into personalized drug treatment and opening new avenues for developing more effective treatments.²⁷⁴

Ferroptosis, distinct from other forms of cell death, is triggered by the accumulation of intracellular iron and subsequent lipid peroxidation.²⁷⁵ Increasing evidence supports ferroptosis as a target for overcoming chemotherapy resistance, as some chemotherapeutic drugs synergize with ferroptosis by promoting iron metabolism or inhibiting antioxidant mechanisms such as glutathione synthesis.^{276, 277} miRNAs play a pivotal role in this process; for example, miR-522 secreted by CAFs blocks lipid-ROS accumulation, inhibiting ferroptosis and promoting drug resistance in gastric cancer.²⁷⁸

In the realm of cancer treatment, drug repurposing strategies are also opening new possibilities. Several existing drugs, including nonsteroidal anti-inflammatory drugs (NSAIDs), metformin, dichloroacetate, enalapril, ivermectin, bazedoxifene, melatonin, and S-adenosylmethionine, are being explored for new applications in cancer therapy.²⁷⁹ miRNAs play a central role in these strategies by modulating tumor suppressor and oncogenic miRNAs. For instance, dichloroacetate (DCA) and let-7a have been shown to work together to induce apoptosis in triple-negative breast cancer cells, while reducing reactive oxygen species and stimulating mitochondrial biogenesis.²⁸⁰ Metformin has demonstrated anticancer effects through miRNAs such as let-7, miR-26, and miR-200, which regulate tumor metabolic pathways.²⁸¹ Similarly, aspirin can improve lung cancer outcomes by targeting the miR-98/WNT1 axis.²⁸² Other examples include simvastatin, which reduces let-7 levels to inhibit cell viability,²⁸³ and miR-192, which enhances osteosarcoma chemosensitivity to methotrexate.²⁸⁴

Multidrug combinations can be used to reverse resistance, PTX and miRNA co-delivery play a crucial role in breast cancer management. These strategies reduce the toxicity of free PTX by improving cellular uptake and bioavailability, achieving tumor site accumulation. Molecular docking studies showed the enhanced anticancer activity of nanoformulations like ANG1005 and Abraxane, which significantly interact with α-tubulin and Bcl-2 proteins. Therefore, ANG1005 and Abraxane may be more suitable for breast cancer therapy than free PTX.²⁸⁵

In recent years, the scientific community has focused on developing recombinant RNA technology to produce large quantities of miRNA preparations economically and efficiently. These miRNAs are naturally produced in cells, with structures consisting of unmodified or post-transcriptionally modified nucleobases, more accurately mimicking natural miRNAs' structure, function, and safety.^{286, 287} The tRNA/pre-miRNA-based method in recombinant RNA technology is particularly noteworthy for its power and efficiency, enabling continuous and efficient production of miRNA reagents.²⁸⁸ Additionally, a study successfully developed a recombinant miRNA vector that significantly increased recombinant miRNA expression in bacteria.²⁸⁹

RNA nanoparticles can also be used for specific cancer treatment. A novel multivalent RNA nanoparticle was designed to store three copies of hepatocyte targeting ligands, one copy of miR122, and 24 copies of PTX to overcome drug efflux and chemoresistance for synergistic treatment of HCC. These nanoparticles enhance tumor localization and effectively deliver miR122 and PTX to liver cancer cells, enhancing their EPR and receptor endocytosis mediated by hepatocyte targeting ligands. miR122 silences drug efflux proteins and oncoproteins, synergizing with PTX to enhance cytotoxicity.²⁹⁰