Extensive research has indicated that certain miRNAs are significantly downregulated in the cartilage tissue of OA patients, thereby promoting disease progression. The overexpression of these miRNAs can protect against OA by regulating the biological functions of chondrocytes and immune cells (31, 32). For example, Lin et al. (33) fully confirmed through clinical samples and animal experiments that miR-653-5p inhibits the JAK/STAT3 signaling pathway by targeting and suppressing IL-6 expression. Inactivation of this pathway alleviates chondrocyte senescence by downregulating expression of the cyclin-dependent kinase inhibitors p21 and p16^INK4a, while simultaneously suppressing matrix metalloproteinase-13 (MMP13) and a disintegrin and metalloproteinase with thrombospondin motifs 5 (ADAMTS5), thereby significantly mitigating the progression of OA pathology. Research has revealed that miR-149 directly targets and suppresses VCAM-1 expression, thereby inhibiting the activation of the PI3K/AKT signaling pathway. The inactivation of this pathway reduces inflammation by downregulating NF-κB activity, regulates the balance between Bax and Bcl-2 to inhibit chondrocyte apoptosis, and promotes the synthesis of ECM components such as collagen II and aggrecan, thereby collectively alleviating OA. Moreover, upregulating miR-149 or inhibiting VCAM-1 in in vivo models can significantly improve the pathological manifestations of OA, suggesting that the miR-149/VCAM-1/PI3K/AKT axis may represent a potential therapeutic pathway for OA treatment (34). Lu et al. (35) reported that the expression of miR-22-3p is downregulated in OA, promoting disease progression, and they demonstrated through in vitro and in vivo experiments that miR-22-3p alleviates OA progression by inhibiting the expression of the inflammasome NLRP3. Additionally, bioinformatics analysis and dual luciferase reporter assays confirmed that HMGA1 is a target of miRNA-4701-5p. By targeting and suppressing HMGA1 expression, miRNA-4701-5p alleviates IL-1β-induced inflammatory damage, apoptosis, and cytotoxicity in chondrocytes (36). Another study reported that miR-214-3p directly targets IKKβ to inhibit the NF-κB signaling pathway, thereby reducing the expression of MMP3, MMP13, and the apoptotic executor protein cleaved-caspase-3, while concurrently elevating the levels of the anti-apoptotic protein Bcl-2. This mechanism alleviates the degradation of the cartilage ECM and apoptosis; moreover, animal experiments have confirmed that intra-articular injection of miR-214-3p agonists can slow the progression of OA (37). Furthermore, miR-548d-5p, miR-485-3p, and miR-140-5p also promote chondrocyte proliferation and inhibit apoptosis (38–41).

In the pathological environment of OA, multiple miRNAs amplify inflammatory responses, promote chondrocyte senescence and apoptosis, thereby disrupting joint homeostasis and driving OA progression. For example, Liu et al. (42) reported that miR-33-5p targets and suppresses SIRT6, thereby leading to the upregulation of cell cycle–dependent kinase inhibitors such as p16 and p21 and a weakening of the inhibitory control over the NF-κB signaling pathway. These changes induce chondrocyte senescence, promote inflammation, and trigger cell cycle arrest, ultimately leading to cartilage degeneration and OA progression. This effect was further validated in animal experiments. Similarly, miR-5581 is upregulated in the cartilage tissue of OA patients. Bioinformatics prediction and dual luciferase reporter assays confirmed that its targeting reduces nuclear respiratory factor 1 protein levels, thereby inhibiting cell proliferation, promoting apoptosis, and disrupting ECM homeostasis (43). Tavallaee et al. (44) reported that miR-27b-3p is upregulated in OA, accelerating OA progression by promoting synovial fibrosis via the PPARG/ADAMTS8 axis. In addition, miR-128a exacerbates OA progression by targeting and suppressing nuclear receptor subfamily 1 group D member 2 (NR1D2), leading to decreased expression of chondrogenic factors and reduced ECM synthesis, whereas cartilage-specific knockout of miR-128a can restore NR1D2 expression and significantly alleviate OA severity (45).

Owing to the complex interaction networks between miRNAs and RNA, some miRNAs do not exert purely protective or inhibitory effects but instead exert differential influences on the progression of OA through distinct signaling pathways. For example, miR-18a-3p can target and inhibit PDP1, thereby reducing the expression of inflammatory cytokines and matrix metalloproteinases, improving cartilage matrix remodeling, and suppressing inflammatory responses (46). However, Ding et al. (47) reported that miR-18a-3p can suppress HOXA1 expression, leading to the upregulation of cleaved caspase-3 and cleaved PARP and the downregulation of BCL-2 expression. This promotes chondrocyte apoptosis and drives the pathological progression of OA. In addition, studies have shown that miR-146a-5p can simultaneously target and inhibit the expression of CXCR4 and NUMB, the former of which blocks SDF-1/CXCR4-mediated chondrocyte autophagy, thereby reducing cartilage degeneration and OA progression (48). Conversely, the latter promotes chondrocyte apoptosis and inhibits autophagy, consequently accelerating OA progression (49). In summary, the same miRNA may target different genes, which in turn have different effects on the progression of OA.

Notably, certain miRNAs exhibit consistent expression alteration patterns in both human OA and small animal OA models. For example, protective miRNAs such as miR-140-5p, miR-214-3p, and miR-548d-5p are downregulated in surgically induced OA joints of rodents (37, 38, 41). Conversely, disease-promoting miR-128a is upregulated in both human OA and rodent OA models, and its functional intervention can alter OA progression in animals (45). This suggests that the pathological roles of these miRNAs are conserved across species. Beyond rodent models, large animal OA models have gained increasing attention in miRNA research in recent years. Large animal joints more closely resemble human joints in terms of size, weight-bearing capacity, cartilage thickness, and natural disease progression, enabling better simulation of the physiological effects of miRNAs within the complex joint microenvironment (50). For example, in equine OA models, multiple studies have reported abnormal expression of miR-223, miR-199a-3p, miR-27b, and other related miRNAs in synovial fluid or extracellular vesicles, closely associated with inflammatory responses, cartilage matrix degradation, and synovial fibrosis (51–53). miR-146a is upregulated in the synovial tissue of dogs with spontaneous OA and is likewise upregulated in early-stage human OA (54, 55), indicating a shared upregulation pattern that provides a valuable foundation for subsequent clinical translation. However, the trends of miRNA changes in large animals do not always align with those in humans. For instance, miR-146a is downregulated in late-stage human OA (54). Furthermore, taking miR-140-5p as an example: it is considered a protective miRNA in human OA with decreased expression in synovial fluid (40). In contrast, in equine OA models, researchers observed a significant increase in miR-140-5p within synovial fluid, accompanied by a decrease in plasma levels, and this phenomenon was interpreted as a compensatory protective response at the joint site against cartilage degradation (56).

Extracellular vesicles (EVs) are cell-derived nanoparticles encapsulated by a lipid bilayer membrane that can encapsulate various bioactive substances, such as nucleic acids, proteins, and lipids (57). Exosomes are a subtype of EVs derived from the endosomal pathway that play a crucial role in intercellular communication and signal transduction (58, 59). EVs offer considerable potential for targeted delivery of miRNA therapies for OA owing to their inherent biocompatibility, low immunogenicity, and cytotoxicity (60). Liang et al. (61) fused cartilage-targeted peptides (CAPs) with lysosome-associated membrane glycoprotein 2b on the surface of exosomes, achieving efficient encapsulation and specific delivery of miR-140, effectively alleviating cartilage degeneration and disease progression. Additional researchers have utilized CAP-modified exosomes derived from subcutaneous fat-derived mesenchymal stem cells (MSCs^SC-Exos) to deliver miR-199a-3p, significantly enhancing the effect of cartilage repair (62). Furthermore, Liu et al. (63) constructed a dual-engineered EVs by loading exogenous miR-223 into human umbilical cord mesenchymal stem cell-derived EVs (hUC-EVs) using electroporation and combining it with surface-modified collagen II-targeted peptides, which significantly enhanced its cartilage targeting ability and therapeutic efficacy. Recently, Wu et al. (64) constructed an engineered stem cell cluster to continuously deliver miRNA-874-3p using EVs secreted by bone marrow-derived mesenchymal stem cells to achieve chondrocyte repair and regeneration while effectively mitigating the progression of OA.

Microfluidic technology has shown great potential in the field of drug delivery because of its ability to achieve precise control of droplet volume, morphology, and internal structure (65). Yang et al. (66) prepared degradable gelatin microspheres (CAP-EXOs/miR-148a@GAM) loaded with CAP-modified exosomes on the basis of microfluidic technology for the continuous delivery of miR-148a for the treatment of OA, significantly promoting the synthesis of ECM and delaying the progression of OA by inhibiting ROBO2 expression and the MAPK signaling pathway. Hydrogel microparticles (HMPs) based on hyaluronic acid prepared by microfluidic technology encapsulate extracellular vesicles derived from subcutaneous adipose stem cells loaded with miR-99b-3p to form the HMPs@Exos^ScAT-99b-3p system; research by Yin et al. (67) reported that this system not only addresses the issues of a short in vivo half-life and poor targeting of EVs but also promotes cartilage self-repair and homeostasis restoration by specifically inhibiting ADAMTS4, thereby demonstrating significant therapeutic potential for OA both in vivo and in vitro.

In the field of biological therapy research for OA, large animal studies based on secretome analysis are gradually emerging. Presently, multiple studies in dogs and horses have employed intra-articular administration of MSC secretome or blood cell secretome (BCS) to reduce inflammatory responses, alleviate joint swelling, and improve joint function (68–70). Other studies have shown that injecting human MSC-derived exosomes into porcine knee weight-bearing osteochondral defect models can significantly improve the morphological and mechanical repair of cartilage and subchondral bone, indicating that they exert cross-species effects (71). As a key component of the secretome, EVs contain abundant endogenous miRNAs (72), which may constitute one mechanism underpinning their therapeutic effects. These studies provide robust translational evidence for future engineered EV delivery systems loaded with specific miRNAs.

In comparison with EV delivery systems, polymeric nanoparticles (PNPs) demonstrate highly controllable physicochemical properties, along with higher production efficiency and enhanced drug loading capacity (73, 74). Polyamidoamine (PAMAM) dendrimers are extensively employed in drug delivery because of their highly branched architecture and internal cavities (75), and polyethylene glycol (PEG) modification can optimize their surface charge and stability (76). Chen et al. (77) delivered miR-224-5p using fifth-generation PAMAM (G5-PAMAM) functionalized with arginine, phenylalanine, and histidine, thereby alleviating cartilage degeneration and synovial inflammation by inhibiting pentraxin 3. Long et al. (78) utilized 50% PEG-modified G5-PAMAM dendrimer macromolecules loaded with miR-455-3p and combined with CAPs and minimal self-peptides (MSPs), which were then further complexed with a thermosensitive hydrogel. This composite not only achieved sustained release and precisely targeted delivery of miR-455-3p but also suppressed chondrocyte hypertrophy and apoptosis while promoting matrix synthesis by regulating key pathogenic pathways. Ji et al. (79) employed PEGylated PAMAM dendritic carriers modified with chondrocyte-specific nucleic acid aptamers (tgg2) to deliver miR-141/200c cluster inhibitors. This approach relieved the suppression of SIRT1, enabling SIRT1 to inhibit the IL-6/STAT3 pathway. Consequently, inflammation was suppressed, the ECM and cartilage were protected, and cartilage degeneration was even reversed, thereby effectively delaying the progression of OA.

Polyvinylamine (PVAm), as a hydrophilic cationic polymer, can effectively bind negatively charged RNA and exhibits good transfection efficiency, making it highly promising for use in the field of targeted therapy (80). Poly(lactic-co-glycolic acid) (PLGA) is a hydrophobic polymer that has been extensively studied because of its excellent biocompatibility, biodegradability, and good mechanical properties (81). Zhao et al. (82) achieved chondro-targeted delivery of miR-140 via CAP-PVAm-PLGA, a copolymer formed from CAP-modified PVAm and PLGA, thereby suppressing inflammation and matrix degradation. Another study employed triple polysaccharide nanoparticles composed of chondroitin sulfate, chitosan, and hyaluronic acid to deliver miR-149-5p. This approach not only effectively transfected cells and significantly reduced the expression of the target gene FUT-1, but also promoted cartilage differentiation through the components released upon degradation (83).

Lipid nanoparticles (LNPs) possess excellent biocompatibility and ability to improve drug stability and targeting efficiency (84), and they are widely used for delivering nucleic acids and small-molecule drugs (85). The presence of positively charged hydrophilic head groups in cationic lipids promotes electrostatic interactions with negatively charged nucleic acids, thereby enhancing the encapsulation of nucleic acids and promoting intracellular escape (86). Zheng et al. (87) encapsulated AgPEI nanoparticles loaded with miR-200c-3p within liposomes to form the Lipo-AgPEI-miR-200c-3p complex, which effectively blocked the inflammatory-apoptotic-matrix degradation malignant cycle of chondrocytes in vitro. However, its in vivo efficacy and safety require further verification. Additionally, He et al. (88) utilized lornoxicam cationic liposomes to load miR-140, obtaining cationic liposomes coloaded with lornoxicam and miR-140, which combine anti-inflammatory and gene therapy functions. This significantly improved the stability of miR-140 and its uptake efficiency by chondrocytes, effectively alleviating joint swelling, inhibiting inflammation, and promoting cartilage repair. This finding encourages the combination of other drugs with miRNA delivery for synergistic treatment of OA.

Compared to organic nanoparticles, inorganic nanocarriers exhibit superior structural stability and enhanced mechanical properties, thereby attracting significant attention from researchers (89). Yang et al. (90) utilized ZIF-8, a metal–organic framework material featuring noncytotoxic zinc(II) as its metal coordination center, as a miRNA delivery vehicle. Through a Y-shaped microfluidic chip, the researchers efficiently prepared miR-200c-3p@ZIF-8, significantly reducing the expression of inflammatory factors in OA. However, traditional spherical nanoparticles have limited loading capacity and insufficient transfection efficiency (91); thus, researchers have developed new inorganic nanocarriers. For example, Li et al. (92) delivered miR-17-5p using streamlined ZnO nanoparticles combined with methyl acrylate-modified gelatin (GelMA). In this way, bone marrow mesenchymal stem cells are recruited via zinc ions, promoting matrix synthesis while inhibiting catabolic processes. Chen et al. (93) used urchin-like cerium dioxide nanoparticles to efficiently load miR-224-5p, and their surface area increased, which not only improved their miRNA loading capacity and antioxidant capacity but also enhanced their therapeutic efficacy. Taken together, these findings indicate that the novel nanoparticle-based miRNA carriers demonstrate superior loading capacity and can effectively enhance the therapeutic efficacy of OA treatment. However, the long-term biological safety of such materials requires further investigation.

EVs exhibit exceptional biocompatibility, low immunogenicity, and excellent tissue penetration. However, their large-scale production remains constrained by technical challenges, with complex purification processes that are difficult to standardize (60). Furthermore, EVs possess limited cargo capacity, and the potential risks associated with engineering modifications remain uncertain. EVs derived from different cell sources may carry bioactive molecules, potentially triggering uncontrollable biological effects (94). In comparison, PNPs are relatively simple to prepare, less costly, and exhibit higher loading efficiencies (73). However, many cationic polymers carry positive charges that may cause cell membrane damage and cytotoxicity, potentially triggering inflammatory responses (95). Additionally, degradation products from certain polymers may also induce cytotoxicity and immune reactions (96). LNPs represent one of the most mature nucleic acid delivery platforms currently available and have been extensively utilized in mRNA vaccines (85). LNPs offer advantages such as stable preparation processes, ease of large-scale production, and high delivery efficiency (84). However, they still carry potential toxicity risks. In addition, due to the susceptibility of lipid membranes to oxidation and hydrolysis, they typically require cryopreservation or cold storage to ensure stability (97). Inorganic nanoparticles possess unique physicochemical properties such as large surface area, excellent stability, and precise controllability of morphology. Their primary concern lies in insufficient long-term biosafety because some inorganic nanoparticles degrade slowly or not at all within the body, potentially accumulating in tissues and triggering toxicity or inflammation (98).