Progress in siRNA therapy and delivery platforms for osteoarthritis

Summary

Osteoarthritis (OA) is driven by a cycle of cartilage decay, synovial inflammation, and bone remodeling. This progression is fueled by the self-perpetuating synergy between matrix-degrading enzymes and pro-inflammatory cytokines, posing a formidable barrier to effective therapy. Small interfering RNAs (siRNAs) offer a transformative precision medicine paradigm, harnessing the RNA interference (RNAi) machinery to silence pathogenic drivers. However, their clinical application hinges on navigating the complex intra-articular microenvironment. This review synthesizes current progress in siRNA delivery architectures, spanning viral vectors, lipidic nanoplatforms, polymeric assemblies, and hybrid biomimetic systems. The spatiotemporal retention and intracellular bioavailability of siRNAs have been augmented by key engineering breakthroughs, including ligand-mediated active targeting, precise surface topography tuning, and integrated stimulus-responsive functionalities. This review, by exploring future directions like multi-target silencing and the use of large animal models relevant to clinical practice, highlights how siRNA-based treatments are moving beyond experimental ideas to applications as real anti-OA drugs.

Graphical abstract

Molecular biology; Medical biotechnology; Biomedical materials

Introduction

Osteoarthritis (OA) stands as a typical degenerative joint disorder, characterized by the progressive degeneration of articular cartilage and chronic inflammation of the periarticular microenvironment. Beyond simple wear-and-tear, the pathogenesis of OA involves a complex interplay of subchondral bone remodeling, meniscal lesion, osteophyte formation, and synovial fibrosis (Figure 1).^1,2,3,4,5 This pathological progression is further exacerbated by a dysregulated homeostatic balance, where accelerated chondrocyte senescence and apoptosis coincide with a catabolic dominance in cartilage matrix metabolism.^6,7 Clinically, these molecular aberrations manifest as debilitating pain, joint stiffness, and impaired mobility, eventually culminating in functional collapse and joint deformity. Given the global aging trend and rising obesity rates, the escalating prevalence of OA, notably within weight-bearing joints, has imposed an unprecedented socioeconomic burden while positioning the disease as a leading driver of adult disability worldwide.^8,9,10 Despite the profound impact of OA, current therapeutic interventions remain largely palliative. First-line strategies, including physical therapy and weight management, offer limited structural protection. Pharmacological options, primarily centered on non-steroidal anti-inflammatory drugs (NSAIDs), target symptomatic relief, rather than disease modification. Biologic agents targeting nerve growth factor (NGF), exemplified by tanezumab, alleviate OA pain by blocking the NGF signaling pathway. Clinical trials have confirmed that both intravenous and subcutaneous administration can significantly reduce pain in knee and hip OA. Notably, however, this drug may precipitate rapidly progressive OA and peripheral sensory abnormalities. These safety concerns have limited its clinical application.^11,12 While surgical procedures like arthroscopic debridement or total joint replacement are indicated for end-stage OA-related patients, they fail to address the underlying biological drivers of early-to-mid stage degeneration.^3,13 In this context, RNA interference (RNAi) has emerged as a transformative frontier for precision medicine. By harnessing the biological effect of the RNA-induced silencing complex (RISC), small interfering RNAs (siRNAs) can execute highly specific, post-transcriptional gene silencing.^14,15 The clinical viability of siRNAs has been validated by the landmark approvals of hepatic-targeted therapeutics such as patisiran and givosiran, which utilize lipid nanoparticles (LNPs) or GalNAc-conjugation to achieve excellent gene knockdown in the liver. However, translating the success of hepatic RNAi to the musculoskeletal system necessitates overcoming unique physiological barriers, including the dense extracellular matrix (ECM) of cartilage and the rapid clearance of molecules from the synovial fluid.^16,17 As a degenerative joint disorder characterized by articular cartilage degradation, synovial inflammation, and subchondral bone remodeling, OA is closely associated with the activation of inflammatory pathways, excessive cartilage catabolism, and elevated levels of pain-related mediators during its pathogenesis.¹⁸ The activation of these aberrant pathways is primarily attributed to the dysregulated expression of genes at the molecular level, whereas current therapeutic modalities are confined to symptomatic relief or surgical intervention.¹⁹ Therefore, siRNA therapy based on the RNAi mechanism provides an exact breakthrough solution for OA therapy, where disease-specific siRNA sequences can be rationally designed to target the key pathogenic genes involved in the core pathological processes of OA, thus enabling the precise silencing of key molecules in inflammatory pathways as well as genes associated with cartilage degradation.²⁰ This allows for the inhibition of articular inflammatory responses and cartilage degeneration processes at their source, which is expected to achieve precise and causal treatment of OA and emerge as a pivotal direction to break through the bottlenecks of traditional clinical therapies.¹⁶ However, its in vivo application relies on efficient delivery platforms to overcome its inherent drawbacks, such as poor stability and insufficient cellular uptake, before clinical translation can be achieved.²¹ Bridging this gap requires the development of sophisticated delivery vehicles capable of enhancing intra-articular retention and facilitating deep ECM penetration. This review systematically delineates the emerging landscape of siRNA therapeutic targets in OA, evaluates the engineering strategies for advanced delivery platforms, and discusses the critical transitions from preclinical proof-of-concept to clinical reality.

Figure 1.

Basic characteristics and treatment strategies of OA

(A) Risk factors, pathogenic mechanisms, and common treatments for OA. Adapted from Zahir-Jouzdani et al.²¹

(B) Schematic drawing of an osteoarthritic joint. Adapted from Bijlsma et al.¹³⁰

(C) Barriers toward intra-articular delivery of DMOADs. Adapted from Gao et al.³⁷

(D) Pathological changes in OA progression. Adapted from Hunter et al.²

(E) Therapeutic strategies for OA by targeting ion channels. Adapted from Zhou et al.¹⁹

Therapeutic targets and synergistic strategies for siRNA intervention in OA

The fundamental premise of siRNA therapeutics lies in the highly specific interrogation and silencing of pathogenic gene expression. The complex pathophysiology of OA, characterized by cartilage lesion, synovial inflammation, and aberrant subchondral bone remodeling, offers distinct therapeutic advantages for siRNA intervention, potentially arresting the disease trajectory at its molecular origins.

Targeting chondrocyte senescence and autophagy dysfunction

Within the degenerative joint microenvironment, cartilage degeneration is intrinsically linked to maladaptive chondrocyte behaviors, notably accelerated cellular senescence and compromised autophagy. Targeting these pathological signaling axes represents a potent therapeutic avenue to preserve cartilage integrity. Accumulating evidence implicates the accumulation of senescent chondrocytes in driving OA progression through the senescence-associated secretory phenotype (SASP). In the progression, the SASP released by senescent chondrocytes is pivotal in mediating inflammation and remodeling the joint microenvironment. The diverse array of bioactive factors secreted through this phenotype act via paracrine signaling and gap junctions to directly induce a senescent phenotype in neighboring cells. This process establishes a self-perpetuating cycle of senescence and inflammation within the joint, thereby exacerbating OA pathogenesis.^22,23 Zhao et al.²⁴ identified fibroblast activation protein (FAP) as significantly upregulated in OA-derived chondrocytes, acting as a critical regulator of senescence via the NF-κB signaling pathway. By engineering an LNP delivery system to load FAP-targeting siRNAs, they achieved potent gene silencing that inhibited the FAP-NF-κB-SASP axis, thereby attenuating cartilage degeneration. Complementing this approach, Wang et al. identified ADAM19 as an effective target for rejuvenating senescent cells. Delivering siADAM19 effectively attenuated the senescent phenotype and promoted hyaline cartilage regeneration, suggesting a potential for reversing cellular aging. Parallel to senescence, defective autophagy leading to the accumulation of damaged cellular components precipitates chondrocyte dysfunction.²⁵ Fu et al.²⁶ reported that interleukin-33 (IL-33) induces senescence by inhibiting autophagy via the p38 MAPK pathway. Consequently, they developed neutrophil membrane-camouflaged nanoparticles loaded with siIL-33 to actively target inflamed tissues and mitigate cartilage degradation. Similarly, knockdown of Tribbles homolog 3 (TRB3), which exhibits age-dependent expression, has been shown to restore autophagic flux and alleviate senescence.²⁷ Moving beyond single-target interventions, recent research has pivoted toward synergistic strategies that address multiple dysregulated pathways simultaneously. For instance, a Zn²⁺-coordinated nanosystem co-loading metformin and p65 siRNA was developed to exert dual chondroprotective effect by activating autophagy and inhibiting NF-κB signaling.²⁸ In another combinatorial approach, Zhang et al. achieved co-delivery of the anti-inflammatory agent curcumin and HIF-2α-targeting siRNA. This dual-modal system concurrently downregulated pro-inflammatory cytokines and silenced the pathogenic HIF-2α gene, exerting a synergistic therapeutic effect on OA progression.²⁹

Inhibiting catabolic enzymes and matrix degradation

A hallmark of OA pathology is the catastrophic imbalance between cartilage anabolism and catabolism, driven by the excessive secretion of degradative enzymes, including matrix metalloproteinases (MMPs, particularly MMP-13), ADAMTS-5, and hyaluronidases. Targeted silencing of these key factors of cartilage degeneration has shown significant therapeutic promise. To overcome delivery barriers, Wang et al.³⁰ engineered a multi-functional hydrogel comprising ROS-responsive, phenylboronic acid-modified hyaluronic acid (HA). This delivery system encapsulated polyethyleneimine-Poly(Ethylene Glycol) (PEI-PEG)-modified Fe₃O₄ nanoparticles bearing siMMP-13, creating a dual-action system that simultaneously combats oxidative stress and precisely silences the primary collagenase MMP-13. Further integrating therapy with diagnostics, MMP13 siRNA-loaded micelles were developed for early post-traumatic OA (PTOA), capable of providing fluorescence-based diagnostic feedback on MMP13 levels while delivering therapeutic siRNA doses.³¹ Leveraging endogenous transport mechanisms, Zhang et al.³² constructed cartilage-targeted exosomes by modifying their surface with chondrocyte-affinity peptides. This biomimetic strategy enhanced cellular uptake and utilized the inherent biocompatibility of exosomes to protect encapsulated siMMP13 from enzymatic degradation, highlighting the potential of biological vehicles in enhancing drug efficacy.

Modulating the inflammatory microenvironment and macrophage polarization

Chronic, low-grade inflammation is a critical driver that fosters a self-perpetuating feedback loop of cartilage degeneration. Proteases from degraded cartilage activate synovial inflammation, which, in turn, releases cytokines that cause further destruction. Breaking this cycle is important. The transcription factor Stat3 is hyperactivated in the osteoarthritic joint, amplifying the inflammatory cascade and upregulating MMPs. To interrupt this, Lv et al.³³ deployed a highly precise DNA origami-based nanocarrier modified with CD44 aptamers for targeted delivery of Stat3 siRNA to chondrocytes. This specific silencing significantly downregulated degradative factors, demonstrating the power of structurally defined nanomaterials in overcoming siRNA stability and targeting challenges.

Beyond direct chondroprotection, recalibrating the immune microenvironment, specifically macrophage phenotype, is crucial. M1-polarized macrophages secrete pro-inflammatory cytokines and acidify the local microenvironment, which induces pathogenic carbonic anhydrase IX (CA9) expression. Yan et al.³⁴ developed a synergistic nanoplatform co-delivering CA9 siRNA and a nitric oxide (NO) scavenger. This system simultaneously silenced CA9 to ameliorate the acidic niche and scavenged NO to drive macrophage reprogramming from the pro-inflammatory M1 to the reparative M2 phenotype. Similarly, targeting the Notch signaling pathway, which dictates macrophage inflammation, Chen et al.³⁵ constructed a photothermal-responsive NO nanogenerator combining Notch1 siRNA. This synergistic approach achieved anti-inflammatory efficacy superior to those of monotherapies. Furthermore, addressing oxidative stress, which is closely linked to inflammation, delivery of siRNA against p47phox via PLGA nanoparticles has proven effective in alleviating ROS-induced chondrocyte damage.³⁶

Challenges impeding siRNA therapeutics in OA treatment

Despite the transformative potential of RNAi, the clinical translation of siRNA for OA is severely impeded by a formidable array of biological barriers, stemming from both the physicochemical properties of RNA molecules and the complex physiological environment of the joint.³⁷ The primary hurdle lies in the inherent instability of a naked siRNA, which is rapidly degraded by ubiquitous nucleases in serum and tissue fluids before reaching target cells. Furthermore, unmodified siRNAs can act as a pathogen-associated molecular pattern (PAMP), triggering Toll-like receptors (TLRs) and inducing undesirable innate immune responses.²¹ The polyanionic nature and high hydrophilicity of the phosphate backbone create additional obstacles, preventing passive diffusion across the hydrophobic, negatively charged cellular membranes.³⁸ Even if intracellular delivery is achieved, sequence-dependent off-target effects remain a critical concern, where partial complementarity to non-target mRNAs or supratherapeutic concentrations may precipitate unintended gene silencing and cellular dysfunction.²¹ Beyond these intrinsic molecular limitations, systemic delivery via carriers encounters significant physiological hurdles. Upon administration, nanocarriers are prone to rapid opsonization by plasma proteins, forming a protein corona that facilitates clearance by the reticuloendothelial system (RES), thereby drastically reducing bioavailability at the target site.³⁹ Most critically for OA therapy, the articular joint presents a unique anatomical barrier. The avascular nature of cartilage, combined with its dense, negatively charged ECM network, severely restricts the deep penetration of therapeutic agents.²⁰ Consequently, achieving therapeutically relevant siRNA concentrations within chondrocytes constitutes a paramount challenge, while non-specific biodistribution raises concerns regarding long-term systemic toxicity. Even following successful tissue accumulation and cellular internalization, siRNA therapeutics face a final, critical bottleneck in intracellular trafficking. Internalized carriers are typically entrapped within endosomes.⁴⁰ Failure to engineer efficient endosomal escape mechanisms results in rapid cargo degradation. Moreover, subsequent to cytosolic release, the siRNAs must evade further degradation by cytoplasmic nucleases to effectively engage the RNAi machinery.¹⁶

Classification of siRNA delivery

The therapeutic efficacy of siRNAs is fundamentally contingent upon navigating a series of formidable biological barriers: evading rapid nuclease degradation in the circulation, achieving specific accumulation within target tissues (while minimizing off-target hepatic/renal clearance), and executing efficient endosomal escape into the cytoplasm to engage the RISC process.⁴¹ Overcoming these obstacles necessitates the development of sophisticated delivery vectors. Fueled by advances in bioengineering and nanotechnology, a diverse array of delivery systems has emerged. Broadly categorized into viral and non-viral vectors, these platforms differ fundamentally in their mechanisms, biosafety profiles, payload capacities, and translational viability, constituting the core framework of current siRNA delivery research (Figure 1 and Table 1).

Table 1.

Comparative analysis of delivery platforms for osteoarthritis

Delivery platform type	Core design	Key advantages	Limitations
Viral vectors	modified natural viral infection mechanism	high transfection efficiency	insertional mutagenesis risk
Lipid nanoparticles	ionizable cationic lipids	high encapsulation, strong endosomal escape, and lubrication function	unverified long-term safety
Polymer-based nanocarriers	polymer charge interaction	tunable degradation and multi-drug co-delivery	potential cytotoxicity
Nucleic acid vector systems	Watson-Crick base-paired self-assembled 3D structures	high programmability and precise targeting	in vivo enzymatic degradation
Peptide-based nanomaterials	cationic peptide + siRNA self-assembly	simple synthesis and strong cell penetration	short half-life
Inorganic and hydrogel-based hybrid systems	inorganic carriers + responsive hydrogels	structural stability and multi-modal therapy	inorganic accumulation risk
Exosome vectors	cell-derived exosomes + surface modification	excellent biocompatibility and low immunogenicity	low loading efficiency

Viral vectors

Viral delivery systems re-engineer natural viral infection pathways for gene transfer. By stripping viruses of their pathogenicity while retaining functional domains essential for cell recognition, endocytosis, and nuclear trafficking, these vectors achieve high-efficiency siRNA delivery. Primary viral vectors include retroviruses (RVs), adenoviruses (Ads), and adeno-associated viruses (AAVs). Among these, AAVs have garnered significant attention due to their reduced immunogenicity, sustained gene expression capabilities, and broad tropism, demonstrating potent therapeutic effects in chronic conditions like neurodegenerative diseases.⁴²

In the context of OA, AAV vectors have shown promise in preclinical, large animal models. A recent study utilized AAVs to deliver key therapeutic targets, including interleukin-1 receptor antagonist (IL-1Ra) and insulin-like growth factor-1 (IGF-1), achieving sustained symptom alleviation and tissue protection in equine knee OA models.⁴³ This success points toward translational potential for both veterinary and human medicine. Clinical trials exploring AAV-mediated delivery of IL-1Ra have provided preliminary evidence of durable expression (over one year) without severe systemic adverse events.⁴⁴ Furthermore, local rAAV-mediated IGF-1 gene therapy has demonstrated long-term overexpression, effectively retarding early pathological changes and promoting osteochondral repair. Such localized approaches offer the dual benefit of minimizing systemic toxicity and facilitating clinical translation.⁴⁵ However, the widespread clinical application of viral vectors remains encumbered by significant limitations. Their restricted payload capacity hampers the co-delivery of multiple siRNAs or large nucleic acid complexes. More critically, persistent concerns regarding potential insertional mutagenesis, immune-mediated inflammatory toxicity, and the scalability challenges of high-quality production continue to hinder their broader adoption.²¹

Non-viral delivery systems

Non-viral systems have emerged as a pivotal frontier in siRNA delivery, capitalizing on their superior biosafety, versatile payload capacities, ease of functionalization, and scalability for mass production.⁴⁶ Unlike their viral counterparts, non-viral vectors self-assemble with siRNAs through physicochemical interactions, such as electrostatic attraction, hydrophobic effects, or biomolecular recognition, to form stable nanocomplexes designed to negotiate biological barriers. These platforms span diverse material classes, including nucleic acid-based, lipid-based, polymer-based, inorganic, and bio-inspired nanocarriers.⁴⁷ The central advantage of non-viral vectors lies in their high tunability. Through precise chemical modification and structural engineering, properties such as biocompatibility, targeting specificity, and endosomal escape efficiency can be meticulously optimized without the risk of insertional mutagenesis. Nevertheless, these systems generally exhibit lower transfection efficiency than viruses, primarily due to challenges in achieving robust circulatory stability, sufficient target tissue enrichment, and potent endosomal escape.⁴⁸

Nucleic acid vector systems

Nucleic acid-based nanocarriers represent an ingenious convergence of nanomedicines and structural biology, with DNA nanostructures like tetrahedrons and origami as key examples.⁴⁹ These carriers exploit the strict Watson-Crick base-pairing rules to self-assemble into precisely defined 3D architectures in a “bottom-up” manner.^50,51 For siRNA delivery in OA, nucleic acid nanocarriers employ multi-faceted strategies for cargo loading and protection: (1) chemical conjugation anchors siRNAs to specific sites; (2) spatial encapsulation cages siRNAs within hollow structures; and (3) electrostatic adsorption complexes siRNAs via cationic mediators. These approaches synergistically shield siRNAs from nuclease degradation and reduce renal clearance, significantly extending their in vivo half-life.^52,53,54 To achieve precision delivery, ligand modification is paramount. By displaying specific aptamers or peptides, these nanocarriers can actively home to pathological cells. This strategy ingeniously integrates passive targeting (enrichment via the enhanced permeability and retention effect, EPR effect, due to optimal size) with active targeting (ligand-receptor binding on inflamed synovial cells or chondrocytes), establishing an effective platform for precise genetic intervention.

Liao et al. addressed siRNA instability and poor uptake by developing a tetrahedral framework nucleic acid (tFNA)-based system (Tsi) loaded with NF-κB-targeting siRNA. Tsi effectively inhibited the NF-κB pathway while upregulating NRF2/HO-1, synergistically reducing inflammation and ROS to alleviate oxidative stress and apoptosis. In vivo, Tsi demonstrated superior joint retention and exerted protective effects on both articular cartilage and subchondral bone.⁵⁵ Furthermore, DNA origami, with its high programmability, allows for the creation of complex polyhedral structures that not only protect siRNAs but also enhance cellular uptake through multi-valent binding (Figure 2).^52,56,57,58 Lv et al. developed a chondrocyte-targeted DNA origami system (OCS), core loaded with Stat3 siRNA and surface functionalized with anti-CD44 aptamers. This design ensured precise recognition and enrichment within pathological chondrocytes, leading to efficient Stat3 silencing, downregulation of MMP13, and attenuation of inflammatory responses. In OA models, OCS significantly reduced ROS, promoted anabolic factor expression, and facilitated cartilage regeneration without overt side effects.³³ In summary, nucleic acid nanocarriers, characterized by their outstanding programmability and versatility in cargo loading and targeting, offer a powerful and precise platform for advancing siRNA therapies in OA.

Figure 2.

Nucleic acid and LNP delivery carriers

(A) The Tsi system effectively alleviates cellular oxidative stress and apoptosis by enhancing nucleic acid stability and cellular uptake, while inhibiting NF-κB activation and activating the NRF2/HO-1 pathway. Adapted from Liao et al.⁵⁵

(B) Schematic of the cartilage-targeting and S100A4-silencing nanoparticles for OA therapy. OA hallmarks encompass cartilage erosion, synovial inflammation, and neurovascular infiltration. By leveraging COLII-targeted nanoparticles for S100A4 silencing, key catabolic and inflammatory biomarkers are downregulated. Consequently, joint degeneration is arrested, and associated pain is effectively relieved. Adapted from Song et al.⁶⁷

(C) An overview of OA gene therapy approaches. Adapted from Grol et al.⁴⁶

(D) Schematic of the design and construction of a DNA origami-based, chondrocyte-targeted delivery system (OCS) for in vivo OA treatment. Adapted from Lv et al.³³

(E) Schematic of the mechanism of chondrocyte-derived FAP in regulating OA progression. PET/CT imaging and molecular assays confirmed FAP overexpression in OA cartilage. Intra-articular delivery of LNP@FAP siRNA effectively eliminates senescent cells and attenuates OA progression, revealing a precise mechanism by which FAP promotes OA and underscoring its potential as both a senescence biomarker and a therapeutic target. Adapted from Zhao et al.²⁴

LNPs

Lipid-based nanoparticles have established themselves as the dominant, clinically validated nucleic acid delivery platform. At their core, LNPs utilize ionizable cationic lipids, a critical component that confers distinct advantages for siRNA delivery.⁵⁹ These lipids mediate the high-efficiency encapsulation of nucleic acids via electrostatic interactions and subsequent self-assembly into stable nanoparticles. Crucially, their pH-dependent ionization facilitates endosomal membrane disruption upon cellular uptake, ensuring efficient cytosolic release of the nucleic acid payload to exert gene silencing.^60,61,62 While early LNP formulations showed promise in primates as early as 2006, their clinical utility was initially limited by low potency and narrow therapeutic indices.⁶³ The field took a quantum leap with the development of next-generation ionizable lipids, such as DLin-MC3-DMA, which enhanced siRNA delivery efficiency by approximately 100-fold, paving the way for widespread therapeutic application.^64,65 Building on this progress, Wang et al. developed an LNP-based siRNA delivery system for OA. In vitro, these LNPs efficiently transfected chondrocytes, with minimal cytotoxicity. In a rat OA model, LNPs loaded with Indian Hedgehog-siRNA (Ihh-siRNA) significantly attenuated cartilage degeneration, preserved COLII levels, and concurrently downregulated MMP13 and type X collagen, confirming their potent chondroprotective effects.⁶⁶

Further advancing targeted therapy, our research team identified the pro-fibrotic factor S100A4 as a driver of OA progression via the MAPK and NF-κB pathways. We subsequently engineered cartilage-targeted LNPs (CT-LNP-siA4) to deliver S100A4 siRNA (Figure 2). This formulation effectively silenced S100A4 expression, inhibiting downstream inflammatory and fibrotic biomarkers (IL-1β, ADAMTS, and MMP-13) to delay cartilage degradation and alleviate pain.⁶⁷ Similarly, addressing the elevated expression of FAP in OA cartilage, Zhao et al. utilized LNPs to deliver FAP-targeting siRNA. By inhibiting the FAP/NF-κB signaling axis, LNP@FAP-siRNA markedly suppressed chondrocyte senescence and reduced SASP factor production, thereby retarding cartilage wear in rat models.²⁴ Innovatively integrating therapy with diagnostics for PTOA, Zhang et al. developed MMP13-responsive theranostic micelles (ERMs@siM13). In this smart system, high MMP13 levels in pathological cartilage trigger the cleavage of the PEG shell, exposing cyclic arginine-glycine-aspartic acid (cRGD)ligands for enhanced uptake by diseased chondrocytes to achieve efficient gene silencing. Simultaneously, the unmasking of Cy5 fluorescence provides real-time feedback on disease progression.³¹ Notably, beyond their role as delivery vehicles, the phospholipid components of LNPs can act as endogenous biolubricants, significantly reducing friction at the cartilage interface and improving joint mechanical function.^68,69,70 In conclusion, LNPs offer a dual-functional strategy for OA, seamlessly integrating the excellent gene silencing capabilities of nucleic acid drugs with intrinsic joint lubrication properties.

Polymer-based nanocarriers

Polymer-based delivery systems constitute a cornerstone of siRNA therapeutics, distinguished by their outstanding synthetic versatility (Figure 3). The ability to precisely tune molecular weight, chemical composition, and architecture allows for the optimization of physicochemical properties to navigate key biological barriers: protecting siRNAs from enzymatic degradation, enhancing cellular uptake, and facilitating endosomal escape. Polymer-based delivery systems can be categorized into two major classes: natural polymers and synthetic polymers. Natural polymers derived from biological sources offer inherent advantages in biocompatibility, biodegradability, and often possess intrinsic bioactivity. Key examples widely utilized in siRNA delivery research are cyclodextrin derivatives, chitosan, and collagen.^71,72,73 Cyclodextrin derivatives, cationic polysaccharides extracted from bacterial cellulose, self-assemble into nanoparticles that effectively shield siRNAs and enhance cellular targeting.⁷⁴ Chitosan, a widely studied cationic polysaccharide, complexes with anionic siRNAs through electrostatic interactions. Its inherent mucoadhesivity and ability to transiently open tight junctions make it particularly suited for overcoming mucosal barriers, showing promise in both cancer and OA therapy.^75,76 Collagen, the primary structural component of cartilage, serves as a biomimetic carrier. While it has demonstrated excellent efficacy in siRNA delivery for anti-tumor studies,⁷⁷ its application in OA is particularly relevant. In a murine OA model, collagen-mediated delivery of ROR2-siRNA achieved targeted gene silencing and successfully delayed cartilage degeneration.⁷⁸ In contrast, synthetic polymers offer a broader design space for functionalization. Cationic polymers like PEI,⁷⁹ polyamidoamine (PAMAM) dendrimers,⁸⁰ and poly(methyl methacrylate) (PMMA) co-polymers^81,82,83 are highly efficient at condensing siRNAs and promoting endosomal escape. However, balancing their transfection efficiency with potential cytotoxicity remains a key challenge. To address this, hybrid systems are often employed. Wang et al. developed self-assembled micelles from polycaprolactone-PEG and PEI-PEG to deliver p65-siRNA, effectively alleviating inflammation and cartilage degeneration.⁷⁹ For precision OA therapy, targeting strategies are essential. Bedingfield’s team engineered chondrocyte-targeted nanoparticles by using a DMAEMA-based cationic polymer conjugated with an anti-type II collagen monoclonal antibody (mAbCII). This active targeting achieved specific enrichment in degenerative cartilage, followed by pH-responsive lysosomal escape. The system silenced MMP13 expression by ∼80% and restored proteoglycan levels, demonstrating superior efficacy compared with glucocorticoids in inhibiting subchondral bone remodeling.⁸¹ To enhance intra-articular retention and achieve stimulus-responsive release, Qiao et al. constructed a sophisticated core-shell system.⁸⁴ They coated dexamethasone (DIA)-loaded gold nanoclusters (AuNCs) with cationic poly (β-amino ester) (PBAE) to complex with siNGF and further encapsulated this into a phase-change material (lauric acid/stearic acid). This architecture improved joint retention via electrostatic and physical barriers. Upon near-infrared (NIR) irradiation, the photothermal effect of AuNCs triggered the phase change, releasing DIA for anti-inflammation and siNGF for sustained analgesia, synergistically promoting functional recovery. Addressing the complex OA microenvironment requires multi-target strategies. Cui et al. developed core-shell nanoparticles to modulate both inflammation and regeneration.⁸⁵ A calcium phosphate (CaP) core condensing siCA9 was shielded by an HA shell grafted with alendronate and kartogenin (KGN). HA provided lubrication, while alendronate stabilized the core. In macrophages, siCA9 drove M1-to-M2 repolarization; in mesenchymal stem cells (MSCs), released KGN induced chondrogenic differentiation. Combined with exogenous BMSCs, this system showed remarkable efficacy in advanced OA models. Furthermore, to prolong therapeutic residence, Bedingfield et al. encapsulated siMMP13-loaded nanoparticles within shape-defined PLGA microplates (siNP-μPLs). A single intra-articular injection maintained MMP13 silencing for over 28 days, alleviating comprehensive pathological changes in a PTOA model, thus underscoring its potential as a disease-modifying osteoarthritis drug (DMOAD).⁸² In conclusion, polymer-based delivery systems exhibit potent clinical application potential in siRNA delivery, and their biodegradability, stability, and biocompatibility render them a highly promising delivery platform for OA treatment.

Figure 3.

Polymer nanoparticle carriers

(A) Schematic of the long-acting siMMP13-μPL delivery system for PTOA therapy. Encapsulation of siNPs within PLGA microplates (μPLs) significantly extends intra-articular retention. In a mouse PTOA model, siMMP13-μPLs sustained potent MMP13 knockdown over 28 days, effectively attenuating cartilage degradation, synovial hyperplasia, and osteophyte formation. This system demonstrates the therapeutic potential of long-term gene silencing for PTOA management. Adapted from Bedingfield et al.⁸²

(B) Schematic of the design of polymeric nanoparticles (TP/siMMP-13) for targeted OA therapy through redox modulation and gene silencing. Adapted from Wu et al.⁷³ TP/siMMP-13, 2,2,6,6-Tetramethylpiperidoxyl and MMP-13-targeting siRNA.

(C) Schematic of the therapeutic efficiency of AHK-CaP/siCA9 nanoparticles and MSCs for OA treatment. AHK-CaP/siCA9 nanoparticles were synthesized using an AHK polymer shell to condense a CaP/siRNA core via coordination interactions. In the joint, these nanoparticles exhibited a dual function: siCA9 release induced M1-to-M2 macrophage repolarization to reverse the inflammatory microenvironment, while KGN release promoted chondrogenesis of MSCs for cartilage regeneration. The synergistic therapeutic efficacy of these nanoparticles combined with exogenous BMSCs was further validated in an MIA-induced OA model. Adapted from Cui et al.⁸⁵ MIA, Monoiodoacetate.

(D) Schematic of Dex and siRNA co-encapsulation in micelles and the intracellular process of micelles. Adapted from Wang et al.⁷⁹

Peptide-based nanomaterials

Cationic peptides represent a distinct class of delivery vectors defined by their facile synthesis, precise sequence control, and inherent biological functions. Their ability to be engineered with specific amino acid sequences allows for the integration of multiple functionalities, including nucleic acid condensation, cellular targeting, and endosomal disruption.⁸⁶ A typical example is the p5RHH peptide, derived from melittin. It self-assembles with siRNAs into stable nanoparticles and possesses potent endosomal escape capabilities, without the cytotoxicity associated with parent melittin. Its well-defined structure-activity relationship makes it a highly translatable platform.^{87,88,89,90,91} In OA applications, Yan et al. utilized p5RHH to deliver siRNA targeting NF-κB, effectively dampening early-stage inflammatory responses and cartilage damage.⁸⁷ Similarly, Duan’s team employed p5RHH-mediated siRNA delivery to silence periostin in a PTOA model, significantly improving cartilage integrity and reducing osteophyte formation.⁸⁸ Beyond serving as primary carriers, peptides are frequently used as functional ligands to modify other lipid or polymer nanocarriers, endowing them with enhanced cell-specific targeting capabilities to maximize therapeutic effect.^92,93,94 Therefore, optimization of peptide sequences can not only enhance the targeting capability of siRNA delivery but also enable it to exert more precise and efficient biological effects in OA treatment.

Inorganic and hydrogel-based hybrid systems

Inorganic nanoparticles and hydrogels offer unique physicochemical properties that complement traditional organic carriers, creating powerful hybrid platforms for synergistic OA therapy.

Inorganic nanocarriers

Materials like CaP, mesoporous silica nanoparticles (MSNs), and gold nanoparticles (AuNPs) provide robust scaffolds for siRNA loading via adsorption, encapsulation, or coordination^28,95,96 Their rigid structures effectively protect siRNAs, while specific material properties enable unique functionalities (Figure 4). For instance, CaP nanoparticles rapidly dissolve in acidic lysosomes, triggering osmotic swelling and highly efficient endosomal escape.³⁴ To achieve precision therapy, these nanocarriers can be optimized via surface functionalization. For example, conjugation with specific targeting peptides or aptamers enables them to actively recognize and bind to specific receptors on the surface of target cells, thereby enhancing accumulation at lesion sites and reducing off-target effects.⁹⁷ Inorganic carriers are particularly suited for multi-modal synergistic therapy. He et al. designed a Zn (II)-dipicolylamine coordination complex to self-assemble metformin and p65 siRNA into a hybrid nanosystem. The Dpa-Zn complex enabled stable siRNA loading and cartilage imaging, while metformin and si-p65 synergistically activated autophagy and inhibited NF-κB signaling, respectively.²⁸ Similarly, Zhang et al. utilized a pH-responsive metal-organic framework (MOF), MIL-101-NH₂, to co-deliver curcumin and HIF-2α siRNA. The MOF protected the agents and degraded in the acidic OA microenvironment, releasing curcumin for anti-inflammation and siRNA for gene silencing (Figure 4).²⁹ Wang et al. engineered a multi-functional, star-shaped AuNP platform for si-DNA Damage-Inducible Transcript 3 (si-DDIT3) delivery and photothermal therapy. This system simultaneously inhibited ferroptosis via siDDIT3 and leveraged mild hyperthermia to suppress angiogenesis and subchondral bone hyperplasia, while also providing bioimaging capabilities.⁹⁸

Figure 4.

Inorganic, and hydrogel delivery carriers

(A) Schematic of star-like AuNPs for DDIT3-targeted gene therapy and theranostics of OA. Star-like AuNPs were synthesized to deliver siDDIT3, a ferroptosis-related biomarker identified via bioinformatics. This nanocomposite enables NIR/SERS dual-modal imaging and effectively suppresses ferroptosis while restoring cartilage matrix homeostasis. In OA mouse models, the treatment significantly alleviates disease progression by reducing joint space narrowing, subchondral bone sclerosis, and pathological neovascularization. Adapted from Wang et al.⁹⁸ under CC BY 4.0 license.

(B) Schematic of the pH-responsive MIL-101-NH2@CCM-siRNA system for synergistic OA therapy. A hybrid nanocomposite, MIL-101-NH2@CCM-siRNA, was developed for the localized and sustained co-delivery of CCM and siHIF-2α. This pH-sensitive system ensures controlled drug release and protects siRNA from degradation, synergistically inhibiting hypoxia-induced chondral dysfunction. Intra-articular injection of this biocompatible platform effectively alleviates OA progression. Adapted from Zhang et al.,²⁹ under CC BY 4.0 license.

(C) Schematic of the ROS-responsive si-Fe-HPP nanocomposite hydrogel for synergistic OA therapy. A nanocomposite hydrogel (si-Fe-HPP) was developed by embedding siMMP-13-loaded Fe₃O₄ nanoparticles into an H₂O₂-scavenging HPP hydrogel. This platform synergizes MMP-13 silencing with ROS neutralization to suppress inflammation and promote cartilage repair. Notably, the hydrogel exhibits intelligent H₂O₂-responsive behavior, accelerating si-FeNP release under high oxidative stress while prolonging retention as inflammation subsides, thereby providing a smart, controlled delivery system for OA management. Adapted from Wang et al.,³⁰ under CC BY 4.P license.

(D) Schematic of the synthesis of NanoMet/siRNA and the autophagy regulation and anti-apoptotic effects on chondrocytes by anchoring to arthritic cartilage, thereby alleviating the inflammatory process. Adapted from He et al.²⁸

Hydrogel-based macroscopic depots

Hydrogels serve as excellent macroscopic drug depots for sustained intra-articular delivery. By encapsulating siRNA-loaded nanocomplexes within a hydrogel matrix, researchers can achieve prolonged retention and stimulus-responsive release at the joint. Wang et al. constructed an ROS-responsive injectable hydrogel (si-Fe-HPP) loaded with Fe₃O₄-PEI-PEG/siMMP-13 nanocomplexes (Figure 4). The hydrogel matrix, cross-linked via ROS-sensitive borate bonds, actively scavenged H₂O₂ while providing sustained release of the gene-silencing nanoparticles, thus targeting both inflammation and matrix degradation.³⁰ Addressing the infrapatellar fat pad’s role in OA, Dai et al. developed an injectable RGD-functionalized nanogel for sustained delivery of Cd61 siRNA targeting integrin β3. This system demonstrated excellent targeting and effectively halted cartilage degeneration.⁹⁹ To enhance cartilage penetration, Chen et al. utilized microfluidics to encapsulate pH-responsive dendrimer/siHIF-2α complexes into HA hydrogel microspheres. These microspheres responsively released smaller nanoparticles in the acidic joint environment, improving deep tissue penetration and anchoring.¹⁰⁰ Furthermore, Yang et al. reported a thermosensitive pNIPAAM/LDH composite hydrogel. Layered double hydroxides (LDHs) efficiently protected siRNAs, while the pNIPAAM network formed a stable depot at body temperature, showing superior performance over conventional alginate or fibrin hydrogels.¹⁰¹ In conclusion, hydrogel-based siRNA delivery systems have paved an innovative avenue for OA treatment by virtue of their distinctive smart-responsive properties and synergistic effects. In particular, hydrogel-based siRNA delivery systems not only overcome the limitations of conventional delivery modalities, enabling sustained release and efficient transfection of the siRNA at lesion sites, but also significantly enhance therapeutic efficacy through the synergy between gene silencing and microenvironmental modulation. With the continuous advancement of materials science and nanotechnology, hydrogel-based siRNA delivery platforms are expected to evolve into a pivotal tool for the precision OA treatment, thereby unlocking new possibilities for clinical translation.

Exosome vectors

Exosomes, small endogenous extracellular vesicles (EVs), have emerged as highly promising next-generation delivery vehicles due to their inherent biocompatibility, low immunogenicity, and natural ability to traverse biological barriers.^102,103 Their lipid bilayer structure protects nucleic acid cargoes, and their surface protein profiles can be engineered for capability. Using recipient-derived exosomes further minimizes immunological risks.¹⁰⁴ Strategies to enhance exosome functionality for OA therapy involve surface engineering. Kim et al. developed “IGF-si-EVs” by modifying siMMP13-loaded umbilical cord MSC-EVs with positively charged IGF-1 (Figure 5). This modification improved joint retention through electrostatic interactions with the cartilage matrix, while IGF-1 promoted regeneration and siMMP13 inhibited catabolism.¹⁰⁵ To specifically target senescent chondrocytes, Feng et al. engineered MSC-sEVs loaded with siMDM2 and modified with a cartilage-targeting peptide (WPD). These “WPD-sEVs-siMDM2” precisely eliminated senescent cells via the MDM2-p53 pathway, restoring matrix homeostasis in aging and PTOA models.¹⁰⁶ Furthermore, Zhang et al. engineered exosomes with cartilage-affinity peptides to develop a nanosystem for the targeted delivery of siMMP13. This nanosystem could efficiently deliver siRNAs to chondrocytes and specifically silence the MMP13, which effectively attenuated cartilage degeneration in OA models, thereby providing a precise strategy for targeted gene therapy in OA.³² In addition, Feng et al. developed cartilage-targeting exosomes loaded with STING siRNA. This approach simultaneously provided joint lubrication and inhibited the cGAS-STING pathway to delay chondrocyte senescence, offering a synergistic strategy.¹⁰⁷

Figure 5.

Exosome delivery vehicle

(A) Strategies for drug loading into exosomes. Adapted from Lai et al.¹⁰²

(B) Schematic of engineered WPD-sEVs^siMDM2 for targeted senescence elimination and OA therapy. WPD-sEVs^siMDM2 were developed to enhance cartilage penetration and specifically eliminate senescent chondrocytes. Intra-articular injection of this platform effectively reduced the senescent cell population and suppressed SASP production, thereby retarding cartilage degeneration in OA mice. Adapted from Feng et al.¹⁰⁶

(C) Schematic of biolubricating hydrogel delivering CAP-EXO-si-STING for OA therapy. A biolubricating hydrogel was formed in situ via intra-articular click chemistry to anchor CAP-EXO-si-STING onto cartilage. Following internalization by chondrocytes, the exosome-mediated si-STING delivery suppresses the STING signaling pathway triggered by excessive mechanical loading. This targeted intervention reverses the aging microenvironment and retards OA progression. Adapted from Feng et al.¹⁰⁷

(D) Schematic of IGF-si-EV fabrication and application to OA therapy. Adapted from Kim et al.,¹⁰⁵ under CC BY 4.0 license.

Engineering paradigms for optimized siRNA delivery in OA

The pathogenesis of OA is orchestrated by a multi-faceted cascade involving the dysregulation of inflammatory cytokines, chondrocyte senescence, and the catabolic breakdown of the ECM. However, therapeutic intervention via siRNAs faces a formidable biological barrier: the intra-articular microenvironment. This niche is defined by rapid synovial fluid turnover, immune cell sequestration, and a localized acidic shift, all of which jeopardize the integrity of exogenous nucleic acids.⁴⁶ To circumvent these barriers, recent research has pivoted toward sophisticated engineering paradigms. While siRNAs offer exquisite precision in gene silencing, their intrinsic physicochemical architecture, characterized by polyanionic charges and a large molecular weight, imposes a “delivery ceiling.” Unmodified siRNAs suffer from rapid renal clearance, nuclease susceptibility, and negligible membrane permeability. Even when encapsulated in high-performance vectors, issues such as endosomal entrapment and low cytosolic bioavailability remain critical bottlenecks.¹⁰⁸ To surmount these barriers, recent research has pivoted toward advanced engineering paradigms. Current optimization strategies employ a dual-track approach: (1) molecular tailoring of the siRNA backbone to enhance its inherent robustness and reduce immunogenicity, and (2) the rational design of functional vectors to ensure spatiotemporal precision in delivery. This section discusses these advancements, spanning chemical modifications, targeted delivery strategies, and multi-modal combination therapies.^109,110

Chemical modification: Bolstering stability and mitigating immunogenicity

Unmodified siRNAs are highly susceptible to rapid degradation by ubiquitous nucleases in serum and tissues, with half-life ranging from minutes to an hour, which severely limits their accumulation at target sites.^111,112 Moreover, exogenous RNAs can act as a PAMP, triggering innate immune responses via TLRs.¹¹³ To enhance the stability and therapeutic index of siRNAs, various chemical modifications have been developed.

Modification of the ribose and phosphate backbone

Modifications to the ribose sugar ring and phosphate backbone are pivotal for improving siRNA druggability. Studies have demonstrated that 2′-sugar modifications, such as 2′-O-methyl (2′-O-Me), 2′-fluoro (2′-F), and 2′-methoxyethyl (2′-MOE), significantly enhance duplex stability and binding affinity to the RISC complex, while simultaneously dampening innate immune responses.⁴⁰ These modifications are integral to approved siRNA therapies like Onpattro (patisiran). Replacing the 2′-hydroxyl group with these substituents alters the siRNA’s recognition profile by innate immune sensors, facilitating immune evasion without compromising biological activity.¹¹⁴ The combined use of DNA analogs, 2′-F RNA, and locked nucleic acids (LNAs) has shown synergistic potential in enhancing gene silencing efficiency while minimizing immune stimulation. Base modifications, such as the incorporation of 5-methylcytidine and 5-methyluridine, further alleviate immune responses.¹¹⁵ Additionally, sequence optimization to avoid uridine- and guanosine-rich motifs helps attenuate immune activation.¹¹⁶ To counter nuclease degradation, modifications like replacing uridine with 2,4-difluorotoluyl ribonucleoside enhance resistance to serum nucleases.^117,118 Strategic placement of 2′-O-Me or 2′-F modifications on both sense and antisense strands can robustly protect against endonuclease attack, thereby enhancing therapeutic effect.¹¹⁹ At the backbone level, replacing phosphodiester bonds with phosphorothioate linkages or morpholino oligomers significantly prolongs the in vivo half-life of siRNA duplexes.¹²⁰

Terminal modifications and conjugates

Conjugating siRNAs with functional molecules can dramatically improve their pharmacokinetics and biodistribution.

Cholesterol/polyethylene glycol conjugation

Cholesterol conjugation enables systemic delivery by promoting binding to serum albumin/lipoproteins, extending the circulatory half-life and enhancing cellular uptake. In murine OA models, intra-articular injection of cholesterol-conjugated MMP13-siRNA resulted in a 5-fold increase in joint accumulation compared with naked siRNA.^121,122 While cholesterol improves cellular internalization, its primary benefit lies in enhancing pharmacokinetic profiles.¹²³ Similarly, conjugation with PEG polymers significantly prolongs the circulation time and improves the overall pharmacokinetic profile of siRNAs.¹²⁴

GalNAc conjugation

Although primarily designed for liver targeting via the asialoglycoprotein receptor (ASGPR), N-acetylgalactosamine (GalNAc) conjugation technology exemplifies the power of receptor-mediated delivery. Approved drugs like givosiran utilize trivalent GalNAc ligands to achieve highly efficient hepatocyte uptake, enhancing RNAi activity by 5-fold and significantly lowering the required dosage.^125,126 While ASGPR is liver specific, the success of GalNAc conjugation has inspired the search for analogous receptor-ligand pairs for targeted delivery to joint tissues.

Targeted delivery: Achieving spatiotemporal precision

Chondro-targeted modification

To maximize therapeutic efficacy and minimize off-target effects, delivering siRNAs specifically to chondrocytes within the cartilage ECM is crucial. Advances in cartilage-targeted delivery systems have focused on functionalizing vectors with affinity ligands.

Aptamer-based targeting

Lv et al. constructed a DNA origami nanostructure loaded with Stat3 siRNA and functionalized with anti-CD44 aptamers. This design enabled precise recognition of CD44-overexpressing chondrocytes, leading to effective MMP13 silencing and alleviation of OA progression.³³

Antibody-mediated targeting

Bedingfield’s team developed cationic polymer nanoparticles modified with mAbCII. This approach achieved cartilage-specific accumulation and pH-responsive release, reducing MMP13 mRNA levels by ∼80% and improving cartilage matrix metabolism.⁸¹

Peptide- and lipid-based strategies

Song et al. developed cartilage-targeted LNPs to deliver siRNA targeting the profibrotic factor S100A4, significantly downregulating inflammatory markers and delaying degeneration.⁶⁷ Similarly, exosome systems modified with cartilage-affinity peptides have been employed for efficient siMMP13 delivery, effectively attenuating disease progression in OA models.³² These strategies collectively highlight the potential of ligand-functionalized vectors to enhance the site-specific bioavailability of siRNAs in cartilage.

Response to the inflammatory microenvironment

The unique physicochemical cues of the OA microenvironment, such as elevated enzyme levels (e.g., MMPs) and localized acidosis, can be exploited to trigger site-specific drug release.

Enzyme-responsive systems

Zhou et al. developed MMP13-responsive theranostic micelles (ERMs@siM13) that shed a protective PEG layer upon cleavage by MMPs, exposing targeting ligands and releasing siRNAs specifically at sites of high disease activity. This system enables simultaneous gene silencing and fluorescence-based disease monitoring.³¹

ROS-/pH-responsive systems

An anti-oxidant hydrogel system (si-Fe-HPP) integrated with siRNA nanocomplexes synergistically alleviates inflammation and promotes repair by scavenging H₂O₂ and silencing MMP-13.³⁰ Leveraging the acidic microenvironment, researchers have developed pH-responsive dendritic polymer vectors to enhance cartilage penetration and endosomal escape, effectively modulating the HIF-2α/MMP-13 pathway.¹⁰⁰ Furthermore, Zhang et al. utilized a pH-responsive MOF (MIL-101-NH₂) for co-delivery of curcumin and siHIF-2α, achieving synchronized release under acidic conditions for synergistic anti-inflammatory and gene-silencing effects.²⁹ These intelligent, stimulus-responsive systems represent a new frontier in OA therapy.

Gene-physical therapy synergy

Polymer-coated AuNCs enabled photothermal-controlled delivery of dexamethasone and siNGF, achieving a synergistic anti-inflammatory and analgesic effect.⁸⁴ Similarly, star-shaped AuNPs integrated siDDIT3-mediated gene silencing with mild photothermal therapy to effectively inhibit ferroptosis and preserve the cartilage matrix.⁹⁸

Mechanistic synergy

He et al. constructed a self-assembled nanosystem co-encapsulating metformin and p65 siRNA. This combinatorial approach exerted synergistic effects by simultaneously activating autophagy and inhibiting the inflammatory NF-κB pathway (via siRNA), reducing chondrocyte apoptosis.²⁸ The co-delivery system based on the MIL-101-NH₂ MOF synchronously releases curcumin and siHIF-2α in the acidic microenvironment, achieving dual regulation of inflammation and target gene expression.²⁹ Furthermore, engineered exosomes modified with IGF-1 (IGF-si-EV) enhanced cartilage targeting and retention, synergistically suppressing inflammation while promoting regeneration.¹⁰⁵ Collectively, these studies demonstrate that combinatorial delivery systems, rationally designed to target multiple pathways and cell types and to leverage diverse therapeutic modalities, provide a more robust and comprehensive strategy for the clinical management of OA.

Challenges and future horizons

Translating siRNA therapeutics from bench to bedside faces formidable hurdles. Realizing its clinical potential demands paradigm shifts in disease stratification, technological convergence, and translational modeling. Future research must pivot on three axes—precision, practicality, and scalability—to drive the clinical implementation of siRNA technology for OA treatment.

Precision stratification and personalized therapy

Moving beyond a monolithic view of OA toward molecular-based precision stratification is a prerequisite for effective intervention. Distinct pathophysiological endotypes require tailored siRNA strategies. For instance, metabolically driven OA might respond optimally to CH25H targeting,¹²⁷ whereas phenotypes characterized by chondrocyte ferroptosis may necessitate Nupr1 intervention.¹²⁸ The identification of robust biomarkers and corresponding companion diagnostics for these subtypes is crucial for clinical decision-making.

Technological convergence and smart vectors

Technological innovation is the linchpin for enhancing therapeutic indices. Next-generation vector design must, therefore, prioritize “smart” multi-functionality. This can be achieved by engineering vectors with stimulus-responsive moieties that are activated by intra-articular cues such as ROS, specific enzyme activities, or acidic pH, thereby enabling spatiotemporally controlled siRNA release at lesion sites.^29,30,129 Concurrently, multi-stage targeting architectures designed to sequentially negotiate physiological barriers and achieve intracellular specific accumulation within chondrocytes represent the frontier of delivery platform development.

Optimizing translational pathways

Clinical translation requires a pragmatic, stepwise evolution. Given safety considerations and the chronic nature of OA, initial clinical inroads may target localized, end-stage disease, progressively expanding toward disease-modifying therapies for early-to-moderate stages. While technically demanding, developing non-invasive administration routes (e.g., oral or transdermal) remains a highly desirable goal to improve patient compliance.

Industrialization and scalability

Commercial viability hinges on cost effectiveness and robust manufacturing. Streamlining siRNA synthesis/purification and developing scalable carrier manufacturing processes are essential. Given the high prevalence of OA, developing “universal” siRNA cocktails, rather than fully personalized sequences, may offer a balance between R&D costs and therapeutic efficacy. Furthermore, engineering long-acting depot formulations to reduce administration frequency will indirectly lower overall treatment burdens.

Outlook

Over the next decade, the interdisciplinary convergence of materials science, molecular biology, and clinical medicine is poised to catalyze substantial breakthroughs in siRNA-based OA therapy. Initial clinical applications will likely focus on well-defined single targets as adjuncts or alternatives to conventional care. As technologies mature and evidence accumulates, multi-gene modulation systems and combinatorial strategies will gradually shift the treatment paradigm from symptomatic relief to authentic disease modification. While siRNA therapy is at a critical translational juncture, overcoming current limitations in delivery efficiency, long-term biosafety, and cost effectiveness requires relentless collaborative innovation. Ultimately, the synergy of personalized precision medicine, intelligent vector materials, and optimized clinical strategies will shape the future of siRNA therapeutics in OA.

Conclusion

siRNA-based gene silencing has garnered significant attention for its high targeting specificity toward the molecular pathogenic mechanisms of OA. Advanced delivery platforms can now effectively overcome multiple intra-articular physiological barriers, enabling spatiotemporally regulated “smart” local delivery and optimized pharmacokinetics. However, the transition from preclinical success to clinical application remains hindered by critical bottlenecks: the long-term safety of repeated administration has yet to be fully established, and scalable manufacturing processes still require breakthroughs. Moreover, the marked molecular heterogeneity of OA necessitates a shift toward precision treatment strategies, where target selection is guided by patient stratification. Future progress lies in the development of multi-modal synergistic therapies, such as combining siRNAs with biologics or physical therapeutic factors via intelligent nanocarriers. Ultimately, realizing gene therapy as a mainstream option for OA management will depend on sustained interdisciplinary collaboration across materials science, biology, and clinical medicine.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (82572742) and Key Clinical Projects of Peking University Third Hospital (BYSY2022065).

Author contributions

Conceptualization, S.S., W.L., N.G., and S.W.; writing – original draft, S.S., W.L., M.L., and J.X.; visualization, S.S., W.L., M.L., and J.X.; writing – review & editing, X.Y., X.L., B.X., and Y.X.; supervision and project administration, Y.X.

Declaration of interests

The authors declare no conflict of interest.