In recent years, natural polymers, synthetic polymers, and composite materials have been extensively employed in tissue engineering research for tendon-bone healing (56). Multilayer scaffolds can mimic the complex structure of the tendon-bone interface, providing an optimal microenvironment for cellular infiltration and growth, thereby enhancing biocompatibility and promoting tissue integration (57, 58). Among these, polycaprolactone (PCL) has emerged as an ideal scaffold material due to its excellent biocompatibility, controllable degradability, and mechanical properties matching those of tendon and bone tissues (59, 60). Three-dimensionally printed PCL scaffolds loaded with bFGF and BMSCs (PCLMF) modulate macrophage gene expression by downregulating pro-inflammatory genes and upregulating anti-inflammatory genes, thereby promoting M1-to-M2 polarisation to suppress inflammation and facilitate repair. This effect is closely associated with the inhibition of the MAPK signalling pathway. BMSCs further promote M2 polarisation by modulating the IL-10/STAT3 pathway, synergising with the immunomodulatory effects of PCLMF to significantly enhance tendon-bone interface repair capacity.

As illustrated in Figure 3, the Wnt3a-modified bionic scaffold remodels the immune microenvironment through sustained release of Wnt3a protein, enhancing M2 polarisation to promote tendon regeneration. The mechanisms comprise: (1) Activation of the classical Wnt/β-catenin signalling pathway: As illustrated in Figure 2, Wnt3a acts as a key ligand that specifically binds to the LRP5/6 receptor, simultaneously interacting with the seven-transmembrane receptor Frizzled (Fzd) and the single-transmembrane co-receptor LRP5/6 on the cell membrane. Formation of this ternary complex (Wnt-Fzd-LRP6) leads to phosphorylation of the LRP6 intracellular domain, thereby recruiting the cytoplasmic scaffolding protein Dishevelled (Dvl). In the absence of Wnt3a, cytoplasmic β-catenin is captured and degraded by a ‘destabilising complex’. This complex comprises Axin, APC, GSK-3β, and CK1. GSK-3β phosphorylates β-catenin, leading to its ubiquitination and degradation. Upon recruitment to the membrane, Dvl inhibits GSK-3β activity, causing the disruptor complex to disassemble. This prevents β-catenin phosphorylation and degradation, thereby activating its transcriptional activity to regulate monocyte differentiation towards the M2 phenotype. Stabilised β-catenin accumulates in the cytoplasm and translocates to the nucleus. Within the nucleus, β-catenin displaces transcriptional repressors, binds to transcription factors of the TCF/LEF family, and initiates downstream gene transcription. This includes key transcription factors promoting M2 macrophage polarisation, such as Irf4 and Jmjd3; (2) Polarisation synergistic amplification effect: IL-4 in the microenvironment binds to the IL-4Rα receptor on the macrophage surface, activating JAK1/JAK3 kinases, which in turn lead to the phosphorylation and dimerisation of STAT6. Nuclearised β-catenin can act as a co-activator for STAT6. β-catenin and STAT6 proteins form complexes within the promoter regions of M2 marker genes. This interaction may recruit histone acetyltransferases, altering chromatin structure to enhance transcription accessibility for M2-associated genes. Under IL-4 microenvironmental stimulation, Wnt3a reinforces the M2 polarisation state by upregulating Arg-1 and CD206 expression, thereby establishing a pro-regenerative molecular phenotype; (3) Paracrine signalling: M2 macrophages secrete factors such as VEGF and BMP-2, which stimulate osteogenic differentiation of tendon stem cells through Wnt–Notch signal interaction, thereby promoting fibrocartilage reconstruction. This scaffold achieves precise regulation of inflammatory and regenerative signals (61).

Milk fat globule-derived epidermal growth factor-8 (MFG-E8) and its recombinant protein (rMFG-E8) modulate macrophage phagocytic function and polarisation status via the efferocytosis pathway, thereby promoting tendon healing. The mechanisms involved include: (1) enhancing macrophage recognition of the ‘eat me’ signal from apoptotic cells, clearing apoptotic cells and promoting M2 polarisation; (2) Inhibiting TLR4 signalling via binding to integrin αvβ3, partially reversing iNOS expression in M1 macrophages; (3) Enhancing CD206 expression and reducing iNOS expression at the early interface stage, thereby mitigating inflammation and promoting osseointegration. This study reveals MFG-E8’s bidirectional signalling for comprehensive anti-inflammatory and pro-regenerative regulation, providing novel immunotherapeutic targets for tendon healing (62, 63).

The chitosan scaffold loaded within the joint-mimetic adhesive layer (JTA) promotes gradient healing of traumatic brain injury (TBI) by remodelling the immune microenvironment through modulating macrophage (Mφ) phenotype. The deacetylation activation effect of positive charges on the chitosan surface drives M1-to-M2 polarisation, upregulating expression of CD206 and factors including IL-10 and TGF-β. M2 macrophages promote angiogenesis and collagen remodelling through a VEGF–BMP-2 signalling cascade, enabling interface regeneration driven by the ‘macrophage–stem cell coupling’. This strategy reveals a novel direction for designing gradient-integrated biomaterials based on immune homeostasis regulation (64).

Decellularised extracellular matrix (aECM) scaffolds with ordered microchannel structures regulate regeneration through biophysical signals, primarily via three mechanisms: (1) Topography-dependent polarisation: The RhoA/ROCK pathway mediates morphological stretching of macrophages, triggering YAP/TAZ nuclear translocation and upregulating Arg-1/IL-10 expression; (2) Mechanical transduction regulation: 3% cyclic stretching promotes Ca²⁺ influx via the PIEZO1 channel, inhibiting NFAT5-driven IL-1β/iNOS expression; (3) Paracrine network remodelling: polarised M2 macrophages secrete PDGF-AA/TGF-β1 to recruit tendon progenitor cells, facilitating ordered ECM deposition (65).

Bismuth-magnesium hydroxyapatite particles (BMnPs) can specifically induce Mφ to polarise towards the M2 phenotype, activating the transcription factor cMaf to upregulate IL-10 expression, thereby forming an anti-inflammatory–pro-osteogenic synergistic effect via both autocrine and paracrine mechanisms. PGE2 released by MSCs binds to macrophage EP2/EP4 receptors, forming a positive feedback loop that further amplifies IL-10 signalling. Compared to conventional micron-scale hydroxyapatite, BMnPs scaffolds not only reduce inflammatory responses by promoting M2 polarisation but also stimulate the release of pro-angiogenic factors, accelerating bone regeneration and mineralisation through both mechanical and molecular mechanisms. This mechanism reveals the pivotal role of immunomodulatory materials in osseointegration (66, 67).

Biodegradable magnesium-based materials promote tendon-bone interface healing through dual mechanisms of immunomodulation and tissue regeneration by releasing Mg²⁺. Research indicates that Mg²⁺ significantly enhances levels of osteogenesis-related factors in periosteal-derived stem cells (10). At the immunological level, Mg²⁺ activates the STAT6 signalling pathway to drive Mφ towards M2 phenotype polarisation, upregulating CD206/Arg expression while suppressing TLR/NF-κB-mediated TNF-α and IL-6 secretion. This shapes an anti-inflammatory microenvironment and enhances alkaline phosphatase activity in mesenchymal stem cells. At the regenerative level, Mg²⁺ promotes BMSC adhesion and osteogenic differentiation via the α5β1 integrin/FAK pathway. Furthermore, it activates the Hif-1α/Sox9 cascade through TRPM7 channels, inducing Acan and Col II synthesis to achieve chondro-like tissue reconstruction. This synergistic ‘immune amplification-signal decoding’ network increases new bone formation by 58%, providing a multi-target intervention strategy for TBI repair (68, 69). Furthermore, porous magnesium or surface-etched structures can similarly induce Mφ to polarise towards the M2 phenotype, accelerating tendon-bone healing (70–74). Magnesium-based materials are biodegradable and are completely metabolised and excreted, gradually disappearing as tissue regenerates, thereby further promoting repair (75–79). Demonstrating the potential of magnesium-based materials and magnesium-pretreated periosteum in tendon healing research (80).

High-purity magnesium screws (HP Mg) exert precise immune regulation via the Mg²⁺ released through the AKT1/PI3K pathway: (1) Mg²⁺ inhibits AKT2 phosphorylation and enhances AKT1 activation via the TRPM7 channel, inducing MRC1/CD163 expression mediated by STAT6; (2) M2 macrophages secrete PDGF-BB/TGF-β3 to activate the SOX9/RUNX2 pathway in hBMSCs, promoting synthesis of type II collagen and aggrecan; (3) At 12 weeks post-surgery, interfacial tensile strength reached (32.5 ± 3.1) MPa, comparable to autologous grafts, achieving conversion from fibrovascular scar to functional fibrocartilage (81, 82).

For the promotion of osseointegration, micro-surface topography is essential when the bone-immune tissue microenvironment is suitably adjusted (83–86). Additionally, when cells are stretched in particular ways, microsurface topography can cause M2 macrophage polarisation (87–89). The construction of TiO₂ honeycomb topologies spanning a gradient scale (90–5000 nm) reveals a mechanism whereby nanomorphology modulates Mφ polarisation through mechanical signal transduction. Nanostructures HC-90, fabricated via thin-film transfer, significantly enhanced Mφ pseudopod formation and promoted spatial morphoplasticity, thereby inducing M2 polarisation and secretion of IL-4, IL-10, and BMP-2 to establish a pro-regenerative immune microenvironment. Mechanistic studies indicate the RhoA/ROCK pathway plays a central role in nanotopology-mediated immune regulation: HC-90 upregulates cytoskeletal genes, including RhoA and Arp2/3, while suppressing TNF-α and NF-κB signalling. Concurrently, elevated p-MLC/p-MYPT1 phosphorylation enhances actin contractility, promoting footplate formation. Furthermore, PPARγ signalling activation and MAPK inhibition jointly drive M1-to-M2 macrophage conversion. This immunoregulatory effect significantly enhances mesenchymal stem cell alkaline phosphatase activity and mineralization capacity in in vitro co-culture models, while in vivo experiments confirm its ability to increase new bone volume fraction and bone-implant contact rate. This study proposes an integrated mechanism of ‘nanoscale confinement-mechanical transduction-immune regulation’, providing theoretical foundations and design principles for surface topography-engineered immune-modulatory bone implant materials (74).

Reactive oxygen species (ROS) encompass superoxide anion (O₂^-), hydroxyl radical (HO₂), peroxyacetyl radical (RO₂), alkoxy radicals, and non-radical compounds readily converted into radicals such as hydrogen peroxide (H₂O₂) and hypochlorous acid (HOCl), which may impede tendon-to-bone healing (93). As normal by-products of oxygen metabolism, ROS can function as signalling molecules to rapidly respond to stimuli during acute inflammation, whilst also potentially triggering oxidative stress (94, 95). Research indicates that reactive oxygen species (ROS) promote M2 polarisation via the ROS/PI3K/Akt pathway or H₂O₂-mediated mechanisms, whilst antioxidants suppressing endogenous ROS reduce M2 marker expression. Part of this effect is mediated through the inactivation of Stat3 during IL-4-induced M2 polarisation (96). Overall, high ROS environments tend to induce M1 polarisation, whereas moderate or low ROS levels favour M2 polarisation, suggesting that ROS exerts a bidirectional regulatory role in immune modulation and tendon repair (94).

ROS has been demonstrated to be one of the most critical factors inducing inflammatory states, capable of rapidly differentiating macrophages into the M1 phenotype, recruiting substantial production of inflammatory cytokines, and triggering neuropathic responses at injury sites, thereby causing persistent pain. The scavenging of ROS remains a prominent research focus, with numerous ROS-targeting therapeutics such as Prussian blue nanozymes now widely employed. However, research into ROS should not be confined solely to its role as an inflammatory trigger; it must also be recognised as an integral component of the inflammatory microenvironment. Regulation of the immune microenvironment should therefore focus on holistic environmental control. Providing optimal ‘soil conditions’ for ‘seed cells’ such as BMSCs and HUVECs during tendon healing is essential to promote their healthy proliferation, thereby achieving near-complete, highly efficient repair at the tendon interface.

In orthopaedic surgery, natural or synthetic biomaterial patches are widely employed to enhance the strength of soft tissue repair (97–100). Surface topography can induce M0 macrophages to polarise towards the M2 phenotype while suppressing excessive pro-inflammatory cytokine secretion, whereas randomized fibrillar structures combined with dynamic mechanical stretching promote M1 polarisation. Mechanically stimulated Mφ induce fibroblast pro-inflammatory responses via NF-κB-p65 nuclear translocation; however, pre-mechanical stimulation confers ‘mechanical protection’ to fibroblasts through the Rho/ROCK pathway, attenuating IL-1β-induced inflammatory cascades. Dynamic mechanical loading interacts with fibroblasts to significantly reduce the pro-inflammatory CCR7⁺ subpopulation while activating the MRC1⁺ M2 phenotype, suppressing NF-κB overexpression and establishing an anti-inflammatory microenvironment (101).

Appropriate mechanical loading remodels the immune microenvironment through multi-level signalling: it inhibits the TLR4/NF-κB pathway in macrophages while activating the IL-4/JAK1/STAT6-PPARγ repair axis, thereby promoting M2 polarisation and enhancing TGF-β1 secretion. Concentration-dependent activation of the SMAD2/3 pathway in mesenchymal stem cells (MSCs) by TGF-β1 released from M2 macrophages induces fibrocartilage cell differentiation. STAT6-PPARγ and YAP/TAZ signalling synergistically regulate TGF-β1 biosynthesis and spatial gradients; excessive mechanical loading reverses repair via NLRP3 inflammasome activation (102, 103). In vivo experiments demonstrate that specific depletion of macrophages completely abolishes the wound-healing-promoting effect of mechanical stimulation, resulting in reduced fibrocartilage layer thickness. Furthermore, early postoperative immobilization diminishes M1 macrophage accumulation, alleviates acute inflammation, and enhances the continuity of fibrous scar tissue and collagen (104). Overall, the biomaterial’s topology and mechanical loading precisely modulate the interaction between monocytes/macrophages and fibroblasts, thereby optimizing healing at the tendon-bone interface.

Mesoporous bioactive glass can modulate the immune microenvironment by incorporating bioactive components (105, 106). During degradation, Li-MSN scaffolds continuously release lithium and silicon ions. Lithium activates the Wnt/β-catenin pathway to promote osteogenic differentiation at the tendon-bone interface, while synergistically inducing the transition of macrophages from pro-inflammatory M1 to anti-inflammatory M2 cells, thereby suppressing local inflammation. M2 macrophages accelerate regeneration by secreting anti-inflammatory and angiogenic factors. Furthermore, silicon ions inhibit RANKL-mediated osteoclast activation, thereby creating a favourable environment for bone regeneration (51).

The Col@PDA&T-Exos scaffold activates important fibrogenic factors and modifies macrophage polarisation to dramatically improve rotator cuff tendon-to-bone repair. According to in vitro tests, the TSPC-Exos scaffold created a beneficial anti-inflammatory milieu for healing by inhibiting M1 macrophage polarisation and encouraging M2 macrophage migration. Additionally, in vivo research shows that the TSPC-Exos scaffold promotes the distribution of M2 Macrophages near the tendon-bone interface in the early postoperative phase and inhibits M1 Macrophages’ activity, which speeds up the healing process. Furthermore, miRNA-Seq research shows that the TSPC-Exos scaffold releases miR-21a-5p, which may target the PDCD4/AKT/mTOR signalling pathway and induce fibrosis. In particular, miR-21a-5p probably inhibits the production of PDCD4, which triggers the AKT/mTOR signalling pathway and encourages BMSC migration, proliferation, and fibrotic differentiation. In addition to demonstrating the TSPC-Exos scaffold’s possible function in controlling macrophage polarisation, this method offers essential theoretical justification for comprehending its possible uses in regenerative medicine (107).

The M2M@PLGA/COX-siRNA biomimetic delivery system employs the homotargeted properties of M2Mφ membranes to achieve precise enrichment at injury sites. The PLGA core enables controlled intracellular release of COX-siRNA, inhibiting PGE2-mediated NF-κB pro-inflammatory signalling and regulating macrophage polarisation from M1 to M2. The system further promotes secretion of anti-inflammatory factors alongside VEGF and TGF-β, accelerating angiogenesis, collagen remodelling, and interfacial mineralization to achieve multi-level intervention encompassing ‘inflammation targeting-immune regulation-regeneration promotion’ (108).

The CHA scaffold, combined with antagomiR-138/nHA and CuBG, promotes TBI healing by relieving inhibition across four key pathways to regulate macrophage polarisation: (1) The Wnt/β-catenin pathway enhances osteogenic gene expression; (2) The HIF-1α/VEGF axis promotes angiogenesis, with Cu²⁺ mildly inducing ROS to amplify signalling; (3) The ERK1/2-MAPK pathway amplifies osteogenic and β-catenin activity; (4) The BMP/SMAD pathway facilitates osteogenesis and endochondral ossification. This multi-pathway synergy, combined with an antimicrobial copper ion microenvironment, achieves rapid repair of large bone defects (109).

Owing to the inherent low metabolic activity of tendons, their minimal oxygen uptake, sparse vascularisation, reduced cellularity following rupture, and the gap created by retraction of the torn tendon, the restored function of an injured tendon rarely fully recovers its original capacity after healing (64, 110, 111). HUVECs-derived exosomes (HUVECs-Exos) incorporated into injectable hydrogels significantly enhance tendon repair. This system improves the local immune microenvironment by suppressing pro-inflammatory M1 macrophages and promoting anti-inflammatory M2 macrophage polarisation. Concurrently, it regulates the NF-κB, Wnt/β-Catenin, JAK-STAT and VEGF signalling pathways to enhance tendon stem cell proliferation, inhibit apoptosis, and support angiogenesis and tissue regeneration, thereby accelerating healing at the tendon-bone interface (112, 113).

As shown in Figure 3, CP@SiO₂ creates a favourable regenerative microenvironment for tendon repair by inhibiting M1 macrophage polarisation and promoting M2 macrophage polarisation. This reduces the release of pro-inflammatory factors while increasing the secretion of anti-inflammatory factors and tissue repair-related factors. Under normal conditions, GSK-3β degrades β-catenin. When GSK-3β is inhibited by ions released from the material, β-catenin escapes the degradation complex, accumulating extensively in the cytoplasm before translocating to the nucleus. Nuclearised β-catenin directly binds to NF-κB subunits, primarily p65. This interaction forms an inactive complex, preventing p65 from binding to the promoter regions of pro-inflammatory genes. Consequently, even in the presence of inflammatory stimuli within the microenvironment, the transcription factor remains ‘hijacked’ by β-catenin, preventing macrophages from initiating M1-type gene expression and thereby exerting an anti-inflammatory effect. Intranuclear β-catenin binds to TCF/LEF transcription factors, initiating transcription of downstream target genes. Growth factors and extracellular matrix components secreted by M2 macrophages further promote proliferation and differentiation of tendon stem/progenitor cells, enhancing tendon mechanical properties and functional recovery. This achieves a dual effect of inflammation suppression and tissue regeneration (114, 115).

Exosomes are extracellular vesicles secreted by cells, measuring approximately 30–100 nm in diameter. Their formation commences with the invagination of the membrane of early endosomes, maturing into multivesicular bodies (MVBs) before being released extracellularly via exocytosis (116). Exosomes can regulate target cell gene expression and functional molecules through autocrine, paracrine, and endocrine modes, and significantly modulate the tendon-bone healing process by influencing macrophage (Mφ) polarisation (117). Exosomes exhibit a two-stage immunomodulatory mechanism, commencing with the local macrophage polarisation regulation phase. Exosomes suppress pro-inflammatory M1 macrophage polarisation while promoting the recruitment of anti-inflammatory M2 macrophages, thereby reducing pro-inflammatory factor levels, inhibiting cartilage matrix degradation, and diminishing scar tissue deposition. Secondly, the M2 macrophage-mediated regeneration phase occurs. M2 macrophages secrete factors such as TGF-β3 and IGF-1 to induce mesenchymal stem cells towards fibrocartilage differentiation. Concurrently, IL-10 exerts anti-inflammatory effects and promotes extracellular matrix remodelling, thereby achieving both inflammation suppression and functional tissue regeneration (118).

Exosomes are diverse in type, and those derived from different sources exhibit distinct regulatory mechanisms and modes of action within the immune microenvironment. Dendritic cell-derived exosomes (DEXs) induce M1→M2 polarisation by suppressing STAT1/NF-κB and activating the STAT6 pathway. This reduces inflammatory cytokine expression while enhancing anti-inflammatory and reparative factors, thereby promoting tendon cell differentiation, type I collagen synthesis, and inhibiting type III collagen. Consequently, DEXs enhance tendon mechanical strength (119). Tendon sheath progenitor cell exosomes (TSPC-Exos) upon phagocytosis by macrophages enrich miR-21-5p and miR-23a-3p, activating the PI3K/AKT and STAT3 pathways whilst inhibiting NF-κB. This promotes M2 polarisation, modulates the immune microenvironment and fibrosis, thereby accelerating vascularisation, ordered collagen fibre arrangement, and fibrocartilaginous transition zone regeneration (107). As illustrated in Figure 2, bone marrow mesenchymal stem cell exosomes (BMSC-Exo) regulate the immune microenvironment via two distinct pathways. Firstly, they suppress M1 polarisation by targeting IRF1 through miR-23a-3p, whilst inhibiting NF-κB signalling via ANXA1-FPR2 to facilitate M2 phenotypic conversion. Secondly, they modulate the Hippo/YAP signalling axis to promote VEGF-dependent angiogenesis, establishing an anti-inflammatory homeostasis. This activates the TGF-β/VEGF paracrine network, ultimately enhancing fibrocartilage regeneration and improving interfacial biomechanical properties (120, 121).

By controlling macrophage polarisation, R-dECM facilitates tendon-bone repair. Rosiglitazone is sustainably released by R-dECM for more than seven days. In order to mitigate inflammatory reactions, rosiglitazone, a PPARγ agonist, efficiently inhibits M1 macrophage polarisation, lowers inflammatory cytokine release, and stabilizes M2 phenotype. Additionally, R-dECM preserves the M2 macrophage phenotype, supports the production of tendon-related genes (e.g., collagen I), prevents detrimental interactions between inflammatory cells and tendon stem/progenitor cells (TDSCs), and shields TDSC migration. In the end, this dual regulation system promotes tendon-bone repair and tissue regeneration by improving TDSC functional activity and optimizing the local microenvironment (122).

Research indicates that human bone marrow-derived mesenchymal stem cell conditioned media (hBMSC-CM) significantly promote tendon-bone interface (TBI) regeneration in a rat rotator cuff injury model by modulating macrophage (Mφ) polarisation. hBMSC-CM coordinates the local immune microenvironment, inhibiting pro-inflammatory M1 macrophage polarisation while promoting anti-inflammatory/repair-oriented M2 macrophage polarisation. This effect is dependent on the Smad2/3 signalling pathway, manifested by a significant increase in phosphorylated Smad2/3 (p-Smad2/3) levels. The Smad2/3-specific inhibitor SB431542 completely blocked the macrophage-modulating effects of hBMSC-CM.

Mechanistically, amplified M2Mφ promote fibroblast recruitment to TBI through paracrine effects, accelerating the early formation of dense fibrocartilage and ultimately establishing regenerated attachment sites with ordered layered organisation. Histological and biomechanical analyses demonstrate that the newly formed tissue exhibits not only significantly improved structural remodelling but also mechanical strength approaching that of natural tendon-bone interfaces. This study reveals that hBMSC-CM initiates a cascade of tissue regeneration by modulating the immune microenvironment via the Smad2/3 signalling axis, providing a theoretical basis for immune-regulated bone-cartilage tissue regeneration strategies (123).

Both in vitro and in vivo studies demonstrate that the paracrine activity of mesenchymal stem cells (MSCs) and their matrix can induce macrophage (Mφ) polarisation (124–133). In mammalian models of ligament or tendon healing, MSC therapy enhances endogenous M2 macrophages and associated anti-inflammatory factors, thereby improving tissue repair (134–137). MSCs regulate wound healing by secreting multiple growth factors and cytokines, with a key mechanism influencing macrophage phenotype achieved via extracellular vesicles (EVs). Engineered EVs achieve M1/M2 rebalancing through a multi-target strategy. Membrane surface signalling: CD47 integration on EV membranes inhibits macrophage SIRPα-mediated phagocytic activation, reducing TNF-α secretion; Content delivery: EVs carrying miR-146a target IRAK1/TRAF6 to suppress the NF-κB pathway, while miR-21 promotes IL-10 release via the PDCD4/AP-1 pathway; Angiogenesis synergy: Intra-EV semaphorin 4D activates VEGFR2 phosphorylation via Plexin B1, forming positive feedback with VEGF-A secreted by M2 macrophages to promote neovascularisation. This ‘immune-vascular coupling’ strategy elevated the M2/M1 ratio at the sheep tendon-bone interface from 0.7 to 3.2, significantly accelerating tendon-bone healing (138).

Suramin is a prescription drug used to treat human African trypanosomiasis (sleeping sickness). It possesses anti-inflammatory, anti-fibrotic and antioxidant properties, alongside other potent biological functions. The drug blocks post-injury extracellular adenosine triphosphate (ATP) activation of the immune system by inhibiting purinergic receptor signalling. It suppresses chemically induced atopic dermatitis by reducing levels of pro-inflammatory cytokines. As illustrated in Figure 2, suramin promotes tendinous bone healing by regulating macrophage polarisation and activating the NF-κB pathway. The drug diminishes pro-inflammatory activity by suppressing pro-inflammatory factor expression in M1 macrophages while activating their reparative functions; concurrently, it enhances the reparative capabilities of M2 macrophages. As a broad-spectrum purinergic receptor antagonist, suramin directly competitively binds to or blocks P2X7 receptors on macrophage surfaces. This step directly disrupts the pathway whereby extracellular ATP transmits the ‘inflammatory initiation signal’ intracellularly, a prerequisite for subsequent NF-κB regulation. Typically, ATP activation of P2X7 receptors induces intracellular potassium efflux and calcium influx, activating NLRP3 inflammasomes and inducing IKK phosphorylation. This leads to degradation of the NF-κB inhibitory protein IκB, releasing NF-κB (p65/p50 dimer) into the nucleus. The NF-κB dimer remains ‘locked’ in the cytoplasm, unable to enter the nucleus and bind DNA. Consequently, pro-inflammatory genes associated with M1 macrophages cannot initiate transcription. By suppressing the M1 pro-inflammatory phenotype, suramin creates an environment conducive to M2 polarisation. M2 macrophages subsequently upregulate reparative secretory functions. Furthermore, suramin promotes articular cartilage repair and angiogenesis by increasing COL1A1, COL3A1, and VEGF expression, thereby accelerating tendon-bone union. Concurrently, it synergistically enhances tendon stem cell migration, proliferation, and differentiation capacity, further driving tissue regeneration and repair (139).

Through three main methods, the dual-loaded LPN controls macrophage polarisation to facilitate tendon-bone healing: (1) Increasing the expression of the M2 marker CD206 and significantly decreasing the expression of the M1 macrophage marker CD86, which helps macrophages shift from the pro-inflammatory M1 phenotype to the anti-inflammatory and reparative M2 phenotype. (2) Dual-loaded LPN further suppresses inflammatory responses by downregulating pro-inflammatory genes and proteins. By downregulating the pro-fibrotic gene serpine1 and its associated PAI-1 protein, it concurrently slows the progression of fibrosis. (3) Dual-loaded LPNs promote tissue regeneration and repair by upregulating the expression of extracellular matrix (ECM) remodelling factors. In addition to reducing fibrosis and inflammation, this complex regulatory mechanism also fosters a favourable microenvironment for tendon-bone mending, which speeds up the healing process (140).

Disulfiram (DSF), an alcohol abuse medication approved by the FDA in 1950, has recently been found to possess potential therapeutic effects in multiple diseases beyond its action on aldehyde dehydrogenase, including anti-tumour and anti-septicaemic properties. In the latest investigation, DSF was found to significantly promote the shift of macrophages (Mφ) from the M1 phenotype to the M2 phenotype, thereby improving fibrosis at the tendon-bone interface through GSDMD-dependent pyroptosis regulation. In a mouse tendon injury model, macrophages exhibited pronounced M1 pro-inflammatory polarisation, characterised by upregulation of iNOS and TNF-α expression, activation of α-SMA+ fibroblasts, and abnormal extracellular matrix (ECM) deposition. DSF treatment markedly promoted the differentiation of CD206+ M2 macrophages while suppressing LPS/IFN-γ-induced expression of M1 markers. Animal studies revealed that oral DSF administration not only reduced F4/80⁺ macrophage infiltration but also inhibited inflammasome activation and pro-inflammatory cytokine release by blocking caspase-1/11-mediated cleavage of GSDMD-N, without affecting GSDMD precursor protein expression. Furthermore, DSF exerts cross-cell effects by modulating the biological actions of macrophage exosomes. Exosomes from DSF-pretreated BMDMs, when applied to fibroblasts, inhibited proliferation and reduced expression of α-SMA, FN, and Col-I. Concurrently, they delayed the G2/M phase of the cell cycle, thereby slowing wound healing and reducing excessive ECM deposition. This study first reveals the pathological association between macrophage pyroptosis and tendon fibrosis, proposing an innovative strategy for tendon repair intervention through a triple mechanism involving ‘immune phenotype remodelling—GSDMD inhibition—exosome regulation’ (141).

As discussed previously, the co-regulation of both macrophage phenotypes represents the ultimate objective in modulating the immune microenvironment. This encompasses the entire process of regulating inflammatory responses and numerous pathological processes, from accelerating inflammation during its initial phase to expediting healing and repair in its late stages—all of which necessitate the coordinated regulation of both phenotypes. From external stimuli such as mechanical, electrical, photothermal, and surface microtopography, to internal microenvironmental factors including reactive oxygen species (ROS), pH, and glucose levels; from fundamental hydrogels to emerging materials like engineered acoustic bodies (EABs), metallic scaffolds, and hydroxyapatite; From traditional, extensively validated nanoenzymes and hormonal therapeutics to the modern validation of millennia-old traditional Chinese medicine, evidence demonstrates that co-regulation of both macrophage phenotypes is feasible and highly effective. However, all materials and drugs must ultimately be grounded in specific studies of mechanisms and pathways. Numerous shared pathways—including NF-κB, Wnt/β-catenin, STAT3, and PI3K/AKT—not only modulate macrophage phenotypes but also regulate fibrosis, osteogenesis, angiogenesis, and antioxidant responses at the tendon-bone healing interface. Therefore, while we utilise macrophage polarisation as our entry point, our focus must extend beyond mere macrophage regulation to encompass functional studies pertinent to tendon-bone healing. This integrated approach will yield greater efficiency with less effort. Regarding existing regulatory strategies, we must pursue material innovation by advancing nanoscale and intelligent materials. Capitalising on the AI revolution, we should integrate AI with immune microenvironment-modulating materials. AI could determine drug delivery timing for materials, enabling truly personalised, round-the-clock therapeutics. Concurrently, we must leverage AI for efficient trial-and-error to identify optimal therapeutic agents, minimising adverse effects while maximising therapeutic efficacy. This approach must not only facilitate healing at the tendon-bone interface but also minimise sequelae such as fibrosis and associated complications. The rapid advancement of AI is poised to unlock entirely new perspectives in immune microenvironment regulation.