Bee sting-shaped microneedles for accelerating Achilles tendinopathy healing via enhanced regulating macrophage polarization

Abstract

Achilles tendinopathy (AT) is a common overuse injury to the Achilles tendon, and its progression is closely associated with an imbalance in M1/M2 macrophages and the inflammatory microenvironment. Traditional AT treatments are often limited by unwanted side effects, poor tissue penetration, and low therapeutic efficacy. Therefore, the targeted modulation of macrophage polarization could be a useful approach to relieve the inflammatory microenvironment of the Achilles tendon. In this study, we developed bee sting-shaped microneedles (DT-Exo-TMNs) to accelerate AT healing by enhancing the regulation of macrophage polarization. The DT-Exo-TMNs effectively reversed M1 macrophage polarization to M2 macrophage with a high efficiency of 43.4% by activating mitochondrial transcription factor A (TFAM). In addition, the photothermal agent, tannic acid /Fe³⁺, embedded in the DT-Exo-TMNs rapidly increased the temperature under near infrared (NIR) irradiation, significantly enhancing the antibacterial activity, promoting the release of therapeutic agents, and facilitating angiogenesis. In vivo experiments confirmed that DT-Exo-TMNs possess sufficient mechanical strength to insert the skin and deliver DT-Exo deep into tendon tissue in AT rat, thereby accelerating tendon repair through the synergistic alleviation of the inflammatory microenvironment and the regulation of immune homeostasis. In summary, these antibiotic-free, biocompatible, and multifunctional DT-Exo-TMNs hydrogel microneedles exhibit substantial potential for AT immunotherapy, providing a safe and effective therapeutic alternative for clinical applications.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04364-8.

Introduction

Achilles tendinopathy (AT) is a common overuse injury of the Achilles tendon, resulting in substantial pain in ankle and foot among athletes and individuals who engage in high-intensity sports [1, 2]. AT-related injuries can considerably impair an individual’s ability to participate in sports and daily activities, potentially leading to complications, extended rehabilitation periods, and hindered functional restorations [3]. Surgery is the predominant method for treating severe Achilles tendon injuries. However, it does not completely restore tendon functionality and poses a risk of re-rupture [4]. Moreover, the risk of post-surgical infection underscores the importance of incorporating antibacterial properties for successful regeneration [5, 6]. Clinically, conservative treatment for AT, including physical therapy and oral anti-inflammatory medications, is often preferred [7, 8]. However, these treatments are limited by low efficiency, low penetration and side effects [1, 9]. Achilles tendon plays a crucial role in facilitating mobility, necessitating the development of materials that can withstand stress and facilitate effective repair [10]. As a consequence, there is a pressing demand for new therapeutic approach that not only enhances tendon regeneration but also significantly improves patient recovery outcomes.

The pathological mechanisms underlying AT remain poorly understood, contributing to limited therapeutic options. Emerging evidence suggests that M1 macrophages play a role in engulf pathogens and promote inflammatory responses in AT progression, whereas M2 macrophages are linked to the reduction of inflammation and the process of tissue formation [11, 12]. Notably, mitochondrial dysfunction in M1 macrophages, characterized by impaired oxidative phosphorylation and accumulation of metabolites, significantly impedes their ability to repolarize into the M2 phenotype [13, 14]. Therapeutic strategies aimed at restoring mitochondrial function through targeted modulation of mitochondrial metabolism could offer a new therapeutic treatment for regenerative medicine. In addition, tendon healing depends on the capacity of tenocytes to respond to oxidative stress and inflammatory signals [15, 16]. Reactive oxygen species (ROS) are implicated in stress-induced apoptotic pathways, potentially contributing to the onset of tendinopathy [17]. Upon tissue injury, tenocytes also secrete pro-inflammatory cytokines, such as IL-1β and IL-6, which modulate the remodeling of the extracellular matrix of the tendon [18, 19]. Concurrently, elevated levels of ROS and the inflammatory microenvironment led to an increase in M1 phenotype macrophages, which promote the progression of AT [20, 21]. Therefore, there is an urgent need to develop intelligent therapeutic agents to address the inflammatory microenvironment and regulate macrophage polarization in AT treatments.

To overcome the challenge of regulating macrophage polarization, stem cell-derived exosomes (Exos) have gained traction in biomedical research due to their biosafety, translational potential, and engineering flexibility [22, 23]. Exos effectively reprogram M1 macrophages toward the M2 phenotype, as evidenced by significant upregulation of CD206, highlighting their therapeutic potential in resolving inflammation through targeted polarization modulation [24, 25]. Considering mitochondrial dysfunction in M1 macrophages, Exos with mitochondrial-targeting ability may enhance M1 macrophage polarization efficacy. Therefore, mitochondria targeted-ligand modified stem cell-derived Exos may promote tissue repair by repolarizing M1 macrophages to the M2 phenotype. Additionally, ROS sustain the pro-inflammatory state of M1 macrophages via activation of the HIF-1α/NF-κB signaling cascade in AT [26, 27]. Conversely, ROS scavenging or pharmacological blockade of this signaling pathway effectively promotes M2 polarization, thereby ameliorating inflammatory responses [27]. It has been reported that natural polyphenolic compounds, such as tannic acid and ferric ions (TA/Fe³⁺), have excellent photothermal conversion ability, remarkable antioxidant capacity, and excellent ROS scavenging ability [28, 29]. Leveraging the anti-inflammatory and antioxidant properties of TA/Fe³⁺, Exos has the potential for promoting tendon healing through enhanced regulation of macrophage polarization [30, 31]. The synergistic effect of the two not only significantly enhances tendon healing but also overcomes the limitations of the traditional treatment in terms of bioactivity modulation and long-term functional restoration, and provides a good therapeutic paradigm for tendon repair. However, the targeted delivery of this combination to the Achilles tendon lesion site has yet to be adequately optimized.

Nonsurgical AT treatment often requires months to achieve pain relief, and the efficacy of pharmacological agents is markedly limited by suboptimal therapeutic outcomes and poor tissue permeability [32, 33]. Microneedle (MNs) mediated transdermal drug delivery has been extensively used in the management of AT owing to its substantial enhancement of the absorption of therapeutic compounds [34, 35]. After encapsulating the therapeutic drugs, MNs insert the stratum corneum to reach the Achilles tendon area by creating numerous reversible microchannels with minimal invasiveness [36]. Currently, most existing microneedle systems have been developed for dermatological applications, with very few being adapted for AT [37, 38]. Moreover, preventing undesirable tissue adhesion remains a significant challenge in the design of microneedles for AT treatment. Despite the promising potential of MNs in clinical applications, their further development has been hindered by high costs, potential toxicity, and suboptimal drug release profiles, compounded by their tendency to detach during ankle movements, owing to soft tissue compression [39, 40]. Consequently, it is imperative to develop a novel MNs system capable of overcoming these limitations, thereby enhancing the clinical translation potential of AT therapy and achieving superior therapeutic outcomes.

In this study, a bee-sting-like tip was constructed to strengthen the structural integrity of a MNs array, inspired by the bee sting needle [40]. This unique array ensured low insertion resistance and high attachment strength during penetration, which may offer more potential benefits in resisting body movement compared with conventional smooth conical microneedles. Application of the drug-loading capacity of MNs can establish transdermal delivery pathways to enhance drug bioavailability and delivery efficiency [41, 42]. Our work pioneers the application of exosome and TA/Fe³⁺ loaded microneedles, engineering a system for deep tissue penetration and sustained release, thereby targeting the deep-seated and structurally complex pathology of Achilles tendinopathy. To further enhance the repolarization efficacy of M1 macrophages, Exos were functionalized with dual-target ligands (DT-Exo by metabolic glycoengineering (MGE)-mediated biorthogonal copper-free click chemistry using dextran sulfate (DS) modification of membranes and introduction of DSPE-PEG₂₀₀₀-TPP using the post-insertion technique (Fig. 1a) [43, 44]. Hence, the bee-sting-like shaped MNs loaded with exosomes and TA/Fe³⁺, termed as DT-Exo-TMNs, was synthesized by mixing gelatin methacryloyl (GelMA), DT-Exo, and TA/Fe³⁺ with a PDMS-negative mold (Fig. 1b). When TA/Fe³⁺ was combined with engineered DT-Exo via photopolymerization, a stable 3D network was formed [45, 46]. DT-Exo-TMNs could enhance M1 macrophage repolarization and reduced tenocyte inflammation in vitro. In addition, DT-Exo-TMNs inhibited bacterial growth to reduce the risk of infection under NIR irradiation, maintaining a sterile environment for tissue regeneration and repair. After inserted in the Achilles tendon region of AT rats, DT-Exo-TMNs accelerated tendon repair by synergistically relieving the inflammatory microenvironment and regulating immune homeostasis (Fig. 1c). Thus, our study highlights a highly promising therapeutic strategy to effectively halt AT progression and demonstrates its broad clinical application in AT treatment.

Fig. 1.

Schematics illustrating bee sting-shaped MNs for accelerating Achilles tendinopathy healing via enhanced regulating macrophage polarization. (a) Schematics illustrating preparation of DT-Exo. HMSC-Exo were isolated from the serum-free supernatant of DS-HMSC. DSPE-PEG₂₀₀₀-TPP were conjugated on the surface of HMSC-Exo though click chemistry, then DT-Exo were obtained. (b) Schematics illustrating preparation of DT-Exo-TMNs. The body solution was added to the mold using a pipette to fill the cavity by heating and vacuum, and the tips were solidified using UV irradiation. Exo-based hydrogel MNs (DT-Exo-TMNs) were constructed after encapsulating them in a native solution containing TA/Fe³⁺. Hydrogel MNs DT-Exo-TMNs were obtained by demolding. (c) Schematics illustrating efficiency evaluation of DT-Exo-TMNs in AT rat. DT-Exo-TMNs accelerated the healing process of Achilles tendinopathy by promoting M2 polarization and exerting anti-inflammatory effects. AT, Achilles tendinopathy; Exo, exosome; HMSC, human mesenchymal stem cell

Materials and methods

Patient sample

Human AT samples were obtained from the Affiliated Hospital of Wannan Medical College. All participants signed an informed consent form and the study was approved by the Ethics Committee of Wannan Medical College (No. 2023 − 129). The patient information is shown in Table S1. Our staging is consistent with previous literature: the early stage is stage I, which is reactive tendinopathy; the middle stage is Phase II, which means tendon disrepair; the late stage is stage III, which is degenerative tendinopathy [47, 48]. Staging is determined based on a comprehensive assessment of the patient’s symptoms, signs, and imaging examinations. Specimens of early and middle stages were obtained during the process of small needle knife treatment, which is a traditional Chinese medicine treatment method [49]. Concurrently, specimens of late stage were obtained through the arthroscopic surgery. Patients who needed amputation for seriously damaged were regarded as the control group [21].

The Achilles tendinopathy samples were extracted and prepared for HE, Masson, immunofluorescent (iNOS, 1:2000, 18985-1-AP; CD206, 1:1000, ab300621), and immunohistochemical staining (collagen Ⅰ, GB1102-3; collagen Ⅲ, GB111629, Servicebio, Wuhan, China). The expression of the inflammation cytokines IL-1β (KGC1103-48, Keygen Biotech, Jiangsu, China), IL-6 (KGC1111-48, Keygen Biotech, Jiangsu, China), and TNF-α (EK182-96, MULTI SCIENCES, Zhejiang, China) was analyzed using ELISA kits (Figs. 2, 3, 4, 5).

Fig. 2.

M1 macrophage positively related to the progression of Achilles tendinopathy in human samples. (a) Schematics illustrating M1 macrophage related to the progression of Achilles tendinopathy in human samples. (b) Immunohistochemistry of COL1 and COL3, Masson staining, and HE staining in human Achilles tendons at different stages of tendinopathy; scale bar: 100μm. (c) The positive Masson collagen fiber of Masson staining human Achilles tendons at different stages of tendinopathy measured by using ImageJ. Data are shown as means ± SD. (n = 3). The significant differences are determined by the one-way ANOVA, **p < 0.01, ***p < 0.001, compared with the Early (I) group. (d) The relative fluorescence intensity of iNOS in human Achilles tendons at different stages of tendinopathy, scale bar: 100 μm. (e) The relative immunofluorescence intensity analysis of iNOS in human Achilles tendons at different stages of tendinopathy calculated by ImageJ. The immunofluorescence intensity of iNOS in the Late (III) group was set as 1. Data are shown as means ± SD. (n = 3). The significant differences are determined by the one-way ANOVA, *p < 0.05, ***p < 0.001, compared with the Late (III) group. (f–h) The expression of (f) IL-1β, (g) IL-6, and (h) TNF-α in human Achilles tendons at different stages of tendinopathy measured using ELISA (n = 3). The significant differences are determined by the one-way ANOVA, *p < 0.05, **p < 0.01, ***p < 0.001, compared with the Late (III) group

Fig. 3.

Synthesis and characterization of DT-Exo-TMNs hydrogel MNs. (a) Schematics illustrating the synthesis of DT-Exo-TMNs hydrogel MNs. (b) Representative TEM images of HMSC-Exo and DT-Exo. Scale bar: 50 nm. (c) Nanoparticle tracking analysis of DT-Exo. The insert image indicated the live stream views of DT-Exo. Scale bar: 800 nm. (d) The expression of surface protein markers (CD9, CD63, CD81, TSG101, Alix and calnexin) in HMSC-Exo and DT-Exo. (e) The morphology of DT-Exo-TMNs hydrogel MNs measured using SEM. Scale bar: 800 μm. (f) The morphology of single tip of DT-Exo-TMNs hydrogel MNs measured using SEM. Scale bar: 300 μm. (g) The pore size distribution of different hydrogel MNs. Scale bar: 50 μm. (h) The EDX mapping of DT-Exo-TMNs hydrogel MNs. Scale bar: 200 μm. (i) The fluorescence image of DT-Exo-TMNs hydrogel MNs loaded with DiR-labeled Exos. (j) The tensile modulus of DT-Exo-TMNs hydrogel MNs. (k) The compressive stress of DT-Exo-TMNs hydrogel MNs. SEM, scanning electron microscopy; TEM, transmission electron microscopy

Fig. 4.

DT-Exo-TMNs promoted the repolarization of M1 macrophage into M2. (a) Schematics illustrating DT-Exo-TMNs promoted the repolarization of M1 macrophage into M2. (b) The intracellular accumulation of Dil-labeled DT-Exo in M1-polarized RAW264.7 cells within 9 h stained with F-actin cytoskeleton. Scale bar: 100 μm. (c) The fluorescence co-localization analysis of mitochondrial staining (green) and Dil staining (red) in M1-polarized RAW264.7 cells after treated with Dil-labeled DT-Exo within 9 h (yellow line in the merged image). Scale bar: 50 μm. (d) Flow cytometric analysis of M1 macrophage polarization in M1-polarized RAW264.7 cells after treated with DT-Exo-TMNs within 12 h. (e) The expression of iNOS, CD206, and TFAM in M1-polarized RAW264.7 cells after treated with DT-Exo-TMNs measured using western blotting. (f) Flow cytometric analysis of M2 macrophage polarization in M2-polarized RAW264.7 cells after treated with DT-Exo-TMNs within 12 h. (g) The expression of iNOS and CD206 and the M2 markers, including Arginase-1, IL-10 and transforming growth factor-beta (TGF-β), in M2-polarized RAW264.7 cells after treated with DT-Exo-TMNs measured using western blotting. (h) Confocal laser scanning imaging merged of M1 macrophages labeled with iNOS (red) and CD206 (green) after treatment with DT-Exo-TMNs for 24 h. Scale bar: 50 μm. (i) Confocal laser scanning imaging merged of M1 macrophages and M2 macrophages labeled with iNOS (red) and CD206 (green) after treatment with DT-Exo-TMNs 24 h. Scale bar: 50 μm

Fig. 5.

Photothermal properties, sustained release and antimicrobial properties of DT-Exo-TMNs hydrogel microneedles. (a) Schematics illustrating antimicrobial properties of DT-Exo-TMNs hydrogel microneedles under NIR irradiation (1.7 w cm^-2). (b) Photothermal properties of hydrogel microneedles under NIR irradiation (1.7 w cm^-2). (c) In vitro cumulative release of DiR-labeled exosome from hydrogel microneedle under irradiation. Data are shown as means ± SD. (n = 3). (d) The viable colony units (E. coli and S. aureus) grew on the plates incubated with DT-Exo-TMNs hydrogel microneedles under NIR irradiation (1.7 W cm^-2) for 5 min (n = 3). (e) The number of viable colonies of E. coli and S. aureus incubated with DT-Exo-TMNs hydrogel microneedles under NIR irradiation. The number of viable colonies of E. coli and S. aureus in the MNs group was set as 1. Data are shown as means ± SD. (n = 3). The significant differences are determined by the one-way ANOVA, ***p < 0.001, in comparison with the MNs group. (f) The bioTEM of E. coli and S. aureus incubated with DT-Exo-TMNs hydrogel microneedles under NIR irradiation, scale bar: 1 µm.

The preparation and characterization of DT-Exo

HMSCs were seeded at the bottom of cell culture dishes and incubated with Ac₄ManNAz (1 µM) to generate azide groups on the surface through the sialic acid pathway [43]. After 48 h of incubation, the cells were incubated with DBCO-DS (1 µM) for 2 h. Finally, DS-modified HMSCs were obtained after washing with PBS twice. To prepare Exos derived from DS-modified HMSC, the competed culture media (HUXMA-90011, OriCell, Guangdong, China) were changed into Exos-free culture medium (HUXMA-90012, OriCell), then the supernatant was collected and DS-modified HMSC Exos (HMSC-Exo) were isolated by using ultracentrifugation techniques [50]. The cells and cellular debris were removed at 10,000 × g for 20 min. Then, larger extracellular vesicles were discarded at 100,000 × g for 30 min. Finally, the HMSC-Exo pellet was obtained and protein concentrations were quantified using the Micro BCA Protein Assay Kit. HMSC-Exos (1 mg mL^{− 1}) were incubated DSPE-PEG₂₀₀₀-TPP (5%) at 4 °C for 6 h, and the DSPE-PEG₂₀₀₀-TPP was inserted on the surface of HMSC-Exos using post-insertion techniques [43, 51]. Subsequently, the DT-Exo were obtained. WB was performed to detect exosome biomarkers using Exos anti-CD9 (1:500, AF5139) and anti-TSG 101 (1:5000, AB125011) antibodies. The morphology of Exos was examined using a transmission electron microscope (HT-7800, HITACHI, Tokyo, Japan) at an accelerating voltage of 100 kV, this work was partially carried out at WanNan Medical College Electron Microscopy Center. The hydrodynamic sizes and zeta potentials of HMSC-Exo and DT-Exo were measured using DLS. In addition, the HMSC-Exo (1.9 × 10¹⁰ particles/mL) and DT-Exo (1.2 × 10¹⁰ particles/mL) size distributions were analyzed using nanoparticle tracking analysis (NTA) (Fig. 3c and Fig. S4). Moreover, the polydispersity index (PDI) of HMSC-Exo and DT-Exo is 0.21 and 0.17.

NanoFCM analysis

To evaluate the hybridization efficiency of two components (DBCO-DS and DSPE-PEG₂₀₀₀-TPP) in DT-Exo using the Flow NanoAnalyzer (NanoFCM Co., Ltd., China) [52, 53]. We selected suitable dyes laser emitters, emission wavelengths of 670 nm and 520 nm, corresponding to the DBCO-Cy5 and DSPE-PEG₂₀₀₀-FITC dye markers, respectively. HMSCs were seeded at the bottom of cell culture dishes and incubated with Ac4ManNAz (1 µM) to generate azide groups on the surface through the sialic acid pathway [43]. After 48 h of incubation, the cells were incubated with DBCO-Cy5 (1 µM) (Same concentration as DBCO-DS) for 2 h. Finally, Cy5-modified HMSCs were obtained. The competed culture media were changed into Exos-free culture medium, then the supernatant was collected and Cy5-modified HMSC Exos (Cy5-HMSC-Exo) were isolated by using ultracentrifugation techniques [54]. HMSC-Exos (1 mg mL^{− 1}) were incubated DSPE-PEG₂₀₀₀-FITC (5%) (Same concentration as DSPE-PEG₂₀₀₀-TPP) at 4 °C for 6 h, and the DSPE-PEG₂₀₀₀-FITC was inserted on the surface of HMSC-Exos using post-insertion techniques [43], the FITC/Cy5-Exo were obtained with a final concentration of 4.6 × 10¹⁰ particles/mL. By using NanoFCM to measure concentrations and molecular weights of the fluorescent markers of DT-Exo.

Fabrication and characterization of MNs (DT-Exo-TMNs)

The MNs array was fabricated using a negative mold [55]. GelMA (Catalog No.: EFL-GM-90, China) was dissolved in 20 mL of PBS containing 0.25% LAP (Catalog No.: EFL-LAP, China) solution at 50 ℃ for 15 min. The tip of each MNs hole was loaded with a needle solution (20%), TA/Fe³⁺ (mass ratio 2:1, 250 µg mL^{− 1}), and DT-Exo (30 µL, 1 mg mL^{− 1}) in 1.8 mL of PBS. Using a pipette, 400 µL of the subject solution was injected into the mold (Y543barb) and allowed to fully fill the mold cavity by heating and vacuuming. After removing the excess solution, the tip was irradiated with UV light for 30 s to cure the solution. Hydrogel MNs can be obtained by demolding after adding 300 µL of base solution to cover the tip of the needle and drying at 37 °C overnight. Subsequently, DT-Exo and TA/Fe³⁺ co-loaded MNs (DT-Exo-TMNs) were obtained. The morphology of freeze-dried DT-Exo-TMNs was observed using scanning electron microscopy (SEM: sigma300, ZEISS, Oberkochen, Germany; Mapping: Xplore30, Oxford Instruments, Oxford, UK). As shown in Fig. 5b, DT-Exo-TMNs exhibited good photothermal properties under NIR irradiation (1.7 W cm^{− 2}) within 5 min. The increased surrounding temperature were more than 40 ℃ in DT-Exo-TMNs hydrogel, which was attributed to bacterial death. DT-Exo-TMNs did not significantly compromise tendon cell viability compared to the non-irradiated control under this dose of NIR irradiation (Fig. S31). In addition, no damage was observed in the Achilles tendon tissue of rats treated with DT-Exo-TMNs followed by NIR irradiation in the tendon region over a one-month period (Fig. S46). Therefore, the transient high temperature induced by NIR irradiation does not cause damage to Achilles tendon tissue in vitro and in vivo. Then, TA/Fe³⁺ exhibited good NIR-responsive photothermal activity and the photothermal capacity was measured under NIR irradiation (1.7 W cm^{− 2}) for 5 min.

In vitro release of Exos from hydrogel MNs

To monitor the release profile of Exos from hydrogel MNs, DiR (KGMP0026; Keygen Biotech, Jiangsu, China) was incubated with DT-Exo to obtain DiR@DT-Exos. After loaded DiR@DT-Exo into hydrogel MNs, DiR@DT-Exo-TMNs was dispersed into PBS at 37 ℃, then the supernatants were collected. The release profile of DiR@DT-Exos was analyzed using an IVIS imaging system (IVIS Lumina LT, PerkinElmer, Massachusetts, USA).

Mechanical testing of hydrogel MNs

The mechanical properties of the DT-Exo-TMNs hydrogel MNs were evaluated using a universal mechanical testing system (WDW-05G, LINGC TEST, Shandong, China). DT-Exo-TMNs with dimensions of 16-mm length, 16-mm width, and 4.4-mm height were subjected to compression testing at a compression rate of 1 mm min^− 1. The Young’s modulus was calculated from the slope of the stress-strain curve.

Swelling properties of hydrogel MNs

To evaluate the swelling ratio of hydrogel MNs under physiological environment, DT-Exo-TMNs were incubated in PBS at 37 ℃. The DT-Exo-TMNs were weighed after removal from the PBS solution at specified times and dried with filter paper. The hydrogel MNs samples were weighed until the same weight was attained. The swelling ratio of DT-Exo-TMNs were calculated by the following equation: . where W₀ and W_t denote the initial mass of the hydrogel MNs and that after swelling at a predetermined time, respectively.

Photothermal antimicrobial properties

To evaluate the photothermal antimicrobial properties, hydrogel MNs were mixed with E. coli (gram-negative) or S. aureus (gram-positive), the sample was continuously irradiated with an 808 nm laser (1.7 W cm⁻²) for 5 min, and the photothermal temperatures were recorded every 60 s using a thermal imager. (FLIR E53, 84504415, USA).

Cytotoxicity assay of hydrogel MNs for HUVEC cells

CCK8 was used to assess the viability of HUVEC incubated with DT-Exo-TMNs. HUVEC (1 × 10⁴ cells per well) were inoculated in 96-well plates, and the hydrogel MNs leachate after 5 min of irradiation with an 808 nm NIR laser was added to the 96-well plates and incubated for 24 h. Cell viability was evaluated using the CCK-8 Cell Proliferation Detection Kit (KGA9306, Keygen iBioTECH, Jiangsu, China).

Live/Dead assay of hydrogel MNs for HUVEC

The proliferative capacity of HUVEC was detected using live/dead staining. HUVECs were incubated in six-well plates, and DT-Exo-TMNs leachate was added to the six-well plates and incubated for another 24 h. HUVEC were then stained using a live/dead assay kit (KGA9501, Keygen BioTECH) and imaged using a fluorescence microscope.

Cell scratch assay

In vitro scratch wound healing assay was performed to evaluate the migration ability of the hydrogel MNs. Cells were plated in a 24-well plate with a 1-mL tip. DT-Exo-TMNs was irradiation conducted for 5 min with an 808-nm NIR laser and after co-incubation with DT-Exo-TMNs for 6, and 12 h after the scratch was made to assess HUVEC migration. The migration of HUVEC was imaged and analyzed at 0, 6, and 12 h by using a microscope.

Tube formation assay

HUVEC were seeded into a 24-well plate coated with Matrix-Gel Basement Membrane Matrix (C0372-1 mL, Beyotime, Shanghai, China). DT-Exo-TMNs was irradiation conducted for 5 min with an 808-nm NIR laser and after co-incubation with DT-Exo-TMNs for 6 and 12 h. Tube formation in HUVEC under different treatment conditions was imaged and analyzed.

Cytotoxicity assay of hydrogel MNs for Achilles tendon stem cells

Primary Achilles tendon stem cells were isolated from rat tendons using the collagenase I (Catalog No.: 17018029, Gibco, United States) method [56]. Achilles tendon stem cells pretreated with tert-butyl hydroperoxide (TBHP) (30 µM) were seeded in 96-well plates. Then, after 5 min NIR irradiation (1.7 W cm^{− 2}) the hydrogel MNs leachate was inoculated into 96-well plates and incubated in an incubator for 24 h. Cell viability was evaluated using the CCK-8 cell proliferation assay kit (KGA9306, KeyGen BioTECH).

The uptake of DT-Exo in Achilles tendon stem cells

Achilles tendon stem cells pretreated TBHP (30 µM) were seeded in confocal dishes, and Dil-labeled DT-Exo were co-incubated for another 3, 6, and 9 h. Hoechst 33,342 staining was conducted to visualize the cell nucleus. Subsequently, images were captured using a laser confocal microscope.

Mitochondrial membrane potential detection in activated Achilles tendon stem cells

The protective effects of the hydrogel MNs on activated Achilles tendon stem cells were detected using JC-1 staining. Achilles tendon stem cells pretreated with TBHP (30 µM) were seeded on six-well plates, and incubated with the hydrogel MNs leachate after 5 min of NIR irradiation (1.7 W cm^− 2) for 24 h. Subsequently, the cells were stained using an Apoptosis Mitochondrial Membrane Potential Detection Kit (JC-1, KGA1904, keygen BioTECH). Subsequently, the Achilles tendon stem cells were imaged under a fluorescence microscope.

ROS assay in activated Achilles tendon stem cells

The ROS-scavenging capacity of the hydrogel MNs in activated Achilles tendon stem cells was examined using ROS staining. Achilles tendon stem cells pretreated with TBHP (30 µM) were seeded on six-well plates, and incubated with the hydrogel MNs leachate after 5 min NIR irradiation (1.7 W cm^− 2) for 24 h. The cells were stained with an ROS dye (H2DCFDA) (KGAF018, Keygen BioTECH). Subsequently, Achilles tendon stem cells were imaged under a fluorescence microscope.

TFAM expression of activated Achilles tendon stem cells

The expression of TFAM protein in activated Achilles tendon stem cells was detected using immunofluorescence staining. Achilles tendon stem cells pretreated with TBHP (30 µM) were seeded on confocal dishes, and incubated with the hydrogel MNs leachate after 5 min NIR irradiation (1.7 W cm^− 2) for 24 h. Cells were stained with anti-TFAM antibody (1:1000, ab307302) and the cell nucleus was stained with Hoechst 33,342. Subsequently, images were captured using a confocal laser-scanning microscope (TCS SP8; Leica, Wetzlar, Germany). The expression of TFAM were also detected using WB.

Macrophage polarization detection

RAW264.7 cells pretreated with LPS (1 µg mL^− 1) plus IFN-γ (10 ng mL^− 1) or IL-4 (20 ng mL^− 1) were co-cultured with the hydrogel MNs leachate after 5 min irradiation with an 808 nm NIR laser for 24 h. For flow cytometry quantitative analysis, typical surface markers of macrophages, such as PE-F4/80 (BD, 565410), PE/Cy7-CD86 (BD, 560582), AF647-CD206 (BD, 565250) were used. Antibody labeling of the macrophages was followed by flow cytometry. For immunofluorescence staining, the cells were rinsed three times with PBS and fixed using 4% paraformaldehyde solution, followed by incubation with 0.2% Triton X-100 in PBS solution for 20 min, and finally cells were closed with 1% BSA in DPBS solution for 45 min. Subsequently, cells were incubated with an anti-iNOS (1:2000, 18985-1-AP) or anti-CD206 antibody (1:1000, ab300621) at 4 °C for 30 min. Fluorescence imaging of cells was performed using a laser confocal microscope. The proteins of RAW264.7 cells with different treatments were extracted using a whole protein extraction kit, Whole Cell Lysis Assay (KGB5303, keygen BioTECH). The expression of iNOS and CD206 was detected by WB.

Small Interfering RNA (siTFAM) Transfection

We have performed experiments to more directly investigate the role of mitochondrial metabolism in the DT-Exo-TMNs-induced macrophage repolarization. SiTFAM (Table S2) was synthesized by Sangon Biotech (Shanghai). M1-polarized RAW264.7 macrophages were seeded in 6-well plates (2 × 10⁵ cells/well). Transfection was performed in serum-free DMEM. Briefly, siRNA was dissolved in DEPC-treated water (1 OD/125 µL). Then, 7 µL of siTFAM transfection reagent diluent was mixed with 690 µL DMEM, combined with 700 µL RNATransMate reagent diluents, incubated for 10 min at room temperature, and added to cells. After 8 h, medium was replaced with DMEM containing DT-Exo-TMNs leachate for 24 h. Total protein was extracted using the Whole Cell Lysis Assay kit (Keygen BioTECH, KGB5303).

Mitochondrial co-localization of DT-Exo in RAW264.7-differentiated M1 macrophages

RAW264.7 cells were pretreated with LPS and IFN-γ (for 24 h) were seeded into confocal dishes, and Dil-labeled DT-Exo were co-incubated for another 3, 6, and 9 h. The cells were stained with a Mito-Tracker Green FM and the cell nuclei were stained with Hoechst 33,342. Subsequently, images were captured using a confocal laser scanning microscope (TCS SP8; Leica, Wetzlar, Germany).

Rat AT healing with hydrogel MNs

All procedures strictly adhered to the adult male SD rats used for the experiments according to the Chinese National Guidelines for the Breeding and Use of Laboratory Animals (250–290 g) were obtained from Jiangsu Qing Longshan Biotechnology Co. Surgical procedures and perioperative handling were performed in accordance with the approved experimental program by the Ethics Committee of Wannan Medical College’s Medical School (WNMC-AWE-2023489). Rats were fed separately in separate cages for a week prior to surgery to acclimatize to the environment. A collagenase injection-induced AT rat model [4, 57] has been used to evaluate the therapeutical effects of hydrogel MNs. It has been reported that intratendinous or peritendinous injection of Type 1 collagenase is the most commonly used method for AT animal model [58]. The collagenase injection have been shown to induce similar histopathological findings to human tendinopathy and are still relied upon in preclinical studies. Intratendinous collagenase injections have replicated features such as collagen structure degradation, increased vascularity, hypercellularity, loss of matrix organization, and partial fragmentation of collagen fibers [59]. Col1 (ST2294, Beyotime, Shanghai, China) was prepared at a concentration of 5 mg mL^{− 1} and injected into the rat Achilles tendon epithelial tissue at a volume of 50 µL for every 2 days for 14 days. Subsequently, a rat AT model was established and MNs were inserted. After the MNs inserted the Achilles tendon, the rats (n = 6) were subjected to NIR irradiation (1.7 W cm^{− 2}) for 5 min. After 14 days, all tendons subjected to the different treatments were harvested and processed for HE, Masson, Sirius, and immunohistochemical staining.

In vitro Exos release assay

To evaluate the retention time of DiR-labeled Exo-TMNs or DT-Exo-TMNs in vivo after treatment in rat AT models, the reservation of MNs was monitored using an IVIS imaging system. The rats were anesthetized using isoflurane and images were captured using the IVIS imaging system at various time points post-insertion. Regions of interest were delineated around the tendons and fluorescence intensity was quantified using Living Image Software (PerkinElmer).

Statistical analysis

Normalized data are detailed in the figure captions. Normally distributed data are described as mean ± SD and analyzed using two-tailed unpaired Student’s t tests for two group comparisons or one-way analysis of variance (ANOVA) for multiple-group comparisons. Statistical significance was set at p < 0.05. differences. Data analysis was performed using SPSS software (version 27.0; Social Inc. IL, Chicago, USA).

Results and discussion

M1 macrophage promote the progression of Achilles tendinopathy in human samples

Macrophages play pivotal roles in tendon healing and tendinopathy. M1 macrophages dominate early stages of inflammation by releasing pro-inflammatory factors (e.g., IL-1β, IL-6, and TNF-α), exacerbating inflammation and tissue damage, whereas M2 macrophages promote anti-inflammatory processes and tissue repair. In the pathological microenvironment of AT (Fig. 2a), an imbalance in M1/M2 macrophages caused persistent inflammation and delayed tissue repair [60–62]. Concurrently, excessive ROS production contributes to mitochondrial dysfunction and apoptosis of tendon cells, ultimately exacerbating tissue damage [63, 64]. To identify the importance of M1 macrophages and the inflammatory microenvironment of AT, we collected human Achilles tendons at different stages of tendinopathy. In AT, a decrease in COL1 and an increase in COL3 lead to an imbalance in collagen ratios, and a decrease in the mechanical properties of the Achilles tendon [65]. As shown in Fig. 2b, Achilles tendon from the late stage of AT exhibited the highest expression of COL3 compared with that of the Early (I) and Middle (II) tendon rupture groups. In contrast, COL1 expression was the highest in the Achilles tendon of the early tendon rupture group and COL1 expression was lowest in the late tendon rupture group. Masson staining was used to evaluate collagen arrangement and fibrosis in Achilles tendinopathy. The results showed that collagen fibers displayed a significant proliferation phenomenon with fibrous encapsulation in the Early (I) group, which was 2.7-fold higher in the expression of collagen fiber deposition than that in the Late (III) group (Fig. 2c). Meanwhile, HE staining (Fig. 2b) revealed that in the Late (III) group, blood vessels were markedly dilated, necrotic cells exhibited abnormal morphology, and tissue structure was significantly more disrupted than in the early (I) and middle (II) groups. Therefore, these results characterize the pathogenesis of human Achilles tendons at different stages of tendinopathy.

Considering the critical role of macrophages in tendon healing, we evaluated the numbers of M1 and M2 macrophages in human Achilles tendons at different stages of tendinopathy. As shown in Fig. 2d and e, the number of M1 macrophages stained with iNOS in human Achilles tendons showed a gradual increase with the development of AT. The M1 macrophages in Early (I) group accounted for 49% in Late (III) group, however, the CD206 fluorescence intensity signal of M2 macrophages in the Late (III) group was substantially lower compared with that of the other two groups (Early (I) and Middle (II) groups). Therefore, an imbalance in M1/M2 macrophages were observed in the progress of AT.

M1-type macrophages are capable of releasing inflammatory factors, including IL-1β, IL-6 and TNF-α, to promote the development of AT [66, 67]. Meanwhile, the elevated ROS and inflammatory microenvironment can also promote the polarization of M0 macrophages to M1 [68]. We proceeded to examine the protein expressions inflammatory factors (IL-1β, IL-6, and TNF-α) in Achilles tendon from human patients using ELISA kits. As shown in Fig. 2f, the expression of IL-1β was the lowest in the normal group, which is 63% compared with that of the Late (III) group. In addition, the expression of IL-6 was significantly decreased in normal group, accounting for 40.2% compared with that of the Late (III) group (Fig. 2g). Similar results were also observed for the TNF-α expression levels (Fig. 2h). The TNF-α expression in Late (III) group was the higher than that in other groups, which is 4.3-fold compared with that of normal group. These results indicated that the inflammatory microenvironment is positively related with the progression of AT. These findings suggest that targeting M1 macrophages and reshaping the inflammatory microenvironment may offer an effective therapeutic approach for Achilles tendinopathy.

Preparation and characterization of DT-Exo-TMNs hydrogels

Given the therapeutic potential of promoting M1 to M2 macrophage repolarization in AT, stem cell-derived exosomes (Exos) offer a promising approach by effectively inducing this phenotypic switch in inflamed tissues. However, mitochondrial dysfunction in M1 macrophages, characterized by impaired oxidative phosphorylation and accumulation of metabolites, significantly impedes their ability to repolarize into the M2 phenotype [13, 14]. Thus, restoring mitochondrial function via targeted metabolic modulation may represent a novel regenerative therapeutic approach [43, 69]. DSPE-PEG₂₀₀₀-TPP modification enhances the mitochondrial targeting capability of nanomedicines, thereby improving reprogramming efficiency, restoring mitochondrial function, and promoting M1 to M2 macrophage repolarization [43]. Considering that elevated ROS and inflammatory factors in AT can induce the polarization of M0 macrophages to M1 [4, 70] modulating the inflammatory microenvironment may facilitate M1-to-M2 repolarization. TA/Fe³⁺ exhibits antioxidant, photothermal, and ROS-scavenging activity with good biocompatibility, effectively ameliorating the inflammatory milieu and promoting M1-to-M2 repolarization [71]. Therefore, the combination of TA/Fe³⁺ and DSPE-PEG₂₀₀₀-TPP modified stem cell-derived Exos may be a potent therapy for AT [43, 72].

Nonsurgical AT treatment for pain relief can take a few months to initiate conservative treatment [73], and the therapeutic efficacy of Exos remains limited by suboptimal effects [4, 74]. Encapsulated TA/Fe³⁺ and DSPE-PEG₂₀₀₀-TPP modified Exos, and the MNs array could insert the stratum corneum to reach the Achilles tendon area with hundreds of reversible microchannels in a minimally invasive manner [4]. However, MNs arrays are easily dislodged when soft tissue compression occurs with ankle motion [75, 76]. In this study, a triple bee-sting-shaped MNs was fabricated using a negative-pressure molding and drying method to enhance the stability of the MNs array (Fig. S1 and Fig. 3a). First, the macrophage-targeted ligand (DS) and DSPE-PEG₂₀₀₀-TPP modified Exo were prepared to regulate the M1-M2 balance of macrophages. Human mesenchymal stem cells (HMSCs) were incubated with Ac4ManNAz, followed by DS functionalization using bio-orthogonal copper-free click chemistry [43, 77]. Meanwhile, DSPE-PEG₂₀₀₀-TPP was inserted onto the surface of HMSC-Exos to obtain DT-Exo using the post-insertion technique [43, 44]. TEM images showed a round cup-shaped morphology, and well-dispersed appearance in both HMSC-Exos and DT-Exo (Fig. 3b). Compared with HMSC-Exos, the average diameter of DT-Exo increased to 125 nm (Fig. S2), whereas the zeta potential remained essentially unchanged (Fig. S3). Additionally, the size distributions were analyzed using nanoparticle tracking analysis (NTA) (Fig. 3c and Fig. S4), which indicated that the primary peaks for both profiles were at approximately 138 nm, consistent with the DLS measurements. Furthermore, we also performed NanoFCM experiment to quantify DS and TPP modification efficiency. By using NanoFCM to measure concentrations and molecular weights of the fluorescent markers of DT-Exo, we calculated that each exosome carried an average of ~524 FITC-TPP molecules and ~297 Cy5-DS molecules. Importantly, Western blotting (WB) confirmed the presence and concentration of key surface antigens, including CD9, CD63, CD81, TSG101 and Alix, and negative for calnexin on both HMSC-Exo and DT-Exo (Fig. 3d). These results validated that the modifications to DS and DSPE-PEG₂₀₀₀-TPP did not affect the bioactivity of HMSC-Exo.

Hydrogel MNs were then synthesized by mixing GelMA, DT-Exo, and TA/Fe³⁺ with the PDMS-negative mold [78], and DT-Exo-TMNs were obtained (Fig. S5). As shown in Fig. 3e and Fig. S6, the DT-Exo-TMNs have uniform tips with a bee-sting-like tip structure (Fig. 3f). Each hydrogel MNs exhibited a consistent, interconnected, and stable 3D network. Figure 3g shows that decreases pore size and denser pore structure were observed with the addition of TA/Fe³⁺ and DT-Exo, TA/Fe³⁺ forms chelates through the combined action of coordination bonds, hydrogen bonds, and electrostatic forces [79, 80]. Upon crosslinking, the polymerization of -C = C- groups induce the crosslinking and solidification of GelMA molecular chains, while the TA/Fe³⁺ system interacts with the GelMA network through hydrogen bonding, among them, Fe³⁺ and phenolic hydroxyl group forms Fe-O bonds [79, 81]. In addition, DT-Exos reversibly bond with both GelMA and TA/Fe³⁺ chelates via hydrogen bonding, enabling synergistic therapeutic effects [82, 83]. This non-covalent binding mechanism exhibits robust stability under physiological conditions, constructing a tightly integrated hydrogel network in DT-Exo-TMNs. Meanwhile, the EDX mapping of the DT-Exo-TMNs further confirmed the loading of TA/Fe³⁺ in the MNs (Fig. 3h and Fig. S7). Fe³⁺ helps hydrogel formation by facilitating polymerization as well as forming ionic cross-linking networks with TA. Therefore, this hydrogen-bonding-based binding allowed the loading of Exos into the hydrogel [82, 84]. Importantly, a strong DiR signal was observed in hydrogel MNs loaded with DiR-labeled Exos(Fig. 3i), and TEM images of DT-Exo-TMNs revealed Exos in the pores of the MNs (Fig. S8), suggesting the attachment of Exos in DT-Exo-TMNs. These results demonstrate the successful fabrication of DT-Exo-TMNs with a bee sting-like tip structure.

DT-Exo-TMNs, featuring a bee-sting-like tip structure, should not only resist dislodgement caused by soft tissue compression during ankle joint motion but also possess sufficient mechanical strength to penetrate the skin without fracturing prior to or during insertion [40]. We therefore evaluated the Young’s modulus of the hydrogel MNs. As shown in Fig. 3j, the TMNs had better mechanical properties than MNs without the addition of TA/Fe³⁺. In addition, DT-Exo-TMNs exhibited stronger compression resistance without significant fracture compared with that of MNs (Fig. 3k). These results indicated that TA/Fe³⁺ enhanced the mechanical properties of MNs, and the incorporation of DT-Exo did not affect the mechanical properties of the hydrogel MNs. Importantly, HE staining of rat skin tissues treated with DT-Exo-TMNs hydrogel showed that DT-Exo-TMNs had inserted the stratum corneum and perforated into the epidermis to a depth of over 300 μm (Fig. S9). In addition, the water absorption capacity of DT-Exo-TMNs was determined using the swelling ratio according to the increase in mass. As shown in Fig. S10, DT-Exo-TMNs exhibited quick swelling behavior within the initial 2 h and reached dissolution equilibrium at 8 h, which was attributed to the water molecules rapidly drilling into the gel and interacting with the hydrophilic groups (e.g., -OH) of the prepared gel network to form a hydration layer. A major network of hydrogels containing GelMA consists of -C = C- bonds crosslinked by photoinduced polymerization, with bioactive agents connected to each other inside the hydrogel by H-bond interactions [79]. In addition, the enhanced complexation effect between the hydrophilic group (-OH) of TA and Fe³⁺ [68] can reduce hydrophilic interactions and strengthen the cross-linking of the network structure. Upon swelling, the swollen interlocking features of a bee-sting-like tip structure could help secure the DT-Exo-TMNs in place.

The adhesion performance of DT-Exo-TMNs in vivo were evaluated by using universal mechanical testing system [85, 86]. Securing rat skin and MNs after inserting DT-Exo-TMNs into rat skin, then slowly stretched them at a rate of 1 mm/min (Fig. S11 a). The cone-shaped MNs without bee-sting shape was taken as the control (Fig. S11 b). In the pull-out phase, a tensile force first increases as the MNs starts to recede from the tissue model due to the static friction between the MNs and the tissue model. The tensile force then drops after a critical point where the MNs starts to slide out of the tissue model. In this study, tension performance of a MNs is assessed via the maximum pull out force (T_max, green point in Fig. S11 c) observed at the critical point. Meanwhile, a cone-shaped microneedles as control group underwent the same procedure. The maximum tension force of DT-Exo-TMNs is 3.7-folder than cone-shaped microneedles (Fig. S11 d). These results suggested that bee-sting-like structure of DT-Exo-TMNs offer a promising anti-detachment capability during AT treatment.

Considering the fatigue resistance and long-term stability in the highly repetitive ankle joint region, we therefore systematically monitored the structural integrity of the DT-Exo-TMNs in rat Achilles tendon regions over 24 days. As is shown in Fig. S12, DT-Exo-TMNs underwent gradual biodegradation, thinning over time, the tip of the DT-Exo-TMNs ultimately disintegrated on the 10 day, and DT-Exo-TMNs gradually lose their adhesion and almost degraded at 24 days. Therefore, DT-Exo-TMNs show potential for treating chronic Achilles tendinopathy, a degenerative condition linked to overuse.

DT-Exo-TMNs hydrogels promoted the repolarization of M1 macrophage into M2

M1 macrophages exhibit significant mitochondrial energy metabolic dysfunction [43]. Mitochondrial energy metabolic reprogramming can effectively enhance the repolarization efficiency of M1 macrophages toward the M2 phenotype. Accordingly, in this study, dual-targeting engineered exosomes enhance mitochondrial energy metabolism by co-targeting M1 macrophages and their mitochondria, thereby achieving efficient phenotypic conversion of macrophages from M1 to M2 state [87, 88]. In the pathological environment of AT, we confirmed the coexistence of inflammatory status and M1/M2 macrophage imbalance in human samples. Elevated ROS and the inflammatory microenvironment increase M1 macrophages, which drive Achilles tendinopathy progression [20, 21]. In contrast, reducing ROS or blocking the related signaling pathway promotes M2 macrophages and alleviates inflammation [27]. To tackle this, we developed DT-Exo-TMNs, in which TA/Fe³⁺ first alleviates the local inflammatory microenvironment [79], followed by DT-Exo-mediated reversal of M1-to-M2 macrophage polarization, synergistically enhancing M1-M2 reprogramming.

Since DT-Exo-TMNs process anti-detachment ability, fatigue resistance and long-term stability during AT treatment, we therefore evaluated the therapeutical effects for AT. Macrophages play a key role in sustaining immune balance and facilitating tendon healing [89, 90]. DT-Exo was designed to regulate mitochondrial dysfunction to promote macrophage polarization to the M2 type, thus accelerating the tendon healing process. Leveraging the anti-inflammatory effect of TA/Fe³⁺ [91], DT-Exo-TMNs has the potential for enhancing the regulation of macrophage polarization by synergistically relieving the inflammatory microenvironment and regulating immune homeostasis (Fig. 4a). Firstly, we have performed intracellular localization to clearly distinguish between internalized exosomes or those merely attached to the cell surface. As is shown in Fig. 4b, the results clearly show a time-dependent increase in red fluorescence signals inside the cells, corresponding to the internalized Dil-positive exosomes. DT-Exo was clearly observed surrounding the nucleus region, and did not attach to the cytoskeleton. These results confirms that DT-Exo were successfully internalized by M1 macrophages. In addition, Dil-labeled DT-Exo exhibited a stronger red signal in M1 macrophage differentiated from RAW264.7 cells within 9 h-incubation (Fig. S13), which was 2.6-fold higher in fluorescence intensity than HMSC-Exo. These findings suggest that DS modification improves the ability of DT-Exo to target macrophages. These findings suggest that DS modification improves the ability of DT-Exo to target macrophages. We further performed co-localization experiments to determine the mitochondrial-targeting ability of DT-Exo in M1 macrophages. M1 macrophages differentiated from RAW264.7 cells, were stained with Mito-Tracker Green FM with green signal and Dil-labeled DT-Exo, respectively (Fig. S14). As illustrated in Fig. 4c, the overlap between red and green fluorescence signals was higher in the DT-Exo group than that in the HMSC-Exo group. In addition, the Pearson coefficient of the 9 h-DT-Exo group was approximately 0.62 (Fig. S15), indicating that DT-Exo specifically targets the mitochondria in macrophages.

The repolarization efficiency of DT-Exo-TMNs on M1 macrophages derived from RAW264.7 cells was then evaluated by flow cytometry. DT-Exo-TMNs treatment showed a significant decrease in iNOS and an increase in CD206 (Fig. 4d) in M1 macrophages derived from RAW264.7 cells. The number of F4/80⁺iNOS⁺ cells (M1 macrophages) was decreased from 85.15% to 48.15%, indicating that DT-Exo-TMNs repolarized M1 macrophages with a high efficiency of 43% (Fig. S16). Moreover, the conversion rate of the DT-Exo-TMNs group was significantly higher than that of the Exo-TMNs group, which is 1.6-time higher than that of Exo-TMNs group. This results indicates that dual-targeted modification more effectively restores mitochondrial function in M1 macrophages and promotes their conversion to M2 macrophages. WB assay also confirmed that DT-Exo-TMNs can efficiently reduce iNOS expression in M1 macrophages (Fig. 4e). In comparison with DS-modified Exos, the transformation efficiency of M1 macrophages into M2 of Exo-TMNs is down to 60% of DT-Exo-TMNs group. Additionally, the proportion of F4/80⁺CD206⁺ cells (M2 macrophages) rose from 6.13% to 16.60% after DT-Exo-TMNs treatment (Fig. 4f). WB also indicated (Fig. 4g) that DT-Exo-TMNs increased the expression of CD206 in M2 macrophages (Fig. S17). As shown in the Fig. 4g, the expression of M2 macrophages markers, including Arginase-1 and IL-10, were significantly upregulated by the DT-Exo-TMNs in contrast to the MNs group. The expression levels of Arginase-1 or IL-10 in the DT-Exo-TMNs group were 6.8-fold or 1.87-fold higher respectively compared to the MNs group, indicating that DT-Exo-TMNs can improve the polarization of M2 (Fig. S18). These findings indicate that modulation of mitochondrial metabolism effectively enhances the transformation capability of M1 macrophages. In addition, the protein expression of TGF-β was dramatically upregulated in the DT-Exo-TMNs group in contrast to MNs groups (Fig. S18 c), indicating that DT-Exo-TMNs have the potential to facilitate tissue repair. Importantly, a consistent trend in transformation was verified using immunofluorescence staining for iNOS and CD206 (Fig. 4h and i), even when RAW264.7 cells in the M2 polarized state were stimulated with LPS in combination with IFN-γ, which mimics the AT microenvironment, with a high level of M2 marker (CD206, green) after DT-Exo-TMNs treatments (Fig. 4i). Thus, DT-Exo-TMNs seemed to prevent M2-polarized RAW264.7 cells from undergoing pathological conversion into M1 macrophages.

In addition, the knockdown of TFAM significantly weakened the activation effect of DT-Exo-TMNs on TFAM, DT-Exo-TMNs cannot reduce iNOS expression and increase CD206 in M1 macrophages with TFAM knockdown (Fig. S19 b and Fig. S20). It has been reported that TFAM is able to promote mitochondrial biosynthesis by regulating the transcription and replication of mitochondrial DNA, thereby supporting the anti-inflammatory function of M2 macrophages [92]. The results demonstrated that TFAM was highly expressed in M2-polarized macrophages compared to M1-polarized group (Fig. S21). After treated with DT-Exo-TMNs, the expression of TFAM were upregulated in M1 macrophages (Fig. S22). These results indicated that the DT-Exo-TMNs effectively reversed M1 macrophage polarization to M2 macrophage by TFAM.

The antimicrobial activity of DT-Exo-TMNs hydrogels

Antimicrobial biomaterials offer substantial advantages in aiding tendon repair by inhibiting bacterial growth, reducing the risk of infection, and ensuring aseptic conditions [93]. TA/Fe³⁺ has good photothermal conversion performance and exhibits broad-spectrum antibacterial effects [94], we therefore investigated the infection-inhibiting features of hydrogel MNs with NIR irradiation (Fig. 5a). As shown in Fig. 5b, DT-Exo-TMNs exhibited good photothermal properties under NIR irradiation (1.7 W cm^{− 2}) within 5 min. The increased surrounding temperature were more than 40 ℃ in DT-Exo-TMNs hydrogel, which was attributed to bacterial death. In addition, the heating curves of DT-Exo-TMNs and Exo-TMNs groups similar to those of the TMNs groups, indicating that the combination with DT-Exo did not affect the photothermal properties of DT-Exo-TMNs (Fig. S23). Moreover, we have reperformed photothermal imaging experiment for different components TA/Fe³⁺, HMSC-Exo and DT-Exo with or without NIR irradiation. As shown in Fig. S24 a, TA/Fe³⁺ exhibited good photothermal properties under NIR irradiation. The temperatures of HMSC-Exo and DT-Exo showed no significant change within 5 min. Meanwhile, the microneedle temperatures for each group remained unchanged over 5 min without irradiation (Fig. S24 b). Importantly, DT-Exo-TMNs exhibited obviously NIR photothermal response after three rounds of NIR radiation and natural cooling, which is consistent with MNs hydrogel with TA/Fe³⁺ concentration ratio at 0.25 mg mL^{− 1} (Fig. S25). The outcomes outlined above reveal that DT-Exo-TMNs hydrogels possess excellent photothermal transition capacities.

The sustained release of exosomes from DT-Exo-TMNs hydrogels therapeutic agents was also evaluated under NIR irradiation. As shown in Fig. 5c, DiR-labeled DT-Exo was released from the DT-Exo-TMNs hydrogels within 14 days, and NIR irradiation promoted the release of Exo after 5 days. In addition, no obvious DT-Exo release was observed in the control group (Fig. S26), which clearly demonstrate a positive correlation between irradiation time and the cumulative release percentage. This finding further underscores the controllability of our NIR-triggered release system. Moreover, the key surface protein expression levels of CD81 and CD63 in DT-Exo released from the DT-Exo-TMNs after UV exposure remained comparable to those of DT-Exo before UV exposure (Fig. S27). These results indicates that the structural integrity and protein composition of the exosomes were well preserved throughout the UV exposure process. Previous research reports that exosomes contain cytokines with antimicrobial properties and innate response signaling molecules that are important in response to viral and bacterial infections [95]. Therefore, the released Exos have the potential to improve the antimicrobial activity of DT-Exo-TMNs hydrogels. As shown in Fig. 5d, DT-Exo-TMNs hydrogels had significant antibacterial activity for both Escherichia coli and Staphylococcus aureus after NIR radiation (1.7 W cm^{− 2}) compared with that in TMNs groups. The numbers of E. coli and S. aureus in the DT-Exo-TMNs group were decreased to 30% and 14% of TMNs, respectively (Fig. 5e). Importantly, bacterial damage following the DT-Exo-TMNs treatment was observed by bioTEM images. As shown in Fig. 5f, both E. coli and S. aureus in the DT-Exo-TMNs group exhibited the most severe damage compared with those in the other groups. These results indicated that DT-Exo-TMNs can achieves efficient bactericidal effects through synergistic effects of TA/Fe³⁺ and DT-Exo.

DT-Exo-TMNs hydrogels promote angiogenesis in vitro

Following injury, tendon healing does not result in regeneration of the natural tendon structure, instead, it involves excessive and disorganized deposition of the extracellular matrix triggered by early angiogenesis [96]. An inadequate healing response is often attributed to a paucity of cells and blood vessels within the tendon [97]. Since DT-Exo-TMNs holds the potential for tissue healing, we therefore investigated whether the DT-Exo-TMNs hydrogel could promote angiogenesis (Fig. 6a). Firstly, an MTT (Fig. S28) assay and Live/Dead staining were used to evaluate the biocompatibility of the DT-Exo-TMNs hydrogels. The DT-Exo-TMNs did not exhibit cytotoxicity during the observation periods of 24 and 48 h. In addition, the live/dead staining results (Fig. 6b) indicate that live HUVEC with green staining did not change in the DT-Exo-TMNs group within 24 h of incubation (Fig. S29). Meanwhile, negligible red signals of dead HUVEC were observed in DT-Exo-TMNs and MNs group. These results indicated that DT-Exo-TMNs possessed good biocompatibility. To evaluate the proangiogenic capacity of DT-Exo-TMNs, we then performed in vitro tube formation experiments and wound healing test. As shown in Fig. 6c, HUVEC formed tubular structures radiating from the cellular aggregates and extensive primary vascular-like networks after DT-Exo-TMNs treatment. Extensive primary vascular-like networks characterized by radiating tubular structures were observed in the DT-Exo-TMNs group, which were 7.2-fold higher than those in the MNs group (Fig. 6d). In addition, the DT-Exo-TMNs group outperformed the MNs group in terms of total master segment (Fig. 6e) and cumulative vessel length (Fig. 6f). The number of tubes also increased in the DT-Exo-TMNs, and was 13.2-fold higher than that in the MNs (Fig. 6g), confirming the efficacy of DT-Exo-TMNs for promoting tube formation in vitro. Moreover, the wound healing test (Fig. 6h) showed that the scratched area of HUVEC in the MNs (6 h) group, which was similar to that in 12 h in the same group, was 4.3-fold higher than that in the DT-Exo-TMNs (12 h) group (Fig. 6i). These results suggested that DT-Exo-TMNs could enhance the migration of endothelial cells [98, 99]. It has been reported that enhancing angiogenesis during the early stages can accelerate tendon healing by facilitating extrinsic tendon repair [100, 101]. Therefore, DT-Exo-TMNs may be effective for treating tendon injury by promoting angiogenesis.

Fig. 6.

DT-Exo-TMNs hydrogel microneedles promote angiogenesis in vitro. (a) Schematics illustrating DT-Exo-TMNs hydrogel microneedles promote angiogenesis in vitro. (b) Live/Dead staining of HUVEC cultured with the hydrogel microneedles for 12 and 24 h. Scale bar: 100 µm. (c) Tube formation of representative image networks in HUVEC with DT-Exo-TMNs treatments. Scale bar: 100 µm. (d-g) (d) Total of junction, (e) master segments, (f) length and (g) tubes of HUVEC with different treatments. Data are shown as means ± SD. (n= 3). The significant differences are determined by the one-way ANOVA, **p < 0.01, ***p < 0.001, in comparison with the MNs group. h) Representative images of a scratch migration assay for 6 h and 12 h in HUVEC with different treatments. Scale bar: 500 µm. i) The scratched area of HUVEC with different treatments measured using Image J. The scratched area of 6 h-MNs group was set as 1. Data are shown as means ± SD. (n=3). The significant differences are determined by the one-way ANOVA, ***p < 0.001, in comparison with the 6h-MNs group

DT-Exo-TMNs hydrogels reduced TBHP-induced oxidative stress in tenocytes

The repair of tendon rupture relies on the intrinsic ability of tendon cells to respond to surrounding oxidative damage stimuli [102]. Therefore, we evaluated the protective effects of DT-Exo-TMNs for tenocytes upon oxidative stimulation with TBHP (Fig. 7a). DT-Exo-TMNs exhibited no toxicity to tenocytes and demonstrated significantly high biocompatibility during long-term co-culture for up to 24 h (Fig. S30). Meanwhile, DT-Exo-TMNs with NIR irradiation did not significantly compromise tendon cell viability compared to the non-irradiated control (Fig. S31). These results indicated the good biosafety of DT-Exo-TMNs during AT treatments. As shown in Fig. S30, the viability of tenocytes was markedly reduced by over 35% after pretreatment with TBHP, according to the CCK-8 assay. However, cell viability was rescued by the DT-Exo-TMNs leachate within 24 h. These results indicate that the DT-Exo-TMNs might alleviate TBHP-induced oxidative stress in tenocytes.

Fig. 7.

DT-Exo-TMNs reduced TBHP-induced oxidative stress in tenocytes. (a) Schematics illustrating DT-Exo-TMNs reduced TBHP-induced oxidative stress in tenocytes. (b) The intracellular accumulation of Dil-labeled DT-Exo in tenocytes within 9 h. Scale bar: 100 μm. (c) The JC-1 staining of in TBHP-pretreated tenocytes with the treatments of DT-Exo-TMNs. Scale bar: 100 μm. (d) The fluorescence images of ROS generated in TBHP-pretreated tenocytes with DT-Exo-TMNs treatments. Scale bar: 100 μm. (e) The fluorescence analysis of ROS generated in TBHP-pretreated tenocytes with DT-Exo-TMNs treatments. Data are shown as means ± SD. (n = 3). The significant differences are determined by the one-way ANOVA, ***p < 0.001, compared with the control group (Con). (f) Immunofluorescence images of TFAM in TBHP-pretreated tenocytes with DT-Exo-TMNs treatments. Scale bar: 100 μm. (g) The expression of TFAM in TBHP-pretreated tenocytes with DT-Exo-TMNs treatments measured using western blotting

Lines of evidence have shown that HMSC-Exos possess anti-inflammatory properties [103, 104]. Therefore, we further investigated the anti-inflammatory properties of DT-Exo in tenocytes. Dil-labeled DT-Exo were taken up by the tenocytes in a time-dependent manner (Fig. 7b). In addition, a stronger red signal of DT-Exo was observed compared to that of HMSC-Exo after 9 h of incubation, suggesting that the DSPE-PEG₂₀₀₀-TPP modification increased the accumulation of DT-Exo (Fig. S32). Mitochondria-associated membranes play a key role in triggering inflammatory responses during the cellular defense against oxidative damage [105, 106]. Changes in the mitochondrial membrane potential during cellular physiology and pathology reflect the functional state of mitochondria. As shown in Fig. 7c. JC-1 in tenocytes exists within the internal matrix of mitochondria as aggregates (J-aggregates), emitting red fluorescence, whereas in TBHP-induced oxidative stress in tenocytes, JC-1 emits green fluorescence in response to apoptosis or impaired mitochondrial function. As shown in Fig. S33, DT-Exo-TMNs rescued the damage caused by TBHP to the mitochondria in tenocytes, which generated an 18.6-fold in the intensity of red signal compared to MNs. Moreover, DT-Exo-TMNs treatment decreased the changes in cellular ROS levels following the TBHP challenge in tenocytes (Fig. 7d). Fluorescence analysis showed a markedly decreased density of TBHP-induced ROS in tenocytes after DT-Exo-TMNs treatment, which was 28% of that of MNs (Fig. 7e). These results indicated that DT-Exo-TMNs protected tenocytes from oxidative stress. Mitochondrial transcription factor A (TFAM) can help eliminate leaked mtDNA to restrict inflammation [107, 108]. Considering the anti-inflammatory properties of DT-Exo-TMNs, we investigated whether they could relieve mitochondrial damage and inflammation by activating TFAM. As shown in Fig. 7f, the intensity of the TFAM green fluorescent signal of tendon sheath cells in the MNs group was significantly weakened compared with that in the Control group, indicating that MNs alone did not rescue mitochondrial dysfunction from the damage caused by TBHP-induced oxidative stress (Fig. S34). However, higher expression of TFAM was observed in the DT-Exo-TMNs group, which was 1.5-fold higher than that in the Exo-TMNs group. Importantly, upregulation of the TFAM protein was observed (Fig. 7g) in the DT-Exo-TMNs group, showing a 1.32-fold increase compared to that in the MNs group (Fig. S35), indicating that DT-Exo-TMNs relieved mitochondrial damage by activating TFAM. Therefore, DT-Exo-TMNs has the potential for tendon repair in vivo.

In vivo evaluation of DT-Exo-TMNs hydrogels for AT treatment

Considering that DT-Exo-TMNs hydrogels have an advantageous performance for Achilles tendinopathy repair in vivo, we therefore evaluated the therapeutic effects of DT-Exo-TMNs for AT at a collagenase injection-induced AT rat model. Since DT-Exo-TMNs has long-term stability in the highly repetitive ankle joint region, the in vivo retention time of the DiR-labeled DT-Exo-TMNs were investigated using an IVIS imaging system. The accumulation of DiR-labeled DT-Exo-TMNs in the Achilles tendon region of AT rat was imaged after inserted in the tendon region and the DiR signal was quantified within 21 days (Fig. 8a). As shown in Fig. 8b, significant fluorescence of the DiR signal was observed within 21 days after inserted with DT-Exo-TMNs under NIR irradiation, which was slightly higher than that in the Exo-TMNs groups, suggesting that DT-Exo-TMNs has long retention time in the tendon region of AT rats. Therefore, DT-Exo-TMNs have the potential to promote repair of the Achilles tendon in vivo (Fig. 8c).

Fig. 8.

In vivo evaluation of DT-Exo-TMNs hydrogel MNs for Achilles tendinopathy rupture treatment. (a) In vivo imaging of DiR-labeled exosome loaded hydrogel MNs after inserted in the Achilles tendon region of rats. (b) Relative fluorescence intensity of DiR in rat Achilles tendon after inserted with DiR-labeled exosome loaded hydrogel MNs. The fluorescence intensity of Achilles tendon epithelial region inserted with DiR-labeled formulation at day 1 was set as 1. Data are shown as means ± SD. (n = 3). (c) Schematics illustrating DT-Exo-TMNs hydrogel microneedling treatment to alleviate AT in model rats. (d) Thermography of Achilles tendon region from AT rat with DT-Exo-TMNs treatment under NIR irradiation (1.7 W cm^− 2). (e) Macroscopic images, immunohistochemistry of HE staining in Achilles tendon tissue from AT rat with DT-Exo-TMNs treatment, scale bar: 100μm. (f) Immunofluorescence of iNOS and CD206 in the Achilles tendon from AT rat with DT-Exo-TMNs treatment; scale bar: 100 μm. (g) The expression of β-actin, iNOS, CD206 and TFAM in rat Achilles tendon tissues upon different treatments using western blotting. (h–k) The expression of (h) IL-1β, (i) IL-6, (j) TNF-α, and (k) TGF-β in Achilles tendon tissue from AT rat with DT-Exo-TMNs treatment by ELISA (n = 6). The significant differences are determined by the one-way ANOVA, ***p < 0.001, in comparison with the AT group. AT, Achilles tendinopathy

Some studies have reported the exosomes process of good diffusion capacity. Odessa Schillaci and his coworkers have reported Exosomes from the metastatic line (SW620Exos) exhibited higher ability to cause endothelial hyperpermeability than exosomes from the non-metastatic line (SW480Exos). Spindle versus amoeboid cells demonstrated that exosomes derived from amoeboid cells have a more remarkable ability to increase endothelial permeability [109]. In addition, the good penetration ability of ultrasmall Prussian blue nanoparticles (uPB-Exo) in inflammation cartilage, despite the surface of cartilage, DiR@uPB-Exo as well as DiR@NEs-Exo penetrated cartilage matrix deeply. The enhanced cartilage penetration of DiR@uPB-Exo is probably attribute able to the adhesion interactions between NEs-Exo and chondrocytes, which suggested that DiR@uPB-Exo could penetrated cartilage matrix deeply in vivo [110]. Therefore, exosomes are well tolerated and stable, and have been considered as ideal nanocarriers that can insert deep tissues and evade immune cell attacks to deliver drugs to the injured site [111]. We therefore performed CLSM experiment on tendon tissue sections after treated the TMNs loaded with Alexa Fluor 488-CD63 labeled DT-Exo. After inserted in the tendon region in AT model rats for 21days, the Achilles tendons were harvested to process into frozen sections and then visualized by using CLSM. As shown in Fig. S36, distinct green fluorescent signals with CD63-labeled exosomes were distributed within the tendon tissue sections with the depth of 800 μm in both DT-Exo-TMNs and Exo-TMNs groups. In addition, these signals of CD63-labeled exosomes were observed interstitially among the collagen fibers, providing direct visual confirmation that the exosomes released from the microneedles not only remained at the local site but also successfully penetrated into the deep of tendon tissue. The deep penetration of exosomes from DT-Exo-TMNs could do great help for enhancing AT treatments.

Given the favorable photothermal performance of DT-Exo-TMNs in vitro, we further examined their thermal profile in the tendon region under NIR irradiation following local insertion. (Fig. 8d). As shown in Fig. S37, thermal imaging showed that DT-Exo-TMNs increased the temperature of the tendon rupture region to more than 40 ℃ under NIR irradiation (1.7 W cm^{− 2}) in vivo, which is similar to the temperature change profile of TMNs. Considering that NIR irradiation could improve antimicrobial activity in vitro, a sterile environment might be maintained during treatment with DT-Exo-TMNs. The treatment regimen for AT was established based on a review of the pertinent literature. Liu and his coworkers have reported that nitric oxide nanomotor driving exosomes-loaded microneedles can significantly promote the healing process of AT within two weeks [4]. Yu and his coworkers have verified that, 2 weeks later, experimental group showed improved integration of the healing tissue with the host tendon while the defect area of the control group displayed poor regularity with structure disruptions. Furthermore, tenomodulin (Tnmd) is a tendon-specific marker important for tendon maturation and exerts a positive effect on TSPCs by supporting their self-renewal and preventing senescence, and it is also essential for the prevention of fibrovascular scar formation during early tendon healing. As noted in the text, the expression of Tnmd was to be elevated at week 1 and 2 in the tendon healing process [112]. Therefore, we conducted in vivo effectiveness evaluation of DT-Exo-TMNs hydrogels for Achilles tendon rupture treatment within two weeks. After two weeks post-treatment, rat Achilles tendon specimens were collected and subjected to HE, Masson, and immunohistochemical staining. The surface of the MNs-treated Achilles tendon and AT group were both encapsulated by extensive inflammatory proliferative tissue (yellowish and dark yellow tissue) in the macroscopic view (Fig. 8e). However, the inflammatory areas were significantly decreased after treatment with DT-Exo-TMNs for 14 days. Moreover, the degree of inflammatory proliferation was significantly higher in the MNs group and AT group than that in the DT-Exo-TMNs group. As observed by HE staining (Fig. 8e), the collagen fibers in the DT-Exo-TMNs group showed a regular and densely arranged structure as well as Normal group. Masson staining showed that tendons in the MNs group thickened, not only by producing more inflammatory tissue but also by edematous collagen fibers (blue-stained area), which was reversed in the DT-Exo-TMNs groups (Fig. S38). In addition, we performed immunohistochemical staining for type I collagen (Col 1) and type III collagen (Col 3). Type III collagen (Col 3) is expressed at low levels in healthy Achilles tendon tissue, whereas Col 1 was highly expressed and neatly ordered in the MNs group. As shown in Fig. S38, Col 3 was significantly higher and Col 1 was significantly lower in the MNs group, which was reversed by DT-Exo-TMNs. These results indicated that DT-Exo-TMNs promoted tendon repair in vivo.

Since DT-Exo-TMNs could promote angiogenesis in vitro, we performed co-localization studies of CD31 and α-SMA in inflamed Achilles tendon area to evaluate the angiogenic ability of the DT-Exo-TMNs in vivo. CD31 (red signal) is an endothelial marker while α-SMA (green signal) is a smooth cell marker, and the combined expression of these two markers can help with accurate assessment of the formation of blood vessels [113]. As shown in Fig. S39, newly formed CD31⁺ vascular structures in DT-Exo-TMNs group exhibit significant α-SMA⁺ pericellular coverage (Fig. S39 a), indicating that DT-Exo-TMNs can facilitating the production of mature blood vessels [114]. Moreover, only a few newly formed blood vessels were observed in the AT and MNs groups. However, there was a large amount of neovascularization in the Achilles tendinopathy area of the DT-Exo-TMNs group, which was more than that in the TMNs and Exo-TMNs groups. Importantly, the number of mature vessels in DT-Exo-TMNs group was comparable to that in the normal group and 7 -times more than that in the AT group. These results indicated that DT-Exo-TMNs can promote angiogenesis in vitro.

Since DT-Exo-TMNs have the enhanced ability to promote the repolarization of M1 macrophages into M2 macrophages in vitro, the M1-M2 balance in the Achilles tendon region of the rat AT after DT-Exo-TMNs treatments were investigated using immunofluorescence staining (Fig. 8f). The CD206 fluorescence intensity signal in the AT group was weaker compared with that in the DT-Exo-TMNs group, accounting for only 4.8% of the DT-Exo-TMNs group. Conversely, the iNOS fluorescence intensity signal of the AT group was stronger than that of the DT-Exo-TMNs, which is 7.1-fold higher than the group of DT-Exo-TMNs (Fig. S40). These results suggested that more M2 macrophages were observed in the DT-Exo-TMNs group. Meanwhile, the expression of CD206 in Achilles tendon tissues after DT-Exo-TMNs treatment was upregulated, generating 2.05-fold higher in protein expression than that in the TMNs group, and the expression of iNOS was lower than that in other groups (Fig. 8g and Fig. S41). Importantly, higher tissue expression of TFAM was detected in tendons after DT-Exo-TMNs treatment, displaying a 0.75-fold higher expression than that in the TMNs group (Fig. S41). These results indicate that DT-Exo-TMNs can repolarize M1 macrophages into M2 in vivo by activating TFAM. As expected, the expression levels of inflammatory cytokines, such as IL-1β, was significantly lower in the DT-Exo-TMNs group compared with that in the AT group, DT-Exo-TMNs group accounting for 18.5% in the AT group (Fig. 8h). Similar results were also observed in the expression levels of IL-6 (Fig. 8i) and TNF-α (9j). The IL-6 in DT-Exo-TMNs group accounting for 42.5% in the AT group, the TNF-α in DT-Exo-TMNs group accounting for 8.9% in the AT group. Therefore, DT-Exo-TMNs group could effectively relieve inflammation. Many reported had shown that M2 macrophages can release transforming growth factor-beta (TGF-β), which can promote angiogenesis and tissue repair [115, 116]. As shown in Fig. 8k, the expression of TGF-β in the DT-Exo-TMNs group was 1.8-times higher than that in the AT group. These findings indicate that DT-Exo-TMNs facilitated tendon repair in vivo through the repolarization of M1 to M2 macrophages and the suppression of inflammation.

While previous studies have also explored immunomodulatory and alleviating inflammation strategies for tendinopathy, the therapeutic efficiency of DT-Exo-TMNs compares favorably in terms of treatment duration. For instance, Wang et al. reported that rosmarinic acid-loaded microneedles could induce M2 polarization through NLRP3 inflammasome inhibition [117]. However, the treatment cycle for AT rat lasts up to four weeks, which is two weeks longer than DT-Exo-TMNs. In addition, Shen et al. demonstrated that BHNPs@GA achieved a remarkable ROS clearance efficiency, yet required 8 to 12 weeks to significantly inhibit tendon calcification and restore collagen architecture [118]. Beyond treatment duration, the design of DT-Exo-TMNs also addresses key limitations of conventional exosome-based therapies. Although BMSC-derived exosomes have reported its ability for promote tendon regeneration by facilitating the proliferation and migration of endogenous tendon stem/progenitor cells [112], their delivery often relies on passive strategies, resulting in uncontrolled release and limited retention at the injury site. In contrast, the DT-Exo-TMNs system employs a dual-targeting engineered exosome strategy that enables active and precise delivery, ensuring selective uptake by M1 macrophages at the inflammatory site. Furthermore, the biomimetic bee-sting-like microneedle array enhances mechanical stability during body movement, potentially improving therapeutic efficacy in dynamic tendon environments. Collectively, these features position DT-Exo-TMNs as a more efficient and targeted approach for tendinopathy repair. The promising tissue repair observed with DT-Exo-TMNs in the collagenase-induced rat model of AT is encouraging. However, it is important to note that this acute injury model does not completely mirror the multifactorial pathophysiology of chronic, overuse-induced AT in humans, warranting further validation in more clinically relevant models.

Moreover, we performed gait analysis to assess pain management of DT-Exo-TMNs in AT. The separation of the prints of the hind and forepaws on the affected side and the changes in step length and stride length were used to reflect the claudication of the rat during walking. As shown in Fig. S42 a, after DT-Exo-TMNs treatment, the prints of the front and hind paws relatively overlapped, and the step length and stride length were more similar to those of the Normal group compared to other groups (Fig. S42 b), indicating that claudication was greatly relieved by DT-Exo-TMNs treatment. Moreover, we have performed biomechanical testing on Achilles tendons from after DT-Exo-TMNs treatment. As shown in Fig. S43, the results demonstrate that tendons in the DT-Exo-TMNs group exhibited significantly strong load-bearing capacity compared to other groups, which is 2.5-times higher in tensile strength than that in AT group. In addition, the tensile strength of tendons in DT-Exo-TMNs group are comparable to Normal tendons, which further confirms the restored of Achilles tendons mechanical properties following DT-Exo-TMNs treated.

Importantly, though HE staining of major organs (Fig. S44), blood tests (Table S3) and serum biochemical analysis (Fig. S45), we found that the DT-Exo-TMNs had good in vivo biological safety. We conducted an extended in vivo long-term biocompatibility study of DT-Exo-TMNs. After inserted DT-Exo-TMNs in the Achilles tendon region of rats for one-month, comprehensive blood tests showed no significant abnormalities in all key parameters compared to MNs (two weeks) groups (Table S4). These results indicated the absence of systemic inflammatory responses induced by the degradation products of DT-Exo-TMNs within one month. In addition, H&E staining of the Achilles tendon tissue after one-month insertion remained intact and healthy tissue structure compared to health group (Fig. S46), confirming that the gradual degradation of the DT-Exo-TMNs did not elicit local toxicity or adverse tissue remodeling. In conclusion, these extended in vivo results provide robust evidence that the degradation products of the DT-Exo-TMNs are biocompatible and do not induce systemic or local toxicity over the long term, supporting its safety profile for potential clinical application. Furthermore, serum biochemical analysis showed that liver function markers (GPT and GOT) and kidney function parameters (BUN) remained within normal ranges and were comparable to those of normal (Fig. S47). Collectively, these findings confirm that DT-Exo-TMNs exhibits no detectable immunotoxicity or adverse effects on major organs at the administered dose, highlighting its potential for clinical application.

Conclusion

In this study, we designed a specialized bee sting-shaped MNs (DT-Exo-TMNs) for joint application. This unique array ensured the stability of the MNs tip during body movement. By leveraging the transdermal delivery capabilities of the MNs, anti-inflammatory agents can be released to enhance the penetration depth of DT-Exo, offering a highly advantageous solution for treating AT. The DT-Exo-TMNs hydrogel MNs effectively stimulated M2 macrophage polarization by activating TFAM. Additionally, the photothermal agent, TA/Fe³⁺, embedded in the hydrogel MNs rapidly increased the temperature under NIR irradiation and significantly enhanced antibacterial activity against E. coli and S. aureus. In vivo experiments demonstrated that DT-Exo-TMNs accelerated the healing process of Achilles tendinopathy by promoting M2 polarization and exerting anti-inflammatory effects. In summary, antibiotic-free, biocompatible, and multifunctional DT-Exo-TMNs hydrogel MNs exhibit substantial potential in the immunotherapy of Achilles tendinopathy, offering a promising and effective therapeutic method for AT treatment.

In this study, we evaluated the numbers of M1 and M2 macrophages in human Achilles tendons at different stages of tendinopathy. While our preliminary data reveal that targeting M1 macrophages and regulating the inflammatory microenvironment may be potent therapies for AT, limited sample size (n = 3) and heightened sensitivity to interindividual differences should be considered. Future studies with larger human cohorts and rigorous selection criteria for AT patients are warranted to validate the functional alterations of M1 macrophages and their potential as therapeutic targets in AT.

We have evaluated the therapeutic effects of DT-Exo-TMNs for AT at a collagenase injection-induced AT rat model. Mechanical overuse methods that better simulate chronic overuse to more closely reflect the clinical reality of AT [33]. In addition, larger animal models offer superior physiological relevance to humans. From the perspective of translational medicine, non-human primates are undoubtedly the most ideal species for shoulder joint research, as they are the most similar to humans in anatomical structure and physiology. Future studies should evaluate DT-Exo-TMNs in larger AT animal models with Mechanical overuse methods to comprehensively assess the therapeutic efficacy and biosafety for clinical translation.

Finally, while we demonstrated that DT-Exo-TMNs promote M2 macrophage polarization via TFAM activation, the detailed molecular pathways and downstream effectors remain incompletely elucidated. Further mechanistic studies, such as transcriptomic or proteomic analyses, are warranted to uncover the comprehensive signaling networks involved in DT-Exo-mediated tendon repair.

Author contributions

**Muwei Hao: ** Writing - original draft, Writing - review & editing, Data curation, Formal analysis, Methodology, Project administration. **Kai Xiang: ** Software, Supervision, Visualization. **Keyi Zhang: ** Data curation, Validation. **Zheng Zhang: ** Software, Supervision. **Jinlong Li: ** Data curation, Software. **Han Sun: ** Writing-review & editing, Funding acquisition, Investigation. **Lei Zhang: ** Writing-review & editing, Conceptualization, Funding acquisition, Investigation, Project administration, Resources, Supervision.

Funding

This work was supported by Anhui Provincial Natural Science Foundation (2308085Y45), The Open Project of State Key Laboratory of Natural Medicines (SKLNMKF200310), Program for Excellent Sci-tech Innovation Teams of Universities in Anhui Province (2023AH010073), Wuhu Science and Technology Department project (2023jc22), Natural Science Research Project of Anhui Educational Committee (2024AH040237), High-level Talent Scientific Research Startup Fundation of Wannan Medical College (WYRCQD2023016), National Nature Science Foundation of China (32371408), China Postdoctoral Science Foundation (2023M740374), Anhui Province Young Key Teachers Overseas Visiting Scholars Project (JWFX2025029).

Data availability

Data will be made available on request.

Declarations

Competing interests

The authors declare no competing interests.