LIPUS combined with mild hyperthermia promotes rotator cuff healing via TRPV1/Ca²⁺/CaMKII-driven BACH2 nuclear translocation and inhibition of M1 polarization

Abstract

Background

Rotator cuff injuries often lead to impaired shoulder function, and tendon–bone healing remains challenging due to the complex structure of the enthesis and persistent remodeling-associated pro-inflammatory signaling. Low-intensity pulsed ultrasound (LIPUS) and mild hyperthermia (MH) have shown potential in tissue repair, but the molecular mechanisms underlying their combined effects are unclear.

Objective

This study aimed to investigate whether LIPUS and MH synergistically promote tendon–bone healing by regulating macrophage polarization through the TRPV1/Ca²⁺/CaMKII/BACH2 signaling pathway.

Methods

A rabbit supraspinatus tendon–bone injury model and in vitro macrophage polarization assays were employed. Histology, immunohistochemistry, MRI, and biomechanical testing were performed to assess repair outcomes. RNA-seq with qPCR and Western blot validation was used to identify key signaling pathways. Functional assays, including Ca²⁺ flux detection, pharmacological inhibition, and siRNA knockdown, were conducted to examine the role of TRPV1, CaMKII, and BACH2.

Results

LIPUS and MH significantly suppressed M1 polarization (CD86, iNOS) and enhanced M2 markers (CD163, Arg1), reducing ROS levels and inflammatory cytokine expression. Combined intervention yielded the most pronounced effects, showing improved fibrocartilage formation, collagen alignment, and bone remodeling, with superior biomechanical strength approaching native tissue. Transcriptomic and protein analyses revealed increased TRPV1 activation, enhanced Ca²⁺ influx, CaMKII phosphorylation, and promoted BACH2 nuclear translocation. Mechanistically, TRPV1 activation promoted Ca²⁺ entry, CaMKII activation, and BACH2 nuclear translocation, thereby inhibiting M1 polarization. Inhibition of TRPV1, Ca²⁺ influx, or CaMKII significantly reduced BACH2 nuclear translocation and diminished the anti-inflammatory effects.

Conclusion

Combined LIPUS and MH synergistically enhance tendon--bone healing, potentially through modulating the TRPV1/Ca²⁺/CaMKII/BACH2 pathway, which appears to modulate macrophage polarization during the later stages of tendon–bone remodeling.

The translational potential of this article

This study identifies the TRPV1/Ca²⁺/CaMKII/BACH2 signaling axis as a novel mechanistic target for non-invasive physical therapy in tendon-bone healing. Our findings provide a strong preclinical rationale for combining LIPUS and mild hyperthermia as a promising, non-pharmacological strategy to improve immune microenvironment and outcomes after rotator cuff repair. This paves the way for future clinical trials to translate this combinatorial biophysical therapy into adjuvant treatment protocols.

Graphical abstract

1. Introduction

Rotator cuff injury is one of the most common causes of shoulder pain and dysfunction, accounting for nearly 50% of shoulder disorders, and its prevalence increases markedly with age [1,2]. Although arthroscopic repair combined with rehabilitation has advanced in recent decades, high retear rates, poor structural integrity, and limited functional recovery remain major clinical challenges [3,4]. Successful tendon–bone healing depends on both the suppression of excessive inflammation and the promotion of angiogenesis and osteogenesis [5,6]. Following injury, M1 macrophages infiltrate the repair site and release pro-inflammatory cytokines such as IL-6 and TNF-α to initiate inflammation [7]. While this early response is necessary for repair initiation, sustained remodeling-associated M1-like signaling disrupts the local microenvironment, exacerbates tissue damage, and delays healing [8,9]. Therefore, strategies that regulate macrophage polarization and optimize the inflammatory milieu are critical for improving rotator cuff repair outcomes.

Physical therapies have attracted increasing attention due to their non-invasive nature and regenerative potential. Among them, low-intensity pulsed ultrasound (LIPUS) has been widely applied in fracture healing and shown to facilitate multiple stages of bone repair, including reducing inflammation, enhancing angiogenesis, promoting chondrogenesis, ossification, and bone remodeling [10]. LIPUS is thought to accelerate tissue repair partly by modulating inflammatory cytokines [11]. Recent studies further demonstrated its anti-inflammatory capacity; for example, Engelmann et al. [12] and Nagata et al. [13] reported that LIPUS combined with dimethyl sulfoxide effectively suppressed TNF-α expression in muscle injury models.

Mild hyperthermia (MH), by moderately increasing local temperature, improves microcirculation and cellular metabolism, thereby showing potential in inflammation regulation and tissue repair [14]. MH has been shown to increase local blood flow, enhance antioxidant capacity, and reduce the release of pro-inflammatory cytokines, thereby alleviating inflammation and pain [15]. In tissue repair, MH promotes cell proliferation, differentiation, and regeneration, facilitating wound healing, osteogenesis, and angiogenesis [16,17]. Moreover, MH plays an immunomodulatory role, as heat stimulation at 40–43 °C suppresses macrophage-derived inflammatory mediators and reduces inflammasome activation and mitochondrial stress [[18], [19], [20]]. However, its role and mechanisms in tendon–bone healing remain largely unexplored.

At the molecular level, the transient receptor potential vanilloid 1 (TRPV1) channel, a non-selective cation channel responsive to heat, acidity, and mechanical stress [21], has emerged as a key regulator of inflammation and tissue repair. TRPV1 activation induces Ca²⁺ influx, activating downstream kinases such as CaMKII, which influence autophagy, inflammation, and cell survival [[22], [23], [24]]. Recent evidence indicates that TRPV1 can attenuate M1 macrophage polarization via the Ca²⁺/CaMKII/Nrf2 pathway in the synovial microenvironment of osteoarthritis [25]. Given that TRPV1 is a polymodal receptor responsive to both thermal and mechanical stimuli, we postulated that it might also serve as a key mediator in the tendon-bone interface, where similar inflammatory and reparative events occur during healing. Another important regulator, BACH2, is a transcription factor essential for immune homeostasis. It has been implicated in CD8⁺ T cell differentiation, Th17 regulation, and repression of pro-inflammatory gene expression, thereby contributing to tissue regeneration [26]. Our transcriptomic analysis revealed a concurrent upregulation of TRPV1 and BACH2 alongside enriched calcium signaling. This led us to hypothesize that BACH2 might be functionally linked to the TRPV1/Ca²⁺/CaMKII axis. Nevertheless, whether BACH2 is directly regulated by this specific signaling pathway during tendon–bone healing remained unclear and constituted the focus of our mechanistic investigation. Nevertheless, whether BACH2 is directly regulated by TRPV1/Ca²⁺/CaMKII signaling during tendon–bone healing remains unclear [27]. Collectively, the TRPV1/Ca²⁺/CaMKII pathway plays a crucial role in injury repair, but the involvement of BACH2 and its precise regulatory mechanisms in tendon–bone healing remain to be fully elucidated.

Given their distinct yet potentially convergent mechanisms, the combination of LIPUS and MH may offer a synergistic strategy to enhance tendon-bone healing. LIPUS, as a source of controlled mechanical stimulation, can directly influence cellular behaviors and activate mechanotransduction pathways, including those mediated by ion channels such as TRPV1. In parallel, MH, as a localized thermal stimulus, can promote vasodilation and metabolic activity while directly activating thermosensitive channels like TRPV1. We hypothesize that their combined application synergistically amplifies TRPV1 activation and downstream Ca²⁺ signaling more effectively than either modality alone. This concerted action is postulated to create a potent immunomodulatory stimulus, robustly driving anti-inflammatory responses—such as BACH2-mediated inhibition of M1 macrophage polarization—to optimize the tendon–bone healing microenvironment.

In this study, we investigated the role of TRPV1 in tendon–bone healing and macrophage polarization using in vivo and in vitro models. We found that combined intervention with LIPUS and MH significantly enhanced TRPV1 expression and activation, leading to reduced M1 macrophage polarization and improved tendon–bone integration. Mechanistically, TRPV1 activation facilitated Ca²⁺ influx and triggered the CaMKII/BACH2 signaling cascade, promoting BACH2 nuclear translocation and downregulating pro-inflammatory markers such as iNOS and IL-6. These findings highlight TRPV1 as a key mechanosensitive and thermosensitive regulator of inflammatory responses and provide evidence for a novel immunomodulatory strategy to accelerate tendon–bone interface healing.

2. Materials and methods

2.1. Animal model and ethical compliance

2.1.1. Rationale for the use of the rabbit model

The New Zealand white rabbit was selected because it is one of the most widely used and well-characterized models for rotator cuff research. The rabbit supraspinatus tendon–bone insertion exhibits enthesis architecture and fibrocartilage zonal structure similar to those of humans, supporting translational investigation of tendon–bone interface biology [28,29]. This species also permits the creation of a reproducible supraspinatus detachment–repair model with stable postoperative healing and rapid fibrocartilage and mineralized interface formation [30]. In addition, rabbits demonstrate predictable biomechanical recovery patterns that parallel key phases of human rotator cuff healing, making them suitable for evaluating mechanical outcomes [31]. Owing to these anatomical and biological advantages, the rabbit model has been extensively applied to study therapeutic strategies and mechanistic regulation of tendon–bone healing [32].

All procedures were approved by the XXX University Institutional Animal Care and Use Committee (XXX) and adhered to the NIH guidelines for the care and use of laboratory animals. Forty-eight male SPF-grade New Zealand white rabbits (12 weeks old, 2.5–3.0 kg) were obtained from the Peking University Medical Animal Facility and housed in a barrier-controlled environment with a 12-h light/dark cycle, 24 ± 1 °C, 50 ± 10% humidity, and ad libitum access to standard chow and sterile water. A unilateral supraspinatus tendon–bone insertion injury model was established, followed by transosseous repair. Animals were randomized into four groups (n = 12 per group): Control (sham), mild hyperthermia (T, 41.5 °C for 20 min/day), low-intensity pulsed ultrasound (LIPUS, 1.5 MHz, 100 mW/cm², 20 min/day), and combination therapy (LIPUS + T). Treatments commenced 3 days postoperatively to target the transition phase from acute inflammation to early tissue repair, a key window for immunomodulation. Animals were euthanized at 6 weeks and 12 weeks postoperatively for histological, immunohistochemical, and molecular analyses. At each sacrifice time point (6 and 12 weeks), the animals from each group (n = 6) were allocated for analysis: three were used for biomechanical testing, and three for histological, immunohistochemical, and imaging evaluations. Gross morphological observation, histological staining, immunohistochemical staining, biomechanical testing, and imaging evaluation were performed to compare the effects of different interventions on supraspinatus tendon-to-bone healing in rabbits.

2.2. Surgical procedure: supraspinatus tendon-to-bone injury model

Animals were anesthetized using 3–5% isoflurane delivered via a small animal anesthesia machine. The operative limb was shaved, disinfected with povidone-iodine, and draped in sterile fashion. A 3 cm longitudinal skin incision was made along the axis from the acromion to the greater tuberosity. Muscle layers were bluntly dissected to expose the supraspinatus tendon insertion. The tendon was sharply detached from the greater tuberosity using a #11 scalpel blade. A transosseous tunnel was created in the greater tuberosity using a 1.0 mm drill bit. The detached tendon was re-sutured using 4–0 absorbable sutures (Vicryl®), passed through the bone tunnel and tied on the lateral humeral cortex under physiological tension. The wound was irrigated with saline and closed in layers. Postoperative analgesia was provided with 10% gentamicin injection (1 mL/kg, intramuscularly, once daily for 3 days).

2.3. Physical therapy protocols

2.3.1. Rationale for parameter selection

The specific parameters for LIPUS and MH were chosen based on systematic in vitro screening performed in our laboratory prior to this in vivo study. For LIPUS, a comparison between 50 and 100 mW/cm² indicated that 100 mW/cm² was more effective in promoting reparative cellular responses. For MH, a temperature gradient study (37–42 °C) identified 41 °C as the optimal temperature for enhancing cell proliferation. Therefore, in this study, the LIPUS intensity was set at 100 mW/cm² (1.5 MHz frequency ensured deep tissue penetration). The MH surface temperature was set at 41.5 °C to achieve an estimated intramuscular temperature of ∼41.0 °C, delivering this optimized thermal stimulus to the tendon-bone interface. This combination was hypothesized to synergistically create a pro-regenerative microenvironment conducive to healing.

Mild hyperthermia (T): A thermostatically controlled local heating pad (±0.1 °C, 5 × 5 cm area) was applied over the surgical site for 20 min/day. To deliver a mild hyperthermic stimulus, the surface temperature was set at 41.5 °C. Based on established tissue heat transfer modeling (Q = ΔT × λ × S/L), this setting was calculated to yield an estimated intramuscular temperature of approximately 41.0 °C, which is within the standard therapeutic range for mild hyperthermia. LIPUS: Delivered using a signal generator and custom coupling probe with output parameters of 1.5 MHz frequency, 100 mW/cm² intensity, 1 kHz pulse repetition, 200 μs pulse duration, and 20 min exposure. The ultrasonic intensity was set at 100 mW/cm², a level higher than the commonly used 30 mW/cm² for fracture healing, to ensure adequate energy delivery to the deeper tendon-bone interface and to provide a potent stimulus for tissue regeneration and immunomodulation based on preliminary evidence [33]. Coupling gel was applied to ensure energy transmission. The two treatments were applied sequentially each day: LIPUS was delivered first for 20 min, followed immediately by mild hyperthermia for 20 min, with no interval between modalities. The specific operation of the intervention is shown in Supplementary Fig. 1.

2.4. Rationale for intervention timing

All therapeutic interventions commenced on postoperative day 3. This time point was selected based on the following considerations: (1) Pilot histological analysis indicated that the acute inflammatory phase in the rabbit supraspinatus repair model peaked around 48 h and began transitioning to a proliferative stage by day 3, defining a critical window for modulating early immune responses. (2) The “Mild hyperthermia” (41.5 °C) applied in this study is conceptualized as a biophysical co-stimulus alongside LIPUS, primarily intended to synergistically activate intracellular signaling pathways (e.g., TRPV1/Ca²⁺) involved in macrophage polarization, rather than as a conventional thermal therapy for vasodilation. (3) Emerging evidence suggests that transitioning to localized heat therapy in the early postoperative period can be beneficial for pain management and tissue perfusion [34]. This combined regimen aims to explore the potential of targeting this early phase to actively guide the healing process towards improved tendon-bone integration. This early intervention (postoperative day 3) targets the peak acute inflammatory window, ensuring that immunomodulatory effects are initiated before sustained pro-inflammatory signaling during the remodeling transition.

2.5. Cell culture and polarization of macrophages

RAW264.7 murine macrophages (Beijing Beyotime Biotechnology, C7505) were cultured in high-glucose DMEM (Gibco) with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin in 6-well plates at 1 × 10⁶ cells/well under standard conditions (37 °C, 5% CO₂). M1 polarization: Induced by 50 ng/mL LPS (Sigma–Aldrich) for 24 h. M2 polarization: Induced by 50 ng/mL recombinant mouse IL-10 (PeproTech) for 24 h. Cells were then exposed to: Mild hyperthermia: 41 °C in incubator for 20 min. LIPUS: Same protocol as above. LIPUS + T: Sequential application of both.

2.6. Quantitative real-time PCR

Total RNA was extracted from cells using TRIzol reagent (Invitrogen) and reverse-transcribed using the PrimeScript™ RT Reagent Kit (Takara). qPCR was performed using TB Green® Premix Ex Taq™ II (Takara) in a StepOnePlus™ system (Applied Biosystems). mRNA expression levels of CD86, iNOS, CD163, and Arg1 were normalized to GAPDH, and relative expression was calculated using the 2^^–ΔΔCt method. Each sample was tested in triplicate. Primer sequences are provided in Supplementary Table 1.

2.7. Flow cytometry for macrophage surface markers

Cells were stained with FITC or PE-conjugated antibodies against: M1 markers: CD80, CD86. M2 markers: CD163, CD206. Following 30 min incubation at 4 °C in the dark, cells were washed with PBS and analyzed using BD FACSCanto™ II. Data were processed using FlowJo V10.

2.8. ROS detection

Intracellular reactive oxygen species were measured using DCFH-DA probe (Beyotime, S0033). After staining, fluorescence was captured with a Leica DMi8 fluorescence microscope and quantified using ImageJ. To ensure that the recorded Fluo-4 AM signals originated from intracellular compartments of viable and adherent cells, each field was briefly examined under phase-contrast or bright-field mode prior to fluorescence acquisition to confirm cell confluency and morphology. Images showing significant cell detachment or abnormal morphology were excluded from analysis.

2.9. Immunofluorescence imaging

Cells were fixed with 4% paraformaldehyde, permeabilized with 0.2% Triton X-100, blocked with 5% BSA, and incubated with primary antibodies (anti-CD86, anti-CD163), followed by Alexa Fluor® 488/594-labeled secondary antibodies. Nuclei were counterstained with DAPI. Images were acquired with PerkinElmer Operetta CLS.

2.10. Histology and immunohistochemistry

Tissue samples were harvested at 6 weeks and 12 weeks post-surgery for immunohistochemical analysis. Harvested tendon–bone interface tissues were fixed in 4% paraformaldehyde, decalcified in 10% EDTA, and embedded in paraffin. 5 μm sections were stained with: H&E for cellularity and collagen architecture. Immunohistochemistry was performed using antibodies against IL-1β, TNF-α (M1 markers), IL-10, and TGF-β (M2 markers), followed by HRP-conjugated secondary antibodies and DAB development. Quantification was done by integrated optical density analysis (Image-Pro Plus).

2.11. Biomechanical testing

Biomechanical testing was performed on n = 3 supraspinatus–humerus complexes per group at each time point (6 and 12 weeks). Supraspinatus–humerus complexes were prepared by dissecting excess tissue and fixing the humeral shaft with two K-wires. The tendon was clamped in a customized soft tissue holder. A preload of 5 N was applied for 10 min, followed by 10 preconditioning cycles (5–50 N, 15 N/s). Samples were then loaded to failure at 1 mm/min on a uniaxial testing machine (Instron 5543). Maximum failure load (N) was recorded.

2.12. MRI and signal-to-noise quotient analysis

MRI scans were performed on a 3.0T system (United Imaging OMEGA) using T2-weighted imaging. Signal intensity was measured within a standardized circular ROI (area: X mm²) placed at the center of the repair site and at the mid-substance of the native supraspinatus tendon, avoiding areas of obvious artifact or blood vessels. All ROI placements and intensity measurements were performed by two independent researchers blinded to the treatment groups. The average of three measurements per region was used for calculation. The SNQ was calculated as:

Where S_repair, S_native, and S_background represent signal intensities at the repair site, native tendon, and background, respectively.

2.13. RNA sequencing and bioinformatics

Total RNA (≥1 μg/sample) was extracted, and quality assessed using Nanodrop 2000 and Agilent Bioanalyzer 2100. RNA-seq libraries were prepared using NEBNext® Ultra II Directional RNA Library Prep Kit (Illumina-compatible), and sequenced using Illumina PE150 platform. Clean reads were aligned to the Oryctolagus cuniculus reference genome using HISAT2 (v2.1.0). Gene expression was normalized as FPKM using HTSeq. Differential expression analysis was performed with DESeq2 (v1.38.3). Genes with |log₂FoldChange| > 1 and P < 0.05 were considered differentially expressed. Gene Ontology (GO), KEGG, and GSEA enrichment were performed using cluster Profiler and top GO packages. Primer sequences are provided in Supplementary Table 2.

2.14. Western blotting and siRNA interference

Nuclear and cytoplasmic proteins were extracted using NE-PER kit (Thermo Fisher). Western blotting was conducted for TRPV1, CaMKII/p-CaMKII, BACH2, with GAPDH or Lamin B1 as internal controls. For knockdown experiments, RAW264.7 cells were transfected with 50 nM BACH2-specific siRNA (sense: 5′-UUCGUCUGUCCAACCAGUG (dT) (dT)-3′) using Lipofectamine 3000. Effects were validated by qPCR and immunoblotting.

2.15. Co-immunoprecipitation (Co-IP) assay

To examine whether CaMKII associates with BACH2 in a TRPV1- and CaMKII-dependent manner, co-immunoprecipitation assays were performed. RAW264.7 macrophages were divided into four groups: Control, LIPUS + mild hyperthermia (LIPUS + T), Capsazepine (CAP) + LIPUS + T, and the CaMKII inhibitor KN-93 + LIPUS + T. Following the indicated treatments, whole-cell lysates were prepared using IP lysis buffer supplemented with protease and phosphatase inhibitors. Equal amounts of protein were incubated overnight at 4 °C with antibodies against CaMKII, followed by incubation with protein A/G agarose beads. Immunoprecipitates were washed thoroughly, eluted with SDS sample buffer, and subjected to SDS–PAGE and immunoblotting using anti-BACH2 antibodies. Input lysates were analyzed in parallel to verify protein expression.

2.16. Statistical analysis

All data are presented as mean ± standard deviation (SD) with each group tested at least in triplicate. One-way analysis of variance (ANOVA) was used for comparisons among three or more groups. Prior to ANOVA, normality was assessed using the Shapiro–Wilk test and homogeneity of variances was evaluated using Levene's test. If variances were homogeneous, the least significant difference (LSD) method was used for post hoc comparisons; if not, Dunnett's T3 test was applied. Groups showing significant differences were further analyzed using Tukey's post hoc test. Data analysis was performed using SPSS version 27.0, GraphPad Prism version 10.1.2, and ImageJ. Figures were edited and arranged using Microsoft PowerPoint and Adobe Illustrator. A P-value <0.05 was considered statistically significant.

3. Results

3.1. LIPUS combined with mild hyperthermia synergistically suppresses M1 macrophage polarization and promotes M2 macrophage polarization in vitro

In vitro assays reflected early macrophage polarization (24 h), while in vivo analyses at 6 and 12 weeks evaluated sustained phenotypic regulation during the remodeling phase. Both qPCR and flow cytometry analyses demonstrated that the combined intervention of LIPUS and MH significantly modulated macrophage polarization (Fig. 1A–H). Specifically, the combined treatment most effectively suppressed M1-type polarization, as evidenced by decreased mRNA and protein expression of CD86 and iNOS, reduced CD80+/CD86+ cell populations, lower ROS production, and diminished CD86 immunofluorescence intensity compared to other groups (all P < 0.01). Conversely, M2-type markers, including CD163 and Arg1 mRNA expression, CD163+/CD206+ cell ratios, and CD163 fluorescence intensity, were significantly elevated in the combined group, indicating enhanced anti-inflammatory polarization (all P < 0.001). While LIPUS or MH alone showed moderate effects, their combination yielded the most pronounced regulation of macrophage phenotype and inflammatory status (Fig. 1I–N). The above results indicate that LIPUS and mild hyperthermia effectively suppress pro-inflammatory responses and enhance anti-inflammatory functions by modulating macrophage polarization.

Fig. 1.

Induction Results of Macrophage Polarization by LIPUS and Mild Hyperthermia Therapy.

A, B, C, D represent the mRNA expression statistical analysis of CD86, iNOS, CD163, and Arg1 in each group, respectively. E and F show the polarization status and statistical analysis of M1 polarization markers CD80 and CD86 in each group. G and H show the polarization status and statistical analysis of M2 polarization markers CD163 and CD206 in each group. I and J display immunofluorescence staining images of CD86 and CD163 in each group. K shows the reactive oxygen species (ROS) fluorescence images for each group. L and M represent statistical analysis of the fluorescence intensity of CD86 and CD163. N is a bar chart of the positive rate statistical analysis. Scale bars: 20 μm. N = 3, * indicates P < 0.05, ** indicates P < 0.01, *** indicates P < 0.001.

3.2. Combined LIPUS and mild hyperthermia enhances histological structure, reduces inflammation, and improves biomechanical strength during tendon-to-bone healing in rabbits

To further investigate the synergistic effects of LIPUS and MH on tendon-to-bone healing, we established a rabbit supraspinatus repair model. Animals were euthanized at 6 and 12 weeks postoperatively for comprehensive evaluation, including gross morphology, histology, immunohistochemistry, biomechanical testing, and MRI. The results below illustrate the differential healing outcomes among control, LIPUS alone, MH alone, and the combined intervention groups.

Histological, immunohistochemical, biomechanical, and MRI analyses collectively demonstrated that the combination of LIPUS and MH significantly promoted tendon-to-bone healing in a rabbit supraspinatus repair model. At both 6 and 12 weeks postoperatively: Histology revealed that the combined group exhibited minimal inflammatory infiltration, organized collagen alignment, and well-formed fibrocartilage transition zones, significantly better than the control, MH, and LIPUS groups (Fig. 2A). Although these histological findings are encouraging, they reflect structural improvement rather than complete functional regeneration, and should be interpreted in conjunction with biomechanical and molecular analyses. Immunohistochemistry showed that the combined group had the lowest expression of M1 markers IL-1β and TNF-α (P < 0.01 vs. other groups), and the highest expression of M2 markers IL-10 and TGF-β (P < 0.01 vs. other groups), indicating effective immunomodulation (Fig. 2B–E), and representative images are shown in Supplementary Fig. 2. MRI findings indicated improved soft tissue continuity and reduced interface signal intensity in the combined group. SNQ values were significantly lower than those in the control group at both time points (P < 0.001), and also lower than in MH (P < 0.01) and LIPUS groups (P < 0.05), reflecting superior healing and reduced inflammation (Fig. 2F and G). Biomechanical testing demonstrated significantly higher maximum failure load in the combined group at 6 weeks (Control: 45.00 ± 5.00 N; MH: 67.67 ± 2.52 N; LIPUS: 73.33 ± 1.53 N; LIPUS + MH: 89.67 ± 4.51 N; P < 0.001 vs. all groups) and 12 weeks (Control: 75.00 ± 5.00 N; MH: 105.33 ± 5.03 N; LIPUS: 114.33 ± 8.14 N; LIPUS + MH: 126.67 ± 1.53 N; Normal: 133.00 ± 2.65 N). At 12 weeks, the LIPUS + MH group was significantly stronger than control (P < 0.001), MH and LIPUS alone (both P < 0.01), and although the mechanical strength remained slightly lower than that of the normal tendon-bone complex (P = 0.0678 vs. Normal) (Fig. 2H). In summary, the combined LIPUS + MH intervention provided synergistic effects, significantly enhancing histological structure, anti-inflammatory response, mechanical strength, and imaging-based healing quality (P < 0.05 or better across assessments.

Fig. 2.

Combined therapy improves structural and biomechanical tendon–bone healing.

A shows the HE staining results at different time points for each group. B, C, D, and E represent the immunohistochemical results for IL-1β, TNF-α, IL-10, and TGF-β, respectively. F shows the MRI scan images for each group at 6 and 12 weeks. G presents the statistical results of SNQ. H shows the statistical analysis of biomechanics (maximum failure load) at different time points for each group. Scale bars: 20 μm. N = 3, ns indicates no statistical significance, * indicates P < 0.05, ** indicates P < 0.01, *** indicates P < 0.001.

3.3. Transcriptomic profiling identifies calcium signaling and TRPV1/BACH2 axis as key regulators underlying the therapeutic effects of LIPUS plus mild hyperthermia

Previous findings demonstrated that while LIPUS or MH alone can modestly enhance tendon-to-bone healing, their combined application exerts a significantly greater therapeutic effect. To explore the molecular basis of this synergy, we performed RNA sequencing (RNA-seq) to identify key differentially expressed genes (DEGs) and enriched signaling pathways involved in the healing process.

Transcriptomic profiling revealed distinct gene expression signatures among the three groups. Compared with the control group, 993 DEGs were identified in the MH group, 163 in the LIPUS group, and 921 in the LIPUS + MH group, with 38 overlapping DEGs (35 upregulated, 3 downregulated). GO enrichment analysis indicated that DEGs in the LIPUS + MH group were primarily involved in biological processes such as extracellular matrix organization, tissue development, and cell differentiation. KEGG pathway analysis highlighted significant enrichment in calcium signaling, PI3K-Akt, and TGF-β pathways (Table 1, Table 2, Table 3, Fig. 3A–F).

Table 1.

Differentially expressed genes analysis results.

Group Comparison	Upregulated Genes	Downregulated Genes	Differentially Expressed Genes
C vs. T	533	460	993
C vs. LIPUS	115	48	163
T vs. LIPUS	729	192	921

Table 2.

Top differentially expressed genes identified in the LIPUS + MH vs Control comparison.

No.	Gene Name	Regulation	Function
1	TRPV1	Upregulated	Sensory receptor for detecting external high temperatures, acidic environments, and mechanical stimuli.
2	COL1A1	Upregulated	Major component of tendon and bone tissue.
3	LOXL4	Upregulated	Involved in collagen crosslinking, contributing to tissue stability and healing.
4	WNK3	Upregulated	Involved in ion transport and cell signaling.
5	COL1A2	Upregulated	Works synergistically with COL1A1 in tendon-bone healing.
6	BACH2	Upregulated	Immune regulation and inflammation control.
7	ADAMTS4	Upregulated	Plays a role in matrix remodeling, involved in tissue repair after injury.
8	HOXD11	Upregulated	Plays a critical role in regulating limb skeletal development.
9	SSTR2	Upregulated	Regulates intracellular calcium concentrations by inhibiting calcium channels, indirectly controlling calcium ion signaling.
10	CAMK1G	Upregulated	Calcium-dependent kinase, regulates intracellular calcium signal transduction.

Table 3.

KEGG pathway enrichment analysis of DEGs in the LIPUS + MH vs Control comparison.

Pathway ID	Pathway	Upregulated Genes	Downregulated Genes	Total Up/Downregulated Genes	Total Genes	P-value
map04020	Calcium signaling pathway	7	17	24	172	0.000190255
map04060	Cytokine-cytokine receptor interaction	22	3	25	190	0.000355837
map05412	Arrhythmogenic right ventricular cardiomyopathy	5	6	11	60	0.00118037
map05205	Proteoglycans in cancer	10	11	21	168	0.00201472
map04151	PI3K-Akt signaling pathway	18	11	29	263	0.00226061
map04512	ECM-receptor interaction	6	3	9	70	0.00490243
map04974	Protein digestion and absorption	6	4	10	88	0.0074913
map04024	cAMP signaling pathway	6	9	15	164	0.00967022
map04640	Hematopoietic cell lineage	5	1	6	239	0.0115226
map04080	Neuroactive ligand−receptor interaction	13	10	23	242	0.0324538

Fig. 3.

Transcriptomic profiling identifies TRPV1/Ca²⁺ signaling as a key pathway activated by combined therapy.

A and B show the volcano plots of differentially expressed genes for the T group and LIPUS group compared with the control group, where red and blue points represent upregulated and downregulated genes, respectively, and gray points represent genes with no significant differential expression. C and D display the GO enrichment results for the T group and LIPUS group compared with the control group. E and F present the KEGG enrichment results for the T group and LIPUS group compared with the control group. The vertical axis represents pathway names, while the horizontal axis represents the corresponding P value for each pathway. The size of the points indicates the number of differentially expressed genes in each pathway, and the color of the points represents the Rich factor, with larger values shown in red. G - P represent statistical analysis of the relative mRNA expression levels for TRPV1, BACH2, COL1A1, LOXL4, WNK3, COL1A2, ADAMTS4, HOXD11, SSTR2, and CAMK1G. Q shows the protein blot bands for TRPV1, BACH2, and GAPDH. R and S represent the normalized relative expression levels of proteins BACH2 and TRPV1 (with the Control value set to 1). N = 3, ns indicates no statistical significance, * indicates P < 0.05, ** indicates P < 0.01, *** indicates P < 0.001, **** indicates P < 0.0001.

To validate key transcriptomic findings, qPCR was performed on representative genes. TRPV1, BACH2, COL1A1, COL1A2, LOXL4, WNK3, ADAMTS4, HOXD11, and CAMK1G showed significant upregulation in the LIPUS and MH groups compared to control (P < 0.01 or P < 0.001), consistent with RNA-seq results (Fig. 3G–P). Among these, TRPV1 and BACH2 were particularly elevated, suggesting roles in sensory signaling and transcriptional regulation. Further Western blot analysis confirmed that both TRPV1 and BACH2 protein levels were significantly higher in the LIPUS and MH groups than in control (P < 0.01), with LIPUS exerting a greater effect (Fig. 3Q–S). These findings support the notion that LIPUS and MH modulate tendon-bone healing at the molecular level by enhancing matrix remodeling, reducing inflammation, and activating regenerative signaling cascades.

3.4. TRPV1/Ca²⁺/CaMKII/BACH2 signaling axis mediates the inhibitory effects of LIPUS plus mild hyperthermia on M1 macrophage polarization

TRPV1 is a calcium-permeable channel sensitive to heat, acidic pH, and mechanical stimuli, playing key roles in inflammation and pain regulation [25]. Its activation promotes Ca²⁺ influx and triggers downstream pathways such as CaMKII, with activity modulated by kinases including PKA and PKC [35,36]. Beyond sensory functions, TRPV1 contributes to immune regulation, including T cell activation and macrophage polarization [[37], [38], [39]]. Its role in M1 polarization remains debated, though some studies suggest TRPV1 activation suppresses iNOS and IL-6 expression, thereby attenuating inflammation [40]. BACH2 is a transcriptional repressor in immune cells that normally resides in the cytoplasm but translocates to the nucleus upon stimulation, where it inhibits inflammatory gene transcription [[41], [42], [43], [44], [45], [46], [47]]. However, its role in tendon–bone healing and its relationship with TRPV1–calcium signaling remain unclear.

Based on our RNA-seq and in vivo experimental findings, we hypothesize that the combination of LIPUS and MH activates TRPV1 channels, leading to increased intracellular Ca²⁺ influx. This in turn may activate downstream CaMKII signaling, promoting the nuclear translocation of BACH2. Consequently, this cascade may suppress M1 macrophage polarization and reduce local inflammation, ultimately facilitating tissue repair at the tendon–bone interface. To test this hypothesis, the study was structured into four experimental modules: (1) TRPV1 and M1 Polarization: To evaluate whether activation of TRPV1 by mild hyperthermia and LIPUS suppresses M1 macrophage polarization and associated inflammatory responses. (2) TRPV1–BACH2 Axis: To determine whether TRPV1 activation exerts anti-inflammatory effects by promoting nuclear translocation of BACH2 and inhibiting M1 macrophage polarization. (3) Ca²⁺ Dependency: To investigate whether calcium influx triggered by TRPV1 activation is required for BACH2 nuclear translocation. (4) Role of CaMKII: To explore whether CaMKII acts as a key mediator in the TRPV1/Ca²⁺-dependent regulation of BACH2 nuclear translocation.

• (1)TRPV1 activation is required for LIPUS- and MH-mediated suppression of M1 macrophage polarization

To investigate the role of TRPV1 in the regulation of M1 macrophage polarization induced by LIPUS and mild hyperthermia, we utilized LPS-stimulated RAW264.7 macrophages and employed capsazepine (CAP), a specific TRPV1 inhibitor, for functional validation. The modulatory effects of LIPUS and mild hyperthermia were systematically evaluated at the mRNA, protein, and phenotypic levels. Both LIPUS and mild hyperthermia significantly suppressed the expression of M1 polarization markers, including CD86, iNOS, and CD80, while upregulating TRPV1 mRNA levels (Fig. 4A–C, G, H, K, L). Furthermore, Western blot analysis revealed increased expression of TRPV1 and BACH2 proteins in both treatment groups, with LIPUS showing the most pronounced effect (Fig. 4D–F). The inhibitory effects on M1 polarization were partially reversed following co-treatment with the TRPV1 inhibitor CAP, confirming the central regulatory role of TRPV1 in this process. Notably, expression of CaMKII protein was also markedly reduced in the LIPUS and mild hyperthermia groups (Fig. 4I and J), suggesting its potential involvement in a downstream negative feedback mechanism of TRPV1 signaling.

Fig. 4.

TRPV1 activation is required for suppression of M1 macrophage polarization.

A, B, C represent the statistical analysis of mRNA expression for CD86, iNOS, and TRPV1 in each group. D shows the electrophoresis bands of BACH2 and TRPV1 protein expression levels in each group. E and F represent the statistical analysis of protein expression levels of BACH2 and TRPV1 in each group. G shows the electrophoresis bands of iNOS protein expression levels in each group. H represents the statistical analysis of iNOS protein expression levels in each group. I shows the electrophoresis bands of CaMKII protein expression levels in each group. J represents the statistical analysis of CaMKII protein expression levels in each group. K shows the statistical analysis of CD80 by flow cytometry in each group. L represents the flow cytometry results for CD80 in each group. N = 3, ns indicates no statistical significance. * indicates P < 0.05, ** indicates P < 0.01, *** indicates P < 0.001.

Collectively, these findings indicate that both LIPUS and mild hyperthermia inhibit pro-inflammatory M1 macrophage polarization by upregulating TRPV1 expression and activating its downstream signaling pathway. LIPUS exhibited superior efficacy in modulating macrophage polarization, highlighting TRPV1 as a key mechanosensitive and thermosensitive target mediating the immunomodulatory effects of physical therapies.

To further provide genetic evidence supporting the specificity of TRPV1 involvement and to exclude potential limitations associated with pharmacological inhibition, we next performed siRNA-mediated knockdown of TRPV1 in LPS-stimulated RAW264.7 macrophages and assessed the responses to LIPUS combined with mild hyperthermia (LIPUS + T). As shown in Fig. 5A and B, TRPV1 protein levels were markedly reduced in the siTRPV1 group compared with the negative control (siNC), confirming efficient knockdown. Under siNC conditions, LIPUS + T robustly increased CaMKII phosphorylation while significantly suppressing iNOS expression, whereas these effects were largely abolished following TRPV1 knockdown (Fig. 5A–D), indicating that TRPV1 is required for LIPUS + T–induced activation of downstream signaling and inhibition of M1-associated markers. Given that BACH2 nuclear translocation represents a key mechanistic event in our proposed pathway, we further performed nuclear–cytoplasmic fractionation and found that LIPUS + T promoted BACH2 nuclear accumulation in siNC cells; however, this nuclear translocation was substantially blocked in siTRPV1-transfected cells (Fig. 5E–G). Consistent with these protein-level changes, qPCR analysis demonstrated that LIPUS + T significantly downregulated the M1-associated genes iNOS and CD86 in siNC cells, while TRPV1 knockdown eliminated this transcriptional suppression (Fig. 5H–J). Collectively, these results provide genetic evidence that TRPV1 is indispensable for LIPUS + T–induced CaMKII activation, BACH2 nuclear translocation, and subsequent inhibition of M1 macrophage polarization.

• (2)TRPV1 activation promotes BACH2 nuclear translocation to inhibit M1 macrophage polarization

Fig. 5.

TRPV1 promotes BACH2 nuclear translocation to inhibit M1 polarization.

RAW264.7 macrophages were transfected with negative control siRNA (siNC) or TRPV1-specific siRNA (siTRPV1), followed by LPS stimulation and subsequent treatment with low-intensity pulsed ultrasound combined with mild hyperthermia (LIPUS + T), as indicated. A Representative Western blot images showing the expression of TRPV1, phosphorylated CaMKII (p-CaMKII), total CaMKII, and iNOS under different conditions. GAPDH was used as a loading control. B, C, D Densitometric quantification of TRPV1, p-CaMKII, and iNOS protein levels, normalized to GAPDH and expressed relative to the siNC + LPS group. E Representative Western blot images of BACH2 expression in cytoplasmic and nuclear fractions. GAPDH and Histone H3 were used as cytoplasmic and nuclear loading controls, respectively. F, G Quantitative analysis of cytoplasmic and nuclear BACH2 protein levels. H, I, J Quantitative PCR analysis of TRPV1, iNOS, and CD86 mRNA expression levels. Data were normalized to GAPDH and expressed relative to the siNC + LPS group using the 2^–ΔΔCt method. N = 3, ns indicates no statistical significance. ns, not significant, * indicates P < 0.05, ** indicates P < 0.01, *** indicates P < 0.001.

Based on the aforementioned findings, we further investigated how TRPV1 mediates the anti-inflammatory effects of LIPUS and mild hyperthermia, with a particular focus on its regulatory role on the transcription factor BACH2. As previously reported, BACH2 exerts anti-inflammatory effects by modulating immune cell activity: it inhibits excessive T cell activation, suppresses the differentiation of pro-inflammatory cells, downregulates pro-inflammatory cytokine expression, maintains immune tolerance, and regulates B cell function, thereby mitigating autoimmune responses.

Under basal conditions, BACH2 predominantly resides in the cytoplasm. Upon activation, it translocates to the nucleus to exert its anti-inflammatory functions. To evaluate this process, we examined nuclear BACH2 levels following TRPV1 activation. Western blot analysis revealed that nuclear BACH2 expression was significantly increased in both the mild hyperthermia (T) and LIPUS groups compared with the control group (P < 0.05), whereas siRNA-mediated TRPV1 knockdown markedly suppressed this nuclear translocation (P < 0.01) (Fig. 6A–C). These results suggest that TRPV1 activation promotes BACH2 nuclear translocation. Consistently, immunofluorescence analysis demonstrated a significant increase in nuclear BACH2 signal intensity and nuclear localization rate in the T and LIPUS groups (P < 0.001), which was markedly diminished in the siRNA group, returning to levels similar to the control group (Fig. 6D and E). These findings further confirmed that TRPV1 activation facilitates the translocation of BACH2 from the cytoplasm to the nucleus—a process essential for inhibiting M1 macrophage polarization and mediating the anti-inflammatory effects of physical therapies.

Fig. 6.

Calcium influx mediates TRPV1-dependent BACH2 nuclear localization.

A shows the electrophoresis bands of BACH2 protein expression levels in the cytoplasm and nucleus for each group. B and C represent the statistical analysis of BACH2 protein expression levels in the cytoplasm and nucleus for each group. D shows the immunofluorescence analysis of BACH2 nuclear translocation in each group. E presents the quantitative statistical analysis of BACH2 nuclear translocation for each group. F and G represent the statistical analysis of mRNA expression levels for CD86 and iNOS in each group. H shows the electrophoresis bands of iNOS protein expression levels in each group. I represents the statistical analysis of iNOS protein expression levels for each group. Scale bars: 25 μm. N = 3, ns indicates no statistical significance. * indicates P < 0.05, ** indicates P < 0.01, *** indicates P < 0.001.

Subsequent qPCR and Western blot analyses showed that CD86 and iNOS mRNA levels were significantly downregulated in both the T and LIPUS groups (P < 0.001), accompanied by a marked decrease in iNOS protein expression (Fig. 6F–I). Importantly, this suppressive effect was partially reversed by BACH2 siRNA, highlighting the critical role of BACH2 in mediating the anti-inflammatory action.

In summary, LIPUS and mild hyperthermia inhibit M1 macrophage polarization by activating TRPV1, which in turn promotes the nuclear translocation of BACH2. This TRPV1-BACH2 axis represents a key mechanistic pathway underlying the anti-inflammatory effects of non-pharmacological physical therapies, providing novel insights into inflammation modulation strategies.

• (3)TRPV1-mediated Ca²⁺ influx is essential for BACH2 nuclear translocation induced by LIPUS and MH

To investigate the role of calcium influx in TRPV1-mediated regulation of BACH2 nuclear translocation, we used the Fluo-4 AM calcium indicator to monitor intracellular Ca²⁺ levels in RAW264.7 cells. Subcellular localization of BACH2 was analyzed by Western blot and immunofluorescence. Under calcium-free culture conditions, the addition of CaCl₂ in combination with LIPUS or mild hyperthermia (T) significantly enhanced Ca²⁺ influx at 10, 30, and 60 min (all P < 0.001), with fluorescence intensity markedly higher than that of single-intervention or control groups (Fig. 7A–D). Notably, the LIPUS and T groups alone also induced moderate Ca²⁺ elevation, suggesting that in the presence of extracellular calcium, both interventions synergistically enhance TRPV1-mediated Ca²⁺ influx. Western blot analysis showed that nuclear BACH2 expression was significantly upregulated in the LIPUS and T groups (P < 0.001), while EDTA treatment or culture in calcium-free medium markedly suppressed this nuclear translocation, reducing nuclear BACH2 levels to near-control levels (Fig. 7E–I). Cytoplasmic BACH2 expression remained unchanged across groups. Immunofluorescence staining further confirmed that both LIPUS and T promoted BACH2 translocation to the nucleus, whereas calcium depletion significantly inhibited this effect (Fig. 7J).

Fig. 7.

CaMKII links TRPV1 signaling to BACH2 nuclear translocation.

A shows the fluorescence images of Ca²⁺ at different time points for each group (Scale bars: 20 μm). B, C, and D represent the statistical analysis of Ca²⁺ expression levels at 10 min, 30 min, and 60 min for each group. E shows the electrophoresis bands of BACH2 protein expression levels in the cytoplasm and nucleus for each group. F and G represent the electrophoresis bands of BACH2 protein expression levels in the cytoplasm and nucleus for the EDTA group. H and I show the statistical analysis of BACH2 protein expression levels in the cytoplasm and nucleus for the Ca²⁺ group. J shows the representative images for mild heat therapy, LIPUS, combined EDTA, and calcium-free medium treatment (Scale bars: 25 μm). N = 3, * indicates P < 0.05, ** indicates P < 0.01, *** indicates P < 0.001.

Collectively, these results demonstrate that LIPUS and mild hyperthermia facilitate BACH2 nuclear translocation via TRPV1-mediated Ca²⁺ influx. This process is highly dependent on extracellular calcium, further confirming the pivotal role of the TRPV1-Ca²⁺ signaling axis in mediating the anti-inflammatory response.

• (4)CaMKII functions as a critical downstream mediator linking TRPV1 to BACH2 nuclear translocation

Based on the preceding findings, we further investigated whether calcium influx triggered by TRPV1 activation during LIPUS and mild hyperthermia (T) interventions mediates BACH2 nuclear translocation via the CaMKII signaling pathway. Given that CaMKII is a key downstream effector of Ca²⁺ signaling, we applied the specific CaMKII inhibitor KN-93 (20 μM) to evaluate its role. Western blot analysis showed that phosphorylated CaMKII (p-CaMKII) levels were significantly elevated in both the LIPUS and T groups compared to the control group (P < 0.01), indicating CaMKII activation. Upon co-treatment with KN-93 (LIPUS + KN-93 and T + KN-93 groups), p-CaMKII levels were markedly reduced, confirming the inhibitory effect (Fig. 8A and B). Total CaMKII levels were slightly decreased in the stimulated groups but were restored after KN-93 treatment (Fig. 8C). Further protein analysis revealed that both LIPUS and T interventions significantly enhanced nuclear BACH2 expression (P < 0.001). However, KN-93 treatment significantly suppressed BACH2 nuclear localization (P < 0.001), reducing it to near-control levels (Fig. 8D–F), indicating a crucial role for CaMKII in BACH2 nuclear translocation. Immunofluorescence assays corroborated these findings: BACH2 translocated to the nucleus in the LIPUS and T groups, as shown by increased nuclear fluorescence intensity. In contrast, KN-93 treatment diminished nuclear localization and redistributed BACH2 to the cytoplasm, resembling the control group (Fig. 8G and H). Notably, the T + KN-93 group retained slightly more nuclear BACH2 than the LIPUS + KN-93 group, suggesting that mild hyperthermia may exert a somewhat stronger effect. Together, these findings support the mediating role of CaMKII in TRPV1-induced BACH2 nuclear translocation.

Fig. 8.

Role of CaMKII in Mild Hyperthermia Therapy and LIPUS Activation of TRPV1-Induced BACH2 Nuclear Translocation.

A shows the electrophoresis bands of p-CaMKII and CaMKII protein expression levels for each group. B and C represent the statistical analysis of p-CaMKII and CaMKII protein expression levels for each group. D shows the electrophoresis bands of BACH2 protein expression levels in the cytoplasm and nucleus for each group. E and F represent the statistical analysis of BACH2 protein expression levels in the cytoplasm and nucleus for each group. G shows the immunofluorescence analysis of BACH2 nuclear translocation for each group. H presents the quantitative statistical analysis of BACH2 nuclear translocation for each group. I Co-immunoprecipitation analysis of the association between CaMKII and BACH2. Scale bars: 25 μm. N = 3, ns indicates no statistical significance. * indicates P < 0.05, ** indicates P < 0.01, *** indicates P < 0.001.

To further investigate the molecular link between CaMKII and BACH2, we performed co-immunoprecipitation assays. As shown in Fig. 8I, BACH2 was detectable in CaMKII immunoprecipitates, indicating that CaMKII associates with BACH2 in macrophages. Notably, the CaMKII–BACH2 association appeared enhanced following combined LIPUS and mild hyperthermia stimulation. Pharmacological inhibition of TRPV1 (CAP) or CaMKII (KN-93) did not completely eliminate the co-precipitation signal, suggesting that the interaction may be partially preserved under these conditions and/or that inhibitor treatment may not fully disrupt complex formation. Together, these data provide molecular evidence supporting a physical association between CaMKII and BACH2, complementing our functional results showing that CaMKII activity is required for BACH2 nuclear translocation.

• (5)Combined LIPUS and MH synergistically activate the TRPV1/Ca²⁺/CaMKII/BACH2 anti-inflammatory axis

To investigate the molecular mechanism underlying the anti-inflammatory effect of combined LIPUS and mild hyperthermia, Western blot analysis was performed to assess TRPV1, CaMKII, phosphorylated CaMKII (p-CaMKII), and BACH2 expression in RAW264.7 macrophages following different treatments. Results showed that the combination group (LIPUS + T) exhibited significantly elevated expression of TRPV1, CaMKII, p-CaMKII, and BACH2 compared to both the control and single-treatment groups (P < 0.05), indicating enhanced activation of the TRPV1/Ca²⁺/CaMKII/BACH2 pathway. Concurrently, the M1 polarization marker iNOS was markedly reduced in the combination group (P < 0.01), suggesting a synergistic inhibitory effect on pro-inflammatory macrophage activation (Fig. 9A–F). The mechanism by which LIPUS combined with mild hyperthermia promotes rotator cuff healing in rabbits via the TRPV1/Ca²⁺/CaMKII/BACH2 pathway is illustrated in Fig. 9G.

• (6)In vivo validation of TRPV1 and BACH2 signaling at the rotator cuff tendon–bone interface

Fig. 9.

Combined therapy synergistically activates the TRPV1/Ca²⁺/CaMKII/BACH2 anti-inflammatory axis.

A shows the electrophoresis bands of protein expression levels for each group. B, C, D, E, and F represent the statistical analysis of protein expression levels for TRPV1, iNOS, CaMKII, p-CaMKII, and BACH2 in each group. G provides a schematic diagram illustrating the mechanism by which Mild Hyperthermia Therapy and LIPUS activate TRPV1 to promote BACH2 nuclear translocation. N = 3, ns indicates no statistical significance. * indicates P < 0.05, ** indicates P < 0.001.

To validate whether the TRPV1/CaMKII/BACH2 signaling axis is involved in tendon–bone healing in vivo. Double immunofluorescence staining demonstrated the co-localization of TRPV1 with the M2 macrophage marker CD163, and BACH2 with the M1 macrophage marker CD86. Compared with the control group, the LIPUS + T group exhibited markedly increased TRPV1 and CD163 signals at the bone–tendon interface, with a higher proportion of TRPV1⁺CD163⁺ cells observed in the merged images. Quantitative analysis confirmed a significant increase in TRPV1⁺CD163⁺ macrophages in the LIPUS+T group (P < 0.01) (Fig. 10A and B). In contrast, staining for BACH2 and CD86 revealed a notable reduction in BACH2⁺CD86⁺ cells following LIPUS+T treatment compared with controls, which was also supported by quantitative analysis (P < 0.01) (Fig. 10C and D). These findings demonstrate that both TRPV1 and BACH2 are expressed in macrophages at the rabbit rotator cuff bone–tendon interface in vivo and that LIPUS + T treatment is associated with increased TRPV1-related M2 macrophage polarization and decreased BACH2-associated M1 macrophage presence. This in-situ evidence supports the involvement of the TRPV1–BACH2 regulatory axis in modulating macrophage polarization during rotator cuff healing in the animal model.

Fig. 10.

In vivo validation of TRPV1-BACH2 signaling at the tendon–bone interface.

A Representative images of TRPV1 (green), CD163 (red), and DAPI (blue). B Quantification of TRPV1⁺CD163⁺ macrophages. C Representative images of BACH2 (green), CD86 (red), and DAPI (blue). D Quantification of BACH2⁺CD86⁺ macrophages. N = 3, ** indicates P < 0.001.

4. Discussion

The present study demonstrates that combined LIPUS and mild hyperthermia significantly improved tendon–bone healing in a rabbit rotator cuff repair model. Histological organization, fibrocartilage regeneration, MRI continuity, and biomechanical strength were all enhanced compared with single-modality treatment. Mechanistically, these benefits were associated with favorable modulation of macrophage polarization during healing, together with activation of the TRPV1/Ca²⁺/CaMKII pathway.

It is important to emphasize that the present study did not aim to characterize the acute inflammatory phase after rotator cuff injury. Classical postoperative inflammatory responses are generally reported to peak within the first 48–72 h after injury and gradually decline during the subsequent 1–2 weeks [48,49]. We therefore agree with the reviewer that acute inflammatory responses are predominantly early events rather than processes occurring at 6 or 12 weeks. In the present study, macrophage-related markers were intentionally evaluated at 6 and 12 weeks to investigate the remodeling-stage immune microenvironment during tendon–bone healing. At these later stages, macrophages are increasingly recognized as participating in extracellular matrix turnover, fibrocartilage maturation, angiogenesis, and bone remodeling rather than representing unresolved acute inflammation [50,51]. Importantly, the current rabbit model involved transosseous tendon repair and continuous tendon–bone interface remodeling, which may prolong local immune-associated signaling beyond the classical acute inflammatory window reported in soft-tissue injury models [49,50]. This may partly explain the discrepancy between our observations and studies focused primarily on early inflammatory responses after acute injury. Moreover, the relatively persistent CD86⁻or cytokine-associated signals observed in some groups likely reflect remodeling-associated immune activity rather than chronic inflammatory pathology. Nevertheless, because early postoperative time points were not included, the present study cannot fully define the temporal dynamics of macrophage polarization during the acute healing phase, which represents an important limitation and warrants further investigation.

From the perspective of inflammation regulation, the combined application of LIPUS and MH effectively inhibited M1-type macrophage polarization while with increased expression of reparative macrophage-associated markers, as evidenced by the downregulation of CD86 and iNOS expression and the upregulation of CD163 and Arg1. These changes significantly improved the local inflammatory microenvironment and facilitated tendon–bone healing [52,53]. Compared with single interventions, the combined treatment yielded more pronounced improvements across multiple inflammatory parameters, suggesting a synergistic role in immune regulation. Previous studies support these findings. Liu et al. reported that LIPUS exerts anti-inflammatory effects by modulating microcirculation and promoting M2 polarization [54], and its combination with MH can further enhance immunoregulatory potential [55]. Similarly, Gouda et al. found that LIPUS suppresses M1-associated genes while promoting M2-related expression, thereby ameliorating inflammatory conditions in renal injury [56]. In addition, Xu et al. demonstrated that hyperthermia can influence macrophage polarization through exosome-mediated transfer of HSPB8 [57], further supporting the role of MH in inflammation regulation.

Healing at the tendon–bone interface is a critical determinant of functional recovery after rotator cuff repair, involving complex cellular and tissue remodeling as well as mechanical adaptation. In vivo experiments showed that the combined intervention of MH and LIPUS was superior to either modality alone in both structural repair and functional enhancement. Histological and biomechanical data indicated that the combination treatment markedly improved collagen fiber organization and fibrocartilage transition zone integrity, with mechanical strength approaching normal levels, outperforming LIPUS or MH monotherapy. Previous studies have reported that LIPUS facilitates chondrocyte differentiation and collagen synthesis via mechanical stimulation, thereby contributing positively to tendon–bone healing [58,59]. While LIPUS shows pronounced efficacy in promoting early chondrogenesis [60,61], its impact on late-stage bone structural restoration remains limited. MH, on the other hand, primarily alleviates inflammation by enhancing blood flow and metabolic activity, consistent with previous reports [62,63], yet its reparative capacity is also inferior to that of the combination therapy. Immunohistochemical analysis revealed that the combined intervention more effectively modulated macrophage polarization, downregulating M1 markers while upregulating M2 expression, thereby optimizing the local immune microenvironment and synergistically promoting regeneration [28,[64], [65], [66]]. Mechanistically, LIPUS and MH may exert their effects through the TRPV1 and FAK/PI3K/AKT pathways, respectively. MRI signal analysis also confirmed the dual benefits of the combined therapy, showing early anti-inflammatory effects and later tissue remodeling [30,[67], [68], [69]]. Compared with previous literature focusing solely on LIPUS [31], the present study provides systematic evidence for the synergistic advantages of LIPUS combined with MH, offering a novel, non-invasive therapeutic approach with promising clinical potential for tendon–bone healing. Moreover, the superiority of the combined intervention may stem from its integration of multiple mechanisms—such as the orderly alignment of collagen fibers, restoration of the osteochondral transition zone, and increased bone mineral density—thereby achieving maximal mechanical performance at the tendon–bone junction [29,70,71]. These findings offer new strategic insights for the treatment of tendon–bone injuries, particularly in scenarios where accelerated repair and enhanced stability are desired.

Although the same LIPUS and mild hyperthermia parameters were applied in both in vivo and in vitro experiments, the intention was not to assume identical physical exposure at the cellular level. Rather, the in vitro model was designed to isolate direct cellular responses to controlled mechanical and thermal stimuli, whereas the in vivo model reflects integrated tissue-level effects, where signal transmission is inevitably modulated by extracellular matrix, tissue thickness, and local biomechanics. Previous studies have demonstrated that although tissue structures may attenuate the magnitude of mechanical or thermal stimuli, the direction and qualitative nature of cellular responses remain comparable, allowing in vitro findings to serve as mechanistic support rather than quantitative replication of in vivo exposure. Therefore, using the same protocol across models facilitates mechanistic linkage while acknowledging differences in absolute stimulus transmission.

At the molecular level, transcriptome sequencing revealed that LIPUS + MH activated several repair-related pathways, including calcium signaling, PI3K–Akt, and TGF-β, which are essential for collagen synthesis and extracellular matrix remodeling [[72], [73], [74]]. Notably, TRPV1 and BACH2 were markedly upregulated, particularly with LIPUS intervention. Given that TRPV1 mediates Ca²⁺ influx, cell migration, and inflammation regulation [75,76], and BACH2 modulates immune cell differentiation and Ca²⁺-related pathways [27,77,78], their co-activation may synergistically enhance tissue regeneration. These findings suggest that LIPUS accelerates tendon–bone healing via TRPV1–BACH2–mediated regulation of cellular function, inflammation, and remodeling [79,80], supporting its potential as an efficient, non-invasive therapeutic approach.

Further validation experiments demonstrated that both LIPUS and MH significantly activated TRPV1, upregulated its mRNA and protein expression, promoted Ca²⁺ influx and CaMKII phosphorylation, enhanced BACH2 nuclear translocation, inhibited M1-type pro-inflammatory polarization, attenuated inflammatory responses, and facilitated tissue repair. Notably, the activation of TRPV1 by LIPUS was particularly prominent, consistent with the findings reported by Francesco et al. [81]. As a mechanosensitive and thermosensitive calcium channel, TRPV1 activation triggers Ca²⁺ influx, which in turn activates downstream CaMKII, thereby promoting the translocation of the transcription factor BACH2 from the cytoplasm to the nucleus and suppressing the expression of pro-inflammatory genes such as CD86 and iNOS [[82], [83], [84]]. The TRPV1 inhibitor CAP markedly attenuated these effects, confirming its pivotal regulatory role. Notably, considering the inherent limitations of pharmacological inhibition, our TRPV1-specific siRNA experiments provided complementary genetic evidence showing that LIPUS + T–induced CaMKII activation, BACH2 nuclear translocation, and suppression of M1-associated markers were largely abolished upon TRPV1 knockdown, thereby strengthening the specificity of the proposed TRPV1/Ca²⁺/CaMKII/BACH2 signaling axis. As a critical anti-inflammatory transcription factor, BACH2 exhibited significantly increased expression and nuclear translocation in both the LIPUS and MH groups, and BACH2 knockdown by siRNA substantially weakened the observed effects, indicating its essential role in this pathway [82,85].

Previous studies have indicated that the nuclear translocation of BACH2 plays a pivotal regulatory role in M1 macrophage polarization and inflammatory responses, effectively suppressing their pro-inflammatory activity [85,86]. The findings of the present study are consistent with these reports and further substantiate the critical role of BACH2 in inflammation regulation and mild hyperthermia–mediated immunomodulation [87]. BACH2 is recognized as a key transcription factor governing immune responses and tissue repair, and its subcellular localization within the nucleus is of substantial importance for cellular functional regulation [88]. To date, there have been no reports describing the promotion of nuclear translocation of BACH2 or similar transcription factors by mild hyperthermia to attenuate excessive immune responses. Research on the role of BACH2 in the combined application of mild hyperthermia and LIPUS remains limited. Our findings suggest that BACH2 may participate in the observed immunomodulatory response.

Finally, the present study examined the regulatory effects of LIPUS, MH, and their combination on the TRPV1/Ca²⁺/CaMKII/BACH2 signaling pathway. The combined intervention resulted in significantly greater TRPV1 activation and enhanced nuclear translocation of BACH2, without a consistent change in total BACH2 protein levels, along with marked inhibition of M1 polarization and its marker iNOS compared with either treatment alone, indicating a synergistic advantage in promoting tendon–bone healing. Importantly, in vivo immunofluorescence analysis further demonstrated that TRPV1 and BACH2 were expressed at the tendon–bone interface and were closely associated with macrophage polarization, providing additional evidence supporting the mechanistic involvement of the TRPV1/CaMKII/BACH2 signaling axis in tendon–bone repair.

Mechanistically, MH may directly activate TRPV1 via thermal stimulation, whereas LIPUS enhances Ca²⁺ influx through mechanical stimulation. These combined effects amplify TRPV1 signaling, promote CaMKII activation and BACH2 nuclear translocation, and subsequently suppress M1 macrophage polarization and inflammatory responses, thereby improving the local immune microenvironment and accelerating tissue repair. It should be noted that macrophage polarization during tendon–bone healing is a dynamic process that evolves across different stages of tissue repair. In the present study, we focused on representative mid-to-late healing stages (6 and 12 weeks), which mainly reflect remodeling-phase responses. Although a more comprehensive temporal analysis would further clarify the dynamic transition of macrophage phenotypes, such investigations were beyond the scope of the current study. Future studies incorporating additional post-operative time points and quantitative approaches, such as flow cytometry, may help to further characterize the temporal immunomodulatory effects of combined LIPUS and MH therapy. Collectively, LIPUS and MH promote tendon–bone healing through an integrated signaling cascade—“TRPV1 sensing of mechanical and thermal stimuli → Ca²⁺ influx → CaMKII activation → BACH2 nuclear translocation → inhibition of M1 polarization and inflammatory responses”—highlighting a novel immunomodulatory strategy. Future studies should further explore the translational potential of targeting this pathway in chronic inflammation and tissue regeneration [89,90].

4.1. Limitations

This study has several limitations. First, the current rabbit model represents an acute supraspinatus tendon detachment and repair model and therefore cannot fully recapitulate the pathological features of chronic degenerative rotator cuff disease in humans, such as muscle atrophy, fatty infiltration, and long-standing inflammatory alterations. In addition, early postoperative inflammatory dynamics (e.g., within the first 1–14 days) were not evaluated in the present study. Accordingly, the current findings primarily reflect remodeling-stage immune regulation rather than acute inflammatory responses during tendon–bone healing. Second, only two postoperative time points (6 and 12 weeks) were assessed, which limited dynamic characterization of time-dependent healing events, including early inflammation resolution, fibrocartilage maturation, and trabecular bone remodeling. Future studies incorporating earlier and multiple sequential time points will help provide a more comprehensive understanding of the temporal immunological and structural changes during tendon–bone healing. Third, although our in vitro siRNA experiments provided genetic evidence supporting the specificity of TRPV1 signaling, macrophage-specific conditional deletion of TRPV1 in vivo would provide more definitive mechanistic validation. However, such conditional genetic approaches are currently not feasible in the rabbit model used in this study. Future investigations using macrophage-specific TRPV1 knockout mice or other advanced genetic strategies may further strengthen causal inference regarding the TRPV1/Ca²⁺/CaMKII/BACH2 signaling axis. Fourth, the present study primarily focused on M1 macrophage polarization. Although we demonstrated that the TRPV1–BACH2 axis regulates macrophage-associated inflammatory responses, its potential effects on other immune cell populations (such as T cells and B cells), fibroblasts, and tendon-derived cells remain unclear. Additional studies are needed to clarify potential cell type-specific regulatory mechanisms within the tendon–bone healing microenvironment. Finally, while we identified CaMKII as an important downstream mediator of TRPV1 signaling and observed distinct calcium flux patterns across treatment groups, the precise downstream molecular interactions remain incompletely understood. Moreover, future fluorescence imaging studies should incorporate nuclear or bright-field counterstaining to allow more intuitive visualization of intracellular localization and calcium dynamics.

5. Conclusion

This study suggests that LIPUS combined with mild hyperthermia improves tendon--bone healing, which is associated with favorable modulation of macrophage polarization and activation of the TRPV1/Ca²⁺/CaMKII/BACH2 signaling axis observed during the remodeling stage. These findings provide mechanistic support for further investigation of this non-invasive strategy to improve the immune microenvironment impaired by persistent inflammatory signals.

Future studies should further investigate the role of TRPV1 in different immune cell populations, particularly its interactions with other signaling pathways, to elucidate the complex molecular networks underlying tendon–bone healing and to inform the optimization of individualized therapeutic strategies in clinical settings.

AI and AI-assisted technologies

We hereby confirm that no generative AI or AI-assisted technologies were used in the writing of this manuscript.

Ethical approval

All experimental protocols were approved by the Peking University Institutional Animal Care and Use Committee (PUIRB-LA2023529) and conducted in compliance with the NIH guidelines for the care and use of laboratory animals.

Author contributions

Xiali Xue: Conceptualization, Investigation, Formal analysis, Writing – original draft. Xingzuan Lin: Investigation, Data curation, Visualization. Aofei Gao: Methodology, Investigation, Validation. Hao Fu: Methodology. Amila Kuati: Methodology, Writing – review & editing. Guoqing Cui: Conceptualization, Funding acquisition, Supervision, Writing – review & editing. Bingbing Xu: Methodology, Supervision, Resources, Writing – review & editing. Qingfa Song: Conceptualization, Funding acquisition, Project administration, Writing – review & editing.

Funding

This study was supported by the Clinical queue construction project of Peking University Third Hospital (BYSYDL2022006), the China Postdoctoral Science Foundation (2025M781879, 2026M792098), and the National Natural Science Foundation of China (82503093).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article.