Dual-responsive PDA-HP hydrogel enables mitochondria-targeted mild photothermal therapy for spinal cord repair

Abstract

Spinal cord injury (SCI) is a devastating neurological disorder with substantial economic and psychological burdens, underscoring the urgent need for effective therapeutic strategies. Here, we developed a dual-responsive hydrogel composed of polydopamine (PDA) and heparin-poloxamer (HP) that enables controllable mild photothermal stimulation under near-infrared (NIR) irradiation. The PDA-HP hydrogel exhibited excellent biocompatibility, biodegradability, and stable photothermal conversion. In a mouse SCI model, in situ administration of PDA-HP combined with NIR irradiation markedly improved locomotor recovery and mitigated tissue damage. Mechanistically, PDA-HP/NIR therapy reduced oxidative stress, preserved mitochondrial structure, restored ATP production, and—most notably—normalized the maladaptive overexpression of heat-shock protein 70 (HSP70) induced by SCI, thereby decreasing apoptosis and promoting neuronal survival. Quantitative proteomics further identified stress-chaperone and mitochondrial pathways as major targets of this intervention. To our knowledge, this is the first study demonstrating that PDA-HP-mediated mild photothermal modulation restores mitochondrial function through HSP70 normalization in SCI. These findings highlight a mitochondria-targeted mild photothermal strategy as a promising and clinically translatable approach for spinal cord repair.

Graphical abstract

1. Introduction

Spinal cord injury (SCI) is a devastating condition that ranks among the most disabling diseases of the central nervous system [1,2]. It imposes profound economic and psychological burdens on both affected individuals and society [2]. Despite continuous research and medical advances, an effective treatment for SCI remains elusive.

Mitochondria, as the primary energy producers and key regulators of apoptosis, play a pivotal role in neuronal survival [3]. Following SCI, damaged axons degenerate and large numbers of neurons undergo apoptosis [1]. Mitochondrial function is severely impaired, disrupting both energy and substrate metabolism [[3], [4], [5]]. This results in reduced ATP production and the accumulation of harmful by-products such as reactive oxygen species (ROS), lactate dehydrogenase (LDH), and malondialdehyde (MDA) [3,5]. As the injury progresses, secondary ischemia, hypoxia, and metabolic disturbances exacerbate mitochondrial damage, creating a vicious cycle of energy deficiency, mitochondrial dysfunction, and neuronal death [6,7]. Therefore, restoring mitochondrial function is crucial for SCI repair. Previous studies, including our own [8], have demonstrated that improving mitochondrial performance can effectively promote recovery after SCI [9,10].

Mitochondria are highly sensitive to moderate thermal stress. Sublethal heating within a physiological tolerance range (<43 °C) can enhance mitochondrial respiration, promote ATP synthesis, and reduce oxidative damage without inducing cytotoxicity [3,11]. These beneficial effects are largely mediated through transient activation of heat shock proteins (HSPs) and stress-adaptive signaling pathways that maintain mitochondrial proteostasis and redox homeostasis [3,12].

Recent preclinical studies have extended these findings to SCI models, showing that controlled light-induced thermal stimulation can restore mitochondrial homeostasis, suppress neuroinflammation, and improve motor recovery [13]. For instance, Zhu et al. demonstrated that low-level near-infrared irradiation enhanced neuronal mitochondrial bioenergetics via the AMPK/PGC-1α/TFAM pathway, while Li et al. reported that mild photonic stimulation alleviated mitochondrial fission imbalance and neuronal apoptosis after SCI [14,15]. Together, these findings suggest that spatially confined, sublethal thermal modulation—whether photobiomodulative or photothermal—can strengthen mitochondrial resilience and facilitate neural repair.

Photothermal therapy (PTT) is a light-controlled thermal strategy that converts absorbed near-infrared (NIR) light into heat through photothermal agents [[16], [17], [18]]. It offers advantages such as noninvasiveness, precise spatiotemporal control, and minimal collateral damage [19]. Compared with conventional heat therapy, PTT allows more localized and tunable temperature elevation, enabling selective activation of beneficial stress responses without triggering protein denaturation or apoptosis [19,20].

Among various NIR-responsive materials, polydopamine (PDA) stands out for its high photothermal conversion efficiency, ROS-scavenging capacity, biodegradability, and excellent biocompatibility [[21], [22], [23]]. Hydrogels are also widely used in tissue engineering due to their injectability, hydration, and biocompatibility, and stimulus-responsive hydrogels further enable precise spatiotemporal control under external stimuli such as temperature or light [24,25]. However, many traditional NIR agents (e.g., graphene oxide, gold nanoparticles, or carbon nanotubes) have limited biocompatibility and long-term safety concerns [26].

To overcome these challenges, we employed heparin–poloxamer (HP), a thermosensitive hydrogel previously proven safe and effective for neural repair [8]. HP remains liquid at room temperature, allowing for minimally invasive injection, and rapidly forms a stable three-dimensional network at physiological temperature (37 °C). Its biodegradability and anticoagulant properties also help reduce perilesional thrombosis and glial scarring [8,27].

In this study, we utilized HP as a biocompatible carrier to develop an in situ injectable hydrogel incorporating PDA nanoparticles, forming a dual-responsive PDA-HP system sensitive to both temperature and light. Prior to gelation, PDA-HP exists as a nanoscale suspension in which uniformly dispersed PDA nanoparticles endow the system with photothermal responsiveness. Upon NIR irradiation, the PDA-HP hydrogel generates controlled mild hyperthermia (<43 °C), aiming to restore mitochondrial function and facilitate spinal cord repair. While previous studies have applied photothermal biomaterials primarily for angiogenesis promotion or anti-inflammatory effects in SCI, the potential of mild NIR-induced hyperthermia to directly modulate mitochondrial function through HSP70 regulation has not been explored. This study uniquely demonstrates that PDA–HP/NIR treatment normalizes maladaptive HSP70 elevation, thereby restoring mitochondrial homeostasis and promoting functional recovery. Proteomic analysis was further employed to elucidate the molecular mechanisms underlying its therapeutic effects in vitro and in vivo. We hypothesized that this thermosensitive PDA-HP hydrogel provides localized and tunable photothermal stimulation to reestablish mitochondrial homeostasis and modulate HSP70-mediated stress responses following SCI. Overall, this approach offers a convenient, precise, and controllable therapeutic strategy for spinal cord injury repair.

2. Materials and methods

2.1. Experimental design overview

To improve readability and logical flow, all experiments were organized into three main parts: first, material synthesis and characterization (Sections 2.2-2.3); second, in vitro studies evaluating photothermal performance, cytocompatibility, oxidative stress resistance, and mitochondrial protection (Section 2.4-2.5); third, in vivo experiments assessing therapeutic efficacy, histology, immunofluorescence, and proteomic mechanisms in SCI mice (Sections 2.6-2.11).

2.2. Design and preparation of PDA-HP hydrogel

PDA nanoparticles were synthesized via alkaline oxidative self-polymerization [28]. Briefly, a mixture of 40 mL absolute ethanol, 90 mL deionized water, and 2 mL ammonium hydroxide (28–30 %) was magnetically stirred in a round-bottom flask. Dopamine hydrochloride (250 mg) dissolved in 10 mL deionized water, was then added, and the reaction was allowed to proceed for 24 h at room temperature. The resulting PDA nanoparticles (average diameter ≈ 193 nm) were collected by centrifugation (15,000 rpm, 10 min), thoroughly washed, and lyophilized.

Heparin-poloxamer (HP) conjugates were prepared via EDC/NHS-mediated coupling. Poloxamer 407 was first activated with 1.3 mM 4-nitrophenyl chloroformate, followed by reaction with ethylenediamine to produce an amine-terminated intermediate. This intermediate was subsequently reacted with heparin in MES buffer overnight at room temperature. The resulting HP conjugate was purified by extensive dialysis over 3 days and then lyophilized to yield a dry powder.

PDA-HP hydrogels were prepared by dispersing PDA nanoparticles in PBS (pH 7.4) and mixing with HP at final concentrations of 10 mg/mL and 160 mg/mL, respectively. The mixture was gently stirred for 24 h to yield a uniform composite hydrogel, which was stored at 4 °C under sterile conditions until use. The PDA nanoparticles (∼193 nm) were homogeneously distributed within the thermosensitive HP matrix prior to gelation, forming nanoscale photothermal domains embedded in the macroscopic hydrogel network. This structural configuration enables localized, controllable light-to-heat conversion while maintaining overall gel integrity and injectability.

2.3. Microstructural and rheological characterization

2.3.1. Scanning electron microscopy (SEM)

The morphology of the HP hydrogel and PDA-HP hydrogel was examined using SEM. Samples were freeze-dried for 48 h, carefully sectioned, and mounted onto conductive adhesive prior to sputter-coating with a thin layer of gold. SEM imaging was performed using a VEGA3 TESCAN instrument (Czech Republic) to assess the microstructural features of the hydrogels.

2.3.2. Determination of steady-state viscosity of HP hydrogel and PDA-HP hydrogel

HP hydrogel and PDA-HP hydrogel samples (60 mL each) were equilibrated in a thermostatic bath with a temperature control precision of ±0.1 °C. A digital rotational viscometer (NDJ-1S, Shanghai, China) equipped with a T-bar guard and spindle No. 1 was used. The rotor was lowered until the engraved line aligned with the meniscus of the sample. The temperature was increased from 20 °C to 50 °C in increments of 5–7 °C, fully covering the mild photothermal range (37–43 °C) relevant to in vivo PTT conditions. After each 10-min equilibration period, viscosity measurements were recorded at 30 rpm. Each measurement was performed in triplicate, and the mean values were used for statistical analysis.

2.4. Photothermal performance measurement

The photothermal behavior of the PDA-HP hydrogel was evaluated under 808 nm near-infrared (NIR) laser irradiation (1.0 W/cm²) to assess its heating capacity under physiologically relevant conditions. Hydrated PDA–HP hydrogel samples (1 mL) were immersed in phosphate-buffered saline (PBS, pH 7.4) rather than tested in the dry state, ensuring thermal conduction consistent with the in-vivo aqueous environment.

Temperature changes were continuously recorded using both an infrared thermal camera (Fluke TiX560, USA) and a K-type thermocouple probe positioned at the center of the hydrated sample to verify the accuracy of surface and internal temperature measurements. Each heating-cooling cycle was conducted under identical irradiation parameters (808 nm, 1.0 W/cm², 10 min per cycle), followed by natural cooling to room temperature before the next cycle, ensuring consistent thermal exposure across all tests.

Heating-cooling cycles were repeated six times to evaluate photothermal stability, and temperature-time curves were plotted to assess concentration-dependent heating behavior. For in-vivo thermographic monitoring, infrared imaging was performed at the spinal lesion site during irradiation. The surface temperature was maintained below 43 °C, consistent with the definition of mild photothermal therapy, and no visible tissue damage was observed.

2.5. In vitro cellular experiments

The ND7/23 cell line, derived from dorsal root ganglia, was obtained from the Stem Cell Bank of the Chinese Academy of Sciences and cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10 % fetal bovine serum (FBS) under standard conditions.

2.5.1. Cellular toxicity induced by tert-butyl hydroperoxide (T-BHP)

To assess T-BHP-induced cytotoxicity in ND7/23 cells, cell viability was measured using the CCK-8 assay (C0037, Beyotime, China). ND7/23 cells were seeded at 8 × 10³ cells per well in 96-well plates and allowed to stabilize overnight. Culture medium was then replaced, and cells were exposed to increasing concentrations of T-BHP (0, 10, 20, 50, 100, or 200 μM) for 6 or 12 h (Fig. S1). Following treatment, 10 μL of CCK-8 solution was added to each well, and plates were incubated at 37 °C with 5 % CO₂ for 1 h. Absorbance at 450 nm was measured using a Tecan Spark microplate reader (Switzerland) to calculate relative cell viability.

2.5.2. Photothermal treatment and hydrogel co-culture

For co-culture experiments, ND7/23 cells were seeded in 24-well plates (1 × 10⁵ cells/well) and incubated with 100 μL of PDA-HP pre-gel solution per well. The mixture was maintained at 37 °C for 10 min to allow complete gelation, forming a thin hydrogel layer over the cells. After 24 h of incubation, the cultures were irradiated with an 808 nm NIR laser (1 W/cm²) for 10 min at a fixed distance of 5 cm. The temperature of the culture medium was continuously monitored with a thermocouple and remained below 43 °C to ensure mild photothermal treatment. Control groups were handled identically without laser exposure.

2.5.3. Live-dead assay

Cell viability within the hydrogel was assessed using a Live/Dead Viability/Cytotoxicity Kit (C2015S, Beyotime, China) following the manufacturer's instructions. ND7/23 cells were co-cultured with PDA-HP hydrogel for 48 h, after which the staining solution was applied, and samples were incubated at 37 °C for 15 min. Fluorescence images were acquired using a Zeiss Axio Vert.A1 microscope (Germany).

2.5.4. Measurement of mitochondrial membrane potential (MMP)

MMP was assessed using the JC-1 fluorescent probe (C2006, Beyotime, China), which selectively accumulates in mitochondria and exhibits a shift in fluorescence emission depending on ΔΨm. Cells were incubated with 5 μM JC-1 for 30 min, washed thoroughly with PBS, and transferred to fresh culture medium. Fluorescence images were captured using a Leica DMIL LED microscope (Germany).

2.5.5. LDH release cytotoxicity assay

Cytotoxicity was evaluated using a commercial LDH assay kit (C0016, Beyotime, China) following the manufacturer's instructions. After T-BHP treatment, ND7/23 cells were incubated with the supplied LDH reaction mixture, and extracellular LDH activity was measured at 490 nm using a Tecan Spark microplate reader (Switzerland).

2.5.6. Detection of intracellular reactive oxygen species (ROS)

Intracellular ROS levels in ND7/23 cells were assessed using the cell-permeable probe DCFH-DA (S0033, Beyotime, China). Cells were seeded at 1.5 × 10⁵ per well and allowed to stabilize for 24 h, followed by exposure to 100 μM T-BHP for 6 h. Cells were then incubated with 10 μM DCFH-DA at 37 °C for 30 min, washed twice with PBS, and immediately imaged using a Zeiss Axio Vert.A1 microscope (Germany). Green DCF fluorescence, proportional to ROS levels, was quantified using ImageJ software.

2.6. Animal model of SCI

All experimental procedures were approved by the Institutional Review Board of Gannan Medical University. C57BL/6 mice were purchased from SLAC and housed under controlled conditions at 22–26 °C with 35–55 % relative humidity, with free access to food and water. Female C57BL/6 mice (8 weeks old, 20–22 g) were used in all experiments. Following complete depilation and sequential disinfection of the surgical site with povidone-iodine and 70 % ethanol, general anesthesia was induced with 5 % isoflurane in 1 l/min O₂ and maintained at 2 %. Body temperature was continuously monitored and maintained at 37 °C on a heated surgical platform, and adequate anesthesia was confirmed by the absence of a tail-pinch reflex. A midline skin incision was made, and the paravertebral muscles were bluntly dissected to expose the T9-T10 vertebrae. The T9 spinous process, angled caudally, and the neutral to slightly caudally tilted T10 process served as anatomical landmarks. Following complete laminectomy at T9-T10, the spinal cord was fully exposed. A moderate contusion was induced at the midline using a 10 g stainless-steel rod dropped from a height of 6.5 mm with the NYU-III (Keck) impactor. Immediately after injury, PDA-HP hydrogel was applied to cover the lesion site, and muscles and skin were sutured in layers. Mice were randomly assigned to four groups: Sham, SCI, PDA-HP, and PDA-HP/NIR. Sham-operated animals underwent identical surgical exposure without spinal impact, whereas the remaining groups received both laminectomy and contusion injury. For animals in the PDA-HP/NIR group, near-infrared (NIR) irradiation (808 nm, 1 W/cm², 30 min/day) was applied for 7 consecutive days post-surgery. During irradiation, the laser beam was directed over the healed incision site at a distance of ∼1 cm, allowing light penetration through the thin post-surgical tissue layer to reach the underlying PDA-HP hydrogel. Previous optical measurements indicate that 808 nm NIR light can penetrate several millimeters of soft tissue in small animals [29], which is sufficient to reach the dorsal spinal cord. Post-operative care included manual bladder expression twice daily until reflexive voiding was restored.

2.7. Behavioral assessment

Motor recovery following SCI was evaluated using three complementary behavioral tests, conducted by three examiners blinded to the experimental groups in a quiet environment. Locomotor function was assessed using the Basso Mouse Scale (BMS, 0-9) [30] during 5-min free-exploration sessions, with scores assigned based on consensus for each hind-limb movement.

For footprint analysis, hind paws were painted with red dye and forepaws with blue dye, and mice traversed a 50 cm runway. Stride length, base of support, and inter-limb coordination were measured from three consecutive steps.

Motor coordination and balance were further evaluated using an accelerating rotarod (4–40 rpm over 300 s). Mice were pre-trained for three consecutive days to establish stable baseline latencies. After SCI, the average latency to fall was recorded over three trials per day, with 15-min intervals between trials.

2.8. Histological staining

2.8.1. Hematoxylin and eosin (HE) staining

For comprehensive histological evaluation, spinal cord, heart, liver, spleen, lung, kidney, and knee-joint (femur-tibia) specimens were fixed in 4 % paraformaldehyde at 4 °C for 48 h. Bones were decalcified in 10 % EDTA (pH 7.4) for 14 days, followed by dehydration, paraffin embedding, and sagittal sectioning at 5 μm. Sections were deparaffinized, rehydrated, and stained with hematoxylin for 5 min and eosin for 2 min.

Soft organs and spinal cord samples were cryoprotected in 30 % sucrose for 48 h, embedded in optimal cutting temperature (OCT) compound, and sectioned transversely at 8 μm. After brief air-drying, cryosections were stained following the same HE protocol. After dehydration, clearing, and mounting, images were captured using a Leica DMIL LED optical microscope (Germany).

2.8.2. Masson's trichrome staining

To assess collagen deposition in the mouse spinal cord, 1-cm spinal segments were immediately fixed in 4 % paraformaldehyde at 4 °C for 24 h, cryoprotected in 30 % sucrose for 48 h, and embedded in OCT compound. Transverse cryosections (8 μm) were mounted on gelatin-coated slides, air-dried, and rehydrated in PBS prior to staining.

Masson's trichrome staining was performed as follows: nuclei were stained with Weigert's iron hematoxylin for 5 min, cytoplasm and myofibers with Biebrich scarlet–acid fuchsin for 3 min, and collagen fibers with aniline blue for 5 min after differentiation in 1 % phosphomolybdic acid. Following sequential dehydration, clearing, and mounting, images were captured using a Leica DMIL LED optical microscope (Germany).

2.9. Immunofluorescence analysis

For immunofluorescence analysis, ND7/23 cells grown on coverslips were blocked with 5 % bovine serum albumin (BSA) at 37 °C for 30 min. Cells were then incubated with primary antibodies against cleaved caspase-3 (C-cas3, 1:500, Proteintech, China) or HSP70 (1:500, Proteintech, China) at 4 °C for 10 h. After three washes with PBS, slides were incubated with appropriate Alexa Fluor-conjugated secondary antibodies for 1.5 h at room temperature, followed by counterstaining with DAPI for 8 min. Fluorescence images were acquired using a Leica TCS SP8 confocal microscope (Germany) and quantified with ImageJ software.

To assess the neuroimmune microenvironment after treatment, paraffin-embedded spinal cord sections (5 μm thickness) were used for immunofluorescence staining. Sections were deparaffinized in xylene and rehydrated through graded ethanol, followed by antigen retrieval in citrate buffer (pH 6.0) at 95 °C for 15 min. After cooling to room temperature, sections were washed three times with PBS and blocked with 5 % BSA for 30 min at 37 °C. Samples were then incubated overnight at 4 °C with the following primary antibodies: GFAP (astrocyte marker, 1:500, Proteintech, China), Iba1 (microglia marker, 1:500, Proteintech, China), Arg-1 (M2 phenotype marker, 1:500, Proteintech, China), iNOS (M1 phenotype marker, 1:500, Proteintech, China). After rinsing, sections were incubated with Alexa Fluor 488- or 594-conjugated secondary antibodies (1:1000, Proteintech, China) for 1 h at room temperature in the dark, followed by counterstaining with DAPI (1 μg/mL, 8 min). Fluorescence images were captured using a Leica TCS SP8 confocal microscope (Germany), and quantification of positive cell ratios and fluorescence intensity was performed using ImageJ software.

2.10. Western blot analysis

Equivalent amounts of protein lysates from spinal cord tissue were separated on 10 % SDS-polyacrylamide gels and electro-transferred onto 0.22 μm PVDF membranes. Membranes were blocked with 5 % non-fat milk in TBST for 1 h at room temperature, followed by incubation overnight at 4 °C with primary antibodies against HSP70 (1:1000, Proteintech, China), Bcl-2 (1:1000, Proteintech, China), Bax (1:1000, Proteintech, China), and β-actin (1:1000, Proteintech, China). After three washes with TBST, membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies for 2 h at room temperature. Immunoreactive bands were visualized using the Bio-Rad ChemiDoc XRS + imaging system. Each experiment was independently repeated six times to ensure reproducibility.

2.11. Proteomics and bioinformatics

Thoracic spinal cord segments (T9-T10) were harvested from each group, immediately frozen in liquid nitrogen for 15 min, and stored at −80 °C. Tissues were lysed in lysis buffer (8 M urea in 1 × PBS, pH 8.0, supplemented with 1 × protease and phosphatase inhibitor [Thermo Scientific, New Jersey, USA]) and sonicated. Lysates were centrifuged at 12,000×g for 30 min at 4 °C, and the supernatant was collected. Protein concentration was determined using a BCA assay kit (Thermo Fisher, NJ, USA).

For each sample, 100 μg of protein was reduced with 200 mM DTT at 37 °C for 1 h, diluted fourfold with 25 mM ammonium bicarbonate (ABC) buffer, and digested overnight with trypsin at a 1:50 enzyme-to-protein ratio. Digestion was terminated with 0.1 % formic acid (FA), and peptides were desalted using C18 columns, eluted with 70 % acetonitrile (ACN), and lyophilized.

Data-dependent acquisition (DDA) was performed on a Q Exactive HF-X mass spectrometer (Top-40, 120 k/15 k resolution) to generate the spectral library. Data-independent acquisition (DIA) employed 60 k MS1 resolution and 42 × 30 k MS2 windows, with an 8–40 % B gradient over 80 min at 600 nL/min and a column temperature of 60 °C. DDA and DIA files were processed using Spectronaut to construct a hybrid library, searched against the UniProt Mus musculus database with a 1 % false discovery rate (FDR). Functional annotation, including GO, InterPro, COG, and KEGG analyses, as well as protein-protein interaction networks, were analyzed using STRING.

2.12. Statistical analysis

All data are presented as mean ± standard deviation (SD). Statistical analyses were performed using SPSS 24.0 (Chicago, USA). Differences among multiple groups in immunofluorescence and Western blot experiments were assessed by one-way analysis of variance (ANOVA) followed by Tukey’s post-hoc test. Graphs were generated using GraphPad Prism 8.0 (GraphPad, California, USA). A p-value <0.05 was considered statistically significant.

3. Results

3.1. Fabrication and characterization of PDA-HP hydrogel

PDA nanoparticles were synthesized via the oxidative self-polymerization of dopamine hydrochloride under alkaline conditions (Fig. 1A). The morphology of the resulting nanoparticles was examined using scanning electron microscopy (SEM) and a nanoparticle size analyzer. PDA nanoparticles exhibited a monodisperse distribution with smooth surfaces and uniform morphology, appearing spherical (Fig. 1B). The average particle size of the PDA nanoparticles (Fig. 1C) remained remarkably stable at approximately 193 nm over a period of 21 days, with minimal fluctuations (193.42 ± 1.53 nm). Additionally, the zeta potential (Fig. S2) exhibited only a slight change, shifting from −34.1 ± 0.26 mV to −32.6 ± 0.46 mV during this timeframe. These observations collectively indicate the excellent physical stability of the formulation. In contrast, the polydispersity index (PDI) gradually increased from 0.18 ± 0.01 to 0.28 ± 0.01 during the same period, indicating a modest decline in colloidal uniformity. This slight increase in heterogeneity is still within an acceptable range for nanocarrier systems, implying that the formulation maintains favorable stability for potential biomedical applications (Fig. 1C).

Fig. 1.

Fabrication and characterization of PDA-HP hydrogel. (A) Schematic illustration of PDA nanoparticle synthesis via oxidative self-polymerization of dopamine. (B) SEM image of PDA. Scale bar is 0.5 μm. (C) The changes in particle size and polydispersity index (PDI) of PDA over 21 days by nanoparticle size analyzer. (D) Schematic representation of HP hydrogel synthesis via EDC/NHS-mediated heparin conjugation to poloxamer. (E) ¹H NMR spectra confirming the chemical composition of HP hydrogel. (F) SEM images of HP hydrogel and PDA-HP hydrogel after freeze-drying following in situ gelation. White dashed circles indicate representative regions containing PDA nanoparticles embedded within the HP matrix; the nanoparticles are uniformly distributed throughout the hydrogel. Scale bars are 50 μm. (G) Photographs of HP and PDA-HP hydrogels at 4 °C (sol state) and 37 °C (gel state). (H) Steady-state viscosity profiles of HP and PDA-HP hydrogels as a function of temperature (20–50 °C). Both hydrogels maintained stable viscosity between 37 °C and 43 °C, confirming good structural stability under mild photothermal conditions.

HP copolymer was synthesized using the EDC/NHS coupling method (Fig. 1D) [8]. The chemical composition of the HP hydrogel was confirmed by ¹H nuclear magnetic resonance (¹H NMR) spectroscopy (Fig. 1E). The characteristic chemical shifts of the polyoxypropylene and polyoxyethylene chains were observed at δ 1.14 ppm (–CH₃), δ 3.40 ppm (–CH), δ 3.50 ppm (–CH₂), and δ 3.65 ppm (–CH₂), while peaks corresponding to heparin appeared at δ 1.10 ppm, δ 1.80 ppm, and δ 4.20 ppm, confirming successful conjugation of heparin to poloxamer.

The PDA-HP hydrogel was prepared by dissolving PDA in phosphate-buffered saline (PBS, pH 7.4) and mixing it with HP powder, followed by gentle stirring for 24 h to form the hydrogel. The final concentrations of PDA and HP were 10 mg/mL and 160 mg/mL respectively, which ensures that PDA-HP hydrogel can still form gel at 37 °C and have mild hyperthermia (<43 °C) under NIR (808 nm, 1 W/cm²) (Fig. 1G and H). As demonstrated in our previous work [8], HP hydrogel exhibits a temperature-sensitive sol-gel phase transition: it remains liquid at 4 °C but rapidly transforms into a gel at physiological temperature (37 °C). This property enables convenient in situ gelation upon injection, allowing the material to adapt to irregular defect geometries and form a three-dimensional supportive network that integrates with surrounding tissues.

To evaluate whether PDA incorporation altered the thermosensitive behavior of the HP hydrogel, we examined the phase-transition characteristics of the PDA–HP formulation. Steady-state viscosity analysis confirmed that the PDA–HP hydrogel maintained the same sol-gel transition temperature (∼37 °C) as the pristine HP hydrogel (Fig. 1G and H). As shown in Fig. 1H, both HP and PDA-HP hydrogels exhibited a sharp viscosity increase near 37 °C, corresponding to sol-gel transition, and remained mechanically stable up to 43 °C, indicating robust structural integrity within the mild photothermal range relevant to SCI repair. Furthermore, SEM imaging revealed that PDA-HP retained a porous three-dimensional microarchitecture comparable to HP (Fig. 1F), facilitating homogeneous PDA distribution and providing a favorable scaffold for potential drug delivery and tissue repair applications.

3.2. Photothermal performance of PDA-HP hydrogel

PDA-HP hydrogel exhibited robust and tunable photothermal performance under 808 nm irradiation (1.0 W/cm²). Photothermal heating profiles (Fig. 2A) were obtained from hydrated PDA-HP hydrogel samples immersed in PBS (pH 7.4, 37 °C), with a thermocouple probe positioned at the sample center to ensure accurate temperature monitoring under physiologically relevant aqueous conditions. Each heating-cooling cycle was conducted for 10 min per irradiation period, followed by natural cooling to room temperature before the next cycle. The nearly identical temperature-time curves across six cycles confirmed the hydrogel’s excellent photothermal stability and reproducibility.

Fig. 2.

Photothermal performance of PDA-HP hydrogel. (A) Photothermal stability of PDA–HP hydrogel over six consecutive heating–cooling cycles under 808 nm laser irradiation (1 W/cm², 10 min per cycle), showing negligible attenuation of heating capacity. (B) Infrared thermal images of SCI mice treated with PDA-HP hydrogel (PDA-HP group) or without treatment (CON group) under identical ambient conditions (22 °C) and irradiation parameters (808 nm, 1 W/cm²) for 1, 3, and 5 min. The PDA-HP group exhibited a significant and localized temperature rise at the lesion site, while only minimal heating was observed in the control group. The color scale (25–45 °C) spans basal body temperature (∼36 °C) to the photothermal peak (∼42 °C). (C) Photothermal heating curves of PDA-HP hydrogels with varying PDA concentrations (1, 10, and 20 mg/mL) under NIR irradiation (808 nm, 1 W/cm²). Temperature elevation exhibited a clear concentration-dependent trend, confirming the tunable and stable photothermal behavior of the material in aqueous (in vitro) conditions. Note that these in vitro heating profiles reflect material performance in solution and may differ from intramedullary temperatures in vivo due to optical scattering, blood perfusion, and tissue heat dissipation. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Infrared thermal imaging (Fig. 2B) was performed under identical ambient temperature (22 ± 1 °C) and irradiation conditions. The lower baseline temperature observed in the control group reflects normal body temperature under anesthesia, whereas the PDA-HP-treated mice displayed localized heating confined to the lesion area. The surface temperature during irradiation remained below 43 °C, consistent with mild photothermal therapy, and the actual intramedullary temperature was likely slightly lower due to convective and perfusive heat dissipation within spinal tissue.

Concentration-dependent heating behavior was further confirmed by the temperature-time curves (Fig. 2C): higher PDA concentrations produced faster and greater temperature elevations. The 10 mg/mL formulation was selected as the optimal condition, as it achieved a controllable thermal rise within the mild hyperthermia window—an important safety threshold for promoting mitochondrial recovery while avoiding tissue damage. These in vitro measurements, performed in hydrated PBS rather than dry gels, align with previous photothermal hydrogel studies [[31], [32], [33]] and thus provide physiologically relevant insight into heat diffusion and safety behavior.

Collectively, the PDA-HP hydrogel demonstrated reliable, repeatable, and tunable photothermal effects suitable for localized and controllable hyperthermia in SCI therapy, balancing efficacy with thermal safety.

3.3. Biosafety of PDA-HP hydrogel

The biosafety of PDA-HP hydrogel was evaluated both in vivo and in vitro. For the in vivo assessment, PDA-HP hydrogel was injected into mice, and their general health status was monitored for 28 consecutive days. Parameters including secretions, excreta, autonomic nervous activity (tearing, piloerection, pupil size, respiratory pattern), body weight, skin/eye/mucous membrane appearance, behavioral activity, responsiveness, and food intake were recorded. No significant differences were observed between the treated and control groups throughout the observation period. After 4 weeks, major organs (heart, liver, spleen, lungs, and kidneys) were collected for HE staining, which revealed no histopathological abnormalities or tissue damage (Fig. 3A–E), confirming the in vivo safety of PDA-HP hydrogel.

Fig. 3.

Biosafety of PDA-HP hydrogel. (A–E) HE staining of major organs—heart (A), liver (B), spleen (C), lungs (D), and kidneys (E)—in SCI mice treated with PBS or PDA-HP/NIR for 4 weeks. Scale bars are 100 μm and 50 μm, respectively. (F) Live/dead staining of ND7/23 cells in the CON and PDA-HP/NIR groups, the number of live cells (mean ± SD, n = 6, ^nsp = 0.335), and the percentage of live cells (mean ± SD, n = 6, ^nsp = 0.303). Scale bars are 100 μm.

In vitro biocompatibility was further assessed using a live/dead assay in ND7/23 cells. Most cells were stained green (viable cells), with no statistically significant difference in live cell counts between the PDA-HP/NIR group and the CON group, indicating excellent cytocompatibility. A few scattered red-stained (dead) cells were observed, confirming minimal cytotoxicity of the PDA-HP hydrogel and demonstrating that the images shown in Fig. 3F are representative of the overall results (Fig. 3F).

3.4. PDA-HP/NIR promotes SCI repair

After confirming the photothermal performance and biosafety of the PDA-HP hydrogel, its therapeutic efficacy was evaluated in SCI mice under NIR laser irradiation (808 nm, 1 W/cm²). Mice were randomly assigned to four groups: Sham, SCI, PDA-HP, and PDA-HP/NIR. The Sham group underwent sham surgery only, while the other groups received PBS, PDA-HP hydrogel, or PDA-HP hydrogel combined with NIR irradiation (30 min/day) following spinal cord injury.

Motor recovery was assessed using the Basso Mouse Scale (BMS), footprint analysis, and the rotarod test (RRT). Sham mice exhibited normal motor function. As shown in Fig. 4A, BMS scores in the PDA-HP/NIR group were significantly higher than those in the SCI group from day 7 post-injury (p = 0.0035). Footprint analysis (Fig. 4B) revealed that Sham mice walked with regular gait and stable hind-limb placement. In contrast, SCI mice displayed complete hind-limb paralysis, dragging their hind limbs and leaving long, continuous marks on the paper. The PDA-HP/NIR group demonstrated markedly improved gait coordination, with partial paw support, more stable hind-limb placement, and relatively coordinated steps compared with the SCI group. Similar trends were observed in the RRT (Fig. 4C and D), where the PDA-HP/NIR group showed longer latency to fall and slower rod-drop speed than SCI mice, indicating enhanced motor coordination and balance. Overall, these results suggest that PDA-HP/NIR treatment accelerates and enhances motor function recovery after SCI.

Fig. 4.

PDA-HP/NIR promotes SCI repair. (A) BMS scores of various groups 28 days after SCI. The data are reported as mean value ± SD (n = 6, ∗∗p < 0.01, ∗∗∗p < 0.00l vs. SCI group). (B) Results of footprint in various groups 28 days after SCI (dpi: days post injury). (C, D) Results of the rota rod test (RRT) in various groups 28 days after SCI. The data are reported as mean value ± SD (n = 12, ∗p < 0.05, ∗∗p < 0.0 l). (E) HE staining results of spinal cord tissues in various groups of mice. Scale bars are 200 μm and 50 μm, respectively. (F) Masson’s Trichrome staining results of spinal cord tissues in various groups of mice. Scale bars are 200 μm and 50 μm, respectively. (G) HE staining results of the knee-joint specimens (femur-tibia) in various groups of mice. Scale bars are 200 μm.

Histopathological evaluation was performed using HE (Fig. 4E) and Masson’s trichrome staining (Fig. 4F). Sham mice exhibited normal spinal cord morphology, dense tissue structure, large and round nuclei, and no obvious pathological changes. SCI mice showed extensive neuronal necrosis and atrophy around the lesion site, severe loss of white and gray matter, disorganized nerve fibers, and glial scar hyperplasia, resulting in markedly larger cavity areas. Both PDA-HP and PDA-HP/NIR groups demonstrated varying degrees of histological improvement, with the PDA-HP/NIR group showing the smallest cavity area, more organized nerve fibers, and reduced collagen fiber deposition, indicating superior recovery of spinal cord structure.

To further evaluate limb pathology, HE staining of the right femur was performed 28 days post-treatment. SCI mice exhibited severe osteoporosis, including reduced bone density, trabecular thinning, sparsity, fractures, and disorganized trabecular structure. In contrast, the PDA-HP/NIR group displayed significantly improved trabecular architecture, increased bone density, and enhanced trabecular thickness, suggesting that PDA-HP hydrogel treatment under NIR irradiation mitigates osteoporosis in paralyzed limbs and further supports motor function recovery (Fig. 4G).

In summary, these findings indicate that PDA-HP hydrogel treatment under NIR laser irradiation effectively promotes motor function recovery following SCI.

3.5. PDA-HP/NIR treatment suppresses glial activation and promotes a shift toward anti-inflammatory polarization following SCI

To evaluate whether PDA-HP–based mild photothermal therapy modulates the post-injury neuroinflammatory microenvironment, we performed immunofluorescence staining for astrocytic (GFAP), microglial (Iba1) [34], and macrophage/microglial polarization markers (Arg-1 and iNOS) [35]. As shown in Fig. 5A–C, SCI induced pronounced astrogliosis and microglial activation, evidenced by markedly elevated GFAP and Iba1 expression in the SCI group compared with the Sham group. PDA-HP treatment alone moderately reduced GFAP and Iba1 intensity, whereas the PDA-HP/NIR group exhibited a significantly greater reduction (p < 0.0001), consistent with the reduced fibrosis and improved tissue preservation observed in Masson staining (Fig. 4F). Quantitative analysis confirmed that PDA-HP/NIR decreased GFAP (66.2 %) and Iba1(68.6 %) levels by approximately 60–70 % relative to SCI alone, indicating effective suppression of reactive gliosis.

Fig. 5.

PDA-HP/NIR treatment suppresses glial activation and promotes a shift toward anti-inflammatory polarization following SCI. (A) Representative fluorescence images and quantification showing GFAP (green) and IBA1 (red) staining of spinal cord sections in each group. Scale bars are 10 μm. Quantification of mean fluorescence intensity is shown in (B, C). Data are presented as mean ± SD (n = 6, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001). (D) Representative fluorescence images and quantification showing Arg-1 (green) and iNOS (red) staining of spinal cord sections in each group. Scale bars are 10 μm. Quantification of mean fluorescence intensity is shown in (E, F). Data are presented as mean ± SD (n = 6, ∗p = 0.011, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

We also examined polarization states of infiltrating macrophages and activated microglia (Fig. 5D–F). SCI resulted in a strong pro-inflammatory response, characterized by high iNOS (M1-like) expression and markedly reduced Arg-1 (M2-like) signal. PDA-HP treatment partially restored Arg-1 expression and attenuated iNOS positivity. Notably, PDA-HP/NIR treatment produced the most favorable polarization profile, showing a significant increase in Arg-1 (p < 0.0001) and a pronounced reduction in iNOS (p < 0.0001). This indicates that mild photothermal modulation promotes a shift toward an anti-inflammatory, tissue-repairing phenotype.

Collectively, these findings demonstrate that PDA-HP/NIR therapy not only mitigates reactive gliosis but also rebalances the inflammatory microenvironment by enhancing M2-like polarization and suppressing harmful M1-like responses. This immunomodulatory role likely contributes to the overall neuroprotective effect observed in functional and histological outcomes.

3.6. PDA-HP hydrogel under NIR laser irradiation improves mitochondrial function

Local tissue hyperthermia has been widely reported to enhance mitochondrial activity [12,36]. As the central hub regulating diverse cellular processes, mitochondria are highly sensitive to stimuli such as heat, ischemia, hypoxia, and oxidative stress, and are particularly vulnerable to temperature and redox perturbations. To assess mitochondrial functional changes following PDA-HP hydrogel treatment with NIR irradiation, we performed a series of in vivo and in vitro experiments.

In spinal cord tissues, adenosine triphosphate (ATP) (Fig. 6A) and malondialdehyde (MDA) (Fig. 6B) levels were quantified. Compared with the sham group, SCI markedly reduced ATP content and elevated MDA levels, indicating mitochondrial dysfunction. These alterations were significantly ameliorated in the PDA-HP/NIR group (p = 0.0002, p = 0.0005), suggesting improved mitochondrial activity.

Fig. 6.

PDA-HP hydrogel treated under NIR laser irradiation could improve the mitochondrial function. (A) ATP level assay of spinal cord tissue samples of various groups. The data of our experiments are reported as mean value ± SD (n = 6, ^nsp = 0.6466, and ∗∗∗p < 0.001). (B) MDA assay of spinal cord tissues in various groups. The data of our experiments are reported as mean value ± SD (n = 6, ∗p = 0.0427, and ∗∗∗ p < 0.001). (C) LDH assay of ND7/23 cells in various groups. The data of our experiments are reported as mean value ± SD (n = 6, ∗ p = 0.0443, and ∗∗∗∗ p < 0.0001). (D) Results of MMP measured by JC-1. Scale bars are 50 μm. (E) The statistical chart of the ratio of JC-1aggregates to JC-1 monomers in various groups. The data of our experiments are reported as mean value ± SD (n = 6, ∗∗∗ p = 0.0004, and ∗∗∗∗ p < 0.0001). (F) Immunofluorescence of ROS in various groups. Scale bars are 100 μm. (G) The statistical chart of DCF fluorescence. The data of our experiments are reported as mean value ± SD (n = 6, ∗ p = 0.0182, and ∗∗∗∗ p < 0.0001).

To further corroborate these findings in vitro, ND7/23 cells were subjected to oxidative damage induced by 100 μM T-BHP for 6 h (Fig. S1), as optimized from our previous work [8]. The effects of PDA-HP/NIR treatment were evaluated by lactate dehydrogenase (LDH) release (Fig. 6C) (p < 0.0001), mitochondrial membrane potential (MMP) (Fig. 6D and E) (p < 0.0001), and reactive oxygen species (ROS) levels (Fig. 6F and G) (p < 0.0001). PDA-HP/NIR treatment significantly reduced LDH leakage and ROS accumulation compared with the T-BHP group. Notably, MMP—an early and sensitive indicator of mitochondrial function—was markedly restored in the PDA-HP/NIR group (p < 0.0001), as assessed by JC-1 staining (Fig. 6D and E).

Collectively, these results demonstrate that PDA-HP hydrogel under NIR irradiation attenuates oxidative stress, preserves mitochondrial integrity, and improves bioenergetic function, thereby exerting anti-apoptotic and neuroprotective effects after SCI.

3.7. PDA-HP/NIR uniquely suppresses SCI-driven HSP70 surge

Quantitative proteomic profiling using 4D-DIA identified a total of 7938 proteins across Sham, SCI, PDA-HP, and PDA-HP/NIR groups (Fig. 7F). Volcano plots revealed that SCI markedly altered the proteome compared with Sham, with 54 proteins up-regulated and 33 down-regulated (Fig. 7A). Among the most prominent changes was a strong induction of heat shock protein 70 (HSP70), showing a log₂ fold change (log₂ FC) of 2.34 and a false discovery rate (FDR) of 3.2 × 10⁻⁴.

Fig. 7.

Quantitative proteomic profiling of spinal cord tissue 7 days after SCI and treatment with PDA-HP hydrogel ± NIR irradiation. (A–D) Volcano plots of differentially expressed proteins (DEPs; |log₂ FC| ≥ 1, FDR <0.05) for the indicated comparisons: (A) SCI vs. Sham, (B) PDA-HP/NIR vs. SCI, (C) PDA-HP vs. SCI, and (D) PDA-HP/NIR vs. PDA-HP. (E) Upset plot showing overlap of DEPs among the four comparisons. (F) Venn diagram of up-regulated proteins for each treatment relative to SCI. (G) Bar chart summarizing the number of up- and down-regulated proteins in each comparison. (H) Heat map of the top 50 most variable proteins (z-score) across all groups. Proteomic data were acquired by 4D-DIA LC-MS/MS (n = 4 biological replicates per group). Notably, SCI induced a marked up-regulation of HSP70, which was almost completely reversed by PDA-HP/NIR treatment.

Treatment with PDA-HP/NIR reversed most SCI-induced alterations, yielding 81 up-regulated and 41 down-regulated proteins relative to SCI (Fig. 7B), and markedly reduced the HSP70 overexpression toward the physiological baseline observed in the Sham group (log₂ FC = −1.97, FDR = 1.1 × 10⁻³). PDA-HP alone provided a more modest rescue, with 59 up-regulated and 25 down-regulated proteins (Fig. 7C), partially suppressing HSP70 expression (log₂ FC = −0.82). The direct comparison of PDA-HP/NIR versus PDA-HP revealed 14 proteins up-regulated and 16 down-regulated, again identifying HSP70 as the most significantly down-regulated protein (log₂ FC = −1.15, FDR = 9.5 × 10⁻⁴; Fig. 7D).

GO enrichment analysis of the differential proteins (Fig. 7E) showed significant enrichment in biological processes related to protein folding, response to unfolded proteins, and chaperone-mediated activities—categories in which HSP70 is a central component. KEGG pathway analysis (Fig. 7G) indicated that PDA-HP/NIR preferentially affected stress-response pathways, including the protein processing in endoplasmic reticulum and HIF-1 signaling pathways. The Venn diagram (Fig. 7F) and upset plot (Fig. 7E) positioned HSP70 within the PDA-HP/NIR-specific rescue set.

Unsupervised hierarchical clustering of the top 50 most variable proteins (Fig. 7H) clearly segregated all samples into four distinct groups, with expression patterns largely driven by the gradient of HSP70 and associated chaperone proteins. These findings collectively establish that selective and robust suppression of the SCI-driven HSP70 surge is a defining molecular hallmark of NIR-activated PDA-HP therapy.

3.8. Convergent in vitro-in vivo validation confirms HSP70 as a principal target of PDA-HP/NIR therapy

To corroborate the proteomic finding that heat-shock-protein 70 (HSP70) is a key downstream effector modulated by PDA-HP/NIR treatment, we conducted complementary in vitro and in vivo assays. In ND7/23 cells exposed to T-BHP, HSP70 expression was markedly elevated, consistent with a stress-response phenotype (Fig. 8A–C). This induction was significantly attenuated by PDA-HP and further suppressed upon NIR activation, indicating that photothermal stimulation amplified the inhibitory effect.

Fig. 8.

Convergent in vitro-in vivo validation confirms HSP70 as a principal target of PDA-HP/NIR therapy. (A, C) Immunofluorescence staining of HSP70 in different treatment groups. Quantification of mean fluorescence intensity is shown in (C). Data are presented as mean ± SD (n = 6, ∗∗∗∗ p < 0.0001). Scale bars are 50 μm. (B, D) Immunofluorescence staining of cleaved caspase-3 (C-cas3) in different treatment groups. Quantification of mean fluorescence intensity is shown in (D). Data are presented as mean ± SD (n = 6, ∗∗ p = 0.0062, and ∗∗∗∗p < 0.0001). Scale bars are 20 μm. (E–H) Western blot analysis of HSP70, Bcl-2, Bax, and β-actin in spinal cord tissues. Quantification of protein expression levels are shown in (F–H). Data are presented as mean ± SD (n = 3, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001).

Similarly, cleaved caspase-3 (C-cas3), a hallmark of apoptosis [37], was strongly upregulated in the T-BHP group but reduced by PDA-HP, with the most pronounced suppression observed in the PDA-HP/NIR group (Fig. 8B–D). Given the close relationship between mitochondrial dysfunction and apoptosis—both key features of traumatic injury—we further assessed C-cas3 levels in SCI tissue by immunofluorescence. The SCI group exhibited a robust increase in C-cas3 fluorescence intensity, which was markedly decreased following PDA-HP treatment (p = 0.0062) and further diminished under NIR irradiation (p < 0.0001), suggesting enhanced inhibition of apoptosis.

Western blot analysis further substantiated these findings. Compared with the SCI group (Fig. 8E–H), it revealed that PDA-HP/NIR not only downregulated HSP70 (p < 0.0001) but also reduced the expression of the pro-apoptotic protein Bax (p = 0.0193), while simultaneously restoring the levels of the anti-apoptotic protein Bcl-2 (p = 0.0026).

Collectively, these findings substantiate the proteomic evidence that NIR-activated PDA-HP specifically and potently suppresses SCI-induced HSP70 expression, thereby shifting the apoptotic equilibrium toward neuronal survival.

4. Discussion

Spinal cord injury (SCI) remains an intractable neurological disorder whose clinical burden is compounded by persistent mitochondrial dysfunction and progressive neurodegeneration. In this study, we developed a polydopamine–heparin poloxamer hydrogel (PDA-HP) capable of in situ gelation at physiological temperature and mild photothermal heating (∼42 °C) under 808 nm near-infrared (NIR) irradiation. Through integrated in vitro, in vivo, and quantitative proteomic analyses, we demonstrated that PDA-HP/NIR therapy promotes motor recovery, preserves mitochondrial homeostasis, restores ATP synthesis, and—most notably—normalizes the maladaptive elevation of HSP70 following SCI. These data establish a controllable, minimally invasive photothermal approach that targets mitochondria-specific stress pathways (Fig. 9).

Fig. 9.

Schematic illustration of the preparation and therapeutic mechanism of PDA-HP hydrogel for spinal cord injury (SCI).

Unlike prior photothermal strategies that rely primarily on nonspecific metabolic activation, our work identifies HSP70 normalization as a mechanistic mediator linking mild hyperthermia to mitochondrial recovery. This represents a conceptual shift from passive heating toward active modulation of stress-chaperone networks. Compared with conventional heating pads or water baths—which yield only modest enhancement of mitochondrial respiration [12]—the NIR-induced, PDA-mediated mild hyperthermia confines the thermal field to the lesion epicenter with minimal systemic impact. The absence of histopathological changes in peripheral organs reinforces the favorable biosafety profile of polydopamine-based materials [12,38] and extends it to a thermoresponsive hydrogel platform adaptable to irregular CNS architectures.

A separate NIR-alone control group was not included because prior studies have consistently demonstrated that 808 nm irradiation at 1 W/cm² for ≤30 min does not induce neuroregeneration or significant heating in the absence of a photothermal absorber [39,40]. In our design, the SCI group served as an appropriate reference for spontaneous recovery under matched surgical and anesthetic conditions. Thus, the therapeutic benefits observed in the PDA-HP/NIR group can be attributed primarily to the photothermal contribution of PDA rather than to NIR exposure itself.

Although infrared thermography was employed to monitor surface lesion temperature, direct intramedullary temperature mapping was not performed. Heat-transfer dynamics in vivo may differ from in vitro measurements due to tissue scattering, blood perfusion, and convective dissipation. Nonetheless, the spatial heating pattern we observed aligns with previously reported NIR-responsive hydrogels and deep-tissue photothermal systems [[41], [42], [43]]. Future work will integrate MR thermometry, microprobe mapping, and computational modeling to validate intramedullary heat distribution. Additionally, NIR-II irradiation (>1000 nm), which offers reduced scattering and deeper penetration, may further enhance translational potential [44]. Notably, NIR light effectively penetrated sutured skin and superficial tissues to reach the lesion center in our T9-T10 laminectomy model. This is consistent with the known penetration depth of 808 nm light in small animals and supports the feasibility of using minimally invasive optical fiber systems for targeted heating in future clinical applications [41,45,46].

Beyond mitochondrial protection, PDA-HP/NIR treatment exerted profound effects on the neuroinflammatory microenvironment. Immunofluorescence analysis demonstrated significant attenuation of GFAP⁺ astrogliosis and Iba1⁺ microglial activation, consistent with the reduced fibrosis and improved tissue preservation observed in Masson staining. Furthermore, Arg-1/iNOS staining revealed a clear shift toward an M2-like reparative macrophage/microglial phenotype, suggesting that photothermal modulation promotes a pro-regenerative immune landscape. These findings indicate that the therapeutic benefits of PDA-HP/NIR arise not only from mitochondrial restoration but also from coordinated regulation of glial reactivity and inflammatory polarization.

Our proteomic analysis identified HSP70 as the most significantly downregulated protein following PDA-HP/NIR therapy, correlating with Bax/Bcl-2 rebalancing and reduced apoptosis. Importantly, this effect should be interpreted as normalization of pathological HSP70 overactivation rather than its complete suppression. Although transient HSP70 induction is beneficial for early proteostasis, sustained overexpression is known to stabilize pro-apoptotic client proteins and exacerbate inflammatory cascades [47]. The combined mitochondrial, inflammatory, and proteomic findings support a context-dependent model wherein PDA-HP/NIR restores physiological HSP70 dynamics to promote neuroprotection. Importantly, numerous studies have demonstrated that HSP70 exerts dual and cell-type–dependent effects in CNS injury [6,48]. In neurons, transient or moderate HSP70 upregulation promotes mitochondrial stability and inhibits apoptosis, supporting its well-recognized neuroprotective role [49,50]. However, in astrocytes and microglia, prolonged or excessive HSP70 activation can amplify CD14/TLR-mediated inflammatory signaling, thereby contributing to glial reactivity and secondary injury [51,52]. These contrasting actions indicate that the consequences of HSP70 elevation depend strongly on cellular context and temporal patterns. Because our study quantified only bulk-tissue HSP70 levels, we were unable to distinguish neuron-specific from glia-specific regulation, which represents an important limitation. Future work incorporating cell-type–resolved analyses (e.g., co-localization staining or conditional Hspa1a/b deletion in neurons vs. astrocytes) will be needed to clarify how PDA-HP/NIR modulates HSP70 dynamics across distinct cell populations. Future studies incorporating HSP70-specific genetic or pharmacological manipulation will help delineate its cell-type–specific roles in the photothermal response.

The PDA-HP system also provides intrinsic ROS-scavenging activity and is likely biodegradable under physiological conditions [8,53], consistent with prior reports on poloxamer-based hydrogels and polydopamine composites [54,55]. Although in vivo degradation and retention were not quantified in this study, existing literature suggests that PDA-HP hydrogel likely exhibits a transient degradation pattern suitable for subacute neural repair [53,54], and fluorescence-labeled tracking and ex vivo analysis will be pursued to confirm kinetic profiles and long-term biosafety. The relatively soft mechanical modulus (∼500 Pa) compared with native spinal tissue (2–5 kPa) may limit its load-bearing capacity [56,57]; future designs may incorporate composite networks for enhanced mechanical stability.

This study has several limitations. The sample size was powered to detect major behavioral differences but may overlook subtle functional changes. The mouse model’s limited tissue thickness also facilitates NIR penetration, posing challenges for translation to humans with larger spinal diameters. Upstream regulatory events governing HSP70 activity (e.g., HSF-1 phosphorylation or CHIP-mediated ubiquitination) were not explored [58]. In addition, direct gain- or loss-of-function validation (e.g., HSP70 overexpression or inhibition) was not performed in this study; however, the proteomic and immunoblot results consistently support a mechanistic link between HSP70 normalization and mitochondrial recovery. Future work will include targeted HSP70 modulation in ND7/23 or primary neuron-glia co-culture systems to confirm its causal role in PDA-HP/NIR-mediated protection. Similarly, although this revision partially validates the immunomodulatory effects of PDA-HP/NIR through GFAP/Iba1 and Arg-1/iNOS staining, deeper mechanistic dissection of glial-neuronal interactions still necessary [34,35]. Additionally, electrophysiological or tract-tracing analyses were not conducted to directly assess axonal regeneration.

Future work will include large-animal studies with real-time thermometry, optical-fiber–based NIR delivery, and quantitative gait analysis; the use of conditional knockout models to dissect cell-type–specific contributors; integration of multimodal theranostic platforms; and combination strategies with mitochondrial antioxidants or neurotrophic factors to maximize therapeutic synergy. Conceptually, this study reframes mild photothermal therapy from a passive metabolic enhancer to an active modulator of stress-chaperone networks and the immune microenvironment, highlighting its potential for precision neuromodulation tailored to the metabolic and inflammatory profiles of individual SCI patients.

Dopamine (DA) undergoes alkaline oxidative self-polymerization to form PDA nanoparticles, which are subsequently combined with heparin-poloxamer (HP) hydrogel precursor to construct the PDA-HP hydrogel. The hydrogel exhibits a sol-gel transition (sol at room temperature, ∼25 °C; gel at body temperature, ∼37 °C), enabling in situ injection at the lesion site. Upon 808 nm NIR irradiation, PDA-HP hydrogel generates mild photothermal effects that reduce ROS accumulation, attenuate LDH leakage, restore mitochondrial membrane potential (ΔΨm) and ATP production, and normalize the maladaptive surge of HSP70 expression. These processes collectively preserve mitochondrial integrity, inhibit apoptosis, and ultimately promote neuroprotection and functional recovery after SCI.

5. Conclusion

In summary, we developed a dual-responsive PDA-HP hydrogel that enables controllable mild photothermal stimulation to promote functional recovery after SCI. PDA-HP/NIR therapy restored mitochondrial homeostasis, normalized maladaptive HSP70 overactivation, reduced apoptosis, and reshaped the neuroinflammatory landscape by attenuating GFAP⁺/Iba1⁺ glial activation and promoting M2-like polarization. These multifaceted protective effects translated into significant behavioral improvement and tissue preservation. Together, our findings identify mitochondria-targeted mild photothermal modulation as a promising and clinically translatable therapeutic strategy for spinal cord repair and establish stress-chaperone regulation as a key mechanistic target for future SCI interventions.

CRediT authorship contribution statement

Yi Li: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Jiaxuan Hu: Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Bing Ran: Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Huisheng Zhong: Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Nayin Zhong: Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Yi Zhong: Visualization, Validation, Supervision, Software, Resources, Project administration. Xinyu Fu: Visualization, Validation, Supervision, Software, Resources, Project administration. Xinying Liu: Visualization, Validation, Supervision, Software, Resources, Project administration. Guanghua Wu: Visualization, Validation, Supervision, Software, Resources, Project administration. Qinwen Zhong: Visualization, Validation, Supervision, Software, Resources, Project administration. Juan Li: Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization.

Ethics approval and consent to participate

All the animal experiments in this work were performed in accordance with the Ethical Committee of Care and Use of Laboratory Animals at Gannan Medical University and approved by the First Affiliated Hospital of Gannan Medical University (Permit No. LLSC-2023031601).

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:Yi Li reports financial support was provided by Jiangxi Provincial Natural Science Foundation. Yi Li reports financial support was provided by Ganzhou Municipal Health Commission. If there are other authors, they 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 is the Supplementary data to this article.

Data availability

Data will be made available on request.