Near-infrared light-driven nanomotors-based microneedles for the active therapy of bacterial infected acne

Abstract

Acne, caused by Cutibacterium acnes, triggers inflammatory lesions. The hypoxic biofilm microenvironment exacerbates acne progression, while inadequate hydrogen peroxide and dense biofilm barriers hinder oxygen generation and nanomedicine penetration. Here, we develop microneedles patch loaded with near infrared-driven self-oxygenating Z@P-M nanomotors for photothermal therapy of acne. Zinc peroxide nanoparticles are asymmetrically modified with polydopamine, followed by in-situ manganese dioxide growth on polydopamine to form Z@P-M nanomotors. Z@P-M nanomotors loaded microneedles penetrate the stratum corneum to deliver antibacterial nanoparticles into the dermis. In female murine acne, Zinc peroxide slowly releases hydrogen peroxide in acidic biofilm, catalyzed by manganese dioxide to generate oxygen, thus alleviating hypoxia and skin inflammation. After near infrared laser irradiation, the thermal gradient generated by the asymmetrically modified polydopamine coating endows the nanomotors with enhanced diffusion to promote biofilm penetration, further improving photothermal therapy efficacy and showing a potential for active acne treatment.

Introduction

Acne is a common skin disease with the highest prevalence between the ages of 15 and 20. The erythematous lesions and scarring caused by acne have a profoundly negative impact on the emotional and mental health of adolescents^1,2. In general, the anaerobic Cutibacterium acnes (C. acnes) are the main cause of acne. C. acnes can proliferate and release lipase, which decompose the accumulated sebum and produce free fatty acids, resulting in hypoxia, and eventually stimulating the secretion of pro-inflammatory factors from the skin cells, such as tumor necrosis factor-alpha (TNF-α) and interleukins (ILs)^3,4,5. Moreover, skin-resident innate lymphoid cells (ILCs) participate in the regulation and maintenance of skin microbial homeostasis⁶. In particular, the reduction of retinoic acid receptor-related orphan receptor gamma t (RORγt⁺) ILCs leads to an abnormal proliferation of sebaceous glands and increases sebum secretion, which is remarkably similar to the acne phenotype as well as to the adaptive survival environment of C. acnes⁷. Interestingly, bacterial extracellular vesicles (bEVs) of C. acnes can significantly reduce the amount of the skin ILCs, which may suggest that excessive accumulation of C. acnes could disrupt the homeostatic balance of skin-resident ILCs, thus leading to inflammatory skin lesions⁸. Therefore, developing a new strategy to alleviate hypoxia and inflammation induced by C. acnes infection and to restore the homeostasis of ILCs is significant for acne treatment. In clinic, mild or moderate acne cases are generally treated with topical medications, while moderate or severe acne cases are treated with oral antibiotics. However, frequent administration of antibiotics can result in severe side effects^2,9. Moreover, biofilm formation by C. acnes is generally resistant to conventional antibiotics due to the protective effect of extracellular polymeric substances (EPS) of C. acnes biofilm⁴. Consequently, localized application of antimicrobial agents directly offers a promising therapeutic strategy for combating acne formation. Microneedles (MNs) systems have the advantage of high transdermal efficiency, favorable patient compliance, and low frequency of drug delivery^10,11. Nanoparticles (NPs) loaded MNs can penetrate easily into the stratum corneum and subsequently deliver antibacterial nanomedicines to the dermis for the treatment of acnes¹². However, dense EPS in bacterial biofilm further prevents the penetration of antibacterial materials and reduces the therapeutic efficacy compared to pelagic bacteria¹³. Photothermal therapy (PTT) is non-invasive, highly controllable and wide-spectrum bactericidal. Biocompatible photothermal agents (PTAs) such as polydopamine (PDA) can convert light into thermal energy, thereby generating high temperature that dissociates the dense structure of the biofilm and leads to bacterial death, showing great potential in treating drug-resistant bacteria and eradicating biofilms^14,15. However, photothermal NPs still have limited penetration in bacterial biofilms, which hinders the therapeutic efficacy of PTT¹³.

Self-propelled nanomotors can overcome these drawbacks by converting external energies such as chemical reaction, light, magnetic field, or ultrasound into mechanical motion, which demonstrate significant potential in biomedicine for applications like targeted drug delivery and constructing mechanical-antibiofilm systems^16,17,18,19, especially near-infrared (NIR) light-driven nanomotors have aroused increasingly widespread research interest in eradicating bacterial biofilms²⁰. Under 808 nm NIR laser irradiation, nanomotor generates a thermal gradient thereby triggering the self-thermophoresis, leading the nanomotor to propel in the direction opposite to the temperature gradient²¹, which endows the nanomotor with a strong power to penetrate the biofilm. Although biofilm-infected tissues show relatively higher concentration of hydrogen peroxide (H₂O₂) compared to normal tissues, the concentration of H₂O₂ in the biofilm microenvironment is still insufficient to maintain a continuous supply of oxygen (O₂)²². Therefore, it is necessary to selectively increase the level of H₂O₂ in the biofilm, and converts them into O₂ to alleviate hypoxia of biofilm microenvironment. Zinc peroxide (ZnO₂) can degrade slowly and continuously in an acidic microenvironment and produce large amounts of H₂O₂²³, which provides sufficient fuel for oxygen production. With the development of nanotechnology, manganese dioxide (MnO₂) with catalase (CAT) mimetic activity is one of the most promising nanozymes. It can effectively catalyze the continuous supply of O₂ from microenvironmental H₂O₂, alleviating hypoxia and oxidative stress to reduce inflammatory responses²⁴. Therefore, it is essential to design and fabricate oxygen self-supplying and self-thermophoretic nanomotors to overcome the biofilm barriers and alleviate hypoxia for active acne treatment.

Herein, we present an intelligent design of MNs encasing self-thermophoretic nanomotors driven by 808 nm NIR laser to alleviate hypoxia and eradicate biofilms for active acne treatment. Precise asymmetric regulation of Janus nanomotors enables the integration of materials with different functions into the single system, which is the basis for the self-propelled behavior of the nanomotor. Janus nanomotors are synthesized by asymmetrically wrapping photothermal PDA onto ZnO₂ (Z), following by MnO₂ growing on PDA in situ. The resulting nanomotors are then loaded into sodium hyaluronate MNs, here referred to Z@P-M MNs (Fig. 1). Upon pressing onto the skin, Z@P-M MNs are dissolved rapidly to release Z@P-M nanomotors in dermis. Asymmetrically grown PDA absorbs 808 nm NIR laser to generate a thermal gradient to induce self-thermophoresis, which endows the nanomotor with enhanced diffusion to penetrate the dense bacterial biofilm. The released Z@P-M nanomotors produce excess H₂O₂ in the acidic biofilm microenvironment and continuously supplies oxygen catalyzed by MnO₂, which not only alleviates the biofilm hypoxia and reduces inflammation, but also synergizes with PTT effect to eliminate the C. acnes biofilm, thereby restoring the ILCs homeostasis of acne.

Fig. 1: Fabrication of Z@P-M MNs and the active treatment of acne.

Z@P-M MNs were dissolved and the loaded Z@P-M nanomotors were rapidly released in the dermis. The asymmetrically grown PDA absorbed 808 nm NIR light to generate self-thermophoresis, which allowed motion-promoted enhanced penetration of dense bacterial biofilms. The released Z@P-M nanomotors generated excess H₂O₂ in the acidic biofilm microenvironment and sustained oxygenation catalyzed by MnO₂, which not only alleviated biofilm hypoxia and reduced inflammation, but also cleared C. acnes biofilm synergistically with the PTT effect, thus restoring the homeostasis of ILCs in acne.

Results

Fabrication and characterization of Z@P-M MNs and Z@P-M nanomotors

As shown in Fig. 1, ZnO₂ NPs were first synthesized via a simple one-pot method. Briefly, zinc chloride (ZnCl₂), poly (viny pyrrolidone) (PVP), and ammonia were dispersed in anhydrous ethanol, and hydrogen peroxide was then slowly added to form monodisperse ZnO₂. By regulating the interfacial energy difference²⁵, polyacrylic acid (PAA) network was asymmetrically wrapped on ZnO₂ to synthesis ZnO₂@PAA (Z@PAA), followed by PDA formation on PAA via dopamine hydrochloric polymerization to obtain Janus ZnO₂@PDA NPs (Z@P). The oxidation of potassium permanganate (KMnO₄) was then utilized to grow MnO₂ in situ on the PDA layer, referred as ZnO₂@PDA-MnO₂ (Z@P-M) nanomotors. To fabricate MNs, sodium hyaluronate with favorable biocompatibility and solubility was selected as raw materials. Hyaluronic acid (HA) is a natural polysaccharide, and can maintain skin moisture and promote skin absorption⁴. Therefore, sodium hyaluronate solution containing Z@P-M nanomotors was filled into the mold, vacuumed, dried and demolded to obtain Z@P-M MNs. As shown in Fig. 2a, b and Supplementary Fig. 1, The Z@P-M MNs patch size is 15 mm × 15 mm, featuring a grid of 225 MNs arranged in a uniform pattern. Scanning electron microscope (SEM) mapping images showed the elemental distribution of C, N, O, Zn, and Mn, indicating that Z@P-M was homogeneously distributed in sodium hyaluronate. Each MN stood at 637 μm in height, a dimension sufficient to traverse both the stratum corneum and epidermal layers (~100 μm) for targeted delivery to the dermal tissue²⁶. The diameter of MN tip was 28.19 μm and this size also ensured the sufficient pressure for the skin penetration²⁷. Sufficient mechanical strength is also important for MNs to penetrate the skin. In Fig. 2c and Supplementary Fig. 2, the stress-strain behavior of each MN was derived from the real-time monitoring of displacement-force variations during mechanical testing. A gradual increase in mechanical strength was clearly observed when sodium hyaluronate concentration was increased from 50 mg/mL to 150 mg/mL, and the compressive force of each MN at the 600 μm displacement was increased from 0.12 N to 0.26 N. In addition, the mechanical strength also gradually increased with increasing concentration of Z@P-M (ZnO₂ concentration = 50 − 100 μg/mL). The curve of Z@P-M MNs showed a continuous rising trend and the stress load was much larger than the minimum mechanical force required for MNs to penetrate the skin (0.098 N per needle)²⁸. MNs did not fracture during the compression, indicating that the patch had sufficient mechanical strength to penetrate the skin. In addition, the water absorption of MNs patch in a humid environment (70 % humidity) was also evaluated. The needle gradually absorbed water and eventually dissolved within 30 min (Supplementary Fig. 3). The results indicated that the MNs patch has promising hygroscopic properties and can provide a favorable environment for tissue repair.

Fig. 2: Fabrication and characterization of Z@P-M MNs.

a SEM images and EDS mapping of Z@P-M MNs (scale bar = 200 μm). n = 3 independent experiments. b Optical image of Z@P-M MNs (scale bar = 1 mm). c Mechanical properties of MNs patches with and without different concentrations of Z@P-M (ZnO₂ concentration = 50 − 100 μg/mL) containing sodium hyaluronate concentration of 150 mg/mL. d TEM images of Z, Z@PAA, Z@P, and Z@P-M, and the corresponding EDS elemental mapping of Z@P-M (scale bar = 100 nm). n = 3 independent experiments. e Zeta potential of Z, Z@P, and Z@P-M. n = 3 independent experiments. f Diameter distribution of Z, Z@P, and Z@P-M. g XRD of Z, Z@P, and Z@P-M. JCPDS cards of ZnO₂ (No. 13-0311) and MnO₂ (No. 44-0141) were provided for comparison. h Mn 2p and O 1 s XPS spectra of Z@P-M. Data in (e) were presented as Mean ± SD. Source data are provided as a Source Data file.

For the characterization of the loaded Z@P-M, transmission electron microscopy (TEM) images showed that asymmetric PDA and MnO₂ layers were clearly observed on one side of ZnO₂. Energy-dispersive X-ray spectroscopy (EDS) mapping further determined the elemental composition of Z@P-M. It was noteworthy that Zn was highly concentrated in the center, and C, N, and Mn were asymmetrically wrapped around Zn, demonstrating the asymmetric structure of MnO₂-modified PDA shell (Fig. 2d, and Supplementary Fig. 4). The spherical ZnO₂ particles were distributed uniformly with a particle diameter of 167.00 nm. When PDA layer was modified, the particle diameter increased to 221.82 nm, and the zeta potential also changed from 1.62 ± 0.08 mV to -13.27 ± 0.31 mV. After the successful deposition of MnO₂ on the surface of PDA, the particle diameter and zeta potential were 240.87 nm and -8.21 ± 0.06 mV, respectively (Fig. 2e, f). In addition, the diameters of Z@P-M before and after 808 nm NIR irradiation (1 W/cm², 5 min) were 242.03 ± 1.89 nm and 240.90 ± 0.40 nm, respectively, and the zeta potentials were -8.49 ± 0.22 mV and -8.42 ± 0.07 mV, respectively, which was almost similar. TEM images exhibited no obvious degradation of Z@P-M before and after 808 nm NIR irradiation, and still maintained the Janus structure, indicating that Z@P-M kept structural stability (Supplementary Fig. 5). Raman spectra of ZnO₂ and ZnO were showed in Supplementary Fig. 6, where ZnO was the control. It was observed that the characteristic peak of ZnO was observed around 437 cm^-1, associating with the E₂ vibrational modes of O atoms. While the characteristic peak of ZnO₂ was around 840 cm^-1, which belonged to the O-O bond stretching band of the peroxide ion^29,30. The above results indicated that ZnO₂ was successfully synthesized. Supplementary Fig. 7 showed the Fourier transform infrared spectrometer (FTIR) analysis of Z, Z@P, and Z@P-M. Z, Z@P, and Z@P-M exhibited an infrared absorption band at 1420 cm^-1, attributed to the vibrational mode of peroxide ions³¹. The FTIR spectrum of Z showed a sharp absorption band at 440 cm^-1, matching the Zn-O vibration band^32,33, confirming the successful synthesis of ZnO₂. Z@P and Z@P-M exhibited absorption peaks at 1595 cm^-1, corresponding to the C=O bond of PDA³⁴. Z@P-M showed an absorption peak at 538 cm^-1 corresponding to the Mn-O bond in MnO₂³⁵. Figure 2g presented the X-ray Powder diffractometer (XRD) pattern of Z, Z@P, and Z@P-M. The yellow and blue lines represented Z and Z@P, respectively, which matched the joint committee on powder diffraction standards (JCPDS) card of ZnO₂. The pink line represented Z@P-M, showing a diffraction peak at 18.75°, corresponding to (200) of MnO₂ crystallographic faces. In addition, the pattern of Z@P-M showed diffraction peaks at 31.38° and 62.86°, corresponding to (111) and (311) of ZnO₂ crystallographic faces, indicating that Z@P-M was successfully synthesized and ZnO₂ exhibited no apparent degradation during the synthesis of Z@P-M. C, N, O, Zn, and Mn of Z@P-M were further investigated by X-ray photoelectron spectroscopy (XPS), with elemental contents of 43.11 %, 2.83 %, 37.83 %, 6.36 %, and 9.87 %, respectively (Supplementary Fig. 8). The peaks at 642.08 eV and 653.68 eV corresponded to the Mn 2p_3/2 and Mn 2p_1/2 spin-orbit peaks of MnO₂, respectively. The Mn 2p_3/2 and 2p_1/2 spectra were each divided into two characteristic peaks, with those at 641.58 eV and 653.18 eV corresponding to Mn³⁺, and those at 642.68 eV and 654.28 eV corresponding to Mn⁴⁺³⁶. The peaks in the O 1 s spectrum mainly belonged to the O-H bond (532.78 eV), C-O (531.28 eV), and Mn-O-Mn bond (529.58 eV)³⁷, thus confirming the presence of MnO₂ in Z@P-M, consistent with the XRD and FTIR results (Fig. 2h).

After successful fabrication of Z@P-M, the photothermal properties were then evaluated by real-time infrared (IR) thermal camera. As shown in Fig. 3a, the IR photothermal images presented a clear temperature difference between PBS, Z, Z@P, and Z@P-M during 10 min of 1 W/cm² 808 nm NIR irradiation, which was consistent with the photothermal curves shown in Fig. 3b. After 10 min of 1 W/cm² 808 nm NIR irradiation, the temperatures of Z, Z@P, and Z@P-M were increased to 32.1 °C, 42.1 °C, and 45 °C, respectively, while the PBS group only increased by 0.8 °C. It was notable that under the same 808 nm NIR irradiation (10 min, 1 W/cm²), Z@P-M exhibits better photothermal effects compared to Z@P. It was attributed to Z@P-M combining PDA and MnO₂, two excellent photothermal agents, thereby exhibiting a significant cumulative photothermal effect³⁸. Z@P-M with 100 μg/mL of ZnO₂ concentration was also irradiated with 808 nm NIR laser with different intensities for 10 min (0.5, 1, and 1.5 W/cm²), and the temperatures were increased to 35.8 °C, 45.1 °C and 51.2 °C, respectively (Supplementary Fig. 9a). When the 808 nm NIR power of 1 W/cm² was applied for 10 min, the temperatures increased sequentially (38.6 − 49.6 °C) with increasing concentrations of Z@P-M (ZnO₂ concentration = 50 − 200 μg/mL), indicating that the photothermal effect of Z@P-M was proportional to the sample concentration (Supplementary Fig. 9b). In addition, Z@P-M presented almost consistent heating/cooling curves after 8 cycles of 808 nm NIR irradiation (1 W/cm²) (Fig. 3c), which proved that Z@P-M had good photothermal stability. The photothermal conversion efficiency (η) of Z@P-M was also calculated, and the η value was 38.66 %, indicating that the developed Z@P-M was a potential photothermal agent (Fig. 3d). Moreover, the IR photothermal images and temperature rise curves of the MNs patches showed that Z@P-M MNs patch had favorable photothermal conversion properties under 1 W/cm² 808 nm NIR irradiation for 10 min (Fig. 3e, f). Z@P-M MNs showed almost consistent heating/cooling curves after 8 cycles of 808 nm NIR irradiation (1 W/cm²), illustrated that Z@P-M MNs patch could withstand prolonged laser irradiation with high photothermal stability (Fig. 3g).

Fig. 3: Photothermal capabilities and acidic release of Z@P-M and Z@P-M MNs.

IR photothermal images (a) and temperature increase curves (b) of PBS, Z, Z@P, and Z@P-M under 10 min of 1 W/cm² 808 nm NIR irradiation. c 8 heating/cooling cycles of Z@P-M suspension (ZnO₂ concentration = 100 μg/mL) under 1 W/cm² 808 nm NIR irradiation. d Photothermal conversion efficiency of Z@P-M. IR photothermal images (e) and temperature increase curves (f) of blank MNs and Z@P-M MNs under 5 min of 1 W/cm² 808 nm NIR irradiation. g 8 heating/cooling cycles of Z@P-M MNs under 1 W/cm² 808 nm NIR irradiation. H₂O₂ (h) and O₂ (i) release profiles of Z, Z@P, and Z@P-M under pH 7.4 and 5.5. n = 3 independent experiments. Data in (h) was presented as Mean ± SD. Source data are provided as a Source Data file.

The endogenous H₂O₂ concentration in biofilm-infected tissues is generally higher than that in normal tissues, but is still insufficient to trigger the rapid and efficient oxygen release²². Therefore, it is necessary to selectively increase the H₂O₂ level in the biofilm. Z@P-M nanomotor could release massive H₂O₂, and the H₂O₂ concentration at pH 5.5 and pH 7.4 was detected by a H₂O₂ detection kit. At pH 7.4, H₂O₂ release levels of Z and Z@P reached 43.38 ± 0.81 μM and 43.17 ± 1.59 μM during 12 h, respectively, indicating that ZnO₂ was decomposed slowly to produce H₂O₂ under neutral conditions (pH = 7.4). When the pH was 5.5, the H₂O₂ release levels of Z and Z@P within 12 h increased sharply to 192.58 ± 1.91 μM and 188.46 ± 2.43 μM, respectively, which were 343.94 ± 5.13 % and 336.82 ± 11.02 % higher than those under neutral condition (pH = 7.4). It was demonstrated that H₂O₂ in the suspension mainly originated from the acid decomposition of ZnO₂, and the modification of PDA barely affected the H₂O₂ release levels. Under pH 7.4 and 5.5 conditions, the H₂O₂ levels of Z@P-M reached to 3.19 ± 0.76 μM and 29.26 ± 3.18 μM within 12 h, respectively. This significant difference in H₂O₂ concentration indicated that the MnO₂ from Z@P-M decomposed the generated H₂O₂, leading to a decrease in H₂O₂ levels in Z@P-M (Fig. 3h). Moreover, the O₂ release level of Z, Z@P, and Z@P-M were examined by dissolved oxygen analyzer. O₂ release levels of Z (pH 7.4), Z (pH 5.5), Z@P (pH 7.4), and Z@P (pH 5.5) within 6 h were 3.74 ± 0.10 mg/L, 4.49 ± 0.09 mg/L, 3.56 ± 0.29 mg/L, and 4.32 ± 0.23 mg/L, respectively. Their release levels were all higher than those of the control group (dd H₂O, 2.28 ± 0.07 mg/L), attributed to the slow decomposition of H₂O₂ from Z and Z@P at room temperature. Compared with Z@P, Z@P-M showed increased levels of O₂ release under neutral and acidic conditions, reaching 5.37 ± 0.44 mg/L and 7.31 ± 0.14 mg/L, respectively. These results suggested that MnO₂ from Z@P-M catalyzed the generated H₂O₂, thereby continuously producing O₂. The microenvironment of bacterial biofilms is generally acidic (pH = 4.5 − 6.5) and hypoxic^39,40. Therefore, acid-responsive ZnO₂ could supply a large amount of H₂O₂ via acid decomposition in biofilm microenvironment. The produced H₂O₂ can be further catalyzed by MnO₂ to generate O₂, thereby alleviating hypoxia of biofilm and reducing inflammation⁴¹ (Fig. 3i). Similarly, the release rate of Zn²⁺ under acidic condition (pH 5.5) was 86.55 ± 7.14 % during 24 h, which was significantly higher than that under pH 7.4 (Supplementary Fig. 10). The results demonstrated that Z@P-M could be degraded under acidic condition, effectively releasing Zn²⁺ and O₂.

Z@P-M with enhanced diffusion promotes skin tissues and biofilm penetration

The asymmetric distribution of PDA-MnO₂ shell in Z@P-M nanomotors leads to asymmetric absorption of 808 nm NIR laser, thus generating a temperature gradient that produces self-thermophoresis for the actuation of the nanomotors⁴². The motion of Z@P-M in PBS under 808 nm NIR irradiation (0, 0.5, 1, and 1.5 W/cm²) was recorded by inverted optical microscopy, and the motion trajectories were analyzed subsequently using Image J. As shown in Fig. 4a and Supplementary movie 1, enhanced diffusion of Z@P-M nanomotors was clearly observed with lengthened trajectories with the gradual increase of 808 nm NIR powers (0 − 1.5 W/cm²). Under 808 nm NIR irradiation (0 − 1.5 W/cm²), the motion of Z@P-M was positively correlated with the 808 nm NIR power. Z@P-M showed classical Brownian motion without 808 nm NIR irradiation. Under 1.5 W/cm² 808 nm NIR irradiation, the velocity of Z@P-M was 4.16 ± 0.26 μm/s, which was 1.42 ± 0.07 folds higher than that of Brownian motion (2.94 ± 0.25 μm/s) (Fig. 4b). The diffusion of NIR light-driven nanomotors was further investigated by mean-square displacement (MSD) analysis. The MSD exhibited an increase proportional to laser intensity escalation (0 − 1.5 W/cm²), indicating enhanced propulsion capacity of Z@P-M nanomotors (Fig. 4c). The enhanced diffusion of Z@P-M nanomotors actuated by 808 nm NIR laser further provided the basis for skin tissues and biofilm penetration.

Fig. 4: Enhanced diffusion, penetration and antibiofilm effect of Z@P-M and Z@P-M MNs.

Tracking trajectories (a) and velocity (b) of Z@P-M nanomotors under 30 s irradiation of different 808 nm NIR intensities (0, 0.5, 1, 1.5 W/cm²). n = 20 independent experiments. c Average MSD calculated from the tracking trajectories under 30 s irradiation of different 808 nm NIR intensities (0, 0.5, 1, 1.5 W/cm²) (Δt = 1.5 s). n = 20 independent experiments. d Fluorescence images of porcine skin treated with RhB labelled Z@P-M-loaded MNs patches at varied depths (0 − 600 μm) after Z@P-M MNs insertion and 808 nm NIR irradiation, where 1 W/cm² 808 nm NIR irradiation was applied for 3 min or 5 min. n = 3 independent experiments (scale bar = 200 μm). Schematic diagram was created in BioRender by Hu, Z. (https://BioRender.com/wd2r0cf). e CLSM images of 5-CFDA - stained C. acnes biofilms after treating with RhB labelled Z@P-M and Z#P-M with or without 808 nm NIR irradiation, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. The red and green fluorescence represented RhB-labeled NPs and 5-CFDA-stained C. acnes biofilms, respectively. n = 3 independent experiments. f Fluorescence intensity of RhB labelled Z@P-M and Z#P-M between 0 − 20 μm of section. n = 3 independent experiments. Crystalline violet staining of C. acnes biofilms (g) and biomass of C. acnes (h) after treating with PBS, blank MNs, Z@P MNs, Z@P-M MNs, and NIR + Z@P-M MNs, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. n = 3 independent experiments. i SEM characterization of C. acnes biofilms structural modifications after different treatments (scale bar = 2 μm), where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. n = 3 independent experiments. j Representative photographs of C. acnes bacterial colonies after different treatments, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. n = 5 independent experiments. TUNEL fluorescence images (k) and fluorescence quantification (l) of C. acnes after different treatments, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. The green and blue fluorescence represented TUNEL and DAPI, respectively. n = 3 independent experiments (scale bar = 50 μm). Data in (b), (f), (h), and (l) were presented as Mean ± SD. P values in (h) and (l) were analyzed by one-way ANOVA with Tukey’s multiple comparisons test (**P < 0.01, ****P < 0.0001). Source data are provided as a Source Data file.

To verify the dissolution and penetration of Z@P-M MNs patches in the skin, a penetration experiment was performed by porcine skin with similar structure to human skin⁴³. Z@P-M MNs patches were pressed onto the porcine skin for 5 min, followed by 808 nm NIR irradiation (1 W/cm²) for 5 min. After the treatment, the patches were carefully removed. As shown in Supplementary Fig. 11, MNs on the patches were completely dissolved during the 10-min treatment, indicating that Z@P-M MNs exhibited favorable skin solubility. To further investigate the dynamic dissolution and penetration process of Z@P-M MNs in the skin, Z@P-M was labelled with Rhodamine B (RhB) and loaded in the MNs patches, and z-stack fluorescence images of the skin were captured by confocal laser scanning microscopy (CLSM). After inserting Z@P-M MNs patch into the skin for 5 min, strong fluorescence was observed within the 0 − 300 μm depth range. The patch was then exposed to 1 W/cm² 808 nm NIR irradiation for 3 or 5 min. At 3 min of 1 W/cm² 808 nm NIR irradiation, the fluorescence intensity in the 0 − 200 μm depth range decreased sharply, while the fluorescence intensity increased in the 400 − 500 μm depth range. At 5 min of 1 W/cm² 808 nm NIR irradiation, the depth range of fluorescence further expanded to 400 − 600 μm, confirming that Z@P-M exhibited enhanced permeability in the skin under 1 W/cm² 808 nm NIR irradiation. This was attributed to the effective motility caused by self-thermophoresis of Z@P-M, thereby enhancing deep penetration in skin tissues (Fig. 4d, and Supplementary Fig. 12).

In order to demonstrate the effect of the asymmetric structure on penetration, ZnO₂#PDA-MnO₂ (Z#P-M) NPs were also developed by encapsulating ZnO₂ with PDA-MnO₂ layer completely. The penetration efficiency of Z@P-M and Z#P-M NPs into methicillin-resistant staphylococcus aureus (MRSA) and C. acnes biofilms was investigated by scanning with CLSM, where Z@P-M and Z#P-M NPs were labelled with RhB (red). Without 808 nm NIR irradiation, Z@P-M and Z#P-M NPs stayed on the surface of the C. acnes biofilm and the penetration was significantly delayed. Whereas, the penetration of Z@P-M and Z#P-M NPs into the biofilm was notably observed under 5 min of 1 W/cm² 808 nm NIR irradiation (Fig. 4e, and Supplementary Fig. 13a). The red fluorescence intensity of each cross-section was analyzed by Image J (Fig. 4f, and Supplementary Fig. 13b). Under 1 W/cm² 808 nm NIR irradiation, Z@P-M had maximum fluorescence intensity at 10 μm and 8 μm depths in MRSA and C. acnes biofilms respectively, while the maximum fluorescence intensity of Z#P-M in both biofilms was stopped at 6 μm. Later, the fluorescence intensity of the biofilm sections (0 − 20 μm) was further analyzed. At the lowest bottom (20 μm) and under 1 W/cm² 808 nm NIR irradiation, the Z@P-M fluorescence intensity of MRSA and C. acnes biofilms were 13.08 ± 3.70 and 3.61 ± 0.09 folds higher than that of Z#P-M, which indicated that NIR light-actuated Z@P-M nanomotors could penetrate the biofilms efficiently.

Antibacterial, antibiofilm and antiinflammation evaluation of Z@P-M MNs

C. acnes infections are the main source of acne, and antibiotic therapy is now the primary treatment for acne⁴. However, the misuse of antibiotics leads to potential drug resistance⁴⁴, so we selected MRSA and C. acnes as models for acne. A systematic anti-biofilm evaluation of NIR light-actuated Z@P-M MNs was carried out. 1 mL of bacterial suspension (1 × 10⁸ CFU/mL) was seeded in 24-well plates for 48 h incubation to form mature biofilms. Figure 4g-h and Supplementary Fig. 14 showed the crystal violet (CV)-stained C. acnes and MRSA biofilms after incubation with Z@P-M MNs. CV staining showed that NIR + Z@P-M MNs group (5 min of 1 W/cm² 808 nm NIR irradiation) disrupted the biofilm most significantly, attributed to PTT effect and motility of NIR light-endowed Z@P-M nanomotor. Compared with the blank MNs group, the biomass of C. acnes and MRSA biofilms in the Z@P MNs group decreased by 56.96 ± 2.09 % and 60.70 ± 1.34 %, respectively, suggesting that the H₂O₂ produced by Z@P MNs inhibited biofilm formation and antimicrobial activity to a certain extent. As reported, H₂O₂ with concentration of 10^-5 − 10^-4M indeed inhibited bacterial and biofilms activity to some extent in vitro and in vivo, and H₂O₂ concentration as high as 166 mM − 1 M were necessary to achieve ideal antibacterial effects^45,46. Biomass of C. acnes and MRSA biofilms in Z@P-M MNs group was decreased by 20.43 ± 2.21 % and 41.88 ± 3.11 % compared with that in Z@P MNs group, due to the O₂ release in response to acidic biofilm microenvironment inhibiting anaerobic C. acnes⁴⁷. Moreover, the surface of Z@P-M was covered with MnO₂ nanosheets, which inhibited the formation of C. acnes and MRSA biofilms by inserting into bacteria to disrupt the cell membranes due to their extremely high aspect ratio edges⁴⁸. Under 5 min of 1 W/cm² 808 nm NIR irradiation, the biomass of C. acnes and MRSA biofilms was significantly reduced by 58.76 ± 0.69 % and 73.00 ± 0.72 % in the NIR + Z@P-M MNs group compared to that of the Z@P-M MNs group. Moreover, the effect of different exposure durations at 1 W/cm² 808 nm NIR irradiation on biomass in the NIR + Z@P-M MNs group was also investigated. Compared with the Z@P-M MNs group, after 3 min and 5 min of 1 W/cm² 808 nm NIR irradiation, the biomass of C. acnes biofilms decreased by 31.04 ± 4.64 % and 57.34 ± 3.41 %, respectively, while the biomass of MRSA biofilms decreased by 47.68 ± 0.58 % and 69.99 ± 0.36 % (Supplementary Fig. 15). It indicated that 5 min of 1 W/cm² 808 nm NIR irradiation had the best anti-biofilm effect. In addition, SEM images showed that the group of Z@P MNs, Z@P-M MNs, and NIR + Z@P-M MNs exhibited some degree of damage to the bacterial structure. In particular, the NIR + Z@P-M MNs group (5 min of 1 W/cm² 808 nm NIR irradiation) exhibited the most obvious bacterial structural distortion and collapse (Fig. 4i, Supplementary Fig. 16). Compared with the control group, TEM images of bacteria revealed that the membrane structure integrity of C. acnes and MRSA bacteria in the NIR + Z@P-M MNs group (5 min of 1 W/cm² 808 nm NIR irradiation) was severely disrupted, accompanied by obvious cytoplasmic leakage (Supplementary Fig. 17). Standard plate counting showed the similar results (Fig. 4j, and Supplementary Fig. 18). Compared with the blank MNs group, the survival rates of C. acnes and MRSA were reduced in the Z@P MNs group and Z@P-M MNs group, indicating that Z@P MNs and Z@P-M MNs had antibacterial effects. After 5 min of 1 W/cm² 808 nm NIR irradiation, the survival rates of C. acnes and MRSA in the NIR + Z@P-M MNs group were decreased to 0.09 ± 0.04 % and 0.59 ± 0.22 %, respectively. Compared with the Z@-M MNs group, 1 W/cm² 808 nm NIR irradiation for 3 min and 5 min resulted in survival rates of C. acnes decreasing by 65.31 ± 7.01 % and 94.26 ± 1.41 %, respectively, while the survival rates of MRSA decreased by 53.45 ± 9.11 % and 94.53 ± 1.55 %, respectively (Supplementary Fig. 19), demonstrating that 1 W/cm² 808 nm NIR irradiation for 5 min exhibited the optimal antibacterial effect. Terminal-deoxynucleotidyl transferase mediated nick end labeling (TUNEL) assay was also used to detect DNA damage in bacteria after different treatments. Compared with the blank MNs group, TUNEL fluorescence was enhanced in the Z@P MNs group, Z@P-M MNs group, and NIR + Z@P-M MNs group (5 min of 1 W/cm² 808 nm NIR irradiation). Among these, the NIR + Z@P-M MNs group (5 min of 1 W/cm² 808 nm NIR irradiation) exhibited the most intense TUNEL fluorescence, confirming that Z@P-M MNs can effectively damage bacterial DNA under 808 nm NIR light irradiation (Fig. 4k-l, Supplementary Fig. 20). It was suggested that the NIR light-driven Z@P-M nanomotor not only disrupted the bacterial structure via PTT effect, but also enhanced the diffusion and penetration in the biofilms due to the motility conferred by its Janus structure, which is promising for treating infections of drug-resistant bacterial biofilms in vivo.

We further investigated the therapeutic effects of Z@P-M MNs in vitro. The cytotoxicity of Z@P-M with different ZnO₂ concentrations was first evaluated by CCK-8 assay. As shown in Fig. 5a, Z@P-M with ZnO₂ concentration less than 100 μg/mL had no significant toxicity. To further explore the biocompatibility of different materials on NIH3T3 cells, HaCaT cells, and Raw 264.7 cells, calcein acetoxymethyl ester/propidium iodide (Calcein-AM/PI) was used to stain the live and dead cells, which showed green and red fluorescence, respectively (Fig. 5b, and Supplementary Fig. 21). Comparing with control group, no significant cytotoxicity of blank MNs, Z@P MNs, Z@P-M MNs, and NIR + Z@P-M MNs (5 min of 1 W/cm² 808 nm NIR irradiation) was clearly observed in NIH3T3 cells, HaCaT cells, and Raw 264.7 cells, indicating that the synthesized MNs patches were indeed biocompatible. Cellular internalization was evaluated in NIH3T3 cells co-cultured with RhB-labelled Z@P-M MNs and Z#P-M MNs, followed by 5 min 808 nm NIR irradiation (1 W/cm²). Cytoskeletal architecture was visualized using actin-tracker green fluorophore. Supplementary Fig. 22 showed that weak fluorescence was observed in the Z@P-M MNs group and Z#P-M MNs group without NIR light treatment. Under 5 min of 1 W/cm² 808 nm NIR irradiation, the fluorescence intensity of the Z@P-M MNs group was 39.73 ± 13.06 % higher than that of the Z#P-M MNs group, indicating that NIR light-triggered enhanced diffusion promoted the cellular uptake of the released Z@P-M nanomotors, which supported a better therapeutic effect of the nanomotor in vitro.

Fig. 5: Evaluation of Z@P-M MNs in vitro.

a Cytotoxicity after treating with Z@P-M (ZnO₂ concentration: 0 − 200 μg/mL). n = 4 independent experiments. b Live/dead cells staining of NIH3T3 cells after different treatments, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. The green fluorescence indicated live cells, and red fluorescence was dead cells (scale bar = 100 μm). n = 3 independent experiments. c Images of NIH3T3 cells scratch after different treatments, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min (scale bar = 200 μm). n = 3 independent experiments. d Migration rates of NIH3T3 cells after different treatments, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. n = 3 independent experiments. e O₂ sensing probe [(Ru(dpp)₃)]Cl₂ staining of NIH3T3 cells after different treatments, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min (scale bar = 100 μm). n = 3 independent experiments. f HIF-1α fluorescence staining of NIH3T3 cells after different treatments, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. The red and green fluorescence represented HIF-1α and actin, respectively (scale bar = 100 μm). n = 3 independent experiments. g ROS fluorescence staining of NIH3T3 cells after different treatments, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. The green and blue fluorescence represented ROS and cell nucleus, respectively (scale bar = 100 μm). n = 3 independent experiments. h IL-1β, IL-6, and TNF-α gene expression in NIH3T3 cells after treating with (1) normoxia, (2) hypoxia + LPS (HL), (3) HL + blank MNs, (4) HL + Z@P MNs, (5) HL + Z@P-M MNs, (6) HL + NIR + Z@P-M MNs, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. n = 3 independent experiments. Data in (a), (d), and (h) were presented as Mean ± SD. P values in (d) and (h) were analyzed by one-way ANOVA with Tukey’s multiple comparisons test (n.s. represented no significance, *P < 0.05, **P < 0.01, ****P < 0.0001). Source data are provided as a Source Data file.

To confirm the repairing effect of MNs patches on the infected tissues, scratching assay of NIH3T3 cells was performed in a hypoxia environment. In Fig. 5c,d, the migration rate of the Z@P MNs group increased to 52.05 ± 2.34 % and 70.92 ± 1.83 % at 18 and 36 h, respectively, which was higher than that of the blank MNs group. It was attributed to the fact that Zn²⁺ derived from ZnO₂ could promote fibroblast proliferation and migration, thereby promoting wound healing⁴. Compared with the Z@P MNs group, the migration rates of Z@P-M MNs group at 36 h increased by 13.03 ± 0.51 %, demonstrating that oxygen released from Z@P-M MNs under hypoxic conditions promoted the skin cell migration⁴⁹. When compared with the Z@P-M MNs group, the migration rate at 18 h and 36 h further increased by 36.07 ± 3.23 % and 11.01 ± 1.47 % after 5 min of 1 W/cm² 808 nm NIR irradiation, indicating that 808 nm NIR irradiation promoted the migration of fibroblasts. NIR light not only promoted the release of active Z@P-M nanomotors from MNs, but also improved the cell metabolism and regulated inflammatory responses, thereby accelerating the fibroblasts migration and the repair of the infected tissues⁵⁰. The O₂ release level of Z@P-M MNs in vitro was further assessed by oxygen sensing probe [(Ru(dpp)₃)]Cl₂. As shown in Fig. 5e and Supplementary Fig. 23a, the probe fluorescence intensity of the Z@P-M MNs group was decreased by 60.64 ± 2.67 % compared with that of the Z@P MNs group. The fluorescence intensity of Z@P-M MNs was further decreased by 46.83 ± 3.47 % after 5 min of 1 W/cm² 808 nm NIR irradiation. Hypoxia-Inducible Factor 1α (HIF-1α) was generally activated in hypoxic skin tissues, and its expression was decreased by continuous delivery of oxygen⁵¹. The enhanced diffusion of NIR light-actuated Z@P-M nanomotors promoted the accumulation of nanomotor into the cells⁵², which enhanced cellular oxygen delivery and further reduced the expression of HIF-1α. Therefore, the oxygenation capacity of Z@P-M MNs was further confirmed by HIF-1α immunofluorescence in vitro. Lipopolysaccharide (LPS) and hypoxia successfully induced high HIF-1α expression in the cells, whereas Z@P-M MNs treatment with or without 808 nm NIR irradiation both showed decrease in HIF-1α expression, in which the most significant decrease in HIF-1α expression was observed for NIR + Z@P-M MNs group, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min (Fig. 5f, and Supplementary Fig. 23b). It was illustrated that the catalytic effect of the coated MnO₂ synergized with NIR light-actuated nanomotor enhanced diffusion to increase the continuous oxygen delivery. In general, improvement of hypoxia and scavenging of reactive oxygen species (ROS) can alleviate the inflammatory response of biofilm infection⁵³. As expected, NIH3T3 cells showed an increase in ROS fluorescence stimulated by LPS and hypoxia, indicating that an increased level of ROS production was highly correlated with inflammation severity. Compared with Z@P MNs, the cells exhibited a 70.66 ± 0.65 % decrease in ROS fluorescence intensity after Z@P-M MNs treatment, which was attributed to the powerful ROS scavenging ability of MnO₂⁵⁴. Z@P-M MNs further decreased the ROS fluorescence intensity by 30.81 ± 1.62 % under 5 min of 1 W/cm² 808 nm NIR irradiation, indicating that active Z@P-M nanomotors could improve the ability to scavenge ROS (Fig. 5g, and Supplementary Fig. 23c). Subsequently, the anti-inflammatory effects of Z@P-M MNs were also evaluated by the gene expression levels of interleukin-1β (IL-1β), interleukin-6 (IL-6), and TNF-α (Fig. 5h). Z@P-M MNs significantly reduced the levels of IL-1β, IL-6 and TNF-α inflammatory factors, which were further decreased under 5 min of 1 W/cm² 808 nm NIR irradiation. It indicated that the released Z@P-M nanomotors had superior motion and diffusion effects and exhibited efficient anti-inflammatory effects in the treatment of C. acnes infections.

Anti-acne ability and alleviation of ILC inhibition by Z@P-M MNs

Encouraged by the efficient antibacterial, anti-biofilm and anti-inflammation of Z@P-M MNs, acne models were established through intradermal C. acnes (2 × 10⁹ CFU/mL) inoculation in dorsal tissue of female balb/c mice, subsequently randomized into 6 experimental cohorts (n = 5/group): acne, erythromycin, blank MNs, Z@P MNs, Z@P-M MNs, and NIR + Z@P-M MNs, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. The female mice with different treatments were monitored with the changes of acne on day 0, 3, 5, and 7 (Fig. 6a). As shown in Fig. 6b, two days after the injection of C. acnes (day 0), acne-like protrusion was formed at the injection site, and the proliferation of C. acnes resulted in the localized infection, ulceration, and skin lesions. Erythromycin group effectively reduced the size of acne, but mild desquamation and skin dryness were still observed, which was attributed to the skin irritation caused by erythromycin ointment⁵⁵. RhB labelled nanomotors were detected in skin tissues after pressing Z@P-M MNs patches on the acne surface and 5 min of 1 W/cm² 808 nm NIR irradiation, indicating that the developed MNs could effectively penetrate the skin to release nanomotors, which was remained in the skin for at least 72 h. It was noteworthy that Z@P-M selectively accumulated in the hair follicles, which was the main colonization site of C. acnes, leading to excessive keratinization and microbial proliferation⁵⁶. Therefore, Z@P-M can effectively target acne lesions (Supplementary Fig. 24).

Fig. 6: Anti-acne ability and alleviation of ILCs inhibition by Z@P-M MNs in vivo.

a Schematic diagram of the in vivo experiment was created in BioRender by Hu, Z. (2025) https://BioRender.com/wd2r0cf. b Visual images of C. acnes infected acnes during different treatments, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min (scale bar = 5 mm). n = 5 mice. c Inflammatory cells of the skins after different treatments, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. n = 5 mice. d Staining of skin tissue sections including H&E (scale bar = 250 μm), IL-6 (scale bar = 100 μm), TNF-α (scale bar = 100 μm), CD31 (scale bar = 100 μm), and HIF-1α (scale bar = 100 μm) after treating with (1) C. acnes, (2) erythromycin, (3) blank MNs, (4) Z@P MNs, (5) Z@P-M MNs, (6) NIR + Z@P-M MNs, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. The green and blue fluorescence represented HIF-1α and cell nucleus, respectively. n = 5 mice. e Immunofluorescence of RORγt and IL-22 in skin tissues after different treatments, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. The yellow, red, and blue fluorescence represented RORγt, IL-22, and cell nucleus, respectively (scale bar = 20 μm). n = 3 mice. Proportion of RORγt⁺ (f) and IL-22⁺ (g) cells after different treatments, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. n = 3 mice. Data in (c), (f), and (g) were presented as Mean ± SD and P values were analyzed by one-way ANOVA with Tukey’s multiple comparisons test (n.s. represented no significance, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). Source data are provided as a Source Data file.

After pressing the Z@P-M MNs patch on mouse skin for 5 min, followed by 5 min of 1 W/cm² 808 nm NIR irradiation, IR photothermal images showed that the temperature rose to 45.22 °C during 5 min (Supplementary Fig. 25, 26). After treatment, the skin healing at the puncture site was assessed. As shown in Supplementary Fig. 27, after removing the patch, a noticeable pinhole appeared on the skin surface, which recovered within 30 min, with no obvious redness, swelling, and ulceration. H&E staining of the skin showed similar results, observing a distinct conical depression in the skin tissues, enhancing the skin delivery of Z@P-M. Moreover, hemolysis assay of Z@P-M (ZnO₂ concentration = 100 μg/mL) was also conducted. Under 5 min of 1 W/cm² 808 nm NIR irradiation, the hemolysis rate of Z@P-M was 4.13 ± 0.49 % (Supplementary Fig. 28). H&E staining of the major organs including heart, liver, spleen, lungs, and kidneys (Supplementary Fig. 29) showed no evidence of organ damage. The above results demonstrated the favorable biosafety of Z@P-M MNs.

To further evaluate the effect of MNs on acnes, the female mice were euthanized on day 7, collection of bacterial and histological and immunohistochemical analysis of the skin at the acnes site were performed. On day 7, bacterial burden analysis showed that erythromycin group, Z@P MNs group, Z@P-M MNs group, and NIR + Z@P-M MNs group (5 min of 1 W/cm² 808 nm NIR irradiation) exhibited a significant reduction in C. acnes colony counts compared with the acne group. Among these, NIR + Z@P-M MNs group (5 min of 1 W/cm² 808 nm NIR irradiation) and erythromycin group exhibited comparable antibacterial efficacy, which was attributed to the synergistic effects of penetration and PTT effects provided by the Z@P-M (Supplementary Fig. 30). The level of inflammation in the skin was assessed by counting the number of inflammatory cells by H&E staining. In Fig. 6c and Supplementary Fig. 31, there was a large infiltration of inflammatory cells and follicle destruction in the skin tissues of the acne group and blank MNs group. For erythromycin, Z@P MNs, Z@P-M MNs, and NIR + Z@P-M MNs groups (5 min of 1 W/cm² 808 nm NIR irradiation), the number of inflammatory cells was relatively reduced, but some extent of inflammatory infiltration was still present. The skin structures were more intact in the NIR + Z@P-M MNs group (5 min of 1 W/cm² 808 nm NIR irradiation) compared with other groups. Moreover, the NIR + Z@P-M MNs group (5 min of 1 W/cm² 808 nm NIR irradiation) showed normalized hair follicle structure and had the highest number of hair follicles. Histological analysis revealed significant inflammatory cells reduction in NIR + Z@P-M MNs group (5 min of 1 W/cm² 808 nm NIR irradiation), correlating with favorable therapeutic outcomes (Fig. 6d). Then we further analyzed IL-6 and TNF-α by immunohistochemistry. Anaerobic C. acnes overproliferated in sebaceous gland cells of mouse skin to form biofilms, which further caused hypoxia, promoted bacterial proliferation, led to inflammation, and upregulated the expression of IL-6 and TNF-α, both of the proteins were expressed at acne^4,57,58. Optical density analysis showed that the expression levels of IL-6 and TNF-α were reduced by 37.03 ± 1.30 % and 54.66 ± 4.09 % in the Z@P-M MNs group compared to the Z@P MNs group. The expression of IL-6 and TNF-α in Z@P-M MNs was the lowest after NIR light treatment (Fig. 6d, and Supplementary Fig. 33a-b). As reported, the anti-inflammatory factor interleukin 10 (IL-10) could inhibit the production of pro-inflammatory cytokines, thereby reducing the inflammation⁵⁹. The expression level of IL-10 in the Z@P-M MNs group was increased by 31.92 ± 5.08 % compared with that in the Z@P MNs group, and the level of IL-10 was further increased after 5 min of 1 W/cm² 808 nm NIR irradiation (Supplementary Fig. 32, and Supplementary Fig. 33c). These findings suggested that NIR + Z@P-M MNs treatment effectively activated the expression of anti-inflammatory genes, thereby downregulating pro-inflammatory factors and alleviating tissue inflammation.

Although it has been demonstrated that Z@P-M MNs could efficiently eliminate C. acnes and inflammation in vivo, their effects on the skin microbiome were not negligible. C. acnes and Staphylococcus epidermidis are commensal bacteria of the skin, exhibiting a relationship of mutual inhibition^7,60. Inflammation and microbial imbalance caused by the over-proliferation of C. acnes can lead to acne⁶¹. Regarding the above issues, the sustained oxygen release of Z@P-M MNs alleviated the anaerobic environment in acne lesions, thereby inhibiting C. acnes proliferation. Most skin microbiota, including Staphylococcus epidermidis, were aerobic bacteria, oxygen generation contributed to restoring abundance of them, thereby further inhibiting the proliferation of C. acnes and re-establishing the microbial homeostasis of skin⁶². Z@P-M exhibited follicle-targeting properties, enabling more precise and safer elimination of C. acnes to promote the restoration of skin microbiota homeostasis. In future clinical applications, monitoring the skin microbiome through techniques such as 16S rRNA sequencing, metagenomic sequencing, and spatial omics will advance our understanding of the relationship between microneedling and skin microbiome⁶³. The hypoxia level of MNs-treated tissues was also explored. HIF-1α, a classical marker of tissue hypoxia⁶⁴, was expressed at a lowest level in the NIR + Z@P-M MNs group (5 min of 1 W/cm² 808 nm NIR irradiation) compared with the other groups (Fig. 6d, and Supplementary Fig. 33d). It was attributed to the large amount of H₂O₂ released from ZnO₂ being converted into adequate oxygen supply by MnO₂, which attenuated the acne inflammation. In addition, the favorable propulsion and PTT effect of NIR light-actuated Z@P-M nanomotors could efficiently clear C. acnes biofilm, thus further reducing the inflammation level of acnes. Acnes caused by C. acnes infection often ruptured leading to wound infection, resulting in severe scarring⁶⁵. Angiogenesis played an essential role in wound healing, and CD31 was an important marker for neovascularization⁶⁶. In Fig. 6d and Supplementary Fig. 33e, The expression level of CD31 in the Z@P MNs group was increased by 35.92 ± 4.99 % compared with that in the blank MNs group, owing to the fact that Zn²⁺ produced by the decomposition of ZnO₂ could promote tissue healing⁵⁵. expression level of CD31 in the Z@P-M MNs group was increased by 32.08 ± 3.59 % compared with that in the Z@P MNs group, and the expression level of CD31 in the Z@P-M MNs group was further enhanced by 46.01 ± 3.87 % after 5 min of 1 W/cm² 808 nm NIR irradiation, suggesting that 808 nm NIR irradiation accelerated the release of O₂ from Z@P-M MNs to alleviate the hypoxic microenvironment of acnes, thus promoting the healing of acne ulcer wounds.

ILCs are considered as one of the potential immune targets for microbial host interactions in acne. Previous studies have shown that bEVs of C. acnes can significantly reduce the number of ILCs, especially RORγt⁺ ILCs⁸. Therefore, exploring the effect of Z@P-M MNs on the restoration of ILCs homeostasis in C. acnes-infected acnes may be an important target for acne treatment. We carried out the immunofluorescence staining of acne skin for IL-22 and RORγt (Fig. 6e). The cells releasing IL-22 and expressing RORγt were identified as ILC3⁶⁷. Immunofluorescence results showed that IL-22 and ILC3 were almost not expressed in the acne and blank MNs groups, whereas the cells co-localized with IL-22 and RORγt were observed in the NIR + Z@P-M MNs group. It indicated that ILC3 was significantly inhibited in the acne and blank MNs groups, while the NIR + Z@P-M MNs group (5 min of 1 W/cm² 808 nm NIR irradiation) could restore the homeostasis of ILC3 (Fig. 6f). It was attributed that enhanced diffusion of Z@P-M nanomotor synergized with PTT effect to further eliminate the over-accumulation of C. acnes and ultimately restored the ILC3 homeostasis. Moreover, IL-22 released from ILC3 was also significantly increased in the NIR + Z@P-M MNs group (Fig. 6g). Since IL-22 is a key cytokine for anti-infection and tissue reparation⁶⁸, it made sense that acne lesions were ideally healed in the NIR + Z@P-M MNs group.

Discussion

In our study, we successfully fabricate a NIR light-driven nanomotors loaded MN system for enhanced C. acnes biofilm penetration, which significantly facilitates acne treatment. Z@P-M nanomotors are prepared by asymmetric modification of in situ grown MnO₂/PDA layer on ZnO₂ through interfacial energy regulation. Then Z@P-M MNs are successfully fabricated by mixing Z@P-M nanomotors with sodium hyaluronate. The resulting MNs patch release active Z@P-M nanomotors in acne tissues through transdermal delivery. In the acidic biofilm microenvironment, Z@P-M nanomotors slowly release H₂O₂ to generate sufficient O₂ supply catalyzed by MnO₂, which alleviates the hypoxia of the biofilm microenvironment, eliminated inflammation, and ultimately promotes the healing of infected tissues. The asymmetric structure of the Z@P-M nanomotor generates self-thermophoresis actuated by NIR light. It further increases the diffusion of the nanomotor to achieve deep biofilm penetration synergizing with PTT effect to eradicate the C. acnes biofilm, thereby restoring the ILCs homeostasis in the infected tissues. In summary, the developed MN system provides a potential platform for acne treatment and regulatory application of MN system by in situ self-supplying of O₂ synergizing with PTT. However, most studies on MNs still remain in preclinical exploration, lacking sufficient clinical data and unpredictable obstacles to clinical translation⁶⁹. Although our Z@P-M MNs patch contains biocompatible sodium hyaluronate approved by US Food and Drug Administration⁷⁰, it is still necessary to evaluate the potential allergenicity and irritation of Z@P-M in clinic. The degradation and clearance of Z@P-M in vivo should also be monitored and evaluated through long-term toxicity evaluation.

Method

Cell lines and animals

Mouse embryonic fibroblast cells (NIH3T3 cells, CL-0171) and mouse monocytic macrophage leukemia cells (RAW 264.7 cells, CL-0190) were obtained from Procell. Human keratinocytes cells (HaCaT cells, AW-CH0096) were purchased from AnWei-Sci. NIH3T3 cells, RAW 264.7 cells, and HaCaT cells were cultured in high glucose Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10 % FBS and 1 % penicillin-streptomycin. The animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Southern Medical University (License No. SMUL202311048). Female balb/c mice (6 weeks old, 25 - 30 g) were purchased from the Laboratory Animal Center of Southern Medical University. These female balb/c mice were allowed free access to food and water and were fed under conditions of ambient temperature (21 − 26 °C), humidity (50 − 60 %), and a 12 h dark/light cycle.

Material

Polyvinylpyrrolidone (PVP Mw ≈ 58,000), isopropanol (IPA, ≥ 99 %), ammonia solution (28 %), dopamine hydrochloride (DA, ≥ 99 %), crystal violet (CV, ≥ 98 %), and Rhodamine B (RhB, ≥ 98 %) were purchased from Innochem (China). Polyacrylic acid (PAA) was obtained from Sigma Aldrich. H₂O₂ (30 %) was bought from Guangzhou Chemical Reagent Factory (China). Fluid thioglycollate medium was bought from Huankai Microbial (China). [Ru(dpp)₃]Cl₂ was obtained from Maokang Biotechnology (China). Zinc chloride (ZnCl₂, ≥ 99.99 %) and sodium hyaluronate was bought from Macklin (China). Cell counting kit-8 (CCK-8), Calcein-AM/PI, H₂O₂ detection assay kit, ROS detection assay kit, 4’,6-diamidino-2-phenylindole (DAPI), and Hoechst 33342 were obtained from Biyotime (China). 5-carboxyfluorescein diacetate (5-CFDA, ≥ 98 %) was obtained from YuanyeBio-Technology (China). One-step TUNEL In Situ Apoptosis Kit was obtained from Elabscience (China). Hematoxylin-Eosin (H&E) Stain Kit were purchased from Solarbio Life Sciences (China). Primers of IL-1β, IL-6 and TNF-α were obtained from Sangon Biotech (China). Total RNA isolation kit was bought from Foregene (China). Rever Tra Ace qPCR RT kit and SYBR Green Master were obtained from Yeasen (China). HIF-1α monoclonal antibody was obtained from Cell Signaling Technology (America). IL-6 polyclonal antibody, TNF-α polyclonal antibody, CD31 polyclonal antibody and IL-10 recombinant antibody were bought from Proteintech (China). APC-RORγt monoclonal antibody was obtained from Invitrogen (America). PE-anti-mouse IL-22 antibody was obtained from Biolegend (America). Chromogen was bought from ZSGB-BIO (China). AnaeroPack and anaerobic culture box (2.5 L) were obtained from Mitsubishi (Japan).

Instruments

The structure of NPs and bacterial samples was observed using a JEM 1400 TEM under 120 kV acceleration voltage (JEOL, Japan). Elemental analyses of Z@P-M were measured by Oxford X-MaxN 80 T IE250, UK. Elemental composition of Z@P-M MNs was conducted via EDS (Phenom ProX, 15 kV, Netherlands). The structure of Z@P-M MNs and bacterial samples were observed by scanning electron microscope (SEM) (Phenom, Netherlands). Mechanical properties of MNs were measured by electronic universal testing machine (SANS, China). The size and zeta potential were characterized by Zetasizer Nano ZSE (Malvern, UK). XPS spectra was performed by X-ray photoelectron spectroscopy (Thermo SCIENTIFIC ESCALAB 250Xi, America) and analyzed by Avantage software (version 5.9931). FTIR spectra were measured by FTIR spectrometer (Nicolet 380, Thermo). The crystalline profile was measured by X-ray diffractometer (Bruker D2, Germany). Raman spectra were measured by laser confocal microscope raman spectrometer (RENISHAW inVia, UK). The release rate of Zn²⁺ was quantified by Inductively coupled plasma mass spectrometry (ICP-MS) (Agilent, America). Samples of Zn²⁺ detection were digested by microwave digestion system (Preekem, China). Thermal images were captured by infrared thermography (Fluke, America). Heating curves were obtained by infrared thermography (YOSEEN INFRARED, China). UV-Vis-NIR spectra were acquired using UV-2600 spectrophotometer (SHIMADZU, Japan). Absorbance of UV-Vis was measured by multifunctional microplate inspection system (Tecan, Switzerland). The release level of oxygen was quantified by dissolved oxygen analyzer (LEICI, China). Fluorescence images of ROS, HIF-1α, and [Ru(dpp)₃]Cl₂ in vitro and the motion of Z@P-M nanomotors were captured by an inverted microscope (Nikon, Japan). Three-dimensional porcine skin was reconstructed through z-stack imaging via CLSM (Nikon, Japan). Three-dimensional biofilms was reconstructed through z-stack imaging via CLSM (Zeiss, Germany). H&E and immunohistochemical slides were scanned by Slide Scan System SQS-1000 (Teksqray, China).

Synthesis of Z

ZnCl₂ (135 mg) and PVP (690 mg) were completely dissolved in 30 mL anhydrous ethanol, then 2 mL ammonia (10 M) was added. When the solution turned clear, 2 mL of 30 % H₂O₂ was slowly dropped under stirring, and the reaction lasted for 1 day. The collected ZnO₂ was wash 3 times by ethanol and stored in anhydrous ethanol.

Synthesis of Z@P-M nanomotors

Z@P-M were synthesized based on the modified method^21,71. Briefly, 10 mg ZnO₂, 1 mL ammonia (2 M), and 7.5 mg PAA were dispersed in 15 mL ultrapure water. Subsequently, 96 mL IPA was slowly dropped uder stirring. After 1 h, 2 mL ammonia (10 M) and 1 mg DA were added in turn and the reaction was conducted for 4 h at 50 °C. Z@P was then collected and washed 3 times with ethanol. 1 mg Z@P and 500 μL KMnO₄ (2 mg/mL) were dispersed in ultrapure water (10 mL), stirring for 20 min. After ultrapure water and ethanol washing, Z@P-M was collected and stored in ethyl alcohol. As control group, Z#P-M was ZnO₂ fully coated by PDA and MnO₂. The synthesis of Z#P-M was carried out in ultrapure water without IPA, and the remaining steps were consistent with above.

Preparation of MNs patches

For Z@P-M MNs patch, 150 mg sodium hyaluronate was dissolved in 1 mL ultrapure water containing Z@P-M (ZnO₂ concentration = 100 μg/mL), then the solution was dropped in MN mold. After 15 min vacuum drying, the mold was dried at 40 °C for 5 h. Blank MNs patch shared the same fabrication process without adding Z@P-M.

Degradation of Z@P-M MNs

Z@P-M MNs was incubated in 70 % humidity box at 25 °C for 30 min. Bright-field morphologic images were captured at 0, 10, 20, 30 min, respectively.

Detection of Zn²⁺ level

ICP-MS was used to detect Zn²⁺ level. Z@P-M NPs (ZnO₂ concentration = 100 μg/mL) were incubated in 5 mL buffer at pH 5.5 and 7.4 during 24 h. The centrifuged supernatant samples (100 μL) at different time points (2 h, 4 h, 8 h, 12 h, and 24 h) were collected and treated by 10 mL 10 % nitric acid (HNO₃). The samples were digested by microwave digestion system, with the temperature increasing to 180 °C during 30 min, and maintained at that temperature for 60 min. Finally, the samples were diluted to a solution containing 3 % HNO₃, filtered through a 0.22 μm filter, and Zn²⁺ concentration was measured by ICP-MS.

Release of H₂O₂

1 mL of Z, Z@P, and Z@P-M NPs (ZnO₂ concentration = 1 mg/mL) were dispersed in 9 mL buffer (pH 5.5 or 7.4) at 37 °C. H₂O₂ release levels at different time points was measured using a hydrogen peroxide assay kit. H₂O₂ levels were detected by mixing the samples with dimethylphenol orange, incubating for 20 min at 37 °C, finally measuring absorbance at 560 nm. Standard curve was ${A}_{560{nm}}=0.0014{C}_{H2O2}+0.0733$ (Supplementary Fig. 34).

O₂ release

1 mL of Z, Z@P, and Z@P-M NPs (ZnO₂ concentration = 1 mg/mL) were dispersed in 9 mL buffer (pH 5.5 or 7.4) at 37 °C, with deoxygenation in a nitrogen atmosphere. Oxygen generation was monitored by dissolved oxygen analyzer during 6 h and the data were recorded per 10 s.

Photothermal properties

Temperature curves of Z@P-M NPs (ZnO₂ concentration = 0 − 200 μg/mL) were evaluated under 808 nm irradiation (1 W/cm², 10 min). Z@P-M samples (ZnO₂ concentration = 100 μg/mL) was exposed to 808 nm NIR light (0.5, 1, and 1.5 W/cm², 10 min) and the temperature curves were recorded. Z, Z@P, and Z@P-M solution with the same ZnO₂ concentration (100 μg/mL) was collected in a 1.5 mL EP tube with PBS as control. All samples were exposed to 808 nm NIR radiation at 1 W/cm² for 10 min and the IR photothermal images and temperature curves were recorded. 8 heating/cooling cycles of Z@P-M (ZnO₂ concentration = 100 μg/mL) and Z@P-M MNs were performed with 1 W/cm² 808 nm NIR irradiation to investigate the photothermal stability. Z@P-M MNs and blank MNs were treated with 808 nm NIR irradiation (1 W/cm², 5 min), the temperature curves and IR photothermal images were recorded.

Photothermal conversion efficiency

Z@P-M NPs (ZnO₂ concentration = 100 μg/mL) were treated with 1 W/cm² 808 nm NIR irradiation for 600 s. And the temperature curve and conversion efficiency (η) were analyzed. The η was derived from Eq. (1):

$$\eta=\frac{{hS}\left({T}_{\max }-{T}_{{surr}}\right)-{Q}_{{dis}}}{I(1-{10}^{{-{\mbox{A}}}_{808}})}$$

(1)

h was heat-transfer-coefficient, S represented the surface area of the container, T_max indicated the equilibrium temperature, T_surr signified the ambient temperature of the surroundings, Q_dis represented the heat associated with the light absorbance by the solvent, I was the 808 nm NIR light power (1 W/cm²), and A₈₀₈ represented the absorbance of Z@P-M NPs at 808 nm (Supplementary Fig. 35). hS was derived from Eq. (2):

$${\tau }_{s}=\frac{{m}_{D}{C}_{D}}{{hS}}$$

(2)

m_D and C_D were the weight (0.5 g) and heat capacity (4.2 J/g) of pure water, was calculated from Eq. (3):

$$t={\tau }_{s}{{\mathrm{ln}}}\left(\frac{{T}_{{RT}}-{T}_{{surr}}}{{T}_{{Max}}-{T}_{{surr}}}\right)$$

(3)

T_RT was calculated from time-temperature curves during 900 s passive cooling phase.

Evaluation of skin penetration of Z@P-M MNs

Fresh porcine skin was purchased from the supermarket, placed in a clean Petri dish with the stratum corneum facing upward, and kept moist. Z@P-M MNs (loaded with RhB labelled Z@P-M) were inserted into the porcine skin for 5 min, followed by 1 W/cm² 808 nm NIR irradiation for 5 min. Finally, the treated porcine skin (insert 5 min, 1 W/cm² 808 nm NIR light irradiated for 3 min or 5 min) was z-stack scanned by CLSM.

Antibiofilm assay

The used bacterial strains were MRSA (ATCC 43300) and C. acnes (ATCC 6919). MRSA and C. acnes were proliferated in Luria-Bertani (LB) medium and thioglycollate medium to mid-log phase, respectively. 1 mL of bacterial suspension (1 × 10⁸ CFU/mL) was seeded in 24-well plates. C. acnes were incubated under hypoxic conditions and MRSA was cultivated under aerobic conditions. Biofilms were formed through 48 h incubation with medium replenishment at 24 h interval, followed by gentle PBS washing to remove planktonic bacteria. The culture medium was removed, and the formed biofilms were gently washed twice with PBS. Different treatments were performed on the biofilms, including (1) control, (2) blank MNs, (3) Z@P MNs, (4) Z@P-M MNs, (5) NIR + Z@P-M MNs, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. The control group was treated with PBS. After 5 min co-culture, NIR + Z@P-M MNs group was treated with 1 W/cm² 808 nm NIR irradiation for 5 min. To evaluate the antibiofilm effects of different duration of 808 nm NIR irradiation, the samples were divided into 4 groups: (1) control, (2) Z@P-M MNs, (3) 3 min NIR + Z@P-M MNs, and (4) 5 min NIR + Z@P-M MNs, where 1 W/cm² 808 nm NIR irradiation was applied. The following procedures were consistent with those described above. After 6 h co-incubation, CV staining was performed on biofilms. The resulting biofilms were washed twice with PBS, then stained with 0.1 % CV for 10 min. After gently washing, the biofilms were photographed and dissolved with 500 μL of anhydrous ethanol. The biomass of the biofilms was then quantified by recording the absorbance at 595 nm.

Biofilm morphology

Biofilms were fixed with 4 % paraformaldehyde at 4 °C for 2 h. Washing with PBS three times, biofilms were dehydrated with gradient ethanol (20 %, 40 %, 60 %, 80 %, and 100 %) and tert-butyl alcohol for 15 min, respectively. Finally, the dehydrated samples were freeze-dried and sprayed with a gold layer for SEM observation.

Penetration of Z@P-M in biofilms

The culture medium of the mature biofilms was removed, and the formed biofilms were gently washed twice with PBS. RhB-labelled Z#P-M and Z@P-M NPs (ZnO₂ concentration = 100 μg/mL) were gently added to the biofilms and incubated for 20 min. Group NIR + Z#P-M and group NIR + Z@P-M were irradiated with 1 W/cm² 808 nm NIR light for 5 min, followed by gently removing the supernatant. Subsequently, the biofilms were stained with 5-CFDA and carefully washed twice with PBS. Finally, Z-stack scanning was performed by CLSM (green represents biofilm, and red represents NPs).

Standard plate counting assay

Antimicrobial rates of all samples were analyzed by standard plate counting assays. 1 mL of C. acnes suspension (2 × 10⁵ CFU/mL) was added to sterile 5 mL EP tube and different treatments were performed, including (1) control, (2) blank MNs, (3) Z@P MNs, (4) Z@P-M MNs, (5) NIR + Z@P-M MNs, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. The control group was treated with PBS. Samples were incubated for 6 h. After the treatment, 200 μL of 100-fold diluted bacterial solution was collected, spread on thioglycollate agar plates and incubated anaerobically at 37 °C for 2 days. Similarly, after MRSA was subjected to the above treatment, the diluted bacterial solution was spread on Luria-Bertani agar plates and incubated aerobically at 37 °C for 1 day. Microbial viability was calculated as: Survival rate (%) = (CFU_{experimental groups} / CFU_{control group}) × 100. To evaluate the antibiofilm effects of different duration of 808 nm NIR irradiation, the samples were divided into 4 groups: (1) control, (2) Z@P-M MNs, (3) 3 min NIR + Z@P-M MNs, and (4) 5 min NIR + Z@P-M MNs, where 1 W/cm² 808 nm NIR irradiation was applied. Other procedures were consistent with those described above.

TUNEL staining

The treated biofilms were fixed with 4 % paraformaldehyde for 10 min and washed 3 times with PBS. TUNEL fluorescence was detected by the One-step TUNEL In Situ Apoptosis Kit.

Cell viability

For cytotoxicity assays, 5 × 10³ NIH3T3 cells/well in 96-well plates were exposed to Z@P-M (ZnO₂ concentration = 0 − 200 μg/mL) for 12 h via CCK-8 method, absorbance at 450 nm was measured by microplate reader (Biotek800 TS). 6 × 10⁵ NIH3T3 cells/well, 6 × 10⁵ RAW 264.7 cells/well, and 6 × 10⁵ HaCaT cells/well were seeded in 24-well plates, treated with different groups, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. Washing with PBS and staining with Calcein AM/PI for 15 min, the cells were captured under inverted fluorescence microscope and the number of live and dead cells was calculated by Image J.

Cellular uptake

3 × 10⁵ NIH3T3 cells/well were seeded in 24-well plates and treated with different groups. Cellular uptake of RhB labelled Z@P-M MNs and Z#P-M MNs were assessed following 6 h co-incubation with NIH3T3 cells and subsequent 5 min of 1 W/cm² 808 nm NIR irradiation. Cytoskeleton was labelled with Actin-Tracker Green-488, which was diluted 1:100 and incubated with cells for 1 h. Finally, the fluorescence images were captured by CLSM.

Migration capability of NIH3T3 cells

NIH3T3 cells were digested and planted on 6-well plates (3×10⁶ NIH3T3 cells/well). Using a 200 μL pipette tip, scrape the monolayer of cells when confluence reaches 90 %. Then cells were cultured in hypoxia. The cell pictures were captured by inverted fluorescence microscope and the wound area and healing rates were calculated at different time point (0 h, 18 h and 36 h).

$${{{\rm{Migration}}}}\%=({{{{\rm{S}}}}{{{\rm{cratcharea}}}}}_{0{{{\rm{h}}}}}-{{{{\rm{S}}}}{{{\rm{cratcharea}}}}}_{18{{{\rm{hor}}}}36{{{\rm{h}}}}})/{{{{\rm{S}}}}{{{\rm{cratcharea}}}}}_{0{{{\rm{h}}}}}\times 100.$$

(4)

Intracellular O₂ detection

Intracellular oxygen levels were quantified using [Ru(dpp)₃]Cl₂ oxygen-sensitive nanoprobes. The fluorescence of the probe can be strongly extinguished by O₂. 3×10⁵ NIH3T3 cells/well were seeded in 24-well plates and cultured for 24 h, the cells were treated with PBS, blank MNs, Z@P MNs, Z@P-M MNs, and NIR + Z@P-M MNs for 6 h under hypoxia, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. Then cells were incubated with the oxygen sensing probe for 4 h. Cells cultured under normoxia as control. The fluorescence intensity was recorded by inverted fluorescence microscope.

Intracellular ROS detection

3×10⁵ NIH3T3 cells/well were seeded in 24-well plates and cultured for 24 h. Then cells were treated with PBS, blank MNs, Z@P MNs, Z@P-M MNs, and NIR + Z@P-M MNs under hypoxic condition in medium containing LPS (1 μg/mL) for 12 h, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. Cells cultured under normoxia as control. The cells were then incubated with DCFH-DA (10 μM) for 2 h, while Hoechst 33342 was used to labelled the nuclei. Finally, the fluorescence figures were captured by inverted fluorescence microscope.

Immunofluorescence in vitro

3×10⁵ NIH3T3 cells/well were seeded in 24-well plates and cultured for 24 h. Cells were treated with PBS, blank MNs, Z@P MNs, Z@P-M MNs, and NIR + Z@P-M MNs under hypoxic condition in medium containing LPS (1 μg/mL) for 12 h, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. Cells cultured under normoxia as control. HIF-1α primary antibody and RhB labelled secondary antibody were used to stain the cells. F-actin was stained by Actin-Tracker Green-488.

qRT-PCR

3 × 10⁶ NIH3T3 cells/well were plated in 6-well plates and cultured for 24 h. Cells were treated with PBS, blank MNs, Z@P MNs, Z@P-M MNs, and NIR + Z@P-M MNs under hypoxic condition in medium containing LPS (1 μg/mL) for 12 h, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. Cells cultured under normoxia as control. Total RNA isolation of NIH3T3 cells were extracted by super total RNA extraction kit and reverse transcribed to cDNA using cDNA ReverTra Ace qPCR RT kit. PCR was conducted at 95 °C, 10 min denaturation, 57 °C, 10 s annealing, 60 °C, and 10 s extension. The PCR reaction system is 5 μL of 2×SYBR Green Master, primer (0.3 μL) and cDNA (1 μL) and RNase Free dH₂O (3.4 μL). The relative expression was calculated using the 2^△△Ct method. Normalization was based on GAPDH expression levels. Supplementary Table 1 showed the primer sequences.

Acne model establishment and treatment

After female balb/c mice were anaesthetized, the hair on the backs of the female mice was exfoliated with a shaver and depilatory cream, and sterilized with 75 % ethanol. To establish an acne vulgaris model, 100 μL of C. acnes inoculum (2 × 10⁹ CFU/mL) was administered via intradermal inoculation along the dorsal region on day -2. Experimental cohorts consisting of 5 female mice per group were randomly allocated into 6 groups: acne, erythromycin, blank MNs, Z@P MNs, Z@P-M MNs, and NIR + Z@P-M MNs, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min. On day 0, the acne group and erythromycin group were treated with PBS solution and erythromycin ointment, respectively. The blank MNs group, Z@P MNs group, and Z@P-M MNs group were treated by pressing the MNs patches onto the dorsal skin for 5 min before removal. For NIR + Z@P-M MNs group, Z@P-M MNs patches were pressed onto the dorsal skin of female mice for 5 min, followed by irradiation with 808 nm NIR light (1 W/cm²) for 5 min, and finally the patches were removed. the dorsal skin of female mice was photographed on day 0, 3, 5, and 7. On day 7, the collected bacterial samples from the female mice wounds were spread on thioglycollate agar plates to observe bacterial burden. After the female mice were euthanized, the skin was cut 5 mm along the wound edge to acquire skin tissues. The collected skin tissues, heart, liver, spleen, lung, and kidney tissues were fixed, permeabilized, embedded in paraffin, and sectioned.

The biocompatibility of Z@P-M MNs in vivo

Z@P-M MNs were pressed onto the dorsal skin of the female mice for 5 min, followed by 1 W/cm² 808 nm NIR light irradiation for 5 min. After the treatment, the patches were gently removed, and the images of the dorsal skin and H&E staining of skin tissues at different time points (0 − 30 min) were collected. Optical images of the MNs patches before and after application were also captured.

H&E staining

Histological specimens underwent sequential processing commencing with xylene-mediated deparaffinization, followed by ethanol gradient rehydration. Staining protocols initiated with hematoxylin counterstaining (5 min), succeeded by aqueous differentiation (deionized H₂O, 3 min) and acidic differentiation (1 % HCl/ethanol, 30 s), culminating in eosin counterstaining (0.5 % w/v, pH 5.5) for cytoplasmic visualization. The images were obtained using a Slide Scan System SQS-1000.

Hemolysis assay

Fresh mouse blood was centrifuged at 4 °C at 860 g for 15 min to obtain a red blood cell pellet. The red blood cells were washed 3 times with PBS and diluted to a final concentration of 5 % (v/v). Different treatments were applied to 1 mL of the red blood cell suspension, including: (1) dd H₂O, (2) PBS, (3) Z@P-M, (4) NIR + Z@P-M, where 1 W/cm² 808 nm NIR irradiation was applied for 5 min, followed by incubation at 37 °C for 4 h. After treatment, the samples were centrifuged at 860 g at 4 °C for 10 min to collect the supernatant, and the optical density (OD) value was measured at 570 nm. The hemolysis rate was calculated using the following formula.

$${{{\rm{Hemolysis}}}}\%=[({{{{\rm{OD}}}}}_{{{{\rm{sample}}}}}-{{{{\rm{OD}}}}}_{{{{\rm{negative}}}}})/({{{{\rm{OD}}}}}_{{{{\rm{positive}}}}}-{{{{\rm{OD}}}}}_{{{{\rm{negative}}}}})]\times 100.$$

(5)

IHC

Paraffin-embedded tissue sections underwent antigen retrieval followed by incubation with IL-6, TNF-α, CD31, and IL-10 primary antibodies (4 °C, 16 h) and secondary antibodies (25 °C, 1 h). After staining with chromogen, the images were obtained using a Slide Scan System SQS-1000.

Immunofluorescence in vivo

Histological preparations underwent standard dewaxing in xylene followed by ethanol gradient hydration. Antigen retrieval was performed by sodium citrate buffer (pH 6.0). Subsequently, samples were incubated with 0.5 % Triton X-100 for 15 min, blocking with 5 % BSA (25 °C, 60 min), and primary antibody incubation targeting HIF-1α (1:1000 dilution, 4 °C, 16 h). Following triple PBS washes, the samples were incubated with fluorescent secondary antibodies for 1 h and labelled with DAPI. RORγt and IL-22 detection was incubated 1:100 with RORγt and IL-22 primary antibodies overnight. The images were captured by CLSM.

Statistical analysis

Each experiment was conducted three times or more. The experimental data were presented as mean ± SD. Statistical analysis were used by one-way ANOVA with Tukey’s multiple comparisons test (n.s. represented no significance, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). P < 0.05 was statistically significant.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.