Polydopamine-Pd nanozymes as potent ROS scavengers in combination with near-infrared irradiation for osteoarthritis treatment

Summary Excessive reactive oxygen species (ROS) in joints could lead to gradual degeneration of the extracellular matrix (ECM) and apoptosis of chondrocytes, contributing to the occurrence and development of osteoarthritis (OA). Mimicking natural enzymes, polydopamine (PDA)-based nanozymes showed great potential in treating various inflammatory diseases. In this work, PDA loaded with ultra-small palladium (PDA-Pd) nanoparticles (NPs) was employed to scavenge ROS for OA therapy. As a result, PDA-Pd effectively declined the intracellular ROS levels and exhibited efficient antioxidative and anti-inflammatory capacity with good biocompatibility in IL-1β stimulated chondrocytes. Significantly, assisted with near-infrared (NIR) irradiation, its therapeutic effect was further enhanced. Further, NIR-stimulated PDA-Pd suppressed the progression of OA after intra-articular injection in the OA rat model. With favorable biocompatibility, PDA-Pd exhibits efficient antioxidative and anti-inflammatory capacity, leading to the alleviation of OA in rats. Our findings may provide new insights into the treatment of various ROS-induced inflammatory diseases.


INTRODUCTION
Osteoarthritis (OA), a low-grade inflammatory disease of joints, is prevalent worldwide with the rapid increase of the aging population. 1 Nevertheless, effective treatments for OA are still scarce currently. 2 Generally, OA is typically characterized by progressive damage to articular cartilage including the death of chondrocytes and the degradation of the extracellular matrix. This is mainly attributed to homeostasis dysregulation of the articular cartilage microenvironment, mainly caused by the accumulated excessive reactive oxygen species, such as hydroxyl radical ($OH), singlet oxygen ( 1 O 2 ), superoxide anion radical ($O 2 À ), and hydrogen peroxide (H 2 O 2 ) in the joint. [3][4][5] Excessive ROS also can stimulate the production of matrix metalloproteinases (MMPs) in chondrocytes, resulting in progressive loss of articular cartilage. 3 In view of the crucial role of ROS in the occurrence and progression of OA, ROS scavenging is possibly to be an effective strategy for OA therapy. For practical application, it required that PDA-Pd NPs could retain a certain degree of stability and degradability in physiological conditions. Thus, PDA and PDA-Pd were respectively dispersed in PBS, fetal bovine serum (FBS), DMEM, and 5 mM H 2 O 2 solutions for 14 days. As indicated in Figure 2A, PDA and PDA-Pd could maintain a certain degree of stability and were homogeneously distributed in the solutions for more than 1 day. On day 3, it was observed that PDA NPs had deposited at the bottom of PBS and DMEM solutions, whereas, quite a few of PDA-Pd NPs were still distributed in PBS on day 3. Furthermore, when it came to day 14, almost all PDA NPs had deposited on the bottom while only a few of PDA-Pd were observed on the bottom in the PBS buffer. Specifically, PDA and PDA-Pd were still well dispersed in the FBS buffer until the 14th day. In the meantime, PDA-Pd NPs at predetermined time points were also observed by TEM to evaluate their degradability. On day 7, it was observed that PDA-Pd NPs retained the spherical shape in PBS buffer while they started to degrade in 5 mM H 2 O 2 solution with the damaged shapes ( Figure 2B). The above results confirmed that PDA-Pd NPs could retain relative stability in various physiological conditions. Meanwhile, it also could degrade gradually as time goes by, especially in a 5 mM H 2 O 2 solution.

Photothermal behavior
Generally, NIR stimulation could be helpful to improve the unique function of nanocarriers, including accelerating the release of medicines and electron transfer reactions. 16,17 PDA and PDA-Pd NPs were homogeneously dispersed in PBS buffer with different concentrations under NIR irradiation to test the photothermal behavior. Under 0.8 W/cm 2 NIR irradiation, when increasing the concentration of PDA-Pd from 0 to 20, 50, and 100 mg/mL, the temperature jumped from 24.9 C to 36.5 C, 46.5 C, and 55.4 C within 15 min respectively ( Figure 2D). As shown in Figure 2E, compared to the PBS buffer, the temperature of PDA and PDA-Pd increased with the extension of NIR irradiation time. After statistical analysis, for 50 mg/mL PDA alone, the temperature went up from room temperature to 53 C, where it only increased to 44.5 C for PDA-Pd after NIR irradiation. Meanwhile, the corresponding thermal images of PBS, PDA, and PDA-Pd were captured with the same concentration at different time points ( Figure 2C). Compared to PDA, the slightly lower temperature existed for PDA-Pd, which was possibly attributed to the existence of ultra-small Pd NPs affected the photothermal effects. Similarly, increasing the power intensity from 0.4 to 1.6 W/cm 2 , the corresponding temperature increased from 38 C to 59.7 C ( Figure 2F). Significantly, after 4 cycles, the photothermal properties of PDA-Pd still maintained relatively stable ( Figure 2G). It confirmed that PDA-Pd NPs possessed good photothermal effects and photothermal stability. Increasing the concentration and power intensity contributed to a stronger photothermal effect.

ROS-scavenging capacity
The ROS-scavenging capacity was initially evaluated using ROS detection kits. As shown in Figure 3A, PDA presented a certain degree of the scavenging capacity of $O 2 À , H 2 O 2 , $OH, and DPPH$ with the scavenging ratio of 21.05%, 7.14%, 6.44%, and 72.78% detected by ROS detection kits respectively. After Pd NPs loading, the corresponding scavenging effects increased to 49.79%, 13.37%, 8.42%, and 78.25% for PDA-Pd. With NIR irradiation the H 2 O 2 scavenging capacity of PDA-Pd significantly increased from 13.37% to 31.09% compared to without NIR. And NIR irradiation could promote the $OH scavenging capacity from 8.42% to 12.36%, while the ability to scavenge DPPH $ was only slightly increased, from 78.25% to 81.64%. Conversely, NIR irradiation did not improve the $O 2 À scavenging ability of PDA-Pd.
And the ROS-scavenging capacity was also concentration dependent. As indicated in Figure S6, it was observed that increasing the concentration of PDA-Pd could be equal to the improved scavenging capacity of $O2-, $OH, H 2 O 2 , and DPPH free radical.
Besides, we also determined the total ROS-scavenging ability by ROS detection kits. As indicated in Figure S7, the best scavenging ability of total ROS was found when PDA-Pd NPs were added with NIR irradiation ( Figure S7A). And increasing the concentration of PDA-Pd could improve the scavenging capacity of total ROS ( Figure S7B).
At last, the ROS-scavenging capacity was also assessed by electron spin resonance (ESR). As indicated in Figure 3B,

Cell viability
For biomedical applications, it was expected that PDA-Pd NPs possessed good biocompatibility. From the results of the cell counting kit-8 (CCK-8) kit, it displayed that PDA and PDA-Pd had favorable biocompatibility below 50 mg/mL with cell viability above 90% ( Figure 4A). PDA has a similar structure and chemical properties to melanin, a natural biopolymer, with low cytotoxicity and good biocompatibility. 11,18 And multiple previous experiments have confirmed the low cytotoxicity of Pd. 13,19,20 Combined with previous studies as well as the results of this experiment, we confirmed that PDA-Pd has low cytotoxicity.
In the meantime, the protective effects of PDA and PDA-Pd NPs were explored by live/dead staining. As indicated in Figure 4B, for the normal chondrocytes, few dead cells (red) were observed and lots of live cells were observed. After IL-1b stimulation, a lot of dead cells existed. The incubation of PDA and PDA-Pd NPs could effectively protect the chondrocytes avoiding the toxicity of ROS with the increased number of live chondrocytes. After statistical analysis, the percentage of dead chondrocytes was 17.05% for IL-1b stimulated chondrocytes; it decreased to 12.96%, 6.82%, and 6.62% respectively for PDA, PDA-Pd, and PDA-Pd with NIR irradiation treatment ( Figure 4C). It confirmed that PDA-Pd NPs could effectively improve the viability of IL-1b-induced chondrocytes. Specifically, PDA-Pd and PDA-Pd + NIR irradiation presented the best protection effects. The protective effect of PDA-Pd with or without NIR irradiation on chondrocytes was not statistically different.

ROS-scavenging and anti-inflammatory ability
The ROS-scavenging capacities of PDA and PDA-Pd NPs at cellular levels were evaluated using the reactive nitrogen species (RNS) and ROS probes. Compared with normal chondrocytes, the total ROS levels significantly increased for IL-1b-induced chondrocytes (control group) with high-intensity green fluorescence over a large area observed. PDA and PDA-Pd NPs could decrease the total ROS levels significantly. Particularly, the fluorescence intensity decreased more obviously for PDA-Pd with NIR irradiation than that for PDA-Pd alone. It displayed the same tendency for RNS levels where the strongest fluorescence intensity was the control group and PDA-Pd+NIR possessed the lowest fluorescence intensity. After statistical calculation, the mean fluorescence intensity of total ROS levels was 6.01 for the normal chondrocytes, increased to 25.94 for the control group while it became 18.73 for PDA, 12.17 for PDA-Pd, and 9.46 for PDA-Pd+NIR respectively. Similarly, the mean fluorescence intensity for RNS was in the order of the control group> PDA> PDA-Pd> PDA-Pd+NIR ( Figure 5A).
The scavenging capacity of hydroxy radicals of PDA and PDA-Pd NPs at cellular levels was evaluated using the hydroxyl radical detection kit (hydroxyphenyl fluorescein, Maokangbio, China). Compared with normal chondrocytes, the hydroxyl radical levels significantly increased for IL-1b-induced chondrocytes (control group) with high-intensity green fluorescence over a large area observed. PDA and PDA-Pd NPs could decrease the hydroxyl radical levels significantly ( Figure S9).
The levels of inflammatory cytokines (MMP-13 and IL-6) and chondrocyte-specific protein (Col2a1) of cell supernatant were analyzed by ELISA. As indicated in Figure 5B, the protein level of IL-6 was super low (53.73 pg/mL) for the normal chondrocytes, and significantly jumped to 1120.10 pg/mL for the control group. After treatment, it was 736.32 pg/mL for PDA and 140.40 pg/mL for PDA-Pd respectively. The gene expression situation was assessed by qRT-PCR (The genes and the corresponding primer sequences were shown in Table 1). From Figure 5C, compared to normal chondrocytes, the levels of TNFa, IL-6, MMP-13, and MMP-3 significantly ascended for chondrocytes after IL-1b induction. However, PDA and PDA-Pd NPs could descend the levels of inflammatory cytokines, especially for PDA-Pd NPs.
Significantly, under NIR irradiation, it could more effectively decrease the gene expression of IL-6, MMP-3, and MMP-13. Besides, the chondrocyte-specific genes (ACAN and Col2a1) went up by PDA and PDA-Pd NPs treatment compared to the control group, especially for PDA-Pd NPs. And the growing trend became more obvious for PDA-Pd with NIR irradiation.
Finally, the protein levels of inflammatory cytokines (IL-6 and MMP-13) and chondrocyte-specific protein (Col2) were also characterized by immunofluorescence staining. After IL-1b stimulation, the levels of IL-6 and MMP-13 were high while they declined after PDA and PDA-Pd NPs treatment. By statistical calculation, the mean fluorescence intensity of IL-6 was in the order of control group> PDA> PDA-Pd> PDA-Pd+NIR, similar to the iScience Article tendency of mean fluorescence intensity of MMP-13. The expression levels of Col2 decreased significantly after IL-1b stimulation in chondrocytes, however, PDA treatment significantly increased the expression of Col2, and PDA-Pd could further improve the therapeutic effect, especially after NIR irradiation ( Figure 5D).
From the above, it gave proof that PDA-Pd NPs could effectively downregulate the expression of inflammatory cytokines and upregulate the level of chondrocyte-specific proteins due to their excellent antiinflammation and ROS-scavenging capacity. Among them, the optimum therapeutic effects happened to PDA-Pd with NIR irradiation.

In vivo therapy effect
The in vivo OA therapy was implemented by PDA and PDA-Pd NPs intra-articular injection with or without NIR irradiation. First, we explored the photothermal effects of NPs in vivo. As shown in Figures 6B and 6C, it was found that the temperature increased versus NIR irradiation time for SD rats with PDA or PDA-Pd injection. Compared to PBS injection, the equilibrium temperature jumped from 38.2 C to 48.1 C for PDA injection while it changed to 43.2 C for PDA-Pd injection during NIR irradiation. Therefore, it offered the feasibility of in vivo photothermal OA therapy.  iScience Article Meanwhile, the macroscopic observation of joints was shown in Figure 6D. For the OA group, the degree of inflammation gradually became intensified with a rough and erosive surface of joints as time went on. Nevertheless, it had an obvious therapeutic effect after PDA and PDA-Pd NPs intra-articular injection, with a significant reduction of deterioration. Specifically, PDA-Pd with NIR irradiation significantly reduced the degree of inflammation with almost no defect and erosion existing, close to the normal joint at week 8. After macroscopic scoring based on Pelletier's methods, it was 1.1 and 1.0 for the sham group while it increased to 13.6 and 14.3 for the OA group at week 4 and week 8 respectively. For PDA-Pd+NIR, the scores significantly decreased to 3.4 at week 4 and 2.4 at week 8 ( Figure 6E).
To evaluate the OA therapy effect, the proteins in synovial fluid after treatment were analyzed by ELISA. As illustrated in Figure 6F, the level of IL-6 and MMP-13 in synovial fluid was relatively low for the sham group while it was at a high level for the OA group. After PDA and PDA-Pd NPs treatment, the expression of inflammatory factors declined. Specifically, for PDA-Pd with NIR irradiation, the downtrend became more significant. Conversely, the expression of Col2a1 had significantly improved after PDA and PDA-Pd NPs treatment. Among them, PDA-Pd with NIR irradiation possessed the best improvement of the expression of Col2a1.
Besides, the results of histological staining were indicated in Figures 7A and 7B. For H&E staining compared to the sham group, obvious OA features like matrix loss, fissures, and fibrosis existed for the OA group at week 4 and week 8. For PDA, it was observed that the matrix loss was slightly restored. Significantly, the obvious restoration of matrix happened to PDA-Pd, especially for PDA-Pd+NIR at week 4 and week 8 ( Figure 7A). Similarly, for Safranin O staining, the destroyed cartilage layers happened to the OA group while it was partially restored after PDA and PDA-Pd NPs treatment. Specifically for PDA-Pd+NIR, the repaired thick cartilage layer was observed infinitely close to the sham group ( Figure 7B). In order to quantitatively assess the degree of cartilage inflammation, we also did the cartilage pathological scoring based on the Osteoarthritis Research Society International (OARSI) scoring system. 21 It was 1.00 and 1.33 for the sham group while it increased to 5.22 and 5.67 for the OA group at week 4 and week 8 respectively. For PDA-Pd+NIR, the scores significantly decreased to 2.67 at week 4 and 2.22 at week 8 ( Figure S8).
Finally, to evaluate the toxicity of PDA and PDA-Pd in vivo, the major organs were collected after intra-articular injection for 8 weeks. As indicated in Figure 7C, the H&E images of all organs of different groups presented almost the same conditions. It confirmed that it had no significant cytotoxicity for PDA and PDA-Pd NPs intra-articular injection with or without NIR irradiation. And we also completed the blood routine test and blood biochemistry test to study the toxicity of NPs. As shown in Figure S10, there was no significant difference in red blood cell count (RBC), hemoglobin (HGB), hematocrit (HCT), platelet count (PLT), and  iScience Article white blood cell count (WBC) among the four groups in the blood routine test. As indicated in Figure S11, there was no significant difference in the results of aspartate aminotransferase (AST), alanine aminotransferase (ALT), creatinine (CREA), and urea among the four groups. In our study, PDA and PDA-Pd NPs were administered locally in the joint at a quite low dose of 10 mg/rat each time, and no significant toxicity was observed.

DISCUSSION
ROS plays an important role in the occurrence and progression of OA. 22 ROS-scavenging has emerged as an important strategy for treating OA. 23 Among the numerous ROS-scavenging materials, PDA has attracted lots of attention due to its excellent ROS-scavenging ability and biocompatibility. However, PDA, as an antioxidant, also shares some common drawbacks of antioxidants. Antioxidants scavenge ROS by direct redox reactions with ROS, and the content of their reducing groups will be gradually consumed with the clearance of ROS, so their effective therapeutic concentrations are difficult to maintain for a long time, and the frequently repeated administration is required, which brings great difficulties to its practical application. Contrary to antioxidants, nanozymes, a new type of catalytic nanomaterials developed in recent years, such as Fe 3 O 4 , Mn 3 O 4 , and CeO 2 , have quite longlasting and efficient catalytic activity. [24][25][26] But, these nanozymes are also difficult to be widely applied in biomedicine because of their inherent disadvantages of easy agglomeration and difficulty in entering cells. 27 In view of the advantages and disadvantages of PDA and nanozymes, some scholars tried to combine PDA and nanozymes in order to obtain more perfect ROS-scavenging materials. Xinyue Guo et al. developed a complex nanozyme PDA@CeO 2 to maintain normal cell morphology by scavenging ROS. 28  NIR has gradually attracted much attention due to its excellent tissue-penetrating ability, excellent photothermal therapy, and photodynamic therapy effects. PDA, as an analog of melanin, has high light absorption in the NIR region. 7 Palladium nanoparticles also have strong light absorption in the NIR region. 30,31 As a conjugate of PDA and palladium, PDA-Pd theoretically should have good light absorption as well as the excellent photothermal effect in the NIR region. In this study, we found that PDA-Pd has a good photothermal effect both in vitro and in vivo.
Recently, some studies also found that NIR has good photocatalytic activity. 32,33 In this study, both the ROS detection kit and ESR results confirmed that NIR could significantly promote the catalase-like activity of PDA-Pd. The anisotropy of the shape of the Pd nanoparticle shows strong localized surface plasmon resonance (LSPR) in the NIR region of the electromagnetic spectrum. 34 Hot electrons generated by LSPR greatly promote the enzymatic activity of metal nanomaterials. Yue Yang et al. confirmed that Pd has a catalase enzyme-like effect. Using NIR to irradiate Pd can trigger the LSPR effect of Pd nanoparticles, which in iScience Article turn generates hot electrons, resulting in enhanced CAT-like enzyme activity. 15 Shanshan Li and coworkers also found that strained palladium nanosheets had CAT-like activity. Meanwhile, the CAT-like activity of strained palladium nanosheets can be further enhanced by NIR. 35 And PDA can produce a synergistic effect with photocatalysts through p-p* electron transitions, thereby greatly enhancing the photocatalytic activity of the catalysts. 7 From the above, it indicated that PDA-Pd+NIR irradiation possessed the optimum OA therapy effects. Previous research found that PDA had a certain degree of antioxidative function in vitro but was limited by its short effective treatment time in vivo. Therefore, PDA loading with Pd could be helpful to retain long-term therapy in vivo. NIR irradiation also accelerated the catalytic effect of PDA-Pd, further leading to OA therapy with high efficacy.
In summary, giving full-play efficient ROS-scavenging activity of nanozyems is of significance for suppressing OA progression. The novel PDA-Pd NPs were explored to act as nanozymes to alleviate OA. The results demonstrated that PDA-Pd could effectively scavenge ROS, downregulate the expression of inflammatory cytokines, and upregulate the expression of the chondrocyte-specific protein.
With the addition of NIR irradiation, better therapeutic results were achieved. The excellent antioxidative ability and cartilage protection ability of PDA-Pd indicated a very effective method for suppressing OA progression, finally resulting in excellent OA therapeutic effects.

Limitations of the study
There are many kinds of ROS in the inflammatory chondrocyte, such as hydroxyl radicals and superoxide anions. To explore the ability of PDA-Pd to scavenge ROS in the biological environment, we have done the ROS detection kit (DCFH-DA) and the hydroxyl radical detection kit on chondrocytes.
Because of the limited funds and time, we have not done the experiment of superoxide anion detection kit at cellular levels. So the scavenging ability of the intracellular superoxide anion of PDA-Pd is not yet clear.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:   Figures S8, S10, and S11.

Synthesis of PDA and PDA-Pd
PDA and PDA-Pd NPs were synthesized based on previously described work with slight modification. 36,37 In brief, 160 mL anhydrous ethanol, 360 mL deionized (DI) water, and 12 mL NH 4 OH were added into a flask and stirred magnetically at room temperature. Then, 2g DA dissolved in DI water (40 mL) was added into the above mixture dropwise. Subsequently, the reaction lasted for 24 h before the black product was collected by centrifugation for 8 min at 11000 rpm. PDA NPs were finally obtained after vacuum drying. Subsequently, 100 mg PDA NPs were dispersed in 80 mL anhydrous ethanol and 6 mL DI water, following by the dropwise addition of 2 mL Na 2 PdCl 4 aqueous solution (10 mg/mL). The mixture was magnetically stirred and reacted for 3 h before adding ascorbic acid solution (10 mg/mL, 6 mL) for another 2 h. The final product PDA-Pd was obtained after centrifugation and vacuum drying.  To determine the chemical and molecular structures of PDA and PDA-Pd, the FTIR spectrum (IRAffinity-1S, Shimadzu, Japan), UV-vis spectra (UV-2700, Shimadzu, Japan), and Raman spectrum (Horiba, France) were applied respectively. The valence state, element composition, and content of nanoparticles were studied by XPS (Thermo ESCALAB 250Xi, Thermo Scientific, USA). The zeta potential of PDA and PDA-Pd was detected with a zeta sizer (Nano-ZS, Malvern, UK) in DI water. Additionally, the size, morphology, structure, and elementary composition of NPs were observed by TEM coupled with energy-dispersive X-ray spectroscopy (EDS) (FEI talos f200s, FEI, USA). Furthermore, the three-dimensional structure of PDA and PDA-Pd was evaluated by atomic force microscopy (AFM) on a multimode microscope (Bruker, Karlsruhe, Germany). In addition, to analyze the thermal properties of the nanoparticles, thermogravimetric analysis (TGA) was detected on a TGA instrument (TGA Q500) under N 2 atmosphere using a TGA instrument (STD650, TA, USA). Finally, the crystallization and molecular structure of PDA and PDA-Pd NPs were measured by XRD on an X-ray diffractometer (Miniflex 600, Rigaku, Japan), and the detailed Pd content of PDA-Pd was analyzed by ICP-OES (Thermo, USA).

Photothermal effect investigation
To systematically assess the photothermal effects, PDA and PDA-Pd were placed under NIR irradiation. In brief, PDA-Pd NPs dispersed in PBS buffer (50 ug/mL) were irradiated by a NIR laser with varied power densities of 0.4, 0.8, and 1.6 W/cm 2 . Then PDA-Pd NPs were dispersed in PBS buffer with different concentrations (20, 50, 100 mg/mL), and then irradiated at 0.8 W/cm 2 respectively. Subsequently, PBS buffer, PDA, and PDA-Pd NPs dispersed in PBS buffer (50 ug/mL) were respectively irradiated with the NIR laser (0.8 W/ cm 2 ). A FLIR NIR camera was applied to collect the corresponding temperatures. Specifically, for the''on and off'' experiment, the PDA-Pd solution was irradiated for 15 min and cooled for 15 min. The above steps were repeated 4 times, and the corresponding temperatures were collected.

In vitro degradability
To investigate the stability and degradability, PDA and PDA-Pd were dispersed in PBS, FBS, DMEM (Thermofisher, China), and 5 mM H 2 O 2 solution with the same concentration of 50 mg/mL respectively. At predefined time points, their dispersion conditions were imaged and their morphology was observed by TEM.

ROS scavenging capability
The ROS scavenging abilities of the nanoparticles were examined by the detection kit. Briefly, to detect the CAT-like activity, we first detected the H 2 O 2 scavenging capability of PDA-Pd at different concentrations (20, 50, 100mg/mL) by a catalase assay kit (Beyotime, China), the absorbance at 520 nm was detected by a microplate reader (Thermo Scientific, USA). Following the same method, we detected the CAT-like activity of PDA, PDA-Pd, and PDA-Pd with NIR irradiation at a concentration of 50 mg/mL respectively, NIR irradiation was implemented at 0.8 W/cm 2 for 5 min during incubation.
To detect the SOD-like activity, we detected the $O 2 scavenging activity of PDA-Pd at different concentrations (20, 50, 100 mg/mL) by a total superoxide dismutase assay kit (Beyotime, China). Different concentrations of PDA-Pd NPs were added into the working solution. Then the absorbance at 450 nm was detected after standing for 30 min and calculated the $O 2 scavenging ratio by comparing it with the blank. Following the same method, we detected the SOD-like activity of PDA, PDA-Pd, and PDA-Pd with NIR irradiation at a concentration of 50 mg/mL.
Besides, the absorbance at 515 nm, 536 nm, and 734 nm was respectively measured to investigate the DPPH free radical, $OH, and total ROS scavenging capacity respectively following the similar above steps by a DPPH free radical scavenging ability detection kit (Solarbio, China), a micro hydroxyl free radical scavenging capacity assay kit (Solarbio, China) and total antioxidant capacity assay kit with ABTS method (Beyotime, China).
At last, the ROS scavenging ability was also evaluated by electron spin resonance (ESR, Bruker A300, Germany). In brief, after the working solution was treated with PDA, PDA-Pd, and PDA-Pd + NIR irradiation (0.8 W/cm 2 , 5 min) at the same concentration of 50 mg/mL respectively, the ESR signals of the residual $OH, $O 2 -, and 1 O 2 were measured.

Inflammatory factors expression
It was well known that the levels of biomarkers IL-6 and MMP-13 reflect the degree of inflammation in chondrocytes and Col2a1 reflects the functional status of chondrocytes. To determine the protein expression levels, the supernatant of chondrocytes was analyzed by ELISA. Briefly, chondrocytes were seeded into 6-well plates with a density of 1310 6

Photothermal effects in vivo
For the in vivo photothermal effect detection of materials, four 8-week-old SD rats were bought from the Animal Research Committee of Guangxi Medical University. After the rats were anesthetized, 100 mL of PBS buffer, PDA solution (50 mg/mL), and PDA-Pd solution (50 mg/mL) were injected into the right knee joints of different rats. After 12 h, the rats were anesthetized again and then irradiated with NIR (808 nm, 0.8W/cm 2 ) on the right knee joint. An FLIR NIR camera was used to capture the images and temperature values during NIR irradiation.

In vivo OA therapy
A total of 45 SD rats (male, 150$180 g) were applied for in vivo experiments. The OA models were established via anterior cruciate ligament transection (ACLT) 4 weeks before further treatment and all rats were divided into 5 groups randomly: 1) Sham group: SD rats with only incising the skin and capsule of the knee joint, without further treatment; 2) OA group: OA models with intra-articular injection of PBS (0.2 mL); 3) OA + PDA group: OA models with intra-articular injection of PDA solution (50 mg/ml, 0.2 mL); 4) OA+ PDA-Pd group: OA models with intra-articular injection of PDA-Pd solution (50 ug/mL, 0.2 mL); 5) OA+ PDA-Pd+NIR group: OA models with intra-articular injection of PDA-Pd (50 ug/mL, 0.2 ml) and NIR irradiation. The intra-articular injection was implemented twice a week for 4 or 8 weeks, and NIR irradiation was implemented once per day with an intensity of 0.8 W/cm 2 for 5 min.
Finally, rats were sacrificed by intraperitoneal injection of an overdose of sodium pentobarbital at pre-designed time points (4 and 8 weeks post-treatment). Immediately after euthanasia, 100 mL of PBS buffer was injected into the knee joints of rats, then the syringe was quickly withdrawn after mobilizing the knee joint of the rat to collect the synovial fluid. Subsequently, the knee joints and major organs (liver, spleen, heart, lung, kidney) were harvested. And the macroscopic observation of knee joints was scored according to Pelletier's macroscopic scoring methods with some modifications. 39 In detail, the depth of the erosion of the articular cartilage was graded on a scale of 0 to 4, where 0 represented the normal surface, 1 represented the slight fibrillation or yellowish discoloration of the surface, 2 represented the erosion extending to the surface or middle layer, 3 represented the erosion extending to the deep layer, and 4 represented the erosion extending to the subchondral bone. By adding the grades of the medial condyle, the lateral condyle, the medial plateau, and lateral plateau, the macroscopic score of the total joint was obtained. 39 Besides, the inflammatory factors (IL-6 and MMP-13) and Col2a1 contents of synovial fluid were measured by the ELISA kit (Solarbio, China), and the absorbance was measured at 450 nm. Finally, the harvested knee joints were fixed by 4% PFA for 2 days and decalcified in 0.5 M EDTA (pH= 7.2) for 1 month. Then the joint samples were embedded in paraffin and sectioned (5 mm thickness) for hematoxylin and eosin (H&E) staining, and Safranine O-fast green (Solarbio, China) staining. And the OA situation was evaluated according to the OARSI scoring system as described previously by three blinded observers.
At last, the major organs (liver, spleen, heart, lung, and kidney) were immersed in 4% PFA for fixation, embedded, sectioned, and then stained with H&E for observation.
The study of the toxicity of NPs.
A total of 12 SD rats (male, 150$180 g) were used to detect the toxicity of the nanoparticles. All rats were randomly divided into the following four groups: 1) Control group: Intra-articular injection of PBS (0.2 mL).
2) PDA group: Intra-articular injection of PDA solution (50 ug/mL, 0.2 mL). 3) PDA-Pd group: Intra-articular injection of PDA-Pd solution (50 ug/mL, 0.2 mL). 4) PDA-Pd + NIR group: Intra-articular injection of PDA-Pd (50 mg/mL, 0.2 mL) and NIR irradiation. The intra-articular injection was implemented twice a week for 4 weeks, and NIR irradiation was implemented once per day with an intensity of 0.8 W/cm 2 for 5 min. iScience Article Finally, at the pre-designed time point (after 4 weeks of treatment), the rats were fasted overnight prior to sampling. Then the rats were anesthetized by intraperitoneal injection of pentobarbital sodium. The venous blood was collected from the rats in all groups prior to sacrifice. Approximately 1.5 mL blood sample from each rat was collected into the tube containing heparin as the anticoagulant for the blood routine test (RBC, WBC, HGB, HCT, and PLT). And another 3 ml of venous blood was coagulated in a clean dry tube at room temperature for 1 h. Then the blood samples were centrifuged at 3500 rpm for 10 min at 4 C, then the supernatant serum was collected for the blood biochemistry tests (AST, ALT, CREA, and urea).

STATISTICAL ANALYSIS
All statistical comparisons of means were performed using GraphPad Prism 8 software (GraphPad, USA). Multiple comparison tests were analyzed by one-way analysis of variance (ANOVA