Synthesis and characterization of the nanoparticles
The processes to synthesize the modified Zn2+ NPs were illustrated in Fig. 1a. ZnO2 NPs were first prepared by a wet chemistry synthesis method. XRD patterns (Additional file 1: Figure S1a) proved the high crystallinity of ZnO2, with all peaks assigned to the ZnO2 phase (PDF NO.13-0311), implying the desirable synthesis of pH-sensitive ZnO2 NPs. In addition, the acid-triggered properties of Zn2+ release and H2O2 production were investigated. Because of their pH-dependent destruction, the ZnO2 NPs released 95.5% and 7.5% of Zn2+ after incubation at pH 5 and 7.4 for 24 h, respectively, as measured by ICP–OES (Additional file 1: Figure S1b). Similarly, a large amount of H2O2 was generated with the acidic decomposition of ZnO2 (pH = 5), whereas little H2O2 was detected under an alkalescent environment (pH = 7.4) (Additional file 1: Fig.S1c). Collectively, these results indicated the effective pH-responsive release of Zn2+ ions and H2O2 from ZnO2 NPs.
ZnO2 NPs were then encapsulated in a mannose-decorated lipid bilayer (liposome, lipo) with catalase (Cat) to form functional ZnCM NPs . As shown in Fig. 1b, monodispersed ZnO2, ZnC and ZnCM NPs with uniform morphologies were successfully synthesized, as evidenced by the evident membrane structures after encapsulation onto ZnO2 . The zeta potentials of ZnO2, ZnC and ZnCM NPs were also determined to examine the changes in the NP surface characteristics. After lipid coating, the zeta potential changed from + 2.77 mV for ZnO2 to − 9.66 mV (ZnC) and then to − 14.1 mV (ZnCM NPs) (Fig. 1c), indicating successful lipid encapsulation. Herein, the increased negative surface potential of ZnCM compared with ZnC could be attributed to the proton dissociation of the targeting mannose moiety, which implied the successful decoration.
Previous studies have shown that ZnO2-based nanoplatforms could enhance oxidative damage in cancer cells by increasing mitochondrial H2O2 and releasing endogenous reactive oxygen species (ROS), thereby promoting cell death [32, 34,35,36]. Therefore, it appeared that detrimental H2O2 could directly destroy DCs instead of inducing immune-tolerant tDCs. Hence, it became of great significance to decrease the cytotoxicity of ZnO2 NPs while maintaining their Zn2+-associated DC tolerogenic ability. It has been shown that DC-involved immune responses take place under oxygen-limited conditions , which reminds us that inducing H2O2 to generate O2 could not only impede the oxidative reactions of ROS but also alleviate the hypoxic state of the activated DCs, altogether contributing to the induction of tDCs. Therefore, we coencapsulated ZnO2 and Cat with liposomes to develop an acid-sensitive catalytic oxygen (O2)-producing cascade nanosystem (ZnCM NPs). To further confirm that Cat was encapsulated into the liposomes, an SDS–PAGE experiment was performed (Fig. 1d). By comparing free Cat with acidolysed ZnCM NPs, a single molecular weight protein blot was observed at the same location, which indicated the successful encapsulation of Cat into the nanosystem. To explore the acid-dependent O2 release from ZnC and ZnCM NPs, we determined the dissolved oxygen concentrations by immersing the NPs in different PBS solutions (pH = 7.4, pH = 5) using a portable dissolved oxygen meter. As shown in Fig. 1e, both ZnC and ZnCM NPs showed rapid O2 production immediately after immersion in an acidic solution (pH = 5). In contrast, negligible O2 was detected at pH 7.4. These results were consistent with the acid-triggered H2O2 production nature of the ZnO2 NPs. The Zn2+ release characteristics of ZnC and ZnCM NPs under different pH conditions were also studied (Fig. 1f). As expected, Zn2+ showed distinct acid-sensitive release properties from ZnC and ZnCM NPs. Collectively, these results showed that biofunctional Zn2+ and O2 could be released effectively in an acidic environment from an acid-sensitive catalytic ZnCM cascadic nanosystem.
Biocompatibility of the ZnCM nanoparticles
Next, we investigated the biocompatibility of the various Zn2+ NPs (Fig. 2a). After ZnO2 NP treatment, the results showed that the DC viability decreased significantly with increasing ZnO2 concentrations, as evidenced by the minimal cytotoxicity at 6.25 μg/ml ZnO2 compared with the control (Ctrl), which was significantly different. In contrast, after coencapsulation with Cat and lipo, Zn2+ from ZnC NPs in the range of concentrations of 0.75 to 25 μg/ml barely showed significant cytotoxicity against DCs in vitro, indicating the noncytotoxicity of the modified NPs via the elimination of toxic H2O2 in vitro. After further loading with both Cat and mannose, the ZnCM NPs ranging from 0.75 to 25 μg/ml hardly exhibited any ability to kill DCs. These results demonstrated that ZnO2 NPs exerted oxidative damage in DCs, which could be alleviated by coencapsulation with Cat, implying the potential biocompatibility of ZnCM NPs in DC-targeted immune responses.
Cellular uptake of the nanoparticles in vitro
DCs have been deemed to be vital during immune response initiation; therefore, targeted drug delivery into DCs not only presents efficient immune regulatory functionality but also, more importantly, avoids the undesirable leakage of ZnO2. Numerous DC-targeted ligands have been explored in previous DC-mediated immunotherapies, such as LSECtin, DC-SIGN , and mannose [26, 38]. Herein, mannose was further anchored onto our nanocarrier for targeted delivery into DCs. In Fig. 2b, after 2 h of incubation, we found that compared with the nontargeted ZnC NPs, fluorescent ZnCM NPs showed better intracellular localization within the DCs, indicating the DC-targeting ability of ZnCM. Additional in vitro Bio-TEM observations confirmed the CLSM results, as evidenced by the significant ZnCM NP uptake in DC intracellular lysosomes (Fig. 2c), while few ZnC NPs were found in DC sections. Further CLSM observations confirmed these lysosome-colocalization results, showing the ZnCM NPs were uptaken by DC lysosomes (Additional file 1: Figure S2). These results showed that the ZnCM NPs specifically entered DC lysosomes in vitro, which enabled the pH-dependent release of Zn2+ and O2 during the immune response of ZnCM NPs.
Furthermore, we aimed to unravel the underlying internalization mechanisms of ZnCM NPs into DCs. Several possible routes were found to participate in the uptake of the exogeneous NPs, such as clathrin-mediated endocytosis, caveolin-mediated endocytosis, micropinocytosis and clathrin-caveolin-independent endocytosis . Hence, various pharmacological inhibitors (amantadine, genistein, amiloride and cytochalasin D) were used. Amantadine prevented the budding of clathrin-coated pits; genistein retarded the activation of the Src family of tyrosine kinases; amiloride inactivated Na+/H+ channels; and cytochalasin D destroyed the cytoskeleton to inhibit NP transportation . Figure 2d showed that cytochalasin D significantly suppressed the internalization efficiency of ZnCM NPs compared to amantadine, genistein and amiloride after 2 h of incubation, indicating the feasible participation of the DC cytoskeleton in the process of ZnCM NP ingestion, which was independent of clathrin-caveolin-mediated endocytosis.
Intracellular pH-dependent Zn2+ and O2 release from the nanoparticles
Acidic lysosome-dependent uptake of external stimuli was found to play vital roles in the biological behaviour of DCs [41, 42]. Herein, ZnCM NPs were taken up by DC lysosomes in a manner that was dependent on the cellular cytoskeleton transition. NPs characterization showed efficient pH-dependent Zn2+ and O2 extracellular release from the ZnCM NPs in vitro. Therefore, the intracellular Zn2+ and O2 release of ZnCM NPs in DCs was investigated.
Treatment with Zn2+ ions has attracted great attention in comparison with other metal ions due to its distinct therapeutic characteristics . Various biological metabolic pathways have been found to include zinc homeostasis, such as enzymatic activity, DNA transcription, protein synthesis and vascularization . Zinc oxide (ZnO) NPs have been widely acknowledged as biocompatible candidates for desirable applications in tissue repair, antibacterial, anti-inflammatory, biomedical imaging and nanocarrier development [45, 46]. Nonetheless, the performance of ZnO was significantly restrained owing to its crystalline stability, leading to a lower level of active Zn2+. Therefore, we synthesized ZnO2-based ZnCM nanoclusters via a green, portable, and convenient process that not only released Zn2+ effectively to increase DC intracellular Zn2+ levels for immune response suppression but also further decomposed the toxic byproduct H2O2 into O2, which ingeniously alleviated the hypoxic state of activated DCs. Figure 3a demonstrated that compared with ZnC-treated igDCs, ZnCM NPs effectively released zinc ions in igDCs, as exemplified by the increased CLSM signals, which indicated that the intracellular Zn2+ concentration was significantly increased after biocompatible ZnCM NP administration. Additionally, the O2 levels in DCs were augmented significantly after NP treatment, as shown by the decreased fluorescence of the O2 indicator [Ru(dpp)3]2+Cl2, which was used to detect the hypoxic state  (Fig. 3b, Additional file 1: Figure S3). Notably, the ZnC NPs showed reduced red fluorescence in DCs in comparison with the vehicle, signifying the intracellular decomposition of H2O2 from the NPs catalysed by Cat. More importantly, the [Ru(dpp)3]2+Cl2 fluorescence in the DCs was alleviated more significantly after treatment with ZnCM NPs than ZnC NPs, indicating that targeted delivery into DCs via mannose was able to ameliorate hypoxia effectively. Hence, these results indicated that compared with vehicle and ZnC, ZnCM NPs could release Zn2+ and O2 efficiently within the targeted DCs in a pH-dependent manner, which encouraged us to explore the subsequent immune effects of these NPs to induce tDCs from igDCs.
In vitro induction of tDCs by nanoparticles
During immune activity, DCs express pattern recognition receptors (PRRs) in response to extracellular signals, which is followed by a series of phenotypic and functional transformations termed the DC activation phase. These activities lead activated DCs to differentiate into immunogenic APCs, propelling the downstream proliferation and differentiation of antigen-specific CD4+ T cells into their corresponding effector cells  to initiate immune activity. The surface molecules presented on DCs are normally major histocompatibility complex (MHC) markers, which deliver the first-line activating stimulus to T cells and are therefore referred to as “signal 1”. In addition, a variety of costimulatory molecules, such as CD80 and CD86, are engaged in mediating signals that are vital for T-cell fate, known as “signal 2”. Finally, igDCs release a number of extracellular mediators (“signal 3”) to stimulate downstream CD4+ T cells, such as IL-1, IL-6, IL-12α, IL-12β, and TNF-α, to initiate the development of the T-cell response, altogether contributing to the origination of the activated adaptive immune response in RA . Hence, after elucidating the intracellular Zn2+ and O2 release from the ZnCM NPs in DCs, we sought to uncover the potential NP immune induction of tolerogenicity in DCs by exploring the possible repression of signals 1–3.
To investigate the role of ZnCM NPs in the regulation of zinc ion homeostasis towards DC tolerogenicity, we first explored the expression of surface markers (signals 1 and 2) on igDCs via flow cytometry analysis (Fig. 3c). Upon stimulation with LPS, igDCs (vehicle) exhibited higher expression of CD86, CD80 and MHC class II on CD11c + cells, which indicated the activation of igDCs. However, after administration of both ZnC and ZnCM NPs, the DC levels of CD86, CD80 and MHC class II (signals 1 and 2) decreased significantly compared with the vehicle group, indicating the suppressed immunogenicity of DCs after zinc supplementation. Further immunofluorescent staining confirmed the above findings (Additional file 1: Figure S4), as demonstrated by the decrease in signal 1- and signal 2-positive staining of the DC maturation surface markers after Zn2+ NP treatments compared with Vehicle treatment. All these data showed that ZnO2-based NPs could switch igDCs towards tDCs, which was attributed to the increased intracellular Zn2+ and O2 concentrations.
Upon LPS stimulation of surface ligands, DCs activate NF-κB phosphorylation to promote targeted mRNA transcription, including IL-1, IL-6, IL-12α, IL-12β, and TNF-α (signal 3), thereby triggering a subsequent immune response . Herein, we found that after stimulation with LPS, igDCs were activated, as demonstrated by the increased mRNA levels of IL-1, IL-6, IL-12α, IL-12β, and TNF-α compared with the those of the Ctrl. However, ZnC NPs decreased the mRNA levels of IL-1, IL-6, and IL-12β, while ZnCM NPs significantly inhibited those of IL-1, IL-12β, and TNF-α in contrast with LPS-induced vehicle (Additional file 1: Figure S5a). Cytometric bead array (CBA) immunoassays also showed that igDCs could produce larger amounts of IL-6, IL-12, and TNF-α than Ctrl cells, which were reduced significantly in response to ZnC and ZnCM NP administration (Additional file 1: Figure S5b). These data indicated that ZnCM NPs potently suppressed LPS-induced DC immunogenicity by repressing signals 1, 2 and 3, implying the potential immune tolerogenicity of ZnCM NPs in vitro.
Repression of T cells by tDCs after nanoparticle treatment
The interaction between DCs and CD4+ T cells was shown by an increase in a series of key biomarkers, such as RANKL, IL-12, IFN-γ, and TGF-β , which promoted the development and progression of RA. igDC stimulation activated CD4+ T-cell homeostasis by facilitating the release of inflammatory cytokines from functional T cells . Therefore, we analysed the proliferation of CD4+ T cells and their CBA levels in vitro after coculture with NP-treated igDCs to identify the role of ZnCM NPs in the modulation of tDC capability towards T cells [49, 52]. First, the proliferation of splenic OT-II CD4+ T cells was assessed by flow cytometry, which showed that coculture with LPS-treated DCs (igDCs) promoted the proliferation of CD4 + T cells. However, ZnC/ZnCM NP-treated cocultured tDCs had a reduction in the numbers of CD4 + T cells and their proliferation compared with those of LPS-treated igDCs, indicating that both ZnC and ZnCM NPs induced the ability of tDCs to silence T cells (Additional file 1: Figure S6a). Furthermore, CBA analysis of CD4+ T cells after DC coculture demonstrated that despite the increase in IL-17 and IFN-γ in T cells after LPS-treated igDC coculture, both ZnC and ZnCM NPs decreased this tendency, among which ZnCM showed more significantly decreased levels of IL-17 and IFN-γ than ZnC NPs (Additional file 1: Figure S6b). Therefore, these results suggested that ZnCM NPs could affect the proliferation and activation of CD4+ T cells by modulating the immune crosstalk between tDCs and functional CD4+ T cells.
Repression of the deubiquitinase OTUB1 induced tDCs by CCL5 degradation via NF-κB signalling
The data reported above suggested that ZnCM NP targeted delivery into DCs could induce tDC transition to suppress CD4+ T-cell activation, thereby inhibiting the progression of RA. To further explore the mechanism by which hyperoxic Zn2+ homeostasis regulates igDC generation, label-free quantitative proteomics analysis of igDCs treated with ZnCM NPs was performed with an RPLC–MS/MS system. Among the differentially expressed proteins in the various groups, protein intensity based on the MaxLFQ algorithm was used to analyse all the proteins in each sample, as shown by the normalized LFQ intensity ratio > 1 or < 1 between groups. The proteomics results showed that the numbers of upregulated proteins and downregulated proteins were 133 and 149 in the Vehicle vs. Ctrl groups, 154 and 154 in the ZnCM vs. Ctrl groups, and 124 and 106 in the ZnCM vs. Vehicle groups. Among all the positive protein candidates screened, the proteins IL-1α, IL-1β, CCL5, Rnaseh2c, Vwde, Cd14, and Bst2 showed the most significant vehicle/Ctrl ratios > 1, indicating their increased protein level in igDCs compared with Ctrl. More importantly, these candidates also exhibited a ZnCM/vehicle ratio ≤ 0.85, indicating significantly decreased protein levels in igDCs after ZnCM NP treatment (Fig. 4a). Herein, CCL5 [chemokine (C–C motif) ligand 5], which leads immune cell migration towards targeted inflammatory sites, was found to play an important role in regulating the local immune response . Moreover, CCL5 can mediate the recruitment of T cells and eosinophils , resulting in the activation of the T-cell response. Therefore, we selected CCL5 as the target molecule for Zn2+ homeostasis regulation against igDCs in RA. Our proteomics study suggested that ZnCM downregulated the expression of CCL5 in igDCs, inducing the activation of tDCs to further inhibit the T-cell response.
Next, we sought to identify the mechanisms regulating the decreased level of CCL5 in tDC transformation by ZnCM NPs. The ubiquitin–proteasome system (UPS) is an indispensable regulatory system for cellular protein homeostasis that balances intracellular protein degradation . Particularly in DCs, the suppression of deubiquitinase function in the UPS cascade leads to alterations in DC phenotype and function, which has propelled the development of novel immunotherapies [56,57,58]. Based on these findings, we screened a series of ubiquitinases and deubiquitinases in igDCs treated with NPs. Figure 4b revealed several active candidates that regulated protein degradation, among which the OTU domain of ubiquitin aldehyde binding 1 (OTUB1) showed a significant decrease in ZnCM-treated igDCs compared with vehicle-treated igDCs. Previously, OTUB1 had emerged as an effective deubiquitinase to cleave K48-linked ubiquitin chains to inhibit protein degradation . It was reported that OTUB1 deubiquitination is closely associated with the biological functions of DCs, which inhibit CD4+ T-cell proliferation and IFN-γ cytokine secretion in allogenic mixed lymphocyte reactions (allo-MLRs) . Our results suggested that ZnCM NPs significantly downregulated the expression of the deubiquitinase OTUB1 in igDCs. Therefore, we selected OTUB1 as the target kinase for Zn2+ homeostasis regulation by ZnCM NPs to induce tDCs in RA.
However, the possible crosstalk between CCL5 and OTUB1 has yet to be explored. Hence, immunoprecipitation (IP) analysis was employed and showed the intimate connection between OTUB1 and CCL5 in DCs (Fig. 4c), indicating that CCL5 could serve as the target molecule for OTUB1 deubiquitination. Specifically, by inhibiting the activation of OTUB1 deubiquitination, ZnCM NPs could promote CCL5 degradation to induce tDCs and further inhibit the T-cell response, exhibiting Zn2+ homeostasis regulation by ZnCM NPs in RA.
In addition, it was shown that the expression of CCL5 was closely related to the activation of NF-κB signalling, leading to regulatory effects on cell fate . Previous studies have shown that extracellular stimuli can activate NF-κB signalling, which contributes to the increased expression of CCL5 . In addition, OTUB1 was found to modulate Toll-like receptor (TLR)-induced igDC maturation via NF-κB signalling, as exemplified by activation of the NF-κB kinases TRAF6 and TAK-1 . These studies implied the possible participation of NF-κB signalling in Zn2+ homeostasis regulation against RA. Herein, we initially verified the downregulated expression of CCL5 in NP-treated DCs compared with vehicle-treated igDCs (Fig. 4d). Next, we found that compared to vehicle-treated igDCs, ZnCM-treated igDCs showed decreased expression of p-IKKα and p-p65 and increased levels of iκBα, indicating inhibited activation of NF-κB signalling. Moreover, the Akt and AMPK pathways hardly showed significant effects after NP intervention in comparison with vehicle-treated igDCs, implicating that the Akt and AMPK pathways were not actively involved in mediating Zn2+ homeostasis regulation in RA (Additional file 1: Figure S7). Collectively, the data above demonstrated that ZnCM NPs could inhibit the activation of OTUB1 deubiquitination, thereby promoting CCL5 degradation via NF-κB signalling to induce tDCs for further T-cell response inhibition, which explained the molecular mechanisms of Zn2+ homeostasis regulation by ZnCM NPs to treat RA.
Nanoparticle localization in vivo
It is unanimously believed that the spleen is the largest immune organ of the whole body and accumulates the greatest number of immune cells. Hence, it is proposed that splenic DCs could serve as an ideal target for ZnCM localization in vivo. Herein, collagen-induced arthritis (CIA) mice representing an RA model were injected with various NPs. Additional file 1: Figure S8 showed the normal physiological morphologies of the heart, liver, spleen, lung, and kidney after CIA RA mice were injected with NPs for 4 weeks, indicating the in vivo biosafety of the ZnC and ZnCM NPs. Next, we sought to uncover the in vivo localization of the ZnCM NPs. As shown in Fig. 5a, the Zn2+ contents in 1 g of heart, liver, spleen, lung, and kidney tissue (µg/g) from ZnC- and ZnCM-treated mice were determined, and a higher concentration of Zn2+ was found in the spleen after ZnCM NP injection than in the liver, heart, lung, or kidney after ZnC NP injection. Furthermore, the ex vivo fluorescence data of the targeted organs from RA mice injected with ZnC NPs demonstrated that these NPs showed significantly more accumulation in the liver than in other organs (Fig. 5b). However, after administration of ZnCM NPs, the fluorescence signal in the spleen increased significantly, leading to a relative decrease in fluorescence intensity in the liver, indicating the possible spleen-targeting ability of ZnCM NPs in vivo. These results indicated that ZnCM NPs most significantly targeted the spleen in RA mice, thereby exerting tolerogenicity on DCs in vivo. This effect could be explained by the greater accumulation of immune cells in the spleen than in the liver and the effective DC-targeting effects of ZnCM NPs, which resulted in the localized splenic targeting of ZnCM NPs in vivo. Next, we used Bio-TEM to observe the intracellular localization of the NPs in the spleen in vivo. Compared with ZnC NPs, ZnCM NPs were internalized in the lysosomes of splenic DCs (Fig. 5c). The above in vivo data indicate that in RA mice, ZnCM NPs significantly targeted DCs in the spleen in comparison with the targeting of ZnC NPs.
Alleviation of rheumatoid arthritis manifestations in vivo by nanoparticles
After NP localization was confirmed in splenic DCs, the in vivo anti-RA effects of the ZnCM NPs were studied. Figure 5d and e show that 4 weeks after NP administration, compared to Ctrl mice, RA mice (vehicle-treated) developed significant swelling of the ankle in the hind ankle, which peaked at 1–2 weeks and was followed by a slight decline for the subsequent 2 weeks; this result indicated a certain amount of self-relief after the excessive immune response in RA CIA mice. In addition, after ZnC NP treatment, the RA mice developed an increase in the degree of swelling in the hind leg ankle at 1 week, which was followed by a significant decrease for the following week compared to vehicle-treated mice. However, no significant difference was found in ankle swelling between vehicle- and ZnC-treated mice in the 3rd and 4th weeks. In contrast, after ZnCM NP treatment, the hind limb ankle experienced a consistent and gradual reduction in the swelling perimeter for 4 weeks compared with RA mice, denoting the significant RA treatment efficacy of ZnCM NPs in vivo.
In addition, inflammation in the hind ankle can cause pain while walking, thus affecting walking gait posture. Therefore, we performed gait analysis to provide a quantitative assessment of behavioural walking changes in RA mice injected with various NPs. Walking gait data were first collected by recording the footprints on the sensor screen. It was found that compared with Ctrl mice, RA mice tended to shift their body weight to the front and avoid stepping on their inflamed hind limbs, resulting in reduced pressure from the hind pawprints compared with the front pawprints. However, after effective treatments, prediseased mice induced an even, balanced pressure on all four paws. Figure 5f showed the changes over the time course of the walking single-limb stance time in the different treatment groups. The RA mice had a reduction in the single time stance of the hind limbs (shown in blue and green), which was significantly recovered by treatment with ZnCM NPs. The quantitative hind limb pressure results indicated that the hind paw pressure of both the ZnC- and ZnCM NP-treated mice increased significantly compared to that of RA mice, despite the evident decrease in hind paw pressure in RA mice compared with Ctrl mice (Fig. 5g). These results indicated that ZnCM NP treatment ameliorated the changes in walking posture caused by RA, which might be attributed to the alleviation of inflammatory pain by NP treatment. Furthermore, the hind-base width, which represents the average distance between the two hind paws, was analysed to evaluate the coordination of RA mice (Fig. 5h). Diseased mice were inclined to place their hind paws farther apart than Ctrl mice, which has been labelled the hallmark of unsteady gaits . After treatment with ZnC and ZnCM NPs, the hind-base widths were partially restored compared with those of RA mice, indicating that the ZnCM NPs could ameliorate walking imbalance to a certain extent compared to RA mice. The mean rotation angle is the average value of the angle between the axis from the mouse mouth tip to the tail tip along the central longitudinal axis. This angle represents a twisted posture of the mouse body while walking caused by the inflamed ankle . Figure 5i demonstrates that there was an increase in the mean rotation angle in RA mice compared with Ctrl mice, which was attributed to severe inflammation in the hind ankle. In contrast, compared with RA mice, the mean rotation angle decreased significantly after ZnCM NP treatment, indicating the efficient alleviation of RA ankle inflammation. In summary, ZnCM NP treatment improved the pathological walking patterns of RA mice in terms of hind limb pressure, mean rotation angle and stance time course, thereby suggesting the alleviation of RA manifestations in vivo and indicating the potential clinical efficacy of ZnCM NPs for RA treatment.
Effects of the nanoparticles on osteolysis and inflammation in vivo
Bone destruction and cartilage damage have emerged as the key milestones during RA progression and indicate its severity and prognosis [65, 66]. Therefore, the diseased ankles of RA mice after NP injection were analysed by μCT (Fig. 6a). Quantitative results showed that RA decreased the Tb.N, BV/TV, and BMD compared with those of Ctrl mice, indicating significant bone destruction caused by RA. However, both ZnC and ZnCM NP treatments failed to protect bone destruction after 4 weeks of administration, as exemplified by the Tb.Th, Tb.N, BV/TV, and BMD data that hardly exhibited significant differences compared with those of RA mice. These results indicated that although ZnCM NPs could alleviate RA immune inflammation, they affected limited osteoclast-mediated osteolysis in vivo after 4 weeks of treatment. Thus, longer treatment and observation times may be needed.
Furthermore, HE and safranin O/fast green (SO/FG) staining (Fig. 6b) showed that RA mice had significant cartilage erosion and synovial inflammation compared with Ctrl mice. Both ZnC and ZnCM NPs not only protected against cartilage damage but also inhibited synovial inflammation, and the ZnCM NPs exhibited a more significant effect than ZnC NPs. IL-17 (signal 3), a crucial cytokine secreted from immune cells that reflects the local inflammatory response, was also observed in stained sections from the ankle and spleen. Figure 6c showed that in RA mice, the levels of IL-17 were elevated in both the ankle and spleen, which was attenuated by ZnCM NP treatment, indicating that ZnCM NPs significantly inhibited inflammatory activity in both the ankle and spleen. Collectively, these data from the RA model mice indicated that ZnCM NPs were capable of inhibiting cartilage erosion and inflammation in the ankle and spleen despite the insignificant effects on osteolysis after 4 weeks of treatment.
In vivo induction of tDCs by nanoparticles
Next, we aimed to examine the effects of ZnCM NPs on DC tolerogenicity in vivo. After mice were administered the appropriate treatments, CD11c+ DCs from the spleen were harvested for flow cytometry analysis. Figure 7a showed that RA mice developed a significant increase in the expression of CD86, CD80, MHC class I and MHC class II in their CD11c+ DCs compared with expression in Ctrl mice. However, treatment with ZnC NPs reduced the level of CD80, while ZnCM NP treatment decreased the expression of CD80, CD86, MHC class I and MHC class II (signals 1 and 2). This result was consistent with those from CD molecule staining in vivo (Fig. 7b), demonstrating that the number of CD80+ and CD86+ CD11c+ DCs increased significantly after RA establishment but decreased after ZnCM NP treatment, indicating the suppressed immunogenicity of DCs after Zn2+ and O2 supplementation in vivo. Furthermore, Fig. 7c showed that the ZnCM NPs significantly affected the staining of CD4+ T cells in the ankle synovium compared with that of RA mice, indicating that ZnCM NPs effectively suppressed the immune crosstalk between tDCs and T cells in vivo to impede the progression of RA-mediated inflammation. All these data showed that modified ZnO2-based NPs could switch igDCs towards tDCs in vivo, indicating that Zn2+ homeostasis controlled by ZnCM NPs effectively regulated RA in vivo (Fig. 8).