Preparation and Characterization of the siIRF1-Loaded Nanoparticles
To simulate the potential mechanisms driving the migration of immune cells to the myocardium, transcriptional sequencing analysis of the heart tissue was conducted. KEGG pathway analysis revealed significant enrichment in the cell adhesion molecules and the chemokine signaling pathway (Figure S1A). The heatmap of the identified chemokines predominantly highlighted the upregulation of CXCL9/10/11 (Figure S1B). Notably, CXCL9/10/11 are known to interact with the chemokine receptor CXCR3, which is primarily expressed on T lymphocytes[32–34]. Moreover, previous studies demonstrated that macrophages constituted the primary immune cell population in myocarditis[35]. Consequently, a hybrid membrane (HM) derived from RAW 264.7 macrophages and activated EL4 T lymphocytes (Figure S2A) was utilized to fabricate biomimetic nanocarriers for the targeted therapy of myocarditis.
The siIRF1@ZIF@HM nanoparticles were synthesized through a phased procedure according to previous studies with some modifications[22, 36]. Initially, the ZIF-8 nanoparticles (ZIF-8 NPs) were precipitated via the dropwise addition of a zinc ion solution to a vigorously stirred 2-methylimidazole (2-MIM) solution. The cyanine 5 NHS ester fluorescent dye (Cy5) labeled siIRF1 (referred to as siIRF1) was then mixed with the prepared ZIF-8 NPs in an aqueous medium and stirred overnight to produce the siIRF1@ZIF nanoparticles (siIRF1@ZIF NPs). Subsequently, the siIRF1@ZIF NPs were camouflaged with hybrid membrane vesicles (Fig. 2A) for further biomimetic modification. Transmission electron microscopy (TEM) showed a rhombic shape of siIRF1@ZIF NPs with an average size of 113.8 ± 1.5 nm (Fig. 2B), while an additional dim layer (17.71 ± 3.1 nm) was observed around the siIRF1@ZIF@HM NPs after HM coating (Fig. 2C). Further, the element mapping images confirmed the uniform distribution of C, N, O, Zn, and P elements within the nanoparticles (Fig. 2D). Dynamic light scattering (DLS) assessments verified the hydrodynamic size and zeta potential of the nanoparticles. The hydrodynamic sizes of ZIF-8, siIRF1@ZIF, and siIRF1@ZIF@HM NPs were approximately 115.7 nm, 120.1 nm, and 133.8 nm, respectively, consistent with the TEM results (Fig. 2E). Zeta potential analysis showed a decrease from 8.7 mV in ZIF-8 NPs to -21.8 mV after siIRF1 encapsulation, and a further change to -11.6 mV after HM modification (Fig. 2F). UV-vis absorbance spectra presented a siIRF1-specific absorption peak at approximately 650 nm due to Cy5 conjugation, which was not observed in pure ZIF-8 (Fig. 2G). These results confirmed the successful encapsulation of siIRF1 into ZIF-8, as well as the HM coating on the surface of siIRF1@ZIF NPs. The stability test in PBS buffer (pH 7.4) over 0, 1, and 7 days demonstrated the persistent hydrodynamic diameter of the siIRF1@ZIF@HM NPs, indicating their dispersion stability without significant aggregation (Figure S3).
To confirm the functionalization of siIRF1@ZIF with HM, western blot analysis revealed that the typical markers of macrophages (ITGA4 and ITGB1[37]) and T lymphocytes (TCR-β[38]) were transferred from cell membranes to the nanoparticles (Fig. 2H). Furthermore, SDS-PAGE analysis indicated that siIRF1@ZIF@HM NPs exhibited dual protein profiles of both T cells and macrophages (Fig. 2I). To ensure maximal expression of CXCR3 on T cells, PMA was utilized to activate the EL4 T lymphocytes prior to membrane isolation[39] (Figure S2B). The Zn2+ ion forms acid noncovalent bonds with 2-MIM, leading to pH-responsive biodegradation of the ZIF-8 nanoparticles[40]. As illustrated in Fig. 2J, the siIRF1@ZIF NPs demonstrated a pH-dependent release of siIRF1 into the acidic milieu (pH 5.5 and 6.0), while the physiological condition (pH 7.4) resulted in a minimal release of siIRF1. The loading efficiency of siIRF1 was calculated to be 99.8% according to the calibration curve shown in Figure S4, with a drug loading capacity (DLC) of 7.3%.
Cellular Uptake and Endo-Lysosomal Escape of siIRF1@ZIF@HM Nanoparticles
We have identified M1 macrophages as the target cell type due to their pathological and pyroptotic properties in myocarditis. Consequently, we sought to explore the targeting capability of siIRF1@ZIF@HM NPs towards M1 macrophages in vitro. The cytotoxicity of siIRF1@ZIF NPs on mouse macrophages was first determined by a CCK-8 assay. Cell viability remained stable at concentrations of siIRF1@ZIF NPs up to 40 µg/mL (Figure S5), indicating good biocompatibility of the nanoparticles. Thus, the nanoparticles were administered at concentrations not exceeding 40 µg/mL in the following experiments. After inducing M1 macrophage differentiation, M0 (unpolarized) and M1 macrophages were incubated with 40 µg/mL siIRF1@ZIF NPs or siIRF1@ZIF@HM NPs for 4 h. The Cy5 fluorescent dye enabled the monitoring of the cellular uptake process, whereas the immunofluorescence staining of CD68 and iNOS confirmed the macrophage differentiation status. As depicted in Fig. 3A-C, M0 macrophages treated with either type of nanoparticles displayed minimal mean fluorescence intensity (MFI), with negligible differences between the groups. In contrast, M1 macrophages exposed to siIRF1@ZIF NPs exhibited a relatively higher MFI compared to M0 macrophages, implying an enhanced endocytic activity in M1 macrophages. However, M1 macrophages treated with siIRF1@ZIF@HM NPs showed a significantly elevated MFI, in stark contrast to other groups, suggesting a preferential uptake of siIRF1@ZIF@HM NPs by M1 macrophages. These findings, consistent in both RAW 264.7 macrophages and bone marrow-derived macrophages (BMDMs), demonstrated the specific targeting capability of siIRF1@ZIF@HM NPs towards M1 macrophages, attributed to the selective uptake rather than a generalized increase in endocytosis. The M1 macrophages are commonly induced by IFN-γ stimulation, which has also been reported to promote the expressions of CXCL9, CXCL10, and CXCL11 in various cell types, including macrophages[23, 41]. In alignment with these findings, we verified a substantial enhancement in the expression levels of these chemokines in macrophages upon IFN-γ stimulation with western blot and qPCR (Fig. 3D-H). These results suggest that the enhanced targeting capability of the HM coating on nanoparticles could be attributed to the interaction between the CXCL9/10/11 chemokines and the CXCR3 receptor.
After confirming specific targeting to the M1 macrophages, we proceeded to investigate the subsequent release of siIRF1 from the endo-lysosomal compartments, a critical step for effective gene silencing by siRNA within the cytoplasm[42]. To visualize the dynamic escape of siIRF1, the M1 macrophages were incubated with 40 µg/mL siIRF1@ZIF@HM NPs for various time durations (2 h, 4 h, 12 h, and 24 h). Confocal microscopy was utilized to capture the fluorescent colocalization images of siIRF1 and lysosomes, labeled with the Cy5 and Lyso-Tracker Green, respectively. As shown in Fig. 3I and J, the signals of Cy5 and Lyso-Tracker were separated upon 2 h incubation, representing the initial stage of endocytosis. After 4 h incubation, a significant overlay of the Cy5 and Lyso-Tracker signals was observed in macrophages, reflecting the entry of siIRF1 into lysosomes. The subsequent spatial dispersion of the Cy5 signal from Lyso-Tracker at 12 h, which was more evident at 24 h, denoted the release of siIRF1 into the cytosol. The escaping process was quantified using mean Pearson's correlation coefficients, with values at 2 h, 4 h, 12 h, and 24 h calculated to be 0.22, 0.52, 0.39, and 0.13, respectively. Overall, these data indicated that the siIRF1 was successfully discharged into the cytosol as a result of the pH-responsive dissociation of the MOF scaffold, thereby facilitating the endo-lysosomal disruption[24].
Therapeutic Effects of siIRF1-Loaded Nanoparticles In Vitro
Our previous research has demonstrated M1 macrophage pyroptosis as the pivotal factor in the pathogenesis of myocarditis[12]. IFN-γ, a key inducer of M1 differentiation, is implicated in the lytic programmed cell death, including pyroptosis and PANoptosis, which may involve IFN-γ-driven upregulation of IRF1[43–45]. To validate the anti-pyroptotic effects of siIRF1 in vitro, we demonstrated that the siIRF1 effectively decreased the protein expression of IRF1 in IFN-γ treated J774A.1 macrophages via siRNA transfection (Figure S6). Subsequently, mRNA levels of IRF1 and its downstream genes, including Caspase-1, ZBP1, and GBP1, were determined by qPCR to assess the gene silencing capability of the siIRF1@ZIF NPs and siIRF1@ZIF@HM NPs. As a result, the administration of siIRF1@ZIF@HM NPs significantly reduced the mRNA expression of IRF1 and its downstream genes, whereas siIRF1@ZIF NPs showed less pronounced gene silencing ability due to lower uptake (Fig. 4A-D). These findings indicated a high gene silencing efficiency of the siIRF1 in this study. Subsequently, the anti-pyroptotic effects of the nanoparticles were determined in both J774A.1 macrophages and BMDMs subjected to pyroptosis induction via IFN-γ incubation followed by ATP stimulation. As shown in Fig. 4E-H, the siIRF1@ZIF@HM NPs significantly inhibited the expression of IRF1 upon IFN-γ stimulation, thereby suppressing the pyroptosis in both macrophages, as evidenced by the decreased levels of the N-terminal of GSDMD (GSDMD-NT) and Cleaved Caspase-1. Notably, both low and high concentrations (5 µg/mL and 10 µg/mL) displayed high therapeutic efficiency with no obvious discrepancy between the two groups. In contrast, the siIRF1@ZIF NPs exhibited relatively lower anti-pyroptotic capacity only at high concentration (10 µg/mL) in J774A.1 macrophages, with its anti-pyroptotic effects diminished in BMDMs. These findings demonstrate that targeting IRF1 in M1 macrophages with siIRF1@ZIF@HM NPs is an effective approach to inhibit IFN-γ-mediated macrophage pyroptosis in vitro.
In Vivo Pharmacokinetics and Targeting Profiles of siIRF1-Loaded Nanoparticles
Our in vitro findings confirmed that siIRF1@ZIF@HM NPs are effective in targeting pro-inflammatory M1 macrophages and attenuating the IFN-γ-induced pyroptosis. To extend these findings to an in vivo context, we assessed the circulation time of siIRF1@ZIF NPs and siIRF1@ZIF@HM NPs after intravenous injection. Serial blood samples were collected at designated intervals, and the pharmacokinetic evaluation was performed by assessing the fluorescence intensity (Fig. 5A). Notably, the clearance rates of siIRF1@ZIF@HM NPs showed no significant difference compared to those of siIRF1@ZIF NPs. Subsequently, the experimental autoimmune myocarditis (EAM) mouse model was employed to evaluate the targeting efficacy of the nanoparticles toward inflamed myocardium and their overall biodistribution. Consistent with the results of transcriptional sequencing, a marked upregulation in the mRNA expression of chemokines CXCL9, CXCL10, and CXCL11 was observed in EAM cardiac tissues (Fig. 5Β). Following intravenous administration of PBS, siIRF1@ZIF NPs, and siIRF1@ZIF@HM NPs to the EAM mice, major organs (heart, liver, spleen, lung, and kidney) were harvested at 4 h, 12 h, and 24 h post-injection for analysis. Ex vivo imaging revealed pronounced cardiac fluorescence intensity in the group treated with siIRF1@ZIF@HM NPs, observed at both gross and sectional levels at all time points (Figs. 5C-E). Quantitative biodistribution assessments of the nanoparticles in major organs are presented in Figure S7A-C. Intriguingly, both nanoparticles predominantly accumulated in the kidneys rather than the liver. Specifically, kidneys from mice receiving siIRF1@ZIF NPs showed a peak in fluorescence intensity at 4 h which diminished over time. In contrast, the siIRF1@ZIF@HM NPs group exhibited a relatively stable renal fluorescence, implying initial preferential cardiac retention followed by renal elimination. Immunofluorescence staining was also employed to evaluate the targeted delivery of the nanoparticles to M1 macrophages. Fluorescence imaging of cardiac cryosections revealed that siIRF1@ZIF@HM NPs were able to accumulate within M1 macrophages marked by CD68 and iNOS expression. In comparison, the siIRF1@ZIF NPs group displayed a less intense and non-specific signal (Fig. 5F). Furthermore, the fluorescence of Cy5-labeled siIRF1 was also observed in other organs with confocal microscopy (Figure S7D), the results of which coincided with the ex vivo imaging. These data indicate that the HM modification substantially enhances the targeting precision of siIRF1@ZIF@HM NPs towards the inflammatory myocardium, particularly favoring M1 macrophages in vivo.
Therapeutic Efficacy and Biosafety of siIRF1 Nanotherapies in EAM Mice
Following the demonstration of targeted in vivo delivery, the therapeutic effects of the nanoparticles against pyroptosis and myocarditis were evaluated. The EAM mice were randomly allocated into five groups. On day 7 post EAM induction, each group received intravenous injections of either PBS, free siIRF1, siIRF1@ZIF NPs, and siIRF1@ZIF@HM NPs, at a dosage of 0.73 mg/kg siIRF1 per administration for five times (Figure S8). On day 21, corresponding to the acute phase of EAM, major organs were harvested for analysis. Compared with the control group, hematoxylin and eosin (HE) staining of myocardial sections revealed extensive inflammation in the EAM group, characterized by leukocyte infiltration exceeding 50% of the cardiac section (Fig. 6A-C). The groups treated with siIRF1 and siIRF1@ZIF NPs exhibited a downward trend in inflammation scores and ratios, albeit without significant statistical differences. In contrast, the administration of siIRF1@ZIF@HM NPs substantially reduced cardiac inflammation. Additionally, EAM induction notably increased the serum level of CK-MB, which was effectively suppressed by siIRF1@ZIF@HM NPs treatment, thereby alleviating the myocardial injury (Fig. 6D). Furthermore, the EAM group showed significant upregulation in the transcriptional levels of IRF1 and pro-inflammatory cytokines, namely IL-1β, IL-6, and TNF-α, primarily produced by M1 polarized macrophages[46]. Consistent with the histological findings, the expression levels of IRF1 and pro-inflammatory cytokines in the siIRF1@ZIF NPs group showed a decreasing trend, while the siIRF1@ZIF@HM NPs administration significantly downregulated the mRNA expression of these cytokines and IRF1 (Fig. 6E-H). These findings demonstrated that siIRF1@ZIF@HM NPs effectively silenced cardiac IRF1 expression, thereby mitigating the inflammatory responses in myocarditis.
To elucidate the therapeutic mechanisms, immunofluorescence staining and western blot analysis were performed. The cardiac cryosections from each group were stained with CD68 and iNOS antibodies to assess the polarization status of macrophages within myocarditis. As depicted in Fig. 6I and J, the EAM and siIRF1 groups showed substantial colocalization of CD68 and iNOS at the inflammatory loci. In contrast, macrophages from the siIRF1@ZIF NPs group exhibited decreased double-positive staining, while the iNOS signal was notably diminished in the siIRF1@ZIF@HM NPs group, indicating an inhibited macrophage polarization. In addition, western blot analysis revealed that the pyroptosis-associated proteins were significantly activated in the EAM group, as evidenced by the cleavage of GSDMD-FL, Pro-Caspase-1, and Pro-IL-1β (Fig. 6K-O). The administration of siIRF1 and siIRF1@ZIF NPs had no obvious effect on pyroptosis. However, siIRF1@ZIF@HM NPs treatment suppressed the cleavage of the GSDMD-FL, Pro-Caspase-1, and Pro-IL-1β, thereby impeding the pyroptotic cascade. Notably, the expression levels of IRF1 were in line with the intensity of pyroptosis, suggesting the underlying mechanisms of pyroptosis in myocarditis. Collectively, these data demonstrated that siIRF1@ZIF@HM NPs downregulated IRF1 expression, suppressed macrophage polarization and pyroptosis, and consequently alleviated the progression of myocarditis.
To evaluate the safety profiles of the nano-formulations, the body weights of mice within each group were monitored throughout the therapy. As illustrated in Figure S9A, no significant fluctuations in body weight were observed. Furthermore, biochemical analysis of the blood samples indicated that the serum levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and creatinine (CREA) remained stable in comparison with the control group (Figure S9 B-D), suggesting unimpaired hepatic and renal function. Concurrently, analysis of the serum levels of total cholesterol (TC), triglyceride (TG), and creatine kinase (CK) demonstrated normal values comparable to those of the control group (Figure S9E-G), indicating metabolic stability. Histological assessment of the major organs, including liver, spleen, lung, and kidney, revealed no significant tissue abnormalities with HE staining (Figure S10), thus confirming the biocompatibility of the nanoparticles.