Abstract
Disruption in the dynamic stability of macrophage pro/anti-inflammatory phenotypes within plaques significantly impacts chronic vascular inflammation and exacerbates atherosclerosis. Reprogramming macrophages from pro-inflammatory to anti-inflammatory phenotype mitigates atherosclerosis progression. However, chronic inflammatory stimulation induces a closed chromatin state in atherosclerotic macrophages, inhibiting their phenotype reprogramming. This study demonstrates that restoring mitochondrial respiration increases histone acetylation (AcH3) and enhances chromatin accessibility in atherosclerotic macrophages, restarting macrophage phenotype reprogramming. Additionally, we identified that miR-10a can facilitate mitochondrial respiration and reorganize macrophage reprogramming. To optimize delivery, prolong circulation time, and target pro-inflammatory macrophages, we developed red blood cell membrane-modified liposome nanoparticles (miR-10a@H-MNP) to deliver miR-10a. Hyaluronic acid was conjugated to the liposomes to specifically target pro-inflammatory macrophages in plaque. Intravenous administration of miR-10a@H-MNP significantly alleviated atherosclerosis progression in male mice. Thus, the epigenomic priming approach developed here effectively triggers macrophage reprogramming in atherosclerosis, presenting a promising metabolically based epigenetic modulation method for plaque clearance.
Introduction
Atherosclerosis is a leading cause of death worldwide. It is characterized by the accumulation of lipids, immune cells, and inflammatory cytokines in the vascular intima1,2. Increasing evidence suggests that macrophages retain memories of past infections and make them more active when trained by endogenous atherogenic factors (such as oxidized lipids, cholesterol crystals, and inflammatory mediators), thereby contributing to the progression of atherosclerosis3,4. Evidence has established that epigenetic reprogramming at the level of chromatin structure, occurring primarily through histone modifications, is a core mechanism supporting the enhanced functional state of well-trained innate immune cells5,6. Previous studies have demonstrated that lipopolysaccharide cannot induce an inflammatory phenotype in histone deacetylase 3 (HDAC3)-deficient macrophages7, and the loss of HDAC3 leads to macrophages skewing towards anti-inflammation phenotype8. Therefore, the reason for the current inefficiency of macrophage phenotypic reprogramming-based treatment of atherosclerosis may be the epigenetic silencing of macrophages caused by sustained hyper-inflammatory training.
Mitochondrial metabolic reprogramming and epigenetic remodeling are closely intertwined and mutually regulate each other. Acetyl-CoA (Ac-CoA), primarily derived from glucose-produced pyruvate in mitochondria, is the primary donor of acetyl groups in eukaryotic cell histone acetylation modifications9. In this process, acetyl groups donated by Ac-CoA are added to lysine residues of histones by histone acetyltransferases (HATs), enhancing transcription factor accessibility. Conversely, HDACs diminish transcription factor accessibility by eliminating histone lysine acetylation and suppressing gene expression through chromatin condensation. Previous research has indicated that M2 phenotype macrophages rely on mitochondrial oxidative phosphorylation (OXPHOS) and fatty acid oxidation (FAO) pathways, whereas M1 phenotype macrophages predominantly depend on glycolysis10. Impairment of OXPHOS functionality results in reduced Ac-CoA supply, potentially exacerbating the constraints of macrophage epigenetic remodeling caused by HDACs11,12. Thus, reprogramming mitochondrial metabolism may overcome the constraints of HDACs-mediated epigenetic silencing of inflammatory macrophage and switching the phenotype to an M2 phenotype, potentially treating atherosclerosis.
The progression of atherosclerosis is hemodynamically regulated, with stable laminar flow exerts atheroprotective effects, whereas low shear stress or turbulent flow contributes to disease development13,14. Emerging evidence indicates that alterations in flow conditions modulate the expression of miRNAs both in vitro and in vivo, and that these flow-sensitive miRNAs, termed mechano-miRNAs, play a crucial role in the regulation of atherosclerosis15,16. Notably, Fang et al. first reported the beneficial role of atheroprotective flow-sensitive miR-10a in cardiovascular disease17. Their study found that pro-atherogenic human atherosclerotic lesions exhibited a reduced abundance of miR-10a. In contrast, overexpression of miR-10a in human aortic endothelial cells inhibited the expression of pro-inflammatory factors vascular cell-adhesion molecule 1 (VCAM-1) and E-selectin17. Furthermore, previous studies have demonstrated that miR-10a can inhibit monocyte activation and restore the impaired mitochondrial oxidative metabolic function of M1 phenotype macrophages, thereby reducing inflammation18,19. These findings suggest that miR-10a has the potential to regulate macrophage mitochondrial energy metabolism and macrophage-mediated atherosclerosis. However, the systemic delivery of miRNAs faces challenges, including rapid clearance and poor targeting of atherosclerotic plaques20,21. Recent advances in biomimetic nanocarriers, including modifications with platelet, red blood cell (RBC), and macrophage membranes for immune evasion, offer promising solutions to these challenges22,23,24. For instance, RBC membrane-coated PLGA nanoparticles, as reported by Wang et al., effectively evade macrophage phagocytosis and preferentially accumulate within atherosclerotic plaques23.
In this study, we address these challenges by engineering reactive oxygen species (ROS)-responsive nanoparticles modified with RBC and hyaluronic acid (HA) (designated as miR-10a@H-MNP) (Fig. 1). These nanoparticles: (1) evade reticuloendothelial system (RES) clearance via biomimetic coating, (2) target activated macrophages via HA binding, and (3) release miR-10a in ROS-rich plaques to reprogram macrophages phenotype and disrupt the immune training by restoring mitochondrial metabolic function and reshaping epigenetics25,26. Our results demonstrated that the miR-10a@H-MNP effectively targeted atherosclerotic macrophages upon intravenous administration and reprogrammed them to an M2 phenotype in vivo. This study presents a therapeutic strategy for halting the progression of atherosclerosis through precise regulation of mitochondrial metabolism, with immense potential for clinical applications.
Figure 1 was created in BioRender. Fang, F. (2025) https://BioRender.com/u6enwo5.
Results
Epigenetic silencing and mitochondrial dysfunction of M1 phenotype macrophage in atherosclerotic plaque
Macrophages in atherosclerotic plaques exhibit immune-trained properties, which hinder the dynamic transition of macrophage phenotypes and cause them to remain in the M1 phenotype state (Fig. 2A). Epigenetic modifications play a crucial role in regulating macrophage phenotypes. To investigate which epigenetic modifications prime the immune-trained macrophage phenotype blockade induced in atherosclerotic plaques, we analyzed human coronary artery plaque scRNA sequencing dataset GSE184073 in this study27. As shown in Fig. 2B, six CD45+ clusters were identified in plaques, including CD4+ T cells, CD8+ T cells, natural killer (NK) cells, B cells, and macrophages. Further analysis of the macrophage clusters revealed that the quantity of M1 phenotype macrophages in plaques exceeds that of the M2 phenotype. Given the important role of HDACs modifications in regulating macrophage phenotypic polarization in atherosclerosis28, we detected the expression of HDACs in two macrophage clusters. As shown in Fig. 2C, the gene expression levels of HDAC1, HDAC2, and HDAC3 were significantly higher in M1 phenotype macrophages compared to M2 phenotype macrophages. Furthermore, our examination of atherosclerotic samples from patients confirmed lower histone acetylation levels (H3K9Ac) compared to the control group, specifically within M1 phenotype macrophages present in plaques (Fig. 2D). These results demonstrate that histone acetylation plays a significant role in the conservation of atherosclerotic-induced macrophages.
a Schematic representation of crosstalk between macrophage epigenetics and mitochondrial function in atherosclerosis. b Uniform manifold approximation and projection (UMAP) plot of scRNA sequencing showed the six identified CD45+ cell types from human atherosclerotic plaques, along with the UMAP plot of two macrophage subpopulations distinguished by different transcriptional profiles. c Violin plots depicting the differential expression of HDAC1, HDAC2, and HDAC3 in M1 and M2 phenotype macrophage clusters within human atherosclerotic plaques. Data analyzed by two-tailed Wilcoxon rank-sum test (Mann-Whitney U test). d Representative immunofluorescence image showing H3K9Ac (red) and M1 phenotype macrophage (CD86, green) in human atherosclerotic plaque samples, three times this experiment was repeated independently with similar results. scale bar: 20 μm. e Gene expression heatmap of mitochondrial function-related genes for M1 and M2 phenotype macrophage subpopulations. Yellow indicates high expression; pink indicates low expression. f Representative immunofluorescence imaging of M1 phenotype macrophage (CD86+) co-stained for iNOS and TFAM in the healthy and atherosclerotic human aorta samples, three times this experiment was repeated independently with similar results. scale bar: 20 μm. Source data are provided as a Source Data file. Panel (a) was created in BioRender. Fang, F. (2025) https://BioRender.com/cf3cxnw.
Mitochondrial metabolic reprogramming and histone acetylation modifications are closely intertwined and mutually regulate one another9. To investigate how mitochondrial function correlated with histone acetylation regulation, we analyzed differential gene expression changes related to the mitochondrial function in two types of macrophage clusters. As shown in Fig. 2E, genes related to mitochondrial energy metabolism and biogenesis, such as MT-CO1, MT-CO2, MT-ND3, MT-CYB, MT-ATP6, etc., were downregulated in the M1 phenotype macrophage cluster. The violin plot also confirmed that, compared with M2 phenotype macrophages, the expression of key genes MT-CYB, MT-ND3 and MT-CO3 associated with mitochondrial energy metabolism and biogenesis was significantly reduced in M1 phenotype macrophages (Fig. S1). Additionally, we found that the inducible nitric oxide synthase (iNOS) protein expression in human and mouse atherosclerotic plaque macrophages is significantly higher in the M1 phenotype macrophages compared to the control group (Fig. 2F and Fig. S2A). We also examined the relationship between M1 phenotype macrophages and mitochondrial function in human and mouse atherosclerotic plaques. As shown in Fig. 2F and Fig. S2B, the expression of mitochondrial transcription factor A (TFAM, a transcription factor for mitochondrial DNA) was significantly decreased. These findings suggest an increase in M1 phenotype macrophage within atherosclerotic plaques, accompanied by epigenetic silencing and mitochondrial dysfunction.
miR-10a reprogramming macrophage phenotype by rescuing mitochondrial function and reshaping epigenetics
Given the critical role of miRNAs in inflammation and immune response, particularly in macrophage polarization29,30, we analyzed an RNA-seq dataset of differentially expressed miRNAs in inflammatory macrophages. As shown in Fig. S3, inflammatory macrophages significantly upregulated 30 miRNAs and downregulated 31 miRNAs, respectively. A literature review of the 31 significantly downregulated miRNAs revealed that 9 were associated with macrophage repolarization, and only miR-10a-5p was linked to macrophage mitochondrial metabolic remodeling17,18,31 (Fig. S4). Therefore, we investigated the effect of miR-10a on mitochondrial biogenesis and mitochondrial energy metabolism in macrophages to determine whether miR-10a affects mitochondrial function in the inflammatory macrophages (Fig. 3A). Seahorse mitochondrial stress tests were conducted to evaluate mitochondrial respiration. Compared to the LPS-induced inflammatory M1-type macrophage group and the miR-10a negative control (miRNC) transfection group, miR-10a transfection significantly enhanced basal respiration, ATP production, maximal respiration, and spare respiratory capacity (Fig. 3B-F). Additionally, MitoTracker Deep Red staining was performed to assess the effect of miR-10a on mitochondrial respiration in inflammatory macrophages. As shown in Fig. S5A, B, miR-10a transfection notably enhanced mitochondrial activity compared to the untreated M1-type macrophage group. Next, flow cytometry analysis confirmed that inflammatory M1-type macrophages exhibited decreased respiring mitochondria and an increased non-respiring mitochondrion, while miR-10a transfection significantly increased mitochondrial respiration in macrophages (Fig. 3G, H and Fig. S6). Finally, we found that miR-10a transfection significantly increased the expression of genes related to mitochondrial biogenesis, such as Sirt1, Ppargc1b, FAO-related genes Acox1, Acox3, and OXPHOS-related genes Idh3a, Ogdh (Fig. S7). These results confirmed that miR-10a can restore the mitochondrial energy metabolism of M1 phenotype macrophages.
a Schematic representation of the crosstalk between epigenetic changes and mitochondrial function mediated by miR-10a in macrophages. b Oxygen consumption rate (OCR) profile of M1 phenotype macrophages after miR-10a transfection. Data are presented as the mean ± SD (n = 3 biological replicates). c-f Basal respiration, ATP production, maximal respiration, and spare respiratory capacity of M1 phenotype macrophages following miR-10a transfection. Data are presented as the mean ± SD (n = 3 biological replicates). g, h Flow cytometry was used to analyze the ratio of respiratory and non-respiratory mitochondria to total mitochondria in M1 phenotype macrophages following miR-10a transfection. Data are shown as the mean ± SD (n = 3 biological replicates). i Immunofluorescence images of H3K9Ac (red) and F-action (green), with quantitative analysis of their average fluorescence intensity (j), in M1 phenotype macrophages after miR-10a transfection or treatment with the FAO inhibitor etomoxir, scale bar: 20 μm. Panel (j) data are presented as the mean ± SD (n = 20 fields from 5 independent samples). Contour plots and quantification of flow cytometry analysis showing the numbers of CD86-positive (k) and CD206-positive (l) cells following different treatments. Data are shown as the mean ± SD (n = 3 biological replicates). P values in (c-h and k-l) were determined by One-way ANOVA analysis followed by Tukey’s multiple comparisons test. P values in (j) were determined by One-way ANOVA analysis followed by Dunnett’s T3 multiple comparisons test. Source data are provided as a Source Data file. Panel (a) was created in BioRender. Fang, F. (2025) https://BioRender.com/4u7xkq0.
Subsequently, we investigated the effect of miR-10a on histone acetylation levels in M1 phenotype macrophages. As shown in Fig. 3I, J, the expression of H3K9Ac in the M1 phenotype macrophages was significantly lower than in the M0 phenotype macrophage. However, miR-10a transfection restores H3K9Ac expression in macrophages (Fig. 3I, J). To validate the association between mitochondrial function and histone acetylation, we assessed the H3K9Ac expression level in M1 phenotype macrophages treated with etomoxir, a mitochondrial FAO inhibitor. As shown in Fig. 3I, J, etomoxir treatment abolished the upregulation of H3K9Ac expression in M1 phenotype macrophages induced by miR-10a. Furthermore, the results of western blot also confirmed that the expression level of H3K9Ac in M1 phenotype macrophages significantly decreased compared with that in M0 phenotype macrophages, while miR-10a transfection significantly increased the expression level of H3K9Ac (Fig. S8A, B). These results indicate that miR-10a increases histone acetylation levels and alters the epigenetic state of macrophages.
To determine whether miR-10a can reprogram macrophage phenotype via regulating histone acetylation, we examined its effect on the expression of M1 and M2 phenotype markers in the presence or absence of a histone acetylation inhibitor. As shown in Fig. 3K, L, M1 phenotype macrophages exhibited elevated CD86 (M1 phenotype marker) expression and minimal CD206 (M2 phenotype marker) expression. Quantitative analysis confirmed that miR-10a significantly upregulated the expression of CD206 and downregulated CD86 expression (Fig. 3K, L). Furthermore, the macrophage repolarization mediated by miR-10a transfection was inhibited by anacardic acid (AA, inhibitor of HATs) treatment and enhanced by valproic acid (VA, inhibitor of HDACs) treatment. These results suggest that miR-10a can drive the repolarization of the M1 phenotype to the M2 phenotype, a process that can be enhanced by histone acetylation or attenuated by histone deacetylation.
Preparation and characterization of miR-10a-loaded biomimetic liposome
To overcome the off-target effect and enhance the therapeutic efficacy of miR-10a, we have designed and fabricated a multifunctional miR-10a@H-MNP biomimetic nano-system for anti-atherosclerotic treatment. Firstly, ROS-responsive liposomes encapsulating miR-10a were fused with RBC membranes to obtain miR-10a@MNP (Fig. 4A). Subsequently, tris(2-carboxyethyl) phosphine (TCEP) was used to reduce the disulfide bonds of RBC membrane to sulfhydryl groups, and HA-PEG-MAL was conjugated to these sulfhydryl groups to obtain miR-10a@H-MNP (Fig. 4A)32,33,34. As shown in Fig. 4B, C, the particle sizes of NP (DSPE-TK-PEG-based liposome without RBC membrane and HA-PEG modification) and H-MNP are 140.7 ± 2.5 nm and 145 ± 4.2 nm, respectively. Transmission electron microscopy (TEM) images showed that NP and H-MNP exhibited roughly spherical morphology (Fig. 4B, C). In addition, the zeta potential of NP and H-MNP were −12.4 ± 0.8 mV and −14.2 ± 0.9 mV, respectively (Fig. 4D). To validate the co-localization of the RBC membrane and liposome, we labeled the RBC membrane and the liposome loaded with cy5-miR-10a using a green fluorescent dye, Dio. After extruding via a 200 nm film, the RBC membrane-modified liposome was incubated with inflammatory macrophages. As shown in Fig. 4E, the RBC membrane and liposome were co-localized, with a Pearson co-localization coefficient of 0.77 (Fig. S9). Additionally, Coomassie brilliant blue staining showed a close protein distribution between the H-MNP and RBC membrane (Fig. 4F), and western blot results further confirmed that H-MNP carries the surface markers CD47 and Ter-119 of RBCs (Fig. 4G). These results indicate that the RBC membrane successfully fused with miR-10a-loaded liposomes.
a Schematic diagram illustrating the design and preparation of miR-10a@H-MNP. b, c TEM and NTA were used to observe the morphology and size distribution of NP and H-MNP, respectively. These experiments were repeated three times and similar results were obtained. d Zeta potential measurements of NP and H-MNP. Data are shown as the mean ± SD (n = 3 biological replicates). e Co-localization of RBC membrane and cy5-miR-10a@H-MNP, scale bar: 20 μm. This experiment was repeated independently three times with similar results. f Coomassie brilliant blue staining showing the protein profile of the RBC membrane and H-MNP (RBC-M denotes RBC membrane). This experiment was repeated independently three times with similar results. g Western blot analysis detecting the expression of CD47 and Ter-119 in RBC membrane and H-MNP. This experiment was repeated independently three times with similar results. h miR-10a encapsulation efficiency of H-MNP. Data are shown as the mean ± SD (n = 3 biological replicates). i Stability of miR-10a@H-MNP against RNase A. j Release profile of cy5-miR-10a from H-MNP. Data are presented as the mean ± SD (n = 3 biological replicates). k Fluorescence images and mean fluorescence intensity of cy5-miR-10a@H-MNP or cy5-miR-10a@NP incubated with M1 phenotype macrophage with different times, scare bar: 20 μm. Data are shown as the mean ± SD (n = 5 biologically replicates). l Flow cytometry of cy5+ cells in M1 macrophages after incubation with cy5-miR-10a@H-MNP or cy5-miR-10a@NP. m IVIS imaging and quantification of radiation efficiency (n) in atherosclerotic mouse aortas post-intravenous injection of cy5-miR-10a@NP or cy5-miR-10a@H-MNP. Panel (n) data are shown as the mean ± SD (n = 3 independent samples). o Immunofluorescence staining image of aorta sections from atherosclerotic mice, 6 h post intravenous injection of cy5-miR-10a@NP (red) or cy5-miR-10a@H-MNP (red), with CD86 labeled macrophages (green) and DAPI labeled nuclei, scale bar: 20 μm. This experiment was repeated independently three times with similar results. Source data are provided as a Source Data file. P values in (d, h, k, and n) were determined by two-tailed unpaired Student t test. Panel (a) was created in BioRender. Fang, F. (2025) https://BioRender.com/u6enwo5.
To assess miR-10a loading efficiency, we measured the fluorescence intensity of cy5-labeled miR-10a in H-MNP according to the previously described protocol35. As shown in Fig. 4H, the miR-10a loading efficiency in NP and H-MNP was 74.03 ± 1.6% and 74.33 ± 2.6%, respectively, with no significant difference between the two groups. We then evaluated the stability of miR-10a in H-MNP according to the previous36. H-MNP demonstrated robust protection against miR-10a degradation in RNase A-containing environments for up to 8 h (Fig. 4I). It is well known that H2O2 hydrolyzes thioketal bonds. Therefore, in the presence of H2O2, DSPE-TK-PEG, the essential component of the liposomes is destroyed37,38. To validate the ROS responsiveness of H-MNP, we assessed the release profile of cy5-miR-10a at 37 °C in different media, including 0 mM H2O2 and 1 mM H2O2. The H2O2 concentrations used in this study were determined based on previous reports, which mimicked the in vitro H2O2 levels in atherosclerotic lesions24,39. In vitro release data showed that about 20% of cy5-miR-10a was released within 24 h in a PBS environment; however, the release ratio significantly increased upon exposure to H2O2, reaching an impressive value of 73.3 ± 2.1% over 24 h (Fig. 4J).
Cellular uptake by macrophages in vitro
Initially, a CCK-8 assay was conducted to evaluate the safety of blank H-MNP. No cytotoxicity was observed in both RAW264.7 cells at 500 μg/mL blank H-MNP for 24 h (Fig. S10). A major challenge in nanoparticle drug delivery is clearance by the mononuclear phagocyte system before reaching the lesion site20,21. Therefore, we first investigated the uptake of RBC membrane-mimetic cy5-miR-10a@H-MNP by macrophages. As shown in Fig. S11, distinct red fluorescence was observed in RAW264.7 cells after incubation with cy5-miR-10a@NP for 8 h. However, the red fluorescence intensity in RAW264.7 cells incubated with cy5-miR-10a@H-MNP modified with RBC membranes was markedly low, significantly less than that observed in the cy5-miR-10a@NP treatment group (Fig. S11). These results suggest that RBC membranes modified H-MNP provide an effective strategy to evade macrophage clearance. We next assessed the ability of the HA-modified H-MNP to target LPS-induced M1 phenotype macrophages. As shown in Fig. 4K, the internalization of cy5-miR-10a@H-MNP by M1 phenotype macrophages following LPS stimulation is time-dependent. Furthermore, we found that the internalization rate of cy5-miR-10a@H-MNP was significantly higher compared to the cy5-miR-10a@NP group over the same period (Fig. 4K). Flow cytometry further confirmed enhanced cellular internalization of cy5-miR-10a@H-MNP in M1 phenotype macrophages compared to the cy5-miR-10a@NP group (Fig. S12 and Fig. 4L). These findings suggest that the HA modification of H-MNP enhances its active targeting capability towards M1 phenotype macrophages.
To verify the contribution of HA modification to the targeting ability of H-MNP towards M1 phenotype macrophages, M1 phenotype macrophages were pre-incubated with free HA and then exposed to cy5-miR-10a@H-MNP for 8 h. As shown in Fig. S13, the uptake of cy5-miR-10a@H-MNP by M1 phenotype macrophages pre-incubated with free HA was significantly reduced than that by M1 phenotype macrophages not incubated with HA. Once the nanoparticles circulate in the bloodstream, their effective accumulation in macrophages of lesions relies on extravasation through the leaky blood vessels in atherosclerotic lesions. To assess this, we established a Transwell model to simulate the process according to the previous1. HUVECs were seeded in the upper chamber of the Transwell, and macrophages were seeded in the lower chamber. Once the cells grew and coalesced, LPS was added for 24 h to induce inflammation in both endothelial cells and macrophages. Subsequently, cy5-miR-10a@NP and cy5-miR-10a@H-MNP were added and incubated for 8 h. As shown in Fig. S14A-C, in contrast to the group without LPS treatment, cy5-miR-10a@H-MNP was able to traverse the endothelial layer and be internalized by inflammatory macrophages in the lower chamber. Collectively, these results validate the active targeting potential of H-MNP towards M1 phenotype macrophages in vitro.
In vivo assessment of H-MNP blood retention and target ability
Before conducting in vivo experiments, we conducted a preliminary assessment of the biocompatibility of H-MNP. Blood compatibility results showed that even at a concentration as high as 500 μg/mL, the hemolysis rate of H-MNP remained below 5%, indicating excellent blood compatibility (Fig. S15). Next, we investigated the in vivo pharmacokinetics of cy5-miR-10a@H-MNP. Blood samples were collected from each group at designated time points, and fluorescence intensity was detected using an in vivo fluorescence imaging system (IVIS). The cy5-miR-10a@H-MNP treatment group exhibited a significantly prolonged blood circulation half-life (t1/2 ≈ 6.2 h) compared to cy5-miR-10a@NP (t1/2 ≈ 3.8 h) over 48 h. Additionally, approximately 10.45% of cy5-miR-10a@H-MNP remained in the blood 48 h post-injection, whereas cy5-miR-10a@NP, lacking HA and RBC membrane modification, was almost eliminated after 48 h (Fig. S16A and B).
Additionally, the in vivo targeting capability and tissue biodistribution of cy5-miR-10a@H-MNP in ApoE−/− mice with atherosclerosis plaques was evaluated. 6 h after the injection of cy5-miR-10a@NP or cy5-miR-10a@H-MNP into atherosclerotic mice via tail vein, the aorta and major organs were sacrificed and analyzed by IVIS. As shown in Fig. 4M, N, the fluorescence intensity in the aortas of the cy5-miR-10a@H-MNP treated group was significantly higher than in the cy5-miR-10a@NP treated group. Furthermore, 6 h after injection, the fluorescence intensity in the liver of the cy5-miR-10a@H-MNP treated group was notably lower than that in the cy5-miR-10a@NP treated group (Fig. S17). It is well-established that a key pathological feature of atherosclerosis involves aberrantly activated macrophages, which exhibit high expression of receptors such as CD4440. The HA modification on the surface of H-MNP acts as a natural ligand for the CD44 receptor40. Therefore, we performed frozen sections of aortic tissue and immunofluorescence staining to investigate the cellular localization of H-MNP within plaques. The results revealed enhanced fluorescence signals within plaques in the cy5-miR-10a@H-MNP-treated group compared to the cy5-miR-10a@NP group (Fig. 4O). Additionally, it was observed that cy5-miR-10a@H-MNP co-localized with M1 phenotype macrophages within the plaque (CD86 marker, green) (Fig. 4O). In conclusion, our findings demonstrate that RBC membrane-modified H-MNP exhibits prolonged circulation in the bloodstream and actively targets and accumulates macrophages in atherosclerosis lesions. The functional HA and RBC membranes also enable the nanoparticles to evade phagocytosis by the RES, thereby minimizing non-specific toxicity and side effects.
In vitro assessment of miR-10a@H-MNP mediated macrophage repolarization
Atherosclerosis is a chronic inflammatory disease in which inflammatory macrophages play a critical role in its progression40,41,42. Thus, LPS-induced macrophage inflammation was chosen as an in vitro model to assess the therapeutic effect of miR-10a@H-MNP. Initially, we evaluated the effect of miR-10a@H-MNP on the repolarization of M1 phenotype macrophages. As shown in Fig. S18A-C, incubation of miR-10a@H-MNP with M1 phenotype macrophages for 24 h significantly reduced the fluorescence intensity of the M1-type marker CD86 and increased the fluorescence intensity of the M2 phenotype marker CD206. Moreover, qRT-PCR analysis revealed that treatment with miR-10a@H-MNP significantly downregulated the expression of M1 phenotype marker-related genes IL-1β, TNF-α, and NOS2 (Fig. S19A) while significantly upregulating the expression of M2 phenotype related genes Arg1 and Mrc1 (Fig. S19B). Next, we examined the effect of miR-10a@H-MNP on macrophage oxLDL uptake by BODIPY staining. As shown in Fig. S20, the red fluorescence intensity in macrophages treated with miR-10a@H-MNP was significantly lower than that in the blank control, miRNC@H-MNP, and miR-10a@NP groups. Flow cytometry analysis further confirmed a significant reduction in the content of neutral lipid droplets in macrophages across all treatment groups, with the miR-10a@H-MNP treatment group exhibiting the most pronounced inhibitory effect (Fig. S21). Additionally, dihydroethidium (DHE) and ROS staining were used to evaluate the ROS levels in macrophages after miR-10a@H-MNP treatment. As shown in Fig. S22 and Fig. S23, the miR-10a@H-MNP treatment significantly reduced ROS levels in macrophages. These results indicate that miR-10a@H-MNP can reprogram M1 phenotype macrophages into M2 phenotype while reducing the oxLDL uptake and intracellular ROS levels.
Anti-atherosclerotic effects of miR-10a@H-MNP in vivo
Given the promising in vitro efficacy and in vivo results of targeting atherosclerotic plaques, we further investigated the therapeutic effects of miR-10a@H-MNP in vivo. As shown in Fig. 5A, ApoE−/− mice were fed an 8-week high-fat diet and then administered therapeutic agents, including miRNC@H-MNP, atorvastatin, miR-10a@NP, and miR-10a@H-MNP twice weekly for 4 weeks. Throughout the dosing period, no significant difference in the average body weight was observed among the treatment group (Fig. 5B). Three days after the final injection, the mice were sacrificed, and their entire aortas were collected. Oil red O (ORO) staining was performed on aorta samples from different treatment groups to assess the content of neutral lipid droplets. As shown in Fig. 5C and Fig. S24, the ORO en face staining results showed that the atorvastatin, miR-10a@NP, and miR-10a@H-MNP administration can reduce the aortic neutral lipid deposition. These findings support the potential anti-atherosclerotic effects of miR-10a and suggest its viability as a novel clinical drug candidate. Notably, the aorta from mice treated with miR-10a@H-MNP exhibited the smallest plaque area at 4.91% of the total aortic tissue area compared to 13.64% for miRNC@H-MNP treatment, 9.08% for atorvastatin treatment, and 8.83% for miR-10a@NP (Fig. 5D).
a Schematic illustration of the treatment protocols. b Body weight changes of ApoE−/− mice treated with various formulations. Data are presented as the mean ± SD (n = 10 independent samples). c Representative images of en face ORO staining of the aorta from ApoE−/− mouse treated with different formulations (saline, miRNC@H-MNP, atorvastatin, miR-10a@NP, and miR-10a@H-MNP). d Quantification of plaque area using ImageJ. Data are shown as the mean ± SD (n = 10 independent samples). e Representative images of ORO staining in aortic root from different treatment groups and quantification of the area of the lesion (black dashed box) by Image J software (scale bar: 200 µm). Data are shown as the mean ± SD (n = 5 independent samples). f Representative images of H&E staining of aortic root in different treatment groups and quantification of the necrotic core area (dark blue solid line frame) relative to lesions area (black dashed line frame) using Image J software (scale bar: 200 µm), Data are shown as the mean ± SD (n = 5 independent samples). g Representative images of Masson’s trichrome staining of the aortic root from different treatment groups and quantification of fibrous cap thickness using ImageJ software (scale bar: 200 µm). Data are shown as the mean ± SD (n = 5 independent samples). h Representative images of Picrosirius red staining of aortic root from different treatment groups and quantification of collagen area (red) relative to lesion area (white dotted box) using ImageJ software (scale bar: 200 µm). Data are shown as the mean ± SD (n = 5 independent samples). P values in (d) were determined by One-way ANOVA analysis followed by Dunnett’s T3 multiple comparisons test. P values in (e-h) were determined by One-way ANOVA analysis followed by Tukey’s multiple comparisons test. Source data are provided as a Source Data file. Panel (a) was created in BioRender. Fang, F. (2025) https://BioRender.com/pg60q9m.
Additionally, we stained aortic root sections to assess the therapeutic potential of each treatment group. ORO staining results of aortic root sections were consistent with those of en face aorta. Treatment with atorvastatin, miR-10a@NP, and miR-10a@H-MNP significantly reduced neutral lipid deposition in the aortic root (Fig. 5E). Treatment with miR-10a@H-MNP notably reduced the neutral lipid deposition area from 60.46 × 104 μm2 in the control group to 28.15 × 104 μm2, demonstrating its strong anti-atherosclerotic therapeutic effect (Fig. 5E). Hematoxylin and eosin (H&E) staining of the aortic root showed that miR-10a@H-MNP can significantly reduce the area of the necrotic core (dark blue solid box) in the lesion plaque (black dashed box) (Fig. 5F). Masson’s trichromatic staining further confirmed that miR-10a@H-MNP treatment significantly increased the thickness of the fibrous cap in the plaque from 16.10 μm in the control group to 33.12 μm (Fig. 5G). Additionally, we used picrosirius red staining to quantify the type I collagen content in the aortic roots of each treatment group. As shown in Fig. 5H, the miR-10a@H-MNP treatment significantly increases the type I collagen content in the plaque. These results confirm that miR-10a@H-MNP treatment enhances plaque stability and reduces the risk of plaque rupture.
In vivo assessment of miR-10a@H-MNP mediated macrophage repolarization
The characteristic biological behavior of inflammatory macrophages within plaques involves the uptake of significant amounts of oxLDL, which transforms them into foam cells, thereby accelerating the progression of atherosclerosis40. Thus, we evaluated the neutral lipid droplet content in the aortic root using BODIPY staining. As shown in Fig. S25, the neutral lipid droplet content was markedly reduced in the miR-10a@H-MNP treatment group compared to the control group. Subsequently, we conducted DHE staining of the aortic root and quantified the ROS levels using fluorescence microscopy. Figure 6A demonstrates intense red fluorescence in the atherosclerosis model, indicating significant ROS generation within the aortic tissue. As discussed in the preceding section, miR-10a@H-MNP is responsive to elevated H2O2 levels (Fig. 4J) and effectively reduces elevated ROS levels in macrophages in vitro (Fig. S22 and Fig. S23), demonstrating promising anti-atherosclerotic efficacy. Notably, compared to other groups, the miR-10a@H-MNP group exhibited a significantly reduced fluorescence intensity indicating lower ROS levels (Fig. 6A, B). This favorable result suggests that miR-10a@H-MNP has the potential to alleviate systemic oxidative stress and mitigate tissue damage caused by excessive ROS production.
a CLSM images and fluorescence intensity quantification (b) of aortic root DHE staining (L denotes lumen). Data in panel (b) are shown as the mean ± SD (n = 20 sections from 5 independent samples). c IHC staining of aortic root macrophage markers CD68, M1 phenotype macrophage marker CD86, and M2 phenotype macrophage marker CD206 after various treatments, quantification of positive areas (red dotted frame) using image J software (d-f) (scale bar: 200 µm). Data in (d-f) are shown as the mean ± SD (n = 5 independent samples). g, h Flow cytometry analysis of the percentage of CD86- and CD206-positive macrophages in aortic tissue of ApoE−/− mice after treatment with miR-10a@H-MNP. Data are shown as the mean ± SD (n = 5 independent samples). The levels of IL-1β (i), TNF-α (j), and IL-6 (k) in the serum of ApoE−/− mice after different treatments. Data are shown as the mean ± SD (n = 5 independent samples). P values in (b and d-j) were determined by One-way ANOVA analysis followed by Tukey’s multiple comparisons test. P values in (k) were determined by One-way ANOVA analysis followed by Dunnett’s T3 multiple comparisons test. Source data are provided as a Source Data file.
Subsequently, we assessed the effect of miR-10a@H-MNP on macrophage phenotypic transformation in vivo. As shown in Fig. 6C, D, miR-10a@H-MNP treatment significantly decreased the CD68-positive area (red dashed box) in the aortic root compared to the saline, miRNC@H-MNP, atorvastatin, and miR-10a@NP treatment groups. Additionally, immunohistochemistry (IHC) staining results further confirmed that miR-10a@H-MNP treatment decreased the proportion of M1 phenotype macrophages (CD86-positive, red dashed box) and increased the proportion of M2 phenotype macrophages (CD206-positive, red dashed box) positive areas (Fig. 6C, E, and F). Next, we collected aortic tissues from each treatment group to prepare single-cell suspensions, which were then analyzed using flow cytometry to evaluate the macrophage profile. As shown in Fig. 6G, H and Fig. S26, compared to the control group, miR-10a@H-MNP treatment significantly reduced the proportion of M1 phenotype macrophages (CD86-positive) in the aorta of atherosclerotic mice and, on the contrary, significantly increased the proportion of M2 phenotype macrophages (CD206-positive) in the aorta. Moreover, we measured IL-1β, TNF-α and IL-6 levels in the serum of ApoE−/− mice in each treatment group. The results showed that the atorvastatin, miR-10a@NP and miR-10a@H-MNP treatment groups significantly reduced IL-1β, TNF-α and IL-6 levels in the serum of ApoE−/− mice (Fig. 6I-K). Among these, the levels of IL-1β, TNF-α and IL-6 in the serum of ApoE−/− mice in the miR-10a@H-MNP treatment group were significantly lower than those in the atorvastatin and miR-10a@NP treatment groups (Fig. 6I-K).
We subsequently evaluated the effects of miR-10a@H-MNP on macrophage mitochondrial function and histone acetylation levels by immunofluorescence staining. As shown in Fig. S27 A, B, treatment with atorvastatin, miR-10a@NP, and miR-10a@H-MNP significantly increases the TFAM-positive macrophage (CD68 positive) cell number within aortic root plaques. Notably, the miR-10a@H-MNP treatment group exhibited the most pronounced effect in restoring mitochondrial function in macrophages. Additionally, H3K9Ac staining further substantiated that both the miR-10a@NP and miR-10a@H-MNP treatment groups significantly enhanced the histone acetylation levels of macrophages (CD68 positive) in the aortic root plaques. Importantly, compared to the miR-10a@NP treatment group, the histone acetylation level in macrophages treated with miR-10a@H-MNP was significantly increased (Fig. S27 C, D). These findings confirm that miR-10a@H-MNP can effectively restore mitochondrial function and elevate histone acetylation levels in macrophages within atherosclerotic plaques in vivo.
In vivo safety evaluation
After 4 weeks of administration, we evaluated the biosafety of miR-10a@H-MNP. The results showed that the blood cells (RBC, WBC, PLT, lymphocytes, neutrophils) and blood biochemistry in the blood of mice in the miR-10a@H-MNP, atorvastatin, miRNC@H-MNP, miR-10a@NP and saline groups. The indicators have no significant differences (ALT, ALP, ALB, CREA, UREA, UA). In addition, H&E staining results of major organs (heart, liver, spleen, lung, and kidney) also confirmed no significant pathological changes after 4 weeks of administration of miR-10a@H-MNP (Fig. S28).
Discussion
The dynamic imbalance of macrophage phenotypes within atherosclerotic plaques promotes chronic inflammation and disease progression. In this study, we introduce a biomimetic nano-strategy that utilizes miR-10a to reprogram mitochondrial metabolism and remodel epigenetic states in macrophages, presenting a novel approach to attenuate atherosclerosis. Our findings highlight the pivotal role of mitochondrial respiration in maintaining macrophage epigenetic plasticity and phenotype switching, while demonstrating that targeted delivery of miR-10a via H-MNP effectively restores these processes, reshaping the plaque microenvironment and halting disease progression.
At the core of our hypothesis is the interplay between mitochondrial metabolism and histone acetylation. We observed that M1 phenotype macrophages within atherosclerotic plaques display suppressed mitochondrial OXPHOS and reduced acetyl-CoA production, leading to diminished H3K9Ac and a closed chromatin state. This epigenetic silencing aligns with previous studies emphasizing the dependence of macrophage polarization on metabolic reprogramming and chromatin accessibility10,43,44. Notably, atheroprotective flow-sensitive miR-10a emerged as a key regulator, restoring OXPHOS by upregulating mitochondrial biogenesis genes (e.g., Sirt1, Ppargc1b) and FAO/OXPHOS-related genes (e.g., Acox1, Ogdh), thereby replenishing acetyl-CoA pools. The subsequent increase in H3K9Ac levels promoted chromatin remodeling, enabling the transcriptional activation of anti-inflammatory pathways. These results corroborate the paradigm that metabolic intermediates, such as acetyl-CoA, act as epigenetic modifiers, directly linking mitochondrial function to macrophage plasticity9,12.
A key innovation of this work is the design of miR-10a@H-MNP, which integrates dual biomimetic modifications to overcome both systemic and cellular barriers. The RBC membrane cloak significantly prolonged circulation time by evading reticuloendothelial system clearance, while HA-mediated targeting leveraged CD44 overexpression on plaque macrophages. Unlike existing cell membrane biomimetic strategies, cell membrane wrapping alone lacks cell targeting capabilities23,24. The RBC membrane and HA dual-modified miR-10a@H-MNP demonstrated M1 phenotype macrophage targeting both in vitro and in vivo. Furthermore, the ROS-responsive release mechanism ensured precise miR-10a delivery within the inflammatory niche, minimizing off-target effects. This approach builds on recent advances in biomimetic nanocarriers but advances the field by combining immune evasion, active targeting, and stimulus-responsive payload release into a unified platform.
The therapeutic efficacy of miR-10a@H-MNP was robust, both in vitro and in vivo. In plaques, miR-10a@H-MNP reduced M1 phenotype macrophage dominance, suppressed oxLDL uptake, and lowered ROS levels, collectively attenuating lipid deposition and necrotic core expansion. Increased collagen content and fibrous cap thickness further indicated stabilized plaques, which are critical for preventing acute complications41,45. Statins have emerged as a cornerstone in the management and prevention of atherosclerosis, primarily due to their potent cholesterol-lowering and anti-inflammatory effects46,47. Notably, miR-10a@H-MNP was significantly more effective at inhibiting atherosclerosis than free atorvastatin injections. Moreover, the therapeutic efficacy of miR-10a@H-MNP exceeded that of free miR-10a, underscoring the necessity of targeted delivery to overcome the challenging biological barriers of atherosclerotic lesions.
Despite these advances, several limitations warrant consideration. First, while murine models provide valuable insights, human plaques exhibit greater heterogeneity in macrophage subsets and extracellular matrix composition. Translating this strategy to clinical settings will require validation in human-derived macrophages and more advanced preclinical models. Second, long-term safety assessments of chronic miR-10a@H-MNP administration, including potential immune reactions to RBC membrane components, need to be explored. Third, the effects of miR-10a delivery on the epigenetic status or functions of other cells within the plaque, including endothelial cells, vascular smooth muscle cells, and other immune cells, remain unclear. Finally, the broader applicability of this platform, such as co-delivering miR-10a with other anti-inflammatory agents, could further enhance therapeutic outcomes, which merits investigation.
In conclusion, our study establishes miR-10a@H-MNP as a potent nanotherapeutic that bridges mitochondrial metabolism and epigenetic regulation, effectively reprogramming macrophage behavior. By resolving the dual challenges of targeted delivery and microenvironment-responsive drug release, this strategy not only mitigates atherosclerosis but also offers a blueprint for treating other chronic inflammatory diseases driven by maladaptive immune memory.
Methods
Human samples
Healthy artery samples were collected from heart transplant donors, while atherosclerotic artery tissue samples were obtained from heart transplant recipients. The collection process strictly adhered to approved guidelines, with informed written consent from patients. The study received ethical approval from the Ethics Committee of Sichuan University, ensuring compliance with ethical standards (approval No. K2021015). The collected aorta samples were fixed in 10% neutral buffered formalin for 18 to 24 h and then embedded in a paraffin.
Analysis of scRNA-seq data
The single-cell RNA sequencing (scRNA-seq) data was filtered using the GEO dataset (http://www.ncbi.nlm.nih.gov/geo). The scRNA-seq datasets GSE184073 from Takuo et al. 27 and Bo et al. 48 were subsequently analyzed. The raw data was processed using R (version 4.3.3) and the R package “Seurat V5.0” to filter out cells expressing fewer than 200 genes and genes expressed in fewer than 3 cells. Cells expressing <200 genes and containing > 20% mitochondrial genes were considered poor quality and discarded. Subsequently, the data was normalized for library size, wherein the raw gene counts for each cell were normalized relative to the total count. The resulting expression data was then scaled to 10,000 and log-transformed. Next, the normalized data was summarized using principal component analysis and graph-based clustering methods. Subsequently, UMAP plots were used to visualize the resulting clusters in two dimensions. Finally, cells were classified based on the expression of known biomarkers.
Immunofluorescence staining of aorta samples
ApoE−/− mice were purchased from GemPharmatech Co., Ltd (Jiangsu, China). All animals were housed at 24 °C, 50% relative humidity, 12-12 h light–dark cycle and provided with free food and water. All mice were fed a high-fat diet for 8 weeks to establish an atherosclerosis mouse model. After the mice were sacrificed, the aorta was isolated, and serial paraffin sections were taken from the aortic arch. The study received ethical approval from the Ethics Committee of Sichuan University, ensuring compliance with ethical standards (approval No. K2021015).
The sections of mouse and human aorta samples were deparaffinized, followed by antigen retrieval and blocking with 5% goat serum. The sections were stained with the antibodies, including H3K9Ac (#9649, CST, USA), CD86 (#sc-19617, Santa Cruz, USA), iNOS (#33424, SAB, USA), and TFAM (mitochondrial transcription factor A, #29688, SAB, USA) at 4 °C overnight. Subsequently, the samples were incubated with corresponding fluorescently labeled secondary antibodies for 2 h at room temperature. The nuclei were stained with DAPI for an additional 15 min and observed using a confocal laser scanning microscope (CLSM). The fluorescence intensity was analyzed using ZEISS ZEN BLUE 3.8 software.
Analysis of RNA-seq data
The inflammation macrophage RNA-Seq dataset GSE143845 was utilized using the online analytical tool GEO2R (GEO’s online tool for analyzing GEO data, available at http://www.nci.nlm.nih.gov/geo/geo2r). Differentially expressed genes (DEGs) were compared with a pairwise FDR < 0.05 in transcript levels.
Evaluation of mitochondrial respiration
RAW264.7 cells (#TIB-71, ATCC) were pre-treated with 500 ng/mL LPS for 24 h. Subsequently, RAW264.7 cells were transfected with miR-10a (10 nM, synthesized and provided by Sangon Bioengineering (Shanghai) Co., Ltd.) or miRNC (10 nM, synthesized and provided by Sangon Bioengineering (Shanghai) Co., Ltd.) using the LipoRNAi™ transfection reagent (Beyotime Biotechnology, China) in Opti-MEM (#31985-062, Gibco) for 48 h according to the manufacturer’s instructions. Active and total mitochondria were labeled using the MitoTracker Deep Red probe (#40743ES50, Yeasen, China) and Mitotracker Geen FM probe (#40742ES50, Yeasen, China), respectively. Then, CLSM and flow cytometry were used to observe mitochondrial morphology and detect the population of activated mitochondria. The sequences of miR-10a or miRNC are provided in Table S1.
Measurement of mitochondrial respiration by Seahorse
RAW264.7 cells were pre-treated with 500 ng/mL LPS for 24 h. Then, RAW264.7 cells were transfected with miR-10a (10 nM) or miRNC (10 nM) using LipoRNAi™ transfection reagent (Beyotime Biotechnology, China) in Opti-MEM (#31985-062, Gibco) for 48 h following the manufacturer’s instructions. Next, RAW264.7 cells were seeded at a density of 10,000 cells per well in the Mito Stress Test Plate. The plate was then transferred to a Seahorse instrument (Agilent, Seahorse XF) and sequentially treated with 15 μM oligomycin, 10 μM FCCP, and 5 μM rotenone/antimycin A. Each experimental group was analyzed in triplicate.
Measurement of mitochondrial respiration by qRT-PCR
RAW264.7 cells were pre-treated with 500 ng/mL LPS for 24 h. Subsequently, RAW264.7 cells were transfected with miR-10a (10 nM, synthesized and provided by Sangon Bioengineering, Shanghai, China) or miRNC (10 nM, synthesized and provided by Sangon Bioengineering, Shanghai, China) using the LipoRNAi™ transfection reagent (Beyotime Biotechnology, China) in Opti-MEM (#31985-062, Gibco) for 48 h, following the manufacturer’s instructions. Total RNA was extracted using TRIzol reagent (Invitrogen, USA). Quantitative reverse transcription PCR (qRT-PCR) was performed after synthesizing the cDNA sample through reverse transcription. The primers used for qRT-PCR are listed in Table S2.
Detection of H3K9Ac levels in vitro
RAW264.7 cells were pre-treated with 500 ng/mL LPS for 24 h. Subsequently, RAW264.7 cells were transfected with miR-10a (10 nM) or miRNC (10 nM) using the LipoRNAi™ transfection reagent in Opti-MEM, following the manufacturer’s instructions for 48 h. To further investigate the role of mitochondrial respiration in miR-10a regulation of histone acetylation, cells were treated with Etomoxir (Eto, 200 μM; #HY-50202, MCE), a fatty acid oxidation inhibitor. For immunofluorescence staining, the cells were stained with H3K9Ac antibody (#9649, CST, USA) at 4 °C overnight. These samples were then incubated with corresponding fluorescently labeled secondary antibodies for 2 h at room temperature. The nuclei were stained with DAPI for 15 min and observed using a confocal laser microscope. Fluorescence intensity was quantified using ZEISS ZEN BLUE 3.8 software. The sequences of miR-10a and miRNC are provided in Table S1. For the western blot, the total histones of cells in each treatment group were extracted by histone extraction kit (#PK10022, Proteintech, China). Lysates were quantified and normalized using BCA. Proteins were separated using SDS-polyacrylamide gel electrophoresis gel and transferred to polyvinylidene difluoride membranes. After blocking with 5% nonfat dry milk in Tris-buffered saline, PVDF membranes were incubated with primary antibodies H3K9Ac antibody (#9649, CST, USA) and Histon H3 antibody (#4499, CST, USA) overnight. Then, HRP-conjugated secondary antibodies were added and incubated for 1 hour at room temperature. Molecular imaging ChemiScope 6000Touch (CLINX, China) system was used for imaging.
Detection of macrophage repolarization in vitro
RAW264.7 cells were pre-treated with 500 ng/mL LPS for 24 h. Then, RAW264.7 cells were transfected with miR-10a (10 nM) using LipoRNAi™ transfection reagent in Opti-MEM for 48 h according to the manufacturer’s instructions. Furthermore, to determine the role of histone acetylation in miR-10a regulation of macrophage polarization, cells were treated with HDAC inhibitor valproic acid (VA, 0.5 mM; #HY-10585, MCE) and histone acetyltransferase inhibitor anacardic acid (AA, 1 mM; #HY-N2020, MCE). Macrophages of the M1 phenotype and M2 phenotype were labeled with PE-conjugated CD86 (#553692, BD Biosciences) and AF647-conjugated CD206 (#565250, BD Biosciences), respectively. The ability of different treatments to induce the transformation of M1 phenotype macrophages to M2 phenotype in vitro was assessed by flow cytometric (BD FACSCelesta™). Data were analyzed using FlowJo software (version 10.0.7r2). The sequences of miR-10a or miRNC are provided in Table S1.
The preparation of miR-10a@H-MNP
Liposomes were prepared according to previously established methods36. A mixture of 12 mg DSPE-TK-PEG2000 (#RJ0232633, Xian Ruixi, China), 4 mg DPPC (dipalmitoylphosphatidylcholine), 4 mg cholesterol was dissolved in 16 mL dichloromethane. After rotary evaporation, a solution containing 50 nmol miR-10a in 2 mL DEPC water was added. The resulting mixture was centrifuged at 12,000 g for 10 min after hydration for approximately one hour to collect the sediment. The obtained products were referred to as miR-10a@NP. To obtain cy5-miR-10a@NP, cy5-labeled miR-10a was used instead of miR-10a.
Whole blood was collected from C57BL/6 mice and centrifuged at 500 g for 10 min to isolate RBCs. RBC membrane fragments were obtained using a hypoosmotic lysis method. Subsequently, the miR-10a@NP and RBC membrane fragments were extruded through a 200 nm extruder for at least 20 cycles to obtain RBC membrane-fused miR-10a@NP (miR-10a@MNP). To reduce sulfhydryl groups on the surface of RBC membranes, the resulting miR-10a@MNP was treated with 1 mM TCEP (tris(2-carboxyethyl) phosphine) at 37 °C for 30 min. Subsequently, it was reacted with 1 mM HA-PEG-Mal (Hyaluronic acid-PEG2000-Maleimide, #R-0055, Hangzhou Xinqiao Biotechnology Co., Ltd.) for 1 hour to obtain miR-10a@H-MNP.
Characterization of the morphology and size of miR-10a@H-MNP
A TEM (Hitachi, HT7800) was employed to observe the morphology of the obtained miR-10a@H-MNP. A nanoparticle tracking analyzer (NTA, PTM ZetaView) determined the particle size distribution and zeta potential of miR-10a@H-MNP.
Characterization on co-localization of RBC membrane and miR-10a@NP
RBC membrane fragments were prepared and incubated with 10 μL Dio fluorescent dye solution for 30 min. After washing 3 times with PBS, the fragments were extruded with cy5-miR-10a@NP. Following the hyaluronic acid modification described above, the resulting cy5-miR-10a@H-MNP was incubated with LPS pre-treated RAW264.7 cells for 4 h before being observed under a CLSM.
Coomassie bright blue stain and western blotting
Total proteins from RBC membranes and miR-10a@H-MNP were extracted and denatured. An appropriate loading buffer was added, and the sample was loaded onto an SDS-PAGE gel for electrophoresis. Protein distribution profiles were characterized using Coomassie Brilliant Blue dye. Additionally, proteins were transferred to PVDF membranes and probed with antibodies against the RBC-specific protein Ter-119 (#14-5921-82, Thermo Fisher, USA) and CD47 (#DF6649, Affinity). After incubation with an HRP-labeled secondary antibody, signals were detected using the ChemiDoc XRS+ system (Bio-Rad Inc., USA) following incubation with an enhanced chemiluminescence (ECL) solution.
miR-10a encapsulation efficiency
In order to detect the encapsulation efficiency of miR-10a, cy5-miR-10a@NP (DSPE-TK-PEG based nanoparticles without RBC membrane and hyaluronic acid modifications) and cy5-miR-10a@H-MNP were freeze-dried and then dissolved in dichloromethane. The cy5-miR-10a content within liposomes was determined using a standard curve for cy5-miR-10a, while the encapsulation efficiency was calculated as the ratio between miR-10a loading and the initial dose of miR-10a.
In vitro stability of miR-10a
Free miR-10a or miR-10a@H-MNP containing equal amounts of miR-10a were incubated in RNaseA and subsequently subjected to 15% urea-PAGE gel electrophoresis at 100 V for 1 hour. The gel was then treated with NA-red (#D0128, Beyotime Biotechnology, China) staining solution for 1 hour to facilitate miRNA labeling. miRNA distribution was visualized using a gel imaging system (CLINX 6200Touch).
miR-10a release profile in vitro
5 mg of cy5-miR-10a@H-MNP was dispersed in PBS containing 0 and 1 mM H2O2 at 37 °C. The supernatant was collected by centrifugation at 12,000 g for 10 min at time points of 6, 12, 18, 24, 30, 36, 42, and 48 h, followed by the addition of fresh PBS (with or without H2O2). The amount of miR-10a released was quantified using a fluorescence standard curve for cy5-miR-10a.
Cell cytotoxicity evaluation
RAW264.7 cells were incubated with varying concentrations of miR-10a@H-MNP (0, 50, 100, 200, 300, 400, 500 μg/mL) for 24 h. Subsequently, the CCK-8 reagent was added, and the mixture was incubated for 4 h. The absorbance of the supernatant was measured at a wavelength of 460 nm using a microplate reader.
Cellular uptake in vitro
RAW264.7 cells were incubated with either 0 or 500 ng/mL LPS for 24 h. 100 μg/mL of cy5-miR-10a@H-MNP or cy5-miR-10a@NP was incubated with RAW264.7 cells for 1, 2, 4, and 8 h. Subsequently, the distribution of nanoparticles in macrophages was observed using CLSM. The average fluorescence intensity of cy5-miR-10a@H-MNP was quantified using ZEISS ZEN BLUE 3.8 software.
Additionally, we examined the population behavior of nanoparticle cellular uptake. After incubating cy5-miR-10a@H-MNP with macrophages (with or without LPS pretreatment) for 8 h, the cy5-positive cell was determined by flow cytometry. The results were analyzed using FlowJo software (version 10.0.7r2).
Validation of targeting ability in vitro
RAW264.7 cells were pre-treated with 500 ng/mL LPS for 24 h. In the experimental group, 10 mg/mL HA-PEG2000 was added and incubated for 4 h, followed by the addition of 100 μg/mL cy5-miR-10a@H-MNP, which was then incubated for 4 h. The nuclei were labeled by DAPI staining and observed under a CLSM. The fluorescence intensity was quantified using ZEISS ZEN BLUE 3.8 software.
Blood retention assay
cy5-miR-10a@NP or cy5-miR-10a@H-MNP (100 mg/kg) was injected into male C57BL/6 mice via the tail vein. Subsequently, at predetermined time intervals (5 min, 0.5 h, 1 h, 2 h, 4 h, 8 h, 12 h, 24 h, 48 h, and 72 h), 20 μL of blood was collected from the retinal vein plexus and mixed with 20 μL of heparin sodium solution. Fluorescence intensity was quantified using the IVIS system (Perkin Elmer, USA).
Target ability and distribution of H-MNP in vivo
cy5-miR-10a@NP or cy5-miR-10a@H-MNP (100 mg/kg) was injected into the atherosclerotic male ApoE−/− mice, which were fed a high-fat diet for 8 weeks, via tail vein. After 6 h, the mice were euthanized, and the aorta and major organs, including the heart, liver, spleen, lungs, and kidneys, were collected. The distribution of the Cy5-miR was observed using the IVIS system (Perkin Elmer, USA). The aortic arch was collected for serial cryosectioning, and the M1 phenotype macrophages were labeled with CD86 antibody (#sc-19617, Santa Cruz, USA). The sections were incubated with FITC-labeled secondary antibody for 2 h, followed by DAPI staining to label the cell nuclei. Finally, the sections were observed and photographed using CLSM.
Immunofluorescence staining of cell sample
RAW264.7 cells were pre-treated with 500 ng/mL LPS for 24 h. Subsequently, 100 μg/mL of miRNC@H-MNP, miR-10a@NP, and miR-10a@H-MNP were added and incubated for 24 h. The samples were stained with the indicated antibodies, including CD86 (#sc-19617, Santa Cruz, USA) and CD206 (#ab64693, Abcam, USA), at 4 °C overnight. Subsequently, the samples were incubated with the corresponding fluorescently labeled secondary antibodies for 2 h at room temperature. The nuclei were stained with DAPI and observed using a CLSM. The fluorescence intensity was quantified using ZEISS ZEN BLUE 3.8 software.
qRT-PCR evaluation of the therapeutic effect of miR-10a@H-MNP in vitro
RAW264.7 cells were pre-treated with 500 ng/mL LPS for 24 h. Then, 100 μg/mL miRNC@H-MNP, miR-10a@NP, and miR-10a@H-MNP were added and incubated for 24 h. Total RNAs were extracted using TRIzol reagent (Invitrogen, USA). The qRT-PCR procedure was performed following reverse transcription to obtain the cDNA sample. The upstream and downstream primers used in the qRT-PCR are listed in Table S2.
Evaluation of oxLDL uptake ability
RAW264.7 cells were pre-treated with 500 ng/mL LPS for 24 h. Then, 100 μg/mL miRNC@H-MNP, miR-10a@NP, and miR-10a@H-MNP were added and incubated for 24 h. Subsequently, 50 μg/mL of oxLDL was added and incubated for an additional 12 h. Next, 10 μM of BODIPY dye (#GC42960, Glpbio, China) solution was added to label lipid droplets inside cells, and the cells were observed using CLSM and flow cytometry. The results were analyzed using FlowJo software (version 10.0.7r2).
DHE and ROS staining
RAW264.7 cells were pre-treated with 500 ng/mL LPS for 24 h. Subsequently, 100 μg/mL of miRNC@H-MNP, miR-10a@NP, and miR-10a@H-MNP were added and incubated for 24 h. Next, 5 μM of dihydroethidium dye (#S0063S, Beyotime, China) or 10 μM of DCFH-DA dye (#S0033S, Beyotime, China) solution was added and incubated for 30 min at 37 °C. Finally, the sample was observed using CLSM. Fluorescence intensity was quantified using ZEISS ZEN BLUE 3.8 software.
Blood compatibility test
10% of the RBCs were collected from male C57BL/6 mice and incubated with different concentrations (50, 100, 200, 300, 400, 500 μg/mL) of miR-10a@H-MNP at 37 °C for 3 h. Then, the centrifuge was placed at 10,000 g for 1 minute, the supernatant was collected, and the absorbance peak was measured at 541 nm. Calculate the hemolysis rate using Eq. 1.
At, An, and Ap are the absorbance values of samples, negative control, and positive control. PBS is a negative control, and deionized water is a positive control, respectively.
Treatment for atherosclerosis mice and ORO staining of the aorta
Fifty male ApoE−/− mice were randomly divided into five groups after being fed a high-fat diet for 8 weeks (n = 10 per group). Mice in each group were injected with saline, atorvastatin (2 mg/kg), miRNC@H-MNP (100 mg/kg), miR-10a@NP (100 mg/kg), or miR-10a@H-MNP (100 mg/kg) via the tail vein twice a week for 4 weeks. Three days after the final injection, the mice were euthanized, and the aorta was isolated from the aortic arch to the iliac bifurcation. Adventitial fat tissue was carefully separated, and the aorta was radially cut and fixed in 4% paraformaldehyde solution for 4 h. The aorta was then stained with Oil Red O (ORO) dye for 10 min to assess neutral lipid deposition within the arterial walls. Images of stained tissue were acquired using a PANNORAMIC MIDI slide scanner. ORO-positive areas and total aortic tissue areas were quantified using ImageJ software.
Histological and IHC staining of the aortic root
For ORO staining of the aortic root, 3 days after the last tail vein injection, the aortic root of 5 randomly selected mice in each group was used to prepare 10-μm serial cryosections, following the previous description49. Only sections with intact and continuous valvular leaflets structures were used for subsequent staining. After 15 min of staining with ORO dye, images of the stained sections were captured using a PANNORAMIC MIDI slide scanner. Finally, the average positive area from 3 non-consecutive sections of 5 biological samples was quantified using Image J software.
The ROS levels in the aortic root cryosections were quantified using DHE staining (5 μM of dihydroethidium dye, #S0063S, Beyotime, China), the neutral lipid droplets of the aortic root cryosections were quantified using BODIPY staining (10 μM, #GC42960, Glpbio, China). For quantification, the fluorescence intensity of 4 non-consecutive sections of 5 biological samples was quantified using ZEISS ZEN BLUE 3.8 software.
For the histological staining of the aortic root, 6-μm serial paraffin sections were prepared from the aortic roots of the remaining 5 mice in each group. Sections with intact and continuous valvular leaflet structures were used for subsequent staining. The necrotic core and lesion area were quantified using H&E staining, fibrous cap thickness was quantified using Masson staining, and collagen content was quantified using Picrosirius Red staining50. Then, bright-field images of the H&E- and Masson-stained sections were acquired using a PANNORAMIC MIDI slide scanner. Polarized light images of the Picrosirius Red stained sections were then acquired using Olympus SLIDEVIEW VS200. Finally, the average positive area of 3 non-consecutive sections of 5 biological samples was quantified using Image J software.
For IHC staining, paraffin sections were incubated with CD68 antibody (#ab303565, Abcam, USA) to label macrophage-positive areas. The sections were incubated with CD86 antibody (#sc-19617, Santa Cruz, USA) to label the M1-type macrophage-positive area. The sections were incubated with CD206 antibody (#ab64693, Abcam, USA) to label the M2-type macrophage-positive area. Then, images of the stained sections were acquired using a PANNORAMIC MIDI slide scanner. Finally, the average positive area of 3 non-consecutive sections of 5 biological samples was quantified using Image J software.
For Immunofluorescence staining, paraffin sections of the aortic root were deparaffinized, followed by antigen retrieval and blocking with 5% goat serum. The sections were stained with the antibodies, including H3K9Ac (#9649, CST, USA), TFAM (#29688, SAB, USA), and CD68 (#sc-20060, Santa Cruz, USA) at 4 °C overnight. Subsequently, the samples were incubated with corresponding fluorescently labeled secondary antibodies for 2 h at room temperature. The nuclei were stained with DAPI for an additional 15 min and observed using a CLSM. Image J software was used to quantify the number of TFAM, H3K9Ac, and CD68 positive cells within the plaques.
Measurement of inflammatory cytokines in serum
Blood samples were collected from ApoE−/− mice 3 days after the last tail vein injection. The blood samples were allowed to stand at room temperature for 30 min and then centrifuged at 2,000 g for 10 min to obtain serum. The ELISA kit measured the inflammatory cytokines including IL-1β, TNF-α and IL-6 (Beijing Solarbio Science & Technology Co., Ltd).
Flow cytometry assay for macrophage polarization in vivo
25 male ApoE−/− mice were randomly divided into 5 groups after being fed a high-fat diet for 8 weeks (n = 5). Mice in each group were injected with saline, atorvastatin (2 mg/kg), miRNC@H-MNP (100 mg/kg), miR-10a@NP (100 mg/kg) and miR-10a@H-MNP (100 mg/kg) via the tail vein twice a week for 4 weeks. ApoE−/− mice were euthanized three days after the last administration and perfused with heparin solution. The aorta was isolated from the aortic arch to the iliac bifurcation, and the adventitial fat tissue was carefully removed. The aortic tissue was incubated with an enzyme solution (1 mg/mL Collagenase Ⅱ (#C8150, Solarbio, China), 6.67 mg/mL trypsin inhibitor (#T8031, Solarbio, China), and 0.744 units/mL Elastase) at 37 °C for 10 min and then separate the vascular adventitia. Subsequently, the aorta tissues were minced and incubated with the same enzyme solution at 37 °C for another 40 min. After passing through a 100-mesh cell strainer, the suspension was centrifuged at 200 g for 5 min to collect the precipitation. The single-cell suspensions were incubated with BV421-conjugated F4/80 antibody (#565411, BD Biosciences), PE-conjugated CD86 (#553692, BD Biosciences) and AF647-conjugated CD206 (#565250, BD Biosciences) for 10 min. After removing unbound antibodies, the cell suspension was analyzed by flow cytometry (BD Accuri™ C6), and the data were analyzed using Flow Jo software (version 10.0.7r2).
Statistical analysis
All data are expressed as this study’s mean ± standard deviation (SD). GraphPad Prism version 9 was used for statistical analysis. The normality of data distribution was rigorously assessed using the Shapiro-Wilk test. For normally distributed datasets involving more than two group comparisons, one-way analysis of variance (ANOVA) was implemented to evaluate intergroup mean differences, preceded by joint verification of variance homogeneity through Brown-Forsythe test and Bartlett’s test. Post-hoc analyses were performed using Tukey’s test for homogeneous variances and the Dunnett’s T3 test for heterogeneous variances. Non-normally distributed datasets underwent nonparametric evaluation via Kruskal-Walli’s test, followed by Dunn’s multiple comparison test. In the comparison between the two groups, an independent samples t-test was applied to normally distributed data with equal variances, while Welch’s t-test addressed variance heterogeneity confirmed by F-test. Mann-Whitney U test was used for non-parametric data. p < 0.05 was considered to be indicative of statistically significant differences.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
All data supporting the findings of this study are available within the Article and its Supplementary Information. Source data are provided with this paper.
References
Fang, F. et al. Inflammatory endothelium-targeted and cathepsin responsive nanoparticles are effective against atherosclerosis. Theranostics 12, 4200–4220 (2022).
Björkegren, J. L. M. & Lusis, A. J. Atherosclerosis: Recent developments. Cell 185, 1630–1645 (2022).
Groh, L., Keating, S. T., Joosten, L. A. B., Netea, M. G. & Riksen, N. P. Monocyte and macrophage immunometabolism in atherosclerosis. Semin. Immunopathol. 40, 203–214 (2018).
Riksen, N. P., Bekkering, S., Mulder, W. J. M. & Netea, M. G. Trained immunity in atherosclerotic cardiovascular disease. Nat. Rev. Cardiol. 20, 799–811 (2023).
Saeed, S. et al. Epigenetic programming of monocyte-to-macrophage differentiation and trained innate immunity. Science 345, 1251086 (2014).
van Tuijl, J., Joosten, L. A. B., Netea, M. G., Bekkering, S. & Riksen, N. P. Immunometabolism orchestrates training of innate immunity in atherosclerosis. Cardiovascular Res. 115, 1416–1424 (2019).
Chen, X. et al. Requirement for the histone deacetylase Hdac3 for the inflammatory gene expression program in macrophages. Proc. Natl Acad. Sci. 109, E2865–E2874 (2012).
Mullican, S. E. et al. Histone deacetylase 3 is an epigenomic brake in macrophage alternative activation. Genes Dev. 25, 2480–2488 (2011).
Sun, L., Zhang, H. & Gao, P. Metabolic reprogramming and epigenetic modifications on the path to cancer. Protein Cell 13, 877–919 (2021).
Dumont, A., Lee, M., Barouillet, T., Murphy, A. & Yvan-Charvet, L. Mitochondria orchestrate macrophage effector functions in atherosclerosis. Mol. Asp. Med. 77, 100922 (2021).
Fusagawa, H. et al. Skeletal muscle endurance declines with impaired mitochondrial respiration and inadequate supply of acetyl-CoA during muscle fatigue in 5/6 nephrectomized rats. J. Appl. Physiol. 135, 731–746 (2023).
Matilainen, O., Quirós, P. M. & Auwerx, J. Mitochondria and Epigenetics - Crosstalk in Homeostasis and Stress. Trends Cell Biol. 27, 453–463 (2017).
Tamargo, I. A., Baek, K. I., Kim, Y., Park, C. & Jo, H. Flow-induced reprogramming of endothelial cells in atherosclerosis. Nat. Rev. Cardiol. 20, 738–753 (2023).
Fang, F. et al. Cathepsin K contributed to disturbed flow-induced atherosclerosis is dependent on integrin-actin cytoskeleton–NF–κB pathway. Genes Dis. 10, 583–595 (2023).
Kumar, S., Kim, C. W., Simmons, R. D. & Jo, H. Role of Flow-Sensitive microRNAs in Endothelial Dysfunction and Atherosclerosis. Arteriosclerosis, Thrombosis, Vasc. Biol. 34, 2206–2216 (2014).
Baek, K. I. & Ryu, K. Role of Flow-Sensitive Endothelial Genes in Atherosclerosis and Antiatherogenic Therapeutics Development. J. Cardiovascular Transl. Res. 17, 609–623 (2024).
Fang, Y., Shi, C., Manduchi, E., Civelek, M. & Davies, P. F. MicroRNA-10a regulation of proinflammatory phenotype in athero-susceptible endothelium in vivo and in vitro. Proc. Natl Acad. Sci. 107, 13450–13455 (2010).
Wei, Y. et al. Dicer in Macrophages Prevents Atherosclerosis by Promoting Mitochondrial Oxidative Metabolism. Circulation 138, 2007–2020 (2018).
Njock, M.-S. et al. Endothelial cells suppress monocyte activation through secretion of extracellular vesicles containing antiinflammatory microRNAs. Blood 125, 3202–3212 (2015).
Li, Z. et al. Mechano-boosting nanomedicine antitumour efficacy by blocking the reticuloendothelial system with stiff nanogels. Nat. Commun. 14, 1437 (2023).
Bourquin, J. et al. Biodistribution, Clearance, and Long-Term Fate of Clinically Relevant Nanomaterials. Adv. Mater. 30, 1704307 (2018).
Hu, C.-M. J. et al. Nanoparticle biointerfacing by platelet membrane cloaking. Nature 526, 118–121 (2015).
Wang, Y. et al. Biomimetic Nanotherapies: Red Blood Cell Based Core–Shell Structured Nanocomplexes for Atherosclerosis Management. Adv. Sci. 6, 1900172 (2019).
Gao, C. et al. Treatment of atherosclerosis by macrophage-biomimetic nanoparticles via targeted pharmacotherapy and sequestration of proinflammatory cytokines. Nat. Commun. 11, 2622 (2020).
Li, C. et al. Engineering extracellular vesicles derived from endothelial cells sheared by laminar flow for anti-atherosclerotic therapy through reprogramming macrophage. Biomaterials 314, 122832 (2025).
Ma, Q. et al. Hyaluronic Acid-Guided Cerasome Nano-Agents for Simultaneous Imaging and Treatment of Advanced Atherosclerosis. Adv. Sci. 10, 2202416 (2023).
Emoto, T. et al. Single-Cell RNA Sequencing Reveals a Distinct Immune Landscape of Myeloid Cells in Coronary Culprit Plaques Causing Acute Coronary Syndrome. Circulation 145, 1434–1436 (2022).
Jin, F. et al. Targeting epigenetic modifiers to reprogramme macrophages in non-resolving inflammation-driven atherosclerosis. Eur. Heart J. Open 1, 1–22 (2021).
Ghafouri-Fard, S. et al. The impact of non-coding RNAs on macrophage polarization. Biomedicine Pharmacother. 142, 112112 (2021).
Liu, G. & Abraham, E. MicroRNAs in Immune Response and Macrophage Polarization. Arteriosclerosis, Thrombosis, Vasc. Biol. 33, 170–177 (2013).
Chiu, J. J. & Lee, D. Y. Hemodynamics-based strategy of using retinoid acid receptor and retinoid X receptor agonists to treat atherosclerotic diseases. Atherosclerosis 263, e50–e51 (2017).
Kim, H. et al. General and Facile Coating of Single Cells via Mild Reduction. J. Am. Chem. Soc. 140, 1199–1202 (2018).
Koh, I. et al. Tris(2-carboxyethyl)phosphine-Mediated Nanometric Extracellular Matrix-Coating Method of Mesenchymal Stem Cells. ACS Biomater. Sci. Eng. 6, 813–821 (2020).
Hao, M. et al. Sensitizing Tumors to Immune Checkpoint Blockage via STING Agonists Delivered by Tumor-Penetrating Neutrophil Cytopharmaceuticals. ACS Nano 17, 1663–1680 (2023).
Liu, Y. et al. Neutrophil-Membrane-Coated Biomineralized Metal–Organic Framework Nanoparticles for Atherosclerosis Treatment by Targeting Gene Silencing. ACS Nano 17, 7721–7732 (2023).
Su, D. et al. MicroRNA-195 liposomes for therapy of Alzheimer’s disease. J. Controlled Release 365, 583–601 (2024).
Wang, J. et al. A one-two punch targeting reactive oxygen species and fibril for rescuing Alzheimer’s disease. Nat. Commun. 15, 705 (2024).
Zhang, B. et al. Precise RNA Editing: Cascade Self-Uncloaking Dual-Prodrug Nanoassemblies Based on CRISPR/Cas13a for Pleiotropic Immunotherapy of PD-L1-Resistant Colorectal Cancer. Adv. Funct. Mater. 33, 2305630 (2023).
He, Z. et al. Resolvin D1 delivery to lesional macrophages using antioxidative black phosphorus nanosheets for atherosclerosis treatment. Nat. Nanotechnol. 19, 1386–1398 (2024).
Fang, F., Xiao, C., Li, C., Liu, X. & Li, S. Tuning macrophages for atherosclerosis treatment. Regen. Biomater. 10, rbac103 (2022).
Fang, F. et al. Macrophage-based pathogenesis and theranostics of vulnerable plaques. Theranostics 15, 1570–1588 (2025).
Fang F. et al. Extracellular Vesicles in Atherosclerosis: From Pathogenesis to Theranostic Applications. Small, 2504761 (2025).
Sowers, M. L. et al. Multi-OMICs analysis reveals metabolic and epigenetic changes associated with macrophage polarization. J. Biol. Chem. 298, 102418 (2022).
Guo, R. et al. Macrophage Polarisation in the Tumour Microenvironment: Recent Research Advances and Therapeutic Potential of Different Macrophage Reprogramming. Cancer Control 32, 1–19 (2025).
Tao, W. et al. siRNA nanoparticles targeting CaMKIIγ in lesional macrophages improve atherosclerotic plaque stability in mice. Sci. Transl. Med. 12, eaay1063 (2020).
Duivenvoorden, R. et al. A statin-loaded reconstituted high-density lipoprotein nanoparticle inhibits atherosclerotic plaque inflammation. Nat. Commun. 5, 3065 (2014).
Collins, R. et al. Interpretation of the evidence for the efficacy and safety of statin therapy. Lancet 388, 2532–2561 (2016).
Liang, B., Liang, W.-L. & Liao, H.-L. Single-cell and bulk characterisation of the distinct immune landscape and possible regulatory mechanisms in coronary plaques vulnerability. Clin. Transl. Med. 13, e1281 (2023).
Centa M., Ketelhuth D. F. J., Malin S., Gisterå A. Quantification of Atherosclerosis in Mice. 1940-087X (2019).
Rittié L. Method for Picrosirius Red-Polarization Detection of Collagen Fibers in Tissue Sections. In: Fibrosis: Methods and Protocols (ed Rittié L.). Springer New York (2017).
Acknowledgements
This study was supported by the National Natural Science Foundation of China (Grant numbers 12372315 to X.H.L, and 32301089 to F.F).
Author information
Authors and Affiliations
Contributions
F.F. and X.L. proposed and designed this project. F.F. performed most of the tests and analyses and wrote the manuscript, and X.L. and Y.S. revised it. E.W., H.Y., and T.Z. are responsible for data analysis. Q.W. and Z.Z. are responsible for data visualization. All authors discussed the results and provided suggestions.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks Noemi Rotllan, Ruibing Wang and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Source data
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Fang, F., Wang, E., Yang, H. et al. Reprogramming mitochondrial metabolism and epigenetics of macrophages via miR-10a liposomes for atherosclerosis therapy. Nat Commun 16, 9117 (2025). https://doi.org/10.1038/s41467-025-64201-8
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-025-64201-8
