Results

Isolation of functional mitochondria and fabrication of MQgel@Mito

We isolated donor mitochondria from the skeletal muscles of Sprague Dawley rats. To ensure the success of mitochondrial transplantation, the quality of isolated mitochondria was strictly detected using different methods. Western blot analysis confirmed the purity of the isolated naked mitochondria (Mito), with robust expression of the mitochondrial marker cytochrome c oxidase subunit Ⅳ (COX Ⅳ) and the absence of the cytoplasmic marker β-tubulin (Fig. 2A). Transmission electron microscopy (TEM) revealed characteristic double-membrane ultrastructure with well-organized cristae, indicating preserved morphological integrity (Fig. 2B). Functional assessment demonstrated that, immediately after isolation, the naked mitochondria exhibited strong respiratory competence. JC-1 staining confirmed high mitochondrial membrane potential (ΔΨm), with the naked mitochondria displaying predominantly red fluorescence (JC-1 aggregates) indicative of polarized organelles, which could be converted by the mitochondrial uncoupler carbonyl cyanide-4-(trifluoromethoxy) phenylhydrazone (FCCP) to green fluorescence (JC-1 monomers) emitting principally (Fig. 2C and D). The synthesis of ATP, relying on the relatively high ΔΨm, directly reflects the functional viability of mitochondria, as evidenced by substantial ATP production in the presence of respiratory substrates, which was abolished upon addition of the ATP synthase inhibitor oligomycin (Fig. 2E). These quality control measures collectively established that our isolation protocol yielded structurally intact and functionally active mitochondria suitable for transplantation studies.

Fig. 2: Isolation of functional mitochondria and fabrication of MQgel@Mito. (A) Expression levels of COX Ⅳ and β-tubulin were evaluated in input, isolated mitochondria suspension (Mito), and cytoplasm (Cyto) to assess the purity of the isolated mitochondria. (B) Representative TEM image of isolated mitochondria. (C) Representative fluorescence images of JC-1 staining in isolated mitochondria. Agg, JC-1 aggregates (red); Mono, JC-1 monomers (green). (D) Quantification of red/green ratio of JC-1 staining in isolated mitochondria (n = 3). (E) ATP production of isolated mitochondria (n = 3). Sub, respiratory substrates; Oly, oligomycin. (F) Representative JC-1 staining fluorescence images of mitochondria at 8 h post-isolation. (G) JC-1 fluorescence intensity ratio (red/green) of mitochondria at 8 h post-isolation (n = 3). (H) Relative fold change in ATP levels of isolated mitochondria protected by MQgel at different concentrations (n = 3). (I-J) Activity of mitochondrial respiratory chain Complex Ⅰ (I) and Complex Ⅳ (J) of isolated mitochondria protected by MQgel at different concentrations (n = 3). (K) Rheological properties of 30 mg mL−1 MQgel and MQgel@Mito under frequency sweeps ranging from 0.1 to 34 rad s−1. G′ represents storage modulus; G′′ represents loss modulus. (L) Rheological properties of 30 mg mL-1 MQgel and mg mL-1 MQgel@Mito under strain sweeps. (M) Self-healing properties of 30 mg mL-1 MQgel@Mito under alternating oscillatory strains: high strain (100%) and low strain (1%). (N) Representative 3D z-stack confocal fluorescence images of MQgel@Mito. (O) The degradation profile of MQgel@Mito in vitro (n = 3). (P) Mitochondrial release kinetics from MQgel@Mito in vitro (n = 3). Data are presented as mean ± standard deviation (SD); ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001.

Subsequently, MQgel was fabricated by covalently conjugating the mitochondria-derived peptide MOTS-c (MRWQEMGYIFYPRKLR) with the self-assembling peptide Q11 (QQKFQFQFEQQ) at a 1:1 stoichiometric ratio, after which an appropriate amount of NaCl was added to the peptide solution to induce the formation of the hydrogel. The circular dichroism (CD) spectra revealed that MOTS-c exhibited a random coil conformation, evidenced by a negative band near 200 nm. In contrast, Q11 displayed a characteristic β-sheet signature with a negative band at around 216 nm. The MQ peptide exhibited a mixed CD profile containing both α-helical and antiparallel β-sheet contributions (Fig. S1A). Deconvolution of the spectra indicated that the antiparallel β-sheet content accounted for 52.3% of the total secondary structure, confirming that the antiparallel β-sheet conformation, the structural prerequisite for self-assembly, was preserved after MOTS-c conjugation (Fig. S1B). The TEM images further demonstrated that both Q11 and MQ formed well-defined nanofibrous networks, whereas MOTS-c did not assemble into ordered structures under the same conditions (Fig. S2).

We fabricated MQgel at concentrations of 20, 30, and 40 mg mL−1 as described in a previously published protocol (Fig. S3) [26]. The mechanical properties of MQgel were characterized via rheological testing, and the storage modulus (G′) values exceeded the loss modulus (G″) across all concentrations in the angular frequency sweep testing, confirming successful hydrogel formation (Fig. S4). Given that the storage modulus of 40 mg mL−1 MQgel exceeded 1000 Pa, which potentially limited the intramyocardial injectability, we continued to investigate the protective effects of 20 and 30 mg mL−1 MQgel via functional assessments of mitochondria. At 8 h post-isolation, JC-1 staining revealed that mitochondria encapsulated in 30 mg mL−1 MQgel exhibited the strongest red fluorescence and had a significantly higher red/green fluorescence ratio than the other groups, while naked mitochondria showed the lowest ratio (Fig. 2F and G). Consistently, ATP synthesis assays demonstrated that 30 mg mL−1 MQgel sustained mitochondrial ATP production for up to 8 h — a marked extension over unprotected mitochondria — whereas 20 mg mL−1 MQgel had little effect. Beyond 8 h, all groups exhibited irreversible loss of synthetic capacity, defining the practical storage window (Fig. 2H). Assessment of respiratory chain complex activities further confirmed that 30 mg mL−1 MQgel preserved both Complex Ⅰ and Complex Ⅳ function after 8 h (Fig. 2I and J). According to these results, 30 mg mL−1 MQgel was selected for mitochondrial protection. To distinguish the contributions of Q11gel, MOTS-c, or MQgel to mitochondrial protection, we compared their effects on ATP synthesis of isolated mitochondria at the 8-h time point. All three treatments improved ATP production relative to naked mitochondria, with MQgel displaying the strongest protective effect, as reflected by the highest ATP synthesis activity (Fig. S5).

Rheological characterization of the composite system (MQgel@Mito) demonstrated that G′ consistently exceeded G″ across the angular frequency range tested, confirming elastic gel-like behavior (Fig. 2K, Fig. S6). Strain sweep testing revealed shear-thinning properties essential for injection through narrow-gauge needles (Fig. 2L). In time-sweep rheological tests, the values of G′ exceeded those of modulus (G″) at a low strain of 1%, but fell below G″ at a high strain of 100%. This strain-dependent viscoelastic behavior demonstrated the self-healing capability of both MQgel@Mito and MQgel, a critical property for maintaining structural integrity after injection into cardiac tissue (Fig. 2M; Fig. S7). To directly visualize the distribution of mitochondria within the hydrogel, MQgel was labeled with fluorescein isothiocyanate (FITC) for green fluorescence, and mitochondria were labeled with MitoTracker Red (MTR) for red fluorescence. Three-dimensional confocal z-stack scanning of the resulting MQgel@Mito provided direct visualization of the isolated mitochondria encapsulated within the hydrogel matrix rather than merely on its surface (Fig. 2N). Scanning electron microscopy (SEM) revealed that 30 mg mL−1 MQgel exhibited a porous three-dimensional architecture resembling an extracellular matrix scaffold (Fig. S8). A gravimetric degradation assay showed that MQgel@Mito underwent gradual degradation in vitro in approximately 10 days at 37 °C (Fig. 2O). The mitochondrial release profile closely paralleled the hydrogel degradation kinetics, with 78.6% ± 4.1% of encapsulated mitochondria released by day 4, after which the release rate slowed substantially and ceased by day 10 (Fig. 2P). This concordance between degradation and release confirmed that mitochondrial release was governed by hydrogel matrix erosion and that MQgel@Mito could achieve sustained mitochondrial delivery. Overall, these results demonstrated that we successfully isolated functional mitochondria with intact ultrastructure and biological function, and fabricated a hydrogel system for the protection of isolated mitochondria (MQgel@Mito).

MQgel shielded isolated mitochondria from oxidative stress and calcium overload

The post-infarction myocardial microenvironment presents formidable challenges to transplanted mitochondria, characterized by pathological accumulation of ROS and calcium overload, both of which severely compromise organelle function. We therefore evaluated the protective capacity of MQgel under conditions simulating these stressors, using tert-butyl hydroperoxide (TBHP) to model oxidative stress and CaCl2 to simulate calcium overload, respectively. First, we assessed the functional consequences, which were striking: compared with untreated mitochondria, naked mitochondria exposed to TBHP showed a 17.85% ± 1.30% reduction in ATP production at 30 min, and the reduction increased to 22.99% ± 4.76% at 2 h. In contrast, MQgel@Mito exhibited strong resistance to TBHP-induced damage, with ATP production decreasing by merely 2.46% ± 0.81% and 5.46% ± 3.78% at the two corresponding time points (Fig. 3A). Longitudinal analysis with data normalized to the 30-min baseline confirmed that MQgel@Mito exerted sustained protective effects over time. After 2 h of TBHP exposure, ATP synthesis in MQgel@Mito retained 0.84 of its 30-min level, whereas unprotected mitochondria retained only 0.57 (Fig. 3B). Calcium overload also induced distinct differences between groups. Naked mitochondria had ATP synthesis reduced by 38.63% ± 1.66% (30 min) and 39.55% ± 1.87% (2 h), while the reductions for MQgel@Mito were merely 19.00% ± 6.30% and 8.06% ± 3.31% (Fig. 3C). Two hours post calcium overload, ATP levels in MQgel-protected and unprotected mitochondria fell to 0.77- and 0.55-fold of their 30-min values (Fig. 3D).

Fig. 3: MQgel shields isolated mitochondria from oxidative stress and calcium overload. (A) ATP relative reduction induced by TBHP at 0.5 and 2 h (n = 3). (B) Fold change of ATP production under TBHP exposure from 0.5 to 2 h (n = 3). (C) ATP relative reduction induced by calcium overload at 0.5 and 2 h (n = 3). (D) Fold change of ATP production under calcium overload from 0.5 to 2 h (n = 3). (E-F) Activity of mitochondrial respiratory chain Complex I and Complex III in isolated mitochondria with or without TBHP exposure (n = 3). (G-H) Activity of mitochondrial respiratory chain Complex I and Complex III in isolated mitochondria with or without calcium overload (n = 3). (I) Representative JC-1 staining fluorescence images of mitochondria under different stimuli. (J) Fluorescence intensity ratio (red/green) of mitochondria stained with JC-1 (n = 3). (K) Representative TEM images of isolated mitochondria under TBHP exposure. Data are presented as mean ± SD; ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001.

To elucidate the mechanistic basis of this protection, we assessed the activities of mitochondrial respiratory chain complexes. Exposure to either stressor markedly inhibited respiratory chain function of naked mitochondria, yet MQgel substantially preserved the activities of Complex Ⅰ and Complex Ⅲ under both conditions (Fig. 3E–H). Besides, we detected the mitochondrial membrane potential using JC-1 staining. TBHP exposure caused substantial depolarization, evidenced by diminished red fluorescence and enhanced green fluorescence, an effect largely reversed by MQgel, indicating the preserved membrane potential. Similar protective effects were observed under calcium overload alone, as well as under combined TBHP and calcium overload, demonstrating that MQgel consistently protects mitochondria across these adverse conditions (Fig. 3I and J). Quantification of mitochondrial superoxide using MitoSOX Red demonstrated significantly lower fluorescence intensity in MQgel@Mito compared to naked mitochondria following TBHP exposure (Fig. S9). From a structural perspective, TEM revealed that TBHP induced morphological abnormalities such as mitochondrial vacuolization and cristae disruption, which were partially rescued by MQgel encapsulation (Fig. 3K). Overall, these findings indicated that MQgel constructed a favorable microenvironment to maintain mitochondrial homeostasis under adverse conditions encountered during mitochondrial transplantation, such as elevated oxidative stress and calcium overload.

MQ-mediated mitochondrial internalization modulated macrophage polarization and improved anti-inflammatory responses

Macrophages have emerged as central orchestrators of the post-MI response, governing the transition from acute inflammation to tissue repair. Therefore, macrophages were applied as recipient cells and the donor mitochondrial internalization was first investigated. In vitro, MQ peptide solution was used to mimic the degradation-released functional products of MQgel. CCK-8 and live/dead staining demonstrated that MQ exhibited excellent biocompatibility with bone marrow-derived macrophages (BMDMs) (Fig. S10). Additionally, fluorescence imaging further confirmed that MQ could be internalized by BMDMs (Fig. S11). Subsequently, we incubated BMDMs with MTR-labeled donor mitochondria in the presence of MQ with different concentrations (0, 10, and 25 μM) for several periods. Flow cytometry analysis revealed that MQ significantly enhanced donor mitochondrial internalization compared with naked mitochondria, as evidenced by increased mean fluorescence intensity (MFI) of MTR (Fig. S12). Quantitative analysis of MFI revealed a time-dependent rise in MFI in BMDMs. Notably, high-concentration MQ markedly promoted the internalization of donor mitochondria (Fig. 4A). Fluorescence images confirmed efficient cytoplasmic colocalization of donor and recipient mitochondria following co-incubation with high-concentration MQ (Fig. 4B).

Fig. 4: MQ-mediated mitochondrial internalization modulates macrophage polarization and improves anti-Inflammatory responses. (A) Time-course analysis of donor mitochondrial uptake in BMDMs. Cells were treated with Mito, 10-MQ@Mito (MQ, 10 μM), or 25-MQ@Mito (MQ, 25 μM). Uptake was quantified by flow cytometry based on the MFI of MTR-labeled mitochondria (n = 3). MFI, mean fluorescence intensity; MTR, MitoTracker Red. (B) Colocalization of MTG-labeled recipient mitochondria (green) and MTR-labeled donor mitochondria (red) in BMDMs after 2 h and 12 h of mitochondria transplantation. MTG, MitoTracker Green. Cell nuclei were counterstained with Hoechst (blue). (C) Time-course analysis of donor mitochondrial uptake in BMDMs. Cells were treated with Mito, MQ@Mito, or MQ@Mito+si-AMPKα1. Uptake was quantified by flow cytometry based on the MFI of MTR-labeled mitochondria (n = 3). (D) Representative flow cytometry scatter plots showing the proportion of F4/80+CD86+ M1 macrophages under different treatments. (E) Quantification of the proportion of F4/80+CD86+ M1 macrophages (n = 3). (F-G) ELISA analysis of pro-inflammatory cytokines IL-1β and TNF-α in the culture supernatant (n = 3). (H) qPCR analysis of Tnf-α mRNA expression (n = 3). (I) Representative flow cytometry plots showing F4/80+CD206+ M2 macrophages under different treatments. (J) Quantification of the percentage of F4/80+CD206+ M2 macrophages (n = 3). (K-M) qPCR analysis of M2-associated marker genes Arg1, Il10, and Tgf-β (n = 3). Data are presented as mean ± SD; ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001.

We next investigated the underlying molecular mechanism that MQ enhanced mitochondrial internalization. Evidence has been found that mitochondria can be internalized via macropinocytosis, a low-efficiency, non-receptor-mediated process [16]. Macrophages, as immune cells, exhibit high levels of macropinocytic activity, which is closely connected with energy metabolism [27]. AMPK acts as a cellular energy sensor and drives the cytoskeletal remodeling necessary for macropinosome formation by activating the actin-severing protein cofilin, thereby enhancing macropinocytosis [[27], [28], [29]]. Meanwhile, it is reported that MOTS-c, the degradation product of MQ, is an MDP that activates AMPK to regulate metabolic homeostasis [21]. Here, we hypothesized the mechanism by which MQ facilitated the internalization of donor mitochondria was related to AMPK activation. AMPK is a heterotrimeric complex (α, β, γ subunits), with the α subunit (α1/α2 isoforms) as the catalytic core [30]. qPCR analysis showed that AMPKα1 was much more highly expressed than AMPKα2 in BMDMs (Fig. S13A), so we constructed AMPKα1-specific siRNA (si-AMPKα1). Both qPCR and western blot verified significant si-AMPKα1 knockdown at mRNA and protein levels (Fig. S13B–D). We then assessed the effect of AMPKα1 knockdown on mitochondrial uptake in BMDMs at different time points by quantifying the MFI of MTR-labeled mitochondria. Mitochondrial uptake was persistently higher in the MQ@Mito group relative to the Mito group. This promoting effect was largely abrogated following AMPKα1 silencing (Fig. 4C, Fig. S14). Western blot analysis further validated the mechanism. Compared with the Mito group, the MQ@Mito group showed elevated p-AMPKα/AMPKα ratio. Such upregulation was obviously reversed upon AMPKα1 knockdown (Fig. S15) and the AMPK inhibitor Compound C (CC) treatment (Fig. S16). To directly confirm the involvement of macropinocytosis, we performed a FITC-dextran uptake assay. MQ@Mito elevated the MFI of FITC-dextran in BMDMs, reflecting enhanced macropinocytic activity, which was abolished by si-AMPKα1 (Fig. S17). In addition, MQ also significantly promoted mitochondrial uptake in lipopolysaccharide (LPS)-pretreated M1 macrophages and this effect was completely abrogated by si-AMPKα1 (Fig. S18). Collectively, these results demonstrated that MQ promoted mitochondrial uptake by activating AMPK-mediated macropinocytosis.

Given the emerging role of mitochondria in regulating macrophage phenotype, we next examined whether MQ@Mito could modulate macrophage polarization [31,32]. LPS-stimulated BMDMs induced robust M1 polarization, characterized by elevated CD86 expression. MQ@Mito pretreatment visually attenuated this pro-inflammatory phenotype, as evidenced by diminished CD86 immunofluorescence (Fig. S19) and a significant reduction in the F4/80+CD86+ population (Fig. 4D and E). To further evaluate the therapeutic potential of MQ@Mito beyond preventive intervention, we examined its effect on M1 polarization under two distinct administration regimens: co-treatment (MQ@Mito added simultaneously with LPS), and post-treatment (MQ@Mito added after LPS-induced M1 polarization). Flow cytometry analysis revealed that MQ@Mito significantly reduced the proportion of F4/80+CD86+ M1 macrophages under these two conditions (Fig. S20), suggesting that MQ@Mito can reverse, rather than merely prevent, the established pro-inflammatory macrophage phenotype. Furthermore, ELISA measurements of supernatant revealed suppressed secretion of pro-inflammatory cytokines including interleukin-1β (IL-1β) and tumor necrosis factor (TNF-α) (Fig. 4F and G). In line with these findings, qPCR analysis of BMDMs demonstrated reduced Tnf-α mRNA expression (Fig. 4H). In parallel, MQ@Mito significantly increased the proportion of F4/80+CD206+ M2 macrophages (Fig. 4I and J), and upregulated expression of Arg__1, il10 and Tgf-β mRNA (Fig. 4K–M), demonstrating that MQ@Mito treatment could promote M2 polarization. Taken together, these results demonstrated that MQ@Mito effectively suppressed LPS-induced M1 polarization and enhanced M2 polarization.

MQ@Mito modulated macrophage polarization via metabolic reprogramming

Accordingly, the immunometabolism of macrophages has garnered significant interest, with studies highlighting subtype-specific pathways that dictate both their phenotype and function, far beyond mere energy provision. Pro-inflammatory macrophages (M1) and anti-inflammatory macrophages (M2) exhibit divergent metabolic programs: the former rely on enhanced glycolysis with impaired OXPHOS, while the latter are characterized by an intact tricarboxylic acid (TCA) cycle and enhanced mitochondrial OXPHOS [33]. In addition, mitochondrial dysfunction in macrophages exacerbates the post-MI inflammatory response and limits regenerative capacity through mechanisms involving oxidative stress [5]. These reports collectively suggest that macrophage immunometabolism could be a potential alternative therapeutic target for MI treatment. Metabolic reprogramming has emerged as a defining feature of macrophage activation states [33]. To elucidate the impacts induced by MQ@Mito on the cellular energy metabolism, we next investigated the effect of MQ@Mito on the metabolic reprogramming of LPS-stimulated macrophages using Seahorse extracellular flux analysis, which allows simultaneous assessment of glycolysis and mitochondrial respiration. In the glycolysis stress test, LPS stimulation significantly increased the extracellular acidification rate (ECAR), indicating a marked shift toward enhanced glycolysis (Fig. 5A). Notably, MQ@Mito treatment exerted the most pronounced inhibitory effect, significantly reducing both glycolysis and glycolytic capacity (Fig. 5B). Conversely, the mitochondrial stress test demonstrated that LPS exposure severely impaired mitochondrial respiration, as reflected by reduced basal and maximal oxygen consumption rates (OCR), while MQ@Mito treatment resulted in the most robust recovery of both basal and maximal respiration (Fig. 5C and D). Consistent with the Seahorse data, the activities of the key glycolytic enzymes hexokinase (HK) and phosphofructokinase (PFK) were significantly upregulated in LPS-stimulated macrophages. This LPS-induced increase in glycolytic enzyme activity was attenuated by treatment with MQ or mitochondria alone, and most effectively suppressed by MQ@Mito (Fig. 5E and F). Furthermore, LPS stimulation led to a significant reduction in intracellular ATP levels, which was most potently restored by MQ@Mito treatment (Fig. 5G).

Fig. 5: MQ@Mito modulates macrophage polarization by metabolic reprogramming. (A) ECAR determined by the glycolysis stress test using Seahorse (n = 3). (B) Quantitative analysis of glycolysis and glycolytic capacity derived from glycolysis stress test (n = 3). (C) OCR determined by the mitochondrial stress test using Seahorse (n = 3). (D) Quantitative analysis of basal and maximal mitochondrial respiration (n = 3). (E-F) Activity of critical glycolytic enzymes HK (E) and PFK (F) of BMDMs with different treatments (n = 3). HK, hexokinase; PFK, phosphofructokinase. (G) ATP production of BMDMs with different treatments (n = 3). (H) OPLS-DA of metabolic profiles among the Control, LPS, and MQgel@Mito groups (n = 3). (I) Venn diagram illustrating the unique and overlapping metabolites among the Control, LPS, and MQgel@Mito groups. (J) Volcano plot showing metabolites with abundance changes. The red dots refer to the upregulated metabolites, and the blue dots refer to the downregulated metabolites. (K) Clustering heatmap of representative DEMs. (L) Metabolism-related KEGG pathway enrichment analysis of representative DEMs. (M) Abundance of glycolysis-related metabolite (G6P) (n = 3). G6P, D-Glucose-6-phosphate. (N-O) Abundance of TCA-related metabolites cis-aconitate (N) and succinate (O) (n = 3). (P) Scheme illustrating that MQ@Mito regulated the polarization of BMDMs by modulating the metabolic reprogramming. Data are presented as mean ± SD; ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001.

In addition, to elucidate the metabolic basis of MQ@Mito-mediated anti-inflammatory effects, we further performed targeted energy metabolomics analysis using liquid chromatography-tandem mass spectrometry (LC-MS/MS). Orthogonal projections to latent structures-discriminant analysis (OPLS-DA) revealed clear separation among the Control, LPS, and LPS + MQ@Mito (abbreviated as MQ@Mito) groups, confirming substantial metabolic remodeling (Fig. 5H). The Venn diagram illustrated the unique and overlapping metabolites across the three groups, showing that a total of 48 energy metabolism-related metabolites were detected (Fig. 5I). Volcano plot analysis identified significantly up- and down-regulated metabolites (Fig. 5J). Hierarchical clustering heatmap of representative differentially expressed metabolites (DEMs) demonstrated that MQ@Mito treatment substantially reversed LPS-induced metabolic alterations (Fig. 5K). Metabolism-related Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway classification analysis was performed on these representative DEMs (Fig. 5L). The heatmap and KEGG pathway classification both confirmed the involvement of glycolysis/gluconeogenesis, TCA cycle, oxidative phosphorylation pathways, amino acid biosynthesis, and D-amino acid metabolism. Specific metabolic intermediates were further examined to reveal the mechanistic insights of MQ@Mito. LPS stimulation elevated the abundance of D-glucose-6-phosphate (G6P), a key glycolytic intermediate, which was normalized by MQ@Mito treatment (Fig. 5M). Analysis of TCA cycle intermediates revealed that LPS disrupted the cycle, causing accumulation of cis-aconitate and succinate, while MQ@Mito treatment alleviated this metabolic disruption (Fig. 5N and O). The mechanistic basis likely involved dual contributions: the MQ peptide, as an MDP, enhanced mitochondrial function and promoted metabolic normalization, while exogenous healthy mitochondria directly augmented cellular OXPHOS capacity. In summary, MQ@Mito exerted anti-inflammatory effects by reprogramming macrophage metabolism, specifically reversing the LPS-induced glycolytic shift while restoring TCA cycle integrity and OXPHOS levels (Fig. 5P).

MQ@Mito preserved mitochondrial homeostasis and attenuated oxidative injury

Having established the metabolic reprogramming effects, we next examined whether MQ@Mito could maintain mitochondrial homeostasis of macrophages and protect against oxidative stress-induced damage. First, we assessed ΔΨm via JC-1 staining. Flow cytometry showed that TBHP significantly reduced the JC-1 red/green ratio, indicating severe depolarization. Although treatment with MQ or mitochondria alone partially restored ΔΨm, MQ@Mito produced the most robust improvement, confirmed by fluorescence (Fig. 6A and B). Next, we measured intracellular ROS (cROS) and mitochondrial superoxide (mtROS) using DCFH-DA and MitoSOX fluorescent probes. TBHP dramatically increased ROS and superoxide levels. Although MQ or mitochondria monotherapy partially attenuated oxidative stress, MQ@Mito exerted the strongest scavenging effect, significantly reducing both cROS and mtROS (Fig. 6C–H).

Fig. 6: MQ@Mito preserves mitochondrial homeostasis and attenuates oxidative injury. (A) Representative JC-1 fluorescence images of BMDMs after different treatments. (B) Quantification of the red/green JC-1 fluorescence intensity ratio of BMDMs after different treatments (n = 5). (C) Representative fluorescence images of cROS levels in BMDMs, stained with DCFH-DA (green) after different treatments. Cell nuclei were counterstained with Hoechst (blue). (D) Representative fluorescence images of mtROS levels in BMDMs, stained with MitoSOX (red) after different treatments. Cell nuclei were counterstained with Hoechst (blue). (E-F) Representative flow cytometry histograms of DCFH-DA (E) and MitoSOX (F) levels in BMDMs after different treatments. (G-H) Quantitative analysis of MFI of DCFH-DA (G) and MitoSOX (H) in BMDMs after different treatments (n = 3). (I) Representative flow cytometry zebra plots of proportion of viable (Q4, Annexin V−, PI−), early phase apoptotic (Q3, Annexin V+, PI−), and late phase apoptotic (Q2, Annexin V+, PI+) BMDMs after different treatments. (J) Quantitative analysis of total apoptosis of BMDMs after different treatments (n = 3). (K) Representative TEM images of mitochondria in BMDMs after different treatments. (L) Scheme illustrating that MQ@Mito preserved mitochondrial homeostasis and attenuated oxidative injury. Data are presented as mean ± SD; ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001.

Sustained oxidative stress characteristically culminates in programmed cell death [34]. We analyzed apoptosis using Annexin V/PI double staining; total apoptosis was defined as the sum of early apoptotic (Annexin V+/PI−) and late apoptotic (Annexin V+/PI+) populations. Flow cytometry revealed that TBHP exposure induced substantial apoptosis in BMDMs, with a total apoptotic rate of 42.43% ± 6.39% compared with 7.66% ± 1.57% in the control group. MQ@Mito treatment significantly rescued this apoptotic cascade, reducing the total apoptosis rate to 15.96% ± 2.13%, notably lower than either MQ or mitochondria alone (Fig. 6I and J). TEM further confirmed that TBHP caused mitochondrial swelling and cristae disruption, while MQ@Mito substantially preserved normal mitochondrial ultrastructure (Fig. 6K). Collectively, these results demonstrated that MQ@Mito mitigated TBHP-induced oxidative stress, restored mitochondrial function, and inhibited apoptosis (Fig. 6L).

Given the above favorable characteristics, we next investigated the protective effects of MQ@Mito on cardiomyocytes. First, CCK-8 assays and live/dead staining confirmed the excellent biocompatibility of MQ in H9c2 cardiomyocytes (Fig. S21). Fluorescence imaging further demonstrated that cardiomyocytes could internalize exogenous mitochondria, and this uptake process was significantly enhanced by MQ (Fig. S22). Notably, MQ@Mito treatment markedly rescued the viability of TBHP-challenged H9c2 cardiomyocytes, confirming its potent cardioprotective effects (Fig. S23). Collectively, these findings demonstrated that MQ@Mito exerted robust cellular protection by maintaining mitochondrial homeostasis, preserving ΔΨm, reducing mitochondrial damage, scavenging both cROS and mtROS, and attenuating oxidative stress-induced apoptosis. Consistently, MQ@Mito protected cardiomyocytes against oxidative stress injury.

MQgel@Mito improved cardiac function in a rat MI model

MQgel was designed to fabricate a mitochondrial protection and delivery platform (MQgel@Mito). Following confirmation of protective effect in vitro, we next verified the absorption period of MQgel@Mito in vivo. FITC-labeled MQgel-encapsulated donor mitochondria were injected into the cardiac infarct border zone following left anterior descending (LAD) coronary artery ligation in rats. Fluorescence intensity decreased gradually, with 89.28% ± 2.62% of the system absorbed by day 5 (Fig. S24). This gradual absorption profile enabled sustained release of mitochondria rather than burst release, optimizing organelle utilization while protecting transplanted mitochondria from the harsh post-infarction microenvironment during the critical early period.

We next evaluated the therapeutic efficacy of MQgel@Mito in a rat acute MI model (Fig. 7A). Following permanent LAD ligation, rats received intramyocardial injections of saline, MQgel, mitochondria (Mito), or MQgel@Mito within the infarct border zone. Echocardiographic assessment was performed on day 7 and day 28 post-MI to evaluate cardiac function in rats (Fig. 7B). We first quantified left ventricular ejection fraction (LVEF) across all experimental groups and time points (Fig. 7C). On day 7, the Sham group exhibited normal LVEF (84.92%±2.48%), whereas saline-treated MI rats displayed severe systolic dysfunction (32.48%±4.92%). Separate treatment with MQgel (50.06%±9.97%) or Mito (48.41%±4.14%) yielded limited functional recovery. In contrast, the MQgel@Mito group achieved a markedly higher LVEF of 63.22%±6.24%, which was superior to all other MI cohorts. On day 28, the intergroup functional hierarchy remained consistent. In the Saline group, LVEF remained low at 30.87%±6.70%; by contrast, monotherapy with MQgel or Mito yielded only moderate recovery, raising LVEF to 50.84%±4.80% and 49.60%±7.95%, respectively. The combined MQgel@Mito therapy exerted the most potent cardioprotective effect, with the LVEF reaching 67.98%±8.28%. Longitudinal trend analysis showed gradual LVEF deterioration in the Saline group, whereas MQgel@Mito treatment drove sustained, progressive cardiac functional improvement throughout the observational period (Fig. S25A). Left ventricular fractional shortening (LVFS) recapitulated the identical intergroup ranking observed for LVEF on day 7 and day 28 (Fig. 7D). On day 7, MQgel@Mito treatment produced the highest LVFS (35.24%±4.55%) among all MI groups, far outperforming the Saline, MQgel, and Mito cohorts. On day 28, the MQgel@Mito group further differed from single-agent treatments, attaining an LVFS of 39.42%±6.89%, while separate treatment with MQgel or Mito generated negligible LVFS improvements between the two time points. Longitudinal comparison confirmed that progressive LVFS enhancement was uniquely detected in the MQgel@Mito group (Fig. S25B). Quantitative analysis of left ventricular chamber dimensions—including left ventricular internal diameter in diastole (LVIDd), left ventricular internal diameter in systole (LVIDs), left ventricular end-diastolic volume (LVEDV), and left ventricular end-systolic volume (LVESV)—further corroborated the outstanding anti-remodeling capacity of MQgel@Mito. Post-MI saline-treated rats developed severe ventricular dilation relative to Sham controls, reflecting maladaptive remodeling and progression toward dilated cardiomyopathy. At both time points, MQgel@Mito significantly attenuated ventricular dilation (Fig. 7E and F; Fig. S25C-H).

Fig. 7: MQgel@Mito improves cardiac function in a rat MI model. (A) Schematic illustration of animal experiments. (B) Representative images of M model echocardiography of different treatments on day 7 and day 28 post-MI (n = 5). The longer line indicates the LVIDd; the shorter line indicates the LVIDs. (C-F) Quantification of LVEF, LVFS, LVIDd, and LVIDs on day 7 and day 28 post-MI (n = 5). (G) Representative images of H&E and Masson's trichrome staining of the rat hearts on day 7 and day 28 post-MI (n = 5). (H-I) Quantification of cardiac collagen area (H) and LV wall thickness (I) on day 7 and day 28 post-MI (n = 5). Data are presented as mean ± SD; ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001.

The rat hearts were harvested on day 7 and day 28 for a series of investigations. Histological analysis using H&E and Masson's trichrome staining corroborated the findings detected by echocardiography. Saline-treated hearts displayed severe fibrosis with substantially increased collagen area (30.60% ± 2.59% on day 7, 37.58% ± 6.66% on day 28) and marked LV wall thinning (1.09 ± 0.08 mm on day 7, 0.81 ± 0.24 mm on day 28). All treatments reduced fibrosis and preserved wall thickness, but MQgel@Mito demonstrated the most pronounced effects, with the lowest collagen area (19.03% ± 1.94%) and thickest LV wall (1.68 ± 0.24 mm) on day 28 (Fig. 7G–I). Safety assessment by histological staining revealed no tissue damage or inflammation in liver, spleen, lung, or kidney tissues (Fig. S26). Blood biochemical analysis revealed that serum levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatinine (Cr), and urea remained within normal physiological ranges in MQgel@Mito-treated rats, with no significant differences compared to the Sham group (Fig. S27), indicating the absence of hepatic or renal toxicity. Overall, these results confirmed that MQgel@Mito facilitated cardiac function and attenuated the adverse remodeling following MI without influencing other organs.

MQgel@Mito attenuated inflammation and facilitated cardiac repair following MI

Tracking studies using MTR-labeled donor mitochondria demonstrated that the fluorescence of transplanted mitochondria gradually diminished over time. Donor mitochondria protected by MQgel exhibited slower fluorescence quenching, with MTR fluorescence still detectable on day 28 post-transplantation. In contrast, donor mitochondria without MQgel protection, once transplanted into cardiac tissue, showed a significantly faster rate of fluorescence quenching. These findings indicated that MQgel could prolong the in-situ retention of mitochondria following in vivo transplantation (Fig. 8A and B). Immunofluorescence images revealed the colocalization of transplanted mitochondria with CD68+ macrophages and cardiac troponin (cTnT)+ cardiomyocytes, respectively, confirming internalization by these two cell types (Fig. 8C and D).

Fig. 8: MQgel@Mito attenuates inflammation and facilitates cardiac repair after MI. (A) Representative fluorescence images of MTR-labeled donor mitochondria with or without MQgel (Mito vs MQgel@Mito) in cardiac tissue at different time points. (B) Quantitative analysis of MTR density in cardiac tissue at different time points (n = 3). (C) Representative fluorescence images of colocalization of MTR-labeled donor mitochondria and macrophage marker (CD68) on day 7 post-MI. (D) Representative fluorescence images of colocalization of MTR-labeled donor mitochondria and cardiomyocyte marker (cTnT) on day 7 post-MI. cTnT, cardiac troponin T. (E) Representative images of CD86/CD68 immunofluorescence staining on day 7 post-MI. (F) Quantitative analysis of the percentage of CD68+CD86+ pro-inflammatory (M1) macrophages of total CD68+ macrophages (n = 5). (G-H) Concentrations of IL-1β (G) and TGF-β (H) in the infarct border zone on day 7 post-MI (n = 3). (I-J) Serum concentrations of CK-MB (I) and cTnT (J) on day 28 post-MI (n = 3). CK-MB, creatine kinase MB. (K) Representative images of TUNEL staining of the infarct border zone of rat myocardium. (L) Quantitative analysis of TUNEL fluorescence ratio in cardiac tissue (n = 5). (M) Representative images of WGA staining. (N) Quantitative analysis of myocyte size (n = 10). Data are presented as mean ± SD; ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001.

Given the anti-inflammatory effects observed in vitro, we examined macrophage polarization in the infarct border zone. Immunofluorescence analysis revealed substantial infiltration of CD68+CD86+ pro-inflammatory (M1) macrophages following LAD ligation. Among total CD68+ macrophages, M1 percentage reached 62.02% ± 7.66% in the Saline group on day 7 post-MI. MQgel@Mito dramatically reduced this proportion to 17.38% ± 16.10%, substantially lower than MQgel (49.67% ± 4.02%) or mitochondria alone (43.17% ± 4.50%) (Fig. 8E and F). Cytokines in the infarct border zone confirmed these findings: pro-inflammatory cytokines IL-1β and TNF-α were markedly elevated at day 7, while anti-inflammatory TGF-β was reduced relative to the Sham group. MQgel@Mito exhibited the most pronounced anti-inflammatory effect and promoted myocardial repair (Fig. 8G and H; Fig. S28).

Subsequently, myocardial injury markers CK-MB and cTnT were significantly reduced in MQgel@Mito-treated animals compared to the Saline group (Fig. 8I and J). TUNEL staining revealed extensive apoptosis in the border zone of infarcted myocardium, consistent with hypoxia and energy depletion. MQgel@Mito substantially reduced TUNEL-positive cells (Fig. 8K and L). Furthermore, wheat germ agglutinin (WGA) staining demonstrated that cardiomyocyte cross-sectional area increased dramatically following MI (883.20 ± 106.30 μm2), representing more than 3.3-fold hypertrophy compared to the Sham group, indicating maladaptive compensatory remodeling. MQgel@Mito significantly attenuated this pathological hypertrophy, reducing cardiomyocyte area to 340.10 ± 57.65 μm2 (Fig. 8M and N). TEM imaging of the infarct border zone revealed severely impaired mitochondrial ultrastructure in saline-treated animals, characterized by swelling, disrupted cristae, and architectural disorganization. All treatments partially rescued cristae organization, with MQgel@Mito displaying the most well-preserved mitochondrial morphology (Fig. S29). Taken together, these findings demonstrated that MQgel@Mito facilitated cardiac repair by sustaining mitochondrial homeostasis, promoting inflammation resolution, and protecting cardiomyocytes from hypoxic injury.

Transcriptomic analysis confirmed metabolic regulation and anti-inflammatory effects

To gain insight into the molecular mechanisms underlying MQgel@Mito therapeutic efficacy, we performed bulk RNA sequencing (RNA-seq) on cardiac tissue from saline- or MQgel@Mito-treated MI rats. The three-dimensional principal component analysis (PCA) revealed clear separation between the two groups, with PC1 accounting for 75.77% of transcriptomic variance, indicating substantial global changes in gene expression (Fig. 9A). The Venn diagram delineated the unique and common expressed genes between the MQgel@Mito and Saline groups, among which 9859 genes were shared (Fig. 9B). DESeq2 was used to analyze differentially expressed genes (DEGs) (MQgel@Mito vs Saline), among which a total of 2356 upregulated and 1755 downregulated genes were found (Fig. 9C).

Fig. 9: Transcriptomic analysis confirms the metabolic and anti-inflammatory effects of MQgel@Mito. (A) Three-dimensional PCA plot (n = 3). PCA, principal component analysis. (B) Venn diagram (MQgel@Mito vs Saline). (C) Volcano plot of DEGs (MQgel@Mito vs Saline). The pink dots indicate to upregulated DEGs, and the blue dots indicate to downregulated DEGs. |log2FC | > 0.585, p value < 0.05. DEGs, differentially expressed genes; FC, fold change. (D) Clustered heatmap of the expression levels of DEGs associated with inflammation, energy metabolism, myocardial contraction, and apoptosis (MQgel@Mito vs Saline). p value < 0.05. (E) GO enrichment analysis of up- and downregulated DEGs in biological process categories (MQgel@Mito vs Saline). p value < 0.05. GO, gene ontology. (F) KEGG pathway enrichment analysis of up- and downregulated DEGs (MQgel@Mito vs Saline). p value < 0.05. (G) GSEA plots showing the enrichment of gene sets associated with NADH dehydrogenase activity and apoptosis (MQgel@Mito vs Saline). GSEA, gene set enrichment analysis; NES, normalized enrichment score. p value < 0.05; adjusted p value < 0.25; |NES| > 1.

Analysis of functionally relevant DEGs revealed patterns consistent with our mechanistic studies: genes associated with inflammation (Il3ra, Il1rap, Crlf1) and apoptosis (Casp3, Bcl2l11) were significantly downregulated, while genes associated with energy metabolism (Mt-nd4, Cox4i1, Ndufa8, Atp5me) and myocardial contraction (Tpm1, Actc1, Myh6) were significantly upregulated in the MQgel@Mito group (Fig. 9D). Gene ontology (GO) enrichment analysis identified 10 biological processes that were either upregulated or downregulated following MQgel@Mito treatment (Fig. 9E). Specifically, biological processes associated with mitochondrial respiration and ATP synthesis were upregulated, whereas those related to fibrillation, inflammatory response, and apoptosis were downregulated. Fig. 9F showed 5 upregulated KEGG pathways (pink) and 5 downregulated KEGG pathways (blue), which were consistent with and corroborated the results of the GO enrichment analysis. Among the KEGG pathways, oxidative phosphorylation and the TCA cycle were significantly upregulated. Downregulation of the HIF-1 pathway was closely associated with improved cellular metabolism and inhibited immune-inflammatory responses, while downregulation of the PI3K-AKT pathway was related to reduced cellular stress and alleviated pathological cardiac remodeling. Gene set enrichment analysis (GSEA) further demonstrated that MQgel@Mito upregulated NADH dehydrogenase activity while downregulating apoptosis-related gene sets (Fig. 9G). To validate the transcriptomic findings, we assessed myocardial ATP production and respiratory chain function in the infarct border zone on day 28 post-MI. Compared with saline, MQgel@Mito significantly enhanced the activities of Complex Ⅰ (NADH dehydrogenase) (Fig. S30A) and Complex Ⅴ (ATP synthase) (Fig. S30B), along with increased tissue ATP levels (Fig. S30C). These functional data are consistent with the upregulation of energy metabolism pathways and Complex Ⅰ-related gene sets identified by GO enrichment analysis (Fig. 9E) and GSEA (Fig. 9G).