Macrophage Membrane‐Cloaked, ROS‐Triggered Quercetin Nanocarriers Target Ovarian Lesions to Treat Polycystic Ovary Syndrome

Polycystic ovary syndrome (PCOS) involves oxidative stress-driven ovarian dysfunction and remains difficult to treat due to drug side effects and poor target engagement. We engineered a reactive oxygen species (ROS)-responsive, macrophage-membrane-camouflaged quercetin nano-therapy (MM@PCD@QNPs) to enhance ovarian delivery and mitigate toxicity. The core comprises a pinacol phenylboronate-dextran conjugate that encapsulates quercetin and undergoes ROS-triggered release; a surface M0 macrophage membrane confers immune evasion and lesion tropism via retained proteins (e.g., CD11b and CD47-SIRPα). MM@PCD@QNPs displayed nanoscale dimensions and stability (133.63 ± 14.60 nm; −33.13 ± 1.52 mV) and released drug under elevated ROS. In DHT-injured granulosa cells and a DHEA-induced PCOS mouse model, the formulation promoted granulosa cell proliferation, suppressed apoptosis, reduced ROS, and preferentially accumulated in ovaries, with negligible in vitro and in vivo toxicity. Transcriptomics and validation implicate activation of the MAPK7-Nrf2-NQO1 axis as a principal mechanism; pharmacologic MAPK7 inhibition abrogated therapeutic effects. By coupling ROS-triggered release with macrophage-mimetic targeting, MM@PCD@QNPs overcome quercetin's low bioavailability and off-target exposure and provide a safe, effective nanoplatform for PCOS therapy. 1 Introduction Polycystic ovary syndrome (PCOS) is a prevalent young-set reproductive endocrine disorder, affecting up to one in five women of childbearing age [1]. Defined by ovulatory dysfunction, hyperandrogenism, and polycystic ovarian morphology, PCOS is frequently accompanied by insulin resistance, dysmetabolism, and adverse pregnancy outcomes [1, 2]. Current therapies (hormonal modulation, antiandrogens, insulin sensitizers) alleviate symptoms but carry adverse effects, show variable long-term efficacy, and offer limited disease modification. Converging evidence implicates oxidative stress (OS), excessive reactive oxygen species (ROS) and impaired antioxidant defenses, as a central driver of granulosa-cell dysfunction, follicular arrest, mitochondrial injury, and ferroptosis [3-6]. Some natural plant extracts and Chinese traditional medicines have shown great potential in recuperation PCOS [7]. Quercetin (QUR), a dietary flavonoid, shows broad antioxidant and anti-inflammatory activity and has improved endocrine and ovarian phenotypes in PCOS models [8-10]. However, its clinical translation is restricted by unfavorable aqueous solubility, rapid metabolism, and low oral bioavailability, necessitating high doses that increase systemic exposure and reduce ovarian targeting [11]. These limitations motivate formulation strategies that both boost bioavailability and deliver quercetin to ovarian tissue while limiting off-target toxicity [12]. Functional nanomaterials overcome the inherent limitations of small molecule antioxidants to some extent. For example, melatonin-loaded polydopamine nanoparticles were used as ROS scavengers for the treatment of dry eye syndrome [13]. Resveratrol-loaded cerium dioxide nano-enzymes significantly improved hormone disruption and inflammation in dehydroepiandrosterone (DHEA)-induced PCOS mice [14]. Boron esters are reported to be oxidatively unstable in response to ROS, aiding in targeted drug release at disease sites [15-17]. Normally, 4-(hydroxymethyl) phenylboronic acid pinacol ester (PBAP) is stable, with its hydrophobic pinacol group protecting against hydrolysis. However, in the presence of excess free radicals, it facilitates borate bond cleavage, neutralizes ROS [18], which is often used as an antioxidant prodrug [17, 19, 20] due to its mild chemical reaction, easy conjugation, and low toxicity of metabolites. Therefore, the nanomaterials constructed by borate-modified polymers have stable ROS response effects. Despite extensive advances in the application of these materials, clinical approval is limited, mainly due to concerns related to biosafety, complexity of synthesis processes, and high costs. Recently, biomimetic nanoparticle delivery systems based on cell membranes or micro-vesicles exhibited promising potential in treating various diseases, including diabetes, non-alcoholic fatty liver disease, and anti-atherosclerosis [21-23] due to their high biocompatibility and plasticity, lesion targeting, and excellent immune evasion properties [24]. For example, red blood cell membrane-encapsulated nanoparticles could prolong blood circulation and reduce immunogenicity [25], while macrophage membrane-encapsulated nanoparticles specifically target inflammation and cancer lesions, avoiding clearance by the immune system and reducing drug toxicity and side effects [21]. However, the application of this technology in the treatment of gynecological diseases, especially PCOS, remains unexplored. Given the central role of oxidative stress in PCOS pathogenesis, this study utilized macrophage membranes and pinacol phenylboronic acid ester to develop a macrophage membrane-encapsulated ROS-responsive biomimetic nanoparticle delivery system loaded with QUR, to overcome the limitations of traditional QUR administration, and achieve precise targeting of polycystic ovary inflammatory lesions, representing an efficient and safe innovative treatment strategy for the clinical treatment of PCOS. 2 Results 2.1 Synthesis Process of MM@PCD@QNPs and Its Stable Characterization Network pharmacological analysis identified QUR as the promising ROS-responsive compound in treating PCOS (Figure S1). To facilitate its clinical translation, we developed a biomimetic nano-delivery system encapsulated within a macrophage membrane (MM). QUR-loaded PABP conjugated with dextran (DEX) polymer nanoparticles (PCD@QNPs) were fabricated using nanoprecipitation-dialysis across various PCD-to-QUR mass ratios (10:1, 5:1, 2:1, 1:1, and 1:2). Subsequent optimization of the MM coating was performed via ultrasonic hydration and liposomal extrusion at PCD@QNP-to-MM mass ratios of 100:1, 10:1, 1:1, and 1:10. Stability assessments showed that PCD@QNPs at 10:1 and 5:1 ratio maintained a stable, nanoscale size, whereas other ratios led to aggregation and precipitation (Figure 1A; Figure S2). The MM@PCD@QNPs at a 1:1 ratio demonstrated optimal stability in both size and zeta potential (Figure S3). The 5:1 PCD@QNPs were selected for further development, exhibiting a drug loading efficiency of 11.44 ± 0.84% and an encapsulation efficiency of 72.75 ± 2.79% (Figure 1B). These nanoparticles had a hydrodynamic size of 87.65 ± 4.14 nm and a zeta potential of -16.53 ± 0.90 mV. Coating with MM of equivalent quality increased the size to 133.63 ± 14.60 nm (Figure 1C) and shifted the zeta potential to -33.13 ± 1.52 mV (Figure 1D), reflecting the highly negative surface charge of the native membrane (-46.33 ± 2.00 mV). Given the high reactivity of aryl boronic esters with ROS, we investigated the drug release profile in response to hydrogen peroxide (H2O2) [26]. In a buffer containing 500 nM H2O2, PCD@QNPs released 69.69 ± 1.19% of QUR cumulatively over 7 days, significantly higher than the 58.00 ± 1.49% released in its absence. Under the same conditions, MM@PCD@QNPs showed cumulative release rates of 59.02 ± 1.49% and 53.86% ±1.22%, respectively (Figure 1E). which suggested H2O2 accelerated QUR liberation from the nanocarrier. The macrophage membrane coating has a certain delaying effect on ROS-responsive drug release. Figure S8 showed that MM@PCD@QNPs exhibited extremely high stability, with the lowest cumulative release rate at 24 h. The drug release amount of MM@PCD@QNPs was significantly increased under H2O2 condition, but still lower than that of non-MM coating NPs. The MM coating acted as a secondary barrier, slightly delaying release and providing a more controlled profile. These properties make MM@PCD@QNPs highly suitable for targeted drug delivery to the ROS-rich ovarian microenvironment in PCOS. Dynamic light scattering (DLS) in phosphate buffered saline (PBS) confirmed the strong colloidal stability of both formulations over three weeks, with PCD@QNPs degrading slightly earlier than their membrane-coated counterparts (Figure 1F). Fourier transform infrared spectroscopy (FTIR) revealed a characteristic peak at 1748 cm−1 (ester group, C = O stretching vibration) for both the synthesized PCDs and PCD@QNPs, while the stretching vibration peak of the carbon-oxygen double bond (carbonyl group) on the phenyl ring of QUR was located at approximately 1680 cm−1 (Figure 1G). Appropriate solvents based on the solubility of the substances were selected to perform proton nuclear magnetic resonance (1H-NMR) investigating the intermolecular non-covalent interactions. PCDs were dissolved in deuterated dimethyl sulfoxide (DMSO) /methanol (v/v = 1:1). Due to interference from the solvent peaks, the PCDs proton spectrum was concentrated between 1–8.5 ppm, with the signals at 3–4 ppm attributed to the pyran ring protons. This assay verified PBAP was successfully grafted onto the hydroxyl groups of DEX via chemical coupling, with a grafting efficiency of 70%, forming the PCDs nanocore, despite solvent peak interference (Figure 1H). Figure 1I depicted the 1H-NMR spectrum of QUR in deuterated DMSO, where signals >9 ppm corresponded to phenolic hydroxyl protons, while 6–8 ppm were attributed to aromatic ring hydrogens. Figure 1J showed that PCD@QNPs retained characteristic proton signals from both QUR and PCDs, but no phenolic hydroxyl peak was observed; PCDs exhibited no corresponding signals at 5.31–6.61 ppm, whereas QUR displayed aromatic ring hydrogens at 6.00–6.89 ppm. 2D 1H-NMR (Figure 1K) further revealed a correlation between 7.25 and 6.17 ppm protons, suggesting their mutual interaction. The results indicated successful self-assembly via non-covalent interactions and multi-functional group coordination. Characterization confirmed the integrity of the MM coating. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) demonstrated a protein profile for MM@PCD@QNPs similar to pure MM, and western blot analysis confirmed the presence of key surface proteins, including the M0 marker cluster of differentiation 11b (CD11b) and the immune evasion molecule signal regulatory protein α (SIRPα) (Figure 2A,B). Transmission electron microscope (TEM) revealed a spherical morphology for MM@PCD@QNPs, with a clearly visible MM layer encapsulating the nanoparticles (NPs) core, unlike the uncoated PCD@QNPs (Figure 2C). Previous studies have shown that immune cell membrane-mimetic NPs can evade capture by the monocyte-macrophage system, thereby reducing drug clearance [27-29]. In vitro uptake experiments showed that LPS-stimulated macrophages (murine macrophage cell line, Raw 264.7) phagocytosed dioctadecyl-3, 3, 39, 39-tetramethyl-1, 3-dicarboxycyanine perchlorate (DiD) labeled-nanoparticles (DNPs) readily, but significantly fewer macrophage membrane-coated DiD labeled-nanoparticles (MM@DNPs), compared with M0 cells (Figure 2D,E). Lipopolysaccharide (LPS)-induced Tsuchiya human monocytic leukemia cell line (THP-1) cells phagocytosed fewer MM@DNPs, with no significant difference compared with other groups (Figure S4A,B). Conversely, using a 500 nM dihydrotestosterone (DHT)-induced chronic inflammation [30], the uptake of MM@DNPs was significantly greater than that of DNPs in granulosa cells (primary mouse ovarian granulosa cells and human ovarian granulosa-like tumor cell line) (CP-M050 and KGN) (Figure 2F,G; Figure S4C, D) indicating enhanced targeting to the desired cells [31]. 2.2 The High Biosafety, Low Toxicity and Few Side Effects of MM@PCD@QNPs Cytotoxicity was assessed to determine safe working concentrations. Cell Counting Kit-8 (CCK-8) assays established 500 nM DHT as a suitable in vitro intervention concentration (Figure S4E), consistent with several previous studies [32]. Treatment with MM@PCD@QNPs (10 ng/mL to 100 µg/mL) for 24 h showed no significant cytotoxicity in KGN cells, though reduced viability was observed at 1 µg/mL and 100 µg/mL after 48 h (Figure 3A). Consequently, 10 µg/mL was selected for subsequent in vitro studies. Live-dead staining corroborated that MM@PCD@QNPs components were non-toxic to KGN cells compared to DHT treatment alone (Figure 3D). For in vivo safety, mice were intraperitoneally injected with MM@PCD@QNPs at concentrations of 1, 5, and 15 mg/mL at a rate of 3 µL per g of body weight (n = 3). Preliminary hemolysis analysis indicated concentration-dependent effects, with MM@PCD@QNPs exhibiting lower hemolysis rates than PCD@QNPs (Figure S4F,G). A concentration of 5 mg/mL was chosen for a two-week treatment in PCOS mice. At this dose, the hemolysis rate of all components was significantly lower than the positive control (H2O), and MM@PCD@QNPs showed a lower rate than free QUR (Figure 3B,C). Analysis of serum biomarkers for liver (alanine aminotransferase, aspartate aminotransferase) (ALT, AST) and kidney (blood urea nitrogen, creatinine) (BUN, CRE) function revealed no significant abnormalities (Figure 3E–H), and hematoxylin and eosin (H&E) staining showed normal histology in the heart, liver, spleen, lung, kidney, and uterus (Figure 3I). Furthermore, MM@PCD@QNPs treatment counteracted the dyslipidemia characteristic of PCOS, normalizing serum levels of (triglyceride) TG, TC (total cholesterol), LDL (low-density lipoprotein), and HDL (high-density lipoprotein) (Figure 3J–M). These results collectively demonstrated the high biocompatibility and low systemic toxicity of the developed nanocomposites, which provided a potential therapeutic strategy for PCOS. 2.3 MM@PCD@QNPs Improve Granulosa Cell Function in a Hyperandrogenic Environment by Promoting Proliferation, Inhibiting Apoptosis, and Reducing ROS Levels Cell proliferation and apoptosis are key functions of granulosa cells in the ovary. 5-ethynyl-2′-deoxyuridine (EdU) staining revealed that 500 nM DHT for 24 h significantly suppressed KGN proliferation, an effect that was reversed by co-treatment with MM@PCD@QNPs (Figure 4A,B). CCK-8 assays confirmed that both QUR and MM@PCD@QNPs significantly promoted cell proliferation after 72 h (Figure 4C). Flow cytometry analysis demonstrated that MM@PCD@QNPs effectively attenuated DHT-induced apoptosis, showing the lowest percentages of early and total apoptotic cells (Figure 4D,E). This was consistent with a decreased ratio of Cleaved-Caspase3 to Caspase3 (C-CAS3/CAS3) (Figure 4F,G). The anti-apoptotic effect of the nano-formulation was more potent than free QUR at an equivalent concentration. Since mitochondrial dysfunction and OS are hallmarks of PCOS, we assessed the mitochondrial membrane potential using 5, 5′, 6, 6′-Tetrachloro-1, 1′, 3, 3′-tetraethylbenzimidazolylcarbocyanine iodide (JC-1) staining. The DHT group exhibited severe mitochondrial depolarization, which was significantly restored by MM@PCD@QNPs (Figure 4H,I). Intracellular ROS levels, measured by 2′, 7′-dichlorodihydrofluorescein diacetate (DCFH-DA) fluorescence and flow cytometry, were markedly elevated by DHT but substantially reduced by MM@PCD@QNPs, with free QUR showing a more modest effect (Figure 5A–D). The activity of the antioxidant enzyme superoxide dismutase (SOD) was increased by PCDs, MM@PCDs, PCD@QNPs, and MM@PCD@QNPs, but only MM@PCD@QNPs significantly reduced the levels of malondialdehyde (MDA) (Figure 5E,F). Western blot analysis further showed that MM@PCD@QNPs significantly upregulated the expression of the antioxidant enzymes NAD(P)H:quinone oxidoreductase 1 and heme oxygenase-1 (NQO1 and HO-1) compared to the DHT group (Figure 5G,H), while the improvement effect of QUR at the same concentration was less significant. It also increased the protein expression of cytochrome p450 family 19 subfamily A member 1(CYP19A1), a key enzyme for estrogen synthesis (Figure 5G,H), indicating a potential restoration of hormonal balance. In addition, we preliminarily evaluated the regulatory effects of our NPs on inflammation and macrophage polarization at the cellular level. Figure 5I showed that interleukin-1 beta and interleukin 6 (IL-1B and IL-6) were significantly reduced in the MM@PCD@QNPs-treated group compared to the DHT-treated KGN group (Figure 5I). Flow cytometry and western blot analysis illustrated that MM@PCD@QNPs treatment significantly reduced cluster of differentiation 86 (CD86) expression and increased cluster of differentiation 206 (CD206) expression following LPS treatment (Figure 5J–M). 2.4 MM@PCD@QNPs Could Target Ovaries Lesions in a Hyperandrogenic Environment A PCOS-like model was established in mice via subcutaneously injecting 6 mg/100 g of dehydroepiandrosterone (DHEA) for 21 consecutive days, followed by intraperitoneal nanoparticle administration every three days for 2 weeks. While DHEA-induced mice gained significant weight, nanoparticle treatment did not significantly alter body weight compared to controls (Figure 6A,B). To evaluate the relative targeting of the NPs, we administered DiD-labeled NPs intraperitoneally and observed the effects in vivo (Figure S5) or ex vivo in vital abdominal organs at 1, 2, 6, 24, and 48 h after successful PCOS model establishment (n = 3). Figure S5 showed that the NPs initially accumulated within the peritoneal cavity, subsequently entering various organs through peritoneal microvessels and lymphatic vessels to exert their effects [33]. Compared to the PCD@QNPs group, MM@PCD@QNPs were cleared by the liver and spleen around 24 h later, and exhibited higher fluorescence signal intensity in the ovaries (Figure 6C). To further compare the effects of macrophage membrane encapsulation, DiD-labeled PCDs, MM@PCDs, PCD@QNPs, and MM@PCD@QNPs were intraperitoneally injected into mice for 24 h (n = 3). Ex vivo imaging revealed that the MM@PCDs and MM@PCD@QNPs groups exhibited relatively strong fluorescence signals in the ovaries (Figure 6D). Immunofluorescence co-staining for the granulosa cell marker follicle-stimulating hormone receptor (FSHR) and the DiD label revealed a stronger signal for MM@PCD@QNPs, suggesting that MM-coated NPs can be taken up more by ovarian granulosa cells than non-coated NPs (Figure 6E). These results suggested that macrophage membrane encapsulation may help improve the stability of NPs in vivo and, to some extent, alter their tissue distribution, thereby promoting their delivery to targeted organs. 2.5 MM@PCD@QNPs Improve the Symptoms of DHEA-induced PCOS Mice by Promoting Proliferation, Inhibiting Apoptosis, and Reducing ROS Levels For therapeutic assessment, a concentration of 5 mg/mL was selected based on a preliminary dose-ranging study (n = 3, Figure S4H–K). DHEA-induced PCOS mice exhibited classic polycystic morphology, hormonal imbalance, and disrupted estrous cycles (Figure 6K,L), while MM@PCD@QNPs treatment significantly ameliorated these pathologies: it reduced the number of cystic follicles, increased corpora lutea count, and decreased primordial follicles (Figure 6F,G). Hormonally, it lowered free testosterone (T) levels, normalized luteinizing hormone/follicle-stimulating hormone receptor (LH/FSH) ratio, and increased CYP19A1 protein expression, a key estrogen synthase, to some extent, modulating ovarian estrous cycling (Figure 6H,I). However, a notable enhancement in intraperitoneal glucose tolerance was not observed (Figure 6J). Crucially, this morphological and hormonal improvement translated to enhanced fertility. Under natural mating conditions, MM@PCD@QNPs significantly increased pregnancy rates and the number of embryos in pregnant PCOS mice (Figure 6M–O). Consistent with in vitro findings, MM@PCD@QNPs promoted ovarian proliferation evidenced by Ki67 staining (Figure 7A,B) and suppressed apoptosis in vivo, as shown by decreased levels of apoptosis-related markers (C-CAS/CAS3 and bcl2-associated x protein/ b-cell lymphoma 2) (BAX/BCL2), along with reduced apoptosis staining intensity (Figure 7C–F). Furthermore, MM@PCD@QNPs boosted the expression of HO-1 and NQO1 (Figure 7G–J), reduced fluorescence intensity of 8-ohdg (Figure 7K,L), confirming the activation of antioxidant defenses. In addition, we found that serum antioxidant and anti-inflammatory levels were reduced in PCOS mice, and MM@PCD@QNPs could significantly increase serum SOD levels and reduce MDA, IL6 and IL-1B levels (Figure 7M–P). 2.6 MM@PCD@QNPs May Exert Therapeutic Effects Through MAPK7/Nrf2/NQO1 To explore the potential mechanism of MM@PCD@QNPs in treating PCOS, granulosa cells from the negative control group, those treated with 500 nM DHT for 24 h, and those treated with DHT plus MM@PCD@QNPs (n = 3) were subjected to RNA transcriptome sequencing. Principal component analysis (PCA), volcano plots, and heatmaps illustrated distinct gene expression profiles among the groups (Figure 8A–D). A total of 607 differentially expressed genes (DEGs) were altered between the comparisons, with 60 overlapping genes respectively, with P 0.5 (Figure 8E). Gene ontology (GO) and kyoto encyclopedia of genes and genomes (KEGG) enrichment analyses identified key pathways, including the “extrinsic apoptotic signaling pathway, ” “reactive oxygen species metabolic process, ” and “regulation of mitotic nuclear division” (Figure 8F,G). Among the top intersecting genes, MAPK7 (DHT_vs_NC: log2FC = -1.45, DHT_vs_MM@PCD@QNP: log2FC = -1.30, P<0.05) was significantly downregulated by DHT and restored by MM@PCD@QNP treatment (Figure 8I). Protein-protein interaction (PPI) network analysis suggested a strong functional link between mitogen-activated protein kinase 7 (MAPK7), NFE2L2 (which encodes Nrf2), and NQO1 (Figure 8H). This led us to hypothesize that MM@PCD@QNPs act through the MAPK7-Nrf2-NQO1 axis. Western blot analysis confirmed that DHT suppressed the protein levels of phosphorylated mitogen-activated protein kinase 7 (p-MAPK7), MAPK7, Nrf2 (nuclear factor erythroid 2-related factor 2), and NQO1, which were effectively restored by MM@PCD@QNPs (Figure 9A,B). Immunofluorescence and nuclear-cytoplasmic fractionation further demonstrated that MM@PCD@QNPs promoted the translocation of Nrf2 from the cytoplasm to the nucleus (Figure 9C–F). As MAPK7 functions upon phosphorylation, we assessed its activation state in animal models, MM@PCD@QNPs increased the phosphorylation level of MAPK7, thereby activating the entire MAPK7-Nrf2-NQO1 signaling cascade (Figure 9G,H). To confirm the pathway's necessity, we inhibited MAPK7 phosphorylation using ERK5-IN-1 (Figure 10G,J). This inhibitor reversed the beneficial effects of MM@PCD@QNPs on proliferation, apoptosis, and ROS suppression (Figure 10A–G) in cellular models. In vivo, the inhibitor partially abrogated the therapeutic effects, leading to increasing number of atretic follicle and decreasing number of corpora lutea (Figure 10I), as well as elevating the expression of apoptotic proteins (C-CAS3/CAS3) (Figure 10J), though hormonal changes were not statistically significant (Figure 10H). This may be because ovarian hormone regulation relies on multiple cell types, and the effects of MAPK7 inhibitors were not strong enough. Western blot confirmed that MAPK7 inhibition downregulated Nrf2 and NQO1 expression in cell and animal models (Figure 10G,J). MAPK7 has been identified as a potential oncogene and prognostic marker, and its overexpression can promote the proliferation, invasion, and migration of ovarian cancer cells [34]. ROS, acting as a hub, can activate the MAPK signaling pathway and, by oxidizing cysteine residues in Keap1, trigger the release of Nrf2 from Keap1 and its translocation to the nucleus, activating the expression of antioxidant genes such as NQO1 [35], while the direct relationship between MAPK7 and ROS remains unclear. Our data preliminarily suggested that MM@PCD@QNPs ameliorated granulosa cell dysfunction and PCOS symptoms, at least in part, by activating the MAPK7-Nrf2-NQO1 signaling pathway (Figure 10K). 3 Discussion Based on the complexity and diversity of PCOS treatment [36, 37], this study innovatively developed a macrophage membrane-coated quercetin nano-delivery system (MM@PCD@QNPs) with low toxicity, minimal side effects, ROS-sensitive controlled drug release and relatively targeted delivery. It demonstrated that MM@PCD@QNPs effectively alleviate PCOS symptoms by activating the MAPK7-Nrf2-NQO1 pathway, promoting granulosa cell proliferation, inhibiting apoptosis, and reducing oxidative stress. This work suggests a safe, effective, and targeted treatment for PCOS and provides new insights into its molecular mechanisms, with potential for clinical application. QUR, although promise for PCOS [10], has shown some limitations in bioavailability, stability, and targeting, restricting its clinical application [7]. Generally, the oral bioavailability of QUR is only around 10%, and its poor water solubility renders intravenous injection extremely challenging [38]. Notably, the development of functional delivery systems has been proven to significantly improve the bioavailability of QUR [39]. To date, intragastric gavage has remained main administration route of QUR in PCOS models, with doses ranging from 15 to 100 mg/kg and treatment durations ranging from 1 to 45 days [11]. The development of QUR nano-therapy for PCOS remains inadequate. Thus, designing appropriate nano-delivery systems has become a pivotal step to enhance its therapeutic potential for PCOS. Nanoparticle drug delivery systems could overcome these challenges by promoting intracellular penetration and enhancing drug absorption [36, 37, 40]. The nano-delivery system we designed was a uniform-particle-size formulation coated with M0 macrophage cell membranes, with PCDs as the nanocore and QUR as the loaded drug. Aryl boronates, characterized by mild chemical reactivity, low toxicity, and ROS sensitivity, can be chemically coupled to DEX to form PCD nanocore. In vitro drug release experiments revealed that the presence of H2O2 enhanced the QUR release from PCD@QNPs, suggesting that our NPs possessed ROS-responsive properties. Different types of macrophages exert distinct functions: M0 macrophage cells primarily mediate immune evasion [41, 42], whereas M1-type ones exert anti-infective, anti-tumor and pro-inflammatory effects [43], and M2-activated ones are characterized by anti-inflammatory properties [44]. Notably, these membranes share common advantages-enhancing their circulation time and targeting efficacy [45, 46]. Our MM@PCD@QNPs preserved most M0 membrane proteins C11b and expressed SIRPα. These biomimetic NPs were less up-taken by LPS-induced macrophages and more by DHT-induced granulosa cells compared to uncoated NPs, suggesting potential for targeted delivery to lesions. As far as we know, the preferential accumulation of MM@PCD@QNPs at the inflammatory sites and their subsequent therapeutic action require the coordination of multiple components. First, membrane-bound proteins (cluster of differentiation 47 (CD47)/ SIRPα), programmed cell death protein 1/programmed cell death protein ligand 1, and integrin family) on MM act as a “stealth shield”, transmitting “don't eat me” signals to effectively evade phagocytosis by the systemic mononuclear-phagocyte system [47]. Subsequently, under inflammatory stimulation, integrins, selectins, and chemokine receptors on the MM exert synergistic effects to drive directional homing to inflamed sites [48, 49]. Cho et al. reported that high expression of ITGA2 on MM has been verified to facilitate nanoparticle migration toward ovarian cancer [50]. The ovarian microenvironment of PCOS is often accompanied by elevated ROS levels (such as O-2 and. OH) and inflammation [34, 35, 45, 51]. Upon reaching target organs, adhesion molecules on the MM specifically interact with ligands (fibronectin, clathrin) on target cells, such as human liver cancer cells HepG2 [52], human colorectal adenocarcinoma cell (Caco-2) [53], and human neuroblastoma SH-SY5Y cells [41], thereby triggering receptor-mediated endocytosis and promoting efficient nanoparticle internalization [51, 54, 55]. Our in vivo and ex vivo imaging results demonstrated, that MM@PCD@QNPs were cleared by the liver more slowly, achieving targeted accumulation in ovarian inflammatory tissue, to a certain extent. Subsequently, the PCD nano-shells underwent chemical reactions with excessive ROS, leading to ester bond cleavage and thus triggering the release of QUR [56]. As a result, PCDs and QUR exert a synergistic therapeutic effect. Impaired granulosa cell proliferation, excessive apoptosis and elevated ROS are common in PCOS [35, 57], and some free radical scavengers have been proven to alleviate PCOS symptoms by enhancing antioxidant enzymes and reducing inflammatory markers [7, 58]. Our MM@PCD@QNPs significantly mitigated granulosa cell damage and PCOS-related symptoms in mice, with a dosage far lower than conventional oral doses (100 mg/kg for 28 days) [11], showing high biocompatibility (low toxicity, minimal side effects). However, our PCOS model did not exhibit significant metabolic abnormalities-a finding consistent with previous studies-likely due to the absence of a high-fat diet regimen in our experimental design [34, 59, 60]. As is well established, intermittent administration can reduce the burden of drug administration and dampen the risk of immunogenicity. Furthermore, RNA transcriptomics suggests that the MAPK7-Nrf2-NQO1 signaling pathway may explain the potential therapeutic mechanism of MM@PCD@QNPs. MAPK7 (also known as ERK5), a downstream molecule in the MAPK family, plays a key role in cell proliferation, differentiation, and survival [61]. Previous literature has shown that the MAPK/ERK signaling pathway is closely associated with cell proliferation, apoptosis, and ROS levels in many diseases [62-64]. However, few studies have focused on the mechanisms involved in MAPK7. Nrf2 is a crucial transcription factor for the anti-oxidative stress response, which separates from Keap1, moves to the nucleus, and antioxidant response element (AREs) to regulate enzymes like NQO1, protecting cells from oxidative damage [65]. This study found that DHT treatment reduced MAPK7, Nrf2, and NQO1 expression in PCOS cells and mice, but MM@PCD@QNPs treatment restored these levels, suggesting it may alleviate PCOS symptoms via the MAPK7-Nrf2-NQO1 pathway. While MAPK7 inhibitor improved granulosa cell function, they did not significantly affect hormone production, possibly due to the far greater complexity of animal models than in vitro cell models, and in vitro results may not fully reflect the complex physiological and pathological processes in vivo [66]. In summary, these results suggest a promising initial advance in the mechanism of effective QUR delivery by MM@PCD@QNPs for PCOS. This study has achieved several innovative breakthroughs in the treatment of PCOS. First, to address the objective limitations of QUR, this study designed and synthesized a novel, low-toxicity, low-side-effect, and highly biosafe ROS-sensitive macrophage membrane-coated nano-delivery system. Second, we experimentally screened the optimal dosage of MM@PCD@QNPs, providing a certain reference for the clinical transformation of our nano-drug delivery systems. More importantly, this study innovatively proposed that the MAPK7-Nrf2-NQO1 signaling pathway may be a key pathway in the pathogenesis and treatment of PCOS, highlighting its potential role in cell proliferation, apoptosis, and ROS regulation. Furthermore, the supplemented fertility studies enrich the research landscape of PCOS therapeutics. These findings not only provide new insights and potential targets for PCOS treatment but also lay the foundation for future PCOS nano-drug development. However, there are some limitations to that need to be discussed. For example, only intraperitoneal injection was used to conduct experiments on the mouse PCOS model and did not directly compare it with the commonly used intravenous injection in clinical practice. Only cellular-level transcriptome sequencing was done, missing animal tissue insights. The therapeutic mechanism via the MAPK7-Nrf2-NQO1 pathway is not fully explored. Macrophage membranes derived from the laboratory are affected by factors such as cell culture conditions and extraction processes, making it difficult to ensure that each batch of cell membranes is completely consistent in protein composition and content, resulting in some batch effects. There is a lack of exploration on long-term reproductive safety. Our future work will focus on comparison of administration methods, establishing a standardized identification and extraction process for macrophage membrane proteins, more in-depth mechanistic research and reproductive safety evaluation to enhance understanding and clinical application. 4 Conclusion This study developed a novel, low-toxicity macrophage membrane-coated biomimetic QUR nano-delivery system with stable nanoscale size, ROS sensitivity, and targeted, sustained release. The MM@PCD@QNPs effectively improved PCOS symptoms by activating the MAPK7-Nrf2-NQO1 pathway, promoting granulosa cell growth, reducing apoptosis, and lowering ROS levels. This approach offers a safe, targeted treatment for PCOS and introduces the MAPK7-Nrf2-NQO1 pathway as a new mechanism for understanding and treating the disease. 5 Methods 5.1 Materials and Reagents Quercetin was purchased from MedChemExpress (New Jersey, USA), N,N′-carbonyldiimidazole (CDI), 4-dimethylaminopyridine (DMAP), and formamide were purchased from Shanghai McLean Biochemical Co., Ltd. (Shanghai, China); 4-(hydroxymethyl)phenylboronic acid pinol ester and dextran (molecular weight:3500) were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. Methanol, dichloromethane, anhydrous magnesium sulfate, hydrogen peroxide (H2O2), dimethyl sulfoxide, deuterated dimethyl sulfoxide (DMSO-d6), and deuterated methanol (methanol-d4) were purchased from Sinopharm Chemical Reagent Co., Ltd. Flasks, dialysis clips and dialysis bags (3500 kilodalton), magnetic rotors were purchased from Shanghai Yuanye Bio-Technology Co., Ltd. Cell membrane extraction kits were purchased from Beyotime Biotechnology Research Institute Co., Ltd. 1, 19-Dioctadecyl-3, 3, 39, 39-tetramethyl-1, 3-dicarboxycyanine perchlorate was purchased from Biotium Inc (Fremont, USA). 5.2 Extraction of Macrophage Membranes Raw 264.7 cells (ATCC, USA) were cultured to confluence in DMEM supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin at 37°C in a 5% CO2 atmosphere. Cells were counted using a counting chamber and 10*7 cells/dish were harvested for subsequent membrane extraction. Following the manufacturer's instructions, cells were washed with pre-chilled PBS (Servicebio, 0.01 M, pH 7.4) and lysed on ice for 15 min using Membrane Protein Extraction Reagent A containing 1% phenylmethanesulfonyl fluoride, and then triturated 50 times with a glass homogenizer. After ensuring sufficient cell disruption, the cell suspension was centrifuged at 3000 rpm at 4°C for 20 min to keep the supernatant and centrifuged at 13000 rpm at 4°C for 30 min to collect the cell pellet. After adding Membrane Protein Extraction Reagent B, the membrane protein solution was purified by brief vortexing and centrifuging at 14 000 rpm on ice for 5 min for three times and stored at −80°C. Protein concentration and quality were determined using the BCA assay (Elabscience, China). 5.3 MM@PCD@QNP Synthesis Dissolve 1947 mg of CDI and 1404 mg of PBAP in 20 mL of dichloromethane in a 25 mL dry round-bottom flask and stir at room temperature for 1 h. The collected organic phase was further washed with saturated saline for three times, dried over anhydrous magnesium sulfate, and then subjected to rotary evaporation in a vacuum dryer for 2 h to obtain activated CDI-PBAP. Then, dissolve 226.8 mg of DMAP, 135 mg of DEX, and 540 mg of CDI-PBAP in 15 mL of DMSO and stir magnetically at 37°C for 24 h to obtain the PABP conjugated with DEX polymer, which was further placed in an activated dialysis bag (MWCO of 3500 Da) and dialyzed against double-distilled water(overnight), and collected and centrifuged at 12, 000 rpm for 15 min to drying using a freeze dryer (Millrock Technology, USA). Different ratios of PCD and QUR were dissolved in formamide and methanol (1:1, v/v), and then added dropwise to double-distilled water under magnetic stirring, and finally dialyzed using a 3500 Da dialysis bag to remove unloaded drug and organic reagents, obtaining membrane-free PCD@QNPs. M0 macrophage membranes and the PCD@QNPs solution were sonicated on ice for 15 min at a mass ratio of 1:1, and then then repeatedly extruded through a liposome extruder (Avanti-610000, Merck, USA) containing a 200 nM filter to obtain a uniform MM@PCD@QNPs. The synthesis steps and exploration process were shown in Figures S2 and S3. 5.4 Cell Culture and Treatment Primary mouse ovarian granulosa cells (CP-M050, Procell, China), KGN, THP-1, and Raw 264.7 cells were cultured in DMEM/F12, 1640, and DMEM supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin, respectively, at 37°C in a 5% CO2 incubator. KGN cells were treated with 500 nM DHT (m6033, Abmole, China) for 24 h, and macrophages were treated with 100 ng/mL phorbol 12-myristate 13-acetate (PMA), followed by 100 ng/mL LPS (L8880, Merck, USA) for 24 h. 5.5 Animal Modeling All animal experiments were conducted in compliance with the Experimental Animal Ethics Committee of Wuhan University (approval number: 20240402B). A total of 91 female C57BL/6 mice (3 weeks old) and 18 male C57BL/6 mice (8 weeks old) were obtained from Jiangsu Jicui Pharmaceutical Biotechnology Co., Ltd. The animals were housed at 20 ± 5°C, 43–47% relative humidity, and a 12-h light-dark cycle, with a standard laboratory diet and water provided. Mice were housed for 3 days prior to the experiment to acclimate to the conditions. DHEA was dissolved in olive oil and ethanol (50:1) and injected subcutaneously for 21 consecutive days at a dose of 6 mg/100 g body weight to create PCOS mice as previous reported [14]. Control group received the same dose of olive oil injected subcutaneously. The animal experiments were divided into the following 7 groups: Control (olive oil via subcutaneous + normal saline via i.p), PCOS (PCOS modeling + normal saline via i.p), PCD (PCOS modeling + 15 mg/kg weight PCDs via i.p), MM@PCD (PCOS modeling + 15 mg/kg weight MM@PCDs via i.p), QUR (PCOS modeling +15 mg/kg weight QUR via i.p), PCD@QNP (PCOS modeling + 15 mg/kg weight PCD@QNPs via i.p), MM@PCD@QNP (PCOS modeling + 15 mg/kg weight MM@PCD@QNPs via i.p), ERK5-IN-1 (PCOS modeling + 15 mg/kg weight MM@PCD@QNPs via i.p + 1 mg/kg body weight ERK5 inhibitor via i.p). All animals were divided into three batches for evaluation of treatment efficacy, fertility, and drug distribution. Reference to previous studies [67], treatment was administered every three days for two weeks, and mice were sacrificed under isoflurane anesthesia. Blood samples were collected from the eyes, centrifuged, and stored at −80°C for biochemical analysis and ELISA. 5.6 Nanoparticle Characterization The chemical structures of PCDs and PCD@QNPs, as well as QUR, were characterized by 400 MHz 1H nuclear magnetic resonance (1H-NMR) spectroscopy on a Bruker Avance AVII-400 spectrometer (Bruker Corporation, Karlsruhe, Germany), for the purpose of analyzing hydrogen bonding interactions. Specifically, PCDs and PCD@QNPs were dissolved in a mixed solvent of DMSO-d6 and CD6OD (methanol-d4, 1:1, v/v), whereas QUR was solubilized in pure DMSO-d6 for spectral acquisition. The size and zeta potential of the NPs were measured using dynamic light scattering using a Malvern Zetasizer Nano ZS (Nano ZS 90, Malvern, UK). A drop of 10 µg/ml nanoparticle solution was deposited on a copper grid, dried, and negatively stained with 1% uranyl acetate [49]. The morphology and size of PCD@QNPs and MM@PCD@QNPs were determined using a TEM (LIBRA 200 FE, Zeiss, Germany) at 200 kV. FTIR (NICOLET 5700, Thermo Fisher Scientific, USA) was used to obtain infrared absorption spectra of the samples and identify key functional groups and chemical bonds. For the stability assay, 1 mL of a 10 µg/mL nanoparticle solution was mixed with 9 mL of 10 mM PBS (pH = 7.4) to monitor size and zeta potentials of the NPs over a period of three weeks. 5.7 Drug Loading and In Vitro Release Experiments 5.8 DiD-labeled NPs Synthesis and In Vitro Uptake PCDs, QUR, and DiD were dissolved in a formamide/methanol mixture (1:1, v/v) at a mass ratio of 7:2:1 and added dropwise into double-distilled water under magnetic stirring, followed by dialyzing (MWCO 3500 Da) to remove free drugs and organic solvents, ultimately yielding DiD-labeled NPs. These NPs were subsequently mixed with macrophage membranes at a 1:1 mass ratio, sonicated in an ice bath, and repeatedly extruded using a liposome extruder to obtain DiD-labeled NPs coated with macrophage membranes. Raw264.7 and THP-1 cells were incubated with 10 µg/mL DiD-labeled PCD@QNPs and MM@PCD@QNPs for 4 h, while primary mouse ovarian granulosa cells and KGN cells were cultured with the same formulations for 24 h, and then visualized under a confocal microscope (FV1200, Olympus, Japan), with three random fields of view used for analysis. 5.9 Transcriptome Sequencing (RNA-Seq) and Bioinformatics Analysis Following the reagent manufacturer's instructions (APExBIO, USA), the total RNA of NC, DHT and DHT+MM@PCD@QNP treatments (n = 3) for 24 h was extracted to construct libraries, and high-throughput sequencing was performed to obtain raw gene expression data. Bioinformatics analysis was used to identify differentially expressed genes, including GO enrichment analysis, KEGG pathway analysis and Protein-Protein Interaction Networks (PPI) networks. 5.10 Network Pharmacology The ChEMBL (https://www.ebi.ac.uk/chembl/) and SwissTargetPrediction databases (http://swisstargetprediction.ch/) were used to obtain the common predicted targets of quercetin, and the GeneCards database was used to obtain the predicted targets of PCOS. The intersection of the two predicted targets was used to draw a Venn diagram, and then Gene Ontology (GO) enrichment analysis, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed. 5.11 ROS Scavenging Detection The DCFH-DA probe was incubated with KGN cells at a dilution ratio of 1:1000 for 1 h. Intracellular ROS levels were recorded using the 488 channel of fluorescence microscopy and FITC channel of a flow cytometer. Following the protocol in the reagent manual, intracellular SOD enzyme activity (A001-3-2, NanJing JianCheng Bioengineering Institute, China) and MDA (A003-4-1, NanJing JianCheng Bioengineering Institute, China) were measured using a microplate reader (Perkinelmer, Massachusetts, USA), reflecting the nanomaterial's ability to scavenge ROS. 5.12 Hemolysis Analysis Fresh red blood cells from healthy mice were washed three times with PBS by centrifugation at 3000 rpm for 5 min to prepare a 4% red blood cell solution. 1 mL of nanomaterial sample of different concentrations and groups was added to 1 mL of a 4% red blood cell pellet and incubated at 37°C for 3 h, and then centrifuged at 3000 rpm for 10 min to observe whether hemolysis occurred. 100 µL of the supernatant was sampled and transferred to a 96-well plate to measure the absorbance at 577 nm (Perkinelmer, Massachusetts, USA) to quantify the degree of hemolysis. Double distilled water (positive control, complete hemolysis) and normal saline (negative control, no hemolysis) were set up for control to record OD values. Hemolysis rate (%) = (sample—negative control) / (positive control—negative control). 5.13 Proliferation and Viability Assay Cell Counting Kit-8 (E-CK-A362, Elabscience, China) was used to monitor cell proliferation. KGN cells were seeded in 96-well plates (100 µl, 3, 000 cells/well). After cell attachment, DHT and different NPs were treated for 24, 48, or 72 h. At each of these time periods, 10% CCK-8 reagent was added and incubated at 37°C for 1 h. The OD values at 450 nm was measured using a microplate reader. Following the manufacturer's instructions, 10 µM Edu reagent (E-CK-A377, Elabscience, China) was co-incubated with the cells for 2 h. Following fixation, permeabilization, fluorescent labeling, and DNA labeling, cell proliferation was observed under a fluorescence microscope at a wavelength of 594 nm. 200 µL of Calcein AM/PI working solution (E-CK-A354, Elabscience, China) was incubated with the cells at 37 °C for 30 min. The percentages of live and dead cells were observed under a fluorescence microscope at 488 nm (green fluorescence, live cells, Calcein) and 594 nm (red fluorescence, dead cells, PI), respectively. 5.14 Apoptosis Detection Flow cytometry was used to detect cell apoptosis using an Annexin V-FITC/PI detection kit (E-CK-A211, Elabscience, China) according to the manufacturer's instructions. Calibrate compensation for FITC/PI apoptosis flow cytometry by acquiring single-stained FITC+ and PI+ cell samples to eliminate spectral overlap between the two fluorescence channels. Early apoptotic cells were FITC-positive and PI-negative, while late apoptotic cells were both FITC- and PI-positive. And then, ovarian tissue slices are dehydrated, permeabilized, and co-incubated with tunel reaction solution (E-CK-A320, Elabscience, China), followed by DAB (3, 3′-diaminobenzidine) staining and nuclear counterstaining. Finally, after dehydration and mounting, apoptotic cells are observed under a microscope. Fix cells and perform JC-1 staining and imaging (C2003S, Beyotime, China) according to the instructions, where red represents aggregates and green represents monomers. 5.15 Immunofluorescence Cells were seeded onto coverslips or culture dishes or tissue sections were prepared to stain 8-hodg (GB150115, Servicebio, China), Nrf2 (T55136, Abmart, China) and FSHR (ab113421, Abcam, USA). After fixation and permeabilization with 0.3% Triton-X, cells were blocked with 5% bovine serum albumin (BSA) for 1 h, incubated with primary antibodies (overnight at 4°C), washed, and incubated with fluorescently labeled secondary antibodies (such as Alexa Fluor 488 or Alexa Fluor 594-labeled antibodies) at room temperature for 1 h in the dark. Finally, cell nuclei were counterstained with DAPI or Hoechst and imaged using a confocal laser scanning microscope (CLSM) or fluorescence microscope. 5.16 Western Blot A certain amount of tissue and cellular proteins were separated by SDS-PAGE and transferred to a polyvinylidene fluoride membrane (Millipore, USA). After blocking, the membrane was incubated with primary and secondary antibodies and finally visualized by chemiluminescence (BioRad, USA). The antibodies used were as follows: GAPDH (10494-1-AP, Proteintech, USA), BAX (AF0120, Affinity, China), BCL-2 (T40056, Abmart, China), CAS3 (caspase-3) (AF6311, Affinity, China), C-CAS3 (cleave-caspase-3) (GB11532, Servicebio, China), HO1 (43966, Cell Signal Technology, USA), NQO1 ((NAD(P)H quinone oxidoreductase 1)) (DF6437, Affinity, China), CYP19A1(16554-1-AP, Proteintech, USA), MAPK7 (DF6835, Affinity, China), phospho-MAPK7 (TA8146, Abmart, China), Nrf2 (T55136, Abmart, China), CD206 (18704-1-AP, Proteintech, USA), CD86 (ab220188, Abcam, USA). 5.17 Real-time Quantitative Polymerase Chain Reaction (RT-qPCR) Total RNA was extracted from cells or tissues and cDNA was synthesized using a reverse transcription kit (Vazyme, China). Subsequently, real-time quantitative PCR was performed using SYBR Green reagent (Vazyme, China) and normalized using housekeeping genes such as ACTB as internal controls to calculate the relative expression of the target gene. The primer sequences used were listed in Table S1. 5.18 Immunohistochemical Staining 5uM tissue sections were dewaxed, antigen retrieved, blocked, and then incubated according to the primary and secondary antibody incubation steps. Finally, they were developed with DAB and counterstained with hematoxylin. The results were observed under a microscope, and 3 fields of view were randomly selected for statistical analysis. The antibodies used and their dilution ratios were as follows: Ki-67 (GB121141, Servicebio, 1:300), NQO1 (1:100) and HO1 (10701-1-AP, Proteintech, 1:200). 5.19 Enzyme-linked Immunosorbent Assay (ELISA) ELISA kits for mouse IL-1B, IL-6, T, LH, and FSH were purchased from Shanghai Hualianke Biotechnology Co., Ltd. The kits were operated according to the manufacturer's instructions. Color changes were generated by enzymatic reactions, and absorbance values were read at specific wavelengths (usually 450 nm or 650 nm). 5.20 Analysis of Liver and Kidney Function Biochemical indicators were purchased from Nanjing Jiancheng Bioengineering Institute. ALT (C009-2-1) and AST (C010-2-1) were used to evaluate liver function, BUN (C013 - 2 - 1) and CRE (C011 - 2 - 1) were used to evaluate kidney function. TG (A110-1-1), TC (A111-1-1), HDL (A112-1-1) and LDL (A113-1-1) were used to assess obesity in mice. 5.21 Estrous Cycle Staining and Fertility Analysis Vaginal secretions from mice were collected using a clean cotton swab and placed on a slide at 8:00-9:00 AM daily, which lasted 10 days, covering approximately 2 estrus cycles. After fixing with 95% ethanol for 10 min, washing, staining with hematoxylin for 5 min, washing, and staining with eosin for 3 min, the samples were allowed to dry and then observed under a microscope. The estrous cycle stage was determined based on cell morphology, size, presence of nuclei, and the ratio of leukocytes to keratinocytes [67]. Based on previous literature, female and male mice were co-housed at a 2:1 ratio for one estrous cycle and then separated [69]. Nine days later, the female mice were observed for pregnancy and the number of embryos. 5.22 H&E Staining and Follicle Counting Following fixation in 4% paraformaldehyde, heart, liver, spleen, lung, kidney, uterus and ovarian tissues were embedded in paraffin, sectioned at 5 µm, and stained with H&E. Follicles were counted in every tenth section and categorized based on distinct morphological features into primordial, primary, and atretic and luteal stages, as previously described [67]. 5.23 Intraperitoneal Glucose Tolerance Test After fasting mice overnight, a 20% glucose solution was injected intraperitoneally. Blood samples were collected from the tail tip at different time points including 0, 30, 60, 90, and 120 min to test blood glucose levels using a glucometer. A blood glucose-time curve was plotted and the area under the curve (AUC) was calculated to assess glucose tolerance [14]. 5.24 In Vivo and Isolated Organ Imaging Mice were injected with DiD fluorescently labeled NPs at a dose of 15 mg/kg body weight, and in vivo and ex vivo imaging was performed at 1 h, 2 h, 6 h, 24 h, and 48 h. Fasting for 16 h, mice in different groups were anesthetized with gas or their organs were isolated and placed in a small animal 3D in vivo optical imaging system (IVIS Lumina III, PerkinElmer, USA) to capture the fluorescence signals, and normalized analysis was performed to quantify the fluorescence intensity. 5.25 Flow Cytometry of Macrophage Polarization PMA-induced M0 cells were pre-treated with LPS (100 ng/mL) for 24 h, then with normal medium, LPS, PCDs, MM@PCDs, QUR, PCD@QNPs or MM@PCD@QNPs for an additional 24 h. Cells were harvested, resuspended in staining buffer containing 1% BSA, and blocked for 30 min on ice in the dark to abrogate non-specific binding, followed by staining with FITC-conjugated anti-mouse CD86 antibody (1:100) for 45 min under the same conditions. After fixation and permeabilization with a commercial fixative-permeabilization buffer, cells were stained with PE-conjugated anti-mouse CD206 antibody (1:100) for another 45 min under the aforementioned conditions. FITC and PE single-stain tubes were prepared in parallel for channel compensation to eliminate fluorescence spillover. Following three washes with cold staining buffer, cells were resuspended in 300 µL of the buffer and analyzed by flow cytometry to determine CD86 and CD206 expression levels. 5.26 Statistical Analysis All statistical analyses followed standardized procedures. The data were presented as mean ± SD. For numerical variables, normality was tested before using standard ANOVA, Corrected Brown-Forsythe and Welch tests with post hoc multiple comparisons, or nonparametric Kruskal-Walli's test with post hoc test. For categorical variables, the chi-square test was used. All statistical graphs were generated using GraphPad Prism 8.0, OriginLab 2019b, and R (4.5.1). The abstract scheme and flowcharts were beautified using Figdraw (2.0). P < 0.05 was considered significant and was marked with *, P < 0.01 with **, and P < 0.001 with ***. Author Contributions W.Z.L. performed conceptualization (lead), data curation (lead), formal analysis (lead), investigation (lead), methodology (lead), visualization (lead), wrote-original draft, wrote-review, and edited; Y.G. performed investigation (supporting), wrote-original draft (supporting), wrote-review, and edited (supporting); N.S. performed investigation (supporting), methodology (supporting), visualization (supporting); Z.M.D. performed data curation (supporting), visualization (supporting). F.Z. performed investigation (supporting), wrote-review, and edited; T.L.Y. performed investigation (supporting), funding acquisition (supporting), resources (co-lead); Y.B.Y performed resources (co-lead), investigation (supporting), methodology (supporting); J.C. performed conceptualization (co-lead), wrote-review, and edited (supporting); W.Y.J. performed conceptualization (co-lead), methodology (supporting), wrote-review, and edited (supporting); Z.N.Y. performed conceptualization (lead), supervision (lead), wrote-review and editing (lead).

This study would like to thank the Central Laboratory of Wuhan University Renmin Hospital for providing the experimental platform and careful guidance. This study was supported by the grants from National Natural Science Foundation of China (Grant No.82571895, Grant No. 82271672), the Shenzhen Science and Technology Program (Grant No. JCYJ20240813111408012, Grant No. JCYJ20220530140609020). Open access publication funding provided by FCT (b-on). Ethic Statement All animal experiments were conducted in compliance with the Experimental Animal Ethics Committee of Wuhan University (approval number: 20240402B). Conflicts of Interest The authors declare no conflicts of interest. Data Availability Statement The databases used for network pharmacology are public and can be accessed publicly at https://www.ebi.ac.uk/chembl/, http://swisstargetprediction.ch/ and https://www.genecards.org/. The other data that support the findings of this study are available from the corresponding author upon reasonable request.