Abstract Abstract Ovarian cancer remains one of the most aggressive cancers, and resistance to Poly (ADP-ribose) Polymerase inhibitors (PARPi) poses a major therapeutic challenge. SIRT5, a NAD + -dependent desuccinylase, plays a crucial role in regulating fatty acid metabolism, which is often reprogrammed in cancer cells to promote drug resistance. This study aimed to investigate the potential of polydopamine (PDA)-polymerized antioxidant nanozyme-loaded SIRT5-modified human umbilical cord mesenchymal stem cells (hUCMSCs) to overcome PARPi resistance in ovarian cancer. We employed multi-omics approaches, including transcriptomics, metabolomics, and proteomics, to identify key molecular pathways associated with resistance mechanisms. High-throughput sequencing and metabolic profiling revealed that SIRT5 modifies fatty acid β-oxidation and regulates the desuccinylation of Enoyl-CoA Hydratase (ECHA), a key enzyme involved in this process. In vitro and in vivo experiments demonstrated that nanozyme-engineered hUCMSCs effectively enhanced PARPi resistance by promoting fatty acid metabolism and desuccinylation. These findings suggest that SIRT5-modified hUCMSCs loaded with antioxidant nanozymes offer a promising therapeutic strategy to combat PARPi resistance in ovarian cancer. The study provides new insights into overcoming drug resistance through metabolic reprogramming and enhances the potential of engineered stem cells in cancer therapy. Graphical Abstract [34]graphic file with name 12951_2025_3516_Figa_HTML.jpg [35]Open in a new tab Supplementary Information The online version contains supplementary material available at 10.1186/s12951-025-03516-6. Keywords: Antioxidant nanozyme-engineered stem cells, Multi-omics, Ovarian cancer, Poly (ADP-ribose) polymerase inhibitors resistance, SIRT5, Enoyl-CoA hydratase, Desuccinylation Highlights * This study first reports the use of SIRT5-modified nanozyme-loaded hUCMSCs to combat PARPi resistance in ovarian cancer. * We found that SIRT5 mediates resistance to PARPi by regulating fatty acid β-oxidation in ovarian cancer cells. * Proteomics analysis identified ECHA as the key protein involved in SIRT5-mediated desuccinylation, which enhances fatty acid metabolism. * Our research reveals that desuccinylation of ECHA promotes its expression and activity, contributing to PARPi resistance in ovarian cancer. * This study provides novel strategies and molecular targets for overcoming PARPi resistance in ovarian cancer, with potential therapeutic applications. Supplementary Information The online version contains supplementary material available at 10.1186/s12951-025-03516-6. Introduction Ovarian cancer is one of the most aggressive malignancies in the female reproductive system, characterized by high incidence and mortality rates [[36]1–[37]3]. According to global cancer statistics, the five-year survival rate for ovarian cancer remains low, particularly among patients with advanced-stage disease [[38]4, [39]5]. Currently, the primary treatments for ovarian cancer include surgical resection, chemotherapy, radiotherapy, and targeted therapy [[40]6–[41]8]. In recent years, advances in molecular biology and genomics have made targeted therapy increasingly significant in the treatment of ovarian cancer [[42]9–[43]11]. Among these therapies, Poly (ADP-ribose) Polymerase inhibitors (PARPi) are widely used because they can inhibit the DNA repair pathways in tumor cells. However, despite the clinical success of PARPi, the issue of resistance has emerged, significantly impacting treatment efficacy. This study aims to explore the therapeutic potential of polydopamine (PDA)-based antioxidant nanozyme-loaded SIRT5 gene-modified human umbilical cord mesenchymal stem cells (hUCMSCs) in reversing PARP inhibitor (PARPi) resistance in ovarian cancer. We hypothesize that the engineered antioxidant nanozyme-loaded hUCMSCs suppress PARPi resistance by silencing SIRT5 via shRNA, thereby inhibiting ECHA desuccinylation and fatty acid β-oxidation (FAO), ultimately overcoming resistance to PARPi. PARPi resistance has become a significant challenge in the treatment of ovarian cancer [[44]12–[45]14]. Studies have identified multiple mechanisms behind PARPi resistance, including alterations in DNA repair pathways, overexpression of drug efflux proteins, and dysregulation of apoptotic pathways [[46]13, [47]15, [48]16]. Among these, changes in DNA repair pathways are considered one of the primary reasons for PARP resistance [[49]13, [50]15, [51]17]. Specifically, tumor cells can evade the inhibitory effects of PARPi by restoring or enhancing homologous recombination repair (HRR) capabilities, thus reducing their dependency on PARPi [[52]18–[53]20]. Additionally, the overexpression of ATP-binding cassette (ABC) transporters, which can pump drugs out of cells, has been implicated as a crucial factor in PARPi resistance, decreasing the intracellular concentration of the drug [[54]21, [55]22]. In summary, the resistance mechanisms of PARPi are complex and involve multiple factors and pathways, highlighting the urgent need for new therapeutic strategies to overcome this challenge. Recent studies have shown that tumor cells undergo metabolic reprogramming to adapt to harsh environments, with abnormal regulation of fatty acid metabolism pathways potentially linked to tumor growth and drug resistance [[56]23–[57]25]. SIRT5, a NAD⁺-dependent desuccinylase, regulates mitochondrial protein post-translational modifications and plays a vital role in fatty acid metabolism and energy homeostasis [[58]26]. It has been reported to modulate the enzymatic activities of various metabolic enzymes through desuccinylation, particularly those involved in FAO [[59]24, [60]27, [61]28]. Moreover, SIRT5 may also act as a deacetylase, reshaping cellular energy metabolism by regulating key enzymes in FAO and fatty acid synthesis, such as CPT1A and ACC1/ACC2. Beyond its metabolic role, SIRT5 may function as a critical node mediating PARP inhibitor (PARPi) resistance by integrating DNA repair, energy metabolism, and oxidative stress regulation. In the context of DNA repair, SIRT5 promotes homologous recombination repair (HRR) efficiency by deacetylating RAD51, while simultaneously suppressing erroneous non-homologous end joining (NHEJ), thereby ensuring accurate DNA double-strand break (DSB) repair and circumventing the reliance on PARP-mediated single-strand break (SSB) repair. From a metabolic standpoint, SIRT5 activates CPT1A and ACADM, driving fatty acid β-oxidation (FAO) to supply ATP required for HRR. Additionally, it upregulates NAMPT, facilitating NAD⁺ synthesis to sustain PARP enzymatic activity [[62]29]. In terms of oxidative stress regulation, SIRT5 improves mitochondrial electron transport chain efficiency and activates antioxidant enzymes such as SOD2 and GPX4, thereby reducing reactive oxygen species (ROS) levels and protecting cells from drug-induced oxidative damage. These coordinated functions enable tumor cells to maintain genomic stability, energy supply, and redox homeostasis under PARPi treatment, ultimately resulting in the development of resistance. Conversely, targeting SIRT5 can simultaneously impair compensatory DNA repair, disrupt metabolic reprogramming, and diminish antioxidant defenses, offering a novel strategy for overcoming PARPi resistance. Indeed, studies have shown that SIRT5 inhibition significantly reduces HRR efficiency, depletes β-oxidation-derived ATP, and exacerbates oxidative stress, thereby restoring sensitivity to PARPi in resistant cells [[63]30]. Moreover, abnormal activation of SIRT5 has been associated with enhanced fatty acid catabolism, upregulation of drug efflux transporters, and inhibition of apoptosis, collectively conferring resistance to chemotherapeutic agents [[64]31, [65]32]. Therefore, SIRT5 may play a pivotal role in PARPi resistance in ovarian cancer by modulating desuccinylation of key metabolic enzymes and promoting energy adaptation. Omics technologies, including transcriptomics, metabolomics, and proteomics, have been widely applied in cancer research to comprehensively uncover the molecular mechanisms underlying tumor development and progression [[66]33, [67]34]. These technologies enable the systematic analysis of changes in gene expression, metabolites, and proteins, providing an integrated perspective on tumor biology. By combining these omics approaches, researchers can elucidate the multifaceted changes in gene expression, metabolic pathways, and protein modifications in cancer cells, which aids in understanding mechanisms of drug resistance [[68]35–[69]37]. For example, transcriptomics can reveal key gene expression changes in resistant cells, metabolomics can analyze characteristics of cellular metabolic reprogramming, and proteomics can provide functional regulatory information at the protein level. Integrating these omics data can identify new targets and strategies to overcome cancer drug resistance. In this study, we performed high-throughput transcriptome sequencing on parental and PARPi-resistant SKOV-3 cells and identified SIRT5 as a key candidate gene. Its role in PARPi resistance was confirmed via MTT, colony formation, and flow cytometry assays. Subsequent proteomic and metabolomic analyses identified the fatty acid β-oxidation pathway and its downstream enzyme ECHA (Enoyl-CoA Hydratase) as critical effectors. ECHA is a rate-limiting enzyme in FAO, and its desuccinylation by SIRT5 enhances its enzymatic activity. This modification facilitates the oxidative breakdown of fatty acids, providing sufficient ATP to sustain DNA repair under PARPi-induced stress. Thus, SIRT5-mediated activation of ECHA contributes to the development of PARPi resistance. The engineered hUCMSCs loaded with Mn₃O₄@PDA nanozymes and shRNA targeting SIRT5 exhibited effective tracking and antioxidant capability in vivo. Animal studies demonstrated that these engineered cells reversed PARPi resistance by suppressing FAO in tumor tissues. The scientific and clinical significance of this study lies in providing new insights and strategies for treating PARPi resistance in ovarian cancer, with potential application prospects. Materials and methods Cell culture Human ovarian cancer cell lines SKOV-3 (CVCL_0532) and OVCAR-3 (CVCL_0465) were obtained from ATCC (USA). SKOV-3 and OVCAR-3 cells were cultured in McCoy's 5A medium (16600082, Gibco, USA) supplemented with 10% FBS (10099141, Gibco, USA), 10 μg/mL streptomycin, and 100 U/mL penicillin (15140148, Gibco, USA). Subsequently, SKOV-3 and OVCAR-3 cells were exposed to a gradient of Olaparib concentrations (HY-10162, Med Chem Express, USA) for 72 h. Cell viability was assessed using the MTT assay, and colony formation ability was evaluated using the colony formation assay. Additionally, following previously reported methods [[70]38], we successfully established Olaparib-resistant SKOV-3 cell lines (SKOV-3/R) (Figure S2A) for further experiments. HEK293T cell lines were also obtained from ATCC (CRL-3216) and cultured in DMEM medium (11965092, Gibco, USA) supplemented with 10% FBS, 10 μg/mL streptomycin, and 100 U/mL penicillin. All cells were incubated in a humidified incubator (Heracell™ Vios 160i CR CO[2] incubator, 51033770, Thermo Scientific™, Germany) at 37 °C with 5% CO[2]. When cell confluence reached 80–90%, they were passaged. hUCMSC were purchased from Ningbo Mingzhou Biotechnology Co., Ltd. (MZ-2721, China) and cultured in a basal medium containing 10% FBS (MPM150312B, Ningbo Mingzhou Biotechnology Co., Ltd., China). Transcriptome high-throughput sequencing and data analysis We collected parental SKOV-3 cell samples (Sensitive group, N = 3) and SKOV-3/R cell samples (Resistance group, N = 3) for transcriptome high-throughput sequencing. The specific steps were as follows: Total RNA was extracted from each sample using Trizol reagent (T9424, Sigma-Aldrich, USA) according to the manufacturer's instructions. The RNA concentration, purity, and integrity were measured using the Qubit® RNA Analysis Kit ([71]Q32852, Life Technologies, USA) with a Qubit® 2.0 Fluorometer®, a Nanodrop spectrophotometer (IMPLEN, Germany), and the RNA Nano 6000 Analysis Kit (5067–1512, Agilent Technologies, USA) with a Bioanalyzer 2100 system. The A260/280 ratio was between 1.8 and 2.0. Each sample's total RNA content was 3 μg, used as input material for RNA sample preparation. Following the manufacturer's recommendations, we generated cDNA libraries using the NEBNext® Ultra™ RNA Library Prep Kit for Illumina (E7760S, genecompany, China) and assessed their quality with the Bioanalyzer 2100 system (Agilent Technologies, USA). Cluster generation for the indexed samples was performed on the cBot Cluster Generation System using the TruSeq PE Cluster Kit v3 cBot HS (Illumina). Finally, sequencing of the prepared libraries was conducted on the Illumina HiSeq 550 platform, producing 125 bp/150 bp paired-end reads. Sequencing data quality control and analysis The quality of raw sequencing data was assessed using FastQC (v0.11.8). Preprocessing of the raw data was performed with Cutadapt (v1.18) to remove Illumina adapters and poly(A) tails. Reads with more than 5% N content were filtered out using a Perl script. The FASTX Toolkit (v0.0.13) was employed to retain reads, with at least 70% of bases having a quality score above 20. BBMap (v39.01) was used for the correction of paired-end sequences. Finally, high-quality reads were aligned to the human reference genome using HISAT2 (v0.7.12). Differentially expressed genes (DEGs) between the Sensitive and Resistance groups were identified from the transcriptome sequencing data using the R package limma, with thresholds set at |log[2]FC|> 3 and adjp < 0.05. To further elucidate the functions of the DEGs, we conducted enrichment analyses using the R package clusterProfiler, with a significance threshold of p < 0.05. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed, categorizing GO terms into biological processes (BP), cellular components (CC), and molecular functions (MF). The results were visualized using bar plots and cluster dendrograms. Public data download and analysis RNA expression data (TPM) for tumor tissue samples from 427 ovarian cancer patients (TCGA-OV) were obtained from The Cancer Genome Atlas (TCGA) database ([72]https://portal.gdc.cancer.gov/). Additionally, transcriptome sequencing data for 88 normal ovarian tissue samples were downloaded from the Genotype-Tissue Expression (GTEx) database ([73]https://www.gtexportal.org/home/index.html). Since these data were obtained from public databases, ethical committee approval was not required. Based on the TCGA and GTEx data (TPM), we analyzed the expression levels of SIRT5, TFF3, DNALI1, and TFF1 in normal ovarian tissue samples (N = 88) and ovarian cancer tissue samples (N = 427). Using the TCGA-OV dataset, we conducted gene set enrichment analysis (GSEA) to evaluate the distribution of fatty acid metabolism pathways in the high and low SIRT5 expression groups. Furthermore, we analyzed the expression of SIRT5 in ovarian cancer cell lines and its protein expression in tumor tissue samples from ovarian cancer patients using the Human Protein Atlas (HPA) database ([74]https://www.proteinatlas.org/). Construction of lentiviral vectors for silencing and overexpression Based on GenBank data, potential short hairpin RNA (shRNA) target sequences were analyzed for human cDNA. Three sequences targeting SIRT5 were designed, along with a negative control (sh-NC) lacking interference sequences. The specific primer sequences are listed in Table S1. These oligonucleotides were synthesized by GenePharma ([75]C06001, GenePharma, China) and used to construct the lentiviral interference vector LV-1 (pGLVU6/GFP). The packaging virus and target vector were co-transfected into the human renal epithelial cell line HEK293T (with a confluence of 80–90%) using Lipofectamine 2000 (11668019, ThermoFisher, USA). After 48 h of culture, the supernatant was collected, filtered, and centrifuged to obtain virus particles. The virus titer was then determined from the collected supernatant. Cell transfection and grouping When ovarian cancer cells or mesenchymal stem cells (MSCs) reached the logarithmic growth phase, they were digested with trypsin and gently pipetted to prepare a cell suspension at a concentration of 5 × 10^4 cells/mL. The suspension was then seeded into 6-well plates, with 2 mL per well. Prior to grouping, lentiviruses (MOI = 10, viral titer = 1 × 10^8 TU/mL) were added to the cell culture medium containing the surfactant polybrene (TR-1003, Sigma-Aldrich, USA) and incubated for 48 h. Stable cell lines were selected using 2 µg/mL puromycin (HY-K1057, Med Chem Express, USA) for two weeks. After 48 h of transfection, RNA and protein levels were analyzed to verify the silencing efficiency. MSCs were divided into sh-NC and sh-SIRT5 groups. SKOV-3/R cells were divided into sh-NC and sh-SIRT5 groups and treated with 0, 5, or 10 µM Olaparib, followed by assays for cell viability and colony formation. The SKOV-3 and OVCAR-3 cells treated with DMSO or Olaparib were grouped as follows: (1) sh-NC, sh-SIRT5, oe-NC, and oe-SIRT5 groups; (2) sh-NC, sh-ECHA, oe-NC, and oe-ECHA groups; (3) oe-NC + sh-NC + DMSO, oe-SIRT5 + sh-NC + DMSO, oe-SIRT5 + sh-ECHA + DMSO, and oe-SIRT5 + sh-ECHA + Bezafibrate groups. Bezafibrate (HY-B0637, Med Chem Express, USA) was used at a concentration of 100 µM for 24 h. Metabolomics analysis Metabolomics analysis was conducted on SKOV-3/R cell samples from the sh-NC group (N = 6) and the sh-SIRT5 group (N = 6) using a combination of the LC20 Ultra-High-Performance Liquid Chromatograph (UHPLC, Shimadzu, Japan) and the Triple TOF-6600 Mass Spectrometer (AB Sciex). A Waters ACQUITY UPLC HSS T3 C18 column (100 × 2.1 mm, 1.8 μm) was employed for chromatographic separation. The column temperature was maintained at 40 °C, and the flow rate was set at 0.4 mL/min. The mobile phase consisted of acetonitrile (113212, Merck, USA) and water containing 0.1% formic acid (159002, Merck, USA). The gradient elution program for mobile phase B was as follows: 5% (0.0–11.0 min), 90% (11.0–12.0 min), and 5% (12.1–14 min). The eluate was directly introduced into the mass spectrometer without splitting. The mass spectrometry conditions for both positive and negative ion modes were as follows: ionization voltage of 5500 V, capillary temperature of 550 °C, nebulizer gas flow of 50 psi, and auxiliary heating gas flow of 60 psi. Orthogonal partial least squares-discriminant analysis (OPLS-DA) and permutation tests (100 permutations) were utilized to preprocess the data and prevent overfitting. Differential metabolites (DMs) were identified in the OPLS-DA model with a variable importance in projection (VIP) score > 1 and p < 0.05. Additionally, univariate analysis was performed, selecting DMs with a fold change ≥ 2 or ≤ 0.5 and a Student’s t-test p < 0.05 as the final DMs. MetaboAnalyst (v5.0) was used to identify the relevant metabolic pathways. Protein sample preparation and measurement The SKOV-3/R cell samples from the sh-NC group (N = 3) and the sh-SIRT5 group (N = 3) were transferred to 5 cm^3 centrifuge tubes. The samples were subjected to ultrasonic disruption in an ice bath using an ultrasonic cell disruptor (SCIENTZ-IID, Scientz, Ningbo, China). The disruption medium consisted of phenol extraction buffer containing 10 mM DTT (R0861, Solarbio, Beijing, China), 1% protease inhibitor mixture (P6731, Solarbio, Beijing, China), and 2 mM EDTA (E1170, Solarbio, Beijing, China). This step was repeated eight times. Next, an equal volume of pH 8.0 Tris-saturated phenol (HC1380, Van Gene Biotechnology, Beijing, China) was added, and the mixture was vortexed for 4 min. The samples were then centrifuged at 5000×g for 10 min at 4 °C, and the upper phenol phase was transferred to a new centrifuge tube. To the phenol solution, 0.1 M ammonium sulfate (101217, Merck, USA) in methanol (106035, Merck, USA) was added at a volume ratio of 1:5, and the mixture was left overnight to precipitate the proteins. The samples were centrifuged again at 4 °C for 10 min, and the supernatant was discarded. The remaining protein pellet was washed once with cold methanol and three times with cold acetone. The washed protein pellet was then dissolved in 8 M urea (U8020, Solarbio, Beijing, China), and the protein concentration was measured using a Bicinchoninic Acid (BCA) assay kit (P0012, Beyotime, China), following the manufacturer's instructions. Protein digestion, enrichment of succinylated lysine peptides, and nano-LC–MS/MS analysis For each sample, 50 µg of protein was digested. The protein solution was mixed with 5 mM DTT and incubated at 56 °C for 30 min. Subsequently, 11 mM iodoacetamide was added, and the solution was incubated at room temperature for 15 min. The urea concentration was then diluted to below 2 M, and trypsin was added at a ratio of 1:50 (w/w) for overnight digestion at 37 °C. An additional trypsin digestion was performed at a ratio of 1:100 (25200056, Thermo Fisher Scientific, USA) for 4 h. The digested peptides were desalted using HyperSep™ C18 columns (60108-302, Thermo Fisher Scientific, USA) and vacuum-dried. The peptides were reconstituted in 0.5 M TEAB and labeled according to the TMT kit protocol (90064CH, Thermo Fisher Scientific, USA). The labeled peptides were then pooled in equal amounts, fractionated using a high-pH reverse-phase peptide fractionation kit (84868, Thermo Fisher Scientific, USA), combined into 15 fractions, and dried before being reconstituted in 0.1% formic acid. Next, 40 µg of anti-succinyllysine antibody (Rabbit pan-specific antisuccinyllysine) (PTM-401, Biolab) was conjugated to 40 µL of protein A/G PLUS-agarose beads and incubated at 4 °C for 6 h. The supernatant was discarded, and the beads were washed three times with NETN buffer. The lyophilized tryptic peptides were dissolved in 500 µL of NETN buffer and incubated with the antibody-conjugated beads overnight at 4 °C with gentle agitation. The beads were then washed three times each with NETN and ETN buffers. The bound peptides were eluted with 1% trifluoroacetic acid, and the eluted solution was combined, lyophilized, and cleaned using C18 ZipTips before nano LC–MS/MS analysis. The digested peptides were reconstituted in 50 µL of 0.5% formic acid and 2% acetonitrile. A total of 100–200 ng of the sample was analyzed using nano LC–ESI–MS/MS. The analysis was performed on an LTQ-Orbitrap Elite mass spectrometer equipped with a CorConneX nano ion source coupled with a Dionex UltiMate 3000 RSLCnano HPLC system. The samples (5 µL) were injected at a flow rate of 20 µL/min into a PepMap C18 trap column for online desalting, followed by separation on a PepMap C18 analytical column. Peptides were eluted with a 120-min gradient of 5–38% acetonitrile in 0.1% formic acid at a flow rate of 300 nL/min, followed by a rapid increase to 90% acetonitrile in 0.1% formic acid for 5 min, and held for 7 min. The column was re-equilibrated with 2% acetonitrile in 0.1% formic acid for 25 min before the next run. The Orbitrap Elite was operated in positive ion mode with a nanospray voltage of 1.5 kV and a source temperature of 250 °C. Ultramark 1621 was used for calibration of the Fourier transform mass analyzer and the LTQ mass analyzer, with m/z 445.120025 as the lock mass for the FT mass analyzer. The instrument operated in data-dependent acquisition mode, performing a full MS survey scan followed by MS/MS scans of the top 15 most intense ions. Ions with a charge state of + 2 or higher and an intensity of over 10,000 counts were selected for MS/MS analysis using HCD with a normalized collision energy of 35%. The resolution for the MS survey scan was set to 60,000 with a mass range of m/z 375–1800, and the resolution for the MS/MS scan was set to 15,000 with a mass range of m/z 100–2000. Dynamic exclusion parameters were set to a repeat count of 1, a repeat duration of 30 s, an exclusion list size of 500, an exclusion duration of 60 s, and an exclusion mass width of ± 10 ppm. HCD parameters included an isolation width of 2.0 m/z, a normalized collision energy of 35%, an activation Q of 0.25, and an activation time of 0.1 ms. All data were acquired using Xcalibur 2.2 software. Proteomics data analysis All MS and MS/MS raw spectra were processed using Proteome Discoverer 1.4 (PD1.4, Thermo Fisher Scientific, USA). The peptide tolerance was set to 10 ppm, and the HCD MS/MS tolerance was set to 0.1 Da. Fixed modifications included carbamidomethylation of cysteine and light (+ 28.031 Da) or heavy (+ 36.076 Da) dimethylation at any N-terminus. Variable modifications included methionine oxidation, asparagine/glutamine deamidation, succinylation, and light/heavy dimethylation of lysine residues. Data filtering parameters were as follows: (i) FDR ≤ 1%, (ii) high confidence for peptide spectrum matches (PSMs), (iii) PSM delta Cn better than 0.15, and (iv) mass accuracy within 2 ppm. All MS/MS spectra of succinylated peptides near the scoring threshold in the initial database searches were manually inspected and verified using PD 1.4 and Xcalibur 2.2 software. For the relative quantification of succinylated peptides between light and heavy samples, precursor ions' peak areas corresponding to succinylated peptides were generated using the methyl duplex algorithm in PD 1.4. The peak areas of significant succinylated peptides in the raw data files were manually inspected using Xcalibur 2.2, with a mass tolerance of 5 ppm and mass accuracy reported to four decimal places. MTT Cells were seeded into 96-well plates at a density of 3–5 × 10^4 cells/mL and incubated for 48 h. MTT solution (CT02, Sigma Aldrich, USA) was added to the cell suspension and incubated for an additional 4 h. Following incubation, dimethyl sulfoxide (DMSO) was added, and the plates were shaken for 10 min to dissolve the formazan crystals. Absorbance was measured at 490 nm using a spectrophotometer (Laspec, China). Olaparib resistance was expressed as the half maximal inhibitory concentration (IC50). Colony formation assay For the plate colony formation assay, cells in the logarithmic growth phase were collected and digested using standard subculturing techniques to create a single-cell suspension with over 95% viability. The cells were counted and diluted to an appropriate concentration in a culture medium. A volume of 5 mL of this suspension, containing approximately 100 cells, was added to each 60 mm dish. The dishes were gently shaken in a cross pattern to ensure even cell distribution. The dishes were then incubated at 37 °C with 5% CO[2] for 2–3 weeks until visible colonies formed. The culture was then terminated, and the medium was discarded. The dishes were gently rinsed twice with PBS and air-dried. The cells were fixed with methanol for 15 min, the methanol was discarded, and the dishes were air-dried again. The cells were stained with Giemsa stain (HY-D0944, Med Chem Express, USA) for 10 min, followed by gentle rinsing with running water and air-drying. Colonies containing more than 10 cells were counted either visually or under a microscope (low magnification). The colony formation rate was calculated using the formula: Colony Formation Rate = (Number of Colonies / Number of Seeded Cells) × 100%. Flow cytometry Ovarian cancer cells (1 × 10^5 cells/well) were collected and washed in cold PBS. The cells were then stained in the dark for 15 min using an assay kit (APOAF-20TST, Sigma-Aldrich, USA). The pellet was resuspended in 400 μL of binding buffer, and 5 μL of Annexin V from the kit was added for staining. Flow cytometry was used to analyze the cells. The cell populations were identified as follows: Annexin V + PI + cells in the upper right quadrant represent late apoptotic cells; Annexin V + PI− cells in the lower right quadrant represent early apoptotic cells; Annexin V− PI + cells in the upper left quadrant represent necrotic cells; and Annexin V− PI− cells in the lower left quadrant represent live cells. Fatty acid β-oxidation assay When cells reached 80–90% confluence, a medium containing ^3H-labeled palmitic acid at a concentration of 0.5 μCi/mL was added. The cells were incubated at 37 °C in a 5% CO[2] incubator for 4 h to facilitate the uptake and oxidation of palmitic acid. After incubation, the medium was promptly removed, and the cells were washed three times with ice-cold PBS. Cells were lysed on ice using RIPA buffer (P0013C, Beyotime, China) for 30 min. The lysate was then centrifuged at 12,000 rpm for 10 min at 4 °C, and the supernatant was collected. A liquid scintillation counter was used to measure the radioactivity in the supernatant, and the oxidation rate of ^3H-labeled palmitic acid was calculated based on the radioactivity counts of the generated ^3H[2]O. An ATP assay kit (BC0300, Solarbio, Beijing, China) was used to lyse cells and determine the total cellular ATP. ATP was extracted according to the manufacturer's instructions and measured using a UV spectrophotometer (DU720, Beckman, USA). The ADP/ATP ratio was calculated. After the cells reached the appropriate confluence, they were treated with a medium inducing lipid droplet formation for 24–48 h. The medium was then removed, and the cells were incubated with a PBS solution containing 1 μM BODIPY 493/503 dye (HY-W090090, Med Chem Express, USA) at 37 °C for 15–30 min, protected from light. The cells were gently washed three times with PBS for 5 min each to remove unbound dye. Cells were fixed with 4% paraformaldehyde at room temperature for 10 min and washed three times with PBS. The nuclei were stained with 1 μg/mL Hoechst 33342 dye at room temperature for 5 min, followed by PBS washing. Fluorescence microscopy or confocal microscopy was used to observe and capture images of the lipid droplets, with BODIPY 493/503 having excitation/emission wavelengths of 493/503 nm. ImageJ software was utilized to quantitatively analyze the number and area of lipid droplets. Reverse transcription quantitative polymerase chain reaction (RT-qPCR) Total RNA was extracted using the Trizol reagent kit (T9424, Sigma-Aldrich, USA). RNA quality and concentration were assessed with an ultraviolet–visible spectrophotometer (ND-1000, Nanodrop, Thermo Fisher, USA). Reverse transcription was performed using the PrimeScript™ RT-qPCR kit (RR086A, TaKaRa, Mountain View, CA, USA). RT-qPCR was conducted using SYBR Green PCR Master Mix reagents (4364344, Applied Biosystems, USA) with the ABI PRISM 7500 Sequence Detection System (Applied Biosystems, USA). GAPDH was used as the internal control for mRNA quantification. Primers for the amplification were designed and provided by Shanghai General Biological Technology Co., Ltd. The primer sequences are listed in Table S2. The 2^−ΔΔCt method was employed to determine the fold change in gene expression between the experimental and control groups. The ΔΔCt calculation is as follows: ΔΔCt = ΔCt (experimental group) − ΔCt (control group), where ΔCt = Ct (target gene) − Ct (internal control). Western blot First, tissues or cells are collected and lysed using an enhanced RIPA lysis buffer containing protease inhibitors (P0013B, Beyotime, Shanghai, China). Determine the protein concentration using a BCA protein assay kit (P0012, Beyotime, Shanghai, China). Separate the proteins using 10% SDS-PAGE, and then transfer the separated proteins to a PVDF membrane. Block the membrane with 5% BSA at room temperature for 2 h to prevent non-specific binding. Incubate the membrane with diluted anti-succinyl-lysine antibodies and the primary antibodies listed in Table S3 (all primary antibodies are rabbit anti-human) for 1 h at room temperature. After washing the membrane, incubate it with HRP-conjugated goat anti-rabbit secondary antibody (ab6721, 1:2000, Abcam, UK) or goat anti-mouse secondary antibody (ab6785, 1:1000, Abcam, UK) for 1 h at room temperature. Mix equal amounts of Pierce™ ECL Western blotting substrate solutions A and B (32209, Thermo Scientific™, Germany) in the dark, and apply the mixture to the membrane. Capture images using a gel imaging system (BIO-RAD, USA). Analyze the grayscale intensity of the bands using ImageJ software, with β-actin as the internal control. Repeat each experiment three times. Immunoprecipitation (IP) The treated cells (sh-NC, sh-SIRT5, and FLAG-ECHA transfection groups) were lysed using RIPA lysis buffer containing protease and deacetylase inhibitors, incubated on ice for 30 min, with shaking every 10 min. The lysates were centrifuged at 12,000 rpm for 15 min, and the supernatants were collected for immunoprecipitation. Anti-ECHA (ab203114, 1:40, Abcam, UK) or anti-FLAG antibody (ab205606, 1:30, Abcam, UK) was added to the cell lysates and incubated with rotation at 4 °C overnight. Protein A/G magnetic beads were then added and incubated for 2 h to capture the antigen–antibody complexes. The beads were washed three times with cold PBS to remove non-specifically bound proteins. The bound proteins were eluted with SDS-PAGE loading buffer and denatured by boiling. The eluted protein samples were subjected to SDS-PAGE electrophoresis and transferred onto PVDF membranes. The membranes were blocked with 5% non-fat milk for 1 h to prevent non-specific binding. The PVDF membranes were then incubated sequentially with anti-succinyl lysine (PTM-401, Biolab), anti-ECHA, anti-SIRT5, and anti-FLAG antibodies (each for 2 h or overnight, according to the antibody instructions). The membranes were subsequently incubated with HRP-conjugated secondary antibodies for 1 h, and signals were detected using ECL chemiluminescence reagents. The Western blot results were analyzed using ImageJ to calculate relative densities. Co-immunoprecipitation (Co-IP) In HEK293T cells, co-express Flag-ECHA and V5-tagged SIRT5 (R960-25, Invitrogen, USA) and SIRT5-H158Y. An anti-Flag antibody for immunoprecipitation is used to isolate Flag-ECHA protein and determine whether V5-tagged SIRT5 and SIRT5-H158Y co-precipitate. Perform Western blot analysis using an anti-V5 antibody to detect immunoprecipitated V5-SIRT5, thereby assessing the interaction between ECHA and SIRT5. Expression and purification of ECHA and ECHB in Escherichia coli ECHA and ECHB were amplified from HEK293T cell cDNA and cloned into the NdeI/XhoI and EcoRI/SalI restriction sites of the pET-Duet and pET28a vectors, respectively. Sequencing-confirmed plasmids were transformed into BL-21 competent cells (D1013M, Beyotime, China). The cells were cultured in LB medium (ST165, Beyotime, China) containing ampicillin at 37 °C. When the OD600 reached 0.7, protein expression was induced with 300 μM isopropyl-β-D-thiogalactoside (IPTG) (ST098, Beyotime, China), followed by overnight incubation at 23 °C. Cells were harvested by centrifugation at 11,325 × g for 8 min, and the pellets were stored at − 80 °C until further use. Upon thawing, the cell pellets were resuspended in 100 mM Hepes buffer (pH 8.0) (C0215, Beyotime, China) containing 1 mM PMSF (ST505, Beyotime, China) and lysed using an EmulsiFlex-C3 cell disruptor. The lysate was supplemented with octyl β-D-glucopyranoside (ST2546, Beyotime, China) to a final concentration of 0.8% (wt/vol) and incubated at 4 °C for 20 min. The mixture was then centrifuged at 48,384×g for 30 min using a Beckman Coulter refrigerated centrifuge. The supernatant was loaded onto a Ni–NTA agarose column and washed with buffer containing 100 mM Hepes (pH 8.0), 30 mM imidazole, and 0.8% octyl β-D-glucopyranoside. Proteins were eluted with a linear gradient of 50–500 mM imidazole in the wash buffer. The eluted ECHA and ECHB protein complexes were analyzed using SDS-PAGE. Fractions containing relatively pure proteins were pooled, buffer-exchanged to 100 mM Hepes (pH 8.0), 0.8% octyl β-D-glucopyranoside, and 10% (vol/vol) glycerol, concentrated and stored at − 80 °C. The purification method for all ECHA mutants was similar. In vitro chemical succinylation of ECHA and ECHB complexes To perform the in vitro chemical succinylation, 50 nM of purified recombinant E. coli complex (ECHA and ECHB) was incubated with 3 mM succinyl-CoA (HY-137808, Med Chem Express, USA) in a mixture containing 100 mM Hepes (pH 7.4), 100 mM KCl, 10% (v/v) glycerol, 1 mM CoA, and 1 mM NAD at 27 °C for 20 min. Subsequently, the non-enzymatic succinylated ECHA and ECHB complexes were treated with SIRT5, and their enzymatic activities were measured. Measurement of ECHA activity The activity of ECHA was assessed using 2-(E)-decenoyl-CoA as the substrate. The combined activities of 2-ECHA and 3-hydroxyacyl-CoA dehydrogenase were measured by monitoring the formation of reduced nicotinamide adenine dinucleotide (Reduced form) (NADH) at 340 nm. The reaction mixture contained 100 mM Tris–HCl (pH 9.0) (T5941, Sigma-Aldrich, USA), 100 mM potassium chloride (KCl) (P9333, Sigma-Aldrich, USA), 100 μg/mL bovine serum albumin (BSA) (A2153, Sigma-Aldrich, USA), 1 mM free CoA (C3019, Sigma-Aldrich, USA), 120 μM nicotinamide adenine dinucleotide (Oxidized form) (NAD) (N7004, Sigma-Aldrich, USA), and 30 μM 2-(E)-decenoyl-CoA (SH-NF-1115, Nafovanny, China). The reaction was initiated by adding purified Flag-tagged ECHA. The increase in absorbance at 340 nm was monitored using a Ultraviolet–Visible (UV–Vis) spectrophotometer for 10 min. The enzyme activity of ECHA in sh-NC and sh-SIRT5 HEK293T cells, as well as the changes in ECHA activity after co-transfection of Flag-ECHA with SIRT5 or SIRT5-H158Y into HEK293T cells, was compared using this method. Additionally, to evaluate whether lysine acetylation regulates ECHA activity, the ECHA and ECHB complex was chemically acetylated by incubation with acetic anhydride. The same detection method was then repeated to observe any changes in ECHA activity post-acetylation. Preparation of Mn[3]O[4]@PDA nanozyme Mn[3]O[4] Nanozyme was synthesized as follows: Dissolve 20 mg/mL of manganese acetate tetrahydrate (Mn(Ac)2·4H2O) (221007, Sigma-Aldrich, USA) in anhydrous ethanol and place the solution in a polytetrafluoroethylene-lined stainless steel autoclave. Heat the autoclave at 120 °C for 24 h. After cooling to room temperature, wash the resulting brown product several times with ultrapure water and collect the Mn[3]O[4] Nanozyme precipitate for further use. Next, disperse 10 mL of 2 mg/mL Mn[3]O[4] (377473, Sigma-Aldrich, USA) in Tris–HCl buffer (pH 8.5). Gradually add this mixture to 10 mL of 2 mg/mL dopamine (DA) solution (73483, Sigma-Aldrich, USA), and sonicate the mixture in an ice bath for 10 min. Wash the Mn[3]O[4]@PDA Nanozyme product three times with water and then redisperse it in ultrapure water for future use. To facilitate the engineering of MSCs, further synthesis of FITC-labeled Mn[3]O[4]@PDA was conducted. Briefly, disperse 2 mL of 2 mg/mL Mn[3]O[4]@PDA in anhydrous ethanol, then add 1 mL of 2 mg/mL FITC solution. Stir the mixture at room temperature overnight under light-protected conditions. Finally, wash the FITC-labeled Mn[3]O[4]@PDA several times with water and redisperse it in water. Characterization of Mn[3]O[4]@PDA nanozyme Transmission electron microscopy (TEM) (JEM-1011, JEOL, Tokyo, Japan) was used to observe the nanozyme at an accelerating voltage of 200 kV. Fourier-transform infrared (FTIR) spectra were measured using an IRTracer-100 FTIR spectrometer (Shimadzu, Japan). Powder X-ray diffraction (XRD) patterns were recorded on a D8 Advance X-ray diffractometer (Bruker, Germany) using Cu Kα radiation. X-ray photoelectron spectroscopy (XPS) data were collected using a PHI-5000 VersaProbe XPS microscope (UIVAC-PHI, Japan). Zeta potential and hydrodynamic diameter were measured using a Zetasizer Nano-ZS90 particle size analyzer (Malvern, UK). Magnetic resonance imaging (MRI) measurements were conducted on a clinical MRI system (GE Discovery 750 W 3.0 T, USA), and relaxation rate analysis was performed using a nuclear magnetic resonance analyzer (mq60, the minispec, BRUKER). Activity of Mn[3]O[4] and Mn[3]O[4]@PDA nanozyme in removing reactive oxygen species (ROS) The hydrogen peroxide (H[2]O[2]) scavenging activity of Mn[3]O[4] and Mn[3]O[4]@PDA nanozymes was evaluated by measuring the color reaction of horse radish peroxidase (HRP) and 2,2'-Azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) (ABTS). In a typical assay, 3, 6, 10, and 20 µg/mL of Mn[3]O[4] and Mn[3]O[4]@PDA nanozymes were incubated with 100 µM H[2]O[2] at room temperature for 30 min. Subsequently, 10 µg/mL HRP (P6782, Sigma-Aldrich, USA) and 20 mM ABTS (HY-15902, Med Chem Express, USA) were added to the mixture. After 5 min, the absorbance at 740 nm was measured using a microplate reader. The decrease in absorbance indicated the H[2]O[2] scavenging rate. The hydroxyl radical (•OH) scavenging activity of the nanozymes was characterized using the methylene blue (MB) (HY-14536, Med Chem Express, USA) decolorization reaction. In a weakly acidic environment (1 mM PBS, pH 6.0), Fe^2+ reacts with H[2]O[2] to produce •OH radicals, causing MB decolorization. For this assay, 5 mM H[2]O[2], 1.8 mM Fe^2+, and varying concentrations of nanozymes (3, 6, 10, 20 µg/mL) were incubated together for 30 min. Then, 50 µM MB was added to the reaction mixture. After 10 min, the absorbance at 650 nm was measured using a microplate reader. The increase in absorbance during MB decolorization was used to calculate the •OH scavenging rate. The superoxide dismutase (SOD) mimetic activity of the nanozymes was measured using a commercial colorimetric SOD activity assay kit (CS0009, Sigma-Aldrich, USA). According to the manufacturer's instructions, Water-Soluble Tetrazolium (WST) working solution, ultrapure water, nanozymes (3, 6, 10, 20 µg/mL), and enzyme working solution were incubated at 37 °C for 20 min. The absorbance at 450 nm was then measured using a microplate reader. The decrease in absorbance was used to calculate the SOD mimetic activity. Mn[3]O[4]@pda nanozyme engineering of MSCs MSCs were seeded in 6-well plates at a density of 1 × 10^5 cells per well. After 24 h, the culture medium was replaced with fresh medium containing 10 µg/mL Mn[3]O[4]@PDA Nanozyme, and the cells were incubated at 37 °C in a 5% CO[2] atmosphere. Cells were collected and observed at 0, 1, 4, 8, 12, and 24 h. To facilitate the identification of internalized nanozyme, FITC-labeled Mn[3]O[4]@PDA was prepared. MSCs were then labeled with Mn[3]O[4]@PDA-FITC and characterized using fluorescence microscopy and flow cytometry. To verify the internalization of the nanozyme in MSCs, the labeled cells were fixed with 5% glutaraldehyde at 4 °C overnight and examined via biological TEM (Bio-TEM). Additionally, Mn[3]O[4]@PDA-engineered MSCs (E-MSCs) were analyzed using MRI to further characterize the internalized nanozyme. The subcellular localization of Mn[3]O[4]@PDA Nanozyme was studied through co-staining with organelle-specific dyes. First, 2 × 10^4 MSCs (sh-NC or sh-SIRT5) were cultured in 10 mm confocal dishes for 12 h. Then, the cells were incubated with a medium containing 10 µg/mL FITC-labeled Mn[3]O[4]@PDA for 12 h. After incubation, the cells were stained with 50 nm MitoTracker Red (HY-D1783, Med Chem Express, USA) at 37 °C for 20 min, gently washed with PBS, and stained with 1 µg/mL Hoechst 33342 (HY-15559, Med Chem Express, USA) (blue) for 5 min. Finally, the fluorescence of the cells was observed using a laser scanning confocal microscope (LSCFM). To simulate oxidative stress conditions, MSCs were exposed to culture medium containing 200 µM H[2]O[2] and co-cultured with various concentrations of nanozyme (0, 3, 6, 10 µg/mL) for 24 h. After washing with PBS, the cells were incubated with medium containing 0.5 mg/mL MTT at 37 °C for 4 h to assess cell viability and evaluate the protective effect of the nanozyme on MSCs. ROS scavenging was monitored using the DCFH-DA probe (HY-D0940, Med Chem Express, USA). Briefly, MSCs were seeded in 24-well plates at a density of 5 × 10^4 cells per well and cultured overnight. The cells were then stimulated with H[2]O[2] solution containing 10 µg/mL Mn[3]O[4]@PDA Nanozyme for 1 h. After washing three times with PBS, the cells were stained with 10 µM DCFH-DA at 37 °C for 30 min. The cells were washed again with PBS, and fluorescence images were captured using a fluorescence microscope to detect the DCFH-DA reaction products with ROS, using excitation/emission wavelengths of 488/510–560 nm. In vivo animal experiments Four 6-week-old BALB/c nude mice (weighing 17–18 g) and seventy 6-week-old SPF-grade nude mice (weighing 17–18 g) were purchased from the Animal Research Center of our institute. All animal experiments were conducted in compliance with the animal care policies of the Chinese National Medical Products Administration (SFDA). The animals were cared for humanely following the standards outlined in the "Guide for the Care and Use of Laboratory Animals," published by the U.S. National Institutes of Health (NIH). The mice were housed in sterile cages with sterile food and water, and all procedures were conducted under aseptic conditions, with room temperature maintained at 24–26 °C and humidity at 40–60%. Subcutaneous tumor model Four BALB/c nude mice were selected to establish a subcutaneous tumor model. SKOV-3 cells in the logarithmic growth phase were harvested and prepared into a cell suspension with a density of 2 × 10^7 cells/mL. Each mouse received a 0.2 mL subcutaneous injection of the cell suspension into the neck region. After confirming that there was no leakage at the injection site, the mice were returned to their cages. The size of the subcutaneous tumors was regularly measured using calipers to monitor growth. When the tumor diameter exceeded 1 cm, the tumors were excised for subsequent orthotopic transplantation experiments. Establishment of orthotopic transplantation model Six-week-old SPF-grade nude mice (weighing 17–18 g) were acclimated to the experimental environment for one week. Mice-bearing subcutaneous tumors were euthanized via cervical dislocation, and the neck tumors were harvested. After removing the capsule and connective tissues, tumors were sectioned into approximately 1 mm^3 pieces and placed in ice-cold phosphate-buffered saline (PBS) for later use. The recipient mice were anesthetized with sodium pentobarbital and secured on a surgical table. An abdominal incision was made to expose the ovaries under a dissecting microscope (Nikon SMA800). The prepared tumor fragments were directly implanted into the right ovary using 8–0 absorbable sutures for closure. The abdominal incision was then closed in layers with sterile silk sutures. Post-surgery, the mice were placed on a warm pad, given anti-infection treatment, and housed under SPF conditions [[76]39]. To exclude potential non-specific effects of treatment, DMSO and saline were used as solvent controls for Olaparib and the nanozyme formulations, respectively. These controls were incorporated throughout in vivo experiments and are described in the Methods section to ensure experimental rigor. Two weeks post-transplantation, when the tumor volume reached approximately 50 mm^3, the mice were randomly assigned to either the DMSO group or the Olaparib group. The Olaparib group received intraperitoneal injections of Olaparib at a dosage of 50 mg/kg daily for two consecutive weeks, while the DMSO group served as the control [[77]38]. The mice in the DMSO and Olaparib groups were further randomly divided into four groups (n = 5 per group): oe-NC + sh-NC + DMSO group (injected with SKOV-3 cells transfected with oe-NC and sh-NC lentivirus, treated with DMSO or Olaparib for the first two weeks, and administered DMSO orally for the subsequent two weeks); oe-SIRT5 + sh-NC + DMSO group (injected with SKOV-3 cells transfected with oe-SIRT5 and sh-NC lentivirus, treated with DMSO or Olaparib for the first two weeks, and administered DMSO orally for the subsequent two weeks); oe-SIRT5 + sh-ECHA + DMSO group (injected with SKOV-3 cells transfected with oe-SIRT5 and sh-ECHA lentivirus, treated with DMSO or Olaparib for the first two weeks, and administered DMSO orally for the subsequent two weeks); oe-SIRT5 + sh-ECHA + Bezafibrate group (injected with SKOV-3 cells transfected with oe-SIRT5 and sh-ECHA lentivirus, treated with DMSO or Olaparib for the first two weeks, and administered Bezafibrate orally for the subsequent two weeks). The Bezafibrate dose was 200 mg/kg, administered once daily for two weeks, with the control group receiving an equivalent volume of DMSO [[78]40]. Additionally, the Olaparib group mice were further randomly divided into six groups (n = 5 per group): Control group (treated with Olaparib); Mn[3]O[4]@PDA group (treated with Olaparib, followed by tail vein injection of Mn[3]O[4]@PDA); Mn[3]O[4]@PDA-MSCs group (treated with Olaparib, followed by tail vein injection of Mn[3]O[4]@PDA-MSCs); Mn[3]O[4]@PDA-MSCs-sh-SIRT5 group (treated with Olaparib, followed by tail vein injection of Mn[3]O[4]@PDA-MSCs-sh-SIRT5); Mn[3]O[4]@PDA-MSCs-sh-SIRT5 + DMSO group (treated with Olaparib, followed by tail vein injection of Mn[3]O[4]@PDA-MSCs-sh-SIRT5, and administered an equivalent volume of DMSO orally); Mn[3]O[4]@PDA-MSCs-sh-SIRT5 + Bezafibrate group (treated with Olaparib, followed by tail vein injection of Mn[3]O[4]@PDA-MSCs-sh-SIRT5, and administered Bezafibrate orally). The Mn[3]O[4]@PDA was injected at a volume of 100 µl with a concentration of 10 µg/ml once per week for two weeks. The Mn[3]O[4]@PDA-MSCs and Mn[3]O[4]@PDA-MSCs-sh-SIRT5 were injected at a volume containing 5 × 10^5 MSCs, once per week for 2 weeks. Four weeks later, the CRi Maestro in vivo imaging system (Cambridge Research & Instrumentation, Massachusetts, USA) was used to analyze the bioluminescent signals from luciferase-expressing ovarian cancer cell lines implanted in mice. The luciferase labeling process was as follows: the SKOV-3 cell line was transfected with an expression vector containing the firefly luciferase gene (luc2) (E1320, Promega, USA). Stable luciferase-expressing ovarian cancer cell lines were established through G418 antibiotic selection. Transfection efficiency and luciferase expression were verified using fluorescence microscopy and a bioluminescence imaging system [[79]41]. After four weeks, the mice were euthanized using CO[2], and the tumors were excised and measured. Tumor volume was calculated using the formula: tumor volume (mm^3) = d^2 × D / 2, where d is the smallest diameter and D is the largest diameter. Tumor tissues were then collected for Western blot and immunohistochemical analyses. Serum biochemical analysis Peripheral blood was collected from mice, and liver function indicators, including serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST), were measured using a Vitros 5,1 FS automated biochemical analyzer (Ortho Clinical Diagnostics, USA). Additionally, kidney function indicators, such as serum blood urea nitrogen (BUN) and creatinine (Cr), were also assessed. H&E staining Mouse tissue sections were dewaxed in water following the instructions of the H&E staining kit (PT001, Shanghai Bogoo Biotechnology Co., Ltd., China). The sections of the heart, liver, spleen, lungs, and kidneys were stained using the following steps: Hematoxylin staining at room temperature for 10 min, followed by rinsing under running water for 30–60 s; differentiation in 1% hydrochloric acid alcohol for 30 s, followed by rinsing in running water for 5 min; Eosin staining at room temperature for 1 min; dehydration through a graded ethanol series (70%, 80%, 90%, 95%, and 100%) with each step lasting 1 min; clearing in phenol xylene for 1 min, and in xylene I and II for 1 min each; mounting with neutral balsam in a fume hood. Finally, the slides were observed under an optical microscope (BX50, Olympus Corp., Tokyo, Japan) to examine tissue morphological changes. Oil Red O staining The tumor tissues were excised from the mice and immediately placed in 4% paraformaldehyde for fixation at room temperature for 24 h. The tissue samples were then dehydrated through a graded series of ethanol and embedded in paraffin. Once the paraffin had solidified, the tissues were sectioned and mounted on slides. After deparaffinization and rehydration, the sections were stained with Oil Red O solution at room temperature in the dark for 10 min. Following staining, the sections were briefly immersed in 70% ethanol to remove any non-specific staining and then rinsed three times with distilled water. The cell nuclei were counterstained with hematoxylin solution for 1 min at room temperature. Finally, the slides were coverslipped, and the distribution and quantity of lipid droplets were observed and photographed under a microscope. Statistical software and data analysis methods We utilized R version 4.2.1 for our statistical analysis, compiled through the RStudio integrated development environment (IDE), version 2022.12.0–353. For file processing, we employed Perl version 5.30.0. Additionally, we used GraphPad Prism version 8.0 for graphical and statistical analyses. We expressed quantitative data as mean ± standard deviation. For comparisons between the two groups, we conducted independent sample t-tests. For comparisons among multiple groups, we used one-way analysis of variance (ANOVA), and for comparisons across different time points, we employed two-way ANOVA. Post hoc tests were performed using the Bonferroni correction. A significance level of p < 0.05 was considered statistically significant. Results Transcriptomics analysis indicates that SIRT5 may improve ovarian cancer PARPi resistance by regulating fatty acid metabolism Ovarian cancer is one of the most prevalent and deadly gynecologic malignancies. Although targeted therapies such as PARP inhibitors (PARPi) have significantly improved patient outcomes, the emergence of resistance remains a major clinical challenge [[80]42]. To investigate the transcriptomic differences between PARPi-resistant and non-resistant ovarian cancer, we collected samples from parental SKOV-3 cells (Sensitive group, N = 3) and SKOV-3/R cells (Resistance group, N = 3) and conducted high-throughput transcriptome sequencing. Using a threshold of |log[2]FC|> 3 and adjp < 0.05, we identified 68 differentially expressed genes (DEGs), with 39 genes upregulated and 29 genes downregulated in the resistant cells (Fig. [81]1A–C). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses revealed that these DEGs are primarily enriched in pathways related to fatty acid metabolism, particularly fatty acid oxidation, which may be associated with PARPi resistance (Fig. [82]1D, [83]E). Emerging evidence supports targeting lipid metabolism to improve the efficacy of PARPi in triple-negative breast cancer (TNBC), further validating our research direction [[84]43]. Fig. 1. [85]Fig. 1 [86]Open in a new tab Screening of genes related to ovarian cancer PARPi resistance. A–C Heatmap (A), volcano plot (B), and differential gene ranking plot (C) of the differential analysis between the Sensitive group (N = 3) and the Resistance group (N = 3) in SKOV-3 cell samples. D, E Bar plot (D) and clustering dendrogram (E) of GO and KEGG analysis for the 68 DEGs. F GSEA analysis of fatty acid metabolism pathway distribution in SIRT5 high-expression and low-expression groups (N = 427). G Expression levels of SIRT5, TFF3, DNALI1, and TFF1 in normal ovarian tissue samples (N = 88) and ovarian cancer tumor tissue samples (N = 427). H Immunohistochemistry results for SIRT5 in ovarian cancer tumor tissue samples. I Expression of SIRT5 in ovarian cancer cell lines. *** indicates p < 0.001 Notably, the DEGs with the highest |log[2]FC| values were SIRT5, TFF3, DNALI1, and TFF1, which were significantly upregulated in the Resistance group compared to the Sensitive group (Fig. [87]1B, [88]C). Further analysis using the TCGA-OV dataset, consisting of 427 tumor samples, revealed a strong association between SIRT5 and the fatty acid metabolic pathway through GSEA analysis (Fig. [89]1F). Compared to normal ovarian tissue samples, SIRT5 expression was significantly upregulated in ovarian cancer tumor samples, while TFF3 expression was significantly downregulated, which contrasts with the previous trend observed in resistance-related differential expression analysis (Fig. [90]1B). DNALI1 and TFF1 did not show significant differential expression (Fig. [91]1G). Additionally, SIRT5 positivity was observed in ovarian cancer patient tumor samples (Fig. [92]1H), and SIRT5 was widely expressed in ovarian cancer cell lines (F[93]ig. [94]1I). SIRT5 has been shown to promote resistance to cisplatin in ovarian cancer [[95]44] and to sunitinib in renal cancer [[96]45]. However, its association with PARPi resistance has not been reported. This knowledge gap provides a rationale for selecting SIRT5 as the central focus of our study. Therefore, we selected SIRT5 as the focus of our research. SIRT5 drives ovarian cancer cell resistance to olaparib To investigate the impact of SIRT5 on ovarian cancer PARPi resistance, we first constructed SKOV-3 and OVCAR-3 cells with either silenced or overexpressed SIRT5. Specifically, we designed three shRNA sequences targeting SIRT5, verified their silencing efficiency, and selected the most effective sequence, sh-SIRT5-1, for subsequent experiments (Figures S1A-B). Similarly, we used lentiviral vectors to construct ovarian cancer cells overexpressing SIRT5 (oe-SIRT5), with oe-NC serving as the control, and confirmed the transfection efficiency using RT-qPCR and Western blot techniques (Figures S1C-D). We then treated the cells with gradient concentrations of Olaparib for 72 h. The results from MTT and colony formation assays indicated that silencing SIRT5 increased the sensitivity of SKOV-3 and OVCAR-3 cells to Olaparib, whereas overexpressing SIRT5 enhanced the resistance of ovarian cancer cells to Olaparib (Fig. [97]2A–C). Flow cytometry analysis showed that overexpression of SIRT5 significantly reduced the proportion of apoptotic cells after Olaparib treatment (Fig. [98]2D, [99]E). Additionally, SIRT5 overexpression markedly decreased the levels of cleaved caspase 3 and cleaved PARP1 post-Olaparib treatment (Fig. [100]2F, [101]G). These results suggest that SIRT5 mediates the sensitivity of ovarian cancer cells to PARPi. Fig. 2. [102]Fig. 2 [103]Open in a new tab The effect of SIRT5 on PARPi resistance in ovarian cancer cells. A MTT assay of cell viability in different concentrations of Olaparib after 72 h of treatment; B Colony formation assay of SKOV-3 cells (10 μM) and OVCAR-3 cells (2 μM) after Olaparib treatment; C Quantitative analysis of colony formation capacity in each group; D Flow cytometry analysis of apoptosis in SKOV-3 cells (10 μM) and OVCAR-3 cells (2 μM) after treatment with DMSO or Olaparib; E Quantitative analysis of apoptosis rates in each group; F Western blot analysis of cleaved caspase 3, cleaved PARP1, and SIRT5 protein expression in SKOV-3 cells (10 μM) and OVCAR-3 cells (2 μM) after treatment with DMSO or Olaparib; G Quantitative analysis of cleaved caspase 3 and cleaved PARP1 protein expression. All cell experiments were repeated three times. *p < 0.05, **p < 0.01, ***p < 0.001 We further examined the effect of SIRT5 on Olaparib sensitivity in Olaparib-resistant SKOV-3 cells (SKOV-3/R). Both SKOV-3/R and parental SKOV-3 cells were exposed to various concentrations of Olaparib for 72 h, and IC50 values were calculated. We found that the IC50 value for parental SKOV-3 cells was 7.45 µM, while for SKOV-3/R cells, it was 67.82 µM (Figure S2A). SIRT5 mRNA and protein levels were significantly higher in SKOV-3/R cells compared to parental SKOV-3 cells (Figures S2B-D). Moreover, SKOV-3/R cells showed little response to 5 µM and 10 µM Olaparib treatment, whereas silencing SIRT5 restored their sensitivity to Olaparib (Figures S2C-E). Silencing SIRT5 also significantly impaired the colony-forming ability of SKOV-3/R cells following Olaparib treatment (Figures S2F-G). Therefore, it is evident that SIRT5 drives the resistance of ovarian cancer cells to Olaparib, and silencing SIRT5 restores the sensitivity of Olaparib-resistant cells to the drug. Metabolomics analysis reveals that silencing SIRT5 alters the fatty acid β-oxidation pathway in olaparib-resistant cells Previous results have demonstrated a close association between Olaparib resistance in ovarian cancer cells and fatty acid metabolism. Silencing SIRT5 can partially restore the sensitivity of resistant ovarian cancer cells to Olaparib. To investigate whether SIRT5 exerts its effects by regulating metabolic pathways, we collected SKOV-3/R cell samples from both the sh-NC group (N = 6) and the sh-SIRT5 group (N = 6) for detailed metabolite analysis using liquid chromatography-tandem mass spectrometry (LC–MS/MS). The overall analysis process is illustrated in Fig. [104]3A. Fig. 3. [105]Fig. 3 [106]Open in a new tab Metabolomics analysis of SIRT5-silenced SKOV-3/R cell samples. A Schematic diagram of the metabolomics workflow in this study; B Volcano plot of serum metabolite differences between the sh-NC group (N = 6) and the sh-SIRT5 group (N = 6), with red representing upregulated metabolites, blue representing downregulated metabolites, and gray indicating metabolites with no significant difference; C OPLS-DA-PCA-2D plot of sample data from the sh-NC group (N = 6) and the sh-SIRT5 group (N = 6), with the x-axis representing the T score and the y-axis representing the Orthogonal T score; D Functional enrichment analysis of differential metabolites in the MetaboAnalyst database; E Fatty acid β-oxidation rate in cells from each group; F ATP production levels in cells from each group; G ADP/ATP ratio in cells from each group; H Lipid droplet formation assessed by BODIPY 493/503 staining, Scale bar = 20 μm; I Quantification of lipid droplet formation in cells from each group. All cell experiments were performed in triplicate. * indicates p < 0.05, ** indicates p < 0.01, *** indicates p < 0.001 We first performed PCA on the mass spectrometry data using EZinfo software, and differential metabolites were screened based on VIP > 1. The data were then corrected using the MetaboAnalyst platform, with fold change and p-values calculated, setting thresholds at |log2FC|> 2 and p < 0.05. This analysis identified 36 differential metabolites, including 23 downregulated and 13 upregulated (Fig. [107]3B), including six key FAO-related metabolites: Acetyl-CoA (log2FC = − 2.8), Palmitoylcarnitine (C16:0) (log2FC = − 3.1), Myristoylcarnitine (C14:0) (log2FC = − 2.5), Succinyl-CoA (log2FC = − 2.7), Long-chain fatty acyl-CoA (LCFA-CoA) (log2FC = − 3.0), 3-Hydroxybutyrate (3-HB) (log2FC = 2.4). Using the OPLS-DA algorithm, we calculated the VIP scores for each metabolite and generated the OPLS-DA-PCA-2D plot (Fig. [108]3C). Finally, pathway enrichment analysis revealed that the differential metabolites were predominantly enriched in the fatty acid β-oxidation pathway (Fig. [109]3D). Notably, SIRT5 silencing led to a significant reduction in Acetyl-CoA and LCFA-CoA levels, accompanied by accumulation of the incomplete β-oxidation product 3-HB, indicating impaired FAO flux and disrupted energy/lipid catabolism in resistant cells. Fatty acid β-oxidation is a primary source of ATP production, crucial for the survival and proliferation of cancer cells. During fatty acid β-oxidation, long-chain fatty acyl-CoA (LCA CoA) is shuttled into the tricarboxylic acid (TCA) cycle, where it is converted to acetyl-CoA and undergoes several rounds of β-oxidation to generate ATP. Previous studies have indicated that fatty acid β-oxidation may be enhanced in drug-resistant tumor cells, as these cells require increased energy and metabolic flexibility to sustain their growth and survival [[110]46, [111]47]. To evaluate the effect of SIRT5 silencing on fatty acid β-oxidation in Olaparib-resistant cells, we utilized the oxidation of ^3H-labeled palmitic acid to measure ATP production and calculate the ADP/ATP ratio. Additionally, inhibiting fatty acid catabolism typically results in the accumulation of lipids in the form of lipid droplets within the cytoplasm due to reduced fatty acid utilization. Therefore, we also assessed lipid droplet formation using BODIPY 493/503 staining. The results indicated that compared to the sh-NC group, the sh-SIRT5 group exhibited reduced palmitic acid oxidation and ATP production, an increased ADP/ATP ratio, and heightened lipid droplet formation (F[112]ig. [113]3E–I). These findings align with the metabolomics analysis, supporting the hypothesis that SIRT5 modulates fatty acid β-oxidation, thereby contributing to Olaparib resistance. Proteomics analysis reveals ECHA as a key protein in SIRT5-mediated desuccinylation SIRT5, an NAD + -dependent desuccinylase, regulates various metabolic pathways through the desuccinylation of key proteins [[114]48]. To further elucidate the molecular mechanisms by which SIRT5 influences Olaparib resistance, we aimed to identify the critical proteins involved in SIRT5-mediated desuccinylation. Using proteomics techniques, we analyzed SKOV-3/R cell samples from both the sh-NC group (N = 3) and the sh-SIRT5 group (N = 3). The analysis included the reduction dimethylation of tryptic peptides followed by the enrichment of succinylated peptides, as outlined in Fig. [115]4A. Equal amounts of total lysates from the sh-NC and sh-SIRT5 groups were labeled with heavy and light dimethyl tags, respectively. The labeled peptides were then combined and enriched using an anti-succinyl-lysine polyclonal antibody. Subsequent nano-liquid chromatography (LC)-MS/MS analysis compared the levels of succinylated peptides between the two groups. The MS analysis identified 115 succinylated proteins potentially regulated by SIRT5. Fig. 4. [116]Fig. 4 [117]Open in a new tab Proteomics analysis of SIRT5-mediated desuccinylation and mechanism validation. A Schematic diagram of the proteomics workflow used in this study; B Metabolic pathways enriched with lysine-succinylated proteins; C Distribution of the number of lysine-succinylation sites per protein; D Immunoprecipitation of ECHA from sh-NC and sh-SIRT5 groups using an ECHA-specific antibody; E Succinylation levels of Flag-ECHA in sh-NC and sh-SIRT5 HEK293T cells; F Succinylation levels of Flag-ECHA co-transfected with SIRT5 or its catalytic mutant SIRT5-H158Y in HEK293T cells; G Co-overexpression of Flag-ECHA and V5-tagged SIRT5 in HEK293T cells, with immunoprecipitation of Flag-ECHA reducing V5-tagged SIRT5; H Comparison of ECHA activity between sh-NC and sh-SIRT5 groups; I Comparison of ECHA activity in different groups of HEK293T cells; J Restoration of ECHA enzyme activity by SIRT5 desuccinylation, with WT and Succ-WT representing non-enzymatically and enzymatically succinylated ECHA and ECHB complexes, respectively. Succ-WT + SIRT5 indicates succinylated ECHA and ECHB complexes treated with SIRT5; K Chemical acetylation of ECHA and ECHB complexes incubated with acetic anhydride; and subsequent activity assessment. All cell experiments were performed in triplicate. * indicates p < 0.05, ** indicates p < 0.01, *** indicates p < 0.001 We then performed GO and KEGG enrichment analyses. Consistent with previous reports, several metabolic pathways, including fatty acid metabolism, the TCA cycle, oxidative phosphorylation, and lysine degradation, were significantly enriched in the SIRT5 desuccinylation targets (Fig. [118]4B). The number of succinylation sites on each protein varied significantly (Fig. [119]4C). Using proteomics, we identified numerous SIRT5 desuccinylation substrates, with ECHA, a protein involved in fatty acid oxidation, being the primary substrate in SKOV-3/R cells due to its high number of succinylation sites. Western blot analysis with an anti-succinyl lysine antibody revealed that ECHA was highly succinylated in the absence of SIRT5 (Fig. [120]4D). Similarly, in HEK293T cells, Flag-tagged ECHA (Flag-ECHA) showed increased succinylation in the sh-SIRT5 group compared to the sh-NC group (Fig. [121]4E). When Flag-ECHA was co-transfected with either SIRT5 or its catalytic mutant SIRT5-H158Y into HEK293T cells, ECHA succinylation levels decreased with SIRT5 co-expression but remained unchanged with SIRT5-H158Y co-expression (Fig. [122]4F). Additionally, co-expression of Flag-ECHA and V5-tagged SIRT5 demonstrated that Flag-ECHA could immunoprecipitate V5-SIRT5, indicating an interaction between ECHA and SIRT5 (Fig. [123]4G). These results suggest that silencing SIRT5 leads to increased succinylation of ECHA. Next, we aimed to determine whether the succinylation of ECHA regulates its enzymatic activity. By using 2-(E)-decenoyl-CoA as a substrate and monitoring the formation of NADH from NAD at 340 nm, we measured the combined activities of ECH and HACD in ECHA. The results indicated that the activity of ECHA in the sh-SIRT5 group was reduced by 32% compared to the sh-NC group, suggesting that succinylation downregulates ECHA activity (Fig. [124]4H). Similarly, ECHA purified from sh-SIRT5 HEK293T cells showed lower activity than that from control cells (F[125]ig. [126]4I). Co-expression of ECHA with SIRT5 reduced ECHA succinylation (Fig. [127]4J) and significantly increased its enzymatic activity (F[128]ig. [129]4I). However, co-expression with SIRT5-H158Y did not alter ECHA succinylation (Fig. [130]4J) or its activity (F[131]ig. [132]4I). Additionally, we purified the recombinant murine trifunctional protein complex (ECHA and ECHB) from E. coli to test the effect of succinylation on its in vitro activity. We found that non-enzymatic succinylation reduced the activity of the ECHA and ECHB complex by 40%, but this activity was restored after SIRT5 treatment (Fig. [133]4J). To assess whether lysine acetylation also regulates ECHA activity, we chemically acetylated the ECHA and ECHB complex by incubating it with acetic anhydride and then checked its activity. As shown in Fig. [134]4K, no changes in ECHA activity were observed following acetylation. These results further demonstrate that SIRT5 increases ECHA activity through desuccinylation. SIRT5 mediates the desuccinylation of ECHA, promoting its expression and enhancing fatty acid β-oxidation, which leads to ovarian cancer PARPi resistance Next, we investigated whether SIRT5 promotes fatty acid β-oxidation by mediating the desuccinylation of ECHA. First, we designed three shRNA sequences targeting ECHA. We verified the silencing efficiency and selected the most effective sequence, sh-ECHA-2 (hereafter referred to as sh-ECHA), for subsequent experiments (Figure S3A-B). Similarly, we constructed ovarian cancer cells overexpressing ECHA (oe-ECHA) using lentiviral vectors, with oe-NC as the control. The transfection efficiency was validated using RT-qPCR and Western blot techniques (Figure S3C-D). We assessed the impact of ECHA silencing/overexpression on fatty acid β-oxidation. The results indicated that, compared to the sh-NC group, the sh-ECHA group showed reduced palmitic acid oxidation and ATP production, increased ADP/ATP ratio, and enhanced lipid droplet formation. In contrast, ECHA overexpression produced the opposite effects (Figure S4A–E). These findings suggest that ECHA enhances fatty acid β-oxidation. Additionally, compared to the sh-NC group, sh-ECHA cells were more sensitive to Olaparib treatment and exhibited reduced colony formation ability, whereas ECHA overexpression had the opposite effects (Figure S4F-H). Flow cytometry analysis showed that, following Olaparib treatment, ECHA overexpression significantly reduced the proportion of apoptotic cells (Figure S4I, J). These results indicate that ECHA enhances PARPi resistance. To investigate whether SIRT5 regulates fatty acid β-oxidation by promoting ECHA expression and thus influences PARPi resistance, we used the fatty acid β-oxidation activator Bezafibrate while simultaneously interfering with SIRT5 and ECHA expression. The experimental groups were as follows: oe-NC + sh-NC + DMSO, oe-SIRT5 + sh-NC + DMSO, oe-SIRT5 + sh-ECHA + DMSO, and oe-SIRT5 + sh-ECHA + Bezafibrate. We conducted MTT assays, colony formation assays, and flow cytometry to evaluate the results. Compared to the oe-NC + sh-NC + DMSO group, the oe-SIRT5 + sh-NC + DMSO group showed decreased sensitivity to Olaparib, increased colony formation, and reduced apoptosis. In contrast, compared to the oe-SIRT5 + sh-NC + DMSO group, the oe-SIRT5 + sh-ECHA + DMSO group exhibited increased sensitivity to Olaparib, reduced colony formation, and increased apoptosis. Moreover, the oe-SIRT5 + sh-ECHA + Bezafibrate group showed decreased sensitivity to Olaparib, enhanced colony formation, and reduced apoptosis compared to the oe-SIRT5 + sh-ECHA + DMSO group (Fig. [135]5A–E). These results suggest that overexpressing SIRT5 can upregulate ECHA expression, activate fatty acid β-oxidation, and thereby promote PARPi resistance in ovarian cancer cells. Fig. 5. [136]Fig. 5 [137]Open in a new tab Effects of SIRT5 on PARPi resistance through regulation of ECHA expression. A MTT assay measuring cell viability after 72 h of treatment with varying concentrations of Olaparib; B Colony formation assay assessing the colony-forming ability of SKOV-3 cells (10 μM) and OVCAR-3 cells (2 μM) post-Olaparib treatment; C Quantitative analysis of colony formation ability across different groups; D Flow cytometry analysis of apoptosis in SKOV-3 cells (10 μM) and OVCAR-3 cells (2 μM) following treatment with DMSO or Olaparib; E Quantitative analysis of apoptosis rates across different groups. All experiments were repeated three times. * indicates p < 0.05, ** indicates p < 0.01, *** indicates p < 0.001 SIRT5 regulates ECHA expression, influencing fatty acid β-oxidation and subsequently promoting PARPi resistance To investigate whether SIRT5 modulates fatty acid β-oxidation by regulating ECHA expression and thereby affects the response to Olaparib treatment, we utilized an ovarian cancer orthotopic transplantation model. Two weeks post-transplantation, we administered 50 mg/kg Olaparib intraperitoneally for 14 consecutive days, using DMSO as a control. The groups were as follows: oe-NC + sh-NC + DMSO, oe-SIRT5 + sh-NC + DMSO, oe-SIRT5 + sh-ECHA + DMSO, and sh-SIRT5 + sh-ECHA + Bezafibrate (Fig. [138]6A). Western blot analysis revealed that SIRT5 and ECHA expression were significantly upregulated in the oe-SIRT5 + sh-NC + DMSO group compared to the oe-NC + sh-NC + DMSO group. In contrast, ECHA expression was significantly downregulated in the oe-SIRT5 + sh-ECHA + DMSO group compared to the oe-SIRT5 + sh-NC + DMSO group, while SIRT5 expression remained unchanged. Additionally, ECHA expression was significantly upregulated in the sh-SIRT5 + sh-ECHA + Bezafibrate group compared to the oe-SIRT5 + sh-ECHA + DMSO group, with no significant change in SIRT5 expression (Fig. [139]6B, [140]C). Fig. 6. [141]Fig. 6 [142]Open in a new tab Effects of SIRT5 on tumor formation and PARPi resistance in SKOV-3 cells via ECHA regulation. A Schematic diagram of animal grouping; B, C Western blot analysis of SIRT5 and ECHA protein expression levels in tumor tissues from each group; D Monitoring of tumor growth through bioluminescence intensity, with three representative examples from each group; E Morphology of tumor tissues from each group, with three representative examples; F Tumor volume in each group; G, H Oil Red O staining to detect lipid droplet formation in tumor tissues from each group, Scale bar = 25 μm; *p < 0.05, **p < 0.01, ***p < 0.001, n = 5 per group Four weeks later, tumor growth was assessed through bioluminescence imaging. Tumors were excised to evaluate morphology and volume, and Oil Red O staining was employed to assess lipid droplet content within the tumor tissue. Compared to the oe-NC + sh-NC + DMSO group, the oe-SIRT5 + sh-NC + DMSO group exhibited faster tumor growth and a significant increase in lipid droplet formation. The oe-SIRT5 + sh-ECHA + DMSO group showed smaller tumor volumes and significantly fewer lipid droplets than the oe-SIRT5 + sh-NC + DMSO group. The sh-SIRT5 + sh-ECHA + Bezafibrate group demonstrated relatively faster tumor growth and an increase in lipid droplet formation compared to the oe-SIRT5 + sh-ECHA + DMSO group. Notably, these changes were more pronounced in the groups treated with Olaparib (Fig. [143]6D–H). These results indicate that overexpression of SIRT5 promotes tumor growth and PARPi resistance in vivo by upregulating ECHA expression and enhancing fatty acid β-oxidation. Preparation of Mn[3]O[4]@PDA nanozyme-loaded SIRT5 gene-modified MSCs First, Mn[3]O[4] was synthesized using a hydrothermal method. Dopamine (DA), known for its unique biological, physicochemical properties, and antioxidant activity, was utilized as an excellent surface modifier due to its strong adhesion and self-polymerization capabilities. Therefore, under mildly alkaline conditions, DA was further polymerized onto Mn[3]O[4], resulting in Mn[3]O[4]@PDA Nanozyme, which exhibited enhanced SOD-like activity and good biocompatibility (Fig. [144]7A). Fig. 7. [145]Fig. 7 [146]Open in a new tab Characterization of Mn[3]O[4]@PDA nanozyme. A Preparation process of Mn[3]O[4]@PDA nanozyme-loaded SIRT5 gene-modified MSCs; B TEM image of Mn[3]O[4]@PDA, Scale bar = 5 nm; C Hydrodynamic diameter of prepared Mn[3]O[4] and Mn[3]O[4]@PDA measured by DLS; D Zeta potential of prepared Mn[3]O[4] and Mn[3]O[4]@PDA; E FTIR spectra of prepared Mn[3]O[4], Mn[3]O[4]@PDA, and DA; (F) XPS survey spectra of Mn[3]O[4] and Mn[3]O[4]@PDA; G XPS spectrum of Mn 2p in Mn[3]O[4]@PDA; H XRD patterns of Mn[3]O[4] and Mn[3]O[4]@PDA compared with simulated patterns of Mn[3]O[4] from the JCPDS database; I Longitudinal relaxation (r1) fitting curves of Mn[3]O[4] and Mn[3]O[4]@PDA at pH 7.0; J T1-weighted and pseudo-color maps of Mn[3]O[4] and Mn[3]O[4]@PDA at different Mn concentrations (0–72 μm) in water; K Flow cytometry analysis of MSC fluorescence intensity at different co-incubation times; L MSC viability after treatment with different concentrations of Mn[3]O[4]@PDA nanozyme; M Labeling efficiency of Mn[3]O[4]@PDA on MSCs at different co-incubation times; N T1-weighted MRI images of engineered MSCs; O Bio-TEM images of engineered MSCs, Scale bar = 25 μm. All cell experiments were repeated three times TEM revealed that the prepared Mn[3]O[4] and Mn[3]O[4]@PDA displayed monodispersed, regular spherical shapes (Fig. [147]7B). Dynamic light scattering characterization showed that the hydrodynamic diameter increased from 211.2 ± 35.4 nm to 343.7 ± 11.5 nm after DA modification (Fig. [148]7C), and the zeta potential decreased from − 2.63 ± 0.4 mV to − 13.32 ± 0.6 mV (Fig. [149]7D). Fourier-transform infrared spectroscopy of Mn[3]O[4]@PDA showed peaks at 2952 cm^−1 (–NH[2]) and 1540 cm^−1 (C=C), and X-ray photoelectron spectroscopy confirmed that the N atoms from DA were significant components of Mn[3]O[4]@PDA (Fig. [150]7E, [151]F). XPS Peak 4.0 software was used to analyze the XPS spectrum of Mn(2p). As shown in Fig. [152]7G, the molar ratio of Mn^2+ (642.0 eV) to Mn^3+ (639.6 eV) increased from 1:2 to 1:6 after modification. Additionally, X-ray diffraction characterization indicated that Mn[3]O[4]@PDA maintained high crystallinity similar to Mn[3]O[4] (Fig. [153]7H). The longitudinal relaxivity (r[1]) of Mn[3]O[4]@PDA was 5.81 mm^−1 s^−1, nearly nine times higher than that of Mn[3]O[4] at pH 7.0, indicating that DA modification significantly enhanced the MRI capability of Mn[3]O[4] (F[154]ig. [155]7I, [156]J). We evaluated the scavenging abilities of Mn[3]O[4] and Mn[3]O[4]@PDA for H[2]O[2], •OH, and •O^2−. Compared to Mn[3]O[4], Mn[3]O[4]@PDA demonstrated significantly higher ROS elimination rates (Figure S5A-C). As shown in Figure S5D, Mn[3]O[4]@PDA significantly enhanced MSC viability under 200 μM H[2]O[2] conditions; specifically, at a concentration of 10 μg/mL Mn[3]O[4]@PDA, MSC survival increased from 19.7 to 77.8%, whereas survival was only 30.2% at the same concentration of Mn[3]O[4]. Furthermore, we observed that ROS fluorescence almost disappeared when MSCs were co-cultured with 10 μg/mL Mn[3]O[4]@PDA Nanozyme (Figure S5E-F), indicating that Mn[3]O[4]@PDA Nanozyme exhibits efficient ROS scavenging activity within cells. Next, we constructed SIRT5-silenced MSCs (Mn[3]O[4]@PDA-MSCs-shSIRT5) and verified the silencing via RT-qPCR and Western blot (Figure S6A-B). We created Nanozyme-engineered MSCs by co-incubating Mn[3]O[4]@PDA with MSCs. After 24 h of co-incubation, the engineered MSCs showed negligible cytotoxicity (Fig. [157]7K). We further investigated the effect of co-incubation time on MSCs labeling efficiency using FITC-labeled Mn[3]O[4]@PDA. As shown in Fig. [158]7L, [159]M, intracellular fluorescence of FITC-labeled Mn[3]O[4]@PDA in MSCs increased with longer co-incubation times, reaching a peak at 12 h, with a labeling efficiency of 87.5% (Fig. [160]7M). Compared to unlabeled MSCs, Mn[3]O[4]@PDA-MSCs-sh-SIRT5 exhibited significant T1 MRI signals, confirming the successful engineering of MSCs with Mn[3]O[4]@PDA Nanozyme (Fig. [161]7N). Bio-TEM images also demonstrated effective internalization of Mn[3]O[4]@PDA within MSCs (Fig. [162]7O). Additionally, RT-qPCR analysis of four pluripotency markers (Oct4, Sox2, Nanog, and CXCR4) showed that Oct4 and Nanog expression levels in Mn[3]O[4]@PDA-MSCs-sh-SIRT5 were comparable to those in normal MSCs, while Sox2 and CXCR4 expression levels were slightly higher than in normal MSCs (Figure S6C). This indicates that SIRT5 silencing or Mn[3]O[4]@PDA engineering does not impair MSC pluripotency and may even enhance it. Mn[3]O[4]@PDA-MSCs-sh-SIRT5 effectively inhibits in vivo PARPi resistance by suppressing fatty acid β-oxidation To investigate whether Mn[3]O[4]@PDA Nanozyme-loaded SIRT5 gene-modified hUCMSCs could enhance the efficacy of standalone interventions, we administered 50 mg/kg Olaparib intraperitoneally for 14 consecutive days approximately two weeks after establishing the orthotopic tumor model. The mice were divided into four groups: Control, Mn[3]O[4]@PDA, Mn[3]O[4]@PDA-MSCs, and Mn[3]O[4]@PDA-MSCs-sh-SIRT5 (Fig. [163]8A). We used Western blot to assess SIRT5 and ECHA expression, monitored tumor growth through bioluminescence imaging, excised tumor tissues at week 4 to observe morphology and measure volume, and performed Oil Red O staining to evaluate lipid droplet content within tumor tissues. The results showed that compared to the Control group, SIRT5 and ECHA expressions were significantly downregulated in the Mn[3]O[4]@PDA-MSCs-sh-SIRT5 group (Fig. [164]8B, [165]C). Both Mn[3]O[4]@PDA-MSCs and Mn[3]O[4]@PDA-MSCs-sh-SIRT5 groups exhibited smaller tumor volumes and significantly reduced lipid droplet formation, with the latter demonstrating more pronounced effects. Furthermore, compared to the Mn[3]O[4]@PDA-MSCs group, the Mn[3]O[4]@PDA-MSCs-sh-SIRT5 group showed further reductions in tumor volume and lipid droplet formation (Fig. [166]8D–H). These findings suggest that Mn[3]O[4]@PDA-MSCs can inhibit in vivo PARPi resistance to a certain extent, potentially due to the repair capabilities of MSCs and the antioxidant activity of the Nanozyme. Mn[3]O[4]@PDA-MSCs-sh-SIRT5 appears to be even more effective in inhibiting in vivo PARPi resistance. Fig. 8. [167]Fig. 8 [168]Open in a new tab Effects of Mn[3]O[4]@PDA-MSCs loaded with sh-SIRT5 on PARPi resistance in vivo. A Schematic diagram of animal grouping; B, C Western blot analysis of SIRT5 and ECHA protein expression levels in tumor tissues of each group; D Tumor growth monitored by bioluminescence intensity, with three representative examples per group; E Morphology of tumor tissues in each group, with three representative examples per group; F Tumor volume in each group; G, H Oil Red O staining of lipid droplet formation in tumor tissues of each group, scale bar = 25 μm; *p < 0.05, **p < 0.01, ***p < 0.001, with five mice per group Next, we investigated whether Mn[3]O[4]@PDA-MSCs-sh-SIRT5 exerted its effects by inhibiting fatty acid β-oxidation. To this end, we further divided the Mn[3]O[4]@PDA-MSCs-sh-SIRT5 mice into DMSO and Bezafibrate groups (Figure S7A). We observed that, compared to the DMSO group, the Bezafibrate group exhibited larger tumor volumes and significantly increased lipid droplet formation (Figure S7B–F). This indicates that activating fatty acid β-oxidation can significantly reverse the improvement in PARPi resistance achieved by Mn[3]O[4]@PDA-MSCs-sh-SIRT5 in vivo. Finally, we conducted an in vivo safety assessment of Mn[3]O[4]@PDA-MSCs. The results showed that the heart, liver, spleen, lungs, and kidneys of the mice in each group were smooth and normal in color and size. Histological examinations of these organs also revealed normal physiological structures and cellular morphology (Figure S8A). Additionally, compared to the control group, there were no significant differences in the biochemical indicators of liver damage (ALT and AST activity) or renal function (blood urea nitrogen [BUN] and serum creatinine [Cr] levels) among the other three groups (Figure S8B-E). These findings suggest that Mn[3]O[4]@PDA-MSCs possess good biocompatibility in vivo. In summary, the evidence demonstrates that Mn[3]O[4]@PDA-MSCs-sh-SIRT5 effectively inhibits PARPi resistance in vivo by suppressing fatty acid β-oxidation. Discussion Ovarian cancer is one of the most lethal malignancies of the female reproductive system, with PARP inhibitor (PARPi) resistance being a significant therapeutic challenge [[169]49, [170]50]. Previous studies have shown that PARPi exerts substantial antitumor effects by inhibiting DNA repair mechanisms in BRCA1/2-deficient cells [[171]51–[172]53]. However, over time, many ovarian cancer patients develop resistance to PARPi, significantly diminishing the efficacy of the treatment [[173]14, [174]54]. The SIRT5 gene, due to its crucial role in fatty acid metabolism and desuccinylation, has become the focus of this study. We systematically explore the specific mechanisms of SIRT5 in ovarian cancer PARPi resistance using multi-omics technologies and antioxidant nanozyme-engineered stem cells, aiming to provide new insights to overcome this challenge. Previous research on PARPi resistance has primarily focused on alterations in DNA repair pathways, upregulation of drug efflux pumps, and anomalies in cell cycle regulation [[175]13, [176]15, [177]17]. For example, studies have indicated that restoration mutations in BRCA1/2 are a major cause of PARPi resistance [[178]52, [179]55, [180]56]. Additionally, the upregulation of the MDR1 gene has been identified as a key factor in influencing PARPi resistance [[181]57, [182]58]. This study, however, reveals a novel mechanism where SIRT5 improves PARPi resistance by regulating fatty acid metabolism pathways. Through high-throughput transcriptome sequencing, metabolomics, and proteomics analyses, we systematically identified key genes and metabolic pathways potentially influenced by SIRT5, providing a new perspective on the mechanisms of resistance. In our multi-omics analysis, we discovered that SIRT5 might improve ovarian cancer PARPi resistance by regulating fatty acid β-oxidation. Transcriptomics analysis revealed that upregulation of SIRT5 significantly affects the expression of a series of genes related to fatty acid metabolism. Metabolomics further validated this finding, showing that the fatty acid β-oxidation pathway is notably active in cells overexpressing SIRT5. These results suggest that fatty acid metabolism may play a crucial role in PARPi resistance. Unlike previous studies that primarily focused on DNA repair and drug efflux pumps, our research highlights the potential role of fatty acid metabolism in resistance mechanisms, providing new directions for the study of drug resistance. Proteomics analysis further elucidated the specific role of SIRT5 in regulating fatty acid metabolism. The results identified ECHA as the key protein for SIRT5-mediated desuccinylation. ECHA plays a vital role in fatty acid β-oxidation [[183]59, [184]60], and its desuccinylation significantly enhances its enzymatic activity [[185]61]. In vitro experiments confirmed this mechanism, showing that overexpression of SIRT5 promotes ECHA desuccinylation, thereby enhancing fatty acid β-oxidation and improving PARPi resistance, consistent with previous findings [[186]61]. Additionally, recent studies have highlighted FAO—a reverse process of fatty acid synthesis (FAS)—as a crucial driver of tumor progression, metastasis, and chemoresistance in various cancers [[187]46, [188]47]. Our data provide the first direct evidence in ovarian cancer that SIRT5 enhances PARPi resistance by promoting ECHA desuccinylation and upregulating FAO, thereby establishing a novel regulatory mechanism between SIRT5 and ECHA in the context of energy metabolism and therapeutic resistance. The in vitro experiments confirmed our multi-omics analysis findings. Through MTT assays, colony formation assays, and flow cytometry, we observed that SIRT5 overexpression significantly enhanced ovarian cancer cell resistance to PARPi. Conversely, silencing SIRT5 markedly reduced this resistance. These results indicate that SIRT5 plays a crucial role in PARPi resistance by regulating fatty acid β-oxidation, providing a novel target and therapeutic strategy distinct from the previously known mechanisms of DNA repair and drug efflux pumps. To further validate this mechanism, we prepared Mn[3]O[4]@PDA nanozyme-loaded, SIRT5 gene-modified hUCMSCs and tested their effects in vivo. The in vivo experiments demonstrated that Mn[3]O[4]@PDA-MSCs-sh-SIRT5 exhibited good traceability and antioxidative effects in a mouse ovarian cancer xenograft model. Post-Olaparib treatment, bioluminescence imaging revealed significantly inhibited tumor growth in mice treated with Mn[3]O[4]@PDA-MSCs-sh-SIRT5, with oil red O staining showing a marked reduction in lipid droplet formation. These findings suggest that Mn[3]O[4]@PDA nanozyme-engineered stem cells effectively inhibit in vivo PARPi resistance by suppressing fatty acid β-oxidation. This study presents several innovations and clinical implications. Firstly, utilizing multi-omics technologies to comprehensively analyze transcriptomics, metabolomics, and proteomics, we systematically revealed the role of SIRT5 in ovarian cancer PARPi resistance. Secondly, the Mn[3]O[4]@PDA nanozyme-loaded, SIRT5 gene-modified hUCMSCs demonstrated excellent traceability and antioxidative capabilities in vivo, offering new strategies for clinical application. Specifically, the combination of SIRT5 modulation, Mn₃O₄@PDA nanozymes, and MSC-mediated delivery enables multi-level intervention in PARPi resistance. SIRT5 affects both cellular metabolism and DNA repair via regulation of β-oxidation, which is directly linked to resistance. Mn₃O₄@PDA functions as a ROS scavenger to reduce oxidative stress and delay resistance development. Notably, while Mn₃O₄ nanoparticles have been previously studied in cancer treatment [[189]62, [190]63], this is the first study to explore their use in ovarian cancer. Additionally, MSCs confer tumor-targeting capability, enhancing drug accumulation within the tumor microenvironment—a distinct advantage over traditional Mn₃O₄ nanoparticles. This triple-action strategy—targeting metabolism, oxidative stress, and precise delivery—offers a promising approach to overcome resistance mechanisms. Despite the promising results, this strategy faces several challenges for clinical translation. First, although MSCs are known to home to tumor sites, ensuring the effective loading, targeted release of Mn₃O₄ nanoparticles, and regulation of SIRT5 activity remains a technical challenge. Second, the complexity of the nanoparticle delivery system may complicate regulatory approval due to the need for stringent quality and safety standards for combinatorial therapies. Third, cost-effectiveness is critical and must be optimized during development, production, and future clinical trials. To address these issues, future work should focus on systematic optimization of delivery systems, including preclinical evaluation of component dosing and small-scale clinical trials to assess therapeutic efficacy. Interdisciplinary collaboration—spanning nanomaterials, cell therapy, and molecular biology—will be crucial to support the development of this complex therapeutic platform. Furthermore, market strategy and cost–benefit analyses will play essential roles in facilitating clinical adoption. The use of patient-derived organoids (PDOs) would further strengthen the clinical relevance of our findings. However, due to resource limitations, this was not included in the current study. Nonetheless, the results from SKOV3 and OVCAR3 cell lines provide strong preliminary evidence for future validation in HGSOC models. In clinical practice, approved PARP inhibitors such as Rucaparib and Niraparib differ in their mechanisms and efficacy profiles. Testing additional PARPi agents is essential to determine whether SIRT5-mediated resistance is drug-specific (e.g., Olaparib) or class-wide. We acknowledge this as a limitation of our current design and plan to expand our drug screening in future studies to comprehensively understand the role of SIRT5 across different PARPi agents. Despite certain limitations in this study, such as a relatively small sample size, model constraints, and technical challenges, as well as the lack of additional metabolic flux analysis to further strengthen the link between fatty acid metabolism and drug resistance, our research provides new insights for the future treatment of PARPi resistance in ovarian cancer. Future studies should further validate this mechanism, optimize the design of nanozyme carriers, and promote their clinical translation. In conclusion, our research reveals the mechanism by which SIRT5 regulates ECHA desuccinylation and fatty acid β-oxidation to improve ovarian cancer PARPi resistance using multi-omics technology. The successfully prepared Mn[3]O[4]@PDA nanozyme-loaded, SIRT5 gene-modified hUCMSCs showed excellent in vivo traceability and antioxidative effects, indicating potential clinical applications. Despite some limitations, this study offers new avenues and directions for treating ovarian cancer PARPi resistance. Future research will further explore this mechanism, optimize nanozyme design, and promote its clinical translation, providing more therapeutic options and hope for ovarian cancer patients. Conclusion Based on the comprehensive analysis of the study, the following conclusions can be drawn: This research successfully developed Mn[3]O[4]@PDA nanozyme-loaded SIRT5 gene-modified hUCMSC (Mn[3]O[4]@PDA-MSCs-sh-SIRT5). These engineered cells exhibit excellent in vivo tracking, antioxidant capabilities, and biosafety. The specific mechanism by which antioxidant nanozyme-engineered stem cells improve ovarian cancer PARPi resistance involves SIRT5-mediated desuccinylation of ECHA, enhancing its expression and consequently promoting fatty acid β-oxidation, which contributes to overcoming PARPi resistance (Fig. [191]9). Fig. 9. [192]Fig. 9 [193]Open in a new tab Molecular mechanism by which antioxidant nanozyme-engineered stem cells improve ovarian cancer PARPi resistance This study presents a potential therapeutic strategy to overcome tumor drug resistance through metabolic reprogramming. This strategy not only enhances the efficacy of cancer treatment but also provides new theoretical and experimental foundations for metabolism-based cancer therapies. Additionally, this method may address challenges related to in vivo tracking and homing abilities. However, despite using a mouse model to simulate human disease conditions, there are differences in PARPi resistance response and metabolic pathways between mice and humans, which may limit the direct translation of the findings. Furthermore, the long-term effects remain unclear. The research primarily focuses on the short-term impact of antioxidant nanozyme-engineered stem cells on improving ovarian cancer PARPi resistance, and further studies are needed to evaluate their long-term efficacy and safety. Supplementary Information [194]12951_2025_3516_MOESM1_ESM.jpg^ (435.7KB, jpg) Supplementary Material 1. Figure S1. Validation of SIRT5 silencing and overexpression efficiency. (A-B) Expression levels of SIRT5 in cells transfected with shRNA were detected by RT-qPCR (A) and Western blot (B); (C-D) Expression levels of SIRT5 in cells transfected with SIRT5 plasmid were evaluated by RT-qPCR (C) and Western blot (D). All cellular experiments were replicated thrice. * indicates p < 0.05, ** indicates p < 0.01, *** indicates p < 0.001 [195]12951_2025_3516_MOESM2_ESM.jpg^ (495.1KB, jpg) Supplementary Material 2. Figure S2. Enhanced Sensitivity of Olaparib-Resistant Cells by Silencing SIRT5. Note: (A) Cell viability and IC50 of SKOV-3 parental and drug-resistant cells treated with different concentrations of Olaparib for 72 h assessed by MTT assay; (B) mRNA expression of SIRT5 in SKOV-3 parental and drug-resistant cells analyzed by RT-qPCR; (C) Protein expression of SIRT5 in SKOV-3 parental and drug-resistant cells determined by Western blot; (D) Quantitative analysis of SIRT5 protein expression; (E) Cell viability of SKOV-3 drug-resistant cells treated with varying concentrations of Olaparib for 72 h using MTT assay; (F) Colony Formation Assay evaluating the colony-forming ability of SKOV-3 drug-resistant cells treated with different concentrations of Olaparib; (G) Quantitative statistical analysis of the colony-forming ability of cells in each group. Each cellular experiment was performed thrice, with significance denoted as * for p < 0.05, ** for p < 0.01, and *** for p < 0.001 [196]12951_2025_3516_MOESM3_ESM.jpg^ (422.3KB, jpg) Supplementary Material 3. Figure S3. Validation of ECHA Silencing and Overexpression Efficiency. Note: (A-B) RT-qPCR (A) and Western blot (B) were employed to detect the expression levels of SIRT5 in cells after lentiviral transfection with shRNA; (C-D) RT-qPCR (C) and Western blot (D) were used to measure the expression levels of SIRT5 in cells after lentiviral transfection with SIRT5 plasmid. All cell experiments were replicated three times. ** indicates p < 0.01, and *** indicates p < 0.001 [197]12951_2025_3516_MOESM4_ESM.jpg^ (620.7KB, jpg) Supplementary Material 4. Figure S4. The Impact of ECHA on Fatty Acid β-Oxidation and PARPi Resistance. Note: (A) Fatty acid β-oxidation rate in each group of cells; (B) ATP production levels in each group of cells; (C) ADP/ATP ratio in each group of cells; (D) Detection of lipid droplet formation using BODIPY 493/503 staining, with a scale bar of 25 μm; (E) Quantitative analysis of lipid droplet formation in each group of cells; (F) Assessment of cellular viability in different concentrations of Olaparib-treated cells after 72 h using MTT assay; (G) Colony Formation Assay determining colony-forming ability of SKOV-3 cells (10 μM) and OVCAR-3 cells (2 μM) following Olaparib treatment; (H) Quantitative analysis of colony-forming ability in each group of cells; (I) Flow Cytometry analysis of apoptosis in SKOV-3 cells (10 μM) and OVCAR-3 cells (2 μM) after treatment with DMSO or Olaparib; (J) Quantitative analysis of cell apoptosis in each group. All cell experiments were repeated three times, with * indicating p < 0.05, ** indicating p < 0.01, and *** indicating p < 0.001 [198]12951_2025_3516_MOESM5_ESM.jpg^ (498.4KB, jpg) Supplementary Material 5. Figure S5. Activity of Mn[3]O[4] and Mn[3]O[4]@PDA Nanozyme in ROS Elimination. Note: (A) Removal of H[2]O[2]by Mn[3]O[4] and Mn[3]O[4]@PDA; (B) Hydroxyl radical scavenging activity of Mn[3]O[4] and Mn[3]O[4]@PDA; (C) Superoxide anion elimination activity of Mn[3]O[4] and Mn[3]O[4]@PDA; (D) Cell viability of MSCs treated with 200 μM H[2]O[2] and different concentrations of Mn[3]O[4]@PDA Nanozyme for 24 h; (E) Representative fluorescence images of MSCs with ROS levels indicated by DCFH-DA probe, Scale bar = 25 μm; (F) Quantitative analysis of relative fluorescence intensity of ROS. All cell experiments were repeated three times, ** indicates p < 0.01, *** indicates p < 0.001 [199]12951_2025_3516_MOESM6_ESM.jpg^ (282.9KB, jpg) Supplementary Material 6. Figure S6. Validation of SIRT5 Silencing Efficiency and Pluripotency of MSCs. Note: (A-B) The expression levels of SIRT5 in MSCs after lentiviral transfection with shRNA were detected using RT-qPCR (A) and Western blot (B); (C) RT-qPCR was performed to measure the expression levels of pluripotency markers Oct4, Sox2, Nanog, and CXCR4. All cellular experiments were carried out in triplicate, and *** indicates p < 0.001 [200]12951_2025_3516_MOESM7_ESM.jpg^ (877.9KB, jpg) Supplementary Material 7. Figure S7. The Influence of Bezafibrate on PARPi Resistance In Vivo. Note: (A) Schematic of animal grouping; (B) Tumor growth monitored by bioluminescence intensity, with 3 representative samples shown for each group; (C) Morphology of tumor tissues from each group, with 3 representative samples displayed for each group; (D) Tumor volume for each group of mice; (E–F) Oil Red O staining to assess lipid droplet formation in tumor tissues from each group of mice, Scale bar = 25 μm; * indicates p < 0.05, ** indicates p < 0.01, *** indicates p < 0.001, with 5 mice per group [201]12951_2025_3516_MOESM8_ESM.jpg^ (2.5MB, jpg) Supplementary Material 8. Figure S8. In vivo Toxicity Assessment of Mn[3]O[4]@PDA-MSCs.Note: (A) Histopathological examination of the hearts, livers, spleens, lungs, and kidneys of mice in each group using H&E staining, Scale bar = 100 μm; (B-C) Serum levels of ALT (B) and AST (C) in mice from each group; (D-E) Serum levels of BUN (D) and Cr (E) in mice from each group; five mice per group [202]Supplementary Material 9.^ (16.2KB, docx) Acknowledgements