Abstract Black phosphorus quantum dots (BPQDs) have shown potential in tumor therapy, however, their anti-angiogenic functions have not been studied. Although BPQDs are easily degraded to non-toxic phosphrous, the reported toxicity, poor stability, and non-selectivity largely limit their further application in medicine. In this study, a vascular targeting, biocompatible, and cell metabolism-disrupting nanoplatform is engineered by incorporating BPQDs into exosomes modified with the Arg-Gly-Asp (RGD) peptide (BPQDs@RGD-EXO nanospheres, BREs). BREs inhibit endothelial cells (ECs) proliferation, migration, tube formation, and sprouting in vitro. The anti-angiogenic role of BREs in vivo is evaluated using mouse retinal vascular development model and oxygen-induced retinopathy model. Combined RNA-seq and metabolomic analysis reveal that BREs disrupt glucose metabolism, which is further confirmed by evaluating metabolites, ATP production and the c-MYC/Hexokinase 2 pathway. These BREs are promising anti-angiogenic platforms for the treatment of pathological retinal angiogenesis with minimal side effects. Keywords: Exosomes, Black phosphorus quantum dots, Angiogenesis, Cell metabolism, Retinopathy Graphical abstract [45]Image 1 [46]Open in a new tab 1. Introduction Pathological retinal angiogenesis results from diabetic retinopathy, retinal vein occlusion, and retinopathy of prematurity, often leads to blindness [[47]1]. Fundus laser treatment and intraocular injection of anti-Vascular Endothelial Growth Factor (anti-VEGF) agents have proven to be effective for pathological retinal angiogenesis [[48][2], [49][3], [50][4]]. However, these treatments have some limitations, such as repeated injections, drug resistance, and systemic adverse effects due to non-selectivity [[51]5,[52]6]. The emerging role of nanomedicine has provided new therapeutic options. Researchers have found that inorganic nanoparticles of gold, copper and carbon have potential for the treatment of pathological angiogenesis [[53][7], [54][8], [55][9]]. However, phosphorus-based nanoparticles have never been studied, despite the fact that it is an essential quantitative element for the body and has specific potential in anti-angiogenic treatment. Owing to their multiple properties, such as fluorescence, photothermal effects, and photosensitizers, black phosphorus quantum dots (BPQDs) have been extensively studied for biomedical applications [[56][10], [57][11], [58][12]]. However, the toxicity of the BPQDs cannot be ignored. Recent studies have revealed that BPQDs are toxic to vascular endothelial cells, lung-derived cells, bone marrow nucleated cells, and kidneys [[59][13], [60][14], [61][15], [62][16]]. Thus, the proper decoration of BPQDs to improve biocompatibility is indispensable for further applications. Exosomes are cell-secreted nanoparticles with a size ranging from 50 ​nm to 150 ​nm. They are emerging as novel bioactive nanoparticles in various applications such as drug delivery, cell therapy, and disease diagnosis [[63]17]. Exosomes are ideal drug delivery platforms because of their unique physical properties such as low immunogenicity, low toxicity, high stability, and high biocompatibility [[64]18,[65]19]. Iron oxide nanoparticle-incorporated exosomes show potential for cardiac repair and targeted cancer therapy [[66]20]. Gold nanoparticle-conjugated exosomes have been reported to enhance blood-brain barrier penetration and neuroimaging [[67]21,[68]22]. A recent study revealed that BPQD-incorporated exosomes exhibit enhanced photothermal therapy efficiency for tumor ablation in vivo [[69]23]. These exosomes largely improve the biocompatibility, stability, and targeting ability of nanoparticles, indicating the promising application of exosomes in nanoparticle delivery. Cell metabolism is the foundation of all vital movements [[70]24]. Glucose metabolism is a major energy source for biological activities [[71]25]. Hexokinases catalyze the first committed step by phosphorylating glucose. There are five hexokinase isoforms, of which hexokinase 2 (HK2) plays a major role in endothelial cell glucose metabolism [[72]26]. The expression of HK2 is controlled by oncogenes, such as hypoxia inducible factor 1 subunit alpha (HIF1α) and c-MYC [[73]27,[74]28]. Conditional knockout of HK2 in endothelial cells suppresses angiogenesis both in vitro and in vivo, which can be rescued by c-MYC [[75]27]. Meanwhile, black phosphorus has been reported to regulate tumor metabolism by downregulating prostaglandin E2 using untargeted metabolomic analysis [[76]29]. Even though photothermal and photodynamic therapies using BPQDs have been reported, the effects of black phosphorus nanomaterials on vascular cell metabolism are rarely reported. Using Arg-Gly-Asp (RGD) modified exosomes to deliver BPQDs can further improve their biocompatibility and minimize side effects. To overcome the limitations of present medicine for pathological angiogenesis [[77]5,[78]6], it is urgent to develop novel therapeutics. The nanocomposites could serve as alternative for pathological angiogenesis therapy, with potential for clinical translation. To this end, we determined the functions of vascular-targeting exosomes incorporated with BPQDs in pathological angiogenesis therapy ([79]Fig. 1). We first modified exosomes with Arg-Gly-Asp (RGD) on their membranes using a donor cell-assisted membrane modification strategy to improve their blood vessel targeting ability. Then, the BPQDs were packaged into RGD-exosomes (BPQDs@RGD-EXO nanospheres, BREs) using electroporation, which largely reduced the toxicity of BPQDs. BREs attenuate pathological angiogenesis in vitro and in vivo. Combined RNA-seq and metabolomic analysis revealed that BREs mainly disrupted glucose metabolism through the c-MYC/HK2 pathway. Our work indicates that engineered BREs are promising for pathological angiogenesis therapy with high vascular selectivity and low toxicity. Fig. 1. [80]Fig. 1 [81]Open in a new tab Schematic illustration of the preparation and application of BPQDs@RGD-EXO (BREs). a) The construction of BREs. b) The schematic illustration of combined RNA-seq and metabolomics. c) The therapeutic outcomes and underlying mechanism of BREs treated group. 2. Material and methods 2.1. Materials BPQDs were purchased from Nanjing XFNANO Materials Tech Co., Ltd. (XF208, Nanjing, China); Tublin, 4′,6-diamidino-2-phenylindole (DAPI) were purchased from Thermo Fisher Scientific (Shanghai, China). Cell-LightTM EdU Apollo567 in Vitro Kit was purchased from Ribobio (Guangzhou, China); Cell Counting Kit- 8(CCK-8) was obtained from Dojindo Laboratories, (Dojindo, Japan); FITC-conjugated Isolectin B4 (IB4), Triton-X-100 and Methyl cellulose were purchased from Sigma-Aldrich (Shanghai, China). Matrigel Matrix was obtained from BD Biosciences (Shanghai, China); TUNEL Apoptosis Assay Kit, Annexin V-FITC/PI Apoptosis Detection Kit, Enhanced ATP Assay Kit and Actin-Tracker Red-Rhodamine were purchased from Beyotime Biotechnology (Jiangsu, China); Calcein-AM/PI Live Cell/Dead Cell Double Staining Kit was purchased from Solarbio (Beijing, China); Antibodies to TSG101, Calnexin, β-Actin, c-Myc, ERK, p-ERK and HK2 were purchased from Proteintech Group (Wuhan, China); Antibodies to PKM2 and LDHA were purchased from Cell Signaling Technology (Shanghai, China); Antibodies to CD63 and CD81 were obtained from Affinity Biosciences (Jiangsu, China); Antibodies to CD9 and 1,1′-Dioctadecyl-3,3,3′,3′-Tetramethylindodicarbocyanine Perchlorate (DiD) were purchased from Yeasen Biotechnology (Shanghai, China). DSPE-PEG-RGD was synthesized by ChinaPeptides (Shanghai, China). Lactic Acid assay kit and Pyruvate assay kit were purchased from Nanjing Jiancheng Bioengineering Institute (Nanjing, China); all the cell culture plates and dishes were purchased from Corning Life Sciences. 2.2. Cell culture HUVECs and HEK 293 ​cells were obtained from ScienCell (San Diego, USA). HUVECs were cultured in Endothelial Cell Medium (ECM, Cell Research, Shanghai, China) supplemented with 5% fetal bovine serum (FBS), 1% endothelial cell growth supplement (ECGS), and 1% penicillin-streptomycin. HEK 293 ​cells were cultured in 10% Dulbecco's modified Eagle's medium (Gibco) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin. Cells were cultured at 37 ​°C and 5% CO[2]. All cell lines tested negative for Mycoplasma. 2.3. Preparation and characterization of BREs HEK 293 ​cells were cultured in DMEM supplemented with 10% exosome-free FBS. To obtain RGD-modified exosomes (RGD-EXO), HEK 293 ​cells were incubated in DMEM with DSPE-PEG-RGD nanoparticles for 4 ​h, as previously reported [[82]30]. Exosomes were separated using the traditional differential centrifugation method. Briefly, the culture medium was first collected and centrifuged at 300×g for 10 ​min at 4 ​°C to remove cells. The supernatant was then centrifuged twice at 2000×g for 20 ​min at 4 ​°C to remove cell debris. To obtain exosomes, the culture medium was centrifuged at 100 000 ​g for 120 ​min at 4 ​°C. Exosomes were washed once with phosphate-buffered saline (PBS) and resuspended in the appropriate buffers. The successful modification of RGD was evaluated using NMR spectral analysis. BREs were then prepared using electroporation methods, as previously reported [[83]23]. Briefly, RGD-EXO (0.2 ​mg) and BPQDs (0.2 ​mg) were mixed in 250 ​μL PBS in the 0.4 ​cm electric cuvette. The BREs were acquired by electroporation at 400 ​V and 150 ​μF. The size and morphological features were visualized using TEM. The exosome concentration was determined using a BCA protein assay kit (Servicebio, Wuhan, China). DiD-labelled BREs were used to assess exosomes internalization by HUVECs. Briefly, BREs (1 ​μg/μL) were incubated with DiD (1 ​μM/L) at 37 ​°C for 30 ​min, the rest of unbound DiD was then removed using centrifuge at 100 000 ​g for 120 ​min at 4 ​°C. DiD-labelled BREs were co-incubated with HUVECs for 4 ​h and analyzed using confocal microscopy. The in vitro BPQDs release profile of BREs in saline solution were measured through ultracentrifugation method. Five mL of BREs solution, containing BPQDs 1 ​mg, were dispersed in 5 ​mL of saline solution before suspended in a centrifuge tube containing 10 ​mL buffer (pH 7.4). The BPQDs were separated with a centrifugation condition of 20,000 ​g, 20 ​min, and 4 ​°C, and an aliquot of solution (500 ​μL) was collected from the supernatant with same volume renewal of release medium at different time points while oscillated at 37 ​°C. The phosphorus concentration of samples was analyzed by inductively coupled plasma-mass spectrometry (ICP-MS, XSERIES 2, Thermo Fisher Scientific, MA, USA). The release of BPQDs in BREs was calculated by standard curve. The encapsulation efficiency (EE) of BREs were examined by ICP-MS. The concentrations and mass of encapsuled BPQDs were obtained according to the standard curve and known sample volume, and the RGD-EXO nanoparticles were used as a control. The EE were calculated by following the formula: [MATH: EE(wt%)=massofencapsuledBPQDsmassoffeedingBPQDs×100(%) :MATH] 2.4. Dead-live staining Live-dead cell staining was performed to assess the cytotoxicity in vitro. Briefly, cells were seeded in 24-well plates at a density of 1 ​× ​10^4 ​cells/well and cultured for 12 ​h. Then, BREs, EXO, RGD-EXO and BPQDs were suspended in ECM and co-incubated with HUVECs for 48 ​h. After that, cells were stained with calcein-AM (living cells) and propidium iodide (dead/late apoptotic cells) for 20 ​min. Cellular fluorescence was monitored under a microscope (IX81, Olympus). 2.5. Cell proliferation Cellular proliferation was assessed using 5-ethynyl-2-deoxyuridine (EdU) assays. Briefly, cells were seeded at a density of 2 ​× ​10^3 ​cells/well in 96-well plates and treated with BREs for 48 ​h. According to the manufacturer's recommended protocol, cells were co-cultured with EdU working solution (1:1000) for 2 ​h, followed by fixation with 4% paraformaldehyde for 30 ​min. Glycine was then added to wash the cells for 5 ​min, followed by two washes with 0.3% Triton X-100. Subsequently, the cells were incubated with Apollo fluorescent azide for 30 ​min at room temperature in the dark, followed by three washes with 0.3% Triton X-100. DAPI was incubated for another 30 ​min, and the cells were washed three times with PBS. The proliferation was calculated as the number of EdU-positive cells/number of DAPI-stained cells. Images were acquired using a fluorescence microscope (IX81, Olympus), and cell counting was performed using the ImageJ software. 2.6. Cell migration Cell migration was evaluated using a wound healing method. In brief, 1 ​× ​10^5 ​cells were plated in a 2-well ibidi chamber plate to form confluent monolayer, then the cells were pretreated with BREs for 48 ​h. After that, the insert was pulled up and cells were rinsed with PBS three times to remove the suspended cells, and then cultured with the original medium. Images were obtained at 0 ​h and 16 ​h using microscope (IX81, Olympus). All images were processed using ImageJ software. 2.7. Tube formation Firstly, 100 ​μL Matrigel (10 ​mg/mL) was pipetted onto pre-cooled 96-well plate and was solidified at 37 ​°C for 1 ​h. Then, HUVECs treated with either control or BREs for 48 ​h were digested by trypsin and 1 ​× ​10^4 ​cells were seeded into the 96-well plate. After incubation at 37 ​°C for 3 ​h, images were captured by microscope (IX81, Olympus) and the tube length was calculated using ImageJ software. 2.8. Sprouting assay Spheroid sprouting assay was performed to investigate the effects of BREs on angiogenesis in vitro. Briefly, HUVECs were detached from the cell culture plate using trypsin-EDTA and neutralized with cell culture medium. Cells were counted using a hemacytometer and cells were resuspended into control medium or medium containing BREs at a density of 20,000 ​cells/mL. The cells were mixed with 1 ​mL of methocel stock solution and were transferred to a sterile multichannel pipette reservoir. 25 ​μL of solution were pipetted onto a 10 ​cm square petri dish using pipette. Then the dish was inverted and incubated in cell culture incubator for 24 ​h. The spheroids were gently washed with 10 ​mL PBS and transferred into a 15 ​mL conical tube. Spheroids were centrifuged at 200 ​g for 5 ​min and then resuspended in 2 ​mL of methanol containing 20% FBS. 4 ​mL collagen solution were mixed with 0.5 ​mL 10x Medium 199, and the pH value was adjusted by adding sterile ice-cold 0.2 ​N NaOH. 2 ​ml of the collagen/Medium 199 solution were mixed with methanol solution containing spheroids. 1 ​mL of the spheroid-collagen solution was added into per well of a 24-well plate which was incubated at 37 ​°C for 30 ​min to allow the collagen to polymerize. Spheroids were simulated with 200 ​μL of ECM or ECM containing BREs for 24 ​h in a humidified incubator (37 ​°C, 5% CO2). Images were captured using a microscope (IX81, Olympus), and sprouting numbers or vascular length were calculated using ImageJ software. 2.9. Mice C57BL/6 mice (SIPPR-BK Experimental Animal Co. China) were kept in alternate dark-light cycles of 12 ​h at RT. All surgeries were performed under sodium pentobarbital anesthesia, and all efforts were made to minimize suffering. All animal experiments were conducted in agreement with the NIH Guide for the Care and Use of Laboratory Animals, and with the approval of the Institutional Animal Care and Use Committee at Shanghai Changhai Hospital (CHEC (A.E)2022–020). 2.10. Retinal vascular development To assess the effect of BREs on the postnatal development of retinal angiogenesis, control pups and pups received intraocular injections of 0.5 ​μg BREs at P5. The pups were euthanized, and the eyes were harvested at P7. Whole-mount retinas were stained with IB4, ERG, and EdU as described in the in vivo endothelial cell proliferation and retina immunofluorescence sections. The ratio of the vascular area to the retinal area was used to quantify superficial vascular layer development. The morphology and distribution of retinal vessels in P6 were assessed in three non-overlapping and randomly selected microscopic fields per retina and whole-mounted retina. 2.11. In vivo endothelial cell proliferation To observe the proliferation of endothelial cells in vivo, we performed proliferation analysis using the Cell-LightTM EdU Apollo567 Kit. EdU staining was performed according to the manufacturer's instructions. P6 and P17 mice were injected with EdU (5 ​mg/kg) and euthanized 4 ​h later. Eyes were enucleated and fixed in 4% PFA for 0.5 ​h at 30 ​min. The eyes were isolated under a microscope and blocked with 1% BSA buffer containing 0.5% Triton-X-100 for 30 ​min at RT. EdU labelling was performed according to the manufacturer's instructions. In the final step, the retinas were mounted on glass slides and sealed with a fluorescent mounting medium. A confocal microscope system (Leica TCS SP5-II) was used to capture the images, and Image J was used to quantify the avascular and neovascular areas. 2.12. Oxygen induced retinopathy Oxygen-induced retinopathy (OIR) was induced as we previously reported [[84]31]. The oxygen level was continuously monitored using an oxygen analyzer (XBS-03 ​S, HangZhou Aipu Instrument Equipment, Hangzhou, China). The pups were euthanized on P17, and their eyes were enucleated for further immunofluorescence assays. To avoid potential inter-litter variability, pups from each litter were randomized as follows: room air (raised by a surrogate mother P7), OIR neonates receiving RGD-EXO or BREs. A single dose of RGD-EXO or BREs (1 ​μg) was injected into the vitreous cavity of P12 mice. Data were analyzed using neonates from at least four different litters for each group. 2.13. RNA sequencing Cells in culture dishes were treated with 50 ​μg/mL BREs for 48 ​h and total RNA prepared for RNA-seq analyses were extracted from cells using Trizol reagent as described by the manufacturer (Invitrogen). RNA sequencing and analysis were performed at the Beijing Genomics Institute (Shenzhen, China). Pathway enrichment analysis was based on KEGG database ([85]http://www.genome.jp/kegg/) and GO database ([86]http://geneontology.org/), a rigorous algorithm has been developed for identifying differentially expressed genes between two samples based on “The significance of digital gene expression profiles”. Multiple tests and analyses were performed using the P value as a threshold of significance, and pathways with a P value ​< ​0.05 were considered significantly enriched in differentially expressed or modified genes. 2.14. Metabolomics Samples were prepared for metabolomic analysis as previously described [[87]32] Briefly, HUVECs treated with BREs or RGD-EXO for 24 ​h were suspended in 400 ​μL of an extraction solution (methanol: acetonitrile: water (v/v/v, 4:4:1)). Afterwards, cells were frozen in liquid nitrogen for 1 ​min and thawed in a 37 ​°C water bath. The above procedure was repeated three times. Then the samples were sonicated in an ultrasonic water bath for 20 ​min and incubated at −40 ​°C for 1 ​h. Next, samples were centrifuged at 12,000 ​g for 30 ​min at 4 ​°C and the supernatant was collected for analysis. Samples were analyzed using an ultra-high-performance liquid chromatography system (Vanquish, Thermo Scientific), chromatographic separation was achieved on an ultra-performance liquid chromatography (UPLC) system (Waters Corporation, Milford, MA, USA) with a Waters ACQUITY UPLC BEH AMIDE column, and a Thermo Q Exactive HFX mass spectrometer was used through the control software (Xcalibur, Thermo Scientific). 2.15. Western blotting Total protein from HUVECs was extracted using RIPA lysis buffer containing 2% protease inhibitor and 2% protease inhibitor (Beyotime, Shanghai, China). The total amount of protein in the lysate was measured by standard micro bicinchoninic acid (BCA) analysis according to the manufacturer's instructions. The samples were prepared in 5 ​× ​SDS-PAGE Sample Loading buffer and heated at 95 ​°C for 10 ​min before loading onto the gel. Protein samples and the protein marker (20 ​μL samples per lane) were separated on a 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) running gel at 120 ​V for 50 ​min. Then the proteins were transferred onto PVDF membranes at 100 ​V for 120 ​min. Blocking was performed in 5% non-fat milk for 2 ​h. The blots were incubated with primary antibodies overnight on a shaking table at 4 ​°C. The membranes were then rinsed thoroughly in TBST and subsequently incubated with horseradish peroxidase–conjugated goat anti-rabbit or anti-mouse antibodies at 1:5000 for 90 ​min at room temperature. Finally, the signals were detected with ECL Super Signal West Pico Chemiluminescent Substrate, visualized using the Gelview 6000Plus Image Capture System (Guangzhou Biolight Biotechnology Co., Ltd, Guangzhou, China), and quantified by gel analysis using ImageJ software. 2.16. Quantitative real-time PCR RNA was extracted and further purified using the RNase-Free DNase Set (QIAGEN) according to the manufacturer's instructions. cDNA was acquired using RNA reverse transcription kit (Beyotime Biotechnology) according to the manufacturer's protocol and stored at −80 ​°C for future use. RPLP0 was used as the internal reference gene to detect C-MYC, LDHA, and HK2. The following sequences of specific primers were used: C-MYC forward 5′- CCTGGTGCTCCATGAGGAGAC-3’; C-MYC reverse 5′- CAGACTCTGACCTTTTGCCAGG-3’; LDHA forward 5′- CCTGGTGCTCCATGAGGAGAC-3’; LDHA reverse 5′- CAGACTCTGACCTTTTGCCAGG-3’; HK2 forward 5′- GAGTTTGACCTGGATGTGGTTGC-3’; HK2 reverse 5′- CCTCCATGTAGCAGGCATTGCT-3’; PKM2 forward 5′- ATGGCTGACACATTCCTGGAGC-3’; PKM2 reverse 5′- CCTTCAACGTCTCCACTGATCG-3’. Relative gene expression were performed using Power SYBR Green PCR Master Mix (Applied Biosystems) according to the qRT-PCR Mixture reagent specifications. After evenly mixing and instantaneously centrifuging, qRT-PCR was performed at 50 ​°C for 2 ​min, 95 ​°C for 10 ​min, 95 ​°C for 15 ​s, 60 ​°C for 1 ​min, 95 ​°C for 15 ​s, 60 ​°C for 15 ​s, and 95 ​°C for 15 ​s, with a total of 40 cycles on a quantitative PCR instrument (LineGene 9600 Plus, Bioer Technology, Zhejiang, China). All samples were run in triplicates and averaged to define one biological replicate. The expression of target genes was normalized to that of the RPLP0 internal reference gene. 2.17. Retina immunofluorescence The eyes were enucleated and fixed in 4% paraformaldehyde (PFA) for 30 ​min at RT after which they were completely isolated under a microscope. The eyeballs were blocked with 5% BSA containing 0.05% Triton-X-100 for 30 ​min at RT. After being washed three times in PBS, the retinas were incubated with FITC-conjugated Isolectin B4 and indicated primary antibodies (1:250 dilution in 5% BSA and 0.05% Triton-X-100 in PBS) overnight at 4 ​°C. Retinas were then incubated with secondary antibodies (1:250 dilution in 5% BSA and 0.05% Triton-X-100 in PBS) for 2 ​h at RT. The retinas were washed another three times with PBS and incubated with DAPI (1:2000 dilution in 5% BSA and 0.05% Triton-X-100 in PBS) for 15 ​min at RT. In the final step, the retinas were mounted on glass slides and sealed with a fluorescent mounting medium. A confocal microscope system (Leica TCS SP5-II) was used to photograph the slides, and Image J was used to quantify the avascular and neovascular areas. 2.18. ATP production The intracellular ATP levels were determined with a commercial ATP assay kit (Beyotime, S0026B) according to the manufacturer's instructions. Briefly, 2 ​× ​10^5 HUVECs were seeded per well of 6-well plate and treated with BREs for 48 ​h. Then each well was given 200 ​μL RIPA buffer for cell lysis. Next, samples were centrifuged at 4 ​°C for 5 ​min at 12,000 ​g, and the supernatants were collected for further detection. After that, 100 ​μL ATP working solution and 20 ​μL supernatant from each group were added to a 96-well plate, and then the relative light unit value was determined with a luminometer. Finally, the ATP concentration in different groups were calculated according to the standard curve. 2.19. Lactate detection The extracellular lactate levels were determined using Lactic Acid assay kit (Nanjing Jiancheng Bioengineering Institute, A019-2-1) according to the manufacturer's instructions. Briefly, 20 ​μL of conditional medium of HUVECs was added to 1.2 ​mL of working solution prepared in advance and mixed thoroughly. The reaction was terminated with 2 ​mL terminator after being incubated at 37 ​°C for 10 ​min 100 ​μL of reaction mixtures were loaded into a 96-well plate and the absorption peak at 530 ​nm wavelength was measured. 2.20. Pyruvate detection The extracellular pyruvate levels were determined using Pyruvate assay kit (Nanjing Jiancheng Bioengineering Institute, A081-1-1) according to manufacturer's instructions. Briefly, 100 ​μL of conditional medium of HUVECs was added to 0.5 ​mL of working solution prepared in advance and mixed thoroughly. The reaction was terminated by adding 2.5 ​mL terminator after being incubated at 37 ​°C for 10 ​min. Finally, 100 ​μL of reaction mixtures were loaded into a 96-well plate and the absorption peak at 505 ​nm wavelength was measured. 2.21. Statistical analysis Statistical analysis was performed using GraphPad Prism (GraphPad software 9.0, GraphPad, Bethesda, MD, USA) and ImageJ software (Image J 2 ​× ​2.1.5.0, National Institutes of Health, USA). Unpaired Student's t-test was used for the data of two-group analysis, while comparisons between multiple groups were performed by one-way analysis of variance (ANOVA). All results are presented as the mean ​± ​standard deviation (SD) of at least three independent experiments. P values ​< ​0.05 were considered as statistically significant (∗P ​< ​0.05; ∗∗P ​< ​0.01; ∗∗∗P ​< ​0.001; ∗∗∗∗P ​< ​0.0001). 3. Results 3.1. Preparation and characterization of BPQDs@RGD-EXO (BREs) The construction of the BREs is illustrated in [88]Fig. 2a. First, the donor cell-assisted membrane modification strategy was used to improve the vascular targeting ability of exosomes. Integrins are commonly overexpressed on the surface of blood vessels [[89]33]. The Arg-Gly-Asp (RGD) peptide can specifically bind to integrins and is widely used as a vascular targeting molecule [[90]34]. Here, c (RGDyK) functionalized phosphatidylethanolamine 1, 2-distearoyl-sn-glycero-3-hosphoethanolamine-N-[RGD (polyethylene glycol)-2000] (DSPE-PEG-RGD) were used to modify the donor cell membrane, as previously reported [[91]30]. Human embryonic kidney 293 (HEK 293) cells were used as donor cells and cultured with DSPE-PEG-RGD to prepare RGD-modified exosomes (RGD-EXO). A traditional differential centrifugation method was used to isolate RGD-EXO from the cell supernatants. The characterization of EXO and RGD-EXO was performed using transmission electron microscopy (TEM), nanoparticle tracking analysis (NTA), and western blotting ([92]Fig. 2b–d). There were no significant differences in the physical characteristics between the two exosomes. The successful modification of RGD on exosomes was confirmed by H NMR spectra analysis ([93]Fig. S1). Fig. 2. [94]Fig. 2 [95]Open in a new tab Characterization of engineered BREs. a) Schematic diagram of BREs preparation. b) Western blot analysis of CD63, CD81, CD9, TSG101 and calnexin. c) The morphology of EXO, RGD-EXO and BREs determined by TEM. d) The Size of EXO, RGD-EXO and BREs evaluated using NTA. e) The quantity of phosphorus in BREs and control exosomes. Data was presented as means ​+ ​SD, n ​= ​3, two-tailed t-test (∗∗∗∗p ​< ​0.0001). f) Fluorescent images of HUVECs after being treated with BREs, EXO, RGD-EXO and BPQDs (scale bar 50 ​μm). Calcein acetoxymethyl staining live cells (Green) and Propidium iodide staining dead cells (Red). g) Statistical result of dead/live cell staining. Data was presented as means ​± ​SD, n ​= ​3, one-way ANOVA (∗∗∗p ​< ​0.001, ∗∗∗∗p ​< ​0.0001). BPQDs are toxic to vascular endothelial cells [[96]35]. In this study, we used electroporation to encapsulate BPQDs into RGD-EXO to minimize the potential toxicity of BPQDs. The prepared BPQDs exhibited a uniform morphology with a size of 4.49 ​nm, as analyzed by TEM ([97]Fig. S2). The BREs showed a cup-shaped structure with an average size of 105 ​nm ([98]Fig. 2c–d). The exosome-specific biomarkers CD63, CD81, CD9, TSG101, and cell-specific biomarker calnexin were confirmed using western blotting ([99]Fig. 2b). After the incorporation of BPQDs, BREs exhibited no significant physical differences from RGD-EXO ([100]Fig. 2c). The incorporation of the BPQDs was analyzed using inductively coupled plasma mass spectrometry (ICP-MS). The data showed that BREs contained a higher concentration of phosphorus than RGD-EXO, suggesting successful incorporation of the BPQDs in BREs ([101]Fig. 2e). The conjugation of RGD peptide onto BREs was confirmed using SDS-PAGE and there were staining lanes that lower than 10kD in BREs and RGD peptide ([102]Fig. S3). The encapsulation efficiency of BREs were calculated using ICP-MS analysis, which was about 13.6 ​± ​2.6%. The BPQDs release profile from BREs was evaluated by ICP-MS. The profile displayed that a relatively fast rate within first 36 ​h followed by mitigation. The cumulative release of BPQDs was 8.16% at 72 ​h ([103]Fig. S4), which revealed that BREs were sufficiently stable. Compared to the same concentration of BPQDs, a concentration of 50 ​μg/mL of BREs (20 ​μg/mL BPQDs) showed no toxicity to vascular endothelial cells ([104]Fig. 2f–g). 3.2. Good vascular targeting ability of BREs in vitro and in vivo HEK 293 ​cells express relatively low levels of integrins and were used as control cells. Human umbilical vein cells (HUVECs) are a widely used cell line for studying vascular function in vitro which express a high level of integrins. BREs were stained with DiD reagent, and the targeting ability of BREs was assessed using fluorescence microscopy. Since cells can internalize exosomes easily, we incubated cells with BREs for half an hour to minimize the effect of internalization and to better observe the targeting ability. To exclude the possible effect of cell internalization, we first compared the targeting ability of DiD-stained exosomes with that of DiD-stained BREs. The result showed that DiD-stained BREs exhibited much higher fluorescence intensities after incubation with HUVECs ([105]Fig. S5). These results suggest that the cells did not internalize many exosomes in a short time. We then incubated BREs with HEK293 ​cells and HUVECs to assess the vascular targeting ability of BREs in vitro. The data showed that the cytoplasm of HUVECs exhibited much higher fluorescence intensity than that of HEK293 ​cells ([106]Fig. 3a), indicating the good vascular targeting ability of BREs in vitro. Physiological and pathological retinal angiogenesis mouse models were used to study the vascular-targeting ability of BREs in vivo, while DiD-stained exosomes were used as controls. Isolectin B4 is a widely used fluorescent dye for specific vascular staining, and ETS transcription factor ERG is commonly used for vascular endothelial nuclear staining [[107]27]. Our data indicated that in the physiological mouse model, BREs were mainly co-localized with the retinal vasculature ([108]Fig. 3b). In the oxygen-induced retinopathy model, it was shown that BREs mainly colocalized with neovascular tufts and retinal vasculature ([109]Fig. 3c). Neovascular tufts are the major pathological changes in oxygen-induced retinopathy, and the good targeting ability of BREs exhibits potential for targeted therapy. Fig. 3. [110]Fig. 3 [111]Open in a new tab Vascular targeting ability of BREs. a) Confocal fluorescent microscopy images of HEK293 ​cells and HUVECs incubated with BREs (scale bar 50 ​μm). b) Fluorescent images of mouse retinal vascular stained with IB4 (white), ERG (red) and DiD-BREs (green) (scale bar 200 ​μm). c) Fluorescent images of mouse retinal vascular stained with IB4 (white), ERG (red) and DiD-BREs (green) in oxygen-induced retinopathy model (scale bar 200 ​μm). 3.3. Anti-angiogenic function of BREs in vitro An EdU proliferation assay was used to explore the effects of BREs on the growth of HUVECs. Unlike the cell counting kit 8 assay, which examines enzyme activity to reflect cell numbers indirectly, the EdU assay stained the DNA of proliferating cells directly [[112]36]. The effects of BREs, EXO and RGD-EXO on the proliferation of HUVECs were initially assessed. The data showed that compared with EXO and RGD-EXO groups, BREs suppressed the growth of HUVECs ([113]Fig. 4a–b). The toxicity of the BREs was evaluated using cell death/live staining. At the concentration of 50 ​μg/mL, the BREs did not induce apparent cell toxicity ([114]Fig. 2f). Cell migration is a key process in vascular growth, and the effect of BREs on cell migration was evaluated using a wound healing assay. The data indicated that BREs inhibited HUVECs migration significantly ([115]Fig. 4c–d). Furthermore, the assembly of the cell cytoskeleton was investigated because it plays a vital role in cell migration. The data showed that BREs treatment disrupted the morphology of F-actin but had no effect on α-tubulin ([116]Fig. S6). Fig. 4. [117]Fig. 4 [118]Open in a new tab Anti-angiogenic role of BREs in vitro. a) Inhibition of cell proliferation by BREs. EdU positive cells (Red), Hochest positive cells (Blue), (scale bar 100 ​μm). b) Statistical result of cell proliferation. Data was presented as means ​± ​SD, n ​= ​5, one-way ANOVA (∗∗∗p ​< ​0.001). c) Inhibition of cell migration by BREs (scale bar 50 ​μm). d) Statistical result of cell migration. Data was presented as means ​± ​SD, n ​= ​5, one-way ANOVA (∗∗∗∗p ​< ​0.0001). e) Inhibition of cell tube formation by BREs (scale bar 50 ​μm). f) Statistical result of cell tube formation. Data was presented as means ​± ​SD, n ​= ​5, one-way ANOVA (∗∗∗∗p ​< ​0.0001). g) Inhibition of cell sprouting by BREs (scale bar 50 ​μm). h) Statistical result of cell sprouting. Data was presented as means ​± ​SD, n ​= ​5, one-way ANOVA (∗∗p ​< ​0.01). Tube formation is a widely used method to evaluate the angiogenic potential of vascular endothelial cells. Here, we found that BREs significantly attenuated the tube formation ability of HUVECs ([119]Fig. 4e–f). The cell sprouting assay is a more advanced method to assess the angiogenic ability of vascular endothelial cells that involves cell proliferation and migration in a three-dimensional space [[120]37]. Our data showed that BREs apparently inhibited HUVECs sprouting, as they inhibited both the length and number of sprouted cells ([121]Fig. 4g–h). These results suggest that BREs are biocompatible and show good antiangiogenic potential in vitro. 3.4. BREs suppress retinal vascular development The development of the retinal vasculature in mice is largely different from that in humans. When a child is born, the retinal vasculature is fully developed. However, the retinal vasculature in mice has just begun to develop after birth. It takes approximately eight days for the retinal vasculature to fully cover the retina ([122]Fig. 5a) [[123]38]. Therefore, retinal vascular development in mice has been widely used to study the angiogenic potential in vivo. Our data showed that compared with EXO and RGD-EXO, BREs could significantly suppress the development of the retinal vasculature ([124]Fig. S7, [125]Fig. 5b–c). Tip cells are leading cells at the tips of vascular sprouts, which are highly proliferative and guide the growth of vascular [[126]39]. The results revealed that the number of tip cells and vascular length were apparently decreased in the sprouting region of the retinal vasculature in BRE-treated retinas ([127]Fig. 5d–e). Endothelial cell proliferation was assessed using the EdU reagent. The data showed that BREs significantly suppressed endothelial cell proliferation ([128]Fig. 5f–h). Meanwhile, BREs did not induce apparent retinal toxicity in mice at postnatal eight days, as there was no change in retinal morphology ([129]Fig. S8). Fig. 5. [130]Fig. 5 [131]Open in a new tab Suppression of retinal vascular growth by BREs. a) Schematic diagram of mouse retinal vascular development. b) Retinal vascular growth delayed in BREs treated mice. A, artery; V, vein (scale bar 1000 ​μm). c) Statistical result of retinal vascular growth. Data was presented as means ​± ​SD, n ​= ​4, two-tailed t-test (∗∗∗p ​< ​0.001). d) Confocal images showing immunostaining of retinal vasculatures. The arrows indicate staining of tip cells (scale bar 200 ​μm). e) Statistical result of tip cells. Data was presented as means ​± ​SD, n ​= ​4, two-tailed t-test (∗∗∗p ​< ​0.001). f) Immunofluorescence staining of Edu (green), ERG (red) and IB4 (white) in P6 mouse retinas. Proliferating ECs are shown in yellow (EdU and ERG double-positive) (scale bar 200 ​μm); g) Statistical result of vascular density. Data was presented as means ​± ​SD, n ​= ​4, two-tailed t-test (∗∗∗p ​< ​0.001). h) Statistical result of proliferating ECs. Data was presented as means ​± ​SD, n ​= ​4, two-tailed t-test (∗∗p ​< ​0.01). 3.5. BREs improves oxygen induced retinopathy The oxygen-induced retinopathy (OIR) mouse model mimics the pathological progression of retinopathy of prematurity in humans (ROP) [[132]40]. As shown in [133]Fig. 6a, the mice aged postnatal seven days (P7) were put into hyperoxia for five days, which led to the regression of retinal vasculature. Then, the mice aged postnatal twelve days (P12) were placed in room air, which resulted in relatively low oxygen levels. Relatively hypoxic conditions could lead to pathological vascular and non-perfusion areas in the retina of mice, which peaked on postnatal seventeen days (P17). Our data showed that compared with EXO, RGD-EXO and BPQDs incorporated exosomes, BREs significantly attenuated pathological neovascularization and avascular areas ([134]Fig. S9, [135]Fig. 6b–d). The proliferated endothelial cells contributed to the formation of neovascular tufts, and our data indicated that BREs could suppress endothelial cell proliferation ([136]Fig. 6e–f). The toxicity of BREs on the retina of P17 mice was evaluated using hematoxylin and eosin staining. The data showed that BREs did not induce any apparent retinal morphological changes ([137]Fig. S10). Systematic toxicity was further evaluated using hematoxylin and eosin staining. The results indicated that BREs did not induce apparent toxicity in the heart, lungs, liver, or kidneys ([138]Fig. S11). Furthermore, none of the serum indicators of renal and hepatic function were changed by BREs ([139]Fig. S12), which proved the biocompatibility of BREs in vivo. The therapeutic efficiency and toxicity of BPQDs were also studied. The results showed that compared with control group, BPQDs could improve retinal avascular area and inhibit retinal neovascularization. However, the therapeutic effect was worse than that of BREs ([140]Fig. S13a-c). BPQDs treatment also induced retinal toxicity as the retina became thinner in BPQDs group ([141]Fig. S13d). These data indicated that BREs were biocompatible and possess better therapeutic efficiency than BPQDs. Fig. 6. [142]Fig. 6 [143]Open in a new tab Amelioration on oxygen-induced retinopathy by BREs. a) Schematic diagram of oxygen-induced retinopathy in mice (scale bar 1000 ​μm). b) Oxygen-induced retinopathy improved in BREs treated mice (scale bar 1000 ​μm). c) Statistical result of avascular area. Data was presented as means ​± ​SD, n ​= ​5, two-tailed t-test (∗∗p ​< ​0.01). d) Statistical result of neovascular tufts. Data was presented as means ​± ​SD, n ​= ​5, two-tailed t-test (∗∗p ​< ​0.01). e) Immunofluorescence staining of EdU (green), ERG (red) and IB4 (white) in P17 mouse retinas. Proliferating ECs are shown in yellow (EdU and ERG double-positive) (scale bar 200 ​μm). f) Statistical result of proliferating ECs. Data was presented as means ​± ​SD, n ​= ​5, two-tailed t-test (∗p ​< ​0.05). 3.6. Combined analysis of RNA-seq and metabolomics BPQDs exhibited broad absorption across the visible-light region, which shows great potential for photothermal therapy. However, easy oxygenation is a major challenge for further applications. It has been reported that BPQDs@EXO is much more stable than BPQDs, as exosomes prevent the reaction between BPQDs and oxygen [[144]23]. The role of BREs in anti-angiogenic therapy result from the highly active properties of BPQDs. BRE-treated HUVECs showed altered mitochondrial morphology ([145]Fig. 7a), suggesting that BREs might affect cellular energy metabolism. To further elucidate the possible mechanism by which BREs disrupt energy metabolism, we performed RNA sequencing and metabolomics. There were 1871 differentially expressed genes in total after treatment with BREs. Among them, 359 genes were upregulated, and 1512 genes were downregulated ([146]Fig. 7b). Meanwhile, the Glycolysis/Gluconeogenesis pathway was downregulated in BREs treated group ([147]Fig. S14). Further analysis indicated that the expression of HK2 significantly decreased ([148]Fig. 7c). HK2 catalyzes the first committed step through the phosphorylation of glucose, and its inhibited expression leads to an energy crisis in the vascular endothelial cells. Without sufficient energy, proliferation, migration, sprouting, and angiogenesis in vitro and in vivo will be suppressed. The RNA-seq data also showed that the upstream oncogene c-MYC was inhibited after treatment with BREs ([149]Fig. 7c). However, the other rate-limiting enzymes, pyruvate kinase isozyme typeM2 (PKM2) and lactate dehydrogenase (LDHA) were not affected. These results suggest that BREs mainly target the c-MYC/HK2 pathway to disrupt glucose metabolism. Fig. 7. [150]Fig. 7 [151]Open in a new tab RNA-seq and metabolomics suggesting a role of glucose metabolism. a) TEM images of BREs treated HUVECs (scale bar 5 ​μm for up panel, scale bar 500 ​nm for down panel). b) Volcano plot of Log2 fold-changed genes for BREs treatment. c) Heatmap of interested genes. Red for up-regulation; Blue for down-regulation. c-Myc and HK2 were down-regulated. d) Volcano plot of Log2 fold-changed metabolites for BREs treatment. e) Phosphoenolpyruvic acid, galactose 1-phosphate, and arbutin were significantly downregulated. Metabolomics revealed a total of 355 metabolites, including 149 upregulated metabolites and 206 downregulated metabolites, in BREs treated group ([152]Fig. 7d). The glycolysis/gluconeogenesis pathway was also downregulated in BREs treated group ([153]Fig. S15). Among these, phosphoenolpyruvic acid, galactose 1-phosphate, and arbutin were significantly downregulated ([154]Fig. 7e). These metabolites were produced during glucose metabolism and the data indicated that glucose metabolism was suppressed in BREs treated group. Together with the RNA-seq data showing that HK2 was downregulated, it was concluded that BREs might target glucose metabolism to suppress retinal angiogenesis. 3.7. Underlying mechanisms in vivo and in vitro The expression of HK2 and c-Myc after BREs treatment was further confirmed using real-time quantitative PCR (qPCR). The results showed that BREs suppressed the expression of HK2 and c-Myc mRNA significantly ([155]Fig. 8a). The protein levels of HK2 and c-Myc were determined by western blotting. Consistent with the mRNA data, BREs inhibited the expression of HK2 and c-Myc at the protein level ([156]Fig. 8b–c). Furthermore, the upstream signaling pathway of c-Myc was assessed. The data showed that BREs could suppress the phosphorylation of ERK. Meanwhile, the expression of other glucose metabolic enzymes, PKM2 and LDHA, was also evaluated, and the data showed that there was no change after treatment with BREs ([157]Fig. S16). These results are consistent with those of the RNA-seq. Pyruvate and lactate contents were detected using commercial kits. The results indicated that BREs inhibited the production of pyruvate and lactate, which further proved that BREs targeted glucose metabolism ([158]Fig. 8d–e). ATP production was evaluated after treatment with the BREs. The results showed that BREs could significantly decrease the production of ATP in HUVECs ([159]Fig. 8f), confirming that BREs disrupt cellular energy production. Furthermore, the expression of HK2 and c-Myc was evaluated in the retina of the OIR model. [160]Fig. 8g–h shows that HK2 and c-Myc expression were upregulated in the retina of the OIR model, which could be inhibited by BREs. Similar results were obtained using immunofluorescence ([161]Fig. 8i). This data suggests that BREs target cellular glucose metabolism via the c-Myc/HK2 pathway to attenuate retinal angiogenesis. Fig. 8. [162]Fig. 8 [163]Open in a new tab BREs attenuate retinal angiogenesis via disrupting glucose metabolism. a) The mRNA levels of HK2 and c-MYC in BREs treated HUVECs were determined by qPCR. Data was presented as means ​+ ​SD, n ​= ​5, one-way ANOVA (∗∗p ​< ​0.01) b) The protein levels of HK2, c-MYC, ERK and p-ERK in BREs treated HUVECs were determined by western blot. c) Statistical result of the protein levels. Data was presented as means ​± ​SD, n ​= ​5, one-way ANOVA (∗∗∗p ​< ​0.001). d) Pyruvate level in BREs treated HUVECs was assessed. Data was presented as means ​± ​SD, n ​= ​5, one-way ANOVA (∗∗∗∗p ​< ​0.0001). e). Lactate level in BREs treated HUVECs was assessed. Data was presented as means ​± ​SD, n ​= ​5, one-way ANOVA (∗∗∗p ​< ​0.001). f) Statistical result of ATP production in BREs treated HUVECs. Data was presented as means ​± ​SD, n ​= ​5, one-way ANOVA (∗∗∗∗p ​< ​0.0001) g) The protein levels of HK2 and c-Myc in retina of OIR mice were determined by western blot. h) Statistical result of protein levels of HK2 and c-MYC. Data was presented as means ​± ​SD, n ​= ​5, one-way ANOVA (∗∗p ​< ​0.01, ∗∗∗p ​< ​0.001, ∗∗∗∗p ​< ​0.0001). i) Immunofluorescence staining of HK2 (magenta), c-MYC (red) and IB4 (green) in P17 mouse retinas of OIR (scale bar 50 ​μm). 4. Discussion BPQDs have been intensively studied in nanomedicine because of their unique physicochemical properties, and have been used as photothermal reagents, fluorescent dyes, biosensor, and drug delivery platforms [[164]10,[165][41], [166][42], [167][43]]. Physical damage resulting from the photothermal effect has been reported by many groups for the treatment of tumor growth [[168]12,[169]23,[170]44]. BP materials are easily oxidized and degraded to non-toxic phosphates [[171]45]. However, BPQDs have been reported to be toxic to vascular endothelial cells, lung-derived cells, bone marrow nucleated cells, and the kidneys [[172][13], [173][14], [174][15], [175][16],[176]35]. This controversy is due to the chemically active nature of BPQDs, which could cause toxicity to the cell membrane before they are internalized into cells. BPQDs apparently induced cellular lipid peroxidation, partially supporting this hypothesis [[177]16]. Therefore, dedicated decoration with BPQDs to reduce their toxicity to cell membranes is of vital importance for further biomedical applications. Polydopamine modification and soybean phospholipid decoration are shown to reduce toxicity and improve stability of black phosphorus materials [[178][46], [179][47], [180][48], [181][49]]. Meanwhile, phosphorus element is easily oxidized into phosphate and surface oxidation of phosphorus can protect black phosphorus from further oxidation by forming a protective layer [[182]50]. Recently, the use of cell-derived extracellular vesicles to package nanomaterials has shown multiple advantages as they can reduce toxicity, improve stability, and strengthen the physicochemical properties of nanomaterials [[183]23,[184]51,[185]52]. Here, we showed that RGD peptide-modified exosomes could reduce the toxicity and improve the stability of BPQDs. In addition, the presence of the RGD peptide could mediate targeted vascular therapy of BPQDs, which could further reduce toxicity to other tissues. Pathological angiogenesis is a common feature of various diseases, such as tumors, rheumatoid arthritis, degenerative arthritis, retinopathy of prematurity and diabetic retinopathy [[186]53]. Retinal angiogenesis models, including retinal vascular development, oxygen-induced retinopathy, and streptozotocin-induced diabetic retinopathy, are widely used to study angiogenesis. In the present study, we used retinal vascular development and oxygen-induced retinopathy models to evaluate the anti-angiogenic potential of BREs. These results suggest that BREs inhibit retinal angiogenesis. Our data also showed that the BREs reduced avascular area in the OIR model, which represented improvement of the other hallmark of pathological retinopathy. The inhibition of excessive ECs proliferation contributes to the restoration of retinal vascularization in the central retinal zone [[187]54,[188]55]. This is attributed to a reduction rather than complete inhibition of ECs function as reported [[189]56]. Both the inhibition of neovascularization and reduction of avascular area are beneficial for improving pathological retinopathy. Meanwhile, the modification of RGD on the surface of exosomes improved the therapeutic efficiency of BPQDs-incorporated exosomes ([190]Fig. S9). The anti-angiogenic role of BREs in vascular endothelial cells was evaluated using cell proliferation, cell migration, tube formation, and cell sprouting assays. These in vitro results were consistent with those obtained in vivo. These data suggest that BREs could be biocompatible nanomedicines for the treatment of pathological angiogenesis. Glucose metabolism is the major energy source for vascular endothelial cells and is an engine for angiogenesis [[191]27]. Multiple growth factors such as vascular endothelial growth factor (VEGF), epidermal growth factor (EGF), and fibroblast growth factor (FGF) drive angiogenesis progression [[192]57]. These growth factors activate various signaling pathways that contribute to pathological angiogenesis. The sprouting of blood vessels is similar to that of a car that requires a driver and an engine to drive. Different growth factors are similar for different drivers, and an alternate can be found all the time [[193]58]. However, cell metabolism is similar to the engine of a car, the broken part of which can stop the car immediately [[194]58]. Therefore, targeting cell metabolism will be a promising option for anti-angiogenic therapy. In the present study, RNA-seq and metabolomics were used to clarify the possible mechanisms by which BREs inhibit angiogenesis. These data suggest that BREs target ECs glucose metabolism by disrupting the expression of c-Myc/HK2. These results were further confirmed both in vitro and in vivo. Approximately 75% of ATP is produced via glucose metabolism in vascular endothelial cells, the inhibition of which leads to a significant suppression of angiogenesis [[195]27]. HK2 is the most active isozyme of hexokinases, which primes glucose for intracellular utilization. The decreased activity of HK2 disrupts glycolysis, glucose oxidation, and ATP production. BREs target the expression of HK2 and the subsequent glucose metabolism in vascular endothelial cells, which could be an alternative for the treatment of pathological angiogenesis. 5. Conclusion In summary, we designed a vascular targeting and biocompatible BREs and evaluated its anti-angiogenic properties in vitro and in vivo. BREs significantly improve the stability and reduce the toxicity of BPQDs. In addition, BREs exhibited good vascular-targeting ability in vitro and in vivo. BREs inhibited the proliferation, migration, tube formation, and sprouting of ECs and suppressed retinal vascular development and oxygen-induced retinopathy in mice. Combined RNA-seq and metabolomic analysis revealed that BREs mainly disrupted glucose metabolism through the c-Myc/HK2 pathway. Our study suggests that engineered BREs show great therapeutic potential in pathological angiogenesis, which could be a promising alternative for anti-VEGF therapy. Credit author statement Funding and study supervision: H·S, W·S; Study design: H·S, X. G, W·H, H. Z; Animal studies design: X. G, R. Z, Q. L; Data acquisition: X. G, Q. L, H. Z, W. Z, Z. N, J. Z, X. C, X. W; Data analysis: X. G; X. W; R. Z, W·H; Writing—original draft preparation: X. G, X. W; writing—review and editing: H·S, W·S, X.W. All authors have read and agreed to the published version of the manuscript. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements