PROTAC-based nanoassemblies targeting BRD4 for potentiate FLASH radiosensitization therapy

Abstract

Ultrahigh dose-rate radiotherapy (FLASH-RT) represents a rapidly emerging radiotherapy modality with improved normal tissue protection. However, tumor radioresistance remains a major obstacle limiting its clinical potential. To address this, we developed a novel redox-responsive, PROTAC-based nanoplatform (APF) formed by the self-assembly of folate-conjugated PEG₂₀₀₀–ARV-771 (FA-PEG₂₀₀₀-ARV-771). Following FA-mediated targeting and endocytic uptake by tumor cells, APF nanoparticles underwent glutathione-triggered cleavage in the reductive tumor microenvironment, releasing ARV-771 which promoted the proteasomal degradation of BRD4. The reduction of BRD4 disrupted the BRD4-c-Myc-RAD51AP1 signaling pathway and damaged the DNA repair pathway that resulting in a significant sensitizing effect of FLASH radiotherapy using Petal Accelerator platform. Both in vitro and in vivo experiments demonstrated that APF-assisted FLASH-RT markedly induced tumor cell apoptosis and necrosis, effectively inhibiting the malignant progression of tumors. Additionally, APF significantly enhanced FLASH-RT-induced tumor cell killing by promoting intracellular reactive oxygen species (ROS) generation and exacerbating DNA double-strand breaks. Transcriptomic analysis further revealed that APF-mediated BRD4 degradation suppressed key DNA repair-related genes. As a result, the APF@FLASH-RT combination achieved superior tumor inhibition in vivo with minimal systemic toxicity. This work provides a promising nanomedicine strategy for enhancing FLASH radiotherapy efficacy through targeted protein degradation.

Graphical abstract

1. Introduction

Radiotherapy (RT) remains a pivotal modality for the treatment of most localized solid malignancies[[1], [2], [3], [4]]. However, its efficacy is frequently constrained by the narrow therapeutic window, as dose escalation inevitably results in collateral damage to adjacent normal tissues and organs[[5], [6], [7], [8], [9], [10]]. Ultrahigh dose-rate (FLASH) radiotherapy is an emerging radiation technique which can effectively irradiate the tumor target area with ultra-fast radiation of over 40 Gy/s[[11], [12], [13], [14], [15]]. This strategy has attracted attention as it significantly reduces RT time, minimizes side effects for normal tissue while suppressing tumor growth as efficiently as conventional radiotherapy[[16], [17], [18], [19], [20]]. Nevertheless, the intrinsic radioresistance of certain tumors remains a major limitation impairing the effects of radiotherapy [21,22]. This resistance is frequently driven by the compensatory upregulation of homologous recombination (HR) factors, such as RAD51 and RAD51AP1, which accelerate DNA damage repair (DDR) post-irradiation and thereby attenuate radiosensitivity[[23], [24], [25]]. As a critical regulator of tumor transcription and DNA repair, bromodomain-containing protein 4 (BRD4) represents a compelling target for potent radiosensitization[[26], [27], [28], [29]] (see Scheme 1).

Scheme 1.

(A) Schematic illustration of FLASH radiotherapy coupled with folic acid (FA)-functionalized APF nanoparticles for targeted tumor delivery and enhanced therapeutic efficacy. (B) Upon intravenous injection, the APF nanoparticles actively targeted HeLa tumors by recognizing the overexpressed folate receptor (FR) on the surface of tumor cells, and then entered the cells through endocytosis. Under elevated glutathione (GSH) conditions within the tumor microenvironment, APF released the BRD4 degrader ARV-771 which disrupted the BRD4-cMyc-RAD51AP1 signaling axis, impaired DNA repair ability and synergistically inhibited the MAPK and PI3K-AKT as well as Wnt signaling pathways, thereby enhancing the anti-tumor effect of FLASH-RT.

Conventional small-molecule BRD4 inhibitors are primarily limited by incomplete target blockade and rapid compensatory accumulation of the protein, thereby compromising their radiosensitizing potential [30,31]. Proteolysis-Targeting Chimera (PROTAC) technology offers a revolutionary alternative by hijacking the cellular ubiquitin-proteasome system to achieve the irreversible degradation of protein of interest (POI) rather than mere functional inhibition[[32], [33], [34], [35], [36], [37], [38]]. However, the application of PROTACs is often impeded by their hydrophobicity, unfavorable pharmacokinetics and off-target toxicity resulting from insufficient tumor-specific delivery [[39], [40], [41], [42], [43]]. Therefore, it remains a formidable challenge to develop tumor-specific and highly efficient PROTAC strategies with minimal side effects to achieve effective radiosensitization.

Herein, we identified that while FLASH-RT directly inflicted lethal DNA double-strand breaks, it simultaneously triggered a p21-dependent cell cycle arrest, which subsequently drived a compensatory upregulation of RAD51AP1 in cervical cancer, thereby promoting adaptive radioresistance. To address this, we developed a stimuli-responsive, self-assembling PROTAC nanoparticle for the targeted degradation of BRD4 to synergistically sensitize FLASH-RT. Upon glutathione (GSH)-triggered intracellular release, the PROTAC effectively inhibited RAD51AP1 induction, thereby impairing DNA repair kinetics and enhancing tumor-cell sensitivity to radiation-induced DNA damage. This approach presents a novel strategy for the FLASH radiotherapy.

2. Materials and methods

2.1. Materials and cell culture

Detailed information regarding the materials, synthesis and characterization of the compounds in this study were provided in the Supporting Information.

Human cervical cancer cell lines (HeLa) and human umbilical vein endothelial cells (HUVEC) were obtained from American Type Culture Collection (ATCC; Manassas, VA, USA). Both cell lines were cultured in DMEM supplemented with 10% FBS and 1% penicillin-streptomycin (Gibco, Grand Island, NY, USA) at 37 °C in a humidified atmosphere containing 5% CO₂.

2.2. Parameters of FLASH radiotherapy

In this work, FLASH radiotherapy was delivered as a high energy X-ray modality using a petal accelerator developed by the China Academy of Engineering Physics. The accelerator was operated in an 8 MeV electron mode, producing a mean dose rate of 94 Gy s⁻¹. The source skin distance (SSD) was 16.5 cm for HeLa cells and 6.45 cm for HeLa-bearing mice. The Absolute dose measurements were performed with Gafchromic™ EBT3 radiochromic films (Ashland Inc., Covington, KY, USA).

2.3. Preparation and characterization of APF

The targeted PROTAC conjugate, FA-PEG₂₀₀₀-ARV-771 was synthesized following the synthetic pathway illustrated in Scheme S1. Detailed experimental procedures and characterization details were provided in the Supporting Information. The chemical structures of the target compound and key intermediates were confirmed by ¹H NMR and ¹³C NMR spectra, and the corresponding spectra were presented in Figs. S1–S5.

The self-assembled APF was synthesized using a nanoprecipitation method. Briefly, FA-PEG₂₀₀₀-ARV-771 (4.0 mg) was dissolved in 0.5 mL of N,N-Dimethylformamide (DMF) and subsequently added dropwise to 4.0 mL of stirring ultrapure water at room temperature. The mixture was purified by dialysis against ultrapure water (MWCO = 10.0 kDa) for 48 h, obtained APF. The morphology and size of APF were characterized by Transmission Electron Microscopy (TEM; JEM-2100plus, JEOL, Japan). The hydrodynamic diameter and polydispersity index (PDI) of APF were determined at room temperature using dynamic light scattering (DLS; NICOMP N3000, Alpharmaca, China). The chemical structures of the synthesized compounds were further confirmed by ¹H NMR and ¹³C NMR spectra recorded on a 400 MHz spectrometer (Bruker, Billerica, MA, USA).

The stability of APF was investigated in PBS, Simulated Body Fluid (SBF) and Dulbecco's modified Eagle medium (DMEM; Gibco, New York, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, New York, USA) at 37 °C. At different time points (0, 3, 6, 12, 24, 48 and 72 h), the particle sizes were measured by DLS. The reductive-responsive release of ARV-771 from APF was systematically evaluated. Briefly, 2 mL of APF dispersion (1 mg/mL) was dialyzed (MWCO, 10.0 kDa) against 20 mL of phosphate-buffered saline (PBS) containing either 0 or 10 mM glutathione (GSH). At various time intervals (0, 1, 3, 5, 8, 10, 12, 24, 36 and 48 h), 2 mL of the external release medium was withdrawn for quantitative analysis using a UV–Vis spectrophotometer (UH5300, Hitachi, Japan), followed by replenishment with an equal volume of fresh medium to maintain a constant total volume of 20 mL.

2.4. Cell viability assay

HeLa and HUVEC cells were seeded into 96-well plates at a density of 5 × 10³ cells per well and allowed to adhere overnight. The cells were then treated with APF at various concentrations for 48 h. Cell viability was subsequently assessed using a standard Cell Counting Kit-8 (CCK-8) assay.

2.5. Cellular uptake study

Did-labeled APF (Did@APF) was prepared by introducing Did during the self-assembly process. For qualitative visualization, HeLa cells were seeded in confocal dishes at a density of 5.0 × 10⁴ cells per dish and cultured overnight. The culture medium was then replaced with fresh medium containing Did@APF and incubated for 2 h and 5 h, respectively. After incubation, the cells were washed with PBS, stained with DAPI and subsequently imaged using a laser scanning confocal microscopy (CLSM; Leica, stellaris 5, Germany). For quantitative uptake analysis, HeLa cells were seeded in 6-well plates (2.0 × 10⁵ cells per well) and treated under identical conditions. Following treatment, the cells were washed with PBS, harvested to obtain a single-cell suspension and analyzed for fluorescence intensity by flow cytometry (FCM; BD FACS Celesta, USA).

2.6. Live-dead cell assay

HeLa cells were seeded in confocal dishes at a density of 2.0 × 10⁵ cells per dish and allowed to adhere overnight. To evaluate treatment-associated cytotoxicity, the cells were first incubated with APF for 4 h and subsequently subjected to FLASH-RT (FRT) at a dose of 6 Gy. After treatment, cell viability was assessed using a fluorescein diacetate/propidium iodide (FDA/PI) live/dead staining kit (obtained from Beijing Solarbio Science & Technology Co., Ltd.) according to the manufacturer's protocol. Viable cells exhibited green fluorescence (FDA), while dead cells displayed red fluorescence (PI). Representative fluorescence images were acquired using a CLSM.

2.7. Colony formation assay

HeLa cells were seeded in 6-well plates at a density of 2.0 × 10³ cells per well and allowed to adhere overnight. The cells were pretreated with APF for 4 h, followed by irradiation with FRT at a dose of 6 Gy. After treatment, the cells were further cultured for 10 days to allow colony formation. Colonies were then gently washed with PBS, fixed with 4% paraformaldehyde for 20 min and stained with 1% crystal violet for 15 min. Excess dye was removed by rinsing with PBS three times, and the stained colonies were air-dried prior to imaging.

2.8. Scratch assay

HeLa cells were seeded in 6-well plates at a density of 3.0 × 10⁵ cells per well and cultured overnight. The cells were pretreated with APF for 4 h, followed by irradiation with FRT at a dose of 6 Gy. After treatment, a straight scratch was subsequently generated across the cell monolayer in each well using a sterile 200 μL pipette tip. After gently washing twice with PBS to eliminate detached cells, fresh medium was replenished. The migratory progress was monitored using an inverted optical microscope, with images captured at both 0 h and 24 h post-scratching. The wound closure area was quantified using ImageJ software.

2.9. Cell apoptosis assay

HeLa cells were seeded in 6-well plates at a density of 2.0 × 10⁵ cells per well and cultured overnight. The cells were pretreated with APF for 4 h, followed by irradiation with FRT at a dose of 6 Gy. After treatment, the cells were harvested and processed using an Annexin V-FITC/7-AAD apoptosis detection kit (APExBIO, Houston, USA) according to the manufacturer's instructions. Flow cytometric analysis was conducted to quantify the levels of apoptosis.

2.10. Intracellular ROS detection assay

HeLa cells were seeded in confocal dishes at a density of 2.0 × 10⁵ cells dish well and cultured overnight. Cells were preincubated with APF for 4 h and subsequently exposed to FRT at a dose of 6 Gy. After irradiation, the culture medium was removed and replaced with 1 mL of DMEM containing the DCFH-DA probe (APExBIO, Houston, USA), followed by incubation in the dark for 30 min. The cells were then washed with serum-free medium and visualized using CLSM. Semi-quantitative analysis of intracellular ROS levels was performed by measuring the mean fluorescence intensity by ImageJ software.

2.11. Intracellular DNA damage assay

HeLa cells were seeded in confocal dishes at a density of 1.5 × 10⁵ cells dish well and cultured overnight. Cells were preincubated with APF for 4 h and subsequently exposed to FRT at a dose of 6 Gy. After treatment, the cells were collected and stained using a γ-H2AX DNA Damage Detection Kit (APExBIO, Houston, USA) according to the manufacturer's instructions. The samples were imaged by CLSM and the mean fluorescence intensity was quantified with ImageJ software.

2.12. Immunofluorescence staining of BRD4 expression

HeLa cells were seeded in confocal dishes at a density of 1.5 × 10⁵ cells dish well and cultured overnight. Cells were preincubated with APF for 4 h and subsequently exposed to FRT at a dose of 6 Gy. After 24 h, cells were washed three times with PBS, fixed with 4% paraformaldehyde for 10 min, permeabilized with 0.1% Triton X-100 for 10 min and blocked with 5% bovine serum albumin (BSA) for 1 h. For BRD4 expression, the samples were then incubated with Alexa Fluor 647-conjugated anti-BRD4 antibody (1:250; Abcam, ab197608) for 1 h at room temperature, followed by staining with phalloidin-Alexa Fluor 488 (1:200; Beyotime, C2201S) for 30 min in the dark. For RAD51AP1 expression, the samples were then incubated with anti-RAD51AP1 antibody (1:250; Abcam, ab88370) for 1 h at room temperature, washed three times with PBS, and then incubated with Goat Anti-Mouse IgG H&L (iFluor™ 488, HA1125) for 1 h. Nucleus were counterstained with DAPI (Beyotime, C1006) for 10 min. Fluorescence images were acquired using CLSM and the mean fluorescence intensity was quantified with ImageJ software.

For 26S proteasome inhibition experiment, HeLa cells were seeded in confocal dishes at a density of 1.5 × 10⁵ cells dish well and cultured overnight. The cells were pretreated with 10 μM MG132 or 3 μM MLN4924 for 12 h, followed by APF. After 24 h, the samples were then incubated with Alexa Fluor 647-conjugated anti-BRD4 antibody for 1 h at room temperature, followed by staining with phalloidin-Alexa Fluor 488 for 30 min in the dark, immunofluorescence staining was performed using a similar procedure as described above.

2.13. Western blot assay

HeLa cells were seeded in 6-well plates at a density of 2.0 × 10⁵ cells per well and cultured overnight. The cells were subsequently treated with APF, FRT (6 Gy) or the combination APF@FRT (6 Gy), while untreated cells served as controls. After 24 h, total proteins were extracted using RIPA lysis buffer (Beyotime, P0013B) supplemented with 1% protease inhibitor cocktail (Beyotime, P1006). Protein concentrations were determined using a BCA Protein Assay Kit (Beyotime, P0012). Equal amounts of protein were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS–PAGE) and transferred onto polyvinylidene fluoride (PVDF) membranes. Membranes were blocked with 5% BSA in TBST for 1 h at room temperature, followed by incubation with primary antibodies at 4 °C overnight. After washing, membranes were incubated with HRP-conjugated secondary antibodies (1:2000; Abcam, ab6721) for 1 h at room temperature. Protein bands were detected using enhanced chemiluminescence (ECL), with β-actin used as the internal loading control. Band intensities were quantified using ImageJ software.

In the degradation pathway assay, HeLa cells were initially preincubated with 10 μM MG132 or 3 μM MLN4924 for 12 h, followed by APF for an additional 24 h. Next, follow the above steps for subsequent procedures.

2.14. RNA sequencing analysis

HeLa cells were plated in 6-well plates and allowed to attach overnight. Cells were then exposed to APF, FLASH radiotherapy (FRT, 6 Gy) or the combined treatment (APF@FRT), with untreated cells serving as the control group. At 24 h post-treatment, total RNA was isolated for transcriptomic analysis. To maintain RNA integrity, all extraction procedures were conducted on ice. Briefly, cells were rinsed three times with ice-cold PBS, followed by the addition of 1 mL TRIzol reagent per well to achieve complete lysis after 3 min incubation at room temperature. The lysates were transferred into RNase-free cryovials and immediately snap-frozen in liquid nitrogen. Samples were shipped on dry ice to BGI (Shenzhen, China) for RNA purification, library preparation, high-throughput sequencing and downstream bioinformatic processing. Differential expression analysis was performed using the DESeq2 package, with significantly altered genes defined by an adjusted P value < 0.05 and an absolute log₂ fold change ≥1. Identified differentially expressed genes were subsequently subjected to Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses. Visualization of sequencing data was generated using the Dr. Tom bioinformatics platform (BGI-Shenzhen, China).

2.15. Animals tumor model

Female BALB/c-Nude mice (6–7 weeks old) were purchased from GemPharmatech Co., Ltd. (Nanjing, China). All animal experiments were reviewed and approved by the Animal Ethics Committee of Mianyang Central Hospital and all experimental procedures adhered to the relevant guidelines and regulations.

To establish the HeLa xenograft model, 5 × 10⁶ HeLa cells were suspended in a 100 μL mixture of Matrigel and PBS (1:1, v/v) and subcutaneously inoculated onto the right thigh of the mice. Tumor growth was monitored regularly. When the tumor volume reached approximately 50 mm³, the animals were randomly assigned into four experimental groups for subsequent experiments.

2.16. In vivo biodistribution of Cy5.5@APF nanoparticles

HeLa tumor-bearing mice were administered with Cy5.5@AP (without FA targeting), or Cy5.5@APF nanoparticles via the tail vein at a Cy5.5 dose of 0.5 mg/kg. At 24 h post-administration, the mice were anesthetized and imaged using a Tanon ABL X6 imaging system (Shanghai Tanon Life Science Co., Ltd.). Subsequently, the mice were sacrificed, and their tumors and major organs (including heart, liver, spleen, lung and kidney) were harvested for ex vivo imaging to evaluate the biodistribution of the administered nanoparticles.

2.17. In vivo antitumor experiment

HeLa tumor-bearing mice were randomly divided into four groups (n = 6): Control, APF (10 mg kg⁻¹), FRT (10 Gy) and APF@FRT (10 mg kg⁻¹ + 10 Gy). Mice in the Control group received 100 μL PBS via tail vein injection. The APF group was administered APF solution (10 mg kg⁻¹). The FRT group was subjected to FLASH radiotherapy at a dose of 10 Gy. The FLASH radiotherapy schematic diagram and dose evaluation was shown in Fig. S6A–C. In the APF@FRT group, mice were intravenously injected with APF (10 mg kg⁻¹) and followed by FRT irradiation (10 Gy). Tumor volumes were monitored for 10 days. Subsequently, the mice were euthanized and tumors were excised, weighed and photographed. Each tumor was fixed in 4% paraformaldehyde for immunohistochemical (IHC) and immunofluorescence (IF) evaluation. Mean fluorescence intensity and relative integral optical density (IOD) were quantified with ImageJ software.

2.18. Biosafety testing

At the experimental endpoint, whole blood samples were collected for routine and blood biochemical tests. Subsequently, the mice were euthanized. Major organs including the heart, liver, spleen, lung and kidney were collected for hematoxylin and eosin (H&E) staining to assess systemic biosafety.

2.19. Statistical analysis

All data were expressed as the mean ± standard deviations (SDs). One-way ANOVA was used to determine differences among groups. Statistical analyses were performed using GraphPad Prism software (version 8.0). *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001.

3. Results and discussion

3.1. Preparation and characterization of APF

APF were subsequently fabricated through nanoprecipitation-driven self-assembly of FA-PEG₂₀₀₀-ARV-771 in ultrapure water (Fig. 1A). The TEM image revealed that APF possessed a well-defined spherical morphology with a uniform distribution (Fig. 1B). According to DLS analysis, the hydrodynamic diameter of APF was approximately 146.7 nm, exhibiting a monodisperse distribution with a low polydispersity index (PDI = 0.074) (Fig. 1C). Furthermore, DLS analysis revealed that the size of APF nanoparticles changed little over time in different media, including SBF, PBS and DMEM supplemented with 10% FBS (Fig. 1D), indicating good stability under physiological and serum-containing conditions. In addition, the redox-responsive release behavior of ARV-771 from APF nanoparticles was evaluated in PBS containing either 0 or 10 mM GSH. In the absence of GSH, ARV-771 exhibited minimal release, reaching a plateau at 48 h with a cumulative release of only 10.6%. In contrast, under reductive conditions (10 mM GSH), the release of ARV-771 increased markedly over time, achieving a cumulative release of 72.6% at 48 h, which was substantially higher than that observed in the GSH-free group under otherwise identical conditions (Fig. 1E). These findings demonstrated that APF nanoparticles remained relatively stable under normal physiological conditions while enabling rapid drug liberation in a reductive tumor-mimicking microenvironment. Such redox-responsive behavior ensures efficient tumor-site delivery and intracellular release of ARV-771, thereby facilitating its therapeutic activity.

Fig. 1.

Preparation and characterization of APF. (A) Schematic illustration of the self-assembly process for the preparation of APF. (B) TEM image of APF (scale bar: 200 nm). (C) DLS size distribution of APF. (D) In vitro stability of APF in different media. (E) In vitro ARV-771 release profile from APF. (F) Cytotoxicity in HeLa cells evaluated by CCK-8 assay. (G) CLSM images of cellular uptake of DiD-labeled APF (scale bars: 20 μm and 10 μm). (H) Quantitative cellular uptake analyzed by flow cytometry. (I) CLSM images of BRD4 immunofluorescence staining in HeLa cells treated with MG132 or MLN4924 (scale bar: 40 μm). (J) Western blot analysis to study the effects of MG132 and MLN4924 on BRD4 degradation in HeLa cells.

As shown in Fig. 1F, APF significantly reduced the viability of HeLa cells in vitro, while only marginally affecting the proliferation of HUVECs at a concentration of 3 μg/mL. The selective cytotoxicity of APF nanoparticles against HeLa cells depended on the endocytosis mediated by folate receptors (FR) and the drug release induced by GSH within tumor cells. Normal HUVEC cells without FR expression did not undergo active uptake or trigger release, resulting in lower toxicity. Therefore, this concentration was selected for subsequent experiments. HeLa cells have been found to overexpress FR, thereby improving the tumor cell-specific uptake capacity [44]. Cellular uptake studies demonstrated that the Did@APF nanoparticles were efficiently internalized in a time-dependent manner (Fig. 1G). Quantitative analysis by flow cytometry revealed significantly higher intracellular accumulation at 5 h compared to 2 h (Fig. 1H). To further verify the BRD4 protein degradation mechanism, the proteasome-inhibition experiment was performed using proteasome inhibitor MG132 and neddylation inhibitor MLN4924. The immunofluorescence (Fig. 1I) and protein (Fig. 1J and Fig. S7) experiments showed that APF treatment markedly reduced BRD4 protein expression, whereas co-treatment with MG132 or MLN4924 substantially rescued BRD4 protein levels. This finding indicated that the APF-induced decrease in BRD4 was predominantly dependent on proteasomal degradation manner.

3.2. Radiosensitization of APF in vitro

The live-dead assay and 3D reconstruction results (Fig. 2A and Fig. S8) indicated that APF and FRT groups both induced a certain degree of tumor cell death (red fluorescence) compared with the Control group. The synergistic interaction between APF and FRT (APF@FRT) exhibited a marked enhancement in fluorescence, indicating superior antitumor performance in vitro. Colony assay demonstrated that the APF@FRT group significantly inhibited tumor growth, providing the most powerful suppression on colony formation. In comparison, the Control group displayed a large number of densely packed colonies, which grew in prominent clusters (Fig. 2B). Scratch assay and semi-quantitative results exhibited that APF or FRT treatment observably suppressed the migratory ability of HeLa cells at 24 h relative to untreated Control group. Importantly, the migration rate in the APF@FRT group (approximately 43.74%) was reduced to nearly half of that in the Control group (approximately 81.57%), suggesting a synergistic inhibitory effect of the combined treatment on cell migration (Fig. 2C and D). Furthermore, the apoptosis rate was detected by Annexin V-FITC/7-AAD double staining assay after various treatments in HeLa cells (Fig. 2E and F). Different extent of cell death could be triggered by different treatments, among which APF@FRT induced the strongest apoptosis in tumor cells. The APF@FRT group induced DNA damage in tumor cells 3.2 times more efficiently than the Control group in vitro, indicating the effective enhancement of FLASH-RT by the APF nanoparticles.

Fig. 2.

In vitro radiosensitization effect of APF. (A) Live/dead staining of HeLa cells. Green fluorescence (FDA) represented viable cells, whereas red fluorescence (PI) indicated dead cells (scale bar: 50 μm). (B) Colony formation assay. (C) Wound healing assay and (D) corresponding semi-quantitative analysis using ImageJ software. (E) Flow cytometry analysis of apoptosis using Annexin V-FITC/7-AAD staining after APF@FRT treatment and (F) corresponding quantification of apoptotic cells. *P < 0.05, **P < 0.01, ***P < 0.001.

3.3. APF enhanced radiotherapy-induced ROS accumulation and DNA damage

To evaluate the radiosensitizing potential of APF, intracellular ROS levels in HeLa cells were measured using the DCFH-DA fluorescent probe. As shown in Fig. 3A, exposure to FRT alone resulted in a pronounced increase in green fluorescence, indicating radiation-induced ROS generation. Notably, the APF@FRT group exhibited significantly stronger fluorescence intensity than the FRT group. Semi-quantitative analysis demonstrated that the MFI in the APF@FRT group was 1.36-fold higher than that in the FRT group. These findings indicated that APF markedly amplified radiation-induced oxidative stress by promoting excessive intracellular ROS accumulation, thereby enhancing oxidative damage in tumor cells. This effect may be attributed to the BRD4-targeting PROTAC delivered by APF. The degradation of BRD4 has been reported to induce ferroptosis in vitro, thereby elevating ROS levels [45]. Furthermore, as a marker of DNA double-strand breakage, γ-H2AX was used to detect DNA damage. As shown in Fig. 3B, compared with FRT group, the APF@FRT group exhibited a markedly increased fluorescence intensity. Semi-quantitative analysis revealed that the MFI in the combination group was elevated by 33.9% relative to the FRT group. These results confirmed that the addition of APF significantly aggravated radiation-induced DNA damage, consistent with the observed ROS amplification.

Fig. 3.

APF enhanced radiotherapy-induced ROS accumulation and DNA damage. (A) Intracellular ROS levels in HeLa cells and the corresponding semi-quantitative analysis (scale bar: 50 μm). (B) γ-H2AX immunofluorescence staining showing DNA double-strand breaks with 3D reconstruction and semi-quantitative analysis (scale bar: 50 μm). *P < 0.05, **P < 0.01, ***P < 0.001.

3.4. Augmentation the anti-tumor activity of FLASH-RT in vivo through the APF sensitization effect

Firstly, to verify the tumor-targeting capability of APF, the in vivo biodistribution experiment was performed using fluorescently labeled APF. Specifically, Cy5.5-labeled APF (denoted as Cy5.5@APF) was administered to HeLa tumor-bearing nude mice. The fluorescence imaging and quantitative analysis results were shown in Fig. 4A and B. Compared with the fluorescent nanoparticles without FA targeting (Cy5.5@AP), there was a significant fluorescence expression at the tumor site in the Cy5.5@APF group, which was 2.5-fold that of the Cy5.5@AP group, indicating that Cy5.5@APF had good in vivo tumor targeting properties.

Fig. 4.

In vivo antitumor performance of APF@FRT. (A) Ex vivo fluorescence images after the treatment of Cy5.5@APF and Cy5.5@AP for 24 h, respectively. (B) Quantitative analysis of the fluorescence distribution in tumor tissue and major organs by ex vivo imaging. (C) Schematic illustration of the treatment protocol. (D) Tumor growth curves. (E) Representative images of excised tumors and (F) corresponding tumor weights. (G) Immunofluorescence staining of γ-H2AX, TUNEL and Ki-67 in tumor tissues (scale bar: 50 μm). (H) Blood biochemical indicators including ALT, AST, Urea and Crea. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Encouraged by the pronounced radiosensitizing effect of APF observed in vitro, we further evaluated its ability to enhance the antitumor efficacy of FLASH radiotherapy in vivo. A HeLa tumor model was established in BALB/c-nude mice, and corresponding treatments and irradiation (10 Gy, 94 Gy s⁻¹) were initiated when the tumor volume reached approximately 50 mm³ one week after inoculation (Fig. 4C). Compared with the Control group, treatment with APF, FRT and APF@FRT all exhibited significant tumor growth inhibition, with the APF@FRT group exhibited the most pronounced tumor inhibition (Fig. 4D–F). Furthermore, immunofluorescence staining of tumor sections and corresponding semi-quantitative results (Fig. 4G and Fig. S9) revealed that the APF@FRT group displayed the strongest γ-H2AX fluorescence signal, indicating the most severe DNA damage. Meanwhile, enhanced TUNEL-positive green fluorescence suggested markedly increased tumor cell apoptosis. In addition, the decreased expression of Ki-67 fluorescence demonstrated that APF@FRT significantly suppressed tumor cell proliferation and reduced the proportion of actively cycling cells, further confirming its enhanced antitumor efficacy. Regarding biosafety, blood biochemical analysis (Fig. 4H) showed that liver function markers (ALT, AST) and renal function markers (Urea, Crea) remained within physiological ranges across all treated groups. And the body weight of the mice changed steadily (Fig. S10). Histopathological evaluation (H&E staining) confirmed no discernible structural abnormalities in major organs, including the heart, liver, spleen, lung and kidney, compared to the control group (Fig. S11A). Additionally, routine blood parameters (WBC, RBC, HGB, and PLT) showed no significant fluctuations (Fig. S11B). These results collectively indicated that the combined APF and FLASH radiotherapy treatment caused no evident systemic toxicity and did not induce noticeable damage to normal organs, demonstrating favorable biosafety.

3.5. Mechanism of APF nanoparticles enhancing anti-tumor effect of FLASH-RT

To investigate the antitumor mechanism underlying the radiosensitizing effect of APF nanoparticles on FLASH radiotherapy, transcriptomic analysis was performed to identify significantly altered gene pathways and biological processes in HeLa cells (Fig. 5). The Venn diagram revealed that 10979 co-expressed genes among the Control, APF, FRT and APF@FRT groups, alongside a distinct subset of treatment-specific differentially expressed genes (DEGs), indicating specific transcriptional reprogramming induced by different treatments (Fig. 5A). In the volcano plot of the FRT vs Control group (Fig. 5B), a significant upregulation of CDKN1A (cyclin-dependent kinase inhibitor 1A, also known as p21) was observed following irradiation, suggesting that radiation-induced DNA damage activated cell cycle arrest to suspend tumor cell proliferation. This response concurrently triggered the compensatory upregulation of the HR factor RAD51AP1, providing a temporal window for DNA repair and contributing to the adaptive radioresistance of cervical cancer cells. In the APF vs Control group, Gene Ontology (GO) enrichment analysis and the corresponding heatmap (Fig. 5C) revealed the significant downregulation of multiple biological processes, including mitotic cell cycle phase transition, G2/M transition of the mitotic cell cycle and DNA-binding transcription factor binding. These results suggested that APF might enhance radiosensitivity by reinforcing cell-cycle arrest and suppressing transcriptional activity, thereby impairing the ability of cancer cells to repair radiation-induced damage and ultimately inhibiting mitotic progression and proliferative capacity.

Fig. 5.

Transcriptomic analysis revealed the molecular mechanisms of APF-enhanced radiosensitization. (A) Venn diagram of expressed genes among APF@FRT, APF, FRT and Control groups. (B) Volcano plot of differentially expressed genes in the FRT vs Control group. (C) GO enrichment analysis of APF vs Control group and corresponding heatmap of representative biological processes. (D) Gene expression trend analysis of APF@FRT vs Control. (E) Volcano plot of differentially expressed genes in APF@FRT vs Control. (F) KEGG pathway enrichment analysis and (G) corresponding heatmap. (H) Gene interaction network.

Furthermore, we investigated transcriptional changes following the combined treatment of APF and FLASH radiotherapy. The gene expression trend plot (Fig. 5D) showed a widespread downregulation across various gene subclusters in the APF@FRT group, indicating a broad inhibitory effect on cellular activity. This suppression may be associated with disruption of key transcriptional networks or signaling pathways, leading to reduced cellular activity and metabolic capacity. Notably, the volcano plot of the APF@FRT vs Control group (Fig. 5E) showed a pronounced downregulation of MYC, a critical downstream target of BRD4, suggesting that APF@FRT effectively disrupted the BRD4–MYC regulatory axis which drived tumor proliferation and metabolic signaling. KEGG pathway enrichment and heatmaps (Fig. 5F and G) further demonstrated that several oncogenic signaling pathways were significantly suppressed following APF@FRT treatment, including the Wnt signaling pathway, signaling pathways regulating pluripotency of stem cells, PI3K–AKT signaling pathway, and MAPK signaling pathway. These results indicated that APF@FRT simultaneously restrained several key pathways involved in tumor proliferation, stemness maintenance and cell survival, thereby exerting potent antitumor effects. Finally, the gene interaction network analysis (Fig. 5H) revealed that MYC occupied a central hub connecting MAPK, PI3K–AKT and growth factor signaling modules, highlighting its critical role in coordinating cellular proliferation, stemness and signal transduction, further supporting the mechanistic role of BRD4–c-Myc axis inhibition in the enhanced therapeutic efficacy of APF-assisted FLASH radiotherapy.

3.6. APF sensitized HeLa tumor cells to RT via BRD4-c-Myc-RAD51AP1 degradation

Based on the above omics results, we first conducted verification at the cellular level. The CLSM images (Fig. 6A) showed that the red fluorescence distribution of BRD4 in the Control and FRT groups was obvious, while the red fluorescence in the APF group and APF@FRT group was almost undetectable, indicating that APF effectively degraded the BRD4 protein. Western blot analysis (Fig. 6B) and corresponding semi-quantitative results (Fig. 6C) revealed that the FLASH-RT treatment did not affect protein degradation. Following APF nanoparticle treatment, both BRD4 and c-Myc protein expression levels decreased, indicating APF's ability to efficiently promote BRD4 degradation and suppress the downstream c-Myc signaling axis. It is worth noting that the in vitro immunofluorescence results of RAD51AP1 (Fig. 6D) showed that after APF treatment alone and in combination with FLASH radiotherapy, the inhibitory effect on RAD51AP1 expression obviously decreased, indicating that APF treatment significantly downregulated the expression of RAD51AP1. This was consistent with the trend of the results observed for BRD4 and c-Myc in this study. These results verified the findings of the transcriptomics analysis at the protein level, further demonstrating that APF sensitized the anti-tumor effect of FLASH radiotherapy by regulating the BRD4-c-Myc-RAD51AP1 pathway.

Fig. 6.

In vitro validation of the BRD4-c-Myc-RAD51AP1 regulatory axis. (A) CLSM images of BRD4 and F-actin immunofluorescence staining in HeLa cells following different treatments (scale bar: 20 μm). (B) Western blot analysis of BRD4 and c-Myc protein expression in the Control, APF, FRT and APF@FRT groups and (C) corresponding semi-quantitative analysis. (D) CLSM images of RAD51AP1 immunofluorescence staining (red) in HeLa cells following different treatments (scale bar: 20 μm). ****P < 0.0001.

3.7. APF downregulated BRD4-c-Myc-RAD51AP1 and modulated p21 response to radiation in vivo

Subsequently, we employed immunofluorescence techniques to further validate the in vivo expression of BRD4, c-Myc, RAD51AP1 and p21 in mouse tumor tissues. As observed in Fig. 7A–C, compared with the Control and FRT groups where BRD4, c-Myc and RAD51AP1 exhibited diffuse and high expression, the fluorescence signals in the APF and APF@FRT groups were significantly weakened. This indicated that APF could effectively inhibit the expression activity of the BRD4-c-Myc-RAD51AP1 axis in tumor cells. Furthermore, p21 was markedly upregulated in the tumor tissues after FLASH irradiation (Fig. 7D), providing in vivo confirmation that FLASH-RT activated p21-mediated cell cycle arrest by inducing DNA damage, which was highly consistent with the transcriptome analysis results. Notably, the addition of APF partially alleviated the cell cycle arrest in APF@FRT after radiation, suggesting that APF may enhance the antitumor efficacy of FLASH-RT by downregulating the BRD4-c-Myc-RAD51AP1 pathway and interfering with DNA repair progress, thereby reducing the tolerance of tumor cells to FLASH-RT.

Fig. 7.

Histological analysis of tumor tissues. (A1) Immunofluorescence staining of c-Myc in tumor sections and (A2) corresponding semi-quantitative analysis (scale bar: 50 μm). (B1) Immunofluorescence staining of BRD4 in tumor tissues and (B2) corresponding semi-quantitative analysis (scale bar: 50 μm). (C1) Immunofluorescence staining of RAD51AP1 in tumor tissues and (C2) corresponding semi-quantitative analysis (scale bar: 50 μm). (D1) Immunohistochemical staining of p21 and (D2) corresponding semi-quantitative analysis (scale bar: 50 μm). *P < 0.05, **P < 0.01, ***P < 0.001.

4. Conclusion

In conclusion, a redox-responsive PROTAC nanoplatform (APF) was successfully developed to enhance the antitumor efficacy of FLASH radiotherapy. The self-assembled nanoparticles exhibited favorable stability and glutathione-triggered drug release, enabling efficient delivery of the BRD4 degrader ARV-771 to tumor cells. Mechanistically, APF significantly amplified radiation-induced ROS production and exacerbated DNA double-strand breaks, while suppressing the BRD4-c-Myc-RAD51AP1 pathway and interfering with DNA repair progress which resulting in a significant sensitizing effect of FLASH radiotherapy. Consequently, the APF@FLASH-RT treatment achieved superior tumor inhibition in vivo with negligible systemic toxicity. Our findings established a nanomedicine-based strategy for potentiating FLASH radiotherapy through selective protein degradation, with significant implications for improving therapeutic index and overcoming radioresistance.

Funding

This work was sponsored by Natural Science Foundation of Sichuan Province (grant number 2025ZNSFSC1727), Mianyang Science and Technology Bureau (Mianyang Science and Technology Program, grant number 2025ZYDF019), NHC Key Laboratory of Nuclear Technology Medical Transformation (Mianyang Central Hospital, grant number 2024HYX002), Clinical Key Specialty of Gastroenterology Department of Mianyang Central Hospital (grant number XHZDZK005) and Talent Introduction and Research Project of Mianyang Central Hospital (grant number 2024RCYJ-003 and 2025RCYJ-001).

CRediT authorship contribution statement

Ruiling Xu: Funding acquisition, Project administration, Writing – original draft. Xiaowen Han: Validation. Yunfei Sun: Investigation. Chaofan Ma: Data curation. Shuqing Liao: Resources. Ruo Tang: Software. Yinquan Pu: Formal analysis. Xueting Zheng: Validation. Jie Zhang: Methodology. Wen Yang: Writing – review & editing. Xiaoan Li: Funding acquisition, Supervision. Zhi Yang: Conceptualization, Funding acquisition, Writing – review & editing.

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.

Appendix B. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.