Enhanced Tumor Retention and Therapeutic Potency in Radionuclide Therapy Using GSH-Induced Self-Assembling Peptides

Abstract

Radiolabeled small-molecule drugs hold significant potential for tumor radionuclide therapy (TRT). However, the clinical application of TRT is constrained by the rapid clearance of small molecules from tumor sites and inadequate tumoral radiation sensitivity. To address these limitations, we report an in situ strategy to consume glutathione for assembly that enhances TRT by increasing reactive oxygen species and extending therapeutic time. The probe, [⁶⁴Cu]Cu-DOTA-PEP1-c(RGDyK) ([⁶⁴Cu]Cu-DP1R), was designed with a glutathione-responsive self-assembling polypeptide backbone (PEP1), a DOTA chelator, and an integrin α_vβ₃-targeting moiety. Radiolabeling under optimized conditions yielded a highly stable complex, with radiochemical purity exceeding 95%. In vivo PET imaging in U87-MG tumor-bearing mice revealed enhanced tumor-specific distribution and extended retention of [⁶⁴Cu]Cu-DP1R compared to [⁶⁴Cu]Cu-DOTA-c(RGDyK) ([⁶⁴Cu]Cu-DR) at 72 h post-injection (6.20 ± 1.61 %ID/g vs. 1.97 ± 0.75 %ID/g, P < 0.01). Enhanced tumor imaging quality and biocompatibility were also observed. Therapeutic evaluation revealed extended survival rates, with the [⁶⁴Cu]Cu-DP1R group achieving a 3.33-times increase over the [⁶⁴Cu]Cu-DR group, while maintaining high biocompatibility without adverse effects on body weight or physiological health. These results highlight the potential of glutathione-responsive self-assembly to extend tumor retention and improve therapeutic outcomes in TRT. By combining the advantages of small peptide-targeting specificity with the prolonged retention of nanostructures, this approach represents a promising strategy to enhance the efficacy and clinical translation of targeted radionuclide therapies.

Graphical Abstract

Introduction

Recent advancements in cancer treatment have led to the development of precise tumor therapies, such as targeted radionuclide therapy (TRT), which can selectively identify tumors while reducing toxicity to non-target organs^[1]. TRT, involving the use of radiopharmaceuticals labeled with therapeutic radionuclides such as ⁶⁴Cu, ¹⁷⁷Lu, ²²³Ra, and ²²⁵Ac, has emerged as a promising approach for treating malignant tumors. These radionuclides emit β or α rays, producing ionizing radiation effects that directly impact tumor cells and indirectly affect neighboring tumor cells through bystander and crossfire effects^{[2, 3]}. However, the limitations of inadequate tumor uptake and short retention time of therapeutic radiopharmaceuticals may result in insufficient radiation doses delivered to the tumor^[4–9]. Additionally, the tumor microenvironment, such as glutathione (GSH), continuously neutralizes reactive oxygen species (ROS) to reduce tumor sensitivity to radiation, which can further lead to unsatisfactory treatment outcomes^[10].

To address these challenges, various nanomaterials have been developed as carriers for radiopharmaceuticals. These include organic polymers like liposomes, serum albumin, and polydopamine, as well as inorganic nanocarriers such as gold nanoparticles, magnetic nanoparticles, and silica nanoparticles^{[7, 11, 12]}. These nanomaterials leverage the enhanced permeability and retention (EPR) effect to transport radionuclides effectively to the tumor site^[13]. Surface modification with specific antibodies or ligands can further enable precise targeting of tumor cells^{[14, 15]}. Despite their advantages, nanomaterials face challenges such as rapid clearance by the mononuclear phagocyte system and potential biological reactions, including immune activation, cytotoxicity, and long-term accumulation in tissues^{[16, 17]}. To overcome these issues, in situ self-assembly molecular structures responsive to the tumor microenvironment have been explored as an alternative strategy^{[18, 19]}. These structures remain as small peptides under normal physiological conditions but undergo structural changes in response to the tumor’s acidic, redox, or enzyme-rich environment, initiating self-assembly into nanostructures^[20–22]. This approach can help stabilize drugs at the tumor site, prolong the therapeutic period, neutralize specific tumor microenvironment, and reduce toxicity to non-target organs.

Previous studies have validated the feasibility of in vivo self-assembly platforms for drug delivery and tumor treatment^{[19, 23, 24]}. Building on this, we have further explored the use of hydrophilic peptide sequences capable of intracellular self-assembly, covalently linked with therapeutic radionuclides, to construct peptide-drug conjugates that specifically accumulate in tumors, achieving prolonged retention and enhanced TRT efficacy. For targeted delivery, we selected the c(RGDyK) peptide, a cyclic peptide known to target integrin α_vβ₃, which is overexpressed in many tumor cells and associated vasculature. In this study, we also introduce a novel in situ self-assembly strategy employing GSH-induced self-assembling Phe-Phe peptide, PEP1, which selectively assembles within the tumor microenvironment. Elevated levels of GSH in the tumor milieu serve as a trigger for PEP1 to form nanostructures, significantly prolonging the probe’s retention time at the tumor site. Mechanistically, PEP1 containing an S-S bond, consumes GSH and further improves ROS concentration and ROS-induced cell damage^[25].

The study aims to evaluate the efficacy of [⁶⁴Cu]Cu-DOTA-PEP1-c(RGDyK) ([⁶⁴Cu]Cu-DP1R) in enhancing tumor retention and therapeutic outcomes in TRT. By employing PET imaging, we investigate the probe’s biodistribution and retention characteristics in U87-MG tumor-bearing mouse models, followed by therapeutic assessments to observe tumor growth suppression and survival benefits. This approach not only addresses the key limitation of radiolabeled small-molecule drugs but also represents a promising advance in tumor-specific TRT by combining enhanced targeting with prolonged retention.

Materials and Methods

Synthesis of Self-Assembled Peptide Molecules

The synthesis of the phenylalanine peptide (PEP1) was achieved using solid-phase peptide synthesis to yield DOTA-PEP1-c(RGDyK) (DP1R) and DOTA-c(RGDyK) (DR). As illustrated in Figure 2-2, PEP1 was first dissolved in 500 μL anhydrous DMSO, to which EDC (2 equivalents) and NHS (1.2 equivalents) were added and stirred at room temperature for 4 h. DIPEA (20 equivalents) and c(RGDyK) (2 equivalents) were then introduced, followed by stirring for 10 h. Reaction monitoring was performed via HPLC, and purification employed a semi-preparative C4 HPLC column. Afterward, the precipitate was rinsed with acetonitrile and dried using a rotary evaporator. Boc deprotection was conducted with a 1:3 solution of trifluoroacetic acid (TFA) and dichloromethane (DCM), yielding the final DP1R product upon purification.

Figure 2.

Structural Characteristics of DP1R. (A) Determination of Critical Micelle Concentration of DP1R; (B) TEM Imaging of DP1R without GSH; (C) TEM Imaging of DP1R after co-incubation with GSH; (D) Particle Size Distribution of DP1R after co-incubation with GSH; (E) Zeta Potential of DP1R before and after Co-incubation with GSH; (F) Radiochemical Purity of [⁶⁴Cu]Cu-DP1R.

Critical Aggregation Concentration Determination

DP1R solutions were prepared at concentrations ranging from 0.1 to 100 μM, each containing 1 mM GSH. The critical micelle concentration (CMC) was determined using pyrene as a fluorescent probe, and CMC values were calculated by the tangent method.

Morphological Analysis of Self-Assembled Peptide Molecules

To study nanoparticle formation, DP1R was dissolved in water at concentrations above the CMC, with 1 mM GSH added to mimic intracellular conditions and promote disulfide bond cleavage. Samples were incubated for 10 minutes and placed onto copper grids before air-drying overnight. Transmission electron microscopy (TEM) was then used to analyze particle morphology and size. The hydrodynamic diameter and zeta potential of DP1R nanoparticles were measured by dynamic light scattering (DLS) in PBS with GSH added to the samples. DLS parameters were set to record hydrodynamic diameter and zeta potential.

Radiolabeling

For radiolabeling, 5–10 nmol of DP1R peptide was reacted with 37 MBq (1 mCi) of [⁶⁴Cu]CuCl₂ in 300 μL of 0.1 M sodium acetate buffer (pH 5.0). The mixture was incubated at 100 °C for 30 min. Radiochemical purity was verified by HPLC with a C18 column at a 1 mL/min flow rate, with samples demonstrating > 90 % radiochemical purity retained for further analysis. When radiochemical purity was < 90 %, Sep-Pak C18 Light cartridges were used. Cartridges were preconditioned with ethanol and water, followed by loading the labeled product, washing with water, and eluting with 75 % ethanol. Ethanol was evaporated under nitrogen, and the product was diluted in saline (ethanol < 10 %) for in vivo experiments. Radiolabeled compounds, [⁶⁴Cu]Cu-DR and [⁶⁴Cu]Cu-DP1R, were stored at room temperature for 4, 24, and 72 h to assess stability. Samples were also incubated with serum and analyzed using instant thin-layer chromatography (iTLC) (mobile phase: 50 mM EDTA, pH 5.5).

Cytotoxicity Assay

HEK293 cells in the logarithmic growth phase were plated at 1 × 10⁴ cells per well in 96-well plates and incubated for 24 h. Following cell adhesion, cells were treated with various concentrations of DR and DP1R (3.96–250 μM) for 24 h, and cell viability was determined using a CCK-8 assay at 450 nm. Further, U87MG cells were incubated with [⁶⁴Cu]CuCl₂, [⁶⁴Cu]Cu-DR, and [⁶⁴Cu]Cu-DP1R (0.1–0.2 MBq) for 4 h. Cell viability was assessed by CCK-8 assay after a 24-hour incubation in fresh medium.

Intracellular GSH/ROS Content Measurement

U87MG cells were seeded at a density of 2 × 10⁵ cells/well in six-well plates and incubated at 37°C for 24 h. After incubation, 1 mL of serum-free medium containing varying concentrations of PBS, DR, DP1R, [⁶⁴Cu]Cu-DR and [⁶⁴Cu]Cu-DP1R (0, 25, 50, and 100 μM) was added to each well, followed by co-incubation for an additional 24 h. Cells were then washed twice with PBS, digested with 400 μL trypsin solution, and neutralized with 800 μL of culture medium. The cell suspension was centrifuged at 3000 rpm for 3 minutes, washed with 1 mL PBS to obtain a cell pellet. The intracellular GSH (Sigma-Aldrich, CS0260) and ROS levels (AAT Bioquest, 22900) were subsequently measured using the assay kit.

Immunocytochemistry

U87MG-luc and MCF-7 cells were fixed, permeabilized, and blocked as per standard protocols. For immunofluorescence staining, cells were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100. After blocking with 5% BSA, samples were incubated overnight at 4°C with the following primary antibody (1:200 dilution, Abcam ab7166). Following PBS washes, samples were incubated for 1 h at room temperature with secondary antibody (1:500 dilution, Abcam ab205719). Cells were then imaged using the microscope.

In Vitro Uptake

U87MG cells were seeded in 24-well plates at a density of 5 × 10⁴ cells/well in 1 mL complete DMEM medium (10% FBS, 1% penicillin-streptomycin) and cultured for 48 h to reach 80% confluency. Cells were then incubated with [⁶⁴Cu]CuCl₂, [⁶⁴Cu]Cu-DR, and [⁶⁴Cu]Cu-DP1R (0.05–1 MBq) in fresh medium for 4 h at 37°C. After incubation, cells were washed twice with ice-cold PBS to remove unbound radioactivity. Membrane-bound fractions were collected by treating cells with 0.5 mL acidic glycine buffer (50 mM glycine, 100 mM NaCl, pH 2.8) for 5 min on ice. Internalized fractions were subsequently obtained by lysing cells with 0.5 mL 1 M NaOH. Radioactivity of both fractions was quantified using a gamma counter (PerkinElmer, 2480 Wizard²).

Tumor Xenograft Establishment

U87MG or MCF-7 cells (5 × 10⁶ cells/mouse) were suspended in 100 μL of PBS:Matrigel mixture (1:1 v/v) and injected subcutaneously into the right forelimb of 4-week-old female BALB/c nude mice. Tumor growth was monitored until reaching 100–150 mm³ (typically 10–14 days) before initiating experiments.

PET Imaging

U87MG and MCF-7 tumor-bearing mice (n = 4/group) were administered 7.4MBq [⁶⁴Cu]Cu-DR or [⁶⁴Cu]Cu-DP1R via tail vein. PET scans were conducted at 1, 4, 24, and 72 h post-injection. Obtain the uptake information of the imaging agent of interest to calculate the retention index.

$$\mathrm{R}\mathrm{e}\mathrm{t}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{I}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{x}\left(\mathrm{R}\mathrm{I}\right)=\frac{\mathrm{\%}\mathrm{I}\mathrm{D}/{\mathrm{g}}_{\mathit{max}\left(\mathit{d}\mathit{e}\mathit{l}\mathit{a}\mathit{y}\mathit{e}\mathit{d}\mathit{R}\mathit{O}\mathit{I}\right)}-\%\mathrm{I}\mathrm{D}/{\mathrm{g}}_{\max \left(earlyROI\right)}}{\%\mathrm{I}\mathrm{D}/{\mathrm{g}}_{\max \left(earlyROI\right)}}\times 100\%$$

Biodistribution Study

At 72 h post-injection, mice (n = 3/group) were euthanized by CO_₂ asphyxiation followed by cervical dislocation. Blood was collected via cardiac puncture into EDTA-coated tubes. Fourteen tissues were harvested: blood, brain, heart, lungs, liver, spleen, kidneys, stomach, small intestine, large intestine, pancreas, muscle, bone, tumor. Tissues were weighed and radioactivity measured using a γ-counter (PerkinElmer 2470 Wizard²).

In Vivo Radiotherapy

Tumor-bearing mice (n = 5/group) received PBS, [⁶⁴Cu]CuCl₂, [⁶⁴Cu]Cu-DR, and [⁶⁴Cu]Cu-DP1R at 66.6 MBq. Tumor growth and toxicity were monitored. Tumor growth was measured every two days, and tumor volume was calculated as Volume=L×W² /2. Monitoring criteria included maximum tumor volume and other health indicators. Tumor sections were subjected to Ki-67 IHC and Fluorescence staining for TUNEL, γ-H2AX, and ROS, with imaging performed on microscope. Liver and kidney functions were analyzed, and histological assessments of major organs were performed.

Statistical Analysis

Data analysis was performed using GraphPad Prism 8.0 software. All results are expressed as mean ± standard deviation (SD). Statistical differences between groups were evaluated using Student’s t-test or one-way analysis of variance (ANOVA), with a P-value of < 0.05 considered statistically significant.

Results

Preparation and Characterization of the Self-Assembled Probe

In this experiment, we successfully constructed tumor microenvironment-responsive in vivo self-assembling peptide-small molecule conjugates DP1R and DR using efficient peptide solid-phase synthesis and amide condensation techniques (Figure. S1–2). The intermediates and final products were purified and validated by HPLC, achieving a purity of >95%. The mass spectrometry results were as follows: PEP1-c(RGDyK) (Figure S3a) [M + H]⁺: [C₈₀H₁₁₁N₁₉O₂₁S₂ + H]⁺: m/z = 1738.77161, found 1738.77108; DP1R (Figure S3b) [M + 2H]²⁺: [C₉₉H₁₃₆N₂₄O₂₇S₃ + 2H]⁺²: m/z = 1095.46572, found 1095.46491; DR(Figure S4) [M + H]⁺: [C₅₁H₇₄N₁₄O₁₆S + H]⁺: m/z = 1171.5201, found 1171.5193.

The CMC of DP1R was determined to be 13.962 μM (Figure 2A). In the absence of GSH incubation, no nanostructure formation was observed under TEM (Figure 2B). When incubated with GSH, micelle-like structures with an average diameter of approximately 27.55 ± 6.43 nm were observed under TEM (Figure 2C). The hydrodynamic diameter of DP1R after incubation with GSH was measured to be 284.28 ± 27.99 nm (Figure 2D), and the Zeta potential changed from 3.65 ± 0.49 mV before incubation with GSH to −36.43 ± 1.89 mV after incubation (Figure 2E).

We successfully synthesized DR and DP1R, and labeled them with [⁶⁴Cu]Cu to obtain the respective radiolabeled products, achieving a labeling efficiency and radiochemical purity of >95% (Figure 2F). Upon completion of labeling, the specific activities of [⁶⁴Cu]Cu-DR and [⁶⁴Cu]Cu-DP1R were approximately 3.7 MBq/nmol. After incubating [⁶⁴Cu]Cu-DP1R with human serum and PBS for 4 h, 24 h, and 72 h, the radiopharmaceutical maintained high radiochemical purity, indicating excellent in vitro stability (Figure S5). Similarly, the radiochemical purity of [⁶⁴Cu]Cu-DR remained stable after incubation with PBS and serum, demonstrating good in vitro stability of the radiopharmaceuticals (Figure S6).

Cellular Characterization of the Self-Assembled Radiolabeled Probe

Immunohistochemistry was performed to identify the cell lines with high integrin α_vβ₃ expression, U87MG (human glioma cell line), and low integrin α_vβ₃ expression, MCF-7 (human breast cancer cell line) (Figure 3A). In vitro experiments measured the uptake of free [⁶⁴Cu]CuCl₂, [⁶⁴Cu]Cu-DR, and [⁶⁴Cu]Cu-DP1R by U87MG tumor cells in both membrane and intracellular compartments over 4 h (Table S1). Negligible uptake was detected when cells were incubated with free [⁶⁴Cu]CuCl₂ (Figure 3B). The uptake percentage dose (%AD) of 0.05 MBq [⁶⁴Cu]Cu-DP1R (9.09 ± 0.48, intracellular: 8.27 ± 0.44, membrane: 0.95 ± 0.16) was 5.83 times higher than that of the [⁶⁴Cu]Cu-DR group (%AD: 1.56 ± 0.78, intracellular: 0.10 ± 0.02, membrane: 1.37 ± 0.54) (P < 0.05). Moreover, the uptake %AD of 0.05 MBq [⁶⁴Cu]Cu-DP1R was higher than that of the 0.1 MBq dose (9.09 ± 0.48 vs. 7.50 ± 0.32, P < 0.01), with a decreasing trend in cellular uptake as the dose increased. Analysis of the uptake mechanism revealed that 90.98% of the radioactivity of 0.05 MBq [⁶⁴Cu]Cu-DP1R was internalized by U87MG cells after co-incubation, with internalization being the predominant form of cellular uptake across different doses of [⁶⁴Cu]Cu-DP1R. Similarly, 87.82% of 0.05 MBq [⁶⁴Cu]Cu-DR was internalized by the cells.

Figure 3.

Characterization of Probes on Cells. (A) Immunohistochemical staining images of integrin α_vβ₃ in U87MG and MCF-7 cells. Scale bar: 50 μm; (B) Uptake of varying doses of [⁶⁴Cu]CuCl₂, [⁶⁴Cu]Cu-DR, and [⁶⁴Cu]Cu-DP1R in U87MG cells. % AD = Percentage of uptake relative to the added dose; (C) Changes in GSH/GSSG levels after co-incubation of different probe with U87-MG cells for 24 h; (D) Intracellular ROS level assay; (E) Cell viability of HEK293 cells after incubation with different concentrations of materials; (F) Cell viability after incubation of radiolabeled compounds with U87MG cells, compared to the control group. ns indicates no statistical significance, * P < 0.05; **, P < 0.01; ****, P < 0.0001.

Evaluation of intracellular GSH/GSSG levels in U87MG cells after co-incubation with different probes are shown in Figure 3C. Results indicate that treatment with 25 μM DP1R reduced intracellular GSH levels by 17%, while treatment with 100 μM DP1R led to a reduction exceeding 52%, both significantly greater than those observed in the PBS control group (P < 0.0001). These findings suggest that DP1R can be effectively triggered by GSH to initiate in situ self-assembly, resulting in substantial GSH depletion within the cellular environment. As shown in Figure 3C and D, [⁶⁴Cu]Cu-DP1R could effectively reduce intracellular GSH (P < 0.0001) and enhance the ROS level (P < 0.05). Also, the oxidative stress increases with the concentration of [⁶⁴Cu]Cu-DP1R.

Cytotoxicity was assessed using the CCK-8 assay after 24-hour co-incubation with c(RGDyK), DR, and DP1R, showing no significant impact on the viability of human embryonic kidney cells HEK293 at a concentration of 250 μM. c(RGDyK), DR, and DP1R demonstrated good biocompatibility at concentrations below 250 μM (Figure 3E). The impact of different concentrations of free [⁶⁴Cu]CuCl₂, [⁶⁴Cu]Cu-DR, and [⁶⁴Cu]Cu-DP1R radiopharmaceuticals on U87MG cell viability was assessed to examine their radiotherapeutic effects (Figure 3F). Neither 0.1 MBq nor 0.2 MBq free [⁶⁴Cu]CuCl₂ demonstrated notable cytotoxicity against tumor cells. In contrast, both [⁶⁴Cu]Cu-DR and [⁶⁴Cu]Cu-DP1R exhibited significant cytotoxic effects compared to free [⁶⁴Cu]CuCl₂. Specifically, the 0.1 MBq [⁶⁴Cu]Cu-DR group showed a cell viability of 78.00 ± 6.35% (P = 0.003), while the 0.2 MBq group further decreased cell viability to 69.33 ± 0.64% (P < 0.0001). Notably, the 0.1 MBq [⁶⁴Cu]Cu-DP1R group achieved a cell viability of 62.38 ± 1.15% (P < 0.0001), and the 0.2 MBq group further reduced cell viability to 46.50 ± 4.62% (P < 0.0001), indicating a significant reduction in cell survival compared to the control.

PET Imaging of [⁶⁴Cu]Cu-DP1R in Tumor-Bearing Mice

To investigate the in vivo performance of [⁶⁴Cu]Cu-DP1R, we selected the integrin α_vβ₃-targeting peptide c(RGDyK) and the U87MG cell line, which highly expresses integrin α_vβ₃. Subcutaneous tumor experiments were conducted to validate the PET imaging of [⁶⁴Cu]Cu-DP1R at different time points (1, 4, 24, 72 h) in mice. One hour after intravenous injection into the U87MG subcutaneous tumor model, both [⁶⁴Cu]Cu-DR and [⁶⁴Cu]Cu-DP1R detected the tumor, whereas no significant tumor uptake was observed in the MCF-7 subcutaneous tumor model with low integrin α_vβ₃ expression (Figure 4A). The high uptake of [⁶⁴Cu]Cu-DP1R in the tumor persisted up to 72 h, while the tumor uptake of [⁶⁴Cu]Cu-DR showed a significant signal reduction at 72 h. Region of interest (ROI) analysis of PET images at different time points (Figure 4B) revealed that tumor uptake of [⁶⁴Cu]Cu-DP1R was significantly higher than that of [⁶⁴Cu]Cu-DR after 4 h (6.90 ± 0.98 %ID/g vs. 4.57 ± 1.08 %ID/g, P < 0.05), and this difference persisted up to 72 h ([⁶⁴Cu]Cu-DP1R vs. [⁶⁴Cu]Cu-DR: 6.20 ± 1.61 %ID/g vs. 1.97 ± 0.75 %ID/g, P < 0.01). The Retention index (RI) for [⁶⁴Cu]Cu-DP1R is 21.90% (± 41.29%), while the RI for [⁶⁴Cu]Cu-DR is −46.97% (± 14.82%). This study performed longitudinal pharmacokinetic comparison of [⁶⁴Cu]Cu-DR and [⁶⁴Cu]Cu-DP1R through quantitative analysis of tumor-to-blood (T/B) and tumor-to-muscle (T/M) signal ratios (Figure 4C–D). Initial biodistribution profiles at 1 h post-injection showed comparable targeting efficacy between the two radiotracers (T/B: 2.66 ± 0.21 vs. 2.51 ± 0.53; T/M: 9.75 ± 2.81 vs. 12.74 ± 4.96). [⁶⁴Cu]Cu-DP1R exhibited significantly enhanced tumor retention kinetics, achieving 2.7-fold higher T/B ratios than [⁶⁴Cu]Cu-DR by 4 h (5.20 ± 0.75 vs. 2.89 ± 0.36, P < 0.01), with concomitant T/M ratio elevation to 16.42 ± 2.03 versus 8.18 ± 2.89 (P < 0.01). The differential pharmacokinetic profiles became progressively more pronounced, culminating in peak T/B values of 10.12 ± 0.43 for [⁶⁴Cu]Cu-DP1R at 48 h (3.27 ± 0.81 for [⁶⁴Cu]Cu-DR, P < 0.001) and maximal T/M ratios of 24.55 ± 3.40 (vs. 6.92 ± 3.52, P < 0.001). At terminal timepoint (72 h), [⁶⁴Cu]Cu-DP1R demonstrated sustained tumor specificity with T/B and T/M ratios of 19.19 ± 5.54 and 24.47 ± 5.34, respectively, representing 3.9-fold (P < 0.05) and 3.5-fold (P < 0.01) enhancements compared to [⁶⁴Cu]Cu-DR (4.89 ± 1.16 and 7.03 ± 1.72).

Figure 4.

PET Imaging Study in U87MG Tumor-Bearing Mice. (A) PET images showing the distribution of [⁶⁴Cu]Cu-DR and [⁶⁴Cu]Cu-DP1R in U87MG tumor-bearing mice at various time points, as well as the distribution of [⁶⁴Cu]Cu-DP1R in MCF7 tumor-bearing mice; (B) Quantitative analysis of the average PET signal intensity for [⁶⁴Cu]Cu-DR and [⁶⁴Cu]Cu-DP1R within tumors based on the PET images shown in (A); (C) Quantitative analysis of the tumor- to-blood signal ratio (T/B) of [⁶⁴Cu]Cu-DR and [⁶⁴Cu]Cu-DP1R; (D) Quantitative analysis of the tumor-to-muscle signal ratio (T/M) of [⁶⁴Cu]Cu-DR and [⁶⁴Cu]Cu-DP1R; (E) Biodistribution analysis across different organs 72 h post-injection of each probe; (F) Integrin α_vβ₃ immunohistochemical staining images of U87MG and MCF-7 tumor tissues. * P < 0.05; **, P < 0.01; ***, P < 0.001.

Biodistribution results (Figure 4E) showed that 72 h after injection of [⁶⁴Cu]Cu-DP1R, the tumor uptake in U87MG tumor-bearing mice was significantly higher than that in the [⁶⁴Cu]Cu-DR group (tumor uptake: 7.79 ± 1.08 %ID/g vs. 1.87 ± 0.88 %ID/g, P < 0.01). Tumor uptake in U87MG tumor-bearing mice injected with [⁶⁴Cu]Cu-DP1R was also significantly higher than in MCF-7 tumor-bearing mice injected with [⁶⁴Cu]Cu-DP1R (tumor uptake: 7.79 ± 1.08 %ID/g vs. 0.47 ± 0.10 %ID/g, P < 0.01). The distribution of [⁶⁴Cu]Cu-DP1R in other vital organs 72 h post-injection was as follows: liver, 2.95 ± 0.13 %ID/g; kidneys, 1.43 ± 0.09 %ID/g; blood pool, 1.01± 0.5 %ID/g, indicating rapid metabolism in major organs. Immunohistochemistry was performed to identify tumors with high integrin α_vβ₃ expression, U87MG, and low integrin α_vβ₃ expression, MCF-7 (Figure 4F).

The disulfide-free control probe [⁶⁴Cu]Cu-DOTA-PEP2-c(RGDyK) (structurally identical to DP1R except lacking the S-S bond) (Figure S7–8) exhibited significantly reduced tumor retention versus DP1R at all timepoints. In U87MG tumors, [⁶⁴Cu]Cu-DOTA-PEP2-c(RGDyK) showed rapid clearance than [⁶⁴Cu]Cu-DP1R (72h uptake: 3.28 ± 0.45 %ID/g vs. 7.79 ± 1.08 %ID/g, P < 0.01) (Figure S9A). The high uptake of [⁶⁴Cu]Cu-DP1R in the tumor persisted up to 72 hours, while the tumor uptake of [⁶⁴Cu]Cu-DR, [⁶⁴Cu]Cu-DP2R showed a significant signal reduction at 72 hours. Regions of interest (ROIs) were drawn on PET images of U87MG tumor-bearing mice at various time points for [⁶⁴Cu]Cu-DR, [⁶⁴Cu]Cu-DP1R, and [⁶⁴Cu]Cu-DP2R. It was found that the tumor uptake of [⁶⁴Cu]Cu-DP1R was significantly higher than that of [⁶⁴Cu]Cu-DP2R at 4 hours post-injection (6.90 ± 0.98 %ID/g vs. 5.53 ± 0.36 %ID/g, P < 0.05) (Figure S9B). Furthermore, [⁶⁴Cu]Cu-DP1R maintained significantly higher tumor uptake compared to the [⁶⁴Cu]Cu-DP2R group even at 72 hours (6.20 ± 1.61 %ID/g vs. 2.93 ± 0.36 %ID/g, P < 0.01). Biodistribution results showed that the tumor uptake of [⁶⁴Cu]Cu-DP1R in U87MG tumor-bearing mice at 72 hours post-injection was significantly higher than that of the [⁶⁴Cu]Cu-DP2R group (Tumor uptake: 7.79 ± 1.08 %ID/g vs. 3.28 ± 0.45 %ID/g, P < 0.01) (Figure S9C). The uptake distribution of [⁶⁴Cu]Cu-DP1R in other major organs at 72 hours was as follows: liver, 5.94 ± 0.09 %ID/g; kidneys, 1.53 ± 0.35 %ID/g; blood pool, 1.19 ± 0.17 %ID/g.

Radiotherapy with [⁶⁴Cu]Cu-DP1R in U87MG Tumor-Bearing Mice

The therapeutic efficacy of the radiolabeled DP1R with the therapeutic radionuclide [⁶⁴Cu]Cu was assessed in U87MG tumor-bearing mice (Figure 5A). At the end of the treatment, tumors were excised and weighed. The [⁶⁴Cu]Cu-DP1R group had the smallest tumor weight (0.25 ± 0.03 g) compared to the [⁶⁴Cu]Cu-DR (0.87 ± 0.03 g), [⁶⁴Cu]CuCl₂ (2.36 ± 0.35 g, P < 0.001) and PBS groups (2.42 ± 0.35 g, P < 0.001) (Figure 5B and C). Twenty days after injection of [⁶⁴Cu]Cu-DP1R, tumor growth was significantly inhibited compared to the PBS group (Figure 5D and E, P < 0.001), with the tumor volume in the PBS group being 12.1 times larger than in the [⁶⁴Cu]Cu-DP1R group. The tumor volume in the [⁶⁴Cu]Cu-DR group was 3.33 times larger than in the [⁶⁴Cu]Cu-DP1R group (407.00 ± 66.34 mm³ vs. 122.20 ± 19.92 mm³, P < 0.001).

Figure 5.

Efficacy of internal radiation therapy in U87MG tumor-bearing mouse models. (A) Therapeutic observation scheme of tumor-bearing mice; (B) Tumor images from U87MG tumor-bearing mice in PBS, [⁶⁴Cu]CuCl₂, [⁶⁴Cu]Cu-DR, and [⁶⁴Cu]Cu-DP1R groups; (C) Tumor weight quantification across different groups at 20 days post-treatment; (D and E) Longitudinal tumor volume measurements (mm³) in different groups; (F) Ki-67 immunohistochemistry (brown), TUNEL immunofluorescence (green fluorescence), γ-H2AX immunofluorescence (red fluorescence) and ROS immunofluorescence (green fluorescence) of the tumor tissues; (G) Quantitative histopathology: relative protein expression or fluorescence intensity of Ki-67、TUNEL、γ-H2AX、ROS normalized to stained control (set as 1.0). Scale bar: 100 μm. ***, P < 0.001.

The anti-proliferative effects were quantitatively assessed through Ki-67 immunohistochemistry. [⁶⁴Cu]Cu-DP1R groups exhibited reduction in proliferative activity, significantly lower than untreated PBS groups (1.35 ± 0.17 vs. 4.13 ± 0.15, P < 0.001, Figure 5F and G). While TUNEL assays demonstrated 101.6-fold apoptotic induction (140.21 ± 24.88 vs. 1.38 ± 0.15, P < 0.001). DNA damage (γ-H2AX, 59.3 ± 11.49 vs. 1.08 ± 0.07, P < 0.001) and ROS accumulation (7.87 ± 0.70 vs. 1.36 ± 0.30, P < 0.001) were markedly elevated, exhibiting spatiotemporal correlation with treatment response.

During the treatment period, no significant behavioral abnormalities or weight loss were observed in any group (Figure 6A). Additionally, our results indicate that, compared to the PBS group, [⁶⁴Cu]Cu-DP1R effectively inhibited tumor growth and extended the median survival time of U87MG glioma-bearing mice (Figure 6B). The median survival in the [⁶⁴Cu]Cu-DP1R group was 36 days, compared to 26 days in the [⁶⁴Cu]Cu-DR group, 20 days in the free [⁶⁴Cu]CuCl₂ group, and 20 days in the PBS group. Blood samples were collected from tumor-bearing mice in each treatment group to assess liver and kidney function. Serum markers for liver function, including ALT, AST, and ALP, as well as BUN as a kidney function indicator, were measured. Results indicated no statistically significant differences in kidney or liver function among treatment groups, with BUN, ALT, AST, and ALP levels within normal ranges (Figure 6C–F). Major organs, including the heart, liver, spleen, lungs, and kidneys, were examined via H&E staining (Figure 6G). Findings showed no visible damage to these organs, suggesting good treatment tolerance, with no mortality or acute side effects observed at the applied dosage.

Figure 6.

Safety and biotoxicity evaluation of different groups in U87MG tumor-bearing mouse models. (A) Longitudinal body weights measurements (g) of mouse in different groups; (B) Survival curves of U87MG tumor-bearing mice in different groups; (C-F) Serum biochemical markers: BUN (C), ALT (D), AST (E), and ALP (F) levels at day 20 post-treatment; (G) H&E staining of major organs (heart, liver, spleen, lung, kidney) at the end of treatment. Scale bar: 100 μm.

Discussion

This study leverages a tumor microenvironment-induced self-assembling small peptide platform, which combines with radionuclide-targeted internal radiation therapy to design and synthesize the small peptide DP1R. This molecule was structured using the PEP1 peptide as the GSH consuming and self-assembly core (Phe-Phe peptide), incorporating a redox-responsive disulfide bond as a molecular switch. The DP1R molecule remains soluble in the circulation due to constrained conformation and electrostatic repulsion. However, it is specifically capable of self-assembling only upon GSH-triggering in the tumor microenvironment, transitioning from small molecules into micellar nanostructures. This critical selectivity arises because GSH-mediated reduction of the disulfide bond (S-S→ 2-SH) triggers three key molecular events: disruption of the constrained hydrophilic conformation, exposing the cryptic diphenylalanine (FF) motif and adjacent hydrophobic domains^[26]; liberation of the FF motifs, enabling spontaneous and strong π-π stacking interactions that drive β-sheet-rich nuclei formation and initiate fibril growth^[27]; and reduction in net negative charge density, significantly lowering electrostatic repulsion and facilitating molecular association^[28]. The transformation allows for prolonged retention at the tumor site, which not only enhances the local radioactive dose of the therapeutic concentration but also increases the cytotoxicity towards tumor cells due to the localized radiation effects generated by radioactive decay. The findings of this study have substantial clinical implications for improving existing radionuclide-targeted therapy strategies, indicating the potential of optimizing drug delivery systems to enhance intratumoral radiation doses and providing new directions for future therapeutic strategies.

The small molecules DR and DP1R were synthesized and purified with purity >95%. TEM analysis revealed that DP1R formed micelles with an average diameter of 27.55 ± 6.43 nm upon GSH incubation, with a slight discrepancy from DLS measurements, possibly due to hydration effects. Previous studies have reported that self-assembly molecules can produce a cascading effect upon activation, where an initial self-assembly event induces subsequent steps, promoting the assembly of other molecules through specific non-covalent interactions, forming larger aggregates, fibers, or networks^[29]. Further design of self-assembling molecules to generate controlled, progressive self-assembly via cascading reactions could provide new insights into prolonging the therapeutic effect of drugs within tumors to enhance efficacy^[30].

[⁶⁴Cu]Cu-DP1R demonstrated remarkable stability and minimal cytotoxicity, highlighting its potential as an effective agent for radiotherapy. The [⁶⁴Cu]Cu-DP1R group exhibited 90.98% internalization, similar to the [⁶⁴Cu]Cu-DR group, and enhanced therapeutic effects by localizing radiation within tumor cells, reducing radiation exposure to normal tissues. Efficient internalization of radiolabeled peptides improves radiotherapy by keeping the radionuclide inside the cancer cell, boosting treatment efficacy, and reducing radiation exposure to healthy tissues, thus minimizing side effects^[31]. In typical radiotherapy, tumor cells frequently elevate GSH levels as a protective mechanism to mitigate damage induced by ROS, which ultimately reduces the treatment’s effectiveness ^[32]. Previous studies have shown that the depletion of GSH can lead to increased ROS production, thereby potentiating the therapeutic effects of radiotherapy ^{[25, 33]}. In our study, we observed that the DP1R peptide, containing an S-S bond, significantly depleted GSH through its direct interaction with the molecule. This depletion was associated with increased ROS production and improved cytotoxicity in U87MG cells, compared to the group without the PEP1 peptide. Notably, radiation-triggered ROS generation may establish a self-amplifying loop: [⁶⁴Cu]Cu decay produces ionizing radiation that directly induces DNA damage while simultaneously elevating intracellular ROS levels through water radiolysis. This dual action synergizes with GSH depletion by DP1R, creating a lethal oxidative stress microenvironment that potentiates tumor cell death. The strong positive relationship between ROS intensity and γ-H2AX foci density, further supporting radiation-ROS-DNA damage axis as a key therapeutic mechanism.

In vivo, PET imaging showed both [⁶⁴Cu]Cu-DR and [⁶⁴Cu]Cu-DP1R successfully localized tumors in U87MG mice within 1 hour, with DP1R maintaining tumor uptake for over 72 h. The Retention Index for [⁶⁴Cu]Cu-DP1R was 21.90%, compared to −46.97% for [⁶⁴Cu]Cu-DR. The sustained tumor retention of [⁶⁴Cu]Cu-DP1R arises from α_vβ₃-mediated active targeting followed by GSH-triggered hierarchical assembly: rapid extracellular micellization (27.55 nm) enables physical entrapment, while subsequent aggregation (284 nm hydrated clusters) and charge inversion (−36.43 mV) potentiate lysosomal sequestration. This dual-mode retention strategy overcomes limitations of conventional EPR-dependent nanocarriers. Critically, the mechanistic necessity of disulfide reduction was confirmed by a structural control probe [⁶⁴Cu]Cu-DP2R—identical to DP1R but lacking the redox-responsive S-S bond—which exhibited equivalent early tumor targeting yet 58% lower 72h retention. This irrefutably proves that sustained retention depends on GSH-triggered assembly rather than inherent molecular properties. The prolonged retention of radiopharmaceuticals at the tumor site enhanced the therapeutic effects. [⁶⁴Cu]CuCl₂ emits both β⁺ and β⁻ particles as well as high-linear energy transfer Auger electrons, indicating its potential in radiotherapeutic applications^[34]. [⁶⁴Cu]Cu-DP1R demonstrated superior therapeutic outcomes, with the tumor volume being 3.33 times smaller than [⁶⁴Cu]Cu-DR group. Multiplex biomarker analysis revealed critical mechanistic insights into this therapeutic superiority. The 3.1-fold decrease in Ki-67 index (P < 0.01) indicates potent suppression of tumor proliferation, likely mediated by radiation-induced G2/M phase cell cycle arrest. More strikingly, we observed a 101.6-fold increase in TUNEL⁺ apoptotic cells (P < 0.001), significantly exceeding typical radiotherapy-induced apoptosis rates^[35]. This exceptional apoptotic response suggests concurrent activation of both intrinsic (mitochondrial) and extrinsic (death receptor) pathways through sustained oxidative stress potentiated by prolonged radioisotope retention. The therapeutic efficacy is further amplified by a self-reinforcing biochemical cycle: radiation-generated reactive oxygen species (ROS) reaching 7.87-fold elevation (7.87 ± 0.70 vs. 1.36 ± 0.30, P < 0.001) impair DNA repair mechanisms, while accumulated DNA damage reciprocally exacerbates oxidative stress^[36]. Importantly, this sustained biomolecular damage mechanism differs fundamentally from acute radiation effects, as prolonged intratumoral isotope retention maximizes ROS bioavailability to perpetuate DNA damage cascades^{[37, 38]}. Importantly, the absence of significant systemic toxicity in treated animals underscores the clinical translatability of this approach.

However, this study has some limitations. First, the molecular mechanisms by which self-assembly within the cell directs the radiotherapeutic effects and bystander effects in peritumoral stromal regions were not thoroughly explored. Future research should delve deeper into the biological processes underlying the changes in the tumor microenvironment. Additionally, while these radiopharmaceuticals showed good stability and uptake in vivo, their long-term stability and metabolic dynamics in complex in vivo environments are not yet fully understood and require further validation. Future work should also consider further functional optimization of small peptide molecules, such as the development of drug carriers, the introduction of targeting ligands, and improvements in peptide structure to enhance therapeutic efficacy. Moreover, long-term toxicity and safety studies are needed to ensure their safety in future clinical applications.

Conclusion

The in vivo self-assembly strategy described in this study presents a promising approach for enhancing the efficacy of targeted radionuclide therapy. By leveraging the unique advantages of small peptide drugs and nanomedicines, this strategy enables precise delivery of therapeutic radionuclides to tumor tissues, thereby improving therapeutic outcomes and reducing systemic toxicity. These findings have the potential to contribute significantly to the advancement of TRT and its clinical applications.

Supplementary Material

Supplementary Figures: Synthesis scheme; High-resolution mass spectrometry; HPLC results (PDF)

Figure 1.

Schematic diagram of GSH-responsive molecular probes, [⁶⁴Cu]Cu-DOTA-PEP1-c(RGDyK), targeting tumor sites, undergoing self-assembly to enhance therapeutic efficacy.

Highlights.

• GSH-responsive self-assembled probe ⁶⁴Cu-DP1R synthesized with successful radiochemical purity and stability.

• 2-fold higher tumor retention vs. control via PET imaging (P < 0.01)

• 3-fold tumor volume reduction with zero observed toxicity.