Hyaluronic acid-modified polymeric nanoplatform delivering ¹³¹I-Hyp suppresses post-ablation residual lesions in colorectal cancer metastases via necrosis-targeted radiotherapy

Abstract

Despite serving as a radical alternative to surgery for inoperable colorectal hepatic metastases patients, thermal ablation faces local tumor progression rates up to 25% from residual tumors, seriously compromising treatment efficacy and survival of patients. We constructed hyaluronic acid (HA)-modified nanoparticles as carriers for the hydrophobic necrosis-avid agent ¹³¹I-hypericin (¹³¹I-Hyp), enabling tumor necrosis-targeted radiotherapy. ¹³¹I-Hyp was synthesized via iodogen-catalyzed electrophilic substitution and loaded into amphiphilic block copolymer hyaluronan-b-poly(ε-caprolactone) (HA-PCL) using dialysis, yielding HA-PCL@(¹³¹I-Hyp) nanoparticles (HP-NPs). HP-NPs were characterized in terms of size, stability, and drug release. Biodistribution and antitumor efficacy in vivo were evaluated in rodent models (nude mice and SRG rats bearing HT-29 subcutaneous tumors) with residual tumors induced by incomplete microwave ablation. HP-NPs showed 84.32% encapsulation efficiency, a uniform spherical shape with a hydrodynamic diameter of 75.66 nm, and rapid cytosolic degradation, enabling the release of ¹³¹I-Hyp in necrotic regions. After intravenous injection into animals with residual tumors, HP-NPs accumulated in tumor tissue through the enhanced permeability and retention effect and CD44/HA receptor-ligand interactions. The released ¹³¹I-Hyp remained selectively in necrotic areas, delivering localized β-radiation to the surrounding residual tumor tissue and significantly inhibiting tumor growth via induction of apoptosis. In conclusion, HP-NPs enable targeted radiotherapy to residual tumor tissue after ablation for colorectal metastases by leveraging necrosis avidity and CD44-mediated HA endocytosis, effectively reducing post-ablation tumor progression. This nanoplatform shows potential for clinical translation in colorectal metastasis treatment.

Graphical abstract

Highlights

• •Amphiphilic HA-PCL nanoparticles delivering necrosis-avid ¹³¹I-hypericin for radiotherapy of residual lesions post-ablation.
• •HP-NPs are monodisperse spheres (<100 nm) that rapidly degrade in necrotic areas to release ¹³¹I-hypericin.
• •Dual targeting: passive accumulation via EPR effect & CD44 receptor-mediated uptake in residual tumors.
• •¹³¹I-Hyp targets tumor necrosis, emitting β-rays to suppress residual tumor growth by inducing apoptosis.
• •HP-NPs reduce local tumor progression after ablation in rodent models of colorectal metastases.

1. Introduction

Colorectal carcinoma (CRC) is the third most common cancer and has high morbidity and mortality rates worldwide. Approximately 15%–25% of patients with CRC are diagnosed with hepatic metastases at initial presentation, and another 45%–55% will develop hepatic metastases during treatment [1]. Thermal ablation techniques such as microwave or radiofrequency ablation are used to treat cancer by rapidly heating tumor tissues, resulting in coagulative necrosis [2]. Due to its minimally invasive nature and wide range of adaptations, thermal ablation has been recommended as a treatment method for patients with unresectable colorectal liver metastasis [3]. However, colorectal cancer metastases demonstrate the highest post-ablation local tumor progression rate (25%) across all malignancies, primarily attributed to suboptimal ablation margins, an independent predictor of recurrence [4]. This compromises curative-intent strategies, as patients with post-ablation recurrence rarely achieve satisfactory outcomes with traditional chemotherapeutic agents due to drug resistance [5]. Therefore, developing new strategies for the treatment of residual lesions after ablation of hepatic metastases from CRC has recently attracted increasing attention.

Coagulation necrosis is a characteristic change after thermal ablation, and residual lesions are often scattered around necrotic areas. Tumor necrosis therapy, which uses necrosis-affinity substances combined with radionuclides to treat residual tumors, targets the ablated tumor necrosis zone and is expected to solve the clinical issues related to CRC treatment. Hypericin (Hyp) is a natural nonporphyrin small-molecule compound with strong affinity for necrosis. In a neutral environment, hypericin can efficiently couple with iodine-131 in the form of covalent bonds (¹³¹I-hypericin (¹³¹I-Hyp)), thus facilitating necrosis-targeted therapeutic strategies [6]. Li et al. [7] reported that accumulated ¹³¹I-Hyp in the necrotic area could kill residual cancer cells with ionizing radiation and inhibit tumor regrowth. However, after the incorporation of an iodine atom into the Hyp structure, the water insolubility of ¹³¹I-Hyp tends to increase. Consequently, ¹³¹I-Hyp may form aggregates that lack necrosis avidity and remain largely in reticuloendothelial system organs, leading to impaired safety and clinical efficacy. The current commonly used cosolvent solubilization method inevitably introduces dimethyl sulfoxide (DMSO), which not only has potential biosafety hazards but also may lead to drug precipitation due to hemodilution after intravenous injection [8]. Therefore, the development of a suitable platform for delivering radioiodinated hypericin, which ensures efficient aggregation during tumor necrosis and optimal biosafety, is crucial.

Nanotechnology-based systems enable the direct delivery of functional units to tumor sites through passive accumulation, active ligand–receptor targeting, and surface functionalization, thereby achieving precise cancer diagnosis [9,10] and targeted therapy while minimizing systemic toxicity [[11], [12], [13]]. Among these, biodegradable polymeric nanoparticles address the limitations of synthetic carriers through synergistic integration of: (1) enhanced drug stability via encapsulation, (2) spatiotemporally controlled release responsive to pathological stimuli, and (3) precision targeting leveraging both enhanced permeability and retention (EPR) effect and ligand-receptor recognition [[14], [15], [16], [17]]. This paradigm enhances the therapeutic index and reduces systemic toxicity, creating a flexible platform for precision oncology applications [[18], [19], [20]]. Hyaluronan-b-poly(ε-caprolactone) (HA-PCL), an amphiphilic block copolymer synthesized via ring-opening polymerization, spontaneously self-assembles in aqueous solutions into well-defined micellar nanostructures with a hydrophobic PCL core and hydrophilic hyaluronic acid (HA) corona. This architecture allows for efficient encapsulation of hydrophobic therapeutics while utilizing the CD44-targeting ability of HA for tumor-specific delivery. By encapsulating ¹³¹I-Hyp in the core of nanoparticles, hydrophobic drug delivery and tumor necrosis-targeted therapy become possible. Hyaluronic acid nanoparticles have been investigated for tumor therapy and imaging because of their tumor-targeting ability in vivo. On the one hand, theranostic nanoparticles bearing an HA shell can effectively reach the tumor site via the EPR effect as well as an active targeting mechanism via the binding of HA to CD44, the HA receptor that is overexpressed on the surface of many tumor cells [21,22]. On the other hand, HA-PCL nanoparticles allow triggered ¹³¹I-Hyp release under high hyaluronidase (Hyal) levels in the cytosolic environment [23]. Furthermore, real-time information on the drug biodistribution of ¹³¹I-Hyp in the hydrophobic core can be obtained through fluorescence imaging and radionuclide tracing.

In this work, ¹³¹I-Hyp was encapsulated into amphiphilic polymers using the dialysis method to construct HA-PCL@(¹³¹I-Hyp) nanoparticles (HP-NPs), which are capable of necrosis-targeted radiotherapy (Scheme 1). Their self-assembly behavior, drug loading capacity, and drug release kinetics were thoroughly studied. Quantitative gamma counting, fluorescence microscopic imaging, and flow cytometric analysis jointly verified the necrosis-targeting propensity and subsequent internalization of HP-NPs in colon carcinoma cells, confirming their potential as theranostic vehicles for tumor-microenvironment-specific delivery. We further assessed the in vivo biodistribution and anticancer efficacy of necrosis-targeting HP-NPs in residual tumor models after thermal ablation, revealing a strong correlation between nanoparticle accumulation in necrotic foci and tumor growth inhibition. Our findings may provide a promising adjuvant therapeutic approach after ablation for colorectal cancer metastases, offering new strategies to reduce recurrence and improve patient prognosis.

Scheme 1.

Schematic illustration of polymer nanoparticles loaded with ¹³¹I-hypericin (¹³¹I-Hyp) for necrosis-targeted radiotherapy of residual tumors post-ablation. HA-PCL@(¹³¹I-Hyp) nanoparticles (HP-NPs) were constructed by encapsulating ¹³¹I-Hyp into amphiphilic polymer hyaluronan-b-poly(ε-caprolactone) (HA-PCL) microspheres via a dialysis method. These nanoparticles are designed to specifically target and persistently accumulate within necrotic tumor tissues. Through beta-ray irradiation, they effectively trigger apoptosis in residual tumor cells in the surrounding area, thereby providing a therapeutic effect on residual tumors post-ablation.

2. Materials and methods

2.1. Materials

Hypericin (>98%) was purchased from Nakeli Biotechnology Co., Ltd. (Chengdu, China). Iodogen (≥95%) and dimethyl sulfoxide (DMSO; ≥99.7%) were purchased from Sigma-Aldrich, Inc. (St. Louis, MO, USA). Na¹³¹I (radiochemical purity >99%) was obtained from HTA Co., Ltd. (Beijing, China). HA-PCL (≥99%, HA grafting ratio: >90%; molecular weights: 5k–5k Da, 5k–10k Da, and 10k–10k Da) was purchased from Qiyue Biological Technology Co., Ltd. (Xi'an, China). Cyanine-5.5 (98.41%) was purchased from MedchemExpress (Shanghai, China). Pyrene (≥99%), hyaluronidase (≥300 IU/mg), and all other analytical-grade reagents were purchased from Macklin (Shanghai, China).

2.2. Critical micellar concentration of HA-PCL

The critical micelle concentration (CMC) of HA-PCL was determined by the pyrene-based fluorescence probe method on a spectrofluorometer (FS5, Edinburgh Instruments Ltd., Livingston, UK) as previously described [24]. In brief, aliquots of pyrene stock solution (0.6 μM in acetone, 20 μL) were added to Eppendorf (EP) tubes, and the acetone was removed using an evaporator at 60 °C. 2 mL of HA-PCL amphiphilic copolymer diluted with ultrapure water was added to final concentrations ranging from 1.0 mg/mL to 6.0 × 10⁻⁵ mg/mL into each EP tube. The combined solutions of pyrene and block copolymer were incubated on a shaker at 37 °C in the dark for 24 h to reach solubilization equilibrium before measurement. Fluorescence spectrometry was used to determine the emission spectra of pyrene with an excitation wavelength of 336 nm. The ratio between the emission intensities at 338 nm and 333 nm was plotted against the base 10 logarithm of the concentration of each tested sample, and the CMC was obtained from the intersection point of the two slopes.

2.3. Radiolabeling of hypericin

Iodine-131-labeled hypericin was obtained through an electrophilic substitution reaction catalyzed by iodogen as previously described [25]. Briefly, 0.5 mg of hypericin was dissolved in 0.5 mL of DMSO to prepare a hypericin solution with a concentration of 1 mg/mL. Then, the hypericin solution and 0.5 mL of phosphate buffer solution containing Na¹³¹I (pH 7.3, 18.4 mCi radioactivity) were sequentially added to iodogen-coated tubes, and the reaction mixture was incubated at room temperature and protected from light for 30 min. The reaction was terminated by transferring the mixture (total volume 1 mL) to a new EP tube, yielding a¹³¹I-Hyp solution with a concentration of 0.5 mg/mL.

2.4. Radiolabeling yield and stability of ¹³¹I-Hyp

The radiolabeling yield of ¹³¹I-Hyp was determined by ascending paper chromatography using Whatman NO.1 filter paper (Whatman, Maidstone, UK) as the stationary phase and 0.1 M hydrochloric acid as the mobile phase [26]. 10 μL of ¹³¹I-Hyp solution (0.5 mg/mL) was spotted on a filter paper strip (1 cm × 10 cm) and developed for 15 min. After development, the labeled hypericin remained at the origin, and the free ¹³¹I shifted to the front of the filter paper. The filter paper strips were cut into two halves and counted for 1 min (cpm) using an Inspector Alert (IA-V2, International Medcom Inc., Sonoma County, CA, USA). Subsequently, three 100 μL aliquots of the reaction mixture were separately mixed with phosphate-buffered saline (PBS; Biosharp, Hefei, China), Roswell Park Memorial Institute (RPMI) 1640 medium (HyClone, Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS; Biological Industries, Beit Haemek, Israel), and rabbit serum at a 1:9 (v/v) ratio, and incubated at 37 °C in the dark. The stability of ¹³¹I-Hyp in vitro was evaluated by detecting radiolabeling yields at different time points (2, 4, 8, 12, 24, 48, and 72 h) after incubation.

2.5. Preparation of HP-NPs

¹³¹I-Hyp was loaded into the hydrophobic inner core of the amphiphilic polymer HA-PCL via the dialysis method. Briefly, HA-PCL (10 mg) was dissolved in 2 mL of warm DMSO, followed by mixing with 1 mL of radioiodine-131-labeled hypericin solution (0.5 mg/mL). This solution was then added dropwise into 5 mL of ultrapure water under moderate stirring. Subsequently, the mixed solution was dialyzed against deionized water at room temperature for 24 h using dialysis tubing (molecular weight cutoff = 7 kDa; Biosharp, Hefei, China) at 200 rpm (HS-350 Hotplate Magnetic stirrer, Yooning Instrument, Hangzhou, China), with deionized water replaced every 8 h to remove DMSO and unencapsulated ¹³¹I-Hyp. Following dialysis completion, the solution was carefully recovered and subjected to gentle evaporation concentration. Subsequently, it was centrifuged at 10,000 rpm for 3 min. The supernatant was filtered through a 0.22 μm syringe filter (Millex-GP, Merck, Darmstadt, Germany), yielding 3 mL of HP-NPs polymeric nanomicelles.

2.6. Characterization of the HP-NPs

The encapsulation efficiency of the HP-NPs was calculated according to the hypericin loaded in the nanoparticles relative to the total hypericin initially used for synthesis by absorptiometry. Briefly, a standard curve of free hypericin in DMSO was first plotted. The absorption profile of HP-NPs was examined using a NanoDrop 2000c spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Subsequently, the HP-NPs sample (100 μL) was dried under vacuum and resuspended in an equal volume of DMSO to disrupt the HA-PCL shell and dissolve the encapsulated ¹³¹I-hypericin. The absorption profile was then measured via ultraviolet–visible spectrophotometry (UV–Vis). Based on the standard curve, the concentration of ¹³¹I-hypericin in the sample was quantified, and the encapsulation efficiency of HP-NPs was calculated. The morphology of the HP-NPs was observed using cryogenic transmission electron microscopy (cryo-TEM). Briefly, samples (5 μL) were dropped on a 400 mesh copper grid with a thin carbon film and tested on a Themis 300 instrument (Thermo Fisher Scientific, Waltham, MA, USA) operated at an accelerating voltage of 200 kV. Images were collected with a defocus range of 2–4 μm. To measure the hydrodynamic diameter, polydispersity index (PDI), and Zeta potential of the HP-NPs by dynamic light scattering, the samples (10 μL) were diluted with 1 mL of ultrapure water and analyzed via a Zetasizer 3000 HSA particle analyzer (Malvern Instruments Ltd., Malvern, UK). All the parameters were tested in triplicate. The radioactivity of the synthesized nanomicelles was detected with a dose calibrator (CRC-55TR, Capintec Inc., Ramsey, NJ, USA), and ¹³¹I was selected as the detected nuclide.

2.7. Stability evaluation of the HP-NPs

The freshly prepared nanoparticles were stored at 4 °C in a dark environment for 30 days, and the encapsulation efficiency, hydrodynamic size, PDI, and Zeta potential measurements were repeated to evaluate the stability of the HP-NPs. Briefly, aliquots (100 μL) of HP-NPs after 30-day storage were desiccated, redissolved in an equivalent volume of DMSO, and subjected to UV–Vis. The encapsulation efficiency was calculated from the standard curve of absorbance versus concentration. To evaluate the long-term stability of nanoparticles, the hydrodynamic diameter, PDI, and Zeta potential of HP-NPs diluted at a volume ratio of 1:100 with ultrapure water at room temperature were measured using a Malvern laser particle size analyzer and compared with those of freshly synthesized nanoparticles. Furthermore, the stability of nanoparticles in vitro under physiological conditions was evaluated by incubation with cell culture medium or rabbit serum. HP-NPs (100 μL) were incubated with RPMI-1640 or rabbit plasma at a volume ratio of 1:4 at 37 °C in the dark. At fixed time intervals (immediately, 2, 4, 8, 12, 24, 48, and 72 h), the mixed solution was centrifuged (3000 rpm, 5 min), the supernatant was detected via an ultraviolet spectrophotometer, and the absorption curve was recorded. Each sample was tested three times.

2.8. In vitro¹³¹I-Hyp release and fluorescence tests

The in vitro release behavior of ¹³¹I-Hyp from the nanoparticles was assessed using dynamic dialysis. PBS (0.01 M, pH = 7.4) was used as the release medium. The HP-NPs (100 μg/mL) were dispersed in 2 mL of release medium with or without 120 units/mL hyaluronidase. The dispersed nanoparticles were placed into dialysis bags (molecular weight cutoff = 7000 Da) and immersed in 38 mL of PBS containing 0.1 % (w/v) Tween-80, then gently shaken at 120 rpm in a water bath at 37 °C. At predetermined time points (2, 4, 6, 8, 10, and 12 h), the release medium was removed, and the absorbance at 600 nm was measured via UV–Vis. The amount of ¹³¹I-Hyp released from the nanoparticles was measured on the basis of a standard curve of hypericin, and this process was repeated three times.

The fluorescence quenching and dequenching characteristics of the HP-NPs were evaluated in different simulated environments in vitro. Three identical aliquots of nanomicelles (500 μL, 70 μg/mL) were incubated at room temperature for 1 h with PBS (pH 7.4) as a control, 5 wt% sodium dodecyl sulfate (SDS; Macklin, Shanghai, China) to disrupt hydrophobic interactions, or hyaluronidase (120 U/mL) to degrade HA shells, respectively. Post-incubation, the three formulations were monitored using an IVIS KODAK IS4000 IN-VIVO FX system (Carestream Health, Inc., Rochester, NY, USA). The fluorescence excitation and emission wavelengths used were 550 and 600 nm, respectively, and the exposure time was 15 s. The intensity of the fluorescence signal was measured using Molecular Imaging software (version 7.2, Bruker Corporation, Billerica, MA, USA).

2.9. Cell culture and viability assays

HT-29 and HCT-15 human colorectal cancer cell lines were obtained from Procell Life Science&Technology Co., Ltd. (Wuhan, China) and cultured in RPMI-1640 supplemented with 10% FBS and penicillin‒streptomycin (100 U/mL; HyClone, Logan, UT, USA) in an incubator with 5% CO₂ at 37 °C. To determine cell viability, HT-29 cells (5 × 10³ cells/well) were seeded onto 96-well plates and incubated for 24 h. After cell stabilization, the culture medium was replaced with 100 μL of serum-free medium containing blank micelles of HA-PCL (0.03125–0.5 mg/mL) or HP-NPs with radioactive concentrations ranging from 0.6 to 4.8 mCi/mL, followed by incubation for 24 h at 37 °C. The cells were then washed twice, and cell viability was evaluated using a Cell Counting Kit-8 (CCK-8; Beyotime, Shanghai, China) assay.

2.10. Cell necrosis avidity assay in vitro

The affinity of HP-NPs for necrotic cells in vitro was assessed in cell necrosis models as previously described [25]. The cryo-necrosis model was induced using dry ice. Briefly, HT-29 cells were seeded in a six-well plate and grown to confluency. After removal of the culture medium, a cell cryo-necrotic area was created using dry ice pellets with a diameter of 10 mm that contacted the bottom of the culture well for 30 s. Subsequently, PBS solution containing 150 μL of HP-NPs was added to the culture well and incubated for 15 min at room temperature in the dark. Then, 4′,6-diamidino-2-phenylindole (DAPI; Beyotime, Shanghai, China) was added, and the cells were incubated for 5 min to stain the nuclei of dead and permeabilized cells. The cells were gently washed twice with PBS and imaged by a fluorescence microscope (Eclipse Ci-L 50i, Nikon, Tokyo, Japan) using NIS-Elements D software, version 4.6 (Nikon, Tokyo, Japan). Three-dimensional (3D) topographic maps of the fluorescence images were plotted using ImageJ 1.53k with the Interactive 3D Surface Plot plugin.

The thermal necrosis model of cells was constructed by a water bath method. Briefly, HT-29 cells were cultured in 75 cm² culture flasks until they reached confluency and then digested to generate cell suspensions. The suspensions were counted and divided evenly into two test tubes, each containing 1 × 10⁷ cells. After centrifugation (3000 rpm, 5 min), the supernatant was discarded, and the test tubes in the experimental group were incubated in a 60 °C water bath for 10 min, whereas those in the control group were incubated in a 37 °C water bath for the same duration. After incubation, 1 mL of PBS solution containing 100 μL of HP-NPs was added to the test tubes of both the experimental and control groups, which were then incubated at room temperature for 15 min. After the tubes were centrifuged at 3000 rpm for 5 min, the supernatant was discarded, and the cell sediment was washed twice with PBS. The cell sediment readings were measured with a gamma counter, using cpm as the unit.

2.11. Expression of the CD44 receptor

CD44 receptor expression in HT-29 and HCT-15 cells was evaluated by cellular imaging and flow cytometry analysis. HT-29 and HCT-15 cells were seeded in a 24-well plate with microscope slides at the bottom, using a culture medium containing decayed HP-NPs. Cells were fixed in freshly prepared 4 % paraformaldehyde (0.5 mL) at room temperature for 15 min. Nonspecific binding sites were blocked with 0.5 mL blocking buffer (5 % goat serum in PBS) for 60 min on a rocking shaker. CD44 antibody (ab6124, Abcam, Cambridge, MA, USA) was added to each well, and the samples were incubated overnight on a shaker at 4 °C. After the removal of the antibodies, the labeled cells were washed with PBS and PBS with Tween 20 (Biosharp, Hefei, China). Diluted goat anti-mouse IgG conjugated with Alexa Fluor™ Plus 488 (A32723TR; Thermo Fisher Scientific) was added (0.25 mL per well), and the mixture was incubated for 1 h at room temperature in the dark. The microscope slides were carefully removed and mounted with antifade mounting medium containing DAPI (G1407, Servicebio, Wuhan, China). The expression of the CD44 receptor in the two types of colon cancer cells was immediately evaluated by fluorescence microscopy.

The expression of CD44 was also assessed by flow cytometry. Briefly, HT-29 and HCT-15 cells were prepared as cell suspensions and divided into two equal volumes (3 mL; cell density: 1.0 × 10⁶ cells/mL), which were used as the experimental group and the control group. Anti-human/mouse CD44 conjugated with fluorescein isothiocyanate (FITC) (F1104401; Lianke Biotech Co., Ltd., Hangzhou, China) was added to the cells in the experimental groups at a 1:20 dilution in flow staining buffer, and the mixture was incubated at room temperature for 15 min in the dark. The cells were subsequently centrifuged at 300×g for 10 min and resuspended in flow cytometry staining buffer. The samples were then analyzed with a FACSCalibur (BD Biosciences, San Jose, CA, USA) using FL1 (515–545 nm).

2.12. Cellular uptake studies

To evaluate the cellular uptake and targeting efficacy of HA-modified HP-NPs toward different cell lines, HT-29 and HCT-15 cells were cultured in 35 mm culture dishes and stabilized for 24 h. For fluorescence imaging, cells were incubated with 100 μL of HA-NPs suspended in PBS (final volume 500 μL per dish) for 30, 60, or 120 min at 37 °C. For the competitive inhibition study, HT-29 cells were also cotreated with 8 mg/ml HA polymer (molecular weight = 5000 Da; Qiyue Biological Technology Co., Ltd. Xi'an, China) and HP-NPs. Cell nuclei were stained with Hoechst 33342 (Beyotime, Shanghai, China) at 1:250 dilution (final concentration: 4 μg/mL) for 20 min at room temperature in the dark. Unbound dye was removed by three washes with ice-cold PBS. The fluorescence images were observed using a Nikon Eclipse 50i microscope.

The cellular uptake of ¹³¹I-Hyp was also quantified by flow cytometry. HT-29 and HCT-15 cells were seeded in 6-well plates and stabilized for 24 h. After removing the culture media, cells were washed with PBS and incubated with fresh media containing HP-NPs (14 μg/mL) for 30, 60, or 120 min. Cells treated with PBS served as the control. Post-incubation, cells were washed with PBS, trypsinized, and collected. The pellets were resuspended in 1 mL PBS and analyzed on a flow cytometer using the FL3 channel (>650 nm).

2.13. Animal models of residual tumors after ablation

The animal study protocol was approved by the Institutional Animal Care and Use Committee (IACUC) of China Medical University (Approval number: KT20250890) on March 13, 2025. In this study, BALB/cA-nu athymic nude mice and Sprague‒Dawley Rag2/Il2rg double knockout (SRG) rats were obtained from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). Before experiments, the animals were anesthetized using a small animal anesthesia machine (R500, RWD, Shenzhen, China) with isoflurane. Animal models of subcutaneous residual tumors after ablation were used to evaluate the biodistribution and therapeutic efficacy of the HP-NPs in vivo. The successful establishment of the residual tumor model after ablation was verified by magnetic resonance imaging (MRI) and hematoxylin and eosin (H&E) staining. The nude mouse model was prepared by injecting 200 μL of RPMI-1640 medium/matrigel mixture (1:1, v/v) containing HT-29 cells (5.0 × 10⁶ cells) into the right side of the hindlimbs of 4-week-old athymic nude mice. The diameter of the tumor was measured using a vernier caliper, and the tumor volume (TV) was calculated by the following formula: TV (mm³) = 1/2 × A × B², where A and B represent the length and width of the tumor, respectively. A residual colon cancer tumor model was prepared by incomplete microwave ablation (MWA) as described previously [27]. Briefly, when the volume of the subcutaneous tumor reached approximately 200 mm³, a MWA probe (XR-A2015W, Nanjing Great Wall Medical Equipment Co., Ltd., Nanjing, China) was used to puncture the subcutaneous tumor from the caudal side along the long axis of the body, and the needle tip was located 1/3 of the cephalic side of the tumor. The residual tumor model was prepared with a power of 3 W and ablation for 12 s. The probe was gently removed with ongoing MWA emission to prevent tumor metastasis along the needle tract. A tumor-bearing model of SRG rats was constructed according to previous methods [28]. Briefly, 100 μL of the HT-29 cell suspension (1.0 × 10⁷ cells) was mixed with 200 μL of matrigel and injected into the right hindlimbs of 4-week-old SRG rats. When the subcutaneous tumors grew to approximately 1000 mm³, the ablated residual colon cancer tumor model was constructed by MWA as previously described with a power of 3 W for 20 s. MRI scanning was performed 48 h after the ablation of the animal model using a 7.0 T NOVA magnetic resonance system (Time Medical, Shanghai, China) equipped with a surface coil to obtain T1-weighted imaging (T1WI) and T2-weighted imaging (T2WI) with the following scanning parameters: T1WI: TR/TE, 800/12 ms; FOV, 40 × 40 mm; imaging acquisition matrix, 256 × 192, with a cross-sectional thickness of 1.0 mm and a cross gap of 0.3 mm; T2WI: TR/TE, 2500/12 ms; FOV, 40 × 40 mm; and imaging acquisition matrix, 256 × 192, with a cross-sectional thickness of 1.0 mm and a cross-gap of 0.3 mm. The model animals were sacrificed immediately after MRI scanning, and the subcutaneous tumors were collected for H&E staining.

2.14. Biodistribution of HP-NPs in vivo

The biodistribution of nanoparticles can be tracked via an in vivo imaging system (IVIS) and single-photon emission computed tomography/computed tomography (SPECT/CT). For fluorescence imaging, residual tumor models of colorectal cancer metastases were prepared by incomplete MWA in 8 tumor-bearing nude mice, and 200 μL of HP-NPs (140 μg/mL) was injected into the mice via the tail vein. The in vivo biodistribution of the nanoparticles was imaged using a KODAK IS4000 IN-VIVO FX system at different time points after injection, with an excitation wavelength of 550 nm, an emission wavelength of 600 nm, and an exposure time of 15 s. The region of interest (ROI) was plotted, and the fluorescence intensity of the subcutaneous tumor area was measured using Bruker MI SE software (v.7.2, Bruker, Billerica, MA, USA). For SPECT/CT imaging, 5 SRG model rats were injected with HP-NPs containing approximately 2 mCi ¹³¹I through the tail vein, and SPECT/CT scans were performed using a Discovery NM/CT 670 Pro (GE Healthcare, Waukesha, WI, USA) at different time points after injection. SRG rats were fixed on the patient bed in the prone position, and the scanning mode was selected as I¹³¹ TOMO CT (zoom: 1.5, matrix: 128 × 128, step and shoot: 12 s, collimator: HEGP, energy: I¹³¹ SC). After radionuclide scanning, the rats were continuously fixed for CT scanning. After imaging data reconstruction, the volume of interest (VOI) of the signal area of the subcutaneous tumor and the standard tube were delineated using Xeleris Functional Imaging Workstation (v.4.0). The local uptake value of the tumor site was calculated by comparison with that of the standard tube. As a control, cyanine-5.5 (Cy5.5), a near-infrared (NIR) fluorescent dye without necrosis avidity, was loaded into HA-PCL nanoparticles for intravenous injection into a nude mouse model. The preparation procedure for the Cy5.5-loaded nanoparticles was similar to that previously described in the literature [29]. Briefly, Cy5.5 was dissolved in DMSO, mixed with the amphiphilic copolymer HA-PCL, and gently vortexed. Unreacted Cy5.5 molecules were removed by dialyzing the mixture for 48 h, and finally, Cy5.5-loaded nanoparticles were obtained. Nude mice bearing partially ablated HT-29 tumors were intravenously injected with Cy5.5-loaded nanoparticles (Cy5.5 dose of 0.2 mg/kg). At the prescribed time after injection (1, 2, 4, 8, 12, 24, 48, 72, 96 and 120 h), the mice were anesthetized, and NIR fluorescence images were obtained using a 12-bit charge coupled device (CCD) camera (KODAK Imaging Station 4000 MM, Eastman Kodak Co., Rochester, NY, USA) equipped with a Cy5.5 bandpass emission filter set (680–720 nm; Omega Optical, Brattleboro, VT, USA). All the NIR fluorescence intensities were calculated using the ROI function of Analysis Workstation software (ART Advanced Research Technologies Inc., Saint-Laurent, Canada).

2.15. HP-NPs distribution ex vivo

The biodistribution of HP-NPs in organs and tumor tissues after intravenous injection was assessed using animal models of nude mice and SRG rats. Nude mouse models were randomly divided into four groups (n = 3 per group) and administered 200 μL of HP-NPs (140 μg/mL) via tail vein injection. Animals were sacrificed at designated time points post-injection (1, 24, 72, and 168 h). Various organs, including subcutaneous tumors, were collected, placed in petri dishes, and imaged using the KODAK Imaging Station 4000 MM. The fluorescence intensities of the subcutaneous tumors and various organs were measured by Bruker MI SE software. The tumors were subsequently incised along the short diameter and fixed with tissue fixative. H&E staining and fluorescence microscopy were conducted to confirm the necrotic regions in the subcutaneous tumor and the distribution of red fluorescence (G-2A, EX: 510−560 nm) emitted by hypericin in the partially ablated tumor. Moreover, subcutaneous tumors from the Cy5.5 control group were collected at 12 h after injection for ex vivo NIR fluorescence imaging, and sections of tumor tissues were viewed by a focus drift compensating microscope (IX81-ZDC2, Olympus, Tokyo, Japan) with excitation and emission wavelengths of 673 nm and 692 nm, respectively.

The biodistribution of ¹³¹I-Hyp delivered by nanoparticles was also assessed by tissue gamma counting (TGC). Three SRG rats were injected with HP-NPs with a radioactivity of approximately 2 mCi via the tail vein. According to the results of SPECT/CT imaging, the SRG rats were sacrificed when the local uptake value of the subcutaneous tumors reached the highest value. The organs and subcutaneous tumors were removed, weighed, and counted for 1 min (cpm) using a Geiger-Müller counter. The radioactivity uptake of ¹³¹I-Hyp in different tissues is expressed as a percentage of the injected dose per gram (%ID/g). Autoradiography was performed on 30 μm slices of subcutaneous tumors from SRG rats using a Molecular Imager FX system (Bio-Rad Laboratories, Hercules, CA, USA) to explore the distribution of ¹³¹I-Hyp in necrotic tumor regions. The slides were exposed for 12 h to a storage phosphor screen, which was then read by a scanner. The acquired autoradiographic images were analyzed using Quantity One 4.6.6 software (Bio-Rad Laboratories, Hercules, CA, USA). The exposed slides were stained with H&E and digitally photographed to determine the presence or absence of necrosis, and the results were compared with autoradiographs.

2.16. In vivo radiotherapy and pathological analysis of HP-NPs

Residual tumor-bearing nude mice were prepared by incomplete MWA as described above, and the model mice were randomly divided into three groups, with 6 mice in each group. The initial body weights and subcutaneous tumor volumes of the nude mice were recorded as baseline. Nude mice in the HP-NPs group were injected with the nanoparticles at a dose of 8 mCi/kg through the tail vein, with a volume of 200 μL. In the blank micelle group, 200 μL of HA-PCL micelle solution at a concentration of 3.33 mg/mL was injected. The control group was given the same volume of saline via the tail vein. The therapeutic results of each group were evaluated by measuring the tumor volume for 14 days. On the 14th day, the mice were sacrificed, and the tumor tissues were surgically removed. The relative tumor proliferation rate was calculated as T/C (%) and calculated from the relative tumor volume (RTV) on Day 14. T/C (%) = T_RTV/C_RTV × 100%. T_RTV was the treatment group, and C_RTV was the control group. For biological electron microscopy (BEM), a volume of 1 mm³ of tumor tissue around necrotic areas was fixed with 2.5% glutaraldehyde and 1% osmium tetroxide, dehydrated with acetone, embedded in Epon812 resin, cut into 70−90 nm sections, stained with uranyl acetate and lead citrate solution, and assessed by transmission electron microscopy (H-7650, Hitachi, Tokyo, Japan) at 80 kV. For histological analysis, tumor tissues were fixed with 10% neutral formalin solution and embedded in paraffin. The sectioned tumor tissues (5 mm) were prepared for H&E staining, terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining, and immunohistochemical staining. Immunohistochemical indicators included caspase-3, caspase-7, BAX, and γ-H2A.X, and Ki67. The relative protein levels were quantified using average optical density (AOD) analysis with ImageJ software (version 1.53k), and the Ki67 immunohistochemical results were evaluated by calculating the percentage of positive cells.

2.17. Safety evaluation in vivo

Nude model mice were randomly divided into two groups (n = 6 per group). The experimental group was injected with 8 mCi/kg HP-NPs via the tail vein, and the control group was injected with an equal volume of saline. All nude mice were sacrificed at 14 days after administration, and organs, including the heart, liver, spleen, lung, kidney, brain, pancreas, intestine, and thyroid, were harvested. After the samples were fixed with tissue fixative, they were embedded in paraffin blocks and sectioned to evaluate histological changes by H&E staining. Blood samples were collected using EP tubes coated with ethylenediaminetetraacetic acid (EDTA). White blood cells (WBCs), red blood cells (RBCs), hemoglobin (HGB), platelets (PLTs), alanine aminotransferase (ALT), aspartic transaminase (AST), total bilirubin (TBIL), and creatinine (Cr) were detected to evaluate changes in routine blood and serum biochemistry.

2.18. Statistical analysis

Continuous data are expressed as the mean ± standard deviation (SD). The differences between the two groups were compared using independent sample t tests after normality tests and homogeneity of variance tests. Differences between groups were analyzed using one-way analysis of variance (ANOVA). All the statistical analyses were performed using SPSS (version 26.0, IBM, Armonk, NY, USA) and GraphPad (version 9.0, La Jolla, CA, USA) software, and a P value less than 0.05 was considered statistically significant.

3. Results and discussion

3.1. Characterization and optimization of HA-PCL nanomicelles

HA-PCL with different proportions of amphoteric groups could all self-assemble into micelles in water. The blank micelles were colorless and transparent solutions. When the laser beam passed through the micelle system, a clear and bright optical path appeared, indicating a typical Tyndall phenomenon (Fig. S1). According to the dynamic light-scattering (DLS) results, the hydrodynamic diameter of the blank HA-PCL nanoparticles was 120.7 ± 0.45 nm (PDI 0.174 ± 0.003) when the molecular weight of HA:PCL was 5000:10,000 Da. When the molecular weights of HA:PCL were 5000:5000 Da and 10,000:10,000 Da, the hydration kinetic diameter of the nanoparticles in the blank micelles was 245.8 ± 1.55 nm (PDI 0.174 ± 0.026) and 330.2 ± 1.46 nm (PDI 0.148 ± 0.038), respectively (Fig. S2). The surface charges of the HA-PCL micelles with three different amphoteric group ratios were all negative, and the Zeta potentials were all greater than 30 mV (Fig. S3), indicating that the colloid stability was highly stable according to previous drug delivery literature [30]. The average size is an important parameter of nanoparticles in aqueous solutions used for drug delivery and varies considerably depending on the cargo molecules and the length ratio of the hydrophilic and hydrophobic blocks [31]. Nanomedicines in the sub100 nm range are currently regarded as important in the study of tumor penetration [32]. Based on the research findings that the hydrophobic core enhances intermolecular hydrophobic interactions between amphiphilic copolymers, promoting tighter self-assembled structures and thereby reducing particle size [33], the amphiphilic copolymer HA-PCL with molecular weight of 5000:10,000 Da was selected as the carrier material for subsequent experiments to optimize nanoparticle dimensions. The CMC is the minimum concentration of polymer required for micelles to form and is an important index of micellar stability. In our work, the CMC of HA-PCL (molecular weight: 5k–10k Da) was measured through a pyrene fluorescence probe. Fig. 1A shows the relationship between the fluorescence intensity ratio (I₃₃₈/I₃₃₃) of pyrene and the logarithm of the HA-PCL concentration at room temperature. The fluorescence intensity ratio remained stable at low concentrations but gradually increased with increasing copolymer concentration. Pyrene aggregates into the hydrophobic core of the micelles via hydrophobic interactions, leading to a fast increase in the fluorescence intensity ratio. The CMC was estimated from the threshold concentration of self-assembled nanoparticle formation by intra- or intermolecular associations and was calculated to be 0.0179 mg/mL. A low critical micelle concentration suggests greater resistance to disruption caused by the rapid dilution effect after intravenous or arterial injection [34]. This is of particular importance for their use as drug carriers because drug-loaded nanoparticles always travel through a large volume of body fluids before reaching their target site. Therefore, the HA-PCL copolymer has potential as a suitable drug carrier because of its low critical micelle concentration. Notably, although a single CMC value is often measured to characterize a system, the CMC value varies depending on the composition of the vehicle in which the polymer is dissolved, and it cannot be assumed that the CMC in water or buffer is the same as that in plasma or other biological fluids [35]. The ¹H NMR spectrum of HA-PCL (molecular weight: 5 kDa–10 kDa) is shown in Fig. S4.

Fig. 1.

Characteristics of the HA-PCL@(¹³¹I-Hyp) nanoparticles (HP-NPs). (A) The critical micelle concentrations of the hyaluronan-b-poly(ε-caprolactone) (HA-PCL; molecular weight: 5 kDa–10 kDa). (B) Stability of ¹³¹I-hypericin (¹³¹I-Hyp) in different in vitro environments. (C) The ultraviolet spectral absorbance values of the nanoparticles with the same concentration of hypericin were significantly lower than those of free ¹³¹I-Hyp (P < 0.0001). (D, E) HP-NPs presented a uniform spherical morphology under cryogenic transmission electron microscopy (cryo-TEM) (D), and a corona of the hydrophilic shell could be observed around the nanoparticles (E). (F) Nanoparticle size distribution of HP-NPs measured by cryo-TEM. (G) Hydrodynamic size distributions of freshly synthesized HP-NPs and those stored for 30 days after synthesis, as measured by dynamic light-scattering. (H) Changes in the Zeta potential of freshly synthesized HP-NPs and after 30 days of placement. (I) Changes in ultraviolet absorption values at 598 nm of HP-NPs incubated in culture medium or rabbit serum for 72 h. PBS: phosphate-buffered saline; RPMI-1640: Roswell Park Memorial Institute 1640.

3.2. Synthesis and characterization of the HP-NPs

In the present study, hypericin was radioiodinated through an electrophilic substitution reaction catalyzed by iodogen (Fig. S5). The radiolabeling yield immediately after the reaction was determined by chromatography to be 98.33% ± 0.12%. In vitro stability tests revealed that the labeling yield of ¹³¹I-Hyp in phosphate solution was 94.9% ± 0.40% after incubation at 37 °C for 72 h, with only a 3.43% reduction, indicating favorable stability. However, after 72 h of culture in cell culture medium and rabbit plasma, the radiolabeling yields of ¹³¹I-Hyp were 85.7% ± 1.93% and 80.8% ± 1.44%, respectively. Whereas ¹³¹I-Hyp underwent partial deiodination after 72 h of incubation in culture medium and serum, the labeling yield of ¹³¹I-Hyp was still maintained above 80%, indicating the tight binding of iodine atoms to hypericin (Fig. 1B). The interaction between ¹³¹I-Hyp and various complex components in cell culture medium and plasma may be one of the factors underlying the occurrence of deiodination.

HP-NPs were successfully prepared via the loading of ¹³¹I-Hyp into the hydrophobic core of the amphiphilic HA-PCL polymer via a dialysis method. The solution in the dialysis tube was centrifuged and filtered to finally obtain purple‒black nanomicelles (pH = 7.2 ± 0.49) consisting of HP-NPs, and the radioactivity measured by a dose calibrator was 15.14 mCi. Ultraviolet spectroscopy revealed that hypericin in DMSO has a characteristic absorption peak at 600 nm (Fig. S6). A standard curve was drawn with the concentration of hypericin as the abscissa and the absorbance at 600 nm as the ordinate. The standard curve is shown in Fig. S7 with the equation of y = 9.0596x + 0.0101 (R² = 0.999). The HP-NPs were assessed by a UV spectrophotometer after the organic film ruptured, and according to the standard curve, the encapsulation efficiency of the nanoparticles was 84.32% ± 0.57%. Moreover, the UV spectral absorbance values of the nanomicelles with the same concentration of hypericin were significantly lower than those of free ¹³¹I-Hyp, and the absorption peak shifted to 598 nm (Fig. 1C). This quenching effect, which may be caused by the solid and insoluble form of radioiodine-labeled hypericin encapsulated by the HA-PCL shell, has also been reported in previous studies and could be used to detect the stability of HP-NPs [36]. The morphology revealed by cryo-TEM revealed that the HP-NPs were monodisperse, spherical-like particles with a diameter of 66.61 ± 10.78 nm (Figs. 1D–F). The Malvern laser particle size analyzer results indicated that the hydration dynamic diameter of the HP-NPs was 75.66 ± 0.90 nm, the PDI was 0.073 ± 0.022, and the Zeta potential was −53.30 ± 2.88 mV (Figs. 1G and H). The particle size determined by cryo-TEM was smaller than that obtained via DLS, possibly because of the different conditions to which the nanoparticles were subjected. According to the Stokes‒Einstein equation, the hydrodynamic size measured by DLS is inversely proportional to the diffusion coefficient, whereas the modification of hydrophilic ligands decreases the diffusion coefficient, thus increasing the hydrodynamic size of nanoparticles [37]. This particle size, which was smaller than 100 nm, was demonstrated to be advantageous for high accumulation in tumor tissues with fenestrated vascular structures via the EPR effect [38].

For stability evaluation, the HP-NPs were stored in the dark at 4 °C for 30 days, and the tests described above were repeated. The encapsulation efficiency of the nanoparticles after 30 days of storage was 78.50% ± 0.51%, indicating that 5.82% of the hypericin was released from the nanoparticles during storage. DLS and Zeta potential measurements revealed that the hydrodynamic diameter (74.88 ± 0.94 nm), PDI (0.088 ± 0.032), and surface charge (−56.5 ± 5.69 mV) of the nanoparticles did not significantly change, suggesting the good stability of the HP-NPs in vitro (Figs. 1G and H). After incubation of the nanoparticles in RPMI-1640 culture medium or rabbit serum for 72 h, the absorbance at 598 nm did not increase significantly, with P values of 0.161 and 0.104, respectively (Fig. 1I). The above results indicate that HP-NPs exhibited high stability under physiological-mimicking conditions. However, it is unclear whether interactions between nanoparticles and various components in the blood after intravenous injection will affect their stability. More studies on the in vivo stability of HP-NPs are also needed before their clinical application.

3.3. Analysis of HP-NPs drug release behavior

The drug release pattern of ¹³¹I-Hyp from HP-NPs was further evaluated with and without hyaluronidase, which is rich in the cytosol of tumor cells and plays a pivotal role in the degradation of HA. As shown in Fig. 2A, HP-NPs without hyaluronidase exhibited slow drug release, with 32.37% ± 1.22% ¹³¹I-Hyp released from the nanoparticles after 12 h of incubation. However, the nanoparticles released 82.23% ± 1.50% of ¹³¹I-Hyp after incubation in the presence of 120 units/mL hyaluronidase, indicating rapid and sustained drug release. The release rate of ¹³¹I-Hyp was much greater in the presence of Hyal, suggesting that the HP-NPs rapidly released ¹³¹I-Hyp in the cytosol of the tumor cells because the HA shell of the nanoparticles was highly susceptible to Hyal.

Fig. 2.

Multifunctional evaluation of HA-PCL@(¹³¹I-Hyp) nanoparticles (HP-NPs) in vitro. (A) Release profile of HP-NPs with or without hyaluronidase (Hyal; 120 units/mL). ^∗P < 0.05. (B) Fluorescence images of the same concentration of HP-NPs after coincubation with phosphate-buffered saline (PBS), sodium dodecyl sulfate (SDS), or Hyal. (C) In vitro fluorescence signal intensity of three groups of Eppendorf tubes as measured by Bruker Molecular Imaging software. ^∗∗∗P < 0.001, ^∗∗∗∗P < 0.0001. (D) Hyaluronan-b-poly(ε-caprolactone) (HA-PCL) blank micelles maintained good cytocompatibility in the concentration range of 31.3–500 μg/mL. (E) HP-NPs exhibited a dose-dependent inhibitory effect on HT-29 cells. (F) Both the negative control and HP-NPs-treated groups exhibited hemolysis rates <5 % across all tested concentrations. ¹³¹I-Hyp: ¹³¹I-hypericin.

3.4. In vitro fluorescence imaging of HP-NPs

The above data demonstrate that the degradation of HA-PCL causes a significant increase in ¹³¹I-Hyp fluorescence signals, offering potential for real-time tracking of drug distribution. The fluorescence imaging capability of ¹³¹I-Hyp released from HP-NPs in vitro was subsequently determined via the KODAK Imaging Station 4000 MM. Different environments were simulated by fixing the concentration of hypericin (70 μg/mL) but varying different additives in the solutions; 5 wt% SDS is commonly used to imitate the cytoplasmic environment after cell disintegration in the necrotic area [36], whereas hyaluronidase is abundant in the cytosol of tumor cells [23,39]. As shown in Figs. 2B and C, the HP-NPs co-incubated with 5 wt% SDS or hyaluronidase at 120 units/mL exhibited strong fluorescence signals, while the nanoparticles remained stable in the PBS solution and showed significantly lower fluorescence intensity. The fluorescence intensity of ¹³¹I-Hyp encapsulated within the HA-PCL film was significantly diminished due to quenching effects. Hyaluronic acid is an important component of the extracellular matrix. In necrotic tissue, as the extracellular matrix degrades, the content of hyaluronidase typically increases [40]. When the HP-NPs experienced a necrotic environment or were degraded by hyaluronidase, the fluorescence intensity increased significantly because ¹³¹I-Hyp was rapidly released from the nanoparticles. Notably, the cytosolic environment mimicked by SDS appeared to have a stronger disintegration effect on HA-PCL than on hyaluronidase.

3.5. Cytotoxicity and hemocompatibility

The in vitro cytotoxicity of the HA-PCL blank micelles and HP-NPs toward HT-29 cells at various concentrations was evaluated using a CCK-8 assay. As shown in Fig. 2D, when the concentration of HA-PCL was in the range of 31.3–125 μg/mL, the cell viability was greater than 89.15% ± 3.99%. The cell survival rate decreased slightly but remained above 80% at HA-PCL concentrations ranging from 250 to 500 μg/mL, demonstrating that this amphiphilic polymer is substantially nontoxic when used for antineoplastic drug delivery. However, the HP-NPs inhibited the proliferation of HT-29 cells in a dose-dependent manner (Fig. 2E). With an increasing concentration of radioactivity of the HP-NPs (0.6–4.8 mCi/mL), the proliferative capacity of the HT-29 cells gradually decreased. The half maximal inhibitory concentration of HP-NPs on HT-29 cells was 0.9677 mCi/mL, as determined by plotting a dose‒response curve (Fig. S8).

Hemocompatibility tests of the drug-loaded nanoparticles were performed by hemolysis tests. Both the negative control PBS tubes and the tubes containing different concentrations of HP-NPs exhibited transparent supernatants, and no obvious hemolysis was observed, whereas hemolysis was evident in the positive control diH₂O tubes (Fig. S9). The results of multifunctional microplate reader detection indicated that the hemolysis rates of the negative control group and the experimental groups at different concentrations (hypericin concentrations of 6.25–50 μg/mL, radioactivity concentrations of 0.22–1.80 mCi/mL) were less than 5%, indicating that no significant hemolysis occurred (Fig. 2F). The hemolysis evaluation suggested that the polymeric nanoparticles encapsulated with ¹³¹I-Hyp had good compatibility with blood and were suitable for in vivo delivery.

Hyaluronic acid, a polydisaccharide composed of D-glucuronic acid and D-N-acetylglucosamine, is one of the major components of the extracellular matrix in connective tissues [41]. PCL blocks are degradable by certain lipases and are commonly used as hydrophobic tails of amphiphilic polymers [42]. In this study, self-assembled polymeric micelles obtained from HA-PCL exhibited unique biosafety and hemocompatibility. In addition to a strong affinity of hypericin for necrotic tissue, hypericin is also a powerful natural photosensitizer that is applied in photodynamic therapy for various oncological diseases. However, hypericin has minimal or no toxicity without actinization [43]. Consequently, we propose that the observed inhibition of HT-29 cell proliferation is primarily mediated by β-particles emitted from ¹³¹I.

3.6. Necrosis affinity assay in vitro

To construct necrosis models of tumor cells in vitro, freezing and hyperthermia procedures were performed to induce the death of HT-29 cells as previously described [25]. The in vitro necrotic affinity of the HP-NPs was evaluated on the basis of their fluorescence and radiation characteristics. The frozen cell death model was prepared using the lethal freezing temperature of dry ice, with the microscope field of view focused on the boundary of the cryogenic region. Since DAPI does not acutely penetrate viable cells with intact cell membranes, it can be used for nuclear staining as well as dead cell imaging. As shown in Figs. 3A and B, after staining with DAPI for 5 min, a significantly greater blue fluorescence intensity of the nuclei in the cryonecrotic area induced by dry ice was observed, which confirmed the successful construction of the cryogenic necrotic cell model. Moreover, the intensity of red fluorescence in the frozen necrotic area was significantly higher than that in the viable cell zone. Merged fluorescence images and 3D surface plots of the fluorescence images revealed that the red fluorescence emitted by ¹³¹I-Hyp was colocalized with the blue fluorescence of necrotic cell nuclei, confirming the preserved necrosis affinity of ¹³¹I-Hyp in HP-NPs. In addition, the merged images also revealed that hypericin targeting necrotic cells was deposited mainly in the cytoplasm.

Fig. 3.

In vitro necrosis affinity evaluation of the HA-PCL@(¹³¹I-Hyp) nanoparticles. (A) Fluorescence micrographs and three-dimensional (3D) intensity surface plots depicting 4′,6-diamidino-2-phenylindole (DAPI; blue) and hypericin (Hyp; red) co-localization in cryo-induced necrotic HT-29 cells post-incubation. The yellow dashed lines separate the necrotic area on the left and the viable cell area on the right. (B) Comparison of the fluorescence intensities of DAPI and Hyp in the hypothermic necrotic area and the viable cell area. ^∗∗∗∗P < 0.0001. (C) The radioactivity of the cell sediments in the experimental group was significantly greater than that in the control group. ^∗∗∗∗P < 0.0001. HA-PCL: hyaluronan-b-poly(ε-caprolactone).

The thermal necrosis model of cells was prepared using a 60 °C water bath for 10 min. After washing with PBS, the cell sediments in the experimental group appeared pale pink (hypericin was red when dissolved), whereas the cell sediments in the control group were milky white (Fig. S10). The gamma values detected by a Geiger−Müller counter suggested that the radioactivity of the cell sediments in the experimental group was significantly greater than that in the control group (7953 ± 412.6 cpm vs. 3669 ± 212.4 cpm, P < 0.0001; Fig. 3C). These results indicate the strong affinity of HP-NPs for necrosis in vitro. Hypericin is believed to have a specific, strong affinity toward necrotic tissue. The mechanism of this phenomenon has not been fully elucidated, although some hypotheses suggest that hypericin binds to specific components in dead cells [44,45]. Several radiolabeled hypericin derivatives, including ¹³¹I-Hyp and ¹²³I-Hyp, can be used as potent necrotizing contrast agents for the noninvasive detection and imaging of various necrosis-related pathologies and diseases, as well as for tumor diagnosis and therapy [46]. In this study, the necrotic affinity and radiotherapy properties of ¹³¹I-Hyp were utilized for the complementary treatment of residual tumors after ablation.

3.7. Cellular uptake of HP-NPs in vitro

HA-modified drug-loaded nanoparticles can interact with CD44 on the surface of tumor cells, achieving active tumor targeting through a receptor-mediated mechanism. To evaluate the ability of HP-NPs to target colon cancer cells, HT-29 and HCT-15 human colon cancer cell lines were selected, and the level of CD44 expression was evaluated by immunofluorescence and flow cytometry. As shown in Fig. 4A, the nucleus and cytoplasm of colorectal cancer cells were stained with DAPI and hypericin, respectively, and the FITC-conjugated CD44 antibody was clearly observed on the surface of HT-29 cells. However, significantly fewer fluorescent signals were detected in HCT-15 cells at the same time. HT-29 cells are believed to have high levels of CD44 receptor expression, whereas HCT-15 cells have low levels of CD44 receptor expression. These expression phenomena were also observed by flow cytometry. As shown in Fig. 4B, the positive quantitative fluorescence level of CD44 on the surface of HT-29 cells was 86.8%, whereas that of HCT-15 cells was only 0.65%. These results were therefore consistent with those of the immunofluorescence experiments.

Fig. 4.

CD44 receptor expression and HA-PCL@(¹³¹I-Hyp) nanoparticles (HP-NPs) binding affinity in vitro. (A) Immunocytochemistry of CD44 expression in HT-29 and HCT-15 cells assessed by fluorescence microscopy with a fluorescein isothiocyanate (FITC) dye. (B) Flow cytometry analysis of CD44 surface expression in HT-29 and HCT-15 colorectal cancer cell lines. Cells were analyzed using unstained isotype controls for gating. CD44-positive cells represented 86.8% of the HT-29 population versus 0.65% of the HCT-15 population. (C) Intracellular distribution of fluorescence from hypericin (Hyp) at 0.5, 1, and 2 h posttreatment in HT-29 cells, HCT-15 cells, and HT-29 cells treated with free hyaluronic acid (HA). Red fluorescence: Hyp; blue fluorescence: Hoechst. (D, E) Flow cytometry (D) and quantitative analysis (E) of the fluorescence intensity of HT-29 and HCT-15 cells incubated with the same concentration of HP-NPs. ^∗P < 0.05, ns: no significant. HA-PCL: hyaluronan-b-poly(ε-caprolactone); DAPI: 4',6-diamidino-2-phenylindole.

The cellular uptake of the HP-NPs was further assessed by fluorescence imaging. HT-29 and HCT-15 cells were selected as human colon cancer cell lines with high and low expression of CD44, respectively. As shown in Fig. 4C, after incubation with HP-NPs, progressively increased red fluorescence was observed over time in the two kinds of human colon cancer cells, indicating that ¹³¹I-Hyp was transported into the cells. Moreover, stronger red fluorescence was always observed in HT-29 cells with high CD44 expression than in HCT-15 cells at the same time points (0.5, 1, and 2 h) after incubation. Notably, the fluorescence intensity of ¹³¹I-Hyp in HT-29 cells markedly decreased when HP-NPs were incubated with excess free HA, indicating that the specific uptake of the nanoparticles in tumor cells was based on a receptor-mediated mechanism between HA and CD44. Hoechst staining and merged fluorescence images revealed that the intracellular distribution of ¹³¹I-Hyp after cellular uptake was mainly located in the cytoplasm. Quantitative determination of the intracellular uptake of the HP-NPs was conducted using flow cytometric analysis. As shown in Figs. 4D and E, the mean fluorescence intensity of both types of colon cancer cells increased with prolonged incubation with the nanoparticles. Furthermore, the mean fluorescence intensity of the HT-29 group treated with HP-NPs was greater than that of the HCT-15 group at each incubation time point, indicating greater uptake of ¹³¹I-Hyp in HT-29 cells with high expression of CD44. CD44, which can specifically bind to HA, is overexpressed on the surface of many human epithelial tumors and has received attention as a target receptor for cancer therapy [47]. The advantages of HA conjugation are that it not only improves the water solubility of hydrophobic drugs such as ¹³¹I-Hyp but also provides targeted delivery of anticancer drugs to CD44-overexpressing tumor cells.

3.8. Animal models and biodistribution of HP-NPs in vivo

To evaluate the biodistribution and therapeutic effects of HP-NPs, nude mice and SRG rats with subcutaneous residual tumors after ablation of metastatic colon cancer were used as animal models (Fig. 5A). The volume of the subcutaneous tumors in the nude mice reached approximately 200 mm^3, as measured via a vernier caliper, at approximately 20 days after the subcutaneous injection of the HT-29 colon carcinoma cell suspension. It took approximately 15 days for the subcutaneous tumors derived from the HT-29 cells to reach a volume of approximately 1000 mm³ in the SRG rats. The MWA of subcutaneous tumors and MRI scanning at 48 h after ablation were performed under isoflurane gas anesthesia. The results of MRI scanning (Fig. 5B) revealed that the T1WI of both animal models could clearly identify the subcutaneous tumor near the right hind limb, with a clear boundary between the tumor and the adjacent muscle tissue. On T2WI, a low-signal area surrounded by a high-signal band, corresponding to the coverage area of the microwave ablation (white dashed circle), was observed in the cross-sectional images of the subcutaneous tumors. The surrounding residual tumor showed a uniform, slightly high signal. Then, the subcutaneous tumors were removed and fixed with 10% formalin, and the establishment of the residual tumor model after ablation was further verified by histopathological examination. On H&E-stained images (Fig. 5B) and sectional photographs of the subcutaneous tumor from nude mice (Fig. 5C), the boundary of the necrotic area after ablation could be clearly identified (the black dashed line separates the necrotic area from the residual tumor tissue). After incomplete MWA, the necrotic areas of the subcutaneous tumors in the two animal models were pink-stained, and the cell structure disappeared. Both imaging and histopathological examination confirmed the successful construction of residual tumor models after incomplete ablation of colon cancer xenografts.

Fig. 5.

Animal models of residual tumors after ablation and biodistribution of HA-PCL@(¹³¹I-Hyp) nanoparticles (HP-NPs) in vivo. (A) Schematic of a subcutaneous residual tumor animal model generated by incomplete microwave ablation. (B) Validation of the subcutaneous residual tumor animal model by magnetic resonance imaging and hematoxylin and eosin (H&E) staining. On T2-weighted imaging (T2WI), the low-signal area surrounded by a high-signal band could be observed in the subcutaneous tumor, which was within the necrotic range after microwave ablation (white dashed circle), and the surrounding residual tumor showed a uniform slightly high signal. On the H&E-stained images, the boundary of the necrotic area after ablation could be clearly recognized (black dashed line). (C) A representative cross-sectional photograph of a tumor from a nude mouse after incomplete ablation. The necrotic area post-ablation is delineated by the black dashed circle. The yellow arrow points to the ablation needle channel. (D) Representative images of in vivo fluorescence for nude mice after tail injection of HP-NPs. The red dashed circles represent the locations of the subcutaneous tumors. (E) Fluorescence intensity of subcutaneous tumors in nude mice at different time points after intravenous injection. (F) Representative single-photon emission computed tomography/computed tomography (CT) images of the Sprague‒Dawley Rag2/Il2rg double knockout (SRG) rat following intravenous injection of HP-NPs. (G) Local uptake values of subcutaneous tumors in SRG rats at different time points after intravenous injection. (H) Representative near-infrared images at different time points after tail vein injection of cyanine-5.5-loaded nanoparticles in the nude mouse model used as a control. (I) Trend of near-infrared fluorescence (NIRF) intensity in the subcutaneous tumors of the nude mice in the control group over time. HA-PCL: hyaluronan-b-poly(ε-caprolactone).

The biodistribution of nanoparticles in vivo is critical for their efficient use in biomedical applications. Unlike other nanoparticles, HP-NPs not only target tumor tissues through the EPR effect and the binding of HA to CD44, but also allow the ¹³¹I-Hyp they carry to remain in the tumor necrosis area for an extended period of time due to its affinity for necrosis. The fluorescence and radionuclide imaging properties of HP-NPs allow the evaluation of their biodistribution by IVIS and SPECT/CT. The distribution of the fluorescent signal emitted by ¹³¹I-Hyp in vivo was monitored by IVIS at the indicated time points after the tail vein injection of HP-NPs into nude mouse models. As shown in Fig. 5D, no obvious concentration of the fluorescent signal was observed at 1 h after injection. The fluorescent signal of ¹³¹I-Hyp was found to accumulate at the subcutaneous tumor site at 4–12 h after intravenous injection and gradually increased in intensity over time. At 24 h after administration, the fluorescence signal began to be mainly retained in the area covered by MWA of the subcutaneous tumor and slowly faded until 216 h. We mapped the subcutaneous tumors as the ROI using the Bruker MI SE software, and measured their fluorescence intensities over time to identify the period of most significant accumulation within the tumors. The results suggested that a pronounced intratumoral peak in the fluorescence signal occurred at 72 h postinjection (Fig. 5E).

Considering the larger size of SRG rats and the stronger penetration ability of γ-rays than the fluorescence signal emitted by ¹³¹I-Hyp, the biodistribution of HP-NPs in vivo was also monitored by SPECT/CT via the emission of γ-rays by ¹³¹I. Fig. 5F shows representative SPECT/CT images of SRG rats bearing partially ablated HT-29 subcutaneous tumors at different time points after the injection of approximately 2 mCi of HP-NPs through the tail vein. The radionuclide results revealed that the radioisotope was primarily distributed in the liver at 1 h after administration. The radionuclide concentration in the liver decreased at 4 h after injection, while aggregation in the thyroid was observed. At 12 h after intravenous injection, the accumulation in the thyroid faded, and the radioactivity of the HP-NPs in the subcutaneous tumor site began to accumulate selectively and was maintained for a long time, with a high level of retention. The residence time of ¹³¹I-Hyp in the subcutaneous tumor was as long as 216 h. The local uptake values of the subcutaneous tumors were highest at 48 h after tail vein injection, as measured by the Xeleris 4.0 workstation (Fig. 5G). In the control group, the NIR dye was concentrated in the non-ablation coverage area of the subcutaneous tumor after intravenous injection of the Cy5.5-loaded nanoparticles; however, its duration was shorter than that of the HP-NPs (Fig. 5H), and the peak NIR fluorescence signal occurred at 12 h after intravenous administration (Fig. 5I).

Using the fluorescence and radionuclide characteristics of ¹³¹I-Hyp, HP-NPs can be monitored via dual-modal imaging. IVIS has the characteristics of high sensitivity and high resolution, but the fluorescence (590–640 nm) emitted by hypericin has limited tissue penetration. The gamma rays emitted by ¹³¹I have strong penetration with high energy, but they are not easy to collimate, which reduces the clarity of the image. The complementary advantages of these dual properties not only allow a comprehensive evaluation of the metabolism and distribution of HP-NPs in vivo but also suggest that the nanoparticles are nanoprobes that can target regions of tumor necrosis and even play a role in therapeutic response monitoring. The in vivo fluorescence imaging results revealed that the HP-NPs circulated throughout the whole body in the early stage after intravenous injection, which was confirmed by the abundant fluorescence signals from organs with high blood perfusion volumes. At 4–12 h after injection, the fluorescence signal of the subcutaneous tumors began to increase and was distributed mainly in the residual tumor tissue around the necrotic area. The EPR effect and active targeting mediated by HA-CD44 may play important roles in this stage. Nanomedicines of 10–100 nm have been reported to accumulate preferentially in tumor tissues through the EPR effect [48]. HP-NPs with a hydrodynamic diameter of 75.66 nm conformed to this particle size range, ensuring their passive cancer-targeting properties. When the nanoparticles passed through the tumor vasculature to the cytoplasmic environment of the post-ablation necrotic area, the organic membrane degraded. The ¹³¹I-Hyp released by the nanoparticles subsequently aggregated in the post-ablation region due to the affinity of hypericin for necrosis. On the fluorescence images at 24 h post-injection, the fluorescence signal began to fill the necrotic area of the subcutaneous tumor and peaked at 72 h post-injection. In contrast, the NIR fluorescence signal from Cy5.5 was concentrated on subcutaneous tumors in nude mice but did not cover the post-ablation necrotic area and regressed rapidly with time. High-intensity fluorescence signals aggregated in subcutaneous tumor regions up to 216 h after intravenous administration of HP-NPs, indicating that ¹³¹I-Hyp had a strong affinity for necrosis and maintained long-term retention, which provides a basis for necrosis-targeted radiotherapy.

Although IVIS can capture high-resolution fluorescence images with its sensitive CCD lens and closed black box device, the penetration of the fluorescence signal is far less than that of gamma rays because of the influence of wavelength and tissue depth. Therefore, we also used SPECT/CT to evaluate the distribution of HP-NPs in vivo. One hour after intravenous injection of the nanoparticles in the SRG rat model, the radionuclide was distributed mainly in the liver, which was consistent with the results of fluorescence imaging of the organs ex vivo. Significantly increased uptake in the thyroid was observed at 4 h after administration, followed by rapid decay at 12 h, indicating mild deiodination of ¹³¹I-Hyp in vivo. The dissociation of radioiodine from the ortho position of the phenolic group is called deiodination. In the radiolabeling and in vitro stability experiments, partial deiodination of ¹³¹I-Hyp was observed after 72 h of incubation in RPMI-1640 medium and serum, but the radiolabeling yield of ¹³¹I-Hyp was still greater than 80%. In addition, the persistently high uptake in subcutaneous tumors from 12 to 192 h after administration and the significantly higher dose of ¹³¹I-Hyp delivered to ablated tumors than to normal organs also suggested that high levels of ¹³¹I were coupled to hypericin and deposited in subcutaneous tumor tissues. Notably, although both IVIS and SPECT/CT revealed the accumulation and long-term retention of HP-NPs in the subcutaneous tumor region, their peak uptake times were not completely consistent. We hypothesized that the following factors may contribute to this phenomenon: (a) The 8.02-day half-life of ¹³¹I led to progressive decay of radioactivity during nanoparticle biodistribution, while hypericin persistently accumulated in necrotic tissues without signal attenuation in IVIS imaging; (b) Slight variations in nanoparticle metabolism may exist between distinct animal models (nude mice vs. SRG rats); (c) Deiodination of ¹³¹I-Hyp during HP-NPs delivery in vivo may contribute to subtle discrepancies between IVIS fluorescence and radionuclide imaging results. Taken together, the above results suggest that HP-NPs injected via the tail vein can selectively aggregate in the subcutaneous tumor area and remain in the necrotic tissue after ablation for a long period of time.

3.9. Analysis of distribution ex vivo and tumor necrosis affinity

Analysis of HP-NPs biodistribution ex vivo was performed by sacrificing and collecting various organs, including subcutaneous tumors, at different time points after a single intravenous injection. Nude mice bearing residual tumors were injected with the nanoparticles through the tail vein and sacrificed at time intervals of 1, 24, 72, and 168 h. Organs and subcutaneous tumors were harvested for fluorescence imaging on IVIS. As shown in Fig. 6A, the fluorescence signal produced by the HP-NPs was distributed mainly in organs with high blood perfusion, including the heart, liver, lung, and kidneys, at 1 h after intravenous injection. At 24 h after injection, the fluorescence signals largely accumulated in the gallbladder, intestine, and subcutaneous tumors. The fluorescence signal intensity in the gallbladder and intestine diminished at 72 h post-injection, whereas a distinct fluorescence accumulation was observed in the subcutaneous tumor, which was confirmed by ROI analysis and fluorescence intensity measurement (Fig. 6B). At 168 h after administration, the fluorescence signal of the HP-NPs was detected only in the subcutaneous tumors but was cleared in the remaining organs. In addition, the biodistribution of the nanoparticles at 48 h after intravenous administration to SRG rat models was determined via TGC (Fig. 6C). The radioactivity distribution revealed the highest accumulation of radioactive in ablated subcutaneous tumors (5.48% ± 1.25%), followed by the liver, spleen, and thyroid, while the lowest radioiodine accumulation was observed in the brain (0.19% ± 0.15%). Dosimetric analysis revealed a significantly elevated radiation exposure in subcutaneous tumors at 48 h post-administration, demonstrating a dose enhancement ratio of 1.92 compared to the liver, which exhibited the highest radiation absorption among normal organs (Table S1). The remarkable tumor specificity of the radiation was attributed to the high necrotic affinity of hypericin under ¹³¹I-labeling. This property enabled long-time retention of ¹³¹I-radiation in the ablated tumor.

Fig. 6.

Biodistribution of HA-PCL@(¹³¹I-Hyp) nanoparticles (HP-NPs) ex vivo and their tumor necrosis affinity. (A) Fluorescence images of harvested organs and subcutaneous tumors at 1, 24, 72, and 168 h post injection. (B) Fluorescence intensity of various excised organs and tumors at 72 h post-injection. ^∗∗∗∗P < 0.0001. (C) Radioiodine distribution in excised organs and subcutaneous tumors of Sprague‒Dawley Rag2/Il2rg double knockout (SRG) rats at 48 h post-intravenous injection of HP-NPs. The data are expressed as the mean % injected dose ± SD/g (%ID/g). ^∗∗∗P < 0.001. (D) The necrotic region boundary (black dashed lines) of the subcutaneous tumors in the nude mice could be clearly identified on the gross section and hematoxylin and eosin (H&E)-stained images, and the corresponding fluorescence images indicated that ¹³¹I-hypericin was deposited mainly in the necrotic area of the tumors. (E) As a control, near-infrared (NIR) fluorescence signals were distributed mainly in residual tumor areas of subcutaneous tumors after intravenous injection of cyanine-5.5 (Cy5.5)-loaded nanoparticles. The black dashed line separates the necrotic area after ablation from the residual tumor. (F) Autoradiography image of subcutaneous tumor sections after intravenous injection of HP-NPs, along with the corresponding H&E staining and fluorescence microscope images. HA-PCL: hyaluronan-b-poly(ε-caprolactone).

The metabolism and excretion of nanomaterials in vivo are closely related to their biosafety and have attracted much attention. Both HA and PCL have good biodegradability, and biodistribution studies of isolated organs have provided clues for exploring the metabolic pathway of HP-NPs. In the nude mouse animal model, dynamic changes in abundant fluorescence signals in the liver, gallbladder, and intestine were observed via ex vivo organ fluorescence imaging at 1, 24, and 72 h after injection. Therefore, we hypothesized that the metabolic pathway by which the ¹³¹I-Hyp released from the nanoparticles may involve hepatic-bile-intestinal metabolism. In a biodistribution study of radioiodinated hypericin as an antitumor agent, Cona et al. [49] reported that Hypericum's main metabolites were principally excreted via the hepatobiliary pathway, which is consistent with our experimental results. The author emphasized that, owing to the hepatobiliary‒intestinal metabolic pathway of ¹³¹I-Hyp, high levels of radioactivity were found in the intestines 24 h after administration, which might result in an increased risk of developing gastrointestinal syndrome at initial exposure. Although no gastrointestinal symptoms were observed in the animals following intravenous administration in this study, clarifying the metabolic pathways of the nanoparticles was crucial to evaluating their biosafety.

To further evaluate the necrotic affinity of HP-NPs in tumor tissues, subcutaneous tumors were sampled and sectioned as needed. As shown in Fig. 6D, the necrotic area after MWA and the surrounding residual tumor (separated by black dashed lines) could be clearly identified from the section of the subcutaneous tumor. In the corresponding fluorescence images, the fluorescence signal of ¹³¹I-Hyp was mainly localized in the necrotic area after ablation. Similar results were also observed in pathological sections of tumor tissue. When the microscopic view field was focused on the boundary of the tumor necrotic area, the red fluorescence of hypericin colocalized with the necrotic area after ablation, confirming that the ¹³¹I-Hyp carried in the nanoparticles mainly targeted the necrotic area of subcutaneous tumors after intravenous injection. In contrast, the area of coagulative necrosis after ablation could also be clearly identified on the section of the subcutaneous tumor in the control group (inside the black dashed circle), but the near-infrared fluorescence of Cy5.5 was distributed mainly in the residual tumor region. The results of fluorescence microscopy confirmed that the dye Cy5.5, which has no necrotic affinity, was deposited mainly in viable tumor tissues (Fig. 6E). Fig. 6F shows the color-coded autoradiogram of subcutaneous tumor sections after intravenous injection of HP-NPs, along with the corresponding H&E staining and fluorescence microscope images. The highly radioactive region seen on the autoradiogram perfectly matched the necrotic area on the H&E-stained section and the red fluorescence of hypericin on the fluorescence microscope. These results indicate that ¹³¹I-Hyp encapsulated by HA-PCL nanoparticles maintained stable conjugation in vivo without significant deiodination and was specifically deposited in the necrotic areas of subcutaneous tumors.

3.10. Necrosis-targeted therapy of residual tumors in vivo

HT-29 tumor-bearing nude mice were divided into the control group, blank micelle group, and HP-NPs group according to different treatments. The initial volume of the subcutaneous tumors in the three groups of animal models was 200.7 ± 17.72 mm³, 204.0 ± 15.04 mm³ and 205.7 ± 15.27 mm³, without significant differences between any two groups (Fig. 7A). The body weights of the nude mice and the volumes of the subcutaneous tumors in each group were monitored for 14 days after intravenous administration. As shown in Fig. 7B, all the nude mice, including those in the HP-NPs group, showed no significant body weight loss during the 14-day postdosing period and no obvious symptoms of toxicity. The body weight change patterns of the nude mice in the three groups were consistent, indicating that the intravenous therapeutic dose of HP-NPs did not cause significant adverse effects in this study. The antitumor efficacy of the nanoparticles was determined by comparing the volume of subcutaneous tumors in each group. The growth curve of the subcutaneous tumors after drug administration revealed that the tumor volume of the nude mice in the control group and the blank micelle group consistently increased. However, the subcutaneous tumor volume of the nude mice in the HP-NPs group remained stable after an intravenous injection. A slow growth tendency appeared on the 8th day after administration (Fig. 7C). The relative tumor proliferation rates in the HA-PCL and HP-NPs groups were 70.23% ± 13.81% and 32.13% ± 9.93%, respectively (Fig. 7D). Compared with that of the control group, the T/C (%) of the HP-NPs group was significantly lower (P = 0.001). In contrast, the T/C (%) of the HA-PCL group was not significantly different from that of the control group (P = 0.1034). Nude mice from the three groups were sacrificed 14 days after intravenous injection, and subcutaneous tumors were harvested and measured. As shown in Figs. 7E and F, the subcutaneous tumor volume of nude mice in the HP-NPs group was significantly smaller than that in the control group and the blank micelle group (control group vs. HP-NPs group = 628.8 ± 244.5 mm³ vs. 183.6 ± 76.98 mm³, P = 0.0009; blank micelle group vs. HP-NPs group = 439.6 ± 125.8 mm³ vs. 183.6 ± 76.98 mm³, P = 0.0420), while there was no significant difference in the volume of subcutaneous tumors between the control group and the blank micelle group (P = 0.1548). These results show that a single intravenous injection of 8 mCi/kg HP-NPs effectively inhibited the growth of residual tumor tissue after ablation of colon cancer metastases without obvious toxic or adverse effects.

Fig. 7.

Analysis of the in vivo therapeutic efficacy of HA-PCL@(¹³¹I-Hyp) nanoparticles (HP-NPs). (A) There was no significant difference in subcutaneous tumor volume among the three groups before treatment. (B, C) Murine body weight (B) and subcutaneous tumor volume (C) were monitored for 14 days post-administration of different treatments. (D) Inhibitory effects of HP-NPs on residual HT-29 colon cancer tumors post-ablation in nude mouse xenograft models at 14 days after intravenous injection via the tail vein. T/C (%) = T_RTV/C_RTV × 100%. RTV = TV_t/TV₀, where TV₀ is measured on day 0 and TV_t is measured on day 14. (E) Photograph of subcutaneous tumors in nude mice from three different treatment groups after 14 days of drug administration. (F) Subcutaneous tumor volume in nude mice across different treatment groups at 14 days post-administration. ^∗P < 0.05, ^∗∗P < 0.01, ^∗∗∗P < 0.001, ns: no significant. HA-PCL: hyaluronan-b-poly(ε-caprolactone).

Furthermore, to explore the mechanism of HP-NPs in treating residual HT-29 tumors after ablation, the sampled subcutaneous tumors were processed according to different requirements for BEM and histopathological analysis. Fig. 8A displays BEM and H&E staining images of necrotic areas post-ablation and residual tumors surrounding the necrotic regions in subcutaneous tumors of nude mice across three groups. Both BEM and H&E staining images revealed coagulative necrosis in the ablated area, the nucleus ruptured and dissolved, and the cytoplasm was granular and disintegrated. H&E staining revealed that the necrotic cells and the disintegrated intercellular stroma had fused into a piece of granular and unstructured homogenous red-stained tissue. The residual tumor cells around the necrotic area in the control group and blank micelle group were characterized by an intact cell membrane and organelles, a high ratio of nucleus to cytoplasm, a uniform chromatin distribution, and a middle nucleolus. The tumor cells around the necrotic region of the HP-NPs group exhibited swelling and ballooning on H&E-stained images and shrinkage of the cell volume in some areas; the cytoplasm was dense, and the nucleus was pyknotic and blue‒black. The BEM images revealed that the residual tumor cells around the necrotic area in the HP-NPs group exhibited severe cell shrinkage, concentrated cytoplasm, and mitochondrial vacuolation. The chromatin of the nucleus was pyknotic, edge-aggregated, and crescent-shaped, indicating morphological changes associated with early apoptosis. In addition, apoptotic bodies were clearly visible in the electron micrographs of the HP-NPs group (Fig. 8B). The results of BEM and histopathological evaluation indicated that the HP-NPs had apparent biological effects on the viable colon cancer cells adjacent to the necrotic area after ablation, possibly resulting from the apoptosis induced by β-rays released from ¹³¹I.

Fig. 8.

HA-PCL@(¹³¹I-Hyp) nanoparticles (HP-NPs) inhibit residual tumors around necrotic areas after ablation by inducing tumor cell apoptosis. (A) Biological electron microscopy (BEM), and hematoxylin and eosin (H&E)-stained images of the post-ablation necrotic area and residual tumor around the ablation zone following different treatments. Dashed line demarcates the ablation zone (left column) from the peri-ablation margin (columns 2–4). (B) Apoptotic bodies (yellow arrow) were observed in the residual tumor tissue around the post-ablation necrotic area in the HP-NPs group. (C) Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining and immunohistochemical staining images of residual tumor tissues from the different treatment groups. (D) Quantitative analysis of TUNEL staining and immunohistochemical staining. ^∗∗∗∗P < 0.0001, ns: no significant. HA-PCL: hyaluronan-b-poly(ε-caprolactone); AOD: average optical density.

To investigate the above possibility, a TUNEL assay and immunohistochemical staining of sections from the three groups of subcutaneous tumors were performed (Fig. 8C). The results of the TUNEL assay revealed that the tumor nuclei of the HP-NPs group presented significant 3,3′-Diaminobenzidine (DAB) staining compared with those of the control group and the blank micelle group, suggesting that the nanoparticles induced massive apoptosis of residual tumor cells. In addition, the expression levels of the proapoptotic factors BAX, caspase-3, and caspase-7 are key proteins for the process of apoptosis. HP-NPs induced the highest level of cleaved caspase-3 and caspase-7 among all the groups, indicating that the nanoparticles could induce residual tumor cells around the necrotic area to undergo apoptosis, which depends on the activation of cysteine proteases and the regulation of BAX gene products. γ-H2A.X, a phosphorylated form of H2A.X, is a variant histone required for checkpoint-mediated cell cycle arrest and DNA repair after double-stranded DNA breaks. DNA damage caused by ionizing radiation can lead to the phosphorylation of H2A.X at the Ser139 site. The damage to double-stranded DNA was confirmed by γ-H2A.X staining, which revealed significantly increased levels of γ-H2A.X foci in the HP-NPs group compared with those in the control group and blank micelle group, suggesting that ¹³¹I irradiation caused DNA damage in viable colon cancer cells. Finally, Ki67 was used to detect the proliferation of residual tumor cells in the three groups. As shown in Fig. 8D, the proliferation rates of residual colon cancer cells in the control group and the blank micelle group were 43.98% ± 1.057% and 43.78% ± 1.894%, respectively, which were significantly greater than those in the HP-NPs group (15.46% ± 1.465%, P < 0.0001). Taken together, the results of the TUNEL assay and immunohistochemical staining demonstrated the inhibitory effect of ¹³¹I-Hyp conferred by HP-NPs on residual colon cancer cells, possibly via the mechanism of tumor cell apoptosis through radiation-induced DNA damage caused by ¹³¹I. In this process, DNA damage caused by ionizing radiation triggers intracellular signal transduction pathways and causes apoptosis. BAX, a proapoptotic protein among the BCL-2 family proteins, is involved in the regulation of this process and may be involved in the alteration of mitochondrial membrane permeability and the release of cytochrome C. Eventually, the apoptotic effector molecules caspase-3 and caspase-7 are activated, triggering the caspase cascade, which induces the apoptosis of residual colon cancer cells.

In vitro studies as well as the biodistribution of nanoparticles in vivo confirmed that HP-NPs could target tumors after intravenous injection via the EPR effect and the specific binding of HA to CD44 on the surface of tumor cells. After the HP-NPs reached the necrotic area of the tumor via blood flow, the nanoparticles degraded in the cytoplasmic environment and released ¹³¹I-Hyp, which has a specific affinity for necrotic tissue. Moreover, the beta rays emitted from ¹³¹I-Hyp deposited in the necrotic region penetrated and irradiated the surrounding residual tumor tissue. In this study, a single intravenous injection of 8 mCi/kg HP-NPs significantly inhibited subcutaneous residual tumor tissues after partial ablation in nude mice. Furthermore, BEM and histopathological analysis revealed that β-rays emitted by ¹³¹I activated apoptosis by damaging the DNA of residual cancer cells around the necrotic area, resulting in tumor suppression. Limited by the size of animals, we did not prepare a model of hepatic metastases from colon cancer, which is a shortcoming of this study that we will address in future research. On the basis of the positive results obtained in our two animal models using colon cancer cell lines, we believe that there will be corresponding treatment efficacy in hepatic metastases.

3.11. Safety evaluation of the HP-NPs in vivo

The biosafety of the HP-NPs in vivo was evaluated by histopathological studies and blood analysis. According to previous studies, ¹³¹I-Hyp at a dose of 300 MBq/kg has achieved significant efficacy in a combretastatin A-4 phosphate-induced tumor necrosis model [7]. Therefore, we evaluated the biosafety of HP-NPs at a therapeutic dose of 8 mCi/kg (296 MBq/kg) in a nude mouse model. Nude mice in the control group and HP-NPs group were sacrificed after 14 days of intravenous administration, and blood samples and major organs, including the heart, liver, spleen, lung, kidney, brain, thyroid, and intestine, were collected. The tissue samples were fixed with 10% neutral formalin, and paraffin sections were prepared for H&E staining. No significant differences in organ damage or inflammation were observed in the tissue sections between the two groups (Fig. 9A). In addition, blood samples from both groups of animal models were tested for hematology and blood biochemistry. In the control group and HP-NPs group, no significant changes were observed in any of the key blood components, and all the values were within the normal reference range (Fig. 9B). These results suggest that intravenous injection of 8 mCi/kg HP-NPs resulted in good biosafety, which provides a basis for clinical translation. However, the multi-pathway clearance mechanisms of nanoparticles complicate the prediction of their long-term toxicity profiles, which should be focused on by future exploration [50].

Fig. 9.

Safety evaluation of the HA-PCL@(¹³¹I-Hyp) nanoparticles (HP-NPs) in vivo. (A) Hematoxylin and eosin (H&E) staining images of major organs from nude mice treated with HP-NPs at a dose of 8 mCi/kg and age-matched healthy controls. (B) Blood biochemistry and routine blood test results of nude mice treated with HP-NPs and healthy controls. HA-PCL: hyaluronan-b-poly(ε-caprolactone); WBC: white blood cell; RBC: red blood cell; PLT: platelet; HGB: hemoglobin; ALT: alanine aminotransferase; AST: aspartic transaminase; TBIL: total bilirubin; Cr: creatinine.

4. Conclusion

The HA-PCL conjugate formed stable nanoparticles (HP-NPs) consisting of a hydrophilic shell and encapsulating ¹³¹I-Hyp in the hydrophobic inner core, which improved the hydrophobicity of ¹³¹I-Hyp. The nanoparticles had a hydrodynamic diameter of less than 100 nm, high stability, good blood compatibility, and rapidly released ¹³¹I-Hyp in the presence of a hyaluronidase-rich or cytoplasmic environment after cell necrosis. The released ¹³¹I-Hyp eliminated fluorescence dequenching and exhibited intrinsic necrosis avidity in vitro and in vivo. The tumor-targeted delivery of HP-NPs occurred via both passive accumulation via the EPR effect and active targeting by the strong receptor-binding affinity of HA for CD44 with high specificity. In animal models of residual colon cancer after ablation, a single intravenous injection of HP-NPs effectively inhibited tumor growth via cell apoptosis induced by β-rays emitted from ¹³¹I. In conclusion, HP-NPs could be successfully applied for targeted necrosis therapy of residual tumors after ablation of colorectal hepatic metastases.

CRediT authorship contribution statement

Han Bao: Writing – original draft, Visualization, Investigation, Conceptualization. Ning Wang: Resources, Investigation. Xiaowen Zhu: Investigation. Song Chen: Validation, Methodology. Yang Wang: Resources. Xiangjun Han: Writing – review & editing. Hongshan Zhong: Writing – review & editing, Funding acquisition, Conceptualization.

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 A. Supplementary data

The following is the Supplementary data to this article: