A Radiopaque and Thermoresponsive Nanogel-Based Pickering Emulsion for Permanent Arterial Embolization

Abstract

The development of liquid embolic agents has significantly advanced the field of transcatheter arterial embolization (TAE). Conventional solid embolic agents, such as gelatin sponges or poly­(vinyl alcohol) (PVA), are limited by their lack of intrinsic radiopacity and reliance on lipiodol, necessitating a cumbersome two-step embolization process. Consequently, a critical need exists for embolic materials that combine intrinsic imaging functionality with enhanced biocompatibility to advance clinical outcomes. We have developed a novel liquid embolic material based on poly­(N-isopropylacrylamide)-co-acrylic acid nanogels (PNAs). Mixing PNAs with lipiodol using a medical three-way stopcock yields a thermosensitive lipiodol-based gel emulsion (TLGE). This material exhibits excellent temperature sensitivity and biocompatibility. Notably, the TLGE demonstrates injectable fluidity and sufficient X-ray contrast in an ex vivo decellularized liver model. This embolization technique requires only a single injection and allows for rapid mixing, offering significant practical advantages in the clinic. Digital subtraction angiography (DSA) facilitates intraoperative imaging and real-time monitoring of the embolization process, aiding in the prevention of nontarget embolization and ensuring precise treatment. In a preclinical study, following the embolization of the right renal artery in New Zealand rabbits, a 42-day observation period revealed no instances of vessel recanalization or damage to nontarget organs. Consequently, this innovative temperature-sensitive material demonstrates significant potential for clinical application in TAE procedures.

1. Introduction

Embolization is a therapeutic intervention that involves the intentional occlusion of blood vessels through the catheter-based delivery of embolic agents. − Clinically available embolic agents are primarily divided into two categories: solid and liquid forms. Solid embolic agents, such as coils, poly­(vinyl alcohol) (PVA) microspheres, and chitosan microspheres, act primarily through mechanical blockage. − In contrast, liquid embolic agents, including lipiodol, anhydrous ethanol, and butyl cyanoacrylate glues, are pivotal for their ability to penetrate distal vasculature inaccessible to solid agents and achieve complete vascular casting. − This property makes them particularly valuable in transcatheter arterial chemoembolization (TACE) for hepatocellular carcinoma, where they effectively induce tumor necrosis by obstructing blood supply. − As a minimally invasive approach, TACE is generally well-tolerated by patients. However, conventional liquid embolic agents, particularly lipiodol, are fraught with limitations. A major issue is the lack of robust and durable occlusion, which can lead to vascular recanalization and tumor recurrence. Specific drawbacks of lipiodol include its transient embolic effect, high fluidity predisposing to reflux and nontarget embolization (e.g., pulmonary embolism), risks of iodine-related allergy or thyroid dysfunction, and a steep technical learning curve for safe administration. The overarching challenges for liquid embolic agents can be summarized as follows: a propensity for migration causing ectopic embolism, a high risk of reflux, difficult intraoperative control, insufficient embolic strength leading to recanalization, and concerns regarding the biocompatibility and potential toxicity of some materials. − Although current liquid agents are adequate for numerous clinical scenarios, these persistent shortcomings underscore a critical need for the development of next-generation liquid embolic materials. Such innovations must address the limitations in durability, biocompatibility, and controlled vascular casting to advance therapeutic outcomes.

Based on its favorable water solubility and the ability to produce consistent and controllable radical generation, 2,2′-azobis­(2-methylpropionamidine) dihydrochloride was selected as the initiator. Using this initiator in conjunction with an ionic emulsifier, we synthesized poly­(N-isopropylacrylamide)-co-acrylic acid nanogels (PNAs) via aqueous free-radical polymerization. As reported in our previous study, the resulting nanogel dispersion exhibited a distinct thermosensitive sol–gel transition alongside remarkable thermodynamic stability. The system demonstrated high yield stress in the gel state while retaining excellent fluidity and pronounced shear-thinning behavior in the sol state. The ability to precisely trigger gelation at a physiologically relevant temperature of 33 °C through simple thermal regulation significantly enhances its potential for clinical application.

Lipiodol, an iodinated oil widely used as a radiopaque contrast agent in computed tomography (CT), is suboptimal as a stand-alone embolic material due to its tendency to cause vascular recanalization, leading to inadequate long-term embolic efficacy. To overcome this limitation, lipiodol is frequently administered in combination with solid embolic agents such as gelatin sponges or poly­(vinyl alcohol) (PVA). Nevertheless, these solid agents themselves suffer from poor radiopacity and a risk of catheter occlusion during delivery. Moreover, the low viscosity and high mobility of lipiodol, coupled with the technical challenges of superselective catheterization, substantially increase the risk of nontarget embolization. To address these challenges, we developed a thermosensitive lipiodol gel emulsion (TLGE) by leveraging the temperature-responsive properties of our synthesized PNAs and the inherent radiopacity of lipiodol. This hybrid material offers significant improvements in embolic efficacy, biocompatibility, and procedural simplicity. For clinical preparation, PNAs and lipiodol were drawn into separate syringes, which were then connected via a medical three-way stopcock. Simply alternating injection through the stopcock allowed for rapid and efficient mixing to form the TLGE, requiring no specialized mixing instruments or prolonged waiting, thereby greatly facilitating its clinical adoption.

A decellularized liver model was established to evaluate the embolization performance and radiopacity of the TLGE. Decellularization, which employs a combination of physical and chemical treatments to remove all cellular components while preserving the native extracellular matrix (ECM) and its structural proteins, − was utilized to create a translucent liver bioscaffold. This model retains the integrity of the original vascular architecture and mechanical strength, providing a physiologically relevant platform for preliminary in vitro testing. − Injection of the TLGE into this model allowed for direct visualization of its flow behavior under X-ray imaging, confirming both its excellent radiopacity and ease of delivery. The efficacy of the TLGE was further validated in a rabbit renal artery embolization model. The material achieved complete and persistent embolization of the main renal artery. Notably, follow-up over 42 days confirmed the absence of vascular recanalization at the target site. These results highlight the dual advantage of this thermosensitive nanogel-based material: it provides clear intraoperative visibility due to its X-ray contrast and ensures durable vessel occlusion, demonstrating significant potential for clinical translation in transarterial embolization (TAE) procedures (Scheme ).

1. Diagram of Thermosensitive Lipiodol Gel Emulsion with Long-Acting Arterial Embolization and Long-Term X-ray Imaging Ability.

2. Materials and Methods

2.1. Materials and Animals

N-Isopropylacrylamide (NIPAM) and acrylic acid (AA) were procured from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Poppy-seed oil-based lipiodol was provided by Jiangsu Hengrui Pharmaceutical Co., Ltd. Medical three-way stopcocks were supplied by Yangzhou Yangsheng Pharmaceutical Technology Co., Ltd. The following reagents were purchased from Aladdin Chemical Reagent Co., Ltd.: N,N′-methylenebis­(acrylamide) (MBA), stearyl trimethylammonium chloride (STAC), 2,2′-azobis­(2-methylpropionamidine) dihydrochloride (AAPH), and rhodamine B. Sodium dodecyl sulfate (SDS) and ethylene glycol-bis­(2-aminoethyl ether)-N,N,N′,N′-tetraacetic acid (EGTA) were acquired from Aladdin Biochemical Technology Co., Ltd. Additionally, 4% paraformaldehyde and pentobarbital sodium were obtained from McLean Reagent Co., Ltd. Lidocaine hydrochloride and gentamicin sulfate were supplied by Huabei Pharmaceutical Co., Ltd. Heparin sodium injection was sourced from Changzhou Qianhong Biochemical Pharmaceutical Co., Ltd. Sodium chloride injection was provided by Wuhan Binhu Shuanghe Pharmaceutical Co., Ltd. A QIAGEN kit was used for DNA extraction. All aqueous solutions were prepared using Milli-Q ultrapure water (18.2 MΩ·cm).

Adult male New Zealand white rabbits (weighing 2.5–3.0 kg) and Sprague–Dawley rats (male, 300–400 g) were supplied by the Laboratory Animal Center of Hubei University of Science and Technology. All experimental procedures were conducted in strict compliance with the Guide for the Care and Use of Laboratory Animals and were approved by the Animal Ethics Committee of Hubei University of Science and Technology (Approval No.: 2024-02-002).

2.2. Synthesis of PNAs

The nanogel was synthesized via emulsion polymerization with the following optimized formulation. Briefly, MBA (0.03 g, 0.19 mmol), STAC (0.05 g, 0.29 mmol), NIPAM (1.7 g, 15 mmol), and AAPH (0.08 mg, 0.14 mmol) were dissolved in 150 mL of ultrapure water. The solution was transferred to a three-necked flask and stirred at 80 °C (250 rpm) in an oil bath until a blue color appeared, indicating radical initiation. At this point, AA (215 μL, 3 mmol) was added to the reaction mixture. The reaction was allowed to proceed for 4 h after the solution turned milky white. The resulting product was dialyzed against flowing water using a dialysis membrane (MWCO: 14 kDa) for 3 days. The purified sample was then prefrozen at −80 °C for 4 h and subsequently lyophilized for 2–3 days to obtain a dry powder for further use.

2.3. Preparation of TLGE

TLGE was prepared by mixing 6 or 8 wt % PNA nanogels with lipiodol at varying volume ratios. Specifically, lipiodol volumes of 0.1, 0.3, 0.5, 0.7, and 0.9 mL were combined with corresponding volumes of 6 or 8 wt % PNA solution to achieve a total volume of 1 mL, corresponding to lipiodol concentration ratios of 10%, 30%, 50%, 70%, and 90% (v/v). The mixtures were loaded into a 1 mL syringe connected to a medical three-way stopcock and alternately pushed through the connector for 100 cycles to form a homogeneous TLGE. The resulting emulsion was subsequently transferred to a 1 mL glass vial for storage. Notably, blends containing 70% and 90% lipiodol failed to form stable gels due to excessive oil content and were therefore excluded from further experiments. Consequently, only the 10%, 30%, and 50% lipiodol formulations prepared with both 6 and 8 wt % PNAs were selected for subsequent comparative studies.

2.4. Characterization of the PNAs

The lower critical solution temperature (LCST) of the PNAs was determined by monitoring the optical transmittance at 500 nm. A 1 wt % aqueous dispersion of PNAs was heated from 25 to 40 °C at a constant rate of 1 °C/min, and the transmittance was recorded using a UV–vis spectrophotometer (Beijing General Analytical Instruments, China). The morphology of the PNA nanogels was examined by transmission electron microscopy (TEM; Tecnai G2 20, FEI Corp, Netherlands) operating at an acceleration voltage of 200 kV. The nanogel dispersion was diluted with ultrapure water to a concentration of 1 mg/mL and ultrasonicated for 10 min. Subsequently, a 10 μL aliquot of the sample was deposited onto a carbon-coated copper grid (300 mesh) and allowed to dry at room temperature. The sample was then negatively stained with 0.1 wt % phosphotungstic acid solution prior to imaging.

The hydrodynamic diameter and zeta potential of the nanogels were characterized using dynamic light scattering (DLS). A 0.2 wt % nanogel dispersion in ultrapure water was prepared, and 2 mL of the sample was transferred to a disposable sizing cuvette. Measurements were performed at 25 and 37 °C. A He–Ne laser (λ = 633 nm) served as the light source, and the scattered light was collected at a 90° angle. Each sample was equilibrated for 200 s prior to measurement, and the reported size values represent the average of four consecutive runs. For zeta potential analysis, 2 mL of the same dispersion was carefully loaded into an electrophoretic cell, ensuring the removal of air bubbles. The measurement was then conducted using the same DLS instrument equipped with a zeta potential module.

2.5. Characterization of the TLGE Emulsion

The stability of TLGE formulations was evaluated by combining lipiodol at concentrations of 10%, 30%, and 50% with 6 or 8 wt % PNA dispersions. For each group, 1 mL of the resulting emulsion was transferred into a transparent glass vial and stored at room temperature (25 °C). The samples were visually inspected and photographed at predetermined time intervals (1, 2, and 3 h; 1, 3, and 10 days) to assess phase separation and macroscopic homogeneity. This method allowed for the systematic comparison of formulation stability and the identification of the optimal composition.

After mixing lipiodol (10%, 30%, and 50%) with 6 wt % PNAs, the resulting emulsions were transferred into 2 mL glass vials. The microstructure of each emulsion was then examined using fluorescence microscopy at room temperature (25 °C). For the in vitro injection experiment, PNA nanogels were first stained with Rhodamine B via incubation (aqueous nanogel dispersion at 1 mg/mL with Rhodamine B at 0.01 mg/mL, 10:1 v/v) for 4 h at room temperature with gentle agitation. Following dialysis (MWCO: 12 kDa) against deionized water for 24 h to remove unbound dye, the purified nanogels were mixed with 50% lipiodol to form a TLGE. The resulting emulsion was injected through various clinical catheters into a transparent vascular model maintained at 25 and 37 °C. The injection process and vascular casting patterns were recorded in real-time using a high-speed camera, and the emulsion’s morphological changes at the two temperatures were analyzed.

The radiopacity of TLGE was quantitatively assessed using computed tomography (CT, Somatom Definition AS 128 S, Erlangen, Germany), and the corresponding Hounsfield unit values were recorded. Rheological characterization of both TLGE and PNAs was performed using a rotational rheometer (Anton Paar MCR92, Austria). Measurements were conducted with a parallel plate geometry (PP50, diameter = 40 mm) at a fixed gap of 0.5 mm. For temperature sweep tests, the storage modulus (G′) and loss modulus (G″) were measured from 25 to 40 °C at an oscillatory stress of 0.5 Pa, a frequency of 1.0 Hz, and a heating rate of 2.0 °C/min. Additionally, steady-state viscosity was evaluated at 25 °C under shear rates ranging from 150 s^–1 to 2580 s^–1.

2.6. Evaluation of the Ex Vivo Live Model

Twenty male Sprague–Dawley (SD) rats (8–10 weeks old, 300–400 g, specific pathogen-free grade) were housed under standard conditions with a 12 h light/dark cycle, ambient temperature of 22–25 °C, relative humidity of 50–60%, and ad libitum access to water and food. An ex vivo decellularized liver model was employed to evaluate the injectability and radiopacity of the TLGE. Livers were harvested from male SD rats euthanized by carbon dioxide asphyxiation. Each liver was perfused via the portal vein with 20 mL of phosphate-buffered saline (PBS), excised, and immediately frozen at −80 °C for 24 h. After thawing at room temperature, the livers were decellularized by continuous perfusion through the portal vein using a syringe pump (Mindray BeneFusion VP3) at a rate of 120 drops/min with the following sequence: normal saline (0.9% NaCl), 0.25% EGTA for 3 h, and 1% sodium dodecyl sulfate (SDS) for 6 h. Both native and decellularized livers were fixed in 4% paraformaldehyde for 24 h, followed by dehydration through a graded ethanol series, clearance in xylene, and embedding in paraffin for histological preparation.

Following decellularization, tissue sections were prepared at a thickness of 4 μm. Histological analysis was performed using hematoxylin and eosin (H&E) staining to compare nuclear presence and tissue architecture between normal and acellular liver specimens. Immunofluorescence evaluation was conducted on both tissue types using the following primary antibodies: rabbit polyclonal anticollagen I (Abcam, ab254349), antilaminin (Abcam, ab11575), and antifibronectin (Servicebio, GB13091). For scanning electron microscopy (SEM), normal and decellularized liver tissues were fixed in 2.5% glutaraldehyde for 24 h, dehydrated in a graded ethanol series, frozen at −80 °C for 2 h, and freeze-dried for 24 h. Samples were then sputter-coated with a 10 nm platinum layer and imaged using an SU8100 SEM at an accelerating voltage of 3 kV.

Radiopacity was assessed using a digital medical X-ray radiography system (DP528-B) under the following parameters: 55 kV, 320.0 mA, 12.00 ms exposure time, EI 57338, and dose 161.86 μGy. The decellularized liver was first imaged alone, then after injection of lipiodol via the portal vein. Macroscopic vessel distribution was monitored using an SZ680 stereomicroscope, with adjustments to focal length and field position as needed. For dynamic embolization evaluation, TLGE was injected intravenously through the portal vein at 0.5 mL/s, and vascular filling was observed at 1–5 s intervals. Genomic DNA content was quantified and compared between decellularized and normal liver tissues from six rats.

Following 10 h of perfusion, both decellularized and normal liver tissues were lyophilized using a vacuum freeze-dryer. Genomic DNA was extracted from all liver specimens according to the manufacturer’s protocol, and its concentration and purity were determined using a Nanodrop 2000 spectrophotometer. For agarose gel electrophoresis, 0.75 g of agarose was weighed into a conical flask containing 50 mL of buffer solution, then heated in a microwave until completely dissolved to form a homogeneous 1.5% gel. Each DNA sample (3–5 μL) was mixed with an equal volume of loading buffer, and 10 μL of the mixture was loaded into the gel wells using a micropipette. Electrophoresis was conducted immediately at 170 V, during which DNA migrated from the cathode (black) toward the anode (red). After electrophoresis, the gel was carefully removed and visualized under ultraviolet illumination. Distinct fluorescent bands indicated the presence of DNA, and images were captured and archived using a gel documentation system.

2.7. Application Value of TLGE in Renal Artery Embolization in Normal Rabbits

Embolization procedures were performed on the right kidneys of healthy New Zealand white rabbits. Twelve adult male rabbits weighing 3.0–3.5 kg were selected for this study. The animals were housed under controlled environmental conditions: ambient temperature maintained at 25–27 °C, relative humidity 50–60%, ventilation rate 10–15 air changes per hour, airflow velocity 0.1 m/s, colony density ≤30 animals per cage, noise levels below 50 dB, and a 12:12 h light/dark cycle. The rabbits received standard feeding three times daily.

Embolization of the right kidney was performed in healthy New Zealand white rabbits. After 1 week of acclimatization and a 12-h fasting period, the rabbits were randomly allocated into three experimental groups (n = 3 per group) for a 42-day study. Anesthesia was induced by intravenous administration of pentobarbital sodium (30 mg/kg) via the marginal ear vein. Fifteen minutes after anesthesia induction, the rabbits were fixed in a supine position. The groin area was shaved and disinfected, followed by local infiltration anesthesia with lidocaine. The femoral artery was isolated using ophthalmic forceps, and an 18G needle was inserted proximally to introduce a 4F coaxial microcatheter (Terumo, Tokyo, Japan). Baseline renal arteriography was performed using iodixanol (300 mgI/mL) injected at 0.5 mL/s to confirm vascular patency.

The respective embolic agents (1 mL lipiodol, 1 mL PVA, or 1 mL TLGE) were slowly injected into the renal artery via the microcatheter. Five minutes postembolization, 2 mL of iodixanol was administered to evaluate the immediate embolic effect. Upon completion of the procedure, the catheter was withdrawn under aseptic conditions, the incision was sutured, and antibiotic prophylaxis was initiated with intramuscular penicillin sodium (10,000–20,000 IU) for three consecutive days. Computed tomography (SOMATOM Sensation 64-slice, Siemens, Germany) was performed on days 7, 21, and 42 postembolization to assess the embolization outcomes. The scanning protocol included nonenhanced localization followed by contrast-enhanced phases (arterial phase at 15 s, venous phase at 25 s, delayed phase at 45 s) after injection of iodixanol (3 mL) and saline flush (10 mL) at 0.5 mL/s. Scan parameters were set at 80 kV and 60 mA. The CT images were evaluated for vascular occlusion completeness, presence of nontarget embolization, and morphological changes in the contralateral kidney.

After the final CT scan, the rabbits were euthanized, and tissue samples from heart, liver, spleen, lungs, and kidneys were harvested and fixed in 4% paraformaldehyde for 24 h. The specimens were processed through standard paraffin embedding, sectioned at 4 μm, and stained with hematoxylin and eosin (H&E) for histological evaluation. Both left and right kidneys were sectioned longitudinally to compare morphological and pathological changes following embolization.

2.8. Evaluation of Biocompatibility

A 2% red blood cell (RBC) suspension was prepared by mixing whole blood (2 mL) collected from New Zealand rabbits with normal saline, followed by centrifugation at 1500 rpm for 10 min. The supernatant was discarded, and the RBC pellet was resuspended in normal saline to obtain the desired 2% RBC suspension. The experimental group consisted of TLGE emulsion containing 6 wt % PNAs and 50% lipiodol, which was serially diluted to concentrations of 5 mg/mL, 2.5 mg/mL, 1.25 mg/mL, and 0.625 mg/mL using normal saline. Normal saline and ultrapure water were used as the negative and positive controls, respectively.

Five replicate hemocompatibility tests were performed for each experimental group. In each test, 300 μL of the respective test solution was combined with 300 μL of the 2% red blood cell suspension in a 1.5 mL microcentrifuge tube. Following incubation at 37 °C for 1 h, the mixtures were centrifuged at 3000 rpm for 25 min. Subsequently, 100 μL of supernatant from each sample was carefully transferred to a 96-well plate. The hemolysis rate (Hr %) was determined by measuring the hemoglobin absorbance at 540 nm using a microplate reader (1420 MultiLabel Counter, PerkinElmer, USA).

$$r\%=\frac{{\mathrm{O}\mathrm{D}}_{\mathrm{s}}-{\mathrm{O}\mathrm{D}}_{\mathrm{n}}}{{\mathrm{O}\mathrm{D}}_{\mathrm{n}}-{\mathrm{O}\mathrm{D}}_{\mathrm{b}}}\times 100\%$$

OD_s represent the absorbance of the experimental group, OD_n represents the absorbance of the negative control group, and OD_b represents the absorbance of the blank group.

The MTT assay was performed to evaluate the cytotoxicity of the materials on HepG2 cells. When the cells reached the logarithmic growth phase, they were harvested by trypsinization and resuspended in DMEM at a density of 1 × 10⁵ cells/mL. Subsequently, the cells were seeded into 96-well plates at 4–5 × 10³ cells per well and incubated for 24 h to allow attachment. TLGE emulsions (6 wt % PNAs with 50% lipiodol) and PNAs alone were diluted in DMEM to concentrations of 0.4, 0.2, 0.1, 0.05, and 0.025 mg/mL. After removing the original medium, 100 μL of each sample solution was added to the corresponding wells and incubated with the HepG2 cells for 24 h. Cells cultured in DMEM alone served as the negative control, while DMEM without cells was used as the blank.

An MTT solution was prepared by dissolving 50 mg of MTT powder in 10 mL of PBS to a final concentration of 5 mg/mL. Then, 20 μL of this solution was added to each well, and the plates were incubated in the dark for 4 h. After carefully removing the supernatant, 150 μL of DMSO was added to each well to dissolve the formazan crystals. The plates were shaken at a constant temperature for 10 min, and the absorbance was measured at 570 nm using a microplate reader. The cell viability was calculated according to the following formula

$$\mathrm{C}\mathrm{v}\%=\frac{{\mathrm{O}\mathrm{D}}_{\mathrm{s}}-{\mathrm{O}\mathrm{D}}_{\mathrm{b}}}{{\mathrm{O}\mathrm{D}}_{\mathrm{n}}-{\mathrm{O}\mathrm{D}}_{\mathrm{b}}}\times 100\%$$

OD_s, OD_b, and OD_n represent the absorbance values of the sample group, blank group, and negative control group, respectively.

3. Results and Discussion

3.1. Structure and Morphology Characterization of the PNA Nanogels

The FT-IR spectra of acrylic acid (AA) and the resulting PNA nanogel were compared to confirm the polymerization. The spectrum of AA exhibits characteristic vibrations of an unsaturated carboxylic acid: the CO stretch of the carboxylic acid at 1701 cm^–1, the CC stretch at 1636 cm^–1, and the unsaturated C=H stretch at 3066 cm^–1. In contrast, the spectrum of the PNA nanogel shows the disappearance of these diagnostic peaksspecifically the CC and unsaturated C–H stretchesalong with the emergence of new bands. These include an amide-related peak at 1550 cm^–1, saturated C–H stretches at 2972 and 2934 cm^–1, and a shifted carbonyl stretch at 1705 cm^–1, indicative of a different chemical environment. The absence of the original acrylic acid CO peak at 1701 cm^–1 confirms the consumption of the monomer and the successful formation of a new polymeric structure with amide/amino functionalities (Figure S1).

The particle size and zeta potential of the PNAs were measured using dynamic light scattering (DLS) with a 0.2 wt % aqueous dispersion. The zeta potential values were −4.2 mV at 25 °C and −9.0 mV at 37 °C (Figure a). The negative zeta potential of the nanogels originates from the ionized carboxyl groups (−COOH) derived from acrylic acid, which impart a negative surface charge. The observed increase in the absolute zeta potential value with temperature is consistent with the thermally induced chain collapse during the phase transition. As shown in Figure b, the hydrodynamic diameter decreased significantly from 102 nm (PDI = 0.279) at 25 °C to 50.7 nm (PDI = 0.332) at 37 °C, confirming the thermosensitivity of the material. This collapse leads to a more compact particle structure, thereby increasing the surface charge density and enhancing electrostatic stabilization against aggregation.

1.

Temperature sensitivity characterization of the PNAs. (a) Zeta potential of PNAs at 25 and 37 °C. (b) Particle size distribution of PNAs at 25 and 37 °C. (c) Changes in transmission of PNAs at 500 nm in the range of 25 to 40 °C. (d) TEM images of the PNAs at 25 °C.

The optical transmittance of the PNA dispersion was monitored from 25 to 40 °C using UV–Vis spectroscopy. The transmittance decreased progressively with increasing temperature, with a 50% reduction occurring at approximately 28 °C, and stabilized near 0.4% above 37 °C (Figure c), indicating a complete phase transition. Transmission electron microscopy (TEM) images revealed that the PNAs exhibit a spherical morphology with uniform size distribution and smooth surfaces (Figure d), consistent with the DLS observations.

3.2. Structure and Morphology Characterization of the TLGE

Based on preliminary screening, 2 and 4 wt % PNAs were excluded due to inadequate gelation and incomplete coagulation at 37 °C. In contrast, 6 and 8 wt % PNAs achieved complete gelation at 33 °C (Figure S2) and were therefore selected for further study. These were blended with different volumes of lipiodol (0.1, 0.3, 0.5, 0.7, and 0.9 mL) to a total volume of 1 mL, corresponding to lipiodol concentrations of 10%, 30%, 50%, 70%, and 90%, respectively (Table S1). Mixtures containing 70% and 90% lipiodol failed to form stable gels at 37 °C and were subsequently excluded from further evaluation (Figures S3 and S4). The stability of emulsions prepared with 6 and 8 wt % PNAs combined with 10%, 30%, and 50% lipiodol was assessed over time at 25 °C (1 h, 2 h, 3 h, 1 day, 3 days, and 10 days; Figure S5). Emulsions containing 10% and 30% lipiodol showed slight phase separation within 1 h and evident delamination after 3 h, indicating poor stability. In contrast, the 50% lipiodol formulation exhibited only minimal stratification by day 10, suggesting superior emulsion stability (Figure a).

2.

Characterization of TLGE stability in vitro. (a) When the lipiodol content was 50%, TLGE showed long-term stability at room temperature 25 °C for 1 h, 2 h, 3 h, 1 day, 3 days and 10 days. (b) CT values of 6 wt % PNAs mixed with different contents of lipiodol. (c) Image of different oil/water ratio under fluorescence microscope at 6 wt % PNAs concentration (original magnification, ×100, ×200 and ×400). (d) Particle size of different oil/water ratio at 6 wt % PNAs concentration. ***p < 0.001.

Due to the high viscosity and catheter adhesion tendency of the 8 wt % TLGE, which hindered injectability, it was deemed unsuitable for embolization applications. Therefore, the 6 wt % TLGE with 50% lipiodol was selected for subsequent rabbit renal artery embolization experiments. The radiopacity of the materials was evaluated using digital radiography (DR) and computed tomography (CT) (Figure S6). For 6 wt % PNAs blended with 10%, 30%, and 50% lipiodol, CT values were 2328.6 ± 31.26 HU, 3063.31 ± 11.42 HU, and 3067.73 ± 7.41 HU, respectively (Figure b). Similarly, 8 wt % PNAs mixed with the same lipiodol proportions yielded CT values of 2020.67 ± 21.56 HU, 3068.65 ± 1.38 HU, and 3068.96 ± 1.08 HU (Figure S7). These results confirm that all tested formulations provide sufficient radiopacity for in vivo imaging and embolization monitoring.

Fluorescence microscopy was performed to examine the morphological characteristics of emulsions prepared by mixing 6 wt % PNAs with 10%, 30%, and 50% lipiodol. The images revealed a progressive increase in emulsion droplet size with rising lipiodol content (Figure c). Notably, the emulsion containing 50% lipiodol displayed a uniform size distribution and well-defined spherical morphology, suggesting enhanced stability compared to the 10% and 30% formulations. Quantitative analysis of droplet sizes confirmed these observations, with values of 112.8 ± 9.7 μm, 122.4 ± 27.4 μm, and 123.2 ± 7.8 μm for the 10%, 30%, and 50% lipiodol mixtures, respectively (Figure d). The reduced polydispersity in the 50% lipiodol group further supports its improved emulsion stability.

3.3. Temperature Sensitivity of the PNCAA Nanogel

The thermoresponsive sol–gel transition behavior of PNAs and TLGE was characterized using a rotational rheometer. As the temperature approached approximately 33 °C, both the storage modulus (G′) and loss modulus (G″) of TLGE and PNAs increased sharply, with G′ surpassing G″ beyond this point (Figure a,b), indicating a sol-to-gel transition and confirming a critical gelation temperature around 33 °C. Above this temperature, G′ continued to rise and remained higher than G″, demonstrating the formation of a solid-like gel network. The rheological behavior under shear was further investigated by measuring viscosity as a function of shear rate. When the shear rate increased from 105 s^–1 to 2580 s^–1, the viscosity of PNAs decreased from 14 mPa·s to 9.8 mPa·s (Figure c), while that of TLGE dropped markedly from 240 mPa·s to 28 mPa·s (Figure d). The pronounced decrease in viscosityparticularly for TLGE, which exhibited an approximately 8-fold reductionconfirms its strong shear-thinning behavior, a characteristic of non-Newtonian fluids.

3.

PNAs, TLGE rheology and vascular casting analysis. (a) Changes in the G′ and G″ moduli of the PNAs at 25–40 °C. (b) Changes in the G′ and G″ moduli of the TLGE at 25–40 °C. (c) The viscosity changes of the PNAs at 25 °C when the shear rate is 105 s^–1 to 2800 s^–1. (d) The viscosity changes of the TLGE 25 °C when the shear rate is 105 s^–1 to 2800 s^–1. (e) The TLGE was retrieved from the catheter maintained at 37 °C.

To visually assess the temperature-dependent gelling behavior, rhodamine-B-stained TLGE was injected through a syringe connected to catheters of different inner diameters (2 mm, 0.55 mm, and 0.2 mm) and immersed in water baths at 25 and 37 °C. At 25 °C, TLGE remained in a liquid state and flowed freely without gelation. In contrast, at 37 °C, TLGE rapidly formed a stable gel within all catheter sizes, confirming its ability to undergo in situ gelation and conform to vessels of varying diameters, a critical feature for adaptable embolization performance (Figure e).

3.4. Ex Vivo Liver Evaluation Results

Since no standardized protocol exists for liver decellularization, we developed a simplified perfusion-based method to efficiently evaluate the performance of embolization materials. − After 10 h of perfusion, histological analysis by H&E staining confirmed the effective removal of cellular components, including nuclei and cytoplasm, from the decellularized liver. Immunofluorescence staining further demonstrated the preservation of key extracellular matrix (ECM) components, including type I collagen, fibronectin, and laminin (Figures a,b and S8). Laminin, a glycoprotein predominantly synthesized by hepatic endothelial and stellate cells, forms a structural basis of the basement membrane along with collagen. Collagen, the major structural polysaccharide–protein in the ECM, provides mechanical support, while fibronectina multifunctional glycoprotein present in the ECM and basement membraneplays important roles in cellular adhesion and tissue repair.

4.

Fabrication and characterizations of decellularized rat liver. (a) Preparation process of decellularized liver in rats and general picture of normal liver and decellularized rat liver. (b) Comparison of normal liver (top) and decellularized liver matrices (bottom), from left to right: hematoxylin and eosin (H&E), collagen I, laminin, fibronectin (original magnification, ×200). (c) SEM images comparing the extracellular matrix of normal liver (c1–c3) and decellularized liver (c4–c6), (c1,c2,c4,c5) represents the vascular wall and surrounding stroma, (c3,c6) represents the extracellular matrix part. The yellow arrow represents the wall of the blood vessel. (d) The process of TLGE emulsion entering decellularized liver blood vessels photographed at 25 °C using a SZ680 type microscope (×400). (e) Before and after injection of TLGE into decellularized liver under digital development X-ray system. (f) DNA agarose electrophoresis. (g) DNA content in normal liver and decellularization liver. **p < 0.01.

Scanning electron microscopy (SEM) revealed that the decellularized liver retained an intact vascular architecture with a porous and loosely organized matrix surrounding the vessel walls, resulting from the removal of parenchymal cells (Figure c). To evaluate the embolization performance of TLGE, rhodamine-B-labeled TLGE was injected into the vascular network of the decellularized liver. The emulsion successfully reached distal small-caliber vessels, demonstrating favorable mobility and comprehensive vascular casting capability (Figures d and S9). Digital X-ray imaging confirmed the excellent radiopacity of TLGE within the decellularized liver (Figure e), supporting its potential for image-guided embolization procedures.

This ex vivo model closely mimics transarterial chemoembolization (TACE) and provides a reproducible platform for preclinical testing of embolic materials. By controlling physiological temperature and flow conditions, it allows for realistic assessment of agent delivery, distribution, and radiopacity under X-ray guidance. To quantify the efficiency of decellularization, agarose gel electrophoresis showed markedly fainter DNA bands in decellularized liver samples compared to normal liver tissue (Figure f). Quantitative analysis indicated that over 90% of DNA was removed from decellularized livers, with residual DNA content below the international threshold of 50 ng/mg (Figure g), confirming successful decellularization.

3.5. Evaluation of In Vivo Imaging Effect of TLGE

Renal artery embolization is a widely accepted model for assessing the efficacy of embolic agents. − In this study, New Zealand white rabbits underwent a 14-day adaptive feeding regimen and were assessed via CT and HE within 42 days postsurgery (Figure a). Following administration of lipiodol, PVA, or TLGE through a 4F microcatheter, digital radiography (DR) was performed at 1, 3, 5, and 7 s postinjection to visualize the dynamic distribution of each agent. Benefiting from its optimal viscosity and injectability, TLGE effectively penetrated into the distal cortical vasculature, achieving comprehensive vascular casting from the renal aorta to the interlobar and afferent arterioles, which demonstrated superior embolization performance compared to both lipiodol and PVA (Figure b). CT scans were performed on the embolized rabbits on days 7, 21, and 42. Cross-sectional CT imaging of the experimental animals further highlighted differences among groups (Figure c). In the lipiodol group, contrast was rapidly washed out by blood flow, leading to poor imaging persistence. Although the PVA group showed limited visibility on days 7 and 21, faint imaging reappeared by day 42, likely due to tissue calcification following prolonged ischemia. In contrast, the TLGE group maintained consistent and clear CT visibility throughout the entire 42-day period, confirming its ability to provide sustained radiographic monitoring and durable vascular occlusion.

5.

Long-term imaging effects of the TLGE on renal artery in normal rabbits. (a) Embolization flowchart of New Zealand rabbit kidney model. (b) Digital radiography of the TLGE. Yellow arrows indicate that the embolic agent reaches the renal vessels at different time periods. (c) Cross-sectional CT images of the kidney embolized by TLGE before embolization, 7 days, 21 days, and 42 days. The red circle represents the embolization kidney.

3.6. Evaluation of the Embolization Effect of TLGE on the Blood Vessels of Normal Rabbit Renal Arteries at All Levels

Three embolization materials, namely lipiodol, PVA, and TLGE, were used to embolize the right kidney of the experimental rabbits. 1 mL iodixanol was administered 5 min later, and partial recanalization of the right kidney was observed in the lipiodol group. At the same time, PVA particles could not completely embolize the right kidney due to their particle size limitation, so part of the contrast agent could enter the right renal artery. Interestingly, TLGE showed excellent casting embolization effect on the right kidney of rabbits, and the lumen of the blood vessels of the right kidney was completely blocked in 5 min (Figure a). CT scans were performed on the embolized rabbits on days 7, 21, and 42 from the 3D reconstructed CT image (Figure b). Compared to the lipiodol and PVA groups, TLGE group showed a more obvious embolization effect on the right kidney, resulting in permanent embolization and ischemia, atrophy, and calcification of the right kidney. Moreover, the embolized artery did not recanalize in the 42-day experiment. Overall, the TLGE group was significantly superior to the other experimental groups in terms of arteriography and embolization and achieved total arterial occlusion for up to 42 days in New Zealand white rabbits.

6.

Evaluation of the embolization effect of TLGE. (a) Comparison of DSA before and after 4 min TLGE injection into rabbit kidney. The red circle represents the embolization kidney. (b) 3D-CT images of the kidney embolized by TLGE before embolization, 7 days, 21 days, and 42 days. The red circle represents the embolization kidney. (c) Representative gross photographs of the right (embolized) and left (contralateral control) kidneys harvested 42 days postembolization. (d) H&E of the right and left kidneys.

Figure c shows the bilateral renal gross images of the iodized oil, PVA, and TLGE groups 42 days after renal artery embolization in rabbits. In the TLGE group, compared to the contralateral normal kidney, the embolized kidney became smaller and showed obvious atrophy. The surface was uneven, the texture was hard to touch, and the fine structure of the kidney tissue could not be distinguished. The kidneys in the PVA embolization group had spotted necrosis on the surface, and it was slightly atrophied as compared to the normal kidneys. After 42 days, the iodized oil group showed no difference between the embolized kidneys and the contralateral normal kidneys, indicating that they all had been recanalized. Hematoxylin and eosin (H&E) showed that after 42 days of embolization in the TLGE group, the pathological findings of renal tissue were nuclei collapse, glomerular atrophy, unclear renal parenchyma, fibrosis, and calcification; furthermore, the glomerular and tubular tissues were necrotic, and their fine structures could not be discerned (Figure d). After 42 days of embolization in the PVA group, glomeruli were still visible, but partial atrophy and nuclear rupture occurred. The kidneys showed of iodized oil group no abnormal changes in each time period after embolization, and the pathological changes were not obvious.

3.7. Biocompatibility Evaluation

At the 42-day postembolization end point, New Zealand White rabbits were sacrificed, and major organs (heart, liver, spleen, lung, and kidneys) were harvested for gross and histopathological examination (H&E staining). Gross inspection revealed no significant abnormalities in any organs other than the expected necrosis in the embolized right kidney. Histological analysis confirmed that the heart, liver, spleen, lungs, and the contralateral (left) kidney, including its renal vessels, tubules, and glomeruli, all exhibited normal architecture with no signs of pathology (Figure a,b). The biocompatibility of TLGE was further supported by hemolysis and cytotoxicity assays. At concentrations ranging from 0.625 to 5 mg/mL, both TLGE and PNA groups exhibited hemolysis rates below 5% (Figure c), with specific values for TLGE being 0.2 ± 0.95%, 1.1 ± 1.34%, 0.4 ± 0.32%, and 0.7 ± 1.01%, respectively. Furthermore, MTT assays demonstrated that TLGE and PNA nanogels, at concentrations of 0.025–0.4 mg/mL, maintained cell viability above 80% (Figure d), indicating low cytotoxicity. Collectively, these results confirm the favorable biocompatibility profile of TLGE.

7.

Biocompatibility of TLGE emulsion. (a) Genera diagram of the heart, liver, spleen, lung and kidney tissues of the experimental rabbits in TLGE emulsion embolization group. (b) H&E diagram of heart, liver, spleen, lung, left kidney and right kidney in TLGE embolization group (original magnification, ×400). (c) Effects of PNAs and TLGE on hemolysis rate of red blood cells in rabbits. The concentrations of PNAs and TLGE were 5–0.625 mg/mL. (d) Effects of PNAs and TLGE on the activity of HepG2 cells. The concentrations of PNAs and TLGE were 0.4–0.025 mg/mL.

4. Conclusions

In this study, we developed a novel temperature-sensitive liquid gel embolization system (TLGE) using a straightforward and efficient synthetic route. When mixed with the conventional contrast agent lipiodol via a medical three-way connector, TLGE maintains excellent fluidity and homogeneous dispersion. Evaluation in a decellularized liver model demonstrated that TLGE exhibits outstanding flowability and vascular penetration, enabling access to terminal vessel branches unreachable by conventional solid embolics. Furthermore, the system provided clear radiographic visualization under X-ray imaging, establishing the decellularized liver as a highly biomimetic platform for evaluating liquid embolic agents. Subsequent in vivo experiments in a rabbit renal embolization model confirmed the embolic efficacy of TLGE, with the material successfully reaching the renal cortical vasculature. Pathological assessment at 42 days postembolization revealed typical postembolic changes including tissue atrophy, necrosis, and calcification, with no evidence of vascular recanalization. Biocompatibility assessments confirmed that TLGE exhibits minimal cytotoxicity and no significant tissue irritation, indicating a favorable safety profile.

In summary, TLGE demonstrates exceptional long-term embolization stability, controllable temperature-induced phase transition behavior, and reliable biosafety, highlighting its strong potential for clinical translation in interventional therapies such as TACE.

All data generated or analyzed during this study are included in this published article and its Supporting Information.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.5c08513.

• Data used for the figures, as well as complementary analyses and additional technical details (PDF)

#.

L.L., N.C., and G.C. contributed equally to this work. L.L. and Y.C. conceived the project and designed the experiments. L.L., N.C. and G.C. carried out the experiments and analyzed the data. N.C., J.X., A.L. and L.Z. assisted in some experiments. H.Y. and L.L. wrote the manuscript. Y.C. and L.L. contributed to the revision.

This study involving animal subjects adhered to rigorous ethical standards and was conducted in strict accordance with the ARRIVE guidelines. Adult New Zealand white rabbits (weighing 2.5–3.0 kg, hermaphrodite) and male Sprague–Dawley rats (weighing 300–500 g) were procured from the Laboratory Animal Center of Hubei University of Science and Technology (Xianning, Hubei, China). All animals were housed in a regulated room temperature for a minimum of 7 days before initiating the study. All experimental procedures were approved by the Animal Ethics Committee of Hubei University of Science and Technology. The study was approved by the Animal Ethics Committee of Hubei University of Science and Technology (approval number: 2024-02-002). All authors agreed to publish this manuscript.

The authors declare no competing financial interest.