A multifunctional injectable nanocomposite hydrogel for precision keloid therapy via ECM remodeling and local pruritus relief

Jingyu Jin; Chengjie Zhu; Nuoya Wang; Mingji Jin; Yanxin Jin; Heying Jin; Shuang Zhang; Yanhong Zhou; Lei Zhao; Yinli Luo; Zhonggao Gao; Zhehu Jin; Shuangqing Wang

doi:10.1186/s12951-026-04060-7

. 2026 Jan 24;24:167. doi: 10.1186/s12951-026-04060-7

A multifunctional injectable nanocomposite hydrogel for precision keloid therapy via ECM remodeling and local pruritus relief

Jingyu Jin ^1,^2,^#, Chengjie Zhu ^3,^#, Nuoya Wang ^4,⁵, Mingji Jin ⁴, Yanxin Jin ⁶, Heying Jin ⁷, Shuang Zhang ⁸, Yanhong Zhou ², Lei Zhao ¹, Yinli Luo ¹, Zhonggao Gao ^4,^✉, Zhehu Jin ^1,^✉, Shuangqing Wang ^4,^✉

PMCID: PMC12911185 PMID: 41580833

Abstract

Keloids are refractory fibroproliferative disorders. Their treatment faces multiple clinical challenges, including the coexistence of multiple pathological mechanisms, low drug bioavailability, and keloid-related pruritus. This study developed an UV-responsive cross-linked hydrogel (FZRL) using lipoic acid-grafted chitosan (LACS) as the matrix. The hydrogel encapsulates ZIF-8 nanoparticles loaded with 5-fluorouracil (5Fu, denoted as 5Fu@ZIF-8) and the local anesthetic ropivacaine. Before injection, the hydrogel exists as a liquid precursor. It can be precisely delivered to the area around keloids. Under mild UV irradiation, it rapidly cross-links into a stable hydrogel. This enables long-term local retention of 5Fu and ropivacaine, preventing premature drug leakage. In the weakly acidic microenvironment of keloids, FZRL hydrogel degrades specifically and releases 5Fu and ropivacaine on demand. A nude mouse keloid model with subcutaneous inoculation of human keloid fibroblasts (KFs) showed that FZRL hydrogel regulates the TGF-β/Smad pathway. It inhibits KFs proliferation and abnormal angiogenesis, modifies the extracellular matrix structure, and remodels the keloid microenvironment. After treatment, the average keloid volume was reduced by 54.7 ± 12.1%. At the same time, it significantly relieved itching-related scratching and irritable behaviors. In conclusion, FZRL hydrogel has dual delivery properties, including UV-responsive gelation and acid-responsive drug release. It synergistically achieves multi-target anti-keloid effects, as well as pruritus relief and analgesia. It provides a new minimally invasive treatment strategy for refractory keloids and has potential clinical translation value.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04060-7.

Keywords: Keloid, Hydrogel, ZIF-8, TGF-β/Smad, Extracellular matrix

Introduction

Keloid is a fibroproliferative disorder characterized by uncontrolled fibroblast activation, excessive deposition of extracellular matrix (ECM), and persistent pruritus and pain [1]. These pathological manifestations markedly impair patients’ quality of life [2]. Accumulating evidence indicates that keloid progression involves the synergistic dysregulation of multiple cellular interactions and signaling pathways [3, 4]. Among these, the persistent overactivation of the TGF-β/Smad axis acts as a central driver, linking key processes such as fibroblast proliferation, myofibroblast differentiation, extracellular matrix accumulation, and abnormal angiogenesis [5, 6]. The biological complexity and dense ECM barrier highlight the need for therapeutic strategies capable of multi-target intervention and deep tissue delivery.

Current clinical treatments, including surgery, corticosteroid injections, radiotherapy, and chemotherapeutic drugs such as 5-fluorouracil (5Fu), remain limited by poor selectivity, high recurrence rates, and insufficient local bioavailability [7, 8]. Particularly, free 5Fu exhibits rapid clearance and cannot effectively penetrate the dense ECM to reach deep fibroblasts, necessitating frequent injections and increasing systemic toxicity [9]. Therefore, developing a minimally invasive, targeted, and long-acting local delivery system that can address both the biological complexity and the physical delivery barriers of keloids remains an urgent unmet clinical need.

In recent years, metal-organic frameworks (MOFs) have shown unique advantages in drug delivery due to their high specific surface area and controllable degradation properties [10, 11]. Among them, zeolitic imidazolate framework-8 (ZIF-8) has been extensively explored in oncology and anti-infective therapy because its Zn-imidazolate coordination bonds remain relatively stable at physiological pH but undergo accelerated degradation in mildly acidic environments [12, 13]. This property is highly relevant to keloids, which exhibit a weakly acidic microenvironment. In contrast to conventional hydrophilic 5Fu injections that diffuse rapidly and are cleared from the lesion, 5Fu-loaded ZIF-8 (5Fu@ZIF-8) enables high drug loading, protects 5Fu from premature degradation, and preferentially releases it in the acidic tissue [14, 15]. 5Fu@ ZIF-8 is expected to improve local bioavailability and break through the dense ECM barrier to achieve deep delivery.

Hydrogels have shown broad application potential in skin repair due to their biocompatibility, local drug retention ability, and tissue microenvironment compatibility [16–19]. In existing studies, methacrylate-based hydrogels enable local drug loading [20, 21]. However, they rely on expensive and potentially toxic photoinitiators (Irgacure 2959), which limits the safety of clinical translation [22]. Poloxamer-based thermosensitive hydrogels can prolong local drug retention via temperature responsiveness [23]. Yet, they have insufficient mechanical strength, are susceptible to body temperature fluctuations, and tend to deform and leak drugs in high-tension keloid areas. This makes it difficult to maintain long-term effective drug concentrations. pH-responsive hydrogels lack externally controllable gelation mechanisms. They are prone to premature drug release in non-target tissues, resulting in insufficient precision [24]. More importantly, the functions of existing hydrogels are limited to a single anti-keloid effect. Most drug-loaded hydrogels release drugs through passive diffusion driven by matrix swelling. They cannot achieve on-demand release triggered by the keloid microenvironment, leading to low drug utilization and potential damage to normal tissues [25]. Among hydrogel-forming materials, chitosan (CS) shows particularly promising potential [26]. The lipoic acid (LA)-grafted CS (LACS) matrix proposed in this study forms a three-dimensional network through rapid cross-linking under ultraviolet light. It can not only achieve precise gelation and local fixation after injection, but also avoid the fragility of traditional thermosensitive materials, thus better meeting the mechanical and controlled-release requirements for keloid treatment.

Notably, keloid-related pruritus and pain are often overlooked in clinical practice and markedly impair treatment compliance [27, 28]. Ropivacaine can relieve these symptoms, but its action lasts only 4–6 h and it readily diffuses into surrounding normal tissues, making long-term local analgesia difficult. Incorporating ropivacaine into a microenvironment-responsive hydrogel could provide prolonged local antipruritic and analgesic effects via slow degradation and sustained release, thereby improving the treatment experience and matching the duration of anti-keloid therapy. Li et al. encapsulated ropivacaine in carboxymethylcellulose/oxidized hyaluronic acid hydrogels and addressed low loading and poor sustained release of water-soluble anesthetics [29]. In this study, ropivacaine is loaded into an LACS-based hydrogel and is expected to exert synergistic antipruritic, analgesic, and antifibrotic effects with 5Fu, which better aligns with practical clinical needs.

To address the above-mentioned bottlenecks, this study proposes a design concept for a responsive delivery system based on the synergy of pathological mechanisms. Its core logic is to design and construct a composite delivery system of UV-responsive hydrogel (FZRL) loaded with 5Fu@ZIF-8 nanoparticles and ropivacaine. By virtue of the inherent properties of the carrier, this system achieves a multi-component synergy, multi-responsive targeting, and clinically compatible anti-keloid treatment strategy. Among them, LACS serves as the hydrogel matrix material. The thiol groups provided by the LA component of LACS can be rapidly cross-linked under UV light induction to form a three-dimensional network structure. This structure enables minimally invasive and precise delivery of the liquid precursor before injection, and achieves time-controlled and site-controlled gelation via UV light. It prevents local loss caused by premature drug leakage or delayed gelation, ensuring stable retention of the carrier at the keloid site. Building on this, the system incorporates 5Fu@ZIF-8 nanoparticles and ropivacaine into the hydrogel while leveraging ZIF-8’s pH responsiveness and the hydrogel’s controlled release properties, enabling it to simultaneously block two key pathological processes: the abnormal activation of KFs and angiogenesis. It also alleviates local discomfort during treatment. This design overcomes the limitation of “single-link inhibition” in traditional treatments and achieves a synergistic effect of anti-keloid, antipruritic, and analgesic. Meanwhile, relying on the dual regulatory mechanisms of UV-responsive gelation and pH-responsive drug release, the system effectively solves the core bottlenecks of poor drug targeting and short action duration. Ultimately, it provides a new therapeutic strategy with efficacy, safety, and clinical practicality for the precise treatment of keloids.

Materials and methods

Materials

5Fu (F809394) were procured from the macklin bio company (Shanghai, China). Ropivacaine (HY-B0563) was purchased from MedChemExpress (New Jersey, US). Zinc nitrate hexahydrate was obtained from Lanzhou Yellow River Institute of Zinc and Magnesium Nanomaterials (Gansu, China). N-hydroxysuccinimide (NHS, L011830), Ethanol (1170), and 2-methylimidazole (L013241) were supplied by Beijing Tongguang Fine Chemicals Company (Beijing, China). CS (C105799) was acquired from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). Lipoic acid (LA) was purchased from Shanghai Acmec Biochemical Technology Co., Ltd. (Shanghai, China). Dimethyl sulfoxide (DMSO d8371), Crystal violet staining solution, and Radio-Immunoprecipitation Assay (RIPA) lysis buffer (R0010) were obtained from Beijing Solarbio Science & Technology Co., Ltd. (Beijing, China). Cell Counting Kit-8 (CCK-8) was purchased from Dojindo Laboratories (Kumamoto, Japan). Rhodamine B (RhB, PH9213) was supplied by Phygene Biotechnology Co., Ltd. (Fujian, China). Pierce BCA Protein Assay Kit (1859078, ZE391532) was purchased from Thermo Fisher Scientific (Shanghai, China). Smad2 antibody (CY5090) and Smad3 antibody (CY5013) were purchased from Mabwell (Shanghai) Bioscience Co., Ltd. (Shanghai, China). Matrigel (256234) was supplied by Corning Inc. (New York, US). Transforming growth factor-β (TGF-β, GB112134) and collagen type Ⅲ (COL3, WL03186) were obtained from Wanlei Biotech Co., Ltd. (Liaoning, China). Fibronectin (FN, AF5335) and α-SMA (AF1032) antibodies were purchased from Affinity Biosciences Co., Ltd. (Jiangsu, China). Collagen type Ⅰ (Col 1, GB11022-3) was supplied by Servicebio Technology Co., Ltd. (Hubei, China). All reagents were analytical grade and used without further purification.

The KFs was purchased from ATCC. HUVEC cells was bought from the Cell Resource Center, IBMS, CAMS/PUMC. For all in vitro experiments, only early-passage KFs (passages 3–5 after thawing) were used, and all assays were completed within 14 days of thawing. Male BALB/c nude mice (16–20 g) were obtained from Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). All animal experiments were approved by the Laboratory Animal Ethics Committee in the Institute of Materia Medica and Peking Union Medical College (00000761). All procedures followed ethical standards during the experiment. This project has been approved by the Ethics Committee of Scientific Research/Cell Clinical Research of Yanbian University Hospital (20250030). All patients signed the informed consent form.

Proteomic analysis of normal skin and keloid tissues

Sample source and handling

Samples were collected from patients undergoing keloid surgery in the Department of Dermatology, Yanbian University Hospital, including keloid tissues (n = 3) and excess normal skin obtained during surgery (n = 3). Fresh tissues were rinsed with normal saline to remove surface blood stains, then immediately stored at −80 °C for freezing to prevent protein degradation.

Protein Extraction, enzymatic Digestion, and mass spectrometry analysis

After grinding samples into powder with liquid nitrogen, RIPA lysis buffer was added. The mixture was sonicated for 2 min using an ultrasonic cell disruptor, followed by incubation on ice for 30 min. Cell debris was removed by centrifugation (13,000 rpm, 4 °C, 20 min), and the supernatant was collected as the total protein extract. The concentration of the extracted protein was determined using a BCA protein quantification kit. The quantified protein samples were subjected to reduction and alkylation in sequence. Then, the protein was digested with trypsin and incubated overnight at 37 °C. Peptides were separated by reversed-phase high-performance liquid chromatography, and eluents at different time periods were collected for mass spectrometry analysis. The separated peptides were analyzed by LC-MS/MS (APExBIO, Houston, USA), and the mass spectrometry data of each sample were recorded. Bioinformatics software was used to analyze the mass spectrometry data, compare the protein expression differences between normal skin and keloid tissues, and generate protein expression profiles.

Construction and characterization of 5Fu@ZIF-8

Construction of 5Fu@ZIF-8

5Fu@ZIF-8 was synthesized via a one-pot method. The specific steps were as follows: 2 mg 5Fu and 80 mg zinc nitrate hexahydrate were weighed and dissolved in 1 mL DMSO, followed by stirring at 1500 rpm at room temperature for 10 min. A 4 mL aqueous solution containing 800 mg 2-methylimidazole was added dropwise to the above solution. After standing for 5 min, the mixture was stirred at 1500 rpm for 30 min to spontaneously form a 5Fu@ZIF-8 suspension. The precipitate was collected by centrifugation at 14,000 rpm for 10 min, washed with deionized water 3 times, vacuum-dried, and then stored at −20℃.

Characterization of 5Fu@ZIF-8

The particle size, polydispersity index (PDI), and Zeta potential of the nanoparticles were determined using a ZS90 Malvern Zeta Nanosizer (Malvern Instruments Ltd., Malvern, UK) based on the dynamic light scattering (DLS) principle. 10 µL each of methanol/water solutions of ZIF-8 and 5Fu@ZIF-8 were dropped onto carbon-coated copper grids. After natural volatilization and drying, the morphology was observed via transmission electron microscopy (TEM, Hitachi H-600). X-ray powder diffraction (XRD) data of 5Fu, ZIF-8, and 5Fu@ZIF-8 were acquired using a D/max 2200VPC X-ray diffractometer (Rigaku Corporation, Japan) with a scanning angle range of 2θ = 3°−50°. Fourier transform infrared (FT-IR) spectra of 5Fu, ZIF-8, and 5Fu@ZIF-8 were obtained using a Bruker Vertex 70 spectrometer (Billerica, USA). The thermal stability and phase transition properties of 5Fu, ZIF-8, and 5Fu@ZIF-8 were evaluated via thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC) using a Mettler TG-DSC 3 + thermogravimetric analyzer and a TA DSC25 differential scanning calorimeter (Switzerland), respectively.

Stability investigation

5Fu@ZIF-8 powder was stored at 4℃ under refrigeration, and samples were taken daily for 7 d. The stored powder was dispersed in pH 7.4 PBS, and its particle size and PDI were measured to evaluate the effect of storage conditions on the stability of 5Fu@ZIF-8.

Cellular-Level studies

Cytotoxicity of blank ZIF-8

KFs and HUVEC were seeded in 96-well plates at 5 × 10³ cells/well. They were cultured at 37℃ with 5% CO₂ for 24 h. The old medium was discarded. 10% fetal bovine serum (FBS) DMEM medium containing blank ZIF-8 at different concentrations was added (200 µL/well). A control group without ZIF-8 and a zero-adjustment group were set. After 24 h of culture, the medium was replaced with serum-free medium containing 10% CCK-8 (200 µL/well) and incubated for 2 h. The absorbance at 450 nm was measured using a microplate reader (BioTek, Dallas, TX, USA). Each group had 6 replicate wells.

Inhibitory effect of 5Fu@ZIF-8 on cell proliferation

KFs and HUVEC were seeded in 96-well plates at 5 × 10³ cells/well and cultured for 24 h. The medium was discarded. 10% FBS medium containing 5Fu, ZIF-8, or 5Fu@ZIF-8 was added (100 µL/well). A blank control group and a zero-adjustment group were set. After 24–48 h of culture, OD values were measured to calculate cell viability. Each group had 6 replicate wells.

KFs and HUVEC were seeded in 12-well plates with cell climbing slices at 1.5 × 10⁵ cells/well. After 24 h of culture, the old medium was discarded. Solutions containing 5Fu, ZIF-8, or 5Fu@ZIF-8 were added and incubated for 24 h. After washing 3 times, 500 µL Calcein-AM/PI double-staining working solution was added to each well. The cells were incubated at 37℃ in the dark for 30 min. Cell staining was observed using Cytation5 (Biotek, Winooski, VT, USA).

KFs and HUVEC were seeded in 12-well plates with climbing slices at 1.5 × 10⁵ cells/well and cultured for 24 h. Solutions containing 5Fu, ZIF-8, or 5Fu@ZIF-8 were added and incubated for 24 h. After fixation with 4% paraformaldehyde, nuclei were stained with 5 µg/mL Hoechst 33,258 for 15 min. Apoptotic morphology was observed via confocal laser scanning microscopy (CLSM).

KFs and HUVEC were seeded in 6-well plates at 2 × 10⁴ cells/well and cultured for 24 h. After treatment as described above, cells were digested with 0.25% trypsin, centrifuged at 1000 g for 5 min, and resuspended in PBS. Analysis was performed using a FACSCalibur flow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA). Each group was repeated 3 times.

Uptake experiment of 5Fu@ZIF-8

RhB@ZIF-8 was prepared by replacing 5Fu with rhodamine B (RhB) to investigate cell uptake characteristics.

KFs and HUVEC were seeded in 12-well plates with cell climbing slices at 1 × 10⁴ cells/well and cultured for 24 h. The medium was discarded. RhB or RhB@ZIF-8 containing 1 µg/mL RhB (diluted with serum-free medium) was added and incubated for 1, 2, or 6 h, respectively. After fixation with 4% paraformaldehyde for 15 min and nuclear staining with DAPI for 15 min, observations were made via CLSM.

KFs and HUVEC were seeded in 6-well plates at 2 × 10⁴ cells/well and cultured for 24 h. After treatment as described above, unabsorbed particles were washed away with cold PBS. Cells were digested with 0.25% trypsin, centrifuged at 1000 g for 5 min, and resuspended in PBS. Analysis was performed using a FACSCalibur flow cytometer. Each group was repeated 3 times.

Inhibitory effect of 5Fu@ZIF-8 on cell migration

KFs and HUVEC were seeded in 6-well plates at 2 × 10⁵ cells/well and cultured for 24 h. Scratches were made with a sterile pipette tip. PBS was used to wash away detached cells. Solutions containing 5Fu, ZIF-8, or 5Fu@ZIF-8 (diluted with 1% FBS medium) were added. A blank control was set. Images were taken with an inverted microscope at 0, 12, and 24 h. Migration rate was calculated using Image J software. Each group was repeated 3 times.

Medium was added to the lower chamber of Transwell chambers (costar 3422, 8.0 μm pore size). Treated KFs and HUVEC were seeded in the upper chamber and cultured for 24 h. After washing with PBS, cells were fixed and stained with crystal violet. The number of migrated cells in the lower chamber was counted.

Expression of α-SMA and FN

KFs were seeded in 12-well plates with cell climbing slices at 1 × 10⁴ cells/well. After 24 h of culture, solutions containing 5Fu, ZIF-8, or 5Fu@ZIF-8 were added and incubated for 24 h. After fixation with 4% paraformaldehyde at room temperature, permeabilization with 0.1% Triton X-100 for 10 min, and blocking with 5% BSA for 1 h, α-SMA primary antibody and FN primary antibody were added respectively and incubated at 4℃ overnight. The next day, Alexa Fluor 488-labeled goat anti-rabbit secondary antibody (1: 500) was added and incubated at room temperature in the dark for 1 h. Nuclei were stained with DAPI for 15 min. Fluorescence distribution was observed via CLSM. Fluorescence intensities of α-SMA and FN were quantified using Image J software.

Quantification of col 1 and col 3

KFs were seeded in 6-well plates at 5 × 10⁴ cells/well. After 24 h of culture, the medium was replaced with drug-containing medium and incubated for 48 h. Cell supernatants from each group were collected and centrifuged at 12,000 rpm at 4℃ for 10 min to remove impurities. Operations were performed according to the instructions of Col 1 and Col 3 ELISA kits.

TGF-β1/Smad signaling pathway

KFs were seeded in 6-well plates at 1 × 10⁶ cells/well and treated with drugs for 24 h. 200 µL RIPA lysis buffer was added to each well. Cells were lysed on ice for 30 min. Cell lysates were scraped and centrifuged at 13,000 rpm at 4℃ for 20 min. Supernatants were collected. Samples were separated by SDS-PAGE (Dakewe, Shenzhen, China), electrophoresis, and membrane transfer, then blocked at room temperature. PVDF membranes were incubated with different antibodies at 4℃ overnight. After washing 3 times, HRP-labeled goat anti-rabbit secondary antibodies were added and incubated at room temperature for 1 h. ECL chemiluminescence kit was used for color development. Images were taken with a gel imaging system. Gray values of target proteins were quantified using Image J software, normalized to β-actin, and relative expression levels of each protein were calculated.

Construction and evaluation of hydrogels

Synthesis and characterization of LACS matrix

The synthesis procedure of LACS is as described above [30]. Briefly, 0.8 g CS was dispersed in 40 mL of 1.5% (v/v) acetic acid solution to prepare a CS solution. After continuous stirring for 1 h, 20 mL of ethanol solution containing 0.08 g LA was added to the CS solution, and stirred until a homogeneous mixture was formed. Subsequently, 0.534 g N, N-dimethylaminopropyl ethylcarbodiimide (EDC) and 0.1334 g NHS were added to the mixture, and the reaction was carried out for 4 h. Finally, the reaction mixture was dialyzed with a dialysis bag for 3 d, and the product was freeze-dried for later use.

The FT-IR was performed with a scanning range of 4000–400 cm^− 1. Pure potassium bromide sheets were used for background correction to eliminate environmental interference. A nuclear magnetic resonance (NMR) spectrometer (Bruker Avance II, 400 MHz) was used for testing at 25℃, with a chemical shift range of δ = 0–11 ppm collected.

Construction of hydrogels

A certain amount of 5Fu@ZIF-8 was added to pH 7.4 PBS. The mixture was stirred at 150 rpm at room temperature in the dark for 15 min to uniformly disperse the nanoparticles, obtaining PBS containing 5Fu@ZIF-8. 2.0 g of dry LACS powder was slowly added to 100 mL of the above solution and dissolved to prepare a 2% (w/v) LACS stock solution. The solution was irradiated with a 365 nm UV curing device (power: 10 mW/cm², irradiation distance: 5 cm). Gel formation was verified by the inverted test tube method (no liquid flow), and the LACS hydrogel containing 5Fu@ZIF-8 (denoted as FZL hydrogel) was finally formed. The hydrogel without 5Fu@ZIF-8 was denoted as LACS hydrogel; the hydrogel with only ZIF-8 added was denoted as Blank ZL hydrogel.

Characterization of hydrogels

2% LACS hydrogel precursor (containing 5Fu@ZIF-8) was loaded into a 1 mL syringe. Injectability was verified by manual injection to observe fluidity. 1 mL of the precursor was placed in a Petri dish. After irradiation with 365 nm UV light for 10 s, the container was inverted to observe whether gelation occurred. Mechanical strength was evaluated by pinching and stretching to verify UV-responsive gelation performance and local retention ability. The microstructure of the hydrogel was observed by SEM. A rotational rheometer was used for testing: 1 mL of gelled hydrogel was taken, and a plate fixture with a diameter of 20 mm and a spacing of 1 mm was used. Frequency scanning was performed at 37℃ with a range of 0.01–10 Hz. The storage modulus (G’) and loss modulus (G’’) were measured to analyze the stability of the gel network.

The hydrogel was weighed accurately and immersed in pH 7.4 PBS and 37℃. It was taken out at 1, 2, 4, 6, 12, and 24 h respectively. After wiping off surface moisture, it was weighed again to calculate the swelling ratio. Hydrogels with known weight were immersed in PBS at pH 5.5, 6.8, or 7.4 (37℃) respectively. Residual hydrogels were taken out at regular intervals, freeze-dried, and weighed to calculate the degradation rate. Following the above conditions: the hydrogel containing 5Fu@ZIF-8 was immersed in PBS at 37℃ with pH 5.5, 6.8, or 7.4. Samples were taken at regular intervals, and an equal volume of fresh medium was added. The concentration of 5Fu in the release solution was determined using a HPLC (1260 Agilent, Germany), and the cumulative release rate was calculated.

KFs and HUVEC were seeded on the surface of blank LACS hydrogels (1 × 10⁴ cells/well) and cultured continuously. Calcein-AM/PI double staining was performed on days 1, 2, 3, and 4 respectively. The ratio of live to dead cells was observed via CLSM to evaluate cell compatibility.

In vivo distribution and safety

In vivo distribution

RhB@ZIF-8 was encapsulated into 2% LACS hydrogel to obtain RhB@ZIF-8@LACS. Nude mice aged 6–8 weeks were selected, and KFs were subcutaneously inoculated to establish a keloid model. After keloid formation, free RhB, RhB@ZIF-8, and RhB@ZIF-8@LACS (RhB dose: 0.1 mg/kg for all groups) were injected adjacent to the keloids respectively. At 0, 4, 24, 30, 48, 56, and 72 h after injection, a small animal in vivo imaging system was used to capture fluorescence signals of the whole body and keloid area of nude mice. At 72 h, the nude mice were sacrificed. Hearts, livers, spleens, lungs, and kidneys were collected, and fluorescence images of isolated organs were taken to compare the differences in fluorescence distribution among groups.

Penetration depth in keloid tissue

After the above nude mice were sacrificed, keloid tissues were completely stripped and subjected to frozen sectioning. The sections were placed under a CLSM and scanned along the vertical direction of the keloid (from the surface to the deep layer) to record the penetration depth of RhB in keloid tissue.

In vivo biosafety detection

Ropivacaine was added to the PBS solution containing 5Fu@ZIF-8, followed by the addition of 2% LACS stock solution. After curing with 365 nm UV light, FZRL hydrogel was obtained. Nude mice were selected for subcutaneous injection of FZRL hydrogel, and a normal saline control group was set up. At 14, 21, and 30 d after administration, blood was collected from the orbital venous plexus. Then the nude mice were sacrificed, and hearts, livers, spleens, lungs, kidneys, and skin tissues at the injection site were collected. The organs and skin tissues were stained with H&E to observe and evaluate whether there were pathological changes such as cellular edema, inflammatory infiltration, and necrosis. An automatic biochemical analyzer was used to detect liver function indicators (ALT and AST) and renal function indicators (BUN and CREA) in serum. The differences of these indicators between the FZRL group and the control group were compared.

In vivo pharmacodynamics

Keloid model establishment and grouped administration

KFs cultured in vitro to the logarithmic growth phase were mixed with Matrigel at a density of 1 × 10⁷ cells/mL, and subcutaneously inoculated into the back of nude mice to establish an ectopic keloid model. Fourteen days after inoculation, the nude mice were randomly divided into 6 groups (n = 5): Control group, free 5Fu group, 5Fu@ZIF-8 group, Blank ZL group (ZIF-8/LACS hydrogel without 5Fu), FZL group (5Fu@ZIF-8/LACS hydrogel), and FZRL group (5Fu@ZIF-8/ropivacaine/LACS hydrogel). Local administration was initiated on the 14th day after keloid formation, once every 2 d, for a total of 7 administrations.

Pharmacodynamic evaluation and body weight monitoring

At the same time every day, nude mice were placed in a transparent observation cage (15 cm ×22 cm ×20 cm). Scratching behavior on the keloid site was observed for 30 min, and the scratching frequency was recorded at 5-min intervals. Local discomfort was evaluated at the following time points: D0 (one day before the first drug injection), D1 (30 min before the first administration of different formulations), D2 (30 min after the first administration), D3 (1 d after the first administration), and D4 (2 d after the first administration). Behaviors such as wiping, scratching the keloid, and restlessness were observed to assess the degree of local discomfort. Keloid photos were taken every 3 d. The keloid volume were measured with a vernier caliper. The weight growth trends of each group were compared to evaluate the effect of drugs on the overall state of nude mice.

Histopathology and apoptosis detection

Twenty-six days after administration, nude mice were sacrificed, and keloid tissues were completely stripped and made into paraffin sections. H&E staining was performed to observe the inflammatory status and cell morphology of keloid tissues via optical microscopy. TUNEL staining was used to label apoptotic cells. The proportion of apoptotic cells to total cells was quantified using Image J software to analyze the inductive effect of each group on keloid cell apoptosis.

ECM regulation detection

Frozen sections of keloid tissues were prepared and subjected to immunofluorescence staining for Col 1, Col 3, α-SMA, and FN, respectively. Fluorescence distribution was observed via CLSM. Fluorescence intensity and positive expression rate were quantified using Image J software to evaluate the deposition changes of ECM components. Western blot was used to detect the protein expression levels of TGF-β1, P-Smad2, Smad2, P-Smad3, and Smad3 in the tissues.

Statistical analysis

All experimental data in this study were expressed as mean ± standard deviation (mean±SD), and the sample size was clearly indicated in each experiment according to the experimental type. Statistical analysis of data was performed using GraphPad Prism 9.0 software, and graphing was completed using Origin 2023 software. Independent samples t-test was used for comparisons between two groups; one-way analysis of variance (one-way ANOVA) was used for comparisons among multiple groups, and specific differences between groups were further analyzed via Tukey’s multiple comparison test.

Results

Proteomic characteristics of human keloid tissues

Proteomic comparison between normal human skin (Control) and keloid tissues revealed the molecular regulatory features of keloid development. Figure 1A showed the overall distribution of differential proteins: 259 proteins were significantly downregulated and 251 were significantly upregulated in keloids, indicating that keloid formation is accompanied by extensive protein expression remodeling and involves complex molecular network regulation. Figure 1B (volcano plot) presented core differential proteins and their expression patterns: collagen family (COL12A1) and glycosaminoglycan metabolism-related protein (HGSNAT) were significantly upregulated, while immune regulation-related proteins (HLA-F and SASH1) were significantly downregulated. Figure 1C confirmed a systematic difference in protein expression profiles between the two groups, with multiple proteins showing consistent up/downregulation trends in keloids, reflecting their characteristic expression patterns. Molecular Function (MF) enrichment analysis (Fig. 1D) showed significant enrichment of functions related to collagen synthesis and modification (procollagen-lysine 5-dioxygenase activity, procollagen-proline 4-dioxygenase activity), ECM structural composition, and fibronectin binding, which is consistent with the pathological phenotype of excessive collagen deposition. GO function association chord diagram (Fig. 1E) indicated that differential proteins are widely involved in processes such as ECM-receptor interaction, protein transport, and cell adhesion, suggesting that the synergistic abnormality of these processes is an important molecular basis for scar hyperplasia and tissue remodeling. In summary, the proteomic characteristics of keloids are abnormal activation of proteins related to collagen metabolism and ECM remodeling, accompanied by disorders in immune regulation and cell adhesion, providing molecular targets and theoretical basis for subsequent targeted intervention.

Fig. 1 — Differentially expressed proteomics (DEPs) between keloids and normal controls, and functional enrichment analysis. (A) Statistical count of DEGs between keloids and normal controls (blue: downregulated genes; red: upregulated genes). (B) Volcano plot of DEGs between keloids and normal controls. (C) Heatmap of DEGs between keloids and normal controls. (D) Bubble plot of functional enrichment for DEGs. (E) Chord diagram of GO functional enrichment for DEGs

Physicochemical characterization of ZIF-8 and 5Fu@ZIF-8

Particle Size, zeta Potential, and microstructure

DLS directly reflects the dispersion stability of nanoparticles in solution, which is a prerequisite for the in vivo application of drug delivery systems. Notably, the hydrodynamic diameter slightly increased after 5Fu loading. As shown in Fig. 2A, ZIF-8 exhibited a particle size of 109.8 ± 2.3 nm (PDI 0.149 ± 0.035), whereas 5Fu@ZIF-8 displayed a larger size of 137.4 ± 3.1 nm (PDI 0.129 ± 0.029). This moderate increase in size is consistent with the incorporation of 5Fu into the ZIF-8 framework and/or pores, which contributes to the overall hydrodynamic diameter while maintaining a narrow size distribution. Importantly, both formulations remained within the 100–150 nm range with low PDI values, indicating good colloidal stability suitable for subsequent in vitro and in vivo applications. The Zeta potentials of ZIF-8 and 5Fu@ZIF-8 were +21.6 mV and +12.6 mV, respectively (Fig. 2B). The loading of 5Fu slightly altered the surface charge properties of ZIF-8, but did not affect the dispersibility of the nanoparticles.

Fig. 2 — Physicochemical properties and performance characterization of ZIF-8 and 5Fu@ZIF-8 nanoparticles. (A) Particle size distribution of ZIF-8 and 5Fu@ZIF-8. (B) Zeta potential of ZIF-8 and 5Fu@ZIF-8. (C) TEM images of ZIF-8 and 5Fu@ZIF-8 (scale bar: 100 nm). (D) Changes in particle size and PDI of 5Fu@ZIF-8 within 7 d (n = 3). (E) Drug loading capacity and encapsulation efficiency of 5Fu@ZIF-8 (n = 3). (F) XRD patterns of ZIF-8, 5Fu@ZIF-8, and 5Fu. (G) FT-IR spectra of ZIF-8, 5Fu@ZIF-8, and 5Fu. (H) TGA, DTG, and DSC curves of ZIF-8, 5Fu@ZIF-8, and 5Fu. (I) Cumulative release profiles of 5Fu from 5Fu@ZIF-8 under pH 7.4, pH 6.8, and pH 5.5 conditions (n = 3)

TEM enabled direct observation of the morphology, size, and structural integrity of the nanoparticles. ZIF-8 showed a regular polyhedral structure with uniform particle size (Fig. 2C). After loading 5Fu, 5Fu@ZIF-8 still maintained a clear geometric morphology, indicating that 5Fu was successfully loaded without destroying the crystal structure of ZIF-8. The structural integrity of the carrier provided a basis for subsequent targeted drug delivery. A small discrepancy between DLS and TEM is reasonable, because DLS reports the hydrodynamic size of nanoparticles in dispersion, whereas TEM measures their dry, dehydrated size [31].

Drug loading performance of 5Fu@ZIF-8

Stability experiments are important for simulating the dispersion behavior of nanoparticles during in vivo circulation or local retention. 5Fu@ZIF-8 showed no significant change in hydrated particle size within 7 d, and the PDI remained below 0.3 (Fig. 2D). This indicated that the nanoparticles were stably dispersed in solution without obvious aggregation. Determination of drug loading capacity and encapsulation efficiency are core indicators for evaluating the drug loading ability of carriers. 5Fu@ZIF-8 had a drug loading capacity of 8.8 ± 1.3% and an encapsulation efficiency of 91.7 ± 2.5% (Fig. 2E), confirming that ZIF-8 had high-efficiency drug loading ability for 5Fu.

XRD is a classic method for analyzing crystal structure and phase composition, which can verify the successful loading of drugs and the integrity of the carrier’s crystal structure. ZIF-8 showed typical characteristic diffraction peaks, and the diffraction pattern of 5Fu@ZIF-8 was basically consistent with that of ZIF-8, with no characteristic peaks of free 5Fu observed (Fig. 2F). The XRD results indicated that 5Fu was successfully loaded and the crystal structure of ZIF-8 was not destroyed. The FT-IR spectra (Fig. 2G) further confirmed the successful loading of 5Fu into the ZIF-8 framework. ZIF-8 showed its characteristic absorption peaks at 3130 cm^− 1 (C–H stretching), 1580 cm^− 1 (C = N stretching of the imidazole ring), 1145 cm^− 1 (C–N stretching), and 421 cm^− 1 (Zn–N vibration), corresponding to the integrity of the ZIF-8 coordination structure. In contrast, 5Fu displayed typical bands at 1690 cm^− 1 (C = O stretching), 1420 cm^− 1 (C–F bending), and 1250 cm^− 1 (C–N stretching). The spectrum of 5Fu@ZIF-8 contained all the major framework peaks of ZIF-8 while also showing weakened and slightly shifted characteristic peaks of 5Fu, indicating the incorporation of 5Fu molecules into the ZIF-8 structure rather than simple physical adsorption. Thermal analysis (TGA, DTG, and DSC) is an important method for analyzing the thermal stability of materials and component interactions, which can indirectly verify drug loading. The results (Fig. 2H) showed that ZIF-8 and 5Fu@ZIF-8 had similar trends in thermal weight loss and heat flow changes, while 5Fu showed significant weight loss and thermal effect differences in the high-temperature region. After 5Fu was loaded, its thermal behavior was regulated by the ZIF-8 framework, which indirectly verified successful loading.

In vitro drug release

In vitro drug release experiments are key indicators for simulating the release behavior of drugs in different in vivo microenvironments, and can reflect the stimuli-responsiveness of the carrier and drug delivery efficiency. In a physiologically neutral environment (pH 7.4), the release amount of 5Fu was 19.27 ± 1.83% at 72 h (Fig. 2I). Under acidic conditions simulating the keloid microenvironment (pH 5.5), the cumulative release rate of 5Fu was 82.15 ± 1.28% within 72 h. This release behavior indicated that 5Fu@ZIF-8 had pH responsiveness and could specifically release drugs at the keloid site.

Table S1 summarizes the fitting results of the in vitro release profiles of 5Fu@ZIF-8 at different pH values using zero order, first order, Higuchi, and Ritger-Peppas models. At pH 7.4, the Higuchi (R² = 0.993) and Ritger-Peppas (R² = 0.9972) models showed the best fits, indicating a diffusion-dominated release process; the Peppas exponent n = 0.62785 further supports diffusion-controlled behavior. At pH 6.8, the Ritger-Peppas model still provided the highest goodness of fit (R² = 0.9285) with n = 0.34767, suggesting diffusion remains dominant but is increasingly influenced by the acidic microenvironment. At pH 5.5, the first order model gave the best fit (R² = 0.9677), consistent with the acid-sensitive degradation of ZIF-8 and indicating a release mechanism governed by the combined effects of carrier degradation and drug diffusion.

Biological effects of 5Fu@ZIF-8 on KFs

Cytotoxicity and proliferation Inhibition

The cell compatibility test of ZIF-8 carrier is the basis for evaluating the biosafety of the carrier. When ZIF-8 concentration ranged from 0.1 to 50 µg/mL, the viability of KFs remained above 80% after 24–48 h of treatment (Fig. 3A), indicating that ZIF-8 itself had low toxicity to KFs and good biocompatibility.

Further, the proliferation inhibition effects of 5Fu and 5Fu@ZIF-8 on KFs were compared. The results showed (Fig. 3B) that at 24 h, there was no significant difference in the effect on cell viability between 5Fu and 5Fu@ZIF-8 at low concentrations (≤ 5 µg/mL). However, after 48 h of treatment, with the increase of drug concentration (≥ 10 µg/mL), the viability of KFs in the 5Fu@ZIF-8 group was significantly lower than that in the 5Fu group. This indicated that after being encapsulated by ZIF-8, 5Fu had a stronger inhibitory effect on KFs proliferation, with time dependence. Live/dead cell staining is a classic method to directly distinguish live and dead cells, which can quickly evaluate the cytotoxicity of drugs. KFs in the Control group and ZIF-8 group were mainly green (live cells), while the proportion of red (dead cells) increased in the 5Fu group (Fig. 3C). The 5Fu@ZIF-8 group had significantly more red fluorescent cells, further verifying that 5Fu@ZIF-8 had a more significant killing effect on KFs. The 5Fu@ZIF-8 group showed significant nuclear pyknosis and fragmentation, and the apoptotic fluorescent signal was significantly stronger than that in other groups (Fig. 3D). The apoptosis results detected by flow cytometry were consistent with the CLSM results (Fig. 3E). The above results indicated that 5Fu@ZIF-8 might inhibit KFs proliferation by inducing apoptosis.

Cell uptake efficiency

CLSM and flow cytometry are key methods to evaluate cell uptake ability by combining quantification and qualification, which can reveal the cell drug delivery efficiency of the carrier. CLSM showed (Fig. 3F) that in the RhB@ZIF-8 group, the intracellular red fluorescence gradually increased with the extension of incubation time (1, 2, and 6 h), and the fluorescence was the strongest at 6 h. However, only weak fluorescence was observed in the RhB group. The quantitative results of flow cytometry further confirmed (Fig. 3G) that the mean fluorescence intensity of the RhB@ZIF-8 group was significantly higher than that of the RhB group, and the uptake reached the peak at 6 h. This indicated that ZIF-8 could efficiently mediate 5Fu into KFs and increase the intracellular drug concentration, which was one of the core mechanisms for enhancing the anti-scar activity of 5Fu. The ZIF-8 carrier had both low toxicity and high-efficiency drug delivery ability. After encapsulation, the proliferation inhibition effect of 5Fu on KFs was significantly enhanced, and this effect was closely related to the improvement of intracellular drug uptake and the induction of cell apoptosis.

It is noteworthy that although free 5Fu is a much smaller molecule than 5Fu@ZIF-8, its intracellular accumulation is not necessarily higher. As a hydrophilic small molecule, free 5Fu mainly enters cells through passive diffusion or nucleobase transporters, and can be rapidly effluxed or metabolized, leading to limited intracellular retention. In contrast, 5Fu@ZIF-8 nanoparticles are internalized predominantly through endocytosis, resulting in much more efficient and sustained intracellular delivery. This mechanism is consistent with the markedly stronger RhB@ZIF-8 fluorescence intensity observed in KFs compared with free RhB, and explains why 5Fu@ZIF-8 produced more pronounced cytotoxicity and apoptosis than free 5Fu at the same nominal drug concentration.

Biological effects of 5Fu@ZIF-8 on HUVEC

Keloid formation is closely related to angiogenesis. Meanwhile, the toxicity of drugs to vascular endothelial cells directly affects normal skin blood supply and healing [32]. Therefore, evaluating the effect of 5Fu@ZIF-8 on HUVEC is of great significance for clarifying the biosafety and anti-keloid mechanism of the preparation.

Cytotoxicity and proliferation effect

The normal function of vascular endothelial cells is the basis of skin physiological homeostasis, so the toxicity of drugs to them must be strictly evaluated. When ZIF-8 concentration ranged from 0.1 to 50 µg/mL, the viability of HUVEC remained above 80% after 24–48 h of treatment (Fig. 4A), indicating that the ZIF-8 carrier had low toxicity to vascular endothelial cells and good biocompatibility.

The results of proliferation inhibition of 5Fu and 5Fu@ZIF-8 on HUVEC are shown in Fig. 4B. At 24 h, there was no significant difference in the effect on cell viability between the two at low concentrations (≤ 5 µg/mL). After 48 h, with the increase of concentration (≥ 10 µg/mL), the viability of HUVEC in the 5Fu@ZIF-8 group was slightly lower than that in the 5Fu group. The live/dead cell staining results showed (Fig. 4C) that HUVEC in the Control group and ZIF-8 group were mainly green. The proportion of red in the 5Fu@ZIF-8 group was higher than that in the 5Fu group, further verifying the cytotoxicity of 5Fu@ZIF-8 to HUVEC. The apoptosis staining results showed (Fig. 4D and E) that 5Fu@ZIF-8 significantly induced HUVEC apoptosis. This indicated that after being encapsulated by ZIF-8, 5Fu had an enhanced cytotoxic effect on HUVEC, and while effectively inhibiting fibroblasts, it also had a good killing effect on vascular endothelial cells.

Cell uptake efficiency

In the RhB group, the intracellular red fluorescence of HUVEC was extremely weak at 1, 2, and 6 h. In the RhB@ZIF-8 group, the intracellular red fluorescence gradually increased with the extension of incubation time (1, 2, and 6 h) (Fig. 4F). The quantitative results of flow cytometry (Fig. 4G) further confirmed that the mean fluorescence intensity of HUVEC in the RhB@ZIF-8 group was significantly higher than that in the RhB group. This indicated that while ZIF-8 achieved targeted drug delivery to scar fibroblasts, it could also increase the drug uptake by vascular endothelial cells, thereby enhancing the effect on vascular tissues. This provides a basis for its precise action on pathological and abnormal cells during local keloid delivery.

Similar to the observations in KFs, the higher cytotoxicity of 5Fu@ZIF-8 toward HUVEC is attributed to its enhanced cellular internalization. Free 5Fu, despite its small molecular size, undergoes fast diffusion and rapid efflux, which limits its intracellular accumulation. In contrast, the nanoscale and positively charged 5Fu@ZIF-8 particles are readily taken up via endocytosis, leading to significantly higher intracellular fluorescence intensity of RhB@ZIF-8 than free RhB in HUVEC. This enhanced uptake, together with pH-responsive intracellular release of 5Fu, provides a mechanistic basis for the stronger inhibitory and pro-apoptotic effects of 5Fu@ZIF-8 compared with free 5Fu.

Regulation of 5Fu@ZIF-8 on cell Behavior, ECM, and TGF-β/Smad pathway

The progression of keloids is closely related to fibroblast migration, myofibroblast phenotypic transformation, and excessive activation of the TGF-β/Smad pathway. Therefore, it is necessary to clarify the effect of 5Fu@ZIF-8 on these processes to reveal its in-depth anti-keloid mechanism.

Cell migration

Excessive migration of fibroblasts and abnormal angiogenesis of endothelial cells are core pathological features of keloid. After KFs were treated with 5Fu@ZIF-8, the scratch closure rate at 12 and 24 h was significantly lower than that in the Control group, ZIF-8 group, and 5Fu group (Fig. 5A), suggesting that 5Fu@ZIF-8 could strongly inhibit the migration of KFs. In the scratch assay of HUVEC (Fig. 5B), the scratch closure rate of the 5Fu@ZIF-8 group was also significantly lower than that of the Control group.

The results of the Transwell assay (Fig. 5C and D) were consistent with the trend of the scratch assay. The number of transmembrane cells in the 5Fu@ZIF-8 group was significantly less than that in other groups, directly verifying its strong inhibitory effect on KFs migration ability. The number of transmembrane HUVEC in the 5Fu@ZIF-8 group also decreased significantly. This result is consistent with the pathological mechanism of keloids. The excessive growth of keloids depends on abnormal neovascularization to provide nutrients and proliferation-promoting factors for fibroblasts. Therefore, inhibiting the migration and invasion of vascular endothelial cells and blocking local scar angiogenesis can cut off the proliferative drive of fibroblasts from the nutrient supply level, forming a dual anti-keloid mechanism together with the direct inhibition of the malignant biological behavior of KFs.

ECM regulation

Excessive ECM deposition is the core pathological feature of keloid. Among them, Col 1 is the main collagen component in scar tissue; the imbalance of Col 3 proportion (increased Col 1/Col 3 ratio) leads to scar tissue stiffness. α-SMA is the core marker of myofibroblasts. FN is a key component of ECM. In KFs, the fluorescence intensity of α-SMA in the 5Fu@ZIF-8 group was significantly lower than that in the Control group, ZIF-8 group, and 5Fu group (Fig. 5E), suggesting that 5Fu@ZIF-8 could inhibit the phenotypic transformation of KFs into myofibroblasts. The fluorescence intensity of FN in the 5Fu@ZIF-8 group also decreased significantly (Fig. 5F), indicating that 5Fu@ZIF-8 could reduce the synthesis and deposition of FN. The quantitative detection results of collagen further confirmed the regulatory effect of 5Fu@ZIF-8 on ECM. Compared with the Control group, the secretion of Col 1 and Col 3 in the 5Fu@ZIF-8 group significantly decreased, and the inhibitory effect was better than that in the free 5Fu group and ZIF-8 group (Fig. 5G). This indicated that 5Fu@ZIF-8 not only reduced non-collagen ECM components such as FN but also directly regulated the synthesis of core collagen in keloid, inhibiting the excessive deposition of ECM from multiple dimensions. 5Fu@ZIF-8 blocks the material basis of keloid from the level of cell phenotype and matrix composition.

Regulation of TGF-β/Smad signaling pathway

The TGF-β/Smad pathway is the core signal axis regulating fibroblast activation and collagen synthesis [33]. Excessive activation of the TGF-β/Smad pathway is a key molecular mechanism of keloid occurrence. The Western blot results (Fig. 5H) showed that in KFs, the levels of TGF-β1, P-Smad2, and P-Smad3 in the 5Fu@ZIF-8 group were significantly lower than those in other groups, while there was no significant change in total Smad2 and Smad3. This suggested that 5Fu@ZIF-8 blocked the excessive activation of the TGF-β/Smad pathway by inhibiting TGF-β1 secretion and P-Smad2 and P-Smad3.

Under pathological conditions, excessive activation of KFs, abnormal angiogenesis, and excessive ECM deposition form a vicious cycle. We obtained multi-dimensional results, which cover cell migration, invasion, phenotypic transformation, collagen secretion, and signaling pathway. Based on these results, 5Fu@ZIF-8 comprehensively inhibits keloids, and this inhibition works at the levels of cell function, molecular pathway, and ECM. This comprehensive inhibitory effect relies on a dual synergistic mechanism: targeted multi-link regulation of KFs and blocking of abnormal angiogenesis (Fig. 5I). 5Fu@ZIF-8 directly targets KFs and is efficiently taken up by KFs. 5Fu@ZIF-8 responds to the weakly acidic pH of the scar microenvironment for specific degradation, directly inhibiting the abnormal proliferation of KFs. At the same time, 5Fu@ZIF-8 downregulates TGF-β1 secretion and inhibits the phosphorylation of Smad2 and Smad3, blocking the excessive activation of the TGF-β/Smad pathway. 5Fu@ZIF-8 inhibits the phenotypic transformation of KFs into myofibroblasts, reducing the source of functional cells for ECM synthesis. On the other hand, 5Fu@ZIF-8 downregulates the secretion of key ECM components such as Col 1, Col 3, and FN, fundamentally reducing the matrix deposition and structural abnormalities of keloid. The excessive growth of keloids depends on abnormal neovascularization to provide nutrients and proliferation-promoting signals for KFs. Therefore, 5Fu@ZIF-8 inhibits the migration ability of HUVEC, blocks abnormal angiogenesis in keloid tissue, and cuts off the driving force for KFs proliferation and ECM synthesis from the microenvironmental nutrient supply level, forming a synergistic effect with direct targeting of KFs. In conclusion, 5Fu@ZIF-8 acts on KFs, vascular endothelial cells, and ECM synthesis processes simultaneously, breaks the vicious cycle, and shifts keloid tissue towards a therapeutic state, ultimately achieving precise and efficient anti-scar effects.

Preparation and performance characterization of UV-Responsive hydrogels

Synthesis and macroscopic properties of hydrogels

The process of forming LACS hydrogels via LA-grafted CS and UV-induced photopolymerization is shown in Fig. 6A. CS served as the natural polymer backbone, while LA provided thiol groups for photopolymerization. After mixing, the mixture was rapidly cross-linked into a three-dimensional network under UV irradiation, which could encapsulate 5Fu@ZIF-8 nanoparticles and other free drugs. In the ¹H NMR spectrum (Fig. 6B), LACS showed characteristic proton peaks of both CS and LA. In the FT-IR spectrum (Fig. 6C), LACS exhibited peak superposition or shift at the characteristic wavenumbers of LA and CS, including the thiol peak of LA and the amino/hydroxyl peaks of CS. The combined results of ¹H NMR and FT-IR confirmed that LA was successfully grafted onto the CS chains, providing a chemical basis for the formation of the hydrogel network via photopolymerization.

The macroscopic photographs in Fig. 6D directly verified the practicality of the hydrogel. The liquid precursor could be smoothly injected through a syringe, confirming the hydrogel’s injectability. After UV cross-linking, a stable gel was rapidly formed, demonstrating the hydrogel’s UV-responsive gelation property. The gelled hydrogel could be clamped and stretched, proving that it possessed the mechanical strength required for local retention, which is suitable for minimally invasive local injection administration at scar sites. SEM images showed that the LACS hydrogel had a porous network structure (Fig. 6E). This structure facilitates the encapsulation and dispersion of 5Fu@ZIF-8, provides channels for drug release, and supports cell adhesion or tissue ingrowth.

Rheological, Swelling, Degradation, and drug release behaviors

The rheological properties of hydrogels directly affect their clinical application effects [34]. In Fig. 6F, G’ was consistently higher than G’’ and remained stable with changes in frequency. This indicated that the hydrogel had an elasticity-dominated stable network structure, which could maintain its shape at the keloid site and avoid rapid deformation and loss after injection.

The swelling ratio of the hydrogel gradually increased and leveled off within 24 h, reaching 1986.42 ± 141.74% (Fig. 6G). Its porous structure allowed rapid water absorption, providing a swelling-driven force for drug release. The degradation rate of the hydrogel in the acidic environment of pH 5.5 was significantly faster than that in the physiological environment of pH 7.4 (Fig. 6H). At 6 d, the hydrogel degraded by 82.87 ± 2.05% in the pH 5.5 environment. This pH responsiveness enables the hydrogel to degrade preferentially at the keloid site, achieving targeted drug release. The drug release of the hydrogel showed pH dependence. In the acidic environment, the cumulative release rate was much higher than that in the physiological environment (Fig. 6I). At pH 5.5, the cumulative release amount reached 69.93 ± 4.93% at 4 d. The release behavior of the hydrogel further confirmed that it achieved sustained release at the scar site through swelling and degradation, reducing drug exposure to normal tissues. Table S2 summarizes the kinetic fitting of the in vitro release profiles of the FZRL hydrogel at different pH values using zero order, first order, Higuchi, and Ritger-Peppas models. At pH 7.4, the first order model provided the best fit (R² = 0.99437), with the Ritger-Peppas model showing a similarly high fit (R² = 0.99256; n = 1.19662), suggesting that release is governed by concentration-dependent kinetics with contributions from matrix relaxation/erosion. At pH 6.8, both first-order (R² = 0.99592) and Ritger-Peppas (R² = 0.99447; n = 0.41202) fitted well, indicating a predominantly diffusion-controlled process under mildly acidic conditions. At pH 5.5, the first-order model again showed the highest goodness of fit (R² = 0.99302), and the Ritger-Peppas exponent decreased to n = 0.2834 (R² = 0.99246), consistent with faster release driven by the combined effects of acidic microenvironment-induced matrix changes and diffusion.

Hydrogel biocompatibility

When KFs and HUVEC were cultured on blank LACS hydrogels for 4 d, both showed dense green fluorescence, with no obvious red fluorescence observed (Fig. 6J). This indicated that the LACS hydrogel itself was non-cytotoxic and had a good supporting effect on the survival and proliferation of KFs and HUVEC, providing a guarantee for the biosafety of the delivery system after subsequent drug loading.

The LACS hydrogel has advantages such as injectable gelation, pH-responsive degradation and drug release, porous structure, and good biocompatibility, making it an ideal carrier for local drug delivery at scar sites.

In vivo distribution and keloid penetration of hydrogels

In vivo distribution

To clarify the in vivo targeting and biosafety of the drug delivery system, RhB@ZIF-8 was used. In vivo fluorescence imaging showed (Fig. 7A) that compared with free RhB and RhB@ZIF-8, RhB@ZIF-8@LACS had a longer persistence of fluorescent signal at the keloid site, with significant retention even at 72 h. In contrast, the fluorescence of free RhB decayed rapidly and weakened significantly after 4 h. Quantitative analysis (Fig. 7B) further confirmed that the fluorescence retention efficiency at the keloid site in the RhB@ZIF-8@LACS group was significantly higher than that in other groups. This suggested that the local retention effect of LACS hydrogel synergized with the targeting property of ZIF-8, enabling long-term drug accumulation at the keloid site. The nude mice were sacrificed, and major organs were harvested. Fluorescence imaging of the heart, liver, spleen, lung, and kidney showed (Fig. 7C and D) that the fluorescence intensity of each organ in the RhB@ZIF-8@LACS group was significantly lower than that in the RhB group. This indicated that the drug delivery system could reduce systemic distribution and decrease non-specific toxicity to normal organs.

Keloid tissue penetration depth

Keloid tissues were harvested from nude mice. CLSM observation revealed that the penetration range and depth of red fluorescence in the keloid tissue of the RhB@ZIF-8@LACS group were significantly better than those of the free RhB and RhB@ZIF-8 groups (Fig. 7E). Free RhB only showed weak fluorescence on the keloid surface, while the fluorescence of RhB@ZIF-8@LACS could penetrate into the inner tissue of the keloid. This result confirmed that ZIF-8 nanoparticles combined with LACS hydrogel could penetrate the dense ECM barrier of keloids, enabling efficient drug penetration into the core region of the keloid and providing a histological basis for exerting anti-scar effects.

In vivo biosafety

FZRL hydrogel was subcutaneously injected into nude mice. After 14, 21, and 30 d, major organs, skin, and blood were collected. H&E staining of the heart, liver, spleen, lung, and kidney showed (Fig. S1A) no significant difference in tissue morphology between the FZRL group and the Control group. No pathological changes such as cellular edema, inflammatory infiltration, or necrosis were observed, indicating that FZRL hydrogel had no obvious toxicity to major organs. H&E staining of the skin showed (Fig. S1B) that the epidermal layer of the skin in the FZRL group was intact, and no abnormal inflammatory cell infiltration or tissue damage was found in the dermal layer, which was consistent with the morphology of the control group. The H&E results indicated that local injection of FZRL did not cause skin irritative damage. The detection results of ALT, AST, BUN, and CREA showed (Fig. S1C) no significant difference in biochemical indicators between the FZRL group and the control group, and all indicators were within the normal physiological range. This further verified that FZRL hydrogel had no obvious hepatic or renal toxicity. Through the synergistic effect of local hydrogel retention and nanoparticle-targeted penetration, FZRL hydrogel achieves long-term drug accumulation and deep penetration at the keloid site, while maintaining good biosafety. This provides in vivo evidence for the safety and efficacy of its subsequent application in animal experiments.

In vivo Anti-Scar pharmacodynamic evaluation of FZRL hydrogel

To verify the in vivo anti-keloid efficacy of FZRL hydrogel, a keloid model was established by inoculating KFs into nude mice [35] (Fig. 8A). KFs cultured in vitro were mixed with Matrigel and then inoculated into the dorsal subcutaneous tissue of nude mice. After keloid formation, the nude mice were randomly divided into 6 groups: Control, 5Fu, 5Fu@ZIF-8, Blank ZL, FZL, and FZRL. Local administration was performed, and continuous observation was conducted for 26 d.

In vivo Scar Inhibition effect and body weight changes

During the treatment, mice in the 5Fu group developed local discomfort due to drug stimulation, showing frequent behaviors such as wiping and scratching the keloid site. Some mice even appeared in an agitated state, and in severe cases, mutual biting between cage mates occurred (Fig. 8B). In contrast, after encapsulating ropivacaine via the long-acting sustained-release property of the hydrogel, mice in the FZRL group showed almost no signs of restlessness, and their behaviors were consistent with the normal state. This indicated that ropivacaine could alleviate local pruritus and discomfort at the keloid site. Gross photographs of keloids showed (Fig. 8C) that the keloid tissue in the Control group was large in volume, hard in texture, and significantly raised on the surface. The keloid in the 5Fu group was slightly reduced but still showed significant hyperplasia. However, the FZRL group had the smallest keloid volume, with an appearance closer to normal skin and a soft texture. Quantitative analysis of keloid volume further confirmed that compared with the Control group, the keloid volume in the FZRL group was significantly reduced from the 14 d after inoculation (Fig. 8D). At 26 d, the keloid volume in the FZRL group was much smaller than that in the Control group, and the inhibitory effect was significantly better than that in the 5Fu@ZIF-8 group and Blank ZL group (Fig. 8E). The average keloid volume was reduced by 54.7 ± 12.1%. Pharmacodynamic experiments showed that FZRL hydrogel achieved significant inhibition of in vivo keloid growth through the synergistic effects of 5Fu (inhibiting scar hyperplasia), ropivacaine (alleviating local discomfort), and carrier-controlled release (enhancing efficacy). Meanwhile, it improved the behavioral comfort of animals during scar treatment.

Changes in mouse body weight are a core indicator reflecting the systemic toxicity of drugs. The weight growth trend of mice in each administration group was consistent with that of the Control group over time (Fig. 8F). At 26 d, there was no significant difference in body weight between the FZRL group and the Control group, and all weights were within the normal physiological range (Fig. 8G). This indicated that local administration of FZRL hydrogel did not cause systemic toxicity in mice, demonstrating good in vivo safety.

Histological and apoptosis analysis

Histopathological detection was performed on the isolated keloid tissues. H&E staining results showed (Fig. 8H) that in the Control group, the collagen fibers in the keloid were disorderly arranged, dense, and had a large thickness. The arrangement of collagen fibers in the 5Fu group was slightly improved, but local hyperplasia still existed. The FZRL group showed a significant reduction in keloid tissue thickness, and the collagen fibers were arranged more loosely, close to the network structure of normal skin. This indicated that FZRL could effectively improve the pathological morphology of keloid tissue. TUNEL staining is shown in Fig. 8I. Apoptotic cells were stained brown, and the proportion of apoptotic cells in the FZRL group was significantly higher than that in other groups. This result was consistent with the conclusion that the delivery system induces KFs apoptosis in in vitro cell experiments, suggesting that FZRL can achieve anti-scar effects in vivo by inducing apoptosis of scar fibroblasts.

ECM regulation

The excessive hyperplasia of keloid tissue is closely related to the abnormal deposition of Col 1 and Col 3. As key components of ECM, α-SMA and FN also participate in the structural remodeling of scars. The expression changes of each component were quantitatively evaluated by immunofluorescence staining to further reveal the anti-keloid mechanism of FZRL hydrogel. FZRL hydrogel can block the pathological progression of keloid from multiple dimensions. In the keloid tissue of the Control group, Col 1 (Fig. 9A, red fluorescence) and Col 3 (Fig. 9B, green fluorescence) showed dense and strongly positive signal distribution; α-SMA (Fig. 9C, red fluorescence) was highly expressed in bundles; and FN (Fig. 9D, green fluorescence) was also widely deposited. Compared with the Control group, 5Fu group, 5Fu@ZIF-8 group, and other groups, the fluorescence intensity and positive rate of Col 1, Col 3, α-SMA, and FN in the FZRL group were significantly reduced. This indicated that FZRL can inhibit the phenotypic transformation of fibroblasts into myofibroblasts and simultaneously reduce the synthesis and deposition of ECM components such as Col 1, Col 3, and FN. It comprehensively blocks the pathological remodeling process of keloid tissue through the core links of myofibroblast activation and ECM accumulation.

Fig. 9 — Regulatory effect of FZRL hydrogel on ECM in keloid tissues. (A) Col 1 immunofluorescence staining of keloids in each group (scale bar: 250 μm) and quantitative analysis of Col 1 positive rate (n = 3). (B) Col 3 immunofluorescence staining of keloids in each group (scale bar: 250 μm) and quantitative analysis of Col 3 positive rate (n = 3). (C) α-SMA immunofluorescence staining of keloids in each group (scale bar: 250 μm) and quantitative analysis of α-SMA positive rate (n = 3). (D) FN immunofluorescence staining of keloids in each group (scale bar: 250 μm) and quantitative analysis of FN positive rate (n = 3)

In vivo regulation of the TGF-β/Smad signaling pathway

The TGF-β/Smad pathway is the core signal axis regulating the activation of scar fibroblasts and collagen synthesis, and its excessive activation is a key molecular mechanism of keloid occurrence. Immunofluorescence results showed that in the keloid tissue of the Control group, the red fluorescent signal of TGF-β1 was dense and strong (Fig. 10A). After treatment with FZRL, the positive rate of TGF-β1 was significantly reduced, and the inhibitory effect was better than that in the 5Fu group, 5Fu@ZIF-8 group, and other groups. This suggested that FZRL can reduce the secretion of TGF-β1 at the keloid site. Western blot results showed that the expression level of TGF-β1 protein was relatively high in the Control group, and P-Smad2 and P-Smad3 were significantly upregulated (Fig. 10B). In the FZRL group, the expression of TGF-β1 was significantly decreased, and the ratios of P-Smad2/Smad2 and P-Smad3/Smad3 were also significantly inhibited. This indicated that FZRL can block the excessive activation of the TGF-β/Smad pathway in vivo by downregulating the secretion of TGF-β1 and inhibiting the phosphorylation activation of Smad2/3, which is consistent with the mechanism of in vitro cell experiments. This further reveals the core anti-keloid mechanism of FZRL from the perspective of molecular pathways. By targeting and blocking the TGF-β/Smad signaling pathway, FZRL hydrogel inhibits the phenotypic transformation of myofibroblasts and ECM deposition, ultimately achieving significant anti- keloid efficacy in vivo (Fig. 10C).

Fig. 10 — Study on the molecular mechanism of FZRL hydrogel regulating the TGF-β/Smad signaling pathway. (A) TGF-β1 immunofluorescence staining of keloids in different treatment groups (scale bar: 250 μm) and quantitative analysis of positive rate (n = 3). (B) Western blot detection and quantitative analysis of TGF-β1, p-Smad2, Smad2, p-Smad3, and Smad3 protein expression in keloid tissues of each group (n = 3). (C) Schematic diagram of the mechanism by which FZRL hydrogel exerts anti-keloid effect by inhibiting the TGF-β/Smad pathway

Discussion

The treatment of keloids has long been constrained by the core contradiction between their complex pathogenesis and the limitations of single-drug therapy [36]. Current approaches mainly target fibroblast proliferation or angiogenesis, but rarely achieve comprehensive regulation of the pathological cascade involving abnormal KFs activation, neovascularization, and excessive ECM deposition. Additionally, patient discomfort and surgical complexity are often overlooked, leading to poor treatment stability and limited clinical translation. In this study, based on the core pathological features of disordered collagen metabolism and excessive angiogenesis revealed by proteomics of human keloid tissues, we designed a multifunctional FZRL delivery system. This strategy integrates carrier responsiveness, multi-drug synergy, and patient-centered design, realizing an upgrade of anti-scar strategies from single-link inhibition to precise regulation of multiple pathological nodes.

5Fu serves as the core therapeutic drug due to its dual pathological effects. Beyond inhibiting KFs proliferation, α-SMA expression, and ECM deposition, 5Fu also directly suppresses HUVEC activity and migration, thereby cutting off the angiogenic supply that sustains keloid progression. This dual mechanism precisely corresponds to the proteomic signature of simultaneous upregulation of collagen and vascular-related proteins, offering advantages over the reported single-target effects [37, 38]. For effective drug delivery, nanoparticles must penetrate the interior of solid lesions and interact with different pathological components in the microenvironment [39]. The incorporation of pH-responsive ZIF-8 nanoparticles further localizes drug release within the acidic keloid microenvironment, mitigating off-target toxicity. At the signal transduction level, FZRL downregulates the TGF-β/Smad pathway, coordinating the inhibition of ECM synthesis and angiogenesis. These findings are confirmed by TUNEL and ECM staining, verifying tissue-level regulatory pathways beyond phenotypic inhibition. Consistent with recent advances in multifunctional biomaterials, multi-interaction platforms can orchestrate cell-matrix signaling, vascular cues, and immune homeostasis to promote tissue regeneration [40–42]. The results demonstrate that FZRL exerts full-chain regulatory effects across signaling pathways, cellular behaviors, and tissue morphology, addressing the limitation of traditional therapies that only inhibit phenotypes but fail to interfere with the underlying mechanisms.

To address the challenges in local keloid treatment, various local delivery strategies have been proposed [43–47]. Microneedles (MNs) are a widely studied system. They mechanically disrupt the skin barrier by creating transient microchannels, significantly enhancing transdermal drug delivery efficiency [48]. However, drugs primarily rely on passive diffusion to enter tissues, lacking sustained retention and precise release. Moreover, needle length and retention capacity directly affect therapeutic efficacy [49]. Elisa Vettorat et al. innovatively utilized the skin microincision technique combined with MNs to enhance drug penetration [50]. Nevertheless, skin microincisions and MN operations may induce local inflammation, which instead exacerbates scar activation, and patient compliance remains limited. In contrast, injectable hydrogels offer advantages of local drug retention and controlled release, but different systems still have limitations. pH-responsive hydrogels lack controllable external triggering mechanisms [24]. Thermosensitive hydrogels are prone to premature gelation during injection, leading to uneven distribution and limited operability [51, 52]. The polyamino acid hydrogels we previously developed exhibit certain thermosensitive properties, yet their gelation rate and mechanical stability are easily affected by in vivo temperature fluctuations. Drug release still predominantly relies on passive diffusion, making it difficult to achieve precise spatiotemporal control [25]. The UV-responsive LACS hydrogel constructed in this study overcomes the aforementioned limitations. Its thiol-disulfide bond cross-linking mechanism ensures the precursor remains in a liquid state before injection. After injection and localization around the keloid, rapid gelation can be achieved on demand via UV irradiation, enabling spatiotemporal control over drug release. This design avoids the risk of premature gelation and eliminates the need for strict temperature control during clinical operations. Medical personnel can flexibly adjust the injection depth and gelation range according to the keloid location. Furthermore, the LACS hydrogel possesses excellent rheological properties, maintaining a stable morphology even in high-tension skin areas, which significantly enhances local retention and controllability of therapeutic efficacy.

Importantly, this work incorporated patient comfort into keloid therapy design, an aspect often overlooked. Clinical data show that approximately 70% of scar patients scratch the scar or discontinue treatment due to inability to tolerate such discomfort, which indirectly leads to reduced stability of therapeutic efficacy. Even our team’s previous studies failed to incorporate patient comfort into design considerations, resulting in a clinical dilemma where efficacy is disconnected from compliance [53, 54]. By co-delivering ropivacaine and 5Fu, FZRL achieves both antiproliferative and analgesic effects in a single administration. This simplifies the treatment regimen and reduces the local irritation caused by ropivacaine, further enhancing safety and compliance. This integration bridges the long-standing therapeutic gap between efficacy, safety, and comfort, improving the translational relevance of keloid treatment.

This study still has certain limitations. The keloid model established by subcutaneous inoculation of KFs in nude mice provides a useful simulation of keloid, but it differs from human orthotopic scars in several important ways. Human orthotopic keloids are influenced by a more complex microenvironment that includes inflammatory responses, immune cell infiltration, and nerve regeneration during their growth, all of which contribute to the disease’s pathology and response to treatments. Therefore, to better predict clinical efficacy, future studies should aim to incorporate humanized or orthotopic animal models that more closely resemble human keloid pathophysiology. While short-term safety has been confirmed in our mouse model, long-term biocompatibility and systemic toxicity must be systematically evaluated in larger animal models before clinical application. Such evaluation should include dynamic monitoring of serum Zn²⁺. The potential differences between human and mouse keloid tissue, such as variations in skin thickness, microvascular structure, and immune responses, need to be carefully considered. To translate the FZRL hydrogel to human-scale keloid therapy, targeted adaptations are essential. Tuning hydrogel cross-linking density will enhance mechanical strength to withstand the variable tension of human skin. Surface modification of nanoparticles or incorporation of targeting ligands can improve delivery specificity in human tissues. Additionally, optimizing 5Fu release kinetics is critical for clinical translation, accounting for differences in pH and inflammatory microenvironments between human and mouse skin. Large-animal studies (porcine) will be invaluable to address these challenges and bridge the gap to human trials.

Conclusion

Addressing the clinical pain points of keloids, namely disorders in multiple pathological links, low bioavailability, and low patient compliance, this study designed and constructed the FZRL hydrogel. By synergistically integrating 5Fu’s dual-targeted regulation of the pathological cycle, ZIF-8’s pH-responsive precise drug release, LACS’s UV responsiveness, and ropivacaine’s role in enhancing treatment comfort, this design achieves multi-dimensional innovation in anti-keloid strategies. FZRL hydrogel demonstrates significant advantages in mechanism targeting, carrier controllability, and clinical adaptability, breaking through the limitations of existing single therapeutic strategies. This design provides a new preparation with both efficacy and safety for the precision treatment of keloids, and holds important clinical transformation value.

Supplementary Information

Supplementary Material 1^{(18.3MB, docx)}

Author contributions

J.J. and C.Z. contributed equally to this work. J.J. and C.Z. designed and performed the main experiments (including hydrogel synthesis, 5Fu@ZIF-8 preparation, and in vitro biocompatibility tests), analyzed the experimental data, and drafted the main manuscript text. N.W. and M.J. conducted the animal experiments (nude mouse keloid model establishment, treatment, and scar volume measurement) and collected pathological analysis data. Y.J. performed the transcriptome analysis of clinical keloid samples. H.J. and S.Z. assisted in characterizing the physicochemical properties of the nanocomposite hydrogel. Y.Z. and L.Z. helped verify the TGF-β/Smad pathway-related protein expression by Western blot. Y.L. participated in data visualization and figure preparation. Z.G., Zh.J., and S.W. conceived the research project, supervised the experimental process, revised the manuscript critically for important intellectual content, and secured funding support. All authors reviewed the manuscript and approved the final version for submission.

Funding

This work was supported by the Fundamental Research Funds for the Central Universities, Peking Union Medical College (3332025063), the National Natural Science Foundation of China (82260617, 82073778, 82104106) and the CAMS Innovation Fund for Medical Sciences (CIFMS) (2021-I2M-1-026, China).

Data availability

Data is provided within the manuscript files.

Declarations

Ethics approval and consent to participate

All animal experiments were approved by the Laboratory Animal Ethics Committee in the Institute of Materia Medica and Peking Union Medical College (00000761). All procedures followed ethical standards during the experiment. This project has been approved by the Ethics Committee of Scientific Research/Cell Clinical Research of Yanbian University Hospital (20250030). All patients signed the informed consent form.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Jingyu Jin and Chengjie Zhu contributed equally to this work.

Contributor Information

Zhonggao Gao, Email: zggao@imm.ac.cn.

Zhehu Jin, Email: jinzh_621@163.com.

Shuangqing Wang, Email: wangshuangqing@imm.ac.cn.

References

1.Jeschke MG, Wood FM, Middelkoop E, Bayat A, Teot L, Ogawa R, et al. Scars. Nat Rev Dis Primers. 2023;9:64. 10.1038/s41572-023-00474-x. [DOI] [PubMed] [Google Scholar]
2.Menashe S, Heller L. Keloid and hypertrophic scars treatment. Aesthet Surg J. 2024;48:2553–60. 10.1007/s00266-024-03869-7. [DOI] [PubMed] [Google Scholar]
3.Liu S, Yang H, Song J, Zhang Y, Abualhssain ATH, Yang B. Keloid: genetic susceptibility and contributions of genetics and epigenetics to its pathogenesis. Exp Dermatol. 2022;31:1665–75. 10.1111/exd.14671. [DOI] [PubMed] [Google Scholar]
4.Ding J-Y, Sun L, Zhu Z-H, Wu X-C, Xu X-L, Xiang Y-W. Nano drug delivery systems: a promising approach to scar prevention and treatment. J Nanobiotechnology. 2023;21:268. 10.1186/s12951-023-02037-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Kumar AS, Kamalasanan K. Drug delivery to optimize angiogenesis imbalance in keloid: a review. J Control Release. 2021;329:1066–76. 10.1016/j.jconrel.2020.10.035. [DOI] [PubMed] [Google Scholar]
6.Deng Z, Fan T, Xiao C, Tian H, Zheng Y, Li C, et al. TGF-β signaling in health, disease, and therapeutics. Signal Transduct Target Ther. 2024;9:61. 10.1038/s41392-024-01764-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Liu C, Khairullina L, Qin Y, Zhang Y, Xiao Z. Adipose stem cell exosomes promote mitochondrial autophagy through the PI3K/AKT/mTOR pathway to alleviate keloids. Stem Cell Res Ther. 2024;15:305. 10.1186/s13287-024-03928-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Cho MY, Song A, Chung KY, Roh MR. A novel method of steroid delivery to improve the efficacy of intralesional injection in keloid treatment. Dermatol Surg. 2022;48:631–5. 10.1097/DSS.0000000000003432. [DOI] [PubMed] [Google Scholar]
9.Dange YD, Salunkhe VR, Honmane SM, Marale PS. Anticancer efficacy of 5-Fluorouracil-Loaded Chitin nanohydrogel in enhanced skin cancer therapy. Bio I. 2025;6. 10.15212/bioi-2025-0090.
10.Jiang J, Kaysar K, Pan Y, Xia L, Li J. A zeolitic imidazolate framework-based antimicrobial peptide delivery system with enhanced anticancer activity and low systemic toxicity. Pharmaceutics. 2024;16:1591. 10.3390/pharmaceutics16121591. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Zhao X, Chen Z, Zhang S, Hu Z, Shan J, Wang M, et al. Application of metal-organic frameworks in infectious wound healing. J Nanobiotechnol. 2024;22:387. 10.1186/s12951-024-02637-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Liu Y, Wu H, Wang S, Zhang X, Gong L, Xiao C, Liu C, Chen L, Zhao H, Liu C, Jin M, Gao Z, Huang W. Biomimetic multifunctional nanodrugs enable regulating abnormal tumor metabolism and amplifying PDT-induced immunotherapy for synergistically enhanced tumor ablation. Mater Today. 2023;68:125–47. 10.1016/j.mattod.2023.07.003. [Google Scholar]
13.Wu H, Jin M, Liu Y, Wang S, Liu C, Quan X, et al. A self-targeting MOFs nanoplatform for treating metastatic triple-negative breast cancer through tumor microenvironment remodeling and chemotherapy potentiation. Int J Pharm. 2024;664:124625. 10.1016/j.ijpharm.2024.124625. [DOI] [PubMed] [Google Scholar]
14.Padya BS, Fernandes G, Hegde S, Kulkarni S, Pandey A, Deshpande PB, et al. Targeted delivery of 5-Fluorouracil and sonidegib via surface-modified ZIF-8 MOFs for effective basal cell carcinoma therapy. Pharmaceutics. 2023;15:2594. 10.3390/pharmaceutics15112594. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Hao J, Chen C, Pavelic K, Ozer F. ZIF-8 as a pH-responsive nanoplatform for 5-Fluorouracil delivery in the chemotherapy of oral squamous cell carcinoma. Int J Mol Sci. 2024;25:9292. 10.3390/ijms25179292. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Yang X, Guo J-L, Han J, Si R-J, Liu P-P, Zhang Z-R, et al. Chitosan hydrogel encapsulated with LL-37 peptide promotes deep tissue injury healing in a mouse model. Mil Med Res. 2020;7:20. 10.1186/s40779-020-00249-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Zhang M-D, Huang X, Li Z, Song W, Kong Y, Zhang C, et al. White-light-induced synthesis of injectable alginate-based composite hydrogels for rapid hemostasis. Mil Med Res. 2023;10:47. 10.1186/s40779-023-00483-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Nishad M, Verma S, Kumar A, Khurana N. Formulation and evaluation of apigenin-loaded microsponge gel for effective angioedema therapy. Curr Pharm Anal. 2025;21:203–15. 10.1016/j.cpan.2025.03.001. [Google Scholar]
19.Wang H, Li X, Li M, Wang S, Zuo A, Guo J. Bioadhesion design of hydrogels: adhesion strategies and evaluation methods for biological interfaces. J Adhes Sci Technol. 2023;37:335–69. 10.1080/01694243.2021.2020502. [Google Scholar]
20.Lu Y, Chen B, Lou X, Li Y, Wu J, Yao H, et al. Ibuprofen conjugated epsilon-poly-l-lysine methacrylate hydrogel modulates macrophage polarization and mitigates inflammation in vivo. Biomacromolecules. 2025;26:4464–76. 10.1021/acs.biomac.5c00479. [DOI] [PubMed] [Google Scholar]
21.Lin C-W, Liu T-H, Chen V, Chuang E-Y, Fan Y-J, Yu J. Synergistic potential of gellan gum methacrylate and keratin hydrogel for visceral hemostasis and skin tissue regeneration. Mater Today Bio. 2024;27:101146. 10.1016/j.mtbio.2024.101146. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Song Y, Xing J, Ren L, Xu X, Han D, Xu H, et al. Preparation of multi-functional quaternary ammonium chitosan/surfactin hydrogel and its application in wound management. Macromol Biosci. 2023;23:e2300166. 10.1002/mabi.202300166. [DOI] [PubMed] [Google Scholar]
23.Jia S, Zhou H, Chen J, Lin J, Zhu X, Weng J, et al. A comprehensive review of hydrogel strategies for repairing peripheral nerve injuries. Brain-X. 2025;3:e70012. 10.1002/brx2.70012. [Google Scholar]
24.Quan Y, Shao H, Wang N, Gao Z, Jin M. Microenvironment-sensitive hydrogels as promising drug delivery systems for co-encapsulating microbial homeostasis probiotics and anti-inflammatory drugs to treat periodontitis. Mater Today Bio. 2025;32:101711. 10.1016/j.mtbio.2025.101711. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Sheng F, Luo Y, Wu D, Yuan J, Zheng G, Zheng Y, et al. Poly amino acid thermosensitive hydrogel loaded with ICG-001 for inhibiting keloid by down-regulating the Wnt/β-Catenin pathway. Mater Des. 2022;222:111050. 10.1016/j.matdes.2022.111050. [Google Scholar]
26.Dong D, Mao L, Qin Z, Guo Y, Yu J, Hu X, et al. A simple and effective hydrogel dressing for advanced management of full-thickness skin wound by multi-functional strategies. J Nanobiotechnol. 2025;23:745. 10.1186/s12951-025-03803-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Mardhiah J, Halim AS, Heng S, Saipolamin AQ. Scoping review for pain mitigation during intralesional injections of corticosteroid for hypertrophic scar and keloid treatment. BMJ Open. 2025;15:e092800. 10.1136/bmjopen-2024-092800. [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Yang E, Xu R, Zhang H, Xia W, Huang X, Zan T. Deciphering pain and pruritus in keloids from the perspective of neurological dysfunction: where are we now? Biomedicines. 2025;13:663. 10.3390/biomedicines13030663. [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Li Y, Chen Y, Xue Y, Jin J, Xu Y, Zeng W, et al. Injectable hydrogel delivery system with high drug loading for prolonging local anesthesia. Adv Sci. 2024;11:e2309482. 10.1002/advs.202309482. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Wang S-Q, Jin M-J, Guo Z-K, Shen D-R, Jin L-N, Cheng F, et al. Trace element-dictated exosome modules and self-adaptive dual-network hydrogel orchestrate diabetic foot regeneration through complement-mitochondria-autophagy circuitry. Mil Med Res. 2025;12:71. 10.1186/s40779-025-00658-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Li X, Dou N, Zhong L, Wu Y, Cai Z, Zhao Z, et al. mPEG@ELA-11 alleviates atherosclerosis via AKT-ER stress-mediated macrophage modulation. BME Front. 2025;6:0203. 10.34133/bmef.0203. [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Wang S, Liu Y, Wang X, Chen L, Huang W, Xiong T, Wang N, Guo J, Gao Z, Jin M. Modulating macrophage phenotype for accelerated wound healing with chlorogenic acid-loaded nanocomposite hydrogel. J Control Release. 2024;369:420–43. 10.1016/j.jconrel.2024.03.054. [DOI] [PubMed] [Google Scholar]
33.Liu J, Pan Z, Khan A, Li H. Targeted drug delivery system for pulmonary fibrosis: design and development of biomaterials. Bio I. 2025;6:1–23. 10.15212/bioi-2025-0016. [Google Scholar]
34.Wang S, Liu Y, Sun Q, Zeng B, Liu C, Gong L, et al. Triple cross-linked dynamic responsive hydrogel loaded with selenium nanoparticles for modulating the inflammatory microenvironment via PI3K/Akt/NF-κB and MAPK signaling pathways. Adv Sci. 2023;10:e2303167. 10.1002/advs.202303167. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Zhang Y, Bao Y, Li Z, Jin Z. A novel nude mice model for studying the pathogenesis of keloid. J Cosmet Dermatol. 2025;24:e70284. 10.1111/jocd.70284. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Shi X, Xia X, Xiao Y, Shu H, Xu Z, Liu M, et al. Ferroptosis-resistant adipocytes drive keloid pathogenesis via GPX4-mediated adipocyte-mesenchymal transition and iron-cystine metabolic communication. Int J Biol Sci. 2025;21:5097–115. 10.7150/ijbs.114930. [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Lv W, Wu Y, Chen H. Orthogonal upconversion nanocarriers for combined photodynamic therapy and precisely triggered gene silencing in combating keloids. J Control Release. 2025;379:1–13. 10.1016/j.jconrel.2024.12.080. [DOI] [PubMed] [Google Scholar]
38.Zhuang B-Y, Hu F-C, Gao X, Leng Q, Zhang Y, You Y. Development of a simvastatin-loaded copolymer acid-sensitive nanocarrier and its experimental use in the treatment of keloids. J Cosmet Dermatol. 2025;24:e16573. 10.1111/jocd.16573. [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Chan WCW. Principles of nanoparticle delivery to solid tumors. BME Front. 2023;4:0016. 10.34133/bmef.0016. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Pan W-C, Lin Y-H, Iao HM, Chang Y-H, Chen Y-H, Liu H-C, et al. In situ magnetoelectric generation of miRNA sponges and wireless electric stimulus by conductive granular scaffolds for nerve regeneration. Adv Mater. 2025;37:e2500650. 10.1002/adma.202500650. [DOI] [PubMed] [Google Scholar]
41.Iao HM, Chen C-Y, Lin Y-H, Pan W-C, Liang C-Y, Liu H-C, et al. Wireless in situ catalytic electron signaling-mediated transcriptomic reprogramming for neuron regeneration via adaptable antennas. Adv Sci. 2025;12:e2504786. 10.1002/advs.202504786. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Chan Y-C, Lin Y-H, Liu H-C, Hsu R-S, Chiang M-R, Wang L-W, et al. In situ magnetoelectric generation of nitric oxide and electric stimulus for nerve therapy by wireless chargeable molybdenum carbide octahedrons. Nano Today. 2023;51:101935. 10.1016/j.nantod.2023.101935. [Google Scholar]
43.Zhao Y, Wu S, Cai Y, Yang H, Dong X, Yang B, et al. Integration of finite element simulations with 3D printing technology for personalized chitin/PLA microneedle-based drug delivery systems in thoracic keloid treatment. Int J Biol Macromol. 2025;315:144487. 10.1016/j.ijbiomac.2025.144487. [DOI] [PubMed] [Google Scholar]
44.Bao Z, Gao P, Xia G, Wang Z, Kong M, Feng C, et al. A thermosensitive hydroxybutyl chitosan hydrogel as a potential co-delivery matrix for drugs on keloid inhibition. J Mater Chem B. 2016;4:3936–44. 10.1039/c6tb00378h. [DOI] [PubMed] [Google Scholar]
45.Li J, Fu R, Li L, Yang G, Ding S, Zhong Z, Zhou S. Co-delivery of dexamethasone and green tea polyphenols using electrospun ultrafine fibers for effective treatment of keloid. Pharm Res. 2014;31:1632–43. 10.1007/s11095-013-1266-2. [DOI] [PubMed] [Google Scholar]
46.Gao Y, Cheng X, Wang Z, Wang J, Gao T, Li P, et al. Transdermal delivery of 10,11-methylenedioxycamptothecin by hyaluronic acid based nanoemulsion for inhibition of keloid fibroblast. Carbohydr Polym. 2014;112:376–86. 10.1016/j.carbpol.2014.05.026. [DOI] [PubMed] [Google Scholar]
47.Aoki M, Matsumoto NM, Dohi T, Kuwahawa H, Akaishi S, Okubo Y, et al. Direct delivery of apatite nanoparticle-encapsulated SiRNA targeting TIMP-1 for intractable abnormal scars. Mol Ther Nucleic Acids. 2020;22:50–61. 10.1016/j.omtn.2020.08.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Moawad F, Pouliot R, Brambilla D. Dissolving microneedles in transdermal drug delivery: a critical analysis of limitations and translation challenges. J Control Release. 2025;383:113794. 10.1016/j.jconrel.2025.113794. [DOI] [PubMed] [Google Scholar]
49.Zhuang Z-M, Wang Y, Feng Z-X, Lin X-Y, Wang Z-C, Zhong X-C, et al. Targeting diverse wounds and scars: recent innovative bio-design of microneedle patch for comprehensive management. Small. 2024;20:e2306565. 10.1002/smll.202306565. [DOI] [PubMed] [Google Scholar]
50.Vettorato E, Volonté P, Musazzi UM, Cilurzo F, Casiraghi A. Skin microincision technique to enhance drug penetration for the treatment of keloid and hypertrophic scars. Int J Pharm. 2025;671:125259. 10.1016/j.ijpharm.2025.125259. [DOI] [PubMed] [Google Scholar]
51.Chen H, Xu J, Sun J, Jiang Y, Zheng W, Hu W, et al. Recent advances on thermosensitive hydrogels-mediated precision therapy. Asian J Pharm Sci. 2024;19:100911. 10.1016/j.ajps.2024.100911. [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Yew PYM, Lin Q, Owh C, Chee PL, Loh XJ. Current research and future potential of thermogels for sustained drug delivery. Expert Opin Drug Deliv. 2025;22:769–86. 10.1080/17425247.2025.2486350. [DOI] [PubMed] [Google Scholar]
53.Wan H, Wang S, Li C, Zeng B, Wu H, Liu C, et al. LA67 liposome-loaded thermo-sensitive hydrogel with active targeting for efficient treatment of keloid via peritumoral injection. Pharmaceutics. 2023;15:2157. 10.3390/pharmaceutics15082157. [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Wu D, Liu X, Jin Z. Placental mesenchymal stem cells-secreted proenkephalin suppresses the p38 MAPK signaling to block hyperproliferation of keloid fibroblasts. Tissue Cell. 2023;85:102218. 10.1016/j.tice.2023.102218. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1^{(18.3MB, docx)}

Data Availability Statement

Data is provided within the manuscript files.

[CR1] 1.Jeschke MG, Wood FM, Middelkoop E, Bayat A, Teot L, Ogawa R, et al. Scars. Nat Rev Dis Primers. 2023;9:64. 10.1038/s41572-023-00474-x. [DOI] [PubMed] [Google Scholar]

[CR2] 2.Menashe S, Heller L. Keloid and hypertrophic scars treatment. Aesthet Surg J. 2024;48:2553–60. 10.1007/s00266-024-03869-7. [DOI] [PubMed] [Google Scholar]

[CR3] 3.Liu S, Yang H, Song J, Zhang Y, Abualhssain ATH, Yang B. Keloid: genetic susceptibility and contributions of genetics and epigenetics to its pathogenesis. Exp Dermatol. 2022;31:1665–75. 10.1111/exd.14671. [DOI] [PubMed] [Google Scholar]

[CR4] 4.Ding J-Y, Sun L, Zhu Z-H, Wu X-C, Xu X-L, Xiang Y-W. Nano drug delivery systems: a promising approach to scar prevention and treatment. J Nanobiotechnology. 2023;21:268. 10.1186/s12951-023-02037-4. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR5] 5.Kumar AS, Kamalasanan K. Drug delivery to optimize angiogenesis imbalance in keloid: a review. J Control Release. 2021;329:1066–76. 10.1016/j.jconrel.2020.10.035. [DOI] [PubMed] [Google Scholar]

[CR6] 6.Deng Z, Fan T, Xiao C, Tian H, Zheng Y, Li C, et al. TGF-β signaling in health, disease, and therapeutics. Signal Transduct Target Ther. 2024;9:61. 10.1038/s41392-024-01764-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR7] 7.Liu C, Khairullina L, Qin Y, Zhang Y, Xiao Z. Adipose stem cell exosomes promote mitochondrial autophagy through the PI3K/AKT/mTOR pathway to alleviate keloids. Stem Cell Res Ther. 2024;15:305. 10.1186/s13287-024-03928-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR8] 8.Cho MY, Song A, Chung KY, Roh MR. A novel method of steroid delivery to improve the efficacy of intralesional injection in keloid treatment. Dermatol Surg. 2022;48:631–5. 10.1097/DSS.0000000000003432. [DOI] [PubMed] [Google Scholar]

[CR9] 9.Dange YD, Salunkhe VR, Honmane SM, Marale PS. Anticancer efficacy of 5-Fluorouracil-Loaded Chitin nanohydrogel in enhanced skin cancer therapy. Bio I. 2025;6. 10.15212/bioi-2025-0090.

[CR10] 10.Jiang J, Kaysar K, Pan Y, Xia L, Li J. A zeolitic imidazolate framework-based antimicrobial peptide delivery system with enhanced anticancer activity and low systemic toxicity. Pharmaceutics. 2024;16:1591. 10.3390/pharmaceutics16121591. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR11] 11.Zhao X, Chen Z, Zhang S, Hu Z, Shan J, Wang M, et al. Application of metal-organic frameworks in infectious wound healing. J Nanobiotechnol. 2024;22:387. 10.1186/s12951-024-02637-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR12] 12.Liu Y, Wu H, Wang S, Zhang X, Gong L, Xiao C, Liu C, Chen L, Zhao H, Liu C, Jin M, Gao Z, Huang W. Biomimetic multifunctional nanodrugs enable regulating abnormal tumor metabolism and amplifying PDT-induced immunotherapy for synergistically enhanced tumor ablation. Mater Today. 2023;68:125–47. 10.1016/j.mattod.2023.07.003. [Google Scholar]

[CR13] 13.Wu H, Jin M, Liu Y, Wang S, Liu C, Quan X, et al. A self-targeting MOFs nanoplatform for treating metastatic triple-negative breast cancer through tumor microenvironment remodeling and chemotherapy potentiation. Int J Pharm. 2024;664:124625. 10.1016/j.ijpharm.2024.124625. [DOI] [PubMed] [Google Scholar]

[CR14] 14.Padya BS, Fernandes G, Hegde S, Kulkarni S, Pandey A, Deshpande PB, et al. Targeted delivery of 5-Fluorouracil and sonidegib via surface-modified ZIF-8 MOFs for effective basal cell carcinoma therapy. Pharmaceutics. 2023;15:2594. 10.3390/pharmaceutics15112594. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.Hao J, Chen C, Pavelic K, Ozer F. ZIF-8 as a pH-responsive nanoplatform for 5-Fluorouracil delivery in the chemotherapy of oral squamous cell carcinoma. Int J Mol Sci. 2024;25:9292. 10.3390/ijms25179292. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR16] 16.Yang X, Guo J-L, Han J, Si R-J, Liu P-P, Zhang Z-R, et al. Chitosan hydrogel encapsulated with LL-37 peptide promotes deep tissue injury healing in a mouse model. Mil Med Res. 2020;7:20. 10.1186/s40779-020-00249-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR17] 17.Zhang M-D, Huang X, Li Z, Song W, Kong Y, Zhang C, et al. White-light-induced synthesis of injectable alginate-based composite hydrogels for rapid hemostasis. Mil Med Res. 2023;10:47. 10.1186/s40779-023-00483-7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR18] 18.Nishad M, Verma S, Kumar A, Khurana N. Formulation and evaluation of apigenin-loaded microsponge gel for effective angioedema therapy. Curr Pharm Anal. 2025;21:203–15. 10.1016/j.cpan.2025.03.001. [Google Scholar]

[CR19] 19.Wang H, Li X, Li M, Wang S, Zuo A, Guo J. Bioadhesion design of hydrogels: adhesion strategies and evaluation methods for biological interfaces. J Adhes Sci Technol. 2023;37:335–69. 10.1080/01694243.2021.2020502. [Google Scholar]

[CR20] 20.Lu Y, Chen B, Lou X, Li Y, Wu J, Yao H, et al. Ibuprofen conjugated epsilon-poly-l-lysine methacrylate hydrogel modulates macrophage polarization and mitigates inflammation in vivo. Biomacromolecules. 2025;26:4464–76. 10.1021/acs.biomac.5c00479. [DOI] [PubMed] [Google Scholar]

[CR21] 21.Lin C-W, Liu T-H, Chen V, Chuang E-Y, Fan Y-J, Yu J. Synergistic potential of gellan gum methacrylate and keratin hydrogel for visceral hemostasis and skin tissue regeneration. Mater Today Bio. 2024;27:101146. 10.1016/j.mtbio.2024.101146. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR22] 22.Song Y, Xing J, Ren L, Xu X, Han D, Xu H, et al. Preparation of multi-functional quaternary ammonium chitosan/surfactin hydrogel and its application in wound management. Macromol Biosci. 2023;23:e2300166. 10.1002/mabi.202300166. [DOI] [PubMed] [Google Scholar]

[CR23] 23.Jia S, Zhou H, Chen J, Lin J, Zhu X, Weng J, et al. A comprehensive review of hydrogel strategies for repairing peripheral nerve injuries. Brain-X. 2025;3:e70012. 10.1002/brx2.70012. [Google Scholar]

[CR24] 24.Quan Y, Shao H, Wang N, Gao Z, Jin M. Microenvironment-sensitive hydrogels as promising drug delivery systems for co-encapsulating microbial homeostasis probiotics and anti-inflammatory drugs to treat periodontitis. Mater Today Bio. 2025;32:101711. 10.1016/j.mtbio.2025.101711. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.Sheng F, Luo Y, Wu D, Yuan J, Zheng G, Zheng Y, et al. Poly amino acid thermosensitive hydrogel loaded with ICG-001 for inhibiting keloid by down-regulating the Wnt/β-Catenin pathway. Mater Des. 2022;222:111050. 10.1016/j.matdes.2022.111050. [Google Scholar]

[CR26] 26.Dong D, Mao L, Qin Z, Guo Y, Yu J, Hu X, et al. A simple and effective hydrogel dressing for advanced management of full-thickness skin wound by multi-functional strategies. J Nanobiotechnol. 2025;23:745. 10.1186/s12951-025-03803-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR27] 27.Mardhiah J, Halim AS, Heng S, Saipolamin AQ. Scoping review for pain mitigation during intralesional injections of corticosteroid for hypertrophic scar and keloid treatment. BMJ Open. 2025;15:e092800. 10.1136/bmjopen-2024-092800. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR28] 28.Yang E, Xu R, Zhang H, Xia W, Huang X, Zan T. Deciphering pain and pruritus in keloids from the perspective of neurological dysfunction: where are we now? Biomedicines. 2025;13:663. 10.3390/biomedicines13030663. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR29] 29.Li Y, Chen Y, Xue Y, Jin J, Xu Y, Zeng W, et al. Injectable hydrogel delivery system with high drug loading for prolonging local anesthesia. Adv Sci. 2024;11:e2309482. 10.1002/advs.202309482. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR30] 30.Wang S-Q, Jin M-J, Guo Z-K, Shen D-R, Jin L-N, Cheng F, et al. Trace element-dictated exosome modules and self-adaptive dual-network hydrogel orchestrate diabetic foot regeneration through complement-mitochondria-autophagy circuitry. Mil Med Res. 2025;12:71. 10.1186/s40779-025-00658-4. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR31] 31.Li X, Dou N, Zhong L, Wu Y, Cai Z, Zhao Z, et al. mPEG@ELA-11 alleviates atherosclerosis via AKT-ER stress-mediated macrophage modulation. BME Front. 2025;6:0203. 10.34133/bmef.0203. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR32] 32.Wang S, Liu Y, Wang X, Chen L, Huang W, Xiong T, Wang N, Guo J, Gao Z, Jin M. Modulating macrophage phenotype for accelerated wound healing with chlorogenic acid-loaded nanocomposite hydrogel. J Control Release. 2024;369:420–43. 10.1016/j.jconrel.2024.03.054. [DOI] [PubMed] [Google Scholar]

[CR33] 33.Liu J, Pan Z, Khan A, Li H. Targeted drug delivery system for pulmonary fibrosis: design and development of biomaterials. Bio I. 2025;6:1–23. 10.15212/bioi-2025-0016. [Google Scholar]

[CR34] 34.Wang S, Liu Y, Sun Q, Zeng B, Liu C, Gong L, et al. Triple cross-linked dynamic responsive hydrogel loaded with selenium nanoparticles for modulating the inflammatory microenvironment via PI3K/Akt/NF-κB and MAPK signaling pathways. Adv Sci. 2023;10:e2303167. 10.1002/advs.202303167. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR35] 35.Zhang Y, Bao Y, Li Z, Jin Z. A novel nude mice model for studying the pathogenesis of keloid. J Cosmet Dermatol. 2025;24:e70284. 10.1111/jocd.70284. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR36] 36.Shi X, Xia X, Xiao Y, Shu H, Xu Z, Liu M, et al. Ferroptosis-resistant adipocytes drive keloid pathogenesis via GPX4-mediated adipocyte-mesenchymal transition and iron-cystine metabolic communication. Int J Biol Sci. 2025;21:5097–115. 10.7150/ijbs.114930. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR37] 37.Lv W, Wu Y, Chen H. Orthogonal upconversion nanocarriers for combined photodynamic therapy and precisely triggered gene silencing in combating keloids. J Control Release. 2025;379:1–13. 10.1016/j.jconrel.2024.12.080. [DOI] [PubMed] [Google Scholar]

[CR38] 38.Zhuang B-Y, Hu F-C, Gao X, Leng Q, Zhang Y, You Y. Development of a simvastatin-loaded copolymer acid-sensitive nanocarrier and its experimental use in the treatment of keloids. J Cosmet Dermatol. 2025;24:e16573. 10.1111/jocd.16573. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR39] 39.Chan WCW. Principles of nanoparticle delivery to solid tumors. BME Front. 2023;4:0016. 10.34133/bmef.0016. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR40] 40.Pan W-C, Lin Y-H, Iao HM, Chang Y-H, Chen Y-H, Liu H-C, et al. In situ magnetoelectric generation of miRNA sponges and wireless electric stimulus by conductive granular scaffolds for nerve regeneration. Adv Mater. 2025;37:e2500650. 10.1002/adma.202500650. [DOI] [PubMed] [Google Scholar]

[CR41] 41.Iao HM, Chen C-Y, Lin Y-H, Pan W-C, Liang C-Y, Liu H-C, et al. Wireless in situ catalytic electron signaling-mediated transcriptomic reprogramming for neuron regeneration via adaptable antennas. Adv Sci. 2025;12:e2504786. 10.1002/advs.202504786. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Chan Y-C, Lin Y-H, Liu H-C, Hsu R-S, Chiang M-R, Wang L-W, et al. In situ magnetoelectric generation of nitric oxide and electric stimulus for nerve therapy by wireless chargeable molybdenum carbide octahedrons. Nano Today. 2023;51:101935. 10.1016/j.nantod.2023.101935. [Google Scholar]

[CR43] 43.Zhao Y, Wu S, Cai Y, Yang H, Dong X, Yang B, et al. Integration of finite element simulations with 3D printing technology for personalized chitin/PLA microneedle-based drug delivery systems in thoracic keloid treatment. Int J Biol Macromol. 2025;315:144487. 10.1016/j.ijbiomac.2025.144487. [DOI] [PubMed] [Google Scholar]

[CR44] 44.Bao Z, Gao P, Xia G, Wang Z, Kong M, Feng C, et al. A thermosensitive hydroxybutyl chitosan hydrogel as a potential co-delivery matrix for drugs on keloid inhibition. J Mater Chem B. 2016;4:3936–44. 10.1039/c6tb00378h. [DOI] [PubMed] [Google Scholar]

[CR45] 45.Li J, Fu R, Li L, Yang G, Ding S, Zhong Z, Zhou S. Co-delivery of dexamethasone and green tea polyphenols using electrospun ultrafine fibers for effective treatment of keloid. Pharm Res. 2014;31:1632–43. 10.1007/s11095-013-1266-2. [DOI] [PubMed] [Google Scholar]

[CR46] 46.Gao Y, Cheng X, Wang Z, Wang J, Gao T, Li P, et al. Transdermal delivery of 10,11-methylenedioxycamptothecin by hyaluronic acid based nanoemulsion for inhibition of keloid fibroblast. Carbohydr Polym. 2014;112:376–86. 10.1016/j.carbpol.2014.05.026. [DOI] [PubMed] [Google Scholar]

[CR47] 47.Aoki M, Matsumoto NM, Dohi T, Kuwahawa H, Akaishi S, Okubo Y, et al. Direct delivery of apatite nanoparticle-encapsulated SiRNA targeting TIMP-1 for intractable abnormal scars. Mol Ther Nucleic Acids. 2020;22:50–61. 10.1016/j.omtn.2020.08.005. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR48] 48.Moawad F, Pouliot R, Brambilla D. Dissolving microneedles in transdermal drug delivery: a critical analysis of limitations and translation challenges. J Control Release. 2025;383:113794. 10.1016/j.jconrel.2025.113794. [DOI] [PubMed] [Google Scholar]

[CR49] 49.Zhuang Z-M, Wang Y, Feng Z-X, Lin X-Y, Wang Z-C, Zhong X-C, et al. Targeting diverse wounds and scars: recent innovative bio-design of microneedle patch for comprehensive management. Small. 2024;20:e2306565. 10.1002/smll.202306565. [DOI] [PubMed] [Google Scholar]

[CR50] 50.Vettorato E, Volonté P, Musazzi UM, Cilurzo F, Casiraghi A. Skin microincision technique to enhance drug penetration for the treatment of keloid and hypertrophic scars. Int J Pharm. 2025;671:125259. 10.1016/j.ijpharm.2025.125259. [DOI] [PubMed] [Google Scholar]

[CR51] 51.Chen H, Xu J, Sun J, Jiang Y, Zheng W, Hu W, et al. Recent advances on thermosensitive hydrogels-mediated precision therapy. Asian J Pharm Sci. 2024;19:100911. 10.1016/j.ajps.2024.100911. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR52] 52.Yew PYM, Lin Q, Owh C, Chee PL, Loh XJ. Current research and future potential of thermogels for sustained drug delivery. Expert Opin Drug Deliv. 2025;22:769–86. 10.1080/17425247.2025.2486350. [DOI] [PubMed] [Google Scholar]

[CR53] 53.Wan H, Wang S, Li C, Zeng B, Wu H, Liu C, et al. LA67 liposome-loaded thermo-sensitive hydrogel with active targeting for efficient treatment of keloid via peritumoral injection. Pharmaceutics. 2023;15:2157. 10.3390/pharmaceutics15082157. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR54] 54.Wu D, Liu X, Jin Z. Placental mesenchymal stem cells-secreted proenkephalin suppresses the p38 MAPK signaling to block hyperproliferation of keloid fibroblasts. Tissue Cell. 2023;85:102218. 10.1016/j.tice.2023.102218. [DOI] [PubMed] [Google Scholar]

PERMALINK

A multifunctional injectable nanocomposite hydrogel for precision keloid therapy via ECM remodeling and local pruritus relief

Jingyu Jin

Chengjie Zhu

Nuoya Wang

Mingji Jin

Yanxin Jin

Heying Jin

Shuang Zhang

Yanhong Zhou

Lei Zhao

Yinli Luo

Zhonggao Gao

Zhehu Jin

Shuangqing Wang

Abstract

Graphical Abstract

Supplementary Information

Introduction

Materials and methods

Materials

Proteomic analysis of normal skin and keloid tissues

Sample source and handling

Protein Extraction, enzymatic Digestion, and mass spectrometry analysis

Construction and characterization of 5Fu@ZIF-8

Construction of 5Fu@ZIF-8

Characterization of 5Fu@ZIF-8

Stability investigation

Cellular-Level studies

Cytotoxicity of blank ZIF-8

Inhibitory effect of 5Fu@ZIF-8 on cell proliferation

Uptake experiment of 5Fu@ZIF-8

Inhibitory effect of 5Fu@ZIF-8 on cell migration

Expression of α-SMA and FN

Quantification of col 1 and col 3

TGF-β1/Smad signaling pathway

Construction and evaluation of hydrogels

Synthesis and characterization of LACS matrix

Construction of hydrogels

Characterization of hydrogels

In vivo distribution and safety

In vivo distribution

Penetration depth in keloid tissue

In vivo biosafety detection

In vivo pharmacodynamics

Keloid model establishment and grouped administration

Pharmacodynamic evaluation and body weight monitoring

Histopathology and apoptosis detection

ECM regulation detection

Statistical analysis

Results

Proteomic characteristics of human keloid tissues

Fig. 1.

Physicochemical characterization of ZIF-8 and 5Fu@ZIF-8

Particle Size, zeta Potential, and microstructure

Fig. 2.

Drug loading performance of 5Fu@ZIF-8

In vitro drug release

Biological effects of 5Fu@ZIF-8 on KFs

Cytotoxicity and proliferation Inhibition

Fig. 3.

Cell uptake efficiency

Biological effects of 5Fu@ZIF-8 on HUVEC

Cytotoxicity and proliferation effect

Fig. 4.

Cell uptake efficiency

Regulation of 5Fu@ZIF-8 on cell Behavior, ECM, and TGF-β/Smad pathway

Cell migration

Fig. 5.

ECM regulation

Regulation of TGF-β/Smad signaling pathway

Preparation and performance characterization of UV-Responsive hydrogels

Synthesis and macroscopic properties of hydrogels

Fig. 6.

Rheological, Swelling, Degradation, and drug release behaviors

Hydrogel biocompatibility

In vivo distribution and keloid penetration of hydrogels

In vivo distribution

Fig. 7.

Keloid tissue penetration depth