Tip-retained hydrated silk fibroin microneedles for sustained local delivery of TNF-α antagonist in psoriasis treatment

Abstract

Psoriasis is a chronic immune-mediated inflammatory skin disease characterized by keratinocyte hyperproliferation and cytokine-driven immune activation, in which TNF-α plays a central pathogenic role. We developed a layered hydrated silk fibroin (SF) microneedle system (ETA@Hyd MN) with SF tips loaded with etanercept (ETA, a TNF-α inhibitor for psoriasis), for local cutaneous delivery to reduce systemic exposure and rapidly dissolving polyvinyl alcohol (PVA) bases that facilitate retention of the SF tips within the skin. Hydration markedly slowed ETA release, with approximately 80% released at 48 h compared with approximately 75% released within 1 h from microneedles without hydration. In HaCaT cell scratch assays, release media from ETA@Hyd MN attenuated the accelerated wound closure response, resulting in a greater remaining scratch width than that in the model group at 48 h. In an IMQ-induced psoriasis-like mouse model, ETA@Hyd MN improved clinical symptoms, reduced PASI scores, alleviated spleen enlargement, and suppressed epidermal hyperplasia and TNF-α/IL-6 expression. These results demonstrate that ETA@Hyd MN enables sustained ETA release and effectively suppresses inflammation, highlighting its potential for treating inflammatory skin diseases.

Graphical abstract

1. Introduction

Psoriasis is a chronic immune-mediated inflammatory skin disease characterized by excessive keratinocyte proliferation and persistent immune activation (Sieminska et al., 2024). Among the cytokines involved, tumor necrosis factor-α (TNF-α) plays a pivotal role in amplifying inflammatory cascades and sustaining disease progression (Bowcock and Krueger, 2005; Chiricozzi et al., 2011; Tobin and Kirby, 2005). TNF-α antagonists have demonstrated significant clinical efficacy; however, their systemic administration is associated with notable limitations, including systemic immunosuppression, increased infection risk, and poor patient compliance due to repeated injections (Guo et al., 2023; Li et al., 2019a, Li et al., 2019b; Semble et al., 2014). Therefore, developing a localized and sustained delivery strategy for TNF-α antagonists is highly desirable for improving therapeutic outcomes while minimizing systemic side effects.

Transdermal local drug delivery is considered an attractive strategy for the treatment of psoriasis, as it enables high drug concentrations at the lesion site while minimizing systemic side effects (Prausnitz et al., 2004; Rapalli et al., 2020; Zeng et al., 2025). However, the stratum corneum presents a formidable barrier that severely limits the penetration of macromolecular therapeutics (Prausnitz et al., 2004; Tuan-Mahmood et al., 2013). Microneedle (MN) technology can physically bypass this barrier, offering an effective platform for the localized delivery of biologics, and holds significant potential for the development of sustained-release systems (Men et al., 2022; Zeng et al., 2025; Zheng et al., 2024). Currently, sustained delivery of TNF-α antagonists via microneedles largely relies on chemical crosslinking or covalent conjugation strategies. For instance, dissolvable microneedle arrays have been developed to deliver TNF-α antibody inhibitors conjugated to hydrophilic polymers such as hyaluronic acid (HA), enabling intradermal delivery and prolonged release (Korkmaz et al., 2016). In addition, a microneedle-based sustained-release system for TNF-α/IL-6R bispecific antibodies has been reported to alleviate inflammation and promote bone regeneration in a rat model of rheumatoid arthritis (Zhang et al., 2025). This system utilizes methacrylated gelatin, which undergoes photo-crosslinking under ultraviolet or visible light in the presence of photoinitiators, forming a hydrogel network that regulates drug release behavior. Despite their effectiveness in prolonging drug release, these approaches generally depend on chemical crosslinking, covalent modification, or solvent-induced stabilization processes. Such strategies not only complicate fabrication but may also introduce potential irritation and biocompatibility concerns (Prausnitz et al., 2004). For psoriatic skin, which is already in a chronic inflammatory state with a compromised barrier function, these additional chemical treatments are less desirable and may exacerbate local irritation (Kircik et al., 2023; Orsmond et al., 2021; Tortola et al., 2012). Therefore, there is a pressing need to develop milder and more biocompatible delivery systems.

Silk fibroin (SF) has attracted considerable attention in drug delivery due to its excellent biocompatibility and tunable structural properties (Hofmann et al., 2006). Notably, hydration can induce conformational transitions in SF, leading to enhanced mechanical strength and modulated drug release behavior (Johari et al., 2020; Tomeh et al., 2019). This unique feature makes it possible to construct sustained-release systems without the need for additional chemical crosslinkers (Hofmann et al., 2006).

In this study, we developed a tip-retained microneedle system composed of a rapidly dissolving polyvinyl alcohol (PVA) base and drug-loaded hydrated SF tips. Upon administration, the PVA base rapidly dissolves, allowing the SF tips to detach and remain embedded within the skin, where they function as a localized drug depot (See Scheme 1). This design enables sustained release of TNF-α antagonists for the treatment of psoriasis.

Scheme 1.

The pathological mechanism of excessive inflammatory stimulation in psoriatic skin is illustrated. TNF-α acts as a key upstream pro-inflammatory cytokine and promotes the expression of downstream inflammatory factors such as IL-6, thereby exacerbating skin inflammation and lesion formation. ETA-loaded microneedles deliver the TNF-α antagonist locally, thereby suppressing TNF-α activity and reducing downstream IL-6 expression. Treatment with ETA@Hyd MN resulted in a marked decrease in TNF-α and IL-6 levels in the skin, alleviating inflammation and improving psoriasis-like lesions.

2. Materials and methods

2.1. Materials

Polydimethylsiloxane (PDMS) microneedle molds were purchased from Henan Weina Benteng Biotechnology Co., Ltd. (Henan, China). SF was prepared from Bombyx mori cocoons according to a previously reported procedure (Rockwood et al., 2011). Polyvinyl alcohol (PVA) was purchased from Aladdin Reagent Co., Ltd. (Shanghai, China). Etanercept (ETA) was obtained from Shanghai Zhongxin Guojian Pharmaceutical Co., Ltd. (Shanghai, China). FITC NHS ester and Rhodamine B NHS ester were purchased from Thermo Fisher Scientific Inc. (Waltham, MA, USA). HaCaT cells were obtained from Wuhan Pricella Biomedical Technology Co., Ltd. (Beijing, China), and lipopolysaccharide (LPS) was purchased from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany). BALB/c mice (female, 6–7 weeks old) were supplied by Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China), and imiquimod (IMQ) was obtained from Sichuan Mingxin Pharmaceutical Co., Ltd. (Chengdu, China). Phosphate-buffered saline (PBS), hematoxylin and eosin (H&E) staining reagents, and DAPI were purchased from Servicebio Co., Ltd. (Wuhan, China). Methylene blue solution was obtained from Beyotime Biotechnology Co., Ltd. (Shanghai, China). Anti-TNF-α and anti-IL-6 antibodies were purchased from Proteintech Group, Inc. (Rosemont, IL, USA), and the TNF-α ELISA kit was obtained from Elabscience Biotechnology Inc. (Wuhan, China).

2.2. Methods

2.2.1. Fabrication and characterization of ETA@Hyd MN

Silk fibroin aqueous solution was prepared from Bombyx mori cocoons according to a previously reported method (Rockwood et al., 2011). Briefly, the cocoons were boiled in 0.02 mol/L sodium carbonate solution to remove sericin and subsequently rinsed thoroughly with deionized water. After drying at room temperature, the degummed silk fibers were dissolved in 9.3 mol/L lithium bromide solution at 60 °C for 4 h. The resulting SF solution was then dialyzed against deionized water using dialysis tubing with a molecular weight cutoff of 3500 Da for 2 days, during which the dialysis water was replaced at least five times to remove residual lithium bromide. Following dialysis, the solution was centrifuged twice at approximately 12,700 ×g (9000 rpm) for 20 min at 4 °C to remove insoluble impurities and protein aggregates. The purified supernatant was collected, and the SF concentration was determined gravimetrically by drying a known volume of the solution to constant weight. The resulting SF aqueous solution was stored at 4 °C until further use.

ETA loaded microneedles were fabricated using polydimethylsiloxane (PDMS) molds through a sequential casting procedure, as illustrated in Scheme 2. The molds were square in shape with an overall patch size of 14.5 × 14.5 mm and contained a 20 × 20 microneedle array. Each microneedle cavity had a needle length of 600 ± 5 μm, a bottom diameter of 250 ± 3 μm, and a tip to tip spacing of 550 ± 3 μm. The groove depth of the mold was 2 ± 0.5 mm. To prepare the drug loaded tip formulation, 198.4 μL of 4% (w/v) SF aqueous solution was mixed with 1.6 μL of ETA stock solution at 25 mg/mL, yielding a total volume of 200 μL containing 40 μg of ETA. The resulting mixture was carefully introduced into the tip region of the PDMS mold and subjected to vacuum treatment to facilitate complete filling of the microneedle cavities and removal of entrapped air bubbles. The mold was then dried at room temperature for 2 h to form the ETA loaded needle tips.

Scheme 2.

Schematic illustration of the preparation of ETA@Hyd MN.

Subsequently, 100 μL of drug free 4% (w/v) SF aqueous solution was added to reinforce the microneedle structure. After vacuum treatment, the mold was dried at room temperature for an additional 2 h. Thereafter, 100 μL of 40% (w/v) polyvinyl alcohol (PVA) solution was cast onto the mold to form the backing layer, followed by vacuum treatment and drying at room temperature for 12 h. The resulting microneedle arrays were carefully removed from the molds and designated ETA@MN.

To obtain hydrated microneedles, the dried ETA@MN arrays were placed in a vacuum drying chamber containing 20 mL of water and exposed to a pressure of approximately −0.1 MPa for 30 min or 60 min. Under these conditions, the microneedles were exposed to water vapor without direct contact with liquid water, resulting in the hydrated formulation designated ETA@Hyd MN.

The morphology and structural characteristics of the fabricated microneedles were examined using confocal laser scanning microscopy (CLSM) and field emission scanning electron microscopy (FE-SEM).

2.2.2. Assessment of hydration induced structural changes by ATR FTIR

Drug free microneedles were prepared using the same fabrication procedure described above to exclude potential interference from ETA. Attenuated total reflection Fourier transform infrared (ATR FTIR) spectroscopy was employed to characterize changes in the secondary structure of SF before and after hydration treatment. Each sample was placed in direct contact with the ATR crystal, and spectra were recorded over the range of 4000 to 400 cm⁻¹ at a resolution of 4 cm⁻¹. Particular attention was given to the amide I region, where changes in peak position were analyzed to evaluate hydration induced conformational rearrangement of SF and the associated formation of β sheet structures.

2.2.3. Mechanical characterization of microneedles

The mechanical properties of the microneedle patches were evaluated using a universal testing machine. MN, Hyd MN, and ETA@Hyd MN were subjected to axial compression testing to determine the effect of hydration and ETA incorporation on microneedle mechanical performance. Each microneedle patch was placed on the testing platform with the needle tips facing upward. The initial position was defined when the upper compression probe first contacted the microneedle tips without applying an appreciable load. The probe was subsequently moved downward at a constant rate of 0.5 mm/min, while the applied force and corresponding displacement were continuously recorded to generate force versus displacement curves for mechanical comparison among the different formulations.

2.2.4. In vitro release from microneedles

ETA was fluorescently labeled with FITC NHS ester through covalent coupling between the NHS ester group and the primary amino groups of ETA. Briefly, ETA and FITC NHS ester were reacted at a molar ratio of 1:3 under mildly alkaline conditions at pH 8.0 for 2 h at room temperature. Following the reaction, unreacted FITC NHS ester was removed by ultrafiltration, and the purified FITC labeled ETA was incorporated into the microneedles according to the fabrication procedure described above. The resulting microneedles were subsequently subjected to hydration treatment, while nonhydrated ETA@MN containing FITC labeled ETA were used as the control.

Each microneedle sample was immersed in 50 mL of phosphate buffered saline (PBS, pH 7.4) in a 50 mL centrifuge tube and maintained at 37 °C. At predetermined time points of 0, 0.5, 1, 2, 4, 8, 12, 24, 48, and 72 h, the release medium was gently mixed to ensure homogeneous distribution of the released ETA. After brief centrifugation to remove suspended matrix residues, 100 μL of the supernatant was collected, and an equal volume of fresh PBS was immediately replenished to maintain a constant release volume.

The fluorescence intensity of FITC labeled ETA in the collected supernatants was measured using a microplate reader, and the amount of released ETA was determined from a corresponding fluorescence calibration curve. Cumulative ETA release was calculated by correcting for the amount of drug removed during each previous sampling step and was expressed as a percentage of the total amount of ETA initially incorporated into the microneedles.

2.2.5. Insertion capability of ETA@Hyd MN

All animal experiments were reviewed and approved by the Welfare Ethics Committee of Zhongyan Zichuang (Beijing) Biotechnology Co., Ltd., Beijing, China (Approval No. zyzc-20,220,420-01). All procedures involving animals were conducted in accordance with the institutional guidelines for the care and use of laboratory animals, with appropriate measures taken to minimize animal suffering. Mice were anesthetized with isoflurane and the dorsal hair was carefully removed. On the following day, the mice were re-anesthetized, and the microneedle patches were inserted vertically into the dorsal skin and maintained in place for 5 min before removal. Subsequently, the treated skin area was stained with 0.4% (w/v) methylene blue solution to visualize the microchannels more clearly. The skin was then imaged using a digital camera. After imaging, the mice were euthanized, and the treated skin at the microneedle insertion sites was excised, fixed, and subjected to hematoxylin and eosin (H&E) staining. Vertical sections were prepared to evaluate the penetration depth and microchannel formation.

2.2.6. In vitro scratch wound assay

HaCaT cells were stimulated with lipopolysaccharide (LPS, 1 μg/mL) for 24 h to establish an inflammatory cell model. The cells were cultured in 12 well plates until approximately 90% confluence was reached. A straight scratch was then generated across the cell monolayer using a sterile 200 μL pipette tip. The wells were gently washed to remove detached cells, followed by replacement with the corresponding treatment medium.

The cells were divided into five groups: Control, cells without LPS stimulation; Model, cells stimulated with LPS without further treatment; MN, LPS stimulated cells treated with release medium obtained from blank microneedles for 12 h; ETA@Hyd MN 1 h, LPS stimulated cells treated with medium containing the release solution collected after incubation of ETA@Hyd MN in PBS for 1 h; and ETA@Hyd MN 12 h, LPS stimulated cells treated with medium containing the release solution collected after incubation of ETA@Hyd MN in PBS for 12 h.

Representative images of the scratched areas were acquired at 0, 12, 24, and 48 h using an optical microscope. For quantitative analysis, five randomly selected nonoverlapping fields were analyzed in each well using ImageJ software, and the scratch width measured from the five fields was averaged to obtain one value for each well. Three independent wells were included in each group, corresponding to a statistical sample size of n = 3. The remaining scratch width at each time point was normalized to the corresponding initial scratch width at 0 h and expressed as a percentage according to the following equation: Remaining scratch width (%) = Wt / W0 × 100%, where W0 represents the initial scratch width at 0 h and Wt represents the scratch width at the indicated time point.

In addition, cell culture supernatants were collected, and the concentration of tumor necrosis factor α (TNF-α) was determined using an enzyme linked immunosorbent assay (ELISA) kit according to the manufacturer's instructions.

2.2.7. Induction of psoriasis-like skin lesions in mice

All animal experiments were reviewed and approved by the Welfare Ethics Committee of Zhongyan Zichuang (Beijing) Biotechnology Co., Ltd. All procedures involving animals were conducted in accordance with the institutional guidelines for the care and use of laboratory animals, and efforts were made to minimize animal suffering. Mice with preexisting dorsal skin lesions, signs of illness, abnormal behavior, or other conditions that could interfere with model establishment or therapeutic evaluation were excluded before the experiment. After IMQ application, mice that failed to develop characteristic psoriasis like manifestations, including erythema, scaling, and skin thickening, were excluded before therapeutic intervention. The IMQ induced psoriasis like mouse model was used because topical IMQ produces characteristic clinical and histopathological features resembling psoriasis, including erythema, scaling, skin thickening, epidermal hyperplasia, and inflammatory activation (Tortola et al., 2012), providing a suitable platform for evaluating the therapeutic effects of locally administered ETA@Hyd MN.

To establish psoriasis like skin lesions, female BALB/c mice aged 6 to 7 weeks were shaved on the dorsal area one day before IMQ application. A total of 60 mg of IMQ cream was applied evenly to the shaved dorsal skin once daily for 10 consecutive days. The development of psoriasis like lesions was monitored daily based on the appearance of erythema, scaling, and skin thickening. After four consecutive days of IMQ application, evident psoriasis like skin lesions were observed, consistent with previous reports (Li et al., 2019a, Li et al., 2019b; Luo et al., 2016). Therapeutic intervention was subsequently initiated on day 5.

2.2.8. In vivo therapeutic efficacy of ETA@Hyd MN

Female BALB/c mice aged 6 to 7 weeks were randomly assigned to five groups with five mice in each group: untreated healthy control (Control), IMQ treated mice (Model), IMQ treated mice receiving blank microneedles without drug loading (MN), IMQ treated mice receiving ETA@MN, and IMQ treated mice receiving ETA@Hyd MN. Therapeutic administration was initiated on day 5, after the establishment of psoriasis like skin lesions had been confirmed based on erythema, scaling, and skin thickening. A single treatment was administered on day 5. One microneedle patch containing 40 μg of ETA was applied to each mouse in the ETA@MN and ETA@Hyd MN groups, while one blank microneedle patch was applied to each mouse in the MN group. Daily IMQ application was continued throughout the treatment period to maintain the inflammatory stimulus. Body weight and PASI scores were recorded daily throughout the experiment. The severity of psoriasis like skin lesions was assessed based on erythema, scaling, and skin thickening according to previously reported criteria (Sun et al., 2013). Each parameter was scored from 0 to 4 as follows: 0, none; 1, slight; 2, moderate; 3, marked; and 4, very marked. Two blinded independent observers performed the assessments, and the mean score was used as the final value for each mouse. The cumulative PASI score was calculated as the sum of the three parameters, ranging from 0 to 12, with higher scores indicating greater disease severity. On day 10, the mice were euthanized, and dorsal skin and spleen tissues were collected for subsequent histological and related analyses.

2.2.9. Hematoxylin and eosin (H&E) staining

Paraffin-embedded tissue sections (5 μm) were deparaffinized by immersion in xylene and rehydrated through a graded series of ethanol solutions (100%, 95%, 70%) to distilled water. The sections were then stained with hematoxylin for several minutes to visualize cell nuclei. Following staining, the sections were rinsed in water and subjected to a blueing step using a weakly alkaline solution. Subsequently, the sections were counterstained with eosin to highlight the cytoplasm and extracellular matrix. After staining, the sections were dehydrated through a graded ethanol series, cleared in xylene, and mounted with a neutral mounting medium. Stained sections were examined under a light microscope for histological evaluation.

For quantitative analysis of epidermal thickness, one vertical skin section perpendicular to the skin surface was randomly selected from each mouse. Five nonoverlapping fields were randomly selected from each section, and epidermal thickness was measured using ImageJ software. The mean value of the five measurements was calculated and used as the representative epidermal thickness value for each mouse. Five mice from each group were included in the statistical analysis, resulting in five independent biological replicates per group.

2.2.10. Immunofluorescence staining

After rehydration and antigen retrieval, paraffin-embedded skin tissue sections (5 μm) were blocked with 10% serum for 30 min. The sections were incubated overnight at 4 °C with the primary antibodies: TNF-α (Proteintech, 1:500) and IL-6 (Proteintech, 1:500). The following day, the sections were returned to room temperature and washed with PBST. Fluorescently labeled secondary antibodies were then applied and incubated at room temperature for one hour, followed by three washes with PBST. Cell nuclei were stained with DAPI (1:1000) and the sections were mounted. Immunofluorescence images were captured using a Nikon confocal microscope. For quantitative immunofluorescence analysis of TNF-α and IL-6, three mice were randomly selected from each group. One skin section was analyzed for each mouse, and five nonoverlapping fields were randomly selected from each section. Fluorescence intensity was quantified in each field, and the mean value of the five fields was calculated and used as the representative fluorescence intensity value for each mouse. Three individual mouse values were therefore used for statistical comparison between groups, corresponding to three independent biological replicates per group.

2.2.11. Statistical analysis

All statistical analyses were performed using GraphPad Prism version 8.0. Data are presented as mean ± standard deviation. The sample size and experimental unit for each dataset are specified in the corresponding figure legends. Comparisons between two independent groups were performed using a two tailed unpaired Student t-test. Comparisons among multiple independent groups were performed using one way analysis of variance followed by Tukey's multiple comparisons test. Adjusted P values were reported for multiple comparisons. Statistical significance was defined as follows: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

3. Results and discussion

3.1. Preparation and characterization of ETA@Hyd MN

The fabrication process of ETA@Hyd MN is illustrated in Scheme 2. A stepwise casting strategy was employed to achieve spatially controlled drug loading. Briefly, an ETA-loaded SF solution was first cast into the tip region to enable localized drug enrichment, followed by the addition of a second SF layer to reinforce the needle structure. Subsequently, a polyvinyl alcohol (PVA) backing layer was introduced to provide mechanical support and facilitate separation of the needles from the substrate. The resulting microneedles exhibited a uniform 20 × 20 array with well-defined morphology (Fig. 1A). SEM images further confirmed the formation of sharp tips and intact structures, which are essential for effective skin penetration (Fig. 1B). To further determine the spatial distribution of ETA within the microneedle matrix, CLSM imaging was performed using rhodamine labeled SF and FITC labeled ETA. As shown in Fig. 1C, the fluorescence signal corresponding to ETA was predominantly confined to the upper approximately 200 μm of the microneedles, whereas SF fluorescence was distributed throughout the entire needle structure. Considering that the overall microneedle height was approximately 600 μm, these observations confirmed that ETA was preferentially enriched within the tip region rather than uniformly distributed throughout the microneedle matrix. This distribution was consistent with the sequential casting procedure and confirmed that the majority of ETA was incorporated into the portion of the microneedle intended to enter the skin.

Fig. 1.

Characterization of ETA@Hyd MN. (A) Representative optical image of the fabricated microneedle array (20 × 20 needles). (B) Representative scanning electron microscopy (SEM) image showing the surface morphology of the microneedles. (C) Representative confocal laser scanning microscopy (CLSM) images showing the spatial distribution of rhodamine labeled SF (red) and FITC labeled ETA (green), demonstrating preferential localization of ETA within the microneedle tip region. (D) Representative ATR-FTIR spectra of microneedles before and after hydration. (E) Representative force displacement curves of different microneedle formulations under axial compression. (F) Representative image of methylene blue staining of mouse skin following microneedle insertion, indicating successful disruption of the stratum corneum. (G) Representative optical images of microneedles before and after insertion into mouse skin. (H) Representative H&E staining of mouse skin sections after microneedle application, demonstrating penetration and retention of residual SF material at the insertion sites. (I) In vitro release profiles of ETA from hydrated and non-hydrated microneedles in PBS. n = 3, data are presented as mean ± SD. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

The structural changes in the SF matrix induced by hydration were subsequently investigated using ATR FTIR spectroscopy. As shown in Fig. 1D, the amide I band of MN was located at approximately 1638.7 cm⁻¹, whereas the corresponding peaks of 30 min Hyd MN and 60 min Hyd MN shifted to approximately 1618.9 and 1618.2 cm⁻¹, respectively. The pronounced shift toward lower wavenumbers after hydration is consistent with a transition of SF from relatively disordered conformations toward a more ordered β sheet associated structure. Water treatment has previously been reported to promote structural rearrangement and β sheet formation in SF through reorganization of intermolecular hydrogen bonding (Hofmann et al., 2006; Hu et al., 2011). Notably, only a minimal difference in the amide I peak position was observed between 30 min Hyd MN and 60 min Hyd MN, suggesting that the major hydration induced conformational rearrangement of SF had largely occurred within 30 min under the present conditions.

The effect of hydration time on the mechanical properties of the microneedles was further examined by axial compression testing (Fig. 1E). Consistent with the ATR FTIR results, 30 min Hyd MN and 60 min Hyd MN exhibited highly comparable force displacement profiles, indicating that prolonging the hydration treatment from 30 to 60 min produced little additional change in mechanical performance. This finding may be associated with the relatively rapid establishment of a structurally stabilized SF matrix during the initial hydration period. A distinct inflection in the force displacement curves was observed at a compressive displacement of approximately 400 μm, possibly reflecting deformation or fracture of the needle portion while the PVA backing maintained the overall integrity of the array. In addition, ETA@Hyd MN displayed a force displacement profile similar to that of the corresponding drug free Hyd MN, indicating that incorporation of ETA did not appreciably affect the mechanical properties of the hydrated microneedles.

The skin penetration performance of the prepared microneedles was further evaluated using mouse skin. Following microneedle application, methylene blue staining revealed a regular array of distinct blue puncture sites corresponding to the microneedle geometry, confirming effective disruption of the stratum corneum and successful formation of microscopic penetration sites (Fig. 1F). Optical examination showed substantial structural alteration of the microneedles after 5 min of application (Fig. 1G), which was likely associated with rapid hydration and dissolution of the PVA backing upon contact with skin fluid. Histological examination further revealed clear microchannels and residual material at the insertion sites (Fig. 1H), confirming successful penetration. The retained SF matrix may therefore act as a local depot for subsequent ETA release.

The regulatory effect of hydration on ETA release was subsequently evaluated in PBS. As shown in Fig. 1I, ETA@MN exhibited a pronounced initial release, with approximately 75% of ETA released within the first 1 h. In contrast, ETA@Hyd MN showed a substantially prolonged release profile, reaching a comparable release level after approximately 20 h and approximately 80% cumulative release at 48 h. This slower release may be related to the hydration induced increase in SF β sheet structure and matrix compactness, which could increase the diffusional resistance of ETA. These results suggest that hydration contributes to both mechanical reinforcement and prolonged ETA release from the SF microneedle matrix.

Taken together, hydration altered the secondary structure of SF and was accompanied by enhanced mechanical performance and prolonged ETA release. Following skin insertion, dissolution of the PVA backing allowed local retention of the SF needle matrix, which may provide a depot for sustained intradermal ETA delivery.

3.2. In vitro evaluation of ETA@Hyd MN in keratinocytes

The effects of ETA@Hyd MN on keratinocyte behavior were first evaluated using a scratch wound healing assay. Representative images at 0, 12, 24, and 48 h are shown in Fig. 2A. Compared with the control group, inflammatory stimulation in the model group markedly accelerated wound closure. Quantitative analysis (Fig. 2B) showed that approximately 60% of the initial wound width remained in the control group at 48 h, whereas the model group exhibited a significantly lower remaining wound width of approximately 28%, indicating accelerated wound closure under inflammatory conditions. A similar value was observed in the MN group, suggesting that the blank microneedle matrix had no obvious effect on this response. In contrast, treatment with the release media obtained from ETA@Hyd MN at 1 and 12 h partially reversed the accelerated wound closure, with approximately 32% and 50% of the initial wound width remaining, respectively. Previous studies have shown that etanercept suppresses abnormal proliferation of inflammatory HaCaT cells, while inhibition of TNF-α signaling can also reduce keratinocyte migration (Li et al., 2023; Scott et al., 2004). The inhibitory effects observed here therefore indicate that ETA released from the hydrated microneedles retained its biological activity. Moreover, the stronger effect produced by the 12 h release medium may be associated with the greater cumulative amount of ETA released over the prolonged incubation period.

Fig. 2.

Effects of ETA@Hyd MN on keratinocyte wound closure and TNF-α levels in vitro.

(A) Representative images of scratch wound healing assay in HaCaT cells at 0, 12, 24, and 48 h after different treatments. (B) Quantitative analysis of the remaining wound width. n = 3 wells per group, data are presented as mean ± SD. (ns, not significant; *p < 0.05; **p < 0.01;***p < 0.001; ****p < 0.0001). (C) TNF-α levels in cell culture supernatants measured by ELISA after treatment. n = 3, data are presented as mean ± SD.

To further evaluate the biological activity of released ETA, TNF-α levels in the culture supernatants were quantified by ELISA (Fig. 2C). Compared with the control group, TNF-α levels were significantly elevated in the model group, while comparable levels were observed in the MN group. In contrast, treatment with both the 1 h and 12 h ETA@Hyd MN release media significantly reduced TNF-α levels, further supporting the preservation of ETA activity following incorporation into and release from the hydrated SF matrix. The greater reduction observed with the 12 h release medium was consistent with the scratch assay and may reflect the higher cumulative ETA exposure achieved after prolonged release.

Taken together, ETA released from ETA@Hyd MN remained biologically active, as evidenced by attenuation of inflammatory wound closure behavior and reduction of TNF-α levels in HaCaT cells. The stronger effects observed with the 12 h release medium were consistent with the greater amount of ETA released over time, supporting the time dependent release of biologically active ETA from the hydrated SF matrix.

3.3. In vivo therapeutic efficacy in an IMQ-induced psoriasis model

The in vivo therapeutic efficacy of ETA@Hyd MN was evaluated using an IMQ induced psoriasis like mouse model, as illustrated in Fig. 3A. IMQ was applied daily throughout the 10 day experimental period. After psoriasis like lesions were established during the first 4 days, a single microneedle treatment was administered on day 5. Representative photographs obtained on day 10 showed pronounced erythema, scaling, and skin thickening in the Model group, while comparable lesions remained in the MN group, indicating that the blank microneedle matrix alone did not appreciably affect disease progression (Fig. 3B). In contrast, both ETA@MN and ETA@Hyd MN alleviated the visible skin lesions, with ETA@Hyd MN showing the most evident reduction in scaling and erythema, although mild residual lesions remained compared with the Control group.

Fig. 3.

In vivo therapeutic efficacy of ETA@Hyd MN in an IMQ-induced psoriasis model on day 10. (A) Experimental design and treatment schedule. (B) Representative images of mice at the end of the experiment. (C–E) Body weight, spleen photographs, and spleen weight, respectively. n = 5, data are presented as mean ± SD. (F—H) Clinical scores for erythema, scaling, and skin thickness. n = 5, data are presented as mean ± SD. (I) Psoriasis Area and Severity Index (PASI) scores. n = 5, data are presented as mean ± SD.

Body weight changes during the experimental period are shown in Fig. 3C. IMQ treatment resulted in an evident reduction in body weight in the Model and MN groups, whereas partial recovery was observed following ETA@MN and ETA@Hyd MN treatment, with a more pronounced tendency in the ETA@Hyd MN group. Although body weight is a nonspecific indicator, this trend suggests an improvement in the general condition of mice following ETA treatment.

Spleen morphology and weight were further evaluated as indicators associated with systemic immune activation. Marked splenomegaly was observed in the Model and MN groups, whereas spleen enlargement was attenuated following ETA@MN and ETA@Hyd MN treatment (Fig. 3D). Quantitative analysis of spleen weight showed a consistent trend, with a greater reduction observed in the ETA@Hyd MN group (Fig. 3E), suggesting that ETA treatment may partially alleviate IMQ associated systemic immune activation.

Quantitative analysis of erythema, scaling, and skin thickness scores further confirmed the gross observations (Fig. 3F to H). The Model and MN groups maintained similarly high clinical scores, whereas ETA@MN treatment reduced the severity of all three parameters. ETA@Hyd MN showed a trend toward greater improvement than ETA@MN, which may be associated with the prolonged local availability of ETA provided by the hydrated SF matrix. Consistently, PASI scores were markedly reduced following ETA treatment, with ETA@Hyd MN showing the lowest score among the IMQ treated groups (Fig. 3I).

Taken together, local delivery of ETA effectively alleviated IMQ induced psoriasis like lesions, as evidenced by improved gross appearance, reduced clinical scores, partial recovery of body weight, and attenuation of splenomegaly. The trend toward greater therapeutic improvement with ETA@Hyd MN was consistent with its sustained release behavior, suggesting that prolonged local ETA availability may contribute to the observed therapeutic effects.

3.4. Histological analysis and cytokine expression in psoriatic skin

Histological and immunofluorescence analyses were performed to further evaluate the therapeutic effects of ETA@Hyd MN on psoriasis like skin lesions (Fig. 4). H&E staining showed that the Model and MN groups exhibited pronounced epidermal hyperplasia, elongated rete ridges, and thickening of the stratum corneum compared with the Control group (Fig. 4A). In contrast, these pathological alterations were alleviated following ETA treatment, with the ETA@Hyd MN group showing the greatest numerical reduction in epidermal hyperplasia. Quantification of epidermal thickness showed a consistent trend (Fig. 4B), with markedly increased values in the Model and MN groups and progressive reductions following ETA@MN and ETA@Hyd MN treatment, respectively.

Fig. 4.

Histological and immunofluorescence analysis of skin lesions in an imiquimod-induced psoriasis model on day 10. (A) H&E staining of skin sections. (B) Quantification of epidermal thickness using ImageJ. Five random nonoverlapping fields were averaged for each mouse. n = 5, data are presented as mean ± SD. (C) Immunofluorescence staining of TNF-α. (D) Quantification of TNF-α fluorescence intensity. n = 3 mice per group, with five randomly selected nonoverlapping fields analyzed for each mouse. Data are presented as mean ± SD. (E) Immunofluorescence staining of IL-6. (F) Quantification of IL-6 fluorescence intensity. n = 3 mice per group, with five randomly selected nonoverlapping fields analyzed for each mouse. Data are presented as mean ± SD.

To further evaluate the local inflammatory response within psoriatic lesions, immunofluorescence staining was performed to analyze the expression of TNF-α and IL-6, two representative inflammatory cytokines involved in the amplification of psoriatic inflammation. TNF-α plays a central role in activating inflammatory responses in keratinocytes, while IL-6 contributes to the maintenance of the local inflammatory microenvironment and is associated with disease severity. Therefore, changes in their expression levels were used to assess the anti inflammatory effects of ETA delivery in psoriatic skin.

Immunofluorescence staining revealed marked differences in inflammatory cytokine expression among groups. Strong TNF-α fluorescence signals were observed predominantly in the epidermal regions of the Model and MN groups, indicating sustained activation of local inflammatory responses following IMQ induction (Fig. 4C). Quantitative analysis further confirmed elevated TNF-α expression in these groups, whereas ETA@MN treatment reduced TNF-α fluorescence intensity, and ETA@Hyd MN showed a further numerical decrease in fluorescence intensity (Fig. 4D). These results are consistent with the biological function of ETA as a TNF-α antagonist and suggest effective attenuation of TNF-α associated inflammation following local microneedle delivery.

A similar pattern was observed for IL-6 expression (Fig. 4E, F). Strong IL-6 fluorescence was detected in the Model and MN groups, whereas both ETA containing formulations reduced IL-6 expression, with the lowest numerical level observed in the ETA@Hyd MN group. The concurrent reductions in TNF-α and IL-6 suggest that local ETA delivery attenuated the inflammatory microenvironment of psoriatic lesions.

Taken together, ETA@Hyd MN markedly alleviated epidermal hyperplasia and reduced local expression of TNF-α and IL-6 in IMQ induced psoriatic skin. These therapeutic effects were consistent with the sustained release behavior of ETA@Hyd MN, suggesting that prolonged local ETA delivery may contribute to its therapeutic activity.

4. Conclusions

In this study, we developed ETA@Hyd MN for sustained local delivery of the TNF-α antagonist ETA in psoriasis. Hydration-induced structural rearrangement of the SF matrix enhanced microneedle stability and markedly prolonged ETA release, while dissolution of the PVA backing facilitated retention of the SF tips within the skin as a local drug depot. ETA released from the hydrated microneedles retained biological activity, and ETA@Hyd MN effectively alleviated psoriasis-like skin lesions, reduced PASI scores and epidermal hyperplasia, and suppressed local TNF-α and IL-6 expression in the IMQ-induced mouse model.

Recent studies have similarly demonstrated that microneedle-based local delivery can alleviate psoriasis-like skin lesions and reduce inflammatory responses in IMQ-induced models (Men et al., 2022; Zeng et al., 2025). In contrast to these systems, which rely on nanocrystal or nanoparticle formulations to enhance intradermal retention or prolong drug release, ETA@Hyd MN achieves sustained local delivery through hydration-induced structural regulation and retention of the SF needle matrix within the skin. Although ETA@Hyd MN showed a trend toward greater therapeutic improvement than ETA@MN, direct assessment of intradermal ETA retention and local pharmacokinetics is still required. Overall, hydration-regulated SF microneedles provide a simple physical strategy for sustained local delivery of protein therapeutics, with potential application in psoriasis and other localized inflammatory skin diseases.

Declaration of generative AI and AI-assisted technologies in the manuscript preparation process

During the preparation of this work, the authors used ChatGPT (OpenAI) for language polishing and improving the clarity and readability of the manuscript. The authors reviewed and edited the output as needed and take full responsibility for the content of the published article.

CRediT authorship contribution statement

Sai Gao: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Ruili Yin: Writing – review & editing, Methodology, Formal analysis, Data curation, Conceptualization. Jingxuan Shi: Methodology, Formal analysis, Data curation. Zhuang Miao: Methodology. Longyan Yang: Writing – review & editing, Supervision, Resources, Project administration, Conceptualization. Dong Zhao: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.