Hydrogel microneedle delivery of a palmitoylated peptide promotes hair regeneration via dual activation of Wnt/β-catenin and cAMP–PKA signaling

Abstract

Abstract

Bioactive peptides are promising candidates for hair regeneration, but their topical application is often limited by poor skin delivery and the need for frequent dosing. To overcome these challenges, we designed a palmitoylated peptide, Pal-MPAPO, and incorporated it into a hyaluronic acid-based hydrogel microneedle (MN) system for localized transdermal delivery. The fabricated MNs showed adequate mechanical strength for skin insertion and provided a time-dependent, prolonged release of Pal-MPAPO. In human dermal papilla cells (DPCs), Pal-MPAPO activated the Wnt/β-catenin pathway and increased intracellular cyclic adenosine monophosphate (cAMP) levels in a predominantly VPAC1-dependent manner. Additional VPAC2 validation did not reveal a significant contribution to cAMP signaling under the tested conditions. Functional interactions between Pal-MPAPO and VPAC1 were supported by molecular docking, molecular dynamics simulations, receptor knockdown, and fluorescence imaging. In a C57BL/6 mouse model, Pal-MPAPO-loaded MNs significantly promoted hair regrowth, achieving efficacy comparable to minoxidil but at a reduced application frequency. These findings identify Pal-MPAPO as a potential peptide candidate for hair regeneration and suggest that hydrogel MN-assisted delivery provides a localized peptide administration strategy.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04542-8.

Introduction

Hair loss, a condition affecting millions worldwide, significantly impacts physical appearance and emotional health [1, 2]. Currently, the common treatments for hair loss include minoxidil, finasteride, platelet-rich plasma, stem cell therapy, low-level laser therapy and hair follicle transplantation [1, 3–7]. However, these methods often come with challenges such as long treatment times, high costs, variable patient response, and unpredictable side effects, making it difficult for patients to achieve satisfactory results. Consequently, the development of safe, non-invasive, and more effective therapeutic strategies remains an urgent need in hair regeneration research.

Bioactive peptides have recently emerged as promising candidates for treating hair loss due to their ability to regulate cell proliferation, migration, and protein synthesis [8–11]. However, their relatively large molecular weight and hydrophilic nature limit skin retention after topical application, often requiring frequent dosing to maintain biological activity. Conventional topical formulations such as lotions, sprays, and gels do not adequately address these challenges. Although nanocarrier-based encapsulation can enhance bioavailability, it often requires complex synthesis, high production cost, and uncertain long-term biocompatibility. Therefore, for peptide therapeutics, an effective transdermal system should not only improve penetration, but also maintain consistent local exposure while reducing dosing burden.

Over the past two decades, protein S-acylation (commonly known as S-palmitoylation) has been recognized as an important regulator of multiple signaling pathways involved in epidermal homeostasis and hair follicle differentiation [12, 13]. However, the effects of palmitoyl peptides on hair follicles following transdermal application remain largely unexplored. Inspired by this mechanism, we introduced palmitoylation modification to MPAPO, a bioactive peptide previously designed by our group, generating a novel derivative named Pal-MPAPO [14]. The incorporation of a palmitoyl moiety was expected to enhance membrane affinity and receptor interaction, thereby improving local biological activity. Given that MPAPO can activate the classic Wnt/β-catenin signaling pathway and promote corneal nerve axon regeneration through the cAMP–PKA (protein kinase A) signal pathway, we hypothesized that Pal-MPAPO may promote hair regeneration via Wnt/β-catenin and VPAC receptor-mediated signaling [15]. Nevertheless, effective intradermal delivery of such macromolecular peptides remains a critical challenge.

To overcome the delivery limitations of peptide-based therapies, transdermal drug delivery technologies have attracted increasing attention. Among these approaches, microneedles (MNs) have emerged as minimally invasive platforms capable of reproducible intradermal deposition and sustained local exposure of macromolecules [16, 17]. Depending on their design and materials, MNs are typically classified into five types: solid, coated, hollow, dissolving, and hydrogel-forming [18, 19]. Solid MNs are easy to fabricate and enhance transdermal permeation by creating microchannels, but they may cause transient barrier disruption or mechanical failure, such as bending or tip retention within the skin [20]. Coated MNs enable rapid drug delivery to the skin; however, their drug loading is often limited and achieving uniform coatings remains challenging [20, 21]. Hollow MNs can deliver a large dose of drug, yet limited mechanical strength and leakage or clogging during infusion often hinder reliable delivery [21]. Dissolving MNs provide one-step administration and rapid release of macromolecules, but involve relatively complex fabrication processes and may exhibit reduced mechanical stability or delayed dissolution [21]. In contrast, hydrogel-forming MNs combine sufficient mechanical strength for skin insertion with in situ swelling to create hydrated diffusion pathways, enabling controlled and sustained drug release from an attached reservoir without leaving polymer residues in the skin [22, 23]. These features make them particularly suitable to deliver macromolecules, improve local exposure consistency, and reduce administration frequency. Accordingly, to facilitate effective and clinically translatable delivery of Pal-MPAPO for hair loss treatment, we encapsulated Pal-MPAPO within a methacrylate-functionalized hyaluronic acid hydrogel to fabricate an MN patch (Pal-MPAPO@MN).

In this study, we developed a hydrogel MN system for the transdermal delivery of Pal-MPAPO and investigated its hair growth-promoting activity and underlying mechanisms. Specifically, we investigated whether Pal-MPAPO regulates hair regeneration through Wnt/β-catenin and VPAC1-dependent cAMP–PKA signaling, and further evaluated the therapeutic efficacy of Pal-MPAPO@MN in a C57BL/6 mouse model.

Results

Preparation and characterization of Pal-MPAPO-loaded hydrogel MNs

The crosslinked hydrogel network enables sustained diffusion of Pal-MPAPO, representing a nano-scale biointerface for peptide delivery. The Fourier transform infrared (FTIR) spectra of hyaluronic acid methacrylate (HAMA) and hyaluronic acid are shown in Fig. S1 (Supporting Information). HAMA exhibited characteristic absorption peaks at 1637 and 1700 cm⁻¹, indicating the presence of C = O and C = C functional groups introduced by esterification with methacrylic anhydride [1]. The fabrication process of blank HAMA MN (B-MN) and Pal-MPAPO@MN is illustrated in Fig. 1a. In this process, HAMA solution loaded with Pal-MPAPO was injected into MN mold and exposed to UV light for gelation. To improve structural integrity and handling during application, a polyvinyl alcohol (PVA) backing layer was incorporated beneath the HAMA MN matrix. PVA was chosen for its biocompatibility and film-forming properties, which enabled uniform insertion force distribution and intact patch removal after needle swelling. Figure 1b presents a representative photograph of Pal-MPAPO@MN. Figure 1c presents the morphology of the fabricated MN array, showing regular needle arrangement. A layered structure was observed in the fluorescence images of the representative Pal-MPAPO@MN patch prepared using Rhodamine B-labeled HAMA and fluorescein-labeled PVA (Fig. 1d). Although the Pal-MPAPO@MN tips appeared slightly rounded, reliable skin insertion was still achieved, as confirmed by mechanical testing and histological microchannel formation. This is consistent with previous reports showing that insertion efficiency depends on both tip geometry and mechanical strength, rather than tip sharpness alone [24, 25]. Fluorescein isothiocyanate (FITC)-labeled Pal-MPAPO was further visualized by confocal laser scanning microscopy (CLSM), and the fluorescence image showed peptide-associated signal within the MN body, supporting successful peptide incorporation (Fig. 1e). As shown in Fig. 1f, the cumulative release profile showed a clear time-dependent release behavior, with cumulative release rates of 7.12 ± 0.60% at 0.5 h, 58.19 ± 1.55% at 12 h, 71.14 ± 1.04% at 24 h, 91.93 ± 1.11% at 48 h, and 94.98 ± 1.52% at 72 h (n = 3). Circular dichroism spectra of pristine Pal-MPAPO and Pal-MPAPO released from Pal-MPAPO@MN were highly similar, supporting preservation of peptide structural integrity after incorporation into the MN system (Fig. 1g). Failure force measurements showed good repeatability and batch consistency. As shown in Fig. 1h, the failure force of Pal-MPAPO@MN was 0.94 N (batch-wise mean values: 0.9353–0.9503 N; coefficients of variation: 0.44–1.41%), while B-MN showed a failure force of 0.88 N (batch-wise mean values: 0.8793–0.8848 N; coefficients of variation: 0.32–1.98%). Hematoxylin and eosin (H&E) staining further verified successful and uniform microchannel formation in ex vivo porcine skin without detectable needle residues or structural breakage (Fig. 1i). Despite exceeding maximum loading force, the MNs remained intact, displaying only tip blunting rather than brittle fracture or shedding (Fig. 1j). As depicted in Fig. 1k, the MNs gradually swelled after insertion and lost their original needle morphology within 3 h. Since B-MN and Pal-MPAPO@MN exhibited comparable morphology and mechanical performance, Pal-MPAPO loading was not expected to markedly alter the insertion behavior of the MN patch under the present conditions.

Fig. 1.

(a) Schematic diagram of the fabrication process of Pal-MPAPO@MN. (b) Representative photograph of the Pal-MPAPO@MN patch. (c) Scanning electron microscopy (SEM) images of the B-MN and Pal-MPAPO@MN. (d) Fluorescence microscopy images of B-MN and Pal-MPAPO@MN. (e) Confocal fluorescence image of Pal-MPAPO@MN prepared with FITC-labeled Pal-MPAPO. The fluorescence signal corresponds to the labeled peptide rather than the HAMA matrix, indicating peptide localization within the MN structure. (f) Relative cumulative release ratio curves of Pal-MPAPO in PBS solution (pH = 7.4) for 72 h from Pal-MPAPO@MN. (g) Circular dichroism spectra of pristine Pal-MPAPO (black) and Pal-MPAPO released from Pal-MPAPO@MN (red). (h) Mechanical displacement curves of the MNs. (i) H&E staining of porcine skin after MN insertion. (j) Post-test morphology of Pal-MPAPO@MN before and after mechanical loading. (k) Time-dependent morphological changes of Pal-MPAPO@MN after insertion (0, 1, 2, and 3 h)

Pal-MPAPO promotes proliferation of human dermal papilla cells (DPCs)

We investigated the effect of Pal-MPAPO on the proliferation of DPCs. DPCs were treated with different concentrations (0, 1, 10, 100, 200, 600, and 1000 nmol/L) of Pal-MPAPO for 24 h. As shown in Fig. S2 (Supporting Information), the percentages of cell viability were 97.93% (1 nmol/L), 100.1% (10 nmol/L), 104.0% (100 nmol/L), 103.4% (200 nmol/L), 94.5% (600 nmol/L), and 94.33% (1000 nmol/L), respectively. Compared with the control group, cells treated with 600 and 1000 nmol/L of Pal-MPAPO showed slightly reduced viability at high concentrations (p < 0.05), suggesting a safe and effective concentration range below 200 nM.

Pal-MPAPO activates the Wnt/β-catenin signaling pathway in DPCs

To further confirm whether Pal-MPAPO promoted hair growth through the Wnt/β-catenin signaling axis at the cellular level, the mRNA and protein expression levels of Wnt3a, Wnt10b, β-catenin, GSK3β, Lef-1, Cyclin D1, and Bcl-2 were assessed (Fig. 2a–h). Quantitative analysis of the western blot data is shown in Fig. S3 (Supporting Information). We also examined the phosphorylation level of GSK3β by western blotting. Consistent with the in vitro results, Pal-MPAPO treatment inhibited GSK3β activity, promoted β-catenin accumulation, and elevated the protein expression levels of Wnt3a, Wnt10b, Lef-1, Cyclin D1, and Bcl-2 in a dose-dependent manner. As expected, the p-GSK3β was upregulated after Pal-MPAPO treatments, further supporting its role in modulating Wnt/β-catenin signaling in DPCs (Fig. 2h).

Fig. 2.

Pal-MPAPO activated the Wnt/β-catenin signaling pathway in vitro study. (a-g) Relative mRNA expression levels of Wnt3a, Wnt10b, β-catenin, GSK3β, Lef-1, Cyclin D1 and Bcl-2 in the DPCs in different groups. (h) Wnt3a, Wnt10b, β-catenin, GSK3β, p-GSK3β, Lef-1, Cyclin D1 and Bcl-2 protein expression levels in the DPCs, detected by western blotting. The grouping of blots was cropped from different gels. β-actin was used as the control for total proteins. Quantitative analysis of Wnt3a, Wnt10b, β-catenin, GSK3β, p-GSK3β, Lef-1, Cyclin D1 and Bcl-2 protein levels (Fig. S3). Data points represent mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Molecular dynamics simulation and molecular docking

Molecular docking revealed that Pal-MPAPO effectively binds the VPAC1 target.

protein, with a binding energy of -5.6 kcal/mol (Fig. 3a). Hydrogen bonds were observed between specific residues such as Ser3, Ser10 of Pal-MPAPO. His171 and His180 of VPAC1 interact with the side chain of Pal-MPAPO residue Ser3 through π–π stacking and hydrogen bond interactions, suggesting a key anchoring role for these imidazole-containing residues. Tyr176 forms dual hydrogen bonds with Ser3 and Ser10 of Pal-MPAPO, indicative of a strong polar interaction network that reinforces the binding stability. Notably, Thr274 and Lys29 were positioned to potentially form a stabilizing salt bridge, further reinforcing the peptide–receptor interface. Molecular dynamics simulations showed that the root mean square deviation (RMSD) of the Pal-MPAPO–VPAC1 complex rapidly increased to 0.5 nm during the 0–25 ns period, indicating significant conformational relaxation (Fig. 3b). RMSD stabilized at 0.5 nm after 25 ns. Root mean square fluctuation (RMSF) analysis indicated high fluctuations in residues 50–150 and 300–400, suggesting flexible regions (Fig. 3c). The radius of gyration (Rg) value decreased from 2.8 nm to 2.5 nm, indicating structural tightening (Fig. 3d). Free energy landscape revealed multiple stable states and a broad conformational dynamic range, potentially linked to functional conformational changes or flexible interface motions (Fig. 3e). These data support a stable and energetically favorable Pal-MPAPO–VPAC1 complex formation.

Fig. 3.

The complex of Pal-MPAPO with VPAC1 receptor was analyzed by molecular docking and molecular dynamics simulation. (a) The molecular docking of Pal-MPAPO and VPAC1. Pal-MPAPO was depicted in green, while VPAC1 was shown in gray. (b) The RMSD values (nm). (c) The RMSF values (nm). (d) The Rg values (nm) plotted against time (ns) for the complex. (e) Free energy landscape of the Pal-MPAPO–VPAC1 complex derived from molecular dynamics simulations. The landscape was projected onto the first two principal components (PC1 and PC2) obtained via principal component analysis (PCA) of the backbone atomic fluctuations. The color gradient represents Gibbs free energy (kcal/mol), with blue regions indicating low-energy, thermodynamically favorable conformational states, and red regions corresponding to high-energy, less populated states

Pal-MPAPO activates the cAMP–PKA signaling pathways in DPCs

To evaluate whether Pal-MPAPO interacts specifically with VPAC1 expressed in DPCs and activates the cAMP-PKA signaling pathway, we performed the following experiments. Initially, VPAC1 expression knockdown efficacy was assessed using small interfering RNA (siRNA) targeting VPAC1 (si-VPAC1) (Gene ID: 7433). DPCs were treated with Pal-MPAPO (200 nmol/L) alone or following VPAC1 knockdown via si-VPAC1. Compared with the control group, Pal-MPAPO showed clear binding to VPAC1 at the cell surface, with fluorescence localized predominantly around the cell membrane. In the third row, si-VPAC1 reduced Pal-MPAPO binding, as indicated by weaker green fluorescence (Fig. 4a). Subsequent quantification of intracellular cAMP levels via enzyme-linked immunosorbent assay (ELISA) revealed significantly higher cAMP concentrations in Pal-MPAPO-treated cells compared to control (p < 0.0001) and VPAC1-silenced groups (p < 0.0001) (Fig. 4b). Western blot analysis further confirmed significant upregulation of PKA in the Pal-MPAPO group relative to control (p < 0.001) and VPAC1-silenced cells (p < 0.01) (Fig. 4c,f). Additionally, p-GSK3β was significantly elevated by Pal-MPAPO treatment compared to control (p < 0.01) and VPAC1 silencing conditions (p < 0.05), and total GSK3β protein levels showed a significant difference between Pal-MPAPO-treated and control groups (p < 0.05), supporting the hypothesis that Pal-MPAPO activates the cAMP–PKA axis in a predominantly VPAC1-dependent manner in DPCs (Fig. 4d,e,f).

Fig. 4.

Pal-MPAPO binds to VPAC1 and activates the cAMP–PKA signaling pathways in DPCs. (a) Immunofluorescence staining confirmed that Pal-MPAPO specifically binds to the VPAC1 receptor on the cell surface. 4′,6-Diamidino-2-phenylindole (DAPI)-stained nuclei are shown in blue, and FITC-labeled Pal-MPAPO is shown in green. The merged images indicate the spatial localization of the fluorescent signals. Scale bar = 50 μm for all panels. (b) Intracellular cAMP levels were measured using an ELISA kit. (c-f) PKA, p-GSK3β and GSK3β protein expression levels detected by western blotting. β-actin was used as the control for total proteins. Data are presented as mean ± SD from three independent biological replicates. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Given that VIP-derived peptides may also interact with VPAC2, we next examined whether VPAC2 contributes to the signaling response induced by Pal-MPAPO. Fig. S4 showed that Pal-MPAPO treatment significantly increased VPAC1 mRNA expression (p < 0.01), whereas VPAC2 mRNA levels were not significantly altered (p = 0.8634). In addition, si-VPAC2 effectively reduced VPAC2 expression, confirming successful knockdown. ELISA analysis showed that Pal-MPAPO markedly increased intracellular cAMP levels compared with the control group (p < 0.0001).

Pal-MPAPO@MN for hair regeneration therapy in mice

As shown in Fig. 5a, to assess the effect of promoting hair growth in vivo, shaved C57BL/6 mice received daily topical administration of either 0.25%, 0.5%, or 1% Pal-MPAPO formulations, 2% minoxidil, or Pal-MPAPO@MN. Treatments were applied to depilated dorsal skin for 13 consecutive days, with weekly photographic documentation. Pal-MPAPO@MN patches were applied to the dorsal skin of mice on days 1, 3, 5, 7, 9, 11, 13, while solvent, minoxidil and Pal-MPAPO solution groups were topically administrated daily. The blank control group received neither pharmacological nor physical interventions throughout the experimental timeline. Representative photographs illustrated the overall morphology of the eight groups (Fig. 5b). The hair length was also recorded on days 7 and 14 (Fig. 5c,d). By day 7, Pal-MPAPO@MN-treated mice already demonstrated significant hair growth (2.64 ± 0.22 mm vs 1.88 ± 0.14 mm for control group; p < 0.0001), which escalated to 3.58 ± 0.48 mm by day 14, showing no significant difference compared to the minoxidil group (3.51 ± 0.57 mm). Dose-dependent effects were noted in Pal-MPAPO groups (L/M/H): high-dose (H) reached 2.67 ± 0.31 mm at day 7 and 3.40 ± 0.64 mm by day 14, paralleling minoxidil’s performance (p = 0.9988). In addition to hair length, regenerated hair coverage was quantified (Fig. 5e). By day 14, the Pal-MPAPO@MN group (85.79 ± 8.18%) and high-dose Pal-MPAPO (83.32 ± 9.40%) achieved hair coverage comparable to minoxidil (86.76 ± 4.04%), with both showing statistically significant improvements over the control group (55.71 ± 12.38%, p < 0.001).

Fig. 5.

In vivo evaluation of Pal-MPAPO for hair loss treatment. (a) Schematic representation of the hair loss therapy in a mouse model via Pal-MPAPO@MN patch and topical administration. Untreated mice served as the control group. (b) Representative photographs of mice treated with topical solvent, minoxidil, 0.25% Pal-MPAPO, 0.5% Pal-MPAPO, 1% Pal-MPAPO, blank HAMA MN patch (B-MN) and Pal-MPAPO@MN patch. (c) Quantitative analysis of growing hair length in each group on day 7 (n = 6), error bars indicate SD. (d) Quantitative analysis of growing hair length in each group on day 14 (n = 6), error bars indicate SD. (e) The ratio of the regenerated hair coverage area to the depilated area in each group on day 14. Data points represent mean ± SD (n = 6). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Histological evaluation of the treated skin tissues was performed using H&E staining along with immunofluorescence analysis (Fig. 6a). As shown in Fig. 6b, the Pal-MPAPO@MN group demonstrated robust hair follicle regeneration (101.0 ± 14.82), comparable to minoxidil (112.3 ± 36.45; p = 0.9732) and significantly higher than the control group (56.17 ± 13.96, p < 0.01). High-dose Pal-MPAPO (107.5 ± 21.29) also approached minoxidil efficacy (p = 0.9999), while mid-dose (94.17 ± 10.38) and low-dose (72.00 ± 24.78) groups showed a dose-dependent effect. Notably, blank MN patch (B-MN, 68.50 ± 11.47) and solvent control (52.33 ± 9.647) showed no significant improvement over the control group (p = 0.9578 and p = 0.9999, respectively), confirming the efficacy of Pal-MPAPO in promoting hair growth. As shown in Fig. 6c, skin thickness in the control, solvent, and B-MN groups remained relatively low and comparable. The Pal-MPAPO@MN group exhibited a mean skin thickness of 790.3 ± 15.3 μm, which was comparable to that observed in the minoxidil group (823.3 ± 22.7 μm).

Fig. 6.

(a) H&E staining of the treated skin for different groups at day 14 (scale bars = 200 μm for all panels). (b) Quantitative analysis of the number of hair follicles in each group on day 14 (n = 6). (c) Skin thickness on day 14 (n = 6). Data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

The hair growth promoting efficacy was also verified by immunofluorescence staining analysis of CD31 (endothelial marker), K15 (hair follicle stem cell marker) and PCNA (cell proliferation marker) in the tissues from the shaved region (Fig. 7a,b). Quantitative analysis of immunofluorescence staining showed that treatments with high-dose Pal-MPAPO and Pal-MPAPO@MN significantly enhanced the expression of CD31, K15, and PCNA proteins compared to control and solvent groups (Fig. 7c–e). Notably, Pal-MPAPO-H and Pal-MPAPO@MN elicited an equivalent response relative to minoxidil. These findings support the conclusion that Pal-MPAPO stimulates neovascularization, follicular stem cell activation, and cell proliferation.

Fig. 7.

(a) Representative images of immunofluorescence staining of CD31 on depilated skin of different groups on day 14 (red: CD31, blue: DAPI). (b) Representative images of immunofluorescence staining of K15 and PCNA on depilated skin of different groups on day 14 (green: K15, red: PCNA, blue: DAPI). Scale bars represent 50 μm for all panels. Quantitative analysis of CD31 (c), K15 (d) and PCNA (e) positive cells at day 14 in each field (n = 4). Data points represent mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

The Wnt/β-catenin pathway plays an essential role in the induction of hair growth [26]. To determine whether Pal-MPAPO regulates this pathway in developing hair follicles, quantitative real-time polymerase chain reaction (qRT-PCR) and western blotting were performed to examine the mRNA expression and relative protein levels of Wnt/β-catenin signaling intermediaries, including Wnt3a, β-catenin, GSK3β and downstream effectors (Lef-1, Cyclin D1, Bcl-2). The results of the qRT-PCR analyses revealed significantly increased expression of Wnt3a, β-catenin, Lef-1, Cyclin D1 and Bcl-2 in the Pal-MPAPO-H and Pal-MPAPO@MN groups (Fig. 8a–f). According to Fig. 8g, compared with the control group, the protein expression of Wnt3a, β-catenin, p-GSK3β, Lef-1, Cyclin D1 and Bcl-2 also increased in the Pal-MPAPO-H and Pal-MPAPO@MN groups, suggesting effective activation of the Wnt/β-catenin pathway. Quantitative densitometric analysis of the western blot data is shown in Fig. S5 (Supporting Information).

Fig. 8.

Pal-MPAPO activated the Wnt/β-catenin signaling pathway. (a-f) Relative mRNA expression levels of Wnt3a, β-catenin, GSK3β, Cyclin D1, Lef-1 and Bcl-2 from three independent mice skin tissues, determined by qRT-PCR. (g) Wnt3a, β-catenin, GSK3β, p-GSK3β, Cyclin D1, Lef-1 and Bcl-2 protein expression levels in mice skin tissues, detected by western blotting. The grouping of blots was cropped from different gels. β-actin was used as the control for total proteins. Data points represent mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Materials and methods

Materials and reagents

Hyaluronic acid, methacrylic anhydride (MA), PVA, Rhodamine B, FITC, minoxidil, propylene glycol, and ethanol (95%) were purchased from Aladdin Biochemical Technology Co., Ltd (Shanghai, China). A custom-made polydimethylsiloxane (PDMS) mold (Beijing Zhongyuan Taihe Biotechnology Co., Ltd.) featuring a 10 × 10 array was employed for MN fabrication. The mold cavities were characterized by a height of 700 μm, a base diameter of 300 μm, and a groove depth of 300 μm. Isoflurane was purchased from RWD Co, Ltd (Shenzhen, China). Cell Counting Kit-8 (CCK-8) was obtained from Dojindo Laboratories (Kumamoto, Japan). Pal-MPAPO was custom synthesized by GL BioChem (Shanghai, China).

Preparation and characterization of HAMA

HAMA was prepared according to a previously reported method [1, 27]. HA was dissolved in deionized water (1 g in 100 mL) and stirred continuously to ensure complete dissolution. Subsequently, 5 mL of MA was added, and the pH was adjusted to 8.5–9.0 using NaOH solution. The reaction was carried out at 4 °C under constant stirring for 24 h. After that, the reaction solution was dialyzed in deionized water for 3 days, with the water being replaced every 4 h. After dialysis, the product was freeze-dried to obtain HAMA. FTIR spectroscopy (IS 10, Thermo, USA) was used to identify functional groups in HA, MA, and HAMA. Spectra were collected in the range of 2000–500 cm⁻¹.

Preparation and characterization of Pal-MPAPO-loaded hydrogel MNs

A 1% HAMA solution (0.5 mL) was added into a PDMS mold and subjected to negative pressure defoaming for 2–3 cycles. The prepared mold was then placed at 35 °C to heat and condense the HAMA solution. Following this, the mold was exposed to UV light (405 nm, 30 mW/cm²) for photocuring. Subsequently, 0.4 mL of a 5% PVA solution was added into the mold to form the MN base, followed by drying at 35 °C for 12 h. The final MNs were carefully removed from the mold. Pal-MPAPO-loaded HAMA hydrogel MNs (Pal-MPAPO@MN) were prepared by following the steps described above, with the addition of Pal-MPAPO to the HAMA solution in the first step. After fabrication, the MN patches were sealed in dry sterile aluminum foil bags containing desiccant and stored at 2–8 °C until use. The morphology of MNs was observed by SEM (Hitachi Regulus 8230, Japan). For imaging, fluorescent MNs were prepared using Rhodamine B (RB)-labeled HAMA (RB-HAMA) and fluorescein labeled PVA (CLSM, Zeiss LSM800). To visualize peptide localization in the MNs, Pal-MPAPO was labeled with FITC, and peptide-loaded MN patches were imaged by CLSM.

Release behavior of Pal-MPAPO from MNs and molecular structural integrity

The actual peptide loading of Pal-MPAPO@MN patches was first determined by complete extraction and subsequent reversed-phase high-performance liquid chromatography (HPLC) analysis. For the in vitro release study, each patch was incubated in PBS (pH 7.4) at 37 °C with gentle shaking. At specified time points from 0.5 to 72 h, aliquots of the release medium were collected and replaced with an equal volume of fresh prewarmed PBS. Pal-MPAPO in each sample was quantified by HPLC, and the cumulative release was calculated as a percentage of the actual peptide loading per patch. All measurements were performed in triplicate. Furthermore, the structural integrity of the extracted peptide was verified by circular dichroism spectroscopy (Chirascan VX, England) through comparison with the pristine peptide.

Mechanical strength of Pal-MPAPO@MN

The mechanical performance of Pal-MPAPO@MN patches was measured using a displacement − force test station (Anton Paar, Austria). Centrally located, morphologically intact MNs were selected under an optical microscope, and a shielding setup was used to isolate the target needle from neighboring needles. The patch was placed on a rigid platform with the needle facing upward, and the probe approached vertically at 30 mm/min from an initial distance of 1 cm. The displacement and force were recorded as the sensor moved 0.5 mm toward the patch backing. To evaluate repeatability and batch-to-batch consistency, failure force was measured using MNs prepared from three independent fabrication batches, with three samples tested per batch.

Culture and cell viability assay of DPCs

Human DPCs were obtained from iCell (China), and cultured in a commercially optimized human DPC complete medium (iCell, iCell-0163a-002 h, China), consisting of a basal medium (1 ×), a proprietary growth supplement (100 ×), fetal bovine serum (FBS; final concentration 10%), and penicillin–streptomycin (P/S; 100 ×), according to the manufacturer’s instructions. Cells were maintained in a humidified incubator at 37 °C with 5% CO₂. The effect of Pal-MPAPO on the viability of DPCs was assessed using CCK-8 assay (Dojindo, Japan) following the manufacturer’s protocols. DPCs were seeded in 96-well plates (1 × 10⁵ cells/mL, 100 μL/well), treated with experimental compounds, and incubated (37 °C, 5% CO₂). After 24 h exposure, 10 μL CCK-8 reagent was added to each well, followed by 1-h incubation. Absorbance at 450 nm was quantified using a BIORAD microplate reader, with blank-corrected values normalized to untreated controls.

qRT-PCR analysis

Total RNA was extracted from cultured DPCs or mouse dorsal skin tissues using Trizol reagent (Invitrogen, USA). Complementary DNA (cDNA) was synthesized using oligodT and the HelixCript™ Thermo Reverse Transcription System (Nanohelix, USA), according to the manufacturer’s instructions. The genes were selected for real-time PCR analysis using the Power SYBR Green PCR Master Mix (Takara Bio, Japan). The primer sequences and antibody information are provided in Tables S1–S4 (Supporting Information). β-actin was used as the internal reference. The relative expression levels were calculated using the 2^−ΔΔCt method.

Western blot analysis

Western blotting was performed according to a previously reported method [15]. Mouse dorsal skin tissues or DPCs were washed with phosphate-buffered saline (PBS, pH 7.4) and lysed in RIPA buffer (Beyotime, China). After centrifugation, the supernatants were collected, and protein concentrations were quantified using a bicinchoninic acid assay kit (Thermo, USA) according to the manufacturer’s instructions. Equal amounts of protein were separated by 10–12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to the PVDF membrane (Merck Millipore, MA, USA). After blocking with bovine serum albumin (BSA) dissolved in Tris-buffered saline with Tween-20 (TBST) (Beyotime, China), the membrane was incubated with primary antibodies at 4 °C for 24 h, followed by incubation with secondary antibodies for 1.5 h at room temperature. Protein bands were visualized using an ECL detection kit (Beyotime, China), and their intensity was quantified using ImageJ software (v1.54f; National Institutes of Health, USA).

Molecular docking and molecular dynamics simulations

The palmitic acid structure was built using Gaussian 16 (Revision C.01; Gaussian, Inc., USA), and the modified cysteine was reintroduced into the protein model for molecular optimization. The structure of Pal-MPAPO was then obtained. VPAC1(Protein Data Bank [PDB] ID: 8E3Y) receptor structure was downloaded from the Protein Data Bank. Molecular docking between Pal-MPAPO and VPAC1 was performed, and the initial complex conformation was predicted using AlphaFold3 (Google DeepMind, 2024). The obtained docking poses were visualized with PyMOL (v3.1; Schrödinger, LLC), and key interface residues and binding energies were analyzed with PDBePISA. Docking poses were validated by 100-ns molecular dynamics simulations using GROMACS (v2024.3) with the AMBER99SB force field and TIP3P water model. The system underwent standard energy minimization, followed by equilibration under constant volume and temperature (NVT) and constant pressure and temperature (NPT) conditions prior to production runs.

Immunofluorescence staining

Small interfering RNA (siRNA) targeting VPAC1 (si-VPAC1) was synthesized and transfected into DPCs. DPCs were seeded in a 12-well chamber (2 × 10⁴ cells/well) and cultured in the same iCell complete medium described above. Lipofectamine™ 3000 (Invitrogen) was diluted in Opti-MEM (Gibco) and mixed thoroughly. The vector was diluted in Opti-MEM and mixed with P3000™ (Invitrogen). The diluted Lipofectamine™ 3000 was mixed with the vector (1:1) and incubated at room temperature for 5 min. The mixture was added to the culture medium, gently mixed, and incubated at 37 °C for 3 days. For fluorescence imaging, Pal-MPAPO was labeled with FITC before incubation with DPCs. After washing, nuclei were counterstained with DAPI (Sigma-Aldrich, D9542), and cells were observed under an immunofluorescence microscope (Olympus TH4-200, Japan).

Quantitative assay for cAMP

The culture supernatant was harvested from the cultured DPCs. si-VPAC1 (Gene ID: 7433) and si-VPAC2 (Gene ID: 7434) were synthesized by GenePharma (Shanghai, China). Intracellular cAMP levels were measured using an ELISA kit (Elabscience, E-EL-0056) according to the manufacturer’s instructions.. Absorbance was determined at 450 nm using a microplate reader.

Animal experiments

All animal experiments were conducted in accordance with the procedures approved by the Animal Ethics Committee of Southern Medical University. A total of 48 healthy male C57BL/6 mice (SPF grade, 6 weeks old, 20 ± 2 g, purchased from Laboratory Animal Center of Southern Medical University) were used in this study. After a one-week acclimation period, the back skin of the mice was shaved from a 2 × 4 cm² area using an electric hair clipper and depilated with a depilatory cream. Mice without obvious wounds on the back were randomly divided into 8 groups (n = 6 per group): Control group, Solvent group, Minoxidil group, 0.25% Pal-MPAPO group (Pal-MPAPO-L), 0.5% Pal-MPAPO group (Pal-MPAPO-M), 1% Pal-MPAPO group (Pal-MPAPO-H), Blank HAMA MN group (B-MN) and Pal-MPAPO@MN group. In the non-MN groups, propylene glycol and ethanol were used as primary solvents at a concentration of 25%:55% (v/v). B-MN and Pal-MPAPO@MN were topically applied to the depilated site of the mice once every other day. In the other groups, solvent, 2% minoxidil (positive control), and Pal-MPAPO solutions were topically applied to the dorsal skin at a dosage of 20 µL per cm² (twice daily) for 13 consecutive days. The 2% minoxidil concentration was selected as a widely accepted benchmark positive control in C57BL/6 hair-regrowth and anagen-induction models, consistent with previous preclinical studies and U.S. Food and Drug Administration (FDA)-approved topical formulations [28–31]. Hydrogel MN patches were applied every other day to match their sustained-release behavior and to model a reduced dosing frequency under the present experimental design. Hair growth was documented by taking pictures on days 7 and 14 after treatment.

Skin tissue histology

On day 14, mice were anesthetized with isoflurane and sacrificed by cervical dislocation. The dorsal skin tissues were collected and fixed in 4% paraformaldehyde for 24 h. The tissues were then embedded in paraffin and sectioned into 5 μm slices. H&E staining was performed following standard procedures. Immunofluorescence staining was used to examine CD31 (Abcam, ab28364) protein expression in dorsal skin, and double immunolabeling was conducted to detect keratin 15 (K15; Abcam, ab80522) and proliferating cell nuclear antigen (PCNA; Abcam, ab29).

Statistical analysis

Data were presented as mean ± standard deviation. Multiple comparisons between different groups were performed using one-way ANOVA with GraphPad Prism (v10.0; GraphPad Software, USA). Statistical significance is indicated by asterisks (*p < 0.05; **p < 0.01; ***p < 0.001, ****p < 0.0001).

Discussion

Currently, S-palmitoylation is predominantly studied as an intracellular post-translational modification, with its dynamic cycling of palmitoylation and deacylation regulating many cellular and physiological processes, including signal transduction, cell migration and division, inter-organelle and intercellular communication, energy metabolism, innate immunity and tissue function [12, 32, 33]. As a result, disruptions in S-palmitoylation and deacylation dynamics are increasingly found to be associated with various human disorders, including metabolic and inflammatory diseases, cancer, and hair loss disorders such as alopecia [13, 34–36]. These findings suggest that palmitoylated peptides may serve as functional modulators of skin and follicular biology. Based on the theoretical insights of the above biological processes, we introduced a palmitoyl moiety into MPAPO to generate Pal-MPAPO, with the intention of enhancing membrane affinity, receptor interaction, and local retention within the skin microenvironment.

Efficient intradermal delivery of macromolecular peptides remains a major translational challenge. Conventional drug delivery systems, such as transdermal sprays and topical formulations, provide a larger application area but suffer from limited penetration, which reduces therapeutic efficacy [37]. MNs have therefore attracted increasing attention as minimally invasive transdermal drug delivery platforms, offering the advantages of controlled drug release and improved local bioavailability [38, 39]. However, many current MN systems incorporate nanoparticles as drug carriers, which not only increases production complexity and costs, but may also introduce potential biocompatibility concerns [3, 40–43]. To address these limitations, we developed a HAMA-PVA hydrogel MN system for the delivery of Pal-MPAPO, providing a hydrogel-based transdermal delivery approach for hair loss treatment. Compared with nanoparticle-loaded MNs reported in other studies, our HAMA-PVA hydrogel MNs exhibit advantages in fabrication simplicity and structural design. The preparation process of this material is well-established and stable, facilitating reproducible fabrication and potential clinical translation. The post-test morphology further suggested that excessive loading caused mild tip deformation rather than brittle fracture. From a safety perspective, it reduces the risk of polymer fragments being retained in the skin, although the observed deformation also indicates that the MN patches should be considered single-use devices rather than reusable arrays. In vivo, this platform allowed less frequent administration while maintaining therapeutic efficacy, supporting its potential as a sustained local delivery strategy. For clinical use in alopecia, both the mechanical compatibility of MNs with scalp tissue and the potential for sensory irritation should be considered. The scalp’s mechanical properties differ from other skin areas; however, hydrogel-forming MNs are inserted in a rigid, dry state and soften after hydration. Previous studies have demonstrated sufficient insertion strength (as low as 0.03 N/needle for 100% penetration in skin models) before swelling [44–46]. Importantly, the penetration depth of MNs is typically limited to the superficial dermis, while major cutaneous nerve bundles reside deeper in the dermis/subcutis [44, 47]. This reduces the likelihood of persistent neural stimulation. This aligns with clinical experience: repeated scalp microneedling for androgenetic alopecia is generally well tolerated and mainly associated with transient pain, scalp irritation, and mild erythema [48–51]. We similarly observed no sustained irritation or abnormal histological changes in treated mouse skin. However, given differences between murine and human scalp (e.g., thickness, innervation density), future work should evaluate comfort and long-term safety using human scalp-relevant models.

After establishing the delivery framework, we investigated the biological mechanism of Pal-MPAPO. The Wnt/β-catenin signaling pathway plays an important role in the regulation and maintenance of DPC function, with enhanced Wnt activity in DPCs recognized as a critical driver of hair regeneration [52–54]. In the canonical Wnt/β-catenin signaling pathway, Wnt proteins bind to the Frizzled receptor and the low-density lipoprotein receptor-related protein (LRP) co-receptor, leading to the inhibition of GSK3β [26]. By blocking β-catenin phosphorylation and ubiquitin-dependent degradation, this inhibition stabilizes cytoplasmic β-catenin. Stabilized β-catenin translocates to the nucleus and interacts with T-cell factor (TCF)/lymphoid enhancer factor (LEF), driving the transcription of genes that regulate cell proliferation [26, 55]. Among the Wnt family proteins, Wnt3a has been demonstrated to activate β-catenin signaling, significantly enhancing hair growth in nude mice implanted with skin constructs comprising DPCs and keratinocytes [56]. Additionally, Wnt10b plays a key role in facilitating the transition of hair follicles from the telogen phase to the anagen phase, thereby supporting the initiation of new hair growth [57, 58]. In our study, Pal-MPAPO markedly increased the expression of Wnt3a, Wnt10b, β-catenin, and downstream effectors including Lef-1, Cyclin D1, and Bcl-2 in DPCs. Computational analysis suggested a potential interaction between Pal-MPAPO and the VPAC1 receptor, which was further supported by siRNA knockdown and intracellular cAMP measurements. Between the two major receptors (VPAC1 and VPAC2) of vasoactive intestinal peptide (VIP), VPAC1 was selected as the primary target in this study. This choice is based on evidence that VPAC2 primarily governs pigmentary responses to VIP, whereas the anti-inflammatory effects of VIP are predominantly mediated through the VPAC1 receptor [59]. VPAC1 also plays a role in immune modulation and hair follicle homeostasis [11]. Our results indicate that Pal-MPAPO acts as a functional agonist of VPAC1, leading to a significant increase in intracellular cAMP levels in a predominantly VPAC1-dependent manner in DPCs, a key second messenger known to regulate follicular proliferation. To further clarify receptor specificity, we additionally evaluated the potential involvement of VPAC2 in the same cellular model. Pal-MPAPO treatment did not significantly alter VPAC2 expression, whereas si-VPAC2 effectively reduced VPAC2 mRNA levels, confirming successful knockdown. However, under Pal-MPAPO stimulation, VPAC2 knockdown did not significantly reduce intracellular cAMP accumulation. Together with the marked reduction observed after VPAC1 knockdown, these findings support that the cAMP–PKA response induced by Pal-MPAPO in DPCs is predominantly VPAC1-dependent under the present experimental conditions. Nevertheless, because a minor contribution of VPAC2 cannot be completely excluded, future studies using receptor-selective antagonists or comparative binding analyses will be valuable to further define the relative roles of VPAC1 and VPAC2. This mirrors previous findings where VIP binding to VPAC1 induced cAMP-mediated activation of the PKA pathway, enhancing cell survival and function [60]. In hair follicle biology, cAMP signaling is particularly relevant as it influences dermal papilla cell activity and growth factor production [61, 62]. Given that PKA activation leads to GSK3β inhibition, Pal-MPAPO may stabilize β-catenin by integrating cAMP and Wnt signaling. The interaction between Wnt/β-catenin and cAMP–PKA signaling suggests that Pal-MPAPO functions as a dual-pathway activator, amplifying hair follicle regenerative responses. VPAC1 activation triggers the cAMP–PKA pathway, which inhibits GSK3β and helps stabilize β-catenin, thereby indirectly enhancing Wnt/β-catenin signaling. As a molecular node connecting regenerative and immunomodulatory pathways, VPAC1 represents a promising therapeutic target for hair loss. Beyond promoting follicular proliferation, it may also contribute to the restoration of immune privilege within the hair follicle niche. The hair follicle is an immune privileged organ, largely due to the local suppression of antigen presentation and pro-inflammatory signaling [63]. Disruptions in immune privilege mechanisms have been linked to alopecia areata and other autoimmune-related hair disorders [63–65]. Previous studies have shown that VIP, an immunoinhibitory neuropeptide released by perifollicular sensory nerve fibres, possesses strong immunoregulatory properties, protecting hair follicles from immune attack [11, 66, 67]. Therefore, VPAC1 was selected as the primary target in our research, providing a basis for future studies on VPAC1-dependent regulation of immune homeostasis.

Importantly, the hair growth-promoting effect observed in vitro was supported by the in vivo findings. In the depilation-induced C57BL/6 mouse model, topical administration of 1% Pal-MPAPO significantly promoted hair regrowth, increased follicle density, and elevated CD31, K15, and PCNA expression compared with the control group, while showing no significant difference relative to the minoxidil-treated group. Although Pal-MPAPO achieved efficacy comparable to minoxidil, the two treatments act through different mechanisms. Minoxidil primarily acts through vascular modulation and potassium channels, while Pal-MPAPO directly regulates follicular signaling pathways such as Wnt/β-catenin and VPAC1-dependent cAMP–PKA. Thus, Pal-MPAPO may serve as a mechanistically complementary strategy for conventional therapy. The increased expression of CD31, K15, and PCNA indicates enhanced vascularization, stem cell activity, and cellular proliferation, respectively [68, 69]. Consistent with these histological observations, analysis of mouse dorsal skin also showed activation of the Wnt/β-catenin signaling pathway. Pal-MPAPO treatment increased Wnt3a and β-catenin expression, along with p-GSK3β and upregulation of Lef-1, Cyclin D1, and Bcl-2. Given that Wnt/β-catenin signaling is known to induce anagen and promote hair follicle proliferation, these molecular changes are consistent with the observed regenerative phenotype [26]. The B-MN group also showed a slight upregulation of several regenerative markers compared to the control group. This aligns with previous findings that MN-induced micro-injury can activate wound-healing-associated regenerative pathways, including Wnt/β-catenin signaling, and enhance hair growth [70–72]. In our study, this effect was limited at the mRNA level but was detectable for selected proteins, suggesting that B-MN should be viewed as a procedural control rather than a completely inert control. Therefore, the efficacy of Pal-MPAPO@MN should be viewed as an integrated effect that exceeds that of blank MN treatment alone. Although this study demonstrates activation of Wnt/β-catenin and cAMP–PKA signaling during treatment, these changes should be interpreted cautiously. Hair follicle regeneration typically involves transient and spatially restricted Wnt signaling within the follicle, not uniform activation across the epidermis [73]. Sustained or widespread epidermal activation, or disruption of the finely tuned homeostatic process in the skin and hair follicles, has been associated with pathological outcomes, whereas localized and temporally regulated activation is required for anagen induction and normal follicle cycling [73]. Our histological analyses showed no abnormal epidermal thickening or disorganized keratinocyte proliferation in non-follicular skin regions, suggesting that pathway activation was not broadly distributed across the epidermis under the present treatment conditions. Nevertheless, longer treatment duration studies will be necessary to evaluate potential cumulative signaling effects during chronic exposure. The depilation-induced C57BL/6 model primarily shows hair-cycle activation instead of pathological alopecia. Because hair follicle re-entry into anagen generally requires continued stimulation, our findings mainly reflect efficacy during treatment rather than long-term remission. Without a post-treatment withdrawal phase, it remains unclear whether Pal-MPAPO@MN delays relapse after treatment stops. Consequently, future research include a follow-up period are essential to evaluate the sustained therapeutic effects in alopecia models. Several bioactive peptides have been explored for hair loss treatment, including CXXC5 inhibitors and AIMP1-derived peptide. CXXC5 inhibitors, such as PTD-DBM, function by disrupting negative Wnt regulators, thereby promoting the expression of β-catenin, alkaline phosphatase, and PCNA [9]. However, their clinical application is limited by low bioavailability and the need for repeated topical administration. AIMP1-derived peptide has been shown to stimulate dermal papilla cell proliferation, but their mechanism of action does not directly target Wnt signaling, limiting sustained regenerative activity [10]. In contrast, Pal-MPAPO’s dual activation of Wnt/β-catenin and VPAC1–cAMP–PKA signaling represents a broader approach to hair regrowth. The combination of biochemical stability, dual signaling activation, and enhanced receptor targeting gives Pal-MPAPO a potential therapeutic advantage over existing bioactive peptides. Moreover, in combination with hydrogel MN delivery, this strategy enables localized exposure with reduced administration frequency, providing a localized delivery option for peptide-based follicular therapy.

Our study also has some limitations. First, the role of Pal-MPAPO in immune regulation remains to be fully explored. Given that VIP has been implicated in immune suppression and regulatory T-cell activation, Pal-MPAPO may exert broader immunomodulatory effects in alopecia areata, which were not specifically addressed in the present work. Additionally, topical and Pal-MPAPO@MN treatments were administered at different frequencies to match their distinct release characteristics. Future studies with dose- and frequency-normalized controls will further clarify the relative delivery efficiency. Finally, the depilation-induced C57BL/6 mouse model in this study mainly tests hair regrowth, not the complex pathology of clinical alopecia. Although this model is widely used for evaluating hair growth-promoting agents, it does not fully recapitulate immune- or genetically mediated hair loss conditions. Future work will employ disease-specific models to more comprehensively assess the therapeutic potential of Pal-MPAPO and Pal-MPAPO@MN in pathological alopecia.

Conclusion

This study introduces Pal-MPAPO as a novel bioactive peptide that promotes hair growth through dual activation of Wnt/β-catenin signaling and VPAC1-dependent cAMP–PKA signaling. A hydrogel-based microneedle delivery system was employed to facilitate localized transdermal administration of Pal-MPAPO and to achieve comparable therapeutic efficacy under a reduced application frequency relative to the topical regimen used in this study. Compared with previously reported bioactive peptides, Pal-MPAPO exhibited favorable biological activity and multi-pathway modulation, indicating its promise as a peptide candidate for hair regeneration. To further assess its translational potential, future work should evaluate long-term efficacy, pharmacokinetic profiles, and therapeutic effects in disease-relevant alopecia models. Overall, these findings support hydrogel MN-assisted delivery as a feasible localized administration strategy for peptides in hair loss therapy, although long-term durability and dose-normalized delivery efficiency remain to be established.

Author contributions

Tianyu Ma: Writing – review & editing, Writing – original draft, Validation, Methodology, Investigation, Formal analysis, Data curation. Qiang Li: Writing – review & editing, Writing – original draft, Validation, Methodology, Formal analysis, Data curation. Xingwu Chen: Writing – original draft, Validation, Methodology, Data curation. Jixiang Dong: Writing – original draft, Validation, Methodology, Data curation. Yuqing Jiang: Validation, Resources, Methodology, Investigation. Shanshan Li: Validation, Resources, Methodology, Formal analysis.

Funding

This study was supported by the National Natural Science Foundation of China (No. 82373773), the Guangdong Basic and Applied Basic Research Foundation (No. 2024A1515012921), and National undergraduate innovation and entrepreneurship training program (202410559101).

Data availability

The RNAseq datasets analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The use and care of experimental animals were approved by the Animal Ethics Committee of Southern Medical University (Approval No.44002100038716). All animal housing and experiments were conducted in strict accordance with the institutional guidelines for care and use of laboratory animals.

Consent for publication

Not applicable.

Competing interests

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