Hydroxypropyl-β-cyclodextrin co-delivery hydrogel of crisaborole and curcumin for combinational antioxidant/anti-inflammatory therapy of vitiligo

Abstract

Vitiligo is a depigmenting autoimmune disease driven by oxidative stress and immune-mediated melanocyte destruction. Current therapies are limited by poor drug permeability, systemic side effects, and insufficient efficacy. Herein, we developed a novel carbomer-based hydroxypropyl-β-cyclodextrin-encapsulated crisaborole/curcumin hydrogel (Cri/Cur@CD-Gel) for combinational transdermal intervention of vitiligo via dual antioxidant and anti-inflammatory pathways. Hydroxypropyl-β-cyclodextrin (HP-β-CD) inclusion complexation overcame the inherent hydrophobicity of crisaborole (Cri) and curcumin (Cur), significantly enhancing their aqueous solubility, stability, and transdermal permeability. Leveraging its skin-adhesive properties, the carbomer hydrogel matrix enhanced drug retention and enabled sustained release at lesion sites, thereby improving drug bioavailability. In vitro, Cri/Cur@CD, leveraging the solubilizing power of HP-β-CD, restored melanocyte viability under oxidative stress, scavenged ROS, normalized mitochondrial membrane potential, and elevated cAMP levels to enhance melanogenesis and tyrosinase activity. In vivo, Cri/Cur@CD-Gel achieved significant repigmentation in monobenzone-induced vitiligo mice, markedly reducing epidermal ROS, CD8⁺ T-cell infiltration, and inflammatory cytokines (TNF-α, IL-1β, and IL-6). This study demonstrates that the combination of HP-β-CD-based encapsulation and carbomer hydrogel delivery serves as a potential topical administration platform for co-delivering hydrophobic crisaborole and curcumin, and establishes a promising strategy to alleviate vitiligo, that concurrently addresses oxidative damage, inflammatory responses, and melanocyte regeneration.

Graphical abstract

A carbomer-based hydroxypropyl-β-cyclodextrin-encapsulated crisaborole/curcumin hydrogel (Cri/Cur@CD-Gel) was prepared for combinational transdermal treatment of vitiligo via dual antioxidant and anti-inflammatory pathways. Thanks to the inclusion complex hydrogel system, the solubility and transdermal ability of Cri and Cur were significant improved, for which the anti-inflammatory activity of Cri and the antioxidant capacity of Cur were fully exerted on the vitiligo lesion. And significant relief of oxidative stress, reduction of inflammatory response, and restoration of fur color were observed in the treatment of vitiligo mouse models.

1. Introduction

Vitiligo is an autoimmune-related depigmentation disease with a global prevalence of 1 %–2 %, which is characterized by melanin loss in the epidermis and hair follicles, and manifests as white patches on the face, limbs, or entire body [1,2]. These lesions impair patients' appearance, cause severe psychological burdens and social stigma, and significantly compromise quality of life [[3], [4]]. Current clinical drug approaches for vitiligo include topical immunosuppressants (corticosteroids and calcineurin inhibitors), narrowband ultraviolet B (NB-UVB) phototherapy, and short-term high-dose corticosteroids to control acute progression [1,5]. However, these traditional treatments result in unsatisfactory outcomes due to limited skin permeation of topical agents and frequent adverse effects (such as skin atrophy, telangiectasia, acne, and rash) caused by prolonged use of glucocorticoids and calcineurin inhibitors [[5], [6], [7], [8]], alongside the limitations of NB-UVB phototherapy, including non-universal efficacy, failure to repigment all anatomic locations, and cumbersome treatment regimens for patients [9,10]. Therefore, it is crucial to develop safe topical formulations for vitiligo treatment, capable of efficient skin penetration, anti-inflammatory activity, melanocyte protection, and enhanced melanogenesis (see Scheme 1).

Scheme 1.

Schematic Representation of Vitiligo Mice Treated with Cri/Cur@CD. (a) Preparation of Cri/Cur@CD by freeze-drying method and preparation of Cri/Cur@CD-Gel by adjusting pH to 6 in carbomer 940 solution. (b) Cri/Cur@CD therapy for vitiligo model by combinational antioxidant and anti-inflammatory.

Transdermal drug delivery systems (TDDS) are drug administration methods effective for treating dermatological conditions, which deliver medications through the skin, enhance drug permeation across the cutaneous barrier, and reduce dosage requirements while minimizing systemic side effects [[11], [12], [13]]. In recent years, novel transdermal formulations such as liposomes and microneedles used to deliver immunosuppressants for vitiligo have emerged as a research hotspot. Hesham et al. demonstrated that ethanol-modified liposomes transdermally delivered tofacitinib, a Janus Kinase inhibitor, achieving anti-inflammatory and repigment effects in animal models [14]. And Ebrahim et al. employed microneedles to deliver the calcineurin inhibitor tacrolimus for vitiligo treatment, showing superior repigment efficacy compared to ointment formulations [15]. However, immunosuppressants with broad biological activity could easily lead to multiple adverse effects as mentioned earlier, and due to the complex etiology of vitiligo, strategies solely targeting immunomodulation rather than comprehensive therapy fail to meet clinical needs.

It is noteworthy that the two key factors identified in the initiation and progression of vitiligo are oxidative stress damage to melanocytes driven by reactive oxygen species (ROS) accumulation, and immune responses mediated by melanocyte-cytotoxic CD8⁺ T cells [[16], [17], [18]]. Accordingly, a comprehensive therapeutic strategy that transdermally delivers antioxidant and anti-inflammatory drugs represents a superior treatment option for alleviating oxidative stress, modulating inflammatory responses, and promoting melanogenesis [[19], [20], [21]]. Doppalapudi et al. developed deformable liposomes co-loaded with melanogenesis promoter psoralen and antioxidant resveratrol, and demonstrated their efficacy in stimulating melanin production and mitigating oxidative stress in cellular experiments [22]. Furthermore, Sun et al. [23] designed flexible liposomes encapsulating the methylprednisolone and polydopamine for effectively improving vitiligo symptoms in mice, in which the liposomes facilitated transdermal drug delivery, methylprednisolone regulated inflammation and polydopamine scavenged ROS. Nevertheless, the application of these liposome formulations is limited due to their complicated preparation processes, instability of ingredients and stringent storage requirements.

As a natural polyphenolic compound, curcumin (Cur) has potential to protect melanocytes in vitiligo lesions from oxidative stress by scavenging ROS and activating the Nrf2 signaling pathway to upregulate antioxidant enzyme expression [[24], [25], [26], [27]]. Consequently, the combination of curcumin with immunosuppressants exhibits comprehensive synergistic advantages. However, the application of Cur in vitiligo treatment research is severely limited by its extremely poor aqueous solubility and low bioavailability [[28], [29], [30]].

The novel immunosuppressant crisaborole (Cri) approved by the FDA for treating atopic dermatitis in infants as young as 3 months [31], has demonstrated pigment-repopulation effects in both monotherapy and combination therapy during clinical practice [32,33]. This suggests its potential as a safe therapeutic agent for vitiligo. As a phosphodiesterase-4 (PDE-4) inhibitor, Cri elevates intracellular cyclic adenosine monophosphate (cAMP) levels by inhibiting PDE-4-mediated hydrolysis, which exerts dual therapeutic mechanism exerts against vitiligo. For immune, Cri reduces T-cell infiltration and suppresses pro-inflammatory cytokine production, at the same time, elevated cAMP upregulates the expression of tyrosinase (TYR) and tyrosinase-related protein-2 (TRP-2), thereby promoting melanogenesis [[34], [35], [36]]. Therefore, combination therapy of Cri and Cur via effective dermal delivery systems is anticipated to achieve significant therapeutic efficacy.

For delivering small molecule drugs, cyclodextrin inclusion complexes serving as TDDS have been extensively utilized in recent years, in which β-cyclodextrin (β-CD) and its derivatives consist of seven glucose units forming a cylindrical configuration with a hydrophobic internal cavity and a hydrophilic external surface composed of multiple hydroxyl groups [37]. This unique structure enables them to encapsulate hydrophobic drug molecules as guests within the cavity via host-guest supramolecular interactions, thereby enhancing drug solubility, stability, and bioavailability [38]. Furthermore, this mechanism facilitates controlled release of hydrophobic drugs, prevents rapid drug aggregation, and improves therapeutic safety [39]. Additionally, compared to β-CD, structurally modified hydroxypropyl-β-cyclodextrin (HP-β-CD) exhibits further improved water solubility, biosafety, and drug-loading capacity, which significantly enhances the skin permeation efficiency of hydrophobic drugs [40]. For instance, Ventura et al.Ventura et al. demonstrated that HP-β-CD-encapsulated celecoxib enhanced transdermal permeation by accelerating dissolution kinetics and directly modulating stratum corneum permeability [41]. Therefore, this study employs HP-β-CD to co-load Cri and Cur, formulating an inclusion complex (Crisaborole/Curcumin@HP-β-CD, Cri/Cur@CD). The inclusion effect of HP-β-CD is leveraged to improve the solubility, dermal penetration capability, and bioavailability of the hydrophobic drugs Cri and Cur, aiming to simultaneously harness Cri's anti-inflammatory action and melanogenesis-promoting capacity, along with Cur's ability to alleviate oxidative stress.

Carbomer 940 is a polymeric acrylic crosslinked polymer that forms a gel at pH close to the skin [42]. It is widely used in cosmetics, pharmaceuticals, and daily chemical products to enhance the retention time of drugs on the skin surface [43]. Based on this, we developed a Cri/Cur@CD inclusion hydrogel (Cri/Cur@CD-Gel) based on carbomer 940. This formulation not only enhances skin permeability of the drugs, but also enables sustained drug release at lesion sites, thus improving the therapeutic efficacy while reducing the systemic side effects. In a monobenzone-induced vitiligo mouse model, the Cri/Cur@CD-Gel demonstrated significant therapeutic efficacy, including robust skin repigment, effective inhibition of inflammation, potent alleviation of oxidative stress, and restored hair color. This study establishes a proof-of-concept strategy through a novel drug formulation for clinical vitiligo management, potentially providing patients with a safer and more effective treatment approach.

2. Materials and methods

2.1. Materials

Crisaborole (Cri), curcumin (Cur), hydroxypropyl-β-cyclodextrin (HP-β-CD) and carbomer 940 were purchased from Aladdin. The assay kits of superoxide anion radicals (O₂⁻•) and hydroxyl radicals (HO•) were purchased from Nanjing Jiancheng Bioengineering Institute of China. 2,2ʹ-Azino-bis (3-ethylbenzothiazoline 6-sulfonate) radicals (ABTS⁺•) and 2,2-Diphenyl-1-picrylhydrazyl radicals (DPPH) was purchased from Shanghai Beyotime Biotechnology of China. CCK-8 kit was purchased from Digital Molecular Technology (Shanghai, China). cAMP ELISA Kit was purchased from YIFEIXUE Biotech. Co., Ltd. TNF-α, IL-1β, IL-6 ELISA kits were purchased from BioLegend, Inc. ROS detection kit (DFCH-DA), mitochondrial membrane potential detection kit (JC-1), the malondialdehyde (MDA), superoxide dismutase (SOD), and g malondialdehyde (MDA) measurement kits were purchased from Beijing Solarbio Science & Technology Co.,Ltd. Other reagents and solvents were all commercially available and used without further purification.

2.2. Preparation and characterization of inclusion compound

Firstly, 1 mL of the mixture ethanol solution (Cri 20 mg/mL, Cur 10 mg/mL) was syringe-injected into 4 mL of HP-β-CD aqueous (25 mg/mL) at 40 °C. After 24 h dark incubation under continuous stirring and then rotary evaporating to removing ethanol, the mixture was freeze-dried, reconstituted in 5 mL deionized water, and sonicated for 5 min. The resulting solution was filtered through a 0.22 μm nylon membrane, and the filtrate was freeze-dried to obtain Cri/Cur@CD. Single drug complexes (Cri@CD, Cur@CD) were synthesized using similar methods.

The morphology of HP-β-CD, Cri, Cur, physical mixture (Cri, Cur and HP-β-CD), and Cri/Cur@CD were characterized by scanning electron microscopy (SEM, ZEISS Sigma 360). The UV–vis, Fourier transform infrared spectra (FTIR), Nuclear magnetic resonance hydrogen spectrum (¹H NMR) and X-ray diffraction (XRD) were respectively characterized by UV–visible spectroscopy (UV–vis, Cary 60 UV–Vis), Fourier transform infrared spectroscopy (FTIR, Spectrum Two), Nuclear magnetic resonance spectrometer (Bruker 400 MHz) and X-ray diffractometer (Rigaku D/MAX-2600). Drug quantification of Cri/Cur@CD was performed via calibration curves by UV–vis spectroscopy. The 10 mg of Cri/Cur@CD were dissolved in anhydrous ethanol (1 mL) and subjected to sequential dilution for spectral analysis. After accurate dilution, the concentration of Cur was quantified at 425 nm, and that of Cri at 251 nm. The drug loading efficiency was determined by the ratio of the mass of the drug components to the mass of the inclusion complex. The antioxidant activities were evaluated by measuring scavenging effects on DPPH, ABTS⁺•, HO•, and O₂⁻• radicals, using a series of gradient Cri/Cur@CD and Cur ethanol solutions (with equivalent Cur concentrations at 100, 80, 60, 40, 20, and 10 μmol/L).

2.3. Preparation and characterization of hydrogel

The inclusion compound hydrogel (Cri/Cur@CD-Gel) was synthesized by adding 0.5 g Cri/Cur@CD to 5 mL 1 % (w/w) carbomer 940 under 25 °C, following by adjusting pH to 6.0 with 0.1 M NaOH. Rheological properties of Cri/Cur@CD-Gel at different pH values were analyzed using a rheometer (Anton Paar MCR 302) with 0.1–100 Hz frequency, and 1 % fixed stress. The flow behavior of gels with different pH values was monitored by tilt test for 1 min at 25 °C. The morphology of Cri/Cur@CD-Gel was examined by SEM. Drug release was studied via dialysis method, and the cumulative release rate (CRR) was calculated using drug content in dialysis medium. The concentration of Cur and Cri were quantified using a UV–vis spectroscopy respectively at 425 and 251 nm.

2.4. Transdermal penetration test

Dorsal skin of C57BL/6 mice was mounted between the chambers of a Franz diffusion cell, with the stratum corneum facing the donor chamber. The receptor chamber of the Franz diffusion cell was filled with normal saline, while the donor chamber received 1 mL of test solutions (Cri/CD@CD solution, Cri suspension or Cur suspension at equivalent doses). The transdermal permeation experiment was conducted at 37 °C, and at predetermined intervals, 1 mL of sample was collected from the receptor chamber, immediately replenished with fresh normal saline. The sample was filtered through a 0.22 μm membrane filter, and quantified via UV–vis standard calibration curves. Cumulative permeation rates were subsequently calculated based on the quantified data.

2.5. In vitro cytotoxicity and cell uptake

Normal skin melanocytes (PIG1) and immortalized keratinocytes (Hacat) were cultured in DMEM high-glucose medium with 10 % FBS and 1 % penicillin-streptomycin at 37 °C (5 % CO₂, 70–80 % humidity). Cri and Cur solutions (100–800 μmol/L in DMSO) and Cri/Cur@CD at matching concentrations were tested. Cells (1 × 10⁴/well in 96-well plates) were treated with drug-containing DMEM for 24 h, and viability was assessed via CCK-8. For cellular uptake, Cri/Cur@CD was fluorescently labeled with FITC (0.5 mL, 1 mg/mL). PIG1 cells in 6-well plates were treated with 10 % fluorescent complex (0–4 h), then analyzed by flow cytometry (after trypsinization and PBS washing) or confocal microscopy (after fixation with 4 % paraformaldehyde and DAPI staining).

2.6. Antioxidant stress and anti-inflammatory activity in vitro

PIG1 cells were cultured in 6-well plates and stimulated with 1 mg/mL LPS for 12 h to establish the model. After 24 h drug treatment, culture medium was collected and centrifuged (4 °C, 500 rpm, 5 min). TNF-α, IL-1β, and IL-6 levels in supernatant were quantified using ELISA kits. PIG1 cells were seeded in 96-well plates at 1 × 10⁴ cells/well and exposed to TBHP (1–5 mmol/L) for 2 h to establish an oxidative stress model, with the concentration inducing 50 % cell viability reduction (determined by CCK-8 assay) selected for subsequent experiments. Following TBHP exposure, cells were treated with Cri/Cur@CD (0.1–5 mg/mL) for 24 h, and viability was reassessed. To evaluate therapeutic synergy, oxidative stress models were treated with equimolar concentrations of HP-β-CD, Cri@CD, Cur@CD, or Cri/Cur@CD for 24 h before viability measurement. For ROS detection, TBHP-treated cells were incubated with DCFH-DA (5 μmol/L, 60 min) and analyzed by fluorescence microscopy, while mitochondrial membrane potential was assessed using JC-1 staining (20 min, 37 °C) with red/green fluorescence ratios quantified. Oxidative stress markers (SOD, CAT, MDA) were measured in cell lysates after TBHP modeling and drug treatment, with protein levels normalized via BCA assay. To assess anti-inflammatory effects, LPS-stimulated PIG1 cells (1 mg/mL, 12 h) were treated for 24 h, and TNF-α, IL-1β, and IL-6 levels in supernatants were analyzed by ELISA. The effect on PIG1 cell proliferation and migration was evaluated by scratch assay. PIG1 cells were seeded in 6-well plates and cultured to confluence. After 4 h treatment with TBHP to induce oxidative injury, a linear scratch was created using a sterile 1000 mL pipette tip. Following PBS washing, cells were incubated with drug for 48 h, and wound closure was imaged at 0, 24 and 48 h.

2.7. Melanin regeneration in vitro

PIG1 cells (1 × 10⁶ cells/well in 6-well plates) were treated with TBHP and various preparations as described previously. Cell suspensions (1 × 10⁶ cells/mL) were lysed through three freeze-thaw cycles (−80 °C/37 °C), followed by centrifugation (3000 rpm, 20 min, 4 °C), with cAMP levels measured in supernatants via ELISA. Melanin production was assessed by lysing cells in 0.1M NaOH (200 μL) after treatment, using dual freeze-thaw cycles and centrifugation, with absorbance measured at 500 nm. Tyrosinase (TYR) activity was determined in 96-well plates by lysing cells in 1 % Triton X-100, incubating with 0.2 % L-Dopa (37 °C, 4 h), and measuring absorbance at 475 nm.

2.8. Antioxidant stress and anti-inflammatory activity In vivo

All animal experiments were conducted according to the institutional guidelines of Shanxi Medical University and were approved by the Institutional Animal Experiment Committee (Approval No. 2022-025, Taiyuan, China), and in accordance with ARRIVE guidelines [44]. In line with previous studies [23,45], only female C57BL/6 mice were enrolled to maintain consistency with the established vitiligo model and reduce animal numbers in accordance with the 3R principle. Female C57BL/6 mice (6 weeks) were acquired from the experimental animal center of Shanxi Medical University. A vitiligo mouse model was established in 6-week-old C57BL/6 mice using 40 % monobenzone cream applied daily for 5 weeks on depilated dorsal skin (2 × 2 cm), with successful depigmentation confirmed by white hair regrowth. The mice were divided into six groups (n = 3): normal control, model control, and four treatment groups (HP-β-CD, Cri@CD, Cur@CD, and Cri/Cur@CD). After modeling, 100 mg of respective hydrogels were applied twice daily for 6 weeks. Melanin recovery was assessed via photoacoustic imaging (21 Hz, 924 nm) before hair regrowth. After sacrifice, skin samples underwent histological analysis including H&E staining, Fontana-Masson staining for melanin, DHE staining for ROS, and IHC for CD8⁺ T cells. Tissue homogenates were analyzed for CAT, SOD, MDA, cAMP, and inflammatory cytokines (TNF-α, IL-1β, IL-6) using commercial kits.

2.9. Biological safety in vivo

Hydrogel irritancy was evaluated by topically administering the Cri/Cur@CD to normal mice for 6 weeks, followed by skin tissue collection for H&E staining and histopathological examination. Systemic biosafety was assessed after treatment completion through whole blood and serum collection for complete blood count and serum biochemistry analyses, and harvesting of major organs for histopathological examination.

2.10. Statistical analysis

Fluorescent and stained images were quantitatively analyzed using ImageJ software. Grayscale channels were calibrated to eliminate background interference. For each region, total color intensity and area were measured, with their ratio used for quantification. Experimental data are presented as mean ± SD (n = 3). Statistical significance was determined by one-way ANOVA and LSD t-test (∗p < 0.05, ∗∗p < 0.01, ∗∗p < 0.001; n.s. = not significant).

3. Results and discussion

3.1. Preparation and Characterization of Cri/Cur@CD

As is shown in Fig. S1, the power of complexes (Cri@CD, Cur@CD, and Cri/Cur@CD) was obtained by freeze-drying. SEM images shown in Fig. 1a–c reveal distinct microstructural characteristics of different materials. HP-β-CD (Fig. 1a) displayed smooth spherical particles, whereas Cri and Cur exhibited irregular block structures (Fig. S2). The physical mixture of Cri, Cur, and HP-β-CD retained their individual features in SEM (Fig. 1b). In contrast, Cri/Cur@CD exhibited homogeneous HP-β-CD-dominated surfaces without drug aggregates (Fig. 1c), confirming effective microscopic integration [46]. The UV–vis spectra of different materials are shown in Fig. 1d. HP-β-CD had no significant absorption peaks in the wavelength range of 200–550 nm. Cri showed significant absorption peaks at 251 nm because of the π-π∗ and n-π∗ electron transitions in the conjugation system of Cri. Cur had the highest absorption peak at 425 nm, which is attributed to the long-conjugated structure of benzene, -C=C- and -C=O in the molecule. The absorption peaks of the inclusion Cri/Cur@CD appeared at 251 nm and 425 nm, which respectively corresponded to the characteristic absorption peaks of Cri and Cur in its components. FTIR spectra were recorded to identify the characteristic functional groups (Fig. 1e). In the infrared spectrum of Cri/Cur@CD, a strong peak near 3400 cm⁻¹ attributed to hydroxyl stretching vibration and peaks near 2920 cm⁻¹ associated with hydroxypropyl and methylene C–H stretching vibration were consistent with HP-β-CD [47]. Additionally, a peak at 2200 cm⁻¹ indicative of the cyano group (C ≡ N) was corresponding with Cri [48], while absorptions in the 1650–1600 cm⁻¹ range assigned to carbonyl (C=O) and conjugated alkene (C=C) moieties were corresponding with Cur [49]. The formation of an inclusion complex through host-guest interaction alters the chemical microenvironment of the hydrogen atoms within the molecules, a change that can be characterized by ¹H NMR spectroscopy. As shown in Fig. 1f–i, characteristic changes were observed in the ¹H NMR spectrum of HP-β-CD after complexation. The hydrogen atoms located on the exterior of the HP-β-CD cavity (signals at δ 1.02, 3.61, and 4.83 ppm) showed no significant chemical shift changes. In contrast, the chemical shifts of the hydrogen atoms pointing toward the interior of the cavity, specifically H-3 and H-5, moved upfield: H-3 shifted from 4.51 ppm to 4.50 ppm (Δδ = −0.01 ppm), and H-5 shifted from 3.37 ppm to 3.35 ppm (Δδ = −0.02 ppm). This upfield shift confirms that the guest molecule has entered the hydrophobic cavity of HP-β-CD, exerting a shielding effect on the interior hydrogen atoms. The ¹H NMR signals of the guest molecules (Cri and Cur) also exhibited significant changes. On one hand, peak broadening occurred for some signals due to the restricted overall molecular motion imposed by the large HP-β-CD molecule. On the other hand, the observed decrease in the chemical shift (upfield shift) of the hydrogen atoms in the guest molecules is likely attributed to the shielding effect provided by HP-β-CD. Collectively, these changes confirm the successful formation of the host-guest inclusion complexes. Furthermore, the phase structures of the materials powers were researched by XRD. As shown in Fig. 1j, HP-β-CD exhibited typical amorphous characteristics, showing a broad diffuse peak at 18.82° (FWHM = 5.42°) with a crystallite size of only 1.49 nm, indicating a disordered molecular arrangement, while the raw drug substances Cri and Cur exhibited high crystallinity. In Cri/Cur@CD, however, the characteristic crystalline peaks of Cri and Cur completely disappeared, leaving only the broad envelope peak of HP-β-CD that further broadened (FWHM = 6.90°, crystallite size = 1.17 nm), suggesting that the drug molecules were successfully embedded into the hydrophobic cavity of cyclodextrin, the long-range ordered structure was destroyed, and an amorphous molecular inclusion complex was formed, which is the reason for overcoming the solubility of poorly soluble drugs. These data strongly confirmed that the inclusion had successfully encapsulated Cri and Cur. The drug loading efficiency (LE) was determined using UV–vis spectroscopy. The inclusion complex, prepared with 100 mg HP-β-CD, 20 mg Cri and 10 mg Cur, exhibited drug loadings of (13.4 ± 0.32) % for Cri and (2.5 ± 0.17) % for Cur. Although Cri and Cur individually demonstrate negligible aqueous solubility, their inclusion complex displayed significantly enhanced solubility, comparable to that of HP-β-CD. Notably, the complex formed a clear solution even at a high concentration as 500 mg/mL (Fig. S3).

Fig. 1.

Characterization of Cri/Cur@CD and Cri/Cur@CD-Gel. (a)–(c) SEM of HP-β-CD, Powder mixture of Cri, Cur, and HP-β-CD, and Cri/Cur@CD. (d) UV–Vis–NIR spectra and (e) FT-IR spectra of HP-β-CD, Cri, Cur, and Cri/Cur@CD. (f)–(i) ¹HNMR spectra of HP-β-CD, Cri, Cur, and Cri/Cur@CD. (j) XRD spectra of HP-β-CD, Cri, Cur, and Cri/Cur@CD. (k)–(n) Free radical scavenging efficiency of DPPH, ABTS⁺•, HO•, and O₂⁻•.

The occurrence of vitiligo is usually induced by factors such as ultraviolet radiation and chemicals to produce ROS, resulting in damage associated molecular patterns, which in turn trigger immune responses [1,50]. Therefore, clearing free radicals is an important aspect of vitiligo treatment. The ability of Cri/Cur@CD to clear various free radicals was examined in vitro. As shown in Fig. 1k–n, the radical scavenging capacities of both free Cur and the inclusion complex Cri/Cur@CD were evaluated against four radical species, including DPPH (Fig. 1k), ABTS⁺• (Fig. 1l), HO• (Fig. 1m), and O₂⁻• (Fig. 1n). Both substrates demonstrated concentration-dependent scavenging activity within the range of 10–100 μmol/L, with scavenging efficiency increasing proportionally with Cur concentration. When concentrations more than 80 μmol/L, the scavenging rates for DPPH.

and ABTS⁺• radicals exceeded 80 % for both compounds. When the concentration reached 100 μmol/L, scavenging efficiencies surpassed 60 % against both HO• and O₂⁻•. No significant differences were observed between the free radical scavenging capacities of the two compounds at equivalent Cur concentrations. These results demonstrate that the conjugated system and enol moiety in Cur are responsible for its potent antioxidant activity. Crucially, while the encapsulation process significantly improves aqueous solubility, it preserves the intrinsic antioxidant properties of Cur, as evidenced by the maintained radical scavenging efficacy of Cri/Cur@CD.

3.2. Preparation and Characterization of Cri/Cur@CD-Gel

In this study, the pH corresponding properties of Carbomer 940 were used to prepare hydrogels (Fig. 2a). The optimal pH for dermal drug delivery was screened based on rheological analysis. An ideal dermal drug delivery matrix requires high structural integrity, characterized by the storage modulus (G′), and elastic-dominated behavior (G’ > G″) to ensure stable residence on the skin. Simultaneously, its pH should be close to the physiological skin range (4.5–6.5) to guarantee biocompatibility. To determine the optimal pH of 1 % Carbomer 940 hydrogel for dermal drug delivery, this study systematically evaluated the rheological properties of the inclusion compound gels under pH conditions ranging from 3 to 8, using rheological measurements (Fig. 2b and c) and tilt test (Fig. 2d). Analysis of the rheological data (Fig. 2b and c) indicates that pH has a decisive influence on gel formation. The sample at pH 3 exhibited a fluid state, while at pH 4, the gel's G′ was extremely low (only 21.8 Pa at 0.1 Hz), indicating severely insufficient structural strength, failing to meet the requirements for drug delivery. When the pH increased to 5–7, the carboxyl groups of carbomer fully dissociated, forming a robust three-dimensional network and demonstrating significant elastic solid behavior (G’ > G″). At 0.1 Hz, the G′ values for pH 5, 6, and 7 reached 308.0, 318.3, and 349.9 Pa, respectively, and condition G’ > G” (tan δ ≪ 1) was satisfied across all tested frequency ranges, indicating elastic-dominated solid-like behavior. This rheological characteristic ensures that the gel maintains its shape during storage and use, forming a stable film layer upon application to the skin that is resistant to flow due to gravity or minor external forces. For the pH 8 sample, due to the pH affecting the color of curcumin, a significant color change occurred and its G′ value at 0.1 Hz began to show a declining trend (279.5 Pa). Considering the risks associated with long-term skin compatibility and irritation, adjusting the formulation pH closer to the skin's microenvironment (pH 4.5–6.5) is a superior choice. In contrast, the pH 6 sample achieves the optimal balance between gel strength (G’ = 318.3 Pa) and biocompatibility. It ensures good skin adhesion while effectively maintaining skin barrier stability and reducing potential irritation to sensitive skin. Furthermore, throughout the frequency sweep range, the G′ and G″ curves of the pH 6 sample exhibited excellent stability, indicating that its network structure can withstand varying degrees of shear stress. Consequently, pH 6.0 was selected as the ideal condition for skin application.

Fig. 2.

Characteristics of hydrogel. (a) Preparation of Cri/Cur@CD-Gel. (b) Rheological modulus and (c) Loss factor of Cri/Cur@CD-Gel at different pH values. (d) Tilt Test of Cri/Cur@CD-Gel at different pH values. (e)–(f) SEM of blank hydrogel and Cri/Cur@CD-Gel. (g) In vitro release of Cri in Cri suspension, Cri/Cur@CD and Cri/Cur@CD-Gel. (h) In vitro release of Cur in Cur suspension, Cri/Cur@CD and Cri/Cur@CD-Gel. (i) Transdermal permeation of Cri in Cri suspension and Cri/Cur@CD. (j) Transdermal permeation of Cur in Cur suspension and Cri/Cur@CD.

Based on these results, four carbomer-940-based hydrogels (HP-β-CD hydrogel, Cri@CD hydrogel, Cur@CD hydrogel, and Cri/Cur@CD hydrogel) were prepared by adjusting pH to 6. SEM images reveal distinct morphological differences between blank hydrogel (Gel) and drug-loaded Cri/Cur@CD-Gel (Fig. 2e and f). The Gel (Fig. 2e) displayed a highly organized fibrous architecture, consisting of densely packed, uniform fibers that formed a matrix with regularly distributed macro-porous structures. In contrast, the Cri/Cur@CD-Gel (Fig. 2f) exhibited a markedly different morphology, featuring thicker, irregular fibers that formed a more loosely organized network with heterogeneous pore distribution of reduced dimensions. These structural modifications provide clear evidence that the incorporation of inclusion complexes induces significant alterations in the supramolecular organization of Carbomer 940-based hydrogels, particularly affecting their fiber morphology and pore architecture.

The in vitro release profiles of Cri and Cur were systematically evaluated, including free drug's suspensions, inclusion complexes (Cri/Cur@CD), and hydrogels (Cri/Cur@CD-Gel) using dialysis membrane methodology (Fig. 2g and h). The Free drug formulations exhibited limited release capacity, demonstrating less than 50 % cumulative release over 24 h, with Cur showing slower release kinetics compared to Cri due to its lower solubility. The inclusion complex formulation Cri/Cur@CD demonstrated markedly enhanced release characteristics, achieving exceed 60 % drug release within the initial 0.5 h and approaching complete release by 1 h. This rapid liberation profile is attributed to the water-soluble complex-mediated diffusion mechanism. In contrast, the hydrogel formulation (Cri/Cur@CD-Gel) demonstrated sustained release behavior, with cumulative release percentages of 30 % at 1 h, exceed 60 % at 6 h, and 80 % at 24 h. This controlled release pattern is governed by the hydrogel's gel-sol phase transition dynamics during the dialysis process. The release rate of inclusion complex hydrogel is between the inclusion complex solution and hydrophobic drug suspension, that means the Cri/Cur@CD-Gel effectively achieves an optimal equilibrium between drug retention and controlled liberation, making it particularly suitable for transdermal delivery applications where sustained therapeutic action is required while avoiding the potential complications associated with initial burst release phenomena.

3.3. Transdermal penetration of Cri/Cur@CD

The transdermal permeation profiles of Cri/Cur@CD were quantitatively assessed using Franz diffusion cells with excised skin (Fig. 2i and j). Comparative analysis revealed that the Cri/Cur@CD solution exhibited significantly enhanced cumulative skin permeation for both Cri (Fig. 2i) and Cur (Fig. 2j) relative to their non-complexed aqueous counterparts throughout the 24 h experimental period. Quantitative evaluation demonstrated that the Cri/Cur@CD solution achieved more than 1 % permeation within the initial 6 h, progressing beyond 2 % by the 24 h endpoint. In contrast, free Cri and Cur solutions showed substantially reduced permeation kinetics, attaining only approximately 1 % cumulative permeation after 24 h. These results confirm that molecular encapsulation via inclusion complexation significantly improves both the aqueous solubility and transdermal diffusion coefficients of Cri and Cur.

This enhanced permeation profile is particularly noteworthy given that both Cri and Cur are poorly water-soluble compounds that exhibit minimal transdermal penetration in their free forms. The cyclodextrin carrier likely improves delivery through multiple combinational mechanisms: (1) increased apparent solubility providing a larger concentration gradient driving force; (2) transient interaction with stratum corneum lipids facilitating paracellular transport; and (3) controlled release of guest molecules at the skin surface, maintaining thermodynamic activity without overwhelming skin barrier properties. The observed permeation enhancement suggests potential for reduced application frequency compared to conventional topical agents, improving patient compliance—a critical consideration for chronic vitiligo management. Due to physical mixing rather than chemical bonding between HP-β-CD and Gel material [51,52], our final inclusion complex-loaded hydrogel formulation retains this superior permeation capability while additionally providing the benefits of controlled release and ease of application.

3.4. Cytotoxicity and cell uptake

Derived from normal human skin melanocytes, PIG1 cells counted as core of damage in vitiligo disease are able to produce melanin in skin [53], while HaCaT cells are the main constituent cells in epidermal tissue accounting for over 90 %, and are target cells for melanosomes secreted by melanocytes [54,55]. Therefore, we examined the biocompatibility of the inclusion complex on these two types of cells using the CCK-8 method. As shown in Fig. 3a and b, Cri/Cur@CD did not affect the viability of and HaCaT and PIG1 cells at concentrations up to 10 mg/mL, while cell viability remained at approximately 70 % at 20 mg/mL, demonstrating the low toxicity of the inclusion complex, which is crucial for safe therapeutics. To further investigate the impact of inclusion compounds on the safety of Cri and Cur, the effects of two free drugs on the viability of two types of cells were measured separately. As shown in Figs. S4–S5, both cells maintained high viability (>90 %) across the tested concentration range (up to 80 μmol/L), demonstrating excellent biocompatibility of Cri and Cri/Cur@CD. As demonstrated in Figs. S6–S7, distinct cytotoxicity was observed between free Cur and Cri/Cur@CD. Free Cur exhibited significant concentration-dependent cytotoxicity, whereas Cri/Cur@CD showed minimal effects on cell viability across the tested concentration range. The observed differential cytotoxicity profiles likely originate from Cur's dual concentration-dependent effects, antioxidant effects at low concentrations versus cellular damage at elevated levels [56]. Notably, the Cri/Cur@CD maintained cell viability >90 % across equivalent Cur concentrations, confirming enhanced safety through inclusion complexation. These findings indicate that cyclodextrin-based complexation significantly enhance the solubility of Cur and mitigate the original cytotoxicity of Cur.

Fig. 3.

Therapeutic effect in vitro. (a) Cytotoxicity of Cri/Cur@CD for HaCaT cells. (b) Cytotoxicity of Cri/Cur@CD for PIG1 cells. (c) FCM investigation of FITC-labeling Cri/Cur@CD uptake by PIG1 cells. (d) CLSM images of FITC-labeling Cri/Cur@CD uptake by PIG1 cells. (e) TBHP-induced oxidative stress model damage in PIG1 cells. (f) Protective effect of different concentrations of Cri/Cur@CD on cellular oxidative stress model. (g) Protective effect of various formulations on cellular oxidative stress model. Effect of various formulations on intracellular (h) cAMP levels, (i) melanogenesis, and (k) TYR activity in cellular oxidative stress model. (j) Cell migration analysis. Mean ± SD, n = 3, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001, n.s. = not significant.

The cell uptake capacity of FITC-labeled Cri/Cur@CD in PIG1 cells was quantitatively evaluated using flow cytometry analysis (Fig. 3c) and temporal fluorescence intensity profiles (Fig. S8). Untreated PIG1 cells exhibited minimal autofluorescence with intensity values distributed below 10² (< mean fluorescence intensity (MFI) = 55.4 ± 4.69). Following 0.5 h incubation with FITC-Cri/Cur@CD, fluorescence intensity dramatically right-shifted to 10³-10⁴ range (MFI = 7197.3 ± 133), confirming efficient cell uptake of the FITC-labeling Cri/Cur@CD. This high fluorescence level persisted after 4 h incubation (MFI = 8083 ± 95), indicating that the effective taken up of inclusion complex can be expected to alleviate oxidative stress and inflammatory reactions within PIG1 cells.

Confocal laser scanning microscopy (CLSM) analysis (Fig. 3d) revealed distinct spatial distribution patterns of the FITC-labeled Cri/Cur@CD, with DAPI-stained nuclei (blue) surrounded by concentrated FITC signals (green) forming spindle-shaped perinuclear regions. Time-course analysis in CLSM shows progressive fluorescence signal intensification correlating with FCM data. Spatial mapping reveals dual localization pattern, exhibiting both cytoplasmic dispersion and nuclear accumulation. This unique biodistribution pattern facilitates prolonged high-concentration maintenance within cellular compartments, potentially enhancing therapeutic efficacy.

3.5. Protective effect in vitro

As the affected cells of vitiligo, melanocyte death caused by oxidative stress is the direct cause of skin lesions[5]. An oxidative stress model was established in PIG1 cells using tert-butyl hydroperoxide (TBHP), with cell viability quantified by CCK-8 assay. As shown in Fig. 3e, PIG1 cells exposed to TBHP (0.1–0.8 mmol/L) for 4 h exhibited a concentration-dependent decrease in viability. At 0.1 mmol/L TBHP, cell viability remained high (92.4 ± 2.9 %), indicative of minimal oxidative damage. However, increasing TBHP concentrations resulted in progressive cytotoxicity, with viability declining to 57.2 ± 4.1 % at 0.4 mmol/L (moderate damage) and further decreasing to 15.6 ± 2.0 % at 0.8 mmol/L (severe damage). This dose-response relationship effectively recapitulates the oxidative stress-mediated melanocyte injury observed in vitiligo pathogenesis. Based on these findings, a TBHP concentration of 0.4 mmol/L (4 h exposure) was selected for subsequent experiments, as it induces significant yet non-lethal cellular damage, providing an optimal balance between model validity and experimental feasibility.

Following establishment of the oxidative stress model in PIG1 cells via 0.4 mmol/L TBHP incubation for 4 h, therapeutic interventions were performed using Cri/Cur@CD at concentrations ranging from 0.1 to 8 mg/mL for 24 h. As illustrated in Fig. 3f, Cri/Cur@CD exhibited significant concentration-dependent therapeutic effect on the oxidative stress model. The model group showed 63.0 ± 3.2 % viability after 24 h recovery, and at 0.1 mg/mL Cri/Cur@CD, cell viability reaches 68.4 ± 2.5 %, showing no significant difference from the model group. Therapeutic efficacy progressively increased with increase of Cri/Cur@CD concentration, achieving 82.5 ± 1.5 % viability at 1 mg/mL (p < 0.001 vs. model) and 95.1 ± 1.1 % at 8 mg/mL (comparable to normal controls). These results confirm Cri/Cur@CD's concentration-dependent therapeutic capacity against TBHP-induced oxidative damage. Based on these findings, we selected 1 mg/mL as the treatment concentration for subsequent experiments. To assess combinational effects of Cri and Cur, the antioxidant stress activities of Cri/Cur@CD, Cri@CD, Cur@CD, and HP-β-CD were compared (Fig. 3g). The model group maintained 63.1 ± 1.7 % cell viability, while HP-β-CD alone showed no therapeutic effect (66.1 ± 3.4 %, p > 0.05). Cri@CD and Cur@CD monotherapies significantly improved viability versus model (p < 0.001), and their effects were statistically equivalent. Remarkably, Cri/Cur@CD achieved superior efficacy (95.9 ± 2.2 % viability, p < 0.001 vs. Cri@CD, p < 0.01 vs. Cur@CD), demonstrating enhanced synergistic protection through inclusion complex-mediated co-delivery system.

In the scratch wound healing assay shown in (Fig. 3j), normal cells reduced the wound area by half within 48 h through proliferation and migration, whereas oxidative stress injury in the model group prevented wound closure, as indicated by decreased cell density along the wound edge. The drug-treated groups exhibited gradual wound area reduction, with the Cri/Cur@CD group showing the most prominent effect—reducing the wound area to nearly one-third at 48 h—thereby demonstrating the positive role of the inclusion complex in promoting PIG1 cell migration and proliferation.

3.6. Enhancement of melanin synthesis in vitro

As a key intracellular second messenger, cAMP regulates oxidative stress responses pathways, inflammatory signaling cascades, and melanogenesis [35]. As quantified by ELISA in Fig. 3h, TBHP-induced oxidative stress significantly reduced intracellular cAMP levels to 0.68-fold of normal controls (p < 0.001). HP-β-CD treatment showed no cAMP modulation (0.69-fold, p > 0.05). Cri@CD and Cur@CD monotherapies elevated cAMP to 0.91-fold (p < 0.001) and 0.79-fold (p < 0.001), respectively. Notably, Cri/Cur@CD achieved maximal cAMP restoration (0.92-fold, p < 0.001 vs. model, p < 0.001 vs. Cur@CD). This synergistic effect attributed to Cri-mediated PDE-4 inhibition, which effectively prevents cAMP degradation. As a key intracellular second messenger, cAMP regulates oxidative stress responses pathways, inflammatory signaling cascades, and melanogenesis. The elevation of intercellular cAMP levels activates protein kinase A (PKA), enhancing superoxide dismutase (SOD) and catalase (CAT) activities to mitigate oxidative damage while suppressing pro-inflammatory cytokine production [23]. Furthermore, elevated cAMP levels stimulate melanin synthesis through PKA/MITF and MAPK/MITF signaling pathways in PIG1 cells [57].

Under normal physiological conditions, the melanin was synthesized in PIG1 cells and transferred to neighboring keratinocytes via melanosome delivery. However, under vitiligo conditions, oxidative stress disrupts melanocyte metabolic function, leading to significant melanogenesis impairment [58]. As is shown in Fig. 3i, the melanin production was quantified by measuring NaOH-solubilized cell contents at 500 nm. TBHP-induced oxidative stress reduces melanin content to 0.64-fold of normal controls (p < 0.001). HP-β-CD treatment shows no therapeutic effect (0.65-fold, p > 0.05 vs model). Cri@CD treatment elevates melanin content to 0.87-fold (p < 0.001 vs model), while Cur@CD achieves 0.85-fold (p < 0.001). Strikingly, Cri/Cur@CD surpasses normal baseline levels (1.14-fold, p < 0.001 vs all groups). These findings demonstrate that the superior efficacy of Cri/Cur@CD in enhancing melanin production.

Tyrosinase (TYR), the rate-limiting enzyme in melanogenesis, catalyzes the hydroxylation of tyrosine to 3,4-dihydroxyphenylalanine (DOPA) and subsequent oxidation to melanin precursors [59]. Thus, Impairment of TYR activity inhibits melanin biosynthesis. The TYR activity was quantified in Fig. 3k via DOPA oxidation assay. TBHP-induced oxidative stress in PIG1 cells reduces TYR activity to 0.68-fold of normal controls (p < 0.001), and HP-β-CD treatment shows no therapeutic effect (0.67-fold, p > 0.05 vs model). Cri@CD and Cur@CD restored TYR activity to 0.87-fold (p < 0.001) and 0.85-fold (p < 0.001), respectively. Remarkably, Cri/Cur@CD demonstrated exceptional restorative capacity, elevating tyrosinase (TYR) activity to levels surpassing physiological baseline measurements (1.14-fold, p < 0.001 vs all groups). Oxidative stress can disrupt melanogenesis through direct inactivation of TYR via ROS-mediated enzyme oxidation and impairment of melanocyte viability [5,50,60]. The Cri/Cur@CD addresses both pathological components through reactivating melanogenic transcription machinery (cAMP-dependent upregulation of melanogenic signaling pathways) and mitigation oxidative stress to preserve cell functionality.

3.7. Antioxidant stress and anti-inflammatory effects in vitro

Excessive ROS and inflammatory response in the affected area of vitiligo further cause damage to melanocytes, therefore clearing ROS and inhibiting inflammatory response are key to treating the disease [4,5,20]. As is shown in Fig. 4a, ROS levels in differentially treated PIG1 cells were detected by DCFH-DA probe, which diffuses across cell membranes and then undergoes esterase-mediated hydrolysis and reacts with ROS to generate fluorescent DCF. Only weak fluorescence signals were observed in the normal group cells, while the model group showed strong fluorescence signals. The fluorescence intensities of the treatment groups were between the normal and model group, indicating differences in intracellular ROS levels among different groups.

Fig. 4.

Anti-oxidant stress and anti-inflammatory in vitro. (a) Fluorescence images of intracellular ROS levels under different treatment and fluorescence semi quantitative analysis (c). (b) Fluorescence microscopy of mitochondrial membrane potential under different treatment and fluorescence semi quantitative analysis (d). Effects of different treatment on (e) SOD activity, (f) CAT activity and (g) MDA content in cellular oxidative stress model. Effects of different treatment on (h) TNF-α, (i) IL-1β, and (j) IL-6 in cellular inflammatory model. Mean ± SD, n = 3, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001, n.s. = not significant.

The green fluorescence intensity quantitatively reflects ROS levels as analyzed in Fig. 4c. Normal control cells exhibited minimal fluorescence (mean intensity 1.43 ± 0.38), while TBHP-induced oxidative stress significantly elevated ROS levels (8.58 ± 0.43, p < 0.001), and HP-β-CD treatment showed no ROS modulation (8.49 ± 0.41, p > 0.05 vs model). Both Cri@CD (4.12 ± 0.32, p < 0.001) and Cur@CD (4.85 ± 0.35, p < 0.001) monotherapies significantly reduces ROS levels compared to model, and their effects were statistically equivalent (p > 0.05). Remarkably, Cri/Cur@CD combination therapy demonstrates superior ROS suppression (2.57 ± 0.26, p < 0.001 vs all groups), achieving near-normalization of oxidative stress.

The mitochondrial membrane potential (ΔΨm) alterations in differentially treated PIG1 cells were quantitatively assessed using the JC-1 fluorescent probe (Fig. 4b) In this assay, JC-1 forms red-fluorescent J-aggregates (JC-1 A) in mitochondria with intact membrane polarization, whereas mitochondrial depolarization induces its conversion to green-fluorescent monomers (JC-1 M). As exhibited in Fig. 4d, the red/green fluorescence intensity ratio quantitatively serves as a quantitative indicator of mitochondrial depolarization. Normal cells exhibited predominant red fluorescence, indicative of a high ΔΨm, whereas TBHP-treated model cells showed a marked shift to green fluorescence, confirming significant mitochondrial depolarization. Treatment with HP-β-CD did not attenuate this depolarization, as evidenced by a comparable red/green ratio (1.98 ± 0.07 vs model group's 1.96 ± 0.069, p > 0.05). In contrast, Cri@CD and Cur@CD monotherapies partially restored mitochondrial polarization, yielding significantly reduced ratios (0.89 ± 0.11 and 0.92 ± 0.13, respectively, p < 0.001 vs model). Strikingly, Cri/Cur@CD combination therapy achieved the most pronounced protective effect, nearly normalizing ΔΨm (0.23 ± 0.06, p < 0.001 vs monotherapies). These results demonstrate that TBHP triggers substantial ROS generation and subsequent of mitochondrial depolarization in PIG1 cells. While HP-β-CD alone was ineffective, Cri@CD and Cur@CD monotherapies provided partial protection, and the Cri/Cur@CD combination therapy exerted a synergistic effect, effectively mitigating ROS accumulation and restoring mitochondrial function. The significantly lower ROS levels and depolarization ratio in Cri/Cur@CD-treated PIG1 cells, compared to monotherapy groups, further corroborate the synergistic mitochondrial protective action of Cri and Cur.

Superoxide dismutase (SOD) and catalase (CAT) are critical enzymatic antioxidants that cooperatively neutralize ROS, in which SOD catalyzes the dismutation of superoxide radicals into hydrogen peroxide (H₂O₂), while CAT decomposes H₂O₂ into water and oxygen. Malondialdehyde (MDA), a terminal product of lipid peroxidation, serves as a key biomarker of oxidative damage. As shown in Fig. 4e, TBHP-induced oxidative stress severely impaired antioxidant defenses in PIG1 cells, reducing SOD activity from 1218.3 ± 83.3 U/mgprot (untreated) to 366.6 ± 47.08 U/mgprot (70 % decrease, p < 0.001). HP-β-CD had no restorative effect (372.4 ± 41.2 U/mgprot, p > 0.05), whereas Cri@CD (798.5 ± 62.1 U/mgprot) and Cur@CD (723.8 ± 58.7 U/mgprot) monotherapies significantly recovered SOD activity (p < 0.001 vs. model). The Cri/Cur@CD combination therapy nearly normalized SOD levels (1156.4 ± 89.3 U/mgprot, p < 0.001 vs. all treatments). Similarly, CAT activity dropped from 561.6 ± 23.4 U/mgprot (control) to 129.9 ± 24.5 U/mgprot (77 % reduction, p < 0.001) under oxidative stress (Fig. 4f). Neither HP-β-CD (133.1 ± 22.7 U/mgprot) nor control group mitigated this decline (p > 0.05). Monotherapies partially restored CAT activity (Cri@CD: 389.5 ± 31.2 U/mgprot; Cur@CD: 357.8 ± 29.8 U/mgprot, p < 0.001), but Cri/Cur@CD was significantly more effective (538.2 ± 35.6 U/mgprot, p < 0.01 vs. monotherapy). Contrastingly, MDA levels (Fig. 4g) surged from 15.5 ± 1.2 nmol/mgprot (control) to 58.9 ± 1.07 nmol/mgprot (280 % increase, p < 0.001) in the TBHP model. While HP-β-CD (57.3 ± 1.15 nmol/mgprot) was ineffective, Cri@CD (32.1 ± 2.1 nmol/mgprot, 45 % decrease) and Cur@CD (36.7 ± 1.9 nmol/mgprot, 38 % decrease) significantly reduced lipid peroxidation. The Cri/Cur@CD combination therapy exhibited the greatest reduction (18.3 ± 1.5 nmol/mgprot, 69 % below model levels, p < 0.001 vs. all groups). These results demonstrate that TBHP induces severe oxidative imbalance, while HP-β-CD alone lacks therapeutic efficacy. In contrast, Cri/Cur@CD co-encapsulation synergistically restores antioxidant defenses, achieving 95 % of normal SOD activity, 96 % of baseline CAT levels, and near-physiological MDA levels (118 % of control). The superior efficacy over monotherapies (p < 0.01) confirms that HP-β-CD-mediated co-delivery enhances oxidative stress mitigation through complementary pharmacological mechanisms.

Immune dysregulation plays a central role in vitiligo pathogenesis, with TNF-α, IL-1β, and IL-6 serving as key pro-inflammatory mediators [1,61]. To evaluate drug anti-inflammatory efficacy, an LPS-induced PIG1 cell inflammation model was established, with cytokine levels quantified via ELISA (Fig. 4h–j). LPS stimulation significantly increased cytokine levels from baseline (TNF-α 20.9 ± 0.67 pg/mL, IL-1β 12.1 ± 0.58 pg/mL, and IL-6 14.9 ± 0.67 pg/mL) to elevated levels (TNF-α 27.3 ± 0.82 pg/mL, IL-1β 10.3 ± 2.30 pg/mL, and IL-6 29.3 ± 0.73 pg/mL), confirming successful inflammation induction. HP-β-CD showed no anti-inflammatory effect (TNF-α 26.8 ± 0.79 pg/mL, IL-1β 9.8 ± 2.1 pg/mL, IL-6, 28.9 ± 0.69 pg/mL, p > 0.05 vs model). In contrast, all three treatment groups (Cri@CD, Cur@CD, and Cri/Cur@CD) significantly reduced pro-inflammatory cytokines compared to the model group (p < 0.001). Cri@CD monotherapy demonstrated stronger suppression than Cur@CD for all cytokines (TNF-α 21.5 ± 0.90 vs 24.1 ± 0.85 pg/mL, IL-1β 32.2 ± 2.43 vs 35.7 ± 2.1 pg/mL, IL-6 19.9 ± 0.57 vs 22.3 ± 0.63 pg/mL, p < 0.05). Notably, Cri/Cur@CD achieved maximal inhibition (TNF-α 18.2 ± 0.78 pg/mL, IL-1β 8.9 ± 0.51 pg/mL, IL-6 16.4 ± 0.49 pg/mL, p < 0.001 vs monotherapies). These results confirm that HP-β-CD-mediated co-delivery of Cri and Cur synergistically suppresses pro-inflammatory cytokine overexpression, highlighting its potential in mitigating vitiligo-associated inflammation.

3.8. Efficacy evaluation in vivo

In this study, a vitiligo model was established by topical application of 40 % Monobenzone cream on depilated C57BL/6 mice (Fig. 5e). Prior to modeling, these mice exhibited follicular melanocyte activity, with depilated skin appearing pink but occasionally retaining residual black patches due to melanocyte distribution heterogeneity. Monobenzone treatment induced progressive depigmentation of these patches and hair whitening, confirming successful model induction (Fig. 5a). Following model establishment, mice received twice-daily treatments for 6 weeks, followed by a 2-week hair regrowth period. The model group and HP-β-CD hydrogel-treated mice showed minimal repigment, with only sporadic black hair regrowth at lesion margins and persistent central depigmentation. In contrast, Cri@CD and Cur@CD hydrogels significantly enhanced hair repigment and reduced depigmented areas, and the Tacrolimus group showed hair color recovery in over half of the area. Notably, Cri/Cur@CD hydrogel achieved more remarkable hair repigment accompanied by perifollicular skin darkening, demonstrating superior therapeutic efficacy over monotherapies.

Fig. 5.

Therapeutic effect in vivo. (a) Comparison of hair color before and after treatment in vitiligo mice of different treatment groups. (b) Photoacoustic imaging of exposed skin in mice from different treatment groups and semi quantitative analysis (f). (c) H&E staining and (d) Melanin staining of skin from different treatment groups. (e) Animal modeling and treatment schematic diagram. (g) Effects on cAMP content in skin from different treatment groups. Mean ± SD, n = 3, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001, n.s. = not significant. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

To objectively assess melanin recovery during the 6-week treatment period while minimizing confounding effects from repeated depilation, photoacoustic imaging (PA) was employed at week 6 to quantify cutaneous melanin content. As shown in Fig. 5b, ultrasonography (US) confirmed intact skin architecture across all groups, while PA imaging at 924 nm specifically detected melanin deposition, with semi quantitative signal analysis in Fig. 5f. Normal control mice exhibited strong melanin-associated PA signals (230.3 ± 26.59 a.u.), whereas model animals demonstrated marked signal attenuation (14.1 ± 2.01 a.u.), reflecting melanocyte dysfunction with potential minor contributions from residual hemodynamic activity. HP-β-CD hydrogel treatment showed no significant melanin recovery (15.9 ± 2.34 a.u., p > 0.05 vs model), whereas both Cri@CD (68.4 ± 12.7 a.u., p < 0.01) and Cur@CD (54.3 ± 11.9 a.u., p < 0.01) monotherapies induced partial signal restoration. Strikingly, Cri/Cur@CD combination therapy (104.1 ± 19.1 a.u., p < 0.001 vs monotherapies) scored close to the Tacrolimus group (78.5 ± 9.4 a.u.), consistent with observed cutaneous repigment patterns. These quantitative imaging findings confirm that HP-β-CD-mediated dual drug encapsulation synergistically enhances melanocyte functional recovery in this vitiligo model, with therapeutic efficacy exceeding that of individual component therapies.

Histological analysis using H&E staining (Fig. 5c) revealed significant treatment-dependent morphological improvements. Normal skin exhibited well-organized epidermal stratification with uniform melanocyte distribution and preserved follicular architecture. In contrast, vitiligo model skin displayed characteristic pathological features, including– marked melanocyte depletion, loss of epidermal disorganization, and follicular structural abnormalities. HP-β-CD treatment had no significant improvement in melanocyte density, while Cri@CD, Cur@CD and Cri/Cur@CD exhibited partial melanocyte repopulations and moderate restoration of epidermal organization, no significant difference with Tacrolimus group. Masson-Fontana staining was employed to evaluate melanin distribution in mouse skin sections (Fig. 5d). Normal skin exhibited physiological melanin deposition with high density and uniformly distributed melanin particles. The vitiligo model skin showed marked melanin depletion. HP-β-CD treatment group demonstrated no significant melanin restoration, while Cri@CD and Cur@CD monotherapies showed increased melanin particles with improved distribution. Cri/Cur@CD combination therapy achieved more apparent melanin density and physiological follicular patterns exceeding Tacrolimus, demonstrating its superior therapeutic efficacy for melanocyte functional restoration.

cAMP levels in mouse skin tissue were quantitatively analyzed via ELISA (Fig. 5g). Normal skin exhibited physiological cAMP concentrations (19.3 ± 0.53 pmol/g tissue), whereas vitiligo model group displayed a significant 42.5 % reduction (11.1 ± 0.74 pmol/g tissue, p < 0.001), indicative of melanocyte dysfunction. Cri@CD and Cur@CD monotherapies significantly restored cAMP levels to 17.7 ± 0.38 pmol/g tissue (p < 0.001 vs. model) and 17.3 ± 0.55 pmol/g tissue (p < 0.001 vs. model), respectively, and 17.5 ± 0.61 pmol/g in Tacrolimus group. Notably, Cri/Cur@CD combination therapy restored cAMP levels (18.3 ± 0.81 pmol/g tissue) statistically indistinguishable from normal group (p > 0.05). These results indicate that vitiligo-associated melanocyte dysfunction correlates with diminished cAMP signaling, which is effectively reversed by Cri and Cur treatment. The Cri/Cur@CD combination therapy demonstrated superior efficacy via synergistic mechanisms.

3.9. Anti-oxidant stress and anti-inflammatory activity In vivo

ROS levels were evaluated in C57BL/6 mouse skin tissue using fluorescent staining (Fig. 6a and S9), with nuclear counterstaining (DAPI, blue) and ROS-specific fluorescence (red). Epidermal ROS quantification (Fig. 6b) was performed by calculating the red/blue fluorescence intensity ratio. Vitiligo model groups exhibited significantly elevated ROS levels in epidermis (red/green ratio 1.39 ± 0.21 in models vs 2.54 ± 0.23 in normal controls), confirming pathological oxidative stress. The descending order of ROS levels was HP-β-CD treatment group, Cri@CD monotherapy group, Tacrolimus group, Cur@CD monotherapy group, and Cri/Cur@CD treatment group (1.38 ± 0.15, p < 0.001 vs monotherapies), demonstrating that the combination therapy of Cri and Cur in Cri/Cur@CD potentiates antioxidant effects, with a slightly superior efficacy relative to the Tacrolimus group.

Fig. 6.

Anti-oxidant Stress and anti-inflammatory in vivo. (a) ROS fluorescence staining of skin from different treatment groups and fluorescence semi quantitative analysis (b). Effects on (c) SOD activity, (d) CAT activity, and (e) MDA content in skin from different treatment groups. (f) CD8 IHC staining of skin from different treatment groups, and fluorescence semi quantitative analysis (g). Effects on (h) TNF-α, (i) IL-1β, (j) IL-6 content in skin from different treatment groups. Mean ± SD, n = 3, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001, n.s. = not significant.

Moreover, the oxidative stress profile was systematically evaluated through the analyses of SOD activity, CAT activity, and MDA levels in mouse skin tissue (Fig. 6c–e). Vitiligo modeling induced oxidative dysregulation, reducing SOD (from 173.8 ± 12.52 to 372.9 ± 14.38 U/mgprot) and CAT (from 5.7 ± 1.19 to 38.7 ± 1.25 U/mgprot), while elevating MDA (20.9 ± 1.23 vs. 3.91 ± 0.59 nmol/mgprot) (all p < 0.001). HP-β-CD treatment showed no effect (p > 0.05). Cri@CD and Cur@CD monotherapies comparably restored SOD and CAT, and reduced MDA (p < 0.01 vs. model). Cri/Cur@CD combination therapy achieved superior restoration (SOD 290.3 ± 16.44 U/mgprot, CAT 29.2 ± 2.14 U/mgprot, and MDA 14.9 ± 0.83 nmol/mgprot), significantly outperforming monotherapies and tacrolimus (SOD 239.2 ± 9.0 U/mg protein, p < 0.05). These findings demonstrate that Cri/Cur@CD provides comprehensive oxidative stress mitigation beyond individual treatments.

Immunohistochemical analysis revealed significant alterations in CD8⁺ T cell distribution patterns across experimental groups (Fig. 6f), with quantitative assessment of epidermal infiltration intensity demonstrating treatment-dependent immunomodulation (Fig. 6g). Normal skin exhibited physiological CD8⁺ T cell distribution (144.15 ± 3.13 a.u.), characterized by sparse, uniform epidermal localization. In contrast, vitiligo model mice showed marked CD8⁺ T cell accumulation (172.79 ± 2.36 a.u., p < 0.001 versus control), exhibiting pathological aggregation in both epidermal and follicular compartments, consistent with established vitiligo immunopathology. Therapeutic intervention produced distinct immunomodulatory outcomes. Cri@CD monotherapy significantly reduced CD8⁺ T cell density to 150.19 ± 3.23 a.u. (p < 0.01 versus model), while Cur@CD monotherapy showed more moderate effects (155.84 ± 4.98 a.u., p < 0.05 versus model), and Tacrolimus therapy achieved a close score (155.8 ± 3.4 a.u.). Notably, the Cri/Cur@CD combination therapy achieved near-complete normalization of CD8⁺ T cell distribution (143.82 ± 1.74 a.u.), demonstrating superior efficacy compared to monotherapies (p < 0.001).

Furthermore, the inflammatory cytokine levels in mouse skin tissue were systemic evaluated through ELISA quantification (Fig. 6h–j). Compared to normal skin tissue, model group upregulated TNF-α (from 5.5 ± 0.57 to 12.0 ± 0.48 pg/mgprot), IL-1β (from 23.3 ± 3.02 to 113.9 ± 3.53 pg/mgprot), and IL-6 (from 2.6 ± 0.31 to 11.6 ± 0.70 pg/mgprot), all p < 0.001. HP-β-CD showed no significant anti-inflammatory activity (p > 0.05). Both monotherapies demonstrated substantial efficacy, and the Cri/Cur@CD combination therapy (TNF-α 7.8 ± 0.50 pg/mgprot, IL-1β 52.7 ± 1.26 pg/mg protein, IL-6 4.9 ± 0.36 pg/mgprot) exhibited superior anti-inflammatory effects p < 0.001 vs. monotherapies), achieving cytokine levels slightly lower than those in the Tacrolimus group (TNF-α 8.0 ± 0.41 pg/mgprot, IL-1β 58.7 ± 2.85 pg/mg protein, IL-6 5.4 ± 0.13 pg/mgprot). These results demonstrate the enhanced, and comparably effective, anti-inflammatory profile of Cri/Cur@CD.

These results demonstrate that combined intervention with Cri and Cur significantly improves hair repigmentation, alleviates oxidative stress, and suppresses inflammatory responses in a mouse model of vitiligo. This strategy achieves comprehensive treatment by targeting the pathogenic mechanisms of vitiligo: Cur's potent antioxidant properties may first protect hair follicle melanocyte stem cells from oxidative damage, while Cri, by inhibiting PDE4 and elevating cAMP, blocks Th1/Th17-mediated autoimmune attack while simultaneously activating melanin biosynthetic pathways including cAMP/PKA/CREB/MITF, MAPK/ERK, and Wnt/β-catenin to accelerate functional recovery. The above links provide mutually promoting function realization: activation of melanogenic signaling activates melanogenesis and melanocyte proliferation, attenuation of oxidative stress is critical for melanocyte survival and function, and suppression of inflammatory cascade creates an immune-tolerant niche for melanocytes conducive to repigmentation. The HP-β-CD hydrogel ensures the sustained and co-localized delivery of these signals to the lesion tissue, thereby enabling their combinational action. Compared with traditional glucocorticoids, this regimen avoids the side effect of skin atrophy, and relative to single-agent JAK inhibitors or Calcineurin inhibitors [62,63], combination therapy may reduce relapse risk through modulation of oxidative stress.

3.10. Skin irritancy and biological safety in vivo

The biosafety assessment of Cri/Cur@CD-Gel, including local skin irritation, systemic toxicity, hematological and biochemical markers were conducted during the treatment period. As shown in Fig. 7a, after continuous application of the hydrogel on normal C57BL/6 mice for 3 and 6 weeks, hair follicles exhibited physiological cycle changes. Compared with control skin, histology of the treated skin showed no signs of irritation, such as erythema, edema, or significant inflammatory cell infiltration, indicating good local tolerance of our inclusion complex hydrogel. At the end of the treatment period, routine blood tests were performed on mice from each group.

Fig. 7.

Biosafety assessment of Cri/Cur@CD-Gel. (a) H&E staining of skin from normal mice treated with the hydrogel for 3 and 6 weeks. (b) WBC, (c) Lymphocyte, (d) RBC, and (e) PLT counts in peripheral blood. (f) Histopathological examination of major organs (H&E staining). (f) ALT, (g) AST, (h) CREA, and (i) UREA levels. Mean ± SD, n = 3.

As shown in Fig. 7b–d, parameters including white blood cell (WBC) count, lymphocyte count, red blood cell (RBC) count, and platelet count in peripheral blood all fell within normal ranges across all groups, including the model group, suggesting that our induction model may not affect peripheral immune status. As shown in Fig. 7f–H&E staining of major organs (heart, liver, spleen, lungs, and kidneys) revealed no obvious pathological damage or abnormalities in any group, indicating the absence of severe systemic toxicity. Furthermore, as shown in Fig. 7g–j, blood biochemical markers such as alanine aminotransferase (ALT) and aspartate aminotransferase (AST) (liver function indicators) as well as creatinine (CREA) and UREA (kidney function indicators) were all within normal limits, demonstrating that the treatment did not impair liver or kidney function. In summary, our inclusion complex hydrogel exhibits favorable biosafety during the treatment period.

4. Conclusion

This study successfully engineered a novel composite inclusion complex (Cri/Cur@CD) through molecular encapsulation of both Cri and Cur within HP-β-CD, subsequently formulated into a carbomer-based hydrogel delivery (Cri/Cur@CD-Gel) system for vitiligo therapy. Cri/Cur@CD demonstrated high drug loading efficiency (13.4 ± 0.32 % for Cri and 2.5 ± 0.17 % for Cur), enhanced aqueous solubility, improved transdermal permeability, and favorable biosafety, while exhibiting potent free radicals scavenging capacity. The optimized hydrogel formulation displays excellent transdermal drug delivery properties and sustained release characteristics (80 % release over 24 h). Comprehensive evaluation in TBHP-induced oxidative stress models using PIG1 cells demonstrated the complex's ability to significantly improve cellular viability (from 63.0 % to 95.9 %), reduce intracellular ROS accumulation (70.0 % decrease), restore mitochondrial membrane potential (ratio of mitochondrial membrane potential reversal 1.96 to 0.38), and enhance antioxidant defenses (SOD activity increased from 366.6 % to 904.8 U/mg protein). The complex also effectively upregulated cAMP signaling (from 0.68-fold to 0.92-fold) and tyrosinase activity (167.6 % increase), promoting significant melanogenesis (178.1 % increase in melanin content). In monobenzone-induced vitiligo models using C57BL/6 mice, the Cri/Cur@CD hydrogel demonstrated superior therapeutic efficacy, promoting cutaneous melanin regeneration and hair repigment, while attenuating oxidative damage through ROS reduction (45.1 % decrease) and modulating immune responses (CD8⁺ T cell infiltration reduced by 16.8 %). The therapeutic outcomes stem from combinational mechanisms combining potent anti-inflammatory action, robust antioxidant capacity, and cAMP pathway activation. Compared to the clinical drug tacrolimus, Cri/Cur@CD-Gel has better therapeutic indicators, and has the potential to avoid the widespread side effects of a single immunosuppressant. These findings demonstrate that the Cri/Cur@CD hydrogel represents a promising multifunctional therapeutic strategy for vitiligo, simultaneously addressing oxidative stress, inflammation, and melanocyte dysfunction through a comprehensive pharmacological approach, with significantly enhanced efficacy compared to monotherapies (p < 0.001). Although the Cri/Cur@CD-Gel system shows promising efficacy in preclinical studies, there remains room for further enhancement of its clinical translation potential. For example, drug loading efficiency could be improved through various strategies; host–guest interactions could be further investigated to determine the optimal drug ratio and loading concentration; and appropriate excipients could be introduced to enhance the flexibility, comfort, and convenience of the hydrogel. Systematically implementing these optimization strategies is expected to significantly advance the translational prospects of this delivery platform toward clinical application. This composite inclusion complex system provides a new idea for transdermal drug delivery, and provides the possibility of safe and efficient treatment of vitiligo.

CRediT authorship contribution statement

Gang Li: Writing – original draft, Visualization, Methodology, Investigation, Conceptualization. Pengqi Zhu: Writing – original draft, Methodology, Investigation, Conceptualization. Caifang Gao: Investigation. Ye Cai: Investigation. Ru Feng: Writing – review & editing. Jinghua Sun: Writing – review & editing. Ruiping Zhang: Writing – review & editing, Supervision.

Declaration of competing interest

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

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.