PEPS was synthesized with few amendment according to a previously published literature (33, 34). ¹H nuclear magnetic resonance (1H NMR) of PEPS were recorded by spectroscopy with CDCl3 (Agilent, USA). Deucravacitinib was purchased from Fu Ruidi Biotech Co., Ltd (Wuhan, China). 2 mg of PEPS and 20 mg of deucravacitinib (dissolved in DMSO) was slowly added to double distilled water. Next, the solvent was removed by rotary evaporation, washing and dialysis (MWCO = 3 kDa) every 6–8 h for two days. Rhodamine and PEG-b-PPS were dissolved in 10 ml dichloromethane as the oil phase and 20 ml of sodium cholate as the water phase. The oil phase was dropped into the water phase followed by 1min ultrasound, and the dichloromethane was completely removed using rotary evaporation. Then 10 ml sodium cholate was added and diluted, centrifuged at 12000 rpm for 30 min, and finally resuspended in deionized water. Ultraviolet and fluorescence spectroscopy showed effectiveness of Rhodamine markers. Deu@PEPS-Rhb was prepared similarly to Deu@PEPS except for performing in the darkness environment.

The morphology of the SPMs were observed by transmission electron microscopy (TEM), performed on a Talos F200X. Zeta potential and size distribution were detected by dynamic light scattering (DLS, Malvern Zetasizer system). Entrapment efficiency (EE%) and Drug Loading capacity% (DL%) were also examined by highperformance liquid chromatography (HPLC). EE% = Weight of deucravacitinib in polymer micelles/Weight of the feeding × 100%; DL% = Weight of deucravacitinib in polymer micelles/Weight of the feeding deucravacitinib plus polymer × 100%. All measurements were in triplicate. Nicole 50 Fourier transform infrared spectroscopy (FTIR) recorded Deu@PEPS and oxidative product of PEPS from 3500 to 500 cm⁻¹.

Different concentrations of Deu@PEPS containing H₂O₂ (5 mM or10 mM) or PBS were utilized to assess the ROS-responsive capacity by HPLC. For oxidation sensitive drug release, deucravacitinib release profiles were conducted in 20 ml PBS. Dialysis tube was used to collect free Deucravacitinib. 1 ml of the solutions was evacuated at 0, 1, 2, 4, 8, 12, 24 h, which were detected by HPLC at the λ max (254 nm).

Immortalized human keratinocytes (HaCaT cells), purchased from the China Center for Type Culture Collection, were cultivated in RMPI 1640 medium and complemented with 10% Fetal bovine serum (Gibco, USA), penicillin and streptomycin (NCM, China). HaCaT cells were incubated in 5%CO₂ circumstance at 37°C. HaCaT cells were co-stimulated with 25 ng/ml TNF-α and 50 ng/ml IL-17A (Abclonal, China) for 24 h in order to induce psoriatic inflammation and oxidative stress.

Wild mice C57BL/6 (female, 6-8 week old) were obtained from GemPharmatech Co.,Ltd. Mice were hosted in a controlled environment with specific-pathogen-free condition. We used C57BL/6 mice to establish the psoriasiform skin inflammation model because this strain is proven to provide a better genetic background for the pathogenesis of psoriasis than other strains (35). C57BL/6 mice were divided into control and IMQ groups. After 6-days 62.5 mg imiquimod (IMQ) cream (Mingxin Sichuan, China) was applied to the shaven dorsal skin to induce psoriatic dermatitis. Carbopol^®934 was prepared by neutralizing the gel to pH 6 by triethanolamine. 0.75% Carbopol^®934 was chosen as the matrix owing to ease in utilization and enough viscosity for topical application. Car@Deu@PEPS-Rhb were topically used in back skin (1 mg/kg) once and mice were sacrificed at different time intervals. The permeation of Car@Deu@PEPS-Rhb in psoriatic lesions was detected by immunofluorescence. The distribution of Car@Deu@PEPS-Rhb was imaged using In vivo imaging system (IVIS) at 24h, 48h, 72h time points to evaluate accumulation in different organs. Ex vivo optical imaging were conducted using black paper. All of the above performed in a dark environment.

Cellular uptake was performed by HaCaT keratinocytes. Rhb labeled PEPS was used to loading deucravacitinib for fluorescence imaging. Cells were paved onto 12-well flat-bottomed plates (3 × 10⁵ cells/well). After cell adherence, 0.5 μg/ml Deu@PEPS or 0.5μg/ml deucravacitinib were added and pre-incubated for 12h. Next, cells were dealed with 200 μM H₂O₂ (incubated for 6 h) or TNF-α and IL-17A (incubated for 24 h) to stimulate ROS. The whole experiment was conducted keep out of the light.

HaCaT keratinocytes (5 × 10³ cells/well) were plated until cell attachment. For toxicity detection, PEPS or Deu@PEPS (0-400 μg/ml) were cultured for 48 h. Next, HaCaT cells were treated with 10% CCK8 (NCM, China) and incubated. Cell viability was accessed by the microplate reader. For cell proliferation assay, the time when cells adhered to the wall was recorded as point 0, then IL-17A + TNFα and different concentration of Deu@PEPS or deucravacitinib were incubated continued (24, 48, and 72 h). Finally, the cells were managed with CCK8 and calculated as we performed before.

Fluorescent probes 2′,7′-Dichlorodihydrofluoresce in diacetate (Solarbio, China) were used to detect intracellular ROS in cells. DCFH-DA as oxidative sensitive fluorescent probe allowed to assess specific intracellular ROS by being oxidised to fluorescent DCF (36).The loss of MMP was estimated by JC-1 kit according to manufacturer’s introduction (Beyotime, China). To perform assays, HaCaT cells (2 × 10⁵ cells/well) were paved overnight. Then, the cells were pretreated with deucravacitinib or Deu@PEPS (0.5 μg/ml)and co-stimulated by TNF-α and IL-17A for 24 h. The cells were dyed with JC-1 dye or DCFH-DA probes (10 μM) in the sunblock and washed to remove remaining probes. The staining was tested using the fluorescence microscope.

Real-Time PCR was utilized to estimate the mtDNA copy number (37). DNA was isolated from psoriatic HacaT cells with DNA extraction kit (Beyotime, China). Normalizing mtND1 gene levels to nuclear beta-2 microglobulin levels were used to access mtDNA copy number (38). Genomic DNA was amplified by elongase polymerase primers (KOD FX, Toyobo, Japan) and long mtDNA fragments were detected ( Additional file 2: Table S2 )

After 7 days of adaptation in the SPF environment, the hair of 6-8 week female mice C57BL/6 back skin was shaved (n = 6-7 per group). Mice were normally fed for 1 day to restore the stratum corneum, 62.5 mg of IMQ was applied to induce psoriatic dermatitis for 6 consecutive days. Psoriasis Area and Severity Index (PASI) scores for thickening, erythema and scaling were determined as reported (39). Body weights changes of mice were recorded. After mice euthanization, spleen samples were weighed, photographed and spleen/body wt% was measured. Skin tissues were quick freezed by liquid nitrogen. Finally, to evaluate short-term biocompatibility of hydrogels in vivo, mice blood samples were used to assess hepatorenal function by biochemical detection.

Tissues of mice skin and organs were fixed with formaldehyde. Then tissues were sectioned at 7-10 μm for H&E staining, immunohistochemistry and immunofluorescence. H&E staining was conducted for evaluation of epidermal hyperplasia, skin inflammation and major organ toxicity. Means of epidermal thickness were counted using Image J.

Tissue sections were stained with Ki67 (GB121141, Servicebio, China), CD3 (GB11014, Servicebio, China), myeloperoxidase (GB15224, Servicebio, China) for IFC and STAT3 (10253, proteintech, China), Krt17 (A3769, Abclonal, China), Cyclin D1 (A0310, Abclonal, China) for IF. Slices were scanned using 3DHISTECH.

DHE can be oxidised into red fluorescence because it is sensitive to reactive oxygen species (40). To evaluate the ROS levels in IMQ induced skin lesion, skin samples were collected after mice were sacrificed and were frozen instantaneously. The frozen sections were stained with DAPI (Beyotime, China) and DHE (Servicebio, China) at 37°C for 30 mins.

The H₂O₂ scavenging power of PEPS to ROS was evaluated by ABTS radical cation scavenging activity test (T-AOC). The scavenging ability of Deu@PEPS in HaCaT was evaluated after co-stimulated with TNF-α + IL-17A for 24 h by T-AOC and SOD. The levels of T-AOC, SOD, MDA and GSH-Px in mice skin and plasma were detected by assay kits according to the manufacture’s guideline (Nanjing JianCheng, China).

TUNEL assay was utilized to monitor in situ DNA fragmentation. Mice skin lesions were embedded in optimal cutting temperature compound and stained with the TUNEL kit (Servicebio, China).

20 mg RNA of mice skin samples (physical homogenization) and cells were extracted by TRIzol (Invitrogen, USA). After chloroform extraction, isopropyl alcohol was added and allowed to stand before centrifugation. Next, the RNA was washed with ethanol, centrifuged and dried, and 25 μL DEPC water was added. The purity and concentration of RNA were measured by NanoDrop (ThermoFisher, USA). Genomic DNA removal was performed and complementary DNA was reversed transcribed (Accurate Biology, China). Quantitative realtime polymerase chain reaction (PCR) was performed by SYBR Green method (TransGen Biotech, China) using RT-PCR System (Bio-Rad, CA). Primers are described in the ( Additional file 2 : Table S3 ). The mRNA expression levels were figured by 2^−ΔΔCT method. Data were normalized to β-actin or GAPDH and compared with controls.

Different formulas were used on the back of C57BL/6 mice continuing 28 days and control group only managed with Carbopol^®934 (n = 4 per group). The weight of the mice was documented and photographed before the hydrogel was given. Signs of skin irritation were monitored as reported before (41). Scratching bouts were assessed every 7 days. After 28 days, mice were euthanized after orbital vein blood collection under anesthesia. Spleen samples were weighed, photographed and spleen/body wt% was measured as the index of immune activation. Long term biocompatibility in vivo was evaluated like short term before.

Statistical analyses were conducted by GraphPrism 9.0. The data are shown as the mean ± standard deviation (Mean ± SD). One-way or two-way analysis of variance was utilized for multiple comparisons. If not specifically requested, p < 0.05 was considered statistically significant.

The diblock polymerization of PEPS was characterized by the peaks in the ¹H nuclear magnetic resonance (NMR) spectrum ( Additional file 1: Figure S1 ): ¹H NMR (chloroform-d) δ 3.64 (s, 4 mH, OCH₂CH₂), 3.49–3.44 (m, 3H, CH₃), 2.91–2.62 (d, nH, SCHCH₂), 1.36 (d, 3nH, CH₃). The mass ratio of 10/1(PEPS to deucravacitinib) was ultimately selected on the basis of the higher drug encapsulation efficiency (~45.76%) and drug loading rate (~4.16%) ( Figure 2A ), as determined by HPLC. PEPS and Deu@PEPS both showed the Tyndall phenomenon and a spherical in morphology by TEM ( Figure 2B ). DLS exhibited a uniform size distribution, which showed the average hydrodynamic size of the PEPS and Deu@PEPS was 89 nm and 105 nm ( Figure 2C ). Fourier-transform infrared spectroscopy (FTIR) showed a slight peak at 750–830 cm^–1, suggesting the ionization of the benzene ring structure. The addition of deucravacitinib led to conformation changes of the enhanced carbon-carbon and carbon-oxygen double bond vibration at 1480–1650 cm^-1 and 1680–1720 cm^-1, respectively. CH₃ stretching vibration was also enhanced at 2815–2930 cm^–1, indicating that deucravacitinib was successfully loaded on the PEPS ( Figure 2D ). We also synthesized rhodamine b-labeled PEPS (Rhb-PEPS) as a fluorescence probe for monitoring ( Additional file 1: Figure S2 , Additional file 2: Table S1 ).

Next, H₂O₂ was selected to assess the ROS-responsive reaction and drug release behavior ( Figure 2E ). The appearance of an absorption band at 970–1090 cm⁻¹ in the PEPS products indicated conversion of the S=O group to poly (propylene sulphone) after H₂O₂ stimulation ( Figure 2F ). We further evaluated the antioxidant activity of PEPS by a 2,2-azinobis (3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) activity. The PEPS manifested significant clearance of ABTS radical cations in the concentration range of 0.1 to 4.0 mg/ml in a dose-dependent pattern ( Figure 2G ). Finally, The drug release was faster at 0-4 h (43.08% ± 2.08%) under 10 mM H₂O₂ than under 5 mM H₂O₂ (32.31% ± 1.31%). At 24 h, 91.1% ± 7.70% of the deucravacitinib had been released, which was significantly higher than the amount released in the absence of H₂O₂ (14.67% ± 0.3%) ( Figure 2H ). DLS further confirmed the stability of storing the SPMs at room temperature for 1 month by assessing the hydrodynamic diameter ( Additional file 1: Figure S3 ).

To facilitate in vivo skin permeability performance and quantification, Deu@PEPS was labeled with Rhb fluorescent dye and administered topically to IMQ-induced mouse model, which has been widely used for psoriatic physiopathology and drug development owing to its strong dependence on the IL-23/IL-17 axis (39). The SPMs were incorporated in 0.75% Carbopol^®934 to produce the hydrogel (Car@Deu@PEPS-Rhb). A single dose of Car@Deu@PEPS-Rhb (1 mg/kg) was topically administered to the dorsal skin. Immunofluorescence (IF) revealed the similar red fluorescence pattern between the control and IMQ groups under the SC and visible epidermal layer at 6 h. Both groups displayed the strong fluorescent signal in the dermis, which increased in a time-dependent manner during 24 h. After 36 h administration, the deposition of Deu@PEPS remained distinct in the dermis of the IMQ group but was negligible in the control group ( Figure 3A ), indicating enhanced skin retention in the oxidative stress microenvironment of the psoriatic skin. Meanwhile, TNF-α and IL-17A were utilized to establish the psoriasiform and oxidative stress HaCaT model. Red fluorescence was first observed in the TNF-α and IL-17A stimutated group as early as 2 h rather than control group. At 6 h, the fluorescent signal was mainly concentrated around the cells, whereas it was concentrated in the center of the cells at 12 h of incubation ( Figures 3B, C ), indicating earlier cell uptake of the SPMs by TNF-α and IL-17A stimulation.

Finally, The major organs were collected at different times after topical administration. The biodistribution of Car@Deu@PEPS-Rhb in the major organs showed that the SPMs mainly accumulated in liver and kidney. But we observed no significant difference between the IMQ and control groups at 24 h in liver in both groups (p = 0.6874). Meanwhile, both two groups showed a similar metabolism rate in liver from 24 to 48 h (p < 0.0001) ( Figures 3D, E ). By 72 h, the fluorescence intensities in the liver and kidneys of the two groups were similar to those of the PBS group, which manifested a nearly consistent metabolic profile in both two groups.

To confirm the therapeutic effect of Deu@PEPS to restrain the overproduction of ROS to control the inflammation in keratinocytes, we used 2′,7′-dichlorofluorescein diacetate (DCFH-DA) to detect introcellular ROS level. Firstly, Intracellular ROS level were elevated after 200 μM H₂O₂, which were similar to TNF-α and IL-17A co-stimulation. Deu@PEPS exhibited a lower intensity of green fluorescence compared with control group, confirming the ability of the SPMs to obliterate excessive ROS in psoriatic keratinocytes ( Additional file 1: Figure S4 ). T-AOC and SOD levels of HaCaT cells decreased after stimulation with TNF-α and IL-17A, but increased after treatment with Deu@PEPS ( Additional file 1: Figure S5 ). In comparison with control group, the psoriasiform HaCaT group showed negligible red fluorescence and enhanced green fluorescence, manifesting the loss of MMP. However, Deu@PEPS (0.5 μg/ml) recovered mitochondrial polarization under oxidative stress, which manifested as strong red fluorescence and an increased ratio of red/green fluorescence ( Figures 4A, B ). The mtDNA copy number was reduced in HaCaT cells after TNF-α and IL-17A, and this depletion was reversed with Deu@PEPS pretreatment, but not with treatment of deucravacitinib alone, suggesting the ability of PEPS to prevent mtDNA damage ( Figure 4C ). Taken together, our results suggest that Deu@PEPS has antioxidant properties to rescue oxidative stress-caused mtDNA damage and mitochondrial dysfunction.

We next investigated the anti-psoriatic effect of Deu@PEPS among the keratinocytes. PEPS and Deu@PEPS exhibited gradual concentration-dependent cytotoxicity with a half-maximal inhibitory concentration value of 11.66 ± 1.91 μg/ml and 8.23 ± 1.38 μg/ml, respectively ( Figure 4D ). The co-stimulation of TNF-α and IL-17A significantly increased proliferation over time. However, pretreatment of Deu@PEPS resulted in a decrease in cell proliferation in a dose-dependent manner from 24 to 72h. 0.5 μg/ml Deu@PEPS exhibited more powerful anti-proliferative capacity at 72 h compared to the same dose of deucravacitinib alone ( Figure 4E ). Quantitative real-time polymerase chain reaction (qRT-PCR) showed that TNF-α and IL-17A stimulation increased the mRNA expression of the inflammatory cytokines ( Figure 4F ), whereas Deu@PEPS pretreatment decreased their expression levels at both doses, and 0.5 μg/ml deucravacitinib alone also showed a slight but significant attenuating effect (p<0.05).

Before evaluating the therapeutic effects of Deu@PEPS in vivo, a pilot study was carried out to determine the optimal dose or treatment interval of the Deu@PEPS. Compared to the IMQ group, the free deucravacitinib (5 mg/kg, QD) and blank PEPS (10 mg/kg, QD) groups showed similar effective anti-psoriatic effects. However, these effects were inferior to those of Deu@PEPS (0.45 mg/kg deucravacitinib, BID), which exhibited the optimum therapeutic effect to improve the skin clinical phenotype ( Additional file 1: Figure S6 ). H&E staining showed conspicuous depressed epidermal thickness and infiltration with inflammatory cells after treatment with Deu@PEPS (0.45 mg/kg deucravacitinib, BID). This result was consistent when considering the effect on total PASI scores. Our preliminary study showed that a low dose of Deu@PEPS (0.45 mg/kg deucravacitinib, BID) was more effective than a 11.11-times higher dose of deucravacitinib alone (5 mg/kg, QD). The increased spleen body wt% is another symbol of psoriasis severity, which reflects increased immune activation in spleen (39, 42). Deu@PEPS resulted in a reduction in IMQ-induced splenomegaly; however, there was no significant change of spleen body wt%, which is likely due to individual differences and the small number of mice.

Figure 5A shows a schematic diagram of the synthesis and transdermal delivery for the treatment of psoriatic dermatitis. The IMQ-induced macroscopic scaly and thickened skin was apparently reduced of various treatment groups (BID) to some extent. In particular, Car@Deu@PEPS resulted in a similar skin appearance to that of the control group, which was superior to a 2.22-times higher dose of Car@Deu ( Figure 5B ). To our surprise, Car@PEPS group attained the greatest and earliest weight recovery, which was even better than that of the Car@Deu@PEPS group ( Figure 5D , Additional file 1: Figure S7 ). Consistent with the improvements in skin appearance, H&E staining showed the greatest improved parakeratosis and epidermal acanthosis in Car@Deu@PEPS group ( Figure 5C ). The epidermal thickness was reduced by 33.4%, 57.4%, and 44.6% in the Car@Deu, Car@Deu@PEPS and CAL groups, respectively ( Figure 5E ). Car@Deu@PEPS group showed the most significantly atrophied spleens and diminished spleen body wt% after IMQ treatment, similar to those of the CAL group, as we expected ( Figures 5F, G ). After 7 days’ experiment, Car@Deu@PEPS group obtained the lowest total PASI (1.16 ± 0.98) and IMQ group gained the highest total PASI (10.50 ± 0.56). By contrast, the scaling, erythema, and thickness scores were increased in the Car@PEPS and Car@Deu@PEPS group initially, which was likely due to the fact that the less ROS was insufficient to induce micelle disintegration at the beginning ( Figure 5H ). Tissue samples of organs and blood were collected to assess the short-term toxicity. As shown in Additional file 1: Figure S8 , no apparent changes of histological and serum biochemical indicators were observed, indicating that these hydrogels only cause negligible toxicity.

To confirm whether continuous topical application could prevent psoriatic pathologic hallmarks, immunohistochemistry was performed to evaluate hyperproliferating keratinocytes and inflammatory infiltration ( Figures 5I, J ). Staining of the hyperproliferating marker Ki67 was enhanced along the basal layers of the epidermis in IMQ group. However, the number of Ki67-positive cells and the staining intensity were reduced to the greatest extent in the Car@Deu@PEPS group compared to those of the IMQ group (red arrows). A cluster of neutrophils (MPO⁺, green arrows) was found in the dermis of the IMQ group and neutrophil infiltration was decreased to the greatest extent in the Car@Deu@PEPS group. Moreover, the Car@Deu@PEPS hydrogel alleviated the infiltration of T cells (CD3⁺, yellow arrows), particularly in the dermis layer.

Whether Deu@PEPS reduces ROS overload in psoriatic lesions in vivo needs to be further explored by Dihydroethidium (DHE) staining. Compared to the excessive ROS accumulation detected in the IMQ group, DHE fluorescence intensely was substantially reduced in the Car@Deu@PEPS and Car@PEPS groups, with only slight reduction detected in the Car@Deu and CAL groups ( Figure 6A ). These results indicated that PEPS could decrease the generation of ROS in psoriatic skin, which is attributed to the conversion of the S=O group of PEPS nanocarriers. To evaluate whether topical administration influenced antioxidant enzyme system, we monitored the levels of oxidative stress markers and the lipid peroxidation product Malondialdehyde (MDA), for intracellular metabolism fatty acid oxidation could be exacerbated of psoriasis (43). The levels of T-AOC, SOD, and GSH-Px in skin lesions were significantly reduced after the induction of psoriatic dermatitis. And these reductions were markedly blocked by treatment of Car@Deu@PEPS and Car@PEPS. In contrast, MPA was increased in the lesions of the IMQ model group and decreased following Car@Deu@PEPS treatment ( Figure 6B ). The plasma antioxidant oxidase levels (catalase, SOD) also decreased after Car@PEPS and Car@Deu@PEPS treatment, although the decrease in SOD levels was not statistically significant ( Figure 6C ).

Finally, qRT-PCR demonstrated that IMQ topical application profoundly decreased the relative mRNA expression of SOD1, SOD2, HO-1 and NRF2, which were improved to the greatest extent by Car@PEPS treatment, followed by Car@Deu@PEPS treatment, showing similar levels to those of the control group. CAL and Car@Deu showed slight antioxidative capability, but there was no significant difference in the levels of the antioxidant-related genes from those of the IMQ group ( Figure 6D ).

Autoimmune feedback loop mediated by K17 is vital in psoriasis both as the inflammation responder and adjuster (44). Accordingly, we investigated K17 and Cyclin D1 under increased STAT3 expression in psoriatic skin samples by immunofluorescence. IMQ remarkably promoted the expression of K17 and Cyclin D1 by inducing STAT3 expression, whereas the topical administration of Car@Deu@PEPS remarkably reversed these trends. Car@Deu alone also reduced the expression of these factors to some extent, which was likely owing to inhibition of STAT3 as a TYK2 inhibitor ( Figure 7A , Additional file 1: Figure S9 ). The results of qRT-PCR analysis showed that the Deu@PEPS topical formula significantly reduced the mRNA expression of psoriatic elements, including cytokines in the IL-23/IL-17 axis (IL-12, IL-17A, IL-23), cytokines associated with innate immunity (IL-6 and TNFα), edigree specific transcription factors of Th17 cell differentiation (RORγt), and the keratinocyte proliferation marker (Krt17) compared to those of the other groups ( Figure 7B ), including the positive control group.

IMQ treatment upregulated the mRNA levels of NLRP3, IL-1β and IL-18, indicating that the inflammasome of NLRP3 was activated in the skin of IMQ-induced psoriatic mice. AIM2, which was associated with the assembly of NLRP3 and the secretion of the pro-inflammatory cytokines IL-1β and IL-18, was also upregulated in the IMQ group. PEPS could cooperate with deucravacitinib to reduce the IMQ-induced elevations in the mRNA expression of inflammasomes and related cytokines by its antioxidative effects in skin lesions ( Figure 7C ). Deu@PEPS also had an effect on pyroptosis by reducing the GSDMD expression level that was elevated in the IMQ group ( Figure 7D ). Finally, terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) staining was conducted to detect pyroptotic cascades that lead to DNA fragmentation. Consistent with the results of qRT-PCR, the results of the TUNEL assay revealed that the number of apoptotic cells in the IMQ mice was higher than that of the control group; however, no increase in TUNEL-positive cells was detected after treatment with Deu@PEPS ( Figure 7E , Additional file 1: Figure S10 ).

For transdermal delivery of nanomedicines, skin irritation and nanomedicine toxicity assessment are essential for their clinical translation (45). Previous work has demonstrated that PEPS exerts a non-inflammatory (46). To confirm the feasible clinical application of the SPMs hydrogel formulation, we assessed their potential for irritation according to the skin appearance and histological examination after application to the back skin of normal mice continuing 28 days. The surface of mice showed mild erythema in the Car@Deu group and certain level of skin atrophy in CAL group, whereas atrophy was negligible in the other treatment groups( Figure 8A ). Skin H&E staining showed no apparent pathological changes, except for epidermal thickening detected in the CAL group, compared with the Car group ( Figure 8B ). CAL and Car@Deu treatments led to increased scratching bouts compared to the mice in the other groups ( Figure 8C ). There were no evident changes in appearance of the spleen or in the spleen/body wt% of the different groups, indicating no significant activation of the immune system for long-term toxicity ( Figure 8D , Additional file 1: Figure S11 ). Over the 28 days of treatment, the CAL and Car@PEPS groups had the least and greatest weight change compared to the initial weight, respectively ( Additional file 1: Figure S12 ).

Additionally, biological safety evaluation was preformed by blood routine examination and serum biochemical indicators. According to Figure 8E , compared to Car only, both Car@PEPS and Car@Deu@PEPS didnot affect blood cell count, including red and white blood cells, monocytes (Mon), lymphocytes (Lymph), granulocytes (GRAN), platelets (PLT) and hemoglobin (HGB). Simultaneously, both Car@PEPS and Car@Deu@PEPS did not exhibit any hepatotoxicity or nephrotoxicity ( Figure 8F ) except for long term CAL. Histological examination of the main organs showed no obvious pathological changes such as necrosis, hyperplasia, or inflammatory cell infiltration ( Figure 8G ). Therefore, these results could reasonably indicate that Car@Deu@PEPS is a biocompatible product with minimal skin irritation and negligible toxicity, demonstrating its further potential for achieving the topical application of deucravacitinib in clinical practice.