Imiquimod cream was obtained from Mingxin Pharmaceutical Co., Ltd. (Sichuan, China). CU-T12–9 and NIK SMI 1 were obtained from MedChemExpress (Shanghai, China). Recombinant murine IL-15 (rIL-15) was obtained from PeproTech (Rocky Hill, USA).

A recurrent psoriatic mouse model was established according to our previous research (16). Seven-week-old male BALB/c mice were obtained from the Guangdong Medical Laboratory Animal Center (Guangzhou, China). Mice were housed under pathogen-free conditions in a temperature-controlled room with a 12-h light/dark cycle and received humane care. Mice were divided into the control, imiquimod (IMQ)–vaseline, and IMQ–IMQ groups. After shaving the back hair was shaved, all mice except those in the control group were treated with 62.5 mg/day IMQ for 7 consecutive days. After immunologic tolerance for 14 days (a treatment-free resting period following initial IMQ application), mice in the IMQ–vaseline group were received vaseline treatment for another 7 consecutive days. Mice in the IMQ–IMQ group were received 20.8 mg/day IMQ for 7 consecutive days to induce recurrent psoriasis (Figure 1A). The severity of psoriasis was evaluated using the Psoriasis Area and Severity Index (PASI), acanthosis, and papillomatosis. PASI was assessed based on three parameters (skin erythema, scaling, and thickness). Each independent parameter was scored as follows: 0, none; 1, slight; 2, moderate; 3, marked; and 4, very marked. To measure acanthosis, the epidermal area was outlined, and its pixel size was measured. The relative area of the epidermis was calculated using the following formula as follows: area = pixels/(horizontal resolution × vertical resolution). The papillomatosis index was typically measured as previously reported (17).

Skin tissues were fixed in formaldehyde, paraffin-embedded, sectioned, and stained with hematoxylin and eosin.

Skin tissues were homogenized in RIPA lysis buffer. Cutaneous IL-17A, IL-15, IL-6, and TNF-α levels were determined using commercial ELISA kits (Boster, China).

The flow cytometric analysis protocol was performed according to our previous research (16). LNs and spleens were harvested and homogenized. Cells were collected and stained with anti-CD4 PE antibody (Clone GK1.5, eBioscience, USA), anti-CD8a FITC antibody (Clone 53-6.7, eBioscience, USA), anti-CD44 eFluor™ 450 antibody (Clone IM7, eBioscience, USA), and anti-CD62L APC antibody (Clone MEL-14, eBioscience, USA). Skin tissues were digested with collagenase type I (Sigma-Aldrich, USA) and DNase (Solarbio, Shanghai, China). After erythrocyte lysis, cells were stained with anti-CD8a Alexa Fluor™ 700 antibody (Clone 53-6.7, eBioscience, USA), anti-CD103 PE antibody (Clone 2E7, eBioscience, USA), anti-CD69 PE-Cyanine7 antibody (Clone H1.2F3, eBioscience, USA), anti-CD4 Alexa Fluor™ 700 antibody (Clone GK1.5, eBioscience, USA), anti-CLA PE antibody (Clone HECA-452, BioLegend, Beijing, China), anti-CD44 eFluor™ 450 antibody (Clone IM7, eBioscience, USA), anti-CD62L APC antibody (Clone MEL-14, eBioscience, USA), and anti-DLAT antibody (MA5-53986, Thermo Fisher Scientific, USA). Cells were analyzed using a flow cytometer (NovoCyte Quanteon, Agilent, USA).

Skin tissues were homogenized in TRIzol reagent (Invitrogen, Carlsbad, USA). Total RNA was isolated and synthesized into single-stranded cDNA. Finally, Quantitative real-time PCR was performed using a ViiA 7 Dx system (Applied Biosystems, Carlsbad, USA) with SYBR Premix ExTaqTM Π (Takara Bio Incorporation, Tokyo, Japan). The primer sequences are shown below:

CXCL9 Forward: CTCGGCAAATGTGAAGAAGCTGA;CXCL9 Reverse: TTCCTTGAACGACGACGACTTTGG; IL-15 Forward: TCCATCTCGTGCTACTTGTGTTTCC; IL-15 Reverse: TCCCTATGGCCCTCATTCTCACTG; CXCL10 Forward: TGCCGTCATTTTCTGCCTCATCC; CXCL10 Reverse: CACATTCTGGAGGAAGTCCTTGG; CXCR3 Forward: CAGCCCTCTACAGCCTCCTCTTC;CXCR3 Reverse: TACAGCCAGGTGGAGCAGGAAG; β-actin Forward: GGCTGTATTCCCCTCCATCG, β-actin Reverse: CCAGTTGGTAACAATGCCATGT.

Skin tissues were homogenized in RIPA lysis buffer as described above. Total protein was isolated and separated by SDS-PAGE. After transfer to PVDF membranes, membranes were blocked and incubated with anti-GAPDH (Cell Signaling Technology, Boston, USA), anti-phospho-NF-κB p65, (Cell Signaling Technology, Boston, USA), anti-NF-κB-p100/p52, anti-p-IKKα, anti-p-RelB, (Cell Signaling Technology, Boston, USA), anti-phospho-TRAF2, and anti-phospho-p38 MAPK antibodies (Cell Signaling Technology, Boston, USA). After incubation with the appropriate secondary antibodies, membranes were visualized using a Bio-Rad gel imaging system and analyzed with Image Lab software.

To purify CD4⁺CD44⁺CD62L⁺ cells or CD8⁺CD44⁺CD62L⁺ cells for adoptive transfer experiments, spleen cells from naïve wild-type mice were stained with anti-CD4 Abs, anti-CD8 Abs, anti-CD44 Abs and anti-CD62L Abs (BD Biosciences, USA). CD4⁺CD44⁺CD62L⁺ cells or CD8⁺CD44⁺CD62L⁺ cells were then sorted out by FACSAria III (BD Biosciences). The purity of the sorted CD4⁺ Tcm and CD8⁺ Tcm is typically >85%.

Mice were divided into control, IMQ–IMQ, CD8⁺ Tcm, CD4⁺ Tcm, and rIL-15 groups. All mice except those in the control group were induced to develop recurrent psoriasis mice as described above. Mice in the CD8⁺ Tcm or CD4⁺ Tcm groups were injected with exogenous CD4⁺ or CD8⁺ Tcm cells (1×10^{^6}) on day 21, respectively. Mice in the rIL-15 group were received recombinant murine IL-15 (i.p., 0.1 μg/mouse) on days 21, 24, and 27. These Tcm cells were isolated from mice treated with IMQ on day 21.

Mice were divided into control, IMQ–IMQ, NIK SMI 1, and CU-T12–9 groups. All mice except those in the control group were induced to develop recurrent psoriasis as described above. Mice in the NIK SMI 1 and CU-T12–9 groups received 20 mg/kg NIK SMI 1 (i.p.) or CU-T12-9 (60 mg/kg, i.p.) on days 21, 24, and 27, respectively.

To obtain mouse keratinocytes, the skin of 4–7-day-old BALB/c pups was isolated and the tissue was incubated in a dissociation enzyme solution at 4 °C for 16–21 h. After epidermal separation, TrypLE™ Express (Gibco) was added and incubated at 37 °C with shaking for 30 min. Digestion was terminated by adding primary keratinocyte culture medium (PriMed-iCell-010, iCell, China). After filtration and centrifugation, the keratinocytes were resuspended and collected. Primary keratinocytes were cultured in primary keratinocyte medium (PriMed-iCell-010, iCell, China). To induce a psoriasis-like model in vitro, recombinant mouse IL-17A (10 ng/mL) was added to the culture supernatant. For the model group, prior to co-culture, the sorted CD8⁺ T cells were pre-differentiated into tissue-resident memory T cells (Trm cells) by induction with recombinant IL-15 (50 ng/mL), NIK SMI 1 and TGF-β (50 ng/mL) Subsequently, these pre-differentiated CD8+ Trm cells were further used to treat the keratinocytes in the co-culture experiment for 48 h. For the NIK SMI 1 group, the keratinocytes were further treated with CD8⁺ Trm cells which induced with recombinant IL-15, TGF-β, and NIK SMI 1 for 48 h. All cells were maintained in a humidified incubator at 37 °C with 5% CO₂.

All results are expressed as means ± standard deviation. Data were analyzed using the Kolmogorov–Smirnov test for normality. The Mann–Whitney U test was used for non-normally distributed data. Data from more than two groups were analyzed using one-way analysis of variance (ANOVA). Student’s t-test was used to identify differences between two groups. A p value <0.05 was considered statistically significant.

After treating with repeated IMQ treatment (Figure 1A), obvious papulosquamous plaques and bleeding were observed on the backs of mice, along with severe epidermal hyperplasia (Figure 1B). Mice in the IMQ–IMQ group also showed higher PASI scores than the control group (Figure 1C). Histological analysis further showed that mice in the IMQ–IMQ group mice also displayed severe acanthosis and papillomatosis (Figure 1C). These data indicate that recurrent psoriatic mice were successfully established. Meanwhile, cutaneous inflammatory cytokines (IL-17A, IL-6, and TNF-α) were elevated in the IMQ–IMQ group (Figure 1C). These findings suggest that psoriasiform dermatitis emerges in repeated IMQ-induced recurrent psoriatic mice.

To investigate the phenotype of T memory cells, Tcm and Tem cells were determined in the Tcm and Tem in spleen and LNs of IMQ-induced primary psoriatic mice on day 7. Both CD8⁺ Tcm and Tem cells were increased in the spleen and LNs of primary psoriatic mice on day 7 (Figures 2A, B). Similarly, CD4⁺ Tcm and Tem cells were also increased in the spleen and LNs of psoriatic mice (Figures 2C, D). These data indicate that both CD4⁺ and CD8⁺ T memory cells are activated in primary psoriasis.

Tem and Tcm cells can further differentiate into Trm cells and persist in the skin (18). After verifying the Tem and Tcm cells in primary psoriatic mice, we further explored the Trm cells in the skin of recurrent psoriatic mice. CD103, CD69, and CLA are biomarkers of Trm cells. The ratio of CD8⁺/CD4⁺ cells was significantly increased in the skin of recurrent psoriatic mice (Supplementary Figure 1B). Furthermore, the CD8⁺ CD103⁺, CD8⁺ CD69^+,and CD8⁺ CLA⁺ Trm cells both were all increased in the skin of IMQ–IMQ-induced recurrent psoriatic mice (Figures 3A, B). Interestingly, there was no significant difference in CD4⁺ CD69⁺ Trm cells between the control group and IMQ–IMQ groups (Figures 3C, D). These data suggest that CD8⁺ Trm cells, but not CD4⁺ Trm cells, are more important in recurrent psoriasis. This finding is consistent with previous research showing that CD8⁺ Trm cells generate IL-17A to promote local skin inflammation in psoriasis (19). In addition, cutaneous gene expression of IL-15, CXCR3, CXCL9, and CXCL10 was upregulated in recurrent psoriatic mice (Figure 3E). These genes are associated with T cell recruitment and Trm cell generation (3, 20–22). Hence, we speculate that CD8⁺ Trm cells are activated and contribute to the recurrence of psoriasiform dermatitis.

To validate the effects of CD8⁺ memory cells in recurrent psoriasis, we separated the CD8⁺ Tcm and CD4⁺ Tcm cells were isolated from primary psoriatic mice and injected into recurrent psoriatic mice (Figure 4A). We aimed to investigate whether CD8⁺ or CD4⁺ Tcm cells will promote recurrent psoriasis. After injection of CD8⁺ Tcm cells, mice exhibited more severe psoriatic symptoms than IMQ–IMQ-induced recurrent psoriatic mice (Figures 4B, C). Cutaneous IL-17A, IL-15, IL-6, and TNF-α levels were also higher in the CD8⁺ Tcm-treated group than in IMQ–IMQ-induced recurrent psoriatic mice (Figure 4D). However, injection of CD4⁺ Tcm cells resulted in psoriatic symptoms similar to those observed in IMQ–IMQ-induced recurrent psoriatic mice (Figures 4B–D). Meanwhile, mice treated with rIL-15 also exhibited psoriasiform dermatitis, high PASI scores, acanthosis, papillomatosis, and elevated inflammatory cytokines, comparable to the CD8⁺ Tcm group (Figures 4B–D). IL-15 is a key cytokine that promotes the generation and survival of Trm cells in the skin (23–27). Therefore, these findings indicate that IL-15–associated CD8⁺ Trm cells, but not CD4⁺ Trm cells, promote the recurrent psoriasis.

To explore the mechanisms underlying Trm cell activation in recurrent psoriasis, the protein expression of classic NF-κB and MAPK were determined. The phosphorylated regulatory factor of canonical NF-κB and MAPK signaling components was examined. Phosphorylated canonical NF-κB p65 (RELA) and non-canonical p-IKKα, p-RelB and NF-κB p100/p52 were up-regulated in IMQ-IMQ induced recurrent psoriatic mice (Figures 5A–F). These results suggest that both canonical and non-canonical NF-κB signaling pathways may contribute to activation of Trm cell activation. TRAF2 is a TNF receptor–associated factor that can be activated by TNF family members and functions as an adaptor protein in activating NF-κB activation through various downstream signaling mechanisms, including mitogen-activated protein kinase (MAPK) pathways (28, 29). However, the phosphorylation levels of p38 MAPK and TRAF2 in the skin were similar between the control group and IMQ–IMQ groups (Figures 5G–I).

To validate the role of NF-κB in CD8⁺ Trm cells and recurrent psoriasis, IMQ–IMQ-induced recurrent psoriatic mice were received an NF-κB inhibitor (NIK SMI 1) or an NF-kB agonist (CU-T12-9) which served as a mechanistic control for NF-κB activation (Supplementary Figure 1A). After treatment with NIK SMI 1, the papulosquamous plaques, PASI scores, and epidermal hyperplasia both were all normalized in recurrent psoriatic mice (Figures 6A, B; Supplementary Figure 1C, D). Cutaneous IL-17A, IL-6, IL-15, and TNF-α levels were reduced following NIK SMI 1 treatment (Figure 6C). In addition, CD8⁺ CD103⁺ and CD8⁺ CD69⁺ Trm cells in the skin also were decreased by NIK SMI 1 treatment (Figures 6D, E). In contrast, mice in the CU-T12–9 group did not exhibit more severe psoriatic symptoms than those in the IMQ–IMQ group (Figures 6A–E). These data indicate that the generation of CD8⁺ Trm cells and recurrent psoriasis is dependent on NF-κB signaling in IMQ–IMQ-induced psoriatic mice. Meanwhile, the DLAT expression was reduced in Trm cells of recurrent psoriatic mice (Figures 7A, B). However, NF-κB inhibitor treatment significantly restored the expression of DLAT expression (Figures 7A, B). These findings suggest that NF-κB activation of NF-κB may promote the retention of CD8⁺ Trm cells by inhibiting the cuproptosis.

Although the effects of NF-κB signaling on CD8⁺ Trm cells were validated in vivo, these experiments were insufficient to elucidate the specific cellular mechanisms of NF-κB in CD8⁺ Trm cells. Therefore, keratinocytes and CD8⁺ T cells were co-cultured to investigate the direct effects and mechanisms of activating NF-κB activation in Trm cells. The IL-17 was used to induce psoriasis-like model in keratinocytes and the recombinant IL-15 and TGF-β were used to induce the CD8⁺ Trm cells differentiation. After 48 h, CD8⁺ Trm cells were significantly reduced following NIK SMI 1 treatment (Figures 7C, D). Expression levels of IL-6, IL-8, and IL-1β both were also decreased by NF-κB inhibition in the supernatant (Figures 7E–G). These results indicate that inhibition of NF-κB in CD8⁺ T cells suppresses CD8⁺ Trm cell differentiation of CD8⁺ Trm cells and prevent production of pro-inflammatory factors.