Salmeterol (Sal), roflumilast (Rof), and extracellular signal-regulated kinase (ERK) inhibitor U0126 were obtained from Selleck (Shanghai, China). PKA agonist forskolin (FSK) and PKA inhibitor H89 were purchased from LC Laboratories. Tumor necrosis factor-α (TNF-α) was synthesized in the PeproTech (Cranbury, NJ, United States).

Female C57BL/6J mice (7–8 weeks old) were purchased from Sipeifu Biotechnology (Beijing, China). They were fed with water and food ad libitum at 23 ± 2°C in a 12-h light/dark cycle. Animal experiments were approved by the Animal Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology. The investigation conforms to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1985). We followed ARRIVE guidelines when reporting this study.

Human keratinocyte cell line (HaCaT cell line) was purchased from the China Center for Type Culture Collection (Wuhan, China). The HaCaT cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM, HyClone, United States) containing 10% fetal bovine serum (FBS) and 1% penicillin and streptomycin in a humidified incubator with 5% CO₂ at 37°C.

Cell viability was analyzed with Cell counting kit-8 (CCK-8, Beijing Labgic Technology, Beijing, China). Briefly, cells were seeded in 96-well plates. After the indicated treatment, CCK-8 was added and incubated for 4 h. Absorbance was assessed at 450 nm with a microplate reader (Awareness Technology, Inc., United States).

Cell proliferation assay is based on incorporating 5-ethynyl-2′-deoxyuridine (EdU) into genomic DNA, which was analyzed using Cell-Light EdU Apollo567 In Vitro Kit (Riobio, Guangzhou, China). Apollo staining and Hoechst 33342 staining (for nuclear staining) were performed. Fluorescence images were analyzed using ImageJ (NIH, Bethesda, MD, United States). The EdU incorporation rate was equal to the ratio of EdU-positive cells (red)/total number of Hoechst-positive cells (blue).

HEK293 cells expressing nuclear-specific PKA biosensor NLS-AKAR3 were rinsed and maintained in PBS for fluorescence resonance energy transfer (FRET) as described previously (Liu et al., 2012). FRET was recorded by exciting the donor fluorophore at 405–455 nm and measuring emission fluorescence with two filters (470DF30 for cyan and 535DF30 for yellow). The acquisition was set with 200-ms exposure in both channels and 20-s elapses. Images in both channels were subjected to background subtraction, and ratios of yellow-to-cyan color were calculated at different time points. The donor/acceptor FRET ratio was calculated and normalized to the ratio value of baseline.

After 1 week of adaptive feeding, hair was removed from the back regions of mice 24 h before treatment (Figure 1). The mice were randomly assigned into the following 11 groups (n = 5–8 mice per group): vehicle group; IMQ group; Sal therapy groups (5, 10, or 20 μg/kg/d); Rof therapy groups (0.03, 0.1, or 0.3 mg/kg/d); and Sal (10 μg/kg/d) + Rof (0.03, 0.1, or 0.3 mg/kg/d) combination groups. The shaved back skin of the vehicle group mice and IMQ group mice received subcutaneously injections at five sites with the same amount of 1% dimethyl sulfoxide (DMSO; Sigma, United States). The vehicle group mice received topical application of Vaseline jelly, while the other mice received a daily topical dose of 62.5 mg of 5% IMQ cream on the shaved back skin for five consecutive days, 4 h after the subcutaneous injection. The severity of psoriasis-like lesions was scored based on the clinical Psoriasis Area and Severity Index (PASI). On day 6, the mice were sacrificed, and skin tissue was taken for further experiments.

Skin tissue was freshly harvested, fixed in 4% paraformaldehyde solution, and embedded in paraffin. The tissue sections (3 μm) were stained with hematoxylin and eosin (H&E) or incubated with anti-Ki67 (Servicebio, Wuhan, China). Images were captured using a light microscope, and analyzed using ImageJ. The mean thickness value was calculated by three to five random fields of view in every section. Cells positive for Ki67 in every section were counted and expressed as positive cells per mm length of basement membrane.

Terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay was performed on skin sections and HaCaT cells using the One Step TUNEL Apoptosis Assay Kit (Beyotime Biotechnology, China) according to the manufacturer’s instruction. Fluorescence images were visualized with fluorescence microscope (Olymbus, Japan) and randomly selected three to four fields were analyzed by ImageJ. The TUNEL positive rate was equal to the ratio of TUNEL-positive cells (red)/total number of DAPI-positive cells (blue).

Skin tissues and cells were lysed with the RIPA lysis buffer (Servicebio, Wuhan, China) containing protease and phosphatase inhibitor cocktail (Roche, Basel, Switzerland). The protein concentration was measured by BCA assay (Beyotime Biotechnology, Shanghai, China). Lysates (20–40 μg total protein) were resolved by SDS-PAGE. The extracts were then transferred onto a PVDF membrane (Merck Millipore, Billerica, MA). The following primary antibodies were used: pPDE4D (1:500; Abcam, Cambridge, UK); pCREB (1:500), ERK (1:1,000) and phospho-ERK (pERK, 1:500) from Cell Signaling Technology (Danvers, MA, United States); CREB (1:1,000) and Bax (1:1,000) from Bimake (Houston, TX, United States); K17 (1:500; Santa Cruz Biotechnology, United States); PDE4D (1:500), Bcl2 (1:200), and GAPDH (1:10,000) from Proteintech (Chicago, IL, United States). Chemiluminescent detection was performed with horseradish peroxidase–coupled secondary antibody from Proteintech (Chicago, IL, United States) and an ECL chemiluminescence reagent kit (Beijing Labgic Technology, China). Band densities were quantified using ImageJ software.

Cyclic AMP content in skin sections were measured with the cAMP-Glo™ Assay kit (Promega, Madison, WI) following the manufacture’s instruction.

Data were presented as mean ± standard error of the mean (SEM). The sample size for each group is shown in the figure legends. At least three sets of independent experiments were performed in the in-vitro studies. Data were analyzed by unpaired t test, one-way ANOVA followed by Turkey’s post-hoc test, or two-way ANOVA followed by Turkey’s multiple comparisons test. p < 0.05 was considered statistically significant. Statistical analysis was performed with GraphPad Prism8.0 (GraphPad Software, Inc., San Diego, CA, United States).

Salmeterol inhibited cell viability in serum culture conditions, but there was no difference among various concentrations (Figure 2A). Compared to the vehicle group, 100 M salmeterol upregulated the ratio of Bax and Bcl2 (Figure 2B). The TUNEL and EdU incorporation assays showed that salmeterol (100 nM) resulted in a significant decrease in proliferation and increase in apoptosis. Interestingly, forskolin (10 μM), a PKA agonist, induced an effect similar to that of salmeterol (Figures 2C,D). These data suggested that the effect of salmeterol on cell proliferation and apoptosis may be related to PKA signaling.

Inflammatory factors such as TNF-α, IL-23, and IL-17 contribute to the pathogenesis of psoriasis (Lowes et al., 2014). We further tested the effect of salmeterol on cell proliferation and apoptosis under TNF-α stimulation. Treatment of HaCaT cells with TNF-α resulted in significant increases in cell viability (Figure 3A) and cell proliferation (Figure 3B), and these effects were inhibited by forskolin (PKA agonist) or salmeterol. The inhibitory effect of salmeterol on TNF-α-induced hyperproliferation was attenuated by the PKA inhibitor, H89, indicating that activation of cAMP/PKA pathway contributes to the influence of salmeterol on TNF-α induced cell proliferation (Figure 3B). Under TNF-α stimulation, salmeterol or forskolin promoted cell apoptosis, which played a part in the inhibition of TNF-α-induced cell viability (Figure 3C). H89, a PKA inhibitor, also reversed this effect of salmeterol, suggesting that salmeterol-activated cAMP/PKA signaling is involved in the promotion of cell apoptosis.

A PDE4 inhibitor has been approved for the treatment of moderate-to-severe plaque psoriasis and psoriatic arthritis (Papp et al., 2013). The anti-inflammatory effect of PDE4 inhibitors is associated with the enhancement of cAMP-dependent PKA-CREB signaling (Jin et al., 2005; Serezani et al., 2008; Milakovic and Gooderham, 2021). As shown in Figure 4A, salmeterol increased phosphorylation of CREB in HaCaT cells exposed to TNF-α, which was attenuated by the PKA inhibitor, H89. The effect of salmeterol on PKA-CREB signaling was similar to that of roflumilast (Figure 4B). Taken together, these results indicated that salmeterol and roflumilast activated the cAMP/PKA/CREB pathway. By FRET assay (Figure 4C), we also determined that salmeterol and roflumilast activated the cAMP/PKA signaling pathway in a dose-dependent manner. Notably, the combination of salmeterol and roflumilast synergistically increased PKA activity.

To analyze the potential therapeutic effect of topical administration of salmeterol on psoriasis in vivo, we constructed an IMQ-induced mouse model that effectively replicated the characteristics of psoriatic lesions in humans (van der Fits et al., 2009) to investigate whether salmeterol in combination with roflumilast has a synergistic effect. On the fifth day, the IMQ-treated mice showed thickened, erythematous, and scaly back skin compared to the vehicle group mice, reflecting psoriasis-like skin conditions (Figure 5A), and the IMQ-treated mice showed significant increases in lesion severity scores (Figures 5B–D). Histopathological analysis of back skin with hematoxylin and eosin staining showed that IMQ induced epidermal thickening (Figure 5F). The above results demonstrated that we successfully constructed an IMQ-induced psoriasis-like mouse model. The phenotypical changes of the back skin (Figure 5A) and lesion severity score (Figures 5B–D) showed that topical treatment with salmeterol, roflumilast, or combination of salmeterol and roflumilast significantly attenuated IMQ-induced psoriatic traits and reduced IMQ-induced epidermal thickening (Figure 5F). However, there was no significant difference with the combination treatment compared to treatment with either salmeterol or roflumilast alone.

Based on the cumulative score (Figure 5E) and the improvement of epidermal thickness (Figure 5F), mouse skin tissues obtained from the Sal (10 μg/kg) group, Rof (0.03 mg/kg) group or Sal (10 μg/kg) + Rof (0.03 mg/kg) group were used to analyze cell apoptosis, cell proliferation, and cAMP levels. The TUNEL assay showed that compared to the IMQ-treated group, salmeterol treatment and roflumilast treatment significantly promoted cell apoptosis in the epidermis (Figure 6A) and increased the Bax/Bcl2 ratio (Figure 6C). The combination treatment also increased the Bax/Bcl2 ratio but did not result in a statistically significant increase in TUNEL-positive cells (Figure 6A). As assessed by immunohistochemistry staining, IMQ treatment increased the expression of Ki67 in the epidermis, whereas salmeterol, roflumilast, and the combination treatment inhibited the IMQ-induced Ki67 expression (Figure 6B). Consistently, IMQ treatment significantly increased the expression of the epidermal proliferation marker, keratin 17 (K17), whereas salmeterol, roflumilast, and the combination treatment inhibited the IMQ-induced K17 expression (Figure 6C). Consistent with the above results, there was no significant difference in the combination treatment compared to treatment with salmeterol or roflumilast alone.

We further determined the effect of salmeterol on PKA/CREB activation in IMQ-induced psoriasis-like skin lesions by Western blot analysis. CREB phosphorylation was inhibited by IMQ treatment, while salmeterol, roflumilast, and the combination treatment partly reversed the IMQ-inhibited CREB phosphorylation. However, the effect of the combination treatment on CREB phosphorylation was not significantly different compared to treatment with salmeterol or roflumilast alone (Figure 6C). Next, we explored whether the level of CREB phosphorylation is associated with cAMP levels. As shown in Figure 6D, the cAMP levels were slightly decreased in skin tissue after IMQ treatment, but the topical treatments showed higher cAMP levels with no synergistic effect in the combination treatment group. These results partly explained why there was no synergistic effect of the combination treatment in vivo. Thus, these results suggested that salmeterol ameliorates IMQ-induced psoriasis-like skin lesions by inducing apoptosis and negatively regulating proliferation via activating CREB.

ERK pathway is a well-known regulator of cell growth, proliferation, and apoptosis. Several studies suggest the association between activation of ERK signaling and psoriasis development (Takahashi et al., 2002; Johansen et al., 2005). The inhibition of ERK relieved skin symptoms in a mouse imiquimod-induced model (Wu et al., 2018; Huang X et al., 2019). A latest study reported that β₂-AR agonists promote ERK dephosphorylation in human airway epithelial cells by cAMP-PKA signaling (Hamed et al., 2021), although some earlier studies showed that PKA mediates the switch of coupling of β₂-AR from Gs to Gi and initiate the activation of ERK pathway (Daaka et al., 1997; Zamah et al., 2002; Smith et al., 2021). Thus, we examined ERK phosphorylation in psoriasis-like skin lesions. As shown in Figure 7A, both salmeterol and roflumilast alone group showed ERK dephosphorylation compared to the IMQ-treated group, but the combination treatment promoted ERK phosphorylation compared to salmeterol or roflumilast treatment alone. We further found that the synergistic effects of anti-proliferation and induction of apoptosis appeared in the presence of ERK inhibitor U0126 (Figures 7B–D).

As shown in Figure 7E, both salmeterol and roflumilast alone group slightly increased the level of PDE4D phosphorylation, which is consistent with previous studies that PDE4s are activated by PKA phosphorylation, providing a negative feedback mechanism by which cAMP regulates its own level (Sette and Conti, 1996; Rochais et al., 2004). In addition, the combination treatment significantly increased the level of PDE4D phosphorylation (Figure 7E), which could partly explain that there was no synergistic effect on increasing cAMP level in the combination treatment group (Figure 6D). Of note, ERK inhibitor blocked the combination treatment-increased PDE4D phosphorylation (Figure 7E), which is consistent with previous studies that PDE4D activity is regulated by extracellular-signal regulated kinases (Wang et al., 2017; Xu et al., 2020).