Dual targeting of mTOR/IL-17A and autophagy by fisetin alleviates psoriasis-like skin inflammation

Abstract

Psoriasis is a chronic autoimmune inflammatory skin disorder characterized by epidermal hyperplasia and aberrant immune response. In addition to aberrant cytokine production, psoriasis is associated with activation of the Akt/mTOR pathway. mTOR/S6K1 regulates T-lymphocyte activation and migration, keratinocytes proliferation and is upregulated in psoriatic lesions. Several drugs that target Th1/Th17 cytokines or their receptors have been approved for treating psoriasis in humans with variable results necessitating improved therapies. Fisetin, a natural dietary polyphenol with anti-oxidant and anti-proliferative properties, covalently binds mTOR/S6K1. The effects of fisetin on psoriasis and its underlying mechanisms have not been clearly defined. Here, we evaluated the immunomodulatory effects of fisetin on Th1/Th17-cytokine-activated adult human epidermal keratinocytes (HEKa) and anti-CD3/CD28-stimulated inflammatory CD4⁺ T cells and compared these activities with those of rapamycin (an mTOR inhibitor). Transcriptomic analysis of HEKa revealed 12,713 differentially expressed genes (DEGs) in the fisetin-treated group compared to 7,374 DEGs in the rapamycin-treated group, both individually compared to a cytokine treated group. Gene ontology analysis revealed enriched functional groups related to PI3K/Akt/mTOR signaling pathways, psoriasis, and epidermal development. Using in silico molecular modeling, we observed a high binding affinity of fisetin to IL-17A. In vitro, fisetin significantly inhibited mTOR activity, increased the expression of autophagy markers LC3A/B and Atg5 in HEKa cells and suppressed the secretion of IL-17A by activated CD4⁺ T lymphocytes or T lymphocytes co-cultured with HEKa. Topical administration of fisetin in an imiquimod (IMQ)-induced mouse psoriasis model exhibited a better effect than rapamycin in reducing psoriasis-like inflammation and Akt/mTOR phosphorylation and promoting keratinocyte differentiation and autophagy in mice skin lesions. Fisetin also significantly inhibited T-lymphocytes and F4/80⁺ macrophage infiltration into skin. We conclude that fisetin potently inhibits IL-17A and the Akt/mTOR pathway and promotes keratinocyte differentiation and autophagy to alleviate IMQ-induced psoriasis-like disease in mice. Altogether, our findings suggest fisetin as a potential treatment for psoriasis and possibly other inflammatory skin diseases.

Keywords: Akt/mTOR and IL-17A; autophagy; fisetin; keratinocytes RNA-sequencing; psoriasis; psoriasis-like skin inflammation; rapamycin; topical administration.

Figure 1

Fisetin treatment modulates several targets and signaling pathways in IL-17A/TNFα-stimulated primary keratinocytes. (A) Venn diagram showing the genes that were differentially regulated with >2-fold increase (shown as upward arrow) or decrease (downward arrow) from each comparison across the different groups of treatment and activation with cytokines in keratinocytes. (B) Metascape analysis at systems level of transcriptome profiles of 3 groups in circus plot: Activated vs Control (green), Activated + Fisetin vs Activated (red) and Activated + Rapamycin vs Activated (blue) in vitro. The plot shows how 1.5-fold changes differentially expressed genes are overlapped: Each arc represents as individual gene. Dark orange color represents the genes shared by multiple groups which forms purple link. On the contrary, the light orange arc represents unique gene to that list from the group. Light blue color represents different genes but shared ontology terms. There is a trend to see more common functional overlap that means subset of genes of the same biological process. (C) Differential enrichment analysis illustrates the transcriptomic enrichment analysis of genes involved in different important ontology terms (GO/KEGG) in cytokine-activated keratinocytes compared with control represented by circles. Enrichment study of GO/KEGG and hall mark gene sets. The overlap between the input gene lists increases with the thickness of pink linkages. The detailed enrichment study is shown in Figure S1 online. (D) All input gene lists were also merged into one list and resulted in a PPI (Protein-Protein Interaction) network. Network nodes are displayed as pies. Color code for pie sector represents a gene list and is consistent with the colors used for Activated vs Control (green), Activated + Fisetin vs Activated (red) and Activated + Rapamycin vs Activated (blue). To identify neighborhoods where proteins are densely connected, the MCODE algorithm was used to this network with biological meaning, which shows that a cytokine mediated signaling, cellular response to cytokine stimulus and autophagosome organization are related to transcriptome expression in the treatment group differentially. (E) The heatmap shows differential expression of mTOR pathway components and autophagy related transcriptome in each replicate of 3 groups. Overlaps at the gene level, where purple curves link identical genes in three groups, including the shared term level, where blue curves link genes that belong to the same enriched ontology term are shown in a circular 3 plot (C).

Figure 2

Fisetin suppresses mTOR kinase activity and reduces phosphorylated mTOR pathway components in cytokine-activated human epidermal keratinocytes. (A) ELISA analysis of mTOR kinase activity in control, fisetin-treated, or cytokine [rhIL-6/IL-22 (15ng/ml each)]-treated primary keratinocytes lysates. (B) Western blot analysis of differential protein expression levels of p-Akt^Ser473, p-AMPK (Thr¹⁷²)?, p-mTOR^Ser2448, and p-raptor (Ser⁷⁹²) in fisetin (15µM) treated, rapamycin (100nM) or (IL-17A+TNFα)-activated HEKa compared with untreated control cells. (C) Graphs of the semi-quantified intensity of bands for p-Akt, p-AMPK, p-mTOR, and p-raptor. Bar represents means ± SD of three separate experiments each performed in triplicate. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 vs. control for IL-6/IL-22-treated only or vs. IL-22 or IL-17A+TNFα for fisetin-treated cells. ns, not significant.

Figure 3

Fisetin induces the expression of autophagy markers in cytokine-stimulated primary epidermal keratinocytes. (A) Representative photographs of the immunofluorescence staining for actin (red) and LC3A/B (green) and the DAPI staining (blue; for nucleus) in fisetin (15 µM), rapamycin (100 nM) or IL-17A+TNFa)-treated HEKa compared with inactivated control keratinocytes/HEKa. (B) Western blot analysis for differential protein levels of autophagy markers Atg5, Beclin1 and SQSTM1 in fisetin- or rapamycin-treated primary keratinocytes activated with cytokines (IL-17A+TNFα) in vitro. (C) Semi-quantification of the bands for LC3A/B, Atg5, and Beclin1 and SQSTM1. Bar is means ± SD of three separate experiments. *P <0.05, **P < 0.01, and ***P < 0.001 vs. control for activated (IL-17A+TNFα) only and activated with keratinocytes treated with fisetin or rapamycin. When contrasting the treatment groups for fisetin and rapamycin, the # is used. * is used when we compared with the activated group. ns, not significant.

Figure 4

Fisetin reduces IL-17A production by activated CD4⁺ T lymphocyte alone and when co-cultured with keratinocytes in vitro. (A) Binding affinity of fisetin to IL-17A, predicted using Autodock vina and interaction of fisetin in binding pocket of IL-17A shown in stick and mesh model. (B) This interaction involves hydrophilic and hydrophobic interactions with different residues shown in 2D layout. ELISA assay for the levels of IL-17A secretion by anti-CD3/anti-CD28 activated purified human CD4+ T lymphocytes under different conditions: (C) anti-CD3/anti-CD28 activated CD4+ T lymphocytes cultured alone and then primed with IL-1β and IL-23 to enhance IL-17A production in the presence or absence of fisetin. (D) anti-CD3/anti-CD28 activated CD4+ T lymphocytes co-cultured with NHEKs in the presence or absence of fisetin (10 µM) or Vit-D3 (0.1 µM). Cells were cultured for 48 h in the indicated conditions and secreted IL-17A in the condition culture media was assayed by ELISA. Bar represents Means ± SE resulted from at least three separate experiments and paired t test was used to compare values between treatments. **P <0.01 and ***P <0.001 (–). indicates non-activated or treated cells. ns, not significant.

Figure 5

Fisetin improves clinical scores, alleviates epidermal hyperplasia, and reduces Ki67 in epidermis of IMQ-induced skin lesion in C57BLC6 mice. (A) Representative preclinical photographs of ears, and back skin of IMQ-treated C57BLC6 mice treated topically once daily for 10 days with fisetin (50 mg/kg body weight) or rapamycin (25 mg/kg body weight) compared with vehicle (vanicream) control mice. Psoriasis Area and Severity Index (PASI) was used for scoring. Scale bar = 100 µm. Schematic overview of Timeline of IMQ studies. (B: i-iii) Four blinded evaluators scored ears, and skin for erythema, scaling, and ear swelling (thickness). Scoring ranged from 0 to 4, increments of 1, with 0 indicating normal phenotype and 4 indicating a severe psoriasis-like phenotype. (C) Histological H&E-stained ear and skin sections from mice treated with vanicream control (i.e., no Aldara reference), IMQ, fisetin+IMQ, or rapamycin+IMQ. (B; iv) Microscopic quantification of average interfollicular epidermal thickness of vanicream control, IMQ, fisetin_IMQ, or rapamycin+IMQ treated mouse skin sections. (D) Immunohistochemical staining of skin sections for Ki67. (E) Quantified numbers of Ki67-positive cells in basal and suprabasal epidermis from mice with different treatments. Scale bar = 100 µm. The epidermal thickness scoring and Ki67-positive cell average count was calculated from each group (n=6). Four areas per section of a sample were sampled and analyzed. Means ± SE are shown, and paired t test was used to compare values between different treatments. *P <0.05, **P <0.01 and ***P <0.001. (-) indicates control group.

Figure 6

Fisetin inhibits the phosphorylation of mTOR, 4EBP1, Akt and Stat3 and increased the phosphorylation of eIF2α in IMQ-treated skin of mice. (A) Immunohistochemical staining of p-mTOR (Ser²⁴⁴⁸), p-4EBP1(Thr^37/46), p-eIF2α (Ser⁵¹), p-Akt (Ser⁴⁷³) and p-Stat3 (Tyr⁷⁰⁵), in fisetin- or vehicle-treated in IMQ-induced mouse dorsal skin sections. Scale bar = 100 µm. (B) Quantified intensity of p-mTOR, p-4EBP1 and p-eIF2α, p-Akt and p-Stat3 per 20× microscopic field view. The mean intensities and quantifications are shown in bar graphs as mean ± SE, and paired t test was used to compare values between the different treatments. These were calculated from each group (n=6). *P <0.05; **P <0.01 and ***P <0.001. (-) indicates control group.

Figure 7

Fisetin normalizes and enhances the expression of differentiation marker loricrin, and reduces the expression of IL-17 and the infiltration of F4/80+ and pan-CD45+ immune cells in IMQ mice skin lesions. (A) Immunofluorescence staining of loricrin (green) and macrophage cell activation marker F4/80 (red) in vehicle-, IMQ-, and fisetin+IMQ or and rapamycin+IMQ-treated mouse skin. (B) Quantification of the fluorescent intensity of loricrin and F4/80+ cells per 20× field view. (C) Immunofluorescence staining of IL-17 (green) and pan CD45+ T lymphocytes (red) in vehicle-, IMQ, fisetin+IMQ-treated, and rapamycin+IMQ-treated mouse skin. DNA in nuclei was counterstained with DAPI (blue). The merged image shows colocalization of IL-17 with some CD45+ cells, which delineates IL-17⁺CD45⁺ T lymphocytes especially in IMQ-treated mice. (D) Quantification of the fluorescent intensity of IL-17 (green) and pan CD45+ T lymphocytes (red) per 20× field view. Shown in bar graph are mean± SD, and paired t test was used to compare values between different treatments, average reading from each group (n=6). *P <0.05 and **P <0.01. (-) indicates control group and ns indicates non-significant.

Figure 8

Fisetin induces the expression of autophagy marker LC3A/B in IMQ-induced skin lesion in mice. (A) LC3A/B (green) immunofluorescence staining and DAPI (blue) staining (for nucleus) in the skin tissue sections of vehicle-, IMQ-, fisetin+IMQ- and rapamycin+IMQ-treated mice. Red arrow heads in the fisetin+IMQ or rapamycin+IMQ treatment indicates green punctate staining of LC3A/B, which is absent in the control or IMQ treatment. Scale bar = 10 µm. (B) Quantification of fluorescent intensity of LC3A/B-positive puntata per 60x field view. Shown in bar graph are mean ± SD, and paired t test was used to compare values between different treatments, number of mice per group n=6. *P <0.05 and ns is non-significant.