Dihydroartemisinin ameliorates psoriatic skin inflammation and its relapse by diminishing CD8+ T-cell memory in wild-type and humanized mice

Abstract

Conventional immunosuppressants cause side effects and do not prevent the recurrence of autoimmune diseases. Moreover, they may not inhibit autoimmunity mediated by pathogenic memory T-cells. Dihydroartemisinin (DHA) has been shown to regulate autoimmunity. However, it remains unknown whether DHA impacts psoriasis and its recurrence. The objective of this study was to determine therapeutic effects of DHA on psoriasis and its relapse as well as its underlying mechanisms. Methods: We established animal models of imiquimod (IMQ)-induced psoriasis-like wild-type mice and humanized NSG mice receiving lesional human skin from patients with psoriasis. Many immunoassays, including immunohistochemistry, flow cytometry, quantitative RT-PCR and Western blotting, were performed. Results: We found that DHA not only ameliorated acute skin lesion of psoriatic mice, but also alleviated its recurrence by diminishing CD8⁺ central memory T (T_CM) and CD8⁺ resident memory T (T_RM) cells. It attenuated epidermal pathology and T-cell infiltration in the skin of IMQ-induced psoriatic mice while suppressing expression of IL-15, IL-17 and other proinflammatory cytokines in the skin. Surprisingly, DHA reduced the frequency and number of CD8⁺, but not CD4⁺, subset of CD44^highCD62L^high T_CM in psoriatic mice, whereas methotrexate (MTX) lowered CD4⁺, but not CD8⁺, T_CM frequency and number. Indeed, DHA, but not MTX, downregulated eomesodermin (EOMES) and BCL-6 expression in CD8⁺ T-cells. Furthermore, DHA, but not MTX, reduced the presence of CD8⁺CLA⁺, CD8⁺CD69⁺ or CD8⁺CD103⁺ T_RM cells in mouse skin. Interestingly, treatment with DHA, but not MTX, during the first onset of psoriasis largely prevented psoriasis relapse induced by low doses of IMQ two weeks later. Administration of recombinant IL-15 or CD8⁺, but not CD4⁺, T_CM cells resulted in complete recurrence of psoriasis in mice previously treated with DHA. Finally, we demonstrated that DHA alleviated psoriatic human skin lesions in humanized NSG mice grafted with lesional skin from psoriatic patients while reducing human CD8⁺ T_CM and CD103⁺ T_RM cells in humanized mice. Conclusion: We have provided the first evidence that DHA is advantageous over MTX in preventing psoriasis relapse by reducing memory CD8⁺ T-cells.

Keywords: Dihydroartemisinin; Immunosuppression; Memory T cells; Psoriasis; Relapse.

Figure 1

DHA reduces the PASI scores and ameliorates the skin lesion of IMQ-induced psoriasis-like mice. (A) Shown is a timeline of various treatments. BALB/c mice were intraperitoneally administered with DHA (DHA-L: 25 mg/kg; DHA-H: 50 mg/kg) or MTX (1.5 mg/kg) for 7 consecutive days during the topical application of IMQ on the dorsal skin. All mice were sacrificed on day 7. (B) The representative photos of dorsal skin in mice on day 7. (C) The PASI scores of the skin lesion in psoriatic mice on days 3, 5 and 7. (D) The body weight changes of the mice on day 7. All data are presented as the mean ± SD from three separate experiments (n = 6 mice/group, *p < 0.05 and **p < 0.01).

Figure 2

DHA inhibits epidermal hyperplasia and T cell infiltration in the skin of psoriatic mice. (A) H&E and IHC staining of Ki67 or CD3 on psoriatic mouse skin at Day 7. Representative images of skin sections are presented (Scale bar, 100 µm). (B-C) Acanthosis and papillomatosis index of murine skin based on H&E staining. (D) Quantitation of Ki67⁺ cells in epidermis. For counting Ki67⁺ cells, six high-power fields in each section of each sample were calculated. (E) The integrated optical density (IOD) of CD3⁺ T cells in the skin was calculated using ImagePro Plus. Values were expressed as fold change relative to control group that was set as 1.0. All data are presented as the mean ± SD from three separate experiments (n = 6 mice/group, *p < 0.05 and **p < 0.01).

Figure 3

DHA inhibits mRNA expression of pro-inflammatory cytokines and IL-15 as well as NFκB signaling in psoriatic mouse skin. The mRNA levels of IL-17A (A), IL-6 (B), IFN-γ (C), TNF-α (D) and IL-15 (E) in the skin were determined by RT-PCR on Day 7. (F) The representative western blots of p-IKKα, IKKα, p-NFκB p65 and NFκB p65 expression in the skin of mice on Day 7. (G) Quantification of the relative expression of p-IKKα/IKKα and p-NFκB p65/NFκB p65. GAPDH expression was used to normalize data. Values were expressed as fold changes relative to control group that was set as 1.0. Data of column graphs are presented as the mean ± SD from three separate experiments (n = 4-6 mice/group, *p < 0.05 and **p < 0.01).

Figure 4

DHA ameliorates the recurrence of skin lesion of IMQ-induced psoriatic mice and reduces skin-infiltrating CD8+ T cells. (A) The timeline of the treatment. BALB/c mice were intraperitoneally administered with DHA (25 mg/kg or 50 mg/kg) or MTX for 7 consecutive days during the first application of IMQ (62.5 mg). After 14 days of recovery, all mice were re-challenged with secondary application of smaller doses of IMQ (20.8 mg) for 7 more days. (B) The representative photos of dorsal skin lesion of mice on day 24 (peak time for the lesion). (C) The PASI scores of the skin lesion in the psoriatic mice on days 21, 24 and 27. (D) The number of T cells, including total CD4⁺, CD8⁺ T cells and CD8⁺ T_RM cells, in the skin was measured via FACS analysis on day 27. All data are presented as the mean ± SD from three separate experiments (n = 6 mice/group, *p < 0.05 and **p < 0.01).

Figure 5

DHA reduces the frequency of CD4⁺ effector T (T_eff), but not CD4⁺ central memory T (T_CM), cells in IMQ-induced psoriasis-like mice. Lymph node and spleen cells were isolated from psoriatic mice on day 7 following various treatments. The percentages of CD4⁺ T_eff cells (CD4⁺CD44⁺CD62L^-) and CD4⁺ T_CM cells (CD4⁺CD44⁺CD62L⁺) in lymph nodes and spleen of the psoriatic mice were determined via FACS. Shown on the top are representative dot plots of CD4⁺ T_eff and CD4⁺ T_CM cell populations in lymph nodes and spleen of the mice (A). Column graphs display the percentages (B-C) or absolute numbers (D-E) of CD4⁺ Teff and CD4+ T_CM cells in lymph nodes and spleen of the mice. Data of column graphs are presented as the mean ± SD from three separate experiments (n = 4-6 mice/group, *p < 0.05 and **p < 0.01).

Figure 6

DHA decreases the frequency of CD8⁺ effector T (T_eff) cells and CD8⁺ central memory T (T_CM) cells in IMQ-induced psoriasis-like mice. Lymph node and spleen cells were isolated from psoriatic mice on day 7 following various treatments. The percentages of CD8⁺ T_eff cells (CD8⁺CD44⁺CD62L^-) and CD8⁺ T_CM cells (CD8⁺CD44⁺CD62L⁺) in lymph nodes and spleen of the psoriatic mice were determined via FACS. (A) Shown are representative dot plots of CD8⁺ T_eff and CD8⁺ T_CM cell populations in lymph nodes and spleen. The percentages (B-C) or absolute numbers (D-E) of CD8⁺ T_eff and CD8⁺ T_CM cells in lymph nodes and spleen of mice were also determined. (F) On day 24, PASI scores of the skin lesion in psoriatic mice treated with DHA (50 mg/kg) plus administration of CD4⁺ T_CM cells, CD8⁺ T_CM cells or rIL-15 were observed. Briefly, mice received FACS-sorted CD4⁺ or CD8⁺ T_CM cells on day 20, the beginning day of IMQ re-challenging, while others were administered with rIL-15 on days 0, 3 and 6 during initial IMQ priming and DHA treatment. (G) The percentages of CD8⁺ T_CM cells in lymph nodes of psoriatic mice treated without or with DHA (50 mg/kg) plus rIL-15. All data of column graphs are presented as the mean ± SD from three separate experiments (n = 5-6 mice/group, *p < 0.05 and **p < 0.01).

Figure 7

DHA inhibits EOMES and BCL-6 expression in CD8⁺ T cells of IMQ-induced psoriasis-like mice. Lymph node and spleen cells were isolated from psoriatic mice after 7 days of initial treatments. Expression of EOMES and BCL-6 by CD8⁺ T cells in lymph nodes and spleen of mice was determined by FACS following intracellular staining. (A) Representative histogram graphs of EOMES expression in CD8⁺ T cells. (B-C) Column graphs show the MFI of EOMES in CD8⁺ T cells from lymph nodes and spleen of the mice. (D) Representative histogram graphs of BCL-6 expression in CD8⁺ T cells. (E-F) Column graphs show the MFI of BCL-6 in CD8⁺ T cells from lymph nodes and spleen of the mice. Dotted line represents the corresponding isotype controls. Data are presented as the mean ± SD from three separate experiments (n = 4-6 mice/group, *p < 0.05 and **p < 0.01).

Figure 8

DHA decreases the frequency of CD8⁺CLA⁺, CD8⁺CD69⁺ and CD8⁺CD103⁺ T_RM cells in the skin of IMQ-induced psoriatic mice. Skin-infiltrating cells were isolated from psoriatic mice on day 7. The percentages of CLA⁺, CD69⁺ or CD103⁺ cells within CD8⁺ T cell subset were determined via FACS analysis. (A) Representative histograms of CLA⁺, CD69⁺ or CD103⁺ cells after gating on CD8⁺ T cell subset. (B) Column graphs show the percentages of CLA⁺, CD69⁺ or CD103⁺ cells within CD8⁺ T cell population. (C) Shown are the absolute numbers of CD8⁺CLA⁺, CD8⁺CD69⁺ and CD8⁺CD103⁺ T_RM cells as well as total CD4⁺/CD8⁺ T cells. Dotted line represents the corresponding isotype controls. Data are presented as the mean ± SD from three separate experiments (n = 4-6 mice/group, *p < 0.05 and **p < 0.01).

Figure 9

DHA alleviates human skin lesion in humanized mice grafted with lesional skin from psoriatic patients. The lesional or non-lesional skin from psoriatic patients was transplanted to the dorsal flank area of NSG mice that received PBMCs from the same patient. Mice were then treated with DHA at doses of 50 mg/kg/day. (A) Shown are the representatives of lesional and non-lesional human skin grafts in NSG mice with or without DHA treatment 8 days after transplantation. (B-C) Human CD4⁺ or CD8⁺ T_CM cells (CD45RO⁺CCR7⁺) from draining LNs of humanized NSG mice that received both psoriatic human skin and PBMCs were enumerated by FACS on day 8. (D) Representative images of H&E and CD3 stainings of the grafted lesional and non-lesional human skin as well as immunofluorescence staining of CD103 on day 8 (Scale bar, 50 or 100 µm). (E-F) Acanthosis and papillomatosis index of skin based on H&E staining above. (G) Fold changes in CD3 expression (IOD) in the grafted lesional and non-lesional skin. (H) The Fluorescence intensity of CD103 in the skin grafts based on immunofluorescence staining. Data are presented as the mean ± SD of four recipient mice (n = 4 mice/group).