BACKGROUND: Psoriasis is an immune-mediated chronic inflammatory skin disease. Current research suggests that the long-term persistence and recurrence of psoriasis are closely related to the feedback loop formed between keratinocytes and immune cells, especially in Th 17 or DC cells expressing CCR6. CCL20 is the ligand of CCR6. Therefore, drugs that block the expression of CCL20 or CCR6 may have a certain therapeutic effect on psoriasis. Glycyrrhetinic acid (GA) is the main active ingredient of the plant drug licorice and is often used to treat autoimmune diseases, including psoriasis. However, its mechanism of action is still unclear. METHODS: Psoriasis like skin lesion model was established by continuously applying imiquimod on the back skin of normal mice and CCR6-/- mice for 7 days. The therapeutic and preventive effects of glycyrrhetinic acid (GA) on the model were observed and compared. The severity of skin injury is estimated through clinical PASI scores and histopathological examination. qRT-PCR and multiple cytoline assay were explored to detect the expression levels of cytokines in animal dorsal skin lesions and keratinocyte line HaCaT cells, respectively. The dermis and epidermis of the mouse back were separated for the detection of CCL20 expression. Transcription factor assay was applied to screen, and luciferase activity assay to validate transcription factors regulated by GA. Technology of surface plasmon laser resonance with LC-MS (SPR-MS), molecular docking, and enzyme activity assay were used to identified the target proteins for GA. Finally, we synthesized different derivatives of 18beta-GA and compared their effects, as well as glycyrrhetinic acid (GL), on the skin lesion of imiquimod-induced mice to evaluate the active groups of 18beta-GA. RESULTS: 18β-glycyrrhetinic acid (GA) improved IMQ-induced psoriatic lesions, and could specifically reduce the chemokine CCL20 level of the epidermis in lesion area, especially in therapeutic administration manner. The process was mainly regulated by transcription factor ATF2 in the keratinocytes. In addition, GUSB was identified as the primary target of 18βGA. Our findings indicated that the subject on molecular target research of glycyrrhizin should be glycyrrhetinic acid (GA) instead of glycyrrhizic acid (GL), because GL showed little activity in vitro or in vivo. Apart from that, α, β, -unsaturated carbonyl in C11/12 positions was crucial or unchangeable to its activity of 18βGA, while proper modification of C3 or C30 position of 18βGA may vastly increase its activity. CONCLUSION: Our research indicates that 18βGA exerted its anti-psoriasis effect mainly by suppressing ATF2 and downstream molecule CCL20 predominately through α, β, -unsaturated carbonyl at C11/12 position binding to GUSB in the keratinocytes, and then broke the feedback loop between keratinocytes and CCR6-expressing immune cells. GA has more advantages than GL in the external treatment of psoriasis. A highlight of this study is to investigate the influence of special active groups on the pharmacological action of a natural product, inspired by the molecular docking result.