Objective:To investigate changes in circadian gene cryptochrome 2 (CRY2) expression in mouse models of psoriasis and HaCaT cells, and to explore underlying mechanisms.Methods:Imiquimod-induced mouse model experiment: 12 C57BL/6 female mice were randomly and equally divided into imiquimod group receiving topical imiquimod treatment for 5 consecutive days and control group receiving no treatment; these mice were sacrificed on day 6, skin tissues were resected from the back of mice, and immunofluorescence staining was performed to determine the CRY2 expression in the epidermis. HaCaT cell transfection experiment: HaCaT cells with small interfering RNA (siRNA) -mediated knockdown of CRY2 served as siRNA-CRY2 group, and siRNA-NC group as control group; 5-ethynyl-2′-deoxyuridine (EdU) staining was performed to evaluate the proliferative activity of the HaCaT cells, real-time fluorescence-based quantitative PCR (qPCR) to determine the mRNA expression of chemokines in the HaCaT cells, and Western blot analysis to determine phosphorylation levels of extracellular signal-regulated kinase 1/2 (ERK1/2) . Tumor necrosis factor-α (TNF-α) -stimulated animal and cell experiments: 12 C57BL/6 female mice were randomly and equally divided into TNF-α group subcutaneously injected with TNF-α solution in the ear for 6 days, and phosphate buffered saline (PBS) group subcutaneously injected with the same amount of PBS; the mice were sacrificed on day 7, skin tissues were resected from the ear of mice, and immunofluorescence staining was conducted to determine the CRY2 expression in the epidermis; CRY2-knockdown HaCaT cells stimulated with 50 ng/ml TNF-α for 12 hours served as siRNA-CRY2 + TNF-α group, and siRNA-NC + TNF-α group as control group; qPCR was performed to determine the mRNA expression of chemokines in HaCaT cells in the above groups. Statistical analysis was carried out by using two-independent-sample t test. Results:Immunofluorescence staining showed that the CRY2 protein expression was significantly lower in the mouse dorsal epidermis in the imiquimod group (0.94 ± 0.23) than in the control group (2.30 ± 0.25, t = 3.99, P = 0.016) . Compared with the siRNA-NC group, the siRNA-CRY2 group showed significantly increased proportions of EdU-positive cells (48.13% ± 10.97% vs. 38.23% ± 0.81%, t = 5.00, P = 0.007) , mRNA expression levels of chemokines CXCL1 and CXCL8, as well as significantly increased phosphorylated (p) -ERK1/2 protein expression levels (all P < 0.05) , while there were no significant differences in the CCL20 mRNA expression or ERK1/2 protein expression between the two groups (both P > 0.05) . Immunofluorescence staining showed significantly decreased CRY2 protein expression level in the mouse ear epidermis in the TNF-α group (0.37 ± 0.34) compared with the PBS group (2.04 ± 0.17, t = 4.38, P = 0.012) ; the relative mRNA expression levels of chemokines CXCL1, CXCL8, and CCL20 in HaCaT cells were significantly higher in the siRNA-CRY2 + TNF-α group than in the siRNA-NC + TNF-α group (all P < 0.05) . Conclusion:CRY2 was markedly underexpressed in psoriasis, which might promote the proliferation of keratinocytes and expression of chemokines CXCL1, CXCL8 and CCL20, and TNF-α might be an upstream cytokine that could downregulate CRY2 expression.