Abstract: 【Abstract】 Objective To evaluate the effect of metformin on ultraviolet A （UVA）-induced photoaging of an immortalized human keratinocytes cell line （HaCaT）, and to explore its potential mechanisms. Methods Cell counting kit 8 （CCK8） assay was performed to evaluate the effect of metformin at different concentrations （0 - 100 mmol/L） on the viability of HaCaT cells, and 10 mmol/L metformin was selected for subsequent experiments. Cultured HaCaT cells were divided into a blank control group （conventional culture）, a metformin group （treated with culture medium containing 10 mmol/L metformin）, a UVA irradiation group （conventional culture for 24 hours followed by 10 J/cm2 UVA irradiation） and a metformin + UVA group （treated with culture medium containing 10 mmol/L metformin for 24 hours followed by 10 J/cm2 UVA irradiation）; UVA irradiation was performed at a dose of 10 J/cm2 once a day for 3 consecutive days. After 4-day treatment, cells were collected, the β-galactosidase assay was performed to determine the proportion of senescent cells in each group, 2′, 7′-dichlorodihydrofluorescein diacetate assay to detect levels of intracellular reactive oxygen species （ROS）, and the comet assay to detect DNA damage levels. Additionally, some HaCaT cells were divided into the blank control group, metformin group, 1.25 μmol/L dorsomorphin （an adenosine monophosphate-activated protein kinase ［AMPK］ inhibitor） + metformin group, and 2.5 μmol/L dorsomorphin + metformin group, and cells in the latter two groups were treated with 1.25 and 2.5 μmol/L dorsomorphin respectively for 2 hours, followed by the treatment with 10 mmol/L metformin for 24 hours. Western blot analysis was performed to determine the cellular localization and phosphorylation levels of nuclear factor-erythroid 2-related factor 2 （Nrf2）. By using the small-interfering RNA （siRNA）-mediated silencing method, siRNA-Nrf2 was transfected into HaCaT cells to knock down Nrf2 expression （siRNA-Nrf2 group）; 2.5 μmol/L dorsomorphin-treated HaCaT cells or Nrf2-knockdown HaCaT cells were treated with metformin and UVA irradiation （dorsomorphin + metformin + UVA group, siRNA-Nrf2 + metformin + UVA group, respectively）, and the proportions of senescent cells were further calculated in each group. Statistical analysis was carried out by using one-way analysis of variance and two-way analysis of variance, and least significant difference （LSD）-t test was used for multiple comparisons. Results Treatment with different concentrations of metformin for 24 hours could affect the viability of HaCaT cells to varying degrees （F = 5 206.31, P < 0.001）; there were no significant differences in the relative survival rates of HaCaT cells between the 10 - 20 mmol/L metformin groups and the control group （0 mmol/L metformin group, all P > 0.05）, while the relative cell survival rates were significantly lower in the 25 - 100 mmol/L metformin groups than in the control group （all P < 0.05）. After UVA irradiation, HaCaT cells shrank significantly and became narrow and elongated, and the intercellular spaces increased; the relative cell survival rate was significantly lower in the UVA irradiation group （76.13% ± 1.03%） than in the blank control group （100.00% ± 1.24%, LSD-t = 14.86, P < 0.001）, but significantly higher in the metformin + UVA group （106.69% ± 2.45%） than in the UVA irradiation group （LSD-t = 11.55, P < 0.001）. Moreover, the UVA irradiation group showed significantly increased proportions of senescent cells （45.14% ± 4.98%）, intracellular ROS levels （144.61% ± 4.91%）, and percentages of DNA in the tail （75.33% ± 1.77%） compared with the blank control group （23.84% ± 1.89%, 55.49% ± 1.57%, 1.88% ± 0.29%, respectively, all P < 0.001）, while the metformin + UVA group showed significantly decreased proportions of senescent cells （24.26% ± 1.34%）, intracellular ROS levels （58.62% ± 2.17%）, percentages of DNA in the tail （15.83% ± 1.23%） compared with the UVA irradiation group （all P < 0.001）. Western blot analysis showed that the Nrf2 expression in the cytoplasm was lower in the 10 mmol/L metformin group than in the blank control group, while the phosphorylated Nrf2 expression in the nuclei was higher in the 10 mmol/L metformin group than in the blank control group, suggesting that metformin could effectively induce the phosphorylation of Nrf2 and its nuclear translocation; both the pretreatment with 1.25 and 2.5 μmol/L dorsomorphin could significantly reduce the phosphorylation levels of AMPKα and Nrf2 induced by 10 mmol/L metformin. The proportions of senescent cells in the dorsomorphin + metformin + UVA group and the siRNA-Nrf2 + metformin + UVA group were 67.84% ± 2.74% and 65.94% ± 1.33%, respectively, which were significantly higher than those in the metformin + UVA group （37.76% ± 1.64%, t = 14.45, 13.34, respectively, both P < 0.001）. Conclusion Metformin may inhibit UVA-induced photoaging of HaCaT cells by activating the AMPK/Nrf2 signaling pathway, scavenging ROS and reducing DNA damage.

Key words: Skin aging, Cell aging, Ultraviolet rays, Metformin, Photoaging, UVA, Nrf2, AMPK