Fig 2: APE1-dependent inhibition of GSK-3β promoted the increase in NRF2. (A and B) Knockdown of APE1 blocked ABS-induced NRF2 upregulation. CPB (A) and OE33 (B) cells were transfected with si-Ctrl and si-APE1 followed by exposure to 100 μM ABS for 20 min, then recovery for 1, 3, 6 h in complete media. Immunoblot results of APE1, NRF2, p-GSK-3β (S9), total GSK-3β and β-actin levels were shown. (C and D) CPB and OE33 cells were pretreated with or without 10 mM LiCl (GSK-3β inhibitor) overnight and on the next day cells were treated with ABS (100μM/20min) followed by recovery time points. Immunoblots of APE1, NRF2, and β-actin levels were shown. (E) OE33 cells were transfected with 1 μg of pcDNA flag-APE1, whereas control cells received an empty vector. After 24 h transfection, the cells were split and pretreated with 10 mM LiCl overnight. On the next day, the cells were treated with ABS, followed by different recovery time points in the presence and absence of LiCl. Immunoblots of APE1, NRF2 and β-actin levels were shown.

Fig 3: APE1 was required for NRF2 stability. (A, C) OE33 and FLO1 cells were treated with or without ABS (100 µM) for 20 min, followed by recovery in complete media with cycloheximide (CHX, 100 µg/mL) at the indicated time points. (E, G) OE33 (E) and FLO1(G) cells with stable knockdown of APE1 using sh-APE1 or sh-Ctrl were treated with cycloheximide (CHX, 100 µg/mL) at the indicated time points. The levels of APE1 and NRF2 were determined by western blotting. The band intensity of NRF2 was measured and normalized with the actin using the Quantity One software (BioRad Laboratories, USA) and normalized to ß-actin of the same samples. (B, D, F and H) Half-life time (t1/2) of NRF2 was calculated and plotted using GraphPad Prism software, corresponding to A, C, E and G, respectively.

Fig 4: Loss of APE1 resulted in increased oxidative stress, DNA damage, and promoted cell death in response to reflux conditions. (A–B) Flow cytometry analyses of OE33 cells to detect the intracellular ROS levels using CM-H2DCFDA dye in APE1 knockdown (sh-APE1) and control (sh-Ctrl) cells treated with or without ABS (pH 4.0). Representative flow cytometry profiles are shown in (A). The quantitative results from 3 independent experiments are shown in (B). (C–F) OE33 cells were transfected with si-Ctrl and si-APE1 for 48 h, followed by acidic bile salts (ABS) treatment for 20 min and recovery in full medium for 3 h. Cells treated with 100 μM H2O2 were used as a positive control. 8-Oxoguanine (8-OxoG), an oxidative DNA damage marker, was used for immunofluorescence. Representative immunofluorescence images are shown in (C) and the quantitative data, using ImageJ, is shown in panel (D). p-H2AX (S139, γH2AX, green), a DNA double-strand breaks marker, and APE1 (red) were used for immunofluorescence. The representative images are shown in panels, (E) and the quantitative data is shown in panel (F). DAPI (blue) was used as a nuclear counterstain. (G) Flow cytometry analysis of annexin V staining in si-Ctrl and si-APE1 cells with or without ABS treatment. The quantitative analysis data from three independent experiments are shown. (H) Western blotting analysis of cleaved PARP in si-Ctrl and si-APE1 cells with ABS or control PBS treatment. *P < 0.05; **P < 0.01; ***P < 0.001. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig 5: A schematic diagram illustrating the role of APE1 in activating NRF2 under reflux conditions. Exposure of cells to reflux conditions (acidic bile salts, ABS) generates high levels of ROS. High ROS and oxidative stress levels induce APE1-dependent increase in NRF2, leading to its release and translocation into the nucleus. In the nucleus, NRF2 is stabilized and regulated by APE1 redox function that inactivates GSK-3ß dependent degradation of NRF2. Loss of redox function of APE1 mediates inactivation of GSK-3-ß to promote accumulation and activation of NRF2.