APE1 stimulates EGFR-TKI resistance by activating Akt signaling through a redox-dependent mechanism in lung adenocarcinoma

Epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs) have become the standard first-line treatment for advanced lung adenocarcinoma (LUAD) cancer patients with activating EGFR mutations. However, most patients show acquired resistance to EGFR-TKIs, thereby resulting in a modest overall survival benefit. Here, we found that expression level of APE1 was closely associated with TKI resistance in LUAD. Our clinical data show that level of APE1 was inversely correlated with progression-free survival rate and median time to progression in EGFR-mutated LUAD patients. Additionally, we observed increased expression of APE1 in TKI-resistant LUAD cell lines compared to their parental cell lines. Overexpression of APE1-protected TKI-sensitive LUAD cells from TKI-induced cell growth inhibition and cell death. In contrast, inhibition of APE1-enhanced TKI-induced apoptosis, cell growth inhibition and tumor growth inhibition in TKI-resistant LUAD. In addition, we identified that APE1 positively regulates Akt activation and APE1 overexpression-induced TKI resistance was attenuated by inhibition of Akt activity. Finally, we demonstrated that inhibition of the redox function of APE1 enhances the sensitivity of TKI-resistant LUAD cells to TKI treatment and inhibits Akt phosphorylation in TKI-resistant LUAD cells, but not by inhibition of the APE1 DNA repair function. Taken together, our data show that increased expression of APE1 significantly contributes to TKI resistance development in LUAD, and targeting APE1 may reverse acquired resistance of LUAD cells to TKI treatment. Additionally, our data show that APE1 regulates TKI resistance in LUAD cells by activating Akt signaling through a redox-dependent mechanism.


Introduction
Lung cancer is the leading cause of cancer-related mortality worldwide, and lung adenocarcinoma (LUAD) is the most common histologic subtype of lung cancer 1,2 . In LUAD, several oncogenic driver mutations have been detected, including K-Ras, epidermal growth factor receptor (EGFR), and BRAF mutations [2][3][4] , and these activating genetic mutations are now targets for kinaseinhibitor therapy 2,5 . Among them, EGFR is found in 10-40% LUAD patients, occurring most frequently in never-smokers and in East Asian populations [6][7][8] . Notably, EGFR tyrosine kinase inhibitors (TKIs) have become the standard first-line treatment for advanced lung cancer patients with activating EGFR mutations 9 . However, acquisition of resistance to these EGFR-TKIs is almost inevitable at a median of 9-13 months, resulting in a modest overall survival benefit 10 .
T790M secondary mutation of EGFR is the most common acquired resistance mechanism to first-generation and second-generation EGFR-TKIs that account for approximately 50% of EGFR-TKI resistance cases of lung cancer 11 . Additional mechanisms of acquired resistance to EGFR-TKIs include activation of insulin-like growth factor-1 receptor (IGF-1R), amplification of MET and HER2, upregulation of the AXL receptor or its ligand, activating mutations in PIK3CA and BRAF, and SCLC transformation 6,10,11 . However, the TKI resistance mechanism for 15-30% of cases is still unknown 6,10,11 .
Apurinic/apyrimidinic endonuclease/redox factor-1 (APE1/Ref-1) is a multifunctional protein that plays critical roles both as a redox regulator of transcription factor activation and as part of the DNA damage response. Previous studies show that elevated APE1 significantly contributes to the development of therapeutic resistance and is positively correlated with poor clinical outcomes in several cancers 12 . Interestingly, although not in lung cancer, a recent report show that APE1 was involved in EGFR activation 13 . In addition, studies show that APE1 also involved in regulation of Akt activation 14,15 . Akt (protein kinase B) is a serine/threonine protein kinase that plays a key role in cancer by stimulating cell proliferation, inhibiting apoptosis, and modulating protein translation 16 . Notably, studies show that activated Akt signaling is involved in the therapeutic resistance of lung cancer, including both T790M and non-T790M mutation mechanisms of EGFR-TKIs resistance 5,17 . These findings suggest that APE1 may be involved in EGFR-TKIs resistance. However, the effects of APE1 on EGFR-TKIs resistance is unknown.
In this study, we identified that APE1 expression was increased in EGFR-TKI-resistant LUAD cell lines compared to their parental cell lines, and the level of APE1 was inversely correlated with median progression time in LUAD patients with EGFR mutations treated only with TKIs. Overexpression of APE1 reduced the sensitivity of to TKIs treatment in TKI-sensitive LUAD cells, while inhibition of APE1 enhanced sensitivity to TKI treatment in TKI-resistant LUAD cells. In addition, we identified that APE1-induced TKI resistance in LUAD cells by activating Akt signaling through a redox-dependent mechanism.

Results
Increased expression level of APE1 was associated with TKIs resistance in EGFR-mutated LUAD patients who were treated only with TKIs were divided into four groups based on APE1 staining score (Fig. 1a). In EGFR-mutated LUAD patients, APE1 levels were not associated with patient age, gender, smoking status, and TNM stage (Table 1). However, our data show that APE1 expression level were inversely correlated with progression-free survival rate (Fig. 1b) and median time to progression (TTP) in LUAD patients with EGFR mutations (Fig. 1c). In addition, we demonstrated significantly increased expression levels of APE1 in TKI-resistant LUAD cell lines HCC827/IR (resistant to icotinib) and PC-9/ER (resistant to erlotinib) (Supplement Fig. 1) compared to their parental cells at both mRNA ( Fig. 1d) and protein levels (Fig. 1e). These results were further confirmed by immunofluorescence in TKI-resistant LUAD cells and their parental cells, and we observed similar results by Western blot analysis (Fig. 1f). Together, these data suggest that increased expression level of APE1 is closely correlated with TKI resistance in EGFR-mutated LUAD.
Overexpression of APE1 contributes to TKI resistance in EGFR-mutated LUAD Next, to investigate whether the increased expression of APE1 is directly involved in EGFR-TKI resistance, we overexpressed APE1 in TKI-sensitive LUAD cell lines HCC827 and PC-9 ( Fig. 2a, b), and the cells were subsequently treated with TKIs and subjected to cell viability and apoptosis assays. Cell viability assay results show that APE1 overexpression significantly protected HCC827 and PC-9 cells from icotinib-induced and erlotinib-induced cell growth inhibition, respectively, compared to their control groups (Fig. 2c). Similarly, apoptosis assay results also show that overexpression of APE1 significantly protected HCC827 and PC-9 cells from icotinib-induced and erlotinib-induced cell death, respectively (Fig. 2d). Altogether, these data demonstrate that overexpression of APE1 significantly contributes to the development of TKI resistance in LUAD cells by attenuating TKI-induced apoptosis.

Silencing of APE1 enhances the sensitivity of TKI-resistant LUAD cells to TKI treatment
Our observation that overexpression of APE1 contributes to the development of TKI resistance in LUAD cells prompted us to investigate whether the silencing of APE1 could enhance the sensitivity of TKI-resistant LUAD cells to TKI treatment. As expected, our results show that silencing of APE1 dramatically enhanced icotinib-induced and erlotinib-induced cell growth inhibition in TKI-resistant LUAD cell lines HCC827/IR and PC-9/ER, respectively ( Fig. 3a and Supplementary Fig. 2a). Additionally, APE1 silencing significantly stimulated icotinib-induced and erlotinib-induced apoptosis of TKI-resistant LUAD cell lines HCC827/IR and PC-9/ER, respectively ( Fig. 3b and Supplementary Fig. 2b). Consistently, we detected significantly increased proapoptotic protein expression, including cleaved caspase-9 and Bax, while decreased anti-apoptotic Bcl-2 expression in the combination group of APE1 silencing and TKI treatment compared to the TKI only treatment group ( Fig. 3c and Supplementary Fig. 2c), suggesting that silencing of APE1 sensitize TKI-resistant LUAD cells to TKI treatment partially through inducing apoptosis. In fact, MTT and apoptosis analysis showed that treatment of caspase inhibitor zVAD-FMK (carbobenzoxyvalyl-alanyl-aspartyl-[O-methyl]-fluoromethylketone) significantly inhibit silencing of APE1 enhanced cell viability  Fig. 3a, b) in both HCC827/IR and PC-9/ER cell lines. Furthermore, we confirmed the in vitro results using a xenograft model generated by HCC827/IR cells. Here, we used AT101 as an APE1 inhibitor. Because AT101 is R-(-)-enantiomer of gossypol and our previous study show that gossypol can inhibits APE1 activity by directly interacting to APE1 18 . Furthermore, AT101 inhibited APE1 expression in gastric cancer 19 and TKI resistance LUAD cell lines ( Supplementary Fig. 4a). Notably, consistent with APE1 silencing, AT101 treatment can overcome APE1 overexpression-induced TKI resistance in LUAD cell HCC827 and PC-9 ( Supplementary Fig. 4b).
These findings clearly suggesting that AT101 is a inhibitor of APE1. Consistent with the in vitro results, animal experiments show that combination treatment of icotinib and APE1 inhibitor, AT101, significantly suppressed tumor growth (Fig. 4a, b) and cancer cell proliferation (Fig. 4c) compared to the control or single drug treatment group. Additionally, the combined treatment of icotinib and AT101 significantly induced pro-apoptotic protein Bax expression and inhibited anti-apoptotic protein Bcl-2 and Bcl-xL expression compared to control or single drug treatment group (Fig. 4d). Taken together, these findings suggest that inhibition of APE1 can overcome the resistance of TKI-resistant LUAD cells to TKI treatment.

Akt pathway was involved in APE1-mediated EGFR-TKI resistance in LUAD cells
Activated Akt signaling is involved in the development of EGFR-TKI resistance. Interestingly, previous studies show that APE1 is involved in the activation of Akt signaling 15,20 . Thus, we subsequently assessed the Akt and p-Akt expression levels in both TKI-resistant and their parental cells. As shown in Fig. 5a, total Akt levels were not upregulated in TKI-resistant cells, while p-Akt levels were significantly upregulated in both TKI-resistant cell lines HCC827/IR and PC-9/ER compared to their corresponding parental cells. Furthermore, to investigate whether increased levels of p-Akt were associated with APE1, we silenced APE1 in TKI-resistant cells and then measured total Akt and p-Akt expression by Western blotting. The results show that silencing of APE1 dramatically decreased the level of p-Akt in both HCC827/IR and PC-9/ER cells compared to their corresponding control cells, but total Akt levels did not change with APE1 silencing (Fig. 5b). Next, to investigate whether Akt activation is involved in the development of APE1induced TKI resistance, APE1 was overexpressed in HCC827 and PC-9 cells, and the cells were then treated with TKI with or without Akt inhibitor MK2206. The results show that inhibition of Akt significantly attenuated APE1 overexpression-induced inhibition of TKI-caused apoptosis in both HCC827 and PC-9 LUAD cell lines (Fig. 5c). In addition, Western blot analysis showed that MK2206 treatment inhibited APE1 overexpression-induced Akt phosphorylation, anti-apoptotic protein Bcl-2 expression while increasing APE1 overexpressionsuppressed expression of pro-apoptotic proteins, Bax and cleaved caspase-9 (Fig. 5d). Consistent with these data, the cell viability assay showed that MK2206 treatment significantly enhanced the sensitivity to TKI treatment of TKI-resistant LUAD cells (Fig. 5e). Taken together, these results suggest that APE1 induces TKI resistance partially through activation of Akt in LUAD.
APE1 induces TKI resistance through a redox-dependent mechanism in LUAD As described above, APE1 is a multifunctional protein that plays a role as a redox regulator of transcription c Combination treatment of APE1 silencing and TKI significantly increased pro-apoptotic protein expression while suppressing the expression of antiapoptotic proteins in TKI-resistant LUAD cells. Indicated cells were transfected with siRNA of APE1 or negative control oligonucleotides (NC), then treated with indicated TKIs. After 48 h of TKI treatment, cells were subjected to Western blotting. d Caspase inhibitor zVAD-FMK treatment inhibited silencing of APE1 enhanced cell viability inhibition effects of TKIs in TKI-resistant LUDA cells. Indicated cells were transfected with siRNA of APE1 (siAPE1) or negative control oligonucleotides (NC). After 24 h of transfection, cells were reseed in 96-well plates. After overenight, cells were treated with indicated drugs for 48 h, then subjected to cell viability assay. **p < 0.01 factor activation and as an apurinic/apyrimidinic-site endonuclease. To investigate which function of APE1 was involved in Akt activation and development of TKI resistance, we treated TKI-resistant LUAD cells with APE1 redox function inhibitor, E3330, or APE1 DNA repair activity inhibitor, APE1 inhibitor III, and then checked the p-Akt levels and the sensitivity to TKI in TKI-resistant LUAD cells. Our results show that APE1 inhibitor III treatment did not affect TKI-induced cell growth inhibition (Fig. 6a), apoptosis (Fig. 6b) and apoptosis-related protein expressions (Fig. 6c). In contrast, redox inhibitor E3330 treatment significantly enhanced TKI-induced cell growth inhibition (Fig. 7a), apoptosis (Fig. 7b) and expression of pro-apoptotic proteins cleaved caspase-9 and Bax, while it inhibited expression of anti-apoptotic protein Bcl-2 (Fig. 7c) in TKI-resistant LUAD cells. Consistent with these results, APE1 inhibitor III treatment did not change p-Akt levels (Fig. 6c), but E3330 treatment significantly inhibited p-Akt levels in TKI-resistant LUAD cells (Fig. 7c). These findings suggest that APE1 induces TKI resistance and Akt activation through an APE1 redox-dependent mechanism in LUAD.

Discussion
EGFR-TKIs are widely used for the treatment of EGFRmutated LUAD patients. Thus, the occurrence of TKI resistance is a crucial clinical issue with respect to the treatment of LUAD with EGFR mutations 21 . In the present study, we used clinical sample analyses and a series of in vitro and in vivo experiments to identify that increased expression of APE1 was closely associated with TKI resistance development in LUAD. Our results show that APE1 was significantly increased in TKI-resistant LUAD cells compared to their parental cells, and EGFR-mutated LUAD patients with high expression APE1 presented lower progression-free survival rates and shorter median time to progression compared to patients with low APE1 expression when treated with TKIs. Overexpression of APE1 inhibits TKI-induced cell growth inhibition and cell death in TKI-sensitive LUAD cells, while silencing of APE1 significantly stimulated TKI-induced cell growth inhibition and apoptosis in TKIs-resistant LUAD cells. Consistent with in vitro, in a xenograft model generated by TKI-resistant LUAD cells, combination treatment of APE1 inhibitor and TKI markedly inhibited tumor growth compared to the control and single drug treatment groups. Taken together, these findings suggest that increased expression level of APE1 significantly contributes to TKI resistance development, and APE1 may be an ideal target for part cases with acquired EGFR-TKIs resistance.
In this study, we also clarified the mechanisms of APE1 on the regulation of sensitivity to TKIs in LUAD cells. The results of the present study indicated that APE1 contributes to TKI resistance development by activating Akt signaling in LUAD cells. Our data show that p-Akt was increased in TKI-resistant LUAD cells and overexpression of APE1 increases p-Akt level in LUAD cells. In addition, increased expression of p-Akt in TKIresistant LUAD cells was significantly reduced by silencing APE1. Furthermore, the inhibition of Akt activation significantly inhibited APE1 overexpression-induced TKI resistance in LUAD cells. These findings also suggest that inhibition of Akt activation may be an ideal target for APE1 overexpression in TKI-resistant LUAD.
Next, we demonstrated that APE1 induces Akt activation through a redox-dependent mechanism in LUAD cells. Previous studies show that APE1 is involved in the regulation of cell proliferation and apoptotic protein expression by redox-specific functions 22 . Consistent with this report, our results also show that APE1 has a redox function dependently involved in TKI resistance development and Akt activation. Our results show that inhibition of the APE1 redox function can inhibit APE1 overexpression-induced resistance of TKI and Akt activation in LUAD. However, inhibition of the APE1 DNA repair function does not affect APE1-induced TKI resistance and Akt activation in LUAD cells. However, the detailed mechanism of how APE1 regulates Akt activation needs further study.
In conclusion, we demonstrated that aberrant increased expression of APE1 contributes to development of EGFR-TKI resistance in LUAD with EGFR mutations by activating Akt signaling through a redox-dependent mechanism. Our data also provide a rationale for targeting APE1 in EGFR-resistant LUAD patients.

Materials and methods
Human specimens, cell culture, and transfection Human specimens were collected from 61 LUAD patients with EGFR mutation by biopsy or surgery at Daping Hospital and Research Institute of Surgery, Third Military University. The patients' characteristics are summarized in Table 1. The collection and use of human samples were approved by the ethical review committees.
PC-9 and PC-9/ER (Erlotinib-resistant PC-9) were obtained from cancer center of Xinqiao Hospital, Third Military Medical University. HCC827 and HCC827/IR (Icotinib-resistant HCC827) cells were obtained from Cancer Institute, Peking Union Medical College, Chinese Academy of Medical Sciences. Cells were cultured in RPMI with 10% fetal bovine serum, 100 μg/me streptomycin and 100U/me penicillin at 37°C. All materials for cell culture were purchased from Sigma-Aldrich (St. Louis, MO, USA).
Cell transfection was performed using Lipofectamine 3000 according to manufacture's instruction (Invitrogen).

Animal experiments
Animal experiments were conducted using 6-week-old female nude mice. 1 × 10 7 HCC827/IR cells were in 100 μl of phosphate-buffered saline were subcutaneously injected into each mouse. When the mean tumor size reached approximately 100 mm 3  Indicated cells were treated with indicated concentrations of TKI and 5 µM of E3330 for 72 h, then subjected to cell viability assay. b E3330 significantly enhanced TKI-induced apoptosis in TKI-resistant LUAD cells. Indicated cells were treated with indicated concentrations of TKI and 5 µM of E3330 for 48 h, then subjected to apoptosis assay. c Treatment of E3330 with or without TKIs inhibited Akt phosphorylation, anti-apoptotic proteins expression and stimulated pro-apoptotic proteins expression in TKI-resistant LUAD cells. Indicated cells were treated with 1 µM of indicated TKI and 5 µM of E3330 for 48 h, then subjected to Western blot analysis. **p < 0.01; ***p < 0.001 groups (6 mice per group). Control group mice were treated with PBS; Icotinib group mice were treated with Icotinib by IP injection (4 mg/day/Kg body weight) every day; AT101 group mice were treated with AT101 through oral injection every day (35 mg/Kg body weight/day); combination group mice were treated with Icotinib and AT101 every day as described above. The treatment of mice were conducted for 10 days and mice were sacrificed after 10 days of the end drug treatment. Tumor volumes were measured every 4 days.