Gallic acid inhibition of Src-Stat3 signaling overcomes acquired resistance to EGF receptor tyrosine kinase inhibitors in advanced non-small cell lung cancer

Abstract

Tyrosine kinase inhibitors (TKIs) targeting epidermal growth factor receptor (EGFR) have clinically benefited to lung cancer patients harboring a subset of activating EGFR mutations. However, even with the remarkable therapeutic response at the initial TKI treatment, most lung cancer patients eventually have relapsed aggressive tumors due to acquired resistance to the TKIs. Here, we report that 3, 4, 5-trihydroxybenzoic acid or gallic acid (GA), a natural polyphenolic compound, shows anti-tumorigenic effects in TKI-resistant non-small cell lung cancer (NSCLC). Using both in vitro growth assay and in vivo xenograft animal model, we demonstrated tumor suppressive effect of GA was more selective for the TKI-resistant cancer compared to the TKI-sensitive one. Mechanistically, GA treatment inhibited Src-Stat3-mediated signaling and decreased the expression of Stat3-regulated tumor promoting genes, subsequently inducing apoptosis and cell cycle arrest in the TKI-resistant lung cancer but not in the TKI-sensitive one. Consistent with the in vitro results, in vivo xenograft experiments showed the TKI-resistant tumor-selective growth inhibition and suppression of Src-Stat3-dependent signaling in the GA-treated tumors isolated from the xenograft model. This finding identified an importance of Src-Stat3 signaling cascade in GA-mediated tumor-suppression activity and, more importantly, provides a novel therapeutic insight of GA for advanced TKI-resistant lung cancer.

Keywords: EGFR-TKI resistance; Src; Stat3; gallic acid; lung cancer.

Figure 1. Therapeutic selectivity of GA for gefitinib resistant NSCLC

(A) A chemical structure of 3,4,5-trihydroxybenzoic acid or GA. (B,–D) Therapeutic effects of GA and gefitinib were evaluated using MTT as well as colony formation assays in HCC827 clones (B), various NSCLC cells (C) and HBECs (D). Lung cells were treated with gefitinib 0.3 μM or GA 50 μM in colony formation asssay or at indicated concentrations in MTT assay. NSCLC lines were classified into 3 categories upon TKI sensitivity as follows: TKI-sensitive (□), HCC827, H3255; TKI-intermediate (◘), HCC827C1, H1666; and TKI-resistant (■), HCC827C2, H1650, (▲), H1975, (●), H358. In every MTT assay, values are mean ± SEM of five replicate assays. (E) The association between GA and gefitinib sensitivity in NSCLC was evaluated using Pearson analysis (r² = 0.72, p = 0.031). The growth response between GA and gefitinib was negatively correlated to each other in the same panel. The plot represents for growth responses of each cell at gefitinib (0.3 μM) and GA (50 μM) from (B) and (C).

Figure 2. GA inhibits Src-Stat3-mediated signaling in TKIR NSCLC

(A) Gefitinib treatment induces Stat3 phosphorylation. HCC827 and H3255 cells were treated with 0.3 μM of gefitinib in a time-dependent manner, and followed by immunoblot assay for proteins involved in EGFR and Stat3 signaling. Basal EGFR activation and Stat3 phosphorylation were reversely correlated to each other between TKIS and TKIR cells (right panel). (B) GA-mediated Src and Stat3 phosphorylation in TKI-sensitive vs. -resistant NSCLC lines. NSCLC cells were treated with GA (20 μM, 50 μM) for 6 hours and followed by immunoblot assay for phosphorylation of EGFR, Src, and Stat3 proteins. (C) mRNA expression of Stat3 regulated genes upon GA treatment. Cells were treated with 50 μM of GA for 24 hours and followed by QPCR assay for mRNA expression of Stat3 target genes. Values are mean ± SEM of triplicate assays. Difference were analysed using Student's t-test. * p < 0.05; ^Δp < 0.01; ^#p < 0.001.

Figure 3. GA induces apoptosis and cell cycle arrest in TKIR NSCLC

(A, C) NSCLC cells were treated with vehicle or GA (20 μM and 50 μM) for 24 hours and immunnoblot assays were performed for the expression of proteins involved in apoptosis (A) and cell cycle regulation (C). (B) DAPI staining were performed to determine apoptosis. Cells were treated with 50 μM of GA treatment for 24 hours and followed by DAPI staining.

Figure 4. In vivo therapeutic evaluation of GA in xenograft model of TKIR tumor

In vivo xenograft tumor model was established by injecting HCC827 or H1650 lung tumor cells into the flank region of athymic nude mouse. HCC827 and H1650 represent TKI-sensitive and -resistant tumors, respectively. The xenografted mice were intraperitoneally injected with vehicle (n = 5, HCC827 and n = 4, H1650) or GA (200 mg/kg/day) (n = 8, HCC827 and n = 6, H1650) every other day for 24 days (HCC827) or 32 days (H1650). Tumor volumes were measured every other day. Tumor volume represents the tumor size ± SEM and statistical analysis was determined using mixed model, p = 0.0211 (A). At the end of the experiment, tumor tissues were isolated for immunoblot assay (B) and QPCR assay (C) for genes of interest involving cell cycle regulation and Stat3 signaling. Values are mean ± SEM. Difference were analysed using Student's t-test. * p < 0.05; ^Δp < 0.01; ^#p < 0.001.