Acquired Resistance to EGFR TKIs Mediated by TGFβ1/Integrin β3 Signaling in EGFR-Mutant Lung Cancer

This article is featured in Highlights of This Issue, p. 2183

First-generation EGFR tyrosine kinase inhibitors (TKI), including gefitinib and erlotinib, have demonstrated dramatic efficacy in non–small cell lung cancer (NSCLC) patients with EGFR-activating mutations (1). Despite an impressive initial response, almost all patients eventually have a relapse due to the occurrence of acquired resistance. The secondary EGFR mutation T790M has been reported to be the most common mechanism of acquired resistance to first-generation TKIs (2, 3). Recently, third-generation TKIs, such as osimertinib, which was approved by the FDA in 2017 for patients with EGFR T790M-positive NSCLC whose disease progressed during or after EGFR TKI therapy, have been developed to target T790M. However, as expected, an increasing number of cases with acquired resistance to osimertinib are being reported (4, 5). The mechanisms of acquired resistance to third-generation TKIs reported in clinical settings are quite similar to those of first- or second-generation TKIs, including tertiary EGFR mutations, bypass or downstream activation, and histologic transformation (4, 5). Because the clinical efficacy of TKIs is ultimately limited by the development of acquired resistance, further investigation of novel molecular mechanisms is essential to develop strategies to overcome or delay the acquired resistance to TKIs in NSCLC.

Integrins are a major family of cell-surface receptors that are heterodimers of the α and β subunits. Integrin is expressed in a cell-specific and context-dependent manner. Abnormal expression of integrin is often associated with the development and progression of various diseases (6, 7). Integrins have been demonstrated to be regulators of cancer progression, such as tumor growth, metastasis, treatment resistance, and cancer stemness (7). Integrins have also been implicated in acquired resistance to EGFR TKIs in lung (8, 9), breast (10), and liver cancers (11). All of these characteristics have established integrins as a promising target for cancer treatment, including targeted drug delivery. However, even within the same cancer or biological event, contradictory results have been reported (7). Therefore, for the purpose of targeting integrins, the specific integrins that are involved or required in those events need to be further investigated.

The aim of our investigation was to identify which specific integrins are involved and how they are regulated in acquired resistance to EGFR TKIs in NSCLC. We have established in vitro and in vivo EGFR-mutant NSCLC models with acquired resistance to EGFR TKIs, including the third-generation TKI osimertinib. In this study, we reported that integrin β3 expression was dramatically and consistently upregulated in these models. Moreover, the role of integrin β3 in acquired resistance to EGFR TKIs in lung cancer was extensively studied. Mechanistically, integrin β3 was induced by increased TGFβ1 in acquired TKI-resistant lung cancer. Taken together, our data demonstrated that the TGFβ1/integrin β3 axis is a new mechanism and target for acquired EGFR TKI resistance in NSCLC.

Gefitinib, SB-431542, and c(RGDfK) were purchased from Selleck and dissolved in DMSO.

Human NSCLC HCC827, H1975, A549, H292, and H1299 cells were purchased from the Cell Bank of Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China), and PC9 cells were a gift from Dr. Qianggang Dong of the Shanghai Cancer Institute, Shanghai Jiao Tong University School of Medicine. The characterization of cell lines was listed in Supplementary Table S1. The cell lines were cultured as previously described (12) and tested by certified third-party laboratories for authenticity using short tandem repeat analysis and examined for Mycoplasma regularly. Gefitinib- or osimertinib-resistant cells were established by the stepwise escalation method and maintained as previously described (12).

Male athymic BALB/c nude mice were purchased from the Shanghai Laboratory Animal Center (Chinese Academy of Sciences, Shanghai, China) and housed in environmentally controlled, specific pathogen-free conditions for 1 week before the study. All experimental procedures were reviewed and approved in accordance with the guidelines for the care and use of laboratory animals and obtained informed written consent at Shanghai Jiao Tong University.

To establish mouse xenograft models, the same amount of the indicated tumor cells was injected subcutaneously into both flanks of each mouse. The tumor volume was measured 1 week after injection, and then every other day or twice a week. Tumor volumes (mm³) were calculated as length × width²/2.

To establish gefitinib-resistant lung cancer tumors in vivo, when HCC827- or PC9-derived tumors reached ∼100 mm³, gefitinib was given by gavage at 12.5 mg/kg daily until the mice were sacrificed.

For the transient knockdown of integrin β3, cells were transfected with 100 nmol/L siITGB3 or siNC (GenePharma) using Lipofectamine 3,000 according to the manufacturer's instructions. The efficacy of transfection was verified by Western blot assay. The sequences used for ITGB3 siRNA are as follows: siITGB3 1: 5′-CAAGCCUGUGUCACCAUACTT-3′; siITGB3 2: 5′-UUACUGCCGUGACGAGAUUTT-3′.

For the stable knockdown of integrin β3, cells were transfected with pGLVH1/GFP/Puro lentivirus encoding ITGB3-shRNA (shITGB3) or scrambled sequence (shNC; GenePharma). Stable cell clones were selected by puromycin and then verified by Western blot assay.

To overexpress integrin β3, cells were infected with EX-E2219-Lv105 lentivirus encoding the full-length human ITGB3 cDNA sequence (LV-ITGB3; GeneCopoeia) or EX-eGFP-Lv105 lentivirus encoding GFP (LV-NC) as a control. Stable cell clones were selected by puromycin and then verified by Western blot assay.

Protein and mRNA expression levels were measured by Western blot and qRT-PCR assays, respectively. β-Actin was used as a loading control for Western blot, and GAPDH was used as a control for qRT-PCR. The lists of antibodies and primers used are available in Supplementary Tables S2 and S3, respectively.

Cell viability was determined using a Cell Counting Kit-8 (CCK8) colorimetric assay (Dojindo) according to the manufacturer's manual, trypan blue exclusion test or IncuCyte ZOOM system (Essen BioScience).

5-Ethynyl-2′-deoxyuridine (EdU) incorporation assays and the IncuCyte ZOOM system were used to measure cell proliferation. Detailed methods are provided in Supplementary Materials and Methods.

Colony formation and transwell migration and invasion assays were performed as previously described (12). Detailed methods are provided in Supplementary Materials and Methods.

Cells (5 × 10⁵) were seeded on ultra-low attachment 6-well plates (Corning). After 48 to 72 hours, cells were collected, and apoptosis was assessed by Western blot analysis of the apoptosis markers caspase-3 and PARP; cell viability assays were conducted using the trypan blue exclusion test.

The 1.3-kb promoter region of human ITGB3 gene (NM_008332) was cloned and inserted into multiple cloning sites of pGL3-basic vector (Promega) to construct pGL3-ITGB3 and verified by sequencing. The empty pGL3-basic vector was used as a negative control (NC). pRL-SV40 was used as an internal control for transfection efficiency. Dual-Luciferase Reporter Assay System (Promega) was used to measure luciferase activity according to the manufacturer's manual.

All data are presented as the mean ± SEM. Statistical analysis was conducted using GraphPad Prism 7.0 software. Differences between groups were examined using Student t test and two-way ANOVA. Differences were considered significant if the P value was less than 0.05.

The data generated or analyzed during this study are included in this published article and its additional files.

All animal experimental procedures were reviewed and approved in accordance with the guidelines for the care and use of laboratory animals at Shanghai Jiao Tong University.

To identify which specific integrins are involved in acquired EGFR TKI resistance in NSCLC, we first generated in vitro models by growing EGFR TKI-sensitive NSCLC cell lines (HCC827 and PC9) in increasing concentrations of an EGFR TKI (gefitinib or osimertinib) as we previously reported (12). Next, to establish in vivo models, nude mice with subcutaneous HCC827 or PC9 xenograft tumors (one left and one right tumors per mouse) were given saline for 1 week (sensitive) or gefitinib orally for 2 to 3 months to establish gefitinib-resistant tumors as previously reported (8, 13, 14). Daily gefitinib treatment of the xenografts led to initial remarkable tumor shrinkage and later tumor regrowth at 6 to 8 weeks (Supplementary Fig. S1B). We then examined the expression of integrins in EGFR TKI-resistant cells and xenograft tumors by Western blot assays. As shown in Fig. 1A and C and Supplementary Fig. S1A, S1C, and S1D, integrin β3 was the only integrin subunit tested that was remarkably and consistently overexpressed in both resistant cells and tumors compared with that in the respective sensitive cells and tumors. qRT-PCR also confirmed the upregulation of integrin β3 mRNA expression (Fig. 1B and D). Then, HCC827 and PC9 cells were treated with gefitinib or osimertinib for up to 72 hours. As shown in Fig. 1E and Supplementary Fig. S2A, integrin β3 expression was gradually and dramatically upregulated on both mRNA and protein levels by qRT-PCR and Western blot, respectively. The expression levels of other integrin subunits were decreased or unchanged except that integrin β1 was moderately increased (Supplementary Fig. S2B). Analysis of integrin β3 expression in 720 lung adenocarcinoma patients, including all stages and treatment regimes, revealed that high integrin β3 expression was associated with a significant decrease in patient survival rates (Fig. 1F). These results indicated that integrin β3 was involved in the progression of lung adenocarcinoma.

To examine whether overexpression of integrin β3 was associated with acquired resistance to EGFR TKIs in vitro, we antagonized integrin β3 pharmacologically by treating resistant cells with the selective integrin αvβ3 inhibitor c(RGDfK) or genetically by transiently transfecting resistant cells with specific integrin β3 siRNA; then, we assessed the sensitivity to TKIs by cell viability (CCK8) and cell growth (IncuCyte) assays. First, we determined the efficacy of c(RGDfK) or integrin β3 siRNA by examining pFAK, pAkt, and integrin β3 levels by Western blot. As shown in Supplementary Fig. S3A, cells transfected with integrin β3 siRNA for 72 hours showed dramatically decreased integrin β3 levels without any significant changes in other integrin subunits tested and decreased pFAK and pAkt levels. The levels of integrin β3 were dramatically downregulated up to 1 week after transfection (Supplementary Fig. S3E). As shown in Supplementary Fig. S3B, cells treated with c(RGDfK) showed decreased pFAK and pAkt levels, moderately decreased integrin β3 and slightly increased αv levels. CCK8 and IncuCyte assays demonstrated that siITGB3 and c(RGDfK) significantly increased the sensitivity of HCC827GR, HCC827OR, and PC9OR cells to gefitinib or osimertinib (Fig. 2A–D; Supplementary Fig. S4A and S4B). Next, we transfected HCC827 cells with a lentiviral vector expressing ITGB3 (LV-ITGB3), which resulted in the overexpression of integrin β3 in HCC827 cells with no concomitant overexpression of other integrin subunits tested and increased pFAK and pAkt levels (Supplementary Fig. S3C). As shown in Supplementary Fig. S4D, overexpression of integrin β3 conferred resistance to gefitinib in HCC827 cells. To further investigate the role of integrin β3 in the resistance to EGFR TKIs in lung cancer, we examined integrin β3 expression levels in several lung cancer cell lines (Supplementary Fig. S2C). Western blot analysis showed that H1975 cells had the highest integrin β3 level of all cell lines tested. Then, we treated H1975 cells with c(RGDfK) or siITGB3 and examined the sensitivity to osimertinib using the CCK8 assay. As shown in Supplementary Fig. S4C, antagonizing integrin β3 also significantly increased the sensitivity of H1975 cells to osimertinib. We also transfected H1975 cells with LV-ITGB3 to further overexpress integrin β3 and found out that overexpression of integrin β3 conferred resistance to osimertinib in H1975 cells as expected (Supplementary Fig. S4E). Taken together, these results indicated an essential role of integrin β3 in EGFR TKI resistance.

Next, we sought to determine whether antagonizing integrin β3 prevents or delays the emergence of acquired resistance. Low confluence cells (200–500/well) were seeded and treated in a 96-well plate, and wells of 50% or greater confluence were scored as positive weekly (15). First, we determined the efficacy of lentivirus shITGB3 by examining pFAK, pAkt, and integrin β3 levels by Western blot. As shown in Supplementary Fig. S3D, cells transfected with shITGB3 showed dramatically decreased integrin β3 levels without any significant changes in other integrin subunits tested and decreased pFAK and pAkt levels. Then, HCC827 or PC9 cells were transfected with lentivirus shITGB3 (Fig. 2E; Supplementary Fig. S5A) or treated with c(RGDfK) (Fig. 2F; Supplementary Fig. S5B) in the presence or absence of an EGFR TKI. We found that an EGFR TKI in combination with c(RGDfK) or shITGB3 significantly reduced the percentage of positive wells compared with single agents, that is, their combination delayed the emergence of acquired resistance. To extend our findings in vivo, we stably transfected HCC827GR cells with shITGB3 or shNC and established subcutaneous xenograft tumors with HCC827GR-shITGB3 or HCC827GR-shNC. Then, xenograft tumor-bearing nude mice were treated with gefitinib daily (12.5 mg/kg). As shown in Fig. 2G and H, in the first 3 weeks of gefitinib treatment, shITGB3 tumors showed more shrinkage compared with shNC tumors, demonstrating that antagonizing integrin β3 could increase the sensitivity to TKIs in vivo. In addition, we observed that HCC827GR-shITGB3 tumors showed a slower regrowth starting from the fourth week of gefitinib treatment, suggesting that integrin β3 knockdown delayed the development of further resistance to gefitinib, even though HCC827GR cells were resistant to gefitinib in vitro.

As shown in Fig. 2A–D and Supplementary Fig. S4, in the absence of a TKI, c(RGDfK) and integrin β3 siRNA decreased cell viability and inhibited cell growth in resistant cells, while integrin β3 overexpression slightly increased cell viability in sensitive cells. To further determine whether integrin β3 regulated cell proliferation, we manipulated integrin β3 pharmacologically or genetically and then assessed the effects on proliferation using EdU incorporation and IncuCyte proliferation assays. As shown in Fig. 3A and B and Supplementary Fig. S6A–I, c(RGDfK) and integrin β3 siRNA inhibited proliferation of high integrin β3-expressing cells, such as HCC827OR, HCC827GR, PC9OR, and H1975, while overexpression of integrin β3 promoted the proliferation of low integrin β3-expressing cells, such as HCC827 and PC9 cells. We also compared the growth rate of xenograft tumors established from HCC827GR-shNC or HCC827GR-shITGB3 cells and found that HCC827GR-shITGB3 tumors showed significantly slower growth than HCC827GR-shNC tumors (Fig. 3C and D). These results demonstrated that integrin β3 confers a survival advantage to lung cancer cells. Next, we studied the roles of integrin β3 in anoikis, which is caspase-dependent apoptosis induced by cell detachment. Tumor cells were seeded on ultra-low attachment plates to prevent cell attachment and then collected after 48 to 72 hours. Apoptosis was assessed by Western blot analysis of apoptosis markers (caspase-3 and PARP) and cell viability assays using the trypan blue exclusion test. As shown in Fig. 3E, Western blotting showed that HCC827OR cells were more resistant to anoikis than HCC827 cells, as evidenced by lower levels of cleaved caspase-3 and cleaved PARP in HCC827OR compared with HCC827, when both of them were cultured in ultra-low attachment plates. Furthermore, c(RGDfK) and integrin β3 siRNA promoted anoikis by increasing c-caspase-3 and c-PARP (Fig. 3F; Supplementary Fig. S6K and S6L) in HCC827OR, HCC827GR, PC9OR, and H1975 cells and decreasing the number of viable HCC827OR and PC9OR cells (Fig. 3G; Supplementary Fig. S6M). On the contrary, overexpressing integrin β3 in HCC827 cells suppressed anoikis by decreasing c-caspase-3 and c-PARP (Supplementary Fig. S6J). Taken together, these results demonstrated that integrin β3 confers anoikis resistance to tumor cells.

Next, we sought to determine the roles of integrin β3 in the acquisition of the EMT phenotype associated with EGFR TKI resistance in mutant-EGFR NSCLC. First, we compared the expression of EMT markers between parental and OR cells. Consistent with previous reports that gefitinib-resistant cells and tumors acquired an EMT phenotype (12), in this study, Western blot analysis showed that the expression of mesenchymal markers, such as vimentin, Snail, Slug, and Zeb1, was dramatically increased, and that of E-cadherin was decreased in OR cells, indicating the acquisition of an EMT phenotype in OR cells (Fig. 4A). Then, we manipulated integrin β3 pharmacologically or genetically and assessed the effects on the expression of EMT markers using Western blot assays and in vitro migration and invasion abilities using transwell assays. As shown in Fig. 4B–F and Supplementary Fig. S7, c(RGDfK) and integrin β3 siRNA reversed EMT and suppressed in vitro migration and invasion abilities in high integrin β3-expressing cells such as HCC827OR, HCC827GR, PC9OR, and H1975 cells, while overexpressing integrin β3 promoted EMT and increased in vitro migration and invasion abilities in low integrin β3-expressing cells such as HCC827 and PC9 cells. These results demonstrated that overexpression of integrin β3 contributed to the EMT phenotype acquired in resistant cells.

Increasing evidence has suggested a relationship between EMT, drug resistance, and cancer stemness (16). Therefore, we first examined cancer stem cell (CSC) traits in resistant cells by comparing the expression levels of CSC markers and colony formation abilities between parental and their respective resistant cells. As shown in Fig. 5A, CSC markers, such as CD44, CD133, Nanog, Sox2, Oct4, and ALDH1, were dramatically upregulated in both GR and OR cells. Moreover, the colony formation assay showed higher numbers of both HCC827GR and HCC827OR cell colonies than HCC827 cell colonies (Fig. 5B). Furthermore, in acquired gefitinib-resistant xenograft tumors established by HCC827 or PC9 cells, the CSC marker CD44 was also dramatically overexpressed compared with that in the respective sensitive tumors (Fig. 5C). Then, we manipulated integrin β3 pharmacologically or genetically and assessed the effects on the expression levels of CD44 and colony formation abilities. As shown in Fig. 5D and E, c(RGDfK) and siITGB3/shITGB3 decreased CD44 expression and colony formation ability in high integrin β3-expressing cells, including HCC827OR, HCC827GR, PC9OR, and H1975 cells, while overexpressing integrin β3 increased CD44 expression and colony formation ability in low integrin β3-expressing HCC827 cells, suggesting that overexpression of integrin β3 confers CSC traits to lung cancer cells.

Lastly, we asked whether resistant cells or tumors express high levels of integrin β3 due to cytokines and growth factors present in the tumor microenvironment. TGFβ1 is known to stimulate integrin expression in various human cancers (17). Furthermore, in breast cancer and non–small cell lung cancers, TGFβ1 has been reported to specifically upregulate integrin β3 compared with other integrin subunits (18, 19). We examined the TGFβ1 levels in resistant cells and tumors and found that they were significantly higher in both resistant cells and tumors than in their respective sensitive controls (Fig. 6A–D). Moreover, TGFβ1 was gradually upregulated in HCC827 cells treated with gefitinib or osimertinib for up to 72 hours (Fig. 6E–H). To determine whether TGFβ1 induces integrin β3, we first treated HCC827 cells with TGFβ1 in the presence or absence of the TGFβRI inhibitor SB-431542 (20). As shown in Fig. 6I, TGFβ1 induced integrin β3 expression at both the mRNA and protein levels and activated downstream FAK signaling, both of which were abrogated by pretreatment with SB-431542. Then, the promoter region of ITGB3 gene was cloned and inserted into the upstream of luciferase gene of pGL3-basic vector to construct pGL3-ITGB3. We transfected HEK293 cells with pGL3-ITGB3 and then treated cells with TGFβ1 and found that relative luciferase activity was increased about 2-fold (Fig. 6J), which suggesting TGFβ1 signaling increased transcription of luciferase gene by enhancing ITGB3 promoter activity. Moreover, in HCC827 cells, osimertinib-induced integrin β3 expression and FAK activation were alleviated by pretreatment with SB-431542 (Fig. 6K, top). As shown in Fig. 6K (bottom), in osimertinib-resistant, high integrin β3-expressing HCC827OR cells, although osimertinib treatment did not further induce integrin β3 expression and FAK activation, SB-431542 pretreatment also decreased the levels of integrin β3 and pFAK. These results demonstrated that increased TGFβ1 levels in resistant cells and tumors were induced by TKI treatment and mechanistically responsible for integrin β3 overexpression. Moreover, we examined other integrin subunits and found out that integrin β1 was also induced by both TGFβ1 and osimertinib (Supplementary Fig. S8). The induction of integrin β1 by TGFβ1 and osimertinib was both abrogated by pretreatment with SB-431542 (Supplementary Fig. S8). However, the underlying mechanisms why integrin β1 was not upregulated in resistant cells and tumors need further investigation.

Altered expression of specific integrin subunits or integrin heterodimers has been linked to many types of tumors (6, 7). However, many studies have reported contradictory results for the same integrin molecules, even within the same cancer type (7). These discrepancies may be due to many factors, including fluctuation of integrin expression at different stages of tumor progression and heterogeneity in patient samples. Here, we establish in vitro and in vivo models of EGFR-mutant lung cancer with acquired EGFR TKI resistance and identify that integrin β3 is dramatically and consistently overexpressed in resistant cells and tumors. Moreover, the overexpression of integrin β3 is involved in proliferation, anoikis resistance, cancer stemness, and EMT. Importantly, we observed increased expression of TGFβ1 in resistant cells or tumors that is responsible for integrin β3 induction.

The role of integrin β3 in acquired EGFR TKI resistance is consistent with previous reports connecting integrin to the development of EGFR inhibition resistance (8–11). Kanda and colleagues found that in erlotinib-resistant lung cancer cell lines, there was increased expression of integrin β1, α2, and α5, along with activation of Src/Akt signaling (9). In this study, we examined the expression levels of the most common integrin subunits, including integrin β1 and α5, and found that integrin β3 is the only one that is dramatically and consistently overexpressed in both resistant cells and tumors, while neither integrin β1 nor α5 is. This could be due to different cell lines and drug treatment regimens. Seguin and colleagues found that in erlotinib-resistant lung cancer xenograft tumors and patients, integrin β3 expression was significantly upregulated, which is consistent with our results (8). In addition, we observed that during the development of acquired EGFR TKI resistance, EGFR-mutant lung cancer acquired anoikis resistance, EMT phenotype, and cancer stemness, all of which are associated with integrin β3 overexpression. As such, integrin β3 is considered to be an attractive target in advanced lung cancer. Indeed, our results showed that antagonizing integrin β3 pharmacologically or genetically increases the sensitivity of resistant lung cancer to EGFR TKIs and delays the occurrence of acquired resistance in vitro and in vivo. Antagonizing integrin β3 also inhibits resistant cell growth, promotes anoikis resistance, reverses EMT phenotype and suppresses cancer stemness. In contrast, overexpressing integrin β3 confers resistance to EGFR TKIs and promotes cell growth, anoikis resistance, EMT phenotype, and cancer stemness. Taken together, our results indicate that integrin β3 is a promising therapeutic target in advanced lung cancer.

Unfortunately, clinical trials targeting integrins, such as cilengitide and abituzumab, failed to show promising outcomes (21–23). Most current therapeutic strategies have been designed to interfere with integrin–ligand interactions. Considering ECM ligand-independent integrin signaling (8, 24, 25), it would be promising to exploit tumor-specific integrin expression. Our study shows that integrin β3 is the only one of the most common integrin subunits that is dramatically and consistently overexpressed in both EGFR-mutant lung cancer cells and tumors with acquired EGFR TKI resistance. High expression of integrin β3 in lung adenocarcinoma patients is also associated with poor survival. Therefore, to prevent or delay EGFR TKI resistance and tumor progression, selectively targeting integrin β3-overexpressing tumor cells would be a promising strategy (26, 27).

Notably, switching between integrin subunits at different stages of tumor progression could be one mechanism by which tumors evade therapy and the reason for the lack of efficacy of specific integrin inhibitors (28–30). Therefore, the other straightforward approach to target integrin β3-overexpressing tumor cells is to inhibit EGFR TKI-induced integrin β3 overexpression. However, inhibiting integrin β3 overexpression requires a better understanding of how this gene is induced during the development of resistance to EGFR TKIs, which is not yet clear. Previous studies have established that integrin β3 expression can be induced in tumor cells by cytokines and growth factors present in the tumor microenvironment, especially TGFβ1 (18, 19). Indeed, we first found that in both resistant cells and tumors, TGFβ1 is increased. In addition, TGFβ1 is induced by EGFR TKI treatment in a time-dependent manner in HCC827 cells. We found that exogenous TGFβ1 can induce integrin β3 expression through activating ITGB3 promoter. Moreover, both TGFβ1- and osimertinib-induced integrin β3 overexpression can be reversed by the TGFβ1 inhibitor SB-431542. Therefore, we propose that in advanced EGFR-mutant lung cancer treated with EGFR TKIs, TGFβ1 is increased, resulting in integrin β3 overexpression and activation of downstream FAK signaling, which is a key component of the signal transduction pathways activated by integrins and has an essential role in cancer cell survival, EMT, metastasis, and stemness (31). Integrin β3 overexpression and FAK activation lead to decreased sensitivity to TKIs, enhanced anoikis resistance, EMT phenotypes, and cancer stemness. We reason that inhibitors targeting TGFβ1 might be combined with EGFR TKIs to prevent or delay the occurrence of acquired resistance and progression of lung cancer, which needs further investigation. We also found that integrin β1 was induced by both exogenous TGFβ1 and short-term osimertinib treatments. However, the reason why integrin β1 was not upregulated in acquired resistant cells or tumors is not yet clear and needs further investigation.

In this study, we define a role for integrin β3 in acquired EGFR TKI resistance and tumor progression in lung cancer. Targeting integrin β3 pharmacologically or genetically could be able to delay lung cancer progression. Moreover, our results demonstrate that integrin β3 overexpression is due to increased TGFβ1 levels in acquired EGFR TKI-resistant lung cancer, and provide a rationale for the inhibition of TGFβ1/integrin β3 signaling as a therapeutic target to circumvent EGFR TKI resistance and delay lung cancer progression.

In conclusion, we found that integrin β3 was dramatically and consistently overexpressed in EGFR-mutant lung cancer cells and xenograft tumors with acquired resistance to EGFR TKIs, and also involved in the progression of lung adenocarcinoma. Antagonizing integrin β3 genetically or pharmacologically was able to delay lung cancer progression. Mechanistically, integrin β3 was induced by increased TGFβ1 in acquired TKI-resistant lung cancer. Our findings identified the TGFβ1/integrin β3 axis as a promising target for combination therapy in advanced lung cancer.

No potential conflicts of interest were disclosed.

Conception and design: C. Wang, D. Lv, H.-Z. Chen, L. Xu

Development of methodology: C. Wang, D. Lv, J. Yue, L. Xu

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): C. Wang, T. Wang, D. Lv, L. Li, J. Yue, L. Xu

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): C. Wang, T. Wang, D. Lv, L. Li, L. Xu

Writing, review, and/or revision of the manuscript: C. Wang, H.-Z. Chen, L. Xu

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): L. Xu

Study supervision: H.-Z. Chen, L. Xu

We thank Dr. Qianggang Dong for providing PC9 cells. The work was supported by the National Natural Science Foundation of China (81773747 and 81372522 to L. Xu) and Science and Technology Commission of Shanghai Municipality (12ZR1416000 and 12140901400 to L. Xu).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.