Icotinib-resistant HCC827 cells produce exosomes with mRNA MET oncogenes and mediate the migration and invasion of NSCLC

A total of 10 NSCLC patients with EGFR 19del mutation, who were in the Affiliated Hospital of Ningbo Medical School of Ningbo University (Ningbo, China) during the period of August 2017 and December 2018, were included into the present study. All patients have been primarily diagnosed in the above-mentioned hospital. The clinical specimens, including serum and bronchoalveolar lavage fluid (BALF), were collected at the time of primary diagnosis and after the treatment with icotinib within a follow-up period of 3–6 months. The clinical characteristics of these patients are presented in Additional file 1: Table S1. All procedures were approved by the Ethics Committee of the Affiliated Hospital of Ningbo Medical School of Ningbo University (Ningbo, China), and each patient provided an informed consent before the specimens were collected.

The human NSCLC cell line HCC827, which was sensitive to icotinib and contained an EGFR exon 19 deletion (DelE746-A750), and the human normal pulmonary epithelial cell line BEAS-2B were purchased from Nanjing Cobioer Biological Science (Nanjing, China). The HCC827IR cell lines (HCC827IR1 and HCC827IR2) were generated by repeated exposure of HCC827 cells to gradually increased concentrations of icotinib (Dalian Meilun Biotechnology Co., Ltd., China) for over six months and HCC827IR-1 clones were selected for subsequent experiments and referred to as HCC827IR. The HCC827IR cells were cultured in RPMI-1640 medium (Gibco, USA) supplemented with 10% fetal bovine serum (Gibco, USA), penicillin (100 U/mL) and streptomycin (100 μg/mL). Pulmonary epithelial cell lines BEAS-2B were cultured in BEBM complete medium (Nanjing Cobioer Biological Science, China). All cell lines were maintained in a humidified incubator at 37 °C with 5% CO₂.

The HCC827 and HCC827IR cell lines were cultured in media with 10% exosome-free FBS (by ultracentrifugation for 12 h). After 48 h, the cell culture media was collected, and the exosomes were isolated from the cell supernatant by differential centrifugation, as previously described [14]. Finally, the concentration of the exosomal protein was determined using a BCA protein assay kit (Thermo Scientific, USA). Then, CD9, CD63 and CD81 (Cell Signaling Technology, Beverly, MA, USA) expression was measured using western blot analysis. The aliquots were stored at − 80 °C. The extracted exosomes and pellets were sent to Hibio Technology Co., Ltd. (Hangzhou, China) for transmission electron microscope (TEM) observation and validation, and the size distribution analysis. Thus, these exosomes were prepared for protein/RNA extraction, cell treatment, etc.

This assay was performed to verify the internalization of the labeled HCC827IR-derived exosome through HCC827 cells. First, the HCC827IR-exosomes were re-suspended in 500 ul of PBS in a 1.5 ml microcentrifuge tube (Eppendorf, EP), and DiR iodide (Dalian Meilun Biotechnology Co. Ltd., China) was added to the tube with the HCC827IR exosome up to a final concentration of 5 μg/ml. Then, the mixture was incubated at 37 °C for 30 min without shaking. Afterwards, the EP tube was centrifuged at 1000 rpm for three minutes, and the supernatant was carefully filtered with a 0.22-μm filter. Subsequently, the HCC827IR-Exosome-DiR liquid was co-cultured with HCC827 cells for 24 h. Finally, these cells were observed under a fluorescence microscope.

Cell activity was determined using the MTT assay. Cancer cells were seeded on 96-well plates at a density of 1 × 10⁵ in each well. After 24 h, these cells were treated with different concentrations (0, 2, 4, 8, 16 and 32 uM) of icotinib for 48 h. Then, a 15-μl MTT solution (0.5%) was added to the medium, and incubated for four hours at 37 °C. Afterwards, the medium was carefully removed, 150 μl of dimethyl sulfoxide (DMSO) was added to each well to dissolve the insoluble formazan product, and the absorbance of the colored solution was measured at 490 nm, calibration read at 630 nm using a microplate reader (MK3, Thermo, USA). Next, the background absorbance of the medium in the absence of cells was subtracted. All experiments were independently performed in quintuplicate, and the mean for the experiment was calculated.

The total proteins of cells and isolated exosomes were lysed with RIPA buffer supplemented with proteinase inhibitors (Beyotime Biotechnology, China), according to manufacturer’s protocols, and centrifuged at 14,000×g for 10 min at 4 °C. Then, the protein concentrations were determined using BCA protein assay kit. The protein(25μg of protein isolated from cells and 35μg of protein isolated from exosomes) were separated using 10–12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and transferred onto polyvinylidene difluoride (PVDF) membranes. The membranes were blocked with 5% skim milk within TBST for one hour at room temperature, and incubated with the primary antibodies at 4 °C overnight. The primary antibodies used for the western blot were as follows: anti-CD9 (1:1000; CST, USA), anti-CD63 (1:1000; CST, USA), anti-CD81 (1:1000; CST, USA), anti-GAPDH (1:3000; Bios, China), anti-EGFR (1:1000; CST, USA), anti-p-EGFR(1:1000; CST, USA), anti-AKT (1:1000; CST, USA), and anti-alpha-Actinin (1:1000; CST, USA). After incubation with the appropriate horseradish peroxidase-conjugated secondary antibodies, the immune-reactive protein bands were visualized with chemiluminescence reagents (CST, USA), followed by imaging on an electrophoresis gel imaging analysis system (D-Digital, USA).

Cells were treated with exosomes (20 μg/ml) or culture medium without exosomes for the appointed time periods. For the cell invasion and migration assay, about 4 × 10⁴ cells were plated into the upper chamber that were pre-coated with or without Matrigel (Matrigel, BD, USA) The culture medium in the upper chamber was FBS-free RPMI 1640, while the medium in the lower chambers was the RPMI 1640 supplemented with 10% exosome-free FBS. After 24 h, all cells that had transferred to the lower chambers were fixed and stained with 0.5% crystal violet. Then, positive staining cells from six representative fields of chambers in each group were photographed and counted under a microscope (Nikon, Japan). For the wound-healing assay, equal numbers of cells pre-treated with or without exosomes were plated into six-well plates. Then, the cell monolayers were wounded with a pipette tip to draw a gap on the plates. HCC827 cells that migrated into the cleared section were observed under a microscope (Nikon, Japan) after 24 and 48 h.

Equal numbers of cancer cells were seeded and cultured in the bottom of 6-well plates overnight. Then, these cells were co-cultured with HCC827 or HCC827IR cells. After 48 h, cells at the bottom were collected and re-suspended in pre-cold PBS. Afterwards, the re-suspended cells were fixed with 70% cold alcohol for one hour, the suspension was centrifuged, and the cell particles were cultured in PI/RNase staining buffer for 15 min (BD Biosciences, CA, USA). Subsequently, the flow analysis of the stained cells was performed using a C6 flow cytometer (BD Biosciences).

Cells were seeded in six-well plates and cultured for 24 h before being fixed by 4% paraformaldehyde, and incubated with specific primary antibodies vimentin (1:100; AF1975, Beyotime) and cytokeratin-7 (1:100; AF1822, Beyotime, China) at room temperature for one hour. Then, these were incubated in the dark with secondary antibodies against vimentin and cytokeratin-7. The staining was developed using a fluorescence detection system (Beyotime, China). The samples were counterstained with 4′,6-diamidino-2-phenylindole (DAPI, Invitrogen) to visualize the cell nuclei. After washing, representative images were examined under a fluorescence microscope (Leica, FL, USA).

Total RNA was extracted from cultured cells and purified exosomes using TRIzolTM reagent (Life Technologies, USA), according to the manufacturer’s instructions. The RNA was resuspended in RNase-free DEPC-water and stored at − 80 °C. An equal amount of RNA was used for reverse transcription. The cDNAs were synthesized by using a reverse transcription kit, according to manufacturer’s instructions (CWBio, Beijing, China). qRT-PCR for cellular/exosomal mRNA, including CDK1 (Cyclin-dependent kinase 1), CDK2 (Cyclin-dependent kinase 2), CDK4 (Cyclin-dependent kinase 4), CDH1 (Cadherin 1), ICAM1 (Intercellular cell adhesion molecule-1), ITGB1 (Integrin beta-1), ITGB3 (Integrin beta-3), CDC42 (Cell division control gene-42), GAS6 (Growth arrest-specific 6) and MET (hepatocyte growth factor receptor), as well as internal reference GAPDH (glyceraldehyde-3-phosphate dehydrogenase) were performed using RT-PCR Quantitation Kit (CWBio, Beijing, China) according the manufacturer’s instructions. Briefly, after an initial denaturation step at 95 °C for 10 min, the amplifications were carried out with 40 cycles at a melting temperature of 95 °C for 15 s, and an annealing temperature of 60 °C for 35 s. The relative expression levels of mRNAs were calculated with 2^–ΔCt method. PCR productions were tested by 2% agarose. The sequences of the specific primers are presented in Additional file 1: Table S1.

Next, Mixed 7.0 μg of IR exosomes with 0.33 μg of MET siRNA in a citrate buffer solution (prepared with DEPC water), with a final volume of 150 μl. The process of electroporation was performed according to the instructions provided by Scientz-2C (Scientz, Ningbo, China). Size exclusion chromatography (SEC) was performed to wipe off the redundant siRNA. Then, quantitative reverse transcription–polymerase chain reaction (RT–qPCR) were performed to detect the level of MET in IR exosomes.

The statistical analysis was performed using the SPSS software (version 20.0; IBM Corp., Armonk, NY, USA) and GraphPad Prism version 6.0 (GraphPad). Comparisons between pairs were performed using Student’s t-test, while multiple comparisons between the groups were analyzed using one-way analysis of variance, followed by Student Newman-Keuls test. All experiments were performed in triplicate, and the results were presented as the mean ± standard deviation. P < 0.05 was considered statistically significant.

Human NSCLC HCC827 cells with the deletion of exon19 of EGFR was used to establish the new model of acquired resistance to icotinib. These HCC827 cells were initially sensitive to icotinib treatment (the half inhibitory concentration [IC₅₀] of 8.1 μM), and were cultured in the presence of icotinib, with an indicated concentration gradients. The HCC827IR clonal sublines were generated (HCC827IR1 and HCC827IR2) with icotinib IC₅₀ > 80 uM through prolonged exposure time (for > 6 months) (Fig. 1a). In this study, the expression of vimentin and cytokeratin-7 was determined using immunofluorescence imaging, and it was found that fluorescence expression of vimentin was stronger in HCC827 IR cells, when compared with parental cells (HCC827) (Fig. 1b). For the subsequent experiments, HCC827IR cells were intermittently treated with 50 uM of icotinib to sustain its drug resistance feature. In order to determine whether these HCC827 cells sustained the activation and phosphorylation of EGFR after icotinib treatment, the phosphorylation status of EGFR was assessed. In cells treated with a certain concentration gradient of icotinib for 48 h, phosphorylation of EGFR was reduced with the increasing level of icotinib concentration (Fig. 1c). The western blot analysis confirmed that the activities of p-EGFR was inhibited by icotinib in a dose-dependent manner. Furthermore, in the context of acquired resistance to icotinib, the expression of phosphorylated EGFR in the HCC827IR cell line was detected by western blot. It was found that the expression of phosphorylated EGFR was not reduced with the increasing concentration level of icotinib, suggesting that phosphorylated EGFR was probably no longer the icotinib target in the HCC827IR model (Fig. 1c).

According to the microarray result from previous studies [15], three mRNAs (AXL, GAS6 and MET) correlated to EMT were chosen for detection by qRT-PCR, and to whether these were differentially regulated in HCC827IR cells, when compared with parental HCC827 cells (P < 0.05). It was observed that the receptor tyrosine kinase AXL and its ligand GAS6 were highly expressed in cells with acquired icotinib-resistance, while the expression of MET was the same both in parental cells HCC827 and HCC827IR cells (Fig. 1d).

In order to determine whether the acquired icotinib-resistant cells have a potential effect on parental cells, HCC827 cells were co-cultured with HCC827IR cells, and HCC827 cells cultured in normal culture medium were used as a control group, and placed in a 6-well plate for 48 h, as previously described. The Transwell assay further confirmed that more HCC827 cells migrated in the HCC827IR affected group (the cell number of HCC827/HCC827IR was 29.25 ± 1.51 vs. those of HCC827/HCC827 was 10.61 ± 3.81)(P < 0.0001)(Fig. 2a). The wound-healing assay revealed that the HCC827 cells (bottom of the plate), which were co-cultured with HCC827IR, expressed a higher level of migration ability, when compared to the control group in 24 h (P = 0.0317) (Fig. 2b). Nevertheless, the cell cycle and cell apoptosis determined by flow cytometry revealed no significant difference between cells co-cultured with HCC827IR cells and the control group (Fig. 2c).

Since tumor-derived exosomes play a significant role in cancer progression [16], and combined with the above result, it was determined whether the exosomes from HCC827IR cells could affect the biological function of icotinib-sensitive lung cancer cells. The exosomes were isolated from the supernatants of HCC827IR cells, which were previously cultured for 48 h in exosome-free medium through ultracentrifugation, and kept the exosome-depleted supernatant (EDS) as the control group for subsequent experiments. The exosomes presented as rounded particles with a size of approximately 60–150 nm (Fig. 3a), and a double layer membrane, as determined by TEM (Fig. 3b). The isolated exosomes were confirmed using western blot analysis with exosome specific-markers CD9, CD63 and CD81 (Fig. 3c). The DiR-labeled exosomes were incubated with HCC827 cells for 24 h, and the localization of these exosomes was assessed by fluorescence microscopy. Then, the DiR-labeled exosomes were internalized in the cytoplasm of HCC827 cells (Fig. 3d). The result suggested that exosomes derived from HCC827IR cells were taken up by HCC827 cells.

In order to determine whether HCC827IR cell-derived exosomes could regulate the biological function of lung cancer cells, HCC827 cells were co-cultured with exosomes from IR cells and the EDS for 48 h, respectively. The EDS was used as a control group. The MTT assay revealed no significant differences between the treated IR exosomes and EDS on HCC827 cell proliferation (with IC50 9.2 uM, Fig. 4a). The western blot assay used to detect the expression of p-EGFR after HCC827IR-derived treatment, and the expression of p-EGFR was found not to be correlated to the icotinib concentration gradient. This shows that p-EGFR does not clearly inhibit after treatment with IR exosomes, when compared to the icotinib alone treated group (Fig. 1c). Hence, the changes in mRNA expression in IR exosomes treated HCC827 cells were further determined, which were correlated to the cell cycle distribution, using qRT-PCR analysis. The results revealed that the expression of CDK2 and CDK4 was significantly higher in cells treated with HCC827IR, when compared to the control group(P < 0.05). However, genes correlated to cell cycle arrest, such as CDKN1A and CDKIB, were also upregulated in cells treated with HCC827IR(P < 0.001), suggesting that IR-derived exosomes might not influence the cell cycle distribution of HCC827 cells (Fig. 4b). Furthermore, the Transwell assay revealed that the group treated with HCC827IR-derived exosomes exhibited a higher ability of migration and invasion, when compared to the EDS group (P < 0.001) (Fig. 4c). In order to determine the relative mRNAs loaded by exosomes involved in the increasing ability of invasion and migration of HCC827 cells, 10 candidate mRNAs relating to cell proliferation, migration and EMT were chosen to for the PCR. The results revealed that almost no free-RNA was expressed in the icotinib-resistant EDS, and most of the candidate mRNAs were loaded both in HCC827-drived exosomes and HCC827IR-drived exosomes. However, MET exhibited an existing band only in icotinib-resistant exosomes based on the gel electrophoresis, after staining with ethidium bromide (Fig. 4d), implying that exosomal MET probably plays a significant role in accelerating HCC827 migration and EMT.

In order to evaluate the effects of exosomal MET on HCC827 cells, MET was knocked down in HCC827IR-derived exosomes by electroporation. The knockdown efficiency of siRNA-MET in exosomes was confirmed by qRT-PCR (Additional file). Then, HCC827 cells were treated with MET−/− icotinib-resistant exosomes, and the Transwell assay (with or without Matrigel) revealed that the migration and invasion ability of HCC827 cells incubated with MET−/− icotinib-resistant exosomes significantly decreased, when compared to the corresponding negative control (Fig. 4c). In addition, several mRNAs correlated to cell migration and EMT were analyzed by qRT-PCR. The results revealed a higher expression of ICAM1, ITGB1, CDC42, GAS6 and MET in cells, which were co-cultured with HCC827IR exosomes and MET−/− icotinib-resistant exosomes (Fig. 5a). Meanwhile, the western blot analysis revealed that the expression of alpha-actinin 4 was upregulated after the treatment of icotinib-resistant exosomes, but this significantly declined after the treatment of MET−/− icotinib-resistant exosomes, implying that the exo-MET might promote HCC827 migration and invasion by upregulating alpha-actinin 4 (Fig. 5b). This result further confirms that, to some extent, the icotinib-resistant-derived exo-MET might attribute to the migration of lung cancer cells.

In order to further evaluate the role of exosomes on normal lung cells, BEAS-2B lung epithelial cells were co-cultured with HCC827 cells, HCC827IR cells and HCC827-derived exosomes for 24 and 48 h, respectively, and BEAS-2B cells were cultured in normal culture medium as the control group. Initially, the changes in BEAS-2B cell morphology was observed under microscopy after culturing for 48 h (Fig. 6a). In all three groups, some of the BEAS-2B cells were observed to elongate, but marker identification of fibrosis was not performed. However, some lacuna appeared only in the intercellular space of BEAS-2B, which co-cultured with icotinib-resistant and icotinib-resistant exosomes cells (Fig. 6a). Afterwards, the protein expression was analyzed by western blot. It is noteworthy that the phosphorylation status of EGFR in BEAS-2B markedly increased in all three groups. In addition, the expression at 48 h was higher than that at 24 h. Meanwhile, in the HCC827 co-cultured group, the AKT level increased with the extension of co-culture time. However, the AKT level was significantly up-regulated in the HCC827IR and icotinib-resistant exosomes co-cultured group (b).

The clinical characteristics of NSCLC patients (n = 10) are presented in Table 1. These 10 patients were similar in pathological diagnosis, but different in neoplasm stage, and treated with icotinib (except for two patients who were treated with gefitinib). The average follow-up period was approximately 3–6 months. In order to investigate the expression of the tumor cell-cycle and migration-related genes in plasma exosomes from patients with NSCLC, 10 different genes were detected, which were mentioned above, in order to compare the expression at different time points after icotinib treatment. According to PCR and gel electrophoresis analysis, it was found that the expression of 10 candidate mRNAs was the same in plasma exosomes as it was in BALF-derived exosomes (Fig. 7a), and MET was found to be differentially presented, with correspondent bands in four of 10 patients after treatment with icotinib for three or more months. However, these were not exhibited in lung cancer patient plasma exosomes at the previous three months (Fig. 7b). Furthermore, combined with the CT scan results, these four patients resulted in not only no efficiency of icotinib treatment, but also presented with cancer metastasis (Fig. 7b). The expression level of 10 candidate mRNAs (Table 2) in the exosomes of plasma collected from NSCLC patients at the time of primary diagnosis and the second check during the 3–6 months follow-up period was assessed. A score of 1 point represents the expression of one mRNA. Based on this scoring criteria, the difference value of 10 mRNA expression at two different time points was calculated before and after icotinib treatment, respectively. Then, these were further calculated using the formula (log₁₀2^{Score 2nd-Score 1st}) to evaluate the correlation of this formula and cancer progression in patients. The sensitivity and specificity of the formula were analyzed for the assessment of lung cancer cell metastasis through the receiver operating characteristic (ROC) curve analysis. This analysis revealed that the area under the curve (AUC) was 0.98, two cutoff value were − 0.7526 and − 0.4515, respectively, meaning the sensitivity and specificity of formula were significantly higher(p = 0.012), suggesting that the score of this formula could be used as a potential indicator for icotinib resistance in NSCLC treatment (c). Meanwhile, we used progression-free survival curve to calculate this result of formula, suggesting that the score was positively correlated with cancer progression (Median was 4, P-value = 0.0399, Fig. 7d).