The human NSCLC cell lines HCC827 and HCC4006, which harbor EGFR activating mutations (16,17), were purchased from the Chinese Academy of Sciences (Shanghai, China). Both cell lines were cultured in RPMI-1640 medium (BioWhittaker^®; Lonza Group, Ltd., Basel, Switzerland) supplemented with 10 mM HEPES, 1 mM L-glutamine, 100 U/ml penicillin/streptomycin (BioWhittaker^®; Lonza Group) and heat inactivated 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.) and grown at 37°C in a 5% CO₂ atmosphere. Gefitinib (Iressa; AstraZeneca, Macclesfield, UK) was dissolved in dimethyl sulfoxide (DMSO; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) at a concentration of 10 mM and stored at −20°C. Gefitinib-resistant HCC827R and HCC4006R cells were established by initially culturing with 1 µM gefitinib in DMEM plus 10% FBS for 6 weeks. Subsequently, a 2-µM concentration of gefitinib was used to treat the surviving cells for 8 weeks and 5 µM for another 8 weeks. Eventually, the gefitinib-resistant NSCLC cell lines were successfully established by culturing the cells in 10 µM gefitinib.

Exosomes were extracted from culture medium using ExoQuick precipitation kit (System Biosciences, Mountain View, CA, USA) according to manufacturer's instructions. Briefly, the culture medium was thawed on ice and centrifuged at 3,000 × g for 15 min to remove cells and cell debris. Next, 250 µl of the supernatant was mixed with 63 µl of ExoQuick precipitation kit and then incubated for 40 min at 5°C after brief shaking and mixing, followed by centrifugation at 1,500 × g for 30 min. Then, the supernatant was removed by careful aspiration, followed by another 5 min of centrifugation to remove the residual liquid. The exosome-containing pellet was subsequently re-suspended in 250 µl phosphate-buffered saline (PBS). The final pellets, containing exosomes, were collected for characterization and RNA isolations. Size distribution of exosomes was analyzed by Zetasizer (Malvern Panalytical Ltd., Malvern, UK). Purified exosomes were labeled with PKH26 Red Fluorescent Cell Linker Kit for General Cell Membrane Labeling (Sigma-Aldrich; Merck KGaA) as per the manufacturer's protocol.

Extraction of RNA from exosomes was performed using the commercial miRNeasy Serum/Plasma kit (Qiagen Sciences Inc., Gaithersburg, MD, USA), and RNA extraction from cell fraction was performed using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. RNA elution steps were carried out at 12,000 × g for 15 sec, and the extracted RNA was dissolved in RNase-free ultra-pure water.

We used 50 µl PBS to suspend the exosomes pellets and then put one drop of this suspension on the parafilm. A copper mesh coated with carbon was then used to drift on the drop for 5 min at 25°C. Then, the grid was removed, and the excess liquid was drained by touching the grid edge against a piece of clean filter paper. The grid was then placed onto a drop of 2% phosphotungstic acid with pH 7.0 for approximately 5 sec, and the excess liquid was drained off. The grid was allowed to dry for several minutes and then examined using a JEM-1200 EX microscope (JEOL Ltd., Akishima, Japan) at 80-kiloelectron volts.

The cDNA was synthesized from 200 ng extracted total RNA using the PrimeScript RT reagent kit (Takara Biotechnology Co., Ltd., Dalian, China) and amplified by RT-qPCR with a SYBR Green Kit (Takara Biotechnology Co.) on an ABI PRISM 7500 Sequence Detection System (Life Technologies; Thermo Fisher Scientific, Inc.) with the housekeeping gene GAPDH as an internal control by using the ∆∆Cq method (18). The primer sequences are presented in Table I.

The small interfering RNA against H19 (si-H19) and hnRNPA2B1 (si-hnRNPA2B1) were synthesized and prepared by Shanghai GenePharma Co., Ltd. (Shanghai, China). Negative control siRNA was purchased from Invitrogen; Thermo Fisher Scientific, Inc. (Shanghai, China; cat. no. 12935-110). All the vectors were labeled with green fluorescence protein (GFP). Briefly, a total of 1.2×10⁷ cells were plated in 15-cm culture dishes for 24 h and then transfected with the vectors described above using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) for 24 h. The cells were then subjected to RNA/protein extraction and further functional assays. The sequences of small interfering RNAs are presented in Table I.

Gefitinib-induced nuclear apoptosis was evaluated by performing TUNEL analysis. In brief, cells were treated with si-H19#3 or negative control for 24 h and fixed by using 4% formaldehyde. Cells were fixed and stained with TUNEL kit according to the manufacturer's instructions (TUNEL Bright-Red Apoptosis Detection kit; cat. no. A113; Vazyme Biotech Co., Ltd., Nanjing, China). TUNEL-positive cells were counted under fluorescence microscopy (DMI4000B; Leica Microsystems GmbH, Wetzlar, Germany).

NSCLC cells were rinsed with cold PBS and fixed by 1% formaldehyde for 10 min. After centrifugation at 1,500 × g for 5 min at 4°C, cell pellets were collected and re-suspended in the NP-40 lysis buffer. For RIP assay, the supernatant was incubated overnight with beads conjugated with anti-hnRNPA2B1 antibody (1:50; cat. no. ab31645; Abcam, Cambridge UK) or negative control mouse IgG (cat. no. 12-371; EMD Millipore, Burlington, MA, USA). The beads were then rinsed with cold NT2 buffer and cultured with proteinase K at 10 mg/ml (Sigma-Aldrich; Merck KGaA).

Alterations in cell viability following transfection or gefitinib treatment was assayed using CCK-8 kit (Dojindo Molecular Technologies, Inc., Rockville, MD, USA). In brief, cells were seeded into a 96-well plate in triplicate and then treated with si-H19 or (and) gefitinib for different periods of time. The cell cultures were then treated with CCK-8 reagent and further cultured for 2 h. The optical density at 450 nm was measured using a spectrophotometer (Thermo Fisher Scientific, Inc.). The percentage of the control samples of each cell line was calculated thereafter.

Briefly, ~0.3 ml supernatant was loaded into the sample chamber of an LM10 Nano-sight unit (NanoSight Ltd., Amesbury, UK) and three videos of either 30 or 60 sec were recorded for each sample. Data analysis was performed with both NTA 2.1 software (NanoSight). The diffusion coefficient and sphere-equivalent hydrodynamic radius were then determined using the Stokes-Einstein equation, and results were displayed as a particle size distribution.

Cell lysates were prepared with RIPA buffer containing protease inhibitors (Sigma-Aldrich; Merck KGaA). Protein concentrations were assessed with the BCA Protein Assay kit according to the manufacturer's instructions (Beyotime Institute of Biotechnology, Shanghai, China). Equal amounts of protein (25 µg) were separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto polyvinylidene fluoride membranes (EMD Millipore). Then, the membrane was blocked with 5% (5 g/100 ml) non-fat dry milk in Tri-buffered saline plus Tween (TBS-T) buffer for 2 h at room temperature. The membranes were incubated overnight at 4°C with a 1:1,000 solution of antibodies: Anti-TSG101 (cat. no. ab125011), anti-hnRNPA2B1 (cat. no. ab31645), anti-CD63 (cat. no. ab134045) and anti-β-actin (cat. no. ab8226; all from Abcam). The horseradish peroxidase-conjugated (HRP) anti-rabbit antibody (1:5,000; cat. no. 7074; Cell Signaling Technology, Inc., Danvers, MA, USA) was used as a secondary antibody for immunostaining for 1 h at room temperature. The proteins were visualized using a detection system of enhanced chemiluminescence (ECL) by using Immobilon Western Chemiluminescent HRP Substrate (EMD Millipore).

Mann-Whitney U test was used for the comparison of datasets containing two groups. The Kruskal-Wallis test followed by post-hoc test with Bonferroni's was used for evaluating the difference among multiple groups. The survival curves of NSCLC cells were estimated via the Kaplan-Meier method, and the difference in survival rate was analyzed using the log-rank testing. Statistical analysis was performed using Prism 4 (GraphPad Software Inc., San Diego, CA, USA) and a P-value threshold of <0.05 was considered to indicate a statistically significant difference.

To investigate the underlying regulatory mechanism of gefitinib resistance, two gefitinib-resistant sub-lines derived from HCC827 and HCC4006 cell lines were constructed (HCC827R and HCC4006R, respectively). We found that the built gefitinib-resistant cells exhibited characteristic changes, including loss of cell polarity, increased intercellular separation, and increased formation of pseudopodia (Fig. 1A). CCK-8 assay revealed that the cell viability of HCC827R and HCC4006R cells was significantly increased when compared to respective parental cells under the treatment of gefitinib for 48 h (Fig. 1B). A dose-effect curve was built, and we identified that the IC₅₀ value of gefitinib for HCC827R cells was 8.75 µM, whereas HCC827 was 1.00 µM, meaning that the ability of resistance to gefitinib was 8.75 times higher for the HCC827R cells than that of the HCC827 cells. Similarly, the HCC4006R cell exhibited a 7.21 times higher resistance to gefitinib than that of the HCC4006 cells (7.43/1.03; Fig. 1C). Then, we determined the expression level of H19 in NSCLC cell lines by using RT-qPCR, and we identified that H19 was upregulated in most NSCLC cells in contrast to normal epithelial cells, 16HBE (Fig. 1D). In addition, H19 was upregulated in gefitinib-resistant cell lines when compared to the respective parental cells (Fig. 1E). This indicated that H19 may be important for gefitinib resistance of NSCLC cells.

With the verification of aberrant expression of H19 in gefitinib-resistant cells, we sought to determine whether H19 was involved in gefitinib resistance. Three small interfering RNAs against H19 were generated, and we found that H19 expression was mostly silenced in the HCC827R and HCC4006R cells when incubated with si-H19#3 (Fig. 2A), which was used for the following experiments. We transfected gefitinib-resistant cells with si-H19#3 (Fig. 2B). Compared with the response of the control group, silencing of H19 promoted gefitinib-induced cell cytotoxicity (Fig. 2C). FACS apoptosis assay revealed that gefitinib exposure caused an increased proportion of apoptotic cells in H19-knockdown cells in contrast to si-NC-transfected cells (Fig. 2D). Then, we used TUNEL assay to detect whether H19 influenced the nuclear apoptosis induced by gefitinib. We found that knockdown of H19 increased the gefitinib-induced nuclear apoptosis of HCC827R cells (Fig. 2E). Therefore, we demonstrated that H19 was essential for gefitinib resistance in NSCLC.

To investigate how H19 regulates gefitinib resistance, we firstly localized the expression of H19 in NSCLC cells. By using the online lncRNA location prediction software lncLocator (http://www.csbio.sjtu.edu.cn/bioinf/lncLocator/), we identified that intracellular H19 was located in exosomes (Table II). To verify whether H19 is secreted through packaging into exosomes, we detected the expression level change of extracellular H19 after treatment with RNase. As revealed in Fig. 3A, H19 in culture medium was little influenced by the treatment of RNase alone but significantly decreased when treated with RNase and Triton X-100 simultaneously, suggesting that extracellular H19 was protected by the membrane instead of being directly secreted. H19 expression levels in exosomes were almost equal to that in whole culture medium, indicating that extracellular H19 was contained in exosomes (Fig. 3B). We then purified and extracted exosomes from culture medium, and the representative micrograph captured by Transmission Electron Microscopy (TEM) was revealed in Fig. 3C. A similar morphology, size, and number were identified between HCC827R and HCC827 cells by NTA analysis (Fig. 3D). Western blot assays further confirmed their identity by enriched exosome proteins, such as TSG101 and CD63 (Fig. 3E). Then, we determined whether H19 was incorporated into exosomes by isolation of exosomes with the ExoQuick purification kit followed by qPCR. As anticipated, H19 was detectable in exosomes, and the expression level was significantly higher in gefitinib-resistant cells than in the respective parental cells (Fig. 3F), indicating that extracellular H19 was secreted through incorporation into exosomes in NSCLC cells.

Subsequently, we investigated whether the packaging of H19 into exosomes was mediated by specific regulator. Previous literature reported that heterogeneous nuclear ribonucleoprotein A2B1 (hnRNPA2B1), an RNA-binding protein, could control RNA loading into exosomes by binding to the specific motif (GGAG) (19), which is found at the 5′ end region of H19. Herein, we used RIP assay to verify the association between H19 and hnRNPA2B1. As anticipated, a substantial enrichment was identified between H19 and hnRNPA2B1 (Fig. 4A). Moreover, RNA pull-down analysis highlighted that hnRNPA2B1-binding ability was impaired when the ‘GGAG’ sequence is mutated (Fig. 4B and C). We then generated hnRNPA2B1 silencing or overexpressing vector (Fig. 4D). We found that the enhanced expression of hnRNPA2B1 promoted the level of exosomal H19 whereas knockdown of hnRNPA2B1 decreased its expression in HCC827R cells (Fig. 4E). Our results indicated that hnRNPA2B1 mediated the packaging of H19 into exosomes.

To examine whether H19 regulates gefitinib resistance through the delivery of exosomes, we demonstrated the H19-contained exosomes can be taken up by recipient cells by two steps. First, we examined whether secreted exosomes can be taken up by recipient cells by labeling isolated exosomes with PKH26 dye from HCC827R cells. The labeled exosomes were then added and incubated with HCC827 and HCC4006 cells for 24 h. As revealed in Fig. 5A, the recipient cells exhibited a red signal under a confocal microscope. Second, we examined whether these exosomes could deliver H19 to recipient cells similar to the intercellular transfer of other non-coding RNA as previously reported (20,21). As anticipated, RT-qPCR revealed an increased expression of H19 in both recipient cells incubated with exosomes (Fig. 5B). Thus, we ascertained that H19-contained exosomes can be taken up by recipient cells.

Next, we determined whether HCC827 and HCC4006 cells with elevated H19 levels displayed an increased resistance to gefitinib. As revealed in Fig. 5C, both recipient cell lines exhibited a promoted cell viability after treatment of exosomes as compared with the control cells. However, this effect was reversed by the transfection of si-H19#3 (Fig. 5C), indicating that it is exosomal H19 that induced gefitinib resistance.