The GE-sensitive PC9 cell line and H292 cell line were purchased from the Cellular Institute (Cell Bank) of the Chinese Academy of Science (Shanghai, China). The GE-resistant cells PC9-GR and H292-GR were established by exposing them to increasing concentrations of GE. When the cells could not sustain growth similarly to the parental cell lines under GE treatment, the GE was discarded before the cells restored the growth rate. After over 12 months of GE exposure, the PC9-GR and H292-GR cell lines were successfully generated and maintained in 2 μM of GE. MTT assays were conducted to confirm the resistance characteristic of PC9-GR and H292-GR cells. All the above cell lines were cultured in RPMI 1640 medium containing 10% fetal bovine serum (FBS) incubated in a humid atmosphere containing 5% CO₂. The resistant sublines were cultured in a GE-free complete medium for at least 2 weeks to eliminate the interference of GE before all the experiments.

Cell proliferation detections were conducted by MTT assays. The PC9, PC9-GR, H292, and H292-GR cells (5×10³) were seeded into 96-well plates overnight. Then, different concentrations of GE and MA were treated alone or in combination for 48 h. After incubation, MTT (final concentration: 5 mg/ml) was added to the 96-well plates and incubated at 37°C in the dark for 4 h. The supernatants were discarded, and the formazan was dissolved by DMSO and measured by an iMark microplate reader (Bio-Rad, Hercules, CA, United States). The viabilities were calculated by a percentage relative to the DMSO-treated control group. A GraphPad 8.01 software calculated the median inhibitory concentration (IC₅₀) values. The synergistic effects of GE and MA in different NSCLC cells were evaluated by the Chou–Talalay method described previously (18).

Annexin V-FITC and PI synchronous staining were conducted to evaluate apoptotic cells after GE and MA co-administration in GE-resistant NSCLC cell lines. In brief, cells at a density of 5×10⁵ were seeded into 6-well plates; after incubation overnight, different concentrations of GE and MA were co-administrated for 48 h at 37°C. Subsequently, the cells were collected and stained with Annexin V-FITC and PI (5% respectively) for 30 min at RT in the dark. Then, the cells were centrifuged and resuspended in 250 μl of detection buffer and analyzed by MACS Verse flow cytometer (BD Biosciences, San Jose, CA, United States). The qualification of apoptotic cells was analyzed by FlowJo VX software (Tristar, CA, United States).

The accumulation of GE in resistant NSCLC cells was evaluated by tritium-labeled GE [GE-T (G), 10 Ci/mmol], purchased from Energy Chemical (Shanghai, China). The cells were plated into 6-well plates overnight and co-administrated with GE-T (G) and MA for 2 h, 24 h, and 48 h. Then, the cells were collected and the liquid scintillation was detected by a Quantulus™ GCT spectrometer (Perkin-Elmer, Waltham, MA, United States). The accumulation rates of GE were calculated via a custom standard curve.

The expression level of EGFR-related downstream pathway molecules, the BCRP and MRP-7 expression level, and FTO and its downstream molecules were evaluated by Western blot assay. Specifically, cells were plated into a 6-well plate overnight and co-administrated with GE and MA. The Triton-X 100 lysis buffer was used to extract the total protein. The concentrations of proteins were qualified by BCA protein Assay Kit (Pierce™, Thermo Fisher Scientific, Waltham, MA, United States). Equal amounts of proteins (20–60 μg) were subjected to 8% to 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) before being transferred onto polyvinylidene fluoride (PVDF) membranes (Millipore, Billerica, MA, United States). Non-fat milk (5%) was used as a blocking agent to block the non-specific binding sites of PVDF membranes. The primary antibodies of total EGFR, p-EGFR, p-Akt, p-ERK1/2, BCRP, MRP-7, FTO, and c-Myc were incubated as recommended concentrations with the membrane at 4°C for overnight. Then, the membranes were washed with TBST buffer and incubated with HRP-conjugated secondary antibodies at RT for 1 h. After washing with TBST buffer, the ECL was added to the membrane and an Invitrogen iBright (Thermo Fisher Scientific, Waltham, MA, United States) was used to visualized the protein bands. The relative intensity was calculated by catching the gray level using ImageJ software.

The mRNA expression levels of BCRP, MRP7, FTO, and MYC were determined by qPCR. In brief, the total RNA was extracted from the cells in the presence or absence of MA using TRIzol reagent (Invitrogen, Carlsbad, CA, United States) according to the manufacturer’s instructions. The concentration of total RNA was measured by ultraviolet spectrophotometry. According to the manufacturer’s instructions, complementary DNAs (cDNAs) were transcribed using RevertAid RT (Thermo Fisher Scientific, Waltham, MA, United States). PCR was performed on Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA) with SYBR™ Green PCR from Thermo Fisher Scientific (Waltham, MA, United States). The 2^–ΔΔCt methods calculated the copy number, and samples were run in triplicate and normalized to GAPDH. Then, the expression levels of mRNAs were reported as fold changes versus control. The primers were as follows: For BCRP, the forward primer (5’-3’) was GCTACACCACCTCCTTCTGT, and the reverse primer (5’-3’) was GGAAGAAGAGAACCCCAGCT. For MRP7, the forward primer (5’-3’) was AGAGTACACCTGTGACCTGC, and the reverse primer (5’-3’) was GAGCACCAACAACAGGGAAG. For GAPDH, the forward primer (5’-3’) was GCACCGTCAAGGCTGAGAAC, and the reverse primer (5’-3’) was TGGTGAAGACGCCAGTGGA.

The parental and drug-resistant cell lines were seeded into 6-well plates overnight. The cells were incubated with tritium-labeled GE (GE-T) in the presence or absence of MA for 2 h, 24 h, 48 h, or 72 h, respectively, at 37°C containing 5% CO₂. Then, the cells were lysed by Triton X-100 lysis buffer (Beyotime, Shanghai, China). The radioactivity of the total lysis product was measured by a Tri-Card 4910 liquid scintillation analyzer (PerkinElmer, Waltham, MA, United States).

A membrane and cytoplasm protein isolation kit isolated the cell membrane protein from Beyotime (Shanghai, China). In brief, 5×10⁷ cells were seeded into a 6-cm² dish, then were incubated in the presence or absence of GE and MA for 24 h. The cells were collected after treating with EDTA buffer, centrifuged, and homogenized. The nucleus was discarded by centrifuging at 700g for 10 min at 4°C, and the supernatants were collected. The cell membrane was collected by centrifuging at 14,000g for 30 min at 4°C. Then, the manufactured lysis buffer was used to extract membrane proteins. The composed membrane proteins were subjected to SDS-PAGE subsequently as aforementioned.

The dot blot assays for detecting m⁶A levels were conducted as previously described (19). In brief, according to the manufacturer’s instructions, the total RNAs were isolated from the cells in the presence or absence of GE and MA using TRIzol reagent (Invitrogen, Carlsbad, CA, United States). PolyATtract^® mRNA Isolation Systems (Promega, Madison, WI, United States) was then used to extract mRNA. The ultraviolet spectrophotometry method was used to determine the concentration of mRNA. Subsequently, the mRNAs were denatured for 3 min at 95°C, followed by cooling in ice directly. The mRNAs were UV cross-linked after being spotted on an Amersham Biosciences Hybond-N+ membrane optimized for nucleic acid transfer (GE Healthcare, Little Chalfont, United Kingdom). The membranes were shed with TBST buffer, followed by blocking with 5% non-fat milk, and hybridized with anti-m6A antibody (Abcam, Cambridge, MA, United States) overnight at 4°C. Then, the membranes were washed with TBST buffer and incubated with HRP-conjugated secondary antibodies at RT for 1 h. After washing with TBST buffer, the ECL was incubated and an Invitrogen iBright (Thermo Fisher Scientific, Waltham, MA, United States). The relative intensity was calculated by catching the gray level using ImageJ software.

The methylated RNA immunoprecipitation (Me-RIP) was conducted to detect the m⁶A modification of MYC as previously described (20). In brief, total RNAs were isolated, then the mRNA was further isolated and purified using the TIANSeq mRNA Capture Kit (TIANGEN Biotech, Beijing, China). The anti-M⁶A or anti-IgG (1:100, Abcam, Cambridge, UK) were added and incubated with protein A/G agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA, United States) in IP buffer overnight at 4°C. The eluent buffer was conducted to elute RNA. The phenol-chloroform was used to purify the RNAs. The qPCR was performed to determine the mRNA level of m⁶A MYC.

Data are expressed as mean ± SD in at least three independent experiments. The significance was determined using one-way ANOVA. GraphPad 8.00 software was used to determine the statistical analysis. p < 0.05 was considered statistically significant.

The potential synergistic anti-proliferation effects of MA and GE were determined by MTT assays and fitted using Chou–Talalay methods as previously described (18). We first detected the antiproliferation characters of MA and GE on parental and resistant NSCLC cells. As illustrated in Figures 1A–D , the IC₅₀ values of GE on PC9 and PC9/GR cells were 0.32 and 3.45 μM, respectively. In H292 and H292/GR cells, the IC₅₀ of GE was 0.51 and 5.03 μM, indicating the confirmed resistance profile of PC9/GR and H292/GR cells. The IC₅₀ of MA, in all cell lines, was 4–8 μM.

Next, the drug combination of GE and MA was conducted, and the combination index (CI) was calculated. As shown in Figures 1E–H , the different IC₅₀ value fractions were undertaken in the combined administration of GE and MA. The CI value calculation results indicated that GE and MA showed potent synergistic effects in GE-resistant cells (PC9/GR and H292/GR). In contrast, no significant synergistic effects of GE and MA were observed in parental cells ( Figures 1I –L ). Based on the results of the drug combination, we chose related concentrations for the following experiments: for PC9/GR cells, 2.5 μM GE and 4 μM MA were adopted; for H292/GR cells, 4 μM GE and 6 μM MA were conducted.

Based on the combination of GE and MA results, different concentrations of GE and MA were co-administrated in PC9/GR and H292/GR cells to evaluate the apoptotic cell population by flow cytometry. As illustrated in Figures 2A, B , though single use of GE or MA in PC9/GR and H292/GR cells induced poor apoptosis, the co-administration significantly induced cell apoptosis potently, indicating the potential combined effects of GE and MA. Furthermore, Western blot results showed that the apoptotic effects of GE combination of MA were caspase-3-dependent ( Figures 2C, D ).

We next evaluated the potential mechanisms of action on the synergistic effects of GE and MA. As shown in Figure 3A , in PC9/GR cells, the EGFR-related downstream key molecules were detected by Western blot. GE or MA single use did not significantly induce inactivation of EGFR downstream molecules. However, the combination-treated GE and MA significantly inhibited the phosphorylation of EGFR, Akt, and ERK1/2, which indicated that the synergistic effects of GE and MA might be related to the inactivation of the EGFR molecule pathway. In addition, the co-administration of GE and MA in H292/GR cells also induced a potent decrease of phosphorylated modification of EGFR signaling pathways molecules, which further indicated that the EGFR signaling is related to the synergistic effects of GE and MA on drug-resistant NSCLC cells ( Figure 3 ).

Considering the impaired cytotoxicity of GE in resistant NSCLC cells, the synergistic effects may be related to the increased accumulation of GE. Hence, the tritium-labeled GE-based accumulation assays were conducted. As shown in Figures 4A, B , short-term incubation of MA (2 h) did not significantly alter GE accumulation in either PC9/GR or H292/GR cells. Interestingly, more prolonged incubation of MA (24 h, 48 h, and 72 h) potently increased the intracellular tritium-labeled GE in PC9/GR and H292/GR cells, but not in their corresponding parental cells.

We next evaluated whether MA or combination with GE could alter the membrane drug pumps in GE-resistant NSCLC cells. As illustrated in Figure 5 , the cell membrane of both PC9/GR and H292/GR cells was isolated, and SDS-PAGE detected the expression of MRP7 and BCRP. As illustrated in Figure 5A , the expression of MRP7 and BCRP in PC9/GR significantly decreased in MA and MA/GE co-administrated group compared with the control or GE treatment group. In addition, similar results were found in H292/GR cells, where both the BCRP and MRP7 were downregulated by MA or MA/GE co-administrated.

The total M6A modification in PC9/GR and H292/GR cells was detected by dot blot assays of M6A. As shown in Figures 6A, B , the M6A modification of mRNA in the MA- or MA/GE-treated group significantly increased, indicating that MA or MA/GE co-administration may affect the activity of M6A-related proteins. Moreover, demethylase FTO was significantly downregulated in both PC9/GR and H292/GR cells in MA- and MA/GE-treated groups ( Figures 6C, D ). Furthermore, the transcription factor MYC was significantly decreased in the MA or MA/GE co-treatment group. The Me-RIP results indicated that both MA alone and in combination with the GE group significantly increased the M6A modification of MYC ( Figures 6E, F ).