Methyltransferase like 7B is a potential therapeutic target for reversing EGFR-TKIs resistance in lung adenocarcinoma

Identification of potential novel targets for reversing resistance to Epidermal Growth Factor Receptor (EGFR)-tyrosine kinase inhibitors (EGFR-TKIs) holds great promise for the treatment of relapsed lung adenocarcinoma (LUAD). In the present study, we aim to investigate the role of methyltransferase-like 7B (METTL7B) in inducing EGFR-TKIs resistance in LUAD and whether it could be a therapeutic target for reversing the resistance. METTL7B-overexpressed LUAD cell lines, gefitinib and osimertinib-resistant Cell-Derived tumor Xenograft (CDX) and Patient-Derived tumor Xenograft (PDX) mouse models were employed to evaluate the role of METTL7B in TKIs resistance. Ultraperformance liquid chromatography-tandem mass spectrometer (UPLC-MS/MS) was used to identify the metabolites regulated by METTL7B. Methylated RNA immunoprecipitation (MeRIP)-qPCR analysis was performed to measure the N6-methyladenosine (m6A) status of mRNA of METTL7B targeted genes. Gold nanocluster-assisted delivery of siRNA targeting METTL7B (GNC-siMETTL7B) was applied to evaluate the effect of METTL7B in TKIs resistance. Increased expression of METTL7B was found in TKIs-resistant LUAD cells and overexpression of METTL7B in LUAD cells induced TKIs resistance both in vitro and in vivo. Activated ROS-metabolism was identified in METTL7B-overexpressed LUAD cells, accompanied with upregulated protein level of GPX4, HMOX1 and SOD1 and their enzymatic activities. Globally elevated m6A levels were found in METTL7B-overexpressed LUAD cells, which was reduced by knock-down of METTL7B. METTL7B induced m6A modification of GPX4, HMOX1 and SOD1 mRNA. Knock-down of METTL7B by siRNA re-sensitized LUAD cells to gefitinib and osimertinib both in vitro and in vivo. This study uncovered a new critical link in METTL7B, glutathione metabolism and drug resistance. Our findings demonstrated that METTL7B inhibitors could be used for reversing TKIs resistance in LUAD patients.


Background
Lung adenocarcinoma (LUAD) is one of the most prevalent malignant tumors with high morbidity and mortality worldwide [1]. Precision medicine based on targeting epidermal growth factor receptor (EGFR) has contributed greatly to the improvement of LUAD treatment in the past decade [2]. The EGFR tyrosine kinase inhibitors (TKIs) have been widely used in clinical practice and significantly prolonged the survival of patients due to their advantages such as rapid absorption, high efficacy and good safety profiles [3]. However, up to 60% of patients ultimately develop drug resistance over a period of 9 to 13 months of using the first-or third-generation TKIs (gefitinib or osimertinib) [4]. One of the well-known mechanisms of developing drug resistance is the genetic mutation of EGFR (e.g. T790M or C797S), in which the mutated protein (either from acquired under selective stress or pre-existed) escapes from the interaction and inhibition of TKIs, resulting in cancer progression and drug resistance [5]. A number of studies have shown that EGFR-independent signaling pathways, including amplification of various receptor tyrosine kinases, activation of bypass signals and epithelial-mesenchymal transition (EMT), also play significant roles in TKIs resistance [6,7]. It is reported that about 40%-50% EGFR-TKIs resistance is related to EGFR-independent signaling pathway [8][9][10]. Therefore, elucidation of new targets independent of EGFR and in-depth understanding of TKIs resistance in LUAD holds not only basic research value but also great clinical significance.
Reactive oxygen species (ROS), including radicals and ions, are highly reactive molecules. ROS are excessively increased under pathophysiological conditions such as hypoxia, chemical stress, drug treatment, and might cause oxidative damage to DNA and genomic instability [11]. Accumulating evidence indicate that intra-tumoral redox homeostasis is involved in tumor development and drug resistance [12,13]. Previous studies had demonstrated the strong correlations between ROS and EGFR-TKIs resistance in lung cancer [14,15]. Therefore, blocking oxidative stress-mediated signaling pathways by reducing antioxidant enzymes is a reasonable strategy for reversing TKIs drug resistance [16].
Methyltransferase-like 7B (METTL7B) is a member of the methyltransferase-like family which contains a methyltransferase domain [17]. Studies from our group and others had identified that METTL7B was involved in tumor occurrence, development, invasion, and migration in various malignant tumors [18][19][20]. However, the role of METTL7B in tumor drug resistance and the underlying mechanism remains unknown. Here, we found that the expression of METTL7B was significantly higher in TKIs-resistant cells in comparison to TKIssensitive cells. Moreover, overexpression of METTL7B induced resistance to gefitinib and osimertinib in LUAD cells both in vitro and in vivo. METTL7B enhanced the expressions of antioxidant enzymes (including SOD1, GPX4 and HMOX1) as well as their enzymatic activities in the ROS-scavenging signaling pathway. Interestingly, MeRIP-qPCR analysis showed that instead of regulating the classic upstream transcription factor NRF2, MET-TL7B directly upregulated the expression of these antioxidant enzymes through mRNA m 6 A modification. Furthermore, knock-down of METTL7B by GNC-siRNA re-sensitized LUAD cells to gefitinib and osimertinib both in vitro and in vivo. Overall, this study provides new insights into the molecular targeted therapy and potential target for reversing EGFR-TKIs resistance in LUAD.

Patients and tumor specimens
The LUAD tissues used for PDX

Reagents and cell viability assay
Gefitinib and osimertinib (Selleck, USA) were dissolved in Dimethyl sulfoxide (DMSO, Sigma-Aldrich, Germany) at a stock concentration of 20 mM. N-acetyl-L-cysteine (NAC, Beyotime, China) and Glutathione (GSH, Beyotime, China) were reconstituted in ddH 2 O at a stock concentration of 1 M. Cells were plated at a density of 3,000 cells/well in 96 well plates. Drugs, siRNAs or vehicle control were added to the medium and treated for 72 h. Cell viability assays were performed by Cell Counting Kit-8 (Dojindo, Mashikimachi, Japan) according to the manufacturer's instructions. The half maximal inhibitory concentration (IC 50 ) values were generated and compared using GraphPad Prism.

RNA extraction and Real-time quantitative PCR assays
Total RNA was extracted from cells using TRIzol Reagent (Invitrogen, USA), and cDNA was synthesized from 1 μg of RNA with the M-MLV Reverse Transcriptase Kit (Promega, USA) as recommended by the manufacturer. Realtime quantitative PCR were performed with Bio-Rad iQ5 Real Time PCR System. The primers sequences used in this study were listed in Additional file 1: Table S1. The expression level of each individual transcript was normalized to GAPDH.

Metabolomics analysis based on UPLC-MS/MS
Cells were washed with cold PBS and collected using a cell scraper. Next, 100 mg of the sample was transferred to a 2 mL centrifuge tube containing 0.3 mL of ethanol, ultrasonicated for 30 min at 25 ℃ and centrifuged at 12,000 rpm for 10 min. The supernatant was filtered through a 0.22 µm membrane. Thirty microliters of filtrate were obtained from each supernatant and mixed to make the quality control sample. The remaining samples were tested by UPLC-MS/MS. For UPLC, chromatographic separation was accomplished in an Thermo   Figure S1). The tumor tissue was cut into scraps and passaged two generations to stabilize its inheritable information. After the tumor of NCG mice were stably formed, the mice were divided into two groups: one group for control and the other group was treated with TKIs (gefitinib and osimertinib) continuously until the tumors were resistant to TKIs. For Cell-Derived tumor Xenograft (CDX) mouse models, PC9, PC9-GR and PC9-OR cells (1 × 10 7 cells/mice) were subcutaneously inoculated into BALB/c nu/nu mice. After the tumors were stably formed, mice were treated with TKIs (gefitinib and osimertinib) and the tumor volumes (V = L × W 2 /2) was measured at a two-day interval. At the beginning of treatment with TKIs, tumor sizes in both CDX and PDX models were decreased significantly. After continuous treatment with TKIs, the volume of tumors started to regain, indicating that the xenograft mouse model that resistant to TKIs had been successfully established.

Immunohistochemistry
The tissue sections were cultured overnight with Anti-METTL7B (A7200, Abclonal Technology, China); Anti-GPX4 (ab125066, Abcam, England); Anti-SOD1 (A12537, Abclonal Technology, China); Anti-HMOX1 (A19062, Abclonal Technology, China) at 4 °C and then cultured with secondary antibody and horseradish peroxidase. All immunohistochemistry (IHC) samples were evaluated by two independent pathologists who were unaware of the source of samples and the results of the subjects. The immunohistochemical images were captured using the BioTek CYTATION 5 image reader. Each core was given a score from 0 to 3 + depending on the METTL7B, GPX4, SOD1 or HMOX1 expression in tumor cells. 0 was given for expression in less than 5% of the tumor cells, 1 + for expression in 5-50% of the tumor cells, 2 + for expression in 50-75%, 3 + for expression in more than 75% of tumor cells. For each tumor, the final score was based on the mean scores of all tumor cores. The statistical evaluation was based on the product of dyeing rate and dyeing intensity.
The m 6 A dot blot assay was performed as previously described [21]. The total RNA samples were loaded to Hybond-N + membrane (GE Healthcare, UK) and UV crossed with the nylon membrane. The membrane was then blocked with 5% nonfat milk for 1 h and incubated with m 6 A antibody (A19841, Ablconal Technology, China) at 4℃, overnight. After incubating with horseradish peroxidase-conjugated anti-mouse IgG, the membrane was visualized with the ECL detection system. The same amount of total RNA samples was spotted on the membrane and stained with 0.02% methylene blue (MB) in 0.3 M sodium acetate (pH = 5.2). The results of m 6 A level were shown in the form of relative density normalized to methylene blue staining density of the m 6 A dot blot.

Methylated RNA immunoprecipitation-PCR (MeRIP-qPCR)
Total RNA was extracted using Trizol reagent, and mRNA was purified using GenEluteTM mRNA Miniprep Kit (Sigma, Louis, MO). RNA fragmentation reagents (NEB, Hertfordshire, UK) were used to randomly fragment RNA. The specific anti-m 6 A antibody (NEB, Hertfordshire, UK) was applied for m 6 A immunoprecipitation. Anti-m 6 A antibody was pre-bound to Protein G magnetic beads in reaction buffer for 30 min. The fragmented mRNA was incubated with m 6 A-antibody-bound protein G magnetic beads at 4℃ for 1 h and washed with low salt reaction buffer and high salt reaction buffer. m 6 A-antibody-bound RNA was extracted from the Dynabeads using Buffer RLT (Qiagen, Hilden, German) and further incubated with Dynabeads MyOne Silane (Life Technologies, West Palm Beach, FL). The RNA and Dynabeads mixture were precipitated with 100% ethanol and washed with 70% ethanol, and then re-suspend with nuclease-free water. The supernatant was carefully collected after the beads were pulled to the side of the tube by a magnetic field. Real-time PCR was carried out following m 6 A-IP to quantify the changes to m 6 A methylation of a certain target gene. The sequences of primers are presented in Additional file 1: Table S1.

Preparation of Gold nanocluster-assisted delivery of siRNA (GNC-siRNA) complex
The positively charged GNCs (1 μg/mL) were mixed with siRNA solution in ultrapure water, and shaken on a bench-top shaker for 1 h to complete the binding of siRNA onto the GNCs via electrostatic interaction. The METTL7B siRNA was added into the GNC solution in different concentrations to determine the saturated concentration of siRNA solution, with the weight ratio of siRNA to GNCs varied from 0:1 to 100:1. The prepared samples were abbreviated as GNC-siRNA. The process was performed according to the previous study [22].

Statistical analysis
Statistical analyses were performed by GraphPad Prism 7.0 (GraphPad Software, La Jolla, CA, USA) for experimental analyses. The results are represented as the means ± SD of at least three independent experiments of biological replicates. Comparisons between two groups were analyzed by Student's T-tests. One-way analysis of variance (ANOVA) followed by Dunnett's test was used for comparisons among multiple groups. The relationship between METTL7B and SOD1, HMOX1, GPX4 expression levels was determined using Pearson correlation analysis. Differences were considered statistically significant when P < 0.05. Asterisks indicate statistical significance compared to the corresponding control: *, P < 0.05; **, P < 0.01; ***, P < 0.001 and ****, P < 0.0001.

METTL7B was overexpressed in TKIs-resistant LUAD
Previously, we identified that METTL7B, a member of the METTL family characterized with methyltransferase domains, promoted cell growth and tumor progression in LUAD [17]. However, the role of METTL members in TKIs-resistant LUAD was not reported. Here, bioinformatics analysis of mRNA expression of METTL members in gefitinib-sensitive PC9 cells and gefitinib-resistant PC9-GR cells indicated that METTL7B was remarkably up-regulated in PC9-GR cells as compared with PC9 cells (Fig. 1a) [23]. We further validated the expression of METTL7B in PC9, PC9-GR and PC9-OR (osimertinibresistant) cells by qRT-PCR and Western blot analysis. Consistent results to bio-informatics analysis were found in these experiments (Fig. 1b-c).
To explore the correlation between METTL7B expression and TKIs resistance in LUAD in vivo, both LUAD CDX and PDX TKIs-resistant mouse models were established ( Fig. 1d-f, 1i-j, 1m-n and Additional file 4: Figure S1). Xenografts were dissected for gene expression analysis, and the results showed that METTL7B was significantly increased in both mRNA and protein level in gefitinib and osimertinib -resistant xenografts as compared with those in gefitinib and osimertinib -sensitive xenografts (Fig. 1g-h, k-l, o-p). These findings indicate that METTL7B is involved in EGFR-TKIs resistance in LUAD.

METTL7B induced resistance to TKIs in LUAD
To investigate whether METTL7B could induce drug resistance in vitro, gene gain-of-function study was performed in LUAD cells. Cell viability analysis showed that the IC 50 of TKIs (gefitinib and osimertinib) in MET-TL7B-overexpressed PC9 and HCC827 cells significantly increased by 2-3 folds, as compared with their vector control cells (Fig. 2a-b and Additional file 5: Figure S2a, c). These findings were further validated in METTL7B knock-down LUAD cells. The IC 50 of gefitinib or osimertinib in METTL7B knock-down PC9-GR, H1975 and PC9-OR cells decreased to 40%-80%, as compared with their vector control cells (Fig. 2c-d and Additional file 5: Figure S2b-c).
To further investigate whether METTL7B could induce resistance to TKIs in LUAD in vivo, mice inoculated with either PC9 or METTL7B-overexpressed PC9 cells were treated with gefitinib, osimertinib (30 mg/kg) or vehicles. In the three groups treated with vehicles, the tumor volumes in METTL7B-overexpressed PC9 cells group were significantly increased as compared with those in vector control, which was consistent with our previous study [18] (Fig. 2e-h). With the treatment of gefitinib and osimertinib, the growth of tumors with ectopic expression of METTL7B still showed a notable rise (Fig. 2f ), while the sizes of tumors were significantly decreased in vector control group, indicated that METTL7B could induce TKIs resistance in LUAD in vivo. In order to further rule out METTL7B-induced TKIs resistance due to cell growth, we constructed the doxycycline-inducible METTL7B expression system and monitored the sensitivity of tumors to gefitinib and osimertinib in LUAD in vivo. Consistent with the results in Fig. 2e-h, under the treatment of gefitinib and osimertinib, the tumor volumes in dox-inducible MET-TL7B group were significantly increased as compared with those in vector control (Fig. 2i-l). Overall, these findings suggest that METTL7B can induce TKIs resistance in LUAD cells both in vitro and in vivo.

METTL7B promoted glutathione metabolism in LUAD cells
To systematically decipher the mechanism of MET-TL7B induced TKIs resistance in LUAD cells, un-targeted metabolomic analysis was performed in PC9 cells. Forty significantly upregulated and twenty-four significantly downregulated metabolites were found in METTL7B-overexpressed PC9 cells ( Fig. 3a and Additional file 2: Table S2). KEGG pathway analysis showed that the differential metabolites were enriched in glutathione metabolism-related process, indicating METTL7B was involved in ROS-scavenging pathways ( Fig. 3b and Additional file 3: Table S3). The receiver operating characteristic (ROC) analysis indicated that reduced glutathione was increased and oxidized glutathione was decreased in METTL7B-overexpressed PC9 cells as compared with those in vector control cells (Fig. 3c-d). These findings were further validated by examining the activity of GSH and GSSG in METTL7B-overexpressed PC9 cells. Our data showed that overexpression of METTL7B could increase the GSH activity and decreased the GSSG activity in PC9 cells, which was consistent with the results of metabolomic analysis (Additional file 6: Fig These findings indicate that METTLB promoted the glutathione metabolism in LUAD cells.

METTL7B promoted ROS scavenging through upregulation of antioxidant enzymes
Given the functional role of METTLB in promoting glutathione metabolism, its underlying molecular mechanism was further examined both in vitro and in vivo. Glutathione (GSH), a tripeptide thiol antioxidant composed of the amino acids glutamic acid, cysteine, and glycine, is the main ROS scavenger in cells and plays a central role in ROS-induced redox signaling [24,25]. The protein expression of key enzymes that regulating ROS scavenging indicated that only GPX4, SOD1 and HMOX1 were significantly up-regulated in METTL7B-overexpressed PC9 and HCC827 cells as compared with vector control cells ( Fig. 4a and Additional file 7: Figure S4a-b). Increased enzymatic activities of GPX and SOD were detected in METTL7B-overexpressed PC9 and HCC827 cells as compared with vector control cells (Fig. 4b-c). Downregulated expressions of GPX4, SOD1 and HMOX1 were further validated by shRNA knock-down of MET-TL7B in PC9-GR, H1975 and PC9-OR cells ( Fig. 4d and Additional file 7: Figure S4c-e). Similarly, the enzymatic activities were found decreased in METTL7B-suppressed PC9-GR, H1975 and PC9-OR cells, respectively ( Fig. 4e-g). These in vitro results were verified in LUAD CDX mouse model (Fig. 2e). Immunohistochemistry assay identified upregulated protein expression of MET-TL7B, GPX4, SOD1 and HMOX1 in METTL7Boverexpressed PC9 cell-derived tumors (Fig. 5a). Moreover, analysis of the expression data from TCGA database indicated that the expression of METTL7B was positively correlated with these three key ROS scavengers [GPX4 (R = 0.41, P < 2.2e-16), HMOX1 (R = 0.3, P < 8.8e-13), and SOD1 (R = 0.35, P < 2.2e-16)] (Fig. 5b). Our results indicate that METTL7B can accelerate ROS scavenging through upregulation of antioxidant enzymes.

METTL7B induced TKIs resistance in LUAD cells in a ROS-scavenging-dependent manner
Given that METTL7B could increase the enzymatic activity of antioxidants in LUAD, we further evaluated whether METTL7B-induced TKIs resistance was antioxidant-dependent. Buthionine sulphoximine (BSO), a specific inhibitor of ROS scavenging, was used (See figure on next page.) Fig. 6 METTL7B-induced TKIs resistance was associated with ROS scavenging in LUAD cells. a-d FLAG-NC and FLAG-METTL7B was stably transfected into gefitinib-sensitive PC9 cells. The enzymatic activities of GPX (a) and SOD (b) were measured by oxidative stress assay kits with or without BSO (200 μM for 24 h). Cell viability was detected 72 h after treatment of different concentrations of gefitinib (c) and osimertinib (d). e-j shMETTL7Bs were stably transfected into gefitinib-resistant PC9-GR and osimertinib-resistant PC9-OR cells. The enzymatic activities of GPX (e and h) and SOD (f and i) were measured by oxidative stress assay kits with or without NAC (10 mM for 6 h) or GSH (8 mM for 6 h). Cell viability was detected 72 h after treatment of different concentrations of gefitinib (g) and osimertinib (j). k-n PC9 cells with FLAG-METTL7B or FLAG-NC were injected into the flank of BALB/c nude mice to form subcutaneous tumors. The mice with subcutaneously implanted tumors were treated with TKIs (gefitinib and osimertinib) (30 mg/kg, qd, po) or combinations of TKIs and BSO (450 mg/kg, qod, ip) as indicated. The growth of tumors was monitored every 2 d. Tumor volume and body weights were presented as mean ± SD from five mice per group. *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0. in METTL7B-overexpressed PC9 cells. The results showed BSO treatment (200 μM) re-sensitized the METTL7B-overexpressed PC9 cells to gefitinib and osimertinib, accompanied with the decreased enzymatic activities of GPX and SOD (Fig. 6a-d). Moreover, the increased sensitivity toward gefitinib and osimertinib and reduced enzymatic activities of GPX and SOD induced by knock-down of METTL7B in PC9-GR and PC9-OR cells could be reversed by ROS scavengers, either N-acetyl-L-cysteine (NAC) (10 mM) or antioxidant GSH (8 mM) (Fig. 6e-j). Consistent results were found in both METTL7B-overexpressed HCC827 cells and METTL7B-suppressed H1975 cells (Additional file 8: Figure S5a-g).
To verify these findings in vivo, we examined whether activating the ROS system, by suppressing GSH biosynthesis with BSO, could abrogate METTL7B-induced gefitinib and osimertinib resistance in LUAD CDX mouse model. It was found that overexpression of METTL7B conferred resistance to gefitinib and osimertinib, and the effect was abrogated by BSO treatment in these mouse model (Fig. 6k-n). These findings suggest that METTL7B can induce TKIs resistance in LUAD cells in a ROS-scavenging dependent manner.

METTL7B upregulated the protein expression of ROS-scavenging genes mediated by m 6 A modification.
The METTL7B is consisted of nuclear export signal (NES), methyltransferase domain, and SAM binding motif I (GXGXG) (Additional file 9: Figure S6a) [26,27]. We next explored whether METTL7B modulated ROS related genes by m 6 A modification. The m 6 A MS analysis indicated that m 6 A level was significantly increased in METTL7B-overexpressed PC9 cells as compared with vector control cells (Additional file 9: Figure  S6b). The levels of m 6 A modification were significantly increased in METTL7B-overexpressed PC9 and HCC827 cells as compared with vector control cells as shown by the m 6 A-specific antibody dot blot (Fig. 7a). Moreover, knock-down of METTL7B by shRNA significantly reduced m 6 A modification level in PC9-GR, H1975 and PC9-OR cells (Fig. 7a).
Then, we detected the m 6 A modification in m 6 A consensus sequence within CDS and the 3' UTR of GPX4, SOD1 and HMOX1 mRNA (Additional file 9: Figure   S6c). The results revealed a higher m 6 A modification in the 3' UTR of GPX4, the CDS of SOD1 and the 3' UTR of HMOX1 mRNA in METTL7B-overexpressed PC9 cells (Fig. 7b-e). In order to further validate these findings, 3-Deazadenosine (3-DA), a S-adenosylmethionine inhibitor which could deplete the m 6 A modification on mRNA [28], was applied. Treatment of 3-DA (10 μM) reduced the level of m 6 A modification induced by MET-TL7B overexpression in PC9 cells (Fig. 7f ). Furthermore, increased GPX activity, insensitivity to gefitinib and the mRNA m 6 A modification of antioxidant genes induced by overexpression of METTL7B were all abrogated by the treatment of 3-DA in PC9 cells (Fig. 7g-l). Taken together, our results suggest that METTL7B can enhance the protein expression of ROS-scavenger genes in LUAD cells through m 6 A modification eventually leading to TKI resistance.

Knock-down of METTL7B reversed the resistance toward gefitinib and osimertinib in LUAD both in vitro and in vivo.
To examine the potential significance of targeting METTL7B in reversing EGFR-TKIs resistance in LUAD, gold nanocluster-assisted delivery of siRNA was used for both in vitro and in vivo cell growth assay [22]. Gefitinib-resistant PC9-GR and H1975 cells, and osimertinib-resistant PC9-OR cells were treated with the well-characterized GNC-siMETTL7B (Additional file 10: Figure S7) as well as vector control in combination with gefitinib or osimertinib. The results showed that combination of siRNA-METTL7B and gefitinib significantly re-sensitized PC9-GR and H1975 cells to gefitinib with a much lower IC 50 as compared with that of vector control ( Fig. 8a-b). Similar results were found in METTL7B-suppressed PC9-OR cell (Fig. 8c).
Further investigation was performed to examine the effects of GNC-siMETTL7B in LUAD CDX mouse model. Coincided with the in vitro results, combination treatment with GNC-siMETTL7B (6 mg siRNA per mouse equivalent) significantly suppressed tumor growth as compared with treatment of gefitinib or osimertinib (30 mg/kg) alone ( Fig. 8d-g). These findings demonstrate that METTL7B could be a potential therapeutic target for reversing TKIs resistance in LUAD.

Discussion
Previously, our group and others had revealed that increased expression of METTL7B contributed to cancer cell proliferation, migration, and invasion, resulting in advanced stages of tumor development and poor survival [18][19][20]. Here, we found that METTL7B promoted TKIs resistance in LUAD via accelerating the scavenging of ROS. Mechanistically, METTL7B upregulated the protein expression of antioxidant genes via m 6 A modification. In addition, targeting METTL7B by GNC-siMETTL7B re-sensitized LUAD cells to TKIs both in vitro and in vivo. These results indicated that METTL7B could be a promising therapeutic target for reversing TKIs resistance in LUAD. Intra-tumoral hypoxia leads to abnormal metabolism and chromosome instability in cancer cells, which results in excessive ROS generation in tumor microenvironment [11]. In order to maintain redox-homeostasis, tumor cells increase enzymatic activity of antioxidants and activate enzyme-independent pathways to scavenge ROS [29]. In fact, a number of studies have shown that most of the antitumor therapies including chemotherapy [30], radiotherapy [31], immunotherapy [32] induce ROS production and activate death signaling cascades in cancer cells. Because METTL7B enhanced the expression of essential ROSscavenging enzymes, SOD1, GPX4 and HMOX1 as well as their enzymatic activities [33][34][35], it is worthy to explore whether METTL7B could promote resistance to other anti-cancer reagents in addition to TKIs in the future.
Drug resistance caused by EGFR mutations has been a major issue of antineoplastic therapy failure in using EGFR inhibitors [5]. The contribution of EGFR-independent molecules and pathways to TKIs resistance has attracted attention [8][9][10]. A weak correlation between the protein expression of METTL7B and EGFR was found in LUAD tissue microarray [18], which promoted us to explore the role of METTL7B in TKIs resistance in LUAD. However, no significant change was found in the protein expression as well as phosphorylation of EGFR in both METTLB-overexpressed and -suppressed LUAD cells (data not shown). Although METTL7B did not directly regulate the activation of EGFR signaling pathway, it enhanced the ROS scavenging signaling pathway by upregulating the expression of antioxidant enzymes, indicating that METTL7B induced TKIs resistance in an EGFR-independent manner (Fig. 4). By using GNC-siRNA targeting METTL7B, the gefitinib and osimertinib resistance was reversed in vivo. Notably, we also identified low expression of METTL7B in low-grade LUAD and tumor adjacent normal tissues (data not shown). These findings provide a rationale for combined use of METTL7B inhibitors with TKIs in patients with EGFRindependent TKIs resistance.
In this study, we identified three downstream targets of METTL7B, GPX4, SOD1 and HMOX1. GPX4 is a critical inhibitor of ferroptosis, a process during which iron-induced peroxidation of membrane lipids causes programmed cell death that distinct from apoptosis [36]. SOD1 is a well-known enzyme for defending against oxidative stress by catalyzing the disproportionation of superoxide into H 2 O 2 and O 2 [33]. HMOX1 is an antioxidant, anti-inflammatory and anti-apoptotic protein which promotes metabolic reprogramming and antioxidant defense [37]. Previous studies have shown that Kelch-like ECH-associated protein 1 (KEAP1)-nuclear factor (erythroid-derived 2)-like 2 (NRF2) antioxidative pathway is involved in the process of drug resistance in tumors [38,39]. Once Oxidative stress commonly induced the conformation of KEAP1, which led to dissociation of NRF2. The dissociated NRF2 entered the nucleus, leading to upregulation of the expressions of ROS metabolic genes, such as GPX4, SOD1, NQO1 and HMOX1 [38,39]. However, we identified that the expression as well as nuclear-cytoplasmic distribution of NRF2 were not changed in METTL7B inducted PC9 cells (Additional file 11: Figure S8), indicating that METTL7B regulated ROS metabolism in an NRF2-independent manner. In this study, we uncovered a novel role of MET-TL7B in metabolic regulation based on the findings that METTL7B regulated antioxidant genes.
METTL7B was found to be involved in cell growth among various malignant tumors [18][19][20]. However, its underlying regulatory mechanism remains unclear. Our data suggested for the first time that METTL7B played a role in m 6 A modification of mRNAs. The level of m 6 A modification was elevated in METTL7B-overexpressed PC9 and HCC827 cells and knock-down of METTL7B (See figure on next page.) Fig. 8 Knock-down of METTL7B reversed the resistance toward TKIs in LUAD cells both in vitro and in vivo. a-c Gold nanocluster-assisted delivery of METTL7B-siRNA was transfected into gefitinib-resistant PC9-GR, H1975 and osimertinib-resistant PC9-OR cells, and the cell viability was evaluated to measure IC 50 of TKIs after treatment with different concentrations of gefitinib and osimertinib for 72 h. d-g PC9-GR (d-e) and PC9-OR (f-g) cells were injected into the two flanks of BALB/c nude mice to form subcutaneous tumors. The mice with subcutaneously implanted tumors were treated with TKIs (gefitinib and osimertinib) (30 mg/kg, qd, po) and GNC-siNC or GNC-siMETTL7B as indicated. The growth of tumors was monitored every 3 d. Tumor volume and body weights were presented as mean ± SD from five mice per group. h The schema represented the critical link between METTL7B, GPX4, HMOX1 and SOD1 in TKIs-resistant LUAD cells. **P < 0.01 and ****P < 0. significantly reduced the levels of m 6 A modification. The status of m 6 A modification of mRNA in the antioxidant genes (GPX4, SOD1 and HMOX1) was directly regulated by METTL7B, and the effect was eliminated by m 6 A modification inhibitor. Recent studies showed that the m 6 A readers, such as IGF2BPs or RNA-binding protein HNRNPD, could selectively bind to the target mRNA and promote the mRNA stability. In gastric cancer, METTL3 promoted the m 6 A modification of HDGF mRNA, and the m 6 A reader IGF2BP3 directly recognized and bound to the m 6 A site on HDGF mRNA and enhanced HDGF mRNA stability [40]. Another study in H/R-treated cardiomyocytes revealed that METTL3 methylated the 3' UTR of TFEB mRNA, thereby promoting the association of HNRNPD with TFEB pre-mRNA and decreasing the mRNA level of TFEB [41]. Therefore, more physiological target genes of METTL7B and the interaction with m 6 A readers warranted further exploration.
Previous studies had showed the strong correlation between m 6 A modification and drug resistance. The m 6 A demethylase FTO was overexpressed in leukemia cells, which enhanced mRNA stability of proliferation/ survival transcripts, increased protein synthesis and induced TKIs resistance [42]. m 6 A reader IGF2BP2 bound to CDS region of ERBB2 mRNA via m 6 A modification, increased ERBB2 translation efficacy, and induced acquired resistance to TKIs [43]. METTL3 could promote m 6 A modification and translation of YAP mRNA through recruiting YTHDF1/3 and eIF3b and increased mRNA stability of YAP via the MALAT1-miR-1914-3p-YAP axis, thereby inducing cisplatin resistance in NSCLC [44]. Hypoxia-induced YTHDF1 upregulated the expression of KEAP1 via m 6 A modification and induced the sensitivity to cisplatin via inhibiting drug-resistant-associated gene AKR1C1 [45]. In sorafenib-resistant hepatocellular carcinoma, METTL3 was significantly down-regulated and could promote m 6 A modification of FOXO3 3'UTR increasing its stability via recruiting YTHDF1, thereby enhancing sorafenib resistance of HCC [46]. In this paper, we found that METTL7B could promote TKIs resistance in NSCLC. Further mechanism revealed that METTL7B promoted expressions of SOD1, HMOX1 and GPX4 via m 6 A modification, indicating that METTL7B may exercise m 6 A function acting as the RNA methyltransferase. And more biochemical experiments of METTL7B require further verification in the future.

Conclusions
In the present study we uncovered a novel role of MET-TL7B in promoting TKIs resistance in LUAD. Under the surviving stress upon TKIs, elevated METTL7B in LUAD cells accelerated the scavenging of excessive ROS in tumor microenvironment by upregulating the protein levels and enzymatic activities of three antioxidant enzymes GPX4, SOD1 and HMOX1 by m 6 A modification in their mRNAs (Fig. 8h). Our study provides new insights that METTL7B could be a potential therapeutic target for reversing TKIs resistance in LUAD.