N6-Methyladenosine Reader YTHDF2 Enhances Non-Small-Cell Lung Cancer Cell Proliferation and Metastasis through Mediating circ_SFMBT2 Degradation

Objective circ_SFMBT2 was reported to facilitate malignant progression in various cancers, but its function in non-small-cell lung cancer (NSCLC) has not been fully uncovered. This study aimed to investigate the effects of N6-methyladenosine (m6A) methylation of circ_SFMBT2 (circ_0017628) on non-small-cell lung cancer (NSCLC) and its underlying mechanisms. Methods Paired tumor and noncancerous tissues from NSCLC patients were surgically collected from January 2020 to March 2021 in our hospital. The levels of circ_SFMBT2 and LATS2 in NSCLC and human bronchial epithelial cells were assayed with qRT-PCR. Overexpression or silencing of circ_SFMBT2, LATS2, or YTHDF2 was performed in the NSCLC cells. CCK-8, colony-forming, and transwell assays were performed to analyze cell proliferation, viability, and migration, respectively. Meanwhile, the expression of MMP-9, E-cadherin, vimentin, and the Hippo/YAP pathway components was examined by western blotting. The m6A enrichment in circ_SFMBT2 was verified using methylated RNA immunoprecipitation, and interaction between circ_SFMBT2 and YTHDF2 was assessed by RNA pull-down and immunoprecipitation assays. Results Both circ_SFMBT2 and LATS2 were lowly expressed in NSCLC cells and tissues. A positive correlation of circ_SFMBT2 with LATS2 was identified, and circ_SFMBT2 was localized predominantly in the cytoplasm. circ_SFMBT2 overexpression negatively regulated cell proliferation, viability, migration, and epithelial-mesenchymal transition while promoting the Hippo/YAP pathway activation. Notably, knockdown of LATS2 effectively abrogated the inhibitory effects of circ_SFMBT2 overexpression on NSCLC cell malignancies. Besides, m6A was specifically enriched in circ_SFMBT2, and circ_SFMBT2 could bind to YTHDF2. Silencing of YTHDF2 led to an increase in circ_SFMBT2 expression while inhibiting the malignancy of cancer cells. Conclusion Our results showed that YTHDF2 could facilitate NSCLC cell proliferation and metastasis via the Hippo/YAP pathway activation by mediating circ_SFMBT2 degradation.

. Due to its multifocality and complexity, NSCLC is linked to a high risk of recurrence even following multimodal treatment [4]. us, pathogenetic studies of NSCLC may contribute to innovative and effective treatments of the patients, as well as the identification of new and important diagnostic markers.
Circular RNAs (circRNAs) are a group of noncoding RNAs harboring a covalently closed-loop conformation, which is generated from the reverse clipping of exonic mRNAs (pre-mRNAs) [5]. circRNAs are implicated in various pathophysiological processes, such as brain development and tumorigenesis [6]. Meanwhile, circRNAs have been found to function in numerous cancers, suggesting that they may potentially serve as a biomarker or therapeutic target. In these cases, an elevated level of circPRKCI, namely E2F7, ultimately promotes lung adenocarcinoma development by competing for binding to miR-545 and miR-589 [7]. circTP63 competes for binding to miR-873-3p and upregulates FOXM1 to expedite cell proliferation in lung squamous cell carcinoma [8]. circ_SFMBT2, a newly discovered circRNA with a circular structure, is derived from the SFMBT2 gene [9,10]. It has been reported that circ_SFMBT2 positively regulates gastric cancer cell proliferation via the miR-182-5p/CREB1 axis [11]. circ_SFMBT2 has also been demonstrated to facilitate the malignant phenotype of esophageal cancer through regulating SLC1A5 in a miR-107dependent manner [9]. To date, the function of circ_SFMBT2 in NSCLC cell development remains largely unknown.
As the most abundant modification in mRNAs, N6methyladenosine (m6A) constitutes over 80% of all RNA modifications [12]. Strikingly, m6A is involved in the quantitative and qualitative regulation of target RNAs by modulating RNA splicing, stability, translocation, and translation [13]. Aberrant m6A methylation could cause dysregulation of genes that are essential for controlling key cellular processes while disrupting homeostasis, thus incurring disorders. Of note, m6A reader YTHDF2 mainly regulates splicing events and post-transcription in the cytoplasm. YTHDF2 can also transport mRNAs from actively translated mRNA repertoires to decaying mRNA repertoires, resulting in mRNA degradation [14]. Besides, studies have found that YTHDF2 facilitated malignant progression in several cancers, including early NSCLC; in this case, its elevated expression was linked with tumor growth and metastasis [15]. Other studies have suggested the potential of YTHDF1 and YTHDF2 as new prognostic factors and drug targets associated with the tumor immune microenvironment of NSCLC [16]. However, it should be noted that YTHDF2 was found to play opposite roles in different types of cancers as well. For instance, it was shown to be upregulated and acted as an oncogene in cancers including acute myelocytic leukemia (AML), lung cancer, and gastric cancer [PMID: 31711642, 31031138, and 31504235], while its overexpression in osteosarcoma and melanoma was associated with decreased cancer cell proliferation and migration (PMID: 31239444 and 32021563). Currently, the biological role of YTHDF2 in NSCLC has not yet been fully uncovered.
In this study, we performed in vitro experiments to elucidate the relationship between YTHDF2 and circ_SFMBT2 and further characterized their roles in NSCLC to provide new information that could lead to the identification of new biological targets and development of strategies for improving the clinical treatment of NSCLC.

Materials and Methods
2.1. Tissue Specimens. Paired tumor and normal tissues from NSCLC patients were surgically collected from January 2020 to March 2021 at our hospital. All the patients voluntarily provided written informed consent. Ethical approval was granted by the ethics committee of our hospital (2020-IRBQYYS-011). All experiments described as follows were repeated three times.

Quantitative Reverse Transcription. PCR (qRT-PCR)
Nuclear-cytoplasmic fractionation was performed using a nucleoplasmic isolation kit, and total RNA extraction from nuclei, cytoplasm, whole cells, or tissues was conducted using TRizol ( ermo, USA). Total RNA was subjected to detecting the concentration and purity by using NanoDrop and subsequently reverse transcription using a random primer reverse transcription kit. After that, cDNA was PCRamplified with TaKaRa SYBR Green kit using the corresponding primers. A total of six replicates of the experiment was carried out. e expression of target genes was normalized to that of GAPDH, and the relative expression was determined using the 2−ΔΔCt method. e PCR primer sequences are presented in Table 1. (FISH). FISH was carried out using the circ_SFMBT2-targeting probe. After 15 mins of fixation with 4% paraformaldehyde, the cells were washed with PBS and dehydrated in graded alcohols. en, the cells were mixed with denatured DNA probes and hybridized overnight at 37°C in a humid and dark environment. On the following day, the cells were subjected to washing with saline-sodium citrate buffer thrice for 5 min each and blocked for 1 h in PBS with 1% BSA and 3% normal goat serum, followed by incubation with an antibiotin antibody (HRP-conjugated) overnight at 4°C. e images were acquired by using a fluorescent microscope.  Assay. Cell proliferative ability was evaluated using the CCK-8 assay. e cells (5 × 10 3 /well) were seeded in a 96-well plate. After 24 h of culture, 10 μL of the CCK-8 solution was applied to each well, and the plate was subjected to incubation at 37°C for 1 h. e OD values were determined at 450 nm.

Colony-Forming Assay.
e cells were trypsin-digested, resuspended with a DMEM-complete medium, plated in a 6well plate at a density of 7 × 10 2 cells/well, and grown at 37°C with 5% CO2 for 14 days. When the colonies were visible, the cells were fixed in 4% paraformaldehyde, followed by staining using 0.1% crystal violet. e staining was photographed by using an inverted microscope, and the colonies with a single clonal cell number greater than 50 were counted. e colony formation rate was measured based on the following formula: (number of colonies/number of cells plated) × 100%.

Transwell Assays.
Matrigel was thawed in a refrigerator for no less than 12 h, diluted in serum-free cell culture medium at 1 : 8, and spread in the transwell upper chamber at 4°C. e cells were harvested by trypsin digestion, resuspended with serum-free medium, and adjusted to a density of 1 × 10 5 cells/ml. Subsequently, 100 μL cells were applied into the upper chamber coated with or without Matrigel, respectively, while 700 μL of 10% FBS-containing medium was added to the lower chamber. Transwell inserts were removed after 24 h of growth under 5% CO 2 at 37°C, and the cells were then washed thrice with PBS. ereafter, the cells were subjected to 30 min fixation, dried, and stained using 0.1% crystal violet for 30 min. Lastly, the stained cells were washed with PBS, dried, and examined using an upright microscope in randomly selected fields. ree visual fields were randomly selected for counting the positive cells.

Western Blot Analysis.
Tissues or cells were lysed in cell lysis buffer (Gibco, USA), and lysates were subjected to centrifugation (14,000 rpm for 15 min) at 4°C. Proteins were quantified using the BCA assay. Each protein sample (20 μg) was separated on SDS-PAGE and electroblotted onto membranes. e membrane was blocked for 1 h in 5% nonfat dry milk and then incubated overnight at 4°C with specific primary antibodies. e following day, the membrane was washed and then subjected to 1 h of incubation with fluorescence-conjugated secondary antibodies at RT. e blot was developed with chemiluminescence reagents, and an imaging system was used to collect the images. Quantification of the immunoreactive bands was conducted using the Image J software, and the expression level of targets was normalized to that of GAPDH.

Methylated RNA Immunoprecipitation (MeRIP).
PBS-washed cells were collected, and total RNA was extracted with 2 ml QIAzol (QIAGEN, Germany). en, 100 μg of the total RNA was diluted in IPP buffer (150 mM NaCl, 10 mM Tris, pH 7.4, and 0.1% NP-40) with 10 μg of the anti-m6A antibody to 300 mL. After 2 h of incubation at 4°C, the mixture was incubated with 50 μL of Invitrogen G-conjugated Dynabeads for another 2 h. ereafter, the beads were subjected to five washings using IPP buffer, followed by resuspension in 500 μL QIAzol. Finally, the immunoprecipitated RNA fragments were purified and analyzed by qRT-PCR according to QIAzol's instructions. To observe m6A enrichment in circ_SFMBT2, both beads alone and isotype IgGconjugated beads were included as a negative control.

RIP.
Harvested cells were formaldehyde-treated, resuspended in nuclei isolation buffer, and lysed for 20 min at 4°C. After 15 min of centrifugation at 2500 g, the nuclear pellets were resuspended in RIP buffer, sonicated, and subjected to incubation with the anti-Ago2 or anti-IgG antibody overnight at 4°C in a shaker. e next day, pro-teinA/G beads were applied, and the incubation was continued for 1 h. en, the beads were pelleted, washed, and subjected to RNA purification. e immunoprecipitated RNA samples were subjected to qRT-PCR assay to determine how circ_SFMBT2 could bind to YTHDF2.

RNA Pull-Down Experiment.
e biotinylated circ_SFMBT2 probe was obtained from GenePharma (Shanghai, China). C-1 beads (Life Technologies, USA) were coated with the circ_SFMBT2 probe by incubation with the probe for 2 h at RT. In the meantime, the harvested cells were lysed and incubated overnight at 4°C with the circ_SFMBT2 probe or oligo probe. e RNA complexes were isolated using an RNeasy mini kit. YTHDF2-induced enrichment of circ_SFMBT2 was assessed using qRT-PCR.

RNase R Digestion.
Total RNA was isolated from the transfected cells.
en, 3 μg of the extracted RNA was subjected to 30 min incubation with 20 U/μL RNase R at 37°C. Afterward, qRT-PCR was conducted to quantify both circ_ and linear SFMBT2.

Determination of Actinomycin D.
e cells were treated with actinomycin D for 0, 4, 8, 12, and 24 h, respectively. After incubation, the cells were harvested and subjected to total RNA extraction. circ_SFMBT2 expression in the cells was quantified using qRT-PCR.
2.14. Statistics. Statistical analysis was performed using the SPSS 26.0 software. All data were presented as mean ± SD. Comparison between multiple groups or two groups was made using one-way analysis of variance or the independent-sample Contrast Media & Molecular Imaging t-test. e expression correlation between circ_SFMBT2 and LATS2 was assessed by Pearson's correlation analysis. P < 0.05 was indicative of statistical significance.

circ_SFMBT2 Was Significantly Downregulated in NSCLC
Cells and Tissues. We first analyzed circ_SFMBT2 expression in NSCLC tissue specimens using qRT-PCR. As depicted in Figure 1(a), circ_SFMBT2 expression was markedly reduced in the tumor tissues compared to their nontumor counterparts. Likewise, circ_SFMBT2 expression was decreased in the NSCLC cells, with A549 and H1299 cell lines exhibiting the lowest expression level (Figure 1(b)). Meanwhile, we observed that circ_SFMBT2 expression was significantly increased in the cytosol relative to the nucleus (Figures 1(c) and 1(d)). Moreover, the FISH assay revealed a predominant cytoplasmic localization of circ_SFMBT2 (Figure 1(e)). ese results were indicative of an involvement of circ_SFMBT2 in NSCLC. Moreover, expression analysis of epithelial-mesenchymal transition-associated markers showed that circ_SFMBT2 overexpression led to an upregulation of MMP-9 and vimentin while downregulating E-cadherin. Notably, coexpression of circ_SFMBT2 and si-LATS2 restored the expression level of MMP-9, vimentin, and E-cadherin in the NSCLC cells (Figure 2(j)). ese observations suggested that circ_SFMBT2 negatively could regulate NSCLC cell malignancy by upregulating LATS2.

circ_SFMBT2 Participated in Regulating Hippo/YAP Pathway Activation in NSCLC Cell Lines.
To investigate how circ_SFMBT2 affects the malignant behavior of NSCLC cells, we examined the Hippo/YAP pathway component expression in the cells. Western blotting revealed marked upregulation of LATS2 and LATS1 as well as a significant downregulation of YAP in the circ_SFMBT2 group compared with the vector group. Strikingly, the altered expression of LATS1 and YAP in circ_SFMBT2-overexpressing NSCLC cells was reversed by si-LATS2 treatment ( Figure 3). Taken together, these data suggested that circ_SFMBT2 could affect NSCLC cell malignancy by activating the Hippo/YAP pathway.

YTHDF2 May Accelerate the Degradation of m6A-Modified circ_SFMBT2.
To further investigate the mechanism underlying the low expression of circ_SFMBT2 in NSCLC cells, we predicted the m6A locus in circ_SFMBT2 by using the sequence-based SRAMP database of m6A modification site predictors (https://www.cuilab.cn/sramp). Notably, circ_SFMBT2 could be retrieved at multiple methylation sites of m6A (Figure 4(a)). As shown in the MeRIP assay (Figure 4(b)), circ_SFMBT2 was specifically enriched by the anti-m6A antibody. Moreover, circ_SFMBT2 was markedly upregulated in NSCLC cells with a reduced expression of the m6A reader YTHDF2 (Figures 4(c) and 4(d)). Meanwhile, RNA pull-down experiments revealed that YTHDF2 could recognize and bind to circ_SFMBT2 (Figure 4(e)). Besides, the RNA stability assay showed that the circ_SFMBT2 level in the YTHDF2

Discussion
An increasing body of evidence shows that circRNA expression correlates with clinicopathological features of cancer patients [17]. e present study found that both circ_SFMBT2 and LATS2 were markedly downregulated in NSCLC cells and tissues. Further analyses identified a positive correlation of circ_SFMBT2 with LATS2. e large tumor suppressor gene (LATS) encodes a ser/thr kinase LATS1 or LATS2 [18]. LATS2 is abnormally expressed in numerous malignancies, including lung, breast, and prostate cancers. LATS2 downregulation could promote cancer cell growth and migration [19]. Particularly, LATS2 was shown to act as a key factor in regulating proliferation, EMT, invasion, and metastatic ability of NSCLC cells [20]. In this study, circ_SFMBT2 overexpression inhibited the expression of LATS2 in the NSCLC cell lines. While cachexia was significantly reduced in cells with an increased expression of circ_SFMBT2, silencing of LATS2 restored the cachexia. In most cases, circ_SFMBT2 was shown to be oncogenic in cancers [21]. e present study demonstrates that circ_SFMBT2 inhibits cancer, showing functional diversity [22].
LATS2 is considered a vital Hippo/YAP pathway component [23]. e Hippo/YAP pathway is critically involved in organ size control. Multiple studies have found that YAP is upregulated and nuclear-localized in various   cancers, while it is capable of promoting stem cell differentiation and renewal during tumor transformation [24]. In particular, while YAP is expressed in NSCLC tissues, YAP overexpression is linked to cancer development and poor prognosis [25]. In addition, YAP could facilitate tumor invasion and metastasis, and drug resistance in NSCLC [26].
In this study, we observed that circ_SFMBT2 overexpression led to a marked upregulation of LATS2 and LATS1, and a significant downregulation of YAP in the cells, while knocking down LATS2 reversed the altered expression of LATS2, LATS1, and YAP in the circ_SFMBT2-overexpressing cells. ese observations led us to conclude that circ_SFMBT2 could regulate the malignant behaviors of NSCLC cells presumably by affecting LATS2 expression. m6A is the most prevalent and reversible post-transcriptional modification in mRNA and circRNA [27]. It has recently been shown to be involved in the metabolism of circRNAs [28]. An m6A reader YTHDF2 binds to its target RNA molecules, and the m6A level in the target RNA can affect the binding between YTHDF2 and its targets [29]. Herein, we showed that while m6A was enriched in circ_SFMBT2, circ_SFMBT2 was significantly upregulated in YTHDF2-knockdown cells. Moreover, RIP and RNA pull-down experiments revealed that YTHDF2 could recognize and bind to circ_SFMBT2. Collectively, these findings proved that both circ_SFMBT2 and YTHDF2 play a key role in NSCLC development.

Conclusion
is study demonstrated that circ_SFMBT2 was lowly expressed in NSCLC, while it could potentially serve as a biomarker to predict NSCLC development. At the same time, we presented data that YTHDF2 induced the degradation of m6A-modified circ_SFMBT2 and enhanced NSCLC cell proliferation and metastasis by activating the Hippo/YAP pathway. ese findings could facilitate our understanding of the effect of m6A modification on circRNA function and mechanisms of NSCLC.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.
Ethical Approval e study protocol was approved by the Research Ethics Committee of Seventh People's Hospital of Shanghai University of TCM (2020-IRBQYYS-011).

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
Jing Xu, Yan Shang, Xiong Qin, and Yun Gai contributed equally to this work.