The Role of ARL4C in Erlotinib Resistance: Activation of the Jak2/Stat 5/β-Catenin Signaling Pathway

Cancer patients who initially benefit from Erlotinib, a drug targeting EGFR path, eventually develop resistance to the drug. The underlying mechanism is largely unknown. This study investigated the role of ARL4C in Erlotinib resistance development of NSCLC. qRT-PCR and Western blotting were performed to analyze the expression of mRNA and protein of ARL4C in two NSCLC cell lines (HCC827 and PC-9). Several assays (MTS, colony formation, transwell migration, luciferase reporter, and chromatin-immunoprecipitation) were used to explore the role of ARL4C in biofunctional changes of Erlotinib-resistant cells and their associations with Jak2/Stat 5/β-catenin signaling. Results demonstrated that (1) long-term use of Erlotinib resulted in downregulation of ARL4C; (2) overexpression of ARL4C could regain the sensitivity to Erlotinib in the drug-resistant HCC827/ER cells, while downregulation of ARL4C increased HCC827, and PC-9 cells' resistance to the drug; (3) Erlotinib-induced downregulation of ARL4C resulted in phosphorylation of Jak2/Stat5 and upregulation of β-catenin and their related molecules Axin2, CD44, Ccnd1, Lgr-5, and MMP7, which promoted the malignant behaviors of Erlotinib-resistant cells; (4) chromatin immunoprecipitation and luciferase reporter assay revealed that Stat5 could bind to β-catenin promoter to upregulate molecules to maintain the malignant behaviors, which might count for how Erlotinib-resistant cell survived while EGFR path was blocked; (5) the expression of ARL4C was not associated with known EGFR gene mutations in both Erlotinib-resistant cells and NSCLC tissues. Our data suggest that Erlotinib resistance of NSCLCs is associated with downregulation of ARL4C via affecting Jak/Stat/β-catenin signaling. ARL4C could serve as a biomarker to predict the effectiveness of TKI targeting therapy and a potential therapeutic target for overcoming Erlotinib resistance in NSCLC.


INTRODUCTION
Lung cancer is the most commonly diagnosed malignancy and a leading cause of cancer death worldwide (1). Nonsmall cell lung cancer (NSCLC) accounts for ∼80% of all lung cancer cases (2). EGFR (Epidermal Growth Factor Receptor) tyrosine kinase inhibitors (TKIs) are a group of important targeting drugs for the treatment of NSCLCs with EGFR mutations, including exon 19 deletions or L858R substitutions. However, the acquired drug resistance occurs within 1 year after the treatment with the first generation of EGFR-TKI, which is related partially with the T790M secondary gatekeeper mutation (3), or activation of other alternative pathways, such as HGF/Met (hepatocyte growth factor/mesenchymal-epithelial transition factor), HER2 (Erb-B2 Receptor Tyrosine Kinase 2), and PIK3CA (Phosphatidylinositol-4,5-Bisphosphate 3-Kinase Catalytic Subunit Alpha) (4). However, mechanisms of TKI resistance in about 30% cases remain unclear (5). Seeking for the unknown mechanisms is a hot research topic in lung cancer research (6).
With selected methods, we found that downregulation of ARL4C (ADP-ribosylation factor-like 4C) might be associated with TKI resistance. ARL4C, also known as arl7, is a 192-aminoacid protein belonging to a small GTP enzyme. It is one of the subfamily members of ADP ribosylation factor and plays an important role in vesicle transport and signal transduction (7). The biofunction of ARL4C is not very clear yet. A study conducted by Su and his colleagues showed that high-level expression of ARL4C could inhibit the migration of ovarian cancer cells. Patients with high level of ARL4C mRNA had a good prognosis (8). Another study with gastric cancer showed that the expression of ARL4C was abnormal and the molecule was involved in tumor cell growth and cell migration (9). Using immunohistochemical analysis, Fujii et al. showed that ARL4C was abnormally expressed in lung cancer tissues. It was involved in the proliferation and invasion of lung cancer cells in vitro and in vivo (10). However, the role of ARL4C in TKI resistance is unexplored.
In this study, we investigated the role of ARL4C in TKI resistance of NSCLC cells by analyzing its functions with various assays. Our data demonstrated for the first time that ARL4C contributed to TKI resistance by activation of the JAK/STAT (Janus Kinase/signal transducer and activator of transcription) signaling pathway for survival when EGFR path blocked. The results suggest that ARL4C could be a promising biomarker for patients who likely benefit from TKI-based targeting therapy.

NSCLC Tissue Samples
NSCLC tissues from 42 patients (22 men and 20 women with a median age of 60.5 years old) who have undergone lung cancer resection in Fujian Cancer Hospital between January 2008 and June 2009 were collected. None of the patients received chemotherapy before surgery, and cancer tissues were obtained immediately after surgical resection. One part of every cancer specimen was frozen at −80 • C for measuring ARL4C mRNA level. The remaining part of the tissues was fixed in formalin and embedded in paraffin (FFPE) for detecting EGFR mutations.

Selection of TKI Erlotinib-Resistant Cell Line
HCC827 and PC-9 (EGFR 19del) cell lines purchased from ATCC were cultured in H1640 medium, containing 10% FBS (GIBCO BRL, Rockville, MD, U.S.A. Cat No. 10099233), 100 U/ml penicillin, and 100 µg/ml streptomycin in an incubator containing 5% CO 2 at 37 • C. The cells were exposed to gradually increased concentration of Erlotinib (Selleck, Houston, TX, USA. Cat No. S1023) from 0, 100, 200, 400, 800, 1,600, and 3,200 nM in their culture medium. After passing 17 generations in 6 months of the selection, Erlotinib was removed from the medium. The cells growing in 3,200 nM of Erlotinib were labeled as TKI Erlotinib-resistant (ER) cell lines HCC827/ER and PC-9/ER, respectively.
Relative mRNA expression levels of ARL4C were quantified in all lung cancer samples using the comparative 2 − Ct method and lung cancer sample T27 as reference reported previously (11,12). The housekeeping gene glyceraldehyde phosphate dehydrogenase (GAPDH) was used to normalize expression levels of ARL4C.
Protein Levels of ARL4C, β-Catenin, and JAK/STAT Signaling Cells were harvested and lysed using RIPA buffer (50 mM Tris-Cl, pH 8.0, 150 mM NaCl, 5 mM EDTA, 0.1% SDS, and 1% NP-40) supplemented with protease inhibitor cocktail (Abcam, Cambridge, MA, USA. Cat No. ab65621). The cell lysates were centrifuged at 12,000 rpm for 30 min at 4 • C. Supernatants were collected and protein concentrations were determined by BCA protein assay (Thermo Scientific, Rockford, Illinois, USA). The supernatants (each with 25 µg proteins) were electrophoresed on 10-12% polyacrylamide gel with sodium dodecyl sulfate (SDS) and then transferred onto nitrocellulose membranes (Millipore, Burlington, MA, USA) at 100 V for 1. 4322) overnight at 4 • C, respectively. After wash, the membranes were further incubated with horseradish peroxidase-conjugated anti-rabbit antibody. Finally, protein bands were developed with the enhanced chemiluminescence Western blot detection kit Immobilon ECL Ultra Western HRP Substrate (Millipore, Bedford, MA, USA. Cat No. WBULS0500), and images were captured on image station 4,000 mm pro (Carestream, Canada). Image J program was used to quantify the protein band intensity relative to loading control of β-actin.

Effects of Erlotinib on Cell Proliferation and IC50
Cells were plated into 96-well-plates (5,000 cells/well) and grew overnight. Different concentrations of Erlotinib (0, 100, 200, 400, 800, 1,600, and 3,200 nM) were added into the cell cultures in triplicate. After culturing for 72 h, 20 µl of MTS (Promega, Madison, WI, USA. Cat No. g3582) and 100 µl of serum containing medium were added to each well-followed by incubation at 37 • C for 2 h. H1640 only with serum was used as background control. The absorption of the palates at 490 nm was read on a Bio-Rad (Hercules, CA, USA) plate reader (model 680). The proliferation of Erlotinib-treated cells was normalized with the control cells (no Erlotinib). IC50 was calculated with SPSS17.0. The experiment was repeated three times.

Colony Formation Assay
For colony-formation assay, cells (about 500 cells/well) were seeded and grew in six-well-plates for 48 h before adding Erlotinib. After 14 days, the cells were fixed in methanol and stained with 0.2% crystal violet. Cell colonies (>50 cells/colony) were pictured with Image Scanner (GE, Piscataway, NJ, USA). The number of colonies in each well was counted using Image J.

Transwell Invasion Experiment
Cell concentration was adjusted to 7 × 10 5 /ml, and 100 µl of the cell suspension was placed onto the upper chamber of each well on 24-well-Transwell plates coated with 1 mg/ml fibronectin (Millipore). The lower chamber contained medium with 20% FBS. After incubation at 37 • C for 48 h, the cells in the upper chamber were wiped off with cotton swabs, and the cells in the other side of chamber membrane were fixed with methanol for 15 min, dried, stained with 0.1% crystal violet, and randomly pictured with a magnification of ×200 under inverted microscope (Olympus, Japan). The cells were counted with Image J program and the average number of the cells from 15 fields was used. The experiment was repeated three times.

Chromatin-Immunoprecipitation (ChIP)
ChIP assay was carried out using a SimpleChIP Enzymatic Chromatin IP Kit (Cell Signaling Technology, Beverly, MA, USA, Cat No. 9002) following the manufacturer's instruction. Briefly, 5 × 10 6 cells were fixed with 1% formaldehyde and quenched in 0.125 M glycine. Cells were sonicated by Bioruptor Sonication System UCD-300. DNA was immunoprecipitated by either control IgG or phospho-stat5 antibody. Precipitated DNA samples and inputs were amplified by PCR. The primers used for the amplification of stat5 binding site in β-catenin promoter are 5 ′ -cctcttccccgttgtttcca-3 ′ (sense) and 5 ′ -ggggtgattctttgctaattt ca-3 ′ (antisense).

Detection of EGFR Mutations in Both NSCLC Samples and Cell Lines
Two methods were used to detect EGFR mutations. For 42 NSCLC paraffin specimens, DNA was obtained using a paraffin tissue DNA Extraction kit (Qiagen, Hilden, Germany. Cat No. 56404). The concentration of DNA was adjusted to 1 ng/ml, and EGFR mutations were detected using the amplification refractory mutation system (ARMS) with human EGFR Mutations Detection kit (Amoy Diagnostics, Xiamen, China. Cat No. ADx-EG01) according to the manufacturer's protocol as previously described (13). Briefly, ARMS-PCR assay was performed in a 50-µl volume containing 5 µl of PCR buffer, 10 pM forward and reverse primers, 20 pM probe, and 12.5 mM dNTPs. The thermocycling conditions were as follows: 95 • C for 5 min, then 15 cycles of 95 • C for 25 s, 64 • C for 20 s, and 72 • C for 20 s, followed by 31 cycles of 93 • C for 25 s, 60 • C for 35 s, and 72 • C for 20 s.
To determine if there is any association between ARL4C expression and EGFR mutations, eight cell lines with different levels of ARL4C were examined with the next-generation sequencing (NGS). DNA was extracted using the GONOROAD kit (Qiagen, Hilden, Germany) and 200 ng of DNA was used to build the library using NEBNext Ultra II DNA library Prep Kit for Illumina (NEB, Ipswich, MA, USA). Integrated DNA technologies (IDT, Skokie, IL, USA) customized probes were used for hybridization capture. All libraries were performed on an MGISEQ2000 instrument according to the manufacturer's instructions (BGI, Shenzhen, Guangdong, China) with 200 cycles, standing for paired-end 100 bp. Then, IDT 10-hotspot gene panels from all eight libraries were used, which included ALK, BRAF, EGFR, ERBB2, KRAS, MET, NRAS, PIK3CA, RET, and ROS1, and were quantitated using Library Quantification Kit-Illumina/Universal (Kapa Biosystems, Wilmington, MA, USA) on an ABI 7500 Real Time PCR system (Applied Biosystems, Waltham, MA, USA). The mutations were detected by the following methods: the Trimmomatic (version 0.39, parameter: PE -threads 4 -phred33 ILLUMINACLIP: adapter.fa:2:30:10 MINLEN:15), and the adapter sequences were AAGTCGGAGGCCAAGCGGTCT TAGGAAGACAA and AAGTCGGATCGTAGCCATGTCGTT CTGTGAGCCAAGGAGTTG and were used to narrow down the raw sequencing data (Fastq), filtering out the adapter contamination reads and low-quality reads to get clean data. Bwa Aln (Version: 0.7.12-r1039) algorithm was used to align the clean data of the human reference genome (hg19) and to get the Sequence Alignment/Map format (sam) file. For the Binary Alignment/Map format (bam) file, the sam file was sorted and the artificial duplication reads were removed by samtools (Version: 0.1.19-44428cd). According to the bed interval file of the 10-hotspot gene panels, Freebayes (version: v1.0.2-6-g3ce827d, parameter: -j -m 10 -q 30 -F 0.001 -C 1 -t bed.file -f hg19.fa) was used to determine the single-nucleotide polymorphisms (SNPs) and insertions or deletions (indels) and then ANNOVAR was used for the annotation.

Statistical Analysis
All data were presented as mean ± SD. Student's t-test was used for analysis. ANOVA was used to determine the statistical difference between or among different experimental groups. The value P < 0.05 was considered as statistically significant difference.

TKI Resistant NSCLC Cell Lines Expressed a Low Level of ARL4C
Two NSCLC cell lines, HCC827 and PC-9, were subjected to Erlotinib selection in culture. After 6 months, IC50 of Erlotinib was significantly increased from 289 to 1,843 nM for HCC827 and from 71.08 to 5232.12 nM for PC-9. These Erlotinib-resistant (ER) cell lines were named HCC827/ER and PC-9/ER, which were 5.37 and 73.60 times more resistant to Erlotinib than their parental cells, respectively (Figures 1A-C). However, results of qPCR and Western blotting demonstrated that the levels of ARL4C were significantly lower in HCC827/ER and PC-9/ER cells than in their parental cells (Figures 1D-G, P < 0.0001). Furthermore, β-Catenin expression significantly increased in HCC827/ER and PC-9/ER, compared with their parental cells (Figures 1D-G, both P < 0.0001). Collectively, these results showed that the expression of ARL4C was reduced in TKIresistant cells (HCC827/ER, PC-9/ER), while the expression of β-catenin was elevated in TKI-resistant cells.

Knocking Down ARL4C Resulted in Enhancing Erlotinib Resistance and Increasing Cell Migration
To confirm that the downregulation of ARL4C was associated with the increased Erlotinib resistance, sh-ARL4C-knockout lentivirus was used to infect HCC827 and PC-9 cells. Western blotting showed that the expression level of ARL4C in the infected cells HCC827/ARL4C-sh and PC-9/ARL4C-sh were significantly lower than those in control cells infected with lentivirus vector alone (HCC827/Vector and PC-9/Vector) (Figures 2A-C Effect of low ARL4C on Erlotinib IC50 was examined with MTS assay. Results showed that knockdown of ARL4C significantly increased Erlotinib resistance as compared to their parental cells. IC50 values of Erlotinib for HCC827, HCC827/vector, and HCC827/ARL4C-sh were 222.8, 393.2, and 1236.2 nM, respectively. The values for PC-9, PC-9/vector, and PC-9/ARL4C-sh were 71.2, 68.23, and 877.97 nM, respectively (Figures 2D-F The cells with knockdown of ARL4C (HCC827/ARL4Csh and PC-9/ARL4C-sh) had a marked enhancement of proliferation and colony formation, compared to their parental HCC827 and PC-9 cells when treated with Erlotinib (Figures 2G-J, both P < 0.001). These ARL4C low-expressing cell lines HCC827/ARL4C-sh and PC-9/ARL4C-sh cells also These results suggest that knockdown ARL4C enhances TKI Erlotinib tolerance, colony formation, and migration of NSCLC cells.

Overexpression of ARL4C in Erlotinib-Resistant HCC827/ER and PC-9/ER Cells Inhibited Their Tolerance to Erlotinib and Cell Migration
To further confirm the relationship between Erlotinib resistance and ARL4C expression, both Erlotinib-resistant HCC827/ER and PC-9/ER cells were infected with ARL4C overexpression lentivirus to regain the ARL4C expression. Results of qPCR and Western blot showed that the expression levels of ARL4C in HCC827/ER/ARL4C-OE and PC-9/ER/ARL4C-OE were significantly higher than those of vector alone control cells, HCC827/ER/Vector, and PC-9/ER/Vector (Figures 3A-C To examine the effect of high level of ARL4C on cell susceptibility to Erlotinib, the IC50 was measured with MTS assay. Results showed that the overexpression of ARL4C significantly decreased Erlotinib resistance of both HCC827/ER and PC-9/ER cells. IC50 values of Erlotinib for HCC827/ER, HCC827/ER/Vector, and HCC827/ER/ARL4C-OE cells were 1.85, 1.71, and 0.47 µM, respectively. The values for PC-9/ER, PC-9/ER/Vector, and PC-9/ER/ARL4C-OE cells were 5232.1, 5146.1, and 928.1 nM, respectively (Figures 3D,E, both P < 0.05).
Similarly, the effect of high level of ARL4C on cell colony formation and migration in the presence of Erlotinib was tested. Results showed that the overexpression of ARL4C in HCC827/ER/ARL4C-OE and PC-9/ER/ARL4C-OE cells resulted in an reduced sensitivity to Erlotinib as compared to HCC827/ER and PC-9/ER cells (Figures 3F-I, P < 0.01, P < 0.001, and P < 0.0001). The migration of HCC827/ER/ARL4C-OE and PC-9/ER/ARL4C-OE was significantly decreased. The number of cells passing through the polyester membrane decreased,  These data suggest that overexpression of ARL4C in Erlotinibresistant lung cancer cells increases the sensitivity of the cells to TKI Erlotinib and inhibits cell migration.

Enhance of β-Catenin Is Critical for ARL4C-Associated Erlotinib Resistance
It has been shown that cells resistant to Erlotinib expressed a high level of β-catenin (12). Our results also showed that ARL4C was expressed at a significantly low level in Erlotinibresistant HCC827/ER and PC-9/ER cells, compared to Erlotinibsensitive HCC827 and PC-9 cells (Figures 1D-G, P < 0.0001), while β-Catenin expression significantly increased in Erlotinibresistant cells (Figures 1D-G, P < 0.0001). Therefore, we speculate that β-catenin might contribute to ARL4C-associated Erlotinib resistance.
ARL4C Regulates β-Catenin Expression Through JAK2/STAT5A Signal Activation Previous studies showed that several relevant downstream signals, such as ERK, JAK, and AKT, might be involved in tumor-promoting effects of ARL4C (14). To further explore potential mechanisms related to the acquired TKI resistance, phosphorylation levels of JAK were examined in ARL4C knockdown-HCC827 cells. Results showed that the phosphorylation level of JAK2 was significantly increased, compared to that in control cells ( Figure 5A). Consistently, the overexpression of ARL4C significantly decreased the phosphorylation of JAK2 in HCC827/ER cells, compared to that in control cells ( Figure 5B). The JAK-STAT signaling pathway not only regulates tumor development but also is closely related to TKI resistance (14). These results indicated that ARL4C abnormal expression affected the phosphorylation of JAK2.
We further studied if ARL4C regulated STAT5A. In HCC827 cells, the knockdown of ARL4C upregulated the phosphorylation of STAT5A as detected by Western blot analysis (Figure 5C). In contrast, the overexpression of ARL4C in HCC827/ER cells downregulated the phosphorylation of STAT5A (Figure 5D).
Chromatin immunoprecipitation showed that the overexpression of ARL4C resulted in the loss of the promoter of STAT5 binding to CTNNB1 (Figure 5E). We found that luciferase expression directed by a 200-bp fragment of the β-catenin promoter containing this STAT5A site was increased in ARL4C downregulated HEK293T cell (Figure 5F). Similarly, downregulation of ARL4C resulted in a significant enhancement of β-catenin promoter activity in HCC827 cells ( Figure 5G). These results indicated that the β-catenin promoter contained a conserved STAT5A binding site. STAT5A acts as a transcription factor in regulating the mRNA expression of β-catenin by binding and activating the β-catenin promoter.
These data suggest that the regulation of β-catenin expression by ARL4C might be through the JAK2/STAT5A signal pathway in the regulating TKI resistance of NSCLC.

Detection of EGFR Mutations
To explore the association between EGFR mutation and ARL4C, EGFR mutations in 42 NSCLC paraffin samples were detected with amplification refractory mutation system (ARMS) with human EGFR Mutations Detection kit. Among 42, 20 were found with EGFR mutations and only 1 had a T790M mutation, while the other 21 had no EGFR mutation (in Supplementary Table 2). In these 42 tissues, mRNA expression level of ARL4C in the EGFR mutation group was −0.064 ± 0.091 (lgRQ, RQ = 2 − Ct ), while the level in the group without EGFR mutation was −0.023 ± 0.099, indicating that the expression of ARL4C was not associated with EGFR gene mutations (Figure 6C, P = 0.7668).

DISCUSSION
In this study, the underlying mechanism by which the resistance of Erlotinib, an EGFR-TKI, in NSCLC was explored. The study revealed for the first time that ARL4C was downregulated and β-catenin was upregulated in Erlotinib-resistant cells via JAK2/STAT5A signaling pathway. By manipulating the expression of these two genes (overexpression or knocking down expression), the study provided evidences that the inhibition of ARL4C selectively enhanced the resistance of HCC827 and PC-9 cells to Erlotinib, while the overexpression of ARL4C enhanced the sensitivity of the cancer cells to the drug. The regulation of the drug resistance of lung cancer cells by ARL4C was through activating the β-catenin/JAK2/STAT5A signaling pathway. These data indicate for the first time that ARL4C plays an important role in the resistance of NSCLC cells to Erlotinib.
Several mechanisms that mediate TKI resistance have been identified: (A) EGFR exon 20 T790M mutation (14); (B) Met gene amplification (15); (C) HER2 gene amplification (16); and (D) PIK3CA mutation (4), which counts for the majority of gene mutations responsible for the TKI resistance. However, one third of causes of the TKI resistance remains unknown. Our data (Figure 6) showed that Erlotinib-induced ARL4C reduction did not associate with any currently known mutations of EGFR, indicating that ARL4C is an independent factor for the development of Erlotinib resistance.
EGFR path is one of the major oncogenic pathways via Ras/Raf/MEK/ERK signaling to promote the cancer cell proliferation, migration, and invasion. When the EGFR path is blocked by Erlotinib, the cancer cells have to develop another oncogenic path for their continuous survival. Our data supported that the downregulation of ARL4C upregulated β-catenin via activation of JAK2/STAT5, which could help the EGFR-path-blocked cancer cells to regain their malignant behaviors ( Figure 6D).
ARL4C is a member of the ADP-ribosylation factor family of GTP-binding proteins. The abnormal expression of ARL4C possesses carcinogenic effect on many types of tumors (17). Recent studies have indicated that ARL4C could be induced by EGF and promotes the motility of cancer cells by remodeling actin cytoskeleton through Arf6 activation (18), which might be involved in the migration, invasion, and proliferation of cancer cells (18). Consistent with this, our data showed that when EGF path was blocked by Erlotinib, the ARL4C was reduced, which triggered the upregulation of the JAK2/STAT5/β-catenin path to replace the EGFR oncogenic signal path. Similarly, our data supported the idea that the inhibition of ARL4C in wild-type HCC827 and PC-9 cells via sh-RNA could selectively enhance the Erlotinib tolerance, while the overexpression of ARL4C could enhance the Erlotinib sensitivity. The results indicate that ARL4C could be a "switch" between the EGFR path and the JAK2/STAT5/β-catenin path to maintain the proliferation, migration, and invasion of cancer cells when cells are exposed to Erlotinib and EGFR path is blocked.
It is well-known that Wnt/β-catenin signaling is a classic pathway with a crucial role in NSCLC progression. β-catenin is a key component in the Wnt signaling cascade and is involved in cancer cell proliferation and tumor progression (19). In the present study, a significant upregulation of β-catenin was observed in Erlotinib-resistant HCC827/ER and PC-9/ER cells compared with their parental cells HCC827 and PC-9. The result is in accordance with a previous observation that Wnt/β-catenin not only participates in the proliferation, invasion, and metastasis of tumor cells but also induces drug resistance (20). Nuclear accumulation of β-catenin was associated with EGFR mutations (21) and β-catenin overexpression was associated with NSCLC cell resistance to gefitinib (20). DDX17 nucleocytoplasmic shuttling promotes acquired gefitinib resistance in NSCLC cells via activation of β-catenin (22). Rab25 promotes Erlotinib resistance by activating the β1 integrin/AKT/β-catenin pathway in NSCLC (23). Alone this line, our finding added another path, ARL4C/JAK2/STAT5/β-catenin, as a new means for cancer cells to escape the survival inhibition by Erlotinib. Our data showed that in the presence of Erlotinib, the overexpression of β-catenin significantly increased the proliferation and decreased Erlotinib sensitivity in HCC827/ER-OE and PC-9/ER-OE (Figures 4H-K). Thus, ARL4C-induced Erlotinib resistance was via β-catenin signaling.
It is well-known that the JAK-STAT signaling pathway serves a crucial role in cell immunity, division, death, and tumor formation (24). The two families involved in this pathway are Janus kinases (JAKs) and signal transducer and activator of transcription proteins (STATs), encoded by the genes JAK (JAK1, JAK2, JAK3, and TYK2) and STAT (STAT1, STAT2, STAT3, STAT4, STAT5A, STAT5B, and STAT6), respectively (25). The continuous activation of JAK/STAT could promote malignant transformation of cells, leading to the development of cancers including NSCLC (21). The inappropriate activation of Stat3 or IL-7-mediated IL-7R-JAK3/STAT5 pathway is associated with an unfavorable prognosis in NSCLC patients and correlated with chemoresistance and radioresistance (24). Inhibition of the JAK/STAT3 signaling pathway has therefore been recognized as a promising therapeutic strategy for NSCLC (25). In addition, it was reported that (1) the inhibition of gp130-Jak-Stat3 signaling could partially inhibit Wnt-β-catenin-mediated intestinal tumor growth and regeneration (26); (2) Stat3 and βcatenin were involved in tumorigenesis (27). By using chromatin immunoprecipitation and luciferase reporter assays, this study revealed for the first time that the promoter region of β-catenin contained a conserved STAT5A binding site. STAT5A acted as a transcription factor regulating the mRNA expression of β-Catenin by binding to and activating the β-catenin promoter, forming ARL4C/JAK2/STAT5/β-catenin axis for the survival of Erlotinib-resistant cells.
Our data also supported that the biofunctions of ARL4C/JAK2/STAT5/β-catenin axis were carried out by a set of molecules, such as Axin2, CD44, Ccnd1, Lgr-5, and MMP7. Axin2 is responsible for the stability of β-catenin (28); CD44 is a famous stem cell marker and adhesion molecule to enhance the proliferation, migration, and invasion (29); Ccnd1 enhances the proliferation of cancer cells via promoting the cell's G1/S transition (30); Lgr-5 is a biomarker for stem cells, involved in tumor development (31). MMP7 promotes the migration and invasion of cancer cells (32). All these functional molecules were upregulated in Erlotinib-resistant cells when ARL4C was suppressed and JAK2/STAT5/β-catenin axis was activated (Figure 6D), serving as executors for the malignant behaviors when the EGFR path was blocked in NSCLC cells.

CONCLUSION
This study demonstrated for the first time that the ARL4C/JAK2/STAT5/β-catenin axis and their downstream molecules Axin2, CD44, Ccnd1, Lgr-5, and MMP7 could help to bypass EGFR oncogenic path and serve as an alternative new signal/function path to maintain the survival and malignant behaviors for HCC827 and PC9 NSCLC cells under the stress of Erlotinib.

DATA AVAILABILITY STATEMENT
The datasets generated for this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: NCBI BioProject (PRJNA648039) (https://www.ncbi.nlm.nih.gov/ Traces/study/?acc=PRJNA648039).

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Research Ethics Committee of the Fujian Provincial Tumor Hospital. The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
XL, LZ, ZY, and YS designed the study, analyzed data, and wrote the manuscript. ZY, DH, and ZH collected and analyzed the NSCLC lung tissues for ARL4C expression level. JL, ZC, and CZ performed all the experiments, collected data, and created the figures. TH carried out the bioinformatics and statistical analysis. All authors contributed to the article and approved the submitted version. Funders did not have any role in study design, data collection, and data analysis.