JAC4 Inhibits EGFR-Driven Lung Adenocarcinoma Growth and Metastasis through CTBP1-Mediated JWA/AMPK/NEDD4L/EGFR Axis

Lung adenocarcinoma (LUAD) is the most common lung cancer, with high mortality. As a tumor-suppressor gene, JWA plays an important role in blocking pan-tumor progression. JAC4, a small molecular-compound agonist, transcriptionally activates JWA expression both in vivo and in vitro. However, the direct target and the anticancer mechanism of JAC4 in LUAD have not been elucidated. Public transcriptome and proteome data sets were used to analyze the relationship between JWA expression and patient survival in LUAD. The anticancer activities of JAC4 were determined through in vitro and in vivo assays. The molecular mechanism of JAC4 was assessed by Western blot, quantitative real-time PCR (qRT-PCR), immunofluorescence (IF), ubiquitination assay, co-immunoprecipitation, and mass spectrometry (MS). Cellular thermal shift and molecule-docking assays were used for confirmation of the interactions between JAC4/CTBP1 and AMPK/NEDD4L. JWA was downregulated in LUAD tissues. Higher expression of JWA was associated with a better prognosis of LUAD. JAC4 inhibited LUAD cell proliferation and migration in both in-vitro and in-vivo models. Mechanistically, JAC4 increased the stability of NEDD4L through AMPK-mediated phosphorylation at Thr367. The WW domain of NEDD4L, an E3 ubiquitin ligase, interacted with EGFR, thus promoting ubiquitination at K716 and the subsequent degradation of EGFR. Importantly, the combination of JAC4 and AZD9191 synergistically inhibited the growth and metastasis of EGFR-mutant lung cancer in both subcutaneous and orthotopic NSCLC xenografts. Furthermore, direct binding of JAC4 to CTBP1 blocked nuclear translocation of CTBP1 and then removed its transcriptional suppression on the JWA gene. The small-molecule JWA agonist JAC4 plays a therapeutic role in EGFR-driven LUAD growth and metastasis through the CTBP1-mediated JWA/AMPK/NEDD4L/EGFR axis.


Introduction
Non-small-cell lung cancer (NSCLC) is the most common form of lung cancer, accounting for about 85% of lung cancers, and remains the leading cause of cancer death worldwide, especially lung adenocarcinoma (LUAD) [1]. The targeted therapy and immunotherapy using checkpoint inhibitors have shifted the treatment paradigm for NSCLC [2]. There is no evidence that immunotherapy can prolong the overall survival of patients with stage I to III NSCLC who receive radiotherapy or surgery, and adverse events are more likely to occur in those who receive immunotherapy [3]. In addition, compared with chemotherapy alone, the combined therapy (chemotherapy plus immunotherapy) only extends overall

Higher Expression of JWA Is Associated with Better Prognosis of LUAD
To obtain population data on JWA expression in lung cancer, we analyzed transcriptome and proteome data from lung cancer. Analysis using UALCAN (http://ualcan.path. uab.edu/analysis.html, accessed on 22 November 2021) showed that JWA mRNA expression in lung cancer was significantly lower than that in normal lung tissues in TCGA samples ( Figure 1A). Furthermore, by analyzing the Gene Expression Omnibus (GEO), we also found that JWA mRNA expression was higher in normal lung tissues than in paired lung-cancer tissues (GSE19804) and non-paired lung-cancer tissues (GSE19188) ( Figure 1B,C). In the lung squamous-cell-carcinoma (LUSC) dataset, lower JWA mRNA expression was detected at stages ranging from moderate dysplasia to invasive carcinoma compared to normal tissue samples ( Figure S1A). Similarly, we used CPTAC proteomic data to analyze the potential difference of JWA protein expression in different stages of lung adenocarcinoma (LUAD). The data showed that JWA protein expression was significantly reduced in LUAD compared to normal lung tissue, and JWA protein levels gradually decreased with increasing tumor stage in LUAD ( Figure 1D,E). Moreover, JWA protein expression was lower in LUSC than in the corresponding normal tissues ( Figure S1B). Meanwhile, Human Protein Atlas data showed that 65% of lung-cancer patients had low or undetectable expression in 20 lung-cancer samples with JWA staining (Figures 1F,G and S1C). Taken together, these results suggest that JWA is downregulated in lung-cancer tissues both in mRNA and in protein levels. In addition, Western blotting showed that JWA protein expression was significantly higher in human bronchial epithelial (HBE) cells than in the non-small lung-cancer cell lines A549, SPCA1, PC9, and H1299 ( Figures 1H and S1D), which is consistent with JWA mRNA expression in cell lines by qRT-PCR ( Figure 1I). Importantly, receiver-operating-characteristic (ROC) analysis indicated that JWA could be used as an independent prognostic biomarker in lung cancer ( Figure 1J). Moreover, patients with higher JWA expression had better prognostic outcomes ( Figure 1K). The Kaplan-Meier Plotter analysis (http://kmplot.com/analysis/, accessed on 6 January 2022) further showed that patients with high JWA expression had longer times in overall survival (OS), suggesting that high JWA levels are associated with better survival in lung cancer ( Figure S1E). To investigate the localization of JWA in organs, we assessed the protein expression of JWA across normal human tissues by mass spectrometry from the Human Protein Map (www.humanproteomemap.org, accessed on 20 January 2022), and the data showed that JWA was mainly enriched in lung tissues ( Figure S1F). Collectively, these results suggest that higher JWA expression is associated with a better prognosis in individuals with lung cancer and that JWA may be a promising potential biomarker of LUAD.

Screening of JWA Small-Molecule Agonist JAC4 for Suppression of LUAD Proliferation and Metastasis
To determine whether the agonists of the JWA gene can selectively activate the expression of JWA in lung-cancer cells, two small-molecular compound agonists, JAC1 [30]and JAC4 [31] (Figure S2A), were used in experimental models. Considering the tumor heterogeneity and drug sensitivity, the subcutaneous tumor-bearing model was first constructed to screen for a better tumor-inhibition effect in lung cancer. Based on the tumor-

Screening of JWA Small-Molecule Agonist JAC4 for Suppression of LUAD Proliferation and Metastasis
To determine whether the agonists of the JWA gene can selectively activate the expression of JWA in lung-cancer cells, two small-molecular compound agonists, JAC1 [30] and JAC4 [31] (Figure S2A), were used in experimental models. Considering the tumor heterogeneity and drug sensitivity, the subcutaneous tumor-bearing model was first constructed to screen for a better tumor-inhibition effect in lung cancer. Based on the tumorvolume and tumor-weight curves, JAC4 showed better anti-proliferation activity than JAC1 (Figure 2A-D). Moreover, Western blotting showed that tumors in the JAC4 treatment groups exhibited higher JWA and BAX protein expression and lower PCNA and BCL2 protein expression ( Figure 2E). In addition, immunohistochemistry (IHC) staining showed that decreased Ki67 expression, increased JWA, and cleaved caspase 3 expressions were detected in the JAC4 treatment groups compared to those in control groups ( Figure 2F,G). Importantly, JAC4 treatment did not change the body weight of mice ( Figure S2B), and histological evaluation of the heart, liver, spleen, lung, and kidneys showed no signs of toxicity in the JAC4-treated mice ( Figure S2C). Moreover, JAC4 treatment did not cause an elevation of serum levels of ALT, AST, BUN, CK, CK-MB, or LDH ( Figure S2D). Collectively, these findings suggest that JAC4 inhibited lung-cancer progression in vivo, with no obvious toxicity. To investigate whether JAC4 suppresses lung-cancer progression in cells, we treated A549 cells with different concentrations of JAC4. Results show that 10 µM JAC4 treatment for 24 h had the best activation effect on JWA expression ( Figure S2E). Colony formation and 5-ethynl-2 -deoxyuridine (EDU) staining assays demonstrated that the cells with JAC4 treatment reduced proliferation and growth potential compared to DMSO-treated ones ( Figure 2H-K). In contrast, JAC4 treatment had little effect on the growth of normal cells (HBE and BEAS-2B) ( Figure S2F). Moreover, JAC4 was found to significantly downregulate PCNA protein expression in a dose-dependent manner ( Figure S2G). In addition, similar to the previous observation that JWA inhibits lung-cancer metastasis [32], transwell cell migration and invasion assays showed that JAC4 treatment greatly inhibited lung-cancer cell migration and invasion compared to DMSO-treated cells ( Figures 2L-N and S2H). Taken together, these results indicate that JAC4 suppresses lung-cancer cell proliferation and metastasis in vitro. Considering that different chiral molecules may have different effects, we further used a subcutaneous-tumor-bearing model to identify which JAC4 chiral molecule has better antitumor effects ( Figure S3A,B). It was found that both the JAC4-R and the JAC4 prototype could significantly inhibit the growth of lung cancer, and JAC4-S had a weak tumor-inhibitory effect; furthermore, JAC4-R did not show a better tumor-suppression effect than the prototype ( Figure S3C-F). Therefore, the JAC4 prototype was used in all later models.

JAC4 Promotes EGFR Degradation through the Ubiquitination Proteasome Pathway
We previously reported that JWA suppresses cell migration by negatively regulating HER2 expression in HER2-positive gastric-cancer cells [33]; moreover, JWA agonist JAC1 enhances the ubiquitination of HER2 [30]. HER2 is known as a member of the ERBB family, and EGFR is a vital target in lung cancer; however, the role of JWA and related small-molecule activator JAC4 on EGFR remains unknown. To explore the underlying mechanisms of how JWA affects lung-cancer progression, Western blotting was performed, which showed that the molecular markers of EGFR, P-EGFR, P-AKT, and P-STAT3 (the downstream pathways of EGFR) were downregulated when the cells increased JWA expression; however, these markers were upregulated when JWA was knocked down ( Figure 3A). These results indicate that JWA may be a negative regulator of the EGFR pathway in lung cancer. To test whether JAC4 could downregulate EGFR in lung-cancer cells, upon JAC4 treatment, immunoblotting analysis was carried out and showed that the protein expression of JWA was increased, whereas the expression of EGFR was decreased in dose-and time-dependent manners; moreover, the downstream of EGFR-related markers (P-AKT and P-STAT3) were synchronously suppressed ( Figures 3B,C and S4A,B). Importantly, the rescue experiments showed that decreased EGFR expression caused by JAC4 treatment was blocked by JWA silence ( Figure 3D). These results suggest that the inhibitory effect of JAC4 on EGFR expression is achieved partly through upregulating JWA expression. Meanwhile, lower EGFR protein expression was detected in tumor tissues from the mice receiving JAC4 treatment than the vehicle-control mice ( Figure 3E). Moreover, immunofluorescence staining showed that the levels of EGFR were decreased in A549 and SPCA1 cells after JAC4 treatment ( Figures 3F and S4C,D). We next examined whether the reduction of EGFR was regulated at either the transcriptional or post-translational level. qRT-PCR analysis showed that the mRNA expression of EGFR was not reduced in JAC4-treated A549 cells ( Figure S4E). Clinical-patient data from the GEO databases revealed that JWA mRNA expression had no obvious correlation with EGFR expression in cancer ( Figure 3G,H). Therefore, JAC4 may regulate the expression of EGFR post-transcriptionally.    To determine whether JAC4 affects the protein stability of EGFR in cells, pulse-chase analysis using cycloheximide (CHX) was carried out and showed that compared with control cells, JAC4 reduced the half-life time of the EGFR protein in both A549 and SPCA1 cells ( Figure 3I). Moreover, the JAC4-mediated degradation of EGFR could be efficiently blocked by the proteasome inhibitor MG132 ( Figure S4F). To further determine whether JAC4 triggered EGFR ubiquitination modification, in-vitro ubiquitination assays were performed in A549 cells with ectopic expression of his-ubiquitin. Results show that ubiquitinated EGFR was increased in cells after JAC4 treatment ( Figure 3J). Furthermore, enhanced EGFR ubiquitination and reduced EGFR protein levels were observed in JWA-overexpressed A549 cells; conversely, knockdown of JWA reduced EGFR ubiquitination and increased EGFR expression ( Figure S4G,H). Collectively, these findings suggest that JAC4 may trigger ubiquitin-proteasomal degradation of EGFR in lung-cancer cells via upregulating the expression of JWA.

JAC4 Ubiquitinates EGFR by E3 Ubiquitin Ligase NEDD4L
To identify which E3 ubiquitin ligases were involved in process of EGFR ubiquitination, we first used the UbiBrowser database to predict the candidates from the human-ubiquitinligase (E3) substrate interactions [34]. The top 20 E3 ubiquitin ligases were found to be potentially involved in EGFR ubiquitination ( Figures 4A and S5A). Next, we further investigated the expression of E3 ligases by mining 60 pairs of matched lung-cancer microarray data ( Figure 4B). The mRNA level of NEDD4L but not MIB1, WWP1, or other ligases was downregulated in patients with lung cancer. These results helped us further narrow down the candidate cope of E3 ligases. Moreover, immunoblot data showed that among the top five E3 ubiquitin ligases (based on confidence scores in the UbiBrowser database), only NEDD4L could be upregulated by JAC4 treatment in human-lung-cancer cells ( Figures 4C and S5B), which was further confirmed in the JAC4-treated lung-cancer tumor-tissue samples ( Figure 4D). We then analyzed the relationship between the NEDD4L levels and the prognosis of lung-cancer patients. Through our analysis of public data regarding the GEO, TCGA, and CPTAC (Clinical Proteomic Tumor Analysis Consortium) databases, the results show that both mRNA and protein levels of NEDD4L were decreased in patients with lung cancer ( Figure S5C-F). This result is also consistent with NEDD4L staining by IHC from The Human Protein Atlas ( Figure S5G-I). Furthermore, by analyzing the GEO data, we found that patients with high NEDD4L expression had better overall survival, which is consistent with the online Kaplan-Meier Plotter tool ( Figure S5J-K). These results suggest that JAC4 can upregulate NEDD4L expression and that NEDD4L is positively associated with a better prognosis for lung-cancer patients. by knockdown of NEDD4L was reversed by JAC4 ( Figure S5Q,R). Our results suggest that JAC4 ubiquitinates EGFR by upregulating E3 ubiquitin ligase NEDD4L expression.  Substrate binding to ubiquitin ligases is a key event in protein degradation and signal transduction. We analyzed protein interactions between NEDD4L and EGFR. The molecular-docking assay showed that NEDD4L could interact with EGFR physically ( Figure 4E). Moreover, we found that ectopically expressed NEDD4L and EGFR proteins interacted with each other in HEK293T cells ( Figure 4G,H), and their interaction also occurred at the endogenous level in A549 cells ( Figure 4F). To identify which regions in NEDD4L may be bound to EGFR, we constructed a series of truncated plasmids of NEDD4L and transferred them into HEK293T cells ( Figure 4I). Co-IP assays revealed that the WW domain was crucial for NEDD4L binding to EGFR ( Figure 4J). These data suggest that NEDD4L binds to EGFR mainly through its WW domain.
To further investigate whether NEDD4L reduced EGFR stability, we overexpressed or knocked down NEDD4L in A549 cells. Cycloheximide pulse-chase experiments (CHXchase) showed that NEDD4L overexpression shortened but NEDD4L knockdown prolonged EGFR half-life time in A549 cells ( Figure 4K,L). Next, we found that NEDD4L overexpression significantly increased the ubiquitination of EGFR ( Figure 4M). In contrast, knockdown of NEDD4L reduced the ubiquitination levels of EGFR ( Figure 4N). Functionally, both colony formation and transwell assays supported that overexpression of NEDD4L inhibited the proliferation and metastasis of lung-cancer cells, which could be blocked by overexpression of EGFR (Figures 4O,P and S5L-N). In addition, the proteinexpression levels of NEDD4L and EGFR were negatively correlated in lung-cancer cell lines ( Figure S5O,P). Furthermore, to prove that JAC4 ubiquitinates EGFR by regulating NEDD4L, we performed a CHX-chase assay in siNEDD4L-transfected cells with or without JAC4 treatment. Results show that the enhancement of EGFR protein stability caused by knockdown of NEDD4L was reversed by JAC4 ( Figure S5Q,R). Our results suggest that JAC4 ubiquitinates EGFR by upregulating E3 ubiquitin ligase NEDD4L expression.

K716 Is Critical for NEDD4L-Mediated Degradation of EGFR by JAC4
To further identify potential lysine residues in EGFR that were ubiquitinated by NEDD4L, we used the phosphosite web tool (http://www.phosphosite.org/, accessed on 17 January 2022) to conduct a prediction assay. Results showed that there were seven ubiquitination sites (K716, K737, K754, K860, K867, K929, and K970) in the intracellular domain of EGFR ( Figure 5A). We next generated a panel of EGFR mutants by replacing individual lysine (K) residues with arginine (R) and tested their responses to CHX treatment. We found that both K716R and all mutants were resistant to CHX treatment ( Figure 5B). Compared with WT EGFR, the K716R mutation clearly increased the protein stability of EGFR in A549 and SPCA1 cells ( Figure 5C-E). We further performed in-vitro ubiquitination assays to determine whether the mutations affect EGFR ubiquitination levels. Compared with WT EGFR, the K716R mutation significantly reduced the ubiquitination ability of EGFR ( Figure 5F,G). We found that JAC4 failed to downregulate EGFR protein levels in cells that were transfected with Flag-EGFR K716R, indicating that K716 was the actual ubiquitination site targeted by JAC4 ( Figure 5H). Moreover, both colony formation, EDU incorporation, and transwell assays supported the finding that the K716R mutation had higher proliferative and metastatic potential than the EGFR WT (Figures 5I-M and S6A-E). These results suggest that K716 of EGFR is critical for NEDD4L-mediated ubiquitination and degradation of EGFR by JAC4.

JAC4 Suppresses EGFR T790M-Driven LUAD Growth and Metastasis
NCI-H1975 (EGFR L858R/T790M) is a LUAD cell line that is known to be resistant to first-and second-generation EGFR inhibitors. Since the mutations of EGFR in NCI-H1975 did not overlap with ubiquitination site K716, we investigated whether JAC4 also exerted a similar effect on it. Through a colony-formation assay, we found that JAC4 significantly inhibited the growth of NCI-H1975 cells ( Figure 6A,B). To further confirm whether JAC4 reduced EGFR protein stability, we treated the NCI-H1975 cells with CHX. Notably, JAC4 treatment reduced the half-life time of the endogenous EGFR protein ( Figure 6C,D). In addition, JAC4 enhanced ubiquitination to degrade EGFR in in-vitro ubiquitination experiments ( Figure 6E). Consistent with previous findings, the half-life of the transfected EGFR-K716 mutant was longer than that of the EGFR WT in NCI-H1975 cells ( Figure 6F,G). The above results indicate that JAC4 inhibited EGFR-T790M cell growth in vitro by degrading EGFR. To determine the effect of JAC4 in EGFR-mutant LUAD progression in vivo, we created a subcutaneous tumor-bearing mouse model. Once the average volume of the tumors reached 100 mm 3 , the animals were randomly assigned to receive vehicle, AZD9291, and JAC4 in combination with AZD9291 and JAC4 ( Figure 6H). Data show that JAC4 treatment inhibited EGFR-T790M-driven LUAD growth and that the combined treatment of AZD9291 and JAC4 synergistically suppressed LUAD growth compared with that in mice treated with AZD9291 ( Figure 6I-K). Moreover, the expression of both JWA and NEDD4L was increased; however, EGFR, p-EGFR, p-AKT, and p-STAT3 expression levels were reduced in JAC4-treated tumor tissues compared to those in the control group ( Figures 6L and S7D). Likewise, we also observed higher NEDD4L expression and lower tumor-cell proliferation (Ki67 staining) using immunohistochemistry (IHC) (Figures 6M and S7E). Importantly, we did not observe obvious weight loss in mice with either JAC4 alone or the combined treatment ( Figure S7A). Besides, the representative serum biomarkers, including ALT, AST, BUN, CK-MB, and CK, and the histological morphology of the heart, liver, spleen, lung, and kidneys, revealed no apparent changes compared with those in the control group ( Figure S7B,C).
To evaluate the effect of JAC4 on the proliferation and metastasis in EGFR-mutant cells, we constructed an in-vivo lung-metastasis model by tail-vein injection of NCI-H1975 cells ( Figure 6N). The results show that JAC4 significantly inhibited lung metastasis, and the combination of JAC4 with AZD9291 synergistically suppressed lung-cancer metastasis ( Figures 6O and S7F). The H&E-staining data indicate that the relative metastasis burden in the JAC4-treated group was significantly reduced compared to that of the control group ( Figure 6P,Q). Concurrent with these findings, analysis of public clinical data recorded in the TCGA and GEO databases demonstrated that EGFR-mutant lung-cancer patients whose tumors expressed higher JWA expression were associated with better overall survival ( Figure 6R).

JWA/AMPK Axis Stabilizes NEDD4L Expression by Phosphorylating NEDD4L at Thr367
How does JAC4 activate NEDD4L in NSCLC cells? Previous studies have shown that JWA suppresses pancreatic-cancer progression via the AMPK-FOXO3a axis [35]. Moreover, deficiency in GTRAP3-18 (the homologous protein of JWA) in mice results in AMPK inhibition [36]. In addition, NEDD4L protein stability is known to be dependent upon its phosphorylation modification [37]. Thus, we speculated that JWA may stabilize E3 ubiquitin ligase NEDD4L through an AMPK-mediated phosphorylation-signaling cascade. To confirm this hypothesis, we first determined the potential positive correlation between JWA expression and P-AMPK level by analyzing the TCGA-RPPA (reverse-phase pro-tein array) and CPTAC-phosphoproteome in lung-cancer patients ( Figure S8A,B). Upon further analysis of the CCLE (Cancer Cell Line Encyclopedia)-RPPA, a negative correlation between P-AMPK and EGFR protein expression was found in lung-cancer cell lines ( Figure S8C). In addition, RNA-seq analyses of AMPK-KO MEF cells revealed higher EGFR expression compared to that in AMPK-WT cells in the public database ( Figure S8D), which supported a negative correlation between AMPK and EGFR expression. Furthermore, NEDD4L expression was insensitive to JAC4 treatment when knocking down AMPK or using an AMPK inhibitor (Figures 7A and S8E). Subsequently, transwell and EDU assays also confirmed that the effect of JAC4 on lung-cancer proliferation and migration could be inhibited when knocking down AMPK or using an AMPK inhibitor ( Figure 7B-E). Consistently, the protein expression of P-AMPK and NEDD4L were elevated in NCI-H1975 and SPCA1 tumor samples upon JAC4 treatment ( Figure 7F-I). To demonstrate whether Foxo3 expression is involved in the regulation of NEDD4L expression, we constructed the si-Foxo3 A549 cell line, and the data showed that the expression of NEDD4L did not change in si-Foxo3 A549 cells ( Figure S8F). Therefore, the JWA/AMPK axis enhanced NEDD4L stability in a Foxo3-independent manner. Our results were also supported by another study that showed that AMPK phosphorylates NEDD4L in Xenopus and is critical for NEDD4L stability [38]. To further investigate the interaction between AMPK and NEDD4L, we completed an AMPK-NEDD4L docking assay, and the results indicate that AMPK interacted with NEDD4L physically ( Figure 7J). In addition, the activated form of AMPK (P-AMPK T172) was co-immunoprecipitated with NEDD4L ( Figure 7K). To determine AMPK phosphorylation sites on NEDD4L, the Group-based Prediction System 3.0 (GPS 3.0) software and phosphoNET kinase predictor were used ( Figure 7L), and the data show that T367 in NEDD4L was a potential phosphorylation site of AMPK. Human NEDD4L T367 was a highly conserved locus across different species ( Figure 7M). To investigate whether NEDD4L T367 phosphorylation is modified by AMPK, a plasmid encoding NEDD4L WT or a phosphorylation-defective mutant (T367A) was transfected into A549 cells, followed by JAC4 treatment for 24 h and subsequent CHX treatment. The half-life of NEDD4L with the T367A mutation was shortened compared with that of NEDD4L WT ( Figure 7N). These findings suggest that JAC4 enhanced the stability of NEDD4L via AMPK-triggered phosphorylation at Thr367.

JAC4 Upregulates JWA Expression by Binding to Its Transcriptional Suppressor CTBP1 in NSCLC Cells
Since it was determined that JAC1 promotes JWA transcription by binding toYY1 [39], we speculate that JAC4 increased JWA transcriptional activity by affecting JWA-cofactor interaction. To confirm this hypothesis, both A549 and HBE cells were treated with biotin-JAC4 for 24 h; compared with proteins interacting with biotin alone, 44 potential JAC4-interacting proteins were identified by streptavidin-immunoprecipitation assay and mass-spectrometry analysis (MS) ( Figure 8A). We further compared these potential proteins with the predicted JWA-promoter-specific transcription factors in public databases (http://jaspar.genereg.net/, accessed on 19 August 2022). As a result, five transcriptional suppressors, including BCLAF1, CTBP1, ZMYM3, SMC3, and HDAC2, were obtained ( Figure 8B). To further confirm which one mediated the role in JAC4 upregulating JWA expression, we completed siRNA-transfection assays to reduce the relevant transcriptional factors and measure JWA expression. The data show that only inhibition of CTBP1 could upregulate JWA expression ( Figure S9A), which is consistent with the verified data indicating that overexpression of CTBP1 resulted in a reduction of JWA in HBE cells ( Figure 8C). Moreover, we generated CTBP1-knockout A549 cells through the Cas9 gene-editing technique (Figures 8D and S9B). As expected, JWA expression was obviously elevated in CTBP1knockout A549 cells ( Figure 8E). Subsequently, we investigated whether JAC4 activating JWA expression was due to inhibition of CTBP1. The data show that CTBP1 protein levels were unaffected by JAC4 treatment ( Figure S9C). Therefore, JAC4 may interact with CTBP1 and block its transcriptional-repressor function on the JWA gene in both HBE and lung-cancer cells. Molecular-docking and Co-IP assays also verified that JAC4 could interact with CTBP1 ( Figure 8F,G). WT ( Figure 7N). These findings suggest that JAC4 enhanced the stability of NEDD4L via AMPK-triggered phosphorylation at Thr367.  To further confirm the interaction between JAC4 and CTBP1 in cells, we conducted thermal-shift assays (CETSA) and found that JAC4 treatment led to significant thermal stabilization of CTBP1 ( Figure 8H). These results indicate that JAC4 directly interacts with CTBP1. As a transcriptional suppressor, nuclear translocation is necessary for CTBP1 postactivation. Our data show that JAC4 treatment reduced its nuclear levels and increased cytoplasm levels of CTBP1 in A549 cells in a dose-dependent manner, suggesting that JAC4 also reduced CTBP1 translocation ( Figure S9D). Moreover, when CTBP1 was knocked out, its transcriptional suppression on JWA expression was completely removed; therefore, JAC4 was unable to exert subsequent inhibition of lung-cancer growth and metastasis via the CTBP1-JWA pathway in these CTBP1-deficient cells (Figures 8I-L and S9E). Contrary to lower JWA expression in lung cancer, CTBP1 was highly expressed in A549, NCI-H1975, and HCC827 lung-cancer cell lines ( Figure 8M) and lung-cancer tissues compared to corresponding controls ( Figure 8N). Therefore, a negative correlation between CTBP1 and JWA expression was identified in human-lung-cancer tissues ( Figure 8O). These results suggest that CTBP1 might be a valuable target involved in the progression of NSCLC due to its role as a transcriptional suppressor, and that JAC4 targeted CTBP1 and therefore rescued the expression of tumor suppressors like JWA.

Discussion
Gene amplification and mutations of EGFR have been implicated in the pathogenesis and progression of many malignancies, including lung cancer [40]. EGFR tyrosine-kinase inhibitors (TKIs) have been widely used in clinical practice, significantly prolonging the survival of patients [41]. However, up to 60% of patients eventually develop resistance within 10-14 months after first-generation TKIs [42]. Furthermore, many EGFR-mutant patients are insensitive to EGFR TKIs, and patients with EGFR WT are insensitive to TKIs despite overexpression [14]. Recent studies have shown that EGFR stability is a fundamental mechanism for maintaining the homeostasis of EGFR signaling. Therefore, targeting and degrading EGFR is a more effective and complete strategy for lung-cancer treatment. Here, based on public lung-cancer transcriptome and proteome data, we found that JWA was lowly expressed in NSCLC and that patients with high JWA expression had a better prognosis. JAC4 significantly activated the expression of JWA in lung-cancer cell lines and tumor tissues, thereby inhibiting the growth and metastasis of EGFR-driven lung cancer in vivo and in vitro. Mechanistically, JAC4 accelerated the degradation of EGFR by activating the phosphorylation of AMPK signaling and stabilizing the expression of NEDD4L at Thr367. The K716 site of EGFR is required for NEDD4L-mediated ubiquitination to degrade EGFR. In addition, JAC4 removed the transcriptional regression of JWA by binding to CTBP1 directly, and therefore rescued normal JWA transcription. Collectively, the small-molecular agonist JAC4 of the JWA gene inhibited EGFR-driven lung-cancer growth and metastasis via the AMPK-NEDD4L-EGFR axis (Figure 9).
Several EGFR-degradation strategies have been reported. One approach is to deliver specific EGFR small-interfering RNAs in vivo; however, the efficacy is limited by rapid degradation and a short half-life in the body [43]. Additionally, protein-hydrolysis-targeted chimeric (PROTAC) technology has been used for EGFR degradation, membrane permeability, solubility, and metabolic stability, which has added to the challenges of synthetic drugs [44]. In this study, we determined that JAC4, an agonist of JWA, can degrade EGFR by activating NEDD4L. Furthermore, since JWA is predominantly localized in the lung and lowly expressed in tumor tissues, this may greatly reduce the toxicity to the normal organism. Although we performed a series of in-vivo and in-vitro functional assays and reversion experiments to confirm that the expression levels of JWA and CTBP1 were necessary for JAC4 to exert its anti-cancer function, we still cannot exclude other downstream molecules that may be involved in the onset of the anti-tumor mechanism of JAC4, which requires further investigation. Interestingly, the previously screened JWA agonist JAC1 could inhibit breast-cancer proliferation by ubiquitinating the expression of HER2 [30], whereas our current study found that JAC4 was superior to JAC1 in terms of tumor suppression in LUAD, which is possibly attributable to the difference in transcription factors necessary for the activation of JWA by the two different compounds: JAC1 activates JWA expression via the transcription factor Yin Yang 1 (YY1) [39], whereas JAC4 promotes JWA expression mainly by reducing the nuclear translocation of CTBP1. Considering the differences, further therapeutic strategies can be selected based on the background expression levels of CTBP1 and YY1 in different cancer types. Several EGFR-degradation strategies have been reported. One approach is to deliver specific EGFR small-interfering RNAs in vivo; however, the efficacy is limited by rapid degradation and a short half-life in the body [43]. Additionally, protein-hydrolysis-targeted chimeric (PROTAC) technology has been used for EGFR degradation, membrane permeability, solubility, and metabolic stability, which has added to the challenges of synthetic drugs [44]. In this study, we determined that JAC4, an agonist of JWA, can degrade EGFR by activating NEDD4L. Furthermore, since JWA is predominantly localized in the lung and lowly expressed in tumor tissues, this may greatly reduce the toxicity to the normal organism. Although we performed a series of in-vivo and in-vitro functional assays and reversion experiments to confirm that the expression levels of JWA and CTBP1 were necessary for JAC4 to exert its anti-cancer function, we still cannot exclude other downstream molecules that may be involved in the onset of the anti-tumor mechanism of JAC4, which requires further investigation. Interestingly, the previously screened JWA agonist JAC1 could inhibit breast-cancer proliferation by ubiquitinating the expression of HER2 [30], whereas our current study found that JAC4 was superior to JAC1 in terms of tumor suppression in LUAD, which is possibly attributable to the difference in transcription factors necessary for the activation of JWA by the two different compounds: JAC1 activates JWA expression via the transcription factor Yin Yang 1 (YY1) [39], whereas JAC4 promotes JWA expression mainly by reducing the nuclear translocation of CTBP1. Considering the differences, further therapeutic strategies can be selected based on the back- NEDD4L belongs to the NEDD4 family of HECT E3 ubiquitin ligases, which includes four WW domains that can specifically recognize targeted substrates containing PPXY, LPXY, or PPR sequences [45]. NEDD4L functions as a tumor-suppressor gene in some cancers, such as breast, pancreatic, and lung cancer [46][47][48]. Our study also found that NEDD4L was negatively associated with proliferation and migration in NSCLC. NEDD4L can also inhibit the expression of various tumor-associated membrane proteins, including LGR5, beta-catenin, and transforming growth-factor beta (TGFβ) receptor [49][50][51]. Although a negative regulatory relationship between NEDD4L and EGFR protein expression has been reported, the specific mechanism has not been elucidated [52]. In addition, inhibition of EGFR-signaling pathway promotes NEDD4L protein expression, which may be caused by its downstream-signaling pathways [52]. Therefore, the deep regulatory relationship between NEDD4L and EGFR, especially on mutant EGFR degradation, has not been elucidated. Our results suggest that the binding of the WW region of NEDD4L to EGFR plays a role in the ubiquitination and degradation of EGFR, thus inhibiting the activation of downstream PI3K/AKT signaling. In addition, mutations or single-nucleotide poly-morphisms in the human NEDD4L gene are associated with a variety of diseases [53][54][55]. By analyzing the mutation databases in TCGA and COSMIC, we found that the main mutations of tumor-related NEDD4L are G39V, P197S, S230F, and R253W ( Figure S8G). These mutations may affect its binding to EGFR and thus fail to degrade EGFR; further studies are required to experimentally validate this hypothesis.
AMPK, a major energy sensor and regulator, regulates cell metabolism and energy homeostasis [56]. AMPK inhibits lung-cancer growth primarily by inhibiting mTORC1 oncogenic signaling [57]. Meanwhile, AMPKα deletion has been shown to promote KRASmediated lung-cancer growth and metastasis [58]. Therefore, AMPK has emerged as a potential target for cancer therapy. GTRAP3-18, a homologous protein of JWA, and mice's loss of GTRAP3-18 resulted in AMPK inhibition [36]. Our study confirmed that JAC4 significantly increased the expression of P-AMPK, and the effect of JAC4 on the growth and metastasis of NSCLC was obviously inhibited after knockdown of AMPK or use of AMPK inhibitor, indicating that JAC4 also acts as an agonist of AMPK to suppress lung-cancer proliferation. NEDD4L expression can be stabilized through ERK and SGK1 phosphorylation cascades [51,59]. We found that AMPK plays a role in phosphorylating and stabilizing NEDD4L in NSCLC, which may be due to the heterogeneity of different cancer types. Unlike JAC4, previous studies have reported that JWA-mimicking peptide JP1 phosphorylates NEDD4L expression via MEK1/2 signaling [60]. Thus, the regulation of NEDD4L by JWA may involve in different mechanisms. Multiple clinical trials have reported that AMPK agonist metformin combined with EGFR-TKIs can prolong the progression-free survival (RFS) in EGFR-mutated NSCLC compared to EGFR-TKIs alone [61,62]. In this study, the combination of JAC4 and osimertinib synergistically inhibited the tumor-bearing growth and metastasis of EGFR-mutant lung-cancer cells in model mice. Therefore, JAC4 has also proven to be potentially valuable in combination with EGFR inhibitors in cancer therapy.
CTBP1 is a critical transcriptional repressor, and deficiency or overexpression of CTBP1 leads to transcriptional imbalances [63]. CTBP1 induces a wide range of tumorigenic-and cancer-stem-cell-relative functions through transcriptional regulation of gene networks [64]. CTBP1 inhibits the adhesion of molecules such as E-cadherin and is associated with the promotion of epithelial-mesenchymal transition (EMT), a step that contributes to the malignant properties of tumors [65]. Interestingly, JAC4 was also found to promote Ecadherin expression, indicating that JAC4 may be involved in the suppression of the EMT process in lung cancer ( Figure S9F,G). In addition, several studies have found that CTBP1 forms complexes with multiple epigenetic regulators or transcriptional regressors, and these complexes further recruit epigenetic regulators to control gene expression. CTBP1 also undergoes dynamic post-translational modifications that affect its own stability and subcellular localization [66,67]. Our results show that JAC4 directly bound to CTBP1 and significantly reduced its nuclear overload, thereby restoring JWA transcription levels to their normal state. The thermal stability of CTBP1 was obviously increased after JAC4 treatment, which further confirms that the combination of both may result in changing the properties of CTBP1. The background events and mechanisms, such as the fine mapping of specific amino-acid residues of CTBP1 binding to JAC4, need to be further elucidated.

Cell Culture, Transfection, and Treatment
The HBE and HEK293T cell lines were purchased from the American Type Culture The details of shJWA, Flag-JWA, and the corresponding plasmids have been described in a previous study [29]. Small interfering RNAs targeting AMPK, BCLAF1, CTBP1, HDAC2, ZMYM3, BCLAF1, FOXO3, and SMC3 were designed and synthesized by RiboBio (Guangzhou, China). The sequences are listed in Supplementary Materials: Table S1. The plasmids and siRNAs were transfected into cells using Lipofectamine 3000 (Invitrogen, Grand Island, NY, USA) according to the manufacturer's guidelines. For the cycloheximidechase experiments, the cells were treated with 100 µg/mL cycloheximide (CHX) (Selleck Chemicals, Shanghai, China) for different time points. For the ubiquitination assay, the cells were treated with 10 µM MG132 (Selleck Chemicals, Houston, TX, USA) for 6 h. AMPK-selective inhibitor Dorsomorphin (HY-13418A) and EGFR inhibitor osimertinib (AZD9291, HY-15772) were purchased from MedChemExpress (MCE).

Colony Formation and EDU-Staining Assays
For the colony-formation assay, 1 × 10 3 lung-cancer cells were seeded into six-well plates, and 24 h later the cells were treated with JAC4 at the indicated doses and cultured for 10-14 days. The medium was changed every 3 days, then fixed with 4% paraformaldehyde and stained with crystal violet (Beyotime, Shanghai, China). Visualize colonies were counted. For EdU-staining assays, proliferating cells were detected with the BeyoClick TM EDU Cell Proliferation Kit with Alexa Fluor 555 (Beyotime, Shanghai, China) according to the manufacturer's protocol. Cells were seeded in 96-well plates and incubated for 2 h with EDU working solution (20 µM) after the cells had returned to the normal state, followed by fixation with 4% paraformaldehyde for 30 min. Then, they were incubated with PBS containing 0.3% Triton X-100 for 15 min at room temperature. To detect the percentage of cell proliferation, the cell nuclei were stained for 10 min using 2-(4-Amidinophenyl)-6indolecarbamidine dihydrochloride (DAPI). Next, images of cells were acquired with a Nikon Ti microscope (Nikon, Tokyo, Japan).

Transwell Assay
Transwell cell-migration assays were performed in 24-well plates using 8 µM-pore polycarbonate-membrane inserts (Corning, Tewksbury, MA, USA). The bottom of the upper chamber was coated with fibronectin (Merck Millipore, Darmstadt, Germany); for the tumor-invasion assay, the chamber membranes were coated with 50 µL Matrigel (BD Biosciences, San Jose, CA, USA). After 48 h of transfection or JAC4 treatment, A549 cells or SPCA1 cells (2 × 10 4 ) were seeded in the upper chamber on serum-free medium and the lower chamber was added to DMEM containing 10% FBS. After incubation at 37 • C for 12 h, cells were fixed with 4% paraformaldehyde, stained with crystal-violet solution, and counted at a magnification × 200 under a microscope.

Western Blot, Co-Immunoprecipitation (Co-IP), and Ubiquitination Assays
Cell samples were lysed with RIPA lysis buffer (1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 50 mM Tris, 150 mM NaCl, pH 7.4) supplemented with protease inhibitors and phosphatase inhibitors (NCM Biotech, Suzhou, China). Protein concentrations were quantified using the BCA Protein Assay Kit (Beyotime, Shanghai, China). Then, proteins were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene-fluoride (PVDF) membranes (Millipore, Darmstadt, Germany), which were blocked by 5% nonfat milk in TBST for 1 h at room temperature and incubated with primary antibody overnight at 4 • C. After 4 washes with PBST, the membrane was incubated with horseradish-peroxidase (HRP)-conjugated secondary antibody for 1 h at room temperature. The primary and secondary antibodies used in the study are listed in Supplementary Materials: Table S3. The signals were visualized by an enhanced chemiluminescence (ECL)-detection kit (Vazyme, Nanjing, China).
For immunoprecipitation, the cells were lysed with NP40 lysis buffer (50 mM Tris-HCl, 150 mM NaCl, 1% NP40, 0.5% deoxycholate, pH 8.0) supplemented with a proteaseinhibitor cocktail. Immunoprecipitation was performed using the indicated primary antibodies and protein A/G agarose beads (Santa Cruz, CA, USA) at 4 • C. The beads were washed 3 times with IP lysis buffer for 5 min each and the proteins were eluted by adding 2 × loading buffer (Beyotime) and boiling for 5 min. The immunoprecipitates were subjected to standard Western-blot analysis. For the ubiquitination assay, the cells were treated with DMSO or JAC4 for 24 h followed by treating with MG132 or not for 6 h (10 µM). Then, cells were harvested and protein samples were prepared and used for Western blotting and Co-IP assays, respectively.

Protein Half-Life Assays
The cells were treated with 10 µM JAC4 or DMSO for 24 h, then with CHX (100 µg/mL) for the indicated time periods. Cell lysates were collected for Western-blot analysis for EGFR protein levels using anti-EGFR antibody (CST, #4267, 1:1000).

Cellular Thermal-Shift Assay (CETSA)
The CETSA was performed by standard protocols [70]. HBE cells were treated with 10 µM JAC4 or DMSO for 6 h. Cells were suspended in phosphate-buffered saline containing protease inhibitors, heated at the indicated temperature for 3 min, and then immediately snap-frozen using liquid nitrogen. The samples were collected and subjected to Westernblot analysis.

Nuclear-and Cytosolic-Protein Extraction
Cells were extracted according to the instructions for the Nuclear and Cytoplasmic Extraction Kit (Beyotime, P0028). Briefly, A549 cells were treated with different concentrations of JAC4 for 24 h, after which the cells were collected by centrifugation and 200 µL of Cell Pulp Protein Extraction Reagent A were added per 20 µL of cell precipitate, shaken vigorously for 5 s, and placed in an ice bath for 10-15 min. Then, 10 µL of Cell Plasma Protein Extraction Reagent B were added, and the cells were shaken vigorously for 5 s and placed in ice bath for 1 min. They were then centrifuged at 12,000-16,000× g for 5 min and the supernatant was aspirated as the cytosolic protein. For the precipitate, 50 µL of nucleoprotein-extraction reagent were added and the supernatant was aspirated for nuclear protein by centrifugation after several rounds of vigorous shaking and an ice bath.

Molecular-Docking Assay
Three-dimensional protein structures of NEDD4L and EGFR were predicted using the bioinformatics tool I-TASSER (Iterative Threading Assembly Refinement), and protein molecular-docking predictions were made using the Discovery Studio 3.0 server, i.e., NEDD4L and EGFR. Local servers were used for docking-data-file processing and embellishment. The HDOCK server (http://hdock.phys.hust.edu.cn/, accessed on 12 October 2022) is a protein-protein-docking method based on a hybrid algorithm of template-based modeling and template-free docking [71]. The complex structures of AMPK-NEDD4L were predicted using the HDOCK algorithm based on the structures of AMPK and NEDD4L (Protein Data Bank structures).

Biotin-Assisted Pull-Down Assay and Mass-Spectrometry Analysis
Biotin-assisted pull-down assay and mass-spectrometry analysis were performed as described [72]. Cells were lysed in IP lysis buffer (20 mM Tris (pH 7.5), 150 mM NaCl, 1% Triton X-100, protease inhibitor, and 1 mM EDTA) and subjected to protein quantification. Briefly, whole-cell protein lysates (500 µg) were incubated with 10 µM biotin or 10 µM biotin-JAC4 for 6 h, followed by a 12 h incubation with streptavidin-coupled beads (Thermo Fisher Scientific, 65601) at 4 • C. Afterwards, biotin-bound beads and biotin-JAC4-bound beads were carefully washed by the eluent and then subjected to SDS-PAGE, the gels were stained with staining solution (Coomassie Brilliant Blue R-250 dye, Beyotime, P0017F) for 30 min and decolorized, and then the gels were cut carefully and analyzed by LC-MS/MS.

In-Vivo Xenograft Assay and Lung-Metastasis Assay
Six-week-old male BALB/c nude mice were purchased from the JiangSu Jicui pharmaceutical company and maintained in specific pathogen-free facilities. Tumor-xenograft and -metastasis models were created with the nude mice. These lung-tumor models were used in this project. (i) For xenograft models, 5 × 10 6 human-lung-cancer SPCA1 cells suspended in 100 µL of PBS in the logarithmic growth phase were subcutaneously injected into the right flank of nude mice. When the average tumor volume reached 100 mm 3 , different treatment groups were intragastrically administrated: vehicle, JAC4 (100 mg/kg), and JAC1 (100 mg/kg), and tumor growth was measured every 2 days using sliding calipers. Tumor volume was calculated using the following formula: tumor volume = 1/2 length × width × width. The tumor was then removed, weighted, and frozen in liquid nitrogen for further analysis. (ii) For EGFR-mutant-cell xenograft models, 3 × 10 6 NCI-H1975 (EGFR T790M) cells suspended in 100 µL of PBS were subcutaneously injected into the right flank of nude mice. When the average tumor volume reached 100 mm 3 , different treatment groups were intragastrically administrated: vehicle, JAC4 (100 mg/kg), AZD9291 (5 mg/kg), and a combination of JAC4 and AZD9291. On day 15, the mice were euthanized and the tumors were removed for subsequent analysis. (iii) For EGFR-mutant-cell lung-metastasis models, NCI-H1975 cells (1 × 10 6 ) were injected into mice through the tail vein, and the mice were treated daily with different treatments starting the following day. At the end of experiment, the mice were sacrificed by over-anesthesia, and lung tissue was collected for histological analysis. All animal studies were approved by the Institutional Animal Care and Use Committee of Nanjing Medical University (IACUC-2012059).

Statistical Analysis
All data were expressed as mean ± SD. Student's t-tests were used to compare means between two groups and ANOVA was used for comparisons among more than two groups. The X 2 test was used to analyze the relationship between JWA expression and various clinicopathologic characteristics. Survival analysis was performed by Kaplan-Meier analysis with the log-rank test. Correlation between JWA expression and other gene expression was performed by the Spearman rank-correlation test. A p < 0.05 was considered statistically significantly. Additional experimental methods are described in the Supplemental Materials.

Conclusions
In the present study, we provide evidence for the first time that JAC4 as an agonist of the JWA gene effectively inhibits the proliferation and metastasis of NSCLC. Mechanistically, JAC4 reduced the PI3K/AKT overactivation caused by both EGFR overexpression and mutations in NSCLC; JAC4 rescued JWA transcription in lung-cancer cells by binding to CTBP1. In addition, JAC4 phosphorylated and stabilized NEDD4L by JWA-triggered activation of the AMPK-signaling pathway; the phosphorylation of NEDD4L further accelerated the degradation of EGFR through enhanced ubiquitination at K716, therefore suppressing the progression of EGFR-driven lung cancer. Our results may provide a new strategy for EGFR-driven cancer therapy, especially for EGFR-mutant cancers.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ijms24108794/s1. Author Contributions: K.D. and J.Z. designed this study. K.D., X.J., Z.W., J.C. and L.Z. carried out the experiments. K.D., X.L. and A.L. analyzed the data. K.D. and J.Z. discussed and drafted the manuscript. C.S. performed the molecular-docking assays. J.Z. supervised the study. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement:
The data supporting the conclusions of this article are presented within the article and its additional files.