CXCL14 promotes metastasis of non-small cell lung cancer through ACKR2-depended signaling pathway

Background: Lung cancer is a malignant tumor with metastatic potential. Chemokine ligand 14 (CXCL14) has been reported to be associated with different cancer cell migration and invasion. However, few studies have explored the function of CXCL14 and its specific receptor in lung cancer metastasis. This study aims to determine the mechanism of CXCL14-promoted cancer metastasis. Methods: The expression of CXCL14, atypical chemokine receptor 2 (ACKR2), and epithelial mesenchymal transition (EMT) markers was evaluated by the public database of The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO), Western blot, enzyme-linked immunosorbent assay (ELISA), quantitative real-time polymerase chain reaction (qPCR), immunohistochemistry (IHC), and immunofluorescence (IF). Migration and wound healing assays were used to observe the motility of cancer cells. A luciferase reporter assay was performed to analyze transcription factor activity. The metastasis of lung cancer cells was evaluated in an orthotopic model. Results: We have presented that overexpression of CXCL14 and ACKR2 was observed in lung cancer datasets, human lung tumor sections, and lung cancer cells. Furthermore, the migration of CXCL14-promoted lung cancer cells was determined in vitro and in vivo. In particular, ACKR2 knockdown abolished CXCL14-induced cancer cell motility. Additionally, ACKR2 was involved in CXCL14-triggered phospholipase Cβ3 (PLCβ3), protein kinase Cα (PKCα), and proto-oncogene c-Src signaling pathway and subsequently upregulated nuclear factor κB (NF-κB) transcription activity leading to EMT and migration of lung cancer cells. These results indicated that the CXCL14/ACKR2 axis played an important role in lung cancer metastasis. Conclusion: This study is the first to reveal the function of CXCL14 in promoting EMT and metastasis in lung cancer. As a specific receptor for CXCL14 in lung cancer, ACKR2 mediates CXCL14-induced signaling that leads to cell motility. Our findings can be used as a prognostic biomarker of lung cancer metastasis.


Introduction
Lung cancer is one of the leading global causes of cancer mortality [1]. Although chemotherapy and radiation therapy have contributed to improvements in lung cancer treatment, survival rates remain poor, overall survival rates at 5 years of patients with non-small cell lung cancer (NSCLC) and small cell lung cancer are 15% and 6%, respectively [2,3]. Metastasis progression plays a crucial role in advanced lung cancer, leading to a poor prognosis and low survival rates [4]. Therefore, an understanding of the molecular mechanism of cancer cell metastasis has become a major subject of interest in lung cancer treatment [5].
C-X-C motif chemokine ligand 14 (CXCL14), known as breast and kidney-expressed chemokine (BRAK), is a member of the chemokine family and is responsible for the chemotactic effects on neutrophils [6]. CXCL14 is generally expressed at high levels in numerous tissues such as the breast, kidney, lung, and skin [7]. However, the mechanism of CXCL14 is attracting increasing attention because abnormal expression of CXCL14 could play an important role in several cancers. CXCL14 is ambiguous and disparate in different types of cancers. In head and neck squamous cell carcinoma and colorectal cancer, CXCL14 suppresses cancer cell migration and invasion [8]. Furthermore, CXCL14 is overexpressed in the stromata of the breast and pancreas and prostate cancer [9][10][11]. In addition, CXCL14 is associated with metastasis of lung cancer [12]. Therefore, understanding the mechanisms responsible for CXCL14-promoted metastasis in lung cancer and identifying the key factors involved will identify new therapeutic targets.
Increasing evidence has indicated that CXCL14 can bind to the atypical chemokine receptor 2 (ACKR2) [13]. ACKR2 is found in endothelial cells and leukocytes in the intestine, lungs, skin, and lymphoid organs [14]. The abnormal expression of ACKR2 is associated with many human diseases, including hepatitis, psoriasis, and cancer [15][16][17]. ACKR2 is overexpressed in Kaposi's sarcoma and primary effusion lymphoma cells [18]. Moreover, CXCL14 binds to ACKR2 in cancer-associated fibroblasts (CAF) resulting in breast cancer metastasis [13,19]. Therefore, it may be worthwhile to target CXCL14/ACRK2-induced cancer metastasis in NSCLC.
Here, this study provides an overview of CXCL14 progressively upregulated along with the progression of NSCLC and associated with tumor metastasis. We also found that CXCL14 promotes the epithelial-mesenchymal transition (EMT) and subseq-uently improves cell migration by transactivating the ACKR2/PLC/PKC/c-Src signaling pathway. Knockdown of CXCL14 suppressed EMT production in NSCLC cells and attenuated cell migration ability. Importantly, CXCL14 overexpressed profoundly enhanced lung cancer cell metastasis. Our findings offer new insights into CXCL14 as a potential therapeutic target for cancer invasion and metastasis of lung cancer.

Analysis of public data sets of CXCL14 and ACKR2 in normal and NSCLC lung tissues
Lung cancer adenocarcinoma (LUAD) data sets from the meta-analysis were downloaded from Lung Cancer Explorer (LCE), which contains data sets of gene profiles of tumor and normal lung samples from The Cancer Genomics Atlas (TCGA) and other published literature. All data sets for the TCGA gene analysis were downloaded from UALCAN, which contains data sets of the gene profiles of tumor and normal lung samples obtained from TCGA. Transcriptome data were obtained from the GEO database.

Cell lines
NSCLC cells (H1299) and human embryonic kidney cells (293T) were purchased from American Type Culture Collection (Manassas, VA, USA). NSCLC cells (A549), normal lung fibroblasts (MRC-5) and human embryo kidney cells (293T) were purchased from the Bioresource Collection and Research Center (BCRC, Hsinchu, Taiwan). H1299 was maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin, and 100 µg/ml streptomycin. A549 was maintained in F-12 with the same supplement. MRC-5 was maintained in EMEM with the same supplement as mentioned above. 293T was maintained in DMEM with the same supplement as mentioned above. All cells were incubated at 37 °C in a suitable incubator with 5% CO2 in an air atmosphere.

Antibody neutralization
In experiments using antibody neutralization, cells were pre-incubated with ACKR2, CXCR4, GPR85, or IgG antibodies for 1 h prior to subsequent experiments.

Cell migration assay
Cells (2 × 10 4 cells/well) were seeded in the upper chamber of Transwell inserts (Cat. No. 3428, 8-μm pore size; Costar, New York, NY, USA). After 24 h, cells were fixed with 4% formaldehyde for 15 min at room temperature and then stained with 0.05% crystal violet for another 45 min. The cells in the upper chamber were removed by swab and the migrated cells were photographed and counted. All experiments were carried out in triplicate wells and were repeated at least three times analyzed by ImageJ 1.52a (National Institutes of Health, Bethesda, MD, USA).

Wound healing assay
H1299, A549, and MRC-5 (2.5 × 10 4 cells) were seeded in Culture-Insert (Cat No. 80209, ibidi GmbH, Munich, BRD). After incubation of cells for 24 h, the inserts were removed and the effect of CXCL14 of wound healing was visualized by the microscope at 0 and 24 h. The wound healing area was photographed and counted. All experiments were carried out in triplicate wells and repeated at least three times and analyzed with ImageJ 1.52a.

Immunohistochemistry
Lung cancer tissue array (LC483a, US Biomax, MD, USA) and animal tissue sections were deparaffinized in xylene, rehydrated in ethanol and washed in deionized water. Novolink polymer detection system was used to block intrinsic peroxidase activity and nonspecific antibody binding sites. The sections were incubated with appropriate primary antibodies (1:200) specific for CXCL14, ACKR2, and N-cadherin. The IHC images were scored taking into account the percentage of positive detection (0-100) by Image J 1.52a and the intensity of staining (0-3+), providing a final score ranging from 0 to 300.

Enzyme-linked immunosorbent assay (ELISA)
The secretion of CXCL14 from cells was evaluated using human CXCL14/BRAK DuoSet ELISA as described by the manufacturer (Cat No. DY866 R&D systems, Minneapolis, MN, USA). Briefly, FBS-free cultured medium was collected from plates after incubation with 5 × 10 5 cells for 24 h. The samples were observed with a detection antibody and developed by streptavidin-HRP and substrate solution, and the absorbance at 450 nm was determined using a microplate reader (VARIOSKAN LUX, Thermo Fisher, Waltham, MA, USA).

Cell viability assay
Cells (1 × 10 4 ) were seeded in each well in a 48-well plate and changed to a medium containing different concentrations of CXCL14 on day 2. After 24 h, 10 µl of CCK-8 (Cat. No. 96992, Sigma-Aldrich) was added to each well and incubated for 4 h. The microplate was measured at absorbance of 450 nm with the VARIOSKAN LUX microplate reader (Thermo Fisher Scientific, Waltham, MA, USA).

Immunofluorescence
After treatment, cells were fixed and incubated with appropriate primary antibodies (1:200) specific for E-cadherin, ZO-1, N-cadherin, vimentin, or p65 at 4 °C overnight. After washing, the secondary goat anti-rabbit IgG antibody conjugated with FITC was used for 1 h at room temperature. Cells were then incubated with a DAPI solution for 5 min and examined using a Nikon ECLIPSE Ti microscope (NIS Elements AR 5.02.01).

Transient transfections
PLC, PKC, c-Src, p65, ACKR2, CXCR4, GPR85 and control siRNAs were purchased from Sigma-Aldrich (St. Louis, MO, USA) and performed with DharmaFECT1 transfection reagent for 24 h according to the manufacturer's instructions. The siRNA sequences used in this study were PLC

Lentivirus transduction and generation of stable cell lines
CXCL14 human-tagged ORF clone lentiviral particle was purchased from ORiGENE (Rockville, MD, USA). Short hairpin (shRNA) targeting the CXCL14 and eGFP-luciferase plasmid was purchased from the National RNAi Core Facility Platform (RNAi Core, Taipei, Taiwan). The target sequence of the CXCL14 shRNA was 5′-GATCCGCTACAGCGACGT GAA-3′. 293T cells were transfected with an empty vector, CXCL14 shRNA or human-tagged ORF clone lentiviral particle CXCL14 for 24 h in a medium containing lentiviral particles. H1299 or A549 cells were incubated with a medium containing lentiviral particles to generate stable CXCL14 shRNAluciferase-GFP cells expressing CXCL14 shRNA or overexpressing CXCL14-luciferase-GFP cells.

In vivo orthotopic lung cancer model
An orthotopic lung cancer model was modified as described in previous studies [20,21]. Four-week-old male BALB/C nude mice were anesthetized with 2% isoflurane. A cell suspension of H1299 (1 × 10 6 cells) with a total volume of 50 μL (RPMI medium: Matrigel = 1:1) were injected directly into the left lung. Three weeks after injection, the tumor was examined with an IVIS Luminar II in vivo imaging system (PerkinElmer, Waltham, MA, USA). All the right and left lungs were collected and embedded in paraffin. All animal experiments were conducted according to the protocols approved by the Institutional Ethics Committee of Shin Kung Wu Ho Su Memorial Hospital (IACUC Approval No. 111MOST006).

Statistical analysis
Data are presented as mean ± standard deviation (SD). Two groups of data were compared using Student's t test. More than two groups of data were compared using one-way analysis of variance using GraphPad Prism 8.0. Statistical significance is represented in figures by: *, p value < 0.05; **, p value < 0.01; ***, p value < 0.001; ****, p value < 0.0001.

CXCL14 is highly expressed and correlates with the clinical stage of lung cancer
Meta-analysis data sets were analyzed to explore the expression of CXCL14 in normal lung tissue and lung tumor. We first examined the expression profile of CXCL14 in lung cancer tissues by using online lung cancer-specific database-the Lung Cancer Explorer (LCE). Meta-analysis showed that CXCL14 was upregulated in lung cancer tissues containing adenocarcinoma (LUAD) ( Figure 1A). TCGA gene analysis indicated that CXCL14 was highly associated with the clinical stage of the disease and metastatic lung cancer of the nodule ( Figure 1B-D). Furthermore, analysis of the Gene Expression Omnibus (GEO) data set (GSE18842 and GSE11969) indicated that the expression levels of CXCL14 in NSCLC patients were significantly higher than in normal samples ( Figure  1E-F). Interestingly, the IHC scores of lung cancer tissue array were similar to Meta-analysis and GEO datasets indicated that CXCL14 expression was significantly upregulated in the lung cancer tissues ( Figure 1G). These observations were confirmed by ELISA and Western blot analysis of CXCL14 levels in normal and cancer cell lines. The level of CXCL14 expression was higher in the cancer cells ( Figure  1H-I). Collectively, the expression level of CXCL14 was overexpressed in lung cancer and patients with metastases. These results suggest that CXCL14 plays a pivotal role in lung cancer metastasis.

CXCL14 promotes the migration of lung cancer cells via EMT
On the basis of the above findings, we examined the function of CXCL14 in lung cancer cells. CXCL14-dependent stimulation of cells promoted cell motility but did not affect lung cell proliferation (Figure 2A-C and S1A-B). EMT promotes cell migration and contributes to cancer cell metastasis [22]. Our investigation of GSE18842 NSCLC tissue samples from the GEO data set identified significantly higher N-cadherin expressions in NSCLC and we observed a positive correlation between CXCL14 and N-cadherin levels ( Figure 2D-E). Next, to investigate the relationship between CXCL14 and EMT, the expression of epithelial cell markers, E-cadherin, ZO-1; mesenchymal cell markers, N-cadherin, and vimentin were examined in H1299 and A549 cells. Stimulation of A549 and H1299 with CXCL14 promoted N-cadherin, vimentin, and snail1 and decreased the expression of E-cadherin and ZO-1 according to immunofluorescence and Western blot assays ( Figure 2F-H). MMPs enhance cancer cell migration and promote tumor development [23,24]. Several adhesion molecules, such as vascular cellular adhesion molecule (VCAM)-1 and intercellular adhesion molecule (ICAM)-1, exhibit increased expression in metastatic cancer [25,26]. Therefore, we investigated the expression of MMPs and CAMs in lung cancer cells. Treated A549 and H1299 cells with CXCL14 did not affect the expression of ICAM-1, VCAM-1 and MMPs ( Figure S1C-D). These results revealed that EMT was positively associated with the migration of CXCL14-promoted lung cancer cells.

ACKR2 is essential for CXCL14-induced cell migration in lung cancer cells
Several studies have reported that CXCL14-induced molecular signaling and cellular responses depend on ACKR2, CXCR4, and GPR85 in many different tumor cells [19,27,28]. Therefore, we investigated the role of ACKR2, CXCR4, and GPR85 in lung cancer. A meta-analysis in all lung cancer tissues containing adenocarcinoma (LUAD) found that the expression of ACKR2 but not the other candidate receptors, CXCR4 and GPR85, was up-regulated in lung cancer studies ( Figure 3A-C). Results of IHC staining and western blotting for levels of ACKR2 in lung cancer were similar to the analysis of TCGA database samples, showing higher levels of ACKR2 expression in lung cancer tissues and lung cancer cells ( Figure 3D-E). Next, lung cancer cells transfected with the ACKR2 siRNA or incubated with ACKR2 neutralized antibody markedly abolished CXCL14-induced cell migration and cell movement, but not CXCR4, and GPR85 ( Figure 3F-I and S2A-D). These results showed that ACKR2 receptor plays a crucial role for CXCL14 to induce migration in lung cancer cells.

CXCL14 induces cancer cell migration through an ACKR2-dependent signaling pathway
Previous studies revealed that activation of the phospholipase C (PLC), protein kinase C (PKC) and c-Src pathways was involved in metastatic development to promote tumor progression [29][30][31].

CXCL14 initiates EMT and cell migration by activating nuclear factor-κB
Nuclear factor-κB (NF-κB) is considered to be involved in EMT and metastasis in the lung and other types of cancer [32][33][34]. In this study, stimulation of lung cancer cells with CXCL14 promoted IKKα, IκBα, and p65 phosphorylation, which was antagonized by the c-Src inhibitor ( Figure 5A-C). To verify direct CXCL14 activation of NF-κB mediating EMT and migration, we pretreated lung cancer cells with the inhibitors of IκB kinase α (IKKα) and NF-κB inhibitor α (IκBα) or p65 siRNA to evaluate CXCL14-promoted cell migration and EMT. We found that CXCL14-induced cell migration and EMT were suppressed by IKKα and IκBα inhibitors and p65 siRNA ( Figure 5D-J and S4A-D). The results illustrated that NF-κB activation is necessary for CXCL14-promoted migration and EMT in lung cancer cells.

CXCL14 promotes NF-κB transcriptional activity via the PLCβ3, PKCα, and c-Src pathways
To explore the mechanism of CXCL14 in NF-κB activation, treating lung cancer cells with PLCβ3, PKCα, c-Src inhibitors reversed CXCL14-mediated activation of p65 ( Figure 6A-B). Besides, IKKα and IκBα inhibitors showed the dependency of specific signaling on CXCL14-induced activation of p65 ( Figure 6C-D). As shown in Figure 6E-G, CXCL14 stimulation of cells increased p65 translocation into the nucleus and NF-κB luciferase activity. These effects were suppressed by PLCβ3, PKCα, c-Src, IKKα, and IκBα inhibitors ( Figure 6E-G). These results demonstrated that CXCL14 promoted NF-κB transcriptional activity through the PLCβ3, PKCα, and c-Src pathways.

CXCL14 induces tumor metastases in vivo orthotopic lung cancer model
To further investigate the role of CXCL14 in lung cancer metastasis, we generated cells with stable overexpressed CXCL14 (CXCL14-OV) and cells with knocked down CXCL14 (CXCL14-KD). CXCL14 overexpression cells promoted high levels of both the CXCL14 and EMT protein, which significantly increased the migration ability of H1299 and A549 cells (Figure 7A-D, G-H and S5A-B) but not cellular proliferation ( Figure 7E-F). H1299 (Vector), CXCL14-OV, and CXCL14-KD were orthotopically implanted in the left lung and cancer metastasis was analyzed using an in vivo imaging system (IVIS). IVIS data revealed that CXCL14 overexpression significantly exhibited a higher bioluminescence intensity (BLI), and the BLI was reduced in the CXCL14-KO group compared to that of the vector group ( Figure 7J). The H&E stain images showed that CXCL14 overexpression increased the tumor distribution area in the left and right lung lobes, and the area in the CXCL14-KD group was lower than in the vector group ( Figure 7K). Interestingly, a positive correlation between CXCL14 and N-cadherin protein expression was observed in mouse lung tissue (Fig.  7L). Our data indicated that endogenous CXCL14 increased EMT-dependent cancer metastasis of lung cancer in mice.

Discussion
Lung cancer metastasis is the main cause of the treatment dilemma [4,5]. Identifying crucial molecules that affect cancer cell metastasis is vital in therapeutic strategies. CXCL14 is overexpressed in lung cancer in the TCGA database and the GEO dataset ( Figure 1A-B and E-F). However, the function of CXCL14 in lung cancer progression is still unclear. CXCL14 was found to be associated with lymph node metastasis in papillary thyroid carcinoma and to promote proliferation, migration, and invasiveness in colorectal carcinoma cells [35,36]. Additionally, CAF overexpresses and secretes CXCL14 in breast and prostate cancer and promotes tumor growth and invasion compared to normal stromal cells [37][38][39]. Moreover, CXCL14 is associated with a poor prognosis for patients with cervical and endometrial cancer [9,40]. The most interesting observation emerging from our data is that CXCL14 promotes tumor metastasis through ACKR2 in NSCLC cells. CXCL14 is associated with tumor stages and nodal metastasis. These findings support that CXCL14 is an appropriate prognostic and therapeutic marker target.
According to Bal et al., ACKRs are involved in the growth, migration, and invasion in tumor cells [41] and bind to cysteine-cysteine-type and cysteine-X-cysteine-type chemokines. Furthermore, ACKR2 is a receptor responsible for CXCL14-induced EMT and metastasis in breast cancer [19]. Additionally, ACKR2 deficiency affects melanoma cell metastasis to the lung regions in a mouse model [42]. Previous studies have shown that the high expression of CXCL14/CXCL12 and cell surface receptor CXCR4 is highly correlated with tumor metastasis in endometrioid carcinoma [27]. In addition, the activation of cancer-associated fibroblasts through the CXCL14/GPR85 pathway has been shown to leads to EMT and progression in breast cancer [28]. In this study, we explored the expression of three candidate receptors of CXCL14 and their role on migration in NSCLC cells. The meta-analysis data in LUAD found that ACKR2 is overexpressed but not CXCR4 and GPR85. The elimination of ACKR2 decreases the migration of CXCL14-induced cancer cells. This did not markedly inhibit CXCL14-induced cell migration in the group using CXCR4 and GPR85-specific siRNA and neutralized antibodies. The most important aspect of the data was that ACKR2 is an essential receptor for CXCL14-promoted lung cancer metastasis.    Previous studies have shown that ACKRs bind to chemokines and activate the β-arrestin pathway rather than coupling with G proteins to scavenge chemokines and regulate cell homeostasis [43]. However, ACKRs were found to regulate kinase phosphorylation and cancer progression in different cancer cells [44][45][46][47]. In addition, dysregulation of cellular kinases, including, phospholipase, phosphokinase, and c-Src, is involved in proliferation, metastasis, and EMT in cancer progression [48][49][50]. In addition, the transcription factor NF-κB is involved in tumor development and metastasis in lung cancer and other types of cancer [34,51,52]. The present results revealed that ACKR2 plays a critical role in the progression of CXCL14-induced lung cancer. The elimination of ACKR2 abolished the CXCL14promoted phosphorylation of PLCβ3, PKCα, and c-Src and lung cancer cells. Additionally, CXCL14 promoted the transcriptional activity of NF-κB, which is related to tumor migration and EMT of lung cancer cells. Furthermore, the inhibition of PLCβ3, PKCα, and c-Src phosphorylation reversed CXCL14-induced transcriptional activity of NF-κB in lung cancer cells. A somewhat remarkable relationship was that CXCL14/ACKR2 promote EMT and cell migration by triggering the activation of PLCβ3, PKCα, and c-Src signaling and transcription factor p65.
Previous studies have shown that mitogen-activated protein kinase (MAPK), focal adhesion kinase (FAK), and Akt signaling pathways have been involved in cell migration and progression in lung cancer [53][54][55]. In this study, we also investigated the role of MAPK, FAK, and Akt in CXCL14-induced cell migration in A549 cells. Our findings showed that the administration of p38 inhibitor (SB203580), JNK inhibitor (SP600125) and FAK inhibitor (FAKi) reduced the migration of CXCL14-induced cells, but not MEK (PD98059 and U0126) (supplementary Figure S3E). These results suggest that further investigation is needed to fully understand the mechanism by which CXCL14 interacts with these signaling pathways and will be the focus of our future work.

Conclusion
The results revealed that chemokine CXCL14 promotes EMT and migration through ACKR2 in lung cancer for the first time. These results broaden our understanding of the role and molecular mechanism of CXCL14/ACKR2 interdependence in promoting lung cancer migration in vitro and in vivo (Figure 8). These findings may provide important information on novel strategies for the clinical diagnosis and treatment of lung cancer metastases.