CX Chemokine Receptor 7 Contributes to Survival of KRAS-Mutant Non-Small Cell Lung Cancer upon Loss of Epidermal Growth Factor Receptor

KRAS-driven non-small cell lung cancer (NSCLC) patients have no effective targeted treatment. In this study, we aimed to investigate targeting epidermal growth factor receptor (EGFR) as a therapeutic approach in KRAS-driven lung cancer cells. We show that ablation of EGFR significantly suppressed tumor growth in KRAS-dependent cells and induced significantly higher expression of CX chemokine receptor 7 (CXCR7) and activation of MAPK (ERK1/2). Conversely, rescue of EGFR led to CXCR7 downregulation in EGFR−/− cells. Dual EGFR and CXCR7 inhibition led to substantial reduction of MAPK (pERK) and synergistic inhibition of cell growth. Analysis of two additional EGFR knockout NSCLC cell lines using CRISPR/Cas9 revealed genotype dependency of CXCR7 expression. In addition, treatment of different cells with gefitinib increased CXCR7 expression in EGFRwt but decreased it in EGFRmut cells. CXCR7 protein expression was detected in all NSCLC patient samples, with higher levels in adenocarcinoma as compared to squamous cell lung carcinoma and healthy control cases. In conclusion, EGFR and CXCR7 have a crucial interaction in NSCLC, and dual inhibition may be a potential therapeutic option for NSCLC patients.


Introduction
Lung cancer is the leading cause of cancer related death worldwide [1]. Non-small cell lung cancer (NSCLC) accounts for approximately 85% of lung cancers [2]. Different options for the treatment of patients with NSCLC are available, including surgery, radiotherapy, chemotherapy, immunotherapy, A pool of two CRISPR/Cas9 EGFR knockout (KO) plasmids (Santa Cruz Biotechnology, Dallas, TX, USA), each encoding the Cas9 nuclease and a 20-nucleotide guide RNA (gRNA) targeting exons 2 and 3 of the EGFR, were used to establish EGFR −/− lung cancer cell lines (Table S1). For transfection, 3 × 10 5 cells were seeded in a single well of a 6-well plate. The next day, cells were transfected with CRISPR/Cas9 plasmids pool using Lipofectamine 3000 (Invitrogen, Carlsbad, USA) according to the manufacturer's instructions with 3 µg of CRISPR/Cas9 plasmids. These plasmids contain GFP (Santa Cruz Biotechnology, Dallas, TX, USA) and puromycin resistance genes. Puromycin was used to select cells with successful uptake of the plasmid. Culture medium was changed one day after transfection, followed by puromycin selection with 2 µg/mL for the next three days. Single cells were seeded into 96-well plates and after three weeks of culture single cell wells were tested for EGFR knockout by Sanger sequencing, Western blot, and fluorescence-activated cell sorting (FACS) analysis. We could not establish HCC827 EGFR −/− cells, which is probably due to the strong dependence of these cells to EGFR signaling [30,31].

DNA Isolation, Polymerase Chain Reaction (PCR), and Sequencing
Cells were harvested by trypsin and washed with PBS. DNA was isolated using the DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany) according to the manufacturer's protocol. Exons 2 and 3 of the EGFR were amplified by PCR using standard procedures. PCR was performed using 150 ng of genomic DNA in a final volume of 25 µL containing 1× PCR buffer, 0.25 µL of Pfu DNA polymerase (Thermo Fisher Scientific, Waltham, MA, USA) and 100 nM primers (Table S2). The PCR program comprised one cycle of 98 • C for 2 min, 30 cycles of 98 • C for 10 s, of 59 • C for 15 s, and of 72 • C for 30 s, and one cycle of 72 • C for 10 min. PCR products were analyzed on 1.5% agarose gel. PCR products were purified using the Wizard SV Gel and PCR Clean-Up Kit (Promega, Madison, WI, USA) according to the company's protocol and sequenced at Macrogen Europe (Amsterdam, The Netherlands).

Colony Formation Assay
Total numbers of 500 or 1000 cells were seeded in a single well of a 6-well plate and treated with 100 nM cetuximab or 5 µM gefitinib for 12 days at 37 • C in a humidified CO 2 incubator. The medium was then removed and cells were washed with PBS, followed by cell fixation using 4% formaldehyde. Cells were stained with 1% crystal violet and colonies were counted. Experiments were performed in triplicate and repeated at least three times.
To test the effects of EGF and CXCR7 inhibitor on cell proliferation, 10,000 cells were seeded in a 12-well plate and grown for 6 days at 37 • C in a humidified CO 2 incubator. The medium was removed, and cells were washed with PBS, followed by cell fixation using 4% formaldehyde. Cells were stained with 1% crystal violet. For quantification of the staining, 1 mL of 10% acetic acid was used for each well to extract the dye, and the absorbance was measured at wavelength of 590 nm.

Wound Healing Assay
The A549 EGFR wt/wt and EGFR −/− cells were seeded in 6-well plates at a density of 7 × 10 5 cells per well and starved overnight in DMEM containing 1% FBS. A wound was gently made by scraping the cells with a sterile 200 µL pipette tip. The detached cells were removed using PBS. Pictures were taken at 0 h, 12 h, and 24 h using a CK2 inverted microscope (Olympus, Tokyo, Japan). Experiments were performed in triplicate and repeated at least three times.

Cell Migration Assay
A total number of 1 × 10 4 cells in DMEM containing 1% FBS was added to the top Transwell insert. The 24-well plate wells were filled with a 750 µL culture medium. The Transwell insert was gently added to the 24-well plate and incubated at 37 • C with 5% CO 2 for 20-24 h. Medium and remaining cells were carefully removed from the top of the membrane. The insert membrane was stained with 0.5% crystal violet. Invaded cells were photographed randomly and counted. Experiments were performed in triplicate and repeated at least three times.

MTS Assay
A total of 3 × 10 3 cells were seeded per well in 96-well plates and cultured for 24 h. Cells were then treated with appropriate drugs for 72 h. Next, cells were incubated at 37 • C with a medium containing MTS for 90 min following the protocol of CellTiter 96 AQueous One Solution Reagent (Promega, Madison, WI, USA). The absorbance was determined at a wavelength of 490 nm using a Synergy H1 plate reader (BioTek, Winooski, VT, USA). Experiments were performed in triplicate and repeated at least three times.

Flow Cytometric Analysis
Cells were harvested, washed twice with standard FACS buffer (PBS/1%FBS), and incubated with a primary antibody (cetuximab) or a human IgG isotype as a negative control (Invitrogen, Carlsbad, CA, USA) for 1 h on ice. Cells were then washed with FACS buffer, followed by 1 h incubation with a goat anti-human IgG (H + L) Cross-Adsorbed secondary antibody, Alexa Fluor 488. EGFR membrane expression was determined using a FACSCalibur flow cytometer (BD, Franklin Lakes, NJ, USA).
2.11. RNA Isolation and Quantitative Reverse Transcriptase PCR (qRT-PCR) Cells were harvested by trypsin and washed with PBS. RNA was isolated using the Maxwell LEV simply RNA Cells/Tissue Kit (Promega, Madison, WI, USA). RNA concentrations were measured using NanoDrop (Thermo Fisher Scientific, Waltham, USA). cDNA was synthesized from 200 ng of total RNA using the Reverse Transcription kit (Promega, Madison, WI, USA) according to the manufacturer's instruction.
qRT-PCR was performed using 20 ng of cDNA as input and SensiMix SYBRkit (Bioline, Taunton, MA, USA) in an ABI Prism 7900HT Sequence Detection System (Thermo Fisher Scientific, Waltham, MA, USA). Primer sets are listed in Table S2. Data analysis was performed using SDS v.2.3 software (Applied Biosystems, Foster City, CA, USA). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA levels were measured and used as reference for data normalization.

CXCR7 Knockdown Using siRNAs
Approximately 3 × 10 5 cells per well were cultured for 24 h in a 6-well plate. Cells were then transfected with a mixture of CXCR7-specific validated siRNAs (Thermo Fisher Scientific, Waltham, USA) and negative control siRNAs (Thermo Fisher Scientific, Waltham, WI, USA) using Lipofectamine 3000 (Invitrogen, Carlsbad, USA) according to the manufacturer's protocol. Total RNA was isolated 72 h after transfection. Experiments were repeated three times.

EGFR Rescue in A549 EGFR −/− Cells
A total of 3 × 10 5 A549 EGFR −/− cells per well were cultured for 24 h in a 6-well plate. Cells were transfected with pCDNA6A-EGFR wild-type plasmids using Lipofectamine 3000 (Invitrogen, Carlsbad, USA). Culture medium was changed one day after transfection, followed by blasticidin selection with an appropriate concentration for the next three days. EGFR expression level was determined by Western blot. Experiments were repeated three times. pCDNA6A-EGFR wild-type plasmid was a gift from Mien-Chie Hung (Addgene plasmid #42665).

Patient Tumor Samples and Immunohistochemistry (IHC)
A total of 47 patients with lung cancer were included in the study. A tissue microarray (TMA) containing 43 formalin fixed paraffin embedded (FFPE) primary lung tumor samples from the Department of Pathology, University Medical Center Groningen (UMCG), and 4 additional FFPE primary lung tumor samples from The First Hospital of Lanzhou University were used to detect CXCR7 protein expression using IHC. The study was performed in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines.
Briefly, 4 µm FFPE sections were deparaffinized using xylene for 10 min. Next, slides were incubated with Tris/HCl (pH = 9) in a microwave. After blocking endogenous peroxidase activity with hydrogen peroxide, slides were incubated with the anti-CXCR7 primary rabbit polyclonal antibody (GTX100027, GeneTex, Irvine, USA) for 1 h at RT. Slides were then incubated with peroxidase-labeled goat anti-rabbit secondary and rabbit anti-goat tertiary antibodies (Dako, Denmark) for 30 min at RT. Visualization was performed using ImmPACT NovaRED Peroxidase (HRP) Substrate (Vector Laboratories, Burlingame, CA, USA) followed by hematoxylin staining. Scoring of the slides was performed by an experienced pulmonary pathologist (WT).

Statistical Analysis
The data are presented as mean ± SD, except where otherwise indicated. Data are derived from three or more independent experiments (unless otherwise indicated), and statistical analysis was performed using GraphPad software v.5.0 (GraphPad Software, La Jolla, CA, USA). Results were analyzed by a two-tailed unpaired student's t-test unless otherwise noted. A chi-square test was used to determine significance in IHC results. p-values less than 0.05 were considered significant. The synergic analysis was performed using CompuSyn Program (ComboSyn, Paramus, NJ, USA).

EGFR Knockout in A549 Cell Line
A549 is an adenocarcinoma cell line that contains a wild-type EGFR and mutant KRAS. To generate an A549 EGFR knockout cell lines, a CRISPR/Cas9 approach was applied using three gRNAs targeting exons 2 and 3 of EGFR (Table S1). Sanger sequencing revealed a 16 bp deletion in exon 2 of EGFR, which resulted in a premature stop codon, L79X ( Figure 1a). Another independent clone with 1 bp insertion in exon 3 was shown in Figure S1. FACS and Western blot showed expression of EGFR in the wild-type A549 cells, while it was totally absent in the EGFR knockout cell line (Figure 1b,c and Figure S2). The data are presented as mean ± SD, except where otherwise indicated. Data are derived from three or more independent experiments (unless otherwise indicated), and statistical analysis was performed using GraphPad software v.5.0 (GraphPad Software, La Jolla, CA, USA). Results were analyzed by a two-tailed unpaired student's t-test unless otherwise noted. A chi-square test was used to determine significance in IHC results. p-values less than 0.05 were considered significant. The synergic analysis was performed using CompuSyn Program (ComboSyn, Paramus, NJ, USA).

EGFR Knockout in A549 Cell Line
A549 is an adenocarcinoma cell line that contains a wild-type EGFR and mutant KRAS. To generate an A549 EGFR knockout cell lines, a CRISPR/Cas9 approach was applied using three gRNAs targeting exons 2 and 3 of EGFR (Table S1). Sanger sequencing revealed a 16 bp deletion in exon 2 of EGFR, which resulted in a premature stop codon, L79X ( Figure 1a). Another independent clone with 1 bp insertion in exon 3 was shown in Figure S1. FACS and Western blot showed expression of EGFR in the wild-type A549 cells, while it was totally absent in the EGFR knockout cell line (Figure 1b,c and Figure S2).

EGFR Plays a Significant Role in Cell Progression in KRAS-Dependent Lung Cancer Cells
To characterize the phenotypic effect of EGFR loss in KRAS-dependent NSCLC cells, we examined the proliferation and migration ability of the A549 EGFR wt/wt and EGFR −/− cells. We observed the morphological changes in EGFR −/− cells. Compared to the normal spindle shaped A549 cells, A549 EGFR −/− cells showed a more flat and irregular polygonal shape ( Figure 1d). A549 EGFR −/− cells showed significant attenuated proliferation and migration properties (Figure 1e-g and Figure S3). To exclude the interference of endogenous EGF, we examined the expression of EGF by Western blot. We did not observe any obvious change in EGF levels in A549 EGFR wt/wt and EGFR −/− cells ( Figure 1h).

EGFR Knockout Does Not Significantly Affect Drug Sensitivity
To determine drug response changes in A549 EGFR wt/wt and EGFR −/− cells, we tested a series of targeted and chemotherapy drugs, including anti VEGFR, BRAF, PDGFR, MEK, and HDAC/HAT, cisplatin, doxorubicin, and staurosporine. At low to intermediate doses of the drugs, we did not observe substantial differences between EGFR wt/wt and EGFR −/− cells (Figure 2). At higher levels, A549 EGFR −/− cells showed slightly more resistance to doxorubicin and cisplatin than EGFR wt/wt cells did. We also observed more resistance to staurosporine, a well-known apoptosis inducer, in A549 EGFR −/− cells ( Figure 2). However, the EGFR −/− cells were more sensitive to SAHA as compared to the EGFR wt/wt cells at a dose of 10 µM.

Effect of EGFR Loss on Downstream Pathways
To test the effect of EGF on cell growth, we performed a clonogenic proliferation assay in the presence and absence of EGF in A549 EGFR wt/wt and EGFR −/− cells. As expected, EGF can stimulate EGFR wt/wt cell proliferation but did not affect it in EGFR −/− cells (Figure 3a). Thus, EGFR loss resulted in decreased cell proliferation with or without EGF (Figure 3a). EGFR loss also led to a stronger inhibition of cell proliferation as compared to gefitinib and cetuximab treatment ( Figure 3b). As EGFR plays crucial roles in the PI3K-Akt and MAPK/ERK pathways, we determined changes in these two key pathways after EGFR loss. We investigated expression of EGFR, pEGFR, Akt, pAkt (Thr308 and Ser473), and MAPK (pERK) by Western blot. After stimulation with EGF, the induction of pEGFR, pERK, and pAkt (Thr308 and Ser473) was substantially lower in A549 EGFR −/− cells in comparison with EGFR wt/wt cells (Figure 3c). In contrast, pERK levels were slightly higher in A549 EGFR −/− cells than the wild-type cells in the absence of EGF (Figure 3c). For staurosporine, cell viability was determined after 24 h due to its toxicity. Data in bar graphs are represented as mean ± SD (n ≥ 3); two-tailed unpaired student's t-test: * p -values < 0.05; ** p -values < 0.01; *** p -values < 0.001; ns: not significant.

Effect of EGFR Loss on Downstream Pathways
To test the effect of EGF on cell growth, we performed a clonogenic proliferation assay in the presence and absence of EGF in A549 EGFR wt/wt and EGFR −/− cells. As expected, EGF can stimulate EGFR wt/wt cell proliferation but did not affect it in EGFR −/− cells (Figure 3a). Thus, EGFR loss resulted in decreased cell proliferation with or without EGF (Figure 3a). EGFR loss also led to a stronger inhibition of cell proliferation as compared to gefitinib and cetuximab treatment ( Figure 3b). As EGFR plays crucial roles in the PI3K-Akt and MAPK/ERK pathways, we determined changes in these two key pathways after EGFR loss. We investigated expression of EGFR, pEGFR, Akt, pAkt (Thr308 and Ser473), and MAPK (pERK) by Western blot. After stimulation with EGF, the induction of pEGFR, pERK, and pAkt (Thr308 and Ser473) was substantially lower in A549 EGFR −/− cells in comparison with EGFR wt/wt cells (Figure 3c). In contrast, pERK levels were slightly higher in A549 EGFR −/− cells than the wild-type cells in the absence of EGF (Figure 3c). Treatment of the A549 EGFR wt/wt and EGFR −/− cells with different drugs. Cells were treated with drugs at indicated concentrations for 72 h and cell viability was determined with an MTS assay. For staurosporine, cell viability was determined after 24 h due to its toxicity. Data in bar graphs are represented as mean ± SD (n ≥ 3); two-tailed unpaired student's t-test: * p-values < 0.05; ** p-values < 0.01; *** p-values < 0.001; ns: not significant.

CXCR7 Is Significantly Upregulated in A549 EGFR −/− Cells and EGFR wt/wt Cells Treated with EGFR Inhibitors
We performed qRT-PCR on a panel of genes associated with the EGFR/MAPK pathway and epithelial-mesenchymal transition (EMT) (Table S2). Surprisingly, CXCR7 showed a marked upregulation at the RNA and protein levels in A549 EGFR −/− cells (Figure 3d,e and Figure S4). In addition, we observed overexpression of CXCL12 (19-fold), a main ligand for CXCR7, in EGFR −/− cells (Figure 3d). Treatment of the EGFR wt/wt cells with gefitinib or cetuximab for 72 h resulted in an increased CXCR7 expression level (Figure 3f). This shows that inhibition or loss of EGFR induces CXCR7 expression in A549 cells and may subsequently contribute to tumor survival.
In contrast to CXCR7, moderate changes were observed in the expression of HER3 (twofold higher) and HER4 (twofold higher) in A549 EGFR −/− cells as compared to EGFR wt/wt cells (Figure 3d). Differences in HER2 mRNA levels were negligible. We did not observe any significant changes in the expression of EMT-related markers. Only the levels of α-catenin mRNA were slightly increased by 1.6-fold in A549 EGFR −/− cells (Figure 3d).

CXCR7 Is Significantly Upregulated in A549 EGFR −/− Cells and EGFR wt/wt Cells Treated with EGFR Inhibitors
We performed qRT-PCR on a panel of genes associated with the EGFR/MAPK pathway and epithelial-mesenchymal transition (EMT) (Table S2). Surprisingly, CXCR7 showed a marked upregulation at the RNA and protein levels in A549 EGFR −/− cells (Figure 3d,e and Figure S4). In addition, we observed overexpression of CXCL12 (19-fold), a main ligand for CXCR7, in EGFR −/− cells (Figure 3d). Treatment of the EGFR wt/wt cells with gefitinib or cetuximab for 72 h resulted in an increased CXCR7 expression level (Figure 3f). This shows that inhibition or loss of EGFR induces CXCR7 expression in A549 cells and may subsequently contribute to tumor survival.
In contrast to CXCR7, moderate changes were observed in the expression of HER3 (twofold higher) and HER4 (twofold higher) in A549 EGFR −/− cells as compared to EGFR wt/wt cells (Figure 3d). Differences in HER2 mRNA levels were negligible. We did not observe any significant changes in the expression of EMT-related markers. Only the levels of α-catenin mRNA were slightly increased by 1.6-fold in A549 EGFR −/− cells (Figure 3d).

Dual Inhibition of CXCR7 and EGFR Downregulates MAPK (ERK1/2) and Suppresses Proliferation
We investigated the synergistic inhibitory effect of gefitinib, erlotinib, and afatinib with a CXCR7 inhibitor. We observed a significant synergistic inhibitory effect of CXCR7 inhibitor combined with afatinib ( Figure 4 and Table 1). However, a combination of CXCR7 inhibitor with erlotinib or gefitinib did not show an obvious inhibitory effect (Figure 4).

Dual Inhibition of CXCR7 and EGFR Downregulates MAPK (ERK1/2) and Suppresses Proliferation
We investigated the synergistic inhibitory effect of gefitinib, erlotinib, and afatinib with a CXCR7 inhibitor. We observed a significant synergistic inhibitory effect of CXCR7 inhibitor combined with afatinib ( Figure 4 and Table 1). However, a combination of CXCR7 inhibitor with erlotinib or gefitinib did not show an obvious inhibitory effect (Figure 4).     Next, we knocked down CXCR7 by siRNAs in A549 EGFR wt/wt and EGFR −/− cells to explore whether cell proliferation can be further suppressed (Figure 5a,b). We showed that CXCR7 knockdown alone did not change A549 EGFR wt/wt cell growth. However, CXCR7 knockdown significantly decreased proliferation of A549 EGFR −/− cells (Figure 5b). To gain insight into the CXCR7-dependent mechanisms that lead to tumor cell proliferation, we checked MAPK (ERK1/2) expression. CXCR7 knockdown decreased pERK levels in A549 EGFR −/− cells as compared to EGFR wt/wt cells (Figure 5c). Moreover, CXCR7 inhibition can further restrain the proliferation of EGFR −/− cells (p < 0.001) (Figure 5d) and decrease pERK levels ( Figure 5e). Conversely, stimulation of CXCR7 with its ligand CXCL12 promoted proliferation of A549 EGFR −/− cells ( Figure 5d) and increase pERK levels ( Figure 5e). Taken together, our results suggest that CXCR7 induces proliferation by enhancing pERK levels in A549 EGFR −/− cells.

Rescue of EGFR Leads to CXCR7 Downregulation in A549 EGFR −/− Cells
To further confirm the interaction between EGFR and CXCR7, we rescued EGFR in A549 EGFR −/− cells. Overexpression of wild-type EGFR in A549 EGFR −/− cells restored EGFR membrane expression (Figure 5f). In the control groups, EGFR signals were not detected. Interestingly, CXCR7 expression dramatically decreased after the reintroduction of EGFR to levels slightly above those in EGFR wt/wt cells (Figure 5g). These data suggest a tight functional regulatory link between EGFR and CXCR7.

CXCR7 Expression, Altered by the Ablation of EGFR, Is Associated with the Genotype of NSCLC
We next determined whether CXCR7 shows similar effects in lung cancer cells with different genetic backgrounds. Knockout of EGFR in H1299 cells, with wild-type EGFR, led to a dramatic decrease in CXCR7 levels (Figure 6a). Knockout of EGFR in H1650, with exon 19 deletions in EGFR, revealed higher expression of CXCR7.
Next, we treated H1299, H1650, and HCC827 cells with gefitinib to investigate changes in the expression levels of CXCR7. H1299 is resistant to gefitinib (IC50, 40 µmol/L) [32]. In contrast to the H1299 EGFR −/− cells, H1299 parental cells showed a decreased CXCR7 expression level after one-day treatment with gefitinib and an increased expression with a low dose of gefitinib three and seven days after the start of treatment (Figure 6b).
Interestingly, both CXCR7 and EGFR showed a highly dose-and time-dependent expression pattern in H1650 cells after treatment with gefitinib ( Figure 6b). Moreover, EGFR and CXCR7 expression showed a clear negative correlation upon gefitinib treatment. Particularly, CXCR7 expression was decreased, whereas EGFR expression increased after treatment with different doses of gefitinib over time in H1650 (Figure 6b). Both EGFR and CXCR7 expression increased after one day of treatment with gefitinib ( Figure 6b). However, in EGFR-mutant NSCLC cell lines (HCC827 and H1650), EGFR and CXCR7 expression substantially decreased after longer treatment (three and seven days) with gefitinib in a dose-dependent manner (Figure 6b). Thus, in EGFR-mutant NSCLC cell lines (HCC827 and H1650), CXCR7 expression levels decreased with gefitinib treatment. Conversely, in EGFR wild-type NSCLC cell lines (H1299 and A549), CXCR7 expression levels increased after gefitinib treatment (Figures 3f and 6b). Together, these data indicate that CXCR7 and EGFR have a close interaction that is dependent on the EGFR genotype.

CXCR7 Expression, Altered by the Ablation of EGFR, Is Associated with the Genotype of NSCLC
We next determined whether CXCR7 shows similar effects in lung cancer cells with different genetic backgrounds. Knockout of EGFR in H1299 cells, with wild-type EGFR, led to a dramatic decrease in CXCR7 levels (Figure 6a). Knockout of EGFR in H1650, with exon 19 deletions in EGFR, revealed higher expression of CXCR7. Next, we treated H1299, H1650, and HCC827 cells with gefitinib to investigate changes in the expression levels of CXCR7. H1299 is resistant to gefitinib (IC50, 40 μmol/L) [32]. In contrast to the H1299 EGFR −/− cells, H1299 parental cells showed a decreased CXCR7 expression level after one-day treatment with gefitinib and an increased expression with a low dose of gefitinib three and seven days after the start of treatment (Figure 6b).
Interestingly, both CXCR7 and EGFR showed a highly dose-and time-dependent expression pattern in H1650 cells after treatment with gefitinib ( Figure 6b). Moreover, EGFR and CXCR7 expression showed a clear negative correlation upon gefitinib treatment. Particularly, CXCR7 expression was decreased, whereas EGFR expression increased after treatment with different doses of gefitinib over time in H1650 (Figure 6b). Both EGFR and CXCR7 expression increased after one day of treatment with gefitinib ( Figure 6b). However, in EGFR-mutant NSCLC cell lines (HCC827 and H1650), EGFR and CXCR7 expression substantially decreased after longer treatment (three and seven days) with gefitinib in a dose-dependent manner (Figure 6b). Thus, in EGFR-mutant NSCLC cell lines (HCC827 and H1650), CXCR7 expression levels decreased with gefitinib treatment. Conversely, in EGFR wild-type NSCLC cell lines (H1299 and A549), CXCR7 expression levels

CXCR7 Is Highly Expressed in Primary Lung Adenocarcinoma
To investigate the clinical relevance of CXCR7 expression in NSCLC patients, we evaluated CXCR7 expression in 47 primary NSCLC patients by IHC. Immunostaining revealed strong cytoplasmic expression of CXCR7 in 74.5% (35/47), weak expression in 17% (8/47), and heterogeneous expression (weak and strong) in 8.5% (4/47) of the tumor samples ( Figure 6c and Table 2). CXCR7 expression was weak in the healthy lung tissue (Figure 6c). CXCR7 expression was strong in the vast majority of the ADCs (92%). In contrast to ADC cases, only 40.5% of SQCC tumors showed strong CXCR7 expression. Approximately 33.3% had a weak expression, and 22.2% of the SQCC tumors showed an intra-heterogeneous CXCR7 expression pattern. In this heterogeneous group, the expression of CXCR7 was weaker in more differentiated tumor cells as compared to the less differentiated ones (Figure 6c and Table 2). Other subtypes (NSCLC NOS and pleomorphic carcinoma; n = 4) showed strong expression. These results suggest that CXCR7 may play a role in the development of different histological subtypes of NSCLC tumors.

Discussion
KRAS and EGFR are commonly mutated in NSCLC. However, whether targeting EGFR is an option for the treatment of KRAS-dependent NSCLC is not clear. In this study, we showed that EGFR plays a critical role in the proliferation of KRAS-dependent cells. Furthermore, we report a novel genotype-dependent crosstalk between EGFR and CXCR7 in NSCLC cells. We showed that EGFR disruption or blockage can lead to upregulation of CXCR7 and activation of the MAPK signaling pathway, resulting in outgrowth or survival advantage of the KRAS-dependent cells. We also demonstrated that dual inhibition of EGFR and CXCR7 suppressed tumor cell growth and reduced MAPK activity by downregulation of pERK levels. Dual inhibition of EGFR and CXCR7 by inhibitors leads to a synergistic effect, which may provide a novel therapeutic approach for NSCLC patients. We also showed that CXCR7 is overexpressed in ADC patients as compared to the SQCC cases.
KRAS-mutant tumor cells appear to rely on KRAS for cell proliferation [33]. In this study, we showed that EGFR also plays an essential role in the survival of NSCLC cells harboring KRAS mutations. We also observed morphological changes in EGFR −/− lung tumor cells as well as substantial reduction in migration ability of these cells. It is not surprising that previous studies have shown that gefitinib is not efficient for the treatment of NSCLC patients with wild-type EGFR, because gefitinib has a higher affinity to the mutated version of EGFR, such as L858R point mutation, as compared to wild-type EGFR [34]. Anti-EGFR therapy with EGFR blocking antibody cetuximab shows resistance due to dysregulation of EGFR and activation of other HER family receptors [35]. Thus, these studies do not show that wild-type EGFR is not necessary for tumor proliferation. Based on our results, wild-type EGFR is important for proliferation and migration of KRAS-mutant tumor cells.
We used 11 inhibitors, including MAPK associated inhibitors, chemotherapeutic agents, HDAC and HAT inhibitors, and an apoptosis inducer agent to evaluate the drug response in A549 EGFR wt/wt and EGFR −/− cells. However, we did not observe significant differences between the two genotypes. These results indicate the ineffectiveness of clinically available drugs for the treatment of KRAS-mutant NSCLC cells. In contrast, simultaneous inhibition (inhibitors or knockdown) of EGFR and CXCR7 significantly attenuated cancer cell growth. Based on our results, we propose a potential novel strategy for the treatment of KRAS-dependent NSCLC by dual inhibition of EGFR and CXCR7. Further in vivo studies are needed to identify appropriate delivery methods that can efficiently and specifically target EGFR and CXCR7 in tumor cells [36,37].
CXCR7 contributes to angiogenesis and cell growth in lung tumors [21,23,38,39]. In addition, overexpression of CXCR7 was associated with poorer prognosis in lung cancer patients [33]. The latter effect was probably induced by the activation of the MAPK pathway [28,40]. We showed that EGFR blockage or knockout leads to reduced growth and a significant increase in CXCR7 levels in A549 cells. CXCR7 knockdown decreased pERK levels and further suppressed proliferation of A549 EGFR −/− cells. Dual inhibition of EGFR and CXCR7 by inhibitors can achieve a synergistic inhibitory effect. Thus, our observations suggest that CXCR7 overexpression is one of the survival mechanisms in KRAS-dependent tumor cells with EGFR loss or cells treated with EGFR-TKIs. Interestingly, we only observed the inhibitory effect of concomitant combination of CXCR7 inhibitor and afatinib, but not gefitinib or erlotinib. One possible explanation is that afatinib is an irreversible HER family inhibitor. Gefitinib and elortinib only have a strong inhibitory effect on mutant EGFR, but afatinib can effectively inhibit both wild-type and mutant EGFR [41,42]. It has previously been shown that afatinib can suppress tumor growth in KRAS-mutant NSCLCs, but not erlotinib or gefitinib. Interestingly, our in vitro results are in concordance with a previously published in vivo study [43]. These observations demonstrate that dual inhibition of CXCR7 and EGFR have a synergistic therapeutic effect in KRAS-mutant NSCLC cells.
CXCL12 is involved in cell proliferation and brain metastasis in lung cancer patients [39]. CXCL12 is overexpressed in NSCLC cell lines and primary lung cancer [44] and may promote tumor cell migration and growth through the ERK pathway [45]. We observed an increased expression of CXCL12, a shared ligand of CXCR7 and CXCR4 with a higher affinity towards CXCR7, in A549 EGFR −/− cells. CXCR7 knockdown is highly efficient in the suppression of cell proliferation of A549 EGFR −/− cells, as shown in our study. However, other ligands of CXCR7 such as CXCL11, MIF, and Dickkopf-3 may also influence CXCR7-mediated cell proliferation, which requires further investigation. We did not observe significant changes in the expression levels of other HER family receptors including HER2, HER3, and HER4 in A549 EGFR −/− cells. Thus, our data suggest that HER family member receptors are not significantly affected by EGFR loss.
Our results show that MAPK/pERK is continuously activated in both EGFR −/− and EGFR wt/wt cells. This is indicative of an EGFR-independent survival mechanism, which may be partly explained by the presence of KRAS activating mutations in these cells. KRAS mutations can lead to constant activation of RAS/RAF/MEK/ERK [46]. Furthermore, as EGFR is upstream of the RAS pathway, loss of EGFR may not affect activation of RAS/RAF/MEK/ERK introduced by mutant KRAS. Notably, one study showed that KRAS pathways coordinate the transduction of CXCL12/CXCR4/CXCR7 and influence growth of pancreatic cancer [47]. It will be of great interest to explore whether CXCR7 expression, altered by the ablation of EGFR, is associated with KRAS in NSCLC. We showed that EGF can still stimulate pAkt in EGFR −/− cells. This observation indicates that EGF-induced phosphorylation of Akt may be independent of EGFR, which to our knowledge has not been reported previously. One hypothesis is that EGF stimulates pAkt via CXCR7, because previous studies have shown that CXCR7 associated pathways can induce Akt phosphorylation [48,49], and we observed CXCR7 overexpression in A549 EGFR −/− cells. In addition, one recent study reported that Akt activation is related to EGFR drug resistance in lung cancer [50]. Given the importance of Akt for tumor survival and growth [51], future studies are needed to explore the crosstalk between EGF-EGFR and CXCL12-CXCR7 in the MAPK pathway.
We found a close inverse relationship between EGFR and CXCR7 expression (Figure 7). When EGFR is expressed in the A549 EGFR −/− cells, a decreased expression of CXCR7 is observed. Mechanisms leading to CXCR7 upregulation in EGFR −/− cells are still not clear. Previous studies showed that the CXCR7 promoter contains five NF-kB binding sites and activated NF-kB can upregulate CXCR7 [52]. Thus, one potential mechanism can be that CXCR7 is transcriptionally activated by NF-kB. Several studies have shown that NF-kB activation drives survival of tumor cell treated with EGFR-TKIs [53][54][55][56]. Another possibility is that CXCR7 overexpression may be related to the mutational status of KRAS [47].
We showed that upon treatment with gefitinib, CXCR7 expression dramatically decreased in EGFR-mutant NSCLC cell lines, but, conversely, it is increased in EGFR wild-type NSCLC cell lines. Blocking mutant EGFR with TKIs can substantially inhibit MAPK/ERK and suppress tumor cell proliferation [57,58]. This indicates that EGFR-TKIs may reduce pERK and inhibit tumor cell proliferation partially through downregulation of CXCR7. Moreover, insensitivity or acquired resistance of EGFR wild-type NSCLC cells for TKIs might be due to increased CXCR7 levels. Our results show that CXCR7 expression is highly related to the genetic background of the lung cancer cells, at least in the three cell lines we tested, suggesting the importance of personalized therapy in lung cancer treatment. Importantly, we observed higher expression of CXCR7 in lung ADCs (92% strongly expressed CXCR7) as compared to lung SQCC cases (40.5% strongly expressed CXCR7). One study, which performed a meta-analysis in nearly 3000 patients, showed that EGFR was less expressed in ADC (39%) than in SQCC (58%) [59]. Given this study and our results, there may be a correlation between EGFR and CXCR7 in patients with NSCLC. Moreover, higher expression of CXCR7 in the less differentiated SQCC tumors may be due to its critical role during tumor initiation and the differentiation process of this tumor subtype. Higher expression of CXCR7 in lung cancer tissues as compared to the healthy lung tissue suggests CXCR7 may be a target for lung cancer treatment. treated with EGFR-TKIs [53][54][55][56]. Another possibility is that CXCR7 overexpression may be related to the mutational status of KRAS [47]. We showed that upon treatment with gefitinib, CXCR7 expression dramatically decreased in EGFR-mutant NSCLC cell lines, but, conversely, it is increased in EGFR wild-type NSCLC cell lines. Blocking mutant EGFR with TKIs can substantially inhibit MAPK/ERK and suppress tumor cell proliferation [57,58]. This indicates that EGFR-TKIs may reduce pERK and inhibit tumor cell proliferation partially through downregulation of CXCR7. Moreover, insensitivity or acquired resistance of EGFR wild-type NSCLC cells for TKIs might be due to increased CXCR7 levels. Our results show that CXCR7 expression is highly related to the genetic background of the lung cancer cells, at least in the three cell lines we tested, suggesting the importance of personalized therapy in lung cancer treatment. Importantly, we observed higher expression of CXCR7 in lung ADCs (92% strongly expressed CXCR7) as compared to lung SQCC cases (40.5% strongly expressed CXCR7). One study, which performed a meta-analysis in nearly 3000 patients, showed that EGFR was less expressed in ADC (39%) than in SQCC (58%) [59]. Given this study and our results, there may be a correlation between EGFR and CXCR7 in patients with NSCLC. Moreover, higher expression of CXCR7 in the less differentiated SQCC tumors may be due to its critical role during tumor initiation and the differentiation process of this tumor subtype. Higher expression of CXCR7 in lung cancer tissues as compared to the healthy lung tissue suggests CXCR7 may be a target for lung cancer treatment.

Conclusions
Our data show that wild-type EGFR plays a significant role in KRAS-mutant NSCLC cancer cells and revealed CXCR7 upregulation as a potential survival mechanism in KRAS-mutant cells upon EGFR loss (Figure 7). Our findings suggest that dual inhibition of EGFR and CXCR7 may be a promising therapeutic strategy for at least a subset of patients with NSCLC.
Supplementary Materials: The following are available online at www.mdpi.com/xxx/s1, Figure S1: Schematic diagram of gRNAs targeting exon3 of the EGFR gene in A549 cells and validation by Sanger sequencing for clone Figure 7. The proposed model of CXCR7-EGFR crosstalk and its effect on the MAPK signaling pathway. EGFR-ERK and CXCR7-ERK signaling pathways promote tumor cell development (left). Ablation of EGFR leads to overexpression of CXCR7/CXCL12 and subsequently stimulates the ERK pathway, resulting in tumor cell survival (middle). Dual inhibition of EGFR and CXCR7 further suppresses ERK signaling and synergistically inhibits tumor growth (right). Thickness of the arrow shows strength of the signaling. The number of solid dots (in red or green) shows the amount of the endogenous epidermal growth factor (EGF) or CXCL12.

Conclusions
Our data show that wild-type EGFR plays a significant role in KRAS-mutant NSCLC cancer cells and revealed CXCR7 upregulation as a potential survival mechanism in KRAS-mutant cells upon EGFR loss (Figure 7). Our findings suggest that dual inhibition of EGFR and CXCR7 may be a promising therapeutic strategy for at least a subset of patients with NSCLC.
Supplementary Materials: The following are available online at http://www.mdpi.com/2072-6694/11/4/455/s1, Figure S1: Schematic diagram of gRNAs targeting exon3 of the EGFR gene in A549 cells and validation by Sanger sequencing for clone 1, Figure S2: Western blot demonstrating EGFR expression in EGFR wt/wt and absence of EGFR protein in EGFR −/− cells (Clone 1), Figure S3: Wound healing assay to evaluate wound closure and cell migration ability at different time points (Clone 1), Figure S4: qPCR and Western blot analysis of CXCR7 in A549 EGFR wt/wt and EGFR −/− cells (Clone 1), Table S1: List of gRNA sequences for CRISPR/Cas9, Table S2: List of the primer sets used for qRT-PCR.