ENO1 (α-enolase) expression is significantly correlated with reduced survival and poor prognosis in many cancer types, including lung cancer. However, the function of ENO1 in carcinogenesis remains elusive. In this study, we found that high expression of ENO1 is present in metastatic lung cancer cell lines and malignant tumors and is associated with poor overall survival of patients with lung cancer. Knockdown of ENO1 decreased cancer cell proliferation and invasiveness, whereas overexpression of ENO1 enhanced these processes. Moreover, ENO1 expression promoted tumor growth in orthotopic models and enhanced lung tumor metastasis in tail-vein injection models. These effects were mediated by upregulation of mesenchymal markers N-cadherin and vimentin and the epithelial-to-mesenchymal transition regulator SLUG, along with concurrent downregulation of E-cadherin. Mechanistically, ENO1 interacted with hepatocyte growth factor receptor (HGFR) and activated HGFR and Wnt signaling via increased phosphorylation of HGFR and the Wnt coreceptor LRP5/6. Activation of these signaling axes decreased GSK3β activity via Src–PI3K–AKT signaling and inactivation of the β-catenin destruction complex to ultimately upregulate SLUG and β-catenin. In addition, we generated a chimeric anti-ENO1 mAb (chENO1-22) that can decrease cancer cell proliferation and invasion. chENO1-22 attenuated cancer cell invasion by inhibiting ENO1-mediated GSK3β inactivation to promote SLUG protein ubiquitination and degradation. Moreover, chENO1-22 prevented lung tumor metastasis and prolonged survival in animal models. Taken together, these findings illuminate the molecular mechanisms underlying the function of ENO1 in lung cancer metastasis and support the therapeutic potential of a novel antibody targeting ENO1 for treating lung cancer.

Significance:

This study shows that ENO1 promotes lung cancer metastasis via HGFR and WNT signaling and introduces a novel anti-ENO1 antibody for potential therapeutic use in lung cancer.

ENO1 is a metabolic enzyme involved in the synthesis of pyruvate, and it is also known to act as a plasminogen receptor and mediate the activation of plasmin and extracellular matrix degradation (1). On the basis of these mechanisms, ENO1 contributes to a wide variety of cellular and physiologic processes, such as proteolysis, extracellular matrix remodeling, metabolism, growth control, tumor metastasis, and allergic response (2, 3). Overexpression of ENO1 mRNA or protein has been observed in many different tumor types, including lung, brain, breast, cervix, colon, gastric, head and neck, kidney, leukemia, liver, ovary, pancreas, prostate, skin, and testicular cancers (3, 4). In addition to its role as a glycolytic enzyme, ENO1 is expressed at the cell surface of most tumors, where it forms a multiprotein complex with urokinase-type plasminogen activator receptor (uPAR), integrins, and certain cytoskeletal proteins, which are responsible for adhesion, migration, and proliferation (5). In tumors, ENO1 can modulate intravascular and pericellular fibrinolytic activity to promote cell migration and cancer metastasis (6–8). In addition, the expression level of ENO1 may serve as a valuable diagnostic biomarker, as it is associated with patient outcomes, such as cancer survival and prognosis (9–11). For example, among patients with non–small cell lung carcinoma (NSCLC), those with tumor cells expressing high levels of ENO1 have relatively poor disease-free and overall survival (OS) rates (12). Recent findings have suggested that targeting ENO1 could be an effective novel approach to overcome drug resistance (13). As such, tumor-associated antigen (TAA)-ENO1 has become an intriguing target for cancer therapy. However, at time of writing this manuscript, no anti-ENO1 drug candidates have yet progressed to clinical trials, suggesting that there is an unmet need for efficacious anti-ENO1 antibodies.

Epithelial-to-mesenchymal transition (EMT) is a cellular program that naturally occurs in a broad range of tissue types and developmental stages (14). However, it has become apparent that EMT is also important in the pathogenesis of many human diseases and cancers, based on its contributions to organ fibrosis (15, 16), therapeutic resistance (17), inflammatory response (18, 19), immunosuppression (20, 21), and metastatic cell dissemination (22). Furthermore, a growing body of evidence has shown that EMT is involved in many tumor metastasis processes, including local invasion of neoplastic cells at the primary tumor site, intravasation into blood vessels, circulation through the vasculature, extravasation into the parenchyma of distant tissues, and survival of micrometastatic deposits (23, 24). To distinguish between cells on the extreme poles of the epithelial and mesenchymal axis, researchers typically examine certain cellular characteristics. As such, epithelial cells display epithelial cell–cell junctions, apical–basal polarity, and lack of motility, while mesenchymal cells exhibit heightened motility, invasiveness, and resistance to apoptosis, along with spindle-like morphology that lacks apical–basal polarity (25). EMT is often detected as a morphologic switch from an epithelial to a mesenchymal phenotype, including loss of epithelial cell markers (e.g., E-cadherin, α-catenin, and γ-catenin) and gain of mesenchymal markers (e.g., vimentin, N-cadherin, and fibronectin); this switch is executed by EMT-activating transcription factors (EMT-ATF), such as SLUG, SNAIL, and TWIST (23, 26). Therefore, EMT-ATFs have become well known for their pleiotropic roles in cancer formation, growth, and metastasis, which underlie their associations with poor clinical outcomes in many epithelial tumor types.

Wnt signaling is a critical pathway for embryonic development and adult cellular injury and repair processes. Upon Wnt ligand binding to its receptors, the capacity of the β-catenin destruction complex to phosphorylate cytosolic β-catenin is inhibited. Unphosphorylated β-catenin accumulates in the cytosol and translocates to the nucleus, where it activates expression of Wnt target genes. This genetic regulation requires integration with the T-cell factor/lymphoid enhancer factor (TCF/LEF) family of transcription factors and is important for controlling cellular proliferation, differentiation, and survival in many contexts, including carcinoma progression (27, 28). In addition to its role in the nucleus, E-cadherin forms a complex with β-catenin at the cell surface that facilitates the maintenance of cell–cell junctions (29). Similar to Wnt ligand binding, downregulation of E-cadherin can also increase levels of β-catenin in both cytoplasmic and nuclear compartments, stimulating cell growth and providing a mechanistic connection with EMT (30).

Another important signaling pathway governing EMT is stimulated by hepatocyte growth factor receptor (HGFR; also called c-Met). HGFR and its ligand HGF are involved in many normal and pathologic biological processes, such as fetal development of liver, placenta, muscle, and the nervous system (31, 32). After birth, activation of the HGF–HGFR pathway appears to be involved in EMT (33), as well as hepatic, renal, and epidermal regeneration (31). Importantly, HGFR signaling also promotes tumor growth, invasion, drug resistance, angiogenesis, and especially the generation and maintenance of cancer stem cells. Activation of the HGFR signaling pathway can occur by several different mechanisms, including increased HGF expression, enhanced HGFR protein expression, or by alteration of other factors or pathways affecting HGFR activation (34). Upon HGFR activation, phosphorylation of its carboxy-terminal tail creates a multifunctional docking site that recruits intracellular adapters and substrates (PI3K, Src, and others; ref. 35). Thus, several pathways involve in proliferation, survival, cell motility, invasion, and metastasis can be activated by this critical signaling pathway.

Our previous findings indicated that ENO1-targeting liposomes loaded with anticancer drugs have tremendous potential for application as a targeted drug delivery system for cancer therapy (36). In this study, we examined the biological significance of ENO1 in tumor progression and invasion, and we demonstrated that ENO1-promoted lung cell invasion, tumor metastasis, and progression involves the HGFR–Src–PI3K–AKT–GSK3β–SLUG and HGFR–WNT–GSK3β–SLUG signaling axes, suggesting that in addition to its utility as a biomarker, ENO1 is also a promising therapeutic target for lung cancer. We then generated a mAb against ENO1, called chimeric ENO1-22 (chENO1-22), which can bind to ENO1 on the surface of lung cancer cells. Treatment with chENO1-22 reduced lung cancer cell growth and invasion, and importantly, chENO1-22 blocked metastasis and improved the survival of mice in a xenograft model. Because chENO1-22 inhibited ENO1-HGFR–mediated downstream signaling to promote SLUG protein degradation, the mAb may be considered a promising therapeutic candidate for lung cancer.

Cell culture

NL-20 and CL1–5 cells were purchased from ATCC and were authenticated by ATCC based on DNA profile, cytogenetic analysis, and isoenzymology. The cells were cultured in accordance with ATCC protocols and were passaged for fewer than 6 months after resuscitation. NL-20 is an immortalized, nontumorigenic human bronchial epithelial cell line derived from normal bronchus. NHBE cells are normal human bronchial epithelial cells, which were purchased from Applied Biological Materials Inc. Hop62, CL1–0, and CL141 cells were generous gifts from Dr. Pan-Chyr Yang (National Taiwan University). All the cell lines were tested negative for Mycoplasma [Mycoplasma PCR Detection Kit (Applied Biological Materials Inc.)] throughout the study.

RT-PCR and Western blotting

RT-PCR and immunoblotting were conducted as described previously (37). The following primary antibodies were used in Western blotting assay: anti-ENO1 (11204–1-AP; ProteinTech), anti–E-cadherin (24E10; Cell Signaling Technology.), anti-vimentin (D21H3; Cell Signaling Technology), anti–N-cadherin (610920; BD Biosciences), anti-SNAIL (C15D3; Cell Signaling Technology), anti-TWIST (GTX127310; GeneTex), anti-SLUG (C19G7; Cell Signaling Technology), anti–GSK3β (27C10; Cell Signaling Technology), phospho–GSK3β (Y216; 612312; BD Biosciences), Phospho–GSK3β (Ser9; 5B3; Cell Signaling Technology), anti-non-phospho (Active) β-catenin (D13A1; Cell Signaling Technology), anti–β-catenin (D10A8; Cell Signaling Technology), Phospho-Src Family (Tyr416; D49G4; Cell Signaling Technology), phospho-PI3 Kinase p85 (Tyr458)/p55 (Tyr199; Cell Signaling Technology), phospho-Akt (Ser473; Cell Signaling Technology), anti-FLAG (F3165; Sigma-Aldrich) and anti-Myc (9E10; Santa Cruz Biotechnology), and anti-HGFR (D1C2; Cell Signaling Technology). Horseradish peroxidase (HRP)-conjugated rabbit antimouse or goat antirabbit secondary antibodies (Santa Cruz Biotechnology) were used as appropriate. Recombinant human ENO1 was purchased from Sino Biological, Inc. Cells were treated with GSK3β inhibitor (1 μmol/L; Bio; Sigma-Aldrich Inc.) for 8 hours. For rescue experiments, ENO1-knockdown CL1–5 cells were incubated with recombinant ENO1 (1 μg/mL; Sino Biological, Inc.) for 30 minutes. Cells were treated with hHGF (100 ng/mL) for 10 minutes.

Flow cytometry

Flow cytometry was carried out on a Beckman Dickinson LSR-II flow cytometer (Cornell Biotechnology; Flow Cytometry Core Facility) equipped with 407, 488, 633 nm lasers. Typical single-color analysis was conducted. Curve fitting and data analysis were performed using GraphPad Prism (v 6.0). P values were obtained by unpaired Student t test. Cell counting was performed with a Countess II (Invitrogen).

Tissue microarrays and IHC

Tissue microarrays were obtained from commercial sources (SuperBioChips). A total of nine normal lung samples, 40 primary lung cancer samples, and 10 metastatic lung cancer cells were tested. The primary antibody used for staining was anti-ENO1 (Abcam Biotechnology). Intensities of IHC signal were scored as 0 (no expression), 1 (faint expression), 2 (moderate expression), and 3 (intense expression) independently by a pathologist blinded to group-identifying information. Areas of positive staining were also evaluated and expressed as percent of total area.

Functional analysis of the target protein

Lentiviruses (pLKO.1) containing the shRNA and pLKO.1 empty vector or pLKO-luciferase controls were generated and used to infect cancer cells following standard procedures. Stable transfectants were established by puromycin selection as described previously (38). In addition to overexpression of the gene products after transfection, the target gene effects were evaluated by functional assays.

Cell viability

Cells were harvested by trypsinization and were resuspended in RPMI containing 10% FBS. The cells were then plated at a density of 3,000 per well in 96-well plates and incubated overnight. Four replicates were treated with each concentration of drug; PBS buffer was added to the control group. After adding WST-1, the reagent was kept in the 96-well plate for 1 hour. The absorbance at 450 nm was monitored.

ATP luminescence assay

ATP levels were monitored using a CellTiter-Glo Luminescent Cell Viability Assay from Promega. ATP was measured by bringing the plate to room temperature and adding CellTiter-Glo directly to each well, as directed by the manufacturer. The contents of the plate were mixed on an orbital shaker to induce cell lysis. Then, the samples were read on a luminescence plate reader. Data are presented as mean ± SE of luminescent readings from three separate experiments.

Matrigel transwell cell invasion assay

Matrigel invasion assay was performed using a 24-well transwell plates (Costar) with polycarbonate filters (pore size, 8 μm). Matrigel was attenuated to 1 mg/mL with 4°C serum-free medium on ice and was added to the top chamber of the plates. After 30 minutes of incubation in a 37°C incubator, a cell suspension with 1 × 105 cells/mL was added. For the bottom chamber, 800 μL medium with 10% FBS was added. After 16–24 hours the cells were fixed with 95% methanol for 5 minutes. Crystal violet (0.1%) was used to stain cells. The transwell plate was then observed under a microscope.

Apoptosis assay

A total of 1 × 106 cells were seeded in a 10-cm dish overnight. Then, cells were detached by dissociation buffer and stained with FITC–Annexin-V antibody and propidium iodide (PI; BD Biosciences) in 200 μL reaction buffer for 20 minutes at room temperature. The samples were analyzed by FACScan (BD FACSCanto II, BD Biosciences). Cells positive for FITC only or both PI and FITC were counted as apoptotic cells.

Site-directed mutagenesis

To construct the mutant ENO1 plasmid, two specific primer sets were used: S40A (For: 5′ CTG TGC CCA GTG GTG CTG CTA CTG GTA TCT ATG AGG CC 3′; Rev: 5′ GGC CTC ATA GAT ACC AGT AGC AGC ACC ACT GGG CAC AG 3′) and D245R (For: 5′ GGT GGT CAT CGG CAT GAG AGT AGC GGC CTC CGA GTT C 3′; Rev: 5′ GAA CTC GGA GGC CGC TAC TCT CAT GCC GAT GAC CAC C 3′). PCR reactions were performed using KAPA HiFi DNA Polymerase ReadyMix (Roche). After 20 amplification cycles, the PCR products were digested by Dpn1 (NEB) at 37°C for 1 hour to remove template DNA. The samples were then used to transform Stbl3 (ECOS)-competent cells.

In vivo orthotopic lung tumor and metastasis models

Six-week-old NOD/SCID female mice were used for orthotopic lung tumor and metastatic lung cancer xenograft models. For the orthotopic xenograft model, 1 × 105 CL1–5 cells stably expressing luciferase and infected with shENO1 or control shCtrl were orthotopically injected into the lung parenchyma of NOD/SCID mice. For the metastasis model, 1 × 106 CL1–5 cells stably expressing luciferase and infected with shENO1 or control shCtrl were injected into mice by tail vein injection. All animal studies were approved by the Institutional Animal Care and Use Committee of Academia Sinica, Taipei, Taiwan.

In vivo imaging and quantification of bioluminescence

In vivo imaging was performed on a Bruker In-Vivo Xtreme II (Bruker BioSpin MRI GmbH). d-luciferin, a substrate for firefly luciferase, was injected intraperitoneally (200 μL d-luciferin at 15 mg/mL, Gold Biotechnology) 10 minutes before luminescence imaging. After injection, mice were anesthetized with 3% isoflurane in an induction chamber and then placed in a Sealed Optical Imaging (OI) Tray, ventilated with 2.5% isoflurane.

A region of interest (ROI) template of fixed size for each lung was used to ensure that areas of identical size were measured in different animals. Background was measured for each image and subtracted from the obtained signal. The luminescence value (photons/s/mm2) for a given animal represents the measurement made from the lung.

Cycloheximide protein synthesis inhibition assay

The turnover of proteins was monitored using a standard protocol for cycloheximide inhibition of protein synthesis. Cells were treated with 300 μg/mL cycloheximide (Sigma) for the indicated time points.

TOP/FOP Wnt activity assay

CL1–5 cells (2 × 104 cells per well) were seeded in 24-well plates. Plasmids were transfected into cells using Lipofectamine 3,000 (Invitrogen). A luciferase reporter (TOPflash or FOPflash) or was co-transfected with a pGL4.74 hRluc-TK construct (Renilla luciferase). Transcriptional activity was determined as the expression of firefly luciferase normalized to the Renilla luciferase levels at 48 hours posttransfection. Measurements were made with a dual luciferase reporter assay kit (Promega).

Immunofluorescence analysis

Cells were grown to the appropriate level of confluence and treated as indicated. At the conclusion of the experiment, cells were fixed by addition of paraformaldehyde (PFA) at 4°C for 20 minutes. PFA was removed, after which, cells were washed three times with PBS and blocked with 2 mL imaging blocking buffer (3% BSA in PBS) at 37°C for 1 hours. Plates were washed two times with PBS, then treated with primary antibody at room temperature for 1 hour. Primary antibody was removed, and plates were washed two times with PBS. Then secondary antibody was added for 1 hour. Plates were washed one time in PBS containing 1 mg/mL DAPI, then once with PBS. Samples were stored at 4°C until imaging.

Phospho-RTK array analysis

The Human Phospho-RTK Array Kit (R&D Systems) was used to determine the relative levels of tyrosine phosphorylation for 49 distinct RTKs, according to the manufacturer's protocol. Briefly, the arrays were incubated with 300 μg of protein lysate overnight at 4°C after blocking for 1 hour with Array Buffer 1. The arrays were washed and incubated with an anti–Phospho-Tyrosine-HRP Detection Antibody. Signals were generated with a Chemi Reagent Mix and captured using a chemiluminescence imaging system (GE ImageQuant LAS4000). The intensity of each phospho-RTK array signal was quantitated with ImageJ software; relative intensities of the average signal from each pair of duplicate spots were determined in relation to the negative control spots.

GST pull-down

In the GST pull-down assay, His-tagged HGFR was incubated with glutathione bead-bound GST-ENO1 or bead-bound GST alone (control) in pull-down buffer (PierceTM GST Protein Interaction Pull-Down Kit) at 4°C for 1 hour.

Extracellular interactions between ENO1 and HGFR

CL1–5 cells transfected with AS2-FLAG control or ENO1-FLAG were washed with 10 mmol/L EDTA in PBS, and then 2 mmol/L DTSSP (Thermo Fisher Scientific) was used as a cross-linker to stabilize the interaction between ENO1 and HGFR. To stop the cross-linking reaction, Tris, pH 7.5 was added to a final concentration of 20 mmol/L. Membrane proteins were extracted using Mem-PER1 Eukaryotic Membrane Protein Extraction Reagent Kit (Thermo Fisher Scientific). Finally, the ENO1-HGFR complex was pulled-down with anti-FLAG agarose beads (A2220; Sigma-Aldrich), and probed by Western blotting.

Immunoprecipitation–Western blotting

Protein lysate was incubated with anti-FLAG agarose beads (A2220; Sigma-Aldrich), anti-Myc agarose beads (C3956; Sigma-Aldrich) for coimmunoprecipitation (co-IP) experiments, or anti-SLUG (C19G7; Cell Signaling Technology, Inc.) for ubiquitination (Ub) assays. The following primary antibodies were used in immunoblotting assay: anti-ENO1 (11204–1-AP; ProteinTech), anti-SLUG (C19G7; Cell Signaling Technology), anti-FLAG (F3165; Sigma-Aldrich), anti-Myc (9E10; Santa Cruz Biotechnology), anti-ubiquitin (P4D1; Cell Signaling Technology) and anti-HGFR (D1C2; Cell Signaling Technology). HRP-conjugated rabbit anti-mouse or goat anti-rabbit secondary antibodies (Santa Cruz Biotechnology) were used as appropriate.

The Myc-tagged HGFR expression plasmid, pEF1x-HGFR-F932-Myc, was constructed by cloning the extracellular domain of human HGFR from cDNA into the pEF1x-Myc-His vector. The corresponding truncated mutant expression plasmids, pEF1x-HGFR-F567-Myc, pEF1x-HGFR-SEMA-Myc, pEF1x-HGFR-4IPT-Myc, were generated according to the whole genome sequence (WGS).

Generation of mAbs against tumor antigen

Generation of mAbs against tumor antigens was performed according to standard procedures (39), with some modifications (40–42). Briefly, the spleen of the immunized mouse was removed, and splenocytes were fused with NSI/1-Ab4–1 (NS-1) myeloma cells. The splenocytes and the myeloma cells were washed twice with serum-free DMEM medium. The final pellet was mixed in a 15-mL conical tube, and 1 mL [50%; volume for volume (v/v)] polyethylene glycol (GIBCO BRL) was added over 1 minute with gentle stirring. The mixture was diluted through slow (1 minute) addition of 1 mL of DMEM twice, followed by slow addition (2 minutes) of 8 mL of DMEM medium without serum. The mixture was centrifuged, and the fused cell pellet was resuspended in DMEM medium supplemented with 15% FBS, hypoxanthine-aminopterin-thymidine (HAT) medium, and hybridoma cloning factor and distributed in the 96-well tissue culture plates. Hybridoma colonies were screened by ELISA for secretion of mAbs that bound to tumor antigens. Selected clones were subcloned by limiting dilution. Final hybridoma clones were isotyped using an isotyping kit. Ascitic fluids were produced in pristane-primed BALB/c mice. The hybridoma cell lines were grown in DMEM medium with 10% heat-inactivated FBS. The mAbs were affinity purified with protein G sepharose 4B gels. ELISA assays and Western blotting were used to measure the activity and specificity of the antibodies.

ELISA

Cell culture (96-well) plates (Corning Costar) were fixed with 2% PFA and blocked with 1% BSA. Cells were added to the plates and incubated for 1 hour. The plates were subsequently washed with PBS containing 0.1% (w/v) Tween-20 (PBST0.1), followed by incubation with HRP-conjugated antimouse IgG (115–035–062; Jackson ImmunoResearch Laboratories) for 1 hour. After washing, the plates were incubated with substrate solution (o-phenylenediamine dihydrochloride; P6787, Sigma). The reaction was stopped by the addition of 3 N HCl, and signals were detected using a microplate reader at 490 nm.

Immunogold labeling

CL1–5 cells were fixed with 4% formaldehyde and processed for cryosectioning. Sections were blocked with BSA, followed by labeling for ENO1 with the chENO1 antibody-22 (40 μg/mL) or NHIgG. After fixation and staining, the cells were examined with a transmission electron microscope.

Statistical analysis

Data are presented as mean ± SEM. Comparisons of data between two groups were made with Student t test. Statistical analyses were performed with GraphPad Prism software, version 5.0. All statistical tests were two-sided, and P values greater than 0.05 were considered statistically significant.

ENO1 is highly expressed in metastatic cancer cell lines and patients, and its high expression is associated with worse OS of patients with lung cancer

To assess the correlation between ENO1 expression and lung cancer malignancy, we measured ENO1 protein levels in a panel of lung cancer cells and clinical NSCLC specimens. ENO1 protein level was elevated in highly metastatic lung cancer cell lines (CL1–5 and CL141) compared with nonmetastatic cell lines (CL1–0 and Hop62), and only weak expression was detected in normal human bronchial epithelial (NHBE) cells or NL-20 immortalized human bronchial epithelial cells (Fig. 1A). In addition, the surface expression of ENO1 was upregulated in CL1–5 cells compared with NHBE or NL-20 cells, according to flow cytometry analysis (Fig. 1B) with a chimeric ENO1 antibody (Supplementary Fig. S1A–S1C). Moreover, we performed immunogold labeling assays to confirm ENO1 protein localization. Expectedly, ENO1 was detected on the cell surface, in the cytoplasm, and in the nucleus of CL1–5 cells (Fig. 1C). In tumor samples from NSCLC patients, increased ENO1 expression was associated with increased tumor–node–metastasis (TNM) and metastatic stage, and ENO1 mRNA and protein were not abundant in normal lung tissue (Fig. 1D–F; Supplementary Fig. S2). Furthermore, survival analysis of patients stratified by ENO1 protein expression level revealed a significant decrease in the survival rate of patients with NSCLC showing high ENO1 expression as compared to those with low ENO1 expression (Fig. 1G). Moreover, high mRNA expression of ENO1 was strongly associated with worse OS for NSCLC patients according to the Kaplan–Meier plotter database (KM plotter, http://kmplot.com/analysis/; Fig. 1H) and The Cancer Genome Atlas (TCGA) dataset (Fig. 1I). Taken together, these findings indicate a strong correlation between high expression of ENO1 and lung cancer malignancy.

Figure 1.

ENO1 is highly expressed in metastatic cancer cell lines and is associated with worse OS of patients with lung cancer. A, ENO1 expression was analyzed by Western blotting in NHBE, NL-20, and human lung cancer cell lines using an anti-ENO1 antibody. α-Tubulin served as an internal control. B, The binding of chimeric ENO1 mAb (chENO1) or NHIgG to live NHBE, NL-20, or CL1–5 cells was evaluated by flow cytometry using phycoerythrin (PE)-conjugated goat antimouse IgG. x-axis, intensity of phycoerythrin fluorescence; y-axis, number of events. C, Immunogold labeling of ENO1 in CL1–5 cells. Cells were fixed with 4% formaldehyde and processed for cryosectioning. Sections were blocked with BSA, followed by a labeling for ENO1 with chENO1–22 (40 μg/mL) or NHIgG. After fixation, the samples were stained and examined on a transmission electron microscope. N, nucleus; C, cytoplasm; M, membrane. D, The mRNA expression of ENO1 in NSCLC tumor tissues (T) and adjacent noncancerous tissues (N) was analyzed from the TCGA dataset. E, Representative images show ENO1 IHC staining in samples comprising the full spectrum of lung cancer stages (original magnification, ×40; scale bar, 200 μm).F, IHC staining data are expressed as the percentage of area with positive staining for ENO1. The median for each group is indicated by a horizontal bar. G, OS analysis was performed for lung cancer patients stratified by ENO1 protein expression, based on tissue microarray data. OS was analyzed using the Kaplan–Meier plotter (KM plotter, http://kmplot.com/analysis/; H) and TCGA dataset (RNA-sequencing data of lung adenocarcinoma; I). Patients were split according to the median. The HR with 95% confidence intervals (CI) and log-rank P values from the webpage are provided. **, P < 0.01; ***, P < 0.001.

Figure 1.

ENO1 is highly expressed in metastatic cancer cell lines and is associated with worse OS of patients with lung cancer. A, ENO1 expression was analyzed by Western blotting in NHBE, NL-20, and human lung cancer cell lines using an anti-ENO1 antibody. α-Tubulin served as an internal control. B, The binding of chimeric ENO1 mAb (chENO1) or NHIgG to live NHBE, NL-20, or CL1–5 cells was evaluated by flow cytometry using phycoerythrin (PE)-conjugated goat antimouse IgG. x-axis, intensity of phycoerythrin fluorescence; y-axis, number of events. C, Immunogold labeling of ENO1 in CL1–5 cells. Cells were fixed with 4% formaldehyde and processed for cryosectioning. Sections were blocked with BSA, followed by a labeling for ENO1 with chENO1–22 (40 μg/mL) or NHIgG. After fixation, the samples were stained and examined on a transmission electron microscope. N, nucleus; C, cytoplasm; M, membrane. D, The mRNA expression of ENO1 in NSCLC tumor tissues (T) and adjacent noncancerous tissues (N) was analyzed from the TCGA dataset. E, Representative images show ENO1 IHC staining in samples comprising the full spectrum of lung cancer stages (original magnification, ×40; scale bar, 200 μm).F, IHC staining data are expressed as the percentage of area with positive staining for ENO1. The median for each group is indicated by a horizontal bar. G, OS analysis was performed for lung cancer patients stratified by ENO1 protein expression, based on tissue microarray data. OS was analyzed using the Kaplan–Meier plotter (KM plotter, http://kmplot.com/analysis/; H) and TCGA dataset (RNA-sequencing data of lung adenocarcinoma; I). Patients were split according to the median. The HR with 95% confidence intervals (CI) and log-rank P values from the webpage are provided. **, P < 0.01; ***, P < 0.001.

Close modal

ENO1 enhances lung cancer cell viability and invasion

To investigate the biological function of ENO1 in lung cancer cells in vitro, we established a constitutive ENO1-knockdown system. CL1–5 or CL-141 cells were infected with lentivirus harboring either one of two short hairpin RNAs (shRNA) that target distinct coding sequences of ENO1 (shENO1–1, shENO1–2) or control nontargeting shRNA (shLacZ; Fig. 2A and D). Knockdown of ENO1 significantly decreased cell viability according to WST-1 (Fig. 2B and E) and crystal violet assays (Supplementary Fig. S3A), and there was a significant decrease in ATP levels of ENO1-knockdown cells compared with control CL1–5 cells (Supplementary Fig. S3B). Moreover, knockdown of ENO1 significantly decreased cell invasion (Fig. 2C and F). We also stably overexpressed ENO1 in CL1–0 cells, which normally show low ENO1 expression (Fig. 2G). We found that ENO1 overexpression significantly enhanced cell viability in the WST-1 (Fig. 2H) and crystal violet assays (Supplementary Fig. S3C), but there was no change in ATP levels in ENO1-overexpressing cells compared with control CL1–0 cells (Supplementary Fig. S3D). ENO1 overexpression also increased the number of invading cells (Fig. 2I). In addition, we performed an annexin-V-PI staining assay, confirming that ENO1 expression did not affect cell apoptosis (Supplementary Fig. S3E and S3F). Collectively, these results suggest that the ENO1 increases the viability and invasion of lung cancer cells in vitro.

Figure 2.

Identification of the biological functions of ENO1 in lung cancer cells in vitro and in vivo. Expression of ENO1 in CL1–5 (A) or CL-141 cells (D) infected with lentiviruses harboring control shRNA (shLacZ), or ENO1 shRNA (shENO1–1 and shENO1–2), or ENO1-overexpressing CL1–0 cells (G) was analyzed by Western blotting. α-Tubulin served as an internal control. Cell proliferation (WST-1) assays were performed using ENO1-knockdown CL1–5 (B), or CL-141 cells (E), or ENO1 overexpressing CL1–0 cells (H). The absorbance values are presented as mean ± SEM from four independent experiments. Matrigel invasion assays were performed using ENO1-knockdown CL1–5 (C), or CL-141 cells (F), or ENO1 overexpressing CL1–0 cells (I). The invading cells were counted and the results are presented as mean ± SEM from three independent experiments. CL1–5-Luci cells were infected with lentiviruses harboring control shLacZ or shENO1 (each also expressing luciferase) and transplanted into the lung parenchyma (1 × 105 cells; J and K) or injected into the tail vein (1 × 106 cells) of NOD-SCID mice (L–N). Tumor formation in the lungs was monitored by bioluminescence imaging. N, The number of tumor nodules was counted after hematoxylin and eosin staining of lungs excised from the mice shown in L. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 2.

Identification of the biological functions of ENO1 in lung cancer cells in vitro and in vivo. Expression of ENO1 in CL1–5 (A) or CL-141 cells (D) infected with lentiviruses harboring control shRNA (shLacZ), or ENO1 shRNA (shENO1–1 and shENO1–2), or ENO1-overexpressing CL1–0 cells (G) was analyzed by Western blotting. α-Tubulin served as an internal control. Cell proliferation (WST-1) assays were performed using ENO1-knockdown CL1–5 (B), or CL-141 cells (E), or ENO1 overexpressing CL1–0 cells (H). The absorbance values are presented as mean ± SEM from four independent experiments. Matrigel invasion assays were performed using ENO1-knockdown CL1–5 (C), or CL-141 cells (F), or ENO1 overexpressing CL1–0 cells (I). The invading cells were counted and the results are presented as mean ± SEM from three independent experiments. CL1–5-Luci cells were infected with lentiviruses harboring control shLacZ or shENO1 (each also expressing luciferase) and transplanted into the lung parenchyma (1 × 105 cells; J and K) or injected into the tail vein (1 × 106 cells) of NOD-SCID mice (L–N). Tumor formation in the lungs was monitored by bioluminescence imaging. N, The number of tumor nodules was counted after hematoxylin and eosin staining of lungs excised from the mice shown in L. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Close modal

Next, we wanted to test whether the ENO1 effects on cell proliferation and invasion are due to its metabolic role in the glycolysis pathway and whether ENO1 catalytic activity is essential for its role in HGFR signaling. Therefore, we investigated the effects of ENO1 knockdown and overexpression on glycolysis and central carbon metabolism using metabolism assays. We found that compared with controls, the ENO1-knockdown cells (shENO1) had a higher oxygen consumption rate (OCR), which mainly reflects oxidative phosphorylation (OXPHOS) through the electron transfer chain (Supplementary Fig. S4A). In addition, ENO1-knockdown cells had a lower extracellular acidification rate (ECAR), an indicator of aerobic glycolysis suggestive of lower metabolic activity in ENO1-knockdown cells (Supplementary Fig. S4B). The OCR to ECAR ratio was higher in ENO1-knockdown cells, indicating a higher contribution of OXPHOS to overall energy production (Supplementary Fig. S4C). Notably, ENO1-overexpressing CL1–0 cells showed opposite results (Supplementary Fig. S4D-S4F). Furthermore, knockdown-resistant lines with two catalytically dead mutants of ENO1-S40A and ENO1-D245R were established (43). Our results showed that ENO1-S40A and ENO1-D245R mutants both induced cell proliferation (Supplementary Fig. S4G) and cell invasion (Supplementary Fig. S4H), similar to wild-type ENO1 (ENO1-WT). Thus, ENO1 catalytic activity is not essential for its roles in enhancing proliferation and invasion.

ENO1 promotes lung cancer tumorigenicity and metastasis in xenograft models

Tumor growth, invasion, and metastasis are key steps in tumor progression. First to explore the effect of ENO1 on tumorigenicity in vivo, we established an orthotopic xenograft model. One hundred thousand CL1–5 cells stably expressing luciferase and infected with either shENO1 or shLacZ were orthotopically injected into the lung parenchyma of NOD-SCID mice. Using luciferin-based bioluminescence to detect tumor size, we found that mice injected with ENO1-knockdown cells had lower luciferase signal than those injected with shLacZ-infected cells, indicating a lower tumor growth rate for ENO1-knockdown tumors (Fig. 2J and K). The shLacZ and shENO1 cells had similar levels of luminescence according to luciferase activity assays (Supplementary Fig. S5A). Lung tissues were collected 21 days after implantation, and hematoxylin and eosin staining was used to detect lung nodules (Supplementary Fig. S5B).

To further test whether ENO1 may affect lung cancer metastasis, we investigated lung nodule formation in NOD-SCID mice intravenously injected with CL1–5 cells stably expressing luciferase and infected with shENO1 or shLacZ. Mouse lungs were collected 31 days after injection, and lung nodules were counted. As shown in Fig. 2L and M, there were stronger luciferase signals in lungs of mice injected with control shLacZ cells compared with the ENO1-knockdown group. In addition, ENO1-knockdown cells created fewer metastatic lung nodules than control cells (Fig. 2N; Supplementary Fig. S5C). Taken together, these data demonstrate that ENO1 plays a critical role in promoting tumor formation and metastatic behavior of lung cancer cells in xenograft models.

ENO1 enhances cancer cell invasion via HGFR- and WNT-driven EMT

Previous studies have shown that EMT is associated with cancer invasion, metastasis, and progression. Because ENO1-knockdown CL1–5 cells exhibited a more epithelial-like phenotype compared with control cells, which displayed fibroblast-like mesenchymal features (Fig. 3A), we decided to profile the expression EMT-related genes in ENO1-knockdown and control cells. We found that ENO1 knockdown inhibited the expression of the mesenchymal markers, N-cadherin, vimentin, as well as the protein level of the EMT regulator, SLUG, while enhancing E-cadherin expression in CL1–5 or CL141 cells (Fig. 3B; Supplementary Fig. S6A and S6B). ENO1-mediated upregulation of SLUG expression was also observed in CL1–5 cells by immunofluorescence (IF; Supplementary Fig. S6C and S6D). In contrast, ectopic expression of ENO1 in CL1–0 cells increased the levels of N-cadherin, vimentin, and SLUG, while decreasing E-cadherin (Fig. 3B). Interestingly, ENO1 did not affect SLUG mRNA level (Fig. 3C). Treatment with MG132 (inhibitor of the 26S proteasome) increased SLUG steady-state protein levels, indicating that the protein level is largely controlled by degradation (Fig. 3D; Supplementary Fig. S6E). Knockdown of ENO1 led to increased levels of ubiquitylated SLUG compared with control cells (Fig. 3E), and it shortened the SLUG protein half-life, according to the protein levels after cyclohexamide treatment (Fig. 3F and G). Based on these results, we concluded that ENO1 regulates SLUG protein stability but not its expression.

Figure 3.

ENO1 promotes EMT by regulating SLUG stability. A, The morphology of CL1–5 cells infected with lentiviruses harboring control shLacZ or shENO1. Scale bar, 50 μm. B, The expression of EMT markers and regulators was detected by Western blotting using CL1–5 cells infected with shLacZ or shENO1 (left) or CL1–0 cells infected with vector or ENO1-FLAG (right). C, The expression of EMT markers and regulators was detected by qRT-PCR using CL1–5 cells infected with shLacZ or shENO1. D, The protein expression of these genes was analyzed in ENO1-knockdown cells (shENO1–1) by treating with or without 10 mmol/L MG132 (proteasome inhibitor) for 6 hours, followed by Western blotting. E, CL1–5 cells were infected with lentiviruses harboring shLacZ or shENO1. Twenty-four hours after transfection, the cells were treated with 10 mmol/L MG132 for 6 hours before cell collection. The lysates were subjected to IP using anti-SLUG antibody (left) and input (right). Western blotting was performed with the indicated antibodies to detect protein interactions. F, Stability of SLUG protein in CL1–5 cells infected with shLacZ or shENO1. Cells were treated with 100 μg cycloheximide (CHX) at the indicated intervals and subjected to Western blotting analysis. G, Quantification of SLUG half-life in indicated groups. *, P < 0.05; **, P < 0.01; ***, P < 0.001; NS, nonsignificant.

Figure 3.

ENO1 promotes EMT by regulating SLUG stability. A, The morphology of CL1–5 cells infected with lentiviruses harboring control shLacZ or shENO1. Scale bar, 50 μm. B, The expression of EMT markers and regulators was detected by Western blotting using CL1–5 cells infected with shLacZ or shENO1 (left) or CL1–0 cells infected with vector or ENO1-FLAG (right). C, The expression of EMT markers and regulators was detected by qRT-PCR using CL1–5 cells infected with shLacZ or shENO1. D, The protein expression of these genes was analyzed in ENO1-knockdown cells (shENO1–1) by treating with or without 10 mmol/L MG132 (proteasome inhibitor) for 6 hours, followed by Western blotting. E, CL1–5 cells were infected with lentiviruses harboring shLacZ or shENO1. Twenty-four hours after transfection, the cells were treated with 10 mmol/L MG132 for 6 hours before cell collection. The lysates were subjected to IP using anti-SLUG antibody (left) and input (right). Western blotting was performed with the indicated antibodies to detect protein interactions. F, Stability of SLUG protein in CL1–5 cells infected with shLacZ or shENO1. Cells were treated with 100 μg cycloheximide (CHX) at the indicated intervals and subjected to Western blotting analysis. G, Quantification of SLUG half-life in indicated groups. *, P < 0.05; **, P < 0.01; ***, P < 0.001; NS, nonsignificant.

Close modal

Since the Wnt signaling antagonist GSK3β is known to affect SLUG protein turnover via phosphorylation and ubiquitin-mediated proteolysis (44), we hypothesized that ENO1 stabilized SLUG by regulating GSK3β. Indeed, SLUG was upregulated in ENO1-knockdown cells after treatment with a GSK3β inhibitor; meanwhile, the active form of β-catenin was also increased (Fig. 4A; Supplementary Fig. S6F). Interestingly, we found that knockdown of ENO1 enhanced GSK3β (Fig. 4A), and it also decreased Wnt signaling activity, as measured by the TOP/FOP luciferase reporter assays (Fig. 4B). Confocal IF revealed that ENO1 knockdown also reduced nuclear β-catenin levels (Fig. 4C). To further test whether β-catenin and SLUG participate in ENO1-induced cell invasion, we overexpressed a constitutively active β-catenin (β-catenin△45; ref. 45) or vector control and found that overexpression of β-catenin△45 restored invasion capacity of ENO1 knockdown cells, but did not affect invasion of control cells (Fig. 4D). In addition, we also overexpressed mutant SLUG (Slug-4SA; SLUG protein with four GSK3β-targeted serine residues replaced by alanines) or wild-type Slug (Slug-WT; ref. 44) and found that overexpression of mutant Slug restored invasion capacity and cell proliferation in ENO1 knockdown cells, but Slug-WT did not (Supplementary Fig. S7A–S7C). Next, we analyzed SLUG protein levels in a panel of lung cancer cells, finding that SLUG protein level was elevated in highly metastatic lung cancer cell lines (CL1–5 and CL141) compared with nonmetastatic cell lines (CL1–0 and Hop62) and NHBE cells (Supplementary Fig. S7D). The expression panel was reminiscent of ENO1 expression in cell lines (Fig. 1A). Patient samples were stratified in low and high ENO1 and SLUG groups, and survival of each group was analyzed. The analysis showed that patients with high ENO1 and high SLUG expression had the worst OS (Fig. 4E). Together, these data support the idea that ENO1 promotes malignant cell invasion via the GSK3β-SLUG signaling axis.

Figure 4.

ENO1 stabilizes SLUG by decreasing GSK3β activity via activation of HGFR-Src-PI3K-AKT and Wnt-β-catenin signaling. A, Protein expression was analyzed by Western blotting in ENO1-knockdown cells (shENO1–1) after treatment with or without 2 μmol/L GSK3β inhibitor (BIO) for 24 hours. B, One set of cells was transfected with the TOP reporter, CMV: Renilla luciferase; a second set of cells was transfected with the FOP reporter, CMV: Renilla luciferase as indicated. The TOP and FOP reporters are as follows: β-catenin–responsive “T-cell factor (TCF) element driving luciferase” (TOP) and non-β-catenin–responsive mutated element also driving luciferase (FOP). The ratio of signals from these two reporters, each normalized to Renilla luciferase as internal control, reflects β-catenin–specific canonical Wnt signaling. After 72 hours, Wnt signaling was measured in ENO1-knockdown CL1–5 cells or control cells. C, Confocal IF assays were used to detect β-catenin expression in CL1–5 cells. A 63× objective was used to observe the samples. Green fluorescence, β-catenin; blue fluorescence, DAPI stain. Bottom, quantification of the mean intensity for nuclear SLUG. D, Matrigel invasion assays were performed on CL1–5 cells expressing indicated plasmid combinations. E, TCGA data analysis (RNA-sequencing data of lung adenocarcinoma) of OS for NSCLC patients with ENO1 low/SLUG low, ENO1 high/SLUG low, ENO1 low/SLUG high, and ENO1 high/SLUG high. Patients were split according to the median. F, Protein levels were detected by Western blotting in CL1–5 cells infected with shLacZ or ENO1 shRNA or ENO1 shRNA cells were treated with recombinant ENO1 for 30 minutes (1 μg/mL). P, phosphorylated protein levels; t, total protein levels. G, Matrigel invasion assays were performed using CL1–5 cells infected with shLacZ or ENO1 shRNA or ENO1 shRNA cells were treated with recombinant ENO1 for 30 minutes (1 or 2 μg/mL). The invading cells were counted and the results are presented as mean ± SEM from three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001; NS, nonsignificant.

Figure 4.

ENO1 stabilizes SLUG by decreasing GSK3β activity via activation of HGFR-Src-PI3K-AKT and Wnt-β-catenin signaling. A, Protein expression was analyzed by Western blotting in ENO1-knockdown cells (shENO1–1) after treatment with or without 2 μmol/L GSK3β inhibitor (BIO) for 24 hours. B, One set of cells was transfected with the TOP reporter, CMV: Renilla luciferase; a second set of cells was transfected with the FOP reporter, CMV: Renilla luciferase as indicated. The TOP and FOP reporters are as follows: β-catenin–responsive “T-cell factor (TCF) element driving luciferase” (TOP) and non-β-catenin–responsive mutated element also driving luciferase (FOP). The ratio of signals from these two reporters, each normalized to Renilla luciferase as internal control, reflects β-catenin–specific canonical Wnt signaling. After 72 hours, Wnt signaling was measured in ENO1-knockdown CL1–5 cells or control cells. C, Confocal IF assays were used to detect β-catenin expression in CL1–5 cells. A 63× objective was used to observe the samples. Green fluorescence, β-catenin; blue fluorescence, DAPI stain. Bottom, quantification of the mean intensity for nuclear SLUG. D, Matrigel invasion assays were performed on CL1–5 cells expressing indicated plasmid combinations. E, TCGA data analysis (RNA-sequencing data of lung adenocarcinoma) of OS for NSCLC patients with ENO1 low/SLUG low, ENO1 high/SLUG low, ENO1 low/SLUG high, and ENO1 high/SLUG high. Patients were split according to the median. F, Protein levels were detected by Western blotting in CL1–5 cells infected with shLacZ or ENO1 shRNA or ENO1 shRNA cells were treated with recombinant ENO1 for 30 minutes (1 μg/mL). P, phosphorylated protein levels; t, total protein levels. G, Matrigel invasion assays were performed using CL1–5 cells infected with shLacZ or ENO1 shRNA or ENO1 shRNA cells were treated with recombinant ENO1 for 30 minutes (1 or 2 μg/mL). The invading cells were counted and the results are presented as mean ± SEM from three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001; NS, nonsignificant.

Close modal

Because the phenotypes of ENO1-expressing cells are often caused by receptor tyrosine kinase (RTK) activation, we further examined whether any RTKs might be involved in ENO1 regulation of the GSK3β-SLUG axis. Human Phospho-RTK arrays were treated with cell lysates from shLacZ- and shENO1-infected CL1–5 cells, revealing an extreme difference in the level of HGFR phosphorylation between the groups. Specifically, HGFR phosphorylation was reduced in ENO1-knockdown cells (Supplementary Fig. S8A and 8B). This result was confirmed by Western blotting (knockdown of ENO1 decreased the level of phosphorylated HGFR), which was also used to probe the known downstream signals. HGFR phosphorylation was coincident with activation of the Src-PI3K-AKT pathway and inactivation of the GSK3β destruction complex, as well as stabilized SLUG and β-catenin in CL1–5 and CL141 cells (Fig. 4F; Supplementary Fig. S8C). Furthermore, purified recombinant ENO1 could rescue shENO1 knockdown cell phenotypes, such as HGFR and WNT downstream signaling (Fig. 4F), and also rescue shENO1 knockdown-suppressive cell invasive ability (Fig. 4G). HGF is known to be an important transactivator of canonical Wnt signaling, an effect that is mediated by HGFR-stimulated GSK3β-dependent LRP5/6 phosphorylation (46), and our data showed that ENO1 could enhance Wnt signaling (Fig. 4B–D and F). Therefore, we tested whether ENO1 activates Wnt signaling via LRP5/6 phosphorylation. Indeed, knockdown of ENO1 decreased the phosphorylation of LRP5/6, but it did not affect the LRP5/6 protein expression (Fig. 4F). We also found that cells expressing ENO1 catalytically dead mutants showed levels of HGFR, GSK3β phosphorylation, and SLUG similar to those expressing ENO1-WT (Supplementary Fig. S9), indicating that ENO1 catalytic activity is not essential for its role in HGFR signaling.

Surprisingly, we also found that ENO1 is essential for HGFR activation since knockdown of ENO1 prevented HGF activation of HGFR and its downstream signaling (Fig. 5A). Thus, we sought to define the mechanism through which ENO1 induces HGFR phosphorylation. We performed GST pull-down assays, in which GST-ENO1 fusion protein or GST alone were incubated with purified His-tagged HGFR, to check for a direct interaction between purified ENO1 and HGFR protein. The pull-down results showed that under our experimental conditions, His-tagged HGFR bound to GST-ENO1 but not GST alone (Fig. 5B). Furthermore, we found that HGF also bound to ENO1 in pull-down assays (Supplementary Fig. S10). Then we used non-cell permeable crosslinking agents, DTSSP, to further confirm ENO1 bound to HGFR on the cell surface by membrane protein extraction and IP-Western blotting (Fig. 5C). Furthermore, we performed co-IP of exogenous (Fig. 5D) and endogenous ENO1 with HGFR to validate the interaction between ENO1 and HGFR (Fig. 5E). To further verify the interactions, we also generated several Myc-tagged HGFR truncated mutants, as illustrated in the upper scheme of Fig. 5F. Co-IP assays showed that SEMA domain corresponding to amino acids 52–496 of HGFR was required for the interaction with ENO1 (Fig. 5F). The interaction between ENO1 and HGFR was evident, but there was no detectable interaction between ENO1 and LRP5/6 (Supplementary Fig. S11A). Therefore, we propose ENO1 may activate HGFR by a direct protein–protein interaction, and the ENO1-induced phosphorylation of LRP5/6 may be dependent on HGFR activation.

Figure 5.

ENO1 associates with HGFR and is essential for HGF-HGFR activation. A, CL1–5 cells infected with shLacZ or ENO1 shRNA were stimulated with vehicle or human HGF for 10 minutes, followed by Western blotting. B, GST pull-down assay. GST-agarose beads bound GST-ENO1 or GST proteins were incubated with purified His-tagged HGFR. After washing, the bead-bound proteins and 5% input His-tagged HGFR protein were released by boiling with SDS-PAGE loading buffer. Samples were then subjected to Western blotting with anti-His or ENO1 antibodies. Anti-GST antibody was used as isotype control. C, Membrane proteins were extracted and IP of affinity cross-linked ENO1 bound to HGFR from CL1–5 cells. D, Immunoprecipitates were prepared from CL1–5 cells after infection with FLAG-tagged ENO1. Cell lysates were immunoprecipitated with anti-ENO1 (IP: ENO1) or anti-HGFR (IP: HGFR) antibodies and blotted with anti-HGFR and anti-ENO1 antibodies. Normal rabbit IgG (NRIgG) antibody was used as isotype control. E, Immunoprecipitates were prepared from CL1–5 cells. Cell lysates were immunoprecipitated with anti-ENO1 (IP: ENO1) or anti-HGFR (IP: HGFR) antibodies and blotted with anti-HGFR and anti-ENO1 antibodies. Normal rabbit IgG (NRIgG) antibody was used as isotype control. F, Top, schematic diagram of mapping the HGFR structural domains. Bottom, lysates from CL1–5 cells cotransfected with FLAG-tagged ENO1 and different Myc-tagged-HGFR constructs or an empty vector for 72 hours were immunoprecipitated with anti-Myc antibodies and fractionated by SDS–PAGE. Immunoblots were probed with anti-FLAG or anti-Myc antibodies. G, CL1–5 cells infected with shLacZ or ENO1 shRNA were treated with HGFR inhibitor for 24 hours and stimulated with hHGF for 10 minutes, followed by Western blotting. P, phosphorylated protein levels; t, total protein levels.

Figure 5.

ENO1 associates with HGFR and is essential for HGF-HGFR activation. A, CL1–5 cells infected with shLacZ or ENO1 shRNA were stimulated with vehicle or human HGF for 10 minutes, followed by Western blotting. B, GST pull-down assay. GST-agarose beads bound GST-ENO1 or GST proteins were incubated with purified His-tagged HGFR. After washing, the bead-bound proteins and 5% input His-tagged HGFR protein were released by boiling with SDS-PAGE loading buffer. Samples were then subjected to Western blotting with anti-His or ENO1 antibodies. Anti-GST antibody was used as isotype control. C, Membrane proteins were extracted and IP of affinity cross-linked ENO1 bound to HGFR from CL1–5 cells. D, Immunoprecipitates were prepared from CL1–5 cells after infection with FLAG-tagged ENO1. Cell lysates were immunoprecipitated with anti-ENO1 (IP: ENO1) or anti-HGFR (IP: HGFR) antibodies and blotted with anti-HGFR and anti-ENO1 antibodies. Normal rabbit IgG (NRIgG) antibody was used as isotype control. E, Immunoprecipitates were prepared from CL1–5 cells. Cell lysates were immunoprecipitated with anti-ENO1 (IP: ENO1) or anti-HGFR (IP: HGFR) antibodies and blotted with anti-HGFR and anti-ENO1 antibodies. Normal rabbit IgG (NRIgG) antibody was used as isotype control. F, Top, schematic diagram of mapping the HGFR structural domains. Bottom, lysates from CL1–5 cells cotransfected with FLAG-tagged ENO1 and different Myc-tagged-HGFR constructs or an empty vector for 72 hours were immunoprecipitated with anti-Myc antibodies and fractionated by SDS–PAGE. Immunoblots were probed with anti-FLAG or anti-Myc antibodies. G, CL1–5 cells infected with shLacZ or ENO1 shRNA were treated with HGFR inhibitor for 24 hours and stimulated with hHGF for 10 minutes, followed by Western blotting. P, phosphorylated protein levels; t, total protein levels.

Close modal

We also found that ENO1-induced HGFR activation is essential for LRP5/6-mediated stabilization of SLUG and β-catenin; inhibition of HGFR activity in ENO1-expressing cells decreased LRP5/6 phosphorylation (Fig. 5G), and knockdown of both LRP5 and LRP6 decreased ENO1-induced SLUG and β-catenin stabilization (Supplementary Fig. S11B). In total, these data show that ENO1 enhances cancer cell invasion by regulating HGFR- and WNT-driven EMT.

chENO1–22 antibody decreases cancer cell proliferation and invasion in vitro by inhibiting ENO1–HGFR axis-mediated downstream signaling to promote SLUG protein degradation

Although many reports have shown that expression of ENO1 significantly correlates with reduced survival and poor prognosis in many cancer types, such as lung cancer (12), there is currently no therapeutic antibody against ENO1 approved for clinical use. Hence, we generated anti-ENO1 mAbs that may be useful for cancer research and therapy. We identified four candidates of anti-ENO1 mAbs through mouse hybridoma screening. Our results showed that all four anti-ENO1 mAbs possess high binding activity to CL1–5 cells by cellular ELISA assay (Supplementary Fig. S1A). Importantly, flow cytometry experiments indicated that anti-ENO1 mAb-22 (ENO1–22) can reliably detect ENO1 on the surface of CL1–5 cells (Supplementary Fig. S1B). Therefore, we cloned the variable domain of anti-ENO1 mAb-22 antibody into a human IgG Fc vector to generate chimeric ENO1–22 (chENO1–22). We validated the binding activity of chENO1–22, finding that the antibody retained high affinity to live CL1–5 cells (Supplementary Fig. S1C); the binding specificity of ENO1 was then validated with ENO1-knockdown CL1–5 cells (Fig. 6A) or ENO1-overexpressing CL1–0 cells (Fig. 6B).

Figure 6.

chENO1–22 antibody decreases cancer cell proliferation and invasion in vitro by inhibiting ENO1-HGFR axis-mediated downstream signaling to promote SLUG protein degradation. A and B, Quantitation of the binding of anti-chENO1–22 to live CL1–5 cells infected with lentiviruses harboring LacZ or ENO1 shRNA (A) or CL1–0 cells transfected with AS2-ENO1-FLAG or the empty vector (B) was evaluated by flow cytometry using phycoerythrin-conjugated goat anti-human IgG. C, Cell proliferation (WST-1) assays were performed using CL1–5 cells treated with chENO1–22. The concentration of antibodies was 40 μg/mL, and the OD values were measured 8, 24, 48, or 72 hours after antibody treatment. D, Matrigel invasion assays were performed using CL1–5 cells treated with chENO1–22 or mENO1–37. The invading cells were quantified, and the results are presented as the mean ± SEM from three independent experiments. E, Protein levels were detected by Western blotting of CL1–5 cells treated with NHIgG or chENO1–22. F, CL1–5 cells were treated with NHIgG or chENO1–22 for 4 hours and then stimulated with vehicle or hHGF for 10 minutes, followed by Western blotting. G, CL1–5 cells were treated with NHIgG or chENO1–22 for 4 hours and then immunoprecipitated with anti-ENO1 (IP: ENO1) or anti-HGFR (IP: HGFR) antibodies, followed by Western blotting. H, CL1–5 cells were treated with 10 mmol/L MG132 and chENO1–22 for 6 hours before cell collection and subsequent Western blotting. I, Stability of Slug protein in CL1–5 cells treated with NHIgG or chENO1–22. Cells were treated with 100 μg cycloheximide (CHX) at the indicated intervals and subjected to Western blotting analysis. Bottom graph shows quantification of SLUG half-life in indicated groups. J, Matrigel invasion assays were performed using CL1–5 cells expressing indicated plasmids, with or without chENO1–22 treatment. The invading cells were quantified, and the results are presented as the mean ± SEM from three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001; NS, nonsignificant.

Figure 6.

chENO1–22 antibody decreases cancer cell proliferation and invasion in vitro by inhibiting ENO1-HGFR axis-mediated downstream signaling to promote SLUG protein degradation. A and B, Quantitation of the binding of anti-chENO1–22 to live CL1–5 cells infected with lentiviruses harboring LacZ or ENO1 shRNA (A) or CL1–0 cells transfected with AS2-ENO1-FLAG or the empty vector (B) was evaluated by flow cytometry using phycoerythrin-conjugated goat anti-human IgG. C, Cell proliferation (WST-1) assays were performed using CL1–5 cells treated with chENO1–22. The concentration of antibodies was 40 μg/mL, and the OD values were measured 8, 24, 48, or 72 hours after antibody treatment. D, Matrigel invasion assays were performed using CL1–5 cells treated with chENO1–22 or mENO1–37. The invading cells were quantified, and the results are presented as the mean ± SEM from three independent experiments. E, Protein levels were detected by Western blotting of CL1–5 cells treated with NHIgG or chENO1–22. F, CL1–5 cells were treated with NHIgG or chENO1–22 for 4 hours and then stimulated with vehicle or hHGF for 10 minutes, followed by Western blotting. G, CL1–5 cells were treated with NHIgG or chENO1–22 for 4 hours and then immunoprecipitated with anti-ENO1 (IP: ENO1) or anti-HGFR (IP: HGFR) antibodies, followed by Western blotting. H, CL1–5 cells were treated with 10 mmol/L MG132 and chENO1–22 for 6 hours before cell collection and subsequent Western blotting. I, Stability of Slug protein in CL1–5 cells treated with NHIgG or chENO1–22. Cells were treated with 100 μg cycloheximide (CHX) at the indicated intervals and subjected to Western blotting analysis. Bottom graph shows quantification of SLUG half-life in indicated groups. J, Matrigel invasion assays were performed using CL1–5 cells expressing indicated plasmids, with or without chENO1–22 treatment. The invading cells were quantified, and the results are presented as the mean ± SEM from three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001; NS, nonsignificant.

Close modal

Next, we evaluated the anticancer potential of chENO1–22, including antiproliferation and anti-invasion effects. chENO1–22 inhibited tumor cell growth (Fig. 6C) and cancer cell invasion in vitro (Fig. 6D), but mENO1–37, which couldn't detect surface form ENO1, could not inhibit cancer cell invasion (Fig. 6D). We next sought to test whether chENO1–22 anticarcinogenic effects are due to its inhibition of ENO1-downstream signaling pathways. Indeed, chENO1-22 treatment inhibited downstream signals of ENO1-HGFR, including the levels of SLUG and active β-catenin (Fig. 6E). Moreover, chENO1-22 treatment attenuated HGF-triggered phosphorylation of HGFR and its downstream signaling components (Fig. 6F). These results were similar to shENO1 knockdown. The mAb also decreased the association between ENO1 and HGFR, as detected by co-IP of endogenous proteins (Fig. 6G). SLUG steady-state protein levels were reduced by treatment with chENO1–22, and treatment with MG132 increased SLUG steady-state protein levels (Fig. 6H). In addition, chENO1–22 shortened the SLUG protein half-life, as assessed with cycloheximide treatment (Fig. 6I). Moreover, we found that overexpression of mutant Slug-4SA restored invasion capacity upon chENO1–22 treatment, but Slug-WT did not (Fig. 6J). These data show that chENO1–22 inhibition of lung cancer invasion likely occurs as a result of suppressed ENO1-HGFR axis signaling, which allows rapid SLUG protein degradation.

chENO1–22 antibody decreases lung cancer cell metastasis in vivo

We then wanted to further analyze the therapeutic effects on tumor metastasis in vivo, so we injected 1 × 106 CL1–5 luciferase cells into NOD-SCID mice via the tail vein to establish a metastatic xenograft model; the injected mice were subsequently treated with control normal human IgG (NHIgG) or chENO1–22 (10 or 20 mg/kg). After 4 weeks of mAb treatment, weaker luciferase signals were seen in the lungs of mice receiving chENO1–22 compared with NHIgG, no matter whether the dose was 10 or 20 mg/kg (Fig. 7A and B). In addition, chENO1–22-treated mice maintained a normal body weight (Supplementary Fig. S12) and showed improved survival compared with NHIgG-treated mice (Fig. 7C). To further understand whether the anti-ENO1 antibody actually blocks lung colonization, we started the antibody treatment before injecting cancer cells. One day after beginning the treatment, 2 × 106 CL1–5-Luci cells were intravenously injected into NOD-SCID mice. Weaker luciferase signals were seen in the lungs of mice receiving chENO1-22 compared with those receiving NHIgG after 25 days of mAb treatment (Fig. 7D and E), and survival was also improved compared with NHIgG-treated mice (Fig. 7F). Thus, chENO1-22 treatment could indeed inhibit lung tumor colonization and metastasis.

Figure 7.

chENO1–22 antibody decreases lung cancer cell metastasis in vivo. A, 1 × 106 CL1–5-Luci cells were intravenously injected into NOD-SCID mice and the normal human IgG control or chENO1–22 antibodies were intravenously injected two times a week for a total of eight injections. chENO1–22 antibodies were injected beginning 3 days after cancer cell injection. Weekly bioluminescence imaging was performed to monitor the growth of the intravenously-injected tumor cells (median survival day: log–rank, P = 0.0019). B, Tumor formation in the lungs was monitored by bioluminescence imaging. C, Kaplan–Meier curves were generated to analyze survival. D, One day after beginning the treatment (NHIgG control or chENO1–22 antibodies), 2 × 106 CL1–5-Luci cells were intravenously injected into NOD-SCID mice. E, Bioluminescence imaging was performed to monitor the growth of the injected tumor cells and tumor formation in the lungs. F, Kaplan–Meier curves were generated to analyze survival. G, A working model depicts ENO1 promotion of cell invasion through regulation of the HGFR-Src-PI3K-AKT-GSK3β-SLUG and HGFR-WNT-GSK3β-SLUG axes. (i) ENO1 increases the phosphorylation of HGFR and Wnt coreceptor LPR5/6 and decreases GSK3β activity to enhance SLUG stabilization and suppress E-cadherin expression. Ultimately, cancer cell invasion capacity is enhanced; (ii) the chENO1–22 mAb inhibits cancer cell invasion by blocking ENO1–HGFR axis–mediated downstream signaling to promote SLUG protein ubiquitination and degradation. **, P < 0.01; ***, P < 0.001.

Figure 7.

chENO1–22 antibody decreases lung cancer cell metastasis in vivo. A, 1 × 106 CL1–5-Luci cells were intravenously injected into NOD-SCID mice and the normal human IgG control or chENO1–22 antibodies were intravenously injected two times a week for a total of eight injections. chENO1–22 antibodies were injected beginning 3 days after cancer cell injection. Weekly bioluminescence imaging was performed to monitor the growth of the intravenously-injected tumor cells (median survival day: log–rank, P = 0.0019). B, Tumor formation in the lungs was monitored by bioluminescence imaging. C, Kaplan–Meier curves were generated to analyze survival. D, One day after beginning the treatment (NHIgG control or chENO1–22 antibodies), 2 × 106 CL1–5-Luci cells were intravenously injected into NOD-SCID mice. E, Bioluminescence imaging was performed to monitor the growth of the injected tumor cells and tumor formation in the lungs. F, Kaplan–Meier curves were generated to analyze survival. G, A working model depicts ENO1 promotion of cell invasion through regulation of the HGFR-Src-PI3K-AKT-GSK3β-SLUG and HGFR-WNT-GSK3β-SLUG axes. (i) ENO1 increases the phosphorylation of HGFR and Wnt coreceptor LPR5/6 and decreases GSK3β activity to enhance SLUG stabilization and suppress E-cadherin expression. Ultimately, cancer cell invasion capacity is enhanced; (ii) the chENO1–22 mAb inhibits cancer cell invasion by blocking ENO1–HGFR axis–mediated downstream signaling to promote SLUG protein ubiquitination and degradation. **, P < 0.01; ***, P < 0.001.

Close modal

On the basis of these findings, we present a proposed model of how ENO1 promotes cell invasion through HGFR–Src–PI3K–AKT–GSK3β–SLUG and WNT–GSK3β–SLUG signaling axes in Fig. 7G.

Our results showed that the ENO1 protein was highly expressed in NSCLC cells but weakly expressed in normal lung tissue. In addition, high ENO1 expression was positively associated with TMN stage and poor outcomes of patients with lung cancer. These findings suggest that ENO1 is a potentially valuable diagnostic biomarker and/or therapeutic target for lung cancer. The invasion-metastasis cascade can be simplified as two key steps: (i) cancer cell dissemination from the primary tumor to distant tissues, and (ii) colonization of distant organs by invading cancer cells, which form micrometastases (22). Expression of ENO1-targeting shRNA suppressed lung cancer cell invasion capacity in vitro and disrupted the formation of micrometastases and the metastatic colonization of tumor cells in vivo. Thus, ENO1 may play an important role in lung cancer metastasis. Moreover, ENO1-targeting shRNA suppressed lung cancer growth in an orthotopic tumor model, suggesting that ENO1 may also participate in lung cancer progression. Together, these findings confirm that ENO1 may serve as a therapeutic target for lung cancer.

In this study, we uncovered a few novel aspects of how ENO1 participates in carcinogenesis. Although several studies previously showed that ENO1 can modulate intravascular and pericellular fibrinolytic activity to facilitate cell migration and cancer metastasis (6–8), the detailed signaling mechanism was unclear. Here, we show that a direct protein–protein interaction between membrane form ENO1 and HGFR stimulates HGFR signaling. In addition, HGFR was previously known to induce GSK3-dependent LRP5/6 phosphorylation (46), and we further find that ENO1 triggers HGFR activity to activate LRP5/6-GSK3β-β-catenin canonical Wnt pathway via increased phosphorylation of LRP5/6, but does not detect direct association of ENO1 and LRP5/6. Moreover, we discover the interaction between ENO1 and HGFR triggered an HGFR-Src-PI3K-AKT-GSK3β signaling axis and ENO1 is essential for HGF-HGFR activation since knockdown of ENO1 prevented HGF activation of HGFR and its downstream signaling. Furthermore, inhibition of HGFR activity and knockdown of ENO1 decrease LRP5/6 phosphorylation, suggesting that ENO1-triggered HGF activation induces transphosphorylation of LRP5/6 in a Wnt-independent manner. Intriguingly, these two pathways (Wnt and AKT) coordinately stabilize β-catenin and SLUG to promote EMT. These new findings signify ENO1 plays a vital role in carcinogenesis through regulating least two typical oncogenic signaling pathways in direct (HGFR-AKT) or indirect (Wnt-LRP5/6) manner.

An altered cadherin profile is a well-established hallmark of EMT, with loss of E-cadherin representing a key step in the acquisition of invasiveness, while an increase in N-cadherin signifies a proinvasive phenotype; this sort of “cadherin switching” has been associated with a poor clinical prognosis in many cancers (47). In the current study, we showed that knockdown of ENO1 suppressed mesenchymal transformation by reducing mesenchymal genes (N-cadherin, Vimentin, and SLUG) and upregulating the epithelial gene, E-cadherin. These results reveal a previously undescribed molecular mechanism by which ENO1 participates in lung cancer progression. Slug, an EMT-ATF, had been shown to transcriptionally suppress the expression of E-cadherin (22) and promote cancer invasion and metastasis in various types of cancers (48, 49). Previous studies showed that the SLUG-E-cadherin axis was associated with cancer metastasis and poor clinical outcomes in multiple types of NSCLC (48, 50), suggesting that SLUG is critically involved in lung cancer progression. Indeed, our study indicates that knockdown of SLUG decreases lung cancer cell invasion, and it is involved in ENO1-promoted cell invasion (Supplementary Fig. S7A and S7B). Moreover, we found that patients with NSCLC in the ENO1 high and SLUG high group had the worst OS (Fig. 4E), demonstrating the vital combined roles of ENO1 and SLUG in determining NSCLC outcome.

ENO1 has been identified as a plasminogen receptor on the surfaces of several diverse cell types including carcinoma cells (3, 51). We also demonstrated that ENO1 was localized to the cell membrane of lung cancer cells, which was confirmed by flow cytometry and immunogold assays. This observation suggested ENO1 may be a targetable therapeutic target in lung cancer. Therefore, we generated a therapeutic mAb against ENO1, chENO1–22, which could be used to attenuate tumor growth and block tumor metastasis in a lung cancer xenograft model. A potential mechanism of action for chENO1–22 is the inhibition of ENO1-HGFR signaling, which allows rapid SLUG protein ubiquitination and degradation through blocking the association of membrane form ENO1 and HGFR. However, the benefits of antibody therapy are often complex and usually involve multiple molecular and physiologic mechanisms. For example, the Fc function of an antibody is important for tumor-cell killing via antibody-dependent cell-mediated cytotoxicity (ADCC) or complement-dependent cytotoxicity (CDC) in vivo, and effective tumor ADCC requires congenic effector cells, target cells, and mAbs; specific antibody affinity and subtypes may also be required (52). These potential physiologic functions of chENO1–22 will be assessed in future studies.

According to our IHC results, ENO1 protein is highly expressed in NSCLC cells, and even high ENO1 expression is positively correlated with TNM staging. But it is weakly expressed in normal lung tissues. chENO1–22 can recognize membrane form of ENO1, which might be used in cancer detection or drug delivery, such as antibody-drug conjugates (ADC). The clinical success of ADCs has provided a promising strategy for the development of highly effective anticancer drugs. ADC consists of effective cytotoxic drugs, which are linked to antibodies through chemical linkers. The approval of Adcetris (brentuximab vedotin, SGN-35), Kadcyla (ado-trastuzumab emtansine, T-DM1), Mylotarg (gemtuzumab ozogamicin), and Polivy (polatuzumab vedotin) by the FDA in 2011, 2013, 2017, and 2019 (52, 53), respectively, aroused new interest in ADC technology. Because chENO1–22 reduced tumor progression and metastatic colonization in this study, it is highly likely that this mAb may be translatable to clinical application.

H.J. Li reports a patent for US 63/164,137 pending. H.C. Wu reports a patent for US 63/164,137 pending. No disclosures were reported by the other authors.

H.J. Li: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. F.Y. Ke: Methodology. C.C. Lin: Methodology. M.Y. Lu: Methodology. Y.H. Kuo: Methodology. Y.P. Wang: Formal analysis. K.H. Liang: Methodology. S.C. Lin: Methodology. Y.H. Chang: Formal analysis. H.Y. Chen: Formal analysis. P.C. Yang: Methodology. H.C. Wu: Conceptualization, resources, data curation, supervision, funding acquisition, validation, investigation, project administration, writing–review and editing.

This research was supported by grants from Academia Sinica (AS-SUMMIT-108), and the Ministry of Science and Technology (MOST-108–3114-Y-001–002, MOST-108–2823–8-001–001, MOST 109–3114-Y-001–001, and MOST 109–2311-B-001–007-MY3 to H.C. Wu). The authors thank the National RNAi Core Facility at Academia Sinica in Taiwan for providing shRNA reagents, Dr. Wann-Neng Jane (Plant Cell Biology Core lab, IPMB, Academia Sinica) for the help in the immunogold labeling and the use of a TEM, and Dr. Ya-Wen Lin (Graduate Institute of Medical Sciences, National Defense Medical Center) for providing the plasmid for constitutively active β-catenin (△45). The authors also thank the Core Facility of the Institute of Cellular and Organismic Biology (ICOB), Academia Sinica, for the technical support in Flow Cytometry and Immunofluorescence Analysis.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Gatenby
RA
,
Gillies
RJ
. 
Why do cancers have high aerobic glycolysis?
Nat Rev Cancer
2004
;
4
:
891
9
.
2.
Hsiao
KC
,
Shih
NY
,
Fang
HL
,
Huang
TS
,
Kuo
CC
,
Chu
PY
, et al
Surface alpha-enolase promotes extracellular matrix degradation and tumor metastasis and represents a new therapeutic target
.
PLoS One
2013
;
8
:
e69354
.
3.
Diaz-Ramos
A
,
Roig-Borrellas
A
,
Garcia-Melero
A
,
Lopez-Alemany
R
. 
alpha-Enolase, a multifunctional protein: its role on pathophysiological situations
.
J Biomed Biotechnol
2012
;
2012
:
156795
.
4.
Capello
M
,
Ferri-Borgogno
S
,
Cappello
P
,
Novelli
F
. 
alpha-Enolase: a promising therapeutic and diagnostic tumor target
.
FEBS J
2011
;
278
:
1064
74
.
5.
Saldanha
RG
,
Molloy
MP
,
Bdeir
K
,
Cines
DB
,
Song
X
,
Uitto
PM
, et al
Proteomic identification of lynchpin urokinase plasminogen activator receptor protein interactions associated with epithelial cancer malignancy
.
J Proteome Res
2007
;
6
:
1016
28
.
6.
Fu
QF
,
Liu
Y
,
Fan
Y
,
Hua
SN
,
Qu
HY
,
Dong
SW
, et al
Alpha-enolase promotes cell glycolysis, growth, migration, and invasion in non-small cell lung cancer through FAK-mediated PI3K/AKT pathway
.
J Hematol Oncol
2015
;
8
:
22
.
7.
Zhan
P
,
Zhao
S
,
Yan
H
,
Yin
C
,
Xiao
Y
,
Wang
Y
, et al
alpha-enolase promotes tumorigenesis and metastasis via regulating AMPK/mTOR pathway in colorectal cancer
.
Mol Carcinog
2017
;
56
:
1427
37
.
8.
Sun
L
,
Lu
T
,
Tian
K
,
Zhou
D
,
Yuan
J
,
Wang
X
, et al
Alpha-enolase promotes gastric cancer cell proliferation and metastasis via regulating AKT signaling pathway
.
Eur J Pharmacol
2019
;
845
:
8
15
.
9.
Takashima
M
,
Kuramitsu
Y
,
Yokoyama
Y
,
Iizuka
N
,
Fujimoto
M
,
Nishisaka
T
, et al
Overexpression of alpha enolase in hepatitis C virus-related hepatocellular carcinoma: association with tumor progression as determined by proteomic analysis
.
Proteomics
2005
;
5
:
1686
92
.
10.
Tomaino
B
,
Cappello
P
,
Capello
M
,
Fredolini
C
,
Sperduti
I
,
Migliorini
P
, et al
Circulating autoantibodies to phosphorylated alpha-enolase are a hallmark of pancreatic cancer
.
J Proteome Res
2011
;
10
:
105
12
.
11.
Tsai
ST
,
Chien
IH
,
Shen
WH
,
Kuo
YZ
,
Jin
YT
,
Wong
TY
, et al
ENO1, a potential prognostic head and neck cancer marker, promotes transformation partly via chemokine CCL20 induction
.
Eur J Cancer
2010
;
46
:
1712
23
.
12.
Chang
GC
,
Liu
KJ
,
Hsieh
CL
,
Hu
TS
,
Charoenfuprasert
S
,
Liu
HK
, et al
Identification of alpha-enolase as an autoantigen in lung cancer: its overexpression is associated with clinical outcomes
.
Clin Cancer Res
2006
;
12
:
5746
54
.
13.
Qian
X
,
Xu
W
,
Xu
J
,
Shi
Q
,
Li
J
,
Weng
Y
, et al
Enolase 1 stimulates glycolysis to promote chemoresistance in gastric cancer
.
Oncotarget
2017
;
8
:
47691
708
.
14.
Wang
H
,
Unternaehrer
JJ
. 
Epithelial-mesenchymal transition and cancer stem cells: at the crossroads of differentiation and dedifferentiation
.
Dev Dyn
2019
;
248
:
10
20
.
15.
Baulida
J
. 
Epithelial-to-mesenchymal transition transcription factors in cancer-associated fibroblasts
.
Mol Oncol
2017
;
11
:
847
59
.
16.
Yao
L
,
Conforti
F
,
Hill
C
,
Bell
J
,
Drawater
L
,
Li
J
, et al
Paracrine signalling during ZEB1-mediated epithelial-mesenchymal transition augments local myofibroblast differentiation in lung fibrosis
.
Cell Death Differ
2019
;
26
:
943
57
.
17.
Mir
N
,
Jayachandran
A
,
Dhungel
B
,
Shrestha
R
,
Steel
JC
. 
Epithelial-to-mesenchymal transition: a mediator of sorafenib resistance in advanced hepatocellular carcinoma
.
Curr Cancer Drug Targets
2017
;
17
:
698
706
.
18.
Lovisa
S
,
Genovese
G
,
Danese
S
. 
Role of epithelial-to-mesenchymal transition in inflammatory bowel disease
.
J Crohns Colitis
2019
;
13
:
659
68
.
19.
Yan
L
,
Xu
F
,
Dai
CL
. 
Relationship between epithelial-to-mesenchymal transition and the inflammatory microenvironment of hepatocellular carcinoma
.
J Exp Clin Cancer Res
2018
;
37
:
203
.
20.
De Matteis
S
,
Canale
M
,
Verlicchi
A
,
Bronte
G
,
Delmonte
A
,
Crino
L
, et al
Advances in molecular mechanisms and immunotherapy involving the immune cell-promoted epithelial-to-mesenchymal transition in lung cancer
.
J Oncol
2019
;
2019
:
7475364
.
21.
Nieto
MA
,
Huang
RY
,
Jackson
RA
,
Thiery
JP
. 
Emt: 2016
.
Cell
2016
;
166
:
21
45
.
22.
Brabletz
T
,
Kalluri
R
,
Nieto
MA
,
Weinberg
RA
. 
EMT in cancer
.
Nat Rev Cancer
2018
;
18
:
128
34
.
23.
Nieszporek
A
,
Skrzypek
K
,
Adamek
G
,
Majka
M
. 
Molecular mechanisms of epithelial to mesenchymal transition in tumor metastasis
.
Acta Biochim Pol
2019
;
66
:
509
20
.
24.
Lambert
AW
,
Pattabiraman
DR
,
Weinberg
RA
. 
Emerging biological principles of metastasis
.
Cell
2017
;
168
:
670
91
.
25.
Zhang
Y
,
Weinberg
RA
. 
Epithelial-to-mesenchymal transition in cancer: complexity and opportunities
.
Front Med
2018
;
12
:
361
73
.
26.
Loh
CY
,
Chai
JY
,
Tang
TF
,
Wong
WF
,
Sethi
G
,
Shanmugam
MK
, et al
The E-cadherin and N-cadherin switch in epithelial-to-mesenchymal transition: signaling, therapeutic implications, and challenges
.
Cells
2019
;
8
:
1118
.
27.
Baarsma
HA
,
Konigshoff
M
,
Gosens
R
. 
The WNT signaling pathway from ligand secretion to gene transcription: molecular mechanisms and pharmacological targets
.
Pharmacol Ther
2013
;
138
:
66
83
.
28.
Volckaert
T
,
De Langhe
SP
. 
Wnt and FGF mediated epithelial-mesenchymal crosstalk during lung development
.
Dev Dyn
2015
;
244
:
342
66
.
29.
Heuberger
J
,
Birchmeier
W
. 
Interplay of cadherin-mediated cell adhesion and canonical Wnt signaling
.
Cold Spring Harb Perspect Biol
2010
;
2
:
a002915
.
30.
Chang
YW
,
Su
YJ
,
Hsiao
M
,
Wei
KC
,
Lin
WH
,
Liang
CL
, et al
Diverse targets of beta-catenin during the epithelial-mesenchymal transition define cancer stem cells and predict disease relapse
.
Cancer Res
2015
;
75
:
3398
410
.
31.
Parikh
RA
,
Wang
P
,
Beumer
JH
,
Chu
E
,
Appleman
LJ
. 
The potential roles of hepatocyte growth factor (HGF)-MET pathway inhibitors in cancer treatment
.
Onco Targets Ther
2014
;
7
:
969
83
.
32.
Uehara
Y
,
Minowa
O
,
Mori
C
,
Shiota
K
,
Kuno
J
,
Noda
T
, et al
Placental defect and embryonic lethality in mice lacking hepatocyte growth factor/scatter factor
.
Nature
1995
;
373
:
702
5
.
33.
Tsarfaty
I
,
Rong
S
,
Resau
JH
,
Rulong
S
,
da Silva
PP
,
Vande Woude
GF
. 
The Met proto-oncogene mesenchymal to epithelial cell conversion
.
Science
1994
;
263
:
98
101
.
34.
Maroun
CR
,
Rowlands
T
. 
The Met receptor tyrosine kinase: a key player in oncogenesis and drug resistance
.
Pharmacol Ther
2014
;
142
:
316
38
.
35.
Nakashiro
K
,
Hara
S
,
Shinohara
Y
,
Oyasu
M
,
Kawamata
H
,
Shintani
S
, et al
Phenotypic switch from paracrine to autocrine role of hepatocyte growth factor in an androgen-independent human prostatic carcinoma cell line, CWR22R
.
Am J Pathol
2004
;
165
:
533
40
.
36.
Wu
CH
,
Kuo
YH
,
Hong
RL
,
Wu
HC
. 
alpha-Enolase-binding peptide enhances drug delivery efficiency and therapeutic efficacy against colorectal cancer
.
Sci Transl Med
2015
;
7
:
290ra91
.
37.
Tsao
CM
,
Yan
MD
,
Shih
YL
,
Yu
PN
,
Kuo
CC
,
Lin
WC
, et al
SOX1 functions as a tumor suppressor by antagonizing the WNT/beta-catenin signaling pathway in hepatocellular carcinoma
.
Hepatology
2012
;
56
:
2277
87
.
38.
Lin
CW
,
Liao
MY
,
Lin
WW
,
Wang
YP
,
Lu
TY
,
Wu
HC
. 
Epithelial cell adhesion molecule regulates tumor initiation and tumorigenesis via activating reprogramming factors and epithelial-mesenchymal transition gene expression in colon cancer
.
J Biol Chem
2012
;
287
:
39449
59
.
39.
Kohler
G
,
Milstein
C
. 
Continuous cultures of fused cells secreting antibody of predefined specificity
.
Nature
1975
;
256
:
495
7
.
40.
Chen
YC
,
Huang
HN
,
Lin
CT
,
Chen
YF
,
King
CC
,
Wu
HC
. 
Generation and characterization of monoclonal antibodies against dengue virus type 1 for epitope mapping and serological detection by epitope-based peptide antigens
.
Clin Vaccine Immunol
2007
;
14
:
404
11
.
41.
Liu
IJ
,
Chiu
CY
,
Chen
YC
,
Wu
HC
. 
Molecular mimicry of human endothelial cell antigen by autoantibodies to nonstructural protein 1 of dengue virus
.
J Biol Chem
2011
;
286
:
9726
36
.
42.
Tung
KH
,
Lin
CW
,
Kuo
CC
,
Li
LT
,
Kuo
YH
,
Lin
CW
, et al
CHC promotes tumor growth and angiogenesis through regulation of HIF-1alpha and VEGF signaling
.
Cancer Lett
2013
;
331
:
58
67
.
43.
Kang
HJ
,
Jung
SK
,
Kim
SJ
,
Chung
SJ
. 
Structure of human alpha-enolase (hENO1), a multifunctional glycolytic enzyme
.
Acta Crystallogr D Biol Crystallogr
2008
;
64
(
Pt 6
):
651
7
.
44.
Kao
SH
,
Wang
WL
,
Chen
CY
,
Chang
YL
,
Wu
YY
,
Wang
YT
, et al
GSK3beta controls epithelial-mesenchymal transition and tumor metastasis by CHIP-mediated degradation of Slug
.
Oncogene
2014
;
33
:
3172
82
.
45.
Suzuki
H
,
Watkins
DN
,
Jair
KW
,
Schuebel
KE
,
Markowitz
SD
,
Chen
WD
, et al
Epigenetic inactivation of SFRP genes allows constitutive WNT signaling in colorectal cancer
.
Nat Genet
2004
;
36
:
417
22
.
46.
Koraishy
FM
,
Silva
C
,
Mason
S
,
Wu
D
,
Cantley
LG
. 
Hepatocyte growth factor (Hgf) stimulates low density lipoprotein receptor-related protein (Lrp) 5/6 phosphorylation and promotes canonical Wnt signaling
.
J Biol Chem
2014
;
289
:
14341
50
.
47.
Hazan
RB
,
Qiao
R
,
Keren
R
,
Badano
I
,
Suyama
K
. 
Cadherin switch in tumor progression
.
Ann N Y Acad Sci
2004
;
1014
:
155
63
.
48.
Shih
JY
,
Yang
PC
. 
The EMT regulator slug and lung carcinogenesis
.
Carcinogenesis
2011
;
32
:
1299
304
.
49.
Shih
JY
,
Tsai
MF
,
Chang
TH
,
Chang
YL
,
Yuan
A
,
Yu
CJ
, et al
Transcription repressor slug promotes carcinoma invasion and predicts outcome of patients with lung adenocarcinoma
.
Clin Cancer Res
2005
;
11
:
8070
8
.
50.
Chang
TH
,
Tsai
MF
,
Su
KY
,
Wu
SG
,
Huang
CP
,
Yu
SL
, et al
Slug confers resistance to the epidermal growth factor receptor tyrosine kinase inhibitor
.
Am J Respir Crit Care Med
2011
;
183
:
1071
9
.
51.
Lopez-Alemany
R
,
Correc
P
,
Camoin
L
,
Burtin
P
. 
Purification of the plasmin receptor from human carcinoma cells and comparison to alpha-enolase
.
Thromb Res
1994
;
75
:
371
81
.
52.
Abbas
AB
,
Lin
B
,
Liu
C
,
Morshed
A
,
Hu
J
,
Xu
H
. 
Design and synthesis of a PD-1 binding peptide and evaluation of its anti-tumor activity
.
Int J Mol Sci
2019
;
20
:
572
.
53.
Lu
RM
,
Hwang
YC
,
Liu
IJ
,
Lee
CC
,
Tsai
HZ
,
Li
HJ
, et al
Development of therapeutic antibodies for the treatment of diseases
.
J Biomed Sci
2020
;
27
:
1
.