Background
Esophageal carcinoma is the eighth most common cause of cancer-related deaths worldwide [
1]. Esophageal squamous cell carcinoma (ESCC) is the most important histopathological form of esophageal carcinoma [
2]. Despite advances in diagnosis and therapy, the prognosis of patients with ESCC remains poor even after surgery [
3]. ESCC is also a highly aggressive malignancy due to the distant metastasis and invasion of neighboring organs [
4].
Cancer stem cells (CSCs) are known to exist in different types of cancers and contribute to aggressive tumor behavior [
5]. CSCs are responsible for tumorigenicity and resistance to therapies such as chemotherapy and radiotherapy after surgery. Recently, it was reported that CSCs from ESCC can be identified by several cell markers, including CD133 and CD90 [
6,
7]. OV6, an epithelial origin marker, was first found to be abundant in hepatic progenitor cells and bile epithelial cells [
8] and was then identified as a CSC surface marker and correlated with tumor progression in hepatocellular carcinoma [
9,
10]. In our previous study, we demonstrated that OV6 expression was also closely associated with the clinical outcome and prognosis of ESCC patients and contributed to tumorigenesis and chemotherapy resistance [
11]. However, the mechanisms involved in the expansion and function of CSCs are not well understood.
Lymphoid enhancer-binding factor 1 (LEF1), a member of the T-cell factor (TCF)/LEF1 family of high mobility transcription factors, is predominantly involved in the Wnt/β-catenin signaling pathway [
12]. Several studies indicated that LEF1 was overexpressed in lung adenocarcinomas and oral squamous cell carcinoma, and its aberrant expression was closely associated with tumor progression and poor prognosis [
13,
14]. It has also been reported that the Wnt signaling pathway regulates cancer stem cell-like properties and contributes to the tumorigenic properties of many cancers, including breast cancer, esophageal cancer, etc. As a downstream mediator of the Wnt signaling pathway, LEF1 is also essential for stem cell maintenance and organ development in addition to its role in epithelial-mesenchymal transition (EMT) by activating the transcription of markers [
15,
16]. Moreover, epithelial cancer cells can obtain stem-like properties through EMT, a process that also plays a crucial role in cancer progression [
17]. Our previous study has also shown that LEF1 is upregulated in ESCC and could positively regulate the invasion, migration, and EMT of ESCC through the cooperation of the OCT4 transcription factor, which indicates the potential role of LEF1 in mediating the self-renewal properties of CSCs [
18]. However, the role of LEF1 in the transcriptional regulation of CSC regulators during ESCC progression remains unclear.
The transforming growth factor beta (TGF-β) signaling pathway is a critical regulator involved in cell growth, differentiation, and development [
19]. The aberrant activation of the TGF-β signaling pathway is responsible for the self-renewal properties and drug resistance of various cancers, including hepatocellular carcinoma, esophageal squamous cell carcinoma, and colorectal cancer [
20‐
22]. During the initiation and progression of tumors, TGF-β could facilitate EMT, a process that is crucial for the acquisition of CSC-like properties [
23]. Thus, the specific inhibition of TGF-β signaling has been developed for anti-cancer therapy. A recent study reported that the application of a TGF-β inhibitor prevents the development of CSCs, which promotes the chemotherapeutic effect against triple-negative breast cancer [
24]. Taken together, these data suggest that hyperactivation of the TGF-β signaling pathway leads to tumorigenicity and stem-like properties of CSCs in various tumors. Inhibitor of DNA binding 1 (ID1), a member of the ID protein superfamily, belongs to the helix-loop-helix transcription factor family. ID1 is widely expressed in many tissues and functions in a number of biological processes, including cell proliferation and apoptosis, among others [
25]. Many studies indicate that ID1 is an oncogene and is critical in promoting cancer progression. High levels of ID1 are found in human cancers of the breast, lung, and esophagus and are associated with poor patient prognosis [
26]. In the process of cell development, ID1 plays an important role in the maintenance of embryonic stem cell self-renewal and hematopoietic stem cells [
27,
28]. ID1 has also been identified as a key regulator of the CSC phenotype in colon cancer [
29]. Furthermore, the important role of ID1 in CSC phenotypes and its participation in the TGF-β-SMAD2/3-ID1 axis make it a candidate for the study of cancer stem cell properties in ESCC.
In the present study, we report that LEF1 overexpression promotes TGF-β signaling pathway activation by directly binding to ID1 to enhance tumorigenesis and the CSC-like phenotype of ESCC in vitro and in vivo. These results provide novel insight into LEF1-mediated tumorigenesis and the mechanisms by which the TGF-β signaling pathway is activated in ESCC.
Methods
Patients and tissue specimens
The tissue microarrays chips (TMA) of the ESCC cohort (31 patients, 98 patients, 75 patients and 70 patients) were purchased from Shanghai Weiao Biotech Company and Wuhan Servicebio Biotech Company, which were obtained from 4 different medical centers. All specimens were well documented with complete follow-ups for periods from 4 to 5 years. Ninety-five patient specimens were collected from patients who were diagnosed with primary ESCC and who received radical esophageal surgery without preoperative chemoradiotherapy from 2012 to 2013 at Changhai Hospital (Shanghai, China). All samples were fixed in 4% formaldehyde and embedded in paraffin wax. The patient samples were obtained with informed consent according to an established protocol approved by the Ethics Committee of Changhai Hospital. 95 patients in our center were observed until May 2017, with a median observance time of 27 months.
Cell culture
The human Eca109, TE1 cells were purchased from the Shanghai Cell Bank (Shanghai, China). After measured by mycoplasma detection, DNA-Fingerprinting, isozyme detection and cell vitality detection, these cell lines were immediately expanded and frozen such that they could be restarted every 3 to 4 months. All cell lines cultured in Dulbecco’s modified Eagle’s medium (Gibco, CA, USA) supplemented with 10% heat-inactivated foetal bovine serum (Gibco-BRL) and antibiotics (100 U/ml penicillin and 100 U/ml streptomycin, HyClone Laboratories, Inc., USA) at 37 °C in a humidified atmosphere of 5% CO2. Spheroids were cultured in F12/DMEM (Gibco, CA, USA) supplemented with EGF (Sigma, St Louis, USA), FGF (Gibco, CA, USA) and ITS (Sigma, St Louis, USA).
Vector construction and lentivirus infection
We constructed lentiviral vectors encoding the human LEF1 gene or green fluorescent protein (GFP) in the pLenti-EF1a-EGFP-P2A-Puro-CMV-MCS-3Flag vector (HeYuan Bio-technology Co., Shanghai, China) and designated them as LV-LEF1 or LV-GFP. The lentiviral vectors were transfected into the HCC cells with a multiplicity of infection (MOI) 5 in the presence of polybrene (5 μg/ml) for 6 h. Stable Eca109 and TE-1 cells knockdown of LEF1 were generated using lentiviral constructs expressing shLEF1(shLEF11#CCCATCCCGAGAACATCAA; shLEF12#CCTCATCCAGCTATTGTAA; shLEF13#GCTACATATGCAGCTTTAT) and negative control (HeYuan Bio-technology Co., Shanghai, China), and incubated with 2 μg/ml puromycin (Sigma, St Louis, USA).
Magnetic cell sorting and flow cytometry
Cells were labelled with primary OV6 antibody (mouse IgG1; R&D Systems, Minneapolis), magnetically tethered to rat anti-mouse IgG1 microbeads, and sorted with a Mini-MACS™ Cell Sorter Kit (Miltenyi Biotec, CA). All of the procedures were following the manufacturer’s instructions. The sorted cells were evaluated by flow cytometry analysis. The flow cytometry was performed with MoFlo Sorter (Beckman, CA) or ImageStreamx (Millipore, US) using an APC-conjugated-OV6 antibody (R&D Systems, Minneapolis) and following manufacturer’s instruction.
Cells were magnetically sorted, then 3000 OV6+ cells were cultured in Ultra-Low Attachment 6-well plates (Corning Lnc., Coring, NY) for 10–12 days. The tumor spheres was counted under an inversed microscopy and the representative pictures were taken. All experiments were performed in triplicate.
Quantitative real-time polymerase chain reaction (qRT-PCR)
Total RNA was extracted from cultured ESCC cell lines or magnetic sorted cells using Trizol (Invitrogen, Grand Island, NY) according to the manufacturer’s instruction. The cDNA was synthesized using the PrimeScript RT Reagent Kit (TaKaRa Bio, Shiga, Japan) following the manufacturer’s instructions. Real-time PCR was performed on a Roche Light Cycler 480 (Roche) using SYBR Green PCR Master Mix (TaKaRa Bio, Shiga, Japan). Primer sequences are listed in Table
1. Each measurement was performed in triplicate and the results were normalized by the expression of the GADPH gene. Fold change relative to mean value was determined by 2
-△△Ct. All experiments were performed in triplicate.
Table 1
Sequence if PCR Primers Used in This Study
CD133 | GCCACCGCTCTAGATACTGC | TGTTGTGATGGGCTTGTCAT |
CD44 | GAGCATCGGATTTGAGA | CATACTGGGAGGTGTTGG |
ABCG2 | CAGGTTACGTGGTACAAGATGA | GATCAGTGATAAGCTCCATTCC |
KLF4 | CCATTACCAAGAGCTCATGC | GTGCCTGGTCAGTTCATCTG |
SOX2 | CAAGATGCACAACTCGGAGA | GCTTAGCCTCGTCGATGAAC |
NANOG | CTGCTGGACTGAGCTGGTTGCC | GCTGAGGCCTTCTGCGTCACA |
OCT4 | AGTGAGAGGCAACCTGGAGA | ACACTCGGACCACATCCTTC |
EPCAM | AATCGTCAATGCCAGTGTACTT | TCTCATCGCAGTCAGGATCATAA |
CXCR4 | ACTACACCGAGGAAATGGGCT | CCCACAATGCCAGTTAAGAAGA |
LEF1 | AGAACACCCCGATGACGGA | GGCATCATTATGTACCCGGAAT |
COMP | GATCACGTTCCTGAAAAACACG | GCTCTCCGTCTGGATGCAG |
BMP8A | AGAAAAGCAACGAGCTGCCG | GCCGCGGACGTCATCAAA |
ID1 | CTGCTCTACGACATGAACGG | GAAGGTCCCTGATGTAGTCGAT |
ID3 | GAGAGGCACTCAGCTTAGCC | TCCTTTTGTCGTTGGAGATGAC |
SMAD9 | CTAGGCTGGAAGCAAGGAGAT | GGGGAATCGTGACGCATTT |
CHRD | TTCGGCGGGAAGGTCTATG | ACTCTGGTTTGATGTTCTTGCAG |
DCN | ATGAAGGCCACTATCATCCTCC | GTCGCGGTCATCAGGAACTT |
Western blot
Whole cultured cells were homogenized in 0.1% SDS and 1 mM PMSF (phenylmethylsulfonyl fluoride) and centrifuged at 12, 000 g for 15 min. Protein extracts were subjected to SDS-PAGE and analyzed using the following primary antibodies: LEF1 (Abcam, ab137872), TGF-β1 (Proteintech,21,898–1-AP), ID1(Santa, sc-133,104), Smad2(CST,5339), Smad3(CST, 9523), p-Smad2(CST, 3108), p-Smad3(CST, 9520) and GADPH (Abcam, ab8245). Then, the membranes were incubated with secondary antibodies (CST,7076,7074) at room temperature for 1 h. The dilution ratio was determined according to the recommended instructions. Protein levels were detected by the Image-Pro Plus 6.0 system (Bio-Rad,1,708,265). Quantification of bands intensity was measured using ImageJ software (version 1.34). All experiments were performed in triplicate.
Immunohistochemistry (IHC)
The TMAs and 95 ESCC tissues were fixed in 4% methanol, embedded in paraffin, and cut into a thickness of 5 μm. After deparaffinization and rehydration procedures, antigen recovery was performed in a heated citrate buffer (pH 6.0) or EDTA buffer (PH 8.0) for 30 min. Then, the slides were incubated with UltraSensitive Streptavidin Peroxidase Kit (Fuzhou Maixin Biotechnology, Fuzhou, China) and the anti-OV6 (1:50; R&D systems, Minneapolis, MN) and anti-LEF1(1:200, Abcam, ab137876) primary antibodies at 4 °C overnight. Then, diaminobenzidine (DAB) (Dako, Carpinteria, CA, USA) staining was used to image specific markers. The immunostaining staining scores was carried out according to our previous study [
11]. Briefly, the immunostaining levels were scored as 0 (negative), 1 + (weakly positive), 2 + (moderately positive), or 3 + (strongly positive). High expression in tumor cells was defined as score ≥ 2 + .
Chemotherapy drug treatments, soft agar Colony formation assay, and viability assay
ECA109 and TE1 cells from different groups were treated with cis-platinum (2.0 g/mL) for 4 days, and the percentage of OV6+ cells was then measured by flow cytometry. For the soft agar assay, magnetically sorted OV6+ cells were cultured in 1 mL of 0.7% agarose with DMEM-mixed upper gel on 6-well plates. Then, 0.6 g/mL of cis-platinum was added into 6-well plates and incubated for 2 weeks. Colony formation numbers was determined by microscope counting. Cell viability after drug treatment was assessed by a cholecystokinin-8 assay (CCK8).
Xenograft mice
Six-week-old male BALB/c nude mice were purchased from Shanghai Experimental Center (Shanghai, China). Mice were used to evaluate the effects of LEF1 upregulation and downregulation of on tumorigenicity and tumour growth. Briefly, different numbers of LV-LEF1, LV-GFP, and LV-shLEF1 ECA109 cells were suspended in 200 μL of DMEM and Matrigel (1:1) (Corning) and injected into the subcutaneous tissue of mice. Tumor size and incidence were measured weekly. Mice were sacrificed at the indicated time points according to the protocols approved by the SMMU Animal Care Facility and the National Institutes of Health guidelines. Tumors were harvested for assessment of tumour size, tumour incidence, western blot and immunohistochemistry analysis.
RNA-seq
RNA-seq was performed on three biological replicates. LEF1-overexpression ECA109 cells and negative control ECA109 cells were used for RNA-seq analysis by Shanghai Novelbio corporation (Shanghai, China). Total RNA from ECA109 LV-GFP/LV-LEF1 was extracted by Trizol and kept at − 80 °C. DNA was removed using a DNA-free DNA Removal Kit (Thermo Fisher, AM1906). The RNA quantitation and quality measurement were performed using a Bioanalyzer 2200 (Agilent Technologies, USA). RNA with a RIN (RNA integrity number) > 8.0 was considered acceptable for cDNA library construction. Differentially expression genes were considered to be significant between groups when the P-value was < 0.05 and the fold change of expression was ≥1.50-fold or ≤ 0.67-fold. All RNA-Seq files are available from the GEO database (accession number: GSE128914).
Luciferase report assay
Different groups of ESCC cell were cultured in 24-well plates in triplicate. After 24 h, ESCC cells were transfected with the indicated plasmids and pRL-TK Renilla plasmid using lipofectamine 2000 reagent (Thermo Fisher, USA, No.11668019). Luciferase and Renilla signals were measured 48 h after transfection by a Dual-Luciferase Reporter Assay Kit (Promega, No. E1980). Data were normalized by the division of firefly luciferase activity with that of Renilla luciferase to eliminate transfection efficiency difference.
Chromatin immunoprecipitation (ChIP) assays
We identified the LEF1-bingding sites on ID1 promoter region by using JASPAR and also referred to Chip-Seq data of LEF1 on GEO. ChIP assay was conducted with SimpleChIP® Enzymatic Chromatin IP Kit (CST, 9003) following the manufacturer’s instructions. Briefly, ECA109 and TE1 cells (4 × 106) were cross-linked by using 1% formaldehyde and used for each immunoprecipitation experiment. Chromatin was digested with the micrococcal nuclease. 2% aliquots of lysates were used as an input reference. LEF1 antibody (Abcam, ab137872) or normal rabbit IgG (CST, 2729) were incubated with the other immunoprecipitation samples at 4 °C for overnight. Then, the crosslink DNA was reversed by NaCl and proteinase K. Immunoprecipitated DNA was amplified by PCR using their specific primers. The primer sequences for ID1 gene were 5′-CGCCCGCTTTAAATTTCGG-3′ (forward), and 5′- CACAGATGAGAGAAA.
TTGAGGC − 3′ (reverse). The signals were calculated as the percentage of input.
Statistical analysis
SPSS 22 software (SPSS, Chicago, IL, USA) was used to statistically analyse the data. The association between markers and clinical features were analysed by chi-square test, Fisher’s exact test or two-side t-test. Spearman’s rank correlation was used to analyse the association between LEF1 and OV6 expression. Survival curves were analysed by using the Kaplan-Meier method. Multivariate analysis of survival was examined by Cox proportional hazard regression model. The experimental data were obtained in three independent experiments and analysed by ANOVA. P < 0.05 indicated statistical significance.
Discussion
Esophageal cancer (EC) is one of the most malignant tumors worldwide and has a high mortality rate [
31]. Most esophageal cancers are histologically classified as esophageal squamous cell carcinoma (ESCC), which accounts for approximately 80% of all EC [
32]. Despite advances in surgery, chemotherapy, and radiotherapy in esophageal squamous cell carcinoma (ESCC) treatment, the prognosis of ESCC patients is still poor [
33].
Cancer stem cells (CSCs) are correlated with cancer recurrence and metastasis due to their self-renewal properties and capacity for drug resistance [
34]. It was recently reported that CSCs directly contribute to therapy resistance in and tumor progression of ESCC [
35]. Recent studies have verified that OV6, an epithelial marker, can be used as a CSC marker in many cancers such as hepatocellular carcinoma and rectal cancer [
36]. The present study found that OV6 was also upregulated in ESCC, and the high expression of OV6 was closely associated with the clinicopathological characteristics and prognosis of ESCC patients [
11]. The elimination of CSCs might prevent cancer progression and recurrence in patients. Compared with traditional treatment, targeted therapy has shown a unique advantage. Therefore, it is imperative to identify novel cancer targets and investigate their clinical relevance to targeted treatment.
LEF1 is a transcription factor that primarily participates in the Wnt/β-catenin signaling pathway. LEF1 is also a facilitator of epithelial-mesenchymal transition (EMT), a feature of cancer cell migration and invasion, as well as cancer cell proliferation and viability [
37]. A previous study suggested that LEF1, as a candidate CSC marker, was highly elevated during EMT in hepatocellular carcinoma [
38]. Moreover, our previous study also reported that LEF1 was highly expressed in esophageal squamous cell carcinoma and was closely related to tumor progression and poor patient prognosis [
18]. In colorectal cancer patients, the overexpression of LEF1 represented a risk factor for the poor overall survival of CRC patients, and increased expression of LEF1 with decreased expression of Notch2 could be used to facilitate the early detection of colorectal cancer [
39]. We speculated that LEF1 8could modulate gene transcription and tumorigenesis independently by processing the Wnt/β-catenin signaling pathway, which might contribute to maintaining CSC-like characteristics. In this study, we found that the overexpression of LEF1 promoted the stem cell-like properties of CSCs in ESCC, including spherical tumor formation, chemoresistance, and tumorigenicity. Moreover, the combination of high levels of LEF1 and OV6 predicted the aberrant clinicopathological characteristics and poor patient prognosis in ESCC patients. These results indicate the crucial role of LEF1 in the regulation of the CSC-like phenotype in ESCC cells, which might hold promise for the development of novel therapeutic targets for tumor treatment.
The TGF-β signaling pathway is an important effector of a number of pathways that plays complex roles in the development, progression, and metastatic potential of cancers and is correlated with tumor invasion and poor patient prognosis [
40]. Recently, many studies have indicated an important role for the TGF-β signaling pathway in promoting EMT and a CSC-like phenotype [
41]. Moreover, EMT, a key process that is often activated during cancer invasion, has been reported to generate cells with stem cell-like properties [
41]. To investigate the mechanism of LEF1’s effect on the tumorigenesis and self-renewal properties of ESCC cells, we applied the RNA-Seq method and found that the TGF-β signaling pathway is significantly activated after LEF1 overexpression. Then, we selected a series of upregulated genes (DCN, COMP, BMP8A, ID3, ID1, SMAD9, and CHRD) from the RNA-Seq data that are related to the TGF-β pathway and validated their expression by qRT-PCR analysis. Based on a luciferase reporter assay, we confirmed that LEF1 directly binds to the promoter of the ID1 gene, which was shown to regulate the CSC-like phenotype in many tumors [
30]. In addition, a significant correlation between LEF1 and ID1 expression was observed in clinical ESCC patients, strongly suggesting that LEF1 regulation of the CSC phenotype is associated with ID1 expression.
In summary, we demonstrated that the overexpression of LEF1 was closely associated with aberrant clinicopathological characteristics and might be used as an independent prognostic factor of ESCC. The association of LEF1 with OV6 expression in ESCC has been statistically proven, and the concomitant elevated expression of LEF1 and OV6 might contribute to poor overall survival rates in ESCC patients. Moreover, the overexpression of LEF1 directly upregulates ID1 and activates the TGF-β signaling pathway, thereby promoting tumorigenicity and the CSC-like phenotype. In-depth investigation of the function of TFAP4 in ESCC would partially shed light on the mechanisms underlying the high rate of recurrence in ESCC and provide a potential therapeutic target to prevent the recurrence of ESCC.
Conclusions
In the present manuscript, we showed the elevated expression of LEF1 was associated with aberrant clinicopathological characteristics and poor patient prognosis of ESCC patients. We also investigated the clinicopathological significance of the association of OV6 with LEF1 expression. Furthermore, we identified LEF1 as a key regulator of CSC-like phenotype and was responsible for ESCC tumorigenesis. Mechanically, LEF1 overexpression in ESCC directly upregulates ID1 and activates TGF-β pathway.
This work underlines the importance of LEF1 in regulating CSC-like phenotype and proposes potential new therapeutic target to better treat this deadly disease.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.