Interleukin-23 receptor signaling mediates cancer dormancy and radioresistance in human esophageal squamous carcinoma cells via the Wnt/Notch pathway
- 1k Downloads
In the tumor microenvironment, inflammatory cells and molecules influence almost every process; among them, interleukin-23 (IL-23) is a pro-inflammatory molecule that exhibits pro- or anti-tumor properties, but both activities remain poorly understood. In this study, we investigated the effect of extracellular IL-23 in IL-23 receptor-positive (IL-23R+) esophageal squamous cell carcinoma (ESCC) and explored the mechanisms underlying this effect. We analyzed ESCC tumor tissues by immunohistochemical and immunofluorescence staining and found that IL-23, which was highly expressed, co-localized with Oct-4A in IL-23R+ ESCC cells. In addition, IL-23 treatment significantly increased the accumulation of CD133+ cells and activated the Wnt and Notch signaling pathways in CD133−IL-23R+ ESCC cell lines. Consistently, CD133−IL-23R+ cells pretreated with IL-23 showed stronger anti-apoptosis activity when exposed to radiation and higher survival than untreated groups. Moreover, the inhibition of Wnt/Notch signaling by a small-molecule inhibitor or siRNA abolished the effect of IL-23-induced dormancy and consequent radioresistance. Taken together, these results suggested that IL-23 facilitates radioresistance in ESCC by activating Wnt/Notch-mediated G0/1 phase arrest, and attenuating these detrimental changes by blocking the formation of dormancy may prove to be an effective pretreatment for radiotherapy.
IL-23/IL-23R is correlated with the acquisition of stem-like potential in ESCC.
CD133−IL-23R+ ESCCs acquired dormancy via IL-23.
Radioresistance depends on IL-23-mediated Wnt/Notch pathway activation in vitro and vivo.
KeywordsInterleukin-23 M1 macrophage Wnt/notch Radioresistance Biomarker
Diagnosing esophageal squamous cell carcinoma (ESCC) in the early stage is difficult because the clinical manifestations, such as esophageal obstruction and burning or aching with a concurrent dragging sensation, tend to be ignored or even misdiagnosed . Previous reports showed that ESCC ranks as the sixth-leading cause of cancer death worldwide and its top morbidity and mortality all fasten on China [2, 3]. In fact, most patients die of recurrences after decades of remission due to resistance to current therapies and distant organ metastases. Irrespective of the resectability of lesions, radiotherapy is the optimal choice and first-line treatment for this disease, despite its limited curative effect in cases with resistance . Therefore, determining the critical factors and conditions that affect the formation of radioresistance is the primary step in the development of targeted ESCC therapy.
According to the studies, tumors contain a subset of cancer cells that establish colonization from a non-dividing state of dormancy and ultimately become resistant to therapy [5, 6, 7]. Cancer cells with stem cell–like features like dormancy have been identified in hematological diseases and solid tumors widely [8, 9]. However, a better understanding of their mechanism of resistance is needed to isolate, identify, and eliminate these cells, and dissect the responsive signaling pathways. As a matter of fact, cancer cells with stem-like properties are not easily detected because the dormant cell population is often small and thought to be maintained by the resident microenvironment such as stromal cells, secreted growth factors, and local immunoregulation factors [10, 11]. Despite the limited knowledge, cancer cells with stem-like properties are generally accepted to express markers typical of normal stem cells, such as c-myc, Oct-4A, CD133, and CD44 . Although CD133 is the most commonly used marker to isolate cancer stem-like cells in several carcinomas, markers, such as those of therapy resistance, differ by cancer type [5, 8, 11, 12, 13]. Likewise, rather than focusing on the identification of optimal markers in specific tumors, research should concentrate on the acquisition of quiescence, self-renewal, or other stemness, and the expression or upregulation of markers constitutes only one method to identify them. Our previous study showed that IL-23 initiates the epithelial–mesenchymal transition (EMT) in ESCC, which allows cancer cells to become motile and invasive . Accumulating evidence shows that cancer cells possessing stem-like properties exhibit an EMT phenotype that enables cells to survive and actively cause relapse, thus a better understanding of the causes of EMT could provide a foundation for tumor treatment [15, 16]. Specifically, tumor-infiltrating lymphocytes (TILs) and their associated cytokines, which are generally associated with better clinical outcomes, play an immune-editing role in cancer, particularly in therapy resistance [10, 17, 18]. However, the tumor-promoting role of immune factors in the formation of cancer stem-like cells has rarely been reported.
In addition, cancer cells that obtain characteristics of stem-like properties express developmentally regulated genes associated with normal stem cells . Among these crucial mechanisms, the Wnt and Notch signaling pathways have been proven to be critical for various cell events . Moreover, the Wnt/β-catenin signaling pathway has been implicated in the maintenance of self-renewal in various types of stem cells through activating target genes, such as CD44, Cyclin D1, c-myc, and p21 . In comparison, Notch signaling was proven to promote the differentiation of stem cells and the activation of Hes family proteins, Myc, p21, and Cyclin D3. Although accumulating data have underlined that the cooperative or adversarial relationship between the Wnt/Notch signaling pathway correlates with tumorigenesis, dormancy, and progression, their respective roles in cancer stem–like cell development remain uncertain . Furthermore, the crosstalk between these pathways and its effect on therapy, particularly radiotherapy, warrants discussion. Radiosensitivity promotes cell–cell contact, which facilitates radical scavenging, DNA repair, and cell-cycle arrest . In conclusion, cancer cells transform into stem-like cells via the acquisition of quiescence, and their responses to radiation might explain the persistence of radioresistance.
The purpose of this study was to investigate the effect of IL-23 on the stem-like properties of ESCC cells. To this end, we analyzed the specific characteristics of IL-23R+ ESCC cells to detect their capability of radioresistance after IL-23 treatment. In addition, we utilized small-molecule inhibitors or siRNA to explore the role of Wnt/Notch signaling in the development of dormancy, particularly in IL-23-mediated cell-cycle modulation.
Materials and methods
The human ESCC cell lines TE-1 was purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China), and TE-10, KYSE-150, and ECA-109 cells were a gift from Dr. Ye Hua of the Clinical Medicine College at Jiangsu University. The human esophageal epithelial cell line, Het-1A, was purchased from Jennio Biological Technology (Guangzhou, China). Cells were respectively maintained in RPMI 1640 or DMEM (Gibco, Carlsbad, CA, USA) containing 10% fetal bovine serum (HyClone, Logan, UT, USA) and cultured at 37 °C in 5% CO2. IL-23, anti-IL-23, and anti-IL-23R were purchased from R&D Systems (Minneapolis, MN, USA), and WHI-P154, DAPT, and ICG-001 were purchased from Selleck Chemicals Inc. (Houston, TX, USA). For RNA interference, cells were transfected with siRNA oligos (Gene Pharma, Shanghai, China) against human β-catenin and Notch1 using Lipofectamine 2000 (Invitrogen, San Diego, CA, USA). The following siRNA sequences (GenePharma, Shanghai, China) were used: siRNA-β-catenin: sense, 5′-GUCCUGUAUGAGUGGGAACTT-3′; antisense, 5′-GUUCCCACUCAUACAGGACTT-3′. siRNA-Notch1: sense, 5′-GGGCUAACAAAGAUAUGCATT-3′; antisense, 5′-UGCAUAUCUUUGUUAGCCCTT-3′. Negative controls using non-transfected cells and empty vector-transfected cells were performed in parallel. The colony formation assay and soft agar assay were described previously .
ESCC and adjacent non-cancerous tissues were collected from patients at the Affiliated Hospital of Jiangsu University (Zhenjiang, China). The tumors had not been treated with chemotherapy or radiotherapy, as confirmed by the Department of Pathology of the Affiliated Hospital of Jiangsu University. Tumor staging was determined according to the American Joint Committee on Cancer criteria. Immunohistochemical (IHC) and immunofluorescence staining, scoring and statistics were performed as previously described , and primary antibodies for Oct-4A, CD133, IL-23, IL-23R, CD68, and HLA-DR were purchased from R&D Systems (Minneapolis, MN, USA).
Flow cytometry and cell sorting
The number of M1 macrophages was measured in 16 independent ESCC samples and their adjacent normal samples. Surface antigens were labeled by incubation with anti-human HLA-DR/CD68 antibodies (BD Biosciences, San Jose, CA, USA). To identify CD133+/IL-23R+ cell subpopulations, ESCC cells (ESCCs) were incubated with APC-conjugated anti-CD133 (Miltenyi Biotec, Auburn, CA, USA) and PE-conjugated anti-IL-23R antibodies (R&D Systems, Minneapolis, MN, USA) or isotype control antibodies. For the cycle analysis, cells were stained with PI (Sigma-Aldrich, St. Louis, MO, USA). To identify early apoptotic cells after ablative radiation (AR, 10 Gy once) or fractionated radiation (FR, five doses of 2 Gy once every 48 h for up to 10 Gy), cells were stained with Annexin V-FITC and PI (KeyGen Biotech, Nanjing, China). The levels of ROS were determined by staining cells with DCFH-DA (Sigma-Aldrich, St. Louis, MO, USA). CFSE (Sigma-Aldrich, St. Louis, MO, USA) was diluted to a desired working concentration in normal medium to culture and label cells. The degree of DNA damage was determined based on quantity of γH2AX (Millipore, Milford, MA, USA). The data were additionally analyzed using the FlowJo software (Tree Star Inc., Ashland, OR, USA). All experiments were independently repeated at least three times.
Molecular biology experiments
For qRT-PCR, total RNA was extracted from cells using TRIzol reagent (Invitrogen, CA, USA), according to the manufacturer’s instruction. The following primers were used to detect specific mRNAs (Invitrogen, Shanghai, China): c-myc (forward, 5′-ACCACCAGCAGCGACTCT-3′; reverse, 5′-GCTGTGAGGAGGTTTGCTGT-3′); Oct-4A (forward, 5′-GAGAATTTGTTCCTGCAGTGC-3′; reverse, 5′-GTTCCCAATTCCTTCCTTAGTG-3′); Cyclin D1 (forward, 5′-GATCAAGTGTGACCCGGACT-3′; reverse, 5′-TCCTCCTCTTCCTCCTCCTC-3′); Hes 1 (forward, 5′-TGAAGGATTCCAAAAATAAAATTCTCTGGG-3′; reverse, 5′-CGCCTCTTCTCCATGATAGGCTTTGATGAC-3′); β-catenin (forward, 5′-GCCCAGGACCTCATGGAT-3′; reverse, 5′-CCAAAATCCATTTGTATTGTTACTCC-3′); Notch1 (forward, 5′-CACTGTGGGCGGGTCC-3′; reverse, 5′-GTTGTATTGGTTCGGCACCAT-3′). A Western blot analysis was performed as previously described .
Clonogenic survival assays
Clonogenic survival assays data were employed to identify radioresistance. To this end, exponentially growing cells were cultured with IL-23 (50 ng/mL) or isometric sterile PBS for 24 h and then treated with ablative radiation for different doses (0, 2, 4, 6, 8, and 10 Gy). Cells were grown for up to 9 days, then fixed and stained with 2% crystal violet; colonies of at least 50 cells were scored. Cell survival curves based on the mean survival fractions of the cell line were generated using the linear-quadratic equation according to the following formula: SF = exp −(αD + βD2), where SF is the surviving fraction and D is the dose.
Methods for evaluation in vivo
ESCC xenografts were implanted subcutaneously by injecting CD133−IL-23R+ TE-1 cells (1 × 106) into the dorsal anterior flank of nude mice (BALB/c inbred, female, 3–4 weeks old, n = 7 per group) to examine the activity of IL-23. Radiation group tumors were treated with 4 Gy of ionizing radiation every 72 h up to a total of 24 Gy, and human recombinant IL-23 (50 ng) was intratumorally injected 1 h prior to each radiation treatment. For the combined treatment, DAPT (0.2 mg) and ICG-001 (0.5 mg) were injected 6 h apart on the day before radiation through intravenous injection of the caudal vein. The growth curves were calculated along with indicated treatments. The tumor volume was measured with a Vernier caliper and calculated as following: tumor size = (π × length × width × height)/6. Radiation was administered using the small animal radiation resource platform, and tumor volume curves were plotted every week. Xenografts were harvested for molecular analysis, and the standard deviation values were determined for each sample’s normalized values.
The differences between groups in the clone assay, flow cytometry, protein expression, RT-PCR, and animal experiments results were analyzed using one-way or two-way ANOVA. Survival curve were analyzed by the Kaplan–Meier method. The difference was significant for p values of < 0.05. All data were analyzed using the SPSS version 16.0 software (Chicago, IL, USA).
IL-23/IL-23R is correlated with the acquisition of stem-like potential in ESCC
To eliminate the potential effects of innate CD133+ cells, we isolated CD133− cells from TE-1 and ECA 109 cells or Het-1A cells using FACS. The mean proportions of cells that expressed CD133 were approximately 1.7% for TE-1 cells and 1.9% for ECA 109 cells, which is almost invisible in Het-1A cells (Fig. 1c). Moreover, a flow cytometry analysis showed that about 8 to 10% of CD133− ESCCs and less than 3% of Het-1A cells were IL-23R+ cells (Fig. 1d). Therefore, after sorting the CD133−IL-23R−/ CD133−IL-23R+ cells, we found that more CD133+ cells were detected in the CD133−IL-23R+ ESCCs after culturing with IL-23 (Fig. 1e). In addition, the expression of CD133 was upregulated in CD133−IL-23R+ cells cultured with IL-23. Meanwhile, the CD133−IL-23R+ cells were pretreated with IL-23 for 24 h, and the acquired upregulation of CD133 expression was maintained for more than 72 h after the removal of IL-23 in ESCCs (Fig. 1f). Furthermore, we demonstrated that IL-23 induced the expression of the aforementioned Oct-4A and c-myc, which are transcription factors involved in cellular transformation and cell cycle progression (Fig. 1g). We have previously suggested that aberrant IL-23 expression enhances the activity of EMT in ESCCs . Therefore, IL-23 mediates the trans-differentiation not only by affecting the accumulation of stem-like makers but also by directionally inducing the expression of stemness-related genes in IL-23R+ ESCCs.
CD133−IL-23R+ ESCCs acquired partial stem-like properties especially dormancy via IL-23
As mentioned, most cancer cells with stem-like properties are characterized by cell-cycle arrest at G0/1. IL-23 was reported to promote proliferation in various types of cancers, but not in ESCCs, according to our previous study . Specifically, treatment with IL-23 rapidly increased the percentage of cells arrested at G0/1 in TE-1 (from 50.10 ± 5.22% to 71.90 ± 9.44%) and ECA 109 cells (from 49.96 ± 5.34% to 68.57 ± 9.01%), compared to controls (Fig. 2e). As transcription factors that link the immune signaling and tumor activity, STAT proteins regulate the key aspects of cancer cell growth, survival, and differentiation . Our data identified excessive p-Stat3 expression in the cells when exposed to IL-23. Moreover, the accumulation of p-Stat3 has been revealed that promotes dysfunction in NF-κB, Wnt, Notch, or other cell signaling pathways in a variety of cells . Accordingly, we found that treating IL-23R+ ESCCs with extracellular IL-23 stimulated the expression of Wnt3a (> 3.0-fold change, p < 0.01), Notch1 (> 1.9-fold change, p < 0.05), Cyclin D1(the mRNA level, p < 0.01), and Hes-1(p < 0.01), and this effect was attenuated by WHI-P154 (inhibitor of Stat3 phosphorylation, p < 0.01 at both ratios), while the pretreatment of anti-IL-23R blocked the phosphorylation cascades and activated Wnt/Notch that trigged by IL-23 (Fig. 2f, g). Taken together our data indicate that the activation of Stat3 is required for IL-23 binding to IL-23R and that Wnt and Notch may work in a combinatorial manner in the regulation of tumorigenesis and dormancy.
Radioresistance depends on IL-23-mediated Wnt/Notch pathway activation
Abnormal DNA repair and radicals scavenging capacity as well as cell-cycle arrest are talismans of cancer stem-like cells exposed to radiation [21, 24]. Although reactive oxygen species (ROS) only differ between groups exposed to irradiation or not (Fig. 3c), IL-23 markedly suppressed the level of γH2AX in CD133−IL-23R+ ESCCs compared to cells treated with AR alone (Fig. 3). As shown in Fig. 3e, radiation induced the phosphorylation of γH2AX and increased the expression levels of DNA-PKcs, Ku70, and Ku80 within 1 h. However, pretreatment with IL-23 did not influence the accumulation of these factors except p-γH2AX. In addition, knocking down β-catenin or Notch 1 gene expression significantly inhibited the IL-23-induced arrest of CD133−IL-23R+ TE-1 cells, as detected by flow cytometry of CFSE labeling (cells co-transfected for 3 days, 55.17 ± 3.71%; NC cells, 84.50 ± 4.09%; siRNA-β-catenin cells, 66.50 ± 9.67%; and siRNA-Notch1 cells, 70.67 ± 5.72%; cells co-transfected for 6 days, 11.83 ± 5.08%; Fig. 3f). To further verify that IL-23 regulates cell-cycle arrest via the Wnt/Notch signaling pathway, we measured the activation and increase in proteins expression at the G0/1 and G1/S checkpoints, and added ICG-001 (binding to CREB-binding protein) and DAPT (a γ-secretase inhibitor) to CD133−IL-23R+ ESCCs. The key regulator p21 was significantly upregulated in response to IL-23 treatment, whereas the level of Cdk2 decreased, which arrested cells in the S phase that was blocked by ICG-001. After cultured with IL-23, increased p16INK4a expression and downregulated Cdk4/6 expression were observed, which arrested most cells at the G0/1 phase. However, the above changes in p16INK4a and Cdk4/6 were abrogated when cells were pretreated with ICG-001 or DAPT, and IL-23 downregulated the phosphorylation of Rb but no longer produced significant changes in p53, Cyclin D1, and Cyclin D3 expression (Fig. 3g). Thus, we found that IL-23 could induce IL-23R+ ESCCs into dormancy and ultimately radioresistance.
IL-23 mediates radiation resistance in human ESCC xenografts
Comparisons and analyses of the number of stem-like or CD133+ cells in tumor and cancer cell lines have confirmed the existence of the innate cancer stem-like cells [5, 6, 8, 13]. Dormancy that prevents normal stem cell exhaustion is characteristic of cancer stem-like cells and allows proliferation to resume, and also regulates proliferation and differentiation in the tumor [11, 18, 27]. By entering a quiescent state, cancer cells can survive therapy. Moreover, tumor dormancy is a state of clinical remission, and a quantitative analysis of ovarian tumors showed that only approximately 30% of all tumor cells progressed through the cell cycle; similar observations have been made in melanoma, prostate, and breast cancer [28, 29, 30, 31]. Recent studies show that the existing immunotherapies primarily arrest the growth instead of eliminating the cancer, and immuno-environmental disturbances are not only involved in the malignant pathology of tumors but also directly and continuously obstruct the therapeutic efficacy . Among this complex network, TAMs are closely related to tumor masses from either circulating monocytes or tissue-resident macrophages, which clearly interact with a wide range of growth factors and cytokines, such as IFN-γ, IL-6, and IL-23, even their function is controversial. Our data revealed that M1 macrophages infiltrated ESCC tissues, and IL-23 exacerbated the malignancy of IL-23R+ ESCCs. As confirmed in our previous study, excessive IL-23 levels result in EMT, and the biological link between EMT phenotypes and cancer stem-like cells or cancer stem cells (CSCs) has recently been evidenced in many types of cancer, and a recent study of prostate cancer showed that CSCs undergo EMT to become resistant to radiation therapy [14, 32]. In the present research, the binding of IL-23 to IL-23R+ ESCCs strongly activated p-Stat3, which resulted in a signaling cascade that supported stem-like features, such as Wnt and Notch signaling. In contrast, Het-1A cells are more invalid and inertia to the same IL-23 treatment probably for the lower population of IL-23R. The observed self-renewal, tumorigenicity, and dormancy confirms IL-23 as an indispensable factor in the tumor microenvironment. Notably, not all IL-23R+ cells can be transformed into CD133+ cells. In other words, we concluded that the formation of stem-like properties in ESCCs is inextricably linked to the quantity of internal signal influence.
Our previous results suggested that exogenous IL-23 influenced the GSK-3β-mediated distribution of β-catenin . According to other research, the levels of Wnt1 and β-catenin in ESCC patients correlate with advanced TNM staging, lymph node involvement, and poor prognosis . Our study demonstrated that IL-23 not only contributed to the accumulation of β-catenin in the nucleus but also upregulated downstream genes, such as c-myc and Cyclin D1. Although Wnt signaling mainly plays a role in the maintenance of adult stem cells, recent studies have explored the integral role of Wnt signaling in cancer stem-like cells [19, 20]. However, in specific environments, Notch, which was identified to be mutually regulated by Wnt, participates in the stemness and differentiation . For example, Notch signaling is likely to inhibit specific differentiation lineages and to maintain the sustainability of self-renewal tumor populations in intestinal epithelial cells, hematopoietic stem cells and esophageal adenocarcinoma cells, and these features are characteristics of stemness [35, 36, 37]. Moreover, some studies have indicated that Notch1 might be a molecular prognostic marker and an effective therapeutic target for patients with ESCC . Conversely, other studies suggested that Notch1 is a tumor suppressor gene in the esophagus . Interestingly, we found that IL-23 consistently stimulated Notch signaling, which was accompanied by apparent increases in p16INK4a and Hes1. Specifically, the molecular interactions of the Wnt and Notch pathways initiated by IL-23 are likely to be cell-context specific and may be a unique model for resistance.
The clinical outcome of radiation and response to immunogenicity, or biomarkers predicting radiation for tumor has revealed the networks of DNA damage response and repair and immune response signaling, and the activation of response and repair triggers innate and adaptive immune response, leading to cell-cycle arrest and immune defects . CSCs, cancer stem-like cells, or therapy-resistant cells in tumors are undoubtedly difficult to destroy, but all of these cells play a part in the regulation of the cell cycle. In other words, inhibiting the proliferation and reducing the exposure to free radicals is a superior approach that minimizes the damage of radiation to cancer cells. Here, we demonstrated the role of IL-23 in G0/1 phase arrest of IL-23R+ ESCCs; specifically, Notch induction led to cell cycle progression before the S phase, whereas Wnt resulted in cells at the G0/1 phase, and ultimately reduced apoptosis caused by radiation. Thus, we inferred that IL-23 could induce IL-23R+ cells into dormancy with stem-like cell markers rather than typical CSCs. The “love–hate struggle” of immune microenvironment and cancer cells is intertwined already; to be sure, the tumor also contains other stroma cells. Interestingly, irradiated fibroblasts can recover IL-23 secretion from irradiated dendritic cells through COX2-dependent PGE2 release . Furthermore, previous studies showed that IL-23 increased the abundances of intercellular adhesion molecule-1 and vascular cell adhesion molecule-1 on endothelial cells, and improved the endothelial marker CD31 in mammary cancer [42, 43]. In contrast, our findings reveal no significant difference in fibroblast and endothelial cells between indicated treatments, no matter if IL-23 treatment or irradiation, in xenograft models (Supplementary Fig. 4). The above observations suggest that the activation of Wnt and Notch signaling by IL-23 resulted in the stem-like properties of ESCCs, which led to self-renewal, tumorigenesis, prominent G0/1 phase arrest, and resistance to irradiation. Thus, selectively targeting cancer stemloids might improve the therapeutic response of patients with cancer.
The excellent technical assistance of Professor Yichuan Xiao (Shanghai Institutes of Biological Sciences, CAS, Shanghai, China) is gratefully acknowledged. National Natural Science Foundation of China (81572956) and the Jiangsu Provincial Key Research and Development Program (BE2017096) supported this work.
Compliance with ethical standards
The study protocols met the standards set by the Declaration of Helsinki and were approved by the Ethics Committee of the Affiliated Hospital of Jiangsu University (Zhenjiang, China). All methods for humans were performed in accordance with the relevant guidelines and regulations. All animal studies were reviewed and approved by the Institutional Animal Care and Use Committee of Jiangsu University (Zhenjiang, China).
Conflict of interest
The authors declare that they have no conflicts of interest.
- 2.Wu C, Hu Z, He Z, Jia W, Wang F, Zhou Y, Liu Z, Zhan Q, Liu Y, Yu D, Zhai K, Chang J, Qiao Y, Jin G, Liu Z, Shen Y, Guo C, Fu J, Miao X, Tan W, Shen H, Ke Y, Zeng Y, Wu T, Lin D (2011) Genome-wide association study identifies three new susceptibility loci for esophageal squamous-cell carcinoma in Chinese populations. Nat Genet 43:679–684CrossRefPubMedGoogle Scholar
- 11.Sansone P, Ceccarelli C, Berishaj M, Chang Q, Rajasekhar VK, Perna F, Bowman RL, Vidone M, Daly L, Nnoli J, Santini D, Taffurelli M, Shih NN, Feldman M, Mao JJ, Colameco C, Chen J, DeMichele A, Fabbri N, Healey JH, Cricca M, Gasparre G, Lyden D, Bonafe M, Bromberg J (2016) Self-renewal of CD133(hi) cells by IL6/Notch3 signalling regulates endocrine resistance in metastatic breast cancer. Nat Commun 7:10442. https://doi.org/10.1038/ncomms10442 CrossRefPubMedPubMedCentralGoogle Scholar
- 15.Mani SA, Guo W, Liao MJ, Eaton EN, Ayyanan A, Zhou AY, Brooks M, Reinhard F, Zhang CC, Shipitsin M, Campbell LL, Polyak K, Brisken C, Yang J, Weinberg RA (2008) The epithelial–mesenchymal transition generates cells with properties of stem cells. Cell 133:704–715CrossRefPubMedPubMedCentralGoogle Scholar
- 18.Braumuller H, Wieder T, Brenner E, Assmann S, Hahn M, Alkhaled M, Schilbach K, Essmann F, Kneilling M, Griessinger C, Ranta F, Ullrich S, Mocikat R, Braungart K, Mehra T, Fehrenbacher B, Berdel J, Niessner H, Meier F, van den Broek M, Haring HU, Handgretinger R, Quintanilla-Martinez L, Fend F, Pesic M, Bauer J, Zender L, Schaller M, Schulze-Osthoff K, Rocken M (2013) T-helper-1-cell cytokines drive cancer into senescence. Nature 494:361–365CrossRefPubMedGoogle Scholar
- 28.Schindlbeck C, Hantschmann P, Zerzer M, Jahns B, Rjosk D, Janni W, Rack B, Sommer H, Friese K (2007) Prognostic impact of KI67, p53, human epithelial growth factor receptor 2, topoisomerase IIalpha, epidermal growth factor receptor, and nm23 expression of ovarian carcinomas and disseminated tumor cells in the bone marrow. Int J Gynecol Cancer 17:1047–1055CrossRefPubMedGoogle Scholar
- 37.Wang Z, Da Silva TG, Jin K, Han X, Ranganathan P, Zhu X, Sanchez-Mejias A, Bai F, Li B, Fei DL, Weaver K, Carpio RV, Moscowitz AE, Koshenkov VP, Sanchez L, Sparling L, Pei XH, Franceschi D, Ribeiro A, Robbins DJ, Livingstone AS, Capobianco AJ (2014) Notch signaling drives stemness and tumorigenicity of esophageal adenocarcinoma. Cancer Res 74:6364–6374CrossRefPubMedPubMedCentralGoogle Scholar
- 39.Agrawal N, Jiao Y, Bettegowda C, Hutfless SM, Wang Y, David S, Cheng Y, Twaddell WS, Latt NL, Shin EJ, Wang LD, Wang L, Yang W, Velculescu VE, Vogelstein B, Papadopoulos N, Kinzler KW, Meltzer SJ (2012) Comparative genomic analysis of esophageal adenocarcinoma and squamous cell carcinoma. Cancer Discov 2:899–905CrossRefPubMedPubMedCentralGoogle Scholar
- 41.Malecka A, Wang Q, Shah S, Sutavani RV, Spendlove I, Ramage JM, Greensmith J, Franks HA, Gough MJ, Saalbach A, Patel PM, Jackson AM (2016) Stromal fibroblasts support dendritic cells to maintain IL-23/Th17 responses after exposure to ionizing radiation. J Leukoc Biol 100:381–389CrossRefPubMedPubMedCentralGoogle Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.