FRAS1-related extracellular matrix protein 2 drives the progression and poor prognosis of IL-1β-associated esophageal squamous cell carcinoma

Background Chronic inflammation generates a tumor-supporting microenvironment, but it remains unclear whether and how the persistent inflammation drives genetic abnormalities leading to the occurrence and progression of esophageal squamous cell carcinoma (ESCC). Methods By global RNA-sequncing, we screened genes related to ESCC with stimulation of the pro-inflammatory factor, interleukin (IL)-1β, and identified FRAS1-related extracellular matrix protein (FREM)2 as a oncogene. Flow cytometry was used to detect the expression of IL-1β and its receptor IL-1R1 in fresh ESCC specimens. The interaction between FREM2 and IL-1β signal was examined in vitro and in vivo. Levels of FREM2 and IL-1R1 were determined in ESCC tissue arrays derived from 299 patients, and their correlation with clinical outcomes was analyzed. Results Multiple genes-related to ESCC occurrence and recurrence were elevated when exposed to the persistent stimulation of IL-1β. Among them, FRAS1-related extracellular matrix protein 2 (FREM2) was identified as a new oncogene in ESCC. IL-1β and its receptor IL-1R1 highly expressed in ESCC, especially in tumor cells. FREM2 was induced by IL-1β, and in turn bound to and stabilized IL-1R1, facilitating IL-1β signal transduction. The activation of downstream NK-κB and JNK signals mediated the tumor-promoting effect, while the reduction of FoxP1 was responsible for IL-1β-induced FREM2 transcription. High levels of FREM2 and IL-1R1 synergistically indicated the shorter survival time in patients. Conclusion These results suggest FREM2 is an IL-1β-stimulated oncogene, which as a cofactor of IL1R1 promotes IL-1β-induced ESCC progression. Therapeutic

strategies targeting FREM2 is likely to prolong the survival of ESCC patients.

Background
Esophageal squamous cell carcinoma (ESCC) accounts for over 90% of esophageal cancer cases [1,2]. Although previous studies have described complicated genetic and epigenetic alternations associated with ESCC progression [3], the underlying causes and mechanisms for these abnormalities have not been fully elucidated.
There is increasing evidence reveal that factors leading to chronic irritation and inflammation of the esophagus, such as smoking, alcohol, and hot drinks and food, have been considered to initiate esophageal squamous cell dysplasia. [4][5][6][7][8][9] Inflammation is acknowledged as one of important risks to induce ESCC. [10,11] Regardless of its origin, inflammation in the tumor microenvironment has many tumor-promoting effects. The cellular effectors and mediators of inflammation are important constituents of the local environment of tumors. A growing number of reports indicate that chronic inflammation in some types of tumor is positively correlated with genetic alterations, such as DNA damage and gene mutations and amplification. [12]; [13][14][15] In other types of cancer, carcinogenic changes, such as activation of oncogenes or silencing of tumor suppressive genes, can in turn induce an inflammatory microenvironment or modulate inflammation-associated transcriptional programs, thus promoting tumors development. However, it is unclear which inflammatory factors and how to cause genetic abnormalities to initiate ESCC.
Interleukin (IL)-1β, a well-known pro-inflammatory cytokine, is abundant in many types of tumor. Upon IL-1β binding to its receptors (e.g., interleukin 1 receptor, type I [IL-1R1]), a cascade of pro-inflammatory gene expression is activated. Elevated levels of IL-1β are correlated poor prognosis and shorter survival in patients with breast, lung, prostate, or pancreatic cancers as well as an increased risk of carcinogenesis. [16] This has been attributed to the induction of inflammatory responses, inhibition of the maturation and activation of tumor-infiltrating cells, or establishment of an immunosuppressive microenvironment. [17] However, there is little evidence that IL-1β directly causes genetic alterations in tumor. [18] In the present study, we identified IL-1β and its receptor IL-1R1 was abundant in ESCC, especially in advanced patients. By screening for genetic alterations of ESCC, we identified multiple genes associated with IL-1β. Among them, FRAS1-related extracellular matrix protein (FREM)2 as a novel oncogene modulated and was modulated by IL-1β signal. IL-1β induced FREM2 overexpression in ESCC, and FREM2 enhanced IL-1β signal transduction by binding to and stabilizing IL-1R1. Both increased FREM2 and IL-1R1 were related to reduced survival time in ESCC patients, serving as prognostic biomarkers. Separating FREM2-IL1R1 complex may block IL-1β triggered ESCC initiation and progress, and be of promising in ESCC therapy.

Clinical specimens
Paraffin-embedded specimens (n = 299) were collected at Zhongshan Hospital, Fudan University (Shanghai, China) in 2007 with patient consent and with the approval of the local ethics committee. None of the patients received chemotherapy or radiotherapy before surgery. The specimens were selected solely based on the availability of complete clinicopathological and follow-up data for the patients.
Overall survival was defined as the interval between surgery and death or between surgery and the last observation for surviving patients. Data were censored at the last follow-up for living patients.

Western blotting
Esophageal tissue and whole cell extracts were analyzed by western blotting according to a standard protocol. The primary antibodies used are listed in Additional file 1. Supplementary Table 1.

Co-immunoprecipitation (co-IP)
The potential interaction between FREM2 and IL-1R1 was evaluated by co-IP. ECa109 and ECa9706 cells were washed twice with ice-cold PBS and then lysed with radioimmunoprecipitation assay lysis buffer. After removing insoluble material by centrifugation at 12,000 × g, the pre-cleared lysates were incubated overnight at 4 °C with primary antibody against FREM2 or IL-1R1 and pre-absorbed protein A- with Welch's correction was used to analyze differences between two groups, and analysis of variance was used for comparisons between more than two groups.
Pearson's correlation coefficient was used to evaluate correlations between groups.
Cumulative survival time was calculated by the Kaplan-Meier method and was analyzed with the log-rank test. Uni-and multivariate analyses were based on the Cox proportional hazards regression model. P < 0.05 was considered statistically significant.

Results
FREM2 is an IL-1β-stimulated gene and associated with ESCC occurrence and recurrence We compared differentially expressed genes between ESCC tumor and corresponding non-tumor tissues by RNA-seq, and further screened ESCC-recurrence associated genes that were response to IL-1β stimulation. We identified 3111 differently expressed candidate genes in ESCC. Compared with our previous RNAseq data for recurrent vs. non-recurrent cases [19], 34 genes were found to be upregulated and 20 were downregulated in the ESCC tumor and recurrence groups relative to the non-tissue and non-recurrent groups respectively (Fig. 1A, Additional file 3. Supplementary Fig. 1A and Additional file 4. Supplementary Table 3). We searched the 34 upregulated genes in TCGA database and confirmed that 16 protein coding genes were significantly upregulated ( Fig. 1A and Additional file 3. Supplementary Fig. 1B).
To determine the relationship between IL-1β and these genes, we stimulated  (Fig. 1B). FREM2 was selected for further study because it was dramatically upregulated and there is little known about this gene in the context of ESCC. Results showed that IL-1β stimulation increased FREM2 protein level and the percentage of FREM2 + ESCC cells (Fig. 1C). Consistent with these observations, FREM2 expression was found to be increased in ESCC as compared to matched normal tissues no matter in mRNA level (Additional file 3. Supplementary Fig. 1C) or in protein level (Fig. 1D, Additional file 3. Supplementary   Fig. 1D). As well as, FREM2 was overexpressed in five ESCC cell lines relative to normal HEECs (Fig. 1E). The flow cytometry analysis demonstrated that the percentage of FREM2 + cells was higher in tumor as compared to adjacent normal tissues (Fig. 1F). These results suggest that FREM2 has an oncogenic role in ESCC.
IL-1β and its receptor IL-1R1 are abounded in ESCC and mainly expressed by tumor cells An analysis of TCGA data revealed that IL-1β level was elevated in ESCC as compared to normal tissues (Additional file 5. Supplementary Fig. 2A), which was further confirmed by our results of ELISA (Additional file 5. Supplementary Fig. 2B) and western blotting (Additional file 5. Supplementary Fig. 2C). A flow cytometry analysis revealed that IL-1β expression was higher in cluster of differentiation (CD)45 + and CD45 − cells of tumors ( Fig. 2A), and was higher in epithelial cell adhesion molecule (EPCAM) + epithelial cells than in CD45 + lymph cells (Fig. 2B).
These results indicated that IL-1β expression was elevated in ESCC tumor tissues, and that tumor cells, rather than lymphocytes, is the major source of IL-1β.
Since IL-1β exerts its effect through binding to IL-1R1, we evaluated IL-1R1 expression by western blotting and flow cytometry. Consistent with our observations for IL-1β, IL-1R1 was overexpressed in ESCC tissue and was more highly expressed in EPCAM + epithelial cells than in CD45 + lymph cells (Fig. 2C-D). To determine the role of IL-1β in tumorigenesis, we stimulated ECa109 ESCC cells with IL-1β and evaluated cell proliferation and cell migration. Results showed that IL-1β stimulation enhanced tumor cell proliferation as well as cell migration (Fig. 2E). These results suggest that IL-1β enriched in ESCC tissues plays a tumor-promoting role, which may be in both autocrine-and paracrine-manners.

FREM2 as a new oncogene is required for IL-1β-induced ESCC cell growth and migration
To determine the biological function of FREM2 and its relationship with IL-1β-related ESCC development, we constructed FREM2 knockout as well as overexpressed ECa109 and ECa9706 cell lines using the clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated (Cas)9 system (Additional file 6. Supplementary Fig. 3A-E). Functional analysis showed that loss of FREM2 suppressed ESCC cell proliferation and colony formation as well as migration ( Fig. 3A-C). Additionally, FREM2 knockout induced the apoptosis of ESCC cells (Additional file 7. Supplementary Fig. 4A), but it had no effect on the cell cycle (Additional file 7. Supplementary Fig. 4B). The effect of FREM2 knockout on ESCC tumor growth was further examined in a xenograft model. Compared to the control group, tumor growth was markedly inhibited in the FREM2 knockout group (Fig. 3D).
We also stably overexpressed FREM2 in ECa109 and ECa9706 cell lines (Fig. 3E), and confirmed that FREM2 overexpression increased cell proliferation and xenograft tumor growth (Fig. 3F-H). These results indicate that FREM2 functions as an oncogene in ESCC.
To further clarify the relationship between FREM2 and IL-1β, ESCC cells lacking FREM2 were stimulated with IL-1β. We found that ESCC cell proliferation and migration were both reduced under these conditions relative to control cells FREM2 enhances IL-1 signal transduction by binding to and stabilizing IL-1R1 Furthermore, we investigated the mechanism by which FREM2 promotes IL-1β signal and subsequent ESCC progression. Results showed that FREM2 + epithelial cells had consistently higher expression of IL-1R1 in ESCC tissues (Fig. 5A), and cell lines (Fig. 5B). Moreover, the increase in FREM2 expression induced by IL-1β stimulation ( Fig. 1C) was accompanied by upregulation of IL-1R1 (Additional file 9. Supplementary Fig. 6A-C). A scatterplot of FREM2 and IL-1R1 protein levels in ESCC tissues revealed a significant positive correlation (r = 0.7154, P = 0.046; Fig. 5C).
To examine the correlation between FREM2 and IL-1R1 in ESCC cells in greater detail, we performed co-IP using an anti-FREM2 or -IL-1R1 antibody. In reciprocal co-IP assays, the FREM2 antibody pulled down IL-1R1 and vice versa (Fig. 5D). These results suggest that endogenous FREM2 forms a complex with IL-1R1. An immunofluorescence analysis revealed that FREM2 and IL-1R1 were colocalized in ESCC cells. Meanwhile, the fluorescence intensity of the two proteins increased synchronously in ECa109 and ECa9706 cell lines following stimulation with 1.5 nM IL-1β (Fig. 5E, Additional file 9. Supplementary Fig. 6D). Upon IL-1β stimulation, the protein level of IL-1R1 in FREM2-overexpressed cells increased markedly and stayed for a longer time (Fig. 5F, Additional file 9. Supplementary Fig. 6E). These results indicate that FREM2 and IL-1R1 directly interact to promote ESCC progression upon activation of IL-1β signaling.
High levels of FREM2 and IL-1R1 correlate with shorter survival in patients To assess the clinical relevance of our findings, we carried out immunohistochemical analysis using tissue arrays of 299 ESCC patients (Table 1). Elevated expression of FREM2 and IL-1R1 was observed in 154 (51.5%) and 160 (53.5%) patients.
Upregulation of FREM2 was significantly correlated with advanced T stage (P = 0.002), lymph node metastasis (P = 0.041), and advanced clinical stage (P = 0.045). High IL-1R1 expression level was related to T stage (P = 0.022). It is worth noting that high expression of FREM2 was related to upregulation of IL-1R1 (P = 0.026).
FREM2 and IL-1R1 immunopositivity was almost exclusively confined to the tumor cells and was observed in the cell membrane and cytoplasm (Fig. 5G); the signal intensity was higher in ESCC than in non-tumor tissues. We divided the 299 patients into three groups: those with high expression of FREM2 and IL-1R1 (FREM2 high /IL-1R1 high ); high expression of either FREM2 or IL-1R1; and low expression of FREM2 and IL-1R1 (FREM2 low /IL-1R1 low ). Results showed that elevated expression of both FREM2 and IL-1R1 was correlated with T stage (P < 0.001), lymph node metastasis (P = 0.023), and clinical stage (P = 0.018).  promoter in ESCC cells (Fig. 6A). FOXP1 level was decreased relative to that in normal ESCC tissue (Fig. 6B, C). This inverse association between FREM2 and FOXP1 strongly suggests that FOXP1 negatively regulates FREM2 transcription.
The results of the ChIP assay confirmed that the direct interaction between FOXP1 and the FREM2 promoter was abrogated in the presence of IL-1β (Fig. 6D).
Furthermore, luciferase activity was decreased in cells transfected with FOXP1 overexpression or control plasmid under IL-1β stimulation (Fig. 6E). FOXP1 overexpression caused a reduction in FREM2 protein level, which was partly reversed in the presence of IL-1β ( Fig. 6F-G). These data indicate that IL-1β regulates FREM2 expression via FOXP1 in ESCC cells (Fig. 6H).

Discussion
Chronic inflammation contributes to tumor initiation and development. However, little is known about the interaction between inflammatory factors and oncogenes in ESCC. In the present study we identified FREM2 as a novel oncogene in ESCC that is upregulated by IL-1β and whose protein product forms a complex with IL-1R1.
Increased expression of FREM2 enhanced the activation of IL-1β downstream signaling by stabilizing IL-1R1. We also showed that the regulation of FREM2 In this study, IL-1β expression was higher in epithelium-derived tumor cells than in CD45 + lymphocytes in ESCC tissues. We speculate that this is not only due to increased lymphocyte infiltration but also because IL-1β production was enhanced, resulting in genetic alterations in tumor cells.

Conclusion
Clarifying the interaction between IL-1β and oncogenes can lead to improved strategies for ESCC prevention and treatment. In particular, our results suggest that as a novel oncogene that amplifies IL-1β signaling, FREM2 is a promising new target for blocking inflammation-associated ESCC progression.     FREM2 interacts with IL-1R1 to enhance IL-1 signal transduction, and predicts poor prognosis

Supplementary Files
This is a list of supplementary files associated with the primary manuscript. Click to download.