Colorectal cancer is the third most common cancer in men and women and the third leading cause of cancer-related deaths in the western world [
]. While screening programs have led to a reduction in colorectal cancer mortality, there is considerable room for improvement in identifying prognostic markers that predict which patients are at risk for metastatic disease [
]. International Union Against Cancer (UICC) stage III colorectal cancer is characterized by cancer spread to nearby lymph nodes and patients with stage III disease are generally at risk for recurrent disease or distant metastasis [
]. Therefore, most national guidelines recommend perioperative radiochemotherapy for management of patients with stage III colorectal cancer [
The homeodomain-containing transcription factor EMX2 (Empty Spiracles Homeobox 2) belongs to the Homeobox gene family which encodes transcriptional regulatory proteins that are essential for growth and differentiation [
]. EMX2 plays a pivotal role during brain development and homozygous EMX2 mutations in mice are associated with ectopic Wnt expression resulting in cortical dysplasia [
]. Aberrant signaling of homeobox genes has been shown in many types of cancer [
]. Accordingly, recent studies suggest a possible involvement of EMX2 in several human cancers including lung, endometrial and gastric cancer [
]. Moreover, EMX2 has been shown to be a predictive marker for survival in lung cancer [
]. However, to our knowledge, no study has evaluated the role of EMX2 in colorectal cancer thus far. In this study, we analyzed the specific expression of EMX2 transcripts in colorectal cancer and corresponding healthy mucosa from 31 patients and investigated putative clinical correlations. Moreover, we applied over- and underexpression of EMX2 in vitro in order to assess its functional significance in colorectal cancer spread.
Tissue samples of primary colorectal adenocarcinoma and corresponding healthy mucosa from a series of 31 patients suffering International Union Against Cancer (UICC) stage III colorectal cancer were included in this study. Samples from 29 separate patients suffering stage IV colorectal cancer were used for additional expression analyses in liver metastases. All patients underwent surgery at the Department of General, Visceral and Transplantation Surgery, University of Heidelberg, Germany, between November 2005 and March 2013. Written informed consent was obtained from all patients. The study was approved by the local ethics committee. Clinical characteristics like gender, age at surgery, tumor location, histopathologic diagnosis including tumor, node, metastasis classification system and International Union Against Cancer (UICC) stage, R classification, perioperative radiochemotherapy and overall survival (time from operation up to death or last follow-up) were obtained from each patient. Exclusion criteria for tissue samples were a histopathological type that was not adenocarcinoma, patients who had synchronous metastasis or patients who had a histopathology-negative lymph node status.
Total RNA was extracted using RNeasy Mini Kit (Qiagen, Hilden, Germany). cDNA synthesis and real-time PCR were performed with a first strand cDNA Synthesis Kit (Thermo Fisher, Karlsruhe, Germany) and LightCycler 480 SYBR Green I Master (Roche Diagnostic GmbH, Germany) using specific primers. Gene expression levels were normalized to the housekeeping gene GUS for each sample. Expression of transcript levels in human cancers were calculated as the level of gene expression in each sample relative to the level of gene expression in the adjacent normal mucosa using the comparative 2
], whereby “overexpression” indicates overexpression relative to the adjacent normal mucosa, whereas “underexpression” indicates underexpression relative to the adjacent normal mucosa. Primer sequences were as follows: GUS (forward) 5’-GATCCACCTCTGATGTTCACTG-3′; GUS (reverse) 3′-TTTATTCCCCAGCACTCTCG-5′; EMX2 (forward) 5′-GCTTCTAAGGCTGGAACACG-3′; EMX2 (reverse) 3′-CCAGCTTCTGCCTTTTGAAC-5′.
Cell culture, adenoviral infection and siRNA transfection
Human colorectal cancer cell lines DLD1 (ATCC® CCL-221™) and CaCo2 (ATCC® HTB-37™) were obtained from ATCC (Manassas, VA, USA) and maintained in basal medium supplemented with 10% FBS and 1% penicillin/streptomycin at 37 °C and 5% CO
2. Cell lines were routinely tested for mycoplasma. All cell lines were free of contaminants. Adenovirus was used for restoring expression of EMX2. Adenovirus used to express EMX2 (Ad-EMX2) or control (Ad-Null) was purchased from Vector Biolabs (PA, USA). Conversely, for knockdown of EMX2, transfection of cells was performed with EMX2 siRNA using HiPerFect (both Qiagen, Hilden, Germany).
For migration assays, a modified Boyden chamber assay (Greiner Bio-one, Germany) was used. Migration inserts were coated with Matrigel (250 μg/mL; BD Biosciences) and FBS was used as a chemoattractant. Migrated cells were quantified after 24 h by dissolving cell-bound crystal violet in 10% acetic acid. Results were normalized to cell proliferation, which was determined in parallel using the cell proliferation reagent WST1 (Roche Diagnostic GmbH, Germany) according to the manufacturer’s instructions.
Whole cell lysates were prepared in RIPA lysis buffer (Merck Millipore, Germany). Anti-EMX2 antibody (ab174897, Abcam, 1:500) and anti-Vinculin antibody (ab18058, Abcam, 1:2000) was used to detect EMX2 and Vinculin, respectively. Horseradish peroxidase-conjugated goat anti-rabbit secondary antibody (sc-2004, Santa Cruz, 1:1000) and goat anti-mouse antibody (sc-2005, Santa Cruz, 1:2000), respectively, was used to label the primary antibodies. SuperSignal TM West Dura Extended Duration Substrate was used as the chemiluminescence substrate. Chemiluminescence images of the western blots were recorded using an ultra-sensitive camera detection platform from Fusion systems (Vilber Lourmat Deutschland GmbH). Semi-quantitative analyses of the resulting images were performed applying ImageJ software (National Institutes of Health, Bethesda, USA).
For histological assessment of EMX2 expression, paraffin-embedded tissues were sectioned at 6 μm thickness, deparaffinized with xylene and rehydrated in a graded series of alcohols. For immunohistochemical staining, sections were blocked, and incubated overnight using a rabbit polyclonal anti-EMX2 primary antibody (Thermo Fisher, 1:200) at 4 °C in a humidified chamber. The slides were then treated with biotinylated horse anti-rabbit HRP conjugated secondary antibody (Vector Laboratories, 1:200). The sections were counterstained with hematoxylin. Images were captured using a Zeiss Axiostar Plus microscope equipped with an Axiocam MRC camera (Zeiss, Jena, Germany).
Statistical analyses were conducted with Excel 2010 (Microsoft, Redmont, WA, USA) and SPSS version 21 (IBM, Armonk, NY, USA). Expressional changes were assessed by the two-tailed student’s t-test. Univariate analysis was performed using the log-rank test and Fisher’s exact test. The Kaplan-Meier method was used to estimate cancer-related survival. Differences between survival curves were evaluated by log-rank test. The Cox proportional hazard model was used to calculate survival related hazard ratios. Results were considered significant at a
P value ≤0.05.
The homeodomain-containing transcription factor EMX2 was first described as an important mediator in embryonic development as EMX2 has been shown to play a role in neuroblast proliferation, migration and differentiation [
]. While transcriptional targets of EMX2 remain largely unidentified, loss of EMX2 function is associated with impaired development of the cortex [
]. Moreover, homozygously EMX2-deficient mice have been shown to display ectopic Wnt expression [
]. Wnt is an oncogene and its signaling represents an early event in a majority of colorectal cancers [
]. Wnt pathway signaling has also been associated with metastatic spread and stemness in CRC [
Recently, the role of EMX2 has been explored in various solid tumors. Several lines of evidence suggest that EMX2 is down-regulated in lung cancer [
]. While the mode of EMX2 down-regulation in non small-cell lung cancer (NSCLC) has been identified as epigenetic silencing, restoration of EMX2 has been shown to antagonize Wnt and to restore sensitivity to cisplatin [
]. More recently, loss of EMX2 expression has been demonstrated in gastric cancer cell lines and primary gastric cancer. EMX2 down-regulation was associated with promoter hypermethylation and the adenoviral delivery of EMX2 in a mouse model of gastric cancer significantly suppressed tumor growth [
]. Thus, these previous observations support our present finding that down-regulation of EMX2 has protumorigenic effects in colorectal cancer.
To our knowledge, this is the first investigation of the functional role of EMX2 in colorectal cancer. In our patient collective, EMX2 was frequently down-regulated in tumor tissue in comparison to matched normal mucosa samples. Down-regulation of EMX2 was associated with metastatic tumor progression and decreased overall survival. These results differ from a previous study, where loss of EMX2 was found in only 2–5% of colorectal cancers [
]. In contrast to the latter study, we found a down-regulation of EMX2 in 54% of our patient collective. These observed differences can in part be explained by the different methods that were applied to assess EMX2 expression. While Kim et al. used immunohistochemical analysis, we applied qRT-PCR to detect EMX2 expression levels [
]. We then normalized expression levels of EMX2 in tumors to expression levels in adjacent normal mucosa. We believe that qRT-PCR is a robust and sensitive tool for quantification of gene expression levels.
Nevertheless, our study has several limitations. First, the power to make statistical inferences was limited by a modest sample size. The main reason for this is the limited number of cases where both tumor and normal mucosa samples were available for RNA extraction, along with appropriate follow-up data allowing for the detection of metastatic disease and assessment of survival. Second, although colorectal liver metastases showed a further decrease in EMX2 expression levels when compared to primary colorectal cancer samples, it remains unknown whether this is due to an evolutionary decrease of EMX2 expression from primary tumor to metastasis and therefore a possible driver or prerequisite for metastatic outgrowth or whether this is a mere coincidence. Metastatic progression of cancer is a complex process involving a multi-step process, where migration represents a key element in the process of the metastasic cascade [
]. The functional assays in this study assessing tumor cell migration suggest a potential role for EMX2 in metastatic disease progression since EMX2 knockdown resulted in increased migration while restoration of EMX2 using an adenoviral vector led to decreased migration. We used a recombinant human adenovirus type 5 as delivery system since this is the vector of choice for functional genomics research [
]. While there are still many challenges that need to be overcome before adenoviral vectors can be safely used in cancer patients, adenoviral vector-based therapeutic strategies represent a promising tool for cancer gene therapy [
]. Altogether the role of EMX2 expression in metastatic spread would be a valuable area of future research in a larger patient cohort.
In summary, our data encourage a significant role of EMX2 in the progression and metastasis of colorectal cancer. Our study demonstrates that a low EMX2 expression level is an independent prognostic factor and correlates with dismal prognosis, decreased overall survival and the development of metastatic disease in stage III colorectal cancer patients. Thus, the study at hand provides a first evidence for the role of EMX2 as a suppressor of metastasis in colorectal cancer. Further, we provide evidence that EMX2 has predictive value as a prognostic factor in stage III colorectal cancer as well as a possible functional role in metastatic spread. Therefore, restoration of EMX2 via gene therapy may represent a promising therapeutic strategy for tailored anticancer therapy.
We thank Mareen Dupovac, Melanie Höfler, Vishnu Mani and Nicholas Rüdinger for their technical assistance. We also thank Ulf Hinz for his statistical support.
This study was funded by the Heidelberger Stiftung Chirurgie (to B. Aykut) and by the Deutsche Forschungsgemeinschaft, Germany (to M. Schneider, grant SCHN 947/4–2 in the framework of the Clinical Research Group KFO 227). The funders have no role in the study design and data collection, analysis, and interpretation as well as manuscript writing.
Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.
This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (
), 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. The Creative Commons Public Domain Dedication waiver (
) applies to the data made available in this article, unless otherwise stated.