Introduction
Colorectal cancer (CRC) is one of the most common digestive malignant tumors, and over the past decade, the CRC incidence rate has declined because of the gradual increase in colonoscopy examinations. In contrast to the rapid decline in overall CRC incidence, the rates in individuals aged younger than 55 years have increased by almost 2% per year from the mid-1990s to 2014, with changes in lifestyle and diet [
1,
2]. It is well known that metastasis is the leading cause of death in patients with malignant tumors, and metastatic tumors are found in 40–50% of patients at initial diagnosis and treatment [
3‐
6]. The metastasis of CRC is a complex process and is regulated by both oncogenes and suppressor genes. However, at present, more research is focused on the function of metastasis-related oncogenes, while relatively few studies have addressed the mechanism of metastasis-related suppressor genes [
7,
8]. Therefore, it is essential to elucidate the function and molecular mechanisms of more metastasis-related suppressor genes underlying the metastasis of CRC from multiple regulatory levels using multiomics analysis.
An increasing number of studies have shown that multiple genetic and epigenetic changes are required for carcinogenesis and the progression of CRC [
9‐
11]. We now have an initial view of the landscape of genetic alterations, including microsatellite instability (MSI), chromosomal instability (CIN) and methylations, which occur in CRC [
12]. However, the biological functions of these features and how they initiate and promote the tumorigenesis and progression of CRC is largely unknown. Therefore, it is imperative to explore the characterization of transcriptomic subtypes of CRC and to identify genes with required expression for the proliferation and metastasis of CRC cells [
13].
DMTN is a transcriptional differentially expressed gene (DEG) that was identified using CRC mRNA sequencing data from The Cancer Genome Atlas (TCGA), and it maps to a region of the short arm of human chromosome 8p21.1, encoding the DMTN protein [
14]. DMTN is an actin-binding/bundling protein that was originally isolated from the human erythrocyte membrane, which directly binds F-actin through actin binding sites to regulate cytoskeleton remodeling [
15,
16]. The actin cytoskeletal rearrangements occur upon activation of the small family of Rho GTPases, RhoA, Rac1 and Cdc42, and the activity of the GTPases is regulated by guanine nucleotide exchange factors (GEFs), GTPase activating proteins (GAPs) and guanine dissociation inhibitors (GDIs) [
17‐
19]. Rho-Rac guanine nucleotide exchange factor 2 (ARHGEF2), which activates Ras homolog family member A (RHOA) and RAC1, has been implicated in various cellular processes involving the establishment of cell polarity, including epithelial tight junction formation and endothelial permeability [
18], and it is required for oncogenic RAS signaling [
20]. It is known that dysregulation in the structure and function of the cytoskeleton is an important factor in the development and progression of malignant tumors [
21‐
24], and the abnormal expression of actin-related proteins is also similar to the malignant phenotype of the transformation of normal cells [
25,
26].
DMTN is responsible for maintaining the shape and integrity of erythrocyte [
25,
27‐
29]. Biochemical characterization of DMTN shows that it exhibits phosphorylation-dependent actin bundling activity [
30]. DMTN is expressed predominantly in hematopoietic (erythrocytes, platelets, and lymphocytes), cardiac, vascular, endothelial, epithelial, skeletal and muscle components, as well as in kidney cells [
14,
15]. The broad expression of DMTN suggests that it may play an important role in the regulation of the actin cytoskeleton in nonerythroid cells [
31].
Some studies have shown that the dysregulation and deletion of DMTN are closely related to the carcinogenesis and metastasis of cancer. The DMTN gene maps to chromosome 8p21.1, a region that is often accompanied by the loss of heterozygosity in prostate cancer patients [
32]. In prostate cancer PC-3 cells, the overexpression of DMTN restores epithelioid cell morphological phenotypes. Recent studies have shown that DMTN regulates cell shape, motility, and wound healing by modulating RhoA activation [
15]. Our preliminary work suggested that DMTN expression was downregulated in CRC, and the result of a Gene Set Enrichment Analysis (GSEA) assay also suggested that the Rac1 signaling pathway was significantly enriched in CRC tissues with low DMTN expression. Considering the above results, we speculate that the dysregulation of DMTN may function as a tumor suppressor gene by regulating the activity of Rac1 signaling in the carcinogenesis, invasion and metastasis of CRC, which may be similar to the effect of truncated mutant adenomatous polyposis coli (APC) on the activation on Rac1 by relieving binding with Asef and Asef2 [
33,
34]. However, the specific functions and molecular mechanisms of DMTN in the progression of CRC and the mechanisms responsible for the downregulation of DMTN remain unclear.
Therefore, in this study, we detected DMTN expression in CRC tissues, analyzed the relationship between DMTN expression and the clinical pathologic parameters, illustrated the role and associated molecular mechanism of DMTN in the invasion and metastasis of CRC, and explored the upstream regulatory mechanisms underlying the downregulation of the DMTN gene.
Materials and methods
Clinical specimen
The clinical research on the samples was performed according to written approval obtained from the Southern Medical University Institutional Board (Guangzhou, China). All the specimens were collected with the informed consent of the patients. Paraffin-embedded archived CRC tissue samples (n = 200) were collected between 2008 and 2012, and 50 fresh CRC tissues and matched adjacent normal tissues were obtained between 2014 and 2015 from the Department of Pathology, Southern Medical University. The surgically resected tissues were frozen in liquid nitrogen immediately until future analysis. The medical records of the patients were reviewed for acquisition of the clinicopathological information, including age, gender, differentiation, and TNM stage. The survival data were available for the cohort of 200 patients, and the median follow-up time was 65 (range, 3–87) months.
Cell culture, vector construction and lentivirus infection
The human CRC cell lines were purchased from The Global Bioresource Center (ATCC, USA). HCT15 was cultured in RPMI 1640 medium (Gibco, Grand Island, NY, USA) containing 10% fetal bovine serum (Gibco, Grand Island, NY, USA). The SW620 and SW480 cells were cultured in Leibovitz’s L-15 medium supplemented with 10% FBS (Gibco). The HT29 and HCT116 cells were cultured in McCoy’s 5a medium modified with 10% FBS (Gibco). The Ls174t cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco) with 10% FBS (Gibco). All cells were cultured at 37 °C in 5% CO2. The vector construction and lentivirus infection were conducted according to previously described methods [
35,
36]. Further details are provided in the Additional file
1: Supplementary Materials and Methods section.
Transwell, wound-healing assay and three-dimensional morphogenesis assay
The transwell, wound-healing assay and the three-dimensional morphogenesis assay were conducted according to previously described methods [
35,
36]. Further details are provided in Additional file
1.
Real-time quantitative PCR, western blotting and immunohistochemistry
Real-time quantitative PCR (RT-PCR), western blotting (WB) and immunohistochemistry (IHC) were conducted according to previously described methods [
35,
36]. Further details are provided in Additional file
1.
Immunofluorescence
The cells were fixed, permeabilized, and blocked using 4% paraformaldehyde for 10 min, 0.5% Nonidet-P40 (Bioshop) for 20 min, and 10% FBS/PBS for 1 h. The antibody incubations were performed in blocking solution for 1 h, and the slides were mounted in Immuno-mount medium (Thermo Fisher Scientific). The immunofluorescence images were collected using the Olympia Deconvolution fluorescence microscope and softWoRx software (Applied Precision). The images were collected using a 100 × or 60 × 1.4 NA oil objective (Olympus).
Co-immunoprecipitation (co-IP) assay
A total of 1 × 106 cells were seeded on a 10-cm plate, and the DMTN plasmid was transfected into the CRC cells. After a 48-h incubation, the cells were washed with cold PBS. Then, ice-cold RIPA buffer was added to the plate, and the cells were scraped with a precooling spatula. The suspension was transferred to a new EP tube, shaken for 15 min and then centrifuged for 15 min at 1400 g at 4 °C. Protein An agarose was added to the protein (100 μl protein A agarose/1 ml protein). After 10 min of shaking at 4 °C, the samples were centrifuged for 15 min at 1400 g at 4 °C to remove the protein G beads. Then, anti-flag antibodies were incubated overnight with rotation at 4 °C. Protein A was added the following day to capture the antigen-antibody complexes. The complexes were incubated overnight with rotation at 4 °C or were rotated for 1 h at room temperature. After washing with wash buffer, the samples were centrifuged 6 times at 1400 g at 4 °C for 2 min each. The samples were boiled for 5 min and then washed with SDS-PAGE loading buffer. Finally, the samples were separated by SDS-PAGE and analyzed by immunoblotting.
GST-fusion proteins and GST-pull-Down assay
The pGEX-4 T-2 plasmid (Amersham) was used to construct the vectors expressing the GST-DMTN-GEF-binding and GST-DMTN-ARHGEF2-DH fusion proteins. The pcDNA3.1-DMTN and pcDNA3.1-ARHGEF2 plasmids served as the templates. The PCR conditions were 95 °C/1 min, 18 cycles of 95 °C/30 s, 55 °C/1 min, and 68 °C/1 min. After digestion, the fragments were subcloned individually in frame with respect to GST into pGEX-4 T-2. GST-DMTN-GEF-binding and GST-DMTN-ARHGEF2-DH were expressed and purified according to the manufacturer’s instructions (Pierce Biotechnology). The purified proteins were used as bait protein for the pull-down assay.
GTPase activation assay
GST-PBD (p21-binding domain of PAK) was used for the Rac1 activity assays [
17]. The CRC cells were transfected with DMTN overexpression, knockdown or control vector. After 72 h, the cells were washed twice with ice-cold PBS and lysed in ice-cold Mg
2+ lysis buffer. The cell lysates were centrifuged for 5 min at 13,000 g at 4 °C, and 40 μl of the supernatant was removed to determine the total Rac1 levels. The remaining supernatants were incubated with GST-PBD on glutathione-sepharose beads and rotated at 4 °C for 2 h. The beads were washed extensively in lysis buffer, and the bound proteins were separated by SDS-PAGE and immunoblotted with anti-Rac1 antibodies.
Bisulfite genomic sequence (BSP) assay
The DMTN CpG island was searched in the NCBI Gene database and the UCSC Genome Browser on the Human February 2009 Assembly (hg19; ref. [
18]). The genomic DNA was treated with sodium bisulfite and subjected to PCR, using primer sets designed to amplify the regions of interest (Additional file
1: Table S3). Bisulfite-sequencing analysis was carried out.
A surgical orthotopic implantation mouse model of CRC was performed as previously described [
37]. The cells (2 × 10
6 per mouse) were subcutaneously injected into the right dorsal flank of female BALB/c athymic nude mice (4–6 weeks of age, 18–20 g) obtained from the Animal Center of Southern Medical University, Guangzhou, China. Two weeks later, the animals were sacrificed, and the tumors were excised. A portion of the tumor was fixed in 10% formaldehyde, embedded in paraffin, cut into 5 μm sections and subjected to IHC using an anti-DMTN antibody (Abcam, Cambridge, MA, USA) or hematoxylin-eosin (H&E) staining. Another portion of the tumor was divided into small pieces of approximately 1 mm in diameter. The surgical orthotopic implantation of the CRC tumor fragments was performed in nude mice after anesthesia. The mice were killed 100 days after surgery, individual organs were excised, and metastases were observed by histological analysis. The orthotopic mouse metastatic model was carried out according to previously described methods [
35,
36]. All mice were housed and maintained under specific pathogen-free conditions and used in accordance with institutional guidelines with approval by the Use Committee for Animal Care. Further details are provided in Additional file
1.
Statistical analysis
The data were analyzed using SPSS 19.0 for Windows. The Mann-Whitney U-test and Spearman’s correlation analyses were applied to analyze the relationship between the expression of DMTN and the clinicopathological features of the CRC cases. A two-tailed paired Student’s t-test was used to compare the two experimental groups. The 5-year overall survival curves were plotted by the Kaplan-Meier method and compared with the log-rank test. A Cox proportional hazard regression model was established for the multivariate analysis of the combinatorial contribution of DMTN and the clinicopathological features to the survival of the patients. P < 0.05 was considered significant.
Accession numbers for the data sets
The GEO database (GSE17538, GSE17536, GSE17537, and GSE16125) and the TCGA data were used to analyze the relationship between the expression of DMTN and the 5-year overall survival of the CRC patients. The GEO databases (GSE13294, GSE32896, GSE13067, GSE7208, and GSE35896) were used for the GSEA analysis of the “KEGG_COLORECTAL CANCER”, “RHO_GTPASES” and “Rac1 signaling pathways” gene sets in the study.
Discussion
It is necessary to identify genes with required expression for the proliferation or survival of tumor cells, and cancer dependencies represent targets for therapeutic efforts [
13]. DMTN is a transcriptional differentially expressed gene (DEG) that was identified using CRC mRNA sequencing data, but little research has been conducted to examine the relationship between the abnormal expression of DMTN and tumorigenesis. Our results showed that DMTN was significantly downregulated in CRC tissues, and its expression level was closely related to advanced progression and a poor prognosis in CRC patients. The expression pattern suggested that DMTN might function as a tumor suppressor gene during the carcinogenesis and progression of CRC. Consistent with our results, Lutchman, M, et al. observed a frequent loss of DMTN in prostate cancer, even for metastatic high-grade prostate cancer, using fluorescent in situ hybridization (FISH), and concluded that the role of DMTN in tumorigenesis may be similar to that of the tumor suppressor gene NF2 in neurofibromatosis based on DMTN functional studies [
32].
Metastasis is a very complex biological process that involves a decrease in adhesion between tumor cells, an increase in cell migration ability, and a reconfiguration of the cytoskeleton, among others. DMTN plays an important role in maintaining cell morphological integrity and regulating cell movement. Mohseni, M. et al. found that the migration speed of mouse embryonic fibroblasts (MEFs) with DMTN HP domain knockout (HPKO) was significantly slowed, and the skin wound healing of the HPKO mice was delayed [
15]. However, fewer reports have addressed the effect of DMTN dysregulation on the invasion and metastasis of tumor cells. Our results showed that the overexpression of DMTN inhibited the migration and metastasis of CRC cells, while the knockdown of DMTN promoted tumor cell migration and metastasis.
Rho GTP family proteins belong to the Ras superfamily, and GTPase activity is regulated by GEF, GDI and GAP [
19,
38,
39]. Rac1 is an important member of the Rho GTPase family, and the abnormal activation of Rac1 is closely related to initiation, invasion and metastasis in a large number of tumors [
17,
40,
41]. Our previous GSEA assay results suggested that the “RHO_GTPASES” and “Rac1 signaling pathways” gene sets were significantly enriched in CRC tissues with low DMTN expression. Moreover, recent studies have also demonstrated that DMTN regulates the activity of Rho GTPase. Mohseni, M et al. found that the headpiece domain of DMTN regulates cell shape, motility, and wound healing by modulating RhoA activation [
14,
15]. Lutchman, M, et al. revealed that DMTN interacts with the Ras-guanine nucleotide exchange factor Ras-GRF2 and modulates mitogen-activated protein kinase pathways, and the expression level of the active form of Rac1 (Rac1-GTP) was also slightly increased after the knockdown of DMTN [
16]. However, the specific molecular mechanism of DMTN downregulation in the regulation of Rac1 signaling pathway activity remains unclear.
The tumor suppressor APC is mutated in sporadic and familial colorectal tumors, and wild-type APC interacts with the Rac1-specific guanine–nucleotide exchange factor (GEF) Asef and Asef2. In contrast, the truncated mutant APC activates Asef and Asef2 and then induces an increase in the levels of the active forms of Rac1 and Cdc42 to regulate cell adhesion and migration [
33,
34,
42,
43]. Interestingly, our study showed that DMTN may play a similar role to APC during the invasion and metastasis of CRC. Our bioinformatics analysis and in vitro and in vivo experiments showed that DMTN inhibited the activity of Rac1 by interacting with ARHGEF2, but the downregulation of DMTN relieved the binding to the ARHGEF2 protein and then enhanced Rac1 signaling pathway activity, stimulated lamellipodia and protrusion formation, and promoted the CRC cell metastasis.
The inactivation of tumor suppressor genes is an important factor in tumorigenesis. The underlying mechanisms mainly consist of gene deletions and mutations, epigenetic changes, and posttranscriptional regulation. For example, numerous deletions and mutations of tumor suppressor genes, such as APC, p53, DCC, and PTEN, occur during the progression of CRC. Indeed, Lutchman, M, et al. have suggested a frequent loss of DMTN in prostate cancer [
32]. However, our TGCA data analysis indicated that the deletion rate of DMTN was only approximately 6.3% in CRC tissues, which suggested the participation of other factors in the downregulation of DMTN. It is well known that the hypermethylation of tumor suppressor genes is an important factor leading to the decline in their expression. We observed a higher degree of CpG island methylation in DMTN in CRC tissues than in normal intestinal mucosal tissues, and the degree of methylation of the CpG islands was negatively correlated with the expression level of DMTN. These results suggested that hypermethylation of the gene promoter might not be a major factor in the inactivation of DMTN.
Acknowledgments
We thank the Central Laboratory of Southern Medical University for technical supports.