DDA1 promotes stage IIB–IIC colon cancer progression by activating NFκB/CSN2/GSK-3β signaling

Overexpression of DDA1 and activation of NFκB are negatively correlated with stage IIB–IIC colon cancer patient survival

To investigate whether DDA1 alone or in combination with NFκB predicts the risk of tumor recurrence in patients with stage II colon cancer, levels of DDA1 and nuclear p65 were assessed. DDA1 mRNA expression was greater in the tumors of all patients with stage II colon cancer than in adjacent normal tissues of 30 randomly selected patients from the cohort (Figure 1A). DDA1 expression was significantly higher in the tumors of relapsed patients than in the tumors of nonrelapsed patients (Figure 1B). Further, DDA1 protein levels and the nuclear translocation of p65 protein, an indication of activation of NFκB, were also higher in tumor samples than in adjacent normal tissues, and in relapsed patient samples than in nonrelapsed samples (Figure 1C and 1D). Taken together, these results suggest that DDA1 upregulation and NFκB activation may be associated with both development and recurrence of stage II colon cancer.

Figure 1

DDA1 and nuclear p65 levels in relapsed and nonrelapsed stage II colon cancer tumors

(A) DDA1 mRNA expression was higher in tumor tissues (T) of 30 patients randomly selected from the 279 patients with stage II colon cancer than in normal tissues (N) (**P < 0.001). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as a control. (B) DDA1 mRNA expression was higher in tumor tissues of the 30 randomly selected patients with recurring tumors (R) than nonrecurring tumors (NR) (**P < 0.001). (C) DDA1 and nuclear p65 protein levels were higher in the represented tumor tissues (T) than normal tissues (N). (D) DDA1 and nuclear p65 protein levels were higher in recurring tumors (R) than nonrecurring tumors (NR). (E) IHC staining of DDA1 and nuclear p65: negative DDA1 in normal tissues (E1) and positive expression in carcinoma tissues (E2). p65 was more frequently expressed in the nuclei of tumor cells that showed positive cytoplasmic DDA1 staining (E3, E4). In well-differentiated tumors of the same T3 stage, DDA1 showed weak staining in patients with non-recurrence (E5) and strong staining in patients with tumor recurrence (E6). In moderately-differentiated tumors of the same T4a stage, DDA1 showed weak staining in patients with non-recurrence (E7) and strong staining in patients with tumor recurrence (E8). In poorly-differentiated tumors of the same T4b stage, DDA1 showed weak staining in patients with non-recurrence (E9) and strong staining in patients with tumor recurrence (E10) (magnification ×200).

Immunohistochemical (IHC) staining was used to quantify DDA1 levels as well as NFκB pathway activation, as measured by p65 nuclear translocation, in 101 relapsed and 178 nonrelapsed patients with stage II colon cancer. DDA1 showed mainly positive staining in the cytoplasm of tumor cells and was often accompanied by NFκB subunit p65 nuclear translocation in the tumor cells of relapsed patients with stage II colon cancer (Figure 1, 1E1–1E4). In patient tumors with the same degree of differentiation, the relapsed patients had significantly higher expression of DDA1 than nonrelapsed patients (Figure 1, 1E5–1E10). A significantly higher proportion (65.4%) of relapsed patients showed strong tumor DDA1 staining as compared to nonrelapsed patients (29.8%) (Table 1). The number of relapsed patients with weak (28.7%) and negative (5.9%) tumor DDA1 staining was significantly lower than the number of nonrelapsed patients with weak (40.4%) and negative tumor staining (29.8%) (Table 1). Relapsed patients (60.4%) exhibited higher nuclear p65 staining as compared to nonrelapsed patients (36.5%) (P < 0.001, Table 1). No significant association was observed between DDA1 or nuclear p65 staining and age, gender, location of the tumor, or vessel invasion (P > 0.05, Table 2).

DDA1 and nuclear p65 staining were positively correlated with pT stage, American Joint Committee on Cancer (AJCC) stage and tumor differentiation in patients with stage II cancer (P < 0.05, Table 2). Further, 60.5% (72/119), 37.6% (38/101) and 27.1% (16/59) of patients with nuclear p65 staining exhibited high, weak, and negative DDA1 expression, respectively (P < 0.001, Table 2). Taken together, these results indicate that DDA1 upregulation and NFκB activation are related to recurrence in stage II colon cancer, and DDA1 contributes to the activation of NFκB.

Kaplan–Meier analysis with the log-rank test was further used to investigate whether these changes are associated with patient survival. DDA1 expressing patients had lower DFS and OS than those without DDA1 expression (Figure 2A). Nuclear translocation of p65 did not significantly impact patient OS (Figure 2B). These results indicate that DDA1 has more power to predict survival than nuclear p65 translocation. However, patients with both DDA1 expression and nuclear p65 translocation had worse rates of DFS and OS than patients with either DDA1 expression or nuclear p65 translocation, and then patients negative for both (Figure 2C).

Figure 2

DDA1 expression predicts poor prognosis in patients with stage II colon cancer undergoing post-operative 5-FU-based adjuvant chemotherapy

(A) Impact of DDA1 expression on patients with stage II. (B) Effect of nuclear p65 levels on patient DFS and OS. (C) Impact of DDA1 expression in combination with nuclear p65 on patient DFS and OS.

These results prompted us to further investigate whether these markers have any predictive value for substages, namely, AJCC stage. DDA1 expression together with p65 nuclear translocation correlated with reduced survival in patients with stage IIB (DFS: P = 0.005; OS: P = 0.005, Figure 3A) or stage IIC colon cancer (DFS: P = 0.021; OS: P = 0.011, Figure 3B), but not in patients with stage IIA colon cancer (DFS: P = 0.059; OS: P = 0.063, Figure 3C). Univariate and multivariate analyses of DDA1 expression, nuclear p65 translocation, and clinical features such as gender and age, and clinical pathological stages further confirmed that DDA1, either alone or in combination with nuclear p65 translocation is an independent prognostic factor for high risk of tumor recurrence (Table 3).

Figure 3

Kaplan-Meier survival analyses for stage II colon cancer patients undergoing post-operative 5-FU-based adjuvant chemotherapy

Effect of DDA1 in combination with nuclear p65 expression in patients with stage IIB (A), stage IIC (B), and stage IIA (C) colon cancer.

DDA1 accelerates colon cancer cell proliferation and promotes cell cycle progression

Analysis of Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways suggest that DDA1 may be involved in the development of colorectal cancer, focal adhesions, apoptosis, and activation of the NFκB signal pathway (Figure S1 and S2). Therefore, the effects of DDA1 protein on colon cancer cell proliferation and cell cycle progression were investigated using overexpression and knockdown. As no cell lines were from patients with IIB – IIC colon cancer, the protein levels of DDA1 from eight general colon cancer cell lines and two normal colon epithelial cell lines were analyzed. The DLD-1 and HT-29 cell lines with low DDA1 expression and SW480 and SW620 cell lines with high DDA1 expression were studied (Figure 4A). We over-expressed DDA1 in low-expressing lines with lentiviral expressing (LV)-DDA1 vector (LV-DDA1), or knocked down DDA1 in high-expressing lines with lentiviral expressing two different shRNAs (shDDA1#1 and shDDA1#2) (P < 0.05, Figure 4B). In vivo, the overexpression or knockdown of DDA1 in cells generated larger or smaller xenografts, respectively, as measured by tumor weights and volumes in nude mice as compared to controls (P < 0.05, Figure 4C). Consistently, DDA1 knockdown with shDDA1#2 suppressed the cancer cell proliferation in vitro and in vivo (Figure S3A and S3B).

Figure 4

DDA1 facilitates colon cancer cell proliferation and cell cycle progression in vitro and in vivo

(A) DDA1 expression was assessed using western blotting in eight colon cancer cell lines and two normal colon epithelial cell lines. (B) Overexpression or knockdown of DDA1 accelerated or inhibited the proliferation of colon cancer cells, respectively. (C) HT-29 and SW480 cell lines were subcutaneously injected into nude mice; tumor weight and volume growth curves are shown at three weeks post-injection (n = 5, *P < 0.05). (D) FACS analyses of cells in every phase of the cell cycle after transfection with LV-DDA1 or DDA1 shRNA#1 in HT-29 and SW480 cells (*P < 0.05). (E) Detection of cell cycle-related proteins in DDA1 overexpressed and knocked down cell lines. (F) IHC staining of DDA1, Ki-67, c-myc, and CyclinD1 in DDA1 overexpressing and control xenografts (magnification ×200).

As changes in cell cycle are closely related to cell proliferation, cell cycle progression was further studied in the transfected cells. Fluorescence-activated cell sorting analysis showed that DDA1 overexpression or knockdown (with shDDA1#1 or shDDA1#2) resulted in decreased G1 phase and increased S phase cells or increased G1 phase and decreased S phase cells, respectively (Figure 4D and S4A). In addition, overexpression or knockdown induced the increase or decrease, respectively, in c-myc, CyclinD1, CDK4, and CDK6 expression, indicating proliferation and G1 to S phase transition of cell cycle (Figure 4E and S4B). Further, DDA1-overexpressing xenografts from nude mice had significantly higher Ki-67, c-myc, and CyclinD1 levels than controls (Figure 4F and S5). These results indicate that DDA1 promotes tumor cell proliferation by inducing S phase arrest.

DDA1 overexpression inhibits apoptosis in 5-FU-induced colon cancer cells

All patients with stage II colon cancer in this study received 5-FU-based adjuvant chemotherapy. We assessed whether DDA1 was upregulated by 5-FU. After treating colon cancer cell lines with different 5-FU concentrations for 48 h, DDA1 mRNA and protein levels were evaluated. There were no differences among the treated and untreated cells (data not shown), indicating that 5-FU did not upregulate DDA1 expression.

The effects of DDA1 on apoptosis inhibition in 5-FU-induced colon cancer cells was investigated using Annexin-V/PI staining, or by measuring apoptosis-related proteins such as cleaved caspase-3 and cleaved poly (ADP-ribose) polymerase (PARP). Annexin V/PI staining revealed that DDA1 overexpression in HT29 cells or knockdown in both DDA1 overexpressed-HT-29 cells and SW480 cells expressing DDA1 at a high level with shDDA1#1 rendered the cells resistant or sensitive to apoptosis induced by different concentrations of 5-FU, respectively (P < 0.05 for both, Figure 5A and 5B). In addition, the overexpression and knocked down decreased and increased 5-FU-induced caspase-3 and PARP cleavage, respectively (Figure 5C and 5D). Further, the specific target effects of the increase in the cleaved caspase-3 and PARP by shDDA1#1 were also confirmed by using the secondary shRNA, shDDA1#2 in both SW480 and SW620 cells (Figure 5E). These results show that DDA1 prohibited apoptosis in 5-FU-induced colon cancer cells and decreased chemosensitivity to 5-FU by inhibiting the activation of caspase-3 and PARP.

Figure 5

DDA1 knockdown increases apoptosis and 5-FU chemosensitivity

(A) Apoptosis of SW620 and HT-29/LV-DDA1 cells transfected with DDA1 shRNA#1 or Scramble after treatment with 5-FU (0 μm/25 μm/50 μm/100 μm) for 48 h. (B) The histogram shows comparative results of Annexin-V and PI FACS for 5-FU-induced apoptosis (*P < 0.05). (C) Changes in caspase-3 and PARP protein levels in HT-29/LV-DDA1 cells transfected with DDA1 shRNA#1 or Scramble, with or without 5-FU (100 μm) for 48 h. (D) Detection of caspase-3 and PARP proteins in SW620 and SW480 cell lines transfected with DDA1 shRNA#1 or Scramble, with or without 5-FU (100 μm) for 48 h.

DDA1 enhances invasion and induces the epithelial to mesenchymal transition

DDA1 overexpressing cell lines (HT-29 and DLD-1) or the knocked down cells (SW480 and SW620) exhibited enhanced or reduced invasion measured using transwell invasion assay, and wound healing capacity monitored using wound healing assay, as compared to the corresponding control cells, respectively (P < 0.05, Figure 6A–6B and S6A–S6B). DDA1 overexpression or knockdown resulted in increased or decreased morphological transition of cells from epithelial phenotype to mesenchyme-like morphology (EMT), respectively (Figure 6C). This result indicates that DDA1 triggers the EMT in colon cancer cells.

Figure 6

DDA1 induces colon cancer cell invasion and EMT

(A) Overexpression or knockdown of DDA1 increased or decreased the number of invasive cells (*P < 0.05). (B) Wound healing assays showed wound closure was promoted in HT-29/LV-DDA1 and DLD-1/LV-DDA1 cells, and delayed in SW480/shDDA1#1 and SW620/shDDA1#1 cells compared with the vector or Scramble groups (*P < 0.05). (C) Exceptional expression of DDA1 led to morphological alterations in colon cancer cells (magnification ×200). (D) Protein expression of EMT markers in HT-29/LV-DDA1, DLD-1/LV-DDA1, SW480/DDA1 shRNA#1, SW620/DDA1 shRNA#1 cells. (E) IHC staining for Snail, β-catenin, E-cadherin, N-cadherin, and vimentin in nude mouse tumor xenografts. Levels of Snail, nuclear β-catenin, N-cadherin, and Vimentin increased in HT-29/LV-DDA1 tumors, whereas E-cadherin expression declined (magnification ×200).

EMT, an essential mechanism in embryonic development and tissue repair, contributes to the progression of cancer and organ fibrosis [28]. During EMT, epithelial protein E-cadherin is downregulated, while mesenchyme proteins such as N-cadherin and vimentin are upregulated [29]. Overexpression or knockdown of DDA1 resulted in decreased or increased expression of epithelial protein E-cadherin, respectively, while expression of mesenchyme proteins N-cadherin and Vimentin was increased or decreased in vitro (Figure 6D and S6C). Further, DDA1 overexpression upregulated Snail, an EMT inductor (Figure 6D and S7) [30]. In vivo, xenografted DDA1-overexpressing tumor cells had lower E-cadherin levels, but higher expression of nuclear β-catenin, N-cadherin, vimentin, and Snail than controls (Figure 6E and S7). These results suggest that DDA1 promotes invasion and migration, and induces the EMT in colon cancer cells.

DDA1 activates NFκB/CSN2/GSK3β signaling

Classical activation of NFκB is involved in tumor proliferation, anti-apoptosis, and EMT [15, 28]. In addition, NFκB activation is required for the induction of the CSN2/GSK3β pathway, which enhances tumor invasion and metastasis [31]. CSN2, the second and most conserved subunit of the COP9 signalosome in eukaryotes, is also targeted by NFκB [32]. Therefore, we investigated whether DDA1 activates NFκB/CSN2/GSK3β signaling. We found that DDA1 overexpression or the knockdown with shDDA1#1 increased or decreased nuclear p65 translocation, indicating classical activation of NFκB, respectively (Figure 7A). We also demonstrated that the DDA1 overexpression or knockdown with shDDA1#1 promoted or inhibited IKKβ phosphorylation, thereby upregulating p-IκBα, which resulted in nuclear p65 translocation via degradation of p-IκBα, respectively (Figure 7B). These results indicate that DDA1 activates the classical NFκB pathway by increasing p-IKKβ expression.

Figure 7

DDA1 promotes colon cancer progression by stimulating NFκB/CSN2/GSK3β signaling

(A) Immunofluorescence analyses showed increased p65 nuclear translocation in HT-29/LV-DDA1 cells as compared with HT-29/LV-vector, but decreased in SW480/shDDA1 compared with SW480/Scramble. (B) Expression of DDA1/NFκB/CSN2/GSK3β pathway proteins in HT-29/LV-DDA1, DLD-1/LV-DDA1, SW480/shDDA1#1 and SW620/shDDA1#1 cells. (C and D) Changes in DDA1/NFκB/CSN2/GSK3β signaling proteins in HT-29/LV-DDA1 and DLD-1/LV-DDA1 cells after treatment with BAY-11-7082 and Helenalin. (E) Changes in DDA1/NFκB/CSN2/GSK3β pathway proteins in HT-29/LV-DDA1 and DLD-1/LV-DDA1 cells after transfection with IKKβ siRNA and Scramble.

p65 nuclear translocation reportedly induces CSN2 upregulation followed by phosphorylation of GSK3β at serine 9, which promotes tumor metastasis and EMT [31]. Overexpression or knockdown of DDA1 resulted in up- or downregulation of CSN2 and p-GSK3β, respectively (Figure 7B). Increased CNS2 and p-GSK3β levels induced by DDA1 overexpression were markedly attenuated following treatment with BAY-11-7082, an inhibitor of phosphorylation and degradation of IκBα, Helenalin, a p65 inhibitor, or siRNA specific for IKKβ (Figure 7C–7E). To directly assess whether DDA1 activated NFκB/CSN2/GSK3β signaling by regulating IKKβ expression, DDA1 and p-IKKβ were analyzed by immunofluorescence. DDA1 was co-expressed with p-IKKβ (Figure 8A), directly bound to p-IKKβ and stabilized its expression (Figure 8B). Functionally, silencing IKKβ with IKKβ siRNA inhibited the increased proliferation, migration, and invasion promoted by DDA1 (Figure 8C–8E). Lastly, the specific target effects of the DDA1-mediated p65 nuclear translocation and p-IKKβ stabilization were further confirmed by using the secondary shRNA shDDA1#2 to inhibit DDA1 activity in both SW480 and SW620 cells (Figure S8A and S8B). These results indicate that DDA1 promotes colon cancer cell proliferation, migration, and invasion by activating NFκB/CSN2/GSK3β signaling. DDA1 specifically promotes and stabilizes p-IKKβ, which increases p-IκBα, p65 nuclear translocation, and expression of CNS2 and p-GSK3β.

Figure 8

DDA1 activates the NFκB/CSN2/GSK3β pathway by regulating IKKβ phosphorylation

(A) Co-expression assays in HT-29 and SW480 cells: DDA1 (red), p-IKKβ (green), DAPI nuclear staining (blue). Merged images of DDA1 (red) and p-IKKβ (green) with DAPI (blue) are also shown. (B) Co-immunoprecipitation between endogenous DDA1 and p-IKKβ in DLD-1, HT-29, SW480, and SW620 cells. p-IKKβ was detected in the immunoprecipitate using an anti-DDA1 antibody as a bait. (C) In vitro proliferation assays in LV-DDA1-transfected or vector-transfected cells after transfection with IKKβ siRNA. (D) Wound healing assays were performed in LV-DDA1-transfected or vector-transfected cells after transfection with IKKβ siRNA. (E) Transwell invasion assays with LV-DDA1-transfected or vector-transfected cells after transfection with IKKβ siRNA.

Patients and tissue samples

Tissue specimens were obtained from 279 patients with stage II colon cancer (Table S1) who underwent tumor resection at Shanghai Jiaotong University Affiliated First People's Hospital, Shanghai, China, between 2001 and 2007. All patients were categorized into tumor recurrence and non-recurrence groups based on the recurrent tumor conditions. None of the patients underwent anticancer treatment before surgery. Diagnoses of all patients were confirmed without microsatellite instability high (MSI-H) pathologically. All patients had at least one conventional high-risk factor for tumor recurrence and received 5-FU-based chemotherapy (5-FU/LV or FOLFOX4; Table S2) after surgery according to National Comprehensive Cancer Network guidelines. All patients signed informed consent forms before enrolling in the study. The study was approved by the Institutional Research Ethics Committee of the hospital.

Cell culture, reagents, and transfection

Cells (SW480, DLD-1, HT-29, RKO, HCT116, HCT8, LoVo, SW620, NCM460, FHC) used in this study were purchased from the American Type Culture Collection. The cells were cultured in Dulbecco's modified Eagle's medium (DMEM) (GIBCO, Grand Island, NY) supplemented with 10% fetal bovine serum (GIBCO), 1% penicillin, and streptomycin with humidity at 37°C and 5% CO₂. Cell lines overexpressing DDA1 were constructed by transfection with lentiviral vectors encoding human DDA1 (LV-DDA1), and DDA1 knockdown cell lines were established through transfection with either shDDA1#1 or shDDA1#2. IκB kinase complex β (IKKβ) was knocked down by transfection with siRNA IKKβ plasmids. LV-vector and LV-Scramble were used as normal controls. All interference sequences are shown in Table S3.

qRT-PCR, western blotting, and immunohistochemical staining

Quantitative reverse transcription-polymerase chain reaction (qRT-PCR), western blotting, and immunohistochemical (IHC) staining were performed, and IHC staining scores were evaluated as described previously [19]. Primers used in qRT-PCR are summarized in Table S4. Antibodies are provided in Table S5.

In vitro proliferation, invasion and wound healing assays

For proliferation assays, transfected HT-29, DLD-1, SW480 and SW620 cells were seeded at 1 × 10⁵ cells per well into 96-well plates. Cell Counting Kit-8 (Beyotime, Biotechnology, Jiangsu, China) was used to test absorbance at 560 nm for each well at different time points using a micro-plate reader (Bio-Rad, Hercules, CA).

To determine the effects of DDA1 on invasion, 1 × 10⁵ cells were seeded onto matrigel-coated chambers (BD Biosciences, San Jose, CA), with serum-free media in the upper chambers and complete media in the lower chambers, and incubated for 24 h. After fixing and staining with 0.1% crystal violet, stained cells were counted and photos were taken using a light microscope (Olympus Corporation, Center Valley, PA).

In wound healing assays, cells were plated into six-well plates. Wounds were scraped when cells reached 90% confluence. The cells were washed with phosphate-buffered saline (PBS) and wound widths were measured and analyzed at 0, 24, 48 and 72 h.

Immunofluorescence analysis

Cells were plated on glass-bottom dishes and incubated overnight. After washing with PBS three times, cells were fixed by 4% paraformaldehyde for 15 min and washed three times with PBS. Cells were permeabilized with 0.1% Triton X-100 for 10 min, followed by incubation with p65 primary antibody overnight at 4°C. Cells were then incubated with Alexa Fluor 488-conjugated secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA) for 2 h at room temperature. 4′,6-Diamidino-2-phenylindole (DAPI) (Roche, Basle, Switzerland) was used to stain nuclei. Fluorescence images were collected using a confocal laser-scanning microscope (TCS SP8; Leica, Wetzlar, Germany).

Flow cytometry assay for cell cycle and apoptosis

Cells were collected at the logarithmic stage of growth, and were centrifuged and then resuspended at 2 × 10⁶ cells/mL. After 30 min incubation in 80% ethanol at 4°C, propidium iodide (PI) staining was utilized to measure cell cycle stage using flow cytometry (BD Biosciences). The Annexin V-APC/PI Apoptosis Kit (eBioscience, San Diego, CA) was used for the apoptosis assay according to the manufacturer's instructions. Flow cytometery was performed to determine cell number, and data were analyzed using FlowJo9.1 software (Tree Star Inc., Ashland, OR).

Co-immunoprecipitation assay

HT-29 and DLD-1 cell lines were transfected with empty vector or LV-DDA1 plasmids. SW480 and SW620 cell lines were transfected with Scramble or DDA1 shRNA plasmids. Cell lysate was immunoprecipitated with a DDA1 antibody. The p-IKKβ antibody was used to detect the level of p-IKKβ.

Xenografted tumors of nude mice

Colon cancer cells HT-29/vector, HT-29/LV-DDA, SW480/Scramble, and SW480/shDDA1#1 (1 × 10⁷ cells in 100 μL of DMEM) were injected subcutaneously into the flanks of nude mice (4 week old, male, Shanghai SLAC Laboratory Animal Co., Ltd., Shanghai, China). Tumor weights and volumes were measured every three days. All mice were killed after four weeks, and tumor samples were harvested and embedded in paraffin. Procedures involving animals conformed to the guidelines of the Institutional Animal Care and Use Committee of Shanghai Jiaotong University Affiliated Shanghai First People's Hospital.

Statistical analysis

Chi-square test or Fisher's exact test was used to compare the differences in categorical variables. Student's t-test or one-way analysis of variance was used to analyze the differences in continuous variables. Kaplan-Meier analyses with log-rank tests were used to evaluate disease-free survival (DFS) and overall survival (OS). The Cox proportional hazard model was performed to estimate the hazard ratio and 95% confidence intervals for DFS and OS. SPSS 19.0 statistical software (SPSS Inc., Chicago, IL) was used to analyze all data. P < 0.05 was regarded as statistically significant.