Tetraspanin CO-029 inhibits colorectal cancer cell movement by deregulating cell-matrix and cell-cell adhesions

Abstract

Alterations in tetraspanin CO-029 expression are associated with the progression and metastasis of cancers in the digestive system. However, how CO-029 promotes cancer metastasis is still poorly understood. To determine the mechanism, we silenced CO-029 expression in HT29 colon cancer cells and found that the CO-029 knockdown significantly reduced cell migratory ability. The diminished cell migration was accompanied by the upregulation of both integrin-dependent cell-matrix adhesion on laminin and calcium-dependent cell-cell adhesion. The cell surface levels of laminin-binding integrin α3β1 and fibronectin-integrin α5β1 were increased while the level of CD44 was decreased upon CO-029 silencing. These changes contribute to the altered cell-matrix adhesion. The deregulated cell-cell adhesion results, at least partially, from increased activity of cadherins and reduced level of MelCAM. In conclusion, CO-029 functions as a regulator of both cell-matrix and cell-cell adhesion. During colon cancer progression, CO-029 promotes cancer cell movement by deregulating cell adhesions.

Figure 1. Silencing CO-029 expression.

The CO-029 shRNA and nonsilencing shRNA were stably expressed in HT29 cells. (A) Flow cytometric analysis of CO-029 expression at the surface of HT29 cells that were transfected with nonsilencing (NS) or CO-029 shRNA (KD) constructs. The negative control mAb was murine IgG, and CO-029 mAb was NS1116. Mean fluorescent intensity (MFI) of CO-029 on KD cells was 62% less compared to that on NS cells. (B) Western blot analysis of CO-029 proteins in NS and KD HT29 cells. KD cells displayed a 90% of reduction in CO-029 proteins. β-actin: loading control.

Figure 2. CO-029 silencing impaired colorectal cancer cell migration.

(A) Wound healing assay. Compared with NS cells, wound closure was significantly impaired in the KD cells at 72 h after creation of wounds in confluent cell monolayers. (B) Transwell migration assay. The KD cells displayed impaired motility in transwell migration experiments. Data are displayed as mean±SD, n = 4. *P<.05.

Figure 3. CO-029 regulated cell-matrix and cell-cell adhesions.

(A) Cell-matrix adhesion. Adhesion onto laminin 111, laminin 332, and fibronectin of HT29-NS and -KD cells was assayed after incubating at 37^oC in 5% CO₂ for 1 h. The adhesion of the KD cells on laminin 111 and laminin 332 was significantly increased compared with NS cells (P = .042 on laminin 111 and  = .048 on laminin 332). (B) Total cell-cell adhesion. Cell aggregation was measured in the Ca⁺⁺-containing media after a relatively strong mechanical force was applied. (C) Ca⁺⁺-independent cell-cell adhesion. Cell aggregation was measured in the Ca⁺⁺-free media after relatively strong or mild mechanical forces were applied, respectively. The aggregation of NS and KD cells after shear stress were quantified as described in Materials and Methods. P values are 0.03 for the aggregated cells in the presence of Ca⁺⁺, 0.001 for the area of aggregates in the presence of Ca⁺⁺, and 0.0004 for mild stress in the absence of Ca⁺⁺. Images of cell aggregates after the shear-stress treatment were obtained under phase-contrast microscopy. All of the data are projected as mean±SEM (n = 4). *P<.05, **P<.01.

Figure 4. CO-029 silencing altered the surface expression of cell adhesion proteins and tetraspanins.

The expression levels of cell-matrix adhesion proteins (A), active β1 integrins (B), cell-cell adhesion proteins (C), and tetraspanins (D) at the surface of HT29-NS and -KD transfectant cells were measured by flow cytometry. The relative levels of these proteins on KD cells to NS cells are presented as histograms (mean±SD, n = 4∼8). *P<.05, **P<.01. In (B), after normalized by the corresponding relative levels of total integrin β1, the levels of active integrin β1 relative to NS cells are presented in the histogram on the right.

Figure 5. The effects of CO-029 silencing on the formations of focal adhesion, stress fiber, and adherens junction

HT29 transfectants were plated on glass coverslips and cultured in complete media for 2 days (for vinculin and F-actin staining, A) or until confluence (for E-cadherin and β-catenin staining, B). The cells were fixed, permeabilized, and incubated with primary mAbs at 4°C overnight, followed by Alex Fluor 594-conjugated secondary Ab staining. F-actin was stained by Alex Fluor 594-conjugated phalloidin. The images were captured with confocal microscopy. Bar  = 20 µm.

Figure 6. CO-029 silencing did not alter the formation of TEMs.

(A) The surface biotinylated HT29-NS and -KD transfectant cells were lyzed with indicated detergent buffers. Tetraspanins CD9, CD151, and CO-029 were immunoprecipitated from the lysates, separated by SDS-PAGE, and detected by enhanced chemiluminescence. The levels of β-actin proteins in lysates were examined by Western blot and served as lysate input control. (B) The expression levels of integrin α3β1-unbound CD151, total CD151, homoclustered CD9, and total CD9 at the surface of HT29-NS and -KD transfectant cells were measured by mAb TS151r, 5C11, C9BB, and Mab7, respectively, using flow cytometry. The relative levels of CD151 and CD9 on KD cells to NS cells are presented as histograms (mean±SD, n = 3∼5).