Coxsackie and adenovirus receptor is a critical regulator for the survival and growth of oral squamous carcinoma cells


Coxsackie and adenovirus receptor (CAR) is essential for adenovirus infection to target cells, and its constitutive expression in various cancerous and normal tissues has been reported. Recently, the biological role of CAR in human cancers of several different origins has been investigated with respect to tumor progression, metastasis and tumorigenesis. However, its biological function in tumor cells remains controversial. Here we report the critical role of CAR in growth regulation of oral squamous cell carcinomas (SCCs) in vitro and in vivo via the specific interaction with Rho-associated protein kinase (ROCK). Loss of endogenous CAR expression by knockdown using specific small interfering RNA (siRNA) against CAR facilitates growth suppression of SCC cells due to cell dissociation, followed by apoptosis. The consequent morphological reaction was reminiscent of anoikis, rather than epithelial–mesenchymal transition, and the dissociation of oral SCC cells was triggered not by lack of contact with extracellular matrix, but by loss of cell-to-cell contact caused by abnormal translocation of E-cadherin from surface membrane to cytoplasm. Immunoprecipitation assays of the CAR-transfected oral SCC cell line, HSC-2, with or without ROCK inhibitor (Y-27632) revealed that CAR directly associates with ROCKI and ROCKII, which results in inhibition of ROCK activity and contributes to maintenance of cell-to-cell adhesion for their growth and survival. Based on these findings, in vivo behavior of CAR-downregulated HSC-2 cells from siRNA knockdown was compared with that of normally CAR-expressing cells in intraperitoneally xenografted mouse models. The mice engrafted with CAR siRNA-pretreated HSC-2 cells showed poor formation of metastatic foci in contrast to those implanted with the control siRNA-pretreated cells. Thus, CAR substantially has an impact on growth and survival of oral SCC cells as a negative regulator of ROCK in vitro and in vivo.


Coxsackie and adenovirus receptor (CAR) was identified as a primary cellular receptor for adenoviral infection and has a crucial role not only for onset of infectious disease but also for adenovirus-mediated gene therapy.1, 2, 3 As a cellular component, CAR is reported to be involved in formation of the epithelial tight junction and maintenance of the cytoskeletal structure of epithelial cells through association of cell-adhesion molecules or microtubules.4, 5 Further investigations revealed that its endogenous expression in cancer cells of several different origins seemed to be related to tumor proliferation and metastasis, however, the biological function of CAR in these malignancies remains under debate. For example, loss or decline of endogenous CAR expression was observed in primary and/or metastatic colon, gastric, esophageal and bladder cancers.6, 7, 8, 9 In addition, ectopically enforced expression of CAR on these CAR-negative cells caused growth suppression and decrease of metastasis, which was inversely corroborated by knockdown assay using CAR-specific small interfering RNA (siRNA).6, 7, 10 In these cases, for example, overexpression of CAR was reported to induce Snail, which caused downregulation of integrins in melanoma cells,10 or it upregulated p21 expression in bladder and prostate cancer cells.11, 12 On the other hand, the opposite phenomena that lead to growth promotion were reported in some studies. In thyroid tumors, enhanced expression of CAR was prominently observed in thyroid carcinomas in comparison with benign thyroid tumors, which was significantly associated with larger tumor size.13 Similarly, CAR expression was enhanced in invasive mammary gland adenocarcinomas raised from preneoplastic precursor lesions in vivo, and its expression was required for efficient tumorigenesis in lung cancer.14, 15 It was also reported that the overexpression of CAR inhibits caspase3/7 activation triggered by TRAIL in colon cancer cells,6 and CAR-overexpressing cells showed strong resistance against TRAIL-induced apoptosis, probably because of upregulation of apoptosis suppressors such as Bcl-2 or Bcl-XL in invasive adenocarcinomas.14

With respect to concrete CAR-associated molecules, there have been only a few studies reported. ZO-1 and junctional adhesion molecule-like protein, for example, were identified as direct partners, suggesting that CAR has multiple roles—some yet to be identified—in regulating cellular behavior via protein interactions.4, 16 Here we demonstrate that CAR is a novel binding partner of Rho-associated protein kinases (ROCKs), ROCKI and ROCKII,17 and functions as a negative regulator to antagonize their activity, where suppression of CAR ultimately leads to apoptotic cell death, a process known as ‘anoikis’, in oral squamous cell carcinoma (SCC) cells.18


Expression of CAR in human tumor cells with diverse origins

We first examined CAR expression in various human tumor cell lines with different origins by quantitative real-time RT-PCR (Figure 1a) and immunoblotting analyses (Figure 1b). CAR mRNA was detected in all solid tumor lines, including carcinomas and sarcomas, however, its expression level was much lower in hematopoietic and brain malignancies. Among solid tumors, carcinomas such as HepG2, Lovo and HSC-3 are expressing comparable amount of CAR mRNA(Figure 1a). In normal human dermal fibroblasts (NHDFs) and peripheral blood mononuclear cells, CAR signal was undetectable (Figure 1a). This result was also confirmed by the protein expression analysis. CAR protein was prominently detected in HeLa, Lovo, HepG2, PANC-1, A549, HaCaT, HSC-2 and PK-8 cells, but barely detected in MKN-28 and MCF-7 cells, and absent in primary acute myelogenous leukemia (AML), T98 and NHDF cells (Figure 1b). Because quantitative PCR and immunoblotting analyses simply detect the amount of CAR mRNA and protein expressions on these tumor cells, infectivity of adenovirus is considered to be useful to examine if CAR being expressed on these lines are biologically functional or not. Therefore, CAR function on these tumor cells was further assessed by infection with a green fluorescent protein (GFP)-expressing adenovirus vector (Ad-GFP), whose infectivity was essentially consistent with the expression profiles of CAR by quantitative reverse transcription PCR and immunoblotting (Figure 1c). These results indicate that CAR might have critical biological role especially in carcinomas, in comparison with malignancies of other origins.

Figure 1

Expression profile of CAR in human tumor cell lines. (a) Total RNAs prepared from indicated cell lines were analyzed for expression of CAR by quantitative real-time PCR. 18S ribosomal RNA was used as control for the analysis. Relative expression level of each line was calculated and shown by ratio (CAR/18S) against that of HeLa as a standard. (b) Protein extracts from the indicated cell lines were also analyzed for expression of CAR by western blotting. Beta-actin was used to indicate loaded amounts of protein. (c) Ad-GFP was infected to the indicated cancer cell lines to become 1.0 multiplicity of infection (M.O.I.). Twenty-four hours after the infection, expressed GFP (green) in the fixed cells were visualized. Nuclei were stained with 4′, 6-diamidino-2-phenylindole (blue).

Downregulation of CAR induces cell dissociation, followed by apoptosis

To understand the critical physiological role of CAR in cancer cells, we employed RNA interference to inhibit its activity, using specific siRNA against human CAR (siCAR). First, we examined whether application of siCAR could efficiently suppress the expression of endogenous CAR in all four squamous carcinoma cell lines, Hela, HSC-2, HSC-3 and Ca9-22, because they are expressing considerable amount of CAR mRNA (Figure 2a). The validated siRNA, at a concentration of 10 nM, induced prominent detachment of these cells from the bottom of culture vessels, and they began to dissociate from each other, whereas control siRNA (siRNA with scrambled sequence of siCAR (siCont.)) yielded no such behavior in these cells (Figure 2b). We next evaluated the cancer cells treated with siCAR in terms of their in vitro growth potential. We observed that CAR silencing greatly reduced cell proliferation and viability (Figure 2c), leading eventually to apoptotic cell death as shown by increased annexin V positivity in the fluorescence-activated cell sorting analysis (Figure 2d). Cell morphology was characterized by a rounded shape, due to loss of cell-to-cell contact. A similar effect was observed not only in SCC cell lines but also in pancreatic and gastric adenocarcinoma cell lines, such as PK-8, PANC-1, and MKN-28 (Supplementary Figure S1a). Taken together, our results suggest that the reaction triggered by siCAR is similar to ‘anoikis’, which is a type of apoptosis caused by detachment of epithelial cells from the extracellular matrices (ECMs) and/or by loss of cell-to-cell contact among neighboring cells.16 It is also noted that normal human fibroblasts, NHDF cells, are negative for CAR expression and did not exhibit the phenomenon associated with siCAR treatment, suggesting that this phenomenon was specifically caused by targeting CAR mRNA (Supplementary Figure S1b).

Figure 2

Downregulation of CAR reduces cell viability in SCC cells. (a) SCC lines, HeLa, HSC-2, HSC-3 and Ca9-22 cells, were transfected with siRNA with the scrambled sequence (siCont.) or CAR siRNA (siCAR). Seventy-two hours after the transfection, total RNAs prepared from the cell lines were analyzed for expression of CAR by quantitative real-time PCR. 18S ribosomal RNA was used as an internal control for the analysis. (b) Under the same conditions as those described in a, each siRNA-treated cells were morphologically examined under an inverted phase contrast microscope. Images were taken at × 100 and × 400 magnifications. (c) After the transfection with siCAR, cell growth of each sample was evaluated at the different time points by MTT assay. Values are indicated as the means±s.d. of three independent experiments. (d) Annexin V-positive cells were counted under the same conditions as those in a by fluorescence-activated cell sorting analysis. Values are expressed as the means±s.d. of three independent experiments.

siCAR inhibits anchorage-independent growth of HSC-2 cells

To clarify whether the siCAR-mediated cellular phenomenon is actually anoikis, we further examined the influence of siCAR on anchorage-independent growth of oral SCC, HSC-2 cells, in vitro by measuring their colony forming ability. When the cells were cultured in agarose-coated dishes, keeping them in suspension, siCAR abolished colony formation (Figure 3a), resulting in increased numbers of single cells (Figures 3a and b) along with sharply reduced cell proliferation and viability (Figure 3c). Aggregated cells were still viable despite having lost contact with the agarose-coated dish; however, most of the dissociated (single) cells entered apoptosis during the culture period, suggesting that apoptosis was triggered by the loss of intercellular contact. This hypothesis was corroborated by assays using culture dishes coated with representative ECMs, typeI collagen, fibronectin and laminin for the siCAR-treated HSC-2 cells. None of the ECMs had any appreciable effect on rescue of the cell dissociation and inviability (Figures 3d and e). Based on these results, we consider the cellular event caused by siCAR to be anoikis.

Figure 3

Downregulation of CAR abrogates anchorage-independent growth in SCC cells. (a) After transfection of HSC-2 cells with the siCont. or siCAR, the cells were cultured in 0.6% agarose-coated dishes for 48 or 72 h; cell morphologies were observed under an inverted phase contrast microscope at × 100 magnification. (b) Under the same conditions described in a, numbers of single cells were counted at the different time points by trypan-blue dye exclusion. Values are expressed as the means±s.d. of three independent experiments. (c) Under the same conditions as in a, cell-growth activity was evaluated at 72 h by MTT assay. Values are expressed as the means±s.d. of triplicated experiments. (d) HSC-2 cells cultured on the different ECM-coated plates (typeI collagen, fibronectin and laminin) were transfected with siCAR for 72 h. Cell morphologies were observed under an inverted phase contrast microscope at × 100 magnification. (e) Under the same conditions described in d, cell growth of each sample was evaluated at 72 h by MTT assay. Values are expressed as the means±s.d. of triplicated experiments. TC plate, non-coated tissue culture plate); Col plate, typeI collagen-coated tissue culture plate; FN plate, fibronectin-coated tissue culture plate; LN plate, laminin-coated tissue culture plate.

siCAR induces intracellular translocation of E-cadherin

Because the cell-to-cell and cell-to-ECM dissociation observed in the epithelial–mesenchymal transition (EMT) is primarily directed not to anoikis, but to the release of epithelial cells from the surrounding tissue for metastasis, we sought to elucidate the mechanistic distinction between EMT and the anoikis-related, siCAR-induced cellular dissociation. Accordingly, we investigated several major EMT markers in siCAR-introduced HSC-2 and HSC-3 cells. Immunoblotting using several antibodies as representative EMT makers revealed no significant downregulation of E-cadherin, no upregulation of vimentin or fibronectin (Figure 4a) and no altered expression of Snail (Figure 4b), with or without siCAR treatment of these cells. Thus, siCAR is likely to affect the molecular pathways that are fundamentally distinct from EMT-caused dissociation.

Figure 4

Downregulation of CAR induces cytoplasmic translocation of E-cadherin in SCC cells. (a) HSC-2 and HSC-3 cells were transfected with siCont. or siCAR (10 nM concentration). Seventy-two hours after the transfection, total proteins extracted from the cell lines were analyzed for expression of CAR, E-cadherin, vimentin and fibronectin by immunoblotting. Beta-actin was used as control for the analysis. (b) After transfection of HSC-2 cells with siCont. or siCAR, cellular amounts of Snail were evaluated at the different time points by immunoblotting. 81B-Fb, a head and neck SCC line that bears an induced EMT phenotype, is a positive control for prominent Snail expression (Maseki et al.38). (c) After introduction of siCont. or siCAR for 72 h, immunofluorescence staining was performed to visualize E-cadherin (red) in HSC-2 (left) and HSC-3 cells (right). The cells were counterstained with Hoechst33258 for nuclei (blue). Scale bars, 50 μm.

On the other hand, it was found that E-cadherin, as a cell-adhesion molecule, apparently translocated from the cell-surface membrane to cytoplasm in response to downregulation of CAR by the siRNA in both HSC-2 and HSC-3 cells, which is consistent with the loss of cellular contact observed in the siCAR-treated cells (Figure 4c). Consistent with the translocation of E-cadherin, beta-catenin also moved from cell surface to cytoplasm, and abnormal aggregation of actin filaments was observed in the siCAR-treated cells (Supplementary Figure S2).

Effect of CAR downregulation on non-neoplastic counterpart cells

Because CAR is also expressed in normal human epithelial cells,19 but not in non-neoplastic mesenchymal cells such as fibroblasts (NHDFs; Figures 1a and b), we examined whether downregulation of CAR expression by siCAR significantly affects viability of normal human epidermal keratinocytes (NHEKs) as a normal counterpart of SCC. The NHEK cells were cultured in serum-free KG2 medium supplemented with 0.5 μg/ml of hydrocortisone, 10 μg/ml of insulin and 0.1 μg/ml of human EGF, adjusted to low Ca (0.03 mM), or in RPMI medium containing normal Ca (1.5 mM) with 10% fetal bovine serum, because the extracellular calcium and serum concentrations critically affect growth and differentiation of the cells.20 Remarkably, the treatment with siCAR did not substantially affect cell proliferation and viability of NHEKs even with higher concentration of the siRNA (50 nM) under the both medium conditions (Figure 5a), although endogenous CAR expression was abolished (Figure 5b). The live cell images from phase contrast microscopy showed no morphological changes supporting anoikis by treatment with siCAR (Figure 5c). Moreover, the siRNA did not significantly alter functional localization of the E-cadherin at the cell membrane of NHEKs (Figure 5d). This observation in NHEKs was consistent with both conditions between serum-free medium supplemented by EGF, hydrocortisone and insulin, and Roswell Park Memorial Institute containing 10% fetal bovine serum (Supplementary Figures S3a and b). These results suggest that, unlike SCC cells, a representative non-neoplastic counterpart, NHEKs, showed resistance against CAR downregulation, although the precise mechanism remains unclear.

Figure 5

Effect of downregulated CAR on normal keratinocytes. (a) NHEKs were cultured in two different conditions, with KG2 medium adjusted to low Ca (0.03 mM) without serum (left) and Roswell Park Memorial Institute (RPMI) medium containing normal Ca (1.5 mM) with 10% serum (right). Under the two different conditions, the NHEK cells were transfected with siCAR or siCont. (sc) for 72 h, and then their viabilities were evaluated by MTT assay. Values are expressed as the means±s.d. of three independent experiments. (b) Under the same conditions as in a, an inhibitory effect by siCAR on CAR expression in NHEK cells was assessed by immunoblotting. Seventy-two hours after the transfection with the siRNA, cells were collected. (c) Under the same conditions as in a and b, morphological images of the siRNA-treated NHEK cells. Images were taken at × 100, × 200 and × 400 magnifications in the two culture conditions, KG2 medium (left) and RPMI medium (right). (d) Expression and distribution of CAR and E-cadherin on siRNA-treated NHEKs (siCont.: left; siCAR: right) by immunofluorescence. CAR (green) and E-cadherin (red) nuclear counterstain by Hoechst33258 (blue) ( × 400).

ROCK is a novel intracellular binding partner of CAR

To obtain the molecular details about the intracellular function of CAR on SCC cells, we focused on molecular events at the cytoplasmic domain of CAR. The possible adaptor proteins for CAR are still not fully elucidated. In our prior study, we found that 10 μM of a ROCK inhibitor (ROCKI: Y-27632) rescued siCAR-mediated anoikis to some extent (Figure 6a) and prevented E-cadherin cytoplasmic translocation in HCS2 cells (Figure 6b). In addition, immunoprecipitates with the full-length CAR from the HSC-2 cell lysates showed kinase activity, allowing phosphorylation of a representative ROCK substrate, MYP1; this phosphorylation was completely inhibited in the presence of 10 μM of Y-27632 in in vitro assay (Figure 6c). Because the cytoplasmic domain of CAR lacks any potential kinase domain, the phosphorylation is likely induced by ROCK. The immunoprecipitation using myc epitope-tagged CAR expressor, CAR-transfected HSC-2 cell lysates revealed coprecipitation of ROCKI and ROCKII with CAR, which demonstrates that ROCKI and ROCKII are novel binding partners for CAR in cytoplasm (Figure 6d). When an excess amount of purified recombinant CAR protein was added to immunoprecipitated ROCKII obtained from HSC-2 cell lysates, kinase activity of ROCKII was abrogated by addition of CAR in a dose-dependent manner (Figure 6e), indicating that CAR suppresses ROCK activity through direct binding. Based on these findings, we further examined ROCK expression in NHEK cells, insofar as the cells showed comparable resistance to siCAR-induced anoikis. Both ROCKI and ROCKII were poorly expressed in NHEK cells in comparison with those in some neoplastic cell lines, including HSC-2, PC-3 and Saos-2 (Figure 6f). Accordingly, ROCKII in NHEK cells showed only a faint kinase activity when the cells were treated with siCAR (Figure 6g). These results indicate that CAR is a critical suppressor of ROCK activity; specifically, downregulation of CAR induces ROCK activation, leading to cytoplasmic translocation of E-cadherin, following anoikis in SCC cells but not in NHEK cells as a normal counterpart.

Figure 6

ROCKs directly associate with CAR. (a) HSC-2 cells were transfected with siCont. or siCAR (10 nM concentration) with or without ROCK inhibitor (Y-27632: 10 μM concentration). Cell viabilities were evaluated by MTS assay 72 h post transfection. Values are expressed as the means±s.d. of triplicated experiments. (b) HSC-2 cells were treated under the same conditions as described in a. After fixation, the cells were stained to visualize intracellular actin filaments (falloidin: green) and E-cadherin (red). Nuclei were stained with Hoechst33258 (blue). × 400 magnification. (c) HSC-2 cells were transiently transfected with the plasmid expressing CAR-Myc-6His. The exogenously expressed CAR purified from the transfected cell lysate by anti-6His antibody-agarose was incubated with recombinant MYPT1 (1.0 μg) of commercial source in the absence (lane 1) or presence (lane 2) of ROCKI (1.0 μM) in the kinase reaction buffer containing adenosine triphosphate (ATP) (see Materials and methods). Immunoblotting probed with anti-phosphorylated MYPT1 antibody was utilized for determination of the phosphorylation status of the substrate MYPT1. (arrow: phospho-MYPT1 protein). (d) GFP (negative control: lane 1) and CAR-Myc-6His (lane 2) was overexpressed in HSC-2 cells under the same conditions described in c. Binding of endogenous ROCKI and ROCKII to the expressed CAR was demonstrated by co-immunoprecipitation using anti-6xHis antibody agarose followed by immunoblotting probed with the indicated antibodies. (e) Human full-length CAR protein purified from the CAR-myc-6His expresssor-transfected HEK293T cells was subjected to SDS–PAGE and detected by Coomassie Brilliant Blue (CBB) staining (left). Endogenous ROCKII was purified by pull-down method using a biotinylated ROCKII antibody with streptavidin agarose (see Materials and methods) from HSC-2 cell lysates. The prepared cellular ROCKII was incubated with recombinant MYPT1 (1.0 μg) under the absence (lane 1) or presence of ROCKI (1.0 μM: lane 2) and increased concentrations of recombinant CAR (0.1 μg: lane 3, 0.5 μg: lane 4 and 1.0 μg: lane 5) in the kinase reaction buffer containing ATP (right). (f) ROCKI and ROCKII expression in the indicated cells by immunoblotting using their specific antibodies. (g) Endogenous ROCKII proteins from NHEK (lanes 1–3) and HCS2 cells (lanes 4–6) transfected with siCont. (lanes 2 and 5), siCAR (lanes 3 and 6) or the siRNA-untreated (lanes 1 and 4) were detected under the same conditions as in e. Kinase activity by the purified ROCKII from each cell sample was shown by immunoblotting using anti-phospho-MYPT1 antibody.

CAR is a critical regulator of ROCK and inhibits ROCK-mediated apoptosis

To demonstrate the negative regulation of the ROCK activity by CAR in cellular level, we performed transient transfection of the artificial CAR (N-terminal 6His-Myc-tagged CAR) and later magnetic cell isolation system using biotinylated Myc-tag antibody to purify CAR-overexpressed cells. According to this procedure, we succeeded to enrich the artificial CAR-overexpressed HSC-2 cells at a rate of 95–98% positivity without affecting cell viability (data not shown). The artificially expressed CAR protein upwardly shifted above the endogenous one because of fused Myc-tag and 6His-tag sequences (2 kDa larger in size), which was expressed at extremely high level (85-fold in density to that derived from the endogenous CAR: Figure 7a). Using this system, we evaluated intracellular ROCK activity in the enriched CAR-overexpressing HSC-2 cells. Because ROCK activity is positively regulated by RhoA, a small GTPase,21, 22 we further transfected the enriched cells with a constitutively active RhoA expressor to enforce sustained activation on the endogenous ROCK. Although the active RhoA caused marked increase in ROCK activity, allowing phosphorylation of the MYPT1 in the CAR-untransfected cells, this phosphorylation was strongly impeded in the CAR-overexpressing cells as well as in the presence of 10 μM of ROCK inhibitor, Y-27632 (Figure 7b). Under the same condition as that in Figure 7b, we examined cleavages of caspase 8 and Bid on these cells because induction of these apoptotic molecules is widely accepted as a specific molecular event in the process of anoikis.23, 24 Consequently, cleaved form of these molecules were detected when the RhoA-ROCK axis was constitutively activated, however, these events were completely abolished in the CAR-overexpressing cells as in the case treated with ROCK inhibitor (Figure 7c). Furthermore, immunofluorescence using fluorescein isothiocyanate (FITC)-annexin V revealed that overexpression of CAR prevented cells from apoptosis caused by sustained activation of RhoA-ROCK signaling (Figures 7d and e). Taking these together, it was indicated that CAR is a negative regulator for ROCK, which prevents RhoA-ROCK-mediated apoptotic cell death via specific interaction with ROCK in SCC cells.

Figure 7

CAR protects SCC cells from ROCK-mediated apoptosis. (a) HSC-2 cells were transiently transfected with the N-terminal 6His-Myc-tagged CAR, incubated for 24 h, and then the CAR-overexpressing cells were purified by Magnetic-Activated Cell Sorting (MACS) procedure using biotinylated Myc-tag antibody. Overexpression of CAR was assessed in comparison with endogenous CAR by immunoblotting using anti-CAR antibody. Lane 1: nontreated HSC-2 cells. Lane 2: enriched CAR-overexpressed cells. (b) The enriched CAR-overexpressing HSC-2 cells (lane 2, 4 and 6) and intact cells (untransfected, non-sorting cells: lane 1, 3 and 5) were further transfected with RFP (negative control: lane 1 and 2) or constitutively active RhoA (RFP-RhoA (active): lane 3–6) under the presence (lane 5 and 6) or absence (lane 1–4) of ROCK inhibitor (Y-27632: 10 μM concentration). Endogenous ROCKII proteins from the prepared cells (lanes 1–6) were isolated, and their kinase activities were assessed by immunoblotting using anti-phospho-MYPT1 antibody. (c) HSC-2 cells were treated under the same conditions as described in b. Signaling molecules associated with an anoikis-mediated apoptosis were evaluated by immunoblotting probed with the indicated antibodies. (d) HSC-2 cells were treated under the same conditions as was described in b. After fixation, the cells were stained to visualize transfected products (RFP and RFP-RhoA (active): red) and nuclei (Hoechst33258: blue). FITC-annexin V (green) was used as an apoptotic marker. × 400 magnification. (e) Under the same condition of d, to estimate the apoptotic magnitude, fluorescence intensities derived from FITC-annexin V were measured for each sample of the cultured cells by fluorescence plate reader.

Effect of CAR downregulation in tumor metastasis in vivo

To assess the significance of CAR expression on cancer cells in vivo, ectopic-xenograft mice models with intraperitoneal injections of GFP-expressing HSC-2 cells (GFP-HSC-2) were examined for their cellular behavior. We thus compared the distribution of intraperitoneally developed tumor (peritoneal metastasis) in vivo between the siCAR-pretreated and the scrambled siRNA (siCont.)-pretreated GFP-HSC-2 cells. Each group of cells were cultured for 12 h after siRNA introduction and then collected, and mice were inoculated with these cells for the indicated periods. At this time point, the siCAR-introduced cells, as well as the siCont.-introduced cells, well maintained their viability and continued growing (Supplementary Figure S4). The in vivo study using the ectopic oral carcinoma cell-xenografted mouse models resulted in a robust inhibition of tumor colonization, subsequent outgrowth and peritoneal dissemination of the siCAR-pretreated cancer cells, in contrast to the scrambled siRNA-treated cells, both at 21 days and 35 days post implantation (Figure 8a). Thus, CAR signaling is implicated in the progression of SCC cells via inhibition of anoikis in vivo.

Figure 8

CAR expression is important in SCCs for tumor growth in vivo. (a) GFP-HSC-2s were introduced either with siCont. or siCAR before injection of these into mouse peritoneal cavity. Expansion of the i.p. engrafted tumor cells was macroscopically visualized at the 21 and 40 days post implantation of the cells with different cell numbers (0.6 × 106 and 1 × 106 cells/intraperitoneal/mouse). (b) Immunohistochemistry of CAR in lingual and pharyngeal SCCs from the patients. Expression of CAR was detected in indicated lesions including non-neoplastic squamous layer. Insets: hyperview ( × 40). Scale bars, 100 μm.

Based on these findings, we next immunohistochemically examined the expression of CAR in surgically excised tumor samples from the patients with lingual, pharyngeal and laryngeal SCCs. As shown in the figures, CAR was abundantly expressed at the junction regions of tumor cells, yielding cancer nests in well-differentiated SCC (Figure 7b). In the case of poorly differentiated SCCs, CAR was similarly detectable at the cellular membrane of the tumor cells, which formed small invasive or metastatic nests, indicating that they essentially retain CAR expression in their development (Figure 8b). This observation was also consistent with vascular invasive foci of the cancer as a small nest, whereas CAR staining was only faintly positive in basal epithelial cells of the normal squamous layer (Figure 8b). The results from histological examination imply that CAR expression along with functional membrane localization may still be required for invasion and distant metastasis of the SCCs in vivo.


Several lines of evidence indicate that CAR has essential roles not only in adenovirus infection but also in a wide variety of cellular processes, such as cell survival, apoptosis,6, 12, 25 adhesion4, 26 and migration.5, 10, 27 Despite these pivotal, and sometimes controversial, functions, molecular events just downstream of CAR remain unknown. In this study, we focused on the function of endogenously expressing CAR and revealed its relationship with anoikis in SCC cells via specific interaction with ROCK, whose activity is critically regulated by CAR.

E-cadherin is now widely accepted to be a key regulator for cell-to-cell contact through modulating the structural network of the actin cytoskeleton.28, 29 Its function was dependent on not only an expression level but also subcellular localization. From this point of view, we found that the downregulation of endogenous CAR induced cytoplasmic translocation of E-cadherin from the cellular junctions, which is distinct from EMT (Figure 4). This implies that siCAR-mediated cell dissociation might be caused by functional impairment of a cellular junction via loss of E-cadherin, which triggers subsequent anoikis. Indeed, accumulated evidence from other studies support this notion, namely, that targeted expression of the dominant-negative form or application of neutralizing antibody against E-cadherin augments the frequency of anoikis in several cancer cells.30, 31 Taken together, the results strongly indicate that CAR has a significant role in regulation of tumor cell survival through adhesion molecules such as E-cadherin.

The next important question is how CAR involves translocation of E-cadherin and subsequent anoikis. To address this issue, we consider the identification of intracellular binding partners with CAR to be essential. Although CAR was found to associate with several intracellular membrane and junction-associated PDZ scaffold proteins MAGI-1b, PICK and PSD-95,32 MUPP-133 and LNX,34 these proteins are likely not to be direct players in the anoikis signaling axis. Therefore, we tried to identify a possible adaptor molecule that binds to the cytoplasmic tail of CAR, which is expected to be critical for CAR-mediated intracellular events. The Rho-associated kinase, ROCK protein, was then found to associate with CAR. The ROCK regulates subcellular localization of E-cadherin through the modulation of the structural network of the actin cytoskeleton, cellular movement via interaction with myosin motor proteins,17, 35, 36 and activates the intrinsic cell death pathway in dissociation-induced anoikis.15, 37 Indeed, we found that CAR inhibits the ROCK activity by direct association in vitro, and the presence of CAR helps maintain functional localization of E-cadherin (Figure 6), promoting cell survival and growth of HSC-2 cells, whereas hyperactivation of ROCK induces apoptosis on SCC cells.

The ROCK family consists of at least two isoforms, ROCKI and ROCKII, that are differentially expressed in distinct cell types.21 Insofar as NHEK cells, as non-neoplastic counterparts, were not seriously affected in their growth and viability by downregulation of CAR, ROCK expression underwent comparative analysis in NHEK cells. Expression of both ROCKI and ROCKII was detected in NHEKs, but was much lower than that found in malignant tumor lines, such as HSC-2, PC-3 and Saos-2 cells (Figure 6f). Furthermore, the kinase activity was comparably low in NHEK cells as compared with HSC-2 cells (Figure 6g). This may explain why NHEK cells were less sensitive to downregulation of CAR, whereas cancer cells were dramatically affected by it, through alteration of ROCK activity. Cancer cells seem to be more dependent on ROCK activity in their growth.

As was mentioned in the introdution, previous studies left much controversy about the biological function of CAR in malignant tumors.6, 7, 8, 9, 10, 11, 12, 13 Here we demonstrated that the constitutive expression of CAR is required for cell growth and survival of SCC cells, inasmuch as loss of CAR triggered dissociation of the nested tumor cells, leading to anoikis through activation of the ROCK signaling pathway, as distinguished from EMT. Histological examination of tumor tissues from SCC patients revealed retention of CAR expression even in distant metastatic and vascular invasive foci consisted of small cancer nests, indicating a requirement for CAR expression in metastasis of cancers (Figure 8b). However, its biological function may vary in tumor cells with diverse origins because of their differences in intracellular machinery affecting the CAR-interacting pathways.

Our study suggests that CAR is a critical regulator for growth, survival and metastasis, at least, in a certain lineage of cancers such as SCCs, through a direct interaction with Rho-associated kinases. Further analysis of this newly identified molecular connection between CAR and Rho-associated kinases should advance our understanding of cancer progression and offer promising therapeutic strategies for diverse cancers.

Materials and methods

Cells, chemicals and adenovirus

Human cancer cell lines used in the present study were as follows: HEK293T (embryonic kidney cell lines stably express the SV40 large T antigen; RIKEN BioResource Center, Tsukuba, Japan), HeLa (cervix adenocarcinoma cell line; ATCC, Rockville, MD, USA), Ca9-22 (oral SCC; Human Science Research Resources Bank, Japan), HSC-2, HSC-3 (oral SCC; Human Science Research Resources Bank), Lovo, WiDr (colon adenocarcinoma; ATCC), MKN-28 (gastric adenocarcinoma; ATCC), HepG2 (hepatoblastoma; ATCC), KYN-2 (hepatocellular carcinoma kindly provided by Dr Yano, Kurume University), PANC-1 (pancreatic adenocarcinoma; ATCC), PK-8 (pancreatic carcinoma cell line; Cell Resource Center for Biomedical Research, Tohoku University, Sendai, Japan), A549 (lung adenocarcinoma; ATCC), MCF-7 (mammary gland adenocarcinoma; ATCC), PC-3, LNCaP (prostate adenocarcinoma; ATCC), KPK-1 (renal clear cell carcinoma cell lines; Clonetics, San Diego, CA, USA), U2OS (osteosarcoma; ATCC), U-87MG, T98, A172 (glioblastoma; ATCC or kindly provided by Dr Natsume, Department of Neurosurgery, Nagoya University), Jurkat (T-cell leukemia; ATCC), MT-1 (adult T-cell leukemia/lymphoma; ATCC), K562 (chronic myelogenous leukemia; ATCC), Saos-2 (osteosarcoma; ATCC) and FL-18 (follicular lymphoma kindly provided by Dr Ohno, Takeda Hospital). 81B-Fb, an EMT-induced line derived from head and neck SCC line UMSCC81B, was kindly provided by Dr Maseki (Department of Otolaryngology, Head and Neck Surgery, Nagoya City University).38

Primary acute myelogenous leukemia cells and the colonic cancer cells from patients with a respective tumor were purchased from Celprogen Inc. (San Pedro, CA, USA) by material transfer agreement with informed consent. Following the second-to-third subcultures, the resulting primary cells were utilized for the in vitro assays. The study using healthy volunteers’ peripheral blood mononuclear cells was approved by the research ethics committees of Aichi Cancer Center.

Primary normal human keratinocytes (NHEK) were purchased from Kurabo Industries Ltd. (Neyagawa, Japan) with a material transfer agreement, and NHDFs were kindly provided by Hayashibara Biochemical Institute (Japan) with material transfer agreement.

ROCK inhibitor, Y-27632, and premade recombinant adenovirus expressing GFP (Ad-GFP) were purchased from Wako Pure Chemical Industries (Osaka, Japan) and Vector BioLabs (Philadelphia, PA, USA), respectively.

Real-time quantitative RT-PCR

Total RNA was extracted from prepared cells, converted to complementary DNA using an oligo (dT) primer and subjected to reverse transcriptase reaction (1 μg of total RNA as a template using Superscript III, Life Technologies, Invitrogen, Carlsbad, CA, USA) for performing quantitative real-time PCR. The forward and reverse primer sequences for CAR were as follows:

Forward primer: 5′-IndexTermATGAAAAGGAAGGAAGTTCATCACGATA-3′; Reverse primer: 5′-IndexTermAATGATTACTGCCGATGTAGCTT-3′.

For detection of 18S ribosomal RNA as an internal control by real-time PCR analysis, TaqMan MGB probe/primer set was utilized (Eukaryotic 18S rRNA, no. 4319413E; Life Technologies, Applied Biosystems, Carlsbad, CA, USA).

Cell proliferation and viability assay

Cell proliferation and viability was evaluated by 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromid (MTT) assay using a Cell Proliferation Kit I (Roche Applied Science, Mannheim, Germany) according to the manufacturer’s protocol. Flow-cytometric analysis using Cy3-labeled annexin V was employed to determine the percentage of apoptotic cells within the total (10 000 cells per sample) for tumor lines treated with 10 nM of CAR siRNA, 48 h after introduction of the siRNA (BD FACSCalibur, BD Biosciences, Tokyo, Japan). For measuring viable single cell, the number of the cells in each sample were counted using the trypan-blue dye exclusion method (Cellstain-Trypan-blue; Dojindo Lab, Tokyo, Japan). For measuring fluorescence intensity of FITC-annexin V-positive HSC-2 cells, a fluorescence microplate reader (Fluoroskan Ascent FL; Thermo Fisher Scientific Inc., Waltham, MA, USA) was used and evaluated the magnitude of apoptosis induction in each sample.

Anchorage-independent growth in soft agar

A soft agar assay was performed as described previously.39 Briefly, growth in soft agar was measured in 35-mm diameter dishes containing a lower layer of 0.7% agar solution in Dulbecco’s modified Eagle’s medium (DMEM) with 10% FBS and overlaid with a 0.35% agar solution in the same growth medium, in which 1 × 105 cells were resuspended. The soft agar was covered with the culture medium (Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum).

Immunoblotting analysis and immunoprecipitation

Western blot analysis was performed by conventional methods. Antibodies used were as follows: rabbit anti-human CAR (Sigma-Aldrich Co. LLC., St Louis, MO, USA), mouse anti-human E-cadherin (BD Biosciences), mouse anti-human vimentin, mouse anti-human fibronectin, rabbit anti-human beta-catenin (Abcam Plc, Cambridge, UK), rabbit anti-human Snail (Cell Signaling Technology, Beverly, MA, USA), mouse anti-Myc-tag (clone 9E11; ATCC), rabbit anti-GFP (Clontech, Mountain View, CA, USA), mouse anti-human ROCKI and ROCKII (Santa Cruz Biotechnology, Santa Cruz, CA, USA) and mouse anti-human beta-Actin antibody (Millipore Co., Billerica, MA, USA). The second antibody was horseradish peroxidase-conjugated anti-mouse or anti-rabbit IgG antibody (Zymax; Life Technologies Japan Ltd., Tokyo, Japan). Positive signals were detected by a chemiluminescence system (SuperSinal West Pico substrate; Thermo Fisher Scientific Inc.).

Monoclonal Anti-His tag (clone 2D8; ATCC) and streptavidin agarose (Invitrogen, Carlsbad, CA, USA) were used for the co-immunoprecipitation experiments. Monoclonal anti-Myc-tag agarose (clone 1G4; MBL, Nagoya, Japan) was used to purify recombinant human CAR protein expressed in HEK293T cells (Figure 6e, left).

RNA interference

Human CAR siRNA (siCAR: sense, 5′-IndexTermrGrGUrArAUrArGrGrGrACUUrArGrCrArATT-3′; antisense, 5′-IndexTermUUrGrCUrArArGUrCrCrCUrAUUrArCrCTT)-3′40 and negative control scrambled siRNA (siCont.: sense, 5′-IndexTermUUrCUrCrArGrArGrGrGUrArArGrArArArGTT-3′; antisense, 5′-IndexTermrCUUUrCUUrArCrCrCUrCrUGrArGrArATT)-3′. These siRNAs were synthesized by Sigma-Aldrich (Tokyo, Japan). 10 nM of each siRNA were transfected using Lipofectamin RNAiMAX reagent (Life Technologies, Invitrogen).

Plasmid constructs

To construct a mammalian expression vector, the CMV promoter-intron (CMVi) from the phCMV-FSR vector (Genlantis, San Diego, CA, USA) was inserted into the promoter-less pDNR-1r vector (Clontech Takara, Moutain View, CA, USA) and this was named pDNR-CMVi. The vector could efficiently express cargo cDNAs. Human cDNA encoding wild-type full-length CAR was then inserted into the pDNR-CMVi vector.41 The CAR was designed to be express as a C-terminal Myc-6His-tagged form. Transient transfection of the plasmid to cultured cells was performed using FuGENE-HD (Promega Biosciences, San Luis Obispo, CA, USA) for HEK293T and HSC-2.

ROCK activity assay

Before performing ROCK activity assay, goat anti-human ROCKII (Santa Cruz Biotechnology) was biotinylated using a Biotin Labeling Kit-SH (Dojindo Molecular Technologies, Rockville, MD, USA) to recover antibody-free ROCKII protein after immunoprecipitation with streptavidin agarose (Figures 6e, right and g).

In vitro ROCK activity assay was performed according to the protocol as reported elsewhere42 but with slight modification. In brief, endogenous ROCKII was prepared by immunoprecipitation using the biotinylated ROCKII antibody prepared with streptavidin agarose from NHEK or HSC-2 cell lysates. The isolated ROCKII protein was mixed with 1 μg of recombinant MYPT1 (654–880 amino acid; Millipore, Billerica, MA, USA) as a substrate for ROCK, and the mixture was incubated in the kinase reaction buffer (1 mM DTT, 10 mM MgCl2, 5 mM Glycerol-2-Phsophate, 0.1 mM Na3VO4 and 25 mM Tris-HCl/pH 7.5) containing 0.2 mM ATP (total volume 50 μl) for 30 min at 30 °C with gentle agitation. The kinase reaction was stopped by adding 25 μl of 3 × SDS–PAGE Sample Buffer and subjected to SDS–PAGE. Phosphorylation of MYPT1 substrate was detected by rabbit anti-phospho (Thr696)-MYPT1 antibody (Cell Signaling Technology, Inc., Danvers, MA, USA).

Animal experiments

HSC-2 cells stably expressing GFP (GFP-HSC-2 cells) were newly cloned and used in animal experiments. The GFP-HSC-2 cells (0.6 × 106 or 1 × 106 cells) were implanted into the peritoneal cavity of female NOD-SCID mice at the age of 6 weeks (Charles River Laboratories Japan Inc., Yokohama, Japan). At two different time points (21 and 40 days), mice were autopsied and the distribution of the peritoneally disseminated tumors was macroscopically examined by fluorescence stereo microscope. The animal studies in the present work were approved by the Aichi Cancer Center on Animal Research. All mouse procedures and euthanasia, including cell transplantations, were done painlessly or under anesthesia, and within the strict guidelines of the Experimental Animal Committees of Aichi Cancer Center Research Institute.

Immunofluorescence staining

To visualize CAR and E-cadherin, cultured cells were fixed with 4% paraformaldehyde and washed with 0.1% Triton X-100 in phosphate-buffered saline. E-cadherin was stained with rabbit anti-CAR antibody (Priestige Antibody; Sigma-Aldrich Co. LLC.) and mouse anti-E-cadherin antibody (BD Biosciences) as primary antibodies, and further treated with FITC-conjugated goat anti-mouse IgG antibody and Cy3-conjugated goat anti-mouse IgG antibody (Jackson ImmunoResearch Laboratories Inc., West Grove, PA, USA) as secondary antibodies, respectively. Cell nuclei were counterstained with 4′, 6-diamidino-2-phenylindole. For the staining of actin filaments (F-actin), FITC-conjugated phalloidin (Molecular Probes Invitrogen, Carlsbad, CA, USA) was used. The fluorescent signals were detected by an Olympus IX71 fluorescence microscope and analyzed using cell Sens Standerd software (Olympus Corp., Tokyo, Japan). Cy3-labeled and FITC-labeled annexin V were purchased from BioVision Inc. (Milpitas, CA, USA). Briefly, 2 × 105 cells were collected and incubated with 500 μl of the annexin V-binding buffer containing 3 μl of Cy3-labeled (or FITC-labeled) annexin V for 5 min at room temperature. After washing with phosphate-buffered saline (−), they were subjected to fluorescence microscopic analysis.


Human oral, pharyngeal or laryngeal carcinoma tissues were fixed in 10% buffered formalin solution. Paraffin-embedded sections were deparaffinized, and antigen retrieval was performed by autocleaving (Pascal DakoCytomation S2800, Glostrup, Denmark) in sodium citrate buffer, and then endogenous peroxide activity was quenched by incubation in 3% hydrogen peroxide (Kanto Chemical Co., Inc., Tokyo, Japan) for 10 min at room temperature. Slides were incubated with rabbit anti-CAR antibody (at a dilution 1:500, Sigma-Aldrich) as a primary antibody overnight at 4 °C. After the slides were washed with 0.05% Tween-20 in phosphate-buffered saline (−), they were subjected to the DAB detection system (Envision+Kit/horseradish peroxidase; Dako North America, Inc., Carpinteria, CA, USA). The study using the patient-derived tissues was approved by the research ethics committees of the Aichi Cancer Center. Informed consent was also obtained from each patient for use of these materials.

Statistical analysis

Data are expressed as means±s.d. We employed simple pair-wise comparison with Student’s t-test (two-tailed distribution with two-sample equal variance). P<0.05 was considered significant.


  1. 1

    Bergelson JM, Cunningham JA, Droguett G, Kurt-Jones EA, Krithivas A, Hong JS et al. Isolation of a common receptor for coxsackie B viruses and adenoviruses 2 and 5. Science 1997; 275: 1320–1323.

  2. 2

    Mayr GA, Freimuth P . A single locus on human chromosome 21 directs the expression of a receptor for adenovirus type 2 in mouse A9 cells. J Virol 1997; 71: 412–418.

  3. 3

    Tomko RP, Xu R, Philipson L . HCAR and MCAR: the human and mouse cellular receptors for subgroup C adenoviruses and group B coxsackieviruses. Proc Natl Acad Sci USA 1997; 94: 3352–3356.

  4. 4

    Cohen CJ, Shieh CJT, Pickles RJ, Okegawa T, Hsieh J-T, Bergelson JM . The coxsackie and adenovirus receptor is a transmembrane component of the tight junction. Proc Natl Acad Sci USA 2001; 98: 15191–15196.

  5. 5

    Fok PT, Huang KC, Holland PC, Nalbantoglu J . The coxsackie and adenovirus receptor binds microtubules and plays a role in cell migration. J Biol Chem 2007; 282: 7512–7521.

  6. 6

    Stecker K, Vieth M, Koschel A, Wiedenmann B, Rocken C, Anders M . Impact of the coxackievirus and adenovirus receptor on the adenoma-carcinoma sequence of colon cancer. Br J Cancer 2011; 104: 1426–1433.

  7. 7

    Anders M, Vieth M, Rocken C, Ebert M, Pross M, Gretschel S et al. Loss of the coxsackie and adenovirus receptor contributes to gastric cancer progression. Br J Cancer 2009a; 100: 352–359.

  8. 8

    Anders M, Rosch T, Kuster K, Becker I, Hofler H, Stein HJ et al. Expression and function of the coxsackie and adenovirus receptor in Barrett’s esophagus and associated neoplasia. Cancer Gene Ther 2009b; 16: 508–515.

  9. 9

    Matsumoto K, Shariat SF, Ayala GE, Rauen KA, Lemer SP . Loss of coxsackie and adenovirus receptor expression is associated with features of aggressive bladder cancer. Urology 2005; 66: 441–446.

  10. 10

    Yamashita M, Ino A, Kawabata K, Sakurai F, Mizuguchi H . Expression of coxsackie and adenovirus receptor reduces the lung metastatic potential of murine tumor cells. Int J Cancer 2007; 121: 1690–1696.

  11. 11

    Okegawa T, Pong RC, Li Y, Bergelson JM, Sagalowsky AI, Hsieh JT . The mechanism of the growth-inhibitory effect of coxsackie and adenovirus receptor (CAR) on human bladder cancer: a functional analysis of car protein structure. Cancer Res 2001; 61: 6592–6600.

  12. 12

    Okegawa T, Li Y, Pong RC, Bergelson JM, Zhou J, Hsieh JT . The dual impact of coxsackie and adenovirus receptor expression on human prostate cancer gene therapy. Cancer Res 2000; 60: 5031–5036.

  13. 13

    Giaginis C, Zarros A, Alexandrou P, Klijanienko J, Delladetsima I, Theocharis S . Evaluation of coxsakievirus and adenovirus receptor expression in human benign and malignant thyroid lesions. APMIS 2010; 118: 210–221.

  14. 14

    Bruning A, Stickeler E, Diederich D, Walz L, Rohleder H, Friese K et al. Coxsackie and adenovirus receptor promotes adenocarcinoma cell survival and is expressionally activated after transition from preneoplastic precursor lesions to invasive adenocarcinomas. Clin Cancer Res 2005; 11: 4316–4320.

  15. 15

    Qin M, Escuadro B, Dohadwala M, Sharma S, Batra RK . A novel role for the coxsakie adenovirus receptor in mediating tumor formation by lung cancer cells. Cancer Res 2004; 64: 6377–6380.

  16. 16

    Verdino P, Witherden DA, Havran WL, Wilson IA . The molecular interaction of CAR and JAML recruits the central cell signal transducer PI3K. Science 2010; 329: 1210–1214.

  17. 17

    Street CA, Bryan BA . Rho kinase proteins-pleiotropic modulators of cell survival and apoptosis. Anticancer Res 2011; 31: 3645–3657.

  18. 18

    Frisch SM, Screaton RA . Anoikis mechanisms. Curr Opin Cell Biol 2001; 13: 555–562.

  19. 19

    Stoff A, Rivera AA, Banerjee NS, Mathis JM, Espinosa-de-los-Monteros A, Le LP et al. Strategies to enhance transductional efficiency of adenoviral-based gene transfer to primary human fibroblasts and keratinocytes as a platform in dermal wounds. Wound Repair Regen 2006; 14: 608–617.

  20. 20

    Sakaguchi M, Miyazaki M, Takaishi M, Sakaguchi Y, Makino E, Kataoka N et al. S100C/A11 is a key mediator of Ca(2+)-induced growth inhibition of human epidermal keratinocytes. J Cell Biol 2003; 163: 825–835.

  21. 21

    Shi J, Wei L . Rho kinase in the regulation of cell death and survival. Arch Immunol Ther Exp (Warsz) 2007; 55: 61–75.

  22. 22

    Lai JM, Hsieh CL, Chang ZF . Caspase activation during phorbol ester-induced apoptosis requires ROCK-dependent myosin-mediated contraction. J Cell Sci 2003; 116: 3491–3501.

  23. 23

    Marconi A, Atzei P, Panza C, Fila C, Tiberio R, Truzzi F et al. FLICE/caspase-8 activation triggers anoikis induced by β1-integrin blockade in human keratinocytes. J Cell Sci 2004; 117: 5815–5823.

  24. 24

    Valentijn AJ, Gilmore AP . Translocation of full-length Bid to mitochondria during anoikis. J Biol Chem 2004; 279: 32848–32857.

  25. 25

    Dorner AA, Wegmann F, Butz S, Wolburg-Buchholz K, Wolburg H, Mack A et al. Coxsackievirus-adenovirus receptor (CAR) is essential for early embryonic cardiac development. J Cell Sci 2005; 118: 3509–3521.

  26. 26

    Walters RW, Freimuth P, Moninger TO, Ganske I, Zabner J, Welsh MJ . Adenovirus fiber disrupts CAR-mediated intercellular adhesion allowing virus escape. Cell 2002; 110: 789–799.

  27. 27

    Huang KC, Altinoz M, Wosik K, Larochelle N, Koty Z, Zhu L et al. Impact of the coxsackie and adenovirus receptor (CAR) on glioma cell growth and invasion: requirement for the C-terminal domain. Int J Cancer 2005; 113: 738–745.

  28. 28

    Mege RM, Gavard J, Lambert M . Regulation of cell-cell junctions by the cytoskeleton. Curr Opin Cell Biol 2006; 18: 541–548.

  29. 29

    Yonemura S . Cadherin-actin interactions at adherens junctions. Curr Opin Cell Biol 2011; 23: 515–522.

  30. 30

    Hermiston ML, Gordon JI . In vivo analysis of cadherin function in the mouse intestinal epithelium: essential roles in adhesion, maintenance of differentiation, and regulation of programmed cell death. J Cell Biol 1995; 129: 489–506.

  31. 31

    Fouquet S, Lugo-Martinez VH, Faussat AM, Renaud F, Cardot P, Chambaz J et al. Early loss of E-cadherin from cell-cell contacts is involved in the onset of Anoikis in enterocytes. J Biol Chem 2004; 279: 43061–43069.

  32. 32

    Excoffon KJ, Hruska-Hageman A, Klotz M, Traver GL, Zabner J . A role for the PDZ-binding domain of the coxsackie B virus and adenovirus receptor (CAR) in cell adhesion and growth. J Cell Sci 2004; 117: 4401–4409.

  33. 33

    Coyne CB, Voelker T, Pichla SL, Bergelson JM . The coxsackievirus and adenovirus receptor interacts with the multi-PDZ domain protein-1 (MUPP-1) within the tight junction. J Biol Chem 2004; 279: 48079–48084.

  34. 34

    Sollerbrant K, Raschperger E, Mirza M, Engstrom U, Philipson L, Ljungdahl PO et al. The Coxsackievirus and adenovirus receptor (CAR) forms a complex with the PDZ domain-containing protein ligand-of-numb protein-X (LNX). J Biol Chem 2003; 278: 7439–7444.

  35. 35

    Maekawa M, Ishizaki T, Boku S, Watanabe N, Fujita A, Iwamatsu A et al. Signaling from Rho to the actin cytoskeleton through protein kinases ROCK and LIM-kinase. Science 1999; 285: 895–898.

  36. 36

    Shewan AM, Maddugoda M, Kraemer A, Stehbens SJ, Verma S, Kovacs EM et al. Myosin 2 is a key Rho kinase target necessary for the local concentration of E-cadherin at cell–cell contacts. Mol Biol Cell 2005; 16: 4531–4542.

  37. 37

    Schackmann RC, van Amersfoort M, Haarhuis JH, Vlug EJ, Halim VA, Roodhart JM et al. Cytosolic p120-catenin regulates growth of metastatic lobular carcinoma through Rock1-mediated anoikis resistance. J Clin Invest 2011; 121: 3176–3188.

  38. 38

    Maseki S, Ijichi K, Tanaka H, Fujii M, Hasegawa Y, Ogawa T et al. Acquisition of EMT phenotype in the gefitinib-resistant cells of a head and neck squamous cell carcinoma cell line through Akt/GSK-3β/snail signalling pathway. Br J Cancer 2012; 106: 1196–1204.

  39. 39

    Cheng HL, Su SJ, Huang LW, Hsieh BS, Hu YC, Hung TC et al. Arecoline induces HA22T/VGH hepatoma cells to undergo anoikis-involvement of STAT3 and RhoA activation. Mol Cancer 2010; 9: 126.

  40. 40

    Vigl B, Zgraggen C, Rehman N, Banziger-Tobler NE, Detmar M, Halin C . Coxsackie- and adenovirus receptor (CAR) is expressed in lymphatic vessels in human skin and affects lymphatic endothelial cell function in vitro. Exp Cell Res 2009; 315: 336–347.

  41. 41

    Sakaguchi M, Murata H, Yamamoto K, Ono T, Sakaguchi Y, Motoyama A et al. TIRAP, an adaptor protein for TLR2/4, transduces a signal from RAGE phosphorylated upon ligand binding. PLoS One 2011; 6: e23132.

  42. 42

    Xiao L, Eto M, Kazanietz MG . ROCK mediates phorbol ester-induced apoptosis in prostate cancer cells via p21Cip1 up-regulation and JNK. J Biol Chem 2009; 284: 29365–29375.

Download references


We thank Dr Hitoshi Ohno (Takeda General Hospital, Kyoto, Japan), Professor Hirohisa Yano (Department of Pathology, Kurume University School of Medicine, Japan) and Dr Atsushi Natsume (Department of Neurosurgery, Nagoya University Graduate School of Medicine, Japan) for kindly providing us lymphoma cell line, hepatocellular carcinoma cell lines and glioma cell lines, respectively, and Dr Yasushi Yatabe (Department of Pathology and Molecular Diagnostics, Aichi Cancer Center Hospital) for advice.

Author information

Correspondence to E Kondo.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Supplementary Information accompanies this paper on the Oncogene website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Saito, K., Sakaguchi, M., Iioka, H. et al. Coxsackie and adenovirus receptor is a critical regulator for the survival and growth of oral squamous carcinoma cells. Oncogene 33, 1274–1286 (2014).

Download citation


  • CAR
  • ROCK
  • anoikis
  • SCC

Further reading