**3. Role of cytokines in oral cancer cell proliferation**

**Figure 3.** Schematic representation of coxsackievirus and adenovirus receptor on the tumor cells.

Finally, it was suggested that NCAM might be associated with not only a cell-to-cell adhesion mechanism but also tumorigenesis, including the occurrence, development, and perineural/

Further studies will be required to identify the signal transduction pathways by which treatment with cimetidine suppresses the growth of salivary gland tumors and to establish a strat-

Coxsackievirus and adenovirus receptor (CAR/CXADR), a transmembrane glycoprotein, was initially characterized as a viral attachment site on the surface of epithelial cells (**Figure 3**) [46]. Later it was identified as a component of the tight junction (TJ) complex, an interacting partner for a number of other TJ proteins and a regulator of TJ formation [47–52]. Furthermore, CAR is known to be a cell-cell adhesion molecule [53, 54]. In terms of function, loss of CAR has been considered to diminish intercellular adhesion, increase proliferation, and promote the migration as well as invasion of cancer cells [55, 56]. On the basis of these observations, a tumor-suppressive role of CAR in human cancers has speculated. Although it has recently been described [55–58] that CAR is observed in various organs, it is still unclear whether it is expressed in oral cancer. Therefore, we examined the role of CAR in SCC in the oral cavity (data not shown). This revealed that CAR was constitutively expressed in five oral SCC cell lines. To analyze the function of CAR, we then examined the proliferative activity of SAS cells

neural invasion of human salivary gland tumors.

88 Prevention, Detection and Management of Oral Cancer

egy for cimetidine-based therapy for those tumors.

**2.2. Coxsackievirus and adenovirus receptor (CAR/CXADR)**

Cytokines are composed of a large family of secreted proteins that bind to and signal through defined cell surface receptors on a wide variety of target cells, playing an important role in the maintenance of homeostasis. Furthermore, many cytokines share structural features and effects during inflammation, development, or immune responses.

The concept of a control mechanism for cellular growth via regulation of apoptosis has recently been erected in a wide variety of tissue systems. Changes in the balance between cell survival and death are definite signs of emergence of various tumors. Therefore, modulation of apoptosis is required so as to maintain the homeostasis of a living organism. The expression of cytokines and their receptors in human oral cancers has attracted a great deal of interest because of their potential importance in tumor immunity. In particular, it has been described that members of the tumor necrosis factor (TNF) family, including Fas/FasL and TNF-related apoptosis-inducing ligand (TRAIL), regulate the deletion of unnecessary immune cells through induction of apoptosis [59–61]. However, despite their expression of these obvious antigens, tumor evasion by the immune system is often inefficient. It is considered that tumor cells may also evade immune attack by expressing TRAIL, Fas ligand, or other molecules that induce apoptosis in activated T cells [62].

### **3.1. Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)**

TRAIL, also called APO2 ligand (APO2L), is a novel member of the TNF cytokine family that was originally characterized by its ability to induce apoptosis [59, 60]. It is recognized that at least four closely related receptors bind to TRAIL: death receptor-4 (DR4) and DR5/ KILLER, which contain cytoplasmic death domains and signal apoptosis [60, 61]; decoy receptor-1 (DcR1) [61–63], which lacks a cytoplasmic tail and inhibits TRAIL function; and DcR2 [64, 65], which contains a cytoplasmic region with a truncated death domain that does not transduce the death signal [67]. TRAIL interacts with its agonistic receptors DR4 and DR5, inducing apoptosis in a wide variety of cancer cell lines derived from breast carcinoma, lung carcinoma, colon carcinoma, lymphoma, malignant melanoma, and malignant glioma [59, 60, 68, 69]. Although DR4 is expressed in many normal human tissues and cells, including spleen and peripheral blood leukocytes, TRAIL induces apoptosis in various cancer cells, but not in normal cells [70]. This may be explained by the fact that TRAIL also interacts with the antagonistic decoy receptors DcR1 and DcR2, which are expressed in normal tissues but not in cancer cells (**Figure 5**) [66, 67]. Neither DcR1 nor DcR2 receptors induce apoptosis, but they protect cells from TRAIL-induced apoptosis [64–66]. Until now, the biological involvement of the complex TRAIL receptor system has remained unclear, and the existing data are conflicting. Nevertheless, because of its selective cytotoxicity against tumor cells, TRAIL is regarded as a promising anticancer weapon that might be highly effective *in vivo* with few side effects, as it has little or no function on normal tissues.

samples (48%) of HOSCC tissue, and there was no correlation among the WHO grades. These findings suggest that HOSCC has the potential to escape immune surveillance by killing host T lymphocytes via DR4/TRAIL and DR5/TRAIL interactions, as suggested for FasL [72–74].

The Mechanisms of Proliferation and Energy Metabolism in Oral Cancer

http://dx.doi.org/10.5772/intechopen.87091

91

**Figure 5.** Schematic representation of the death receptors and decoy receptors on the tumor cell.

The expression of decoy receptors in cancer cells is, however, a phenomenon that objects against previous reports [62, 63]. Indeed, several authors have currently described that a decoy receptor is expressed in various cancer types and our results are consistent with their findings [75–77]. Therefore, cancer cells may also avoid TRAIL-induced apoptosis by express-

On the other hand, cells differ significantly in their response to TRAIL. As contrasted to HUVEC or other oral cancer cells, only KB cells undergo significant apoptosis following exposure to recombinant human (rh)TRAIL. The reason why KB cells, despite their expression of

Commonly, the most proximal step in suppression of a death receptor pathway is inhibition of ligand binding. This may be acquired by lack, or the presence of decoy receptors [65–68], or mutations of death receptors [78, 79]. However, it has recently been described that there is no correlation between the expression of TRAIL receptor and susceptibility to TRAIL-induced apoptosis in various cancer types [75–77]. Furthermore, the existence of antiapoptotic proteins, such as bcl-2, bcl-xL, and/or fas-like IL-1-converting enzyme (FLICE)-like inhibitory protein [80, 81], also seems to be significant, as they are resistant to death receptor-mediated

ing a decoy receptor.

a decoy receptor, react to rhTRAIL remains unresolved.

TRAIL is expressed in most normal human cells and tissues, including the peripheral blood leukocytes, spleen, lung, and prostate, but not the brain [60]. However, the expression of TRAIL in human neoplasms is largely unknown. Accordingly, we have examined whether TRAIL and its receptors are expressed in HOSCC tissues or cell lines and whether these cell lines are sensitive to TRAIL-induced apoptosis [71]. This revealed that the mRNA and protein levels of TRAIL and its receptors are co-expressed in HOSCC cell lines in the absence of paracrine fratricide or autocrine suicide. Moreover, TRAIL protein was also detected in 24 of 50

**Figure 5.** Schematic representation of the death receptors and decoy receptors on the tumor cell.

the maintenance of homeostasis. Furthermore, many cytokines share structural features and

The concept of a control mechanism for cellular growth via regulation of apoptosis has recently been erected in a wide variety of tissue systems. Changes in the balance between cell survival and death are definite signs of emergence of various tumors. Therefore, modulation of apoptosis is required so as to maintain the homeostasis of a living organism. The expression of cytokines and their receptors in human oral cancers has attracted a great deal of interest because of their potential importance in tumor immunity. In particular, it has been described that members of the tumor necrosis factor (TNF) family, including Fas/FasL and TNF-related apoptosis-inducing ligand (TRAIL), regulate the deletion of unnecessary immune cells through induction of apoptosis [59–61]. However, despite their expression of these obvious antigens, tumor evasion by the immune system is often inefficient. It is considered that tumor cells may also evade immune attack by expressing TRAIL, Fas ligand, or other molecules that

TRAIL, also called APO2 ligand (APO2L), is a novel member of the TNF cytokine family that was originally characterized by its ability to induce apoptosis [59, 60]. It is recognized that at least four closely related receptors bind to TRAIL: death receptor-4 (DR4) and DR5/ KILLER, which contain cytoplasmic death domains and signal apoptosis [60, 61]; decoy receptor-1 (DcR1) [61–63], which lacks a cytoplasmic tail and inhibits TRAIL function; and DcR2 [64, 65], which contains a cytoplasmic region with a truncated death domain that does not transduce the death signal [67]. TRAIL interacts with its agonistic receptors DR4 and DR5, inducing apoptosis in a wide variety of cancer cell lines derived from breast carcinoma, lung carcinoma, colon carcinoma, lymphoma, malignant melanoma, and malignant glioma [59, 60, 68, 69]. Although DR4 is expressed in many normal human tissues and cells, including spleen and peripheral blood leukocytes, TRAIL induces apoptosis in various cancer cells, but not in normal cells [70]. This may be explained by the fact that TRAIL also interacts with the antagonistic decoy receptors DcR1 and DcR2, which are expressed in normal tissues but not in cancer cells (**Figure 5**) [66, 67]. Neither DcR1 nor DcR2 receptors induce apoptosis, but they protect cells from TRAIL-induced apoptosis [64–66]. Until now, the biological involvement of the complex TRAIL receptor system has remained unclear, and the existing data are conflicting. Nevertheless, because of its selective cytotoxicity against tumor cells, TRAIL is regarded as a promising anticancer weapon that might be highly effective *in vivo* with few side effects,

TRAIL is expressed in most normal human cells and tissues, including the peripheral blood leukocytes, spleen, lung, and prostate, but not the brain [60]. However, the expression of TRAIL in human neoplasms is largely unknown. Accordingly, we have examined whether TRAIL and its receptors are expressed in HOSCC tissues or cell lines and whether these cell lines are sensitive to TRAIL-induced apoptosis [71]. This revealed that the mRNA and protein levels of TRAIL and its receptors are co-expressed in HOSCC cell lines in the absence of paracrine fratricide or autocrine suicide. Moreover, TRAIL protein was also detected in 24 of 50

effects during inflammation, development, or immune responses.

**3.1. Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)**

induce apoptosis in activated T cells [62].

90 Prevention, Detection and Management of Oral Cancer

as it has little or no function on normal tissues.

samples (48%) of HOSCC tissue, and there was no correlation among the WHO grades. These findings suggest that HOSCC has the potential to escape immune surveillance by killing host T lymphocytes via DR4/TRAIL and DR5/TRAIL interactions, as suggested for FasL [72–74].

The expression of decoy receptors in cancer cells is, however, a phenomenon that objects against previous reports [62, 63]. Indeed, several authors have currently described that a decoy receptor is expressed in various cancer types and our results are consistent with their findings [75–77]. Therefore, cancer cells may also avoid TRAIL-induced apoptosis by expressing a decoy receptor.

On the other hand, cells differ significantly in their response to TRAIL. As contrasted to HUVEC or other oral cancer cells, only KB cells undergo significant apoptosis following exposure to recombinant human (rh)TRAIL. The reason why KB cells, despite their expression of a decoy receptor, react to rhTRAIL remains unresolved.

Commonly, the most proximal step in suppression of a death receptor pathway is inhibition of ligand binding. This may be acquired by lack, or the presence of decoy receptors [65–68], or mutations of death receptors [78, 79]. However, it has recently been described that there is no correlation between the expression of TRAIL receptor and susceptibility to TRAIL-induced apoptosis in various cancer types [75–77]. Furthermore, the existence of antiapoptotic proteins, such as bcl-2, bcl-xL, and/or fas-like IL-1-converting enzyme (FLICE)-like inhibitory protein [80, 81], also seems to be significant, as they are resistant to death receptor-mediated apoptosis. Holistically, our results also suggest that there may be no correlation between the expression of TRAIL receptor and sensitivity to TRAIL-induced apoptosis in HOSCC cell lines and that TRAIL-resistant cells (HSC-2, HSC-3, HSC-4, and Ca9-22) may express cytoprotective proteins that block TRAIL-induced apoptosis or that the apoptotic effect of TRAIL is regulated by other mechanisms. It has also been described that TRAIL, in combination with an anticancer drug, acts cooperatively to induce apoptosis in various cancer cells that are resistant to TRAIL or chemotherapy [75, 82, 83]. This combination of TRAIL with chemotherapeutic reagents might be a useful therapeutic strategy against TRAIL-resistant cell lines such as HSC-2, HSC-3, HSC-4, and Ca9-22 used in our investigation.

microRNA-mediated dysregulation) of MET [92–94]. c-Met overexpression and MET amplification are thought to be associated with a poorer prognosis in some types of tumors, including non-small cell lung cancer and gastric cancer [94]. In HOSCC, it has also been reported that c-Met expression is associated with cisplatin resistance and a strong propensity for metastasis *in vivo* [95], as well as a poor prognosis [96]. However, details of the involvement of c-Met in oral carcinogenesis are still unclear. Accordingly, we investigated how the relationship between the expression of c-Met and several tumor activation-related markers such as NF-κB is associated with oral carcinogenesis (data not shown). In addition, the expression and distribution of c-Met and NF-κB were also examined in HOSCC tissues (data not shown). The results of real-time qRT-PCR and immunoblot analysis indicated overexpression of c-Met mRNA and protein in SAS cells. Therefore, SAS cells were used in this study. To investigate how c-Met functions in SAS cells, c-Met knockdown analysis was performed. c-Met knockdown appeared to reduce the number of SAS cells. To confirm whether this had been due to apoptosis, caspase activity was then analyzed, and this revealed that apoptosis had indeed occurred via activation of caspases-9 and caspases-3/caspases-7 in SAS by c-Met knockdown. Furthermore, SAS showed cell cycle arrest at S/G2/M phase during this apoptotic cell death. Subsequently, to determine NF-κB expression after c-Met knockdown, we also used a siRNA approach to reduce the expression of c-Met and determine the effects on NF-κB activity. As expected, the level of c-Met mRNA was markedly reduced by c-Met siRNA. Moreover, c-Met knockdown by c-Met siRNA clearly decreased the activation of NF-κB mRNA in SAS cells, in comparison with controls. These data indicated that c-Met upregulated NF-κB activation and consequently that c-Met knockdown led to apoptosis of SAS cells. These combined data suggested that c-Met produced by autologous cancer cells promoted tumor growth. Furthermore, our *in vivo* studies demonstrated c-Metspecific immunoreactivity, consistent with the observation of NF-κB-positive cells in HOSCC biopsy samples (c-Met expression, 9/20 (45%); NF-κB expression, 18/20 cases (90%)). This result suggests that c-Met expression correlates with increased activation of NF-κB. Based on these in vitro and *in vivo* observations, it can be hypothesized that c-Met function leads to NF-κB activation and subsequently anti-apoptosis and that as a consequence, it may be associated

The Mechanisms of Proliferation and Energy Metabolism in Oral Cancer

http://dx.doi.org/10.5772/intechopen.87091

93

with tumorigenesis, including growth, development, and angiogenesis in HOSCC.

**4. Energy metabolism in oral cancer**

**4.1. Glucose metabolism in oral cancer**

Further investigations of the role of c-Met will be required to fully understand c-Met-mediated tumor proliferation and to establish a therapeutic strategy for c-Met-based oral cancer.

In normal tissues, adenosine triphosphate (ATP) is mainly produced in mitochondria via complete oxidative phosphorylation (OXPHOS) of glucose. Conversely, only 10% of ATP is produced from glycolysis in which glucose is replaced to lactate [97]. Interestingly, cancer tissues possess high levels of glycolysis in the cytosol even under aerobic conditions, which is upregulated by PI3K/Akt signaling in the mitochondria, a phenomenon known as the "Warburg effect" or "aerobic glycolysis" [98–103]. More than 80 years ago, Otto Warburg

In fact, we are currently investigating the synergistic effects of α-mangostin and TRAIL on induction of apoptosis via the mitochondrial pathway in squamous cell carcinoma of the oral cavity [84]. To summarize, mangosteen (*Garcinia mangostana*) is a tree discovered in Southeast Asia, and the pericarp of its fruit has been used in folk medicine for the treatment of many human diseases. The rinds of mangosteen fruit contain a high concentration of xanthone, a type of polyphenol. One form of xanthone, α-mangostin, has been described to exhibit chemopreventive effects against chemically induced colon cancer through a decrease of c-Myc expression, suppressing tumor growth in a mouse model of mammary cancer. A recent study has proved the inhibitory effect of α-mangostin on the growth of prostate cancer. However, it is still unclear whether α-mangostin induces cell death in oral cancer. Then, the present study examined the impact of α-mangostin on HOSCC. First, we analyzed the expression of c-Myc in five HOSCC cell lines (HSC-2, HSC-3, HSC-4, Ca9-22, and SAS). The highest level of c-Myc mRNA expression was found in SAS cells and the lowest in HSC-4 cells. Therefore, SAS cells were treated with α-mangostin, which was observed to exert a weak cytocidal effect. Since α-mangostin has been described to exert synergistic effects on cancers when combined with anticancer drugs, we tried to evaluate these synergistic effects of α-mangostin in combination with TRAIL. We found that this combination induced apoptosis in SAS cells through the mitochondrial pathway via activation of caspase-3/caspase-7 and caspase-9, following the release of cytochrome c. In addition, this apoptosis was induced by S/G2/M-phase arrest. Immunoreactivity for c-Myc was revealed in the cytoplasm of cancer cells in 16 (40%) of the 40 cases of HOSCC. These data showed that the combination of α-mangostin and TRAIL may have considerable potential for the treatment of oral cancer.

Further investigation of TRAIL-mediated cell death, including the interaction of TRAIL and its receptors in oral cancer cells under various conditions, will be required to establish a strategy for TRAIL-based oral cancer therapy, which does not cause liver toxicity.

#### **3.2. Hepatocyte growth factor (HGF) and its receptor, c-Met**

The tyrosine kinase receptor c-Met ordinarily binds with hepatocyte growth factor (HGF), which triggers its involvement in processes such as cell differentiation, cell growth, angiogenesis, and embryogenesis [85, 86]. However, c-Met activation is also associated with processes related to malignant transformation, such as invasion, tumor growth, angiogenesis, and metastasis [87–91]. In addition to autocrine or paracrine signaling via HGF, c-Met may also be activated via the mutation, protein overexpression or amplification, or transcriptional alteration (via microRNA-mediated dysregulation) of MET [92–94]. c-Met overexpression and MET amplification are thought to be associated with a poorer prognosis in some types of tumors, including non-small cell lung cancer and gastric cancer [94]. In HOSCC, it has also been reported that c-Met expression is associated with cisplatin resistance and a strong propensity for metastasis *in vivo* [95], as well as a poor prognosis [96]. However, details of the involvement of c-Met in oral carcinogenesis are still unclear. Accordingly, we investigated how the relationship between the expression of c-Met and several tumor activation-related markers such as NF-κB is associated with oral carcinogenesis (data not shown). In addition, the expression and distribution of c-Met and NF-κB were also examined in HOSCC tissues (data not shown). The results of real-time qRT-PCR and immunoblot analysis indicated overexpression of c-Met mRNA and protein in SAS cells. Therefore, SAS cells were used in this study. To investigate how c-Met functions in SAS cells, c-Met knockdown analysis was performed. c-Met knockdown appeared to reduce the number of SAS cells. To confirm whether this had been due to apoptosis, caspase activity was then analyzed, and this revealed that apoptosis had indeed occurred via activation of caspases-9 and caspases-3/caspases-7 in SAS by c-Met knockdown. Furthermore, SAS showed cell cycle arrest at S/G2/M phase during this apoptotic cell death. Subsequently, to determine NF-κB expression after c-Met knockdown, we also used a siRNA approach to reduce the expression of c-Met and determine the effects on NF-κB activity. As expected, the level of c-Met mRNA was markedly reduced by c-Met siRNA. Moreover, c-Met knockdown by c-Met siRNA clearly decreased the activation of NF-κB mRNA in SAS cells, in comparison with controls. These data indicated that c-Met upregulated NF-κB activation and consequently that c-Met knockdown led to apoptosis of SAS cells. These combined data suggested that c-Met produced by autologous cancer cells promoted tumor growth. Furthermore, our *in vivo* studies demonstrated c-Metspecific immunoreactivity, consistent with the observation of NF-κB-positive cells in HOSCC biopsy samples (c-Met expression, 9/20 (45%); NF-κB expression, 18/20 cases (90%)). This result suggests that c-Met expression correlates with increased activation of NF-κB. Based on these in vitro and *in vivo* observations, it can be hypothesized that c-Met function leads to NF-κB activation and subsequently anti-apoptosis and that as a consequence, it may be associated with tumorigenesis, including growth, development, and angiogenesis in HOSCC.

Further investigations of the role of c-Met will be required to fully understand c-Met-mediated tumor proliferation and to establish a therapeutic strategy for c-Met-based oral cancer.
