Metabotropic Glutamate Receptor 4 Expression in Colorectal Carcinoma and Its Prognostic Significance

Glutamate is an essential amino acid that plays important roles in signaling as a major excitatory neurotransmitter at neuronal synapses. The action of glutamate is mediated by the glutamate receptors (1), which are divided into two major groups—ionotropic and metabotropic receptors. The ionotropic receptors are cation-specific ion channels that mediate rapid synaptic transmission (1).

The metabotropic glutamate receptors (mGluR) are much slower in their responses, which occur through a variety of second messenger cascades via G proteins (1, 2). To date, eight mGluRs have been cloned and classified into three subtypes based on sequence information and intracellular effector systems (3). Group I receptors (mGluR1 and mGluR5) are coupled to phospholipase C and stimulate the production of inositol (1,4,5)-triphosphate and diacylglycerol, leading to activation of protein kinase C. Group II (mGluR2 and mGluR3) and group III receptors (mGluR4, mGluR6, mGluR7, and mGluR8) initiate the inhibitory cyclic AMP cascade (1–3).

Abnormal glutamate signaling has been linked to the pathogenesis of several human psychiatric and neurologic disorders (4). In addition, glutamate is involved in a number of important physiologic functions, including sensory perception, memory, and learning; it is also involved in regulating developmental functions, such as proliferation, migration, and survival of neuronal progenitors and neurons (4–6).

Recently, glutamate receptors have also been identified in peripheral nonneuronal tissues, including bone (7), skin (8), and pancreas (9); glutamate signaling has been implicated in differentiation of osteoblasts, proliferation of keratinocytes, and regulation of insulin secretion in these organs, respectively (7–9). Moreover, glutamate and its receptors have been reported to play roles in development of melanoma in mice (10) and to promote growth of malignant glioma cells in vitro (11). In the previous study, we identified overexpression of mGluR4 in a 5-fluorouracil–resistant colon cancer cell line, compared with its parental cell line SNU-769A (12). However, concepts such as the peripheral distribution of mGluR4 in human tissue and its roles in nonneuronal tissue are less well understood. Here, we report the expression of mGluR4 in nonneuronal tissues, such as normal and carcinoma tissues of colon, and altered expression of mGluR4 associated with various clinicopathologic parameters in colorectal carcinoma.

Patients and tissue samples. Twenty-one normal and 312 malignant tissue samples were obtained from 21 organs, including salivary gland, oral cavity, larynx, lung, esophagus, stomach, colon, liver, gallbladder, pancreas, breast, uterine cervix, endometrium, ovary, testis, prostate, kidney, urinary bladder, thyroid, adrenal gland, and skin. All tissue samples were taken from the files of the Department of Pathology, Seoul National University Hospital, South Korea, with written informed consent. The studied malignant tissues consisted of 287 carcinomas, 17 germ cell tumors (12 seminoma of testis, 4 dysgerminoma, and 1 yolk sac tumor of ovary), and 8 malignant skin melanomas.

In addition, 241 surgically resected colorectal carcinoma cases were obtained from Seoul National University Hospital, Seoul, South Korea, in 1998. A review of clinical charts and pathologic reports was done to obtain clinicopathologic data. All cases were adenocarcinomas and were classified according to WHO criteria (13) and staged according to the criteria of the International Union Against Cancer (14). The clinical outcomes of the colorectal cancer patients were followed from the date of operation until death or December 31, 2003. Mean follow-up time was 43 months (range, 2-74 months). Recurrent disease was defined as either local relapse or recurrence with distant metastases. Disease recurrence was detected by sonographic ultrasound, computed tomography, or magnetic resonance imaging, and was confirmed by pathologic examination. The microsatellite instability status of 230 of the enrolled cases had been previously reported (15); two cases were additionally analyzed for microsatellite instability at the BAT-25 and BAT-26 markers as previously described (15).

Immunohistochemistry. All tissues were routinely fixed in 10% buffered formalin and embedded in paraffin blocks. Core tissue biopsies of 1 or 2 mm in diameter were taken from individual paraffin-embedded tissues (donor blocks) and arranged in new recipient paraffin blocks (tissue array block) using a trephine apparatus (Superbiochips Laboratories, Seoul, South Korea). Because it has already been shown that a single sample from each tumor was sufficient to identify protein expression or molecular alteration related to clinical outcome, we sampled a tissue core from each case (16).

Immunostaining was done using the avidin-biotin peroxidase complex method. After antigen retrieval process using a citrate buffer solution (antigen unmasking solution; Vector Laboratories, Burlingame, CA) for 15 minutes in an 800 W microwave oven, polyclonal rabbit anti-mGluR4a (dilution 1:1,000; Upstate Biotechnology, Lake Placid, NY) and monoclonal mouse anti-p53 (clone DO7; dilution 1:100; Vector Laboratories) antibodies were applied. Cytoplasmic or membranous expression was regarded as positive for mGluR4 expression and the percentages of stained tumor cells were assigned the following scores: −, 0%; +, 1 to 50%; ++, >50%. As for p53, staining at >10% of tumor cell nuclei was considered positive (17). Two pathologists (H.J. Chang and W.H. Kim), without knowledge of the clinicopathologic data, did blind analysis of the immunostaining results.

Human colorectal cancer cell lines and Western blot analysis. The human colorectal cancer cell lines, SNU-61, SNU-81, SNU-407, SNU-1033, SNU-1047, SNU-C2A, SNU-C4, and SNU-C5 (18, 19), and the human fibrosarcoma cell line, HT1080, were obtained from the Korean Cell Line Bank (Seoul, South Korea). Total homogenates from these cell lines were analyzed by Western blot analysis as previously described (12) using either anti-mGluR4 (dilution 1:2,000; Upstate Biotechnology) or anti–β-actin (dilution 1:50,000; Sigma, Saint Louis, MO) as the primary antibodies. To quantify the level of protein expression, the intensities of mGluR4- and actin-immunoreactive signals were calculated using GelCompar II software (Applied Maths, Kortrijk, Belgium).

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide and cell invasion assay. The effect of the mGluR4 agonist, l-2-amino-4-phosphonobutyric acid (L-AP 4; Tocris Cookson, Ltd., Avonmouth, United Kingdom), and the mGluR4 antagonist, (S)-amino-2-methyl-4-phosphonobutanoic acid (MAP 4, Tocris Cookson), on cell proliferation was tested using an 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide colorimetric assay done as previously described (12). All experiments were done thrice, and the mean and SD of increased or decreased cell proliferation were calculated.

The cell invasion assay was done using a cell invasion assay kit (Chemicon, Temecula, CA), according to the manufacturer's instructions. Briefly, the assay was done in an invasion chamber consisting of a 24-well tissue culture plate with 12 cell culture inserts. A cell suspension in serum-free RPMI 1640 (Invitrogen, Carlsbad, CA) containing 2 mg/mL sodium bicarbonate, 100 units/mL penicillin, and 0.1 mg/mL streptomycin was added to the inserts, and each insert was placed in the lower chamber containing 10% bovine calf serum RPMI 1640. After 72-hour incubation in a cell culture incubator, invasiveness was evaluated by staining of cells that migrated through the extracellular matrix layer and clung to the polycarbonate membrane at the bottom of insert.

Statistical analysis. The χ² and ANOVA tests were used to determine the correlations between mGluR4 expression and each clinicopathologic parameter. Disease-free survival was estimated by the Kaplan-Meier method with the log-rank test. Multivariate analysis was done using the Cox proportional hazards model. P < 0.05 was considered statistically significant. All statistical analyses were done using SPSS software (SPSS, Chicago, IL).

Expression of metabotropic glutamate receptor 4 in normal and malignant tissues of various organs.Table 1 summarizes the immunohistochemical expression of mGluR4 in the tested normal and malignant tissues. Expression of mGluR4 was identified in the normal epithelia of the upper respiratory tract, gastrointestinal tract, breast, uterine cervix, urinary bladder, and skin, whereas expression was not detected in the thyroid, lung alveoli, liver, testis, or prostate. Normal colonic crypt epithelia showed cytoplasmic expression of mGluR4 in the supranuclear portion. The myenteric plexus also expressed mGluR4 (Fig. 1).

In the corresponding malignant tissues, mGluR4 seemed to be more specifically expressed than in normal tissues. mGluR4 expression was frequently identified in colorectal adenocarcinomas (68%), malignant melanomas of skin (63%), laryngeal squamous cell carcinomas (50%), and breast carcinomas (46%), whereas it was not detected in carcinomas of the salivary gland, esophageal squamous cell carcinomas, endometrial carcinomas, prostatic adenocarcinomas, or seminomas of testis.

Expression of metabotropic glutamate receptor 4 and its relation to clinicopathologic features in colorectal carcinoma. Among the 241 cases of colorectal adenocarcinoma, 122 (51%) cases showed altered expression of mGluR4. One hundred ten (46%) cases showed loss of mGluR4 expression, whereas 12 (5%) cases showed overexpression of mGluR4 with diffuse cytoplasmic staining in >50% of tumor cells (Fig. 1).

When mGluR4 expression was examined compared with clinicopathologic parameters, we observed that loss of mGluR4 expression was more frequently seen in moderately to poorly differentiated types than in well-differentiated types (24% versus 49%; P = 0.028). In addition, overexpression of mGluR4 was more frequent in cases with recurrence than in cases without recurrence (11% versus 3%; P = 0.034; Table 2), and overexpression of mGluR4 was identified in cases of T stage 3 or 4 only (P = 0.199). In contrast, we observed no significant association between mGluR4 expression and other clinicopathologic parameters, including tumor size, location, lymphatic or venous invasion, lymph node or distant metastasis, stage, microsatellite instability, or p53 expression (Table 2).

Analysis of disease-free survival was done in 218 patients who had undergone curative resection (R0 according to the International Union Against Cancer guideline). Patients with mGluR4 overexpression showed significantly poorer disease-free survival than those without mGluR4 overexpression (P = 0.0176; Fig. 2). A multivariate Cox proportional regression model significantly correlated stage (P = 0.001), overexpression of mGluR4 (P = 0.0166), and venous invasion (P = 0.0181) with poor disease-free survival, independent of tumor differentiation, location, depth of invasion (pT stage), lymph node or distant metastasis, lymphatic invasion, microsatellite instability status, or p53 overexpression (Table 3).

Effect of metabotropic glutamate receptor 4 agonist and antagonist on cell proliferation and invasion of human colorectal cancer cell lines. To clarify whether mGluR4 signaling can affect the proliferation and invasion of human colorectal cancer cell lines, mGluR4 agonist (L-AP 4) and antagonist (MAP 4) were used. Before this in vitro assessment, the expression of mGluR4 in various human colorectal cancer cell lines was evaluated using Western blot analysis to select the cell lines (Fig. 3A). Each human colorectal cancer cell line showed different intensity of the immunoreactive mGluR4 signals at 110 kDa. Among the cell lines tested, the strongest immunoreactive signal of mGluR4 was detected in SNU-407, whereas both SNU-61 and SNU-1033 showed moderate signal; the lowest signal was obtained from SNU-81 (Fig. 3A). These four cell lines, and mGluR4-negative human hepatoma cell line Huh-7 (12), were used to verify the effect of mGluR4 agonist and antagonist on cell proliferation. The proliferation of SNU-81 was increased by the addition of mGluR4 agonist, L-AP 4 (Fig. 3B). However, the other three colorectal cell lines and Huh-7 did not show any significant change in proliferation. When cell lines were treated with mGluR4 antagonist, MAP 4, proliferation was suppressed but this suppression was not observed in SNU-81 and Huh-7 (Fig. 3B). Two cell lines, SNU-81 and SNU-407, showing most different levels of immunoreactive mGluR4 signal in Western blot analysis, had been used in invasion assay. SNU-81, SNU-407, and Huh-7 cell lines did not show any invasive nature compared with HT1080, a positive invasion control. However, in the presence of L-AP 4, increased invasiveness was observed in both SNU-81 and SNU-407 but not in Huh-7 (Fig. 3C).

In addition to their expression in the central nervous system, glutamate receptors have been reported to be widely expressed in peripheral, neuronal, or nonneuronal tissues, such as bone (7), skin (8), and pancreas (9). Indeed, recent studies have revealed that peripheral glutamate receptors may be involved in a variety of physiologic functions (20–22). For example, glutamate receptors in the myenteric plexus of the guinea pig ileum were related to modulation of intestinal contractility (20, 21), and mGluRs in the rat heart were suggested to play an important role in cardiac function (22). Furthermore, glutamate-stimulated tumor growth and glutamate antagonist–induced suppression of cancer cell growth strongly suggest that glutamate signaling is likely to be important in nonneuronal cancer cells (6).

In the present study, we found that mGluR4 was widely distributed in a variety of normal tissues, including upper respiratory epithelium and colon crypt. However, in the corresponding malignant tissue, its expression was limited to specific types of organ or histologic types. Especially, mGluR4 expression was the most frequent in colon adenocarcinoma (Table 1). This finding suggests that mGluR4 may regulate the pathophysiology of normal and carcinoma cells originating from the colon epithelium. In the normal colonic epithelium, mGluR4 was expressed in the cytoplasm rather than the membrane, perhaps due to receptor internalization, which is a mechanism for desensitization (23–25).

The physiologic role of mGluR4 is important in both neuronal (26) and nonneuronal tissues (12). Decreased expression of the mGluR4 gene is associated with neuronal apoptosis and selective activation of mGluR4 protects against excitotoxic neuronal death (26). Consistent with these findings, our previous study showed that the mechanism underlying 5-fluorouracil resistance in human cancer cells may involve increased expression of mGluR4 (12). However, mGluR4 expression in pretherapy biopsy samples was not associated with resistance to preoperative chemoradiation therapy in rectal carcinoma (data not shown), and mGluR4 expression was not significantly different between groups with or without neoadjuvant therapy in this study (Table 2). In contrast, mGluR4 expression seemed to be significantly associated with differentiation (P = 0.028), recurrence (P = 0.034; Table 2), and disease-free survival (P = 0.0176; Fig. 2). These results are consistent with previously reported associations in oral squamous cell carcinoma between glutamate receptor expression and poor prognosis (27).

Our data indicate that mGluR4 expression is linked to prognosis of colorectal carcinoma, but it is not yet known how this molecule executes this physiologic role to affect recurrence and disease-free survival. Various regulation pathways are coupled to mGluR signaling, including the mitogen-activated protein kinase, phosphatidylinositol-3-kinase, phospholipase C, and inhibitory cyclic AMP pathways (28–30). All of these pathways have been associated with cell proliferation, differentiation, and/or antiapoptotic survival (31), suggesting that the regulation and final effect of mGluR4 signaling caused by the overexpression of mGluR4 may affect the biological behavior of colorectal carcinoma cells. Our in vitro experiments could not examine the mGluR4 signaling network in detail (Fig. 3), but our results strongly support previous reports describing the involvement of mGluR4 signaling in tumor growth and malignancy (6, 11). In the presence of the mGluR4 agonist, SNU-81 cells showed increased proliferation (Fig. 3B), whereas cell lines SNU-61, SNU-407, and SNU-1033, which had moderate to high mGluR4 immunoreactivities, were not affected by addition of the mGluR4 agonist. In contrast, the mGluR4 antagonist suppressed proliferation of cell lines SNU-61, SNU-407, and SNU-1033, whereas proliferation of SNU-81 was not significantly affected (presumably because the level of mGluR4 signal was already low; Fig. 3B). Taken together, these results suggest that mGluR4 signaling may be necessary but not sufficient for colorectal cancer cell proliferation. Alteration of cell proliferation and invasiveness of colorectal cancer cell lines by the treatment of the mGluR4 agonist and antagonist further supported the notion that mGluR4 signaling plays a role in tumor growth and progression and also provides a possible link between mGluR4 overexpression and the prognosis of colorectal carcinomas (Fig. 2; Table 3).

In conclusion, the role of mGluR4 in colon epithelium and its regulatory mechanisms still remain to be clarified. However, our results confirm the existence of mGluR4 signaling in the colon and indicate its significance as a possible poor prognostic factor in colorectal carcinoma.