A. The Initiation Sequelae of a Meal

In the mucosa of the upper digestive tract, mGlu receptors together with taste receptor type 1 (T1R) are believed to act as chemosensory receptors responsible for the umami basic taste (savoriness attributable to the carboxylate anion of MSG) thus contributing to the early detection of proteinaceous content in a meal (for reviews, see Shigemura et al., 2009; Yasumatsu et al., 2009). mGlu1 and mGlu4 receptor variants have been described in the rat taste buds by several techniques, including RT-PCR, RNase protection assays, and in situ hybridization (Chaudhari et al., 1996, 2000; San Gabriel et al., 2005). It is noteworthy that these truncated forms of mGlu receptors, also known as taste mGlu1 and taste mGlu4, lack a great portion of the glutamate-binding domain, (Chaudhari et al., 2009); consequently, the effective concentration of l-glutamate required to activate the taste mGlu receptor variants is higher than that needed for the full-length receptors (Chaudhari et al., 2000, 2009).

In support of a role for mGlu4 in taste, the group III mGlu receptor antagonist CPPG decreased the amount of aversion for MSG and other amino acids when MSG served as the conditioned stimulus in a taste aversion test (Eschle et al., 2009). However, when CPPG was mixed with these amino acids, the strength of the learned taste aversions and cross-generalization were either decreased or increased, indicating that the antagonist was not only able to reduce the intensity of the stimulus experience but also changed the qualities of the sensory experience (Eschle et al., 2009). These authors propose that multiple receptors are involved in amino acid taste and that taste mGlu4 receptors contribute to the taste of MSG and at least some other l-amino acids.

Other mGlu receptors have also been identified in taste sensing tissues, including mGlu2 and mGlu3 (Toyono et al., 2007). In addition, focal expression of mGlu4 was identified by immunohistochemistry in healthy human salivary glands (Chang et al., 2005), a finding that has not yet been supported by functional characterization.

B. Esophagus and Stomach

The lower esophageal sphincter plays a key role in preventing gastric reflux. The main cause of gastroesophageal reflux disease (GERD) is transient lower esophageal sphincter relaxation (TLESR) (Blackshaw, 2008). l-Glutamate is an important activator of vagal pathways mediating TLESR (Partosoedarso and Blackshaw, 2000); furthermore, retrograde tracing studies indicate that mGlu1–8 receptor proteins are expressed in these gastric vagal afferents (Page et al., 2005). Therefore, it was hypothesized that TLESR may be associated with impairment in mGlu receptor signaling. In a conscious ferret model of TLESR induced by a high gastric load, the prototypical mGlu5 receptor antagonist MPEP prevented sphincter relaxation dose-dependently, reaching a maximal inhibition of 71%. MPEP also reduced reflux episodes and increased basal lower esophageal sphincter pressure (Frisby et al., 2005). Similar results were obtained by administration of MPEP in a dog model of TLESR (Jensen et al., 2005). In the ferret, stronger inhibition of TLESR was achieved by a different mGlu5 receptor selective antagonist, 3-((2-methyl-1,3-thiazol-4-yl)ethynyl)pyridine; on the other hand, the group I receptor agonist DHPG tended to increase TLESR. The group II receptor agonist APDC had no effect, whereas the group III receptor agonist l-AP4 slightly reduced TLESR. It is noteworthy that the mGlu8 receptor agonist (S)-3,4-dicarboxyphenylglycine was able to inhibit TLESR by 54% (Frisby et al., 2005). The promising therapeutic potential of mGlu5 receptor antagonists for patients with gastroesophageal reflux disease has been further exploited in clinical trials: the mGlu5 negative allosteric modulator ADX10059 (Bolea et al., 2004) improved symptoms in patients with GERD who reported fewer and shorter reflux episodes than those treated with placebo (Keywood et al., 2009); however, ADX10059 administration was associated with central side effects such as dizziness, which was attributed to the compound's rapid absorption. Better tolerability was achieved by a modified release formulation (Zerbib et al., 2010).

mGlu1 receptor is expressed in rat esophagus mucosa, and mGlu4 is present in both mucosa and muscle layer (Akiba et al., 2009); likewise, normal human esophagus also presents positive immunoreactivity for mGlu4 (Chang et al., 2005). However, the potential effects of directly targeting these receptors in the esophagus mucosa or muscle are unknown to date. It is noteworthy that group I mGlu receptors seem to also have a role in gastroesophageal pain sensitivity, in that the mGlu5 receptor antagonist 3-((2-methyl-1,3-thiazol-4-yl)ethynyl)pyridine dose-dependently inhibited mechanosensitivity in vitro (Slattery et al., 2006). These data suggest an interesting additional potential for the use of mGlu5-directed compounds for treatment of patients with GERD, who may show increased esophageal perception as a result of chronic tissue damage.

As for the stomach, various degrees of mRNA expression have been described for mGlu1–8 in the different cell components of the rat stomach mucosa (Nakamura et al., 2010). In the case of mGlu receptor protein levels, intense mGlu2/3 staining was found in both parietal and endocrine cells, suggesting a role in the regulation of gastric acid and gastrin secretion (Gill and Pulido, 2001). mGlu1, on the other hand, is present in the apical membrane of chief cells, which suggests that it may be involved in the expression of pepsinogen in the stomach mucosa (San Gabriel et al., 2007). In general, mGlu receptor expression in the stomach mucosa is believed to contribute to sensing levels of luminal l-glutamate, although functional evidence supporting this hypothesis is not available to date.

C. Duodenal Protection

Akiba et al. (2009) studied the effect of l-glutamate in the defense mechanisms of the rat duodenal mucosa. Perfusion of the duodenum with glutamate increased intracellular pH and mucus gel thickness in vivo and decreased acid-induced epithelial cell injury in vivo (Akiba et al., 2009). This is indicative of a protective role for luminal l-glutamate in addition to the physiological role of early detection of a meal. They also showed that mGlu1 and mGlu4 were expressed in the mucosa of both gastric antrum and duodenum. It is noteworthy that the group III mGlu receptor agonist l-AP4 mimicked duodenum-protecting effects produced by l-glutamate; in contrast, the mGlu1/5 receptor agonist DHPG increased intracellular pH but had no effect on mucus gel thickness. Simultaneous administration of the group III mGlu receptor antagonist MAP4 with l-glutamate prevented glutamate-induced alkalinization and mucus secretion, whereas the mGlu1/5 receptor antagonist AIDA partially inhibited alkalinization but had no effect on mucus secretion induced by l-glutamate (Akiba et al., 2009). This recent evidence suggests that patients suffering from duodenal ulcers may benefit from treatments directed against mGlu4 or mGlu1/5 signaling.

D. Regulation of Intestinal Fluid Secretion

Immunohistochemistry studies have demonstrated the presence of mGlu2/3 in submucosal neurons of rat jejunum and ileum (Larzabal et al., 1999) and of mGlu1α and mGlu5 in submucosal neurons of the guinea pig ileum. These neurons have a secretomotor function, which means they regulate the process of absorption and secretion of fluid from the faeces and toward the lumen, respectively (Hu et al., 1999; Liu and Kirchgessner, 2000). The group I mGlu receptor agonist DHPG can induce mGlu5 internalization in isolated enteric neurons, an effect mimicked in ex vivo guinea pig ileum preparations by stimulating villi movement with gaseous bubbling, thus indicating that mechanical mucosal stimulation may be an important mechanism for activation of enteric mGlu5 receptors in vivo (Liu and Kirchgessner, 2000). In submucosal neurons, glutamate and group I mGlu receptor agonists evoke slow depolarizations, an effect that is sensitive to the group I mGlu receptor antagonist (S)-4-carboxyphenylglycine (Hu et al., 1999; Ren et al., 2000). On the other hand, slow excitatory postsynaptic potentials (EPSPs) were not affected by the antagonist (Ren et al., 2000). l-Glutamate acting via group I mGlu receptors has been shown to suppress slow EPSPs and potentiate slow inhibitory postsynaptic potentials in submucosal neurons (Ren et al., 1999). However, conflicting evidence also shows that the mGlu5 receptor antagonist MPEP, as well as (S)-4-carboxyphenylglycine, is able to depress slow EPSPs in submucosal neurons (Liu and Kirchgessner, 2000). It is noteworthy that another piece of work shows that expression of functional mGlu5 receptors in ileum and colon of guinea pig, rat, and mouse is restricted to enteric glia (Nasser et al., 2007). Glial c-fos immunoreactivity and phosphorylated ERK1/2 levels are increased upon mGlu5 activation in guinea pigs, and chemically induced colitis was associated with redistribution of the receptor. On the other hand, a mouse model of colitis (the IL-10 gene-deficient mouse, characterized by spontaneous chronic intestinal inflammation) shows significantly lower levels of glial mGlu5 receptor expression in colon myenteric plexus (Nasser et al., 2007).

The luminal environment in the large intestine is markedly different to that of the small intestine, and one characteristic of colonic epithelium is that there is little or no transfer of amino acids from the colonic lumen to portal blood except for a short period after birth. As such, amino acids including l-glutamate must be taken into colonocytes from arterial blood (Blachier et al., 2009). However, mGlu receptors can be found in colon mucosa, suggesting that l-glutamate plays a role in colon mucosal physiology. mGlu4 receptor has been detected by immunohistochemistry in normal human colon epithelium (Chang et al., 2005). Recently, we have shown that the mouse colon mucosa also expresses both mRNA and protein for mGlu7 (Julio-Pieper et al., 2010). Moreover, we demonstrate that the selective mGlu7 receptor agonist N,N′-bis(diphenylmethyl)-1,2-ethanediamine (AMN082) induces an increase in fecal water content in a stress-induced defecation paradigm. Furthermore, the mGlu7 receptor agonist can act directly on the colon, as shown by ex vivo experiments in which pretreatment of colon tissue with AMN082 induces a higher secretory response to the muscarinic receptor agonist bethanechol (Julio-Pieper et al., 2010). This effect is dependent on the presence of submucosal neurons, because when tissues were treated with a nerve toxin, the secretory response was markedly blunted (Julio-Pieper et al., 2010). Together, these data indicate that glutamatergic activation of colonic mGlu7 receptor could be a component in the pathophysiology of secretory disorders such as diarrhea and a significant factor underlying stress-induced diarrhea. Thus it may play a role in functional intestinal disorders such as irritable bowel syndrome (IBS) that are comorbid with chronic stress states (Quigley, 2006; Clarke et al., 2009; Gros et al., 2009). Stressful life events and experimental stress also induce exacerbation of symptoms and visceral hypersensitivity in patients with functional GI disorders (Whitehead, 1992; Posserud et al., 2004). The foregoing evidence supports the idea of a complex interaction between the GI tract and the brain, namely the brain-gut axis. Thus mGlu7 may not only regulate the central components of physiological stress (Cryan et al., 2003; Mitsukawa et al., 2006) but may also have a role to play peripherally in the regulation of fluid and electrolyte transport, which is significantly disrupted in diseases such as IBS, by enhancing colonic secretory activity.

E. Intestinal Motility

Myenteric neurons in the enteric nervous system are a major site of expression of mGlu receptors in the gut. mGlu2/3 has been found by immunohistochemistry in myenteric neurons of rat jejunum and ileum (Larzabal et al., 1999). In addition, mouse colon homogenates containing the muscle layers together with the enteric nerves contain mGlu7 protein, although the administration of an mGlu7 receptor agonist did not modify fecal pellet output upon stress (Julio-Pieper et al., 2010). RT-PCR showed weak expression of mGlu4 and -6, and a strong expression of mGlu7 and -8 in longitudinal muscle with adherent myenteric plexus obtained from rat duodenum and ileum. The presence of mGlu8 receptor in the enteric nervous system of human, rat, and guinea pig has also been shown by immunohistochemistry (Tong and Kirchgessner, 2003). An accelerating effect on guinea pig colon motility and longitudinal muscle contractions was observed after the application of mGlu8 receptor agonists on the isolated tissue. On the other hand, mGlu8 receptor antagonists used alone were able to decrease motility (Tong and Kirchgessner, 2003), which indicates a probable tonic effect of mGlu8 on colonic contractions.

F. Microbiota Composition

The composition of intestinal flora is considered an important factor for both the maintenance of good health and development of disease states such as inflammatory bowel disease and IBS (Clarke et al., 2009). In the specific case of IBS, patients report recurrent abdominal pain in combination with either diarrhea or constipation, which may arise due to an impairment on intestinal motility (Clarke et al., 2009). The amount of various fecal bacterial strains is different in patients with IBS compared with healthy control subjects (Kassinen et al., 2007; Kerckhoffs et al., 2009); in addition, the composition of their intestinal microflora seems to change more over time (Mättö et al., 2005). Moreover, alterations in the fecal microbiota have also been found in an early life-stress model of IBS (O'Mahony et al., 2009). Gut bacteria contain a relatively high internal pool of glutamate that responds by increasing as the external environment becomes hyperosmotic (Yan, 2007). It is plausible that l-glutamate of bacterial origin has the potential to activate mGlu receptors in the GI tract. Thus, further studies must investigate whether the ability of different bacteria to accumulate/release glutamate plays a role in normal gut function and in GI disorders such as IBS and inflammatory bowel disease. Furthermore, the existence of multiple nodes of communication between the brain and gut is supported by increasing evidence from germ-free animals, probiotic and antibiotic administration, and microflora transplantation studies, which suggest that the enteric microbiota can directly influence brain-gut axis function (Rhee et al., 2009). This results in perturbations of the stress response (Forsythe et al.,; Verdu, 2009), immune system (Sudo et al., 2002), endocrine system (Turnbaugh and Gordon, 2009), and pain processing (Rousseaux et al., 2007; O'Mahony et al., 2010) and in behavioral changes (Forsythe et al., 2010). Understanding whether l-glutamate from microbiota and thus mGlu receptors play a role in brain-gut axis communication is worthy of further investigation.

A. Homeostasis of Glucose

The islet of Langerhans, the pancreatic endocrine unit responsible for homeostasis of blood glucose, possesses a glutamatergic system similar to the one found in the CNS (Moriyama and Hayashi, 2003). Under low-glucose conditions, l-glutamate is cosecreted with glucagon from α cells, where it is believed to act in an autocrine fashion (Uehara et al., 2004). Ionotropic receptors are expressed in both α and β pancreatic cells, and their stimulation seems to modulate glucagon/insulin secretion (Uehara et al., 2004). The expression of mRNAs for all eight subtypes of mGlu receptors has been studied by RT-PCR in rat and human pancreatic islets and in rat pancreatic cells. In the study performed by Brice et al. (2002), mRNAs for mGlu receptors 3, 5, and 8 were detected by PCR in all the samples; mGlu2 was found only in β cells, whereas mGlu4 was identified in rat islets. Group I and II mGlu receptor agonists increased the release of insulin in the presence of glucose, whereas a group III mGlu receptor agonist had the opposite effect (Brice et al., 2002).

A further exploration of the role of group I mGlu receptors in pancreatic function showed that endogenous activation of mGlu5 is required for an optimal insulin response to glucose both in clonal β cells and in mice. Confocal analysis showed that mGlu5 was localized mainly at the surface of β cells under basal conditions. Agonist exposure induced a rapid and transient internalization of surface receptors. mGlu5 was also found in purified insulin-containing granules (Storto et al., 2006). In this study, β cells did not respond to mGlu5 receptor agonists that act extracellularly, but a cell-permeant analog of l-glutamate did induce an increase in intracellular Ca²⁺ and insulin secretion. The mGlu5 antagonist MPEP decreased both effects, whereas an antagonist for mGlu1 was inactive. mGlu5 knockout mice had a blunted insulin response after a glucose pulse, an effect that was replicated by the administration of MPEP to wild-type mice. Mice injected with MPEP or lacking the mGlu5 receptor also presented a defective glucagon response after an insulin challenge. This indicates that insulin secretion is under the control of mGlu5 both in clonal β cells and in vivo (Storto et al., 2006).

Another study reports evidence for functional mGlu4, but not other mGlu receptors, in the rat pancreatic islets. The mGlu4 receptor was colocalized with glucagon and pancreatic-polypeptide producing cells, and stimulation of the receptor inhibited secretion of glucagon in a cascade involving inhibition of cAMP production (Uehara et al., 2004). Together with the data provided by Storto et al. (2006), this suggests new potential pharmacological targets for the treatment of endocrine disorders related to glucose metabolism (Tables 4 and 5). In addition, it adds a new range of possible side effects to the emerging mGlu receptor-based therapies for neurological and mood disorders (Lavreysen and Dautzenberg, 2008).

B. Steroidogenic and Reproductive Tissues

The hypothalamus-pituitary-adrenal axis is the most important regulator of the stress response and l-glutamate plays a key role in its activation (Zelena et al., 2005). As for the adrenal gland, no evidence has yet shown a direct action of l-glutamate on glucocorticoid secretion at the adrenal cortex level. However, l-glutamate stimulates the secretion of catecholamines both in the dog adrenal gland and in chromaffin cells isolated from bovine adrenal medulla (Nishikawa et al., 1982; González et al., 1998). These effects are not caused solely by iGlu receptors; they can be mimicked by the group I mGlu agonists ACPD and DHPG (González et al., 1998; Arce et al., 2004). Arce et al. (2004) also report that most of the bovine chromaffin cells isolated from adrenal medulla express group I mGlu receptors, which are at least of two subtypes: mGlu1 and mGlu5a, as demonstrated by immunocytochemistry. mGlu5 receptor mRNA, as well as mGlu7 receptor mRNA, is also present in the mouse and rat adrenal gland (Scaccianoce et al., 2003). A later study also demonstrated the presence of mGlu2/3 and mGlu4a receptors in large-sized type I ganglion neurons and in the small-sized type II adrenal ganglion neurons by immunohistochemical analysis of the rat adrenal medulla (Sarría et al., 2006). However, the functional relevance of this expression has yet to be determined.

In the testis, the presence of glutamate transporters and glutamate decarboxylase enzymes supports the idea that non-neuronal components in this tissue possess their own glutamatergic system (Tillakaratne et al., 1992; Hayashi et al., 2003). Gill et al. (2000) and Gill and Pulido (2001) described an intense and specific signal for mGlu2/3 in the perinuclear cytoplasm of interstitial cells and myoid (contractile) cells in the testis of adult Sprague-Dawley rats. Immunoreactivity with anti-mGlu2/3 is also present in the head of mature spermatids and spermatozoa (Gill et al., 2000; Gill and Pulido, 2001). Contrary to this finding, Storto et al. (2001) reported the presence of mGlu1, -4, and -5, but not mGlu2 or -3 receptor mRNA in the testis of the same strain of rat, a result that was confirmed by Western blot analysis. The mGlu1/5 receptor agonist ACPD induced inositol phospholipid hydrolysis in rat testes preparations, indicating that these receptors are functional (Storto et al., 2001). In testis of the Wistar strain of rat, on the other hand, all mGlu receptors except mGlu7 have been detected by PCR, including an additional splice variant for mGlu2 that was not present in whole rat brain (Takarada et al., 2004). In this study, [³H]glutamate accumulation by testicular fractions was not inhibited by the group I, II, or III mGlu receptor agonists DHPG, DCG-IV, or l-AP4, respectively (Takarada et al., 2004).

Analysis of human testis showed that immunoreactivity for mGlu1a receptor (a splice variant of mGlu1) was restricted to intertubular spaces, whereas mGlu5 receptor was abundantly expressed inside the seminiferous tubuli and in the mid-piece and tail of mature spermatozoa. However, neither the group I agonist quisqualate nor the mGlu5 receptor antagonist MPEP induced changes in human sperm motility (Storto et al., 2001).

Ovarian production of glutathione depends on l-glutamate (see Fig. 2); this process is modulated by gonadotropins and is critical for limiting the damage induced by ROS under intense cell proliferation (Tsai-Turton and Luderer, 2005). However, glutamate signaling through mGlu receptors remains highly unexplored; studies in the ovary are only descriptive. In the macaque ovary, a strong immunolabeling for mGlu2/3 receptor is present in the oocyte, the theca, and in granulosa cells. No staining was detected in the atretic follicles, corpora albicans, or the stroma (Gill et al., 2008). Likewise, in the rat ovary, the oocyte showed intense staining for mGlu2/3, whereas the corpus luteum was moderately immunoreactive, and the rest of ovarian structures had a very faint signal (Gill and Pulido, 2001). Furthermore, in the rat uterus, anti-mGlu2/3 antibodies show preferential binding to the most superficial layer of the stratified squamous epithelium of the exocervix (Gill and Pulido, 2001). In humans, the presence of estrogens during the first half of the menstrual cycle is associated with predominance in superficial cells in the cervical epithelium. Conversely, with the influence of progesterone after ovulation, this cell type becomes less predominant (Noyes et al., 1975). It is noteworthy mGlu2/3 expression is predominant in proliferating ovarian and uterine structures, which indicates that its production may be cyclically regulated and is also consistent with a differential need for glutamate/glutathione depending on the proliferative status of the tissue. Finally, mGlu4 receptor has also been found in the human cervix and is weakly expressed in the endometrial glands (Chang et al., 2005), whereas the ovary proved to be negative for mGlu4 (Chang et al., 2005).

In summary, most of studies involving mGlu receptors in steroidogenic and reproductive tissues involve only detailed descriptions of their anatomical localization, with little functional data available. The participation of mGlu receptors in events such as steroid hormone production, germinal cell development, and cyclic cell turnover is an exciting field that remains to be fully exploited.

A. Immune Cell Maturation and Activation

Bone marrow-derived T-cell precursors undergo differentiation in the thymus in a process that involves regulation of the expression of various membrane proteins. Group I and II mGlu receptors are expressed in whole mouse thymus, isolated thymocytes, and a thymic stromal cell line, as has been demonstrated by both RT-PCR and Western blot analysis (Storto et al., 2000a). Flow cytometry has also shown that mGlu5 receptor expression is induced along thymocyte maturation, whereas mGlu1a had the opposite pattern of expression. The level of mGlu2/3 expression remained unaltered through the different stages of thymocyte maturation (Storto et al., 2000a). A different study showed a positive immunostaining and Western blot of mGlu receptors 2/3, 4, and 5 in rat thymic cells (Rezzani et al., 2003). Moreover, expression of mGlu 2/3, 4, and 5 receptors was abundant in dendritic cells and lymphocytes of the thymic medulla but was weak in lymphocytes of the cortex. After 2 days of treatment with the immunosuppressant agent cyclosporine, a rapid inhibition on the expression of mGlu 2/3, 4, and 5 was induced in the rat thymus, reaching undetectable levels after 3 weeks of treatment (Rezzani et al., 2003).

mGlu receptor expression is not exclusive to young immune cells, in that mature lymphocytes are activated by selective mGlu receptor ligands. When rat peripheral lymphocytes are exposed to the group III mGlu receptor agonist l-AP4, they respond by producing ROS. Moreover, activation of lymphocytes with both an ionotropic receptor agonist [N-methyl-d-aspartic acid (NMDA)] and l-AP4 caused a synergistic increase in ROS levels and induced necrotic cell death without activating the proapoptotic caspase-3 pathway that was observed in the presence of NMDA alone (Boldyrev et al., 2004). The majority of circulating lymphocytes express the cell surface protein CD3; treatment with anti-CD3 antibodies can induce apoptosis, a response known as activation-induced cell death. It is noteworthy that l-glutamate has an inhibitory effect on apoptosis induced by anti-CD3 treatment, which is believed to be a consequence of decreased expression of Fas Ligand, an important component of immune cell programmed death mechanisms (Chiocchetti et al., 2006). The mGlu receptor agonists ACPD, quisqualate, DHPG, and (R,S)-2-chloro-5-hydroxyphenylglycine can mimic this prosurvival effect. The effects of quisqualate were inhibited by the antagonists MPEP, AIDA, and LY367385 (Chiocchetti et al., 2006). These data indicate that both mGlu1 and mGlu5 receptor signaling may be important for the pro-survival effects of glutamate on peripheral lymphocytes.

Whereas resting peripheral T cells are unable to cross the BBB under normal circumstances, access can be gained subsequent to cellular damage in the CNS (such as after an ischemic attack), where they assist neurons against l-glutamate neurotoxicity by inducing changes in microglial phenotype (Pacheco et al., 2007). Differential expression of glutamate receptors upon T-cell activation could be relevant to this process: Pacheco and colleagues (2007) demonstrated that the mGlu5 receptor is present constitutively in human peripheral blood lymphocytes, whereas mGlu1 is expressed only upon activation via the T-cell receptor-CD3 complex (Pacheco et al., 2004). In that specific study, activation of mGlu5 did not trigger the phospholipase C signaling pathway but instead activated adenylate cyclase, thus being presumably mediated by the long splice-variant of mGlu5 termed mGlu5a. On the other hand, only mGlu1 receptor was linked to ERK1/2 activation. Consistent with their differential expression in resting and activated lymphocytes and different signaling pathways, mGlu5 mediates the inhibition of cell proliferation evoked by glutamate, which is reverted by the activation of inducible mGlu1 (Pacheco et al., 2004). In addition, the mGlu5 receptor antagonist MPEP induced a significant increase in IL-6 secretion, whereas the specific mGlu1 receptor antagonist 7-(hydroxyimino)cyclopropa[b]chromen-1α-carboxylate ethyl ester impaired IL-2, IL-6, IL-10, tumor necrosis factor-α, and interferon-γ production (Pacheco et al., 2006). In addition, dendritic cells are able to secrete glutamate when interacting with T cells, a process that seems to be essential for lymphocyte function, because the absence of glutamate led to impaired Th1 (IL-2 and interferon-γ) and proinflammatory (IL-6 and tumor necrosis factor-α) cytokine production. These changes were not associated with a decrease in T-cell proliferation (Pacheco et al., 2006).

B. Autoimmunity

mGlu receptor expression in peripheral tissues seems to be relevant for the development of autoimmune-related disorders. The production of glutamate receptor autoantibodies is involved in the pathophysiology of some paraneoplastic encephalopathies; two patients with Hodgkin's disease represent an interesting case. Although the cancer was in remission, they developed severe cerebellar ataxia and produced mGlu1 autoantibodies that had the ability to block l-glutamate-induced inositol phosphate formation and to induce ataxia when administered intrathecally to mice (Pleasure, 2008). This may be considered circumstantial evidence for the role of mGlu receptors in the initiation or maintenance of central autoimmune disorders; however, it could be relevant in the context of peripheral immune dysregulation. In the case of CNS-specific antigens, these are unlikely to be exposed to the peripheral immune system unless a cerebrovascular infarction or a severe inflammatory condition occurs. mGlu receptors, on the other hand, are widely expressed in the periphery, so one can speculate that they can induce the generation of autoantibodies upon immune dysregulation, resulting in deleterious effects not only in the CNS but also in the periphery.

mGlu4-deficient mice are highly susceptible to experimental autoimmune encephalomyelitis (EAE), a model of multiple sclerosis (Fallarino et al., 2010). Conversely, increased resistance to EAE is conferred by treatment of wild-type mice with PHCCC, which can act as an mGlu4-positive allosteric modulator but also as a group I mGlu receptor antagonist (Fallarino et al., 2010). This is in agreement with a previous report showing that long-term l-AP4 treatment increases the recovery rate from EAE in Lewis rats (Besong et al., 2002).

IL-17-producing T helper (Th-17) cells are considered mediators of autoimmunity in multiple sclerosis and EAE, and accumulation of Th-17 cells in the CNS as well as in the periphery is associated with the development of demyelinating plaques (Segal, 2010). Fallarino et al. (2010) show evidence indicating that the absence of mGlu4 in dendritic cells is key to induce a differentiation of T helper cells toward the Th-17 phenotype. They also suggest that activation of mGlu4 receptor (as a result of elevated levels of glutamate produced at the neuroinflammatory status) might exert a protective effect by preventing the unbalance in T helper cells. Such mechanism presents a clear therapeutic potential in this and maybe other demyelinating diseases.

In other autoimmune disorders, such as rheumatoid arthritis, synovial cells (the modified fibroblasts residing between the cartilages of a joint) are surrounded by a fluid rich in inflammation mediators (Flood et al., 2007). Synovial fluid from patients with rheumatoid arthritis and from animal models of arthritis also contain increased levels of l-glutamate, in correlation with elevated concentrations of cytokines (Flood et al., 2007). Only mGlu8, but not mGlu receptors 1 to 7, is expressed in rat cultured synovial fibroblasts. mGlu8 is also present in synovial tissues, with no change in expression in a collagen-induced model of rheumatoid arthritis (Hinoi et al., 2005). In contrast, another study shows that the mGlu4 receptor is expressed in healthy rat patella and menisci and in normal human synovial cells but not in synoviocytes obtained from patients with rheumatoid arthritis or in knee cartilage from patients with osteoarthritis (Flood et al., 2007). However, this potential protective role for the mGlu4 receptor warrants further investigation.

A. Endothelial Oxidative Damage

Oxidative status of endothelial cells is strongly dependent on glutathione availability (Kevil et al., 2004), and the first and rate-limiting step in glutathione biosynthesis requires l-cysteine, glycine, and l-glutamate, as shown in Fig. 2 (Martin and Teismann, 2009). However, increasing evidence suggests that l-glutamate may not serve only as a metabolic substrate upon oxidative injury: activation of group I mGlu receptors with the agonist DHPG prevents nitric oxide-induced toxicity, DNA degradation, and the progression of membrane phosphatidylserine exposure in endothelial cells from rat brain (Lin and Maiese, 2001), suggesting that glutamatergic signaling trough mGlu1 or mGlu5 receptors may be protective during episodes of acute oxidative damage.

B. Endothelial Cell Barrier

The vascular endothelial cell monolayer regulates transport of macromolecules and fluid between blood and the interstitium, thus acting as a dynamic and semiselective barrier (Wu, 2005). Some organs, such as the CNS structures, the retina, and testes, are further protected from external stimuli by additional selective cellular barriers such as the BBB and blood-retinal barrier (Hosoya and Tachikawa, 2009; Banks, 2010). A variety of stimuli such as physical or inflammatory injury or bioactive compounds can alter the endothelial barrier and increase vessel permeability, thereby compromising organ function (Mehta and Malik, 2006).

l-Glutamate has been shown to have a role in the maintenance of endothelial barrier as inhibition of glutamate transport from the aqueous humor causes an increase in paracellular permeability in the intraocular endothelium/epithelium (Langford et al., 1997). Cell-free supernatants from stimulated polymorphonuclear leukocytes (PMN) significantly altered endothelial permeability, suggesting the presence of soluble PMN-derived mediators. Mass spectrometry analysis of PMN-derived supernatants indicated that glutamate is one of the components regulating human endothelial barrier function (Collard et al., 2002). It is noteworthy that microvascular endothelial cells from human brain (BMVEC) express mGlu receptors 1, 4, and 5 (Collard et al., 2002). Furthermore, group I (DHPG) or III (l-AP4) mGlu receptor agonists significantly increased the paracellular permeability of BMVEC cultured as a monolayer. Similar treatment of human BMVEC with the group I (PHCCC) or III (CPPG) mGlu receptor antagonists attenuated glutamate-mediated increases in paracellular permeability (Collard et al., 2002). In a mouse model of hypoxia, pretreatment with PHCCC or CPPG significantly decreased the permeability of the blood-brain barrier (Collard et al., 2002). Conversely, group I and III mGlu receptor agonists induced an increased permeability both in the hypoxic and normoxic mice (Collard et al., 2002).

Endothelial cells from other sources also express mGlu receptors: microvascular endothelial cells from human skin are positive for mGlu1, mGlu4, and mGlu5 (Collard et al., 2002). mGlu5 immunoreactive material has been found in the lining of blood vessels both in rat and macaque hearts (Gill et al., 1999; Mueller et al., 2003). An mGlu2/3 signal is present in the endothelial lining of the rat respiratory tissues (Gill et al., 2000). Whether these receptors have a role in modulating endothelial oxidative damage, endothelial barrier integrity, or others is yet to be investigated.

A. Retina

All mGlu receptors, with the exception of mGlu3, have been identified in the retina. Immunohistochemistry and in situ hybridization studies have shown differential expression of mGlu receptors in the different cell layers of retina (Connaughton, 2005). Jensen (2006) used extracellular microelectrodes to study the effects of activation of group II mGlu receptors on the responses of rabbit retinal ganglion cells to a moving light stimulus with responses in these cells shown to be substantially reduced by the group II receptor agonist DCG-IV; this effect was reversed by the group II receptor antagonist ethylglutamic acid. In addition, responses that were direction-selective were prolonged by the use of an acetylcholinesterase inhibitor; this was reversed by DCG-IV, suggesting that postsynaptic group II mGlu receptors have the potential to influence retinal ganglion cells by inhibiting light-evoked acetylcholine release (Jensen, 2006).

Defects in the gene encoding the mGlu6 receptor can lead to congenital stationary night blindness (CSNB), characterized by myopia and impairment of night vision. It is noteworthy that Zeitz et al. (2007) demonstrated that CSNB-associated mutations in three different domains of mGlu6 receptor interfere with proper protein trafficking to the cell surface (Zeitz et al., 2007). Further to this, Beqollari et al. (2009) investigated how the signaling cascade was affected in a cellular model of CSNB carrying an induced mutation of the mGlu6 receptor (E775K). E775K is incapable of activating G_o and primarily couples to G_i proteins, suggesting that the mGlu6 receptor primarily functions through G_o proteins in retinal ON bipolar cells. An inability of the receptor to activate G_o would result in loss of function and lead to CSNB (Beqollari et al., 2009).

B. Inner Ear

Group I mGlu receptors contribute to the neurotransmission between inner hair cells and afferent neurons in the mammalian cochlea. The group I mGlu receptor agonist DHPG and the ionotropic glutamate receptor agonists α-amino-3-hydroxyl-5-methyl-4-isoxazole-propionate and NMDA produced an increase in afferent firing, activation that was reversibly blocked by the group I mGlu receptor antagonist AIDA in a dose-dependent manner (Kleinlogel et al., 1999). Cochlear afferent neurons are sensitive to injury by ischemia and noise exposure; dopamine released from cochlear efferent fibers is believed to exert a protective function by reducing glutamate excitotoxicity on afferent neurons. The group II mGlu receptor agonist APDC dose-dependently increased the release of dopamine in a guinea pig cochlea preparation, whereas group I and III mGlu receptor agonists and antagonists were ineffective. In addition, the GABA_A receptor antagonist bicuculline reversed the effect of APDC, suggesting that the mGlu-mediated increase on dopamine release was due to a disinhibitory mechanism involving local GABAergic fibers (Doleviczényi et al., 2005).

Histochemical studies in human and mouse show that mGlu7 receptor is expressed in hair cells and in spiral ganglion cells of the inner ear (Friedman et al., 2009). Age-related hearing impairment, also known as presbycusis, is the most prevalent sensory impairment in the elderly. Common alleles of GRM7, the gene encoding for mGlu7 receptor, contribute to an individual's risk of developing age-related hearing impairment. Because presbycusis is associated to the loss of sensory cells, polymorphisms in the mGlu7 gene could possibly be part of a mechanism involving altered susceptibility to glutamate excitotoxicity (Friedman et al., 2009).

C. Pain Sensation

mGlu receptors are involved in various stages of central pain processing (Karim et al., 2001; Varney and Gereau, 2002; Goudet et al., 2009) and interact with TRPV1 receptors, which are involved in the modulation of pain sensation (Hu et al., 2002; Yang and Gereau, 2002). It is noteworthy that mGlu5 receptors expressed on the peripheral terminals of sensory neurons are also involved in nociceptive processes and contribute to the hyperalgesia associated with inflammation (for review, see Neugebauer, 2001). When inflammation was induced in the rat hind paw by administration of Freund's complete adjuvant, hyperalgesia was significantly reduced by intraplantar microinjection, but not by intracerebroventricular or intrathecal microinjection, of the selective mGlu5 receptor antagonist MPEP. Immunohistochemistry confirmed the presence of mGlu5 in peripheral nerve fibers in the naive rat hind paw skin (Walker et al., 2001).

A relevant aspect when dealing with inflammatory pain is treating the inflammation itself. Intraplantar injection of bee venom induces spontaneous nociception, heat, and mechanical hyperalgesia together with an inflammatory response. Local pretreatment with the group I mGlu receptor antagonist AIDA or with the group II and III receptor agonists APCD and l-AP4, respectively, inhibited bee-venom induced spontaneous nociception. Only APCD was able to reduce mechanical hyperalgesia. Group II and III mGlu receptor antagonists proved to reverse the effects induced by the respective agonists, and drug treatments contralateral to the bee venom injection had no effect, suggesting the participation of local mGlu receptors in the induction/reversal of venom-induced spontaneous nociception. It is noteworthy that neither AIDA, APCD, or l-AP4 had an effect on inflammatory swelling (Chen et al., 2010).

Jang et al. (2004) studied the contribution of peripheral mGlu receptors to the development of hyperalgesia in neuropathic pain. They used a rat model involving spinal nerve lesion (SNL) preceded by dorsal rhizotomy, which is performed to avoid the potential central effects induced by SNL. By analyzing the paw withdrawal threshold to increasing mechanical stimuli they show that blockade of peripheral group I mGlu receptors with the antagonist d,l-2-amino-3-phosphonopropionic acid prevents the induction of neuropathic pain behavior. When d,l-2-amino-3-phosphonopropionic acid is administered locally to the hind paw before SNL, it delays the onset of SNL-induced mechanical hyperalgesia. Conversely, intraplantar injection of the group II mGlu receptor agonist APDC also delays the onset of mechanical hyperalgesia, suggesting that activation of peripheral group II mGlu receptors prevents the induction of neuropathic pain behavior (Jang et al., 2004).

A. Embryonic Development

The involvement of glutamate in early CNS development has been extensively investigated. Double-knockout mice lacking the glutamate transporters GLAST and GLT1 show multiple brain defects and perinatal mortality (Matsugami et al., 2006). In addition, mGlu3 receptor is known to induce differentiation of neural stem cells (Ciceroni et al., 2010). Less is known about glutamatergic signaling in immature peripheral tissues; activation of mGlu receptors has been addressed using mainly in vitro or ex vivo models of embryonic development.

Embryonic stem (ES) cells can be grown in vitro as undifferentiated, self-renewing cells in the presence of leukemia inhibitory factor, a member of the interleukin-6 family of cytokines (Cappuccio et al., 2005; Spinsanti et al., 2006). Cultured mouse ES cells maintained under this pluripotent condition express mGlu5 receptors, whereas the mRNA of all other mGlu receptor subtypes was not detectable by RT-PCR. The presence of mGlu5 was confirmed by Western blot and flow cytometry (Cappuccio et al., 2005). An increase on intracellular Ca²⁺ was produced when ES cells were treated with quisqualate in the absence of extracellular l-glutamate, which indicates that mGlu5 was activated by endogenous glutamate. Blockade of mGlu5 receptors with MPEP or antisense-induced knockdown of mGlu5 decreased the expression of transcription factors such as c-Myc, important to maintain ES cell-undifferentiated state. Likewise, exposure of ES cells to MPEP reduced alkaline phosphatase activity, another marker of undifferentiated state in ES cells (Cappuccio et al., 2005). Endogenous activation of mGlu5 sustains the increase in c-Myc induced by leukemia inhibitory factor in embryonic stem cells by inhibiting glycogen synthase kinase-3b and phosphatidylinositol 3-kinase. These two effects result from the activation of protein kinase C (Spinsanti et al., 2006).

On the other hand, when cultured ES cells form aggregates or embryoid bodies (EBs), which differentiate in a way that resembles embryogenesis, they start losing mGlu5 and begin to express mGlu4 receptor. This switch could be critical for further development, because l-AP4 induces an increase on mRNA for brachyury and H19 (a mesoderm-specific transcription factor and a noncoding RNA produced by endoderm, respectively), and a decrease on the expression of the mRNA for fibroblast-growth factor-5, a more primordial ectoderm marker. The mGlu4 receptor antagonist (R,S)-α-methylserine-O-phosphate prevented these effects. In addition, when EBs were incubated in a medium that induces an enrichment of the culture on neural precursors, l-AP4 increased the expression of the neural markers nestin and Dlx-2, showing that mGlu4 affects cell differentiation in a context-dependent manner (Cappuccio et al., 2006).

During embryonic development of the spine and long bones, mesenchymal stem cells differentiate by forming a cartilaginous structure that then undergoes ossification to finally induce bone formation (Karsenty, 2003). l-Glutamate seems to be a key component of chondrocyte developmental maturation, because ex vivo data indicate that l-glutamate suppresses chondral mineralization and induces apoptosis in cultured embryonic mouse metatarsals (Wang et al., 2006). In the same model, mineralization is also greatly inhibited by the group III mGlu receptor agonist l-AP4. This effect was independent of apoptosis and also sensitive to CPPG. The metatarsal total length remained unchanged. Similar to l-AP4, a group II mGlu receptor agonist (DCG-IV) was more effective in inhibiting the mineralization than a group I mGlu receptor agonist (DHPG), whereas none of the iGlu receptor agonists (α-amino-3-hydroxyl-5-methyl-4-isoxazole-propionate, kainate, or NMDA) drastically affected the mineralization (Wang et al., 2005).

B. Tumoral Growth

Plasma levels of l-glutamate are generally elevated in patients with carcinoma (Dröge et al., 1987) and seem to correlate with an impairment in immune function. Furthermore, tumor cells show decreased motility and invasiveness upon administration of ionotropic glutamate receptor antagonists (Rzeski et al., 2001). mGlu receptors are also expressed in several cell lines derived from human tumors, including neuroblastoma, rhabdomyosarcoma/medulloblastoma, thyroid carcinoma, lung carcinoma, astrocytoma, multiple myeloma, glioma, colon adenocarcinoma, T-cell leukemia, and breast carcinoma (for review, see Stepulak et al., 2009). Massive mGlu receptor expression is sometimes seen in transformed tissues; this is the case for mGlu3 mRNA, which is increased by 5-fold in aldosterone-producing adenomas compared with normal human adrenal glands (Ye et al., 2007). The implications of such overexpression remain unknown; however, increasing data suggest that mGlu receptors could be novel targets for the treatment of especially aggressive or chemotherapy-resistant tumors (see Tables 3, 4, and 5 for a summary).

A good amount of evidence implicates mGlu1 in the pathogenesis of melanoma. Human melanoma biopsies and cell lines, but not healthy nevi or melanocytes, express mGlu1 (Pollock et al., 2003). In the TG-3 mouse, a transgenic mouse model that is predisposed to develop multiple melanomas, expression of mGlu1 in skin tumors is sufficient for melanoma development in vivo (Pollock et al., 2003). Cell lines derived from TG-3 mouse melanoma lesions express the two members of group I, mGlu receptors 1 and 5. To explore the contribution of mGlu5 to melanoma development, mGlu5-null mice were crossed with TG-3 mice. The resulting offspring still developed tumors with onset, progression, and metastasis very close to that described for TG-3, indicating that mGlu1 can promote melanocyte transformation independently of mGlu5 expression (Marín et al., 2005).

Shin et al. (2008) also showed that immortalized melanocytes express mGlu1 receptor whereas nontransformed melanocytes do not; in addition, the presence of mGlu1 was associated with phorbol ester and anchorage independence, indicating that mGlu1 expression was linked to a higher tumorigenic capacity. mGlu1-expressing immortalized melanocytes are tumorigenic in mice, and continuous expression of the receptor is required for maintenance of tumorigenicity as demonstrated by small interfering RNA studies (Shin et al., 2008). More recently, human melanoma cells were shown not only to express mGlu1 but also to release high levels of l-glutamate (Namkoong et al., 2007). Treatment of mGlu1-expressing human melanoma cells with the mGlu1 receptor antagonists LY367385 or (3aS,6aS)-hexahydro-5-methylene-6a-(2-naphthalenylmethyl)-1H-cyclopenta[c]furan-1-one (BAY36-7620) induced a reduction in cell proliferation. The same result was obtained with the glutamate release inhibitor and glutamate transporter activator riluzole (Namkoong et al., 2007). It is noteworthy that riluzole also decreased tumor growth by 50% in mice receiving a human melanoma cell graft (Namkoong et al., 2007).

Pharmacological blockade of mGlu3 receptors reduces cell proliferation and mitogen –activated protein kinase activation in cultured human glioma explants or glioma cell lines that express mGlu3 (D'Onofrio et al., 2003). Furthermore, systemic administration of the mGlu2/3 receptor antagonist, (2S)-2-amino-2-[(1S,2S)-2-carboxycycloprop-1-yl]-3-(xanth-9-yl) propanoic acid (LY341495), inhibits the growth of glioma cells implanted either under the skin or inside the brain parenchima of nude mice. Treatment with LY341495 also significantly reduced the number of Ki-67-positive tumor cells (Arcella et al., 2005) and tumor cell aggregates (Ciceroni et al., 2008) in the brains of these mice.

Chang et al. (2005) studied the expression of mGlu4 receptor in several healthy and transformed human tissues. mGlu4 receptor expression was identified in 68% of colorectal carcinomas, 50% of laryngeal carcinomas, and 46% of breast carcinomas. In the case of colorectal carcinoma, overexpression of mGlu4 was associated with poor prognosis (Chang et al., 2005), and cell lines derived from human colorectal carcinomas showed increased cell invasiveness when treated with l-AP4 (Chang et al., 2005). In another study, comparative proteomics were used to characterize a human colon cancer cell line that was resistant to 5-fluorouracil (5-FU; a common chemotherapy agent). 5-FU-resistant cells were found to overexpress mGlu4 in comparison with the parental cancer cells (Yoo et al., 2004). In the presence of 5-FU, cell survival was increased by the group III mGlu receptor agonist l-AP4 in the nonresistant parent cancer cells; conversely, survival was synergically decreased by 5-FU and the group III receptor antagonist MAP4 in 5-FU-resistant cells (Yoo et al., 2004). It is noteworthy that 5-FU down-regulated mGlu4 expression, and MAP4 had a dose-dependent cytotoxic effect in both cell lines (Yoo et al., 2004).

In contrast to the above evidence, the expression of mGlu4 receptors seems to be inversely related to the severity of human medulloblastoma. After scoring the extent of immunoreactivity for mGlu4 in human biopsies of medulloblastoma, the absence of spinal metastases, cerebrospinal fluid spread, and tumor recurrence as well as the survival of patients were all shown to be associated with high levels of mGlu4 immunoreactivity. Treatment with PHCCC (which is considered a group I mGlu receptor antagonist but can also act as a positive allosteric modulator of mGlu4 receptor) reduced the proliferation of cultured medulloblastoma cells and inhibited the growth of medulloblastoma implants in mice. In addition, subcutaneous or intracranial injections of PHCCC during the first week of life reduced the incidence of medulloblastoma from 85 to 28% in a mutant mouse model known to develop the disease upon X-ray irradiation. This indicates that activation of mGlu4 receptors also affects early events in tumorigenesis (Iacovelli et al., 2006).

In the case of human oral squamous cell carcinoma, strong immunoreactivity for mGlu5 was also associated with a decreased survival rate (Park et al., 2007). The mGlu5 receptor agonist DHPG increased tumor cell migration, invasion, and adhesion in human tongue cancer cells, an effect that was reversed by the mGlu5 receptor antagonist MPEP (Park et al., 2007).

C. Other Disorders Involving Pathological Cell Proliferation/Differentiation

In the retinal epithelium, the pathological transition of epithelial cells to a mesenchymal phenotype, a condition known as proliferative vitreoretinopathy, can lead to retinal detachment and the loss of vision. Abnormal proliferation and migration of retinal pigment epithelium cells play a key role in the development of this retinopathy. It has been shown that glutamate stimulates human retinal pigment epithelium cell proliferation in vitro as well as ERK and cAMP response element-binding protein phosphorylation. These effects can be mimicked by the group I mGlu receptor agonist ACPD and prevented by the receptor inhibitor (S)-α-methyl-4-carboxyphenylglycine. In addition, inhibition of the MEK/ERK pathway prevents ACPD-induced cell proliferation (García et al., 2008). These authors suggest that retinal detachment or an injury to the blood-retinal-barrier may allow the passage of high levels of glutamate to the retina epithelium, thus contributing to the development of proliferative vitreoretinopathy.