Mechanisms Associated with Activation of Intracellular Metabotropic Glutamate Receptor, mGluR5

The group 1 metabotropic glutamate receptor, mGluR5, is found on the cell surface as well as on intracellular membranes where it can mediate both overlapping and unique signaling effects. Previously we have shown that glutamate activates intracellular mGluR5 by entry through sodium-dependent transporters and/or cystine glutamate exchangers. Calibrated antibody labelling suggests that the glutamate concentration within neurons is quite high (~10 mM) raising the question as to whether intracellular mGluR5 is maximally activated at all times or whether a different ligand might be responsible for receptor activation. To address this issue, we used cellular, optical and molecular techniques to show that intracellular glutamate is largely sequestered in mitochondria; that the glutamate concentration necessary to activate intracellular mGluR5 is about ten-fold higher than what is necessary to activate cell surface mGluR5; and uncaging caged glutamate within neurons can directly activate the receptor. Thus these studies further the concept that glutamate itself serves as the ligand for intracellular mGluR5.

Besides mechanisms by which glutamate can enter the cell, another limitation to the notion that endogenous ligand can activate intracellular mGluR5 is the idea that cytoplasmic glutamate concentrations are in the mM range. Indeed, 10 mM is frequently used as the concentration of cytoplasmic glutamate with levels ranging up to 100-200 mM within vesicles [25]. If cytoplasmic glutamate concentrations are indeed 10 mM then an intracellular glutamate receptor would be maximally activated or possibly desensitized long before a new bolus of glutamate entered the cell. To address these issues we have used cellular, optical and molecular techniques to determine the (1) intracellular localization of glutamate; (2) concentrations necessary to activate cell surface and intracellular mGluR5; and (3) the effects of uncaging caged glutamate within neurons.

Cell Culture and Transfection
Primary cultures of striatal neurons were prepared from postnatal day 1 rat pups as previously described [13]. The cells were plated onto 12-mm poly-d-lysine-coated glass coverslips (60,000/coverslip) for immunostaining or Ca 2+ imaging. Cells were cultured in humidified air with 5 % CO 2 at 37 °C for 11-15 days before use. For experiments using the microplate reader, cultures were plated at 40,000 cells per well in black-walled, clear-bottomed 96-well plates and then cultured as above. Striatal cultures were transfected with plasmid mito-eYFP (gift from Dr. Ian Reynolds; Department of Pharmacology, University of Pittsburgh, Pittsburgh, Pennsylvania) using lipofectamine 2000 (Invitrogen, Carlsbad, CA) on DIV 9 and then immunostained on DIV 10.

Fluorescent Measurements of Intracellular Ca 2+
DIV 11-15 striatal neurons grown on coverslips were loaded with Ca 2+ fluorophore Oregon Green 488 BAPTA-1 AM, neurons, we transfected cultures with mito-YFP to label mitochondria and then fixed and stained for glutamate. In support of EM studies showing anti-glutamate immunogold particles over mitochondria [30], immunofluorescence showed anti-glutamate co-localized with mito-YFP in striatal cell bodies and processes (Fig. 1a). In addition, cultures stained with HSP60, a marker of mitochondria as well as anti-glutamate also showed co-localization (Fig. 1b). Thus the majority of immunoreactive glutamate within these GABA-ergic neurons is compartmentalized in mitochondria.

"Location" Bias Apparent in Receptor-Mediated Ca 2+ Responses
Earlier studies [15] showed no significant differences in glutamate binding at receptors prepared from striatal plasma membrane or intracellular membrane sources. Those studies, however, did not address location-specific receptor responses in terms of function. Therefore, we used real time Ca 2+ imaging to determine half-maximal glutamate 2-Methyl-6-(phenylethynyl)pyridine (MPEP, 10 μM, Tocris). Calcium responses in the ROI and control areas were analyzed using MetaMorph software (Molecular Devises).

Glutamate is Sequestered in Neuronal Mitochondria
Previous studies using techniques such as 13 C-NMR, 13 C-and/or 15 N-GC/MS have provided compelling evidence that glutamate has many fates within the cell. For example, a large proportion of glutamate is taken up by the mitochondria where it is transaminated and serves as a substrate for the TCA cycle [27,28]. Indeed, anti-glutamate immunogold electron microscopy studies indicate that particles representing glutamate are clustered over mitochondria as well as the nucleus [29]. We have previously shown that mGluR5 is highly expressed in GABA-ergic striatal neurons [13]. To determine whether glutamate could be visualized in striatal couple to PLC to induce release of Ca 2+ from intracellular stores [24]. Here, we used this assay to show that the half maximal glutamate concentration for intracellular mGluR5 is 61.3 ± 20.3 μM (Fig. 2d). Presumably, the increased EC 50 value associated with intracellular mGluR5 reflects properties of the uptake mechanisms involved in glutamate transport into the cell [13,15]. Collectively, these data show that intracellular, striatal mGluR5 can function independently of signals originating at the cell surface and thus plays a dynamic role in mobilizing Ca 2+ in a specific, localized manner. In addition these data emphasize that intracellular receptors can be activated with glutamate concentrations far lower than the putative intracellular cytoplasmic concentration, consistent with the notion that glutamate is sequestered in the cell.

Selective Uncaging of Glutamate Activates Intracellular mGluR5 Within the Cell and in the Dendrites
To further demonstrate that glutamate activates intracellular mGluR5, we electroporated caged glutamate (MNI-Glu) into individual neurons along with fluoro-ruby to tag recipient cells. Following electroporation, cultures were loaded concentrations associated with the plasma membrane or intracellular mGluR5-mediated Ca 2+ responses. As shown previously [15], glutamate-mediated Ca 2+ changes consisted of two phases, an initial rapid rise followed by a sustained elevation (Fig. 2a, red trace). Both sets of responses were terminated by the addition of the permeable mGluR5 antagonist, MPEP, whereas cultures pretreated with the impermeable, nontransported antagonist LY393053, only exhibited a sustained Ca 2+ response pattern (not shown). As shown previously, LY393053 by itself had no effect on Ca 2+ responses in striatal cultures [13][14][15]. In contrast, addition of the nontransported agonist, DHPG, led to a rapid transient Ca 2+ peak (Fig. 2a, blue trace), which could be blocked by LY393053 (not shown). The half-maximal glutamate concentration to stimulate a rapid transient Ca 2+ response (cell surface) is 2.21 ± 0.8 μM (Fig. 2b) whereas the halfmaximal concentration to induce a sustained plateau Ca 2+ response (intracellular; [15]) is 21.4 ± 4.0 μM (Fig. 2c).
To extend these results, we used a fluorescence-based Ca 2+ flux plate-reader assay in which cells were loaded with the ratiometric Ca 2+ indicator Fura-2 AM before Ca 2+ flux measurement. Previously we used this assay system to show that mGluR5-expressing spinal cord dorsal horn neurons in fluorescence whereas the surrounding regions did not (Fig. 3c, d). MPEP blocked all Ca 2+ responses (not shown). These data emphasize the notion that some of the signaling originates in the ER, an organelle not easily assayed other than using the selective tools described here. Thus akin to results in the hippocampus [23], our findings indicate that activation of dendritic, intracellular mGluR5 also leads to in situ Ca 2+ changes with neither input to nor output from the cell soma.

Discussion
In addition to functioning as a neurotransmitter, glutamate serves many metabolic roles within the cell such as acting as a building block for protein synthesis, playing a role in energy metabolism, and transferring reducing equivalents from the cytoplasm to the mitochondria [31]. Given these myriad tasks, it is not surprising that intracellular glutamate concentrations are thought to be in the mM range. Contrary to this notion, here we show that glutamate is largely compartmentalized in mitochondria, at least in striatal neurons.
Because ultrastructure studies have also shown large numbers of mGluR5 gold particles on endoplasmic reticulum (ER) membranes [12], we tested whether mGluR5 expressed on dendritic ER or endosomal membranes can mediate local Ca 2+ rises. To do so intracellular receptors were pharmacologically isolated by blocking cell surface mGluR5 with LY393053 as well as mGluR1, ionotropic and Group 2/3 mGluR targets. In addition to the antagonists, the extracellular buffer also contained MNI-glutamate which was uncaged in an ROI on a striatal dendrite at least 20 μm away from the cell body (Fig. 3c). Only the region of the dendrite juxtaposed to the uncaging spot exhibited a change  (c, d ). a, b Striatal neurons were injected with fluoro-ruby and MNI-caged-glutamate. Uncaging of MNI-glutamate at a somal ROI induced a Ca 2+ rise at the red ROI whereas no Ca 2+ changes were seen at the blue ROI (N = 5). c, d MNI-caged glutamate was bath applied to striatal cultures at a concentration of 200 µM.
MNI-glutamate was uncaged on a striatal dendrite at the red ROI in the presence of LY53 (20 µM), the NMDA receptor blocker APV (100 µM), the AMPA/Kainate receptor blocker CNQX (20 µM), the mGluR1 blocker CPCCOEt (20 µM) and the Group 2/3 mGluR antagonist LY341495 (100 nM). Uncaged glutamate generated a Ca 2+ rise at the red ROI whereas no Ca 2+ changes were seen in a different neurite in the same field (white arrows) (N = 10) (Color figure online) and regulation of the many enzymes involved in its production will also play an important role in receptor activation.
Moreover, the EC 50 for glutamate activation of intracellular mGluR5 is ~61 μM, a value inconsistent with high concentrations of "free" glutamate within the cytoplasm. Finally uncaging glutamate within the neuronal soma led to a rapid mGluR5-mediated Ca 2+ response further demonstrating intracellular glutamate activation of intracellular receptors.
These studies do not rule out the possibility that glutamate could also be generated in situ. For example, glutamate may also be generated within the nucleus by phosphateactivated l-type glutaminase [32]. Studies by Aledo and co-workers have shown that throughout the brain of humans, monkeys, rats and many other species, l-type glutaminase can be found primarily in the nucleus where it exhibits kinetics consistent with those of the better characterized liver enzyme [32]. More recently, GIP, a scaffolding protein known to bind to a PDZ domain within l-type glutaminase, has been shown to be associated with ER and nuclear membranes [33]. Thus, in addition to transport from extracellular sources and in situ cytoplasmic production, enzymes and scaffolding proteins exist to localize glutamate production near intracellular mGluR5.
In conclusion, uncertainties regarding the concentrations, fate and function of intracellular glutamate as well as findings showing the highly compartmentalized nature of glutamate production and disposition have obscured studies on intracellular receptors and transporters such as mGluR1 [22], mGluR5, and EAAT3. These uncertainties have delayed progress in discerning the role of intracellular glutamate as well as the functional consequences of its production and/or uptake. For instance, the presence of EAAT3 on postsynaptic processes as well as intracellular membranes suggests that glutamate could be targeted to other receptors/channels that might also be present on intracellular membranes. Because these transporters work in "reverse", glutamate could bind to any ligand binding domain present on the same membrane as a given transporter. Given that such proteins would still maintain their "signaling" domains within the cytoplasm, entirely new signaling molecules might be associated with these intracellular receptors. Akin to β-arrestin-scaffolded GPCRs, one set of second messenger systems might be exchanged for another [34]. In support of this notion, our data show that activated intracellular receptors generate sustained Ca 2+ responses and trigger completely different signaling pathways [15]. As mGluR5 plays important roles in development, synaptic function, and learning and memory as well as pathological roles in Fragile X syndrome, anxiety, addiction, and Parkinson disease, understanding the long term consequences of intracellular receptor activation might lead to novel therapeutic strategies for these disorders. It seems likely that besides receptorspecific characteristics and signaling partners that the many factors that modulate glutamate metabolism such as transporter distribution, fluctuation of metabolite concentration,