The peptide N-acetylaspartylglutamate (NAAG) is present in high concentrations in the mammalian central nervous system. Various mechanisms have been proposed for its action, including selective activation of the metabotropic glutamate receptor (mGluR) subtype 3, its action at the N-methyl-d-aspartate receptor, or the production of glutamate by its hydrolysis catalyzed by an extracellular protease. To re-examine its agonist activity at mGluR3, we coexpressed human or rat mGluR3 with G protein inward rectifying channels in Xenopus laevis oocytes. High-performance liquid chromatography analysis of commercial sources of NAAG showed 0.38 to 0.48% glutamate contamination. Although both human and rat mGluR3 were highly sensitive to glutamate, with EC50 values of 58 and 28 nM, respectively, purified NAAG (100 μM) had little activity (7.7% of full activation by glutamate). Only in the millimolar range did it show significant activity, possibly due to residual traces of glutamate remaining in the purified NAAG preparations. In contrast, the unpurified NAAG sample did produce a full agonist response with mGluR3 coexpressed with Gα15, with an EC50 of 120 μM, as measured by a calcium release assay. This response can be explained by the 0.38 to 0.48% glutamate contamination. Our results suggest that NAAG may not have a direct agonist activity at the mGluR3 receptor. Thus, several in vivo and in vitro published results that did not address the issue of glutamate contamination of NAAG preparations may need to be re-evaluated.
The neurochemistry of the modified dipeptide N-acetylaspartylglutamate (NAAG) has been studied extensively over the past 25 years. It is present in millimolar concentrations in many neuronal pathways throughout the central nervous systems of mammals (Coyle, 1997). NAAG is contained in vesicles and released from neurons in a calcium-dependent manner (Neale et al., 2000). In the brain, the major mode of metabolism of NAAG is via glutamate carboxypeptidase II, also known as either prostate-specific membrane antigen, or N-acetyl-α-linked acidic dipeptidase. In the central nervous system, this enzyme is present on astrocytes, where it cleaves the dipeptide into N-acetylaspartate and l-glutamate (Berger et al., 1999). The development of inhibitors of this peptidase has been an active area of study because they have been shown to produce beneficial effects in a variety of in vitro and in vivo cellular and animal models of neurological disorders, including amyotrophic lateral sclerosis (Ghadge et al., 2003); pain, particularly inflammatory pain (Yamamoto et al., 2007); ischemia (Slusher et al., 1999); and diabetic retinopathy (Neale et al., 2005).
Based on neurochemical profiling, it has been assumed for some time that neuroprotective properties of NAAG stem from its reported ability to act as 1) a partial agonist or antagonist at the NMDA receptor, and 2) a highly selective agonist at the metabotropic glutamate receptor (mGluR) subtype 3 of mGluR, with the latter being the most widely accepted mechanism of action (Wroblewska et al., 1997; Neale et al., 2000). The interaction of NAAG at the NMDA receptor has been controversial, and it is not completely clear how the peptide affects NMDA receptor activity. For example, NAAG was reported to activate NMDA receptors, with an EC50 value of 0.7 mM (Westbrook et al., 1986), whereas others have reported that it possesses antagonist activity at this receptor (Bergeron et al., 2005, 2007). Among G protein-coupled receptors, NAAG has been reported to be a highly selective agonist at mGluR3, with an EC50 value of 65 μM at a chimeric receptor consisting of the extracellular ligand binding domain of mGluR3 and the transmembrane domain and carboxyl terminus of mGluR1a (Wroblewska et al., 1997).
As with glutamate stimulation of mGluR3, NAAG stimulation is assumed to inhibit the release of glutamate from nerve terminals, thereby affording neuroprotection in excitotoxic testing paradigms. Another potential mGluR3-mediated neuroprotective mechanism is via the release of growth factors. Activation of mGluR3, which is expressed in both neurons and astrocytes, by the group II agonists 2S,1′R,2′R,3′R)-2-(2,3-dicarboxycyclopropyl)glycine and 4-carboxy-3-hydroxyphenylglycine has been shown to protect neurons from excitotoxic cell death by inducing the release of transforming growth factor-β from astrocytes (Bruno et al., 1998). In fact, experiments conducted on mGluR2 and mGluR3 knockout mice strongly indicate that much of the antiexcitotoxic properties of group II mGluR agonists are mediated by the mGluR3 receptor subtype rather than the mGluR2 subtype (Corti et al., 2007). Thus, mGluR3 may play an as yet underappreciated role in both normal brain function and disease states.
The two group II mGluRs, mGluR2 and mGluR3, are highly homologous with each other. The rat sequences are 65% identical in the extracellular ligand binding domain and 75% identical in the transmembrane domain region. This high degree of amino acid sequence homology has stymied the development of highly receptor subtype-specific compounds (Conn and Pin, 1997; Braüner-Osborne et al., 2007). Nevertheless, NAAG has been suggested to be a selective agonist for mGluR3 (Wroblewska et al., 1997). However, the molecular basis for this specificity has not been reported. A molecular docking analysis conducted on the mGluRs has shown that the glutamate binding site is rather restricted in terms of the space available for larger ligands, such as peptides, to bind (Wang and Hampson, 2006; Wang et al., 2006). In the present study, we re-examined the interaction of NAAG at group II mGluRs. We found that unpurified NAAG, but not purified NAAG, activated mGluR3 and that commercial preparations of NAAG are contaminated with l-glutamate; the latter finding probably explains the previously reported activity at this receptor.
Materials and Methods
Purification of NAAG. NAAG (Sigma-Aldrich, St. Louis, MO) was dissolved in UltraPure water. Prepacked poly-prep cation exchange columns (2 ml) (AG 50W-X8 resin, 200–400 mesh; Bio-Rad Laboratories, Hercules, CA) were washed with water and then with 5 bed volumes of 10 mM HCl, pH 2.0. One milliliter of 10 mM NAAG was loaded onto the column; the unretained fraction containing NAAG was collected, and the column was washed with 1 bed volume of water followed by 5 bed volumes of 10 mM HCl. The bound glutamate can be eluted with 1 M ammonium.
HPLC Analysis. Amino acid analyses of unpurified commercial samples and purified samples of NAAG were carried out using phenylisothiocyanate-derived amino acids on an Alliance Separation Module (Waters, Milford, MA) equipped with a Hypersil ODS column and a 2487 dual-wavelength absorbance detector (set at 254 nm; Waters). The injection volume was 20 μl, the mobile phase was 60% acetonitrile, and the column temperature was 48°C. The data were analyzed using Empower 2 chromatography software (Waters).
Mass Spectrometric Analysis of NAAG Samples. The NAAG samples were analyzed by electrospray liquid chromatography/mass spectrometry by loop injection and gradient elution on a QTOF-1 mass spectrometer (Waters) and a 6210 TOF system (for accurate mass measurements; Agilent Technologies, Santa Clara, CA) fronted with 1100 and 1200 HPLC systems, respectively (Agilent Technologies). A purified batch from Sigma-Aldrich was diluted (10×) with water/acetonitrile/formic acid (1:1:0.1, v/v) before analysis (1-μl aliquot injections). The compound was studied under both positive and negative ionization conditions. For gradient elution, a Discovery HS C18 75 × 2.1 mm, 3 μM column was used at a flow rate of 0.5 ml/min. The analysis was performed using mobile phases of water/acetonitrile/formic acid at ratios of 98:2:0.1 (v/v) (phase A) and 2:98:0.05 (v/v) (phase B). The gradient was ramped linearly from 2 to 95% B over 10 min.
Functional Analysis of mGluRs in HEK 293 Cells. Human embryonic kidney 293 cells (CRL 1573; American Type Culture Collection, Manassas, VA) were transiently transfected in six-well plates using 2 μg of rat wild-type mGlu3 (or rat mGluR2) cDNA plus 2 μg of Gα15 cDNA using the FuGENE 6 reagent (Roche Diagnostics, Indianapolis, IN) or Lipofectamine 2000 (Invitrogen, Carlsbad, CA) as instructed by the manufacturers. Twenty-four hours after transfection, the cells were plated in polyornithine pretreated 96-well microtiter plates (Costar black microtiter plates, Corning Life Sciences, Lowell, MA) at 100,000 cells/100 μl of Dulbecco's modified Eagle's medium. Forty hours after transfection, cells were washed once with calcium assay buffer (20 mM HEPES, pH 7.4, 146 mM NaCl, 5 mM KCl, 1 mM MgCl2, 1 mM CaCl2, 1 mg/ml bovine serum albumin, and 1 mg/ml glucose) and incubated at 37°C in 100 μl of assay buffer. After a 2-h incubation, the assay buffer was changed, and the cells were incubated for another 1 h at 37°C. The assay buffer was removed, and 30 μl of assay buffer containing 6 μM Fluo-4 (Invitrogen) was added into each well; the cells were incubated for 1 h at room temperature in the dark. The cells were washed three times with assay buffer and then incubated in 150 μl of assay buffer for 30 min at room temperature in the dark. Agonists were dissolved in assay buffer, and the responses were recorded on a FLEXstation benchtop scanning fluorometer (Molecular Devices, Sunnyvale, CA) at room temperature, with settings of 485 nm for excitation and 525 nm for emission. Maximal responses were observed with 1 mM l-glutamate. Thus, in each experiment the results are expressed as a percentage of the response obtained with 1 mM l-glutamate and are described as percentage of activation. Prism software (GraphPad Software Inc., San Diego, CA) was used to plot fluorescence intensities and to calculate the EC50 values.
Preparation of Xenopus laevis Oocytes. Surgically removed ovarian lobes of X. laevis frogs were obtained commercially (Nasco, Fort Atkinson, WI), and teased out in OR2 solution (82 mM NaCl, 2.5 mM KCl, 5 mM HEPES, 1.5 mM NaH2PO4, 1 mM MgCl2, and 0.1 mM EDTA, pH 7.4). The oocytes were defolliculated by incubation in 25 ml of OR2 solution containing 2 mg/ml collagenase 1A (Sigma-Aldrich) two times for approximately 60 min on a platform vibrating at 1 Hz and stored in 0.5× Leibovitz's L-15 medium containing 50 mg/ml gentamicin, 10 units/ml penicillin, and 10 mg/ml streptomycin.
Preparation and Injection of cRNA. Capped cRNAs from the linearized vectors containing human mGluR3, human Kir 3.1 genes, and Kir 3.4 genes were prepared using mMESSAGE mMACHINE T7 Ultra kit (Ambion, Austin, TX). For rat mGluR3 cRNA, N-terminally c-Myc-tagged mGluR3 or mGluR2 DNA was linearized by NotI, and 1 μg of purified linearized DNA was mixed with NTP/antireverse cap analog, T7 reaction buffer, and T7 enzyme mix in total volume of 20 μl, and incubated at 37°C for 2 h. After adding 1 μl of DNase I, the mixture was incubated at 37°C for 15 min, and then 20 μl of 5× E. coli Poly (A) Polymerase I E-PAP buffer (Ambion), 10 μl of 25 mM MnCl2, 10 μl of ATP solution, 4 μl of E-PAP, and 36 μl of RNase-free water were added. After 45-min incubation at 37°C, the mixture was placed on ice. cRNA was purified by phenol/chloroform extraction and isopropanol precipitation. Freshly harvested oocytes (0–1 day old) were injected with 25 to 50 nl of RNA cocktail with approximately 5 to 10 ng of the mGluR3 and 1 ng each of Kir 3.1 and 3.4. The injected oocytes were incubated at room temperature for 3 to 7 days before recording.
Two-Electrode Voltage-Clamping Measurements. All measurements were done in a medium containing ND-96 (96 mM NaCl, 2 mM KCl, 1.8 mM CaCl2·2H2O, 1 mM MgCl2·6H2O, and 5 mM HEPES, pH 7.5). Two-electrode voltage-clamp recording was carried out using OpusXpress amplifier (Molecular Devices), which allows simultaneous recording from eight oocytes. The assays for GIRK-based mGluR2 and mGluR3 activation in oocytes have been described previously (Saugstad et al., 1996; Sharon et al., 1997). Oocytes were impaled with two electrodes of 0.5 to 2 MΩ tip resistance when filled with 3 M KCl. Membrane potential was clamped at -60 mV. Current amplitude was measured from baseline to peak using Clampfit (Axon Instruments). Various other data shown here were fitted and plotted using Prism software (GraphPad Software Inc.). The GIRK currents mediated through Kir 3.1/3.4 are K+ channels that were uncovered by perfusing the oocytes with high K+ buffer (2 mM NaCl, 96 mM KCl, 1.8 mM CaCl2·2H2O, 1 mM MgCl2·6H2O, and 5 mM HEPES, pH 7.5), typically for 120 to 180 s. A pulse of 60 to 120 s of agonist(s) was applied during the high K+ pulse, to measure agonist activation of mGluR3 receptor and Gβγ production, which in turn modulates the GIRK currents. For LY341495 antagonist experiments, the agonist pulse was followed by a solution with added antagonist. Application of 3 mM Ba2+ in the buffer blocked most or all of the currents, showing that the current being measured is largely mediated through GIRK channels (data not shown).
Data Analysis. Dose-response curves for agonists and antagonists were fitted by nonlinear regression to the equations I = Imax/[1 + (EC50/Ag)n]or I = Imax - Imax/[1 + (IC50/An)n], where Imax is maximal normalized current response (in the absence of antagonist for inhibitory curves), Ag is the agonist concentration, An is antagonist concentration, EC50 is the agonist concentration eliciting half-maximal current, IC50 is the antagonist concentration eliciting half-maximal current, and n is the Hill coefficient. Antagonist curves were constrained to Imax = 1 and Imin = 0. For agonist efficacy curves, Imin was constrained to zero, but Imax was not constrained. For the antagonists, IC50 values were converted to Kb values using the Leff-Dougall (Leff and Dougall, 1993) variant of the Cheng-Prusoff equation Kb = IC50/((2 + ([Ag]/[EC50])n)1/n - 1), where Ag is the agonist and n is Hill coefficient.
Purification of NAAG. There have been previous speculations that the commercial NAAG may have significant impurities, including glutamate (Losi et al., 2004). NAAG, obtained from Sigma-Aldrich, was purified on a cation exchange column essentially as described in Losi et al. (2004). Figure 1 shows HPLC chromatograms of unpurified (top) and purified (bottom) NAAG. Quantification of the glutamate peak area showed that the unpurified NAAG from various batches from Sigma-Aldrich were contaminated with glutamate in the range of 0.38 to 0.48%, as seen in the Fig. 1 by the peak marked Glu. The batches purified in this study had significantly lower glutamate levels (0.01%), as shown by the diminished glutamate peak in the chromatogram in Fig. 1. Because NAAG does not possess a free amino group to react with the derivatizing reagent, it is not present in the chromatogram. Losi et al. (2004) reported that all commercial NAAG sources contain between 0.1 and 0.5% glutamate, and they had to purify NAAG from Tocris Bioscience (Ellisville, MO) (to <0.1% in their studies). Our qualitative HPLC analysis supports this observation that NAAG from Tocris Bioscience also contained glutamate. A recent study further confirmed glutamate contamination in NAAG from Tocris Bioscience to be ∼0.5% (Fricker et al., 2009). We focused all our studies starting with commercial NAAG from Sigma-Aldrich.
We went on to establish unequivocally whether the purified NAAG indeed was structurally intact. For this purpose, purified NAAG was spectroscopically confirmed by electrospray mass spectrometric analysis. The ESP(+) and ESP(-) mass spectra obtained from analysis of the purified sample from Sigma-Aldrich are shown in Fig. 2. The electrospray(+) mass spectrum shows the protonated molecular ion at m/z 305 together with the protonated dimer at m/z 609. The corresponding ESP(-) negative ions were observed at m/z 303 and 607, respectively. The spectra confirm the molecular mass of the neutral molecule as 304 Da. The elemental composition of the protonated molecular ions (m/z 305) from the two NAAG samples was confirmed, by accurate mass measurement to within 1.04 ppm, respectively.
Analysis of Purified and Unpurified NAGG in a Calcium Release Assay. The effect of commercial (Sigma-Aldrich) unpurified NAAG on mGluR3 coexpressed with Gα15 was measured by calcium-based fluorescence in a FLEXstation benchtop scanning fluorometer. Unpurified NAAG elicited robust activation, with EC50 value of 120 μM (Fig. 3). In contrast, purified preparations of NAAG elicited no measurable response in this assay up to 5 mM on mGluR3 or mGluR2.
Effect of NAAG on mGluR3 Expressed in X. laevis Oocytes. To assess functional activity of unpurified and purified NAAG, human and rat mGluR3 receptors were coexpressed with GIRK1 and GIRK4. Glutamate elicited robust responses to the GIRK current, which enabled us to characterize effect of NAAG compared with glutamate in a more quantitative manner. Concentration-response curves showed that both human and rat mGluR3 are highly sensitive to glutamate, with 58 and 28 nM EC50 values, respectively (Fig. 4). Application of commercial unpurified NAAG at 10 or 100 μM elicited strong current responses in both human (Fig. 5b) and rat mGluR3 (data not shown). Response to 100 μM NAAG was comparable with that with 1 mM glutamate applied to same at the end of the experiment (Fig. 5b). The glutamate response was slightly smaller than the impure NAAG response, possibly due to desensitization or rundown. Over multiple studies, we did not observe a difference between maximal stimulation by glutamate or unpurified NAAG. Purified NAAG, in contrast, had significantly reduced responses to both human (Fig. 5a) and rat mGluR3 (Fig. 5c). The purified NAAG preparations still had a low level of residual glutamate, which could be responsible for residual activity seen in Fig. 5, a and c. Using the numbers from the HPLC assay for the purified NAAG (0.01% glutamate), 100 μM purified NAAG would still have 10 nM glutamate, sufficient to elicit a small response. An application of 10 nM glutamate elicited a response 5.5 ± 1.1% of the full agonist response on human mGluR3 (data from Fig. 4), which fully explains residual current (∼7.7%) by 100 μM purified NAAG in Fig. 5a. Table 1 summarizes the effect of unpurified and purified NAAG on human mGluR3, showing that purified NAAG had highly diminished responses compared with unpurified NAAG.
If residual glutamate in NAAG preparations is responsible for its activity on mGluR3, a competitive antagonist should inhibit the activation with similar potency. Thus, if we compare the inhibition by mGluR3 directly activated by glutamate (1 μM) to that activated by NAAG, in which the concentration of the calculated contaminating glutamate in the NAAG sample is also 1 μM, the inhibition curves should be the same or very similar. The batch of NAAG used at 250 μM has approximately 1 μM glutamate and is expected to fully activate human mGluR3 receptors (Fig. 4). Human mGluR3 receptors were stimulated with either 250 μM NAAG (Sigma-Aldrich) or 1 μM glutamate and dose-dependently inhibited with the mGluR2/3-selective orthosteric antagonist LY341495 in the X. laevis oocyte assay. The inhibition curves from this study were essentially identical (Fig. 6a). LY341495 inhibited both NAAG- and glutamate-activated mGluR3, with an IC50 value of 57 and 71 nM, respectively (Fig. 6a). This corresponds to a calculated Kb value of 4.8 and 3.0 nM or log(Kb) of -8.3 or -8.5, respectively. (Note that the Kb value used here is the Leff-Dougall variant of the Cheng-Prusoff equation that determines antagonist affinity in functional tests correcting for the agonist EC50 value, Hill number, and agonist concentration; Leff and Dougall, 1993.)
As a final demonstration of the presence of glutamate in unpurified NAAG, we tested NAAG on human mGluR2 coexpressed with GIRK channels in X. laevis oocytes. Application of 1 mM NAAG (Sigma-Aldrich) elicited a response that was blocked by LY341495 in a dose-dependent manner, with an IC50 value of 58 nM (Fig. 6b). The response to 1 mM NAAG was 64% (63.4 ± 0.087; n = 7) compared with 100 μM glutamate, which causes 100% activation. The glutamate contamination in 1 mM NAAG was calculated to be approximately 4 to 5 μM; the activation observed with commercial 1 mM NAAG on mGluR2 was very similar to that with equivalent glutamate (EC50 value for glutamate on human mGluR2 is 2.6 μM; data not shown).
We explored the possibility that the activity of NAAG at mGluR3 is due to glutamate contamination in commercial samples rather than to direct activation of mGluR3 by NAAG, in part, because we had variable results with commercial “off-the-shelf” samples. Other studies have questioned the purity of NAAG (Losi et al., 2004) using purified NAAG to reassess it at native NMDA receptors and concluded that NAAG was not active at NMDA receptor in cerebellar granule cells. Given this, we set out to examine the purity of NAAG and whether the purity of NAAG altered its agonist activity at mGluR3.
When rat mGluR3 coexpressed with Gα15 was tested the in FLEXstation calcium release assay, we observed that unpurified NAAG preparations had low potency (EC50 = 120 μM) consistent with previous publications (Wroblewska et al., 1997). HPLC analysis of these NAAG preparations showed that there was significant contamination with glutamate. We estimated that this contamination is approximately 0.38 to 0.48% NAAG. Commercial NAAG preparation was then purified through an ion exchange column, which substantially reduced the glutamate level to 0.01%. This purified NAAG preparation was inactive at mGluR3 in the FLEXstation assay. Mass spectrometric analysis confirmed that intact NAAG was present in the purified samples.
To independently confirm this observation, we assessed the activity of glutamate, unpurified NAAG, and the glutamate-depleted NAAG in X. laevis oocytes expression rat or human mGluR3 with GIRK. In this preparation, we confirmed the high potency of glutamate (EC50 = 58 nM at human mGluR3 and 28 nM at rat mGluR3). Unpurified NAAG showed robust currents, confirming the FLEXstation observations. In addition, the glutamate-depleted NAAG was nearly inactive; residual activity could be due to the trace remaining glutamate. Thus, in the NAAG dose-response study in the FLEXstation, at the EC50 value (120 μM), there would have been approximately 0.6 μM glutamate. In the oocytes studies, 10 and 100 μM NAAG would have approximately 0.5 and 5 μM glutamate, respectively. Given the high potency of glutamate at mGluR3, this level of glutamate contamination is sufficient to explain the NAAG responses in both assays. The observation that the inhibition curves of NAAG or glutamate by LY341495 overlap, and their corresponding Kb values are similar, constitutes independent support for our hypothesis that the activation of mGluR3 by NAAG is through contaminating glutamate. Finally, the commercial NAAG sample was able to activate mGluR2 at high doses predicted by the level of glutamate contamination. Thus, these data demonstrate that at recombinant rat and human mGluR3, NAAG is not an agonist.
Demonstrating that NAAG is not a selective mGluR3 agonist has implications on studies published previously and on their resulting interpretations. Using [3H]LY354740, Schweitzer et al. (2000) demonstrated that glutamate (from Tocris Bioscience) had an affinity of 0.44 μM at mGlu3 and 1.2 μM at mGluR2, whereas NAAG showed affinities of 19 and 236 μM, respectively. These NAAG affinity determinations can be explained by glutamate contamination in the NAAG preparation. Likewise, many studies have used NAAG to demonstrate an activity in native preparations and ascribe this to activation of mGluR3.
Our findings suggest that conclusions from previous studies investigating the effects of NAAG on mGluR3 need to be re-evaluated. Several studies have used NAAG in native preparations and have shown a lack of activity and interpreted the result as demonstration that mGluR3 is not involved in the observed activity. For example, Neale and Salt (2006) showed that the mGluR2/3 agonist LY354740 modulated the inhibitory postsynaptic currents in the superior colliculus, whereas NAAG (up to 500 μM) was without activity. These results were interpreted to demonstrate that the activity of LY354740 was mediated by mGluR2. Another study (Pöschel et al., 2005) suggested that “mGluR3 is critically required for hippocampal LTP [long-term potentiation]” based on the activity of commercial NAAG in an in vivo preparation. In these studies, and others, the use of NAAG to show the role of mGluR3 needs to be re-evaluated because almost all studies have not have addressed the issue of glutamate contamination. Thus, the in vivo effects of NAAG described previously are probably due to other mechanisms, such as NAAG as a source of glutamate released by the action of NAAG peptidases. The interpretation of different studies is further confounded by the possibility that mGluR3 receptor may have coupling to different G proteins in different systems (Wroblewska et al., 2006). Several studies have linked mGluR3 to the pathophysiology of schizophrenia (Coyle, 2006) through candidate gene association studies (Addington et al., 2004). These hypotheses on the role of mGluR3 in schizophrenia have been, at least in part, based on using NAAG to demonstrate a selective mGluR3 action. A recent study further supported our findings that commercial sources of NAAG are contaminated with glutamate. They showed that purified NAAG had little effect on GIRK currents from HEK cells transfected with human mGluR2 or mGluR3 and human GIRK1/2 (Fricker et al., 2009). They also showed that 1 μM LY341495 abolished glutamate-evoked GIRK. They did not show whether LY341495 inhibits currents stimulated by commercial NAAG. We have explored this in much more detail in an X. laevis oocyte-based assay. First we show that commercial NAAG-induced currents are inhibited by LY341495 in a dose-dependent manner in both human mGluR2 and mGluR3. Detailed concentration-response curves on both human mGluR3 and human mGluR2 allowed us to calculate the affinity of the active moiety in NAAG to be identical to that of glutamate, suggesting that glutamate is indeed a likely pharmacological moiety responsible for this activity in impure NAAG. We further confirmed purity of NAAG in our purified preparations by mass spectroscopic techniques.
In summary, we have shown that NAAG is not an agonist at mGluR3 receptor. We speculate that it may have an alternative role, such as acting as an N-acetyl-α-linked acidic dipeptidase-regulated reservoir of glutamate. The hypothesis that the high concentration of NAAG in the central nervous system may serve as an important source of physiological glutamate requires further investigation.
We thank Geeta Ayer and Alan Robbins for generating the human mGluR3 and Kir constructs.
This work was supported in part by the Canadian Institutes for Health Research [Grant MOP-81179] (to D.R.H.).
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
ABBREVIATIONS: NAAG, N-acetylaspartylglutamate; NMDA, N-methyl-d-aspartate; mGluR, metabotropic glutamate receptor; HPLC, high-performance liquid chromatography; HEK, human embryonic kidney; GIRK, G protein-coupled inwardly rectifying potassium channel; ESP, electrospray; LY341495, 2S-2-amino-2-(1S,2S-2-carboxycyclopropyl-1-yl)-3-(xanth-9-yl)propanoic acid; LY354740, (2S,4S)-2-amino-4-(4,4-diphenylbut-1-yl)-pentane-1,5-dioic acid.
- Received February 19, 2009.
- Accepted April 22, 2009.
- The American Society for Pharmacology and Experimental Therapeutics