Retracted: Chronic NMDA administration to rats up-regulates frontal cortex cytosolic phospholipase A2 and its transcription factor, activator protein-2

Authors

  • Jagadeesh S. Rao,

    1. Brain Physiology and Metabolism Section, National Institute on Aging, National Institutes of Health, Bethesda, Maryland, USA
    Search for more papers by this author
  • Renee N. Ertley,

    1. Brain Physiology and Metabolism Section, National Institute on Aging, National Institutes of Health, Bethesda, Maryland, USA
    Search for more papers by this author
  • Stanley I. Rapoport,

    1. Brain Physiology and Metabolism Section, National Institute on Aging, National Institutes of Health, Bethesda, Maryland, USA
    Search for more papers by this author
  • Richard P. Bazinet,

    1. Brain Physiology and Metabolism Section, National Institute on Aging, National Institutes of Health, Bethesda, Maryland, USA
    2. Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada
    Search for more papers by this author
  • Ho-Joo Lee

    1. Brain Physiology and Metabolism Section, National Institute on Aging, National Institutes of Health, Bethesda, Maryland, USA
    Search for more papers by this author

Errata

This article is corrected by:

  1. Errata: Retraction Volume 140, Issue 6, 979, Article first published online: 13 February 2017

Address correspondence and reprint requests to Jagadeesh S. Rao, PhD, Brain Physiology and Metabolism Section, National Institute on Aging, National Institutes of Health, 9000 Rockville Pike, Bldg. 9, 1S-126, Bethesda, MD 20892, USA. E-mail: jrao@mail.nih.gov

Abstract

Excessive N-methyl-d-aspartate (NMDA) signaling is thought to contribute to bipolar disorder symptoms. Lithium and carbamazepine, effective against bipolar mania, are reported in rats to reduce brain transcription of an arachidonic acid selective calcium-dependent cytosolic phospholipase A2 (cPLA2), as well as expression of one of its transcription factors, activator protein (AP)-2. In this study, we determined if chronic administration of NMDA (25 mg/kg i.p.) to rats would increase brain cPLA2 and AP-2 expression, as these antimanic drugs are known to down-regulate excessive NMDA signaling. Administration of a daily subconvulsive dose of NMDA to rats for 21 days decreased frontal cortex NMDA receptor (NR)-1 and NR-3A subunits and increased cPLA2 activity, phosphorylation, protein, and mRNA levels. The activity and protein levels of secretory phospholipase A2 or calcium-independent phospholipase A2 were not changed significantly. Chronic NMDA also increased the DNA-binding activity of AP-2 and the protein levels of its α and β subunits. These changes were absent following acute (3 h earlier) NMDA administration. The changes, opposite to those found following chronic lithium or carbamazepine, are consistent with up-regulated arachidonic acid release due to excessive NR signaling and may be a contributing factor to bipolar mania.

Abbreviations used
AMPA

α-amino-3 hydroxy-5-methyl-4-isoxazolprorionic acid

AP-2

activator protein-2

cPLA2

calcium-dependent cytosolic phospholipase A2

DTT

dithiothreitol

GRE

glucocorticoid response element

iPLA2

calcium-independent phospholipase A2

mGluRs

metabotropic receptors

NF-κB

nuclear factor kappa B

NMDA

N-methyl-d-aspartate

NR

NMDA receptor

PEA 3

polyoma enhancer element 3

sPLA2

secretory phospholipase A2

The excitatory neurotransmitter glutamate plays important roles in physiological and pathological processes within the central nervous system. Glutamate receptors are classified into two major classes, ionotropic receptors and metabotropic receptors (mGluRs). Ionotropic receptors are classified further based on selectivity to α-amino-3 hydroxy-5-methyl-4-isoxazolprorionic acid (AMPA), kainate, and N-methyl-d-aspartate (NMDA). Whereas mGluRs are classified into three groups, mGluRs I, II, and III and are coupled to G proteins. Activation of glutamate receptors induces signaling cascades involved in learning and memory, synaptogenesis, and nociception (Dingledine et al. 1999). Altered glutamatergic signaling has been suggested in many neurological and psychiatric disorders including Alzheimer (Fang et al. 2005; Snyder et al. 2005), Huntington (Young et al. 1988), and Parkinson disease (Hallett et al. 2005), schizophrenia (Mueller and Meador-Woodruff 2004), and bipolar disorder (Clinton and Meador-Woodruff 2004; Mueller and Meador-Woodruff 2004; Basselin et al. 2006).

NMDA receptors (NRs) are present throughout the brain and are composed of several subunits (NR-1, NR-2A-2D, and NR-3A-3B) which are derived from seven genes (Riva et al. 1994; Blahos and Wenthold 1996; Goebel and Poosch 1999). Fewer NR-1 (Nudmamud-Thanoi and Reynolds 2004; Beneyto et al. 2007) and NR-3A subunits (Mueller and Meador-Woodruff 2004) in brain have been implicated in bipolar disorder. Chronic exposure to therapeutically relevant concentrations of lithium, valproate, or carbamazepine, agents effective in the mania of bipolar disorder, decreases NMDA-induced cytoplasmic vacuolization in rat primary hippocampal cultures (Bown et al. 2003), decreases NMDA-induced currents in hippocampal cultures, decreases phosphorylation of the NR-2B subunit, and decreases NMDA-induced calcium influx in cerebral cortical neurons (Hashimoto et al. 2002). Pre-clinical studies suggest that AMPA receptor levels are increased by drugs with an antidepressant profile and decreased by antimanic drugs (Du et al. 2003, 2007), whereas AMPA receptor levels from post-mortem brain samples of patients with bipolar disorder do not differ significantly from controls (Scarr et al. 2003; Beneyto et al. 2007).

Activation of the NMDA receptor induces an influx of extracellular calcium into cortical and spinal cord neurons (MacDermott et al. 1986). Increased intracellular calcium can activate many calcium dependent signaling enzymes, including calcium-calmodulin kinase (Fukunaga et al. 1992), mitogen-activated protein kinase (Chandler et al. 2001), protein kinase C (Fukunaga et al. 1992), phospholipase C (Dudek et al. 1989), calcineurin (Halpain et al. 1990), and the calcium-dependent cytosolic phospholipase A2 (cPLA2) that selectively releases arachidonic acid from membrane phospholipid (Lazarewicz et al. 1988, 1990; Clark et al. 1995; Weichel et al. 1999). Consistent with cPLA2 activation, acute NMDA administration induces arachidonic acid release in cerebral granular cells (Lazarewicz et al. 1990), hippocampus (Sanfeliu et al. 1990), cortex (Tapia-Arancibia et al. 1992), striatal neurons (Dumuis et al. 1993), and in multiple regions within the brain of the awake rat (Basselin et al. 2005).

Activation of PLA2 by NMDA is of particular interest to bipolar disorder because two effective antimanic agents, lithium and carbamazepine, when administered chronically to rats at therapeutically relevant doses decrease transcription of cPLA2, its protein level and activities, while not affecting brain expression of secretory sPLA2 or calcium-independent iPLA2 (Rintala et al. 1999; Rapoport and Bosetti 2002; Ghelardoni et al. 2004). Additionally, the reduced transcription of cPLA2 following chronic lithium and carbamazepine is accompanied by a reduced DNA binding activity of the cPLA2-transcription factor, activator protein (AP-2), suggesting a mechanism for the cPLA2 change (Rao et al. 2005, 2007a).

Based on our studies with chronic lithium and carbamazepine, we suggested that some antimanic agents are effective in bipolar disorder by down-regulating arachidonic acid release and turnover through inhibition of arachidonic acid-selective cPLA2 (Rintala et al. 1999; Rapoport and Bosetti 2002; Ghelardoni et al. 2004; Bazinet et al. 2006). In this case, an animal model of bipolar disorder might be one in which cPLA2 is selectively up-regulated. The data on NMDA (see above) suggest that rats administered NMDA chronically might represent such a model. Thus, in the present study, we examined the effect of chronic and acute NMDA on brain expression of cPLA2 as well as of sPLA2 and iPLA2 (Dennis 1994), on the DNA binding activity of AP-2, and the brain expression of NR subunits. We analyzed the frontal cortex because clinical studies indicate functional and structural abnormalities in this region in bipolar disorder patients (Rajkowska 2002; Lyoo et al. 2004; Soares et al. 2005) and because we performed our prior studies on frontal cortex tissue (Rao et al. 2005, 2007a). Upon finding that chronic NMDA selectively increased frontal cortex cPLA2 expression, we further examined the transcriptional basis for this increase (Rao et al. 2005, 2007a).

Materials and methods

Animals

The study was conducted following the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals (Publication no. 80-23), and was approved by the Animal Care and Use Committee of the National Institute of Child Health and Human Development. Male CDF-344 rats, weighing 200–215 g (Charles River Laboratories; Wilmington, MA, USA), were acclimatized for 1 week in an animal facility with controlled temperature, humidity, and light cycle, and food (NIH-31) and water were provided ad libitum.

Rats were randomly assigned to one of two chronic treatment groups where they received either saline (control) or drug (NMDA) by i.p. injection. The control group (= 10) received vehicle (0.9% saline) once daily for 21 days, and the NMDA group (= 10) received 25 mg/kg NMDA (Sigma Chemical Co., St Louis, MO, USA) once daily for 21 days. This dose is subconvulsive but can induce paroxysmal EEG activity (Ormandy et al. 1991), and it increases arachidonic acid signaling, which can be blocked by the NMDA receptor inhibitor MK-801, in the brain of awake rats (Basselin et al. 2005). Three hours after the last saline or NMDA injection, a rat was anesthetized with CO2 and decapitated. The brain was rapidly excised and the frontal cortex dissected, cut sagittally, frozen in 2-methylbutane at −50°C, and stored at −80°C until use.

For acute studies, animals were randomized into two groups that received a single i.p. injection of either 25 mg/kg NMDA (= 10), or 0.9% saline (vehicle) (= 10). Three hours after injection, animals were anesthetized with CO2 and decapitated and the frontal cortex was removed and stored as described above.

Preparation of membrane and cytosol fractions

Membrane and cytosolic fractions were prepared from frontal cortex as previously described by (Dwivedi et al. 2000). Frontal cortex from control and chronic NMDA administered rats were homogenized in a homogenizing buffer containing 20 mmol/L Tris-HCl (pH 7.4), 2 mmol/L EGTA, 5 mmol/L EDTA, 1.5 mmol/L pepstatin, 2 mmol/L leupeptin, 0.5 mmol/L phenylmethylsulfonyl fluoride, 0.2 U/mL aprotinin, and 2 mmol/L dithiothreitol (DTT), using a Polytron homogenizer. The supernatant was centrifuged at 100 000 g for 60 min at 4°C. The resulting supernatant was the cytosol fraction, and the pellet was resuspended in the homogenizing buffer containing 0.2% (w/v) Triton X-100. The homogenate was kept at 4°C for 60 min with occasional stirring and then centrifuged at 100 000 g for 60 min at 4°C. The resulting supernatant was used as the membrane fraction. Protein concentrations of nuclear extracts, membrane and cytosolic fractions were determined by using Bio-Rad protein Reagent (Bio-Rad, Hercules, CA, USA).

Preparation of nuclear extracts

Nuclear extracts were prepared from frontal cortex of control and NMDA administered rats, as previously described (Lahiri 1998). Briefly, the frontal cortex was homogenized in 10 mmol/L HEPES, pH 7.9, 0.1 mmol/L EDTA, 0.1 mmol/L EGTA, 1 mmol/L DTT, and 10 mmol/L KCl, buffer with a protease inhibitor cocktail (Roche, Indianapolis, IN, USA) using a Teflon-glass homogenizer. After adding 0.5% NP-40, five additional strokes of homogenization were performed. The suspension was incubated for 10 min on ice, and then centrifuged in a microcentrifuge at 13 000 g for 1 min at 4°C. To the nuclear pellet, solution B (20 mmol/L HEPES, pH 7.9, 1 mmol/L EDTA, 1 mmol/L EGTA, 1 mmol/L DTT, and 0.4 mol/L NaCl) and a protease inhibitor cocktail (Roche) were added. Tubes were mixed and placed on a small rotatory shaker for 30 min. Finally, the mixture was centrifuged at 13 000 g for 3 min at 4°C. The supernatant containing the proteins from the nuclear extracts was transferred to a fresh tube.

Western blot analysis

Proteins from membrane, cytosolic, and nuclear extracts (65 μg) were separated on 10–20% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Bio-Rad). Following sodium dodecyl sulfate-polyacrylamide gel electrophoresis, the proteins were electrophoretically transferred to a nitrocellulose membrane (Bio-Rad). Membrane and cytosolic protein blots were incubated overnight in Tris-buffered saline buffer, containing 5% non-fat dried milk and 0.1% Tween-20. Membranes were probed with specific primary antibodies (1 : 1000 dilution) for the NR-1, NR-2A, NR-2B, and NR-3A or for phosphorylated NR-1(Ser896) or phosphorylated NR-2B (Ser1303) (Upstate Biotech, Charlottesville, VA, USA). Cytosolic blots were incubated with primary antibodies for group IVA cPLA2,, group IIA sPLA2, group VIA iPLA2, (Santa Cruz Biotech, Santa Cruz, CA, USA), and for phospho-cPLA2 (1 : 500) (Cell Signaling, Beverly, MA, USA) (Dennis 1994). Nuclear AP-2α, AP-2β, and AP-2γ protein levels were determined using specific primary antibodies (1 : 1000 dilution, Santa Cruz). Membrane, cytosolic and nuclear protein blots were incubated with appropriate horseradish peroxidase-conjugated secondary antibodies (Bio-Rad) and were visualized using a chemiluminescence reaction (Amersham, Piscataway, NJ, USA) on X-ray film (XAR-5, Kodak, Rochester, NY, USA). Optical densities of immunoblot bands were measured using Alpha Innotech Software (Alpha Innotech, San Leandro, CA, USA) and were normalized to β-actin (Sigma) to correct for unequal loading. All experiments were carried out twice with up to 10 independent samples. Values are expressed as percent of control.

Phospholipase A2 activities

Activities of individual PLA2 enzymes were measured in the cytosolic fraction as described in detail elsewhere (Yang et al. 1999; Lucas and Dennis 2005), with slight modifications to the extraction. For cPLA2 activity, a portion of the cytosolic fraction was incubated in 100 μmol/L 1-palmitoyl-2-arachidonoyl-sn-glycerol-3-phosphorylcholine (Avanti; Alabaster, AL, USA) and phosphatidylinositol 4,5-bisphosphate (97 : 3) containing approximately 100 000 cpm of 1-palmitoyl-2-[1-14C] arachidonoyl-sn-glycerol-3-phosphorylcholine (Perkin-Elmer, Boston, MA, USA) and 4,5 biphosphatidylinositol (Avanti) in 400 μmol/L triton X-100 mixed micelles containing 100 mmol/L Hepes, pH 7.5, 80 μmol/L calcium, 2 mmol/L DTT, and 0.1 mg/mL fatty acid free bovine serum albumin. For iPLA2 activity, a portion of the cytosolic fraction was incubated in 100 μmol/L 1-palmitoyl-2-palmitoyl-sn-glycerol-3-phosphorylcholine (Avanti) containing approximately 100 000 cpm of 1-palmitoyl-2-[1-14C] palmitoyl-sn-glycerol-3-phosphorylcholine (Amersham, Buckinghamshire, UK) in 400 μmol/L Triton X-mixed micelles in 100 mmol/L HEPES, pH 7.5, 5 mmol/L EDTA, 2 mmol/L DTT, and 1 mmol/L ATP. For sPLA2 (IIA) activity, a portion of the cytosolic fraction was incubated in 100 μmol/L 1-palmitoyl-2-linoleoyl-sn-glycerol-3-phosphorylethanolamine (Avanti) and 1-palmitoyl-2-oleoyl-sn-glycerol-3-phosphorylserine (Avanti) (1/1) in the form of small unilamellar vesicles containing approximately 100 000 cpm of 1-palmitoyl-2-[1-14C] linoleoyl-sn-glycerol-3-phosphorylethanolamine (Amersham) in 100 mmol/L Hepes, pH 7.5, with 1 mmol/L calcium and 1 mg/mL bovine serum albumin.

The assays were started by adding reagent to cytoplasm extracts for 30 min at 40°C in a shaking bath. Reactions were terminated by adding Dole’s reagent (2-propanol: heptane: 0.5 mol/L sulfuric acid, 400 : 100 : 20, by volume) followed by vortexing. Released [1-14C] fatty acids were extracted with the addition of heptane and water. One milliliter of the heptane was loaded on a bond elute reservoir with a frit pre-loaded with silicic acid. The unesterified [1-14C] fatty acids were eluted from the silicic acid by adding diethyl ether with the help of a vacuum. Radioactivity of the elutant was determined by liquid scintillation counting and activity was calculated after correcting for the background of blank samples. All samples were run in triplicate and values are expressed in pmol/min/mg of protein.

Total RNA isolation and real time RT-PCR

Total RNA was isolated from frontal cortex of control and chronic NMDA administered rats using RNeasy lipid tissue mini kit (Qiagen, Valencia, CA, USA). Briefly: tissue was homogenized in Qiagen lysis solution and total RNA was isolated by phenol-chloroform extraction. cDNA were prepared from total RNA according to manufacture instructions using high capacity cDNA archives kit (Applied Biosystems, Foster City, CA, USA). Briefly: cDNA was generated in a thermal cycler using 5 μg of total RNA and mixture of multiscribe reverse transcriptase (50U/μL), random primers (10X), dNTPS (25X). Expression of cPLA2 was determined using specific primers and probes for cPLA2, purchased from TaqManR gene expression assays (Applied Biosystems), consisted of a 20x mix of unlabeled PCR primers and Taqman minor groove binder probe (FAM dye-labeled). The fold change in gene expression was determined using the ΔΔCT method (Livak and Schmittgen 2001). Data were expressed as the relative level of the target gene (cPLA2) in the chronic NMDA-administered animals normalized to the level of the endogenous control (β-globulin) and relative to the control rats (saline injected) (calibrator), as previously described (Ghelardoni et al. 2004; Rao et al. 2005). All experiments were carried out twice in triplicate with six independent samples per group.

Electrophoretic mobility shift assay

Gelshift assays were performed on the nuclear extracts from rat frontal cortex to examine the DNA-binding activities of transcription factors AP-1 and AP-2, glucocorticoid response element (GRE), nuclear factor kappa B (NF-κB), and polyoma enhancer element 3 (PEA 3), which are known to regulate cPLA2 mRNA expression (Morri et al. 1994). Nuclear protein extracts (15 μg) were incubated with a non-radioactive (10 ng) biotin labeled DNA oligo consensus (Panomics, Redwood City, CA, USA) in gelshift buffer (10 mmol/L Tris-HCl, pH 7.5, 1 mmol/L NaCl, 1 mmol/L MgCl2, 0.5 mmol/L EDTA, 0.5 mmol/L DTT, 4% glycerol, and 50 μg/mL poly dI : dC) for 30 min on ice. The DNA-protein complex was separated on a 5% Tris-Borate-EDTA gel and electrophoretically transferred to a nylon membrane. The biotin-labeled oligonucleotide complex was visualized using a streptavidin-horseradish peroxidase conjugate coupled with chemiluminescence on X-ray film (Kodak). The following oligonucleotide sequences were used for the gelshift assay: AP-1-CGCTTGATGAGTCAGCCGGAA; NF-κB-AGTTGAGGGGACTTTCCCGGC; AP-2-GATCGAACTGACCGCCCGCGGCCCGT; GRE-GACCCTAGAGGATCTGTACAGGATGTTCTAGATCCAATTCG; and PEA 3-GATCTCGAGCAGGAAGTTA. The specificity of each transcription factor was determined using 100 times excess unlabeled probe with a fixed amount of biotin labeled DNA oligo consensus (10 ng) and nuclear extracts (15 μg). Optical densities of gelshift bands were quantified using Alpha Innotech software (Alpha Innotech). Values were expressed as percent of control.

Statistical analysis

Data are expressed as means ± SEM. Statistical significance was calculated using a two-tailed, unpaired t-test with statistical significance set at < 0.05.

Results

Effect of chronic NMDA on NMDA receptor subunits in rat frontal cortex

Compared with controls, chronic NMDA significantly decreased NR-1 (Fig. 1a, = 0.005) and NR-3A protein levels (Fig. 1d, = 0.01) without detectable differences in NR-2A or NR-2B levels (Fig. 1b and c, = 0.45; = 0.58). There was no significant difference in the phosphorylation of either the NR-1(Ser896) or NR-2B (Ser1303) subunit between chronic NMDA-treated animals and controls (Fig. 1e and f, p = 0.91; p = 0.59).

Figure 1.

 Representative immunoblots of (a) NR-1, (b) NR-2A, (c) NR-2B, and (d) NR-3A protein in frontal cortex of rats chronically administered saline (control, = 8) or N-methyl-d-aspartate (NMDA) (= 8). Representative immunoblots of phosphorylated (e) NR-1 and (f) NR-2B protein in frontal cortex of rats chronically administered saline (control, = 8) or NMDA (= 8). Bar graphs are ratios of optical densities of non-phosphorylated or phosphorylated NR subunits to β-actin, expressed as percent of control, and were compared using an unpaired t-test, mean ± SEM (*p < 0.05, **p < 0.01).

Effect of chronic NMDA on PLA2 activities in rat frontal cortex

Chronic NMDA significantly increased cPLA2 (IV) activity (75%) (Fig. 2ap = 0.02), but not sPLA2 (IIA) or iPLA2 (VI) activity (Fig. 2b and c, p = 0.11; p = 0.24) compared with respective control values.

Figure 2.

 Activities of (a) calcium-dependent cytosolic phospholipase A2 (cPLA2) IV, (b) secretory phospholipase A2 (sPLA2) IIA, and (c) calcium-independent phospholipase A2 (iPLA2) VI in frontal cortex of rats chronically administered saline (= 8) or N-methyl-d-aspartate (NMDA) (n = 8) were compared using an unpaired t-test, mean ± SEM. (*p < 0.05). Representative immunoblots and protein levels of (d) cPLA2 (control, n = 10; NMDA, n = 9), (e) sPLA2 (control, n = 10; NMDA, n = 10), and (f) iPLA2 (control, n = 10; NMDA, n = 8) in frontal cortex of rats chronically administered saline or NMDA. Bar graphs are ratios of optical densities of immunoblots to that of β-actin, expressed as percent of control, and were compared using an unpaired t-test, mean ± SEM (*p < 0.05; **p < 0.01).

Effect of chronic and acute NMDA on PLA2 isoforms in rat frontal cortex

The increased cPLA2 activity after chronic NMDA was associated with a 66% increased cPLA2 protein level in the frontal cortex, compared with the level in control rats (Fig. 2d, p = 0.0024). In contrast, a significant change in cPLA2 protein was not observed 3 h after acute NMDA administration (Fig. 3a). Frontal cortex sPLA2 (IIA) and iPLA2 (VI) protein levels were not significantly different after chronic NMDA treatment, compared with controls (Fig. 2e and f) (p = 0.35, p = 0.54).

Figure 3.

 Representative immunoblots of (a) calcium-dependent cytosolic phospholipase A2 (cPLA2) IV in frontal cortex of rats acutely (3 h earlier) administered saline (control, n = 10) or N-methyl-d-aspartate (NMDA) (n = 10); and of (b) phosphorylated cPLA2 in frontal cortex of rats chronically administered saline (control, n = 10) or NMDA (n = 7). Bar graphs are ratios of optical density of cPLA2 or phosphorylated cPLA2 to that of β-actin, expressed as percent of control. (c) cPLA2 IV mRNA in the cortex of rats chronically administered saline (control, n = 6) or NMDA (n = 6), measured using RT-PCR. Each sample was assayed in triplicate in two independent experiments. Bar graphs are expressed as level of cPLA2 in NMDA-administered animals normalized to endogenous control (β-globulin) and relative to the control rats (calibrator), using the ΔΔCT method. Data were compared using an unpaired t-test, mean ± SEM (*p < 0.05). (d) Specificity of DNA-protein binding activity for AP-1, AP-2, glucocorticoid response element (GRE), nuclear factor kappa B (NF-κB), and polyoma enhancer element 3 (PEA 3) in rat frontal cortex nuclear extracts was tested in the absence (a) and presence (b) of excess (100 times) unlabeled probe, as described in Materials and methods. Addition of labeled probe showed presence of shift and the addition of excess unlabeled probe blocked the binding of labeled probe. (e) Addition of antibody to AP-2α, AP-2β, or AP-2γ generated a supershift.

Effect of chronic NMDA on cPLA2 phosphorylation and mRNA in rat frontal cortex

Chronic NMDA increased (65%) the level of phosphorylated (Ser505) cPLA2 protein in rat frontal cortex compared with controls (Fig. 3b, p = 0.010). We examined the basis for the increased cPLA2 protein level after chronic NMDA, by measuring cPLA2 mRNA. The frontal cortex cPLA2 mRNA level was significantly increased (78%) in rats chronically administered NMDA, compared with controls (Fig. 3c, p = 0.045).

Effect of chronic NMDA on cPLA2-regulating transcription factors

An excess amount (100 times) of cold specific oliogonucleotide blocked the binding of labeled oligonucleotides to nuclear proteins, indicating the specificity of DNA-binding activities (Fig. 3d). The addition of antibodies for AP-2α, AP-2β, or AP-2γ generated a super shift indicating the presence of the respective proteins in the complex (Fig. 3e). We examined whether the increase in cPLA2 mRNA with chronic NMDA was associated with up-regulation of known cPLA2-regulating transcription factors, AP-1, AP-2, GRE, NF-κB, and PEA 3 (Morri et al. 1994). There was a significant increase in the DNA-binding activity of the AP-2 transcription factor following chronic NMDA (Fig. 4ap = 0.004), but not with binding activity of NF-κB, GRE, AP-1, or PEA 3 (Fig. 4b–e).

Figure 4.

 Representative gelshift bands (DNA–protein complex) of AP-2, nuclear factor kappa B (NF-κB), glucocorticoid response element (GRE), AP-1, and polyoma enhancer element 3 (PEA 3) in frontal cortex from rats chronically administered saline (control, n = 8) or N-methyl-d-aspartate (NMDA) (n = 8). DNA-binding activity was measured in brain nuclear extracts as described in Materials and methods section. Data were compared using an unpaired t-test, mean ± SEM (**p < 0.01). Bar graphs are given relative to control.

The effect of chronic NMDA on AP-2α, AP2β, and AP-2γ protein levels

We considered the basis for up-regulated AP-2 DNA-binding activity by examining protein levels of its subunits (Fig. 5). Compared with controls, chronic NMDA significantly increased protein levels of AP-2α (26%, p = 0.002) and AP-2β (30%, p = 0.01) in the cytosolic and nuclear (AP-2α, 33%, p = 0.048; AP-2β, 33%, p = 0.011) fractions. There was no significant difference between chronic NMDA-treated animals and controls in protein levels of AP-2γ in cytosol (p = 0.07) and nuclear extracts (p = 0.64).

Figure 5.

 Representative immunoblots and protein levels of cytosolic AP-2α [control, n = 9; N-methyl-d-aspartate (NMDA), n = 10], AP-2β (control, n = 9; NMDA, n = 10), and AP-2γ (control, n = 10; NMDA, n = 10), and of nuclear AP-2α (control, n = 9 and chronic NMDA, n = 7) and AP-2β (control, n = 6 and chronic NMDA, n = 8), and (f) AP-2γ (control, n = 10 and chronic NMDA) in frontal cortex from rats chronically administered saline (control) or NMDA. Immunolabeling of cytosolic and nuclear AP-2 proteins was performed as described in Materials and methods. Data are ratios of optical density of each AP-2 isoform to that of β-actin, expressed as percent of control. Data were compared using an unpaired t-test, mean ± SEM, (*p < 0.05, **p < 0.01).

Discussion

This is the first study to show that chronic administration of a subconvulsive dose of NMDA (0.25 mg/kg i.p. daily, for 21 days) to rats selectively decreased frontal cortex NR-1 and NR-3 protein levels, while no differences were detected in the levels of NR-2A or 2B subunits or the phosphorylation of NR-1 or NR-2B. In addition, chronic NMDA increased frontal cortex cPLA2 activity, phosphorylation, protein, and mRNA expression, without significantly altering sPLA2 or iPLA2 activities or protein levels. The increased cPLA2 mRNA expression after chronic NMDA was accompanied by increased AP-2 DNA-binding activity and increased protein levels of its AP-2α and AP-2β subunits. There was no significant change in the AP-2γ protein level or in the activity of other transcription factors known to regulate cPLA2 transcription (NF-κB, GRE, AP-1, and PEA 3).

Excitotoxic stimulation of glutamate receptors by NMDA or glutamate decreases NR-1 expression (Gascon et al. 2005). In vitro studies indicate that the NR-3A subunit co-assembles with other NR subunits (NR-1, NR-2A, or NR-2B) to form NMDA receptors with decreased activity and calcium influx (Ciabarra et al. 1995; Sucher et al. 1995), while mice lacking NR-3A subunits have increased NMDA receptor activity (Das et al. 1998). Thus, chronic administration of NMDA decreased expression of frontal cortex NR-3A, and thereby may have increased NMDA receptor activity in frontal cortex. Consistent with Gascon et al. (2005), the decreased in NR-1 subunit after chronic NMDA administration might be due to neuroadaptive changes because of over-activation of NMDA receptors. One Limitation of using the commercially available antibody for NR-1 was that it recognizes four different splice variant proteins of the NR-1 subunit, and it is not clear which NR-1 splice variant protein levels were decreased in the current study. Further studies are needed to examine individual NR-1 subunit levels in response to chronic NMDA exposure.

NMDA receptor activation enhances cPLA2 activity and the release of arachidonic acid from the sn-2 position of choline and inositol glycerophospholipids (Sanfeliu et al. 1990; Farooqui et al. 1997; Weichel et al. 1999). The increased release may be due to the activation of PLA2 or of phospholipase C yielding 1,2 diacylglycerol, which is further catalyzed by diacylglycerol lipase to liberate arachidonic acid and monoacylglycerol (Piomelli and Greengard 1990). Chronic NMDA selectively increased cPLA2 activity, protein and mRNA levels in rat frontal cortex. The increased cPLA2 activity was accompanied by an increased protein level and phosphorylation at serine 505. NMDA can activate MAPK in cultured rat hippocampal neurons (Kurino et al. 1995), which is known to phosphorylate cPLA2 at serine 505 (Borsch-Haubold et al. 1998; Hefner et al. 2000), possibly explaining our observation.

The increased cPLA2 protein level was associated with an increase in cPLA2 mRNA and in the binding activity of its AP-2 transcription factor (Morri et al. 1994), along with increased protein levels of its AP-2α and AP-2β subunits. AP-2 is important for neurodevelopment (Moser et al. 1997) and monoaminergic activity (Damberg et al. 2001), and AP-2α promotes apoptosis by down-regulating Bcl-2 via a bax/cytochrome c/Apaf1/caspase 9 dependent mitochondrial pathway (Wajapeyee et al. 2006). It is not clear from our study if the increase in AP-2 subunit expression occurred pre- or post-transcriptionally. Further studies are necessary to understand how chronic NMDA alters AP-2 expression.

Chronic NMDA administration selectively decreased rat frontal cortex NR-1 and NR-3A expression, possibly because of over-activation of NMDA receptor function. Decreased NR-1 and NR-3A receptor subunit levels also have been reported in post-mortem brain samples of bipolar disorder patients (Vawter et al. 2000; Scarr et al. 2003; Mueller and Meador-Woodruff 2004; Nudmamud-Thanoi and Reynolds 2004). Activation of NMDA receptors stimulates several signaling enzymes, including cPLA2 (Weichel et al. 1999; Basselin et al. 2005). Thus, over-activation of NMDA receptors may be related to bipolar disorder if, as suggested from the effects of antimanic drugs, an increase in the arachidonic acid cascade is implicated in its pathogenesis (Rapoport and Bosetti 2002).

Two agents that are effective against bipolar mania, lithium and carbamazepine, are reported to decrease cPLA2 activity, protein and mRNA expression in brain after chronic administration to rats, while not affecting brain expression of iPLA2 or sPLA2 (Rintala et al. 1999; Ghelardoni et al. 2004); and to decrease DNA-binding activity of AP-2, a transcription factor for the cPLA2 gene (Rao et al. 2005, 2007a). This has suggested that symptoms of bipolar mania reflect a hyperactive arachidonic acid cascade, due perhaps to increased activity of arachidonic acid selective cPLA2 (Rintala et al. 1999). Our findings in this paper, that chronic NMDA up-regulated brain cPLA2 but not iPLA2 or sPLA2 expression, and up-regulated expression of the AP-2 that regulates cPLA2 transcription, suggest that the rat given NMDA chronically may be a potential model of disturbed arachidonic acid signaling in bipolar mania, and used to screen for other potentially effective drugs. In this regard, chronic lithium in rats has been reported to block increased arachidonic acid brain signaling following a subconvulsive dose of NMDA, by down-regulating cPLA2 transcription but also possibly by altering phosphorylation of some of NR subunits to inactivate NMDA receptors (Nonaka et al. 1998; Ma and Zhang 2003; Basselin et al. 2005).

In a similar vein, dietary n-3 polyunsaturated fatty acid deprivation for 15 weeks post-weaning in rats was reported to increase scores on tests of aggression and depression, symptoms found in bipolar disorder (DeMar et al. 2006). This dietary deprivation also increased cPLA2 activity, protein and mRNA in the frontal cortex (Rao et al. 2007b), and thus may be another model of an up-regulated arachidonic acid cascade for bipolar disorder (Rapoport and Bosetti 2002). It would be of interest to apply tests of aggression and depression in rats chronically administered NMDA to compare the results with the n-3 polyunsaturated fatty acid deprivation regimen. Clinical studies also support the hypothesis that arachidonic acid signaling is altered in patients with bipolar disorder. Genetic studies have suggested that alterations in the sPLA2 gene increase the risk of bipolar disorder (Jacobsen et al. 1999; Dawson et al. 2000) and patients with bipolar disorder have increased serum PLA2 activity (Noponen et al. 1993; Hallett et al. 2005). Increased phospholipid hydrolysis has been reported in patients with affective disorders (Lieb et al. 1983; Hibbeln et al. 1989; Sublette et al. 2004), including elevated PGE2 levels in plasma (Lieb et al. 1983), cerebrospinal fluid (Linnoila et al. 1983) and saliva (Nishino et al. 1989), while a post-mortem study has reported decreased cortical cytosolic prostaglandin E2 synthase in patients with bipolar disorder (Maida et al. 2006).

In conclusion, chronic subconvulsive NMDA administration to rats led to several neuroplastic changes in brain NMDA receptor subunits as well as increased cPLA2 activity and expression, possibly through the activation of the AP-2 transcription factor. Acute NMDA administration did not produce such changes. Increased NMDA receptor function and increased cPLA2 activity and expression may be related to the adverse effects of NMDA signaling on brain function and plasticity, specifically in relation to bipolar disorder.

Acknowledgements

This work was entirely supported by the Intramural Research Program of the National Institute on Aging, National Institutes of Health. We thank the National Cancer Institute (NCI), Center for Cancer Research (CCR) Fellows Editorial Board, for proofreading the manuscript.

Ancillary