The Caveolin Scaffolding Domain Modifies 2-Amino-3-hydroxy-5-methyl-4-isoxazole Propionate Receptor Binding Properties by Inhibiting Phospholipase A2 Activity*

Activation of the enzyme phospholipase (PLA 2) has been proposed to be part of the molecular mechanism involved in the alteration of 2-amino-3-hydroxy-5-methyl-4-isoxazole propionate (AMPA) glutamate receptor responsiveness during long term changes in synaptic plasticity (long term potentiation). This study assesses the effect of the caveolin-1 scaffolding domain (CSD) on the activity of the regulatory enzyme PLA2. Caveolin-1 is a 22-kDa cholesterol-binding membrane protein known to inhibit the activity of most of its interacting partners. Our results show that the calcium-dependent cytosolic form of PLA2 (cPLA2) and caveolin-1 co-localized in mouse primary hippocampal neuron cultures and that they were co-immunoprecipitated from mouse hippocampal homogenates. A peptide corresponding to the scaffolding domain of caveolin-1 (Cav-(82-101)) dramatically inhibited cPLA2 activity in purified hippocampal synaptoneurosomes. Activation of endogenous PLA2 activity with KCl or melittin increased the binding of [3H]AMPA to its receptor. This effect was almost completely abolished by the addition of the CSD peptide to these preparations. Moreover, we demonstrated that the inhibitory action of the CSD peptide on AMPA receptor binding properties is specific (because a scrambled version of this peptide failed to have any effect) and that it is mediated by an inhibition of PLA2 enzymatic activity (because the CSD peptide failed to have an effect in membrane preparations lacking endogenous PLA2 activity). These results raised the possibility that caveolin-1, via the inhibition of cPLA2 enzymatic activity, may interfere with synaptic facilitation and long term potentiation formation in the hippocampus.

Phospholipase (PLA 2 ) 1 belongs to a superfamily of enzymes that play a central role in the regulation of arachidonic acid (AA) release from membrane phospholipids and catalyze the production of various metabolites. PLA 2 activity has been postulated to play an important role in key metabolic pathways. In addition to its involvement in signal transduction, membrane repair, neurodegeneration, and apoptosis (1), there is a growing body of evidence suggesting a role in the modulation of neurotransmitter release and long term potentiation (LTP) (2)(3)(4). Changes in synaptic function observed with LTP are thought to be the result of modifications of postsynaptic currents mediated by the 2-amino-3-hydroxy-5-methyl-4-isoxazole propionate (AMPA) subtype of glutamate receptors (5,6). PLA 2 activity was shown to be part of the molecular mechanisms regulating AMPA receptor function during long term changes in synaptic operation (7)(8)(9). It has been proposed that during high frequency stimulation a large entry of calcium mediated by N-methyl-D-aspartate (NMDA) glutamate receptor activation might cause an increase in PLA 2 activity. The enzyme would augment AMPA receptor affinity by changing the lipid environment of AMPA receptors, thereby producing LTP (10). Recently, it has been reported that detergent-insoluble caveolin-rich membrane domains are markedly enriched in AMPA-type glutamate receptors (11), AA and PLA 2 (12,13), suggesting a possible role for PLA 2 in caveolae-dependent cellular functions.
Caveolae have been implicated in the sequestration of many signaling molecules (14). Caveolin is thought to function as a scaffolding protein within these membrane microdomains where it interacts with several signaling proteins (14,25,26). A short (20 amino acids residues 82-101) cytosolic N-terminal region of caveolin, termed the caveolin scaffolding domain (CSD), is involved in the formation of caveolin oligomers and mediates the interaction with signaling molecules, which generally results in the inactivation of signaling (27). Most caveolin-interacting proteins identified so far contain a caveolinbinding motif located within their enzymatically active catalytic domain (28). This caveolin-binding motif is also present within the catalytic domain of the cytosolic form of PLA 2 (cPLA 2 ); yet, virtually nothing is known about the potential relationship of caveolin with cPLA 2 .
In this article, we report evidence demonstrating the presence of the cytosolic (85-kDa) PLA 2 in caveolin-enriched mem-brane fractions isolated from hippocampal preparations. We also show that a CSD peptide can regulate the enzymatic activity of cPLA 2 in these preparations. Finally, we explored the effect of the CSD fragment on PLA 2 -mediated modulation of AMPA receptor binding properties.

EXPERIMENTAL PROCEDURES
Animals, Reagents, and Antibodies-Adult male, 3-month-old CD-1 mice were obtained from Charles River (St-Constant, Quebec, Canada). The anti-caveolin-1 polyclonal antibody (pAb N-20) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-mouse cPLA 2 polyclonal antibody was from Oxford Biomedical Research (Oxford, MI), and the monoclonal anti-glutamate receptor (GluR2) antibody was from Chemicon International Inc. (Temecula, CA). Goat anti-rabbit and sheep anti-mouse biotinylated antibodies were from Calbiochem (San Diego, CA). Texas Red dye-conjugated donkey anti-rabbit IgG and fluorescein isothiocyanate-conjugated donkey anti-mouse IgG were purchased from Jackson ImmunoResearch Laboratories (West Grove, PA). Cell culture supplies were purchased from Invitrogen (Burlington, Ontario). Peptides, corresponding to amino acids 82-101 of the scaffolding domain of caveolin-1 (Cav-(82-101)), were synthesized, purified, and analyzed by reversed-phase high pressure liquid chromatography and mass spectometry by the W. M. Keck Biotechnology Resource Laboratory (Yale University School of Medicine, New Haven, CT). A scrambled version of the caveolin-1 peptide (CSD-X) was also used as a negative control. Recombinant human cPLA 2 was obtained from Roger L. Williams  Isolation of Caveolae-enriched Fractions-Caveolae-enriched fractions were separated by using ultracentrifugation with a discontinuous sucrose gradient system as described previously (29). Briefly, hippocampi from mice were homogenized (ϳ20 strokes in a Dounce homogenizer) into sodium carbonate buffer (500 mM, pH 11) and sonicated (3 ϫ 20-s bursts). The resulting suspension (1.5 mg of protein in 2 ml) was brought to 42.5% sucrose (w/v) by adding 2 ml of carbonate buffer containing 85% sucrose and placed at the bottom of a 12-ml ultracentrifuge tube. A discontinuous gradient was formed above the 42.5% bed by adding 6 and 2 ml, respectively, of 30 and 5% sucrose solutions prepared in carbonate buffer. After centrifugation for 18 h at 35,000 rpm (4°C), 12 ϫ 1-ml fractions were collected starting at the top of each gradient. An equal volume of each fraction was used to precipitate proteins (20% trichloroacetic acid, 1 h) and analyzed by SDS-PAGE. The protein concentration of each fraction was estimated using a BCA protein assay (Pierce) kit.
Synaptoneurosome Preparation-Synaptoneurosomes were prepared according to the method of Hollingsworth et al. (30). Briefly, mice were killed, and the brains were rapidly removed and chilled in ice-cold Krebs-Ringer buffer (124 mM NaCl, 3 mM KCl, 1.25 mM KH 2 PO 4 , 2.5 mM MgSO 4 , 3 mM CaCl 2 , 26 mM NaHCO 3 , 10 mM glucose) pregassed with a mixture of O 2 /CO 2 (95:5). The hippocampus was cut into small pieces and then homogenized with a Dounce homogenizer in Krebs-Ringer buffer. The homogenate was filtered and then centrifuged at 2000ϫ g for 15 min. The pellet was resuspended in freshly oxygenated Krebs-Ringer buffer and recentrifuged at 2000 ϫ g for 15 min. The final pellet was resuspended and immediately used for binding assays, PLA 2 activity measurement, or Western blot analysis.
Co-immunoprecipitation of cPLA 2 with Caveolin-1-Mice hippocampi were homogenized (ϳ20 strokes with a Dounce homogenizer) in octyl glucoside lysing buffer (60 mM n-octyl ␤-glucopyranoside, 50 mM Tris-HCl, pH 7.4, 125 mM NaCl, 1 mM EDTA, 1 mM EGTA) and sonicated three times (10-s bursts). The homogenate was rotated (4°C) for 60 min and centrifuged at 14,000 rpm for 10 min, and the supernatant (1 g/l protein in octyl glucoside buffer) was used for the immunoprecipitation. Cell lysates were precleared with protein G-Sepharose for 45 min at 4°C. Cleared lysates were then incubated with 4 g of caveolin-1 or cPLA 2 antibodies (rotation at 4°C for 2 h). Cell lysates were then incubated with protein G-Sepharose for 1 h at 4°C. Protein G beads were then extensively washed with octyl glucoside buffer. Proteins were eluted from beads, resolved by SDS-PAGE, and subjected to Western blot analysis.
Cell Culture-Primary hippocampal neuronal cultures were prepared from 1-day-old CD-1 mice, essentially as described previously for embryonic rats (31). Briefly, brains were removed, and the hippocampus was dissected and placed in Hanks' balanced salt solution containing 15 mM HEPES. The tissue was incubated with 0.25% trypsin for 10 min at 37°C and then rinsed several times in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, penicillin/streptomycin, and fungizone. Neurons were seeded into 16-mm 2 culture wells containing poly-D-lysine-coated glass coverslips. After 24 h, the medium was changed to serum-free and supplemented with neuronal growth supplement B-27. The neurons were used for immunofluorescence assays 7 days after plating.
Immunofluorescence-Neurons were rinsed briefly in phosphatebuffered saline and fixed in 4% paraformaldehyde for 30 min. The fixed cells were permeabilized and blocked in phosphate-buffered saline containing 25% goat serum, 1% bovine serum albumin, and 1% SDS for 1 h at room temperature. After 1 h of incubation (37°C) with specific antibody, the cells were washed and incubated (1 h, 37°C) with fluorescein isothiocyanate-conjugated donkey anti-mouse IgG (1:100 dilution) for anti-GluR2 monoclonal antibody and Texas Red dye-conjugated donkey anti-rabbit IgG (1:100) for anti-caveolin-1 and anti-cPLA 2 polyclonal antibodies.
PLA 2 Activity Measurement-Determination of phospholipase activity using [ 3 H]arachidonic acid was performed as described by Verity et al. (32). Briefly, synaptoneurosomes were labeled for 1 h in a Krebs-Ringer buffer containing 200 nM [ 3 H]arachidonate and 2 mM CaCl 2 . After [ 3 H] incorporation, synaptoneurosomes were washed three times and resuspended in 500 l of Krebs buffer, with or without 75 g/ml melittin. Following incubation at 33°C for 45 min and centrifugation, supernatants were removed. The pellets were resuspended and subjected to radioactivity measurement. All assays were performed in triplicate.
[ 3 H]AMPA Binding Assays-[ 3 H]AMPA (specific activity of 60 Ci/ mmol; PerkinElmer Life Sciences) binding assays were performed as described previously by Massicotte et al. (33). Briefly, synaptoneurosomes were preincubated in the presence or absence of different concentrations of the CSD peptide (fragment 82-101) or CSD-X for 15 min at room temperature. After preincubation, 20 g/ml melittin or 50 mM KCl were added, and synaptoneurosomes were incubated at 33°C for 45 min. Synaptoneurosomes were then washed twice with a Tris acetate solution (100 mM, pH 7), and [ 3 H]AMPA was added, diluted in a Tris acetate solution containing 100 mM KSCN. The nonspecific binding was defined as that measured in the presence of 0.2 mM quisqualate. After a 45-min incubation on ice, the preparations were centrifuged (15 min, 14,000 rpm), and the supernatants were removed. The pellets were resuspended in a 0.2 M NaOH solution and subjected to radioactivity measurement. All assays were performed in triplicate.

cPLA 2 Localization in Caveolin-rich Microdomains and
Association with Caveolin-1-Homogenates from hippocampal tissue were fractionated by ultracentrifugation using a discontinuous sucrose gradient system that was previously shown to resolve caveolae-enriched subcellular fractions (29). As shown in Fig. 1A, almost all the endogenously expressed caveolin-1 (Cav-1) is recovered in the caveolae fractions 2, 3, and 4, at the 5-30% sucrose interface. By contrast, most of the total cellular protein is distributed in the denser sucrose fractions (fractions 8 -12) (not shown). Importantly, a fraction of the total endogenous cPLA 2 protein (with the expected molecular mass of 110 kDa (34)) was recovered in caveolae fractions (Fig. 1A), demonstrating the presence of the cPLA 2 protein in caveolin-enriched fractions from mice hippocampi.
To determine whether caveolin-1 forms a stable complex with cPLA 2 , hippocampal homogenates were subjected to immunoprecipitation using antibodies directed against the Nterminal domain of caveolin-1 (residues 2-21) or cPLA 2 protein. The immunoprecipitates were separated by SDS-PAGE and subjected to immunoblot analysis, with antibodies directed against either caveolin-1 or cPLA 2 . As shown in Fig. 1B, caveolin-1 protein was found in immunoprecipitates of cPLA 2 , and, conversely, a small amount of cPLA 2 protein was detected in immunoprecipitates of caveolin-1. To further confirm these biochemical findings, the localization of cPLA 2 and caveolin-1 in cultured hippocampal neurons was assessed. Although cPLA 2 and caveolin-1 were located over the entire neuronal cell surface, both proteins localized more substantially to the cell body, and in many cells, high levels of expression were also focally enriched on growth cones (Fig. 2). These results strongly support an association between cPLA 2 and caveolin-1.
Inhibition of cPLA 2 Activity by the CSD Peptide-Caveolin-1 has been shown to down-regulate the activity of several signaling molecules (27). This observation, together with the results described above, led us to examine the possible reduction of cPLA 2 activity in synaptoneurosomes treated with the CSD peptide. Synaptoneurosomes, obtained by low speed centrifugations and filtration of brain homogenates, are a suspension of synaptic plasma membranes that contains synaptosomes with attached postsynaptic densities. The presence of both caveolin-1 and cPLA 2 has been confirmed by Western blots carried out on hippocampal synaptoneurosome preparations (Fig. 1A). Incubation of synaptoneurosomes for 1 h in the presence of [ 3 H]arachidonate resulted in a significant lipid incorporation of AA (123 Ϯ 15 ϫ 10 3 dpm/mg of protein) and a basal release corresponding to approximately 2% of total [ 3 H]arachidonate incorporation. Purified CSD peptide produced a concentrationdependent inhibition of [ 3 H]arachidonate release on PLA 2 stimulation by melittin (Fig. 3). A highly significant effect of CSD peptide treatment was observed at 5 nM, with a 38.9% reduction of arachidonate release when compared with nontreated controls. A 70.9% reduction was reached with a concentration of only 50 nM of the purified CSD peptide.
To confirm specifically that the inhibition of AA release observed is cPLA 2 -dependent, the effect of the CSD peptide was tested on synaptoneurosomes pretreated with the cPLA 2 inhibitor, AACOCF 3 . Fig. 4A shows that the CSD peptide completely lost its effect on melittin-stimulated release of AA in the presence of the cPLA 2 inhibitor. To further identify specifically the cPLA 2 as a target of the CSD peptide, we also investigated the inhibitory action of the CSD peptide on recombinant cPLA 2 activity. As shown in Fig. 4B, the CSD peptide completely FIG. 2. Immunocolocalization of cPLA 2 and caveolin-1 in mouse primary hippocampal neurons in culture. Hippocampal neurons were grown for 7 days, and then antibodies specific to cPLA 2 (top panel) and caveolin-1 (bottom panel) were used to localize the proteins. Note that both cPLA 2 and caveolin-1 are most highly expressed on the soma or cell body and are also focally enriched on growth cones (arrowheads). Experiments using negative controls that omitted primary antibodies were also performed and yielded the expected results (data not shown). Scale bar ϭ 10 m. abolished the effect of recombinant cPLA 2 on AA release, whereas the CSD-X failed to show any effect on cPLA 2 activity.
Inhibition of the PLA 2 -induced Modulation of AMPA Binding by the CSD Peptide-Because several studies indicate that calcium-dependent PLA 2 participates in selective changes in AMPA receptor properties (33,35), the previous results prompted us to examine the possibility that the inhibition of cPLA 2 activity could interfere with AMPA receptor binding. Double immunofluorescence was first used to examine whether caveolin-1 and AMPA receptors are co-localized at the cellular level. Fig. 5 shows double labeling of cultured hippocampal neurons with a mouse monoclonal antibody directed against the GluR2 subunit of glutamate AMPA receptor and a rabbit polyclonal antipeptide antibody generated against caveolin-1. Note that GluR2 immunostaining appears primarily at the cell body and as puncta localized to areas of cellular outgrowth. Although caveolin-1 exhibited a more diffuse fluorescence on projections, the merged image shows a general co-localization of GluR2 and caveolin-1 in hippocampal neurons.
It has been previously shown that potassium-induced depolarization of synaptoneurosomes increased [ 3 H]AMPA binding to membrane fractions (36,37). It has also been reported that treatment of synaptoneurosomes with melittin, a potent activator of endogenous phospholipases, increases [ 3 H]AMPA binding (8). Fig. 6 illustrates the effect of melittin and KCl in the presence of different concentrations of CSD peptide. Preincubation of synaptoneurosomes with CSD peptide produced a marked reduction of the melittin-and KCl-induced increase in [ 3 H]AMPA binding to membrane fractions. However, the inactive CSD-X peptide at the highest concentration (50 nM) did not affect AMPA binding in both situations (not shown).
As shown in Fig. 7, the CSD peptide also reduced the basal [ 3 H]AMPA binding to its receptor. To eliminate the possibility that the CSD peptide directly interacts with AMPA receptors, [ 3 H]AMPA binding was evaluated on purified membranes, which, unlike synaptoneurosome preparations, do not contain

FIG. 5. Co-immunolocalization of caveolin-1 and GluR2 subunit of glutamate receptor in mouse primary hippocampal neurons.
Hippocampal neurons were grown for 7 days and then doubly immunostained with either GluR2 monoclonal or caveolin-1 polyclonal antibodies. GluR2 was detected with fluorescein isothiocyanate-conjugated donkey anti-mouse IgG (green) and caveolin-1 with Texas Red dye-conjugated donkey anti-rabbit IgG (red). The merge is an overlay of the GluR2 (green) and caveolin-1 (red) signals. Note the significant areas of co-localization (yellow) between GluR2 and caveolin-1. Negative controls omitting a given primary antibody were also performed and yielded the expected results (data not shown). Scale bar ϭ 10 m. PLA 2 activity (not shown). Membrane preparations were obtained by a low speed centrifugation of hippocampal homogenates (3500 rpm) followed by two ultracentrifugations (18,000 rpm) of the supernatants. Our results showed no CSD-induced decrease of AMPA receptor binding in these membrane preparations (Fig. 7A). Similarly, the inhibitory effect of CSD peptide was completely abolished when the binding experiments with synaptoneurosomes were carried out at 0°C instead of 33°C (Fig. 7B). In these low temperature conditions, no phospholipase activity is detectable (not shown), indicating again that the CSD peptide modulation of [ 3 H]AMPA binding on synaptoneurosomes reflects an inhibition of PLA 2 enzymatic activity by the peptide rather than a direct action of the compound on AMPA receptors. DISCUSSION In the present study, we confirm the presence of both caveolin-1 and cPLA 2 in mouse hippocampal neurons. Moreover, subcellular fractionation experiments revealed that a significant fraction of cPLA 2 is found in caveolin-rich membrane microdomains from hippocampal preparations. Co-immunoprecipitation experiments also indicated that caveolin-1 and cPLA 2 associate with each other, suggesting that caveolin-1 may functionally interfere with cPLA 2 enzymatic activity. In support of this assumption, we found that the peptide corresponding to the CSD peptide dramatically inhibited cPLA 2mediated release of arachidonic acid. In addition, we showed that the inhibition of cPLA 2 activity by the CSD peptide modified [ 3 H]AMPA binding in synaptoneurosomes. Taken together, these results suggest that the inhibition of cPLA 2 by caveolin-1 may be an important and previously unrecognized mechanism for modulating neuronal AMPA receptor binding properties.
Caveolin-1 is an integral membrane protein, and one of its proposed roles is to regulate the activity of signaling proteins that reside in caveolae. Numerous signaling molecules such as Ha-RAS, c-Src, and eNOS (28) have been shown to be functionally associated with caveolin-1. It was recently reported that caveolae are enriched in AA (12), the release of which is regulated by PLA 2 . There are four major types of PLA 2 , including cPLA 2 (cytosolic), sPLA 2 (secretory), iPLA 2 (Ca 2ϩ -independent) and platelet-activating factor (PAF) acetylhydrolase (38). Although there are a number of types of PLA 2 in the brain, the 85-kDa cPLA 2 (type IV) is unique because it selectively releases AA from phospholipids (39). In resting cells, cPLA 2 is localized to the cytoplasm, but in response to increases in cytosolic free calcium, the enzyme translocates to the plasma membrane and the nuclear envelope (40). Based on our findings, cPLA 2 appears to be, in part, localized in the caveolae. However, final confirmation of this localization must await detailed electron microscopy studies. Moreover, although our data do not establish a direct protein-protein interaction between caveolin-1 and cPLA 2 , co-immunoprecipitation experiments indicate that caveolin-1 and cPLA 2 proteins interact somehow with one another. In this regard, the large number of proteins known to associate with caveolin-1 might explain the small amount of caveolin-1 detectable in cPLA 2 immunoprecipitation (Fig. 1).
Among the proteins that functionally associate with caveolin-1, a common denominator seems to be the presence of a motif that interacts with the scaffolding domain of caveolin-1 (CSD) (amino acids 82-101), which is a juxtamembrane region in caveolin-1 containing numerous aromatic residues. In our study, a peptide derived from the scaffolding domain of caveolin-1 was used to examine the putative functional relationship between caveolin-1 and the activity of the regulatory enzyme cPLA 2 . Synaptoneurosomes, which are pinched-off nerve terminals associated with released post-synaptic structures, are useful tools to study mechanisms of transmitter release, regulation of transmitter receptors and second messenger pathways. We found that the CSD peptide dramatically inhibits melittin-induced AA release from synaptoneurosome preparations. Melittin is a strong activator of endogenous PLA 2 s. With the use of the cPLA 2 inhibitor, AACOCF 3 , we showed that the major proportion of melittin-induced AA release is cPLA 2 -dependent. The absence of any CSD peptide effect on synaptoneurosomes pretreated with the cPLA 2 inhibitor and the complete inhibition of the recombinant cPLA 2 activity by the CSD peptide specifically identified cPLA 2 as a target of the CSD peptide. Moreover, the scrambled version of the CSD peptide (CSD-X) did not affect melittin-induced AA release, showing the specificity of the observed effect of the CSD fragment on endogenous cPLA 2 . It should be noted here that in synaptoneurosome preparations, 50% of the formed vesicles are likely to be inside-out, the remaining being right-side-out. This explains how the CSD peptide we used, which is a cell-impermeable peptide, can interact with intracellular cPLA 2 .
How might caveolin-1 inhibit the activity of cPLA 2 ? One possible physiological mechanism for the CSD inhibition of cPLA 2 activity is that, following the enzyme activation, cPLA 2 translocation to the cell membrane (40) may promote its physical interaction with caveolin-1. Another hypothesis is that cPLA 2 clustering with other signaling molecules inside caveolae may indirectly inactivate cPLA 2 activity. For instance, cPLA 2 can be stimulated by protein kinases such as protein kinase C (41) and mitogen-activated protein kinase (42), which are known to be inhibited by caveolin-1 (43).
Although the exact mechanism for the putative inhibition of cPLA 2 activity by caveolin-1 is still unknown, the modulatory effect observed may have an important homeostatic role in brain cells. For example, it has recently become apparent that activation of the calcium-dependent PLA 2 could be part of the molecular mechanisms involved in alterations of AMPA receptor properties during long term changes in synaptic operation (LTP) (7)(8)(9). In accordance with a recent study providing biochemical evidence for a localization of AMPA-type glutamate receptors to caveolae-like structures (11), our study shows the co-immunolocalization of the GluR2 subunit of AMPA receptors with caveolin-1 in primary hippocampal neuron cultures.
Increased PLA 2 activity has previously been shown to upregulate [ 3 H]AMPA binding to AMPA receptor (44,45). In accordance with these findings, the activation of endogenous PLA 2 activity with KCl or melittin increased AMPA binding to its receptor. This effect was almost completely abolished by the addition of the CSD peptide to synaptoneurosome preparations. It is not excluded that a small fraction of the effect of CSD peptide on AMPA binding might have been mediated by the inhibition of PLA 2 s other than the cPLA 2 . However, as we showed previously with the cPLA 2 inhibitor, the effect of melittin on synaptoneurosomes is mostly cPLA 2 -dependent. This indicates that the inhibition of the cPLA 2 isoform of the enzyme might be principally involved in the reduction of AMPA binding, which is consistent with our experiments showing the highly potent and efficient inhibition of the purified recombinant cPLA 2 by the CSD peptide. We also demonstrated that the inhibitory action of the CSD peptide on AMPA binding depends on its modulation of PLA 2 activity by carrying out experiments in conditions where PLA 2 enzymatic activity is absent. Interestingly, biochemical evidence has shown that [ 3 H]AMPA binding is augmented after LTP induction (46 -48). Moreover, several reports suggest that different compounds that inhibit PLA 2 activity interfere with LTP (4,49,50). Thus, we could hypothesize that an up-regulated physiological level of caveolin-1 expression in the brain, such as that seen in aging (51)(52)(53) or in Alzheimer disease (53), might negatively interfere with synaptic function.
In conclusion, we show for the first time the regulation of cPLA 2 activity and the resulting modulation of AMPA receptor binding properties by the scaffolding domain of caveolin-1. These results raise the possibility that caveolin-1 may interfere with phospholipid-derived metabolite production, synaptic facilitation, and LTP formation. In this regard, future investigations into this area should include experiments dedicated to evaluating the potential involvement of the caveolin scaffolding domain in regulating changes of AMPA receptor properties in synaptic plasticity and LTP formation in vivo.