Pituitary Adenylate Cyclase-activating Polypeptide (PACAP(1–38)) Enhances N -Methyl- D -aspartate Receptor Function and Brain-derived Neurotrophic Factor Expression via RACK1*

We recently identified a novel mechanism for modulation of the phosphorylation state and function of the N -methyl- D -aspartate (NMDA) receptor via the scaffold- ing protein RACK1. We found that RACK1 binds both the NR2B subunit of the NMDA receptor and the nonrecep-tor protein-tyrosine kinase, Fyn. RACK1 inhibits Fyn phosphorylation of NR2B and decreases NMDA recep-tor-mediated currents in CA1 hippocampal slices (Yaka, R., Thornton, C., Vagts, A. J., Phamluong, K., Bonci, A., and Ron, D. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 5710– 5715). Here, we identified the signaling cascade by which RACK1 is released from the NMDA receptor complex and identified the consequences of the dissociation. We found that activation of the cAMP/protein kinase A pathway in hippocampal slices induced the release of RACK1 from NR2B and Fyn. This resulted in the induction of NR2B phosphorylation and the enhancement of NMDA receptor-mediated activity via Fyn. We identified the neuropeptide, pituitary adenylate cyclase activating polypeptide (PACAP(1–38)), as a ligand that induced phosphorylation of NR2B in turn results in an increase in the activity of the NMDA receptor channel. As a result of RACK1 nuclear translocation, BDNF expression is enhanced.

The ionotropic glutamate receptor subtype, N-methyl-D-aspartate (NMDA), 1 plays an essential role in neuronal develop-ment, excitotoxicity, synaptic plasticity, and learning and memory (2). The ligand-gated NMDA receptor channel is a heteromer comprised of NR1 and at least one of four NR2 subunits (A-D) (3). The cytoplasmic tail of the NR2B subunit is phosphorylated on tyrosine residues (4) and is the most abundant tyrosine-phosphorylated protein in the postsynaptic density (5). Phosphorylation of NMDA receptor subunits regulates the activity of the channel (6). Specifically, application of a tyrosine kinase inhibitor causes a progressive decrease in NMDA receptor-mediated currents, and conversely, inhibition of protein-tyrosine phosphatases results in an increase in NMDA receptor-mediated currents (7). Subsequent studies have identified the Src family of protein-tyrosine kinases (PTKs) as enzymes that phosphorylate the NR2 subunits, regulating NMDA receptor activities (6 -8). Hence, modulation of NMDA receptor phosphorylation by Src family protein-tyrosine kinases is likely to play an important role in modulating glutamate-mediated pathways.
In the recent years it has become increasingly apparent that the formation of localized signaling complexes, which include receptors, kinases, phosphatases and their substrates, are highly important for the regulation of signal transduction cascades (9,10). Scaffolding proteins play a major role in the assembly of such signaling complexes (11). Recently, we identified the scaffolding protein RACK1 as a novel regulator of the phosphorylation state and function of the NMDA receptor. We found that RACK1 interacts with both the cytoplasmic tail of the NR2B subunit and Fyn and inhibits the ability of Fyn to phosphorylate the NR2B subunit and consequently inhibits NMDA receptor-mediated excitatory postsynaptic currents (EPSCs) in CA1 hippocampal neurons (1). Peptides that inhibit the interactions between RACK1 and Fyn and RACK1 and NR2B restored the ability of Fyn to phosphorylate NR2B in the presence of RACK1, and enhanced NMDA receptor-mediated EPSCs via activation of the Src family protein-tyrosine kinase Fyn (1). Based on these results we propose a model wherein RACK1 localizes Fyn in close proximity with its substrate the NR2B subunit but prevents the phosphorylation of the subunit until the appropriate signal occurs. This signal should cause the dissociation of RACK1 from the NMDA receptor complex, allowing Fyn to phosphorylate the channel, resulting in enhanced channel activity.
Although it is well established that activation of Src family protein-tyrosine kinases such as Fyn are required for the phosphorylation of NMDA receptor subunits and the regulation of channel activity, the signaling cascades that lead to kinase activation and the subsequent phosphorylation of the NMDA receptor subunits are not fully understood. Here we identify a signaling pathway that induces the release of RACK1 from the NMDA receptor complex and the functional consequences of RACK1 dissociation.

Animals
Fyn Ϫ/Ϫ mice (129SvImJ/C57BL.6J hybrids) were purchased from Jackson Laboratories. Mice were mated in-house with 129/SVJ wild type mice to generate Fyn ϩ/Ϫ mice. Fyn ϩ/Ϫ mice were then mated to generate Fyn Ϫ/Ϫ and Fyn ϩ/ϩ littermate controls, which were used for biochemical and electrophysiological experiments. The genotyping of mice was determined by RT-PCR analysis of products derived from tail mRNA. Only male mice were used in the studies. The age of the individuals ranged from 3 to 5 weeks. For each colony, mice were randomly (with respect to the genotype) housed in groups of five individuals in Plexiglas cages. Food and water were available ad libitum. Animals were kept in conditions of constant temperature (23°C), humidity (50%), and light-dark cycle (light on from 6:00 am to 6:00 pm). 3-4-week-old male Sprague-Dawley rats were purchased from Simonsen and kept under the same conditions.

Immunoprecipitation
The membranal fraction from brain homogenates was precleared by incubation with protein G-agarose (Invitrogen). The samples were centrifuged, and the protein quantity was determined using bicinchoninic (BCA) reagent (Pierce). Immunoprecipitation was performed with 5 g of the appropriate antibodies, with ϳ0.5 mg of protein diluted in 1ϫ immunoprecipitation buffer (1% Triton X-100, 150 mM NaCl, 10 mM Tris HCl, pH 7.4, 1 mM EDTA, 1 mM EGTA, and protease and phosphatase inhibitor mixtures). After overnight incubation at 4°C, protein G-agarose was added, and the mixture was incubated at 4°C for 2 h. The agarose resin was washed extensively and samples resolved by SDS-PAGE on a 10% gel. Membranes were probed with the appropriate antibodies, and immunoreactivity was detected with ECL and processed using the STORM system.

Primary Hippocampal Neurons
Newborn rats at P0 were decapitated, and hippocampi were dissected bilaterally. Cells were dissociated by enzyme digestion with papain followed by brief mechanical trituration and plated on poly-Dlysine-(Sigma) treated chamber slides (Nalgene). Cells were plated (1 ϫ 10 5 cells/chamber) and maintained in Neurobasal medium (Invitrogen) supplemented with B27, penicillin, streptomycin, and Glutamax-1 (Invitrogen) and maintained in culture for 3 weeks.

Immunocytochemistry
After treatment, RACK1 immunofluorescence was performed as described previously (14,15). Briefly, cells were washed in cold phosphatebuffered saline containing 0.3% Triton X-100, fixed in cold methanol, and blocked in phosphate-buffered saline containing 0.3% Triton X-100 and 3% normal goat serum. Immunofluorescence was performed with monoclonal IgM anti-RACK1 (1:100). Staining was detected with secondary antibodies conjugated to Texas Red (1:250). Slides were viewed with a laser scanning confocal microscope (Bio-Rad MRC-1024). Z field images were processed by obtaining the middle Z field sections using NIH Image 1.61 (National Institutes of Health) and Adobe Photoshop (Adobe Systems Inc.).

Nuclear Extraction
Hippocampal slices were homogenized, and nuclear and non-nuclear extracts were prepared using an NE-PER kit in accordance with the manufacturer's protocol (Pierce).

Electrophysiology
Transverse hippocampal slices (400 -450 m) were prepared from 3-5-week-old male Fyn ϩ/ϩ , Fyn Ϫ/Ϫ mice or 3-4-week-old male Sprague-Dawley rats. The slices were allowed to recover for at least 1-2 h in artificial cerebrospinal fluid perfusion medium as described above. After recovery, slices were submerged and superfused continuously with artificial cerebrospinal fluid at 25°C. Field excitatory postsynaptic potentials (fEPSPs) were recorded from stratum-radiatum of CA1 region with glass microelectrodes filled with 2 M NaCl. To obtain NMDA receptor-mediated fEPSPs, 100 M picrotoxin and 10 M NBQX were added to the bath solution to block ␥-aminobutyric acid type A receptorand (AMPA) ␣-amino, 3-hydroxy, 5-methyl, 4-isoxazolepropionic acid receptor-mediated IPSPs and EPSPs, respectively. To evoke fEPSPs, Schaffer collateral/commissural afferents were stimulated with 0.1-Hz pulses using steel bipolar microelectrodes at intensities adjusted to produce an evoked response that was 50% of the maximum-recorded fEPSP for each recording. The maximal rate of change in fEPSP within a time window selected around the rising phase was calculated. Data were collected using an Axopatch-1D amplifier (Axon instruments, Foster City, CA), filtered at 2 kHz, and digitized at 5-10 kHz. To obtain AMPA receptor-mediated fEPSPs, 100 M picrotoxin and 50 M D-AP5 were added to the bath solution to block ␥-aminobutyric acid type A receptor-and NMDA receptor-mediated IPSPs and EPSPs, respectively.

RT-PCR
Total RNAs were isolated using Trizol reagent (Invitrogen) and reverse transcribed using a Reverse Transcription System kit (Promega) at 42°C for 30 min. BDNF and control GPDH expression were analyzed by PCR with temperature cycling parameters consisting of initial denaturation at 94°C for 2 min followed by 35 cycles of denaturation at 94°C for 30 s, annealing at 52°C for 30 s, extension at 72°C for 1 min, and a final incubation at 72°C for 7 min. The BDNF primers were based on a coding frame of the rat BDNF gene: upstream, 5Ј-TTG AGC ACG TGA TCG AAG AGC-3Ј, and downstream, 5Ј-GTT CGG CAT TGC GAG TTC CAG-3Ј. The GPDH primers were based on the rat GPDH gene: upstream, 5Ј-TGA AGG TCG GTG TCA ACG GAT TTG GC-3Ј, and downstream, 5Ј-CAT GTA GGC CAT GAG GTC CAC CAC-3Ј. The PCR products were separated by 1.8% agarose gel and photographed by Eagle Eye II (Stratagene).

Data Analysis
NR2 Phosphorylation-Scanned images of the bands corresponding to NMDA receptor subunits were quantitatively analyzed by densitometry with the NIH Image 1.61 program providing peak areas; values were expressed as a ratio of phosphorylated to total amount of NMDA receptor subunits. At the conditions used, all blots were within the linear range. Statistical analysis was performed using Student's t test for significant differences.
BDNF mRNA Expression-Scanned images of the RT-PCR products were analyzed quantitatively by densitometry with the NIH Image 1.61 program. The intensities of BDNF and the corresponding GPDH controls were measured, and the BDNF:GPDH ratio was calculated. The results are presented as the mean ratio Ϯ S.D. Titration of the GPDH RT-PCR was performed using 5-cycle increments from 20 to 60 cycles. At the conditions used (35 cycles), GPDH amplification was observed to be within the linear range.

Identification of a Signal Transduction Cascade That Induces the Dissociation of RACK1 from the NMDA Receptor
Complex-Several signal transduction cascades such as PKA (13,14) and PKC (15)(16)(17) have been shown to affect the localization and/or function of RACK1. We found previously that the adenylate cyclase activator, forskolin, induces the translocation of RACK1 to the nucleus in glioma and neuroblastoma cell lines (14). We therefore hypothesized that activation of the cAMP signaling cascade which results in the nuclear compartmentalization of RACK1 could release RACK1 from the NMDA receptor complex to enable Fyn to phosphorylate the NR2B subunit. Accordingly, hippocampal slices were treated with vehicle or 10 M forskolin, and the integrity of the complex was determined by immunoprecipitation. Anti-RACK1 and anti-NR2B antibodies were used to immunoprecipitate RACK1 and NR2B, respectively, and Western blot analysis with anti-RACK1, anti-Fyn, and anti-NR2B antibodies was used to detect coimmunoprecipitation of the corresponding proteins. As shown in Fig. 1, in mouse hippocampal slices treated with vehicle, anti-RACK1 coimmunoprecipitated Fyn and NR2B, and anti-NR2B antibodies coimmunoprecipitated Fyn and RACK1. These results are in line with our initial observation of complex formation in rat hippocampal slices (1). However, incubation of hippocampal slices with forskolin, but not its inactive analog dideoxyforskolin, resulted in the dissociation of Fyn and NR2B from RACK1 ( Fig. 1, left panel) and caused Fyn and RACK1 to dissociate from the NR2B subunit ( Fig. 1, right  panel). These results suggest that activation of the cAMP signaling pathway initiates the release of RACK1 from the NMDA receptor complex.
Forskolin and PACAP(1-38) Induce Specific Phosphorylation of NR2B-If forskolin-induced release of RACK1 from the NMDA receptor complex is mediating Fyn phosphorylation of the NR2B subunit, then forskolin should induce the phosphorylation of NR2B in a Fyn-dependent manner. To test the hypothesis, we determined whether forskolin induced the phosphorylation of the NR2B subunit in hippocampal slices from wild type (Fyn ϩ/ϩ ) mice and from mice in which the Fyn gene was deleted (Fyn Ϫ/Ϫ ). Hippocampal slices were treated with vehicle or forskolin, the NR2B subunit was immunoprecipitated, and phosphorylation was determined with anti-phosphotyrosine antibodies. As shown in Fig. 2, forskolin caused a significant increase in the phosphorylation of the NR2B subunit in Fyn ϩ/ϩ mice (Fig. 2, top two panels). Importantly, the tyrosine phosphorylation of the NR2B subunit was not detected in Fyn Ϫ/Ϫ (Fig. 2, middle two panels), suggesting that as predicted, Fyn is the kinase that phosphorylates the NR2B subunit of the NMDA receptor upon cAMP/PKA stimulation. Because RACK1 localizes Fyn specifically to the NR2B subunit but not NR2A (1, and data not shown), we hypothesized that the release of RACK1 from the NMDA receptor complex would not alter the phosphorylation state of the NR2A subunit. Indeed, there was no detectable phosphorylation of the NR2A subunit in hippocampal slices from Fyn ϩ/ϩ mice treated with forskolin (Fig. 2, bottom two panels).
Next, we assessed whether the activation of hippocampal G s -coupled receptors that activate adenylate cyclase would, like forskolin, enhance tyrosine phosphorylation of NR2B. Interestingly, we identified the neuropeptide PACAP(1-38) as a ligand that caused an enhancement of tyrosine phosphorylation of NR2B (Fig. 3a, top two panels) but not NR2A (Fig. 3a, bottom two panels) in rat hippocampal slices. Next we determined which tyrosines were phosphorylated on NR2B upon PACAP(1-38) treatment, using phospho-specific anti-NR2B antibodies raised against phospho-tyrosine residues, previously identified to be phosphorylated by Fyn (18). At concentrations as low as 10 nM, PACAP(1-38) induced an increase in phosphorylation of tyrosines 1336 and 1252 (Fig. 3b) and 1472 (data not shown) on NR2B. We found previously that RACK1 interaction with Fyn and NR2B prevents the ability of Fyn to phosphorylate the subunit (1). Thus, if PACAP-induced phosphorylation of NR2B is indeed mediated via the release of RACK1 from the NMDA receptor complex, then increasing RACK1 intracellular levels should inhibit PACAP(1-38)-induced phosphorylation by the interaction of exogenous RACK1 with Fyn and NR2B. To test the hypothesis, we used the Tat protein transduction method (13,19) to transduce RACK1 into hippocampal neurons (Fig. 3c) and monitored PACAP(1-38)induced NR2B phosphorylation. Preincubation of slices with 1 M Tat-RACK1 inhibited PACAP(1-38)-induced NR2B phosphorylation of tyrosines 1336 and 1252 ( Fig. 3d and data not shown), and there was no significant effect of Tat-RACK1 alone on basal levels of NR2B phosphorylation. In addition, Tat-RACK1⌬N that lacks the binding sites for Fyn and NR2B (1) was inactive (Fig. 3e), suggesting that Tat-RACK1 inhibition of NR2B phosphorylation is specific. Taken together, these results suggest that NR2B phosphorylation is mediated by the release of RACK1 from NMDA receptor, enabling Fyn to phosphorylate the subunit.
The PACAP(1-38) receptors are positively coupled to the cAMP/PKA pathway via G␣ s , and PACAP(1-38) is the most potent endogenous activator of the cAMP signaling pathway (20). However, PACAP(1-38) receptors have also been linked to the activation of the PKC pathway via G␣ q (20). In addition, PKC activation has been shown to increase the function of the NMDA receptor channel via Src (21,22). We therefore also tested the ability of the PKC activator PMA to release RACK1 from Fyn and NR2B and to induce NR2B phosphorylation in rat hippocampal slices. Slices were incubated with vehicle or 500 nM PMA, and the presence of the Fyn, RACK1, and NR2B complex and the phosphorylation state of NR2B were determined. PMA treatment did not result in the dissociation of RACK1 from Fyn and NR2B (Fig. 4a), nor did PMA induce the phosphorylation of NR2B (Fig. 4b), suggesting that activation of PKC is not involved in the release of RACK1 from the NMDA receptor complex which results in NR2B phosphorylation. Taken together, our results suggest that activation of the cAMP/PKA pathway by PACAP(1-38) releases RACK1 from the NMDA receptor complex to induce Fyn phosphorylation of NR2B.

Forskolin and PACAP(1-38) Enhance NMDA Receptor-mediated fEPSPs in Hippocampal Slices of Wild Type but Not
Fyn Ϫ/Ϫ Mice-Tyrosine phosphorylation has been shown previously to increase the activity of the NMDA receptor channel (7,8,23). PACAP(1-38) was found to enhance NMDA receptor function in chick cortical neurons (24), in sympathetic preganglionic neurons (25), and in the CA1 region of the hippocampus (26). However, the molecular mechanism that mediates PACAP induction of NMDA receptor activity is unknown. We proposed that activation of the cAMP/PKA pathway by forskolin or PACAP(1-38) resulting in the phosphorylation of the NR2B by Fyn caused the enhancement of NMDA receptor-mediated activities. To examine this possibility, hippocampal slices from Fyn ϩ/ϩ and Fyn Ϫ/Ϫ mice were treated with 10 M forskolin or 1 nM PACAP(1-38), and NMDA receptor-mediated fEPSPs were recorded from hippocampal CA1 region. A robust increase in fEPSPs was measured in hippocampal slices from Fyn ϩ/ϩ mice when forskolin, but not dideoxyforskolin was bath applied (Fig. 5a). The same increase was observed upon bath application of PACAP(1-38) (Fig. 5b). Importantly, there was no change in fEPSPs in hippocampal slices from Fyn Ϫ/Ϫ mice treated with either forskolin (Fig. 5a) or PACAP(1-38) (Fig.  5b), suggesting that Fyn is required for forskolin-or PACAP(1-38)-induced potentiation of NMDA receptor-mediated fEPSPs. To rule out the possibility that PACAP(1-38)-induced enhancement of NMDA receptor-mediated potentials was cause by enhanced presynaptic glutamate release, AMPA receptor-mediated fEPSPs were recorded in the CA1 region. As shown in Fig.  5b, there was no change in AMPA receptor-mediated fEPSPs after bath application of PACAP(1-38), suggesting that PACAP(1-38)-induced enhancement of NMDA fEPSPs occurred postsynaptically.
Next we verified that PACAP(1-38)-induced enhancement of NMDA receptor function occurred in hippocampal slices from rat. Bath application of 10 nM PACAP(1-38) increased NMDA receptor-mediated fEPSPs in rat slices (Fig. 5c) as was observed in the mouse. The concentration of PACAP(1-38) which produced the enhancement in rat hippocampal slices was 10fold higher than in slices from mice; however, this higher concentration was also required for NR2B phosphorylation in rat slices and therefore is likely to be the result of a species difference in PACAP(1-38) sensitivity. To examine whether PACAP(1-38)-or forskolin-induced enhancement of NMDA receptor-mediated fEPSPs was specific for the NR2B-containing NMDA receptor channels, fEPSPs were recorded in the presence of the NMDA receptor antagonist ifenprodil. At low concentrations, ifenprodil is selective for NR2B-containing receptors (27). Therefore we determined the IC 50 for ifenprodil inhibition on NMDA receptor-mediated fEPSPs in our rat hippocampal slice preparation. We found that 5 M inhibited 50% of NMDA receptor-mediated fEPSPs. Next, we bath applied 5 M ifenprodil, and 20 min later 10 M forskolin or 10 nM PACAP were added to the bath. As predicted, 5 M ifenprodil completely abolished the PACAP(1-38)-or forskolin-induced increase of NMDA receptor-mediated fEPSPs (Fig. 5c,  NR2B then inhibition of Fyn kinase activity should abolish this increase. Therefore, fEPSPs were recorded in the presence or absence of a specific inhibitor for the Src family of proteintyrosine kinases, protein phosphatase 2. As shown in Fig. 5c, PACAP(1-38)-induced enhancement of fEPSPs was abolished when PACAP  was coapplied with 25 nM of protein phosphatase 2. Taken together, these results suggest that activation of the cAMP/PKA pathway results in the dissociation of RACK1 from the NMDA receptor complex, phosphorylation of NR2B subunit of the NMDA receptor by Fyn, and the consequent enhancement of NMDA receptor channel activity in a Fyn-dependent manner.
Forskolin and PACAP  Induce the Translocation of RACK1 to the Nuclear Compartment in Hippocampal Neurons-The intracellular compartmentalization of RACK1 is not fixed, and different stimuli can induce its movement to distinct intracellular compartments (13)(14)(15). Therefore, we hypothesized that the release of RACK1 from the NMDA receptor by activation of the cAMP/PKA pathway would lead to its redistribution to a different site. We have shown previously that forskolin induces the nuclear compartmentalization of RACK1 in glioma and neuroblastoma cell lines (15). To assess whether forskolin could enhance the nuclear compartmentalization of RACK1 in the hippocampus, dissociated hippocampal neurons were treated with vehicle or forskolin, and RACK1 nuclear compartmentalization was determined by immunohistochemistry and confocal microscopy. RACK1 was found in processes and cell bodies but not in nuclei in control cells (Fig. 6a, left  panel). Forskolin treatment resulted in a decrease of RACK1 in the processes and a marked increase in RACK1 staining in the nuclei (Fig. 6a, right panel), suggesting that forskolin induced the redistribution of a portion of RACK1 into the nucleus. We also performed a similar experiment in rat hippocampal slices using subcellular fractionation and isolating nuclear and nonnuclear fractions. Forskolin treatment increased RACK1 in the nuclear fraction and decreased RACK1 in the non-nuclear fraction (Fig. 6b). Next, we hypothesized that if the activation of cAMP/PKA pathway by forskolin induces the translocation of RACK1 to the nucleus, then this effect will also be detected with PACAP(1-38) treatment. Rat hippocampal slices were incubated with PACAP , and the presence of nuclear RACK1 was assessed by subcellular fractionation. As shown in Fig. 6c, the same concentration of PACAP(1-38) (10 nM) which increased NR2B phosphorylation and enhanced NMDA receptor function (Figs. 3 and 5) also increased RACK1 in the nuclear fraction. These results suggest that the dissociation of RACK1 from the NMDA receptor complex may be caused by the translocation of RACK1 to the nuclear compartment induced by forskolin or PACAP .
PACAP  Induces BDNF mRNA Expression via Nuclear RACK1-Finally, we analyzed the functional consequences of PACAP-38-induced RACK1 nuclear compartmentalization. Recently we found that in C6 glioma cells nuclear RACK1 mediates the induction of the immediate early gene c-fos (13). In cerebellar granule cells, PACAP(1-38) induces c-fos expression via the cAMP/PKA pathway (28). PACAP  has been also shown to increase the expression of the brain-derived neurotrophic factor (BDNF) in cortical neurons and in astrocytes (29). We therefore assessed whether PACAP(1-38) induced expression of BDNF in the hippocampus. As shown in Fig. 7a PACAP   (Fig. 7b). However, Tat-RACK1⌬C alone did not affect the basal levels of BDNF mRNA (Fig. 7b). Therefore, our results suggest that the release of RACK1 from the NMDA receptor complex results not only in the enhancement of NMDA receptor activity by Fyn but also in an increase in the mRNA levels of BDNF via RACK1 itself. DISCUSSION Compartmentalization of signaling proteins in close proximity to the NMDA receptor is highly important for the regulation of phosphorylation of the NMDA receptor subunits and the regulation of function of NMDA receptor activity. Recently, we identified RACK1 as a scaffolding protein for Fyn kinase and the NR2B subunit of the NMDA receptor and revealed a novel molecular mechanism by which RACK1 modulates the function of Fyn and the NMDA receptor (1). RACK1 binds directly to both Fyn and the cytoplasmic tail of NR2B, inhibits the phosphorylation of NR2B by Fyn, and decreases NMDA receptormediated EPSCs in CA1 hippocampal neurons (1). Here, we identified the cAMP/PKA pathway as an important signaling mechanism responsible for 1) the release of RACK1 from the NMDA receptor complex, 2) the subsequent tyrosine phosphorylation of the NR2B subunit, 3) the increase in NMDA receptor-mediated fEPSPs in hippocampal slices in a Fyn-dependent manner, and 4) the increase in BDNF expression via RACK1. Based on our results we propose the following model (Fig. 8). RACK1 binds the cytoplasmic tail of NR2B and Fyn. This allows Fyn to be localized in close proximity to its substrate. However, when RACK1 is present in the complex, Fyn cannot phosphorylate NR2B (Fig. 8a). Activation of adenylate cyclase by forskolin or activation of G s -coupled PACAP(1-38) receptors that activate the cAMP/PKA pathway releases RACK1 from NR2B and Fyn and induces its nuclear compartmentalization (Fig. 8b). Once RACK1 is removed from the NMDA receptor, Fyn is then free to phosphorylate NR2B, which in turn results in an increase in the activity of the NMDA receptor channel. Lastly, nuclear RACK1 will induce the expression of BDNF (Fig. 8b).
We found that the activation of the cAMP/PKA signaling pathway resulted in the release of RACK1 from the NMDA receptor complex. Several reports suggest a link among RACK1, cAMP signaling, and NMDA receptors. Yarwood et al. (30) identified a direct interaction between RACK1 and the cAMP-specific phosphodiesterase 4 isoform phosphodiesterase 5 (30), and recently phosphodiesterase 4 was found to regulate the function of the NMDA receptor and thus is likely to be part of the postsynaptic density complex as well (31). Moreover, several reports suggest a link between cAMP/PKA signaling and the phosphorylation state and activity of the NMDA receptor channel. In vitro, PKA phosphorylates the NR1 subunit of the NMDA receptor (32,33). Forskolin enhances NMDA receptor activity in spinal cord dorsal horn neurons and in hippocampal neurons in culture (34,35), and the NMDA receptorassociated protein, Yotiao (36), interacts with PKA holoenzyme and type 1 protein phosphatase (37). Taken together, these findings strongly suggest that the cAMP machinery is compartmentalized within close proximity to RACK1 and the NMDA receptor, allowing the rapid dissociation of RACK1 upon activation of a G␣ s -coupled receptor, and the rapid regulation of the NMDA receptor function.
We identified the neuropeptide PACAP(1-38) as a ligand responsible for the phosphorylation of the NR2B subunit and the enhancement of NMDA receptor-mediated activities via Fyn. PACAP(1-38) mediates its activities via two receptors, type 1 (PAC1), which binds PACAP(1-38) in high affinity, and  8. RACK1 forms a complex with Fyn and NR2B and is released upon adenylate cyclase stimulation. a, RACK1 binds the cytoplasmic tail of NR2B and Fyn. This allows Fyn to be localized in close proximity to its substrate. However, when RACK1 is present in the complex, Fyn cannot phosphorylate NR2B. b, activation of adenylate cyclase by forskolin or activation of G s -coupled PACAP(1-38) receptor PAC1 that activates the cAMP/PKA pathway releases RACK1 from NR2B and Fyn and induces its nuclear compartmentalization. Once RACK1 is removed from the NMDA receptor, Fyn is free to phosphorylate the NR2B subunit, which in turn results in an increase in the activity of the NMDA receptor channel. As a result of RACK1 nuclear translocation, BDNF expression is enhanced. type 2 (PAC2), which binds PACAP  and vasoactive intestinal peptide at equal affinities (38). The PAC1 receptor is highly expressed in several regions of the central nervous system including the hippocampus (38). The hippocampus mainly expresses the PAC1 but not PAC2 receptors (39), and electron microscope studies revealed that the PAC1 receptors are found in postsynaptic densities of synaptic junctions in brain regions including the hippocampus (40). Therefore, it is likely that the increase in tyrosine phosphorylation of NR2B occurs via activation of the PAC1 receptors by PACAP . PAC1 receptors are coupled mainly to the cAMP/PKA pathway via G␣ s , but also to the phospholipase C/PKC pathway via activation of G␣ qcoupled receptors (19). However, it is unlikely that PACAP(1-38) is mediating the phosphorylation of NR2B and the enhancement of NMDA receptor activities via the PLC/PKC pathway. First, we could not detect dissociation of the NR2B⅐RACK1⅐Fyn complex upon activation of PKC with the PKC activator PMA. Second, activation of the phospholipase C/PKC pathway requires higher concentrations of PACAP(1-38) (41) compared with the concentrations used here, which enhanced the phosphorylation and function of the NMDA receptor. Interestingly, PACAP(1-38) has been implicated in learning and memory. PAC1 deletion mutant mice show deficits in contextual fear conditioning, which is a hippocampus-dependent associative learning, and in a passive avoidance learning (42), and thet also show impairment of hippocampal mossy fiber long term potentiation (43). Thus, it is possible that the enhancement of NMDA receptor function by PACAP(1-38) is important for synaptic plasticity, a process that underlies several forms of learning and memory.
In addition to PACAP(1-38) activities on NMDA receptor, we found that the peptide enhanced the expression of BDNF via nuclear RACK1. Recently, we found that in C6 glioma cells nuclear RACK1 mediates the induction of expression of the immediate early gene c-fos in a cAMP/PKA-dependent manner (13). In cerebellar granule cells, PACAP(1-38) was found to induce c-fos expression via the cAMP/PKA pathway (28). Because the promoter region of BDNF contains several AP1 sites (44), it is possible that PACAP(1-38) induces BDNF expression via c-fos. Interestingly, BDNF has also been strongly implicated in the mechanism leading to enhancement of synaptic strength (45). Presynaptically, BDNF enhances glutamate release in hippocampal neurons (46 -48). Postsynaptically, BDNF enhances the activity of NMDA receptors (49,50), at least in part, by inducing the phosphorylation of NR1 and NR2B but not NR2A subunits (51). BDNF is essential for the formation of long-term potentiation, and BDNF knockout mice have impaired long-term potentiation that is restored with viral infection of BDNF (52)(53)(54)(55)(56)(57). In addition, in excised membrane patches from cultured hippocampal neurons, acute exposure to BDNF increases NMDA single channel open probability, and this effect is dependent on the presence of the NR2B subunit (58). These findings raise the possibility that the increase in BDNF levels via the nuclear compartmentalization of RACK1 initiates a positive feedback loop for further enhancing the function of hippocampal NMDA receptors.
In summary, we have identified cAMP/PKA as an important regulatory signaling pathway for the localization and function of the scaffolding protein RACK1. The movement of other scaffolding proteins has been shown to be important for proper transduction of signaling cascades and regulation of gene expression. For example, the scaffolding protein STE5 moves to the nucleus to induce transcription (59). The dual function of RACK1 in response to cAMP/PKA activation, i.e. an enhancement in NMDA receptor phosphorylation and activity as well as an increase in expression of BDNF, leads to the interesting possibility that the dynamic compartmentalization of scaffolding proteins adds yet another dimension to signaling events. Our results also suggest that the dynamic compartmentalization of RACK1 is important for the regulation of processes leading to synaptic plasticity.