Regulation of Neuroprotective Activity of Myocyte-enhancer Factor 2 by cAMP-Protein Kinase A Signaling Pathway in Neuronal Survival*

The transcription factor myocyte-enhancer factor 2 (MEF2) has been shown to be required for the survival of different types of neurons. However, the death- or survival-inducing second messenger pathways that regulate MEF2 activity remain to be fully elucidated. Membrane depolarization by KCl induces neuronal survival that is dependent upon MEF2-mediated gene transactivation. Here we report that membrane depolarizationinduced activation of MEF2 requires the cAMP-protein kinase A (PKA) pathway. Inhibition of the activity of cAMP-PKA pathway attenuates membrane depolarization-induced activation of MEF2 activity and neuronal survival, whereas enhancing the activity of this pathway prevents KCl withdrawal-induced inhibition of MEF2 and neuronal apoptosis. Moreover, PKA directly phosphorylates MEF2 at Thr-20 in vitro to increase MEF2 DNA binding activity. A mutation of Thr-20 to Ala renders MEF2 resistant to PKA phosphorylation in vitro and reduces its DNA binding activity. Transfection of this T20A mutant blocks survival and induces apoptosis in cultured cortical and cerebellar granule neurons. This study identifies the transcription factor MEF2 as a target of cAMP-PKA pathway and demonstrates that PKA phosphorylation of MEF2 is a key step in modulating its DNA binding activity and ability to promote neuronal survival.

Both growth factors and neuronal activity provide trophic support for neurons during their development; neurons that do not receive sufficient support will undergo apoptotic death (1). Although the mechanism is poorly understood, it has long been known that elevation of cyclic AMP (cAMP) levels will promote neuronal survival that is independent of neurotrophic factors (2). This has been observed in cerebellar granule neurons (CGNs), 1 a widely used model system for studying the mecha-nisms of membrane depolarization-induced survival of cerebellar granule neurons (3,4). In this model, cerebellar granule neurons cultured in the presence of a high concentration of extracellular KCl (25-30 mM) survive in the absence of serum or other additional growth factors, whereas lowering the concentration of KCl (5 mM) induces rapid and relatively synchronized apoptosis. Several signaling pathways required to mediate the survival of cerebellar granule neurons have been identified. These include brain-derived neurotrophic factor signaling pathway, mitogen-activated protein kinase cascade, and the phosphatidylinositol 3 (PI3)-kinase/Akt pathway (5)(6)(7)(8).
Myocyte-enhancer factor 2 (MEF2), isoforms A-D, is a family of transcription factors that play critical roles in diverse cellular processes, including neuronal survival (9,10). Recent studies have demonstrated that MEF2 is necessary for activity-dependent survival of cerebellar granule neurons (11)(12)(13). MEF2s are the target for several key intracellular signaling pathways that are known to control cellular survival and apoptosis. These include p38 mitogen-activated protein kinase, extracellular signal-regulated kinase 5, Cdk5, and caspases (11,(13)(14)(15)(16). However, the dynamic changes in MEF2 phosphorylation status upon neuronal survival and apoptotic signaling strongly suggest that regulation of MEF2 is more complex, involving yet unidentified kinases and phosphatases (12,17). To fully understand the mechanisms underlying neuronal survival and the role of MEF2 in this process, it is essential to identify the regulators of MEF2.
The second messenger cAMP regulates a number of physiological processes, including neuronal survival (18). cAMP has been shown to be neuroprotective for dorsal root ganglion neurons, retinal ganglion neurons, and cerebellar granule neurons under several experimental paradigms, including growth factor withdrawal (2, 19 -23). Dependent upon the nature of stress and the types of neurons, cAMP may mediate neuronal survival in part through its major downstream effector, protein kinase A (PKA). cAMP signaling has been shown to stimulate surface expression of the TrKB receptor, inhibit the proapoptotic factors GSK3␤ and BAD, activate the neuroprotective mitogen-activated protein kinase pathway, and phosphorylate cyclic AMP-response elementbinding protein, a survival factor (24 -30).
Our data demonstrate that in cerebellar granule neurons, where MEF2 activity is shown to be required for membrane depolarization-dependent survival, inhibition of the cAMP-PKA pathway correlates with a reduction of transactivation activity of MEF2. We have shown that depolarizing concentrations of KCl activate PKA, which phosphorylates MEF2 on a threonine residue (Thr-20) located within its DNA binding and dimerization domain near the NЈ-terminal region. Phosphoryl-ation at this site by PKA leads to enhanced DNA binding by MEF2 in vitro, and blocking the cAMP-PKA signaling pathway reduces MEF2-dependent DNA binding and gene expression. A MEF2 construct in which the PKA phosphorylation site was mutated to an Ala fails to mediate gene expression in response to survival signals and, when overexpressed in neurons, induces neuronal apoptosis.

MATERIALS AND METHODS
Plasmids, Chemicals, and Antibodies-Constructs for wild type and mutated MEF2-dependent luciferase reporter and MEF2C⅐VP16 were described previously (31). Antibodies were purchased from the following vendors: anti-MEF2C from Cell Signaling Technology, anti-MEF2A polyclonal antibody from Santa Cruz Biotechnology, and anti-MEF2D antibody from BD Transduction Laboratories. Propidium iodide and poly-L-ornithine were purchased from Sigma, H-89, forskolin, myristoylated PKI peptide, and CPT-cAMP from Calbiochem, mammalian transfection system from Promega, Western blot stripping buffer from Pierce, and sulforhodamine caspase detection kit from Biocarta.
Culture of Rat Primary Neurons-Cultures of primary cerebellar granule neurons were established as described by Mao and Wiedmann (12). Briefly, cerebellum from postnatal day 6 rat pups was dissected and subjected to enzymatic dissociation. Dissociated cells were cultured on poly-L-ornithine-coated plates in basal growth medium. Culture of primary cortical neurons from embryo day 17 or 18 Long Evans rats (Harlan or Charles River) was carried out as described by Mao and Wiedmann (12). Cortical neurons were grown in basal growth medium in plates coated with laminin and poly-D-lysine (Sigma).
Transfection of Cells-Primary neurons in Dulbecco's modified Eagle's medium were transfected by the calcium phosphate method at day 4 -6 in vitro as described by Mao and Wiedmann (12). In general, neurons were returned to conditioned full medium after transfection and then treated as indicated when needed. A 3:1 DNA ratio of effector versus ␤-galactosidase or GFP vector was used for survival and luciferase assays, respectively.
In Vitro Kinase Assay-Purified recombinant MEF2 proteins (0.5 g) were incubated with purified catalytic subunit PKA (Promega) or PKA immunoprecipitated from whole cell extracts from neurons by using an anti-PKA antibody in a kinase reaction buffer containing [␥-32 P] and cold ATP. Whole cell extracts from cultured primary cerebellar granule neurons were prepared as described previously (12). In brief, cells were washed once with ice-cold phosphate-buffered saline and lysed with 500 l of buffer (for 100-mm plate) containing 20 mM Tris⅐HCl (pH 7.4), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 2.5 mM sodium pyrophosphate, 1 mM ␤-glycerophosphate, 10 mM sodium fluoride, 1 mM Na3VO4, and 1% Triton X-100. For immunoprecipitation, 100 g of cell lysate were incubated with anti-PKA antibody for 1 h at 4°C and then incubated with Protein A-agarose beads for an additional 2 h. The beads were washed with kinase buffer five times. The kinase reaction was carried out for 10 min at 30°C following the manufacturer's protocol (Promega) and terminated by the addition of Laemmli sample buffer. Reaction products were resolved by SDS-PAGE, and 32 P-labeled proteins were visualized by autoradiography. The stoichiometry was determined as described (32). Briefly, 0.5 g of MEF2C-(1-105) and 80 units of purified PKA were incubated in the kinase reaction buffer. The kinase reaction was terminated at 5, 10, 15, and 30 min. 20 l of the reaction were applied to P81 filter paper (Upstate). The filters were washed three times for 5 min each with 0.75% phosphoric acid and then washed once with ethanol for 5 min. 32 P i incorporation into MEF2C-(1-105) was quantified by Cerenkov counting using a liquid scintillation analyzer (Packard). A control reaction was performed with the same assay mixture containing PKA but without substrate MEF2C-(1-105).
Luciferase Reporter Gene Assay-Primary neurons were transiently transfected with various constructs using the calcium phosphate transfection procedure described by Mao and Wiedmann (12). A ␤-galactosidase expression plasmid was used to determine the efficiency in each transfection. The total amount of DNA for each transfection was kept constant by using control vectors. Cell lysates were analyzed for luciferase and ␤-galactosidase activity according to the manufacturer's instructions (BD Bioscience).
Detection of Caspase Activation-Cerebellar granule neurons were treated with 10 M H89 or the same volume of its solvent Me 2 SO for 5 h. Sulforhodamine caspase detection reagent was used and incubated with the cells for 1 h according to the manufacturer's manual. Cells were observed and photographed using a fluorescence microscope (excitation 550 nm; emission 590 -600 nm).
Cell-counting Kit-8 -Water-soluble tetrazolium salt (WST-8; Dojindo Molecular Technologies, Inc.) was added directly to the cells and bioreduced by the cellular dehydrogenases to give a colored formazan product that is soluble in the tissue culture medium. The intensity of the color reaction was measured using a Packard Spectracount. The relative survival rates were represented with the O.D. value of the control group set as one.
Survival Assays-The survival assays were carried out as described previously (31,33). Neurons were stained with propidium iodide without permeabilization. GFP-positive cells with or without propidium iodide staining were counted using a fluorescence microscope in a blind manner. Three hundred or more transfected cells were counted for each treatment. Apoptotic rates were represented as the percentage of GFP-positive apoptotic cells among the total number of GFP-positive cells counted.
Statistics Method-The results were analyzed using one-way analysis of variance for samples where appropriate.

PKA Activity Is Required for MEF2 Function and Membrane
Depolarization-induced Survival of Cerebellar Granule Neurons-We first established that cAMP-PKA signaling is required for depolarization-mediated survival of cerebellar granule neurons. Cerebellar granule neurons from postnatal day 6 rat brain were cultured in the presence of either depolarizing concentrations of KCl (29 mM) or low concentration of KCl (5 mM) in the absence of serum. Membrane depolarization stimulated the survival of cerebellar granule neurons, indicated by the healthy nuclear morphology following Hoechst staining, while lowering the concentration of KCl-induced neuronal apoptosis as shown by chromatin condensation (Fig. 1A). Several studies have implicated cAMP in the survival of cerebellar granule neurons upon depolarization (20, 34 -37). However, the downstream effectors and mechanisms that mediate cAMP signaling in cerebellar granule neurons have not been fully investigated. Because cAMP may signal via different downstream effectors, including PKA and cAMP-regulated guanine nucleotide exchange factors (GEFs) such as GEF1/II or Epac (exchange protein directly activated by cAMP) (38,39), it was important to determine which downstream effector(s) mediates the trophic effect of cAMP. To distinguish the role of PKA in this process, two widely used potent inhibitors of protein kinase A, H89 and myristoylated PKI peptide, were applied to block PKA activity in cerebellar granule neurons under depolarizing concentrations of KCl. Inhibition of PKA by H89 induced a dose-and time-dependent loss of neuronal viability (Fig. 1B, left and middle panels). The more specific PKA inhibitor, myristoylated PKI peptide, also reduced neuronal viability in the presence of depolarizing concentrations of KCl (Fig. 1B, right panel). Conversely, increasing cAMP levels with forskolin or the cAMP analog CPT-cAMP protected neurons from low concentration KCl-induced apoptosis (Fig. 1C). Together, these data confirm that the activity of cAMP-PKA pathway is required for depolarization-induced survival of cerebellar granule neurons.
Our previous studies show that MEF2 is required for neuronal survival mediated by membrane depolarization (11,12). To test the possibility that the cAMP-PKA pathway may regulate the function of MEF2, we employed a sensitive MEF2-dependent reporter gene assay to correlate survival signal, cAMP-PKA activity, and MEF2 function (11). Cerebellar granule neurons were transiently transfected with a luciferase reporter gene whose expression is under the control of two MEF2 DNA binding elements. MEF2 activity was measured by luciferase assay. Membrane depolarization-dependent survival correlated with high MEF2 activity, whereas 5 mM KCl-induced apoptosis was associated with a decrease in MEF2 function ( Fig. 2A), consistent with a positive role of MEF2 in neuronal survival. Treating neurons with two PKA inhibitors, H89 and myristoylated PKI peptide, reduced MEF2-dependent reporter gene expression (Fig. 2B). In contrast, KCl withdrawal inhibited MEF2 trans-activation activity, and this inhibition was attenuated by stimulation of cAMP-PKA pathway with either CPT-cAMP or forskolin (Fig. 2C). Together, these data indicate that MEF2 activity correlates with the activity of cAMP-PKA pathway, suggesting that cAMP-PKA may target MEF2.
Membrane Depolarization Regulates PKA Kinase Activity and PKA Phosphorylates MEF2 Directly-Our preceding data indicate that cAMP-PKA is required for neuronal activity-dependent survival. Given that cAMP-PKA pathway affects MEF2 transactivation activity (Fig. 2), we investigated the possibility that PKA may directly target MEF2. To test this possibility, purified PKA and a highly purified N-terminal fragment of recombinant MEF2 protein (MEF2C-(1-105)) were incubated in a kinase reaction buffer containing [␥-32 P]ATP as described under "Materials and Methods." The results were viewed after autoradiography. Incubation of MEF2C-(1-105) with PKA resulted in significant phosphorylation of the N-terminal MEF2 fragment (Fig. 3A, top panel).
To determine the stoichiometry of MEF2 phosphorylation, the kinase reaction was stopped at 5, 10, 15, and 30 min. 32 P i incorporation into MEF2C-(1-105) was measured by a scintillation counter in a filter binding assay. A representative result after subtraction of control is shown in Fig. 3A, bottom panel. Upon saturation of P i incorporation, we found that 0.5 mol P i was incorporated into 1 mol MEF2C-(1-105). This NЈ-terminal fragment contains the highly conserved MADS and MEF2 domains. Sequence analysis revealed the presence of a consensus PKA phosphorylation site at position 20 (RQVT) that is conserved among all four isoforms of MEF2 (Fig. 3B, top panel). Mutation of Thr-20 to an Ala almost completely abolished PKA-mediated phosphorylation (Fig.  3B, bottom panel), suggesting that phosphorylation by PKA requires Thr-20. Interestingly, the RQVT sequence is also part of a consensus motif for protein kinase B (Akt), RN-RQVT, making Thr-20 a potential target for Akt (40). To test this possibility, equal amounts of MEF2C-(1-105) were incubated with the same reactive units of PKA and PKB for the same length of time under the optimal conditions for each enzyme, respectively. Under this condition, PKA robustly phosphorylated MEF2C-(1-105), whereas Akt led to a much weaker incorporation of 32 P into MEF2C-(1-105) (Ͼ100-fold less efficient than PKA adjusted for loading) (Fig. 3C), consistent with previous reports that fail to demonstrate efficient phosphorylation of MEF2 by Akt under different experimental conditions (41).
To corroborate the role of PKA in regulation of MEF2, we tested directly whether membrane depolarization regulates PKA kinase activity. Cerebellar granule neurons were exposed to either high or low concentrations of KCl. Endogenous PKA was immunoprecipitated from cellular lysates, and its activity was determined by in vitro kinase assay with MEF2C-(1-105) as a substrate in the presence of [␥-32 P]ATP. PKA activity was high in neurons cultured in medium containing depolarizing concentrations of KCl, and this activity was significantly inhibited by H89 (Fig. 3D). Withdrawal of KCl led to a decrease in PKA kinase activity. Addition of forskolin re-stimulated PKA activity that was inhibited by KCl deprivation. The specificity of the in vitro assay was confirmed by near complete inhibition of the kinase activity present in the immunoprecipitates with the addition of the specific PKA inhibitor PKI. Together, these data indicate that membrane depolarization specifically enhances PKA function in cerebellar granule neurons.
Phosphorylation by PKA at Thr-20 Regulates MEF2 DNA Binding Activity-Our experiments presented in Fig. 3 showed that PKA phosphorylates MEF2 at Thr-20. Because the NЈterminal region included in our phosphorylation assay contains conserved subdomains that are responsible for mediating interaction between MEF2 and specific DNA sequences, it raised the possibility that phosphorylation at Thr-20 may affect MEF2dependent DNA binding activity. We tested the effect of phosphorylation by PKA on the DNA binding activity of MEF2 in electrophoretic mobility shift assay, which we have used previously to characterize the specific interaction between MEF2 and DNA (12). Different amounts of recombinant MEF2C NЈterminal fragment-(1-105) were either phosphorylated or nonphosphorylated with PKA in vitro with cold ATP and then incubated with a 32 P-labeled DNA probe containing a single MEF2 binding site. Autoradiography showed that MEF2C-(1-105) binds to the MEF2 probe (Fig. 4A). This binding is sequence-specific because mutation of the MEF2 DNA binding site present in the probe completely abolished the formation of MEF2C-(1-105) and DNA probe complex. Importantly, phosphorylation by PKA significantly enhanced the signal of MEF2-(1-105) and DNA probe complex, suggesting that phosphorylation by PKA stimulates MEF2 DNA binding activity. On the other hand, mutation of Thr-20 to an Ala severely reduced the binding of MEF2 to DNA following phosphorylation by PKA compared with wild type MEF2 controls. This reduced binding of MEF2C T20A to DNA is due to a specific change of MEF2 DNA binding property as equal amounts of various MEF2s were used in each assay (data not shown). 2 Together, these data suggest that Thr-20 of MEF2 is critically involved in mediating the interaction between MEF2 and DNA. To extend this observation to neurons, cerebellar granule neurons cultured in medium containing 29 mM KCl were treated with PKA inhibitors H89 or PKI and tested for MEF2-dependent DNA 2 X. Tang, X. Wang, and Z. Mao, unpublished observations. binding by electrophoretic mobility shift assay. Membrane depolarization was associated with a robust DNA binding by MEF2. Inhibition of PKA by H89 resulted in a significant decline of MEF2 DNA complex formation, supporting the idea that inhibition of PKA reduces MEF2 DNA binding potential (Fig. 4B, left panel). Similarly, the more specific PKA inhibitor, myristoylated PKI peptide, induced a significant decrease of MEF2 DNA formation in cerebellar granule neurons cultured in 29 mM KCl (Fig. 4B, right panel). Consistent with these, brief exposure of cerebellar granule neurons to 5 mM KCl caused a clear decline of DNA binding activity by MEF2 that could be reversed by increasing cAMP level or stimulating PKA activity with either CPT-cAMP or forskolin. Further, this enhancing effect of CPT-cAMP or forskolin on MEF2 DNA binding in the presence of 5 mM KCl was reduced by either H89 or myristoylated PKI peptide (Fig. 4B). To corroborate these findings, we tested the effect of H89 or PKI on a constitutively active MEF2 fusion protein, MEF2C⅐VP16, using the MEF2 reporter assay in cerebellar granule neurons. MEF2C⅐VP16 is a fusion of the NЈ-portion of MEF2, which contains the DNA binding and dimerization domain of MEF2, with the viral protein VP16, which contains the transactivation domain and is not known to be regulated by cAMP-PKA. Overexpression of MEF2C⅐VP16 led to a 50-to 100-fold increase in MEF2-dependent reporter gene activity. The PKA inhibitors H89 and myristoylated PKI peptide significantly inhibited MEF2-dependent reporter gene expression activated by MEF2C⅐VP16 (Fig. 4C), supporting the notion that PKA inhibitors may reduce the activity of MEF2C⅐VP16 by interfering with its DNA binding potential.
Blocking Phosphorylation by PKA Inhibits MEF2-dependent Gene Transactivation and Neuronal Survival-The increased MEF2 DNA binding and transactivation activity by PKA suggests that the survival-promoting function of MEF2 requires PKA. To demonstrate this directly, we tested the effect of a PKA phosphorylation mutant of MEF2A, MEF2A T20A, on MEF2mediated gene transactivation in reporter assay. Cerebellar granule neurons were transiently transfected with MEF2 luciferase reporter, and the endogenous MEF2 activity of neurons was stimulated by culture medium that contained 29 mM KCl. Co-expression of MEF2A T20A inhibited membrane depolarization-induced luciferase gene expression (Fig. 5A). Similar results were obtained with MEF2C T20A (data not shown). 2 These results suggest that MEF2 T20A mutants that are resistant to PKA phosphorylation inhibit the activation of endogenous function of MEF2 upon stimulation by membrane depolarization in a dominant negative fashion. Consistent with these observations, overexpression of MEF2A T20A also resulted in a significant increase in the apoptotic neurons cultured under 29 mM KCl condition (Fig. 5B). Similarly, although our previous studies show that overexpression of MEF2C⅐VP16 greatly reduced 5 mM KCl-induced apoptosis of cerebellar granule neurons (11), MEF2C⅐VP16 was far less effective in reducing H89-induced neuronal death (Fig. 5C), consistent with the findings that H89 targets the NЈ-terminal DNA binding domain of MEF2. Together, these data support the idea that enhanced DNA binding activity by PKA is required for the survival-promoting function of MEF2 in response to membrane depolarization.
cAMP-PKA Regulates MEF2 Activity under Growth Factormediated Survival Conditions and in Different Types of Neurons-cAMP-PKA signaling pathway plays a role in the survival of several types of neurons under different survival paradigms (2, 19 -23). Similarly, MEF2 has been shown to be required for neuronal survival under several conditions (11,12). To broaden the role of PKA-mediated regulation of MEF2 in neuronal survival, we tested the regulation of MEF2 by PKA in serum-stimulated survival of cerebellar granule neurons. Serum induced a robust stimulation of MEF2 activity that was dependent on PKA, as H89 blocked the serum-induced MEF2mediated gene expression (Fig. 6A). Consistently, serum withdrawal led to an inhibition of MEF2 function. This inhibition was attenuated by increasing the level of cAMP with CPT-cAMP (Fig. 6A). In addition, inhibition of PKA reduced serumstimulated MEF2C⅐VP16 activity in a manner similar to that caused by membrane depolarization (Fig. 6B). Our previous studies show that MEF2 supports the survival of cortical neurons (11,14). Therefore, we also examined whether PKA activ- ity is required for MEF2 function in cortical neurons. Primary cortical neurons isolated from embryo day 18 rat brains were cultured in the presence of serum and assayed for MEF2 ac-tivity as described above in MEF2 reporter assays. Consistent with our observations in cerebellar granule neurons, inhibition of PKA activity with H89 also significantly down-regulated MEF2 function in cortical neurons (Fig. 6C). Together, these data suggest that MEF2 activity is positively regulated by cAMP-PKA upon stimulation by different survival signals and in different types of neurons. DISCUSSION Accumulating evidence indicates that transcription factor MEF2 plays a critical role in the survival of different types of neurons under several experimental conditions (11)(12)(13)(14). In this study, we have identified the cAMP-PKA pathway as an important activator of MEF2 activity in neurons. PKA phosphorylates MEF2 directly at Thr-20 to enhance MEF2 DNA binding in vitro and promotes MEF2-dependent survival of cerebellar granule neurons. Given the recognized role of cAMP in neuronal survival, identifying the nuclear survival factor MEF2 as a downstream effector of cAMP-PKA pathway provides an important molecular basis for its potent trophic action.
cAMP has been shown to promote the survival of different types of neuronal cells, including cerebellar granule neurons, retinal ganglion neurons, and dorsal root ganglion neurons (2, 19 -23). Consistent with these findings, inhibition of PKA by PKI and H89 blocks the survival of cerebellar granule neurons (Fig. 1). Under our experimental conditions, however, myristoylated PKI inhibitor is less potent than H89 in preventing the survival of CGNs in 29 mM KCl. H89 is shown to inhibit other kinases in addition to PKA (42). Therefore, some of the effect of H89 on survival may be due to its inhibition of other kinases. However, PKA activity is clearly required to mediate cAMP survival signal in CGNs. Only a very limited number of nuclear effectors of PKA have been revealed to mediate cAMP-PKA signaling; among those, fewer have been implicated in neuronal survival (43). Therefore, identification of MEF2 as a downstream mediator of PKA-dependent neuronal survival expands the targets by which cAMP-PKA mediates survival signaling. Our data suggest that PKA signaling regulates MEF2 in response to growth factor as well as to membrane depolarizationinitiated survival in at least two types of neurons. Together, these findings raise the possibility that cAMP-PKA-mediated MEF2 response may play an important role in supporting the survival of a wide spectrum of neurons under a broad range of conditions.
Our kinase studies show that Thr-20 is required for phosphorylation of MEF2 by PKA in vitro. Mutation of Thr-20 to Ala renders MEF2 insensitive to cAMP-PKA signaling in neurons. Although the PKA recognition motif was initially defined as RRXS/T (44), further analysis showed that PKA also recognizes substrates containing the consensus site RXXS/T (45,46). The amino acid residues NЈ terminus to Thr-20 are RQVT 20 , fitting the RXXT motif. Based on this, the simplest interpretation of our data is that PKA also regulates MEF2 in cells through phosphorylation of Thr-20. The stoichiometry of MEF2 phosphorylation is 0.5. Coupled with the finding that mutation of a single site (Thr-20) almost completely abolished MEF2 phosphorylation by PKA, these data suggest that under our condition in vitro, about 50% of MEF2-(1-105) were phosphorylated at a single site. This is consistent with observations of others that under similar experimental conditions often only a fraction of protein is phosphorylated (47,48). The less than 100% phosphorylation of MEF2 probably reflects the fact that our in vitro kinase assay conditions were less than optimal.
Several kinases have been identified to regulate MEF2 activity in neurons. These include positive regulators p38 mitogen-activated protein kinase and extracellular signal-regulated kinase 5, as well as the negative regulator Cdk5 (11, 14, 31). These kinases phosphorylate MEF2 at its CЈ-transactivation domain and modulate MEF2 transactivation activity through yet unknown mechanisms. In contrast to these CЈ regulatory kinases, our data presented in this study suggest that regulation of MEF2 by PKA is mechanistically different in that PKA phosphorylates MEF2 at its NЈ-DNA binding and dimerization domain. Consistent with the function of this domain, PKA phosphorylation at Thr-20 enhances DNA binding activity of MEF2 in vitro, and blocking PKA activity reduces MEF2 DNA binding in cerebellar granule neurons. To date, the only other kinase reported to target MEF2 at its NЈ terminus is casein kinase II (CKII), which phosphorylates Ser-59 (49). However, phosphorylation of MEF2 by CKII is thought to be constitutive instead of regulated. Moreover, CKII-mediated phosphorylation of MEF2 has not been shown to occur in neuronal cells. Therefore, PKA-mediated phosphorylation of MEF2 represents the first example of signal-dependent regulation of NЈ-terminal DNA binding function of MEF2. Structural analysis of the NЈ-terminal portion of MEF2A indicates that the ␣1 helix segment, which contains Thr-20 and extends from residues 14 to 38, may be involved in DNA bending and extensive intersubunit interactions between two dimer-forming MEF2 molecules (50). The ␣1 segments from two MEF2 molecules are arranged in a coiled-coil antiparallel configuration that allows proper interaction with DNA. Given that MEF2 dimer formation is critical for its DNA binding, it is thus possible that phosphorylation at Thr-20 may induce a structural change that provides a more favorable interaction either between MEF2 and MEF2 molecules or/and between MEF2 and DNA.
Although cAMP-PKA pathway is generally recognized for its neuroprotective role, recent studies have shown that cAMP can also negatively regulate the survival of cortical neurons by inhibiting brain-derived neurotrophic factor-stimulated PI3kinase (PI3K) signaling (51). Interestingly, the PI3-kinase pathway positively regulates MEF2 activity in muscle cells (41). Consistent with this, our data indicate that PI3K pathway also promotes MEF2 function in neurons. 2 Together, these findings suggest that under certain conditions, cAMP signaling may negatively regulate MEF2 activity by interfering with PI3K signaling to MEF2. In agreement with this, cAMP signaling has been shown to inhibit membrane localization of the p85 regulatory subunit of PI3K in corneal endothelial cells and to inhibit GTPase Rap1-mediated activation of PI3K in rat C6 glioma cells (52,53).
In neurons, the cAMP-PKA pathway also plays a role in pathological responses and in synaptic plasticity (42). For example, during acute cerebral ischemia, a decline of cAMP-PKA activity correlates with ischemic neuronal damage, especially in vulnerable regions such as the CA1 of the hippocampus, and enhancement of this pathway parallels neuronal survival. Various isoforms of MEF2 are expressed in the hippocampus (54 -56). Therefore, studying whether inhibition of PKA-mediated regulation of MEF2 plays a role in ischemia-induced death of hippocampus neurons in the ischemic core would provide important information on how ischemia regulates neuronal survival and death. Furthermore, activation of cAMP-PKA has been implicated in synaptic plasticity, a process that requires transcription and protein synthesis (57). It would be interesting to determine whether MEF2 activity is also involved in the cAMP-PKA signaling that regulates gene expression relevant to learning and memory.