Cyclin-dependent kinase 5 mediates neurotoxin-induced degradation of the transcription factor myocyte enhancer factor 2

Abstract

Regulation of the process of neuronal death plays a central role both during development of the CNS and in adult brain. The transcription factor myocyte enhancer factor 2 (MEF2) plays a critical role in neuronal survival. Cyclin-dependent kinase 5 (Cdk5) mediates neurotoxic effects by phosphorylating and inhibiting MEF2. How Cdk5-dependent phosphorylation reduces MEF2 transactivation activity remained unknown. Here, we demonstrate a novel mechanism by which Cdk5, in conjunction with caspase, inhibits MEF2. Using primary cerebellar granule neuron as a model, our investigation reveals that neurotoxicity induces destabilization of MEF2s in neurons. Destabilization of MEF2 is caused by an increase in caspase-dependent cleavage of MEF2. This cleavage event requires nuclear activation of Cdk5 activity. Phosphorylation by Cdk5 alone is sufficient to promote degradation of MEF2A and MEF2D by caspase-3. In contrast to MEF2A and MEF2D, MEF2C is not phosphorylated by Cdk5 after glutamate exposure and, therefore, resistant to neurotoxin-induced caspase-dependent degradation. Consistently, blocking Cdk5 or enhancing MEF2 reduced toxin-induced apoptosis. These findings define an important regulatory mechanism that for the first time links prodeath activities of Cdk5 and caspase. The convergence of Cdk5 phosphorylation-dependent caspase-mediated degradation of nuclear survival factors exemplified by MEF2 may represent a general process applicable to the regulation of other survival factors under diverse neurotoxic conditions.

Figure 1.

Glutamate-induced neuronal apoptosis correlates with the degradation of MEF2A and MEF2D. A, Cerebellar granule neurons were treated with glutamate (100 μm). Chromatin condensation was revealed by Hoechst 33258 staining; arrows indicate apoptotic cells (top). Loss of neuronal viability was quantified by dehydrogenase activity at various times after glutamate exposure (bottom). B, Glutamate inhibited MEF2-dependent luciferase activity in a time-dependent manner. CGNs were transfected with a MEF2-dependent luciferase reporter construct and then exposed to glutamate. Error bars represent SD. mt, Mutant; wt, wild type. C, Glutamate promoted degradation of MEF2A and MEF2D proteins. CGNs were treated with glutamate; MEF2A and MEF2D protein levels were determined by Western blotting using anti-isoform-specific antibodies. Histone H1 was used as a loading control after reprobing the same membrane. D, The level of c-Jun was not altered by glutamate. CGNs were treated with glutamate for the times indicated. The levels of c-Jun, MEF2D, and actin were determined by Western blotting.

Figure 2.

Caspase-dependent degradation of MEF2 after glutamate treatment. A, Degradation of MEF2 was mediated by both caspase and proteasome-dependent processes. CGNs were pretreated with z-VAD.fmk or proteasome inhibitor I and then treated with (left) or without (right) glutamate. Levels of MEF2D and actin were determined by Western blotting. B, Glutamate-induced MEF2D degradation correlated with the generation of Cdk5 activator p25, activation of caspases, and loss of CGN viability. CGNs were treated with glutamate on different days in culture for the time indicated. MEF2D, p25, and actin levels were determined by Western blotting (top). Activation of caspase was determined by the Poly Caspase Detection kit (middle), and cell viability was determined by dehydrogenase level (bottom graph). Error bars represent SD.

Figure 3.

Correlation and activation of nuclear Cdk5 and caspase by glutamate. A, Glutamate activated nuclear Cdk5 kinase activity. An in vitro Cdk5 kinase assay was performed using purified C′ fragment of MEF2D as a substrate after anti-Cdk5 immunoprecipitation from nuclear lysates (top). Levels of Cdk5 and histone H1 protein were determined by Western blotting (middle and bottom, respectively). The Cdk5 inhibitor roscovitine was included with glutamate where indicated. B, Glutamate stimulated incorporation of [³²P]orthophosphate into MEF2. CGNs were treated with glutamate and metabolically labeled with [³²P]orthophosphate. ³²P incorporation in MEF2D was determined by autoradiography after anti-MEF2D immunoprecipitation (top). The bottom panel shows the MEF2D level after immunoprecipitation. C, MEF2D was phosphorylated at the Cdk5 recognition site, Ser444. The level of MEF2 phosphorylated at Ser444, and total MEF2 were determined by Western blotting using phospho-specific anti-MEF2 antibody and anti-MEF2D antibody (reprobing the same membrane), respectively. Quantitation is shown at the bottom (n = 3). D, Glutamate activated caspase 3 in CGNs. Activated caspase 3 was determined by Western blotting (top). Z-VAD.fmk was added with glutamate where indicated. The bottom panel shows histone H1 protein level. E, Increased phosphorylation and degradation of MEF2 correlated. Equal amounts of protein were loaded under each condition based on histone H1. Western blotting analyses were performed using antibodies to MEF2A, MEF2D, phospho-MEF2, and histone H1 after reprobing the same membrane. F, Inhibition of caspase rescued MEF2-dependent reporter activity. mt, Mutant; wt, wild type. G, Inhibition of caspase reduced glutamate-induced death of CGNs. CGNs were treated with glutamate for 6 h, and cell viability was measured by WST assay. Pre and Post indicate that Z-VAD.fmk was added 30 min before or 4 h after glutamate exposure, respectively. The total length of Z-VAD.fmk treatment was 2 h. n = 3. Error bars represent SD.

Figure 4.

Cdk5 inhibitor roscovitine prevented glutamate-induced MEF2 degradation. A, Glutamate induced fragmentation of MEF2. CGNs were treated with glutamate as indicated. For the top panels, Western blotting with an antibody anti-N′ MEF2 shows fragmentation of MEF2 (top blot). The same membrane was reprobed for MEF2D for equal loading (bottom blot). For the bottom panels, anti-phospho-MEF2 blotting was performed (top blot). The same membrane was probed for MEF2D (bottom blot) and H1 (equal H1 loading data not shown). The double arrowhead and single arrow indicate the positions of the full length of MEF2A/D and MEF2 fragments, respectively. B, Cdk5 and caspase inhibitors reduced glutamate-induced degradation of MEF2. CGNs were incubated with glutamate with or without pretreatment of Cdk5 and/or caspase inhibitors. The same membranes were probed successively with anti-MEF2A, MEF2D, and histone H1 antibodies (top 3 panels) or with anti-MEF2D and actin antibodies (bottom 2 panels). C, Caspase and Cdk5 inhibitors reduced MEF2 fragmentation. CGNs were treated as described in B. Western blotting with anti-N′ and anti-phospho-MEF2 antibodies were performed. p-MEF2₂ is the overexposure of p-MEF2₁ to show fragmentation. The double arrowhead and single arrow indicate the positions of the full length of MEF2A/D and MEF2 fragments, respectively.

Figure 5.

Isoform-specific recognition of phospho-MEF2 fragments. A, Recognition of phospho-MEF2 fragments by isoform-specific MEF2A and MEF2D antibodies. CGNs were treated with glutamate as in Figure 4C, and MEF2 fragments were analyzed by Western blotting using phospho-MEF2 antibody. The same membrane was reprobed successively with antibodies to MEF2A, MEF2D, and histone H1. B, Inhibition of Cdk5 rescued MEF2-dependent reporter activity. MEF2 reporter assay was performed as described in Figure 1 B (n = 3). mt, Mutant; wt, wild type. Error bars represent SD.

Figure 6.

Neurotoxicity induced Cdk5-dependent cleavage of MEF2. A, Expression and degradation of MEF2D in SH-SY5Y neuronal cells. Endogenous MEF2D was determined by Western blotting after hydrogen peroxide treatment (top). The membrane was reprobed for β-actin as a loading marker (bottom). B, Hydrogen peroxide induced phosphorylation of endogenous MEF2D in SH-SY5Y cells. The level and phosphorylation of MEF2D were analyzed as described by immunoblotting. C, Hydrogen peroxide-induced, Cdk5-dependent degradation of MEF2D. Expression of MEF2D and MEF2DS444A by recombinant adenoviruses is equal (top left). SH-SY5Y cells were infected with the indicated recombinant adenoviruses, and MEF2D level was determined by immunoblotting. MEF2DS444A was more resistant to hydrogen peroxide-induced degradation (middle left). SH-SY5Y cells were infected with recombinant adenoviruses as indicated and then treated with hydrogen peroxide. The levels of MEF2D, MEF2DS444A, and β-actin were analyzed by Western blotting. The bottom left graph indicates the relative level of MEF2 under various treatments (n = 3). Dominant-negative Cdk5 blocked MEF2 degradation (right). The experiment was performed as described for the middle left panel. Error bars represent SD.

Figure 7.

Cdk5-mediated phosphorylation of MEF2 facilitated caspase-3-dependent degradation of MEF2. A, Caspase-3 degraded GSTMEF2D phosphorylated by Cdk5 in vitro. Purified GSTMEF2D was incubated with or without purified Cdk5/p25 in the presence of [γ-³²P]ATP and then incubated with or without activated caspase-3 for 90 min and autoradiographed after Cdk5 kinase assay (top left). The same membrane was probed with anti-MEF2D antibody (bottom left). The right panel represents similar experiments as described in the left panel using cold ATP in kinase assay. The GSTMEF2D level after caspase-3 treatment was determined with an anti-GST antibody. The arrow indicates a fragment present only in caspase-3-treated phosphorylated GSTMEF2D. B, Endogenous MEF2D phosphorylated by Cdk5 in glutamate-treated CGNs was targeted for degradation by caspase-3. Glutamate treatment reduced MEF2D and increased phosphorylated MEF2D in CGNs. The same amounts of MEF2D from the lysates were subjected to anti-MEF2D immunoprecipitation (middle). The precipitated MEF2D was incubated with or without caspase-3 as described in A (bottom). C, Increasing Cdk5 activity was sufficient to enhance caspase-3-mediated degradation of MEF2D. SH-SY5Y cells were transiently transfected with control, MEF2D, or MEF2D plus p35/Cdk5 vectors. The susceptibility of overexpressed MEF2D to caspase-3-mediated degradation was determined as described in B. The top panel shows the same amounts of MEF2D and enhanced levels of p35 present in the lysates before caspase-3 treatment. The bottom panel shows that coexpression of p35/Cdk5 enhances degradation of MEF2D by caspase-3.

Figure 8.

MEF2C resisted glutamate-induced phosphorylation and degradation in CGNs. A, Glutamate exposure did not alter MEF2C but reduced MEF2A and MEF2D levels in CGNs. B, Glutamate exposure did not increase MEF2C phosphorylation. CGNs were treated with glutamate as described in Figure 3A. Levels of phospho-MEF2D and phospho-MEF2C were determined by phospho immunoblotting (top). The same membrane was reprobed successively for MEF2D and MEF2C. C, Expression of alternatively spliced variants of MEF2C in CGNs. The alternatively spliced exons A and B in MEF2C are diagramed (top). RT-PCR analysis of MEF2C mRNA shows the expression of the splicing variants expressed by CGN (MEF2C - b and + b are vector controls) (middle). Cdk5 kinase assay of MEF2C immunoprecipitated from CGNs (anti-MEF2C IP) was performed as described in Figure 7A. MEF2C C′ is a bacterially expressed C′ MEF2C fragment used as a positive substrate control (bottom). D, Cdk5/p25 induced phosphorylation and degradation of MEF2C + b but not MEF2C - b in HEK293 cells. Cdk5 induced phosphorylation of MEF2C + b but not MEF2C - b (top). HEK293 cells were transfected with various constructs indicated. The amounts of MEF2C and its level of phosphorylation were determined by successive immunoblotting using anti-MEF2C (indicated by MEF2C) and phospho-MEF2 antibody (indicated by p-MEF2C). The same lysates were used to analyze MEF2C levels. The amounts of protein loaded were quantified using β-actin as a control. The levels of MEF2C - b (middle) and MEF2C + b (bottom) were determined by anti-MEF2C immunoblotting. E, MEF2C - b was more resistant to glutamate-induced inhibition than MEF2C + b. CGNs were transfected with MEF2 reporter and the constructs indicated and treated with glutamate. The change of luciferase activity relative to untreated controls is presented. Error bars represent SD.

Figure 9.

Enhanced MEF2 function rescued CGNs from glutamate-induced apoptosis. A, Blocking Cdk5 rescued CGN from glutamate-induced apoptosis. CGNs were transfected with a construct encoding GFP (green), treated with glutamate, and stained with PI (red). The top panel shows exemplary image of transfected neurons unexposed (top row) or exposed (bottom row) to glutamate. The middle graph quantifies neuronal apoptosis by PI staining (n = 4). The bottom graph shows the quantitative correlation of glutamate-induced neuronal death determined by dehydrogenase activity. B, Blocking Cdk5 restored MEF2 function after glutamate treatment. CGNs were transfected and treated as indicated. MEF2 reporter assay was performed (n = 3). C, D, MEF2DS444A rescued CGN from glutamate-induced apoptosis to a greater degree than did wild-type MEF2D (C; n = 4) and enhanced MEF2 activity in reporter assay (D; n = 3). The experiments were performed as described for A and B, respectively. E, Blocking Cdk5 rescued 12 DIV CGNs from glutamate-induced apoptosis. The experiments were performed as described for the middle panel in A using CGNs after 12 d in culture (n = 3). mt, Mutant; wt, wild type. Error bars represent SD.