Mitogen- and stress-activated protein kinase 1 mediates cAMP response element-binding protein phosphorylation and activation by neurotrophins

Abstract

Activation of the transcription factor cAMP response element-binding protein (CREB) by neurotrophins is believed to regulate the survival, differentiation, and maturation of neurons in the CNS and PNS. Although phosphorylation of Ser133 is critical for the expression of CREB-regulated genes, the identity of neurotrophin-regulated Ser133 kinases has remained controversial. We show here that neurotrophin-induced CREB phosphorylation in CNS neurons depends exclusively on the extracellular signal-regulated kinase 1/2-activated kinase mitogen- and stress-activated protein kinase 1 (MSK1). Small interfering RNA directed against ribosomal S6 kinase 1 (RSK1) and RSK2 reduced phosphorylation of a RSK substrate but did not effect CREB-dependent transcription. However, expression of a selective inhibitory MSK1 mutant markedly attenuated BDNF-stimulated CREB phosphorylation and CREB-mediated transcription. Moreover, the ability of neurotrophins to stimulate CREB phosphorylation was abolished in CNS neurons from MSK1 knock-out mice. Consistent with a role for MSK1 in Ser133 phosphorylation, neurotrophin-induced expression of CREB-regulated genes was attenuated in MSK-deficient neurons. These results indicate that MSK1 is the major neurotrophin-activated Ser133 kinase in CNS neurons.

Figure 1.

Neurotrophin-induced CREB phosphorylation is mediated by an ERK1/2-activated kinase. A, B, Cortical neurons were pretreated with 10 μm KN62 or vehicle and stimulated with BDNF or NMDA for the indicated times (in minutes). Protein extracts were immunoblotted for phospho-cAMP response element-binding protein, ERK2, phospho-CaMKIV, or CaMKIV. C, Cortical neurons were pretreated with 10 μm LY294002 or vehicle and stimulated with BDNF for the indicated times (in minutes). Protein extracts were immunoblotted for phospho-cAMP response element-binding protein, phospho-Akt, and total CREB. D, Rat cortical neurons were treated with 50 ng/ml BDNF or vehicle for 5 min. Akt1 was immunoprecipitated and assayed for kinase activity against GST-cAMP response element-binding protein or GST-Bad. The kinase reactions were separated by SDS-PAGE, and the bands corresponding to the substrate were autoradiographed. Similar results were seen in three experiments. E, Cortical neurons were pretreated with 10 μm PD184352 or vehicle and stimulated with BDNF or NT3 for the indicated times (in minutes). Protein extracts were immunoblotted for phospho-CREB, phospho-ERK, phospho-RSK1/2, phospho-Akt, phospho-ERK5, and total ERK2. F,G, Cortical neurons were pretreated with 5 μm SB202190 or vehicle and stimulated with BDNF or sodium arsenite (As) for the indicated times (in minutes). Protein extracts were immunoblotted for phospho-cAMP response element-binding protein, phospho-HSP27, ERK2, and HSP27.

Figure 2.

Activation of CREB by neurotrophins does not depend on RSK. A, Rat PC12 cells were transfected with empty vector (EV) or kinase-dead RSK2 mutants (RSK2 K100R) and stimulated with NGF for 15 min. Protein extracts were immunoblotted for phospho-MSK1, RSK2, and MSK1. B, Cortical neurons were electroporated with RSK1 (siRSK1) and RSK2 (siRSK2) siRNA DNAs. The cells were stimulated with BDNF (50 ng/ml), and protein extracts were immunoblotted for RSK2 and pCREB. C, Cortical neurons were transfected with a GFP expression construct and an eightfold excess of RSK1 (siRSK1) and RSK2 siRNA (siRSK2) DNAs. The cells were fixed and immunostained for total Rsk. Green fluorescence is GFP, and red fluorescence is staining for RSK. D, Cortical neurons were transfected with HA-tagged LKB1 and a fourfold excess of RSK1 (siRSK1) and RSK2 siRNA (siRSK2) DNAs. LKB1 was immunoprecipitated with an anti-HA Ab, and immunoprecipitates were immunoblotted for phospho-LKB1 Ser431. E, Cortical neurons were transfected with a CRE-regulated luciferase reporter and an eightfold excess of empty vector, siRNA vector RSK1 (siRSK1), or siRNA vector RSK2 siRNA (siRSK2). Neurons were treated with 50 ng/ml BDNF or 10 μm forskolin (forsk) and assayed for luciferase activity. The data are quadruplicate determinations, and the error is SEM. Unless otherwise noted, all data are representative examples of at least three experiments.

Figure 3.

MSK1 is a neurotrophin-regulated CREB kinase. A, Rat cortical neurons were pretreated with 10 μm PD184352 or vehicle and stimulated with 50 ng/ml BDNF (B) or NT3 (N) for the indicated times (in minutes). Protein extracts were immunoblotted for phospho-MSK1 and total MSK1. B, Cortical neurons were transfected with GFP (green) and HA-tagged MSK1. Cells were immunostained for the HA epitope (red) and were also stained with the DNA dye, Hoechst 33258 (blue). C, Mouse cortical neurons were treated with BDNF for 15 min. MSK1 and RSK2 were immunoprecipitated and assayed for kinase activity against full-length CREB. The data are quadruplicate determinations, and the error is SEM.

Figure 4.

Expression of a dominant-negative MSK1 mutant attenuated neurotrophin-induced CREB-dependent transcription. A, PC12 cells were transfected with empty vector (EV) or an HA-tagged MSK1 construct and an excess of either empty vector or kinase-dead MSK1 (MSK1 KD). HA-tagged MSK1 was immunoprecipitated and assayed for kinase activity using purified CREB. Where indicated, PC12 cells were stimulated with 100 ng/ml NGF for 15 min. B, Cortical neurons were electroporated with GFP or kinase-dead MSK1 DNAs and stimulated with BDNF or forskolin (forsk; 10 μm) for the indicated times (in minutes). Protein extracts were immunoblotted for phospho-cAMP response element-binding protein and total CREB. C, Cortical neurons were transfected with a CRE-regulated luciferase reporter and a 10-fold excess of empty vector or kinase-dead MSK1 (MSK1 KD). Neurons were treated with 50 ng/ml BDNF or 10 μm forskolin and assayed for luciferase activity. The data are quadruplicate determinations, the error is SEM, and asterisks denote p < 0.01.

Figure 5.

Neurotrophin-induced CREB phosphorylation is absent in cortical neurons deficient for MSK1. A, Cortical neurons from wild-type or MSK1/2 double knock-outs were treated with 50 ng/ml BDNF (B) or 10 μm forskolin (F) for the indicated times (in minutes). Extracts were immunoblotted for the indicated epitopes. B, Cortical neurons from wild-type mice, MSK1/2 double knock-outs, MSK1 knock-outs, or MSK2 knock-outs were treated with 50 ng/ml BDNF for the indicated times (in minutes). Extracts were immunoblotted for the indicated epitopes. C, Tissue from the indicated brain regions or cultures was used to prepare cDNA. Real-time quantitative PCR was used to determine mRNA levels for MSK1 and MSK2 in these brain regions.

Figure 6.

Neurotrophin-induced expression of CREB-regulated genes is impaired in MSK-deficient cortical neurons. A, Mouse cortical neurons were transfected with fos-luciferase, Mkp1-luciferase, or Nurr1-luciferase, and a 10-fold excess of empty vector or ACREB DNA. The neurons were treated with vehicle or BDNF for 3 hr and assayed for luciferase activity. The data were normalized for CAG-promoter driven β-galactosidase activity. The activity of the CAG promoter is not affected by expression of ACREB. B, Cortical neurons from WT and MSK1 –/– mice were transfected with GalCREB and Gal-luciferase DNAs. The cells were stimulated with forskolin or BDNF for 3 hr and assayed for luciferase activity as in A. C–E, Mouse cortical neurons from WT or MSK1/2 knock-outs were cultured and treated with BDNF, NT3, or forskolin for the indicated times (in minutes). cDNA was prepared, and real-time quantitative PCR was used to determine mRNA levels for c-fos, Mkp1, and Nurr1. The data were normalized to 18s RNA levels also determined by real-time PCR. Error bars denote SEM, and asterisks denote p < 0.01 (n = 4–10).