ERK mediates activity-dependent neuronal complexity via sustained activity and CREB-mediated signaling

Abstract

The pattern of neuronal dendritic arbor reflects numerous functional consequences including electrophysiological properties and connectivity of the neuron. Cortical neurons grown in culture conserve dendrite morphology of neurons in vivo, and they display distinct morphological responses to KCl-induced depolarization. Extracellular signal-regulated kinase (ERK) signaling and subsequent cAMP response element-binding protein (CREB)-mediated gene expression are important mediators of activity-dependent dendrite growth. Interestingly, individual cortical neurons displayed a wide range of (ERK) phosphorylation after KC1 stimulation. To determine if ERK activation mediates distinct morphological responses to depolarization, the dendritic morphology of GFP transfected cortical neurons with high and low levels of pERK intensity was examined after KC1 stimulation. Both nonpyramidal and pyramidal neurons with high pERK intensity showed greater total dendritic length, branch point number, and primary process number than those with low pERK. However, the length of the longest process was not significantly different between high and low pERK neurons. These data suggest that neurons translate different levels of ERK signaling into distinct dendrite complexity by regulating specific aspects of the complexity. This observation further raised a question of ERK signaling duration during depolarization. Immunostaining of time-course KC1 stimulated neurons revealed that neurons with pERK staining after depolarization had sustained ERK activation during the stimulation. The protein level of c-Fos, which is an immediate early gene that indicates sustained ERK signaling, was maintained during prolonged KC1 stimulation, and most pERK positive neurons expressed c-Fos. The c-Fos protein level was suppressed by blocking ERK signaling hours after stimulation. Blocking sustained ERK signaling using the same method inhibited primary dendrite formation after depolarization. In addition, increasing the duration and level of ERK phosphorylation with dominant negative MAP kinase phosphatase-1 increased dendrite complexity. These data indicate that sustained ERK signaling is required and sufficient for the formation of dendrite complexity. To ascertain if the difference in ERK activity and morphology is mediated by nuclear signaling mechanisms, the activation of nuclear downstream targets was examined. p90 ribosomal S6 kinase (RSK) and cAMP response element (CRE)-binding protein (CREB) were phosphorylated at active sites in high pERK neurons. Further, CRE-mediated gene expression was activated in high pERK neurons when measured using the dual promoter reporter construct, SCCG. Furthermore, high pERK neurons in which CRE-mediated gene expression was inhibited with dominant negative CREB (KCREB) showed decreased dendrite complexity when compared to high pERK neurons without KCREB. Interestingly, the level of gene expression was different between neurons. Neurons with high levels of CRE-mediated transcription showed greater dendrite complexity, with increased total dendritic length and branch point number, than neurons with low levels of CRE-mediated transcription. Taken together, these data indicate that ERK signaling contributes to the regulation of activity-dependent dendrite complexity via two mechanisms: induction of CRE-mediated gene expression and stabilization of the gene products expressed.