ACTIVATING TRANSCRIPTION FACTOR 3 UP-REGULATED BY c-Jun NH 2 -TERMINAL KINASE/c-Jun CONTRIBUTES TO APOPTOSIS INDUCED BY POTASSIUM DEPRIVATION IN CEREBELLAR GRANULE NEURONS Y. MEI, a,b Z. YUAN, a,b B. SONG, a,b D. LI, a C. MA, a C. HU, a Y.-P. CHING c AND M. LI a,b * a Department of Pharmacology, Zhongshan School of Medicine, Sun Yat-sen University, 74 Zhongshan Road II, Guangzhou 510080, China b Proteomics Center, Zhongshan School of Medicine, Sun Yat-sen University, 74 Zhongshan Road II, Guangzhou 510080, China c Department of Anatomy, University of Hong Kong, 21 Sassoon Road, Pokfulam, Hong Kong, China Abstract—Cerebellar granule neurons (CGNs) depend on po- tassium depolarization for survival and undergo apoptosis when deprived of depolarizing concentration of potassium. Activating transcription factor 3 (ATF3), a stress-inducible protein, belongs to the ATF/CREB family of transcription factors family and is involved in cell growth and apoptosis. However, the role of ATF3 in neuronal apoptosis remains unknown. Here, we showed that ATF3 was up-regulated un- der potassium deprivation in CGNs, and this induction was preceded by a rapid and sustained activation of c-Jun NH 2 - terminal kinase/c-Jun signaling pathway, which plays a fun- damental role in neuronal apoptosis. Furthermore, ATF3 up- regulation was abolished by inhibition of JNK or knockdown of c-Jun. Finally, knockdown of ATF3 by RNA interference protected CGNs from potassium deprivation–induced apo- ptosis. Taken together, our results indicate that ATF3 is a downstream target of JNK/c-Jun pathway and contributes to apoptosis induced by potassium deprivation in rat CGNs. © 2007 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: ATF3, JNK, c-Jun, apoptosis, cerebellar granule neurons. Activating transcriptional factor 3 (ATF3) is first identified as a transcriptional factor binding to ATF/CRE sites (Hai et al., 1989). As an immediate early regulated gene, ATF3 is rapidly induced upon a wide variety of stress stimuli includ- ing genotoxic agents, ischemia and hypoxia (Yin et al., 1997; Hai et al., 1999; Tsujino et al., 2000; Murata et al., 2006). ATF3 is also shown to be involved in various cel- lular activities such as stress response, cell cycle progres- sion, apoptosis and immune regulation (Hai and Hartman, 2001; Hartman et al., 2004; Pan et al., 2005; Yan et al., 2005; Gilchrist et al., 2006). Accumulating evidences suggest that ATF3 plays a critical role in apoptosis. However, in non-neuronal cells, both pro- and anti-apoptotic functions of ATF3 have been demonstrated. When over-expressed, ATF3 induces apo- ptosis in ovarian cells (Syed et al., 2005) and increases the sensibility of HeLa cells to apoptotic stimuli such as eto- poside or camptothecin (Mashima et al., 2001). Further- more, knocking out ATF3 in embryonic fibroblasts can partially protect cells from genotoxic stress-induced apo- ptosis (Lu et al., 2006), indicating the pro-apoptotic role of ATF3. On the other hand, ATF3 also possesses anti- apoptotic abilities. For example, over-expressed ATF3 can protect endothelial cells from tumor-necrosis factor (TNF) –induced apoptosis (Kawauchi et al., 2002) and rescues cardiac myocytes from apoptosis induced by doxorubicin (Nobori et al., 2002). The dual roles of ATF3 in apoptosis may be explained by its special functions in transcriptional regulation. ATF3 forms homodimer with itself or with other bZip proteins. Homodimeric form of ATF3 functions as a transcriptional repressor (Chen et al., 1994), whereas het- erodimeric form of ATF3 functions as a transcriptional repressor or activator (Hsu et al., 1992; Chu et al., 1994; Nilsson et al., 1997). Thus, the role of ATF3 in apoptosis is largely dependent on the cell context, the complex form and stimuli. So over-expression strategy alone may not reveal the exact function of endogenous ATF3 under cer- tain conditions. In variety of neuronal apoptosis or injury models, ATF3 was reported to be up-regulated (Ohba et al., 2003; Huang et al., 2005; Lindwall and Kanje, 2005), and the induction of ATF3 has even been used as a marker for nerve injury (Tsujino et al., 2000). However, only a few studies have been aimed to investigate the role of ATF3 in neuronal apoptosis. By adenovirus-mediated over-expression anal- ysis, ATF3 was found to prevent superior cervical ganglion neuron from apoptosis induced by nerve growth factor (NGF) withdrawal (Nakagomi et al., 2003) and suppress hippocampal neuron apoptosis induced by kainic acid (Francis et al., 2004), indicating that over-expressed ex- ogenous ATF3 plays an anti-apoptotic role in neuronal apoptosis. Up until now the anti-apoptotic function of ATF3 has been merely observed in research based on such over-expression strategy. However, the role of endoge- *Correspondence to: M. Li, Department of Pharmacology, Zhongshan School of Medicine, Sun Yat-sen University, 74 Zhongshan Road II, Guangzhou 510080, China. Tel: +86-20-87331553; fax: +86-20- 87331653. E-mail address: limt@mail.sysu.edu.cn (M. Li). Abbreviations: ATF3, activating transcriptional factor 3; BP, band pass; CGNs, cerebellar granule neurons; CMV, cytomegalovirus; DIV, days in vitro; DMSO, dimethylsulfoxide; EGFP, enhanced green fluo- rescent protein; -Gal, -galactosidase; JNK, c-Jun NH 2 -terminal ki- nase; MLK, mixed lineage kinase; NGF, nerve growth factor; RNAi, RNA interference; 5 K, serum-free medium containing 5 mM KCl; 25 K, serum-free medium containing 25 mM KCl; 25 K+S, 10% fetal bovine serum and 25 mM KCl. Neuroscience 151 (2008) 771–779 0306-4522/08$32.00+0.00 © 2007 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2007.10.057 771