Neuron Perspective A Brief History of Neuronal Gene Expression: Regulatory Mechanisms and Cellular Consequences Zilong Qiu1 and Anirvan Ghosh1,* 1Neurobiology Section, Division of Biology, University of California, San Diego, La Jolla, CA 92093, USA *Correspondence: aghosh@ucsd.edu DOI 10.1016/j.neuron.2008.10.039 A central goal of cellular and molecular neuroscience is to explain the development and function of the ner- vous system in terms of the function of genes and proteins. Models of gene regulation have evolved from be- ing focused on transcriptional and translational control to include a variety of regulatory mechanisms such as epigenetic control, mRNA splicing, microRNAs, and local translation. Here we discuss how developments in molecular biology influenced the study of neuronal gene expression, and how this has shaped our under- standing of neuronal development and function. Introduction The enormous influence of molecular biology on our understand- ing of nervous system function is reflected in the fact that in vir- tually all areas of neuroscience research we tend to describe mechanisms in terms of genes and proteins. This has been true in developmental neurobiology for a long time and is in- creasingly true in other areas of neuroscience, such as cellular physiology and neurological disease. We routinely describe de- velopmental events in terms of transcription factors and ligand- receptor interactions. Notch and Delta, NGF and Trk, and Sonic Hedgehog and Patched are part of this new vocabulary. The phenomenon is not restricted to developmental neurobiology. It is now common to describe electrophysiological phenomena in terms of regulation of AMPA and NMDA receptors, PDZ pro- teins, TARPs, and CaM kinases. We are in an era where we gauge our understanding of the brain based on our ability to explain neurobiological phenomena in terms of the role of the under- lying genes and proteins. One could very well describe the state of affairs as a ������Neuron Effect,������ as a molecular approach to un- derstanding the nervous system has been a hallmark of papers published in the journal over the past two decades. But it would be a mistake to think that this is a view of neuroscience that has been pushed by Neuron instead, the founders of the journal and first Neuron editorial team from UCSF recognized the incredible impact that molecular biology was having on neuroscience and created a venue for publication of the most exciting work in the field. The experiment has been an unqualified success. It is useful to look back at some of the early discoveries that made molecular biology of the nervous system an area of such great fascination. The discovery of Nerve Growth Factor by Rita Levi-Montalcini and Stanley Cohen was a transforming event and highlighted the great power of understanding developmental events in terms of ligand-receptor interactions. The work from Sydney Brenner, Seymour Benzer, and colleagues illustrated the power of genetics to get to the molecular basis of neuronal function. The discovery of sensory transduction pathways, first forvisionandthenforothersensorysystems,allowedustounder- stand how we perceive the external world. The purification of pro- teinsandcloningofgenesinvolvedinsynapticvesiclereleaseand ion channels transformed the study of cellular physiology. Whereas many of the early discoveries on the molecular basis of neuronal function had their roots in biochemistry, the rapid pace of discovery in molecular biology and the accompanying understanding of gene regulation has driven many of the ad- vances in the past two decades. The central dogma had taught us that genes are encoded in DNA, that DNA was transcribed into mRNA, and that mRNA was translated into protein. Molecu- lar investigations of gene regulation revealed a host of regulatory mechanisms that dramatically expand the ways in which a cell can regulate its protein composition. Not only is the transcription of many genes tightly regulated, but splicing, trafficking, and translation of mRNA can also be exquisitely controlled, which al- lows for incredibly precise control over protein levels and local- ization. In a cell as complex as a neuron, these gene regulatory mechanisms are widely used to facilitate proper development and function of the nervous system and allow the nervous sys- tem to adapt to changes in the environment. In this Perspective we discuss a few examples to illustrate how the discovery of gene regulatory mechanisms over the past 20 years has been closely linked to the emergence of major ideas regarding brain development and function. Stimulus-Dependent Transcription and Neuronal Adaptive Responses Although the relationship between genes and proteins was de- scribed in the 1940s, the first study to show that extracellular sig- nals could acutely regulate eukaryotic gene expression was a 1984 paper by Greenberg and Ziff where they reported that the proto-oncogene c-fos was rapidly induced bygrowth factor stim- ulation of 3T3 cells (Greenberg and Ziff, 1984). This study had a major impact, since it changed the concept of gene regulation from an autonomous property of cells to a process that was an integral part of a cell���s response to changes in the environment. Greenberg, Greene, and Ziff, as well as Curran and Morgan, went on to show that NGF stimulation of PC12 cells also led to the rapid induction of c-fos expression, which suggested that such dynamic regulation of gene expression might be a feature of the nervous system (Greenberg et al., 1985 Curran and Morgan, 1985). Shortly thereafter, Greenberg and colleagues reported that stimulation of PC12 cells with Acetylcholine led to c-fos expression and that this required calcium influx via Neuron 60, November 6, 2008 ��2008 Elsevier Inc. 449

voltage-sensitive calcium channels (VSCC) (Greenberg et al., 1986). This was a critical study, as it showed that neurotransmit- ter-induced calcium influx, which had previously thought only to exert acute effects, could lead to a rapid and robust transcrip- tional response. Thus, in a short period of 2 years, the concept of activity-dependent regulation of gene expression became es- tablished, which had major implications for activity-dependent development and function of the nervous system. A separate line of investigation from Goodman, Montminy, and colleagues, who were studying cAMP regulation of gene expression, led to the identification of CREB, a key transcription factor that mediates stimulus-dependent transcription. In 1986 they reported that cAMP regulated somatostatin mRNA levels and identified a cAMP-responsive element (CRE), which was sufficient to confer cAMP responsiveness (Montminy et al., 1986a, 1986b). Montminy and colleagues isolated the transcrip- tion factor that binds to the CRE and named it cAMP-respon- sive element binding protein (CREB) (Montminy and Bilezikjian, 1987). They showed that elevation in cAMP led to phosphoryla- tion of CREB at Ser-133, and that this modification was re- quired for transcription activation by CREB (Gonzalez and Montminy, 1989). In the meantime, Greenberg and colleagues showed that calcium-dependent induction of c-fos expression was mediated by a calcium-responsive element that also bound CREB (Sheng et al., 1991). Thus CREB was identified as a key mediator of cAMP- and calcium-dependent transcrip- tion in neurons. The Role of CREB A series of observations in the 1990s implicated CREB-mediated transcription as a critical mediator of adaptive responses in the nervous system. One area of active investigation was the poten- tial role of CREB in memory. A study from the Benzer lab had im- plicated cAMP signaling in learning and memory (Dudai et al., 1976), and Kandel and colleagues had shown that synaptic plas- ticity in Aplysia required cAMP signaling, but it was not clear how cAMP signaling might be connected with memory. An intriguing possibility was that cAMP might exert its effects by regulating gene expression, which was supported by pharmacological studies from the 1960s and 1970s from the Flexner, Agranoff, and Barondes labs that suggested that gene expression and protein synthesis were required for the retention of memory (re- viewed in Davis and Squire, 1984). Following the identification of the CRE by Montiminy and colleagues, Kandel���s group showed that injection of a CRE-containing DNA fragment impaired long-term plasticity in Aplysia (Dash et al., 1990), which sug- gested that CRE-mediated gene expression was likely to be important for long-term memory. Kandel and colleagues continued to investigate the role of cAMP signaling in plasticity and reported that cAMP stimulation of hippocampal slices mimics the late phase of long-term poten- tiation (LTP) (Frey et al., 1993). Shortly thereafter, Tully and col- leagues reported that a dominant-negative form of CREB blocks long-term memory in Drosophila, and Silva and colleagues re- ported that mice carrying a mutation in CREB had deficient long-term memory (Yin et al., 1994 Bourtchuladze et al., 1994). While these studies built support for the idea that CREB might play an important role in memory, it was difficult to know if this pathway had a specific role in memory or whether these molecules played a more general role in mediating neuronal re- sponses to environmental changes. Investigation of the role of CREB in other systems suggested that CREB was unlikely to be selectively involved in memory, but rather was likely to be generally involved in mediating long- term neuronal responses to external stimuli. An important set of observations came from Eric Nestler and his colleagues, who examined the role of CREB in addiction. They showed that morphine administration reduces CREB phosphorylation in the rat locus coeruleus, and that opiate receptor antagonists increased CREB phosphorylation (Guitart et al., 1992). Subse- quent work from the Nestler group showed that modulation of CREB could regulate the response to cocaine (Carlezon et al., 1998), and Malenka and colleagues showed that CREB regulates excitability of nucleus accumbens neurons, another structure implicated in cocaine addiction (Dong et al., 2006). In separate studies Ginty, Greenberg, and colleagues showed that CREB phosphorylation in the suprachiasmatic nucleus was regulated by light, and Ginty and colleagues showed that NGF- induced signaling to CREB was important for the cell survival effects of NGF (Ginty et al., 1993 Riccio et al., 1997, 1999). Ghosh and colleagues showed that CREB was involved in activ- ity-dependent dendritic growth (Redmond et al., 2002), and work from the Malenka group showed that CREB activity could regu- late the number of silent synapses (Marie et al., 2005). These observations indicated that CREB-mediated transcription was likely to be involved in regulating a diverse set of neuronal responses. While much of the early investigation of calcium-dependent transcription and its consequences was focused on CREB, it is now clear that calcium signaling targets a number of different transcription factors that mediate different cellular effects of cal- cium signaling. The Lipton, Greenberg, and Bonni labs identified MEF2 as a calcium-regulated transcription factor in neurons and showed that MEF2 was involved in mediating activity-dependent survival and in regulating excitatory synapse number (Leifer et al., 1993 Mao et al., 1999 Flavell et al., 2006 Shalizi et al., 2006). In an effort to identify novel calcium-dependent transcrip- tion factors, Ghosh and colleagues developed a new screen called Transactivator Trap and identified a set of new calcium- regulated transactivators (Aizawa et al., 2004). The first of these factors was CREST, which was shown to be involved in mediat- ing activity-dependent dendritic growth (Aizawa et al., 2004). Two other factors cloned in this screen were NeuroD2 and LMO4, both of which are involved in barrel cortex development (Kashani et al., 2006 Ince-Dunn et al., 2006). It now appears that changes in neuronal activity in response to extracellular sig- nals can lead to the activation of a large number of transcription factors. Some of them, such as CREB, may have a general role in neuronal adaptive responses, whereas others may have more specific roles in mediating specific aspects of activity-depen- dent development and plasticity. Activity-Regulated Genes Ever since c-fos was identified as a calcium-regulated gene, there has been an interest in identifying genes whose expression is regulated by neuronal activity. The earliest in vivo evidence of activity-dependent regulation of gene expression came from Morgan and colleagues, who showed that c-fos and other 450 Neuron 60, November 6, 2008 ��2008 Elsevier Inc. Neuron Perspective