Neurons have the capacity to undergo activity-dependent changes in their molecular composi­tion and structure in order to adjust their synaptic strength. Such synaptic plasticity underlies learning and memory and dysfunctions of synaptic plasticity contribute to brain diseases such as epileptogenesis, responses to ischemia, drug addiction, and neuropsychiatric disorders. Since enduring forms of synaptic plasticity, like long-term potentiation (LTP) and long-term memory require activity-dependent gene induction that is important in defining neuronal connectivity in the brain, it is anticipated that many forms of mental disabilities, including neurodegenera­tive processes and cognitive disturbances will be understood as cortical or limbic cognates of disturbed activity-dependent gene transcription. We therefore have focussed much attention on the identification and functional characteriza­tion of the specific genes that are induced by patterned synaptic activity. In the past we used differential screening and subtractive cloning strategies to identify the first activity-regulated genes (e.g. Nature (1993) 361, 453-457; Proc. Natl. Acad. Sci. USA (1995) 92, 5734-5738; EMBO J. (1999) 18, 3359-3369 and EMBO J. (1999) 18, 20, 5528-5539). More recently we have been using transcriptional profiling tech­nologies to monitor changes in mRNA expres­sion on a whole-genome scale in an unbiased way by comparing gene-expression before and at several time points after neuronal stimulation (Nature (2010) 465:182-187; PLoS One (2013) 8:e76903). This allowed us in a first step to establish a comprehensive library of activity-regulated genes. These genes are then searched for common elements, such as being present in the same pathway, sharing interactions or functions, or having similar DNA-motifs in their promoter region. In this way we can identify pathways of genes involved in synaptic plasticity and tran­scription factors mediating the activity-dependent expression programs. Moreover, by comparing knock-out animals with wild-type animals, or by using time-resolved measurements we are now beginning to unveil causal upstream-downstream relations between these genes. As these tech­niques allow us to discover novel pathways that so far have been elusive, we begin to understand the neuron specific genomic response to synaptic activity.

Small G-Protein (AR-Arl) Daniel Mensching, Ora Ohana, Guido Hermey

One of the genes we identified in the whole genome Chip survey is a novel activity-regulated gene encoding an ADP-ribosylation factor-like small G-Protein (AR-Arl). Transcription of AR-Arl is increased in the hippocampus 1 and 2 hours after kainic acid induced seizures. Moreover, 1 hour after in vivo LTP stimulation AR-Arl transcript levels are elevated in the dentate gyrus. We found that the small GTPase AR-Arl shares many char­acteristics with proto-typical Arf proteins. We demonstrate that N-terminal myristoylation of AR-Arl is required for GTP-dependent attach­ment to membranes of the Golgi apparatus and endosomes. Moreover, we can induce the recruit­ment of AR-Arl to vesicles by Brain-Derived- Neurotrophic-Factor (BDNF) and AR-Arl co-localizes with the BDNF receptor TrkB in late endosomes. In addition, we find that AR-Arl func­tions in long-range axonal retrograde endosomal transport. A dominant-negative AR-Arl variant deficient in GTP-mediated activation impairedfast retrograde trafficking of late endosomes in cultured hippocampal neurons. Conditional AR-Arl knock-out mice that we generated exhibit a reduced activation of ribosomal protein S6, indicating an impairment of BDNF induced translation required for synaptic plasticity. In line with these observations, AR-Arl-deficient mice show a distinct memory deficit and spent less time exploring an unknown object in the novel object recognition task compared to wild type littermates.

SorCS1 Guido Hermey, Sandra Oetjen, Abuzar Kaleem

SorCS1 belongs to the Vps10p-Domain receptor family, which defines a group of receptors binding neuropeptides and trophic factors (Cell. Mol. Life Sci. (2009) 66, 2677-2689). SorCS1 has been related to the genetically complex disorders type 2 diabetes and Alzheimer’s disease (Ann. Neurol., (2011) 69, 8-10). We identified SorCS1 and the highly homologous SorCS3 as activity-regulated genes and showed that SorCS1, SorCS2, and SorCS3 are expressed in a combinatorial mostly non-overlapping pattern in the developing and adult nervous system (J. Neurochem. (2004) 88, 1470- 1476; PLOS One (2013) 8 (10), e76903; J. Comp. Neurol. (2014) 522, 3386-402). One shared function among Vps10p-Domain recep­tors is their proteolytic processing and our recent studies demonstrate that alternative processing can adjust receptor activity. Thus, differential processing of SorCS1 modulates the interac­tion with the receptor Sortilin and thereby its function (Biochem. J. (2014) 457, 277-288). In addition SorCS2 exhibits disparate func­tions in neurons and glia depending on its cell specific proteolytic processing (Neuron, (2014) 82, 1074-87). Another proposed shared char­acteristic of Vps10p-Domain receptors is the sorting and intracellular targeting of ligands. In agreement, we localized SorCS3 to hippocampal dendrites and postsynaptic vesicles and demon­strated that it conveys endocytosis of ligands (J. Comp. Neurol. (2014) 522, 3386-402). Currently we extend our studies by analyzing how adaptor proteins identified by us convey activity depen­dent dendritic targeting of SorCS1 and SorCS3.

Sgk1 Ralf Scholz, Claudia Mahlke

Using subtractive cloning strategies we identified the serum and glucocorticoid-inducible kinase 1 (Sgk1) as an activity-regulated gene in the brain.
Based on the constitutive knockout mouse model generated in our laboratory (J. Clin. Invest. (2002) 110, 1263-1268), we found that the Sgk1 gene product has a very short half-life (Biochem J. (2006), 399, 69-76) and lack of SGK1 is causative for a variety of non-neuronal deficits. In collabo­ration with Florian Lang’s group (University of Tübingen) we identified defects in renal function (e. g. Am. J. Physiol. Renal Physiol. (2009) 297, 704-712), mast cell activation (J. Immunology (2009) 183, 4395-4402), hemostasis (Blood (2012), 119, 251-261) and muscle homeostasis (EMBO Mol. Med. (2013), 5, 80-91) in Sgk1 null mice. In the brain of wild type animals we observed an activity-dependent induction of Sgk1 in two different cell types. In oligodendrocytes induction is dependent on glucocorticoid release, whereas in dentate granule neurons transcriptional activation is independent of glucocorticoids but strictly dependent on synaptic activity. Our initial behavioral studies using complete knockout animals revealed strongly reduced locomotor and exploratory activity. To dissect the cellular basis for the behavioral phenotype and exclude non-neuronal influence we generated conditional Sgk1 knockout mice. These animals will allow us to analyze the consequences of cell type specific deletions of Sgk1. In a complementary approach, we set out to identify novel interaction partners to elucidate the molecular function of SGK1 in the brain. Employing the yeast two-hybrid system to screen mouse cDNA libraries, we identified several interaction candidates. Selected interac­tion partners were validated biochemically and currently we proceed to understand the molecular function underlying these interactions and link them to phenotypes found in the Sgk1 knockout mice.

Arc/Arg3.1 Lars Binkle, Xiaosong Mao, Jerome Gruhlich, Jakob Gutzmann, Joachim Nowock, Guido Hermey, Uwe Borgmeyer, Tiemo Marquarding, Johanne Kläschen, Karin Kähler, Lilianna Kucharczyk, Xiaoyan Gao, Sergio Castro Gomez, Jasper Grendel, Francesca Xompero, Ora Ohana

Among activity-dependent genes Arc/Arg3.1 stands out. Only few show the exquisite regula­tion and breadth of functional importance as this immediate early gene. Arc/Arg3.1 was discov­ered
in our laboratory and independently of us by Paul Worley and colleagues. The implica­tions from the original discovery of Arc/Arg3.1 have now been borne out in studies establishing a function for the protein in multiple forms of protein synthesis-dependent synaptic plasticity, regardless of the polarity of change. Mice in which we have disrupted the Arc/Arg3.1 gene show altered synaptic plasticity and severe defi­cits in hippocampus-dependent and -independent cognitive tasks, which require the consolidation of newly encoded memories (Neuron (2006) 52, 437-444; Nat. Neurosci. (2010) 13:1082-1089; Neuron (2011) 69:437-444). Further evidence demonstrates that the expression of Arc/Arg3.1 is important for homeostatic synaptic scaling (Proc. Natl. Acad. Sci. USA (2011) 108: 816-821;. J. Neurosci. (2010) 30:7168-7178). Conversely, aberrant Arc/Arg3.1 expression has been impli­cated in psychiatric and neurodegenerative diseases, including Alzheimer’s disease (Cell (2011) 147:615-628).

Functional consequences of the subcel­lular localization of Arc/Arg3.1 Xiaosong Mao, Joachim Nowock, Tiemo Marquarding, Johanne Kläschen, Karin Kähler, Lilianna Kucharczyk, Francesca Xompero, Ora Ohana

Previous experiments established a strong link between gene expression and physiological and pathological neuronal plasticity; however, it remains an open question how transcriptional activation taking place in the nucleus can selec­tively modify stimulated synaptic sites in the distant dendritic compartment of the neuron. Such selective modifications of synapses that have experienced coincident activity are required by the Hebbian rule and might be a prerequi­site for input specificity of LTP. The analysis of Arc/Arg3.1 might guide our thinking and provide insights into this problem. Most strik­ingly, following LTP-producing stimulation Arc/Arg3.1 mRNA is localized to the dendrites of neurons that received patterned synaptic activity. Consequently, Arc/Arg3.1 mRNA may be locally translated at activated synapses and may have a key role in synapse specific modifications during plastic events in the brain. To test this hypothesis we have generated a phage artificial chromosome harboring mutations that completely abolish the targeting of Arc/Arg3.1 mRNA but leave all other properties intact. Mice carrying these muta­tions have plasticity and memory deficits. The impairments are severe but distinct from those observed in the complete Arc/Arg3.1 Ko mice. In a complementary approach we generated knock-in mice in which Arc/Arg3.1 protein is only somatically and dendritically localized but not found in the nucleus anymore. The consequences of this mutation on behavior and physiology are currently under investigation.

Arc/Arg3.1 function Lars Binkle, Jakob Gutzmann, Uwe Borgmeyer, Guido Hermey, Francesca Xompero, Xiaoyan Gao, Sergio Castro Gomez, Ora Ohana

We find that Arc/Arg3.1 can regulate AMPA receptor trafficking by binding to proteins of the endocytic machinery. To get a more complete understanding of the post-synaptic protein networks Arc/Arg3.1 interacts with, we gener­ated TAP-tagged animals and conducted conven­tional as well as split ubiquitin Y2H screens. These screens yielded several proteins that are resident in the endosomal system. This is in agreement with the proposed function of Arc/ Arg3.1 in endocytosis and might help to explain the versatile role of Arc/Arg3.1 in synaptic plas­ticity. We focused our attention on a new trans­membrane protein and a so far uncharacterized sorting nexin. Members of this large protein family are of central importance in regulating the endosomal sorting of membrane cargo. As a first step to further study the functional impor­tance of the identified interactions on receptor trafficking and plasticity we will virally express binding-deficient mutants in neurons of condi­tional sorting nexin mice that we have recently generated.

Conditional Arc/Arg3.1 KO mice Xiaoyan Gao, Sergio Castro-Gomez, Jasper Grendel, Francesca Xompero, Ora Ohana

Following memory encoding and retrieval, Arc/ Arg3.1 expression increases in various areas of the hippocampus and cortex, suggesting their mutual contribution to memory formation. In addition, we have recently found that Arc/Arg3.1 is expressed in the brain during the first 4 weeks of the critical period of development. To dissect the spatial- and temporal roles of Arc/Arg3.1 in development and in adult memory-formation we generated conditional Arc/Arg3.1 (cKO) mice and remove Arc/Arg3.1 in specific brain regions or time points by injecting rAAV- Cre viruses or by breeding with Cre-transgenic mice. We investigate the impact of Arc/Arg3.1 ablation on memory performance, network structure and function and on synaptic plasticity in the brain.

Functional circuits in primary sensory cortex Ora Ohana

Sensory information arriving in the cortex is encoded and processed by complex algorithms in the local cortical circuitry. These algorithms require distinct interactions between excitatory and inhibitory neurons and between different layers of the cortex. These interactions had been the focus of our investigations for the last decade. In particular we focused on the thala­mocortical-corticothalamic circuitry entailing L4, L6 and L5 (J Physiol (1998) 513, 135-48 ; J Neurophysiol (2008) 100, 1909-1929; PLoS One ( 2012) 7 e40601). We study these interac­tions using several complementary techniques: multi-electrode patch clamp recordings from cortical neurons in acute slices of sensory cortex, glutamate uncaging, 3-D reconstructions of the recorded neurons and computational modelling. In the future, we plan to further investigate the connectivity within and between these layers and their contribution to specific behaviors.

The several findings described above open up new avenues and pave the way to investigate mechanisms of plasticity, or when disturbed are the cause of mental diseases, psychiatric disor­ders or play roles e.g. in addiction, epileptogen­esis, ischemia, and Alzheimer disease. The main focus of our research, however, will remain on the analysis of learning and memory. Much prog­ress has been made, within discrete levels of analysis, characterizing biophysical, molecular and cellular adaptations associated with plasticity and cognitive functions. However, it has proven difficult to integrate these findings and translate the specific knowledge at each level into an under­standing of informa­tion processing and storage. A long-term goal of our research is to elucidate how mental functions emerge from specific changes at molecular levels. We see the use of mouse genetics as an important means of building bridges between molecular biology and systems neurobiology and between systems neurobiology and behavior. This provides the rationale for an integrated approach to follow the flow of information from excitatory events in the dendrite through neural networks in behaving animals. We hope in this way to extract some of the fundamental rules that govern dendritic information processing in the activity-driven refinement of networks that underlies learning and memory.