With the generous support of the ISN the 1st Westerburg Meeting on Spinogenesis and SynapticPlasticity was organized by the Dept. ofNeurochemistry / Mol. Biology, Leibniz Institute for Neurobiology, Magdeburg, Germany.
The venue took place in Germany's is Germany's oldest and best preserved moated castle, the Westerburg. The number of participants was limited to 100, ensuring a very interesting small conference with intense discussions in a rather uncommon atmosphere. A total of 29 oral presentations were delivered and 20 posters presented to an interested and alert audience thathelped to make this symposium a very succesful meeting. Presentations were focussed on the assembly and disassembly of excitatory and inhibitory synapses and the molecular machinery that controls the number, shape and plasticity of synapses. During formation of synaptic connections, neurons form contacts via trans-cellular adhesive interactions that are mediated by cell adhesion molecules (CAMs). Y. Yamaguchi (San Diego, CA) presented data showing that dendritic protrusions of cultured hippocampal neurons remain filopodial and do not transform into mature spines without functional ephrin-EphB2 receptor signaling. A reduction in the number of spines is found in triple knockout (TKO) mice lacking EphB1-3 receptors. An underlying molecular mechanism might be that EphB receptors are required for tyrosine phosphorylation of the postsynaptic heparin sulfate proteoglycan (HSPG) syndecan-2. Phosphorylation of syndecan-2 is required for clustering which in turn can regulate the of spines.
The importance of trans-synaptic adhesive interactions was further demonstrated in the
presentation of T. Sakisaka / Y. Takai (Osaka, Japan). During development, nectins (namely presynaptic nectin-1 and postsynaptic nectin-3) co-localize in adherens puncta and synapses.
Hence, inhibition of nectin based adhesion decreases the size and number of synaptic contacts in cultured neurons. Moreover, A. Dityatev (Hamburg, Germany) presented data showing that a polysialylated form of the neural cell adhesion molecule NCAM promotes formation of synapses during early synaptogenesis and is required for formation of perforated spine synapses associated with NMDA-receptor dependent LTP. Moreover, the effect requires signaling via fibroblast growth factor and NMDA receptors. The importance of PSA-NCAM for synaptic plasticity is further supported by impaired LTP and LTD in mice deficient in NCAM or PSA. A severe synaptic phenotype was also reported by M. Missler (Göttingen, Germany) for KO mice deficient for three a¨- neurexins. Neurexins bind to postsynaptic CAMs such as neuroligins and syndecans.
Interestingly the TKO mice exhibit a drastic reduction in spontaneous transmitter release with an altered amplitude and use-dependent modulation of evoked EPSCs. Furthermore, it was suggested that a-neurexins couple N-type Ca2+-channels to synaptic vesicle exocytosis. Effective synaptic transmission requires that neurotransmitters are from the extracellular space. In glycinergic synapses these functions are performed by glial and presynaptic transporterd GlyT1 and GlyT2, as was shown by V. Eulenburg (Frankfurt, Geremany). Loss of their functions leads to severe phenotypes and the death of mutant mice. C. Seidenbecher (Magdeburg, Germany) presented data from double KO mice deficient in the extracellular matrix proteoglycans neurocan and brevican that show complete loss of LTP but apparently normal learning.
Synapse formation requires the assembly of large protein complexes which crucially depend on the targeting of mulit-domain scaffolding proteins. E. Gundelfinger (Magdeburg, Germany) presented evidence that that the cytomatrix of the active zone (CAZ) is transported preassembled in a dense core transport vesicle containing two giant scaffolding proteins, Piccolo and Bassoon, to axon terminals. These vesicles carry a number of known components of the CAZ and their vesicular nature implies that maturation of the presynaptic terminals occurs in a quantal manner, with 3-5 PTVs being sufficient to form a functional presynapse. In accord, mice lacking large parts of Bassoon have a severe synaptic phenotype with the majority of brain synapses being inactive and a rapid onset of generalized epileptic seizures. Moreover, synaptic ribbons are not correctly anchored at the retinal photoreceptor synapse, which leads to an impairment in the processing of visual information.
The assembly of the postsynaptic cytoskeleton is less well understood. It is thought that vesicular transport, local protein synthesis and assembly under the control of local signaling molecules contribute in parallel since evidence for corresponding transport vesicles is lacking. In inhibitory synapses, synaptic accumulation of GABAA receptors depends on gephyrin that interacts with the dynein light chain-1 and possibly is transported to synapses by dynein motors (M. Kneussel, Hamburg, Germany). In accord several talks focussed on how assembly and synaptic targeting of scaffolding molecules is brought about in excitatory synapses. In Drosophila, PDZ domains are important to target the disc large (Dlg) protein to the neuromuscular junction (U. Thomas; Magdeburg, Germany). Targeting to the postsynaptic membrane can also require lipid modifications as shown for PSD-Zip 70 (K. Sobue, Osaka, Japan). Synaptic localization of SAP90/PSD-95 is regulated by palmitoylation which is carried out by palmitoyl-acyl-transferases (PATs) (M. Fukata / D. Bredt, San Francisco, CA). Fukata and co-workers could also identify the PAT responsible for membrane anchoring of SAP90/PSD-95 in dendritic spines. Sala and coworkers (Milano, Italy) demonstrated that ProSAPs / Shanks, a group of proteins that are thought to be master scaffolding molecules of the PSD proteins need a complex network of intra- and intermolecular interactions, e.g. between SAP90/ PSD-95, GKAP and ProSAPs / Shanks to induce clustering of these molecules.
The shape and function of spines is determined by small G-proteins from the rho family (including rho, rac, cdc42), as well as rap. PIX (E. Kim; KAIST, South Korea) and kalirin-7 (R. Mains; Farmington, CT) are postsynaptic GEFs for rac and cdc42 which are anchored at the PSD via binding to PDZ domains that have an important role for spine morphogenesis. For kalirin-7 it a contribution to dopamine signaling in chronic cocaine exposure was shown, resulting in an increased number of spines in dopamine target neurons. Further work by Kim (KAIST, South Korea) Kreienkamp (Hamburg, Germany) and Böckers (Ulm, Germany) showed that ProSAPs / Shanks provide a scaffold for G-protein signaling. Shank targets the rac/cdc42 specific exchange factor PIX and the associated kinase PAK to synapses. In addition, shank also binds IRSp53, another effector of cdc42. IRSp53 promotes actin rearrangements which enhance the number of dendritic spines or filopodia-like protrusions. A second interaction of IRSp53 via its C-terminus with SAP90/PSD-95 connects two major PSD-scaffolds, SAP90/PSD-95 and ProSAP/ Shank. Böckers described a linkage of shank to Spar, which is a GTPase activating protein for rap. This interaction involves the intermediate ProSAP interacting protein ProSAPiP. As stressed by Sobue, Spar interacts with PSD- Zip70 a member of the ProSAPiP family, actin filaments and regulates spine actin dynamics, being necessary for spine maturation. The intricate connection between NMDA receptors and regulation of the spine cytoskeleton was further illuminated by A. Matus (Basel, Switzerland) who showed that receptor stimulation leads to a loss of actin motility in spines, due to a translocation of the actin-regulatory profilin from the dendrite to spine heads.
Konnerth and co-workers (Munich, Germany) investigated patterns of synaptic activity during early brain development. Ca2+ imaging in cortical slices revealed the regular appearance of Ca2+ waves propagating through the cortex. Similar recordings in mice confirmed that that these waves are generated in vivo during sleep-like episodes. Synaptic activity may induce bi-directional changes, i.e. either long-term potentiation or depression. The leading hypothesis so far is that the magnitude and duration of changes in intracellular Ca2+ determine the outcome of synaptic activity.
M. Sheng (Cambridge, MA) challenged this view showing that pharmacological block of NMDA receptors containing the NMDA receptor subunit NR2B blocks LTD but not LTP, whereas inhibition of NR2A-containing receptors has opposite effects.
Long-term changes of synaptic efficiency eventually require changes in synaptic structure which depend on regulation of gene expression. protein synthesis. A pivotal role in this process has been attributed to the transcription factor CREB which translocates to the nucleus after enhanced synaptic activity. Expression of an constitutively active kinase results in an increase of glutamate receptor-mediated currents, LTP and the number of dendritic spines (R. Malenka, Stanford, CA).
Another mechanism was presented by M. Kreutz (Magdeburg, Germany) who identified an interaction between the postsynaptic Ca2+-sensor caldendrin and the novel protein Jacob. Activation of NMDA receptors results in a nuclear accumulation of Jacob that is negatively controlled by its binding to caldendrin and leads to a rapid stripping of synaptic contacts and a simplification of the dendritic tree. Another signaling pathway to the nucleus of postsynaptic cells in Drosophila was presented by V. Budnik (Amherst, MA). Presynaptically secreted Wnt proteins binds to the postsynaptic receptor Frizzled, resulting in the phosphorylation of Dishevelled. Phosphorylated Dishevelled prevents the formation of the Axin-Conducin-GSK3ß complex, which would otherwise promote the degradation of ß-catenin/Armadillo. Thus, Wnt stabilizes ß-catenin that is required for Lef/Tcf transcription factors to initiate transcription of Wnt responsive genes important in synapse formation and stabilization.
Another mechanism that might be part of this scheme is the transport of mRNAs into dendrites and local translation beneath synapses (D. Kuhl, Berlin, Germany; S. Kindler, Hamburg, Germany). Kindler showed that transport is mediated by the 3' untranslated regions of mRNAs binding of proteins to specific targeting elements. Kuhl highlighted the role of dendritically synthesized arg3.1 protein in synaptic plasticity. Mice lacking arg3.1 develop a deficit in the maintenance of LTP, which correlates with subtle performance deficits in a spatial learning paradigm. The role of local translational control was further emohasized by Jerry Yin (Madison, WI). Atypical PKC was shown to be a pivotal player in synaptic plasticity and learning in Drosophila probably mediated by controlling protein synthesis in a highly spatio-temporal manner.
Synaptic activity also regulates the synaptic localization of neurotransmitter receptors. D. Choquet (Bordeaux, France) reported that synaptically located AMPA receptors have on average a much lower coefficients of diffusion than extrasynaptic receptors moving long distances before being trapped at synaptic sites. The low mobility of synaptic GluR2 receptors depends on the expression of stargazin. Complementary data on the development of the neuromuscular junction in Drosophila were presented by S. Sigrist (Göttingen, Germany). New synapses in close neighborhood of preformed counterparts (resembling multiple spine boutons increasing in number after induction of LTP in the hippocampus) are mainly assembled from new receptors rather than from receptors derived from the preformed postsynaptic regions. Interestingly, overexpression of different subunits of glutamate receptors differentially affect the size of such junctions. S. Moss (Philadelphia, PA) investigated the lateral mobility of GABAA receptors and showed that protein knockdown of gephyrin reduces clustering of GABAA receptors probably by limiting their lateral mobility. GABAA receptors are rapidly subjected to endocytosis and that association with the Huntingtin-associated protein HAP1 determines whether receptors recycle back to the cell surface or are degraded in lysosomes.

Account of expenses from ISN-support

Young speaker (age below <38)
Dr. M. Fukata (UCSF, San Francisco, CA) 450.-€ (Conf. fee, accomodation)
Dr. S. Sigrist (ENI, Göttingen, Germany) 500.-€ (Trav. Exp. Conf. fee, accomod.)
Dr. T. Sakisaka (Osaka University, Osaka, Japan) 450.-€ (Conf. fee, accomodation)
Dr. E. Kim (KAIST, South Korea) 1.900.-€ (Trav. Exp., Conf. fee, accomod.)
Dr. M. Kneussel (ZMBH, Hamburg, Germany) 550.- € (Trav. Exp.,Conf. fee, accomod.)
Dr. V. Eulenburg (MPI, Frankfurt, Germany) 600.- € (Trav. Exp.,Conf. fee, accomod.)
Dr. C. Sala (Telethon, Milan, Italy) 50.- € (Trav. Exp., Conf. fee, accomod.)

Young poster presenter (age below <30)
Regina Dahlhaus (IfN, Magdeburg, Germany) 160.-€ (Conf. fee, accomodation)
Veronica Saavedra (Univ. Santiago de Chile, Chile)160.-€ (Conf. fee, accomodation)
Rodrigo Sandoval (Univ. de Chile, Santiago, Chile)160.-€ (Conf. fee, accomodation)
Gwendlyn Kollmorgen (Univ. Hamburg, Germany)160.-€ (Conf. fee, accomodation)
Magdalena Blazejzyk (IIMB Warsaw, Poland) 160.-€ (Conf. fee, accomodation)
Adam Sobzack (IIMB Warsaw, Poland) 160.-€ (Conf. fee, accomodation)

Total expenses: 6.060.-€ / roughly 7.500.- US $

Reduction of conference fee from 200.- to 100.-€

Dr. Michael R. Kreutz