The Silver Lab
UCL logo

Synaptic transmission, neural computation and information processing in neuronal networks

Our lab combines in vitro and in vivo experimental approaches with transgenic technologies, advanced analysis methods and computer modelling to study brain function. Experimental approaches we use include patch-clamp recordings, dynamic clamp, functional and structural 2-photon imaging, uncaging and FRAP (fluorescence recovery after photobleaching). These approaches are often combined with statistical analysis and diffusion-reaction modelling to study synaptic function. Neuronal anatomy, electrophysiology, high speed 3D acousto-optic lens (AOL) 2-photon imaging (Microscopy) and uncaging are also combined with compartmental neuronal modelling (Neuroinformatics) approaches to study synaptic integration. While 3D AOL 2-photon imaging and network modelling are combined to study information processing at the network level. The selected examples below highlight how these interdisciplinary approaches have advanced our understanding of brain function...

Synaptic transmission: We have investigated how sensory signals cross the first synapse in the cerebellar cortex. We combined patch-clamp recordings of synaptic currents with quantal statistical analysis (Silver 2003) and models of short term synaptic plasticity to establish that cerebellar mossy fibre terminals have an unusually large resource of vesicles (300 per site) and that vesicles can be docked and primed much faster (~12 ms) than previously thought for central synapses (Saviane and Silver 2006). By developing a quantitative glutamate uncaging method that combines photolysis with 3D reaction-diffusion modelling, we characterized AMPA receptor desensitization within the synapse (Digregorio et al. 2007). This showed that synaptic AMPARs are more resistant to desensitization than previously thought, allowing them to convert glutamate arising from spillover into current without desensitizing. These pre- and post-synaptic properties allow the cerebellar mossy fibre-granule cell synapse to signal over an unusually wide bandwidth. We have recently extended this work to the molecular level, in collaboration with Stefan Hallermann, Eckart Gundelfinger and Jens Eilers, by establishing that the synaptic protein Bassoon underlies this fast reloading of vesicles (Hallermann et al. 2010). We have also investigated how vestibular information is encoded by mossy fibre synapses, in collaboration with Troy Margrie, by performing in vivo patch-clamp recordings from granule cells (Arenz et al. 2008). This demonstrated that angular velocity is linearly related to synaptic charge.
Neural computation: We have investigated the mechanisms underlying synaptic integration using patch-clamp recordings, dynamic clamp and modelling approaches. We have shown that inhibition can perform neuronal gain modulation, allowing rate-coded signals to be multiplicatively scaled - an essential neural computation (Mitchell and Silver 2003). Prior to this it was thought that shunting inhibition has a purely subtractive effect on firing rate. Counter intuitively the mathematical operations performed by inhibition depend on the properties of excitatory synaptic input. More recently, we have shown that short-term depression in excitatory synaptic inputs converts additive operations into multiplicative operations (Rothman et al. 2009). Combining depressing excitatory inputs with inhibition therefore provides a powerful mechanism for controlling neuronal gain and multiplying together synaptic input.
Network processing: We established that a functional feed-forward inhibitory circuit is present within the input layer of the cerebellar cortex and that Golgi cells appear tuned to extract temporal information from the sensory input (Kanichay and Silver 2008). We used paired patch calmp recordings, anatomical methods (in collaboration with Zoltan Nusser's lab) and network modelling to examine how electrically coupled networks of inhibitory interneurons respond to excitatory synaptic input (Vervaeke et al. 2010). This lead to two major discoveries: 1) gap junction potentials can have an inhibitory action rather than an excitatory one, 2) sparse synaptic excitation of inhibitory interneuron networks can trigger desynchronization - a poorly understood network phenomenon. This showed that electrical synapses can either promote network synchronization or trigger rapid desynchronization depending on the synaptic input.
See Publications for more studies from the Silver Lab.

This page last modified 28 March 2011
University College London - Gower Street - London - WC1E 6BT - Telephone: +44 (0)20 7679 2013 - Copyright © 1999-2011 UCL
Disclaimer | Accessibility | Privacy | Advanced Search | Help



Search by Google