Granular layer

Home Lab Home
NEUROIMAGE

Combinining multiple techniques to understand neural code

CNISM

The nervous systems can learn, store and process a huge amount of information, which is encoded in terms of electrical signals called spikes and transmitted between neurons at the synapses. The mechanisms of signal coding and learning are the object of an open debate. In biophysical terms, the main question is how simple processes in neurons can generate cognitive functions and form complex memories like those experienced by humans and animals. In principle, if one would be able to record from all the neurons in a network involved in a given behavior, the route toward reconstruction of the related computations would be greatly facilitated. In fact, this is not possible with current techniques for several reasons. First, the more precise are neuronal recordings (e.g. with the patch-lamp technique) the more limited is the number of simultaneously recorded neurons (1 or 2). Conversely, global recordings (e.g. field recordings) collect activity from many neurons but do not tell much about single neuron computation. Secondly, the activity of neurons in the brain is governed by probabilistic rules and neuronal mechanisms are stochastic in nature. Multiple neurons have therefore to be recorded simultaneously and repeatedly and their correlated activity investigated.
Addressing the present issue bears the appealing opportunity of gaining new insight into the neural code. It is not difficult to imagine the fallout in scientific terms and in potential applications to artificial systems. In order to circumvent the indetermination intrinsic in neural processing and the limitations imposed by available techniques, we propose to develop a multi-level approach facing the question in the cerebellar cortex . This choice is motivated by the simplified network architecture of the cerebellum, by the impressive amount of knowledge that has been accumulated on the properties of its neurons and synapses, and by the availability of leading theories for cerebellar computation. The cerebellar network is composed of neurons, the granule cells, which are activated by numerous extra-cerebellar areas and relay information to Purkinje cells. The granule cells are inhibited through a double feed-back and feed-forward loop by Golgi cells. The computational properties of this circuit are puzzling and go far beyond the sparse coding and spatial pattern separation originally proposed by Marr, one of the fathers of computational neuroscience. The granular layer shows oscillation, resonance and is endowed with specific time-dependent properties and synaptic plasticity.

 


Our investigation exploit a strong level of integration through the European projects SENSOPAC (VI-FP) on cognition in artificial systems and CYBERRAT (VII-FP) on bi-directional bio-electronic interfaces, as well as with a PRIN project on synaptic plasticity, to which the Neurophysiology Laboratory of Pavia takes part. By integrating the contribution of the aforementioned projects, the investigation will exploit the following combination of techniques:
  • Patch-clamp recordings, to achieve high-resolution recordings of electrical activity in single neurons. This is a well established  and broadly used technique (Neher and Sackmann, 1994) that will be combined with single neuron mathematical modeling.
  • Multi-site CHIP recordings, an advanced  technique exploiting FET and CMOS technology to obtain fine-grain local field potential (LFP) and single-unit recordings, therefore potentially allowing  resolution of multiple single neurons. The pitch is about 10 mm  and the band-width >10 KHz.
  • Voltage-sensitive dye (VSD) recordings, another recent technique allowing to obtain high-resolution scanning of circuit activity over large network sections. Here, single cell resolution is not possible due to light diffraction and temporal resolution can range up to 10 KHz with an appropriate camera.
  • Second-harmonic generation (SHG) recordings. This is a very promising technique currently under development at the European Laboratory for Non-linear Spectroscopy (LENS), University of Florence which, once appropriately implemented, will give the best chance of recording multiple single neurons. 
  • Mathematical modeling of single neurons and networks. The computational properties of the network, although testable on an experimental basis, can be reconstructed only through appropriate computational models built from single neuron properties and wired following the rules revealed anatomically and functionally.

The University of Pavia Site

 

Principal Investigator: Egidio D’Angelo

Team Members: Paola Rossi, Francesca Prestori, Jonathan Mapelli, Sergio Solinas, Elisabetta Cesana, Shyam Diwakar, Daniela Gandolfi.

 

European Laboratory for Non-Linear Spectroscopy (LENS)

 

Principal Investigator: Francesco Pavone

Team Member : Leonardo Sacconi

 

Project Description

Neuroimage description

Related Publications

Experimental observations and technical development

    1.   Roggeri L, Rivieccio B, Rossi P, D’Angelo E (2008) Tactile stimulation evokes long-term synaptic plasticity in the granular layer of cerebellum”. J Neuroscience, …

    2.   L. Sacconi, J. Mapelli, D. Gandolfi, J. Lotti, R. P. O’Connor, E. D’Angelo and F. S. Pavone (2008). Optical recording of electrical activity in intact neuronal networks with random access second-harmonic generation microscopy.

Computational reconstruction of single cell behaviors

    1.   Sergio Solinas, Lia Forti, Elisabetta Cesana, Jonathan Mapelli, Erik De Schutter, Egidio D’Angelo. (2007) Fast-reset of pacemaking and theta-frequency resonance patterns in cerebellar Golgi cells. Frontiers in Cellular Neuroscience 1-4:1-9 .

    2.   Sergio Solinas, Lia Forti, Elisabetta Cesana, Jonathan Mapelli, Erik De Schutter, Egidio D’Angelo. (2007) Computational reconstruction of pacemaking and intrinsic electroresponsiveness in cerebellar Golgi cells. Frontiers in Cellular Neuroscience 1-2:1-12.

    3.   Shyam Diwakar, Jacopo Magistretti, Mitchell Goldfarb, Giovanni Naldi, Egidio D’Angelo.  Axonal Na+ channels ensure fast spike activation and back-propagation in cerebellar granule cells. J Neurophysiology, in press.

    4.   R. Carrillo, E. Ros, S. Tolu, T. Nieus, E. D’Angelo. (2007) Event-driven simulation of cerebellar granule cells. Information Processing in Cells and Tissue. …

    5.   Michele Bezzi, Angelo Arleo, Thierry Nieus, Olivier Coenen, Egidio D’Angelo. Quantitative characterization of information transmission in a single neuron.  In preparation.

Stating and testing the hypothesis

    6.   E. D’Angelo (2008). The critical role of Golgi cells in regulating spatio-temporal integration and plasticity at the cerebellum input stage. Frontiers in Cellular Neuroscience.

    7.   E.D’Angelo, C De Zeeuw (2008). Timing and plasticity in the cerebellum: focus on the granular layer. TINS.

    8.   Jesús A. Garridoa, Eduardo Ros a, Richard R. Carrilloa, Egidio D'Angelo (2007) Noise reduction and time slicing facilitated by local network topologies in the cerebellum granular layer network. Information Processing in Cells and Tissue. In preparation.