Determinants of Signal Characteristics in a Retinal Network: Contribution of Local Field Potentials and Spike Trains.

点击翻译按钮获取中文摘要

The teamwork research taking place in my lab share the common objective of a better comprehension of how the neural activity of the various cell populations of the retina translates a visual stimulus into a neural code that can be understood by the brain. We have been using for many years the electroretinogram (ERG), a composite signal collected non-invasively at the surface of the cornea, to decipher the relative contribution of different cell population. More recently, we have gained expertise in the multi electrode array (MEA), a technology that allows the ex-vivo (isolated retina) recording of both the ERG-like signal and the spiking activity of the ganglion cells, the population giving rise to the efferent message of the retina.  Using the classical pharmacopeia of neurotransmitter a-/antagonists, we have documented many types of interactions occurring in the composite ERG signal, particularly at the level of the RGCs and the bipolar cells, two major contributors to the ERG signal. More recently, we have developed a innovative immunological method to selectively target specific isoforms of the Voltage-Gated Sodium Channels. This novel method shall allow us to decipher the role of dendritic integration and spike generation in the overall retinal processing. It is our intention in the next few years to capitalize on this new technique to further our knowledge on how integration takes place in response to diffuse changes in light intensities (serial integration) and how various parts of a visual scene interplay (lateral integration) to form a meaningful efferent message. We will also start investigating how the simple integrative plexus that constitutes the retina is capable of processing the astronomical amount of information generated by a visual scene. For that purpose, we will analyze the retinal activity captured in parallel by the 64 recording channels of the MEA and investigate the data by applying energy-saving processes algorithms such as transfer entropy, coherence estimation or dimension reduction and population activity time-course to the neural code generated by complex stimuli.