Elio Raviola, MD

Theories of brain function as well as interpretations of electrophysiological experiments are suffering from incomplete knowledge of the microcircuitry of the nervous system. In this respect, the retina is a privileged organ, because its physical location, the distinctive morphology of its neurons, the regularity of its architecture, and the properties of its inputs and outputs favor a unique variety of experimental approaches that are not possible elsewhere in the brain. In addition, the retina is one of the few sites in which the shape of neurons is the visible signature of its connections, because each neuronal type has a unique morphology, dictated by the rigorous stratification of their synaptic connections and the need for ordered sampling of the visual world. During the past few years, as a result of the efforts of a number of laboratories including ours, a complete inventory of the anatomical types of horizontal, bipolar, amacrine and ganglion cells has become available, at least for the retina of a few species of mammals. The problem now is to learn the contribution of each type of neuron to the internal economy of the retina or its coding of the visual world. In tackling such a task, first and foremost, one has to identify in the living state the cell type that is the object of physiological or molecular studies. To solve the problem of cell type identification, we pioneered the use of genetic methods to express a visible marker in specific neuronal populations of the mouse retina. As an object of our experiments, we chose retinal dopaminergic neurons (DA cells), which fulfill a fundamental role in vision because they release dopamine, a catecholamine modulator responsible for many of the events that lead to neural adaptation to light. We targeted PLAP to this cell type by linking its cDNA to a regulatory sequence of the gene that codes for tyrosine hydroxylase, the rate-limiting enzyme for dopamine biosynthesis.

In these transgenic animals, we studied the synaptic connections of DA cells and investigated their ligand- and voltage-gated channels with the patch clamp technique. We observed that DA cells possess a pacemaker activity and thus fire in vitro action potentials in a slow, rhythmic pattern, a property also found in other catecholaminergic neurons of the CNS. We studied the constellation of voltage-gated channels responsible for this behavior, as well as the effects of GABA, glycine and glutamate/kainate on the firing pattern. To analyze the composition of the GABAA receptors of DA cells, we developed an RT-PCR assay to identify the transcripts of the various GABAA receptor subunits at a single cell level. We showed that the cells express multiple subunits of the GABAA receptor, assembled into multiple receptors, and we analyzed by patch clamp their pharmacological properties. Next, we studied dopamine release in solitary DA cells by amperometry with carbon fiber electrodes and showed that, as a result of the spontaneous generation of action potentials, quanta of dopamine are released by exocytosis over the surface of the cell body. Since the perikaryon of DA cells does not contain presynaptic active zones, this release is by necessity extrasynaptic and represents one of the sources of the dopamine that acts by volume transmission on distant targets in the retina. In addition to paracrine secretion of dopamine, we have recently shown that DA cells also release GABA extrasynaptically. It is known that DA cells establish synapses onto AII amacrine cells, a neuron that transfers rod signals to cone bipolar endings. By using triple-label immunocytochemistry and confocal microscopy, we showed that these contacts are probably GABAergic, because GABAA receptors are clustered at the postsynaptic active zone. Another accomplishment of our laboratory was the development of a technique to study global gene expression in single DA cells. Solitary DA cells were patch clamped and individually harvested. Their mRNA was amplified by a novel method, SMARTT7, based on the combination of the PCR-based SMART technique with the T7 RNA polymerase. The single cell probes were then used to screen the original RIKEN cDNA microarrays. We found that DA cells secrete insulin, the neuropeptide CART, the cytokine interferon a and the chemokine monocyte chemoattractant protein-1. They contain the COP9/Signalosome, a 500 kDa nuclear protein complex that acts as a transcriptional repressor in the cascade of events regulated by light during seedling development: it was the first time that the presence of this macromolecular complex was described in the nervous system and its functional significance remains to be elucidated. Finally, DA cells contain the most common circadian clock-related proteins, supporting the idea that DA cells have a role in the retinal internal clock.

Thus, DA cells appear to carry out four main functions in the retina, each characterized by a different time course: 1) through their fast GABAergic synapses on AII amacrine cells, DA cells control the transfer of rod signals to ganglion cells on a time scale of the order of the millisecond. 2) They release dopamine over their entire surface. This modulator acts at a distance by volume transmission on a large number of retinal neurons, presiding over the process of transition from the dark-adapted to the light-adapted state over a time scale of seconds to minutes. 3) They contain the most common circadian clock-related proteins suggesting a role in the circadian regulation of retinal function over a time scale of hours.

"During the past few years, as a result of the efforts of a number of laboratories including ours, a complete inventory of the anatomical types of horizontal, bipolar, amacrine and ganglion cells has become available, at least for the retina of a few species of mammals."