RESULTADOS VALOR PERCIBIDO POR EL EMPLEADO

gradient, which drives GABAa and GABAc receptor currents, is controlled in retinal

bipolar cells. Here I review previous work on this topic.

The membrane potential of a cell depends on the concentrations of ions on either side of the membrane and the relative permeability of the membrane to those ions. Whenever ion channels open, the permeability of the membrane to the relevant ion increases, and the membrane potential shifts towards the reversal potential of that ion. As mentioned in section 1.4, ionotropic GABA receptors conduct anions when activated, in vivo mainly chloride and bicarbonate (Bormann et al,, 1987). The

permeability to bicarbonate depends on the subunit composition of the GABA receptor, but is in most cases about 0.1 to 0.3 of that to chloride (for example Kaila, 1994; Feigenspan and Bormann, 1998). So when GABA receptor ion channels open, the membrane potential (Vm) shifts towards the reversal potential of the GABA evoked conductance ( E q a b a ) , which is between the reversal potential for chloride (Eci-) and the reversal potential for bicarbonate ( E h c o s - ) » due to the higher permeability for chloride it is much closer to Eci- than to E h c o 3 - The GABA response is hyperpolarizing when E q a b a is more negative, and depolarizing when it is more positive than Vm. Additionally, GABA can also evoke biphasic responses consisting of a hyperpolarization followed by a depolarization (these biphasic responses required the presence of bicarbonate in the solution), for example during intense neuronal activity or the prolonged presence of GABA. This means that E q a b a shifts from being more negative than Vm to being more positive, which may be explained by chloride accumulating inside the cell during prolonged channel opening, along with a significant bicarbonate permeability to the channel (Staley et al., 1995; Backus et al., 1998; Dallwig et al., 1999; Freeh et al., 1999).

During development GABA responses are often excitatory (for example Cherubini et al., 1990; Cherubini et al., 1991; Chen et al., 1996; Owens et a l, 1996), i.e. E q a b a is significantly more positive than V m , and can depolarize the membrane potential sufficiently to activate voltage gated calcium channels (Reichling et al.

1994; Owens et al., 1996). Calcium entry evoked by such GABA evoked depolarization is believed to be important for neuronal proliferation, migration and synaptogenesis. In the adult animal GABA responses are usually inhibitory (Kaila, 1994): this is either due to Ega b a being more negative than Vm (so that GABA hyperpolarizes the membrane and moves the membrane potential further away from the threshold of many voltage gated cation channels) or due to E q a b a being close to (only slightly more negative or positive) or equal to Vm (so that GABA stabilizes the membrane potential at or around its resting value, this is termed “shunting inhibition”). The developmental shift of E q a b a appears to be mainly due to a shift of Eq], with the intracellular chloride concentration ([Cfji) being higher in younger animals. This has been proposed for some time but was only recently directly demonstrated in cells of mammalian brain slices using the gramicidin perforated patch-clamp technique (for example: Owens et al., 1996; Ehrlich et al., 1999; Kakazu et al., 1999).

The intracellular chloride concentration in neurones is usually not determined by passive chloride movement through chloride conducting channels, but is regulated by specific chloride transporters, such as the Cl'-ATPase (extruding chloride), the Cl* /HCOa'-exchanger (usually accumulating chloride), the Na'*’-dependent Cl'/HCO]'- exchanger (usually extruding chloride) and transporters of the electroneutral cation- chloride co-transporter family (Kaila, 1994; Delpire, 2000). The latter family has seven members, but the two most important ones for neurophysiology appear to be the sodium-potassium-chloride cotransporter NKCC-1 and the brain specific potassium-chloride co-transporter KCC-2 (Payne et al., 1996), both of which have been shown to control [Cl*]i in a variety of neurones (for example Kakazu et al., 1999; Rivera et al., 1999; DeFazio et al., 2000; Sung et al., 2000; Hubner et al., 2001b; Martina et al., 2001). During each transport cycle, NKCC-1 transports 1 sodium, 1 potassium and 2 chloride ions across the membrane (Russell, 2000), whereas KCC-2 co-transports 1 potassium with each chloride ion (Payne, 1997).

Thus, both transporters are electroneutral. During physiological conditions the ion concentrations are such that NKCC-1 accumulates chloride, whereas KCC-2 extrudes it from the cell (see Figure 4.6B for the energy changes associated with chloride transport by these carriers). Thus the presence of NKCC-1 in the cell membrane could aid in depolarizing E q a b a , whereas the presence of KCC- 2 could aid in hyperpolarizing GABA responses. Interestingly the expression for NKCC-1 and KCC-2 show cellular, subcellular and developmental differences. The overall brain expression levels of NKCC-1 are high at birth and decline during early postnatal development (Plotkin et al, 1997). The presence of this transporter in the neurones of young animals explains their high internal chloride level, more positive E q a b a and depolarizing, excitatory GABA responses. By contrast, KCC-2 levels are low at birth and increase during postnatal development (Lu et a l, 1999). The increase in KCC-2 expression was shown to coincide with, and be responsible for, the switch from depolarizing to hyperpolarizing GABA responses, and is regulated by the extracellular GABA concentration (Rivera et a l, 1999; Ganguly et al, 2001). NKCC-1 can be found in neurones and glial cells, whereas KCC-2 is expressed only by neurones (Payne et al, 1996; Hubner et a l, 2001a). When present in the same cell, NKCC-1 and KCC-2 can be found at different subcellular locations; for example NKCC-1 was found mainly in the dendrites, whereas KCC-2 was found mainly in the axon terminals of retinal rod (ON) bipolar cells, suggesting compartmentalized chloride concentration in this neurone (Vardi et al, 2000a; Vu et a l, 2000). As described in detail in section 4.1, this distribution could make Eqi more positive at the dendrites than at the synaptic terminals of the bipolar cells, which would be appropriate for generating the observed receptive field structure and light response transience in these cells. In Chapter 4, I examine whether the intracellular chloride concentration is, as suggested by the chloride transporter distribution, different in the dendrites and the axon terminal of retinal rod bipolar cells.