When stimulation is distributed over all available bipolar cells, but locally weaker, suppression is less effective and gain stays high. Furthermore, this local gain control can be viewed as a dynamic process; it affects the later part of the spike burst, but not its initial phase, which determines the first-spike latency. In the following, we test neuronal mechanisms that may implement such a dynamic local gain control mechanism. A first candidate mechanism for local gain control in

homogeneity detectors is synaptic depression at bipolar cell terminals. Indeed, bipolar cell signals can display substantial depression (Burrone and Lagnado, CH5424802 cell line 2000 and Singer and Diamond, 2006), which could partly

suppress responses to strong local activation. When activation is distributed over more bipolar cells, on the other hand, as in the case of homogeneous receptive field activation, synaptic depression is likely to be less effective and thus should permit longer spike bursts. We therefore tested whether homogeneity detectors are cells with particularly strong local adaptation, as would result from synaptic depression. To do so, we used a stimulus that aimed at predepressing synapses in one half of the receptive field. We assessed the effect of this predepression on the iso-rate curves by a brief activation of one receptive field half shortly before each stimulus of the iso-rate-curve http://www.selleckchem.com/products/chir-99021-ct99021-hcl.html measurement (Figure 6A). As expected, the predepression stimulus reduced sensitivity of the ganglion cells, which is reflected by the increased radius of the iso-rate curves

(Figures 6B and 6C) as compared to the control condition without the predepression stimulus. The reduction in sensitivity may L-NAME HCl contain both global and local components; a symmetric scaling of the predepressed iso-rate-curve radius along all directions reflects a global loss in sensitivity, whereas an asymmetric scaling provides evidence for a local loss in sensitivity and thus a local adaptation mechanism. If the nonconvex iso-rate curves of the homogeneity detectors were to result from particularly strong synaptic depression, this asymmetric scaling should be particularly strong for these cells. However, this was not supported by the experimental data. In fact, homogeneity detectors typically displayed rather global adaptation effects and less local sensitivity loss (Figure 6C) than cells with a convex iso-rate curve (Figure 6B). Synaptic depression is thus not a plausible mechanism for the particular features of homogeneity detectors. As an alternative model, we explored whether local inhibitory signaling could mediate a local gain control.