We are not aware that tuning functions with a triphasic form have been described before in a sensory neuron. A switch in the polarity of the synaptic output of bipolar cells is especially surprising because the electrical response in the soma is determined by the type of glutamate receptor sensing transmitter release from photoreceptors: HIF-1�� pathway a metabotropic receptor in ON cells and an ionotropic receptor in OFFs (Masland, 2001). We therefore investigated synaptic tuning curves in bipolar cells by imaging

a second variable reflecting signal transmission—the calcium signal driving neurotransmitter release. These experiments were carried out using a line of transgenic zebrafish expressing SyGCaMP2 (Dreosti et al., 2009). Use of the ribeye promoter described in Figure 1 allowed us to localize TNF-alpha inhibitor expression of SyGCaMP2

to ribbon synapses. Figure 6G shows examples of responses from individual ON and OFF bipolar cell terminals stimulated with steps of light over the same intensity range used in experiments employing sypHy. The top two traces provide examples of sustained ON cells that generate transient OFF responses at the highest luminance tested (arrowed); the next trace is an OFF cell in which the tuning curve passes through a maximum, and the bottom trace is an example of an OFF cell that generates ON responses at the lowest intensities (arrowed). Collected results using SyGCaMP2 are shown in Figures 6H and 6I and are expanded on in Figures S4, S5C, and S5D (using 100 ON synaptic terminals and 39 OFF). These tuning curves were constructed using the same general approach applied to sypHy measurements, except that the response was quantified as the initial rate of change of SyGCaMP2 fluorescence

normalized to the baseline. The tuning curves of linear (49%) and nonlinear (51%) terminals were described well by Equation 3, with shape parameters σ and h very similar to those estimated by assessing TCL the exocytic response using sypHy (cf. Figures 6C and 6D). How do the “linear” and “nonlinear” tuning curves affect the encoding of a sensory stimulus? A useful way to frame this question is to ask how many different levels of luminance (NL) might be discriminated by observing the output of the bipolar cell terminal, taking into account the variability inherent in the process of synaptic transmission (Jackman et al., 2009 and Smith and Dhingra, 2009). At many synapses, including ribbon synapses of bipolar cells, vesicle release follows Poisson statistics, with a variance equal to the mean (Katz and Miledi, 1972, Laughlin, 1989, Freed, 2000a and Freed, 2000b).