Relative contribution of rod and cone inputs to bipolar cells and ganglion cells in the tiger salamander retina.

1. The relative contribution of rod and cone inputs to bipolar and ganglion cells were studied by comparing the response-irradiance relations, spectral sensitivities, and response waveforms of these neurons recorded from the isolated, flat-mounted tiger salamander retina under dark-adapted conditions. 2. Bipolar cells could be differentiated both on the basis of the polarity of the light response and on their relative rod/cone input. Thus some depolarizing bipolar cells appeared more strongly influenced by rod input (DBCR), whereas others were more influenced by cone input (DBCC). Similarly, hyperpolarizing bipolar cells could be divided into those that received rod-dominant input (HBCR) or cone-dominant input (HBCC). 3. The light onset response of sustained-ON ganglion cells reflected both rod-dominant input from DBCRs and cone-dominant input from DBCCs. 4. OFF ganglion cells displayed both a rod-dominant sustained light offset response and a cone-dominant transient light offset response, suggesting input from both HBCRs and HBCCs. 5. In ON-OFF ganglion cells, the light onset response was strongly rod dominated and was presumably mediated by DBCRs, whereas the light offset response displayed both rod and cone influence, suggesting input from HBCRs and HBCCs. The contribution of cones to the light onset response of ON-OFF ganglion cells was only observed in the presence of a rod-adapting background light. 6. A suppression of the light offset responses of OFF and ON-OFF ganglion cells was observed, which was dependent both on the wavelength and irradiance of the light stimulus. 7. These results indicate that the photoreceptor inputs to bipolar cells in the tiger salamander retina are segregated such that they form separate rod-dominant and cone-dominant pathways. Thus the response properties of the different types of ganglion cells are influenced not only by the excitatory and inhibitory inputs they receive from the bipolar and amacrine cells but also whether these inputs are provided through rod-dominant or cone-dominant pathways. The functional implications of these findings are discussed.