Nonlinearities in retinal bipolar cells shape the encoding of artificial and natural stimuli

The retina dissects the visual scene into parallel feature channels, and bipolar cells are speculated to play a key role in this signal separation. Yet, bipolar cells are traditionally viewed as simple, linear neurons. Here, using the salamander retina, we investigated the hypothesis of linear signal processing in bipolar cells by intracellularly recording their voltage signals under artificial and natural visual stimuli. We observed nonlinear representation of contrast and, unexpectedly, also nonlinear spatial integration in a sizable fraction of bipolar cells. Furthermore, linear receptive field models fail to describe responses of nonlinear bipolar cells to spatially structured artificial and natural stimuli. The nonlinear properties occur in the receptive field center and may be cell-type specific, with stronger effects in transient than sustained bipolar cells. Thus, our data suggest that nonlinear signal pooling starts earlier than previously thought, that is, before signal integration in bipolar cells.