Receptive field center-surround interactions mediate context-dependent spatial contrast encoding in the retina

Abstract

SummaryAlmost every neuron in the early visual system has some form of an antagonistic receptive field surround. But the function of the surround is poorly understood, especially under naturalistic stimulus conditions. Anatomical and functional characterization of the surround of retinal ganglion cells suggests that surround signals combine with the excitatory feedforward pathway upstream of an important synaptic nonlinearity between bipolar cells and postsynaptic ganglion cells. This circuit architecture suggests at least two unexplored consequences for receptive field structure. First, center and surround inputs should interact nonlinearly, even prior to the summation across visual space that forms the full receptive field center. Second, inputs to the surround may alter the sensitivity of the center to spatial contrast in a scene. We test these hypotheses using both synthetic and naturalistic visual stimuli, and find that the statistics of natural scenes promote these nonlinear interactions. This work shows that, unexpectedly, the surround modulates spatial contrast encoding based on visual context.

Inference of Nonlinear Spatial Subunits in Primate Retina with Spike-Triggered Clustering

“…First, the assumptions of the SNMF algorithm are not directly matched to the structure of the subunit response model: the present model assumes nonnegative subunit outputs, while the SNMF factorization method assumes that the subunit filters themselves are nonnegative. The assumption of nonnegative filters is inconsistent with suppressive surrounds previously reported in retinal bipolar cells ( Dacey et al, 2000 ;Fahey & Burkhardt 2003;Turner et al, 2018), and limits the application of the method to other systems such as V1 (Figure 8). However, a recent modification of the SNMF method applies non-negativity constraints on subunit weights instead of the subunit filters (Jia et al, 2018), making it more suitable for other applications.…”

“…Second, the assumption of space-time separable RGC subunits significantly reduces the number of parameters that need to be estimated, but could be relaxed with more data. For example, a rank 2 approximation that allows for separate space-time separable center and surround filters for each subunit may be useful (Schweitzer-Tong, Enroth-Cugell & Pinto, 1970 ), given the center-surround structure of bipolar cell receptive fields (Dacey et al, 2000, Turner et al, 2018.…”

Linear and nonlinear chromatic integration in the mouse retina

“…Finally, many standard models of visual responses in retinal ganglion cells assume linear receptive fields and thus linear integration of visual signals. Yet, recent studies have emphasized the importance of incorporating nonlinear signal integration, in particular nonlinear spatial integration, to predict the responses of ganglion cells to natural stimuli (Heitman et al, 2016;Liu et al, 2017;Shah et al, 2020;Turner and Rieke, 2016;Turner et al, 2018). Commonly, such models are fitted to responses from achromatic (grayscale) stimuli, and it will be interesting to explore how to include nonlinear chromatic integration for predicting response to chromatic stimuli.…”

“…Studies of spatial integration, in particular, refined the idea of receptive fields (Barlow, 1953;Hartline, 1938) and later resulted in the distinction of linear and nonlinear spatial integration, as originally exemplified by the X and Y cells of the cat retina (Enroth-Cugell and Freeman, 1987;Enroth-Cugell and Robson, 1966;Hochstein and Shapley, 1976). Mechanistic investigations of spatial integration then elucidated the role of retinal bipolar cells in shaping signal transmission through the retina (Bolinger and Gollisch, 2012;Borghuis et al, 2013;Demb et al, 2001;Schwartz et al, 2012;Turner and Rieke, 2016) and helped characterize the suppressive receptive field surround (Barlow, 1953;Kuffler, 1953;Takeshita and Gollisch, 2014;Turner et al, 2018).…”

Speed-Selectivity in Retinal Ganglion Cells is Modulated by the Complexity of the Visual Stimulus

“…It has been shown that a non-linear interaction between the center and surround parts of the RF in RGCs modulates the response to spatial patterns in a context-dependent manner, and that this effect is greater during naturalistic stimulation 50 . We think that this effect can explain, at least in part, the results that we observe; the additional signals contained in the complex stimulus would act as non-linear inhibitors, resulting in lower response when the parameters are farther from the preference of the cell, producing narrower tuning curves.…”