Abstract
aspects of visual information from the retina to various parts in the brain. Retinal output is most directly
conveyed to the cortex via the retino-geniculo-cortical pathway, comprised of RGCs, relay cells in the
dorsolateral geniculate nucleus (dLGN) and the primary visual cortex (V1). While it has long been known that this pathway is not homogenous but consists of parallel channels each carrying specific information, it is still debated which RGC types project to the dLGN and how their output is transformed at the level of the dLGN. Here, we started to characterize, in the mouse model, the functional properties of dLGNprojecting RGCs and to compare responses of RGCs and dLGN neurons to the same set of visual stimuli.
We explored two approaches for selective labeling and physiological characterization of dLGN-projecting RGCs. First, we injected a retrograde tracer (“mini-Ruby”, Molecular Probes) into the mouse dLGN. After 7 days, the retina was removed and electroporated with a synthetic calcium indicator (Briggman & Euler, J Neurophysiol 2011). Using two-photon in-vitro imaging, we recorded light-evoked calcium activity from the population of RGCs that had been labelled by the retrograde tracer. Visual stimuli included
frequency/contrast modulated full-field flicker, dense noise, moving bar, and chromatic stimuli. Second, we injected an adeno-associated virus (AAV) encoding the calcium biosensor GCaMP6 into the dLGN.
Through transfection of RGC terminals this leads to retinal biosensor expression, which enabled us to selectively record light-evoked calcium responses in dLGN-projecting RGCs. In a separate set of experiments, we characterized the responses of dLGN neurons to the same visual stimuli using in-vivo extracellular multi-electrode recordings in the dLGN of awake, head-fixed mice. Combining the data sets from the retina and the dLGN, we seek to build computational models that will test if and how dLGN responses can be described as specific combinations of RGC output channels and the influence of local inhibitory circuits. Specifically, we will ask if response features are simply inherited from the RGC input or if they are modified within the LGN.
In conclusion, this study promises to yield a functional characterization of the population of dLGNprojecting
RGCs, and to provide fundamental insights into how the representation of visual information changes along the first stages of the retino-geniculo-cortical pathway.
