Continuous Rod Noise and Nonlinear Synaptic Filtering Shape Signal Detection in the Primary Rod Pathway

Rod photoreceptors are optimized to detect single photons, but their performance near absolute visual threshold is limited by intrinsic phototransduction noise. The two major components are discrete single-photon-like events and continuous fluctuations in outer-segment current. Although spontaneous rhodopsin activation is established as the source of discrete noise, the molecular origin of continuous noise remains controversial, and how these distinct noise sources influence signal processing at the first visual synapse is not fully understood. To address these questions, we combined genetic manipulation of rhodopsin, phosphodiesterase 6 (PDE6), and glutamic acid rich protein 2 (GARP2) with single-cell recordings from rods and rod-bipolar cells (RBCs) and pharmacological manipulation of mGluR6 signaling. Continuous noise was unaffected by reduced rhodopsin concentration but was strongly influenced by PDE6 expression and regulation, supporting basal PDE6 activity as a major source of continuous noise. Reducing either discrete or continuous noise increased signal-to-noise ratio (SNR) in rods, but these changes were differentially processed at the rod-to-RBC synapse. In WT retina, the nonlinear transformation increased SNR, whereas elevated continuous noise limited this increase. Conversely, reduced continuous noise increased rod SNR, which was maintained in RBCs and associated with a lower detection threshold. Pharmacological reduction of the mGluR6-dependent nonlinearity altered these genotype-dependent differences, indicating that nonlinear filtering is well matched to normal signal and noise distributions but does not fully compensate when presynaptic noise is experimentally altered. Together, our study identifies spontaneous PDE6 activity as a major source of continuous noise and demonstrates that continuous noise interacts with nonlinear processing at the first visual synapse to determine the fidelity and detection threshold of dim-light signals.