• gap junctions
• mouse
• photoreceptors
• retina

Rod and cone photoreceptors form gap junctions (GJs) which provide an alternative route for rod signals when light saturates the primary high gain pathway. Indirect evidence suggests that in mammals, as in lower vertebrates, rod-cone coupling is dynamically regulated by light and circadian rhythmicity through endogenous neuromodulators such as dopamine (DA). However, the only direct tests done so far, in macaque, found coupling to be static. Moreover, recordings from the postsynaptic cone bipolar cell, in mouse, suggest that this route may give only a minor contribution to rod signaling.
In my thesis I investigated the functional impact and regulatory latitude of rod-cone coupling by recording, with perforated patch clamp, from mouse cones in an in vitro retinal slice preparation. In the process, I optimized the techniques required to gain intracellular access to these small and challenging neurons. I dissected rod input in the photovoltage of wild type mouse cones by exploiting differences in light sensitivity, kinetics of recovery from bright flashes, and relative spectral preference to green (G) over ultraviolet (UV) light. Most cones expressed rod-like features, including: (1) responses to dim flashes, (2) slow plateaus in response to moderately bright flashes and a transient suppression of dim flash responses, (3) long recovery of dim flash responses and slow plateaus after rod-saturating backgrounds, (4) preference for dim G over dim UV flashes, irrespective of the intrinsic spectral preference of the cone determined with rod-saturating pre-flashes. Dim and bright flash responses had different reversal potentials, consistent with an origin in separate electrotonic compartments. The role of GJs was confirmed pharmacologically. Cones dramatically increased their coupling to rods within minutes after seal formation, revealing mechanisms for rapid plastic change, triggered in my experiments by a perturbation of the intracellular milieu. In fully coupled cones the overall junctional conductance could exceed the light-sensitive conductance. In contrast to wild type animals, in connexin isoform 36 (Cx36) knockout mice cones did not appear to be able to couple to rods, supporting a key role for Cx36 in rod-cone GJs. In disagreement with indirect data, but similarly to what observed in single macaque cones, I found evidence that would rule out a role of the dopaminergic system in the regulation of rod-cone coupling.
My work provides the first direct and conclusive evidence for rod-cone coupling in the mouse retina, an emerging model for studies of early visual processing in health and disease. This coupling is not hardwired but can be rapidly up-regulated, revealing that junctional contacts are adequate for it to play an important role in rod visual signaling and, potentially, also in the biochemical interaction between photoreceptors. The cellular mechanisms leading to a spontaneous coupling increase during patch recordings need to be investigated to reconcile the lack of DAergic modulation in single cell recordings with other indirect evidence.