Research

Our laboratory is interested in the first steps in vision; that is, how photons produce electrical changes in photoreceptors, and how the time course of these signals change with light intensity. Rod and cone photoreceptors signal the presence of light using a G protein cascade that results in a reduction of the cell's inward current. As long as the cascade remains activated, the light response persists, making deactivation essential for maintaining both the sensitivity and temporal resolution of vision. To study deactivation mechanisms, we record from intact photoreceptors in which the functions of various proteins have been specifically perturbed using gene-targeting techniques.

Left: Bright-field image of a captured rod outter segment during a recording; Right: an example of a single-flash current response.
(Click images for larger view)

Recordings from rod outer segments allows us to analyze the cell's responses to flashes or steps of light quantitatively and in real time. From the changes in inward current, we can infer a great deal about the rates of the reactions that underlie both the activation and deactivation of the response elicited from a single activated rhodopsin molecule. Signaling through GPCRs (G protein coupled receptors) underlies many cellular processes, yet it is not known which molecules determine the duration of signaling in intact cells. Two candidates are the GRKs (G protein coupled Receptor Kinases) and RGSs (Regulators of G protein Signaling), deactivation enzymes for GPCRs and G proteins, respectively.

Recently, we have investigated whether GRK1 (rhodopsin kinase) or the RGS of photoreceptors, RGS9-1, governs the overall rate of recovery of the light response in the mammalian rods. We found that overexpression of GRK1 increases phosphorylation of the GPCR, rhodopsin, but had no effect on photoresponse recovery. In contrast, overexpression of the photoreceptor RGS complex (RGS9-1•Gβ5L•R9AP) dramatically accelerated response recovery, revealing that G protein deactivation is normally at least 2.5-times slower than rhodopsin deactivation. This resolves a long-standing controversy: the deactivation of many G protein molecules, rather than the shutoff of a single long-lived photoexcited rhodopsin, dominates the recovery of the rod's flash response.

Diurnally, our visual system operates over range of light intensities that vary by more than eleven orders of magnitude. Over this entire range, visual function critically depends upon the ability of photoreceptor cells to generate output signals. Light adaptation in photoreceptors aids this process by decreasing the cell's sensitivity to light and speeding the incremental response kinetics. These two hallmarks of cascade adaptation stem from decades of physiology experiments that have focused on the changes that occur over the course of a few tens of seconds, in response to relatively dim steady light. Surprisingly, virtually nothing is known about the adaptive changes that occur on a longer time scale, and in response to brighter intensities.

Recently we discovered a novel form of adaptation rod photoreceptors that speeds Gt/PDE deactivation (termed "adaptive acceleration"; Krispel et al 2003). Both the induction of this adaptation and the fading to the dark-adapted state is rather slow (tens of seconds to minutes), suggesting that the adaptation may require post translational modification (e.g. phosphorylation) or intracellular translocation of an additional deactivation factor. Using electrophysiological and biochemical assays for response recovery and cascade deactivation, we are currently trying to pinpoint the molecular event that is altered in this form of adaptation, and how it is regulated. Additional regulatory processes can alter G protein signaling on a longer time scale in vitro, and thus may have important physiological correlates for cell function. For example, light adaptation changes the phosphorylation state of several cascade regulators. We are currently investigating the role of phosphorylation in altering the response kinetics. On a longer time scale, bright light causes cascade regulators like arrestin and recoverin, as well as Gt itself, to dramatically translocate between cellular compartments, and we are interested with the functional consequences of these movements on photoreceptor function.

Our experiments investigating rhodopsin deactivation and its role in reproducibility have revealed that multiple phosphorylation by rhodopsin kinase and high affinity arresting binding is essential for quenching rhodopsin's activity, and doing so reproducibly. These requirements for arrestin may partially explain the poor scotopic vision observed in patients with Oguchi disease, a form of stationary night blindness.

Additionally, experiments on transducin deactivation have shown that the RGS9-1•Gβ5L•R9AP complex are obligate partners in the timely deactivation of transducin and termination of the light response. Recently, human patients with mutations in RGS9-1 or R9AP have been found to have marked loss of acuity to low-contrast moving objects (Nishiguchi et al, 2004). This condition, termed bradyopsia, supports the notion that the temporal resolution of the visual system is ultimately limited by the temporal resolution of the photoreceptors and perhaps explains why the deactivation mechanisms are so tightly controlled.

Our studies of light adapted rods have revealed a novel, long-lasting acceleration of the step that normally limits recovery kinetics. We expect that this will ultimately have widespread ramifications for all of visual science. One function of this form of adaptation may be to help the visual system make a smooth transition from photopic to scotopic vision. If this form of adaptation occurs in humans, it would suggest that human psychophysics experiments that have been thought to isolate cone adaptation may in fact also include rod adaptation effects.