The outer and inner halves of photoreceptor adaptation

After setting the foundation for our understanding of the action potential, Alan Hodgkin turned to the study of photoreceptors. These cells do not discharge in the classical sense, but nevertheless possess a rich complement of voltage‐gated currents. When Hodgkin presented this new research direction to the Royal Society, he underscored one question in particular:

A major problem confronting anyone who attempts to construct a model of a light receptor is to account for the changes in sensitivity which enable photoreceptors to deal with a wide range of light intensities (Hodgkin, 1972).

Such changes in sensitivity are remarkable. Rods signal the absorption of single photons to support vision in starlight, yet tune themselves to work even when illumination is brighter by orders of magnitude (Luo et al. 2008). Furthermore, this tuning follows the Weber–Fechner law – a change in the overall light level is matched by an equal and opposite change in sensitivity – and therefore favours the encoding of contrast. Adaptation by both rods and cones is a large part of how one sees over the ∼10 log‐unit breadth of environmental irradiances, and of how objects appear similar under different lighting conditions (Rieke & Rudd, 2009).

Much has been learned about mechanisms of negative feedback in the phototransduction cascade that generate adaptation (Arshavsky & Burns, 2012). Less attention has been paid to the voltage‐gated currents that shape the output of this cascade and ultimately govern signal transmission. In this issue of The Journal of Physiology, Pahlberg and colleagues demonstrate a potent role of these currents in rod adaptation (Pahlberg et al. 2017). To appreciate the sophistication of their experiments, a summary of rod photoreception is helpful.

Rods capture photons using a molecule called rhodopsin, which activates a G‐protein cascade to close non‐selective cation channels. These steps occur in a compartment called the outer segment. The resulting hyperpolarization propagates through the inner segment, which contains voltage‐gated ion channels, to influence the synaptic output.

Rhodopsin has a purple hue, but its colour is fleeting. Following photon absorption, the activated pigment splits into its constituent opsin protein and chromophore, both of which are transparent to the human eye at physiological concentrations. In other words, the pigment bleaches. Photoreceptor sensitivity is reduced as fewer rhodopsins remain available for photon catch. Sensitivity falls further because of bleaching adaptation, whereby opsin weakly but constitutively activates phototransduction, including elements of negative feedback that reduce gain in the cascade (Cornwall & Fain, 1994). Gain also declines with the number of transduction channels that remain to be closed, due to response compression.

Rods express a palette of voltage‐gated currents, two of which are especially well‐positioned to drive negative feedback during phototransduction. These are a hyperpolarization‐activated, non‐selective cation current (I _h) and an inwardly rectifying potassium current (I _Kx). As phototransduction produces hyperpolarization, I _Kx deactivates and I _h activates to promote depolarization (Barnes, 1994). Thus, these currents produce a smaller, faster, and more tightly regulated response to light.

The question taken up by Pahlberg and colleagues is how voltage‐gated currents might influence adaptation. Their central experiments took advantage of the rod's compartmentalization. Drawing the outer segment into a suction electrode caused phototransduction current to be isolated, while viewing the cell through a patch electrode allowed other currents to be included. They found that phototransduction current displayed a steeper decline in sensitivity with light history than did the overall voltage response. Furthermore, this difference was eliminated by blocking voltage‐gated currents. The exquisite quality of the data and the precision of the analysis make the message clear: voltage‐gated currents support the dynamic range of rod photoreception. Additional evidence for this conclusion was recently provided by in vivo studies of animals lacking the channel that produces I _h in rods (Sothilingam et al. 2016).

Many questions are open for future investigation. For example, how do individual currents flow during the light response to generate its particular waveform? To what degree are their voltage dependencies and kinetics matched to the parameters of phototransduction and synaptic release, especially with changes in adaptation state? How different is the situation in cones and other sensory cells?

Each human retina is estimated to contain ∼120 million rods, which exceeds the number of cones by ∼20‐fold. A common view is that rods are inoperative in daylight because their extreme sensitivity comes at the cost of dynamic range; in this scenario, ∼95% of the photoreceptors consume resources without contributing to vision. In providing an especially sharp view of how voltage‐gated currents extend rod function over high irradiances, Pahlberg and colleagues highlight the possibility that these cells influence perception in ways that have yet to be found.

Additional information

Competing interests

None declared.

Funding

Support was provided by the National Institutes of Health (R01 EY023648, R01 EY025555 and R21 EY025840 to M.T.H.D.; 1U54HD090255 to the Boston Children's Hospital IDDRC; and P30 EY012196 to Harvard Medical School).