Makino Research Projects

What sets the speed of the photoresponse?

Rods contain an enormous number of rhodopsin molecules densely packed in membranous disks, in order to achieve high sensitivity. But dense packing could impair phototransduction by restricting the lateral mobility of rhodopsin as well as that of other proteins in the membrane. To test the hypothesis that the speed of the photoresponse is limited by the rates with which molecules collide with each other on the membrane, we recorded photoresponses of single rods containing fewer rhodopsins in their membranes. This condition was achieved by hemizygous knockout of the rod opsin gene in transgenic mice (R+/-). Photoresponses were accelerated by the reduction in rhodopsin packing density, supporting the hypothesis (Fig. 1). The key step in the activation of the photoresponse was identified as the collision between rhodopsin and transducin. Response recovery appears to be limited by the collision rate between rhodopsin and rhodopsin kinase and/or that between between transducin and RGS9. These results predict that under scotopic conditions, humans heterozygous for a null mutation in rhodopsin will have higher than normal temporal resolution.


Figure 1. Top panels: membrane surfaces of wild type and transgenic mouse rod membranes. Rhodopsins and transducins are depicted by red and blue circles, respectively. The green area shows the surface area “sampled” by a photoexcited rhodopsin (green star) every few milliseconds as it diffuses. The area expands and the number of transducins contacted increases as the packing density of rhodopsins decreases. Bottom panels: flash responses of wild type (left) and R+/- (right) rods recorded with suction electrode. Responses of R+/- rods rise and recover more rapidly than those of the wild type rods.

To find out how much the speed of the photoresponse slows with more rhodopsin in the membranes, recordings were made from mutant mice induced to overexpress rhodopsin. Surprisingly, the rods expanded their diameter to accommodate the extra rhodopsins without increasing membrane crowding. The rising phase of the photoresponse was still slowed, however, due to the increased intracellular volume and the dispersal of transducins over a greater membrane surface area.

What determines the spectral sensitivity of a photoreceptor?

The spectral sensitivity of a photoreceptor is largely determined by the visual pigment that it contains. Although photoreceptors were generally thought to express only one type of pigment, there are a number of examples where a cone has been shown to contain two different types of pigments. However, we discovered that the UV-sensitive cone in tiger salamander contains three different types of pigments, all of which are functional (Fig. 2). In addition to a UV-absorbing pigment, the pigments of blue- and red-sensitive cones are also present. Selectivity of pigment expression is not completely lacking in the UV-sensitive cone because the green-sensitive rod pigment is absent. Interestingly, the blue-sensitive cone pigment is also expressed in a blue-sensitive rod (Fig. 2), the first example of a rod and a cone utilizing the same visual pigment.

Mankino_Research_002Figure 2. Averaged spectral sensitivities of salamander UV-sensitive cones (violet circles), blue-sensitive rods (blue diamonds) and cones (blue circles) and red-sensitive cones (red circles). Continuous lines show fits of the spectra with a template. The fit to the UV-sensitive cone spectrum was made assuming that the long wavelength limb included components that were the same as those giving rise to the spectra of the other two cone types.




Can visual pigments be activated pharmacologically?

Ordinarily, photoisomerization of visual pigment changes it from a quiescent to a catalytically active state. We discovered that pigments can also be activated by the binding of small molecules. In electrical recordings of single photoreceptors, pharmacologic activation mimicked the effect of a background light, leading to reduced circulating current, lower sensitivity, and faster flash response kinetics (Fig. 3). The truncated retinoid, ß-ionone, activates some visual pigments but not others. In contrast, vitamin A appears to be less selective. Based upon results of a biochemical screen of opsin against a chemical library, we are now testing the effects of a number of non-retinoid compounds for their ability to activate visual pigments in rods and cones. Certain inherited retinal degenerations are caused by photoreceptor dysfunction in which there is a mutation that disrupts the phototransduction cascade. Depending upon the nature of the defect, compounds that either up- or down-regulate visual pigment activity may prove useful as a treatment option.


Figure 3. Effect of exogenously applied ß-ionone on salamander rods. A. The structure of the truncated retinoid, ß-ionone. B. Flash responses (top) and stimulus-response plots (bottom) for dark-adapted rods of salamander. Results obtained in Ringer’s solution are shown in black; those obtained during perfusion with ß-ionone are shown in gray. Left. The green-sensitive rod responses were unchanged by ß-ionone. Right. The blue-sensitive rod had a reduced response in the presence of ß-ionone. The sensitivity to flashes also decreased as shown by the rightward shift of the stimulus-response relations (gray triangles). The cell fully recovered after washing away the ß-ionone (black diamonds). B modified from Isayama et al. (2009).

Why do rods employ two Ca2+ binding proteins to regulate cGMP synthesis?

In retinal rods, cGMP serves as the second messenger that links photon capture by rhodopsin on the disk membrane to ion channel activity on the plasma membrane. After rhodopsin photoexcitation, the hydrolysis of cGMP leads to closure of cGMP-gated channels and curtailment of the influx of Na+ and Ca2+. During response recovery, retina-specific guanylyl cyclases (retGCs) replenish cGMP, reopen the channels, and resume the circulating current. The cGMP levels cannot be restored rapidly without facilitation by guanylyl cyclase activating proteins (GCAPs), which sense the decrease in Ca2+ caused by illumination and greatly stimulate the rate of cGMP synthesis.

All vertebrate rods use at least two GCAPs: GCAP1 and GCAP2. Working closely with the Dizhoor laboratory at Salus University, we are exploring the basis for the dual system by studying mutant mice that lack either or both of the two GCAPs. Deletion of GCAP2 did not change the amplitude of the single photon response, but slowed its recovery (Fig. 4, orange trace). Elimination of GCAP1 allowed the photon response to rise for twice as long to reach an amplitude that was twice as large (Fig. 4, blue trace). Although lack of GCAP2 did not affect GCAP1 expression, knockout of GCAP1 did cause an up-regulation of GCAP2 as detected by immunofluorescence and Western blotting. The overexpression of GCAP2 resulted in acceleration of the response recovery rather than the slowdown that was expected from the loss of one component in the feedback loop.


Figure 4. The single photon responses of wild-type control (black trace), GCAP1 knockout (blue), GCAP2 knockout (orange), and GCAPs1&2 knockout (gray) rods.

The physiological properties of these photoresponses support a relay model for GCAPs regulation of retGCs (Fig. 5). GCAPs 1 and 2 change from Ca2+-bound to Mg2+-bound states in a sequential manner. Because of its lower Ca2+ affinity, GCAP1 is the first responder that senses the decrease of Ca2+ initiated by photon absorption and limits response amplitude. GCAP2, with a higher Ca2+ affinity, does not assist retGCs until Ca2+ concentration declines even further, which happens during a single photon response as well as with bright light. GCAP2 stimulation of retGCs provides for a timely response recovery. Together, the two GCAPs grant the essential Ca2+ feedback to set the operating range in vertebrate rod photoreceptors.


Figure 5. Relay model for GCAPs regulation of retGC activity based on Ca2+ concentration change. The physiological free Ca2+ concentration ranges from 250 nM in the dark to 20 nM in the light in a mouse rod. In the dark, both GCAPs bind Ca2+ and suppress retGC activity. Under illumination, GCAP1 responds first to the light-induced Ca2+ decrease. Ca2+ dissociates from GCAP1 and Mg2+ takes its place. As Ca2+ concentration falls further, a similar Mg2+/Ca2+ exchange occurs with GCAP2 to replenish cGMP at the maximal rate. As the channels reopen, Ca2+ concentration returns to the dark level, GCAPs rebind Ca2+ and suppress retGC activity, restoring it to basal levels.