Light, Ca²⁺, and Photoreceptor Death: New Evidence for the Equivalent-Light Hypothesis from Arrestin Knockout Mice

In this issue of Investigative Ophthalmology and Visual Science, Chen et al. ¹ show that light exposure can greatly accelerate light damage in arrestin knockout mice. Arrestin and its splice variant p44 are soluble proteins that bind to phosphorylated rhodopsin and prevent the binding of rhodopsin to the G-protein transducin. Arrestin therefore plays a key role in turning off the visual cascade, and animals without arrestin show greatly prolonged light responses. ²  

Chen et al. ¹ show that when arrestin knockout mice are placed in constant light too dim to produce degeneration in normal animals, the photoreceptors rapidly degenerate; but if animals are kept in darkness, no degeneration occurs. These results provide strong evidence that light itself, in conjunction with the slow turnoff of the photoresponse in the arrestin knockout animals, ² is directly responsible for the degeneration of the photoreceptors. Arrestin knockout mice also show some evidence of degeneration in cyclic (12-hour light–12-hour dark) light, but the loss of photoreceptors is very slow. This slow rate of degeneration is broadly consistent with the progression of Oguchi disease in humans, a form of stationary night blindness that in some patients is apparently produced by mutations in the arrestin gene. ³  

The results of Chen et al. ¹ raise the important question of how light exposure produces photoreceptor death. Although it has long been known that continuous exposure to light can produce degeneration, the significance of this observation has been unclear, first because degeneration is much more pronounced in albino animals than in pigmented animals, ⁴ and second because there is no agreement about how light damage occurs. In many previous reports, the intensity of the continuous light used to produce degeneration was high enough to produce a nonspecific, toxic effect of illumination. The experiments of Chen et al. ¹ show that in arrestin knockout animals, light can produce degeneration even in pigmented animals at intensities that have no effect on normal pigmented mice. It is therefore very unlikely in these experiments that light had some nonspecific effect such as photodynamic damage. A more probable explanation is that in normal pigmented animals, the light was not bright enough to saturate the photoresponse, but in arrestin knockout animals, the pronounced decay of the single-photon response ² produced a more strongly maintained suppression of the photocurrent. As Chen et al. ¹ concluded, the degeneration in arrestin knockout animals is probably caused by constitutive activation of the visual cascade. 

We have previously proposed ^{5 6} that degeneration in continuous light and in certain forms of retinitis pigmentosa may be caused by constitutive activation of the photoreceptor. This equivalent-light hypothesis of retinal degeneration has received recent support, not only from the findings of Chen et al., ¹ but also from similar observations in rhodopsin kinase (RK) knockout animals ⁷ and from the discovery of new forms of photoreceptor dystrophies for which constitutive activation of the visual cascade is the most likely explanation for degeneration. These include mutations of the rod α subunit of the cyclic guanosine monophosphate (cGMP)–gated channel ⁸ and of the retinal guanylyl cyclase (RetGC-1). ^{9 10} Mutations of the channel leave the photoreceptor outer segment with almost no resting influx of Na⁺ or Ca²⁺, much as if the receptor were constantly exposed to a very bright light (Fig. 1) . Similarly, mutations of the cyclase prevent the synthesis of cGMP, resulting in a very low resting cGMP concentration and closure of the cGMP-gated channels. Thus, both these mutations lead to a situation in which the rod is continuously hyperpolarized in the dark, just as it would be during saturating continuous light. 

This equivalent-light signal produced in the photoreceptors is apparently also signaled to other retinal cells. The best evidence for this comes from recent work in the chicken retina, where the pigment granules of the retinal pigment epithelium take a very different position when the animal is in the light or the dark. In chick rd animals in which one form of guanylyl cyclase (RetGC-1) is absent, the pigment granules remain in the “light” configuration, even when the animals are kept in darkness. ¹¹ The photoreceptors in these animals are apparently sending a signal to the pigment epithelial cells that is equivalent to the one that occurs in light. 

Possible Mechanisms by Which Equivalent Light Might Produce Degeneration

In photoreceptors containing mutations for cGMP-gated channels or guanylyl cyclase, or in arrestin knockout mice, the transduction cascade is strongly activated. Because intracellular Ca (Ca²⁺ _i) is heavily dependent on Ca²⁺ influx through the cGMP-gated channels, it appears very likely that Ca²⁺ _i will be low in these animals (Fig. 1) . In the central nervous system, Ca²⁺ _i is known to play an important role in degeneration, particularly in cases for which Ca²⁺ _i is excessively elevated during stroke or trauma. ¹² Elevated Ca²⁺ _i is thought to disrupt the membrane potential and outer membrane of the mitochondria and produce the release of cytochrome c and other proteins that activate caspases, which are proteases that mediate programmed cell death or apoptosis. ¹³ In the retina, elevated Ca²⁺ _i has long been thought to be responsible for some forms of degeneration, for example in animals with the phosphodiesterase rd ¹⁴ or the GCAP1 Y99C mutations. ^{15 16}  

More recent evidence indicates that too low a Ca²⁺ concentration also seems to produce cell death. Cultured neurons deprived of growth factors normally die but can be rescued in medium containing high K⁺, which produces membrane depolarization. ¹⁷ This protection from death has been shown to be produced by the gating of Ca²⁺ channels, leading to an increase in the intracellular free Ca²⁺ concentration ^{18 19} and the activation of CaM-kinase kinase. ²⁰ Apoptosis produced by low Ca²⁺ _i may be an important mechanism of cell death and synapse elimination in the nervous system during development, because neurons that are not depolarized by ongoing synaptic input would not receive sufficient stimulation to keep Ca²⁺ _i above a minimal level. A similar process may produce degeneration in photoreceptors whenever the Ca⁺ _i is maintained at too low a level over a too prolonged a period. 

This Ca²⁺ hypothesis is unlikely to be the only mechanism of photoreceptor death. Travis, ²¹ for example, has suggested that many forms of degeneration, including those produced by real or equivalent light, may be mediated by O₂ toxicity. Loss of the photoreceptor response may decrease O₂ consumption in the outer retina enough to raise O₂ tension to levels that may be toxic. Another possibility is that constant real or equivalent light may disrupt vital circadian processes. ⁵ Degeneration in some forms of retinitis pigmentosa may also be caused by disruption of the structure of the photoreceptor or abnormal transport of protein to the outer segment plasma membrane. It seems possible, however, that disruption of the outer segment plasma membrane may also disrupt the Ca²⁺ economy of the outer segment, perhaps by making the plasma membrane too leaky to Ca²⁺ or by inhibiting the synthesis of cGMP. 

Some mutations that produce constitutive activation lead to stationary night blindness. It may be that humans with Oguchi disease ^{3 22} or with the rhodopsin G90D mutation ²³ have a lowered outer segment Ca²⁺ _i during normal cyclic light exposure, but the Ca²⁺ concentration may not be low enough for a sufficiently long period to trigger rapid degeneration. Now that Ca²⁺ _i measurements from mammalian photoreceptors are feasible, ²⁴ it will be interesting to test possible correlations between Ca²⁺ _i and photoreceptor survival. 

There is now considerable support for the equivalent-light hypothesis, and it might be asked what further experiment would provide a definitive proof. An important prediction of this hypothesis is that degeneration, produced, for example, in arrestin knockout or RK knockout mice by continuous illumination, should be prevented if a second mutation were introduced that blocked the transduction cascade. Thus, ironically, a mutation that “blinded” the rod should block the degeneration. Although this would clearly not be an advisable therapeutic strategy, the introduction of a second blinding mutation in arrestin knockout animals would be a definitive test of whether an equivalent-light signal is responsible for at least some forms of retinal degeneration. 

 Supported in part by the John Simon Guggenheim Memorial Foundation (GLF), and the W. M. Keck Foundation (JEL), and by Grants EY 01844 (GLF) and EY 01496 (JEL) from the National Eye Institute, National Institutes of Health.

 Submitted for publication April 9, 1999; accepted June 14, 1999.

 Commercial relationships policy: N.

Figure 1.

 

Equivalent light and mechanisms of degeneration. Left: a normal, dark-adapted rod outer segment, with Na⁺ and Ca²⁺ entering through cGMP-gated channels, and Ca²⁺ exported through the Na⁺ /K⁺–Ca²⁺ transporter. The mutations PDE rd and GCAP1 Y99C both have been shown to produce high outer segment levels of cGMP that have been postulated to produce high intracellular Ca²⁺, which may lead to degeneration (top: middle and right). Real or equivalent light produced by cyclase or channel mutations, or by knocking out the arrestin or rhodopsin kinase (RK) gene, would lead to low levels of cGMP and closed channels. Because Ca²⁺ entry is blocked, there is a low level of intracellular Ca²⁺ (bottom: middle). Either low Ca⁺ or elevated O₂ may then trigger apoptosis (bottom: right). The loss of the outer segment and eventual death of the photoreceptor would then eliminate the photoreceptor light response and block transmitter release, which may be responsible for spreading the equivalent-light signal to other cells, perhaps as the result of blocking the retinal circadian rhythm. ⁵

 

The authors thank many colleagues for discussion of ideas—in particular, Jeannie Chen, Melvin I. Simon, and Gabriel H. Travis.