Blue Light’s Effects on Rhodopsin: Photoreversal of Bleaching in Living Rat Eyes

Abstract

purpose. To determine whether blue light induces photoreversal of rhodopsin bleaching in vivo.

methods. Eyes of anesthetized albino rats were exposed to either green (550 nm) or deep blue (403 nm) light, and the time course of rhodopsin bleaching was determined. Rhodopsin was isolated from whole retinas by detergent extraction and measured photometrically. To inhibit photoreversal of bleaching, rats were perfused with 70 mM hydroxylamine (NH₂OH), a known inhibitor of photoreversal. To determine whether blue-absorbing, photoreversible photoproducts were formed, rhodopsin was bleached to near completion with green light and then exposed to blue light. Finally, experimental results were simulated on a computer by means of a simple, three-component model involving a long-lived photoreversible photoproduct.

results. Photoreversal of bleaching in blue light occurs in vivo as evidenced by the following: In the absence of NH₂OH, bleaching of rhodopsin by blue light was slow and complex. In the presence of NH₂OH, however, blue light bleached rhodopsin very fast with a simple, pseudo–first-order kinetic. A long-lived bleaching intermediate produced by green light exposure was photoreversed to rhodopsin by exposure to blue light. The three-component computer model, invoking a blue-absorbing, photoreversible, long-lived intermediate accurately described the data.

conclusions. Because of the instantaneous, nonmetabolic regeneration of rhodopsin by the process of photoreversal of bleaching, blue light exposure permits the absorption of large numbers of photons by rhodopsin and by a photoreversible intermediate of bleaching in vivo. These data may have an important impact on resolving mechanisms of blue light–mediated damage to the retina.

The chromophore(s) for blue light–induced retinal damage are still not completely elucidated. For exposure to excessive white light, elegant experiments by Noell et al. ¹ and by Williams and Howell ² pointed to rhodopsin as the chromophore for light damage, and a recent study unequivocally shows that the assumptions of these early works are true. ³ Could rhodopsin also be crucial for blue light–induced retinal injury? 

The visual pigment rhodopsin consists of 11-cis-retinal and opsin, the apoprotein of the pigment. Photon absorption causes bleaching of rhodopsin, a process that is initiated by the photoconversion of 11-cis-retinal to all-trans-retinal. Subsequently, rhodopsin undergoes a series of dark reactions that culminate in the dissociation of retinal, thus completing the process of bleaching. However, before this dissociation occurs, an intermediate of the bleaching process, metarhodopsin II (MII), ^{4 5} interacts with the G-protein transducin. This interaction initiates a cyclic nucleotide cascade that ultimately converts the absorbed light into an electrical response of the rod cell. Quenching of G-protein activation occurs by binding of arrestin to MII. ⁶ Because the absorption maximum of the bleaching intermediate MII is at 380 nm, ⁷ MII is spectrophotometrically virtually indistinguishable from dissociated, free all-trans-retinal, a truly bleached product. ^{7 8 9} However, a significant difference between free retinal and MII is that, in MII, the all-trans-retinal continues to occupy its chromophoric site in opsin, whereas free retinal is not attached to opsin. ⁸ Furthermore, the all-trans-retinal transiently binds to phosphatidyl-ethanolamine and, after reduction to all-trans-retinol, is transported to the pigmented epithelium. ^{10 11}  

Hubbard and Kropf ¹² discovered that normal photobleaching could be prevented if additional light was absorbed by bleaching intermediates. This prevention of bleaching is effected by a trans-to-cis photograph reisomerization of the retinal chromophore after absorption of additional photons, and the process was thus called “photoreversal” of bleaching. ¹³ That photoregeneration created functional rhodopsin from bleaching intermediates was also suggested by Cone, who reported that a complete bleaching of rhodopsin abolishes early receptor potentials in excised rat eyes, but that they can be recorded again after blue light is flashed onto the eye. Thus, rhodopsin must have been restored by the absorption of the additional light of short wavelength. ¹⁴ More light absorbed caused less bleaching. Indeed, working at low temperatures to extend the lifetime of successive intermediates, Hubbard and Kropf ¹² showed that photograph steady states between a given intermediate and a mixture of rhodopsin and isorhodopsin (9-cis-chromophore) were possible and were indefinitely stable as long as the temperature was maintained. That is, the system was “closed.” They further showed that the quantum efficiencies for the cis-trans and trans-to-cis reactions were virtually identical (0.7) for the earliest intermediates. A critical stage of bleaching arose when MII appeared during irradiation. Suddenly, less rhodopsin could be photoregenerated, and the maximum amount of rhodopsin bleached by a flash increased dramatically. Although the system became “leaky,” it was nevertheless concluded that the quantum efficiency for photoconverting MII to stable, light-sensitive pigments was lower than that of the earlier intermediates, but that it was not zero. ^{4 15 16 17}  

It can be deduced from the work of Hubbard and Kropf ¹² that three conditions are necessary for maintaining photograph steady states between rhodopsin and any given intermediate of bleaching: The intermediate absorbing the additional light must retain retinal at the chromophoric site, the incoming photons must be of the proper wavelength to be absorbed by that intermediate, and the intermediate absorbing the light must have a lifetime that is longer than or at least commensurate with the rate at which photons are coming in (i.e., intensity) and being absorbed. 

Except for the flash photolysis study by Hagins ¹⁸ and the observations of Cone, ¹⁴ all the work referenced thus far was performed in vitro. In the present study we show by direct measurements of the course of bleaching of rhodopsin that blue light photoregenerated a relatively long-lived, blue-absorbing intermediate in vivo. Based on our results and the identification of rhodopsin as chromophore for light damage, ^{1 2 3} we show that in experiments of blue light damage, ^{19 20 21 22 23 24} photoreversal of bleaching may have occurred, leading to the absorption of high numbers of photons per unit of time and rendering blue light highly effective in the mediation of light damage. 

Methods

Animals

All experiments conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Male, Sprague–Dawley rats, 10 to 12 weeks old, were maintained for 3 to 7 weeks in 3-lux cyclic light (light/dark, 12 hours/12 hours). Rats were dark adapted overnight (16 hours) and anesthetized with a mixture of ketamine (75 mg/kg) and xylazine (23 mg/kg). 

Light Exposure

Left eyes of anesthetized rats (ketamine 75 mg/kg and xylazine 23 mg/kg) were kept moist with Methocel (CibaVision, Hergiswil, Switzerland) and exposed either to deep blue light (403 ± 10 nm) or to green light (550 ± 10 nm) for up to 30 minutes. The light exposure system consisted of a xenon short-arc reflector lamp (230 V, 50 Hz, 120 W; Intralux MDR 100; Volpi, Schlieren, Switzerland) with interference filters to eliminate UV and infrared radiation and a liquid fiberoptic light guide (8 mm in diameter) to the animal’s eye. The optical system included a switch holder for blue (403 ± 10 nm, bandwidth) or green (550 ± 10 nm bandwidth) interference filters. 

Intensities of exposure were at 300 μW/cm² for 403-nm blue light and 409 μW/cm² for 550-nm green light. These intensities gave equal photon fluxes at the cornea of 6 × 10¹⁴/sec per square centimeter. 

NH₂OH Studies: Inhibition of Photoreversal

NH₂OH actively removes the retinal from the chromophoric site of late-bleaching intermediates including MII, ^{25 26} without affecting rhodopsin itself (data not shown). ^{27 28 29} Although inhibition of photoreversal of bleaching by NH₂OH is very efficient, it may nevertheless be incomplete: some intermediates may still be able to absorb enough photons to be photoreversed before NH₂OH can remove retinal from the chromophoric site in opsin, which will, perforce, prevent photoreversal of bleaching. To apply NH₂OH, deeply anesthetized animals were cardially perfused with NH₂OH solution (70 mM in 0.9% NaCl [pH 7.2], 100 ml) in dim red light immediately before irradiation of the eye with blue light, as described earlier. 

Rhodopsin Measurements and Calculation of the First-Order Rate Constants

Retinas were extruded through slits across the corneas and immersed in 1 ml distilled water for 30 seconds. After centrifugation at 14,000g for 3 minutes at room temperature, supernatants were discarded, and 0.7 ml cetyltrimethylammonium bromide (CTAB, 1% in H₂O) was added to each retina pellet. Retinas were homogenized with a polytron for 20 seconds, and the homogenates were centrifuged at 14,000g for 3 minutes (room temperature). Supernatants were drawn off quantitatively and scanned on a rapid-recording spectrophotometer (Carry 50; Varian, Basel, Switzerland). After scanning, rhodopsin was bleached completely by exposing the retinal extract to 20,000 lux of white light for 1 minute. Extracts were scanned again, and the difference in absorbancy of rhodopsin at 500 nm was determined. This value was used to calculate from the Beer-Lambert law (molar absorbance coefficient = 42,000 M/cm) the concentration of rhodopsin in the extract, as described recently. ³⁰  

Because the intensity was held constant during a run, the rate of bleaching (K _b) is a first-order rate constant and has units of (time⁻¹). Thus, K _b = ln (R _t/R ₀)/t; where (R _t/R ₀) is the fraction of rhodopsin remaining after exposure during a time (t). 

Computer Simulation of Results

The experimental results were simulated on a computer (Excel; Microsoft, Redmond, WA). The authors will supply details of this simulation on request. The model simulated was  

\[R\ {_{K_{\mathrm{pr}}}^{K_{\mathrm{b}}}}\ B\ {_{}^{K_{\mathrm{1}}}}\ P\]

where R is rhodopsin and B is an as yet unidentified blue-absorbing, photoreversible intermediate that decays slowly to P (products). The rate constants are K _b for bleaching, K _pr for photoreversal of bleaching, and K _l for loss of photoreversibility (leak). Numerical values of these rate constants, either measured (K _b) or estimated (K _pr and K _l) from the results, were put into the differential rate equations that describe the model. The set of equations was then solved simultaneously and fit to the data by iteration. ³¹ An important constraint on the iteration was that K _b, the bleaching rate constant in blue light (at 300 μW/cm²) was not allowed to change because its absolute value had been measured (see Fig. 3 ). Thus, all other rates were relative to this anchor, requiring that only two parameters be fit. 

Results

Blue Light–Mediated Photoreversal of Bleaching In Vivo

Bleaching of rhodopsin with green light (550 nm) caused the rapid disappearance of the visual pigment leaving less than 5% of rhodopsin after 5 or 10 minutes of exposure. When blue light (403 nm) was used instead, rhodopsin disappeared slowly, with approximately 20% of the dark value left, even after 30 minutes of exposure. However, when the same exposure was tested in an animal perfused with NH₂OH (70 mM), rhodopsin disappeared fast, almost comparable to green light bleaching (Fig. 1) . In contrast, when animals were exposed to green light, disappearance of rhodopsin was not influenced by NH₂OH (12% ± 3.8% rhodopsin present after 1 minute of green light exposure in the absence of NH₂OH, n = 3; 13% ± 3.9%, rhodopsin present in the presence of NH₂OH, n = 3). Therefore, long-wave green light (550 nm) did not cause photoreversal of bleaching as has been suggested by others. ^{32 33} NH₂OH has no effect on rhodopsin per se (data not shown) but strongly reacts with the retinal chromophore in late intermediates of the bleaching process. ^{27 28 29} The interaction results in the removal of the retinal from its chromophoric site and forms a retinal oxime with a maximal absorption at 365 nm, well out of the range of rhodopsin-visible absorption. ^{27 28 29} In these experiments, preventing potential photoreversal with NH₂OH greatly enhanced the rate at which rhodopsin disappeared in a blue light exposure. 

That NH₂OH, delivered by cardiac perfusion, penetrated the blood–retina barrier and reached the bleaching intermediates in photoreceptors was shown by the different spectra of pigments extracted from retinas of animals that had or had not been perfused with NH₂OH (Fig. 1 ; insert). Whereas the absorption maximum of the final bleached products in retinas without NH₂OH was at 380 nm, the correct wavelength for all-trans-retinal to absorb, ⁷ perfusion with NH₂OH caused a shift of the absorption maximum to 365 nm, the wavelength of retinal oxime absorption. We conclude that the NH₂OH arrived at the photoreceptors, reacted with late intermediates of bleaching, and efficiently inhibited photoregeneration. However, because NH₂OH prevents photoreversal of bleaching only if removal of retinal from the bleaching intermediate is faster than the absorption of additional photons, we cannot exclude that some few intermediates could be photoreversed, even in the presence of NH₂OH. Nevertheless, the rate of blue light bleaching in the presence of NH₂OH proceeded in an almost straight line (Fig. 3) indicating a very efficient inhibition of photoreversal in our experiments. 

Although long-wave green light does not cause photoreversal of bleaching (see above), ^{32 33} rhodopsin was not bleached to completion. When rhodopsin was bleached using a 5-minute green flash (550 nm; 409 μW/cm²), rhodopsin was metabolically regenerated to 13% of the dark value during 15 minutes in darkness (data not shown). Therefore, the metabolic regeneration of rhodopsin in the visual cycle may have restored enough rhodopsin during exposure that bleaching and regeneration reached an equilibrium at very low levels (less than 5%). 

Photoregeneration of Rhodopsin by Blue Light after Green Light Bleaching

Green light bleaches rhodopsin without causing photoreversal (see above). ^{32 33} After 5 minutes of exposure to green light, less than 5% of rhodopsin was left (Fig. 2) . Continuation of the green light exposure decreased this value even further. However, when the green light (after 5 minutes) was rapidly switched to blue light (for 1 minute), rhodopsin was regenerated to 28% on average (Fig. 2) . Obviously, some intermediates of the green light–induced bleaching process were long lived and capable of reversal to rhodopsin after absorption of additional blue light. More light appeared to cause less bleaching, a hallmark of photoreversal of bleaching. 

Bleaching in the Absence of Photoreversal

The bleaching functions (Fig. 1) obtained with green light and with blue light in the absence of photoreversal (that is in the presence of NH₂OH), gave the rates of photon absorptions by rhodopsin at 550 and 403 nm, respectively. To determine the corresponding values of the rate constants for these bleaches, we plotted the bleaching functions as integrated first-order processes and determined their slopes (Fig. 3) . The rate constant (slope) for the bleach in blue in the presence of NH₂OH (K _b), was calculated to be 0.61 per minute. This value was used in the model that will be described. The constant in green light was 2.11 per minute. The rates differed by a factor of 3.5, showing that the photosensitivity of rhodopsin was greater at 550 nm than at 403 nm, consistent with the absorption spectrum of rhodopsin combined with the transmissions of the green and blue filters. 

Of interest is that the lines were straight over 2 and 3 ln units (Fig. 3 ; r ² > 0.99 for both). This implies that photoreversal is virtually completely avoided in green light and inhibited by NH₂OH in blue light. If photoreversal were to occur, as in the bleaching process with blue light in the absence of NH₂OH, the lines would be curvilinear—the more photoregeneration, the greater the curvature. This is further supported by direct measurements of the rhodopsin remaining after a 1-minute green light bleaching in the absence or presence of NH₂OH. The value was identical in both conditions (12%–13%) strongly suggesting that photoreversal of bleaching does not occur in long-wave green light (550 nm) or only very little so. 

A Model to Determine Bleaching

Development of the model was based largely on the value of K _b and on the shape of the blue no-NH₂OH curve in Figure 1 . Inspection of the function without NH₂OH shows that the curve declined very slowly after an initial bleaching burst and that a substantial amount of rhodopsin remained even after 30 minutes of exposure. If bleaching depended only on rhodopsin and a photoreversible intermediate, the system would be closed and the cis-trans isomerizations (bleaching and photoregeneration) would come into steady state, the level of which would be determined by the relative photosensitivities of rhodopsin and the blue absorber. However, the slow decline of the rhodopsin level in blue light indicated that the system was not closed; it was leaky. Photoreversibility was gradually lost, and a steady state could not be maintained, even though the light intensity was held constant. This leak in the system was most likely due to the slow decay of the photoreversible intermediates (MII) to products (opsin and free all-trans-retinal, which is converted rapidly to all-trans-retinol). The products enter the visual cycle to regenerate rhodopsin physiologically through a reisomerization–oxidation step of the all-trans-retinol in the pigment epithelium. 

The measured value of K _b and estimated values of K _pr and K _l were put into a set of differential rate equations as described in the Methods section, and the equations were solved simultaneously. The results of this simulation are given in Figure 4 , along with the experimentally determined data points of the blue bleach (in the absence of NH₂OH, Fig. 1 ). With K _b fixed at the measured value of 0.61 per minute, the best fit values of K _pr and K _l were 0.35 per minute and 0.075 per minute, respectively. This suggests that trans-cis (photoreversal) absorptions at 403 nm occur approximately half as often as do cis-trans (bleaching) absorptions (K _pr/K _b= 0.35/0.61 = 0.57). Also, the calculated leak rate was approximately one fourth the photoreversal rate (K _l/K _pr= 0.075/0.35 = 0.21), large enough to prevent a true photograph steady state between rhodopsin and the blue absorber. The good fit of the measured data points to the calculated curve indicates that the model with only two photoconvertible states and one leak step is sufficient to describe the data. 

Number of Photons Absorbed by Rhodopsin

The results in Figure 1 can be used to calculate the number of photons absorbed by rhodopsin and by the photoreversible blue absorber. The calculation begins with the measured rate of rhodopsin bleaching in blue light in the absence of photoreversal (NH₂OH curve, Fig. 1 ; also, Fig. 3 ). Whether the rod outer segments are dark adapted or partially bleached, rhodopsin (as distinct from the bleaching intermediate that absorbs blue light) in those outer segments absorbs photons at a rate given by K _b R _t/Γ, where K _b is the rate constant for rhodopsin bleaching with NH₂OH in blue light of 300μ W/cm²—namely, 0.61 per minute. R _t is the rhodopsin present at any time (t), and Γ is the quantum efficiency with which rhodopsin is bleached, independent of the wavelength (0.7 molecules bleached/absorbed photon). ¹²  

In Table 1 we summarize the results of the calculations to determine the numbers of photons absorbed by rhodopsin in the presence and absence of NH₂OH. We also present the numbers of photons absorbed by the blue absorber, assuming its time course is as given by the model (Fig. 4) and that it is MII. To obtain these numbers we integrated the curves graphically (and, for confirmation, by numerical means as well; data not shown) from t = 0 to t = 5, 10, and 30 minutes. The total number of absorptions by the average molecule (combining absorptions while in the rhodopsin state with those while in the MII state) during 30 minutes of blue light exposure (403 nm at 300 μW/cm²) was approximately 55. This is a surprisingly large amount of light absorbed by individual molecules. 

Because rhodopsin bleaching has a quantum efficiency of Γ = 0.7, 1.4 (1/0.7 = 1.4) photons must be absorbed, on average, to bleach one molecule. Similarly, the photoreversal of MII has a quantum efficiency of 0.22 for producing one rhodopsin ³⁴ ; therefore, 4.5 (1/0.22 = 4.5) photons, on average, must be absorbed by MII to produce one rhodopsin. 

Figure 5 shows an abbreviated scheme for conceptualizing bleaching and photoreversal in 403-nm light. To complete one bleaching and regeneration cycle, a rhodopsin molecule (and the reversible intermediate) must therefore absorb approximately six photons of blue light at the given wavelength and intensity. In the 30 minutes during which 55 absorptions occur, a single rhodopsin molecule completes on average approximately nine cycles of bleaching and regeneration. 

Discussion

These results show that photoreversal of a blue-absorbing, long-lived intermediate of rhodopsin bleaching could occur in vivo. This conclusion is drawn from four observations: (1) When photoreversal is inhibited by NH₂OH, bleaching with a steady blue light is rapid; (2) in such conditions the bleaching is simple first order, as befits pseudo–first-order photochemical processes; (3) when no NH₂OH is used, the bleaching is slow and when plotted as first order, the function is curved; and (4) when most of the rhodopsin in an eye is bleached with long-wave green, nonphotoreversing light, and the products of that bleach are exposed to blue light, substantial amounts of rhodopsin are regenerated by the blue exposure. 

Therefore, we conclude that the apparent slowness of rhodopsin bleaching by blue light, in the absence of NH₂OH, is not due to a failure of rhodopsin to absorb that light. Rather, the slowness is a result of the rapid regeneration of the rhodopsin by the photograph reisomerization of the all-trans-retinal that resides in the chromophoric site of a long-lived photoproduct. Before bleached rhodopsin can absorb another photon, it usually must be regenerated in the visual cycle, a slow process (15 minutes in darkness regenerated rhodopsin to 13% of dark levels after a 5-minute green light bleaching; data not shown) that involves shuttling of the all-trans-retinol to the pigment epithelium, enzymatic reisomerization to 11-cis-retinol, and the transport back to the photoreceptors. In blue light conditions, regeneration can occur very fast through the process of photoreversal, and a single rhodopsin molecule can therefore absorb many photons and can enter the bleaching pathway several times in a short period. 

Evidence suggests that the long-lived intermediate that is photoreversed by blue light may be MII: (1) The retinal is still attached to the chromophoric site in MII. Thus, any reisomerization of the MII retinal to 11- or 9-cis occurs in the chromophoric pocket of opsin and thus efficiently restores rhodopsin. (2) MII, similar to retinal, has an absorption maximum of 380 nm, ³⁵ which is sufficiently close to the 403 nm used in this study. MII absorbs light at 403 nm approximately 70% to 80% as efficiently as rhodopsin does at 500 nm. ³⁶ Furthermore, Williams ³⁴ found that in vitro 1 of 4.5 absorptions converts MII to rhodopsin, and therefore the quantum efficiency of this process (0.22) is high enough to photoconvert reasonable amounts of the all-trans to 11- or 9-cis, even at 403 nm. (3) MII is a long-lived bleaching intermediate with a lifetime measured in seconds. ³⁷ Our light source projects a large number of photons every second onto the cornea (∼3 × 10¹⁵ photons/sec). Even though approximately 20% of these photons (at 403 nm wavelength) are absorbed by the rat lens and therefore do not reach the retina ³⁸ and a large fraction of the photons is not absorbed by rhodopsin, ³¹ such intensities are sufficient to “hit” substantial numbers of MII molecules before they decay irreversibly. 

Other deep blue–absorbing photoproducts can be ruled out as candidates for the reversible intermediate, even though some of them fulfill some of the necessary criteria. These molecules include retinal, retinol, N-retinylidene-N-retinylphosphatidyl-ethanolamine, N-retinylidene-N-retinylethanolamine, ^{39 40} and N-retinylidene phosphatidylethanolamin. ¹¹ The spectra of these are appropriate for absorbing 403-nm light strongly. Of importance, however, is that the retinoid components of these molecules no longer occupy the chromophoric site on opsin, and they are therefore unlikely candidates for photoreversible molecules. Furthermore, after dissociation, retinal and retinol assume orientations in the disc membrane that are perpendicular to the disc. Thus, their absorption dipoles are orthogonal to the e vector of incoming light. ^{41 42} They are therefore extremely weak absorbers of light in vivo. The orientations in vivo of the other three compounds have not been determined. 

Another blue-absorbing candidate, metarhodopsin III (MIII,λ _max = 480 nm), should be considered separately from the others, because there is discord about its photoreversibility. Williams ³⁴ found that MIII was not photoreversible but Mathews et al. ⁸ reported that it was weakly and indirectly (through MII) reversible with light. Even if weakly reversible, the time course of MIII’s appearance does not fit our observations: Figure 1 shows that photoreversibility in the blue light was present early during the irradiation. This is evidenced by the fact that the two blue curves (with and without NH₂OH) diverged early—i.e., the difference in bleachability appeared soon during the exposures. In addition, the model (Fig. 4) indicates that photoreversibility began to be lost after approximately 1 minute as the product appeared. Taken together, these facts strongly support the notion that the photoreversible intermediate is MII, because MII appears within approximately a millisecond ^{4 13 15 34} and then begins a slow decay to MIII (or retinal). ^{8 9} If MIII were the photoreversible blue absorber, photoreversibility would have increased in long time interval as MIII appeared. Although we cannot completely exclude that a minor amount of rhodopsin is regenerated from MIII, we nevertheless conclude, by a process of elimination, that MII is the main reversible blue absorber in the rat retina. 

It is not clear what the molecular or cellular effects of so many photon absorptions by one molecule might be. It seems certain, however, that photoregenerated pigment can produce electrophysiological responses and thus constitutes functional rhodopsin. ^{14 43} Beyond this, however, a number of questions remain unanswered. (1) Is MII photoreversible when either transducin or arrestin is attached, or must these ligands be dissociated? Arnis and Hofmann, ⁴ using an excess of transducin, found in vitro that the binding to transducin inhibits reversal of the intermediate if no nucleotides are available, but it is not clear what happens in in vivo conditions. (2) What happens to transduction or to response shutoff if absorptions occur while MII is bound to these ligands? (3) Do some of the deleterious effects of blue light derive from interfering with these processes by trans-cis photoisomerization? Answers to such questions could help to elucidate whether damage in blue light is due simply to an excessive number of photon absorptions or to absorptions at certain time points (e.g., during transduction or response shutoff). 

In conclusion, we have shown that the apparent slowness of rhodopsin bleaching in blue light is not due to a failure of rhodopsin to absorb that light ⁴⁴ but is caused by the photoreversal of bleaching. Conclusions about the identity of the chromophore that mediates blue light–induced damage should therefore be reexamined. Further, it is possible from experiments such as these to calculate the numbers of photons absorbed by rhodopsin and by the photoreversible intermediate(s). The calculations indicate that previously unsuspected large numbers of photon absorptions are occurring in blue light, and this insight could be valuable in understanding blue light–induced retinal damage. 

 Each author contributed equally to this work.

 Supported by the Ernst and Berta Grimmke Foundation, Düsseldorf, Germany; the Schweizerische Unfallverhütungsanstalt Research Foundation, Luzern, Switzerland; the Bruppacher Foundation, Zurich, Switzerland; and Paolo Baiocchi, Intercast Europe, Parma, Italy. TPW received an Alexander Humboldt Scientist Award sponsored by R. Paulsen.

 Submitted for publication November 4, 1999; revised March 1 and June 7, 2000; accepted July 5, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

Bleaching of rhodopsin in vivo by blue and green light. Eyes of anesthetized rats were exposed to green light (▴; 550 nm; 409μ W/cm²), to blue light (○; 403 nm; 300μ W/cm²), or to blue light in the presence of NH₂OH to inhibit potential photoreversal of bleaching (•; 403 nm; 300 μW/cm²). Rhodopsin disappeared rapidly in green light and in blue light with NH₂OH, conditions with minimized photoreversal of bleaching. However, in blue light without NH₂OH, rhodopsin disappeared slowly, despite irradiation with the same photon fluxes as in the other two conditions. Each data point represents mean values of one to three independent experiments. Inset: Spectra of photoproducts formed after bleaching of rhodopsin in eyes of animals that were (+NH₂OH) or that were not (no NH₂OH) cardially perfused with 70 mM NH₂OH. The wavelength of maximal absorption in the absence of NH₂OH is at approximately 380 nm, the correct position for normal rhodopsin-bleaching products. However, the maximal absorption shifted to approximately 365 nm in NH₂OH-perfused animals indicating that NH₂OH reached the photoreceptors and formed the retinal oxime.

Figure 2.

 

Photoregeneration of rhodopsin from intermediates of the bleaching process. Eyes of anesthetized rats were exposed to green light (•; 550 nm; 409 μW/cm²) for up to 10 minutes. For some animals, the wavelength of the light was rapidly changed to 403 nm (▪; blue; 300 μW/cm²) after 5 minutes of green light exposure and irradiation continued for an additional 1 minute. Rhodopsin bleaching in green light generated photoproducts from which rhodopsin (28% ± 15%) could be photoreversed by the additional blue light exposure. Data points for green light bleach are from Figure 1 . Data points for photoregeneration are shown as mean ± SD from eight independent experiments.

Figure 3.

 

Determination of the kinetics of rhodopsin bleaching. Data from Figure 1 were plotted as integrated, first-order processes. R ₀ describes the dark value of rhodopsin, R _t stands for the amount of rhodopsin at time (t) during bleaching. The lines of bleaching in green light (▴; 550 nm; 409 μW/cm²) and of bleaching in blue light (•; 403 nm; 300 μW/cm²) in the presence of NH₂OH were straight over several half-lives, indicating that very little or no photoreversal of bleaching occurred in green light or in blue light with NH₂OH present. In contrast, the line of rhodopsin bleaching in blue light (○; 403 nm; 300μ W/cm²) without NH₂OH present was curvy, indicating that bleaching in this condition is not a simple, first-order process. The rate constants for rhodopsin bleaching in vivo were determined from the slopes of the straight lines. The values are K _b = 0.61 per minute (bleaching with blue light in the presence of NH₂OH) and K _g = 2.1 per minute (bleaching with green light).

Figure 4.

 

Computer simulation of the three-state model for bleaching of rhodopsin in blue light. The model involves rhodopsin (R), a photoreversible blue absorber (B), and an irreversible product (P). (▪) Experimentally determined data of rhodopsin bleaching in blue light (403 nm; 300 μW/cm²) without NH₂OH present (see Fig. 1 ). Because the data points are fit very well by the simulation, the result implies that a simple three-component model is sufficient to describe the bleaching of rhodopsin by blue light.

Table 1.

 

Numbers of Photons Absorbed by the Average Molecule in 403-nm Light

	Exposure Time (min)
		5	10	30
Rhodopsin: NH2OH present	1.34	1.37	1.39‡
Rhodopsin: NH2OH absent*	3.6	5.4	10.0
Blue absorber: NH2OH absent†	15.6	24.3	45.0
Totals (*)+ (†)	19.2	29.7	55.0

‡  When bleaching is complete, this value is 1.4, as required by the known photosensitivity of rhodopsin in nonphotoreversing light.

Figure 5.

 

Scheme for calculating the total number of photons absorbed by rhodopsin (R) and the photoreversible intermediate, assuming it is MII. Photoregenerated rhodopsin is shown as R′. The numbers on the arrows indicate the average numbers of photons needed to cause the reactions and reflect the fact that the quantum efficiencies of bleaching and photoreversal are not unity.