Resonance Raman spectroscopy of octopus rhodopsin and its photoproducts

We report here the resonance Raman spectra of octopus rhodopsin and its photoproducts, bathorhodopsin and acid metarhodopsin. These studies were undertaken in order to make comparisons with the well-studied bovine pigments, so as to understand the similarities and the differences in pigment structure and photochemical processes between vertebrates and invertebrates. The flow method was used to obtain the Raman spectrum of rhodopsin at 13 degrees C. The bathorhodopsin spectrum was obtained by computer subtraction of the spectra containing different photostationary mixtures of rhodopsin, isorhodopsin, hypsorhodopsin, and bathorhodopsin, obtained at 12 K using the pump-probe technique and from measurements at 80 K. Like their bovine counterparts, the Schiff base vibrational mode appears at approximately 1660 cm-1 in octopus rhodopsin and the photoproducts, bathorhodopsin and acid metarhodopsin, suggesting a protonated Schiff base linkage between the chromophore and the protein. Differences between the Raman spectra of octopus rhodopsin and bathorhodopsin indicate that the formation of bathorhodopsin is associated with chromophore isomerization. This inference is substantiated by the chromophore chemical extraction data which show that, like the bovine system, octopus rhodopsin is an 11-cis pigment, while the photoproducts contain an all-trans pigment, in agreement with previous work. The octopus rhodopsin and bathorhodopsin spectra show marked differences from their bovine counterparts in other respects, however. The differences are most dramatic in the structure-sensitive fingerprint and the HOOP regions. Thus, it appears that although the two species differ in the specific nature of the chromophore-protein interactions, the general process of visual transduction is the same.     

Photolysis intermediates of the artificial visual pigment cis-5,6-dihydro-isorhodopsin.

The photolysis intermediates of an artificial bovine rhodopsin pigment, cis-5,6-dihydro-isorhodopsin (cis-5,6,-diH-ISORHO, lambda max 461 nm), which contains a cis-5,6-dihydro-9-cis-retinal chromophore, are investigated by room temperature, nanosecond laser photolysis, and low temperature irradiation studies. The observations are discussed both in terms of low temperature experiments of Yoshizawa and co-workers on trans-5,6-diH-ISORHO (Yoshizawa, T., Y. Shichida, and S. Matuoka. 1984. Vision Res. 24: 1455-1463), and in relation to the photolysis intermediates of native bovine rhodopsin (RHO). It is suggested that in 5,6-diH-ISORHO, a primary bathorhodopsin intermediate analogous to the bathorhodopsin intermediate (BATHO) of the native pigment, rapidly converts to a blue-shifted intermediate (BSI, lambda max 430 nm) which is not observed after photolysis of native rhodopsin. The analogs from lumirhodopsin (LUMI) to meta-II rhodopsin (META-II) are generated subsequent to BSI, similar to their generation from BATHO in the native pigment. It is proposed that the retinal chromophore in the bathorhodopsin stage of 5,6-diH-ISORHO is relieved of strain induced by the primary cis to trans isomerization by undergoing a geometrical rearrangement of the retinal. Such a rearrangement, which leads to BSI, would not take place so rapidly in the native pigment due to ring-protein interactions. In the native pigment, the strain in BATHO would be relieved only on a longer time scale, via a process with a rate determined by protein relaxation.     

A dark and constitutively active mutant of the tiger salamander UV pigment

A triple mutant (F86L/T93P/S118T; bovine rhodopsin numbering) of the tiger salamander UV cone pigment appears to be trapped in an open conformation that is metarhodopsin-II-like. The pigment is able to activate transducin in the dark, and the ligand-free apoprotein is also able to activate transducin constitutively. The pigment permits protons and chloride ions from solution access to the active site as it displays a pH- and NaCl-dependent absorption spectrum not observed with the wild-type pigment. However, the wild-type properties of light-dependent activity and a pH-independent absorption spectrum are recovered upon reconstitution of the triple mutant with 11-cis-9-demethyl retinal. These results suggest that binding the native chromophore cannot deactivate the protein because of steric interactions between the protein, possibly residue 118, and the 9-methyl group of the chromophore. Furthermore, the absorption spectrum of the 9-demethyl retinal regenerated pigment exhibits a band broader and with lower extinction at the absorption maximum than either the human blue or salamander UV wild-type pigments generated with the same retinal analogue. The broad spectrum appears to be comprised of two or more species and can be well-fit by a sum of scaled spectra of the two wild-type pigments. Binding the chromophore appears to trap the pigment in two or more conformations. The triple mutant reported here represents the first example of a dark-active cone pigment and constitutively active cone opsin.     

Resonance raman spectroscopy of an ultraviolet-sensitive insect rhodopsin

We present the first visual pigment resonance Raman spectra from the UV-sensitive eyes of an insect, Ascalaphus macaronius (owlfly). This pigment contains 11-cis-retinal as the chromophore. Raman data have been obtained for the acid metarhodopsin at 10 degrees C in both H2O and D2O. The C = N stretching mode at 1660 cm-1 in H2O shifts to 1631 cm-1 upon deuteriation of the sample, clearly showing a protonated Schiff base linkage between the chromophore and the protein. The structure-sensitive fingerprint region shows similarities to the all-trans-protonated Schiff base of model retinal chromophores, as well as to the octopus acid metarhodopsin and bovine metarhodopsin I. Although spectra measured at -100 degrees C with 406.7-nm excitation, to enhance scattering from rhodopsin (lambda max 345 nm), contain a significant contribution from a small amount of contaminants [cytochrome(s) and/or accessory pigment] in the sample, the C = N stretch at 1664 cm-1 suggests a protonated Schiff base linkage between the chromophore and the protein in rhodopsin as well. For comparison, this mode also appears at approximately 1660 cm-1 in both the vertebrate (bovine) and the invertebrate (octopus) rhodopsins. These data are particularly interesting since the absorption maximum of 345 nm for rhodopsin might be expected to originate from an unprotonated Schiff base linkage. That the Schiff base linkage in the owlfly rhodopsin, like in bovine and in octopus, is protonated suggests that a charged chromophore is essential to visual transduction.     

The endogenous chromophore of retinal G protein-coupled receptor opsin from the pigment epithelium

The recent identification of nonvisual opsins has revealed an expanding family of vertebrate opsin genes. The retinal pigment epithelium (RPE) and Müller cells contain a blue and UV light-absorbing opsin, the RPE retinal G protein-coupled receptor (RGR, or RGR opsin). The spectral properties of RGR purified from bovine RPE suggest that RGR is conjugated in vivo to a retinal chromophore through a covalent Schiff base bond. In this study, the isomeric structure of the endogenous chromophore of RGR was identified by the hydroxylamine derivatization method. The retinaloximes derived from RGR in the dark consisted predominantly of the all-trans isomer. Irradiation of RGR with 470-nm monochromatic or near-UV light resulted in stereospecific isomerization of the bound all-trans-retinal to an 11-cis configuration. The stereospecificity of photoisomerization of the all-trans-retinal chromophore of RGR was lost by denaturation of the protein in SDS. Under the in vitro conditions, the photosensitivity of RGR is at least 34% that of bovine rhodopsin. These results provide evidence that RGR is bound in vivo primarily to all-trans-retinal and is capable of operating as a stereospecific photoisomerase that generates 11-cis-retinal in the pigment epithelium.     

A comparative study of the infrared difference spectra for octopus and bovine rhodopsins and their bathorhodopsin photointermediates

Fourier-transform infrared difference spectroscopy has been used to detect the vibrational modes in the chromophore and protein that change in position and intensity between octopus rhodopsin and its photoproducts formed at low temperature (85 K), bathorhodopsin and isorhodopsin. The infrared difference spectra between octopus rhodopsin and octopus bathorhodopsin, octopus bathorhodopsin and octopus isorhodopsin, and octopus isorhodopsin and octopus rhodopsin are compared to analogous difference spectra for the well-studied bovine pigments, in order to understand the similarities in pigment structure and photochemical processes between the vertebrate and invertebrate systems. The structure-sensitive fingerprint region of the infrared spectra for octopus bathorhodopsin shows strong similarities to spectra of both all-trans-retinal and bovine bathorhodopsin, thus confirming chemical extraction data that suggest that octopus bathorhodopsin contains an all-trans-retinal chromophore. In contrast, we find dramatic differences in the hydrogen out-of-plane modes of the two bathorhodopsins, and in the fingerprint lines of the rhodopsins and isorhodopsins for the two pigments. These observations suggest that while the primary effect of light in the octopus rhodopsin system, as in the bovine rhodopsin system, is 11-cis/11-trans isomerization, the protein-chromophore interactions for the two systems are quite different. Finally, striking similarities and differences in infrared lines attributable to changes in amino acid residues in the opsin are found between the two pigment systems. They suggest that no carboxylic acid or tyrosine residues are affected in the initial changes of light-energy transduction in octopus rhodopsin. Comparing the amino acid sequences for octopus and bovine pigments also allows us to suggest that the carboxylic acid residues altered in the bovine transitions are Glu-122 and/or Glu-134.     

Bathoproducts of rhodopsin, isorhodopsin I, and isorhodopsin II.

Bathorhodopsins were prepared by partially (10--15%) photoconverting bovine rhodopsin (11-cis chromophore) or isorhodopsin I (9-cis chromophore) at 77 degrees K; care was taken to avoid establishing photostationary states. The absorption spectra calculated for the bathorhodopsins derived from the two parent pigments are identical in their lambda max 'S, bandwidths, and extinction coefficients. This result provides further support for the hypothesis that bathorhodopsin is a common intermediate between an 11-cis pigment (rhodopsin) and a 9-cis one (isorhodopsin I) and thus probably has an all-trans chromophore. This in turn is strong evidence for the cis-trans isomerization model of the primary event in vision. The spectrum of the bathoproduct of isorhodopsin II (9,13-dicis chromophore) is different from the other pigments' bathoproducts.     

Characterization of the primary photointermediates of Drosophila rhodopsin

Invertebrate opsins are unique among the visual pigments because the light-activated conformation, metarhodopsin, is stable following exposure to light in vivo. Recovery of the light-activated pigment to the dark conformation (or resting state) occurs either thermally or photochemically. There is no evidence to suggest that the chromophore becomes detached from the protein during any stage in the formation or recovery processes. Biochemical and structural studies of invertebrate opsins have been limited by the inability to express and purify rhodopsins for structure-function studies. In this study, we used Drosophila to produce an epitope-tagged opsin, Rh1-1D4, in quantities suitable for spectroscopic and photochemical characterization. When expressed in Drosophila, Rh1-1D4 is localized to the rhabdomere membranes, has the same spectral properties in vivo as wild-type Rh1, and activates the phototransduction cascade in a normal manner. Purified Rh1-1D4 visual pigment has an absorption maximum of the dark-adapted state of 474 nm, while the metarhodopsin absorption maximum is 572 nm. However, the metarhodopsin state is not stable as purified in dodecyl maltoside but decays with kinetics that require a double-exponential fit having lifetimes of 280 and 2700 s. We investigated the primary properties of the pigment at low temperature. At 70 K, the pigment undergoes a temperature-induced red shift to 486 nm. Upon illumination with 435 nm light, a photostationary state mixture is formed consisting of bathorhodopsin (lambda(max) = 545 nm) and isorhodopsin (lambda(max) = 462 nm). We also compared the spectroscopic and photochemical properties of this pigment with other vertebrate pigments. We conclude that the binding site of Drosophila rhodopsin is similar to that of bovine rhodopsin and is characterized by a protonated Schiff base chromophore stabilized via a single negatively charged counterion.     

Spectral tuning in the human blue cone pigment.

The first determination of the absolute absorption maximum of the human blue cone visual pigment is presented. After expression in COS cells, reconstitution with 11-cis-retinal, and purification, the blue pigment exhibits an absolute absorption maximum of 414 nm. The pigment reacts rapidly with hydroxylamine in the dark and is capable of activating bovine rod transducin in a light-dependent manner. Products of mutations of proposed spectral tuning residues in the blue pigment do not behave as predicted when using rhodopsin mutants as a model. Mutations of amino acids in the ring portion of the chromophore binding pocket of rhodopsin serve well as a predictive model for mutations in the blue pigment, but mutations near the Schiff base do not.     

Photoisomerization efficiency in UV-absorbing visual pigments: protein-directed isomerization of an unprotonated retinal Schiff base

A visual pigment consists of an opsin protein and a chromophore, 11-cis-retinal, which binds to a specific lysine residue of opsin via a Schiff base linkage. The Schiff base chromophore is protonated in pigments that absorb visible light, whereas it is unprotonated in ultraviolet-absorbing visual pigments (UV pigments). To investigate whether an unprotonated Schiff base can undergo photoisomerization as efficiently as a protonated Schiff base in the opsin environment, we measured the quantum yields of the bovine rhodopsin E113Q mutant, in which the Schiff base is unprotonated at alkaline pH, and the mouse UV pigment (mouse UV). Photosensitivities of UV pigments were measured by irradiation of the pigments followed by chromophore extraction and HPLC analysis. Extinction coefficients were estimated by comparing the maximum absorbances of the original pigments and their acid-denatured states. The quantum yield of the bovine rhodopsin E113Q mutant at pH 8.2, where the Schiff base is unprotonated, was significantly lower than that of wild-type rhodopsin, whereas the mutant gave a quantum yield almost identical to that of the wild type at pH 5.5, where the Schiff base is protonated. These results suggest that Schiff base protonation plays a role in increasing quantum yield. The quantum yield of mouse UV, which has an unprotonated Schiff base chromophore, was significantly higher than that of the unprotonated form of the rhodopsin E113Q mutant, although it was still lower than the visible-absorbing pigments. These results suggest that the mouse UV pigment has a specific mechanism for the efficient photoisomerization of its unprotonated Schiff base chromophore.     

Comparative study on the chromophore binding sites of rod and red-sensitive cone visual pigments by use of synthetic retinal isomers and analogues

A comparative study on the chromophore (retinal) binding sites of the opsin (R-photopsin) from chicken red-sensitive cone visual pigment (iodopsin) and that scotopsin) from bovine rod pigment (rhodopsin) was made by the aid of geometric isomers of retinal (all-trans, 13-cis, 11-cis, 9-cis, and 7-cis) and retinal analogues including fluorinated (14-F, 12-F, 10-F, and 8-F) and methylated (12-methyl) 11-cis-retinals. The stereoselectivity of R-photopsin for the retinal isomers and analogues was almost identical with that of scotopsin, indicating that the shapes of the chromophore binding sites of both opsins are similar, although the former appears to be somewhat more restricted than the latter. The rates of pigment formation from R-photopsin were considerably greater than those from scotopsin. In addition, all the iodopsin isomers and analogues were more susceptible to hydroxylamine than were the rhodopsin ones. These observations suggest that the retinal binding site of iodopsin is located near the protein surface. On the basis of the spectral properties of fluorinated analogues, a polar group in the chromophore binding site of iodopsin as well as rhodopsin was estimated to be located near the hydrogen atom at the C10 position of the retinylidene chromophore. A large difference in wavelength between the absorption maxima of iodopsin and rhodopsin was significantly reduced in the 9-cis and 7-cis pigments. On the assumption that the retinylidene chromophore is anchored rigidly at the alpha-carbon of the lysine residue and loosely at the cyclohexenyl ring, each of the two isomers would have the Schiff-base nitrogen at a position altered from that of the 11-cis pigments.(ABSTRACT TRUNCATED AT 250 WORDS)     

Asp83, Glu113 and Glu134 are not specifically involved in Schiff base protonation or wavelength regulation in bovine rhodopsin

Site-specific mutagenesis was employed to investigate the proposed contribution of proton-donating residues (Glu, Asp) in the membrane domains of bovine rhodopsin to protonation of the Schiff base-linking protein and chromophore or to wavelength modulation of this visual pigment. Three point-mutations were introduced to replace the highly conserved residues Asp83 by Asn (D83N), Glu113 by Gln (E113 Q) or Glu134 by Asp (E134D), respectively. All 3 substitutions had only marginal effects on the spectral properties of the final pigment (less than or equal to 3 nm blue-shift relative to native rhodopsin). Hence, none of these residues by itself is specifically involved in Schiff base protonation or wavelength modulation of bovine rhodopsin.     

Wavelength regulation in iodopsin, a cone pigment.

The opsin shift, the difference in wavenumber between the absorption peak of a visual pigment and the protonated Schiff base of the chromophore, represents the influence of the opsin binding site on the chromophore. The opsin shift for the chicken cone pigment iodopsin is much larger than that for rhodopsin. To understand the origin of this opsin shift and the mechanism of wavelength regulation in iodopsin, a series of synthetic 9-cis and 11-cis dehydro- and dihydro-retinals was used to regenerate iodopsin-based pigments. The opsin shifts of these pigments are quite similar to those found in bacteriorhodopsin-based artificial pigments. On the basis of these studies, a tentative model of wavelength regulation in iodopsin is proposed.     

The retinal conformation and its environment in rhodopsin in light of a new 2.2 A crystal structure

A new high-resolution structure is reported for bovine rhodopsin, the visual pigment in rod photoreceptor cells. Substantial improvement of the resolution limit to 2.2 A has been achieved by new crystallization conditions, which also reduce significantly the probability of merohedral twinning in the crystals. The new structure completely resolves the polypeptide chain and provides further details of the chromophore binding site including the configuration about the C6-C7 single bond of the 11-cis-retinal Schiff base. Based on both an earlier structure and the new improved model of the protein, a theoretical study of the chromophore geometry has been carried out using combined quantum mechanics/force field molecular dynamics. The consistency between the experimental and calculated chromophore structures is found to be significantly improved for the 2.2 A model, including the angle of the negatively twisted 6-s-cis-bond. Importantly, the new crystal structure refinement reveals significant negative pre-twist of the C11-C12 double bond and this is also supported by the theoretical calculation although the latter converges to a smaller value. Bond alternation along the unsaturated chain is significant, but weaker in the calculated structure than the one obtained from the X-ray data. Other differences between the experimental and theoretical structures in the chromophore binding site are discussed with respect to the unique spectral properties and excited state reactivity of the chromophore.     

Melanin in human irides of different color and age of donors

Melanin is the main chromophore of the human iris. This pigment is considered to be the most important factor that determines the color of the irides. Previous studies based mainly on chemical degradation methods showed that brown irides contain more melanin than blue ones. In our study, we used electron spin resonance (ESR) spectroscopy to detect and characterize melanin free radical centers and associated iron in human irides. Based on this method, we determined the amount of melanin in the irides and the relative content of iron in iridial melanin as a function of their color, shade, and the age of their donors. Chemical degradation of iridial homogenates enabled us to characterize the structure of eumelanin and determine the content of pheomelanin present in human and bovine irides. The ESR amplitude, the normalized intensity obtained by double integration of the ESR signal of melanin, and the content of the pigment in the irides depended on color and shade of the eyes being 40% higher in the brown group of the irides compared with all other groups. On the other hand, the relative iron content normalized to the melanin content in light blue irides showed a small decrease with age of donors. Melanin in human and bovine irides was mostly composed of eumelanin, and pheomelanin content was of the order of a few percent. Although some differences in the structure of eumelanin present in the human and bovine irides are possible, the results obtained in this study suggest that human irides contain eumelanin with very similar chemical properties.     

Synthesis of 11-cis-locked bicyclo[5.1.0]octanyl retinal and an enantioselective binding to bovine opsin

Both enantiomers of 13-(E) and 13-(Z) isomers of 11-cis-locked bicyclo[5.1.0]octanyl retinal were prepared by an improved synthesis and incubated with bovine opsin. The synthesis also establishes the absolute configuration of the enantiomers. Only one of the enantiomers binds to opsin, thus showing the steric restrictions regarding the middle polyene moiety of the retinoid molecule; this is in sharp contrast to the known leniency of the ring moiety binding site of retinoids. However, although one enantiomer is incorporated into the pigment, the circular dichroic spectrum of the pigment incorporating the bound enantiomer yields only a very weak Cotton effect, showing that, once incorporated, the bicyclo[5.1.0]octanyl chromophore is flattened by the opsin binding site. The titled retinoid was synthesized for study of the absolute conformation of the retinal pigment in rhodopsin.     

Raman microscope and quantum yield studies on the primary photochemistry of A2-visual pigments.

The 77-K resonance Raman vibrational spectrum of intact goldfish rod photoreceptors containing 3,4-dehydro (A2) retinal is dominated by scattering from the 9-cis component of the steady state at all excitation wavelengths. Intact goldfish photoreceptors were regenerated with an A1-retinal chromophore to determine whether this behavior is caused by the protein or the chromophore. The resulting Raman spectrum was typical of an A1-pigment exhibiting significant scattering from all three components of the steady state: rhodopsin, bathorhodopsin, and isorhodopsin. Furthermore, regeneration of bovine opsin with A2-retinal produces a characteristic "A2-Raman spectrum" that is dominated by scattering from the 9-cis pigment. We conclude that the differences between the Raman spectra of the A1-and A2-pigments are caused by some intrinsic difference in the photochemical properties of the retinal chromophores. To quantitate these observations, the 77-K adsorption spectra and the photochemical quantum yields (phi) of the native A2-goldfish and the regenerated A2-bovine pigments were measured. In the goldfish A2-pigment, the value of phi 4 (9-cis----trans) is 0.05; phi 3 (trans----9-cis) is 0.10; and phi 2 (trans----11-cis) is 0.35. By contrast, in the bovine A1-pigment, these quantum yields are 0.10, 0.053, and 0.50, respectively. The reduced value of phi 4 and the increased value of phi 3 in the goldfish pigment confirms that the 9-cis isomer is photochemically more stable in A2-pigments.     

Synthesis of the all-trans-retinal chromophore of retinal G protein-coupled receptor opsin in cultured pigment epithelial cells.

Light-dependent production of 11-cis-retinal by the retinal pigment epithelium (RPE) and normal regeneration of rhodopsin under photic conditions involve the RPE retinal G protein-coupled receptor (RGR) opsin. This microsomal opsin is bound to all-trans-retinal which, upon illumination, isomerizes stereospecifically to the 11-cis isomer. In this paper, we investigate the synthesis of the all-trans-retinal chromophore of RGR in cultured ARPE-hRGR and freshly isolated bovine RPE cells. Exogenous all-trans-[(3)H]retinol is incorporated into intact RPE cells and converted mainly into retinyl esters and all-trans-retinal. The intracellular processing of all-trans-[(3)H]retinol results in physiological binding to RGR of a radiolabeled retinoid, identified as all-trans-[(3)H]retinal. The ARPE-hRGR cells contain a membrane-bound NADPH-dependent retinol dehydrogenase that reacts efficiently with all-trans-retinol but not the 11-cis isomer. The NADPH-dependent all-trans-retinol dehydrogenase activity in isolated RPE microsomal membranes can be linked in vitro to specific binding of the chromophore to RGR. These findings provide confirmation that RGR opsin binds the chromophore, all-trans-retinal, in the dark. A novel all-trans-retinol dehydrogenase exists in the RPE and performs a critical function in chromophore biosynthesis.     

Early photolysis intermediates of the artificial visual pigment 13-demethylrhodopsin

Nanosecond time-resolved absorption measurements are reported for the room temperature photolysis of a modified rhodopsin pigment, 13-demethylrhodopsin, which contains the chromophore 13-demethylretinal. The measurements are consistent with the formation of an equilibrium between a BA-THO-13-demethylrhodopsin species and a blue-shifted species (relative to the parent pigment), BSI-13-demethylrhodopsin. The results are compared to those acquired after photolysis of native bovine rhodopsin [Hug, S. J., Lewis, J. W., Einterz, C. M., Thorgeirsson, T. E., & Kliger, D. S. (1990) Biochemistry (preceding paper in this issue)] and to results obtained after photolysis of several modified isorhodopsin pigments in which the BSI species was first observed. It is concluded that in all of the pigments the results are consistent with the formation of an equilibrium between BATHO and BSI, which subsequently decays on a nanosecond time scale at room temperature to a lumirhodopsin intermediate.     

An Alternative Pathway Mediates the Mouse and Human Cone Visual Cycle

One of the fundamental mysteries of the human visual system is the continuous function of cone photoreceptors in bright daylight. As visual pigment is destroyed, or bleached, by light [1], cones require its rapid regeneration, which in turn involves rapid recycling of the pigment's chromophore. The canonical visual cycle for rod and cone pigments involves recycling of their chromophore from all-trans retinol to 11-cis retinal in the pigment epithelium, adjacent to photoreceptors [2]. However, shortcomings of this pathway indicate the function of a second, cone-specific, mechanism for chromophore recycling [3]. Indeed, biochemical [3], [4], [5], [6] and [7] and physiological [8] studies on lower species have described a cone-specific visual cycle in addition to the long-known pigment epithelium pathway. Two important questions remain, however: what is the role of this pathway in the function of mammalian cones, and is it present in higher mammals, including humans? Here, we show that mouse, primate, and human neural retinas promote pigment regeneration and dark adaptation selectively in cones, but not in rods. This pathway supports rapid dark adaptation of mammalian cones and extends their dynamic range in background light independently of the pigment epithelium. This pigment-regeneration mechanism is essential for our daytime vision and appears to be evolutionarily conserved.