Resonance Raman microscopy of rod and cone photoreceptors.

We have constructed a Raman microscope that has enabled us to obtain resonance Raman vibrational spectra from single photoreceptor cells. The laser beam which excites the Raman scattering is focused on the outer segment of the photoreceptor through the epiillumination system of a light microscope. Raman scattering from the visual pigment in the photoreceptor is collected by the objective and then dispersed onto a multichannel detector. High-quality spectra are recorded easily from individual outer segments that are 5 x 50 micrometer in size, and we have obtained spectra from cells as small as 1 x 10 micrometer. We have used the Raman microscope to study photostationary steady-state mixtures in pigments from toad (Bufo marinus) and goldfish (Carassius auratus) photoreceptors; these photoreceptors were frozen in glycerol glasses at 77 degrees K. Comparison of our toad red rod spectra with previously published spectra of bovine rod pigments demonstrates that the conformation of the chromophore in the first photointermediate, bathorhodopsin, is sensitive to variations in protein structure. We have also studied the first photointermediate in the goldfish rod photostationary steady-state. This bathoporphyropsin has a much lower ethylenic stretching frequency (1,507 cm-1) than that observed in the toad and bovine bathoproducts (approximately 1,535 cm-1). Preliminary results of our work on goldfish cone pigments are also reported. These are the first vibrational studies on the vertebrate photoreceptors responsible for color vision.

Vibrationally coherent photochemistry in the femtosecond primary event of vision.

Femtosecond pump-probe experiments reveal the impulsive production of photoproduct in the primary event in vision. The retinal chromophore of rhodopsin was excited with a 35-femtosecond pulse at 500 nanometers, and transient changes in absorption were measured with 10-femtosecond probe pulses. At probe wavelengths within the photo-product absorption band, oscillatory features with a period of 550 femtoseconds (60 wavenumbers) were observed whose phase and amplitude demonstrate that they are the result of nonstationary vibrational motion in the ground state of the photoproduct. The observation of coherent vibrational motion of the photoproduct supports the idea that the primary step in vision is a vibrationally coherent process and that the high quantum yield of the cis-->trans isomerization in rhodopsin is a consequence of the extreme speed of the excited-state torsional motion.

Activation and regeneration of rhodopsin in the insect visual cycle.

Light absorption by rhodopsin generates metarhodopsin, which activates heterotrimeric guanine nucleotide-binding proteins (G proteins) in photoreceptor cells of vertebrates and invertebrates. In contrast to vertebrate metarhodopsins, most invertebrate metarhodopsins are thermally stable and regenerate rhodopsin by absorption of a second photon. In experiments with Rh1 Drosophila rhodopsin, the thermal stability of metarhodopsin was found not to be an intrinsic property of the visual pigment but a consequence of its interaction with arrestin (49 kilodaltons). The stabilization of metarhodopsin resulted in a large decrease in the efficiency of G protein activation. Light absorption by thermally stable metarhodopsin initially regenerated an inactive rhodopsin-like intermediate, which was subsequently converted in the dark to active rhodopsin. The accumulation of inactive rhodopsin at higher light levels may represent a mechanism for gain regulation in the insect visual cycle.

Arrestin function in inactivation of G protein-coupled receptor rhodopsin in vivo.

Arrestins have been implicated in the regulation of many G protein-coupled receptor signaling cascades. Mutations in two Drosophila photoreceptor-specific arrestin genes, arrestin 1 and arrestin 2, were generated. Analysis of the light response in these mutants shows that the Arr1 and Arr2 proteins are mediators of rhodopsin inactivation and are essential for the termination of the phototransduction cascade in vivo. The saturation of arrestin function by an excess of activated rhodopsin is responsible for a continuously activated state of the photoreceptors known as the prolonged depolarized afterpotential. In the absence of arrestins, photoreceptors undergo light-dependent retinal degeneration as a result of the continued activity of the phototransduction cascade. These results demonstrate the fundamental requirement for members of the arrestin protein family in the regulation of G protein-coupled receptors and signaling cascades in vivo.

Site of G protein binding to rhodopsin mapped with synthetic peptides from the alpha subunit.

The interaction between receptors and guanine nucleotide binding (G) proteins leads to G protein activation and subsequent regulation of effector enzymes. The molecular basis of receptor-G protein interaction has been examined by using the ability of the G protein from rods (transducin) to cause a conformational change in rhodopsin as an assay. Synthetic peptides corresponding to two regions near the carboxyl terminus of the G protein alpha subunit, Glu311-Val328 and Ile340-Phe350, compete with G protein for interaction with rhodopsin. Amino acid substitution studies show that Cys321 is required for this effect. Ile340-Phe350 and a modified peptide, acetyl-Glu311-Lys329-amide, mimic G protein effects on rhodopsin conformation, showing that these peptides bind to and stabilize the activated conformation of rhodopsin.

Movement of retinal along the visual transduction path.

Movement of the ligand/receptor complex in rhodopsin (Rh) has been traced. Bleaching of diazoketo rhodopsin (DK-Rh) containing 11-cis-3-diazo-4-oxo-retinal yields batho-, lumi-, meta-I-, and meta-II-Rh intermediates corresponding to those of native Rh but at lower temperatures. Photoaffinity labeling of DK-Rh and these bleaching intermediates shows that the ionone ring cross-links to tryptophan-265 on helix F in DK-Rh and batho-Rh, and to alanine-169 on helix D in lumi-, meta-I-, and meta-II-Rh intermediates. It is likely that these movements involving a flip-over of the chromophoric ring trigger changes in cytoplasmic membrane loops resulting in heterotrimeric guanine nucleotide-binding protein (G protein) activation.

Spectral tuning of rhodopsin and metarhodopsin in vivo.

Color vision is dependent upon the expression of spectrally distinct forms of rhodopsin in different photoreceptor cells. To identify the structural features of rhodopsin that regulate spectral sensitivity and absorption in vivo, we have constructed a series of chimeric Drosophila rhodopsin molecules, derived from a blue- and a violet-sensitive rhodopsin, and used P element-mediated germline transformation to generate transgenic flies that express the modified pigments in the R1-R6 photoreceptor cells of the compound eye. Our analysis of these animals indicates that multiple regions of the opsin protein are involved in regulating rhodopsin spectral sensitivity and that the native and photoactivated forms of rhodopsin can be tuned independently of each other. These results demonstrate the feasibility of designing receptor molecules with specifically modified activated states.

Histidine residues regulate the transition of photoexcited rhodopsin to its active conformation, metarhodopsin II.

The biologically active photoproduct of rhodopsin, metarhodopsin II (M II), exists in a pH-sensitive equilibrium with its precursor, metarhodopsin I (M I). Increasing acidity favors M II, with the midpoint of the pH titration curve at pH 6.4. To test the long-standing proposal that histidine protonation regulates this conformational transition, we characterized mutant rhodopsins in which each of the 6 histidines was replaced by phenylalanine or cysteine. Only mutants substituted at the 3 conserved histidines showed abnormal M I-M II equilibria. Those in which His-211 was replaced by phenylalanine or cysteine formed little or no M II at either extreme of pH, whereas mutants substituted at His-65 or at His-152 showed enhanced sensitivity to protons. The simplest interpretation of these results is that His-211 is the site where protonation strongly stabilizes the M II conformation and that His-65 and His-152 are sites where protonation modestly destabilizes the M II conformation.

A molecular pathway for light-dependent photoreceptor apoptosis in Drosophila.

Light-induced photoreceptor apoptosis occurs in many forms of inherited retinal degeneration resulting in blindness in both vertebrates and invertebrates. Though mutations in several photoreceptor signaling proteins have been implicated in triggering this process, the molecular events relating light activation of rhodopsin to photoreceptor death are yet unclear. Here, we uncover a pathway by which activation of rhodopsin in Drosophila mediates apoptosis through a G protein-independent mechanism. This process involves the formation of membrane complexes of phosphorylated, activated rhodopsin and its inhibitory protein arrestin, and subsequent clathrin-dependent endocytosis of these complexes into a cytoplasmic compartment. Together, these data define the proapoptotic molecules in Drosophila photoreceptors and indicate a novel signaling pathway for light-activated rhodopsin molecules in control of photoreceptor viability.

Resonance Raman spectra of octopus acid and alkaline metarhodopsins.

The resonance Raman spectra of acid and alkaline metarhodopsins of octopus were measured. The acid metarhodopsin exhibited the Schiff base C = N stretching band at 1655 cm-1 in H2O and at 1625 cm-1 in 2H2O, and therefore the Schiff base is shown to be protonated. The C = C stretching band was observed at 1548 and 1572 cm-1 for acid and alkaline metarhodopsins, respectively. Other Raman lines of octopus acid metarhodopsin were assigned from the data of Cookingham et al. (Cookingham, R.E., Lewis, A. and Lemley, A.T. (1978) Biochemistry 17, 4699-4711). Frequencies of some structure-sensitive Raman lines differed among octopus, squid and bovine metarhodopsins; such differences may be important in interpreting the presence of optical activity in octopus metarhodopsins but its absence in squid metarhodopsins.

Circular dichroism of squid rhodopsin and its intermediates.

Circular dichroism (CD) and absorption spectra of squid (Todarodes pacificus) rhodopsin, isorhodopsin and the intermediates was measured at low temperatures. Squid rhodopsin has positive CD bands at wavelengths corresponding the alpha- and beta-absorption bands at liquid nitrogen temperature (CD maxima: 485 nm at alpha-band and 348 nm at beta-band) as well as at room temperature (CD maxima: 474 nm at alpha-band and 347 nm at beta-band). The rotational strength of the alpha-band has a molecular ellipticity about twice that of cattle rhodopsin. The CD spectrum of bathorhodopsin displays a negative peak at 532 nm, the rotational strength of which has an absolute value slightly larger than that of rhodopsin. The reversal in sign at alpha-band of the CD spectrum may indicate that the isomerization of retinal chromophore from twisted 11-cis form to twisted 11-trans form has occurred in the process of conversion from rhodopsin to bathorhodopsin. Lumirhodopsin has a small negative CD band at 490 nm, the maximum of which lies at 25 nm shorter wavelengths than the absorption maximum (515 nm), and a large positive CD band near 290 nm, which is not observed in rhodopsin and the other intermediates. This band may de derived from a conformational change of the opsin. In the process of changing from lumirhodopsin to LM-rhodopsin, The CD bands at visible and near ultraviolet regions disappear. Both alkaline and acid metarhodopsins have no CD bands at visible and near ultraviolet regions.

Effect of phospholipid and detergent on the Schiff base of cephalopod rhodopsin and metarhodopsin.

In the visual pigment, the chromophore retinal is bound through a Schiff base linkage with the protein moiety, opsin. The secondary interaction of retinal with opsin was studied in cephalopod rhodopsin and its photoproduct, metarhodopsin. The pK of the protonation of the Schiff base in metarhodopsin was affected by the phospholipids and detergents surrounding the protein, and varied between 9.2 and 6.9. Among nonionic detergents, the fatty acid ester of sucrose behaved like phospholipids but other detergents changed the protein conformation so that the pK of the Schiff base in metarhodopsin became nearly equal to the pK of N-retinylidene butylamine. This tendency was manifested in rhodopsin as the formation of a 380 nm pigment.

Formation of hypsorhodopsin at room temperature by picosecond green pulse.

Excitation of squid rhodopsin with a single laser pulse (532 nm, 25 ps) at 18 degrees C yielded photorhodopsin, a precursor of bathorhodopsin. In the linear region, no relation between amount of photorhodopsin and excitation-energy hypsorhodopsin was detected, while in a photon saturation region this was observed. The time constant of hypsorhodopsin to bathorhodopsin decay was about 125 ps. Dependencies of formation of photorhodopsin and hypsorhodopsin on the excitation energy suggest that hypsorhodopsins of squid and octopus are formed by a two-photon reaction. No cattle hypsorhodopsin was detected in our experimental conditions.

Formation of hyposorhodopsin in frog retina.

Hypsorhodopsin was formed in frog retina by irradiation at liquid helium temperature and converted into bathorhodopsin above about 29K.

Two forms of intermediates of frog rhodopsin in rod outer segments.

Using frog rod outer segments, we measured changes of the absorption spectrum during the conversion of rhodopsin to a photosteady-state mixture composed of rhodopsin, isorhodopsin and bathorhodopsin by irradiation with blue light (440 nm) at -190 degrees C and during the reversion of bathorhodopsin to a mixture of rhodopsin and isorhodopsin by irradiation with red light (718 nm) at -190 degrees C. The reaction kinetics was expressed by one exponential in the former case and by two exponentials in the latter. These results suggest that rhodopsin is composed of a single molecular species, while bathorhodopsin is composed of two kinds of molecular species designated as batho1-rhodopsin and batho2-rhodopsin. On warming the two forms of bathorhodopsin, each bathorhodopsin converted to its own lumirhodopsin, metarhodopsin I and finally a free all-trans-retinal plus opsin. The absorption spectra of the two forms of bathorhodopsin, lumirhodopsin and metarhodopsin I were measured at -190 degrees C. We infer that a rhodopsin molecule in the excited state relaxes to either batho1-rhodopsin or batho2-rhodopsin, and then converts to its own intermediates through one of the two parallel pathways.

Specific photoisomerization of retinal in squid rhodopsin and metarhodopsin.

The composition of retinal isomers in the photosteady-state mixtures formed from squid rhodopsin and metarhodopsin was determined by high-pressure liquid chromatography. A large amount of 9-cis-retinal was obtained at liquid N2 temperature when rhodopsin was irradiated with orange light, but only small quantities of 9-cis-retinal were obtained at 15 degrees C. Scarcely any 9-cis-retinal was produced from metarhodopsin by irradiation at liquid N2 temperature. A large quantity of 7-cis-retinal was found in the photoproduct of rhodopsin irradiated at solid carbon dioxide temperature, but not at 15 degrees C and liquid N2 temperature. 7-cis-Retinal was not produced from metarhodopsin at any temperatures. These results indicate that the photoisomerization of retinal is regulated by the structure of the retinal-binding site of this protein. The formation of 9-cis- and 7-cis-retinals is forbidden in the metarhodopsin protein.

Effect of cholesterol on rhodopsin stability in disk membranes.

The effect of cholesterol on rhodopsin stability has been investigated in intact disk membranes. Because cholesterol readily equilibrates between membranes, the disk membrane cholesterol content can be altered by incubation with cholesterol/phospholipid vesicles. The effect of membrane cholesterol on rhodopsin was investigated using three independent techniques: thermal bleaching, differential scanning calorimetry (DSC) and activation of the cGMP cascade. Rhodopsin exhibited an increased resistance to thermally induced bleaching as the membrane cholesterol level was increased. DSC also indicated that the protein is stabilized by cholesterol in that the Tm increased in response to higher membrane cholesterol. A similar degree of stabilization was observed in both the unbleached and bleached states in the DSC experiments. These results suggest that cholesterol affects the disk membrane properties such that thermally induced unfolding is inhibited, thus stabilizing the rhodopsin structure. Furthermore, high membrane cholesterol inhibited the activation of the cGMP cascade. This is consistent with the stabilization of the metarhodopsin I photointermediate relative to the metarhodopsin II intermediate.

Effect of GTP on the rhodopsin-G-protein complex by transient formation of extra metarhodopsin II.

The light-induced transient interaction between rhodopsin and G-protein in the presence of GTP has been measured by the formation of extra metarhodopsin II. Disc membranes were recombined with the hypotonic extract containing the G-protein. Without GTP, a flash induces stable rhodopsin-G-protein complexes which dissociate upon addition of GTP. In low GTP (less than 10 microM) transient rhodopsin X G-protein interaction is observed. Rhodopsin X G-protein dissociates the faster, the more GTP is present (rate of dissociation, 0.3/s at 5 microM GTP; T = 3.5 degrees C). The results corroborate that the uptake of GTP terminates the rhodopsin-G-protein complex and allow an estimation of the rhodopsin X G-protein lifetime.

Biochemical aspects of the visual process. XL. Spectral and chemical analysis of metarhodopsin III in photoreceptor membrane suspensions.

The late photointermediates of rhodopsin photolysis have been analyzed spectrally and chemically in bovine rod outer segment membrane suspension at 25 degrees C and pH 6.5. The decay of metarhodopsin II follows two spectrally distinct routes, resulting 40 min after illumination in a stable mixture of photo-products with absorbance maxima around 380 and 452 nm, free retinal and metarhodopsin III, respectively. Chemical analysis shows that three different products are involved: free retinal (approx. 34%), protein-bound retinal (approx. 51%) and lipid-bound retinal (approx. 15%). The latter fraction consists of retinylidene-phosphatidylethanolamine exclusively. Photolysis of membranes reconstituted with various phospholipids gives a qualitatively normal spectral picture, but the production of metarhodopsin III may vary with the phospholipid composition, i.e. with the percent of phosphatidylethanolamine present. Chemical analysis shows that with increasing phosphaatidylethanolamine content of the membrane, the retinylidene phosphatidylethanolamine fraction increases proportionally at the expense of free retinal, while the fraction of protein-bound retinal remains unaffected. The results indicate that under these conditions metarhodopsin III (defined as a long wavelength product of metarhodopsin II decay) is composed of two chemically distinct components: opsin-bound retinal and retinylidene phosphatidylethanolamine.

Activation of phosphodiesterase in frog rod outer segment by rhodopsin analogues.

Activation of guanosine 3',5'-cyclic monophosphate (cGMP) phosphodiesterase (EC 3.1.4.35) in frog rod outer segment membrane by rhodopsin analogues has been investigated. A rhodopsin analogue modified at the Schiff-base linkage (N-retinyl-opsin) or the beta-ionone ring (3-dehydro-rhodopsin) in the retinylidene chromophore of rhodopsin has some ability in activation of the enzyme. In consideration of our previous observation that opsin including a retinal-oxime can activate the enzyme, it seems likely that the Schiff-base linkage is not always necessary for the phosphodiesterase activation. On the other hand, a change in the length of the side chain of retinal (complex of opsin and beta-ionone, beta-ionylideneacetaldehyde or retinylideneacetaldehyde) or dissection of the conjugate double-bond system of the side chain (retro-gamma-rhodopsin) remarkably reduces the activation ability. However, 5,8-epoxy-rhodopsin having a similar dissected conjugate double-bond system induces some enzyme activation because of its rigid conformation around C7-C8-C9 single bonds. Consequently, it is suggested that the necessary portion of rhodopsin chromophore for the activation of the enzyme is the rigid conjugate double-bond system between the beta-ionone ring and the Schiff-base linkage in its all-trans form.