Determination of membrane protein structure by rotational resonance NMR: bacteriorhodopsin.

Rotationally resonant magnetization exchange, a new nuclear magnetic resonance (NMR) technique for measuring internuclear distances between like spins in solids, was used to determine the distance between the C-8 and C-18 carbons of retinal in two model compounds and in the membrane protein bacteriorhodopsin. Magnetization transfer between inequivalent spins with an isotropic shift separation, delta, is driven by magic angle spinning at a speed omega r that matches the rotational resonance condition delta = n omega r, where n is a small integer. The distances measured in this way for both the 6-s-cis- and 6-s-trans-retinoic acid model compounds agreed well with crystallographically known distances. In bacteriorhodopsin the exchange trajectory between C-8 and C-18 was in good agreement with the internuclear distance for a 6-s-trans configuration [4.2 angstroms (A)] and inconsistent with that for a 6-s-cis configuration (3.1 A). The results illustrate that rotational resonance can be used for structural studies in membrane proteins and in other situations where diffraction and solution NMR techniques yield limited information.

Evidence for a vibrational phase-dependent isotope effect on the photochemistry of vision.

Vibronic coupling is key to efficient energy flow in molecular systems and a critical component of most mechanisms invoking quantum effects in biological processes. Despite increasing evidence for coherent coupling of electronic states being mediated by vibrational motion, it is not clear how and to what degree properties associated with vibrational coherence such as phase and coupling of atomic motion can impact the efficiency of light-induced processes under natural, incoherent illumination. Here, we show that deuteration of the H11-C11=C12-H12 double-bond of the 11-cis retinal chromophore in the visual pigment rhodopsin significantly and unexpectedly alters the photoisomerization yield while inducing smaller changes in the ultrafast isomerization dynamics assignable to known isotope effects. Combination of these results with non-adiabatic molecular dynamics simulations reveals a vibrational phase-dependent isotope effect that we suggest is an intrinsic attribute of vibronically coherent photochemical processes.

Low-lying electronic states of carotenoids.

Four all-trans carotenoids, spheroidene, 3,4-dihydrospheroidene, 3,4,5,6-tetrahydrospheroidene, and 3,4,7,8-tetrahydrospheroidene, have been purified using HPLC techniques and analyzed using absorption, fluorescence and fluorescence excitation spectroscopy of room temperature solutions. This series of molecules, for which the extent of pi-electron conjugation decreases from 10 to seven carbon-carbon double bonds, exhibits a systematic crossover from S2----S0 (1(1)Bu----1(1)Ag) to S1----S0 (2(1)Ag----1(1)Ag) emission with decreasing chain length. Extrapolation of the S1----S0 transition energies indicates that the 2(1)Ag states of longer carotenoids have considerably lower energies than previously thought. The energies of the S1 states of spheroidenes and other long carotenoids are correlated with the S1 energies of their chlorophyll partners in antenna complexes of photosynthetic systems. Implications for energy transfer in photosynthetic antenna are discussed.

Towards a vibrational analysis of spheroidene. Resonance Raman spectroscopy of 13C-labelled spheroidenes in petroleum ether and in the Rhodobacter sphaeroides reaction centre.

We report resonance Raman spectra of the carotenoid spheroidene and its 14'-13C and 15'-13C substituted analogues in petroleum ether and bound to the reaction centre of Rhodobacter sphaeroides R26. The spectra in petroleum ether correspond to planar all-trans spheroidene while those of the reaction centres are consistent with a nonplanar 15,15'-cis spheroidene. The effect of 13C labelling is largest in the carbon-carbon double-bond stretching region. The 15'-13C substitution of the reaction centre bound spheroidene, however, hardly changes the C=C band as compared to that for the natural abundance spheroidene apart from a new weak band at 1508 cm(-1). This observation has been interpreted as a decoupling of the C15=C15' stretch from the other double-bond stretches in combination with a small intrinsic Raman intensity of this local mode for 15,15'-cis spheroidene.

Detection of threonine structural changes upon formation of the M-intermediate of bacteriorhodopsin: evidence for assignment to Thr-89.

The behavior of threonine residues in the bacteriorhodopsin (bR) photocycle has been investigated by Fourier transform infrared difference spectroscopy. L-Threonine labeled at the hydroxyl group with 18O (L-[3-(18)O]threonine) was incorporated into bR and the bR-->M FTIR difference spectra measured. Bands are assigned to threonine vibrational modes on the basis of 18O induced isotope frequency shifts and normal mode calculations. In the 3500 cm-1 region, a negative band is assigned to the OH stretch of threonine. In the 1125 cm-1 region, a negative band is assigned to a mixed CH3 rock/CO stretch mode. The frequency of both these bands indicates the presence of at least one hydrogen bonded threonine hydroxyl group in light adapted bR which undergoes a change in structure by formation of the M intermediate. Spectral changes induced by the substitution Thr-89-->Asn but not Thr-46-->Asn or Asp-96-->Asn are consistent with the assignment of these bands to Thr-89. These results along with another related study on the mutant Thr-89-->Asn indicate that the active site of bR includes Thr-89 and that its interaction with the retinylidene Schiff base and Asp-85 may play an important role in regulating the color of bacteriorhodopsin and the transfer of a proton to the Schiff base.

Protein-chromophore interactions in alpha-crustacyanin, the major blue carotenoprotein from the carapace of the lobster, Homarus gammarus. A study by 13C magic angle spinning NMR.

MAS (magic angle spinning) 13C NMR has been used to study protein-chromophore interactions in alpha-crustacyanin, the blue astaxanthin-binding carotenoprotein of the lobster, Homarus gammarus, reconstituted with astaxanthins labelled with 13C at the 14,14' or 15,15' positions. Two signals are seen for alpha-crustacyanin containing [14,14'-13C2]astaxanthin, shifted 6.9 and 4.0 ppm downfield from the 134.1 ppm signal of uncomplexed astaxanthin in the solid state. With alpha-crustacyanin containing [15,15'-13C2]astaxanthin, one essentially unshifted broad signal is seen. Hence binding to the protein causes a decrease in electronic charge density, providing the first experimental evidence that a charge redistribution mechanism contributes to the bathochromic shift of the astaxanthin in alpha-crustacyanin, in agreement with inferences based on resonance Raman data [Salares, et al. (1979) Biochim. Biophys. Acta 576, 176-191]. The splitting of the 14 and 14' signals provides evidence for asymmetric binding of each astaxanthin molecule by the protein.

Photoreceptor rhodopsin: structural and conformational study of its chromophore 11-cis retinal in oriented membranes by deuterium solid state NMR.

Rhodopsin is the retinal photoreceptor responsible for visual signal transduction. To determine the orientation and conformation of retinal within the binding pocket of this membrane bound receptor, an ab initio solid state 2H NMR approach was used. Bovine rhodopsin containing 11-cis retinal, specifically deuterated at its methyl groups at the C19 or C20 position, was uniaxially oriented in DMPC bilayers. Integrity of the membranes and quality of alignment were monitored by 31P NMR. Analysis of the obtained 2H NMR spectra provided angles for the individual labelled chemical bond vectors leading to an overall picture for the three dimensional structure of the polyene chain of the chromophore in the protein binding pocket around the Schiff base attachment site.

Asymmetric binding of the primary acceptor quinone in reaction centers of the photosynthetic bacterium Rhodobacter sphaeroides R26, probed with Q-band (35 GHz) EPR spectroscopy.

The reaction center (RC)-bound primary acceptor quinone QA of the photosynthetic bacterium Rhodobacter sphaeroides R26 functions as a one-electron gate. The radical anion QA.- is proposed to have an asymmetric electron distribution, induced by the protein environment. We replace the native ubiquinone-10 (UQ10) with specifically 13C-labelled UQ10, and use Q-band (35 GHz) EPR spectroscopy to investigate this phenomenon in closer detail. The direct observation of the 13C-hyperfine splitting of the gz-component of UQ10A.- in the RC and in frozen isopropanol shows that the electron spin distribution is symmetric in the isopropanol glass, and asymmetric in the RC. Our results allow qualitative assessment of the spin and charge distribution for QA.- in the RC. The carbonyl oxygen of the semiquinone anion nearest to the S = 2 Fe(2+)-ion and QB is shown to acquire the highest (negative) charge density.

FTIR spectroscopy shows weak symmetric hydrogen bonding of the QB carbonyl groups in Rhodobacter sphaeroides R26 reaction centres.

The absorption frequencies of the C = O and C = C (neutral state) and of the C...O (semiquinone state) stretching vibrations of QB have been assigned by FTIR spectroscopy, using native and site-specifically 1-, 2-, 3- and 4-13C-labelled ubiquinone-10 (UQ10) reconstituted at the QB binding site of Rhodobacter sphaeroides R26 reaction centres. Besides the main C = O band at 1641 cm-1, two smaller bands are observed at 1664 and 1651 cm-1. The smaller bands at 1664 and 1651 cm-1 agree in frequencies with the 1- and 4-C = O vibrations of unbound UQ10, showing that a minor fraction is loosely and symmetrically bound to the protein. The larger band at 1641 cm-1 indicates symmetric H-bonding of the 1- and 4-C = O groups for the larger fraction of UQ10 but much weaker interaction as for the 4-C = O group of QA. The FTIR experiments show that different C = O protein interactions contribute to the factors determining the different functions of UQ10 at the QA and the QB binding sites.

Bioefficacy of beta-carotene dissolved in oil studied in children in Indonesia.

BACKGROUND: More information on the bioefficacy of carotenoids in foods ingested by humans is needed. OBJECTIVE: We aimed to measure the time required for isotopic enrichment of beta-carotene and retinol in serum to reach a plateau, the extent of conversion of beta-carotene dissolved in oil with use of beta-carotene and retinol specifically labeled with 10 (13)C atoms, and the intraindividual variation in response. DESIGN: Indonesian children aged 8--11 y (n = 35) consumed 2 capsules/d, 7 d/wk, for < or =10 wk. Each capsule contained 80 microg [12,13,14,15,20,12',13',14',15',20'-(13)C(10)]beta-carotene and 80 microg [8,9,10,11,12,13,14,15,19,20-(13)C(10)]retinyl palmitate. Three blood samples were drawn per child over a period of < or =10 wk. HPLC coupled with atmospheric pressure chemical ionization liquid chromatography-mass spectrometry was used to measure the isotopic enrichment in serum of retinol with [(13)C(5)]retinol and [(13)C(10)]retinol and of beta-carotene with [(13)C(10)]beta-carotene. The beta-carotene in the capsules used had a cis-trans ratio of 3:1. RESULTS: Plateau isotopic enrichment was reached by day 21. The amount of beta-carotene in oil required to form 1 microg retinol was 2.4 microg (95% CI: 2.1, 2.7). The amount of all-trans-beta-carotene required to form 1 microg retinol may be lower. CONCLUSIONS: The efficiency of conversion of this beta-carotene in oil was 27% better than that estimated previously (1.0 microg retinol from 3.3 microg beta-carotene with an unknown cis-trans ratio). The method described can be extended to measure the bioefficacy of carotenoids in foods with high precision, requiring fewer subjects than other methods.

Bacterial and human cell mutagenicity study of some C18H10 cyclopenta-fused polycyclic aromatic hydrocarbons associated with fossil fuels combustion.

A number of isomeric C18H10 polycyclic aromatic hydrocarbons (PAHs), thought to be primarily cyclopenta-fused PAHs, are produced during the combustion and pyrolysis of fossil fuels. To determine the importance of their contributions to the total mutagenic activity of combustion and pyrolysis samples in which they are found, we characterized reference quantities of four C18H10 CP-PAHs: benzo[ghi]fluoranthene (BF), cyclopenta[cd]pyrene (CPP), cyclopent[hi]acephenanthrylene (CPAP), and cyclopent[hi]aceanthrylene (CPAA). Synthesis of CPAA and CPAP is described. The availability of reference samples of these isomers also proved to be an essential aid in the identification of the C18H10 species often found in combustion and pyrolysis samples. Chemical analysis of selected combustion and pyrolysis samples showed that CPP was generally the most abundant C18H10 isomer, followed by CPAP and BF. CPAA was detected only in pyrolysis products from pure PAHs. We tested the four C18H10 PAHs for mutagenicity in a forward mutation assay using S. typhimurium. CPP, BF, and CPAA were roughly twice as mutagenic as benzo[a]pyrene (BaP), whereas CPAP was only slightly active. These PAHs were also tested for mutagenic activity in human cells. In this assay, CPP and CPAA were strongly mutagenic but less active than BaP, whereas CPAP and BF were inactive at the dose levels tested. Also, the bacterial and human cell mutagenicity of CPAA and CPAP were compared with the mutagenicity of their monocyclopenta-fused analogs, aceanthrylene and acephenanthyrlene. Although the mutagenicities of CPAP and acephenanthrylene are similar, the mutagenic activity of CPAA is an order of magnitude greater than that of aceanthyrlene.

Photoactivation of rhodopsin: interplay between protein and chromophore.

Data in the literature suggest a finely tuned interaction between ligand (11-cis-retinal) and protein (opsin) in order to allow very efficient photoactivation of the ligand and highly vectorial rhodopsin activation with a huge increase in receptor activity. We have further investigated this interaction using ligand homologues, 13C-ligand labelling or 15N-protein labelling, in combination with Fourier transform infrared (FT-IR) and solid-state magic angle spinning (ss-MAS)-NMR spectroscopy. Using 1D rotational resonance (RR) or double-quantum heteronuclear local field (2Q-HLF) ss-MAS-NMR we report the first structure refinement of the rhodopsin chromophore in situ. These measurements yield a specification of the torsional strain in the for isomerization essential C10-C13 segment of the chromophore. This strain is thought to contribute to the high rate and stereospecificity of the photoisomerization reaction. In agreement with previous data, the C10-C13 segment region reaches a relaxed all-trans configuration at the lumirhodopsin photointermediate. MAS-NMR analysis of [15N]lysine-labelled rhodopsin reveals the presence of a 'soft' counterion, requiring intermediate water molecules for stabilization. FT-IR studies on [2H]tyrosine-labelled rhodopsin demonstrate participation of several tyrosin(at)e residues in receptor activation. One of these, probably Tyr268, is already active at the bathorhodopsin stage. Finally, the effect of ligands with single additional methyl substituents in the C10-C12 region has been investigated. They do not affect the general activation pathway, but perturb the activation kinetics of rhodopsin, suggesting steric interference with protein residues. Possible implications of these results for a structural role of water residues will be discussed, as well.

Anomalous rotational resonance spectra in magic-angle spinning NMR.

Magic-angle spinning NMR spectra of samples containing dilute spin-1/2 pairs display broadenings or splittings when a rotational resonance condition is satisfied, meaning that a small integer multiple of the spinning frequency matches the difference in the two isotropic shift frequencies. We show experimental rotational resonance NMR spectra of a 13C2-labeled retinal which are in qualitative disagreement with existing theory. We propose an explanation of these anomalous rotational spectra involving residual heteronuclear couplings between the 13C nuclei and the neighboring 1H nuclei. These couplings strongly influence the rotational resonance 13C spectrum, despite the presence of a strong radiofrequency decoupling field at the 1H Larmor frequency. We model the residual heteronuclear couplings by differential transverse relaxation of the 13C single-quantum coherences. We present a superoperator theory of the phenomenon and describe a numerical algorithm for rapid Liouville space simulations in periodic systems. Good agreement with experimental results is obtained by using a biexponential transverse relaxation model for each spin site.

Changes in structure of the chromophore in the photochemical process of bovine rhodopsin as revealed by FTIR spectroscopy for hydrogen out-of-plane vibrations.

The hydrogen out-of-plane bending (HOOP) vibrations were studied in the difference Fourier transform infrared spectra of lumirhodopsin and metarhodopsin I by use of a series of specifically deuterated retinal derivatives of bovine rod outer segments. The 947 cm-1 band of lumirhodopsin and the 950 cm-1 band of metarhodopsin I were assigned to the mode composed of both 11-HOOP and 12-HOOP vibrations. This result suggests that the perturbation near C12-H of the retinal in the earlier intermediate, bathorhodopsin (Palings, van den Berg, Lugtenburg and Mathies, Biochemistry, 28 (1989) 1498-1507), is extinguished in lumirhodopsin and metarhodopsin I. Unphotolyzed rhodopsin and metarhodopsin I exhibited the 14-HOOP bands in the 12-D derivatives at 901 and 886 cm-1, respectively. Lumirhodopsin, however, did not show the 14-HOOP in the 12-D derivatives. The result suggests a change in geometrical alignment of the C14-H bond in lumirhodopsin with respect to the N-H bond of the Schiff base.

Magic angle spinning NMR studies on the metarhodopsin II intermediate of bovine rhodopsin: evidence for an unprotonated Schiff base.

Magic angle spinning (MAS)13C-NMR spectra of the metarhodopsin II intermediate have been obtained using bovine rhodopsin regenerated with retinal 13C-labeled at the C-13 and C-15 positions to investigate the protonation state of the retinal Schiff base linkage. The 13C-labeled rhodopsin was reconstituted into 1,2-dipalmitoleoylphosphatidylcholine bilayers to increase the amount of meta II trapped at low temperature. Both the 13C-15 (159.2 ppm) and 13C-13 (144.0 ppm) isotropic chemical shifts are characteristic of an unprotonated Schiff base, while the 13C-15 shift is significantly different from that of retinal (191 ppm) or a tetrahedral carbinolamine group (70-90 ppm) previously proposed as an intermediate in the hydrolysis of the Schiff base at the meta II stage. This rules out the possibility that meta II non-covalently binds retinal or is a carbinolamine intermediate and provides convincing evidence that Schiff base deprotonation occurs in the meta I-meta II transition, an event that is likely to be important in triggering the activation of transducin.

Triplet energy transfer between the primary donor and carotenoids in Rhodobacter sphaeroides R-26.1 reaction centers incorporated with spheroidene analogs having different extents of pi-electron conjugation.

Three carotenoids, spheroidene, 3,4-dihydrospheroidene and 3,4,5,6-tetrahydrospheroidene, having 8, 9 and 10 conjugated carbon-carbon double bonds, respectively, were incorporated into Rhodobacter (Rb.) sphaeroides R-26.1 reaction centers. The extents of binding were found to be 95 +/- 5% for spheroidene, 65 +/- 5% for 3,4-dihydrospheroidene and 60 +/- 10% for 3,4,5,6-tetrahydrospheroidene. The dynamics of the triplet states of the primary donor and carotenoid were measured at room temperature by flash absorption spectroscopy. The carotenoid, spheroidene, was observed to quench the primary donor triplet state. The triplet state of spheroidene that was formed subsequently decayed to the ground state with a lifetime of 7.0 +/- 0.5 microseconds. The primary donor triplet lifetime in the Rb. sphaeroides R-26.1 reaction centers lacking carotenoids was 60 +/- 5 microseconds. Quenching of the primary donor triplet state by the carotenoid was not observed in the Rb. sphaeroides R-26.1 reaction centers containing 3,4-dihydrospheroidene nor in the R-26.1 reaction centers containing 3,4,5,6-tetrahydrospheroidene. Triplet-state electron paramagnetic resonance was also carried out on the samples. The experiments revealed carotenoid triple-state signals in the Rb. sphaeroides R-26.1 reaction centers incorporated with spheroidene, indicating that the primary donor triplet is quenched by the carotenoid. No carotenoid signals were observed from Rb. sphaeroides R-26.1 reaction centers incorporating 3,4-dihydrospheroidene nor in reaction centers incorporating 3,4,5,6-tetrahydrospheroidene. Circular dichroism, steady-state absorbance band shifts accompanying the primary photochemistry in the reaction center and singlet energy transfer from the carotenoid to the primary donor confirm that the carotenoids are bound in the reaction centers and interacting with the primary donor. These studies provide a systematic approach to exploring the effects of carotenoid structure and excited-state energy on triplet transfer between the primary donor and carotenoids in reaction centers from photosynthetic bacteria.

Carotenoid-to-bacteriochlorophyll singlet energy transfer in carotenoid-incorporated B850 light-harvesting complexes of Rhodobacter sphaeroides R-26.1.

Four carotenoids, 3,4,7,8-tetrahydrospheroidene, 3,4,5,6-tetrahydrospheroidene, 3,4-dihydrospheroidene and spheroidene, have been incorporated into the B850 light-harvesting complex of the carotenoidless mutant, photosynthetic bacterium, Rhodobacter sphaeroides R-26.1. The extent of pi-electron conjugation in these molecules increases from 7 to 10 carbon-carbon double bonds. Carotenoid-to-bacteriochlorophyll singlet state energy transfer efficiencies were measured using steady-state fluorescence excitation spectroscopy to be 54 +/- 2%, 66 +/- 4%, 71 +/- 6% and 56 +/- 3% for the carotenoid series. These results are discussed with respect to the position of the energy levels and the magnitude of spectral overlap between the S1 (2(1)Ag) state emission from the isolated carotenoids and the bacteriochlorophyll absorption of the native complex. These studies provide a systematic approach to exploring the effect of excited state energies, spectral overlap and excited state lifetimes on the efficiencies of carotenoid-to-bacteriochlorophyll singlet energy transfer in photosynthetic systems.

A resonance Raman study of octopus bathorhodopsin with deuterium labeled retinal chromophores.

The resonance Raman spectrum of octopus bathorhodopsin in the fingerprint region and in the ethylenic-Schiff base region have been obtained at 80 K using the "pump-probe" technique as have its deuterated chromophore analogues at the C7D; C8D; C8,C7D2; C10D; C11D; C11, C12D2; C14D; C15D; C14, C15D2; and N16D positions. While these data are not sufficient to make definitive band assignments, many tentative assignments can be made. Because of the close spectral similarity between the octopus bathorhodopsin spectrum and that of bovine bathorhodopsin, we conclude that the essential configuration of octopus bathorhodopsin's chromophore is all-trans like. The data suggest that the Schiff base, C = N, configuration is trans (anti). The observed conformationally sensitive fingerprint bands show pronounced isotope shifts upon chromophore deuteration. The size of the shifts differ, in certain cases, from those found for bovine bathorhodopsin. Thus, the internal mode composition of the fingerprint bands differs somewhat from bovine bathorhodopsin, suggesting a somewhat different in situ chromophore conformation. An analysis of the NH bend frequency, the Schiff base C = N stretch frequency, and its shift upon Schiff base deuteration suggests that the hydrogen bonding between the protonated Schiff base with its protein binding pocket is weaker in octopus bathorhodopsin than in bovine bathorhodopsin but stronger than that found in bacteriorhodopsin's bR568 pigment.

A resonance Raman study of the C=C stretch modes in bovine and octopus visual pigments with isotopically labeled retinal chromophores.

Previous resonance Raman spectroscopic studies of bovine and octopus rhodopsin and bathorhodopsin in the C-C stretch fingerprint region have shown drastically different spectral patterns, which suggest different chromophore-protein interactions. We have extended our resonance Raman studies of bovine and octopus pigments to the C=C stretch region in order to reveal a more detailed picture about the difference in retinal-protein interactions between these two pigments. The C=C stretch motions of the protonated retinal Schiff base are strongly coupled to form highly delocalized ethylenic modes located in the 1500 to 1650 cm-1 spectral region. In order to decouple these vibrations, a series of 11,12-D2-labeled retinals, with additional 13C labeling at C8, C10, C11 and C14, respectively, are used to determine the difference of specific C=C stretch modes between bovine and octopus pigments. Our results show that the C9=C10 and C13=C14 stretch mode are about 20 cm-1 lower in the Raman spectrum of octopus bathorhodopsin than in bovine bathorhodopsin, while the other C=C stretch modes in these two bathorhodopsins are similar. In contrast, only the C9=C10 stretch mode in octopus rhodopsin is about 10 cm-1 lower than in bovine rhodopsin, while other C=C stretches are similar.

Synthesis of isotopically labelled retinal. Structural and functional studies at the atomic level of the chromophore in visual pigments.

OBJECTIVE: To describe the strategy followed in studying the structure and function of the chromophore in visual pigments at the atomic level. METHODS: A three-step strategy was followed. First, 11-cis retinal with high enrichment (99% 13C or 99% 2H) at pre-determined positions or combinations of positions was synthesized. Second, the rhodopsins with isotope label in the chromophore were prepared combining opsin with the isotopically labelled 11-cis retinal. Third, these systems were studied with isotope-sensitive physical techniques. CONCLUSIONS: Isotope labelling together with isotope-sensitive spectroscopy is a highly promising method for obtaining information at the atomic level of biologically active systems. Strategies based on specific stable isotope labelling are likely to be applicable to many research areas.