Conserved amino acids in F-helix of bacteriorhodopsin form part of a retinal binding pocket.

A 3-dimensional model for the retinal binding pocket in the light-driven proton pump, bacteriorhodopsin, is proposed on the basis of spectroscopic studies of bacteriorhodopsin mutants. In this model Trp-182, Pro-186 and Trp-189 surround the polyene chain while Tyr-185 is positioned close to the retinylidene Schiff base. This model is supported by sequence homologies in the F-helices of bacteriorhodopsin and the related retinal proteins, halorhodopsin and rhodopsins.

Arginine to glutamine substitutions in the fourth module of Xenopus interphotoreceptor retinoid-binding protein.

PURPOSE: Interphotoreceptor retinoid-binding protein (IRBP) is unusual for a lipid-binding protein in that its gene is expressed uniquely by cells of photoreceptor origin and consists of four homologous repeats, each coding for a module of approximately 300 amino acid residues. All-trans retinol binding domains, which appear to be present in each module, are composed of conserved hydrophobic regions [Baer et al, Exp Eye Res 1998; 66:249-262]. Here we investigate the role of highly conserved arginines contained in these regions. METHODS: To study the arginines in an individual module without the interference of ligand-binding activity elsewhere in the protein, we expressed in E. coli the fourth module of Xenopus IRBP by itself as a soluble thioredoxin fusion protein (X4IRBP). Arginines 1005, 1041, 1073 and 1122 were separately replaced by glutamine using PCR overlap extension mutagenesis. The glutamine substitutions were confirmed by liquid chromatography-tandem mass spectrometry. The binding of all-trans retinol and 9-(9-anthroyloxy)stearic acid (9-AS) to X4IRBP and each of the mutants was evaluated by fluorescence spectroscopy. Binding was followed by monitoring the enhancement of ligand fluorescence and the quenching of protein endogenous fluorescence. The ability of the recombinant proteins to protect all-trans retinol from oxidative degradation was evaluated by monitoring absorbance at 325 nm over time. RESULTS: The substitution of Gln for Arg1005 about doubled the amount of ligand necessary to attain saturation and about doubled the level of fluorescence enhancement obtained at saturation for both all-trans retinol and 9-AS. Although there was not a significant change in the Kd, the substitution increased the calculated number of binding sites (N) from approximately 2 to approximately 4 per polypeptide. The other Arg->Gln mutants did not significantly change the Kd or N. None of the mutations compromised the ability of the module to protect all-trans retinol from degradation. CONCLUSIONS: Our data suggest that the function of the conserved arginines in IRBP is fundamentally different from that of other retinoid-binding proteins. These residues do not appear to play a role in defining the specificity of the ligand-binding domain. Rather, Arg1005 appears to play a role in defining the capacity of the domain. Our data suggest that the binding site consists of a single hydrophobic cavity promiscuous for fatty acids and all-trans retinol.

Differences between the photocycles of halorhodopsin and the acid purple form of bacteriorhodopsin analyzed with millisecond time-resolved FTIR spectroscopy.

At pH 1, bacteriorhodopsin (bR) is thought to function as a halide ion pump, in contrast to its biological function as a proton pump at neutral pH. Despite the apparent similarity in function between this 'acid purple' form of bR and the native form of halorhodopsin (hR), their FTIR difference spectra measured ca. 5 ms after photolysis are significantly different. The most striking difference is the appearance of a positive band at 1753 cm-1 and a negative band at 1732 cm-1 in the bRacid purple difference spectrum. These and other spectral features are similar, but not identical, to those of the bR-->O difference spectrum measured at neutral pH. The structure of the bRacid purple longest-lived product therefore corresponds more closely to the O photoproduct of the bR proton-pumping photocycle, rather than the hL photoproduct seen on a similar time scale in the hR photocycle. The 1753- and 1732-cm-1 bands are largely unaffected by the D212N mutation, but both appear to lose a portion of their intensities with either the D85N or D96N mutation. Thus Asp-85 and -96 likely undergo substantial changes in hydrogen-bonding environment during the halide-pumping cycle of bRacid purple. Our FTIR results deepen the distinctions between the hR and bR photocycles. The mechanism of chloride pumping in hR has been thought not to involve protonation or hydrogen bonding changes of carboxylic acid groups. In bRacid purple, however, it seems likely that at least one carboxylic acid might play an important role in the mechanism of chloride pumping, leading to an increase in thermodynamic or kinetic stabilization of the O intermediate.

Nano- and microsecond time-resolved FTIR spectroscopy of the halorhodopsin photocycle.

Step-scan Fourier transform infrared spectroscopy with 50 ns time resolution was applied to the early stages of the photocycle of halorhodopsin (hR) for the temperature range 3-42 degrees C. Kinetic data analysis with global fitting revealed two distinct kinetic processes associated with relaxations of the early red-shifted photoproduct hK; these processes have time constants tau 1 approximately equal to 280 ns and tau 2 approximately equal to 360 microns at 20 degrees C. Spectral features demonstrate that the tau 1 process corresponds to a transition between two distinct bathointermediates, hKE and hKL. The vibrational difference bands associated with both tau 1 and tau 2 transitions are spread throughout the whole 1800-900 cm-1 range. However, the largest bands correspond to ethylenic C=C stretches, fingerprint C-C stretches and hydrogen out-of-plane (HOOP) wags of the retinal chromophore. The time evolution of these difference bands indicate that both the tau 1 and tau 2 decay processes involve principally a relaxation of the chromophore and its immediate environment. The decay of the intense HOOP vibrations is nearly equally divided between the tau 1 and tau 2 processes, indicating a complex chromophore relaxation from a twisted nonrelaxed conformation in the primary (hKE) bathointermediate, to a less-twisted structure in hKL, and finally to a roughly planar structure in the hypsochromically shifted hL intermediate. This conclusion is also supported by the unexpectedly large positive entropy of activation observed for the tau 1 process. The two relaxations from hKE to hL are largely analogous to corresponding relaxations (KE-->KL-->L) in the bacteriorhodopsin photocycle, except that the second step is slowed down by over 200-fold in hR.

Expression and characterization of the fourth repeat of Xenopus interphotoreceptor retinoid-binding protein in E. coli.

Interphotoreceptor retinoid-binding protein (IRBP) is an extracellular glycolipoprotein which in higher vertebrates has a 4-repeat structure and carries endogenous vitamin A and fatty acids. The location of IRBP's 1-2 binding sites for retinol is unknown. To begin to understand which repeat(s) are responsible for ligand-binding, we expressed the fourth repeat of Xenopus IRBP in E. coli to determine if it could by itself bind all-trans retinol. Our expression studies used a polyhistidine fusion domain to purify the recombinant protein directly from inclusion bodies. The fusion protein could be renatured without aggregation if refolded at a sufficiently dilute concentration (< 3 microM). The recombinant fourth repeat of Xenopus IRBP binds [3H]all-trans retinol and the fluorescence of this ligand increases 8-fold upon binding. The binding is saturable with a Kd = 0.4 microM. The expression of recombinant IRBP fragments as fusion proteins in prokaryotes will be useful for defining the structural requirements for ligand binding by this interesting protein.

Fourth module of Xenopus interphotoreceptor retinoid-binding protein: activity in retinoid transfer between the retinal pigment epithelium and rod photoreceptors.

PURPOSE: Interphotoreceptor retinoid-binding protein (IRBP), an extracellular protein believed to support the exchange of retinoids between the neural retina and retinal pigment epithelium (RPE) in the vertebrate eye, exhibits a modular, i.e., repeat, structure. The present study was undertaken to determine whether an individual module of IRBP has activity in retinoid transfer between the RPE and rod photoreceptors. METHODS: The retinoid transfer activity of a recombinant protein corresponding to the fourth module of Xenopus laevis IRBP (X4IRBP) was examined in two ways. First, X4IRBP was tested for its ability to support the regeneration of porphyropsin in detached/reattached Xenopus retina/RPE-eyecups. Following illumination and removal of native IRBP, Xenopus eyecups supplemented with 42 microM X4IRBP or (as a control) Ringer's solution were incubated in darkness and then analyzed for regenerated porphyropsin. Second, toad (Bufo marinus) RPE-eyecup preparations were used to evaluate X4IRBP's ability to promote the release of 11-cis retinal from the RPE. RESULTS: The regeneration of porphyropsin in X4IRBP-supplemented Xenopus retina/RPE-eyecups (0.45 +/- 0.04 nmol; mean +/- SEM, n = 11) exceeded that in controls (0.13 +/- 0.02 nmol, n = 11). For promoting the release of 11-cis retinal from the toad RPE, 42 microM X4IRBP was more effective than equimolar bovine serum albumin although considerably less than that of 26 microM native bovine IRBP. CONCLUSIONS: The results indicate a low but significant activity of IRBP's fourth module in reactions relevant to retinoid exchange.

Resonance Raman evidence for an all-trans to 13-cis isomerization in the proton-pumping cycle of bacteriorhodopsin.

Using a dual-beam flow technique, we have obtained resonance Raman spectra of the M412 photointermediates of both native purple membrane (15H M412) and purple membrane regenerated with 15-deuterioretinal (15D M412). For comparison, we have also obtained Raman spectra of the n-butylamine Schiff bases of the 13-cis and all-trans isomers of 15H and 15D retinal. The 15D model compound spectra, when compared to the 15H spectra, show isotopically induced spectral changes that are markedly different for the two isomers. There is a very close agreement between the frequency and intensity changes which occur upon deuteration of M412 and those which occur upon deuteration of the 13-cis model compound, but not even a qualitative correspondence exists when M412 and the all-trans model compound are similarly compared. These data demonstrate that the chromophore of M412 is an unprotonated Schiff base of 13-cis-retinal rather than all-trans-retinal. An analogous spectral comparison of 15H and 15D light-adapted bacteriorhodopsin (bRLA) with the 15H and 15D protonated Schiff bases of 13-cis- and all-trans-retinal demonstrates that bRLA contains an all-trans chromophore, in agreement with previous extraction experiments. Thus, a trans leads to cis isomerization occurs in the proton-pumping photocycle of Halobacterium halobium.

Resonance Raman spectra of bacteriorhodopsin's primary photoproduct: evidence for a distorted 13-cis retinal chromophore.

We have obtained the resonance Raman spectrum of bacteriorhodopsin's primary photoproduct K with a novel low-temperature spinning sample technique. Purple membrane at 77 K is illuminated with spatially separated actinic (pump) and probe laser beams. The 514-nm pump beam produces a photostationary steady-state mixture of bacteriorhodopsin and K. This mixture is then rotated through the red (676 nm) probe beam, which selectively enhances the Raman scattering from K. The essential advantage of our successive pump-and-probe technique is that it prevents the fluorescence excited by the pump beam from masking the red probe Raman scattering. K exhibits strong Raman lines at 1516, 1294, 1194, 1012, 957, and 811 cm-1. The effects of C15 deuteration on K's fingerprint lines correlate well with those seen in 13-cis model compounds, indicating that K has a 13-cis chromophore. However, the presence of unusually strong "low-wavenumber" lines at 811 and 957 cm-1, attributable to hydrogen out-of-plane wags, indicates that the protein holds the chromophore in a distorted conformation after trans leads to cis isomerization.

Anion-protein interactions during halorhodopsin pumping: halide binding at the protonated Schiff base.

Halorhodopsin (hR), the light-driven chloride pump of Halobacterium halobium, has been studied by Fourier transform infrared (FTIR) spectroscopy. Direct hydrogen bonding of halide ions with the protonated Schiff base (PSB) group was detected by means of halide-dependent perturbations on this group's vibrational frequencies. FTIR difference spectra were obtained of the hR-->hL photoreaction in reconstituted membrane vesicles. Nearly identical results were obtained using either low-temperature static difference spectroscopy at 1-cm-1 resolution or a stroboscopic time-resolved technique with 5-ms temporal and 2-cm-1 spectral resolution. The frequency of the negative difference band due to the PSB C = N stretch mode in the hR state shows a dependence on the type of halide counteranion that is present, 1632 cm-1 in the presence of Cl-, 1631 cm-1 in Br-, and 1629 cm-1 in I-. The C = NH+ stretch frequency thus correlates with the strength of the hydrogen bond formed by the halide. Analogous halide-dependent shifts of the C = NH+ frequency were observed in IR spectra of model compound retinylidene PSB salts. We also observed a significant halide dependence of the visible absorption maximum of hR solubilized in lauryl maltoside detergent. From such halide perturbation effects, we conclude that there is a direct hydrogen-bonded interaction between the Schiff base group and an externally supplied halide ion in the hR state. Halide perturbation effects are also observed for PSB-group vibrations in the hL state. Thus, despite an apparent overall weakening of hydrogen-bonding interactions of the PSB with its environment after chromophore photoisomerization to form hL, the PSB remains hydrogen-bonded to the halide. The results are best explained in terms of a "one-site, two-state" model for anion binding near the chromophore in the hR state, as opposed to a previously proposed two-site model.

Structural comparison of metarhodopsin II, metarhodopsin III, and opsin based on kinetic analysis of Fourier transform infrared difference spectra.

Fourier transform infrared difference spectra were measured at 30-s intervals after a complete bleach of rhodopsin (rho) samples at 20 degrees C and three different pH values. At each pH, all of the spectra could be fit globally to two exponential decay processes. Using a branched unimolecular kinetic model in which metarhodopsin II (meta II) is hydrolyzed to opsin and retinal both directly and through metarhodopsin III (meta III), we calculated rho-->meta II, rho-->meta III, and rho-->opsin difference spectra at each of the pH values and obtained estimates for the microscopic rate constants at each pH. Because of assumptions that had to be made about the branching ratio between the meta II decay pathways, some uncertainties remain in our calculated rho-->meta III difference spectrum at each pH. Nevertheless, our data covering long time ranges, especially those obtained at pH 8, place significant new constraints on the spectrum of meta III and thus on its structure. The rho-->meta II spectrum shows no significant pH dependence over the range examined (pH 5.5-8). However, the rho-->meta III and rho-->opsin spectra each include a limited subset of pH-dependent peaks, which are mostly attributable to titratable amino acid side chains. Our observations can be used to refine an earlier conclusion that the visual pigment refolds to a rhodopsin-like conformation during meta II decay (Rothschild, K.J., J. Gillespie, and W.J. DeGrip. 1987 Biophys. J. 51:345-350). Most of this refolding occurs in the same way at pH values ranging from 5.5 to 8 and whether meta II decays to meta III or opsin. Meta II displays unique spectral perturbations that are mostly attributable to a few residues, probably including three to four aspartic or glutamic acids and an arginine.

Infrared spectroscopic detection of light-induced change in chloride-arginine interaction in halorhodopsin.

A light-induced transient change in the ionic interaction between chloride and arginine in the transmembrane anion pump halorhodopsin (hR) is detected with infrared absorption spectroscopy. In the IR difference spectrum of hR and one of its photoproducts (hL), only a few bands have frequencies that depend on the particular halide ion (Cl-, Br-, or I-) present. Three of the halide-sensitive negative difference bands (at 1695, 1610, and 1170 cm-1) correspond in frequency to arginine C-N vibrations and undergo anion-dependent shifts that match those seen in ethylguanidinium halide model compounds. These shifts reflect the different strengths of the ionic interactions formed with the various halides. We conclude that a halide-arginine ion pair is present in the hR state; this interaction appears to be disrupted by photoconversion to hL.

Protein dynamics in the bacteriorhodopsin photocycle: submillisecond Fourier transform infrared spectra of the L, M, and N photointermediates.

The usefulness of stroboscopic time-resolved Fourier transform IR spectroscopy for studying the dynamics of biological systems is demonstrated. By using this technique, we have obtained broadband IR absorbance difference spectra after photolysis of bacteriorhodospin with a time resolution of approximately 50 microseconds, spectral resolution of 4 cm-1, and a detection limit of delta A less than or equal to 10(-4). These capabilities permit observation of detailed structural changes in individual residues as bacteriorhodopsin passes through its L, M, and N intermediate states near physiological temperatures. When combined with band assignments based on isotope labeling and site-directed mutagenesis, the stroboscopic Fourier transform IR difference spectra show that on the time scale of the L intermediate, Asp-96 has an altered environment that may be accompanied by change in its protonation state. On the time scale of the L----M transition, this Asp-96 perturbation/deprotonation is largely reversed, and Asp-85 becomes protonated. During the M----N transition, Asp-85 appears to remain protonated but undergoes a change in its environment as evidenced by a shift of vC = O from 1761 to 1755 cm-1. The retention of a proton on Asp-85 in the N state indicates that the proton transferred from the Schiff base to this residue in the L----M step is not released to the extracellular medium during the same photocycle, but rather during a subsequent one. Also during the M----N transition, Asp-96 undergoes a deprotonation (possibly for the second time in a single photocycle). Bands in the amide I and amide II spectral regions in the M----N difference spectrum indicate the occurrence of a conformational change involving one or more peptide groups in the protein backbone.

Millisecond Fourier-transform infrared difference spectra of bacteriorhodopsin's M412 photoproduct.

We have obtained room-temperature transient infrared difference spectra of the M412 photoproduct of bacteriorhodopsin (bR) by using a "rapid-sweep" Fourier-transform infrared (FT-IR) technique that permits the collection of an entire spectrum (extending from 1000 to 2000 cm-1 with 8-cm-1 resolution) in 5 ms. These spectra exhibit less than 10(-4) absorbance unit of noise, even utilizing wet samples containing approximately 10 pmol of bR in the measuring beam. The bR----M transient difference spectrum is similar to FT-IR difference spectra previously obtained under conditions where M decay was blocked (low temperature or low humidity). In particular, the transient spectrum exhibits a set of vibrational difference bands that were previously attributed to protonation changes of several tyrosine residues on the basis of isotopic derivative spectra of M at low temperature. Our rapid-sweep FT-IR spectra demonstrate that these tyrosine/tyrosinate bands are also present under more physiological conditions. Despite the overall similarity to the low-temperature and low-humidity spectra, the room-temperature bR----M transient difference spectrum shows significant additional features in the amide I and amide II regions. These features' presence suggests that a small alteration of the protein backbone accompanies M formation under physiological conditions and that this conformational change is inhibited in the absence of liquid water.

Vibrational spectra of individual millimeter-size membrane patches using miniature infrared waveguides.

We have used miniature planar IR waveguides, consisting of Ge strips 30-50 microm thick and 2 mm wide, as evanescent-wave sensors to detect the mid-(IR) evanescent-wave absorbance spectra of small areas of biomolecular monolayers and multilayers. Examples include picomolar quantities of an integral transmembrane protein (bacteriorhodopsin) and lipid (dimyristoyl phosphatidylcholine). IR bands due to the protein and lipid components of the plasma membrane of individual 1.5-mm-diameter devitellinized Xenopus laevis oocytes, submerged in buffer and sticking to the waveguide surface, were also detected. A significant improvement in sensitivity was observed, as compared to previous sizes and geometries of evanescent-wave sensors (e.g., commercially available internal reflection elements or tapered optical fibers). These measurements suggest the feasibility of using such miniature supported planar IR waveguides to observe structural changes in transmembrane proteins functioning in vivo in single cells.

A large photolysis-induced pKa increase of the chromophore counterion in bacteriorhodopsin: implications for ion transport mechanisms of retinal proteins.

The proton-pumping mechanism of bacteriorhodopsin is dependent on a photolysis-induced transfer of a proton from the retinylidene Schiff base chromophore to the aspartate-85 counterion. Up until now, this transfer was ascribed to a > 7-unit decrease in the pKa of the protonated Schiff base caused by photoisomerization of the retinal. However, a comparably large increase in the pKa of the Asp-85 acceptor also plays a role, as we show here with infrared measurements. Furthermore, the shifted vibrational frequency of the Asp-85 COOH group indicates a transient drop in the effective dielectric constant around Asp-85 to approximately 2 in the M photointermediate. This dielectric decrease would cause a > 40 kJ-mol-1 increase in free energy of the anionic form of Asp-85, fully explaining the observed pK alpha increase. An analogous photolysis-induced destabilization of the Schiff base counterion could initiate anion transport in the related protein, halorhodopsin, in which aspartate-85 is replaced by Cl- and the Schiff base proton is consequently never transferred.

Supported planar germanium waveguides for infrared evanescent-wave sensing.

We have fabricated miniature planar IR waveguides with thicknesses of 30-50 mum, consisting of 12-mm long, 2-mm wide strips of Ge supported on ZnS substrates. Evidence for efficient propagation of broadband IR light through these waveguides is provided by the presence of characteristic high- and low-frequency optical cutoffs of Ge; by the observation of an oscillatory interference pattern in the transmittance spectrum, which exhibits a dependence on waveguide thickness and propagation angle that closely matches waveguide theory; and by the detection of strong evanescent-wave absorption from small (2 mm(2)) droplets of liquid, e.g., water, on the waveguide surface. As also predicted by theory, the surface sensitivity (detected light absorbance per unit area of sample-waveguide contact) is shown to increase as a function of incidence or bevel angle.

Fourier transform infrared study of the halorhodopsin chloride pump.

Halorhodopsin (hR) is a light-driven chloride pump located in the cell membrane of Halobacterium halobium. Fourier transform infrared difference spectroscopy has been used to study structural alterations occurring during the hR photocycle. The frequencies of peaks attributed to the retinylidene chromophore are similar to those observed in the spectra of the related protein bacteriorhodopsin (bR), indicating that in hR as in bR an all-trans----13-cis isomerization occurs during formation of the early bathoproduct. Spectral features due to protein structural alterations are also similar for the bR and hR photocycles. For example, formation of the red-shifted primary photoproducts of both hR and bR results in similar carboxyl peaks in the 1730-1745-cm-1 region. However, in contrast to bR, no further changes are observed in the carboxyl region during subsequent steps in the hR photocycle, indicating that additional carboxyl groups are not directly involved in chloride translocation. Overall, the close similarity of vibrations in hR and bR photoproduct difference spectra supports the existence of some common elements in the molecular mechanisms of energy transduction and active transport by these two proteins.

Orientation of the bacteriorhodopsin chromophore probed by polarized Fourier transform infrared difference spectroscopy.

Polarized, low-temperature Fourier transform infrared (FTIR) difference spectroscopy has been used to investigate the structure of bacteriorhodopsin (bR) as it undergoes phototransitions from the light-adapted state, bR570, to the K630 and M412 intermediates. The orientations of specific retinal chromophore and protein groups relative to the membrane plane were calculated from the linear dichroism of the infrared bands, which correspond to the vibrational modes of those groups. The linear dichroism of the chromophore C=C and C-C stretching modes indicates that the long axis of the polyene chain is oriented at 20-25 degrees from the membrane plane at 250 K and that it orients more in-plane when the temperature is reduced to 81 K. The polyene plane is found to be approximately perpendicular to the membrane plane from the linear dichroism calculations of the HOOP (hydrogen out-of-plane) wags. The orientation of the transition dipole moments of chromophore vibrations in the K630 and M412 intermediates has been probed, and the dipole moment direction of the C=O bond of an aspartic acid that is protonated in the bR570----M412 transition has been measured.

Structure-function studies on bacteriorhodopsin. IV. Purification and renaturation of bacterio-opsin polypeptide expressed in Escherichia coli.

Expression of the bacterio-opsin gene in Escherichia coli has been described in the accompanying papers. We now describe rapid and efficient methods for the purification of the E. coli-expressed bacterio-opsin. Bacterio-opsin can be extracted from E. coli membranes in a denatured form by using an organic solvent containing chloroform, methanol, water, and triethylamine. The bacterio-opsin, enriched to 30-50% in the extract, can be further purified to 90% by ion-exchange chromatography on DEAE-Trisacryl or hydroxylapatite chromatography in organic solvents or by preparative sodium dodecyl sulfate gel electrophoresis. In appropriate aqueous phospholipid/detergent mixtures, up to 80% of purified protein refolds and binds retinal covalently to regenerate the bacteriorhodopsin chromophore. When reconstituted into phospholipid vesicles, bacteriorhodopsin from E. coli shows the expected proton pumping activity in response to illumination.