HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 276, Issue 35, 32495-32505, August 31, 2001

This Article Abstract Full Text (PDF) All Versions of this Article: 276/35/32495    most recent M102652200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Brown, L. S. Articles by Spudich, J. L. Search for Related Content PubMed PubMed Citation Articles by Brown, L. S. Articles by Spudich, J. L. Social Bookmarking                      What's this?

Photochemical Reaction Cycle and Proton Transfers in Neurospora Rhodopsin^*

From the ^§ Department of Physiology & Biophysics, University of California, Irvine, California 92697 and the ^¶ Department of Microbiology & Molecular Genetics, The University of Texas Medical School, Houston, Texas 77030

Received for publication, March 26, 2001, and in revised form, May 15, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

It was recently found that NOP-1, a membrane protein of Neurospora crassa, shows homology to haloarchaeal rhodopsins and binds retinal after heterologous expression in Pichia pastoris. We report on spectroscopic properties of the Neurospora rhodopsin (NR). The photocycle was studied with flash photolysis and time-resolved Fourier-transform infrared spectroscopy in the pH range 5-8. Proton release and uptake during the photocycle were monitored with the pH-sensitive dye, pyranine. Kinetic and spectral analysis revealed six distinct states in the NR photocycle, and we describe their spectral properties and pH-dependent kinetics in the visible and infrared ranges. The phenotypes of the mutant NR proteins, D131E and E142Q, in which the homologues of the key carboxylic acids of the light-driven proton pump bacteriorhodopsin, Asp-85 and Asp-96, were replaced, show that Glu-142 is not involved in reprotonation of the Schiff base but Asp-131 may be. This implies that, if the NR photocycle is associated with proton transport, it has a low efficiency, similar to that of haloarchaeal sensory rhodopsin II. Fourier-transform Raman spectroscopy revealed unexpected differences between NR and bacteriorhodopsin in the configuration of the retinal chromophore, which may contribute to the less effective reprotonation switch of NR.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The haloarchaeal retinal proteins, bacteriorhodopsin (BR¹), halorhodopsin (HR), and sensory rhodopsins I and II (SRI and SRII), contain seven transmembrane helices, show extensive sequence homologies, and share similar photochemistry which involves photoisomerization of the retinal from all-trans to 13-cis (for a recent review, see Ref. 1). In these proteins, thermal relaxation of the photoinduced changes drives the "photocycle," and the stored energy of a photon is utilized to perform ion transport (protons for BR, chloride for HR) or signaling for phototaxis (positive or negative, depending on light stimulus color for SRI, negative for SRII). Until very recently, these archaeal rhodopsins were thought to constitute an isolated family. In 1999, a protein with high homology to BR was revealed by partial sequencing of the Neurospora crassa genome (2). This protein, named NOP-1, was heterologously expressed in the methylotrophic yeast Pichia pastoris, and, upon binding of retinal, formed a pigment (NR) with photochemical properties similar to those of SRII (3). At least two distinct intermediates, similar to the M and O states of BR, were described for the NR photocycle (3). Since then, several similar opsins have been found in other fungi (reviewed in Ref. 1) as well as in eubacteria (4). These findings extended the family of archaeal opsins to two more superkingdoms, Bacteria and Eucarya, suggesting that BR-like pigments could serve as energy sources or photoreceptors of ancient origin among species of various taxa.

The physiological role of the fungal opsins is not yet known. They are more likely to be involved in photosensory signaling by light-modulated interaction with signal-transducing proteins, as SRI and SRII (1), than proton pumping as the homologous archaeal (BR) and eubacterial (proteorhodopsin) pigments (5, 4). The photocycle turnover of the transport rhodopsins is typically fast (<50 ms), whereas sensory rhodopsins generate long-living intermediates that serve as signaling states and persist for hundreds of milliseconds. Because Pichia-expressed NR exhibits a slow photocycle similar to sensory rhodopsins, in particular to that of SRII, it was argued (3) that NR would be a very ineffective proton pump. On the other hand, if NR had a photosensory function, a slow photocycle would be beneficial, because the long-lived intermediates are the signaling states in SR proteins (6). However, the properties of the NR photocycle in its natural host are not known. The location of NR in Neurospora and its putative transducer and signaling cascade are likewise not known. Deletion of the nop-1 gene does not lead to any significant abnormalities in the Neurospora life cycle (2), making the role of NR even more enigmatic.

Comparison of the amino acid sequences of NOP-1 (2) and bacterio-opsin (7) indicates that many residues important for proton transport are preserved in the fungal protein, and from this one would expect at least some proton pumping activity. The conserved residues include the complex counter-ion to the retinal Schiff base (i.e. Asp-85, Asp-212, Tyr-57, Tyr-185, and Arg-82, numbering for BR), and the important components of the retinal-binding pocket such as Trp-182, Trp-86, Tyr-83, and Thr-89 (see Ref. 8 for the structural details of BR). The primary proton donor Asp-96 (9, 10) is preserved as Glu, as are its hydrophobic environment and hydrogen-bonding partner, Thr-46 (11). Some of the residues in contact with the retinal (e.g. Ser-141 on helix E) are changed, which may partially account for the large blue shift of the absorption maximum of NR relative to BR (534 versus 568 nm), similar to the shift observed for SRII (12). The most striking differences in the two structures are in the extracellular region. The F-G interhelical loop is quite different from that of BR, and Glu-194, although not Glu-204, is absent. Indeed, the photocycle of NR is reminiscent of the E194Q mutant of BR (13), where a substantial amount of slowly decaying O intermediate accumulates.

An interesting question raised by the overall similarity of the NR photocycle (3) to both SRII and E194Q BR is whether the NR photocycle transports protons. The archaeal sensory rhodopsins do not pump ions when in a tight molecular complex with their cognate transducers, but in some cases light-driven vectorial transport of protons does occur in the absence of transducer. HtrI-free SRI was shown to exhibit single-photon-induced proton pumping out of Halobacterium salinarum cell envelope vesicles at pH  7 (14). Although in cell envelope vesicles H. salinarum SRII exhibits only electroneutral light-induced circulation of protons to and from the extracellular medium (15), the related pigment from Natronobacterium pharaonis was found to have some proton transport activity in such vesicles (16) as well as in black lipid films (17). The proton transport by N. pharaonis SRII was enhanced by factors that favor secondary photoreactions of late photocycle intermediates (presence of sodium azide, low pH, and the F86D mutation) (17). This leaves open a possibility that SRII can transport only by a double-photon mechanism (16, 17). When expressed in Xenopus oocytes, the wild-type N. pharaonis SRII did not show any stationary photocurrent, whereas salinarum SRII exhibited weak, although significant, vectorial proton transport (18).

A suggested explanation for pumping by archaeal sensory rhodopsins and the effects of mutations on SR signaling is that the same conformational changes in BR that contribute to the reprotonation switch in proton transport are responsible for communication of the photoactivation signal to the transducer in the signaling states of SR receptors (6, 19). This unified model is supported by the observation that the opening of the cytoplasmic channel in SRI is blocked by interaction with its transducer HtrI (20), and the weaker proton pumping by SRII is also blocked by interaction with its transducer HtrII (16, 18), very possibly by the same mechanism. A consequence of this view is that under some conditions reprotonation of the Schiff base in the photocycles of sensory rhodopsins may be more rapid from the cytoplasmic side than from the extracellular side, and protons are transported as in BR (21), although with lower efficiency. Thus, to understand the various physiological and non-physiological transport modes, the description of proton acceptors and donors in these photocycles is necessary.

One significant difference between SRII and BR is that SRII lacks the proton donor, Asp-96, on the cytoplasmic side (22). For this reason, the retinal Schiff base of SRII is not reprotonated rapidly from the cytoplasmic direction as in BR but obtains a proton mostly from the extracellular side, in a pH-dependent manner (15). On the other hand, introduction of an aspartic acid, at the position appropriate for a proton donor, into N. pharaonis SRII did not significantly alter the photocycle kinetics (17), and the D96N and D96A mutants of BR lack a donor and have changed kinetics but transport protons (9). It appears that the directionality of Schiff base reprotonation is decided by factors additional to the presence or absence of a cytoplasmic proton donor. Unlike the other sensory rhodopsins, NR does have a cytoplasmic proton donor, Glu-142. Is the reprotonation of its Schiff base similar to the wild-type BR (9) and its D96E mutant (23), or to SRII?

In this paper we report a study of the photocycle of NR by means of visible and infrared spectroscopy, aimed at gaining information about proton transfers in the protein. The measurements were at different pH values, to evaluate if there are any pH-dependent steps associated with proton transfers. Proton release and uptake were followed with a pH-sensitive dye. Results with the E142Q mutant of NR demonstrated that, unlike in BR, the homologue of Asp-96 (Glu-142) does not serve as a proton donor for the retinal Schiff base, and its replacement does not alter the photocycle. It is possible therefore that the reprotonation of the Schiff base is mostly from the extracellular side, decreasing the efficiency of any proton transport, as found for SRII (15, 18).

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Wild-type His-tagged NOP-1 with a truncated N terminus was expressed in P. pastoris as described before (3). Plasmids encoding the E142Q and D131E mutants of NR were constructed by a two-step polymerase chain reaction mutagenesis by a modification of the megaprimer method (24) using as template the NoppHIL-S1 plasmid described previously (3). In the first step the following mutagenic primers were used: 5'-CAAGGCCCACTCGACGTAGC-3' for the D131E mutation and 5'-AGGCACAGCTGCAGCAGC-3' for the E142Q mutation, in combination with the forward primer 5'-TTCGCTCGAGAATTCGAAACA-3' constructed to the beginning of the gene and including an XhoI site immediately upstream of the nop-1 gene. The product of this first reaction was then extended in a second polymerase chain reaction amplification using the reverse primer 5'-ATAGCCTCGAGACACCACG-3' that includes the XhoI site present within the gene. The final 530-base pair product was digested with XhoI and replaced in the NoppHIL-S1 plasmid. The orientation of the replaced fragment and presence of the mutation were confirmed by sequencing.

Mutants of NR were expressed analogously to the wild-type. Membranes containing wild-type or mutant NOP-1 were isolated as described elsewhere (3) with the following modifications. Washed Pichia cells were resuspended in about 100 ml of 1 M sorbitol, 0.5 mM PMSF. Retinal (3 mM stock in ethanol) was added to a final concentration of about 50 µM, and the initially near-white cells were left at room temperature for 1-2 h until the full development of a dark orange color. About 50 mg of lyticase (crude, from Arthrobacter luteus, Sigma Chemical Co.) was added, and the cells were slowly shaken at 30 °C for 3-4 h to digest cell walls. After that, the cells were vortexed several times with 25-30% volume of glass beads (420-600 µm, acid-washed) and centrifuged at low speed (300 × g). The colored supernatant was withdrawn and stored, while the remaining cells and debris were resuspended in a few milliliters of buffer A (7 mM NaH₂PO₄, pH 6.5, 7 mM EDTA, 7 mM DTT, 1 mM phenylmethylsulfonyl fluoride), and vortexing followed by centrifugation was repeated several times until the breakage of the cells was complete. The supernatants were combined and put into several centrifuge tubes on top of 25 ml of 60% sucrose and centrifuged for 90 min at 105,000 × g in a fixed-angle rotor. The colored material on top of the sucrose cushion was collected and washed by centrifugation in buffer A (15 min, 23,000 × g). When further purification was needed, the membranes were treated with 10% DM in buffer A for 30 min and washed several times by centrifugation (15 min, 23,000 × g).

The purple membrane of H. salinarum containing BR was isolated according to the standard procedure (25).

Samples for the FTIR measurements were prepared by drying the protein in 1 mM BTP, pH 7.0, on a CaF₂ window under mild vacuum. After drying, the films were soaked for 60-120 min in 20 mM BTP and succinate buffers, 10 mM NaCl, 5 mM DTT at the specified pH. A 6-µm Teflon spacer (Spectra-Tech Inc., Shelton, CT) was used to fix the sample thickness. This procedure for preparing wet (50-70% water, w/w) films with defined pH was described previously (26). All measurements were done at 25 °C, using a temperature-controlled sample holder (Harrick, Ossining, NY) connected to a water bath (RTE-111, Neslab, Portsmouth, NH).

The FTIR time-resolved measurements were performed on an IFS-66s instrument (Bruker, Germany), at 2-cm¹ resolution. Interferograms were collected in the rapid-scan mode with 85-ms time resolution. The IR detector was equipped with a 2000 cm¹ cut-off filter (Optical Coating Laboratory, Inc., Santa Rosa, CA). Excitation was provided by the second harmonics of the Nd:YAG laser (Minilite II, Continuum, Santa Clara, CA) at 532 nm, with ~7-ns pulse width, and ~2-mJ/cm² pulse energy. The laser pulses were spaced at times greater than 5× the slowest decay time constant. A custom built program provided the triggering of the spectrometer, allowing one full scan before the arrival of the excitation flash from which the baseline was calculated. Kinetic analysis of data was done as described before (13, 26, 27). The data were globally fitted by exponentials using the program FITEXP, and the number of statistically valid transient states was determined by F-test (for details see Ref. 27 and references therein).

FT-Raman measurements were performed as before (28), with a spectral resolution of 2 cm¹. We used a concentrated suspension of Pichia membranes (OD > 10) in the same buffer as for the FTIR measurements. The Raman spectrum of the buffer was subtracted.

Low temperature spectroscopy was performed using a Shimadzu UV-1601 spectrophotometer equipped with an Oxford cryostat with ITC-4 temperature controller (29). Illumination was provided by a 175-watt Cermax xenon lamp (ILC Technology, Sunnyvale, CA) through a 5-mm diameter, 6-ft liquid light guide.

Kinetic measurements in the visible range were done as described earlier (30), using membranes encased in polyacrylamide gels equilibrated with the same buffer as for the infrared measurements. Amplitudes of kinetic traces taken at different wavelength were corrected for progressive photobleaching (less than 20%). Measurements of proton kinetics with pyranine (31) were performed in unbuffered membrane suspensions as before (32) but at low ionic strength (few millimolar of NaCl) to avoid membrane aggregation. The kinetic analyses were performed as for the FTIR data.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Photocycle of NR, as Measured by Spectroscopy in the Visible Range-- Fig. 1 shows time courses of light-induced absorption changes in polyacrylamide gels loaded with membranes of P. pastoris containing NR. These measurements were at four pH values between 5 and 8, to reveal any pH-dependent photocycle steps. Absorption change was followed at 420 nm to monitor the deprotonated retinal Schiff base (the M intermediate), at 500 nm to monitor the depletion of the initial state, and at 600 nm to monitor the appearance of red-shifted states (e.g. the O intermediate). Fig. 1 shows that the photocycle turnover is slow (a few seconds, Table I). In general, it is characterized by intermediates similar to the M and O states of BR and SRII, as noted before at pH 6.5 (3). The improved time resolution allowed us to observe the deprotonation of the Schiff base (rise of the M state), which is somewhat slower (Table I) than the analogous process in BR. A comparison of kinetic traces at different pH values (Fig. 1A-D, Table I) shows that the proton concentration does not greatly affect the Schiff base deprotonation, which is to be expected if this process is an internal proton transfer to a homologue of Asp-85 (Asp-131 for NR) (33), the same way as in BR (34).

View larger version (30K): [in this window] [in a new window]   Fig. 1.   Time courses of absorption changes in the wild-type NR at different pH values. Absorption changes after laser pulse photoexcitation were followed at 420, 500, and 600 nm and are shown on a logarithmic time scale. Conditions: 20 mM BTP and succinate buffers, 10 mM NaCl, 5 mM DTT, pH 5.0 (A), pH 6.0 (B), pH 7.0 (C), pH 8.0 (D), 25 °C. Absorbance of NR at 530 nm was approximately 0.3 (after subtracting light-scattering).

Unlike its deprotonation, the reprotonation of the retinal Schiff base shows marked pH dependence in the pH range between 5 and 8 (Fig. 1A-D). At pH below 5 the protein was unstable. At higher pH, the slow components of the M decay are larger in amplitude and their time constants are somewhat longer (Table I). The increases in the amplitudes are largely responsible for the drastic lengthening of the lifetime of the M state that is evident in Fig. 1. Effectively, the overall decay of M (as determined by a single exponential approximation) is thereby slowed by approximately 300-fold as the pH is increased from 5 to 8. This is accompanied by a dramatic decrease in the amplitude of the red-shifted O-like intermediate(s). Such pH-dependent behavior can be explained as a shift in the equilibrium between the M and N/O intermediates, which should be pH-dependent, because it involves a proton uptake step (see below). pH independence of the M decay is regarded as evidence for an internal proton donor (i.e. Asp-96 in BR) to the Schiff base (9). Strong pH dependence of the M decay in the pH range between 5 and 8 (in Fig. 1A-D) is not a property of the wild-type BR, but it is similar to what was reported for the D96N mutant of BR (35) and for SRII (15) where an internal proton donor is lacking.

We collected kinetic traces similar to those shown in Fig. 1, but at 15-nm intervals in the range 410-665 nm (excluding 530 nm, because of a strong laser artifact), at pH 5 and 7. The traces were globally fitted with a set of five exponentials (Table I). The risk factor for adding the fifth exponential was negligible but would have increased 80-fold for a sixth one. The resulting difference spectra of the species present before each of the five statistically valid kinetic processes are shown in Figs. 2, A and B. These spectra represent mixtures of the true photocycle intermediates. The presence of spectral contributions by M-like and O-like states in more than one kinetic component can be an indication that some steps in the photocycle involve back-reactions with non-negligible rates. An alternative interpretation would call for the existence of parallel reaction pathways, including slow or fast substates of these intermediates, as suggested before (3).

View larger version (33K): [in this window] [in a new window]   Fig. 2.   Difference spectra of the kinetic components of the NR photocycle. Spectra were obtained by global five-exponential fitting of data as in Fig. 1, but at 17 different wavelengths. The calculated points are represented by circles. The solid lines are cubic spline functions, included to guide the eye. Each spectrum represents an absorption change relative to the unilluminated sample, before the indicated kinetic process. A, pH = 5.0; spectrum 1, 1 = 190 µs; spectrum 2, 2 = 4 ms; spectrum 3, 3 = 13 ms; spectrum 4, 4 = 160 ms; spectrum 5, 5 = 1.5 s. Conditions are as in Fig. 1A. B, pH = 7.0; spectrum 1, 1 = 120 µs; spectrum 2, 2 = 0.7 ms; spectrum 3, 3 = 22 ms; spectrum 4, 4 = 190 ms; spectrum 5, 5 = 1.4 s. Conditions are as in Fig. 1C.

The first spectrum (number 1 in Fig. 2, A and B) shows absorption changes before the major rise of the M intermediate. The absorption increase at longer wavelengths (obvious especially at pH 7) signifies the presence of bathointermediate(s). The first possibility is that it is a normal K-like state of the photocycle of NR, originating from the majority of the population that contains all-trans retinal. This idea is supported by spectroscopic data at cryogenic temperatures. When a hydrated film of NR was illuminated at 100 K with blue-green light (410-530 nm), a large red-shift in the spectrum was produced. This low temperature difference spectrum is remarkably similar to the K minus BR spectrum (Fig. 3), but blue-shifted in accordance with the 34-nm shift in the maximum of the spectrum of NR relative to BR. Moreover, as in the case of the K intermediate of BR (36), illumination of the bathointermediate of NR with red light (>610 nm) led to a complete reversal of the changes induced by the blue-green light. Warming up the sample with the K intermediate of NR to 150 K produced species spectrally similar to the L intermediate of BR (not shown). The second possibility is that the red-shifted product in spectrum 1, Fig. 2, originates from a different photocycle, that of a minor fraction of NR containing 13-cis retinal. This possibility cannot be excluded, because retinal extraction (3) and FT-Raman data (see Fig. 10 and its description in the text below) both show the presence of some 13-cis pigment. Similar to what was described for N. pharaonis HR (37), we found no evidence for light- or dark-adaptation that would shift the NR population from 13-cis to all-trans and vice versa. We observed that, similar to BR (38, 39), both dehydration and treatment with detergents increased the amount of the early red-shifted photoproducts, presumably belonging to 13-cis NR (not shown).

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Difference spectra (after illumination minus before illumination) of hydrated films of NR and BR. Spectra taken at 100 K, using a green light (500-550 nm). The spectra are normalized to one another for the sake of comparison.

In search of an L-like intermediate in the photocycle at room temperature, we manipulated the spectra in Fig. 2. The spectra at pH either 5 or 7 could be used to obtain an estimate for the spectrum of L, but by different manipulations. At pH 5, the first spectrum represents a mixture of the M state and the earlier intermediates. We removed the contribution of M by subtracting spectrum 2 from spectrum 1, scaled to eliminate absorption change at 415 nm. The resulting spectrum (Fig. 4A, open circles) appears to be a mixture of an L-like state and a small amount of the red-shifted state (attributed to K) described above. At pH 7, the contribution of the red-shifted state to the first spectrum is greater, but the second and the third spectra are free from red-shifted state(s). Here, the spectra contain mixtures of the L and M intermediates, the third spectrum containing a larger contribution from M than the second one. Their difference, obtained by scaled subtraction to remove the contribution of M and shown in Fig. 4A as closed circles, should represent another estimate of the spectrum of the L-like state. Indeed, the two difference spectra in Fig. 4A look similar. They suggest that, similar to the photocycles of BR and SRII, an L-like intermediate is present in the NR photocycle, as indicated also by low temperature spectroscopy.

View larger version (28K): [in this window] [in a new window]   Fig. 4.   Attempts to calculate the spectra of NR intermediates. Results of arithmetic manipulation of the spectra from Fig. 2 are shown. A, to visualize an L-like intermediate at pH 5 (open circles) and 7 (filled circles), spectrum 2 was interactively subtracted from spectrum 1 in Fig. 2A, and spectrum 3 was subtracted from spectrum 2 in Fig. 2B until the absorbance at 410-425 nm became zero. B, to obtain the spectra of the two last intermediates at pH 7, the contribution of the M intermediate (using spectrum 3 from Fig. 2B) was subtracted either from spectrum 4 (open circles) or 5 (open squares). The normalized spectra of the late intermediates at pH 5 from Fig. 2A are shown for comparison (spectrum 4, closed circles; spectrum 5, closed squares). The lines are included to guide the eye.

The second and third spectra at each pH are dominated by M-like intermediate(s). The fourth and the fifth spectra show a predominance of O-like red-shifted product(s) at lower pH and of M-like blue-shifted products at higher pH (Fig. 2). The two last spectra (spectra 4 and 5) are not identical. Together with the presence of two distinct kinetic components in this time-range, this implies at least two intermediates that arise after the M state. The last of these intermediates appears to be more red-shifted than the preceding one. To compare the spectra of these late intermediates at pH 5 and 7, we subtracted the contribution of M to the spectra at pH 7 (using spectrum 3 from Fig. 2B). The resulting spectra are shown in Fig. 4B along with the corresponding kinetic components at pH 5 (which required no correction because they contained no contribution from M), taken directly from Fig. 2A. After having removed the contribution of M, the spectrum of the fourth kinetic component at pH 7 agrees well with the spectrum at pH 5, and this is so for the spectra of the fifth component as well. Therefore, the amplitude of the O-like intermediate(s) at higher pH is less, mainly because the decay of the M state is slower and not because of redistribution of any N-like and O-like intermediates. On the other hand, the calculated spectra of the fourth and fifth components are not identical. They might contain different mixtures of two O states, as evidenced in the photocycle of E194Q BR (26) or, additionally, an N state, which is suggested by the FTIR spectra (see below).

Proton Uptake and Release by NR-- We employed two different methods. First, we illuminated an unbuffered suspension of NR-containing membranes with steady yellow light (wavelengths > 470 nm) and measured the changes of pH with a microelectrode. We observed rapid alkalinization of the medium (faster than 1 s), which relaxed after the light was switched off. This can be interpreted as the accumulation of a slowly decaying photocycle intermediate, which is formed concomitantly with or after the proton uptake from the bulk and decays along with the proton release. Similar pH changes were observed before for SRII (15) and wild-type BR at low pH (40), where proton uptake occurs in the photocycle before proton release.

To measure the kinetics of proton release and uptake, we used flash illumination and the pH-sensitive dye pyranine. Fig. 5A shows proton kinetics in unbuffered suspension of NR membranes and chromophore kinetics under the same conditions. Proton uptake precedes proton release, as expected from the photostationary measurements. The proton kinetics seem to correlate with the kinetics of the red-shifted intermediate(s), similar to what was observed for SRII (15) and the E194Q mutant of BR (13). Consistent with the occurrence of proton uptake at this time, the kinetics of decay of the M intermediate was found to be markedly pH-dependent (Fig. 1).

View larger version (33K): [in this window] [in a new window]   Fig. 5.   Kinetics of the proton release and uptake. Absorption changes at 457 nm without pyranine were subtracted from those with pyranine. Chromophore kinetics, obtained similar to those in Fig. 1, are shown for comparison. Conditions: A, NR membranes, 5 mM NaCl, pH 5.9, with or without 250 µM pyranine. B, NR membranes treated with DM, 2 mM NaCl, pH 6.3, with or without 150 µM pyranine.

Reliable measurements of proton kinetics require a high signal-to-noise ratio, which is difficult to achieve in the turbid suspensions of NR membranes. To enrich the membranes in NR and improve their optical properties, we treated the membranes with DM, which extracted some of the unwanted proteins and lipids. The DM treatment did change the photocycle of NR (Fig. 5B) but only to the extent that would be caused by an increase of pH (by about 1.5 units). This might be expected from a change of surface pH after removal of acidic lipids. As in the untreated membranes (compare with Fig. 1), the amount of the late red-shifted intermediate(s) decreases dramatically, and the slow phases of M decay have markedly increased amplitudes. Another difference is that the apparent decay of absorption at 600 nm is more rapid than the decay of M. This points out an interesting characteristic of this photocycle, clarified by the spectral analysis (Fig. 2), which showed that there are two red-shifted kinetic components, and the last one is more red-shifted. The proton kinetics in this sample, shown in Fig. 5B, are qualitatively not very different from those measured before the detergent treatment. However, now they do not agree with the absorption changes at 600 nm, which originate from the next to last intermediate and imply that proton uptake and release correlate with the rise and decay of the last intermediate of the NR photocycle instead.

Photocycles of NR Mutants, as Measured by Spectroscopy in the Visible Range-- We replaced the homologues of Asp-85 and Asp-96, the two carboxylic acids crucial for the functioning of BR (10, 34). Asp-131 of NR (the homologue of Asp-85) on the extracellular side was replaced by Glu, and Glu-142 (the homologue of Asp-96) on the cytoplasmic side by Gln. The mutant D131N was not used because it could not be reconstituted with retinal, even upon incubation for as long as 24 h. Fig. 6 shows photocycle kinetics for the E142Q mutant at various pH values under the same conditions as used for wild-type NR (Fig. 1). The photocycles of the mutant and the wild-type proteins have the same overall character (Table I), and the same pH dependence. The only (and minor) difference is that the M decay of the mutant is like that of the wild-type at a somewhat lower pH, by 0.5-0.7 unit.

View larger version (29K): [in this window] [in a new window]   Fig. 6.   Time courses of absorption changes in E142Q NR mutant at different pH values. Absorption changes after laser pulse photoexcitation were followed at 420, 500, and 600 nm, and are shown on a logarithmic time scale. Conditions: same as in Fig. 1, pH 5.0 (A), pH 6.0 (B), pH 7.0 (C), pH 8.0 (D).

To measure photocycle kinetics in the D131E mutant of NR we had to enrich the sample by treating it with DM, because its expression yield was much lower than that for the wild-type. For this reason, its kinetics are compared with wild-type NR (Fig. 7, Table I) treated with DM in a same way. There are two important differences. First, formation of the M intermediate is much more rapid in the mutant (it is not kinetically resolved in these measurements). Second, the relative amplitude of O-like intermediate(s) is much larger in D131E than in the wild-type. Interestingly, both of these kinetic features are characteristic of the homologous D85E mutant of BR also (41).

View larger version (27K): [in this window] [in a new window]   Fig. 7.   Time courses of flash-induced absorption changes in the wild-type NR (A) and its D131E mutant (B), after treatment with DM. Absorption changes after laser pulse photoexcitation were followed at 420, 500, and 600 nm, and are shown on a logarithmic time scale. Conditions: same as in Fig. 1B, pH 6.0.

Photocycles of the Wild-type NR and Its E142Q Mutant, as Measured by FTIR Spectroscopy-- In general, the FTIR difference spectra of NR (Fig. 8) are similar to those of other halobacterial retinal proteins, such as SRII (as shown earlier by low temperature static measurements (33)) or BR at low pH (spectrum 7, Fig. 8). The latter spectrum contains a large contribution from the O state, as indicated by changes in the C-C stretch region characteristic of a distorted all-trans retinal (42, 43). Our rapid-scan FTIR measurements of wet films of NR membranes yielded difference spectra (Fig. 8, spectra 1-6) corresponding to the last kinetic component in the visible range (Fig. 2, spectrum 5). It is not clear how these spectra should be normalized for their comparison. As expected from the visible spectra, these infrared spectra are dominated by O-like features at lower pH (as well as N-like features), whereas M-like features appear at higher pH (Fig. 8, spectra 3-4). When scaled to the amplitude of the C=O stretch band of Asp-131, the spectra at higher pH are characterized by larger negative amplitudes of the ethylenic stretch at 1534 cm¹, characteristic for the M-like state(s), and lower positive amplitudes at 1508 cm¹, characteristic for the O-like state(s) (44). The latter band is distinct from the ethylenic stretch band of the red-shifted photointermediate of 13-cis NR at 1525 cm¹, as readily observed in dry films of DM-treated Pichia membranes (see below).

View larger version (28K): [in this window] [in a new window]   Fig. 8.   FTIR difference spectra of the NR membrane films at pH 5.0 (1), pH 6.0 (2), pH 7.0 (3), pH 8.0 (4). The films were equilibrated with the same buffer as in Fig. 1. The spectra represent changes observed before the last kinetic component of the photocycle (cf. Fig. 2). Curve 5 represents the same spectrum for E142Q mutant at pH 5. Curve 6 shows the spectrum of wild-type NR, in the same buffer but with D2O instead of H2O (pD 6.0). The spectra are normalized according to the summed amplitude of the two protonation bands of Asp-131 (for the N/O-like states at 1752 cm1, and for the M-like state at 1759 cm1). Curve 7 shows a difference spectrum measured approximately 10 ms after laser excitation in the wild-type BR at pH 3.5, 5 mM succinate, 10 mM NaCl, at 25 °C.

The amplitudes of the positive bands at 1186, 1300, and 1393 cm¹ are diagnostic for the N intermediate of BR (45). In NR they are somewhat smaller at pH 8 than at pH 5-6 (Fig. 8). In the fingerprint region, the IR spectra of NR at all pH values (from 5 to 8, spectra 1-4) are remarkably similar to the spectrum of the mixture of late intermediates (M, N, and O) of BR at pH 3.5 (spectrum 7 in Fig. 8). Only two negative bands are present (at approximately 1244 and 1200 cm¹) out of the three bands from depletion of the all-trans retinal usually seen in the L, M, and N states of BR at physiological pH (at approximately 1252, 1200, and 1168 cm¹) (46). The third negative band at 1164 cm¹ is very small, as in BR at pH 3.5 (spectrum 7 in Fig. 8), and is present in NR only at pH above 6 (Fig. 8, spectra 3-4). The positive band at 1186 cm¹, usually seen in N and L states of BR (42), appears in NR at all pH values tested, as in BR at pH 3.5 (spectrum 7). A positive shoulder at 1175 cm¹, which develops into a separate band at pH 8 (spectrum 4), is also in accord with the BR spectrum at pH 3.5. Together with the absence of the negative band at 1164 cm¹ (1168 cm¹ in the case of BR), the latter shoulder/band is a clear indication for the presence of O-like state(s) (42, 43). An additional negative band at 1275 cm¹ is present both in NR and in BR at pH 3.5. This band is very small in BR at physiological pH, although it is greatly increased in the spectra of M/N-mixtures in the E194Q mutant (26).

We note that the frequency of one of the depletion bands from all-trans retinal (1244 cm¹) is downshifted from what is observed in the analogous spectrum in BR (1253 cm¹) (42), as in SRII (47). The chromophore origin of this band is confirmed by our FT-Raman data (Fig. 10 and its description in a text below). This band is D₂O-sensitive (spectrum 6 of Fig. 8) in a non-trivial way, its position being upshifted by 12 cm¹. Normally, deuterium causes a downshift of frequency for bands originating from R-H/D stretching vibrations. In BR this band (at 1255 cm¹) was assigned to contributions from C₁₂-C₁₃ stretch and from Lys-216 rocking motions (44), and its unusual deuterium upshift is known (43). In NR this contribution would be from Lys-263.

The carboxylic stretch region (expanded in Fig. 9) shows several prominent bands, all of which downshift in D₂O by approximately 8 cm¹ (Fig. 9, spectrum 6), as expected for protonated carboxylic stretch bands (48). The positive band at 1752 cm¹ was assigned to Asp-131 before (33), based on its shifted position in the mutant D131E. The homologous Asp-85 band in the M state of BR has a higher frequency, by about 7 cm¹, than in the late intermediates (N and O) (42, 43, 46). This is also true for the NR spectra (Fig. 9, spectra 1-4) where the M intermediate is seen as a shoulder at 1759 cm¹ at higher pH. There is also a pair of negative and positive bands at 1738/1721 cm¹. These bands are similar to those observed for perturbation of Asp-115 of BR (34, 49) and its homologue in SRII (47) but not in N. pharaonis SRII (50), which lacks this aspartate (51). By analogy, these bands may originate from perturbation of Asp-161 of NR, the homologue of Asp-115 of BR. On the other hand, the greater amplitude of the 1721 cm¹ band at higher pH could mean that, similar to the E194Q mutant of BR (26), the proton lost by the homologue of Asp-85 (Asp-131) is transferred to the homologue of Asp-212 (Asp-259), to be released to the bulk during decay of the last intermediate. Finally, there is another broad negative band around 1711 cm¹, which disappears at higher pH. This band could belong to one of the glutamates or aspartates of NR, which deprotonates in the late intermediates of the photocycle. The absence of this band at high pH would mean that it is titrated in the pH range of the measurements (from 5 to 8).

View larger version (22K): [in this window] [in a new window]   Fig. 9.   The same FTIR difference spectra as in Fig. 8, but with the carboxylic stretch region expanded. Wild-type NR, pH 5.0 (1), pH 6.0 (2), pH 7.0 (3), pH 8.0 (4). E142Q, pH 5.0 (5). Wild-type NR in D2O, pD 6.0 (6). Wild-type BR, pH 3.5 (7).

NR contains a small fraction of 13-cis retinal, as noted before (3) and evident also from our FT-Raman spectra (see below). Thus, small contributions from the 13-cis cycle could be expected. The amount of 13-cis species is greatly increased by drying the sample, and, to a lesser extent, by DM treatment. To study the 13-cis photocycle, we measured time-resolved FTIR spectra of dry DM-treated membranes (data not shown). The FTIR spectra of the two last kinetic components in these samples suggest mixtures of two different intermediates. One of these originates from an M-like state, whereas the other comes from a red-shifted state with all-trans retinal, characterized by a 1525 cm¹ positive ethylenic stretch band. The latter component decays well before the final decay of the M-like intermediate and may be similar to the red-shifted photoproduct of 13-cis BR (52). Comparison of this spectrum with the spectra we report in Fig. 8 shows only minor contribution from this intermediate of 13-cis NR.

In accordance with the similarity of the E142Q and the wild-type NR photocycles in the visible spectrum (Fig. 6), the infrared difference spectrum of E142Q (Figs. 8 and 9, spectrum 5) is very similar to the spectrum of wild-type NR. The carboxylic stretch region (Fig. 9) did not reveal the disappearance of protonation/deprotonation bands of Glu-142 that might have been expected from its replacement with Gln. If Glu-142 were deprotonated in the wild-type NR, we would be able to detect its differential band, because the proton uptake is slower (see Fig. 5) than the time resolution of the FTIR measurements. From these results, we can conclude that the negative band at 1711 cm¹ does not belong to Glu-142, and that, in fact, Glu-142 does not appear to undergo the protonation change expected if it were the proton donor to the Schiff base.

Characterization of NR by FT-Raman Spectroscopy-- Fig. 10 shows FT-Raman spectra of the dark-adapted states of BR and NR, containing bands nearly entirely from the chromophore (53). As judged from the relative amplitude of the 1187 cm¹ band in the fingerprint region and a shoulder in the ethylenic stretch band, both proteins are mixtures of all-trans and 13-cis retinal species (54). This correlates with earlier retinal extraction data (3). Despite the similarity of the two spectra, there are some notable differences. The frequency of the ethylenic stretch of NR is upshifted relative to BR, as expected from its blue-shifted absorption maximum in the visible spectrum (55). The C=N stretch band of the protonated Schiff base could not be reliably observed because of a strong amide-I band at 1658 cm¹. The positions of many bands in the fingerprint C-C stretch region coincide in the two proteins. The exceptions include the two bands associated with C₁₂-C₁₃ vibrations, at 1255 cm¹ (56) and 1171 cm¹ (57) in BR. As noticed in the FTIR difference spectra of NR (see above), the frequencies of both bands are downshifted, to 1244 cm¹ and 1166 cm¹, respectively.

View larger version (27K): [in this window] [in a new window]   Fig. 10.   FT-Raman spectra of dark-adapted suspensions of the wild-type BR and NR. Conditions: 0.5 M Na2SO4, 0.1 M phosphate buffer, 0.1 M succinate buffer, pH 5 for BR; the same as in Fig. 1B for NR. Contributions of the buffers were subtracted. The spectra were normalized by the amplitude of their ethylenic stretch band.

Several bands representing in-plane and out-of-plane hydrogen vibration modes are altered in NR. For example, the 15-H wag band (56) at 1004 cm¹, which is nearly absent in BR, now surpasses in amplitude the 1010 cm¹ methyl rock band. The 985-cm¹ wag band (combined 7-H and 8-H) and 962-cm¹ wag band (combined 11-H and 12-H) (56) disappear completely. The 1347-cm¹ band, which represents in-plane rocking vibrations of the hydrogens of the Schiff base nitrogen and C₁₅ (56), is greatly diminished in NR. On the other hand, the amplitude of the 1303-cm¹ band, assigned to motions of hydrogens at C₇ and C₈ (56), is increased. The most prominent difference between FT-Raman spectra of BR and NR is the 1446-cm¹ band, which has an unusually large amplitude in NR (Fig. 10). This band is due to asymmetric deformations of methyl groups of the retinal (56).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

We have studied the properties of NR from N. crassa expressed in P. pastoris by means of visible, FTIR, and FT-Raman spectroscopy. Even though the close similarity of the amino acid sequences of BR and NR suggests that the latter could also have proton pumping activity, the initial spectroscopic characterization of NR (3) had indicated that it is more like the sensory pigment SRII. We employed several techniques to get more information about proton transfers during the photocycle of NR and to learn whether these transfers could lead to proton pumping. First, we studied the details of the NR photocycle by visible spectroscopy at various pH values to detect pH-dependent steps that may be associated with proton transfers. Second, we followed uptake and release of protons by NR using the pH-sensitive dye pyranine. Third, we studied the photocycles of two mutants of NR, E142Q on the cytoplasmic side and D131E on the extracellular side, to test the participation of these carboxylic acids in the deprotonation and reprotonation of the Schiff base and thereby to uncover the vectoriality of the proton transfers. Fourth, we used time-resolved FTIR spectroscopy to follow the protonation states of carboxylic acids and their correlation with changes in the configuration of the chromophore. Finally, we employed FT-Raman spectroscopy to elucidate differences in the retinal configuration in BR and NR, which may be responsible for their different behavior.

The general features of the NR photocycle emerging from this (Fig. 1, Table I) and the earlier (3) studies are similar to those of SRII and the E194Q mutant of BR. The combination of exponential analysis of the kinetic data with the results of low temperature spectroscopy indicates the presence of at least six photocycle intermediates. They are spectrally similar to the major intermediates found in the photocycles of BR and SR-II and seem to occur in the same sequence. The first NR photocycle state described is a K-like bathointermediate, detected at cryogenic temperatures (Fig. 3). After the decay of this K-like state, we detected the presence of at least five intermediates on a statistically valid level. The calculated spectra associated with the apparent transitions between their mixtures are presented in Fig. 2. The presence of M-like and O-like states is evident from positive absorption changes in the blue and red regions of the spectra. The K intermediate decays into an L-like state, which is visualized (Fig. 4A) by removing contribution of M from the spectra in Fig. 2, and is detectable at low temperature. Similar to BR (58) and SRII (59, 60), the retinal Schiff base deprotonates in the submillisecond time range, giving rise to an M-like intermediate. Apparently, the Schiff base proton is transferred to Asp-131, the homologue of BR's Asp-85, as found earlier by low temperature FTIR spectroscopy (33). The phenotype of the D131E mutant supports this idea, showing dramatically accelerated formation of the M intermediate (Fig. 7) similar to the D85E mutant of BR (41). Thus, it seems that the first half of the photocycle is very similar for all three pigments (BR, SRII, and NR) and that the energy stored by the chromophore is sufficient to perform the proton transfer obligatory for pumping, from the Schiff base to the primary proton acceptor. However, the results with NR suggest that this step may not be sufficient for proton transport, as reported in earlier studies of SRII also (15).

The details of the reprotonation of the retinal Schiff base of NR (decay of the M intermediate) are quite different from those in wild-type BR, but similar to SRII (15). The decay of M is strongly pH-dependent (Fig. 1), unlike in BR where reprotonation of the Schiff base is from the internal donor Asp-96 and nearly pH-independent in a wide pH range (35). This fact alone makes it questionable that the homologue of Asp-96, Glu-142, is involved in Schiff base reprotonation. Indeed, the photocycle kinetics of the E142Q mutant shows very little change (Fig. 6), and reprotonation of the Schiff base is in fact faster than in wild-type NR at the same pH. This implies that reprotonation of the Schiff base may be from the extracellular side, as in SRII (15). That would make the NR photocycle a non-transporting one. Alternatively, the reprotonation of the Schiff base might occur from the cytoplasmic side but bypassing Glu-142. The FTIR spectra of the carboxylic stretch region at lower pH (Fig. 9) show a negative band at 1711 cm¹, which implies deprotonation of a Glu or an Asp in late intermediates of the NR photocycle. This band is not from Glu-142, because it is present also in the spectrum of the E142Q mutant. The pH dependence of the amplitude of this band correlates with the pH dependence of the M decay, and it is tempting to assign it to the proton donor for the Schiff base. Its apparent pK_a in the unphotolyzed state of NR is about 6. If the reprotonation occurs from the extracellular side, this band could belong to the homologue of BR's Glu-204, i.e. Glu-251. It should be noted that, in analogy with BR, one would expect to see a deprotonation band of the proton donor to the Schiff base only in the N-like intermediate (43). Even though a presence of the N-like state cannot be unambiguously derived from spectra in the visible range (Fig. 2), it is evident from the time-resolved IR spectra (Fig. 8).

The visible (Fig. 2) and the infrared (Fig. 8) spectra imply that there are at least two late intermediates after the decay of the M-like state. Thus, the photocycle of NR contains all the intermediates (K-, L-, M-, N-, and O-like states) previously described for the other retinal-proteins (BR (58) and SRII (60, 61)). The M- and the O-like intermediates of NR were described earlier (3), but the K-, L-, and N-like states have not been detected before. Comparing the number of kinetic components (five) with the number of the spectrally distinct states after K-like intermediate (four, L-, M-, N-, and O-like) reveals that at least one of those intermediates possesses two kinetically distinct (but spectrally similar) substates (e.g. O₁ and O₂). There are both strong N-like features and O-like features in the FTIR difference spectra of the last kinetic component (Fig. 8). The rise of this component correlates with the proton uptake (Fig. 5), as detected by pyranine. The Schiff base could be reprotonated either directly from the bulk or via its putative primary proton donor (the one with infrared band at 1711 cm¹). The pH dependence of the amplitude of the red-shifted late intermediates is similar to that observed for SRII (15), where increase in the transient accumulation of the O intermediate correlates with faster rate of M decay at lower pH. The larger amplitude of the late red-shifted species in the D131E mutant of NR (as compared with the wild-type) may mean that their decay is determined by the rate of the deprotonation of Asp-131, similar to Asp-85 in BR (62). As in BR at low pH (40) or in its E194Q mutant (13), this deprotonation is accompanied by the final proton release to the bulk (Fig. 5).

It is not easy to determine whether NR transports protons. We tested for such transport directly, using a pH electrode and three enclosed membrane systems: liposomes, membrane vesicles, and whole Pichia cells (not shown). Even if NR can pump protons, the resulting pH changes will be at least 100 times smaller than for wild-type BR, because the NR photocycle is much slower. For that reason we used liposomes with the D96N mutant of BR, which has even slower turnover rates, as a control. We were able to detect the signals from proton pumping by D96N BR but not from NR. Instead, we detected pH changes from the photostationary accumulation of the NR intermediates as was seen also in membrane suspensions (see "Results"). Taking into account the small magnitude of the expected signal, together with uncertainties regarding the orientation of NR in the liposomes and the proton permeability of the yeast membranes, we felt that although the weight of evidence is against proton transport by NR, it does not disprove it. That Glu-142 does not participate in reprotonation of the Schiff base and its replacement changes the photocycle very little gives stronger support to the idea that NR does not transport protons. If so, it is a puzzling question why it does not. There is a potential proton donor on the cytoplasmic side, although even if it did not contain one, that would not prevent BR from transporting protons, as in a case of the D96N mutant (9). One obvious hypothesis would be that NR does not possess what is called a reprotonation switch (63-65), i.e. its Schiff base does not gain access to the cytoplasmic side (30). The reprotonation switch may require conformational changes of the protein backbone (64, 66), the retinal itself (67, 68), and/or changes in the proton affinities of the proton donor and acceptor (21). Conformational changes that contribute to the switch in BR may occur in NR but without breaking the connection of the Schiff base to the extracellular side. Do we obtain any clues about the switch from the properties of NR? The FT-Raman spectra (Fig. 10) show that several bands due to methyl and hydrogen vibrations from the retinal and Lys-263 are strongly altered as compared with BR. This means that the conformation of the retinal and Lys-263 is quite different from that in BR, a consequence of a potentially very important difference in the primary sequence of helix G. As detected by x-ray crystallography (69), helix G of BR contains a non-proline kink, a -bulge, at Lys-216 that binds the retinal. In NR, the homologue of the otherwise well-conserved (70) Val-217 is a proline. Model building indicates that the irregularity of its hydrogen bonding will strongly alter the geometry of the -bulge and restrict the motions of Lys-263. This may be a reason for the different local environment of the retinal and perhaps an altered reprotonation switch.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grants GM29498 (to J. K. L.) and GM27750 (to J. L. S.) and by Department of Energy Grant DEFG03-86ER13525 (to J. K. L.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^§ To whom correspondence should be addressed. Tel.: 949-824-7783; Fax: 949-824-8540; E-mail: lsbrown@uci.edu.

Published, JBC Papers in Press, July 2, 2001, DOI 10.1074/jbc.M102652200

	ABBREVIATIONS

The abbreviations used are: BR, bacteriorhodopsin; HR, halorhodopsin; SRI and SRII, haloarchaeal sensory rhodopsins I and II, respectively; NOP-1, opsin from N. crassa; NR, rhodopsin from N. crassa; K, L, M, N, and O, photocycle intermediates of BR or analogous states in SRII and NR; DM, n-dodecyl -D-maltoside; FTIR, Fourier-transform infrared; DTT, dithiothreitol; BTP, bis-tris-propane.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				V. B. Bergo, M. Ntefidou, V. D. Trivedi, J. J. Amsden, J. M. Kralj, K. J. Rothschild, and J. L. Spudich Conformational Changes in the Photocycle of Anabaena Sensory Rhodopsin: ABSENCE OF THE SCHIFF BASE COUNTERION PROTONATION SIGNAL J. Biol. Chem., June 2, 2006; 281(22): 15208 - 15214. [Abstract] [Full Text] [PDF]

				S. A. Waschuk, A. G. Bezerra Jr., L. Shi, and L. S. Brown From the Cover: Leptosphaeria rhodopsin: Bacteriorhodopsin-like proton pump from a eukaryote PNAS, May 10, 2005; 102(19): 6879 - 6883. [Abstract] [Full Text] [PDF]

				S. Casas-Flores, M. Rios-Momberg, M. Bibbins, P. Ponce-Noyola, and A. Herrera-Estrella BLR-1 and BLR-2, key regulatory elements of photoconidiation and mycelial growth in Trichoderma atroviride Microbiology, November 1, 2004; 150(11): 3561 - 3569. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 276/35/32495    most recent M102652200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Brown, L. S. Articles by Spudich, J. L. Search for Related Content PubMed PubMed Citation Articles by Brown, L. S. Articles by Spudich, J. L. Social Bookmarking                      What's this?