Expression of Xanthorhodopsin in Escherichia coli.

Xanthorhodopsin (XR) from Salinibacter ruber is a light-driven proton pump containing retinal and a light-harvesting carotenoid antenna salinixanthin. Previous structure-functional studies of XR were conducted using a protein isolated from the native host only due to the absence of heterologous expression in Escherichia coli. In this paper, we describe cell-free synthesis and incorporation in lipid-protein nanodiscs of the recombinant XR that demonstrated its principal compatibility with E. coli biosynthetic machinery. To produce XR in E. coli, three C-terminal deletion variants of this protein were constructed. In contrast to the full-length XR, their expression resulted in efficient synthesis in E. coli cells. However, cells producing recombinant XR variants bound retinal only upon growth in minimal medium, not in the rich one. The XR3 variant with deletion of ten C-terminal amino acid residues was obtained and characterized. Its absorption spectrum and photocycle kinetics were close to those reported for XR isolated from S. ruber membranes and bleached from salinixanthin. We have also constructed the first mutants of XR, H62M and D96N, and examined their properties.

Proton transfer reactions in donor site mutants of ESR, a retinal protein from Exiguobacterium sibiricum.

Light-driven proton transport by microbial retinal proteins such as archaeal bacteriorhodopsin involves carboxylic residues as internal proton donors to the catalytic center which is a retinal Schiff base (SB). The proton donor, Asp96 in bacteriorhodopsin, supplies a proton to the transiently deprotonated Schiff base during the photochemical cycle. Subsequent proton uptake resets the protonated state of the donor. This two step process became a distinctive signature of retinal based proton pumps. Similar steps are observed also in many natural variants of bacterial proteorhodopsins and xanthorhodopsins where glutamic acid residues serve as a proton donor. Recently, however, an exception to this rule was found. A retinal protein from Exiguobacterium sibiricum, ESR, contains a Lys residue in place of Asp or Glu, which facilitates proton transfer from the bulk to the SB. Lys96 can be functionally replaced with the more common donor residues, Asp or Glu. Proton transfer to the SB in the mutants containing these replacements (K96E and K96D/A47T) is much faster than in the proteins lacking the proton donor (K96A and similar mutants), and in the case of K96D/A47T, comparable with that in the wild type, indicating that carboxylic residues can replace Lys96 as proton donors in ESR. We show here that there are important differences in the functioning of these residues in ESR from the way Asp96 functions in bacteriorhodopsin. Reprotonation of the SB and proton uptake from the bulk occur almost simultaneously during the M to N transition (as in the wild type ESR at neutral pH), whereas in bacteriorhodopsin these two steps are well separated in time and occur during the M to N and N to O transitions, respectively.

Light-driven Na(+) pump from Gillisia limnaea: a high-affinity Na(+) binding site is formed transiently in the photocycle.

A group of microbial retinal proteins most closely related to the proton pump xanthorhodopsin has a novel sequence motif and a novel function. Instead of, or in addition to, proton transport, they perform light-driven sodium ion transport, as reported for one representative of this group (KR2) from Krokinobacter. In this paper, we examine a similar protein, GLR from Gillisia limnaea, expressed in Escherichia coli, which shares some properties with KR2 but transports only Na(+). The absorption spectrum of GLR is insensitive to Na(+) at concentrations of ≤3 M. However, very low concentrations of Na(+) cause profound differences in the decay and rise time of photocycle intermediates, consistent with a switch from a "Na(+)-independent" to a "Na(+)-dependent" photocycle (or photocycle branch) at ∼60 µM Na(+). The rates of photocycle steps in the latter, but not the former, are linearly dependent on Na(+) concentration. This suggests that a high-affinity Na(+) binding site is created transiently after photoexcitation, and entry of Na(+) from the bulk to this site redirects the course of events in the remainder of the cycle. A greater concentration of Na(+) is needed for switching the reaction path at lower pH. The data suggest therefore competition between H(+) and Na(+) to determine the two alternative pathways. The idea that a Na(+) binding site can be created at the Schiff base counterion is supported by the finding that upon perturbation of this region in the D251E mutant, Na(+) binds without photoexcitation. Binding of Na(+) to the mutant shifts the chromophore maximum to the red like that of H(+), which occurs in the photocycle of the wild type.

Directed evolution of a far-red fluorescent rhodopsin.

Microbial rhodopsins are a diverse group of photoactive transmembrane proteins found in all three domains of life. A member of this protein family, Archaerhodopsin-3 (Arch) of halobacterium Halorubrum sodomense, was recently shown to function as a fluorescent indicator of membrane potential when expressed in mammalian neurons. Arch fluorescence, however, is very dim and is not optimal for applications in live-cell imaging. We used directed evolution to identify mutations that dramatically improve the absolute brightness of Arch, as confirmed biochemically and with live-cell imaging (in Escherichia coli and human embryonic kidney 293 cells). In some fluorescent Arch variants, the pK(a) of the protonated Schiff-base linkage to retinal is near neutral pH, a useful feature for voltage-sensing applications. These bright Arch variants enable labeling of biological membranes in the far-red/infrared and exhibit the furthest red-shifted fluorescence emission thus far reported for a fluorescent protein (maximal excitation/emission at ∼ 620 nm/730 nm).

Breaking the carboxyl rule: lysine 96 facilitates reprotonation of the Schiff base in the photocycle of a retinal protein from Exiguobacterium sibiricum.

A lysine instead of the usual carboxyl group is in place of the internal proton donor to the retinal Schiff base in the light-driven proton pump of Exiguobacterium sibiricum (ESR). The involvement of this lysine in proton transfer is indicated by the finding that its substitution with alanine or other residues slows reprotonation of the Schiff base (decay of the M intermediate) by more than 2 orders of magnitude. In these mutants, the rate constant of the M decay linearly decreases with a decrease in proton concentration, as expected if reprotonation is limited by the uptake of a proton from the bulk. In wild type ESR, M decay is biphasic, and the rate constants are nearly pH-independent between pH 6 and 9. Proton uptake occurs after M formation but before M decay, which is especially evident in D2O and at high pH. Proton uptake is biphasic; the amplitude of the fast phase decreases with a pKa of 8.5 ± 0.3, which reflects the pKa of the donor during proton uptake. Similarly, the fraction of the faster component of M decay decreases and the slower one increases, with a pKa of 8.1 ± 0.2. The data therefore suggest that the reprotonation of the Schiff base in ESR is preceded by transient protonation of an initially unprotonated donor, which is probably the ε-amino group of Lys-96 or a water molecule in its vicinity, and it facilitates proton delivery from the bulk to the reaction center of the protein.

Photocycle of Exiguobacterium sibiricum rhodopsin characterized by low-temperature trapping in the IR and time-resolved studies in the visible.

The photocycle of the retinal protein from Exiguobacterium sibiricum, which differs from bacteriorhodopsin in both its primary donor and acceptor, is characterized by visible and infrared spectroscopy. At pH above pKa ~6.5, we find a bacteriorhodopsin-like photocycle, which originates from excitation of the all-trans retinal chromophore with K-, L-, M-, and N-like intermediates. At pH below pKa ~6.5, the M state, which reflects Schiff base deprotonation during proton pumping, is not accumulated. However, using the infrared band at ~1760 cm(-1) as a marker for transient protonation of the primary acceptor, we find that Schiff base deprotonation must have occurred at pH not only above but also below the pKa ~6.5. Thus, the M state is formed but not accumulated for kinetic reasons. Further, chromophore reisomerization from the 13-cis to the all-trans conformation occurs very late in the photocycle. The strongly red-shifted states that dominate the second half of the cycle are produced before the reisomerization step, and by this criterion, they are not O-like but rather N-like states. The assignment of photocycle intermediates enables reevaluation of the photocycle; its specific features are discussed in relation to the general mechanism of proton transport in retinal proteins.

Structure changes upon deprotonation of the proton release group in the bacteriorhodopsin photocycle.

In the photocycle of bacteriorhodopsin at pH 7, a proton is ejected to the extracellular medium during the protonation of Asp-85 upon formation of the M intermediate. The group that releases the ejected proton does not become reprotonated until the prephotolysis state is restored from the N and O intermediates. In contrast, at acidic pH, this proton release group remains protonated to the end of the cycle. Time-resolved Fourier transform infrared measurements obtained at pH 5 and 7 were fitted to obtain spectra of kinetic intermediates, from which the spectra of M and N/O versus unphotolyzed state were calculated. Vibrational features that appear in both M and N/O spectra at pH 7, but not at pH 5, are attributable to deprotonation from the proton release group and resulting structural alterations. Our results agree with the earlier conclusion that this group is a protonated internal water cluster, and provide a stronger experimental basis for this assignment. A decrease in local polarity at the N-C bond of the side chain of Lys-216 resulting from deprotonation of this water cluster may be responsible for the increase in the proton affinity of Asp-85 through M and N/O, which is crucial for maintaining the directionality of proton pumping.

Molecular dynamics simulation of the unfolding of individual bacteriorhodopsin helices in sodium dodecyl sulfate micelles.

We report molecular dynamics simulations of the trends in the changes in secondary structure of the seven individual helices of bacteriorhodopsin when inserted into sodium dodecyl sulfate (SDS) micelles, and their dependence on the amino acid sequence. The results indicate that the partitioning of the helices in the micelles and their stability are dependent on the hydrophobicity of the transmembrane segments. Helices A, B, and E are stable and retain their initial secondary structure throughout the 100 ns simulation time. In contrast, helices C, D, F, and G show structural perturbations within the first 10 ns. The instabilities are localized near charged residues within the transmembrane segments. The overall structural instability of the helix is correlated with its partitioning to the surface of the micelle and its interaction with polar groups there. The in silico experiments were performed to complement the in vitro experiments that examined the partial denaturation of bacteriorhodopsin in SDS described in the preceding article (DOI 10.1021/bi201769z ). The simulations are consistent with the trends revealed by the experimental results but strongly underestimate the extent of helix to extended coil transformation. The reason may be either that the sampling time was not sufficiently long or, more interestingly, that interhelix residue interactions play a role in the unfolding of the helices.

Secondary and tertiary structure of bacteriorhodopsin in the SDS denatured state.

We characterized the structure of partially unfolded bacteriorhodopsin in sodium dodecyl sulfate (SDS) micelles and compared it with its in vitro refolded structure after reconstitution with dimyristoylphosphatidylcholine/3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (DMPC/CHAPS). Intrahelical and interhelical distances were mapped in the protein using strategically located spin-label pairs at helical ends, assayed by pulsed electron paramagnetic resonance spectroscopy (double electron-electron spin resonance, DEER). We find that in SDS the intrahelical end-to-end distances exhibit broad distributions, suggesting a heterogeneous ensemble of conformations with differing secondary structures. Nevertheless, a majority of the denatured population retains end-to-end distances similar to those in the native state. In contrast, the observed greatly increased interhelical distances, in addition to their very broad distributions, suggest that in the SDS micelles very little of the native tertiary structure remains.

Carotenoid response to retinal excitation and photoisomerization dynamics in xanthorhodopsin.

We present a comparative study of xanthorhodopsin, a proton pump with the carotenoid salinixanthin serving as an antenna, and the closely related bacteriorhodopsin. Upon excitation of retinal, xanthorhodopsin exhibits a wavy transient absorption pattern in the region between 470 and 540 nm. We interpret this signal as due to electrochromic effect of the transient electric field of excited retinal on salinixanthin. The spectral shift decreases during the retinal dynamics through the ultrafast part of the photocycle. Differences in dynamics of bacteriorhodopsin and xanthorhodopsin are discussed.

Structural changes in bacteriorhodopsin during in vitro refolding from a partially denatured state.

We report on the formation of the secondary and tertiary structure of bacteriorhodopsin during its in vitro refolding from an SDS-denatured state. We used the mobility of single spin labels in seven samples, attached at various locations to six of the seven helical segments to engineered cysteine residues, to follow coil-to-helix formation. Distance measurements obtained by spin dipolar quenching in six samples labeled at either the cytoplasmic or extracellular ends of pairs of helices revealed the time dependence of the recovery of the transmembrane helical bundle. The secondary structure in the majority of the helical segments refolds with a time constant of <100-140 ms. Recovery of the tertiary structure is achieved by sequential association of the helices and occurs in at least three distinct steps with time constants of 1), well below 1 s; 2), 3-4 s; and 3), 60-130 s (the latter depending on the helical pair). The slowest of these processes occurs in concert with recovery of the retinal chromophore.

Removal and reconstitution of the carotenoid antenna of xanthorhodopsin.

Salinixanthin, a C(40)-carotenoid acyl glycoside, serves as a light-harvesting antenna in the retinal-based proton pump xanthorhodopsin of Salinibacter ruber. In the crystallographic structure of this protein, the conjugated chain of salinixanthin is located at the protein-lipid boundary and interacts with residues of helices E and F. Its ring, with a 4-keto group, is rotated relative to the plane of the π-system of the carotenoid polyene chain and immobilized in a binding site near the ß-ionone retinal ring. We show here that the carotenoid can be removed by oxidation with ammonium persulfate, with little effect on the other chromophore, retinal. The characteristic CD bands attributed to bound salinixanthin are now absent. The kinetics of the photocycle is only slightly perturbed, showing a 1.5-fold decrease in the overall turnover rate. The carotenoid-free protein can be reconstituted with salinixanthin extracted from the cell membrane of S. ruber. Reconstitution is accompanied by restoration of the characteristic vibronic structure of the absorption spectrum of the antenna carotenoid, its chirality, and the excited-state energy transfer to the retinal. Minor modification of salinixanthin, by reducing the carbonyl C=O double bond in the ring to a C-OH, suppresses its binding to the protein and eliminates the antenna function. This indicates that the presence of the 4-keto group is critical for carotenoid binding and efficient energy transfer.

Reconstitution of gloeobacter rhodopsin with echinenone: role of the 4-keto group.

In previous work, we reconstituted salinixanthin, the C(40)-carotenoid acyl glycoside that serves as a light-harvesting antenna to the light-driven proton pump xanthorhodopsin, into a different protein, gloeobacter rhodopsin expressed in Escherichia coli, and demonstrated that it transfers energy to the retinal chromophore [Imasheva, E. S., et al. (2009) Biochemistry 48, 10948]. The key to binding of salinixanthin was the accommodation of its ring near the retinal ß-ionone ring. Here we examine two questions. Do any of the native Gloeobacter carotenoids bind to gloeobacter rhodopsin, and does the 4-keto group of the ring play a role in binding? There is no salinixanthin in Gloeobacter violaceous, but a simpler carotenoid, echinenone, also with a 4-keto group but lacking the acyl glycoside, is present in addition to ß-carotene and oscillol. We show that ß-carotene does not bind to gloeobacter rhodopsin, but its 4-keto derivative, echinenone, does and functions as a light-harvesting antenna. This indicates that the 4-keto group is critical for carotenoid binding. Further evidence of this is the fact that salinixanthol, an analogue of salinixanthin in which the 4-keto group is reduced to hydroxyl, does not bind and is not engaged in energy transfer. According to the crystal structure of xanthorhodopsin, the ring of salinixanthin in the binding site is turned out of the plane of the polyene conjugated chain. A similar conformation is expected for echinenone in the gloeobacter rhodopsin. We suggest that the 4-keto group in salinixanthin and echinenone allows for the twisted conformation of the ring around the C6-C7 bond and probably is engaged in an interaction that locks the carotenoid in the binding site.

Coordinating the structural rearrangements associated with unidirectional proton transfer in the bacteriorhodopsin photocycle induced by deprotonation of the proton-release group: a time-resolved difference FTIR spectroscopic study.

In the photocycle of bacteriorhodopsin at pH 7, proton release from the proton releasing group (PRG) to the extracellular medium occurs during formation of the M intermediate. This proton release is inhibited at acidic pH, below the pK(a) of the PRG, approximately 6 in M, and instead occurs later in the cycle as the initial state is restored from the O intermediate. Here, structural changes related to deprotonation of the PRG have been investigated by time-resolved FTIR spectroscopy at 25 degrees C. The vibrational features at 2100-1790, 1730-1685, 1661, and 1130-1045 cm(-1) have greater negative intensity in the pure M-minus-BR spectrum and even in the M-minus-BR spectrum, that is present earlier together with the L-minus-BR spectrum, at pH 7, than in the corresponding M-minus-BR spectra at pH 5 or 4. The D212N mutation abolishes the decreases in the intensities of the broad feature between 1730 and 1685 cm(-1) and the band at 1661 cm(-1). The 1730-1685 cm(-1) feature may arise from transition dipole coupling of the backbone carbonyl groups of Glu204, Phe208, Asp212, and Lys216 interacting with Tyr57 and C(15)-H of the chromophore. The 1661 cm(-1) band, which is insensitive to D(2)O substitution, may arise by interaction of the backbone carbonyl of Asp212 with C(15)-H. The 2100-1790 cm(-1) feature with a trough at 1885 cm(-1) could be due to a water cluster. Depletion of these bands upon deprotonation of the PRG is attributable to disruption of a coordinated structure, held in place by interactions of Asp212. Deprotonation of the PRG is also accompanied by disruption of the interaction of the water molecule near Arg82. The liberated Asp212 may stabilize the protonated state of Asp85 and thus confer unidirectionality to the transport.

Low-temperature FTIR study of multiple K intermediates in the photocycles of bacteriorhodopsin and xanthorhodopsin.

Low-temperature FTIR spectroscopy of bacteriorhodopsin and xanthorhodopsin was used to elucidate the number of K-like bathochromic states, their sequence, and their contributions to the photoequilibrium mixtures created by illumination at 80-180 K. We conclude that in bacteriorhodopsin the photocycle includes three distinct K-like states in the sequence bR (hv)--> I* --> J --> K(0) --> K(E) --> L --> ..., and similarly in xanthorhodopsin. K(0) is the main fraction in the mixture at 77 K that is formed from J. K(0) becomes thermally unstable above approximately 50 K in both proteins. At 77 K, both J-to-K(0) and K(0)-to-K(E) transitions occur and, contrarily to long-standing belief, cryogenic trapping at 77 K does not produce a pure K state but a mixture of the two states, K(0) and K(E), with contributions from K(E) of approximately 15 and approximately 10% in the two retinal proteins, respectively. Raising the temperature leads to increasing conversion of K(0) to K(E), and the two states coexist (without contamination from non-K-like states) in the 80-140 K range in bacteriorhodopsin, and in the 80-190 K range in xanthorhodopsin. Temperature perturbation experiments in these regions of coexistence revealed that, in spite of the observation of apparently stable mixtures of K(0) and K(E), the two states are not in thermally controlled equilibrium. The K(0)-to-K(E) transition is unidirectional, and the partial transformation to K(E) is due to distributed kinetics, which governs the photocycle dynamics at temperatures below approximately 245 K (Dioumaev and Lanyi, Biochemistry 2008, 47, 11125-11133). From spectral deconvolution, we conclude that the K(E) state, which is increasingly present at higher temperatures, is the same intermediate that is detected by time-resolved FTIR prior to its decay, on a time scale of hundreds of nanoseconds at ambient temperature (Dioumaev and Braiman, J. Phys. Chem. B 1997, 101, 1655-1662), into the K(L) state. We were unable to trap the latter separately from K(E) at low temperature, due to the slow distributed kinetics and the increasingly faster overlapping formation of the L state. Formation of the two consecutive K-like states in both proteins is accompanied by distortion of two different weakly bound water molecules: one in K(0), the other in K(E). The first, well-documented in bacteriorhodopsin at 77 K where K(0) dominates, was assigned to water 401 in bacteriorhodopsin. The other water molecule, whose participation has not been described previously, is disturbed on the next step of the photocycle, in K(E), in both proteins. In bacteriorhodopsin, the most likely candidate is water 407. However, unlike bacteriorhodopsin, the crystal structure of xanthorhodopsin lacks homologous weakly bound water molecules.

Two bathointermediates of the bacteriorhodopsin photocycle, from time-resolved nanosecond spectra in the visible.

Time-resolved measurements were performed on wild-type bacteriorhodopsin with an optical multichannel analyzer in the spectral range 350-735 nm, from 100 ns to the photocycle completion, at four temperatures in the 5-30 degrees C range. The intent was to examine the possibility of two K-like bathochromic intermediates and to obtain their spectra and kinetics in the visible. The existence of a second K-like intermediate, termed KL, had been postulated (Shichida et al., Biochim. Biophys. Acta 1983, 723, 240-246) to reconcile inconsistencies in data in the pico- and microsecond time domains. However, introduction of KL led to a controversy, since neither its visible spectrum nor its kinetics could be confirmed. Infrared data (Dioumaev and Braiman, J. Phys. Chem. B 1997, 101, 1655-1662) revealed a state which might have been considered a homologue to KL, but it had a kinetic pattern different from that of the earlier proposed KL. Here, we characterize two distinct K-like intermediates, K(E) ("early") and K(L) ("late"), by their spectra and kinetics in the visible as revealed by global kinetic analysis. The K(E)-to-K(L) transition has a time constant of approximately 250 ns at 20 degrees C, and describes a shift from K(E) with lambda(max) at approximately 600 nm and extinction of approximately 56,000 M(-1) x cm(-1) to K(L) with lambda(max) at approximately 590 nm and extinction of approximately 50,000 M(-1) x cm(-1). The temperature dependence of this transition is characterized by an enthalpy of activation of DeltaH(++) approximately 40 kJ/mol and a positive entropy of activation of DeltaS(++)/R approximately 4. The consequences of multiple K-like states for interpreting the spectral evolution in the early stages of the photocycle are discussed.

Reconstitution of Gloeobacter violaceus rhodopsin with a light-harvesting carotenoid antenna.

We show that salinixanthin, the light-harvesting carotenoid antenna of xanthorhodopsin, can be reconstituted into the retinal protein from Gloeobacter violaceus expressed in Escherichia coli. Reconstitution of gloeobacter rhodopsin with the carotenoid is accompanied by characteristic absorption changes and the appearance of CD bands similar to those observed for xanthorhodopsin that indicate immobilization and twist of the carotenoid in the binding site. As in xanthorhodopsin, the carotenoid functions as a light-harvesting antenna. The excitation spectrum for retinal fluorescence emission shows that ca. 36% of the energy absorbed by the carotenoid is transferred to the retinal. From excitation anisotropy, we calculate the angle between the two chromophores as being ca. 50 degrees , similar to that in xanthorhodopsin. The results indicate that gloeobacter rhodopsin binds salinixanthin in a manner similar to that of xanthorhodopsin and suggest that it might bind a carotenoid also in vivo. In the crystallographic structure of xanthorhodopsin, the conjugated chain of the carotenoid lies on the surface of helices E and F, and the 4-keto ring is immersed in the protein at van der Waals distance from the ionone ring of the retinal. The 4-keto ring is in the space occupied by a tryptophan in bacteriorhodopsin, which is replaced by the smaller glycine in xanthorhodopsin and gloeobacter rhodopsin. Specific binding of the carotenoid and its light-harvesting function are eliminated by a single mutation of the gloeobacter protein that replaces this glycine with a tryptophan. This indicates that the 4-keto ring is critically involved in carotenoid binding and suggests that a number of other recently identified retinal proteins, from a diverse group of organisms, could also contain carotenoid antenna since they carry the homologous glycine near the retinal.

Structural snapshots of conformational changes in a seven-helix membrane protein: lessons from bacteriorhodopsin.

Recent advances in crystallizing integral membrane proteins have led to atomic models for the structures of several seven-helix membrane proteins, including those in the G-protein-coupled receptor family. Further steps toward exploring structure-function relationships will undoubtedly involve determination of the structural changes that occur during the various stages of receptor activation and deactivation. We expect that these efforts will bear many parallels to the studies of conformational changes in bacteriorhodopsin, which still remains the best-studied seven-helix membrane protein. Here, we provide a brief review of some of the lessons learned, the challenges faced, and the controversies over the last decade with determining conformational changes in bacteriorhodopsin. Our hope is that this analysis will be instructive for similar structural studies, especially of other seven-helix membrane proteins, in the coming decade.

Structural changes in the N and N' states of the bacteriorhodopsin photocycle.

The bacteriorhodopsin transport cycle includes protonation of the retinal Schiff base by Asp96 (M-->N reaction) and reprotonation of Asp96 from the cytoplasmic surface (N-->N' reaction). We measured distance changes between pairs of spin-labeled structural elements of interest, and in general observed larger overall structural changes in the N state compared with the N' state. The distance between the C-D loop and E-F interhelical loops in A103R1/M163R1 increased approximately 6 A in the N state and approximately 3 A in the N' state. The opposite trend of distance changes in V101R1/A168R1 and L100R1/T170R1 supports counterclockwise rotation of helix F in the N but not the N' state. Small distance increases were observed in S169R1/S226R1, but little change was seen in G106R1/G155R1. Taking earlier published EPR data into account, we suggest that structural changes of the E-F loop occur first, and then helices F and G begin to move together in the late M state. These motions then reach their maximum amplitude in the N state, evidently to facilitate the release of a proton from Asp96 and the formation of a proton-conduction pathway from Asp96 to the Schiff base. The structural changes reverse their directions and decay in the N' state.