The pH dependence of the subpicosecond retinal photoisomerization process in bacteriorhodopsin: evidence for parallel photocycles.

The pH dependence of the subpicosecond decay of the retinal photoexcited state in bacteriorhodopsin (bR) is determined in the pH range 6.8-11.3. A rapid change in the decay rate of the retinal photoexcited state is observed in the pH range 9-10, the same pH range in which a rapid change in the M412 formation kinetics was observed. This observation supports the previously proposed heterogeneity model in which parallel photocycles contribute to the observed pH dependence of the M412 formation kinetics in bR.     

Coupling photoisomerization of retinal to directional transport in bacteriorhodopsin

In order to understand how isomerization of the retinal drives unidirectional transmembrane ion transport in bacteriorhodopsin, we determined the atomic structures of the BR state and M photointermediate of the E204Q mutant, to 1.7 and 1.8 A resolution, respectively. Comparison of this M, in which proton release to the extracellular surface is blocked, with the previously determined M in the D96N mutant indicates that the changes in the extracellular region are initiated by changes in the electrostatic interactions of the retinal Schiff base with Asp85 and Asp212, but those on the cytoplasmic side originate from steric conflict of the 13-methyl retinal group with Trp182 and distortion of the pi-bulge of helix G. The structural changes suggest that protonation of Asp85 initiates a cascade of atomic displacements in the extracellular region that cause release of a proton to the surface. The progressive relaxation of the strained 13-cis retinal chain with deprotonated Schiff base, in turn, initiates atomic displacements in the cytoplasmic region that cause the intercalation of a hydrogen-bonded water molecule between Thr46 and Asp96. This accounts for the lowering of the pK(a) of Asp96, which then reprotonates the Schiff base via a newly formed chain of water molecules that is extending toward the Schiff base.     

Structural changes in bacteriorhodopsin following retinal photoisomerization from the 13-cis form

Bacteriorhodopsin (BR), a light-driven proton pump in Halobacterium salinarum, accommodates two resting forms of the retinylidene chromophore, the all-trans form (AT-BR) and the 13-cis,15-syn form (13C-BR). Both isomers are present in thermal equilibrium in the dark, but only the all-trans form has proton-pump activity. In this study, we applied low-temperature Fourier-transform infrared (FTIR) spectroscopy to 13C-BR at 77 K and compared the local structure around the chromophore before and after photoisomerization with that in AT-BR. Strong hydrogen-out-of-plane (HOOP) vibrations were observed at 964 and 958 cm(-)(1) for the K state of 13C-BR (13C-BR(K)) versus a vibration at 957 cm(-)(1) for the K state of AT-BR (AT-BR(K)). In AT-BR(K), but not in 13C-BR(K), the HOOP modes exhibit isotope shifts upon deuteration of the retinylidene at C15 and at the Schiff base nitrogen. Whereas the HOOP modes of AT-BR(K) were significantly affected by the mutation of Thr89, this was not the case for the HOOP modes of 13C-BR(K). These observations imply that, while the chromophore distortion is localized near the Schiff base in AT-BR(K), it is located elsewhere in 13C-BR(K). By use of [zeta-(15)N]lysine-labeled BR, we identified the N-D stretching vibrations of the 13C-BR Schiff base (in D(2)O) at 2173 and 2056 cm(-)(1), close in frequency to those of AT-BR. These frequencies indicate strong hydrogen bonding of the Schiff base in 13C-BR, presumably with a water molecule as in AT-BR. In contrast, the N-D stretching vibration appears at 2332 and 2276 cm(-)(1) in 13C-BR(K) versus values of 2495 and 2468 cm(-)(1) for AT-BR(K), suggesting that the rupture of the Schiff base hydrogen bond that occurs in AT-BR(K) does not occur in 13C-BR(K). Rotational motion of the Schiff base upon retinal isomerization is probably smaller in magnitude for 13C-BR than for AT-BR. These differences in the primary step are possibly related to the absence of light-driven proton pumping by 13C-BR.     

Photocycle of dried acid purple form of bacteriorhodopsin.

The photocycle of dried bacteriorhodopsin, pretreated in a 0.3 M HCl solution, was studied. Some properties of this dried sample resemble that of the acid purple suspension: the retinal conformation is mostly all-trans, 15-anti form, the spectrum of the sample is blue-shifted by 5 nm to 560 nm, and it has a truncated photocycle. After photoexcitation, a K-like red-shifted intermediate appears, which decays to the ground state through several intermediates with spectra between the K and the ground state. There are no other bacteriorhodopsin-like intermediates (L, M, N, O) present in the photocycle. The K to K' transition proceeds with an enthalpy decrease, whereas during all the following steps, the entropic energy of the system decreases. The electric response signal of the oriented sample has only negative components, which relaxes to zero. These suggest that the steps after intermediate K represent a relaxation process, during which the absorbed energy is dissipated and the protein returns to its original ground state. The initial charge separation on the retinal is followed by limited charge rearrangements in the protein, and later, all these relax. The decay times of the intermediates are strongly influenced by the humidity of the sample. Double-flash experiments proved that all the intermediates are directly driven back to the ground state. The study of the dried acid purple samples could help in understanding the fast primary processes of the protein function. It may also have importance in technical applications.     

Selectivity of retinal photoisomerization in proteorhodopsin is controlled by aspartic acid 227

Similarly to bacteriorhodopsin, proteorhodopsin that normally contains all-trans and 13-cis retinal is transformed at low pH to a species containing 9-cis retinal under continuous illumination at lambda > 530 nm. This species, absorbing around 430 nm, returns thermally in tens of minutes to initial pigment and can be reconverted also with blue-light illumination. The yield of the 9-cis species is negligibly small at neutral pH but increases manyfold (>100) at acid pH with a pK(a) of 2.6. This indicates that protonation of acidic group(s) alters the photoreaction pathway that leads normally to all-trans --> 13-cis isomerization. In the D97N mutant, in which one of the two acidic groups in the vicinity of the retinal Schiff base is not ionizable, the yield of 9-cis species at low pH shows a pH dependence similar to that in the wild-type but with a somewhat increased pK(a) of 3.3. In contrast to this relatively minor effect, replacement of the other acidic group, Asp227, with Asn results in a remarkable, more than 50-fold, increase in the yield of the light-induced formation of 9-cis species in the pH range 4-6. It appears that protonation of Asp227 at low pH is what causes the dramatic increase in the yield of the 9-cis species in wild-type proteorhodopsin. We conclude that the photoisomerization pathways in proteorhodopsin to 13-cis or 9-cis photoproducts are controlled by the charge state of Asp227.     

Crystal structures of acid blue and alkaline purple forms of bacteriorhodopsin

Bacteriorhodopsin, a light-driven proton pump found in the purple membrane of Halobacterium salinarum, exhibits purple at neutral pH but its color is sensitive to pH. Here, structures are reported for an acid blue form and an alkaline purple form of wild-type bacteriorhodopsin. When the P622 crystal prepared at pH 5.2 was acidified with sulfuric acid, its color turned to blue with a pKa of 3.5 and a Hill coefficient of 2. Diffraction data at pH 2-5 indicated that the purple-to-blue transition accompanies a large structural change in the proton release channel; i.e. the extracellular half of helix C moves towards helix G, narrowing the proton release channel and expelling a water molecule from a micro-cavity in the vicinity of the retinal Schiff base. In this respect, the acid-induced structural change resembles the structural change observed upon formation of the M intermediate. But, the acid blue form contains a sulfate ion in a site(s) near Arg82 that is created by re-orientations of the carboxyl groups of Glu194 and Glu204, residues comprising the proton release complex. This result suggests that proton uptake by the proton release complex evokes the anion binding, which in turn induces protonation of Asp85, a key residue regulating the absorption spectrum of the chromophore. Interestingly, a pronounced structural change in the proton release complex was also observed at high pH; i.e. re-orientation of Glu194 towards Tyr83 was found to take place at around pH 10. This alkaline transition is suggested to be accompanied by proton release from the proton release complex and responsible for rapid formation of the M intermediate at high pH.     

Crystallographic structure of the K intermediate of bacteriorhodopsin: conservation of free energy after photoisomerization of the retinal

The K state, an early intermediate of the bacteriorhodopsin photocycle, contains the excess free energy used for light-driven proton transport. The energy gain must reside in or near the photoisomerized retinal, but in what form has long been an open question. We produced the K intermediate in bacteriorhodopsin crystals in a photostationary state at 100K, with 40% yield, and determined its X-ray diffraction structure to 1.43 A resolution. In independent refinements of data from four crystals, the changes are confined mainly to the photoisomerized retinal. The retinal is 13-cis,15-anti, as known from vibrational spectroscopy. The C13=C14 bond is rotated nearly fully to cis from the initial trans configuration, but the C14-C15 and C15=NZ bonds are partially counter-rotated. This strained geometry keeps the direction of the Schiff base N-H bond vector roughly in the extracellular direction, but the angle of its hydrogen bond with water 402, that connects it to the anionic Asp85 and Asp212, is not optimal. Weakening of this hydrogen bond may account for many of the reported features of the infrared spectrum of K, and for its photoelectric signal, as well as the deprotonation of the Schiff base later in the cycle. Importantly, although 13-cis, the retinal does not assume the expected bent shape of this configuration. Comparison of the calculated energy of the increased angle of C12-C13=C14, that allows this distortion, with the earlier reported calorimetric measurement of the enthalpy gain of the K state indicates that a significant part of the excess energy is conserved in the bond strain at C13.     

Structural changes of pharaonis phoborhodopsin upon photoisomerization of the retinal chromophore: infrared spectral comparison with bacteriorhodopsin

Archaeal rhodopsins possess a retinal molecule as their chromophores, and their light energy and light signal conversions are triggered by all-trans to 13-cis isomerization of the retinal chromophore. Relaxation through structural changes of the protein then leads to functional processes, proton pump in bacteriorhodopsin and transducer activation in sensory rhodopsins. In the present paper, low-temperature Fourier transform infrared spectroscopy is applied to phoborhodopsin from Natronobacterium pharaonis (ppR), a photoreceptor for the negative phototaxis of the bacteria, and infrared spectral changes before and after photoisomerization are compared with those of bacteriorhodopsin (BR) at 77 K. Spectral comparison of the C--C stretching vibrations of the retinal chromophore shows that chromophore conformation of the polyene chain is similar between ppR and BR. This fact implies that the unique chromophore-protein interaction in ppR, such as the blue-shifted absorption spectrum with vibrational fine structure, originates from both ends, the beta-ionone ring and the Schiff base regions. In fact, less planer ring structure and stronger hydrogen bond of the Schiff base were suggested for ppR. Similar frequency changes upon photoisomerization are observed for the C==N stretch of the retinal Schiff base and the stretch of the neighboring threonine side chain (Thr79 in ppR and Thr89 in BR), suggesting that photoisomerization in ppR is driven by the motion of the Schiff base like BR. Nevertheless, the structure of the K state after photoisomerization is different between ppR and BR. In BR, chromophore distortion is localized in the Schiff base region, as shown in its hydrogen out-of-plane vibrations. In contrast, more extended structural changes take place in ppR in view of chromophore distortion and protein structural changes. Such structure of the K intermediate of ppR is probably correlated with its high thermal stability. In fact, almost identical infrared spectra are obtained between 77 and 170 K in ppR. Unique chromophore-protein interaction and photoisomerization processes in ppR are discussed on the basis of the present infrared spectral comparison with BR.     

Location of the carboxyl terminus of bacteriorhodopsin in purple membrane.

Purple membrane samples have been prepared by trypsin digestion to have either 10 or 21 residues removed from the carboxyl terminus of the proteins. Electron diffraction of single membranes and x-ray diffraction of unoriented membrane pellets have been carried out on both these specimens and on native purple membranes. the main conclusion from this work is that the carboxyl terminus is almost entirely disordered, being free to take up many positions, and that its removal does not affect the packing in the crystal. The low resolution x-ray diffraction difference map may also suggest the approximate location of the carboxyl terminus.     

Dimeric-like kinetic cooperativity of the bacteriorhodopsin molecules in purple membranes.

The kinetics of the absorption changes accompanying the photocycle of bacteriorhodopsin (BR) strongly depend on the intensity of the exciting short laser pulse. The decrease in the flash intensity dependence of the M kinetics after different extents of bleaching of the purple membranes by hydroxylamine proves the existence of a cooperative interaction between the photocycling BR molecules. The yield of the slow component of the M decay (M(s)) is a quadratic function of the extent of the fraction cycling. The slope of the relative weight of M(s) versus the fraction cycling is 0.5. This slope indicates a dimeric-like cooperative interaction, although the structural units of the purple membranes are the trimers of the BR molecules. For the most probable cooperative mechanism an asymmetric trimeric interaction is suggested, which accounts for the apparently dimeric features. A photocycling molecule may influence only one of its two neighbors in the trimer. From this asymmetric feature a deformative interaction is expected to be the cooperative mechanism, which would be an allosteric regulating mechanism in the purple membrane.     

Structural change of threonine 89 upon photoisomerization in bacteriorhodopsin as revealed by polarized FTIR spectroscopy.

The all-trans to 13-cis photoisomerization of the retinal chromophore of bacteriorhodopsin occurs selectively, efficiently, and on an ultrafast time scale. The reaction is facilitated by the surrounding protein matrix which undergoes further structural changes during the proton-transporting reaction cycle. Low-temperature polarized Fourier transform infrared difference spectra between bacteriorhodopsin and the K intermediate provide the possibility to investigate such structural changes, by probing O-H and N-H stretching vibrations [Kandori, Kinoshita, Shichida, and Maeda (1998) J. Phys. Chem. B 102, 7899-7905]. The measurements of [3-18O]threonine-labeled bacteriorhodopsin revealed that one of the D2O-sensitive bands (2506 cm(-1) in bacteriorhodopsin and 2466 cm(-1) in the K intermediate, in D2O exhibited 18(O)-induced isotope shift. The O-H stretching vibrations of the threonine side chain correspond to 3378 cm(-1) in bacteriorhodopsin and to 3317 cm(-1) in the K intermediate, indicating that hydrogen bonding becomes stronger after the photoisomerization. The O-H stretch frequency of neat secondary alcohol is 3340-3355 cm(-1). The O-H stretch bands are preserved in the T46V, T90V, T142N, T178N, and T205V mutant proteins, but diminished in T89A and T89C, and slightly shifted in T89S. Thus, the observed O-H stretching vibration originates from Thr89. This is consistent with the atomic structure of this region, and the change of the S-H stretching vibration of the T89C mutant in the K intermediate [Kandori, Kinoshita, Shichida, Maeda, Needleman, and Lanyi (1998) J. Am. Chem. Soc. 120, 5828-5829]. We conclude that all-trans to 13-cis isomerization causes shortening of the hydrogen bond between the OH group of Thr89 and a carboxyl oxygen atom of Asp85.     

Photocycles of bacteriorhodopsin in light- and dark-adapted purple membrane studied by time-resolved absorption spectroscopy.

Nanosecond time-resolved absorption spectra have been measured throughout the photocycle of bacteriorhodopsin in both light-adapted and dark-adapted purple membrane (PM). The data from dark-adapted samples are interpretable as the superposition of two photocycles arising independently from the all-trans and 13-cis retinal isomers that coexist in the dark-adapted state. The presence of a photocycle in dark-adapted PM which is indistinguishable from that observed for light-adapted PM under the same experimental conditions is demonstrated by the observation of the same five relaxation rates associated with essentially identical changes in the photoproduct spectra. This cycle is attributed to the all-trans component. The cycle of the 13-cis component is revealed by scaling the data measured for the light-adapted sample and subtracting it from the data on the dark-adapted mixture. At times less than 1 ms, the resulting difference spectra are nearly time-independent. The peak of the difference spectrum is near 600 nm, although there appears to be a slight (approximately 2 nm) blue-shift in the first few microseconds. Subsequently the amplitude of this spectrum decays and the peak of the difference spectrum shifts in two relaxations. Most of the amplitude of the photoproduct difference spectrum (approximately 80%) decays in a single relaxation having a time constant of approximately 35 ms. The difference spectrum remaining after this relaxation peaks at approximately 590 nm and is indistinguishable from the classical light-dark difference spectrum, which we find, in experiments performed on a much longer time scale, to peak at 588 nm. The decay of this remaining photo-product is not resolvable in the nanosecond kinetic experiments, but dark adaptation of a completely light-adapted sample is found to occur exponentially with a relaxation time of approximately 2,000 s under the conditions of our experiments.     

Light activates rotations of bacteriorhodopsin in the purple membrane

To investigate how a photoactivated chromophore drives the proton pump mechanism of bacteriorhodopsin, we have observed how the chromophore rotates during the photocyle. To do this, we examined the dichroism induced in aqueous suspensions of purple membrane fragments by flashes of linearly polarized light. We find that the flash stimulates both the photocycling chromophores and their noncycling neighbors to undergo large (greater than 10 degrees - 20 degrees) rotations within the membrane during the photocycle, and that these two chromophore populations undergo distinctly different sequences of rotations. All these rotations could be eliminated by glutaraldehyde fixation as well as by embedding unfixed fragments in polyacrylamide or agarose gels. Thus, in these immbolizing preparations the chromophore can photocycle without rotating inside a bacteriorhodopsin monomer by more than our detection limit of 2 degrees - 5 degrees. The large rotations we observed in aqueous suspensions of purple membranes were probably due to rotations of entire protein monomers. The process by which a photocycling monomer causes its noncycling neighbors to rotate may help explain the highly cooperative behavior bacteriorhodopsin exhibits when it is aggregated into crystalline arrays of trimers.     

Restricted motion of photoexcited bacteriorhodopsin in purple membrane containing ethanol.

The molecular motion of retinal within the purple membrane was investigated by flash-induced absorption anisotropies with or without ethanol. In the absence of ethanol, the measured anisotropies at several wavelengths exhibited almost the same slow decay. This slow decay was attributed to only the rotation of purple membrane sheet itself in the aqueous suspension. In the presence of ethanol, however, we observed the wavelength-dependent anisotropies. The fluidity of the purple membrane, investigated with a fluorescence anisotropy method, was increased by the addition of ethanol. These facts indicated that the characteristic motion of bacteriorhodopsin is induced in perturbed purple membrane with ethanol. The data analysis was performed, taking account of the overlapping of absorption from ground-state bacteriorhodopsin and photointermediates. The results showed that the rotational motion of photointermediates within the membrane was more restricted than that of nonexcited bacteriorhodopsin. The addition of ethanol facilitated the rotation of nonexcited protein, whereas it did not significantly affect the motion of photointermediates. The restricted motion of photointermediates is probably caused by a conformational change in them, which may hinder the rotation of monomer protein and/or induce the interaction between photointermediate and neighboring proteins.     

Transmembrane location of retinal in bacteriorhodopsin by neutron diffraction

The transmembrane location of the chromophore of bacteriorhodopsin was obtained by neutron diffraction on oriented stacks of purple membranes. Two selectively deuterated retinals were synthesized and incorporated in bacteriorhodopsin by using the retinal- mutant JW5: retinal-d11 (D11) contained 11 deuterons in the cyclohexene ring, and retinal-d5 (D5) had 5 deuterons as close as possible to the Schiff base end of the chromophore. The membrane stacks had a lamellar spacing of 53.1 A at 86% relative humidity. Five orders were observed in the lamellar diffraction pattern of the D11, D5, and nondeuterated reference samples. The reflections were phased by D2O-H2O exchange. The absolute values of the structure factors were nonlinear functions of the D2O content, suggesting that the coherently scattering domains consisted of asymmetric membrane stacks. The centers of deuteration were determined from the observed intensity differences between labeled and unlabeled samples by using model calculations and Fourier difference methods. With the origin of the coordinate system defined midway between consecutive intermembrane water layers, the coordinates of the center of deuteration of the D11 and D5 label are 10.5 +/- 1.2 and 3.8 +/- 1.5 A, respectively. Alternatively, the label distance may be measured from the nearest membrane surface as defined by the maximum in the neutron scattering length density at the water/membrane interface. With respect to this point, the D11 and D5 labels are located at a depth of 9.9 +/- 1.2 and 16.6 +/- 1.5 A, respectively. The chromophore is tilted with the Schiff base near the middle of the membrane and the ring closer to the membrane surface. The vector connecting the two label positions in the chromophore makes an angle of 40 +/- 12 degrees with the plane of the membrane. Of the two possible orientations of the plane of the chromophore, which is perpendicular to the membrane plane, only the one in which the N----H bond of the Schiff base points toward the same membrane surface as the vector from the Schiff base to the cyclohexene ring is compatible with the known tilt angle of the polyene chain.     

-Amino acid contents of mitochondria and some purple bacteria

The endosymbiont most likely to have given rise to mitochondria is an aerobic bacterium belonging to the α subdivision of the so-called purple bacteria such as Rickettsia, Bradythizobium and Agrobacterium [1 and 2]. Contents of the -enantiomers of serine, alanine, proline, glutamate and aspartate in rat liver whole mitochondria, mitochondrial outer membranes, inner membranes and matrix, soluble proteins and free amino acids were detected. These values for -amino acid content were compared with those in soluble proteins and free amino acids from the purple bacteria Paracoccus denitrificans, Pseudomonas aeruginosa and Escherichia coli, members, respectively of the α, β, and γ subdivisions, to find any similarity between mitochondria and these purple bacteria. A similarity was observed in protein -amino acid contents which were low (<1.5%, D-type/D-type+L-type) both in the membrane and soluble protein fractions from mitochondria and in soluble protein from bacteria. Oddly, substantial amounts of free -serine and free -aspartate (around 2%) were found for the first time in mitochondria. The contents of -serine and -aspartate were higher than those of -alanine, -proline and -glutamate. In purple bacteria, the concentration of -serine (<2%) was the lowest of the five amino acids examined, and those of -alanine (27–32%) and -glutamate (7–26%) were high. Therefore, no similarity was shown in the free -amino acid content between mitochondria and any of the three purple bacteria.     

The location of retinal in the purple membrane profile by neutron diffraction.

The trans-membrane location of retinal in the purple membrane of Halobacterium halobium, has been determined by low-angle neutron scattering studies on aqueous dispersions of the membranes. The membrane was bleached and regenerated with deuterated and with hydrogen-containing retinal. The modified retinal was obtained by extraction from bacteria grown in a totally deuterated medium. The determination of the retinal position is based on the differences in neutron scattering between a purple membrane sample with normal, protonated retinal and another sample with deuterated retinal. A distinct scattering density difference between the two preparations was observed. A direct structure determination was used with the retinal localized from a Fourier difference density profile. We conclude that the β-ionone ring portion of the retinal is situated centrally in the membrane.     

Proteolysis and flash photolysis of bacteriorhodopsin in purple membrane fragments

Pronase treatment of aqueous suspensions of purple membrane fragments from H. halobium leads to the cleavage of bacteriorhodopsin. The protein fragments remaining in the membrane after treatment with relatively small concentrations of enzyme (2% w/w) in normal daylight range in molecular weight from 20,000-21,000 daltons, indicating that cleavage occurs mainly near the extremities of the protein chain. At higher enzyme concentrations the relative amounts of protein fragments having smaller molecular weight increase. Generally, the relative loss of retinal chromophore is larger than that of protein and thus the retinal binding site seems to be located near one of the chain ends that is cleaved off by enzyme.Irradiation with white light during the time of proteolysis (at both low and high enzyme concentrations) results in extensive cleavage, so that under certain conditions no high molecular weight components can be detected in SDS-polyacrylamide gels. It, therefore, appears that parts of the bacteriorhodopsin chain become more exposed to enzyme digestion when the purple membrane is illuminated.Enzyme treated aqueous purple membrane fragment suspensions still show photocycle activity. The main consequence of proteolysis is a pronounced appearance of biphasicity in the decay of M₄₁₂ and the regeneration of bR₅₇₀. Simultaneously the yield of O₆₆₀ is reduced. As with untreated purple membrane, the correlation between the rates of decay of M₄₁₂ and regeneration of bR₅₇₀ is greatest when the yield of O₆₆₀ is lowest.     

Replacement effects of neutral amino acid residues of different molecular volumes in the retinal binding cavity of bacteriorhodopsin on the dynamics of its primary process.

We have determined the rate and quantum yield of retinal photoisomerization, the spectra of the primary transients, and the energy stored in the K intermediate in the photocycle of some bacteriorhodopsin mutants (V49A, A53G, and W182F) in which residue replacements are found to change the Schiff base deprotonation kinetics (and thus the protein-retinal interaction). Because of their change in the local volume resulting from these individual replacements, these substitutions perturb the proton donor-acceptor relative orientation change and thus the Schiff base deprotonation kinetics. These replacements are thus expected to change the charge distribution around the retinal, which controls its photoisomerization dynamics. Subpicosecond transient spectroscopy as well as photoacoustic technique are used to determine the retinal photoisomerization rate, quantum yield, and the energy stored in the K-intermediate for these mutants. The results are compared with those obtained for wild-type bacteriorhodopsin and other mutants in which charged residues in the cavity are replaced by neutral ones. In some of the mutants the rate of photoisomerization is changed, but in none is the quantum yield or the energy stored in the K intermediate altered from that in the wild type. These results are discussed in terms of the shapes of the potential energy surfaces of the excited and ground states of retinal in the perpendicular configuration within the protein and the stabilization of the positive charge in the ground and the excited state of the electronic system of retinal.