The isomeric composition of retinal was measured in a number of bacteriorhodopsin (bR) mutants (D85N, D212N, R82A, Y185F, and D115N) under various conditions, using a rapid retinal extraction technique followed by HPLC analysis. Besides the 13-cis and the all-trans retinal isomers observed in wild type (wt) bR under physiological conditions, the 1 1-cis and 9-cis retinal isomers were observed in variable but minor amounts in the bR mutants. In addition, the values of the equilibrium constant at two temperatures and the enthalpy change for the all-trans to 13-cis isomerization process in the dark-adapted state of D212N, DMN, deionized blue bR, and wt bR were determined. We find that perturbation of the retinal cavity (pocket) by residue replacement changes the relative thermal stability of the different retinal isomers, allowing for thermaland/or photoisomenzation of the retinal chromophore along C9-Cl0 and C1 1 -C12 bonds to moderately compete with the isomerization around the C I~-C I~ bond. The bR mutants expressed in Halobacterium salinarium studied in the present work showed normal 13-cis to all-trans light adaptation, in contrast with abnormal all-trans to 13-cis light adaptation observed for D212E, D212A, and D212N expressed in Escherichia coli, suggesting an influence of the purple membrane lattice and/or the lipids on the stability of the different retinal isomers within the protein.

The first section of this paper is a detailed summary of studies made by us and others on metal cations binding to deionized bacteriorhodopsin (dIbR) and its variants. Our studies include the luminescence experiments of Eu 3 + binding to dIbR and potentiometric studies of Ca" binding to dIbR, to deionized bR mutants, to bacterioopsin, and to dIbR with its C-terminus removed. The results suggest the presence of two classes of binding sites, one class has two high-affinity constants, and one has one low-affinity constant. For Ca" binding, there is one metal cation in each of the two high-affinity sites which are coupled to the charged aspartates 85 and 212 (known to be in the retinal cavity) but not coupled to each other. The lowaffinity class can accommodate 0-6 Ca" ions and most of them are bound to the surface. Mg" has a slightly smaller value for its binding constant to the highestaffinity site. Thus, one expects more Ca" than Mg" bound to the two high-affinity sites. In the second section, we summarize our recent study on the effect of metal cation charge density (Ca'", Mg", Eu 3 +, Tb 3 +, H03+, Dy3+) on the kinetics of both Schiff base deprotonation and proton transport to the extracellular surface. For all metal cations, the apparent rate constant of the slow components of the deprotonation process is the same as that for the transport process at 22°C. The temperature studies, however, show this apparent equality to be fortuitous and to result from cancellation of the contribution of the energy and entropy of activation. Thus, while the entropy of activation is positive for the deprotonation process, it is negative for the proton transport process. These kinetic parameters depend weakly on the charge density, but in an opposite sense for the two processes. These results suggest that the deprotonation is not the rate-limiting step for the proton transport process. A possible mechanism is proposed in which a hydrated metal cation is used to induce the deprotonation of the protonated Schiff base and to dissociate one of its HzO molecules to donate the proton in the L~M process.b~-:>(bR)*~I460~J625~~IO~L550~M4l2~N~b- H+ +H+In the L sso~M412 step, a proton is transferred from

In our continuing effort to characterize the metal cation binding in bacteriorhodopsin (bR) using Ca(2+)-specific electrodes, potentiometric titration was carried out on deionized solubilized bR (containing monomeric units) and deionized bacterioopsin (bR with its retinal removed). Scatchard plots were analyzed. The monomer was found to have plots similar to those of the trimer, suggesting that the binding sites in bR are localized within the protein monomer unit and not between the molecules within the trimer structure. This also supports the previous assumption that the curvature in the Scatchard plot of regenerated bR is not due to cooperativity of metal cation within the trimer, but rather due to multiple sites. Recent studies further support the finding that the curved Scatchard plot is not due to the cooperativity between the metal ions in the two high affinity sites, wherever they are. The results of the analysis of the Scatchard plot for deionized bacterioopsin have shown a change in the binding characteristics of the high affinity but not the low affinity sites from that observed in bR. This result supports previous conclusions that metal cations in the high affinity sites are not far from the retinal cavity.

Intravenous injection of a bolus of low molecular weight fluorescent tracer was used to estimate the time delay between changing glucose levels in blood and ISF. The results indicate that the ISF sampling technology utilized here is capable of tracking rapidly rising levels of blood glucose.

The number and identity of the metal cations bound to wild-type bacteriorhodopsin (bR) are determined by using inductively coupled plasma mass spectrometry (ICP-MS) and ICP emission techniques. The results indicate that there at ≈2 total Ca 2+ and Mg 2+ per bR molecule with a ratio of ≈3:1 Ca 2+ to Mg 2+ . This observed ratio is found to agree with the calculated ratio using previously determined binding constants for the two high affinity sites of Ca 2+ to deionized bR et al. Biophys. J. 1992 et al. Biophys. J. , 61, 1201. This suggests that the high-affinity binding sites in deionized bR are similar to those in native bR. Structural perturbation of the native membrane by cleavage of the C-terminus decreases the number of ions per bR to 1.4. The observed ratio of total ions in this sample to total ions in bR is found to agree with that calculated using known binding constants for each. The results on the number of metal cations/bR and their ratio in bacterioopsin agrees with the calculated number using previously observed binding constants in deionized bO (Yang; et al. Biophys J., in press) only if one assumes that the second high-affinity site (not the first) is removed by retinal removal. Removal of 75% of the lipids from the purple membrane is found to greatly reduce the number of metal cations from 2 to 0.16. This suggest that if metal cations are in the two high-affinity sites (which are the only type of binding sites evident in our native bR sample), the removal of lipids, known to change the protein tertiary structure, changes also the metal ion binding sites.X