Deprotonation of lipid-depleted bacteriorhodopsin.
30
Citation
30
Reference
10
Related Paper
Citation Trend
Abstract:
The removal of 75% of the lipid from bacteriorhodopsin caused the following: (i) decreased efficiency and rate of deprotonation of the protonated Schiff base (as monitored by absorption of the M412 intermediate); (ii) increased efficiency of deprotonation of deionized samples; (iii) a decrease by 1 unit in the pH at which deprotonation ceases; (iv) increased intensity of Eu3+ emission in Eu3+-regenerated deionized delipidated samples; (v) increased exposure of the Eu3+ sites to water; and (vi) elimination of the dependence of the deprotonation efficiency on the metal cation concentration. These results are discussed in terms of changes in the protein conformation upon delipidation, which in turn control the deprotonation mechanism.Keywords:
Halobacteriaceae
The removal of 75% of the lipid from bacteriorhodopsin caused the following: (i) decreased efficiency and rate of deprotonation of the protonated Schiff base (as monitored by absorption of the M412 intermediate); (ii) increased efficiency of deprotonation of deionized samples; (iii) a decrease by 1 unit in the pH at which deprotonation ceases; (iv) increased intensity of Eu3+ emission in Eu3+-regenerated deionized delipidated samples; (v) increased exposure of the Eu3+ sites to water; and (vi) elimination of the dependence of the deprotonation efficiency on the metal cation concentration. These results are discussed in terms of changes in the protein conformation upon delipidation, which in turn control the deprotonation mechanism.
Halobacteriaceae
Cite
Citations (30)
Infrared spectroscopy is used to characterize the transitions in the photocycle of bR involving the M intermediate. It has been shown previously that in this part of the photocycle a large protein conformational change takes place that is important for proton pumping. In this work we separate the spectra of the L, M, and N intermediates in order to better describe the timing of the molecular changes. We use the photoreaction of the M intermediate to separate its spectrum from those of L and N. At temperatures between 220 and 270 K a mixture of M and L or N is produced by illumination with green light. Subsequent blue illumination selectively drives M back into the ground state and the difference between the spectra before and after blue excitation yields the spectrum of M. Below about 250 K and L/M mixture is separated; at higher temperatures an M/N mixture is seen. We find that the spectrum of M is identical in the two temperature regions. The large protein conformational change is seen to occur during the M to N transition. Our results confirm that Asp-96 is transiently deprotonated in the L state. The only aspartic protonation changes between M and bR are the protonation of Asp-85 and Asp-212 that occur simultaneously during the L to M transition. Blue-light excitation of M results in deprotonation of both. The results suggest a quadrupolelike interaction of the Schiff base, Asp-85, Asp-212, and an additional positive charge in bR.
Halobacteriaceae
Cite
Citations (43)
During the L-->M reaction of the bacteriorhodopsin photocycle the proton of the retinal Schiff base is transferred to the anionic D85. This step, together with the subsequent reprotonation of the Schiff base from D96 in the M-->N reaction, results in the translocation of a proton across the membrane. The first of these critical proton transfers occurs in an extended hydrogen-bonded complex containing two negatively charged residues (D85 and D212), two positively charged groups (the Schiff base and R82), and coordinated water. We simplified this region by replacing D212 and R82 with neutral residues, leaving only the proton donor and acceptor as charged groups. The D212N/R82Q mutant shows essentially normal proton transport, but in the photocycle neither of this protein nor of the D212N/R82Q/D96N triple mutant does a deprotonated Schiff base (the M intermediate) accumulate. Instead, the photocycle contains only the K, L, and N intermediates. Infrared difference spectra of D212N/R82Q and D212N/R82Q/D96N demonstrate that although D96 becomes deprotonated in N, D85 remains unprotonated. On the other hand, M is produced at pH > 8, where according to independent evidence the L<==>M equilibrium should shift toward M. Likewise, M is restored in the photocycle when the retinal is replaced with the 14-fluoro analogue that lowers the pKa of the protonated Schiff base, and now D85 becomes protonated as in the wild type.(ABSTRACT TRUNCATED AT 250 WORDS)
Halobacteriaceae
Cite
Citations (34)
In the bacteriorhodopsin‐containing proteoliposomes, a laser flash is found to induce formation of a bathointermediate decaying in several seconds, the difference spectrum being similar to the purple–blue transition. Different pH buffers do not affect the intermediate, whereas an uncoupler, gramicidin A, and lipophilic ions accelerate decay of the intermediate or inhibit its formation. In the liposomes containing E204Q bacteriorhodopsin mutant, formation of the intermediate is suppressed. In the wild‐type bacteriorhodopsin liposomes, the bathointermediate formation is pH‐independent within the pH 5–7 range. The efficiency of the long‐lived O intermediate formation increases at a low pH. In the wild‐type as well as in the E204Q mutant purple membrane, the O intermediate decay is slowed down at slightly higher pH values than that of the purple–blue transition. It is suggested that the membrane potential affects the equilibrium between the bacteriorhodopsin ground state (Glu‐204 is protonated and Asp‐85 is deprotonated) and the O intermediate (Asp‐85 is protonated and Glu‐204 is deprotonated), stabilizing the latter by changing the relative affinity of Asp‐85 and Glu‐204 to H + . At a low pH, protonation of a proton‐releasing group (possibly Glu‐194) in the bacteriorhodopsin ground state seems to prevent deprotonation of the Glu‐204 during the photocycle. Thus, all protonatable residues of the outward proton pathway should be protonated in the O intermediate. Under such conditions, membrane potential stabilization of the O intermediate in the liposomes can be attributed to the direct effect of the potential on the pK value of Asp‐85.
Halobacteriaceae
Proton Transport
Cite
Citations (4)
On the molecular mechanisms of the Schiff base deprotonation during the bacteriorhodopsin photocycle
Using optical flash photolysis and time-resolved Raman methods, we examined intermediates formed during the photocycle of bacteriorhodopsin (bR), as well as the bR color change, as a function of pH (in the 7.0-1.5 region) and as a function of the number of bound Ca(2+) ions. It is found that at a pH just below 3 or with less than two bound Ca(2+) per bR, the deprotonation (the L(550) --> M(412)) step ceases, yet the K(610) and L(550) analogues are still formed as in native bR. The lack of deprotonation in the photocycle of both acid blue and deionized blue bR and the similarity of their Raman spectra as well as of their K(610) and L(550) analogues strongly suggest that both blue samples have nearly the same retinal active site. It is suggested that in both blue species, bound cations are removed via a proton-cation exchange equilibrium, either on the cation exchange column for the deionized sample or in solution for the acid blue sample. The proton-cation exchange equilibrium is found to quantitatively account for the pH dependence of the purple-to-blue color change. The different mechanisms responsible for the large reduction ( approximately 11 units) of the pK(a) value of the protonated Schiff base (PSB) during the photocycle are discussed. The absence of the L(550) --> M(412) deprotonation process in both blue species is discussed in terms of the previously proposed cation model for the deprotonation of the PSB during the photocycle of native bR. The extent of the deprotonation and the blue-to-purple color change are found to follow the same dependence on either the pH or the amount of cations added to deionized blue bR. This observed correlation is briefly discussed.
Halobacteriaceae
Flash photolysis
Cite
Citations (46)
The removal of metal cations inhibits the deprotonation process of the protonated Schiff base during the photocycle of bacteriorhodopsin. To understand the nature of the involvement of these cations, a spectroscopic and kinetic study was carried out on bacteriorhodopsin samples in which the native Ca 2+ and Mg 2+ were replaced by Eu 3+ , a luminescent cation. The decay of Eu 3+ emission in bacteriorhodopsin can be fitted to a minimum of three decay components, which are assigned to Eu 3+ emission from three different sites. This is supported by the response of the decay components to the presence of 2 H 2 O and to the changes in the Eu 3+ /bR molar ratio. The number of water molecules coordinated to Eu 3+ in each site is determined from the change in its emission lifetime when 2 H 2 O replaces H 2 O. Most of the emission originates from two “wet” sites of low crystal-field symmetry—e.g., surface sites. Protonated Schiff base deprotonation has no discernable effect on the emission decay of protein-bound Eu 3+ , suggesting an indirect involvement of metal cations in the deprotonation process. Adding Eu 3+ to deionized bacteriorhodopsin increases the emission intensity of each Eu 3+ site linearly, but the extent of the deprotonation (and color) changes sigmoidally. This suggests that if only the emitting Eu 3+ ions cause the deprotonation and bacteriorhodopsin color change, ions in more than one site must be involved—e.g., by inducing protein conformation changes. The latter could allow deprotonation by the interaction between the protonated Schiff base and a positive field of cations either on the surface or within the protein.
Halobacteriaceae
Cite
Citations (40)
ADVERTISEMENT RETURN TO ISSUEPREVArticleNEXTProtonation and deprotonation of the M, N, and O intermediates during the bacteriorhodopsin photocycleGyorgy Varo and Janos K. LanyiCite this: Biochemistry 1990, 29, 29, 6858–6865Publication Date (Print):July 24, 1990Publication History Published online1 May 2002Published inissue 24 July 1990https://pubs.acs.org/doi/10.1021/bi00481a015https://doi.org/10.1021/bi00481a015research-articleACS PublicationsRequest reuse permissionsArticle Views136Altmetric-Citations37LEARN ABOUT THESE METRICSArticle Views are the COUNTER-compliant sum of full text article downloads since November 2008 (both PDF and HTML) across all institutions and individuals. These metrics are regularly updated to reflect usage leading up to the last few days.Citations are the number of other articles citing this article, calculated by Crossref and updated daily. Find more information about Crossref citation counts.The Altmetric Attention Score is a quantitative measure of the attention that a research article has received online. Clicking on the donut icon will load a page at altmetric.com with additional details about the score and the social media presence for the given article. Find more information on the Altmetric Attention Score and how the score is calculated. Share Add toView InAdd Full Text with ReferenceAdd Description ExportRISCitationCitation and abstractCitation and referencesMore Options Share onFacebookTwitterWechatLinked InRedditEmail Other access optionsGet e-Alertsclose Get e-Alerts
Halobacteriaceae
Microsecond
Proton Transport
Stopped flow
Cite
Citations (62)
Upon light adaptation by continuous (or pulsed) illumination, the artificial bacteriorhodopsin (bR) pigments, I and II, derived from synthetic 14F retinal and a short polyenal, respectively produce a long-lived red-shifted species denoted Ol. An analogous phenomenon was observed by Sonar, S., et al. [(1993) Biochemistry 32, 2263-2271], in the case of the Y185F mutant (pigment III). The nature of these Ol species was investigated by studying a series of effects, primarily their red light photoreversibility, the associated proton uptake and release processes, and the effects of pH on their relative amounts, which are interpreted in terms of pH-dependent acid−base equilibria. Experiments were also carried out with pigments I and II derived from the mutants D96A, E204Q, R82Q, and D85N. The Ol species of pigments I and II (and possibly also that of pigment III) are identified as an unusually long-lived (all-trans) intermediate of the photocycle of their 13-cis isomer. It is concluded that in Ol, Asp-85 is protonated, a process associated with proton uptake from the extracellular side. Subsequent proton release (to the same side of the membrane) occurs from Glu-204 (or from a group closely interacting with it) prior to the decay of Ol. At high pH (>9), Ol reversibly converts to a purple form, due to deprotonation of Asp-85, while at still higher pH (>11), a blue-shifted species characterized by a deprotonated Schiff base is generated. These transitions constitute the first demonstration of the titration of a photocycle intermediate of a retinal protein. The respective pKa values are determined and discussed in relation to those pertaining to the unphotolyzed (dark-adapted) pigments. It appears that the pKa values are controlled by a hydrogen bond network involving water molecules, which binds the protonated Schiff base with Asp-85 and Glu-204. The disruption of this network in pigments I-III may also be responsible for the long lifetime of the Ol species, due to the inhibition of thermal trans−13-cis isomerization. The results are relevant to the molecular mechanism of the photocycles of both 13-cis- and all-trans-bR, primarily to the nature and to the deprotonation mechanism of the proton-releasing group.
Halobacteriaceae
Proton Transport
Cite
Citations (8)
Chromophore
Halobacteriaceae
Proton Transport
Cite
Citations (225)