Andrey D. Kaulen

Time resolution of the intermediate steps in the bacteriorhodopsin-linked electrogenesis

Bacteriorhodopsin, the retinal-containing protein from Halobacterium halobium, was discovered by Oesterhelt and Stoeckenius in 197 1 [l] . The study of this compound revealed that it functions as an electrogenic light-dependent proton pump [2-41. Electric potential and current generation mediated by bacteriorhodopsin was directly measured in this group [5-91. Another line of investigation demonstrated several intermediate spectral forms of bacteriorhodopsin, participating in light-induced H’ translocation [IO-131 . The objective of this work was to obtain the time resolution of the early events in bacteriorhodopsinmediated electrogenesis. Electric potential generation accompanying a single turnover of bacteriorhodopsin was estimated by electrodes immersed into solutions on both sides of a phospholipid-impregnated collodion film with bacteriorhodopsin proteoliposomes or membrane fragments attached to one of its surfaces. Potential generation (plus on the bacteriorhodopsin-free side of the film) was found to be composed of two phases, one correlating with the 570 nm + 412 nm spectral transition (r = 25-30 ~.ts) and the other with the reversal to the 570 nm state (7 = 6-l 2 ms). The latter phase was specifically inhibited by La3’. An early photopotential of the opposite direction was also revealed (7 < 0.3 ps). Both direct phases were decelerated by D,O and abolished at pH < 2, while the amplitude of the opposite phase increased. The amplitudes of the direct phases decreased when bacteriorhodopsin was kept in the dark (r = lo-20 min).

Correlation of photochemical cycle, H+ release and uptake, and electric events in bacteriorhodopsin

The kinetics of 3 photoinduced responses of bacteriorhodopsin have been compared: spectral changes, pH shifts in a suspension of open purple membrane sheets and electric potential generation by the sheets incorporated into a lipid‐impregnated collodion film. In the presence of a pH‐buffer, the H+ release by bacteriorhodops in was shown to correlate with the formation of the M412 intermediate and the microsecond phase of the potential generation. The H+/M412 ratio is equal to 0.7±0.1 if the ionic strength of the solution is high. In the absence of the buffer, the H+ release proved to be much slower than spectral and electric responses. The kinetics of H+ uptake by bacteriorhodopsin is close to M412 decay and to the electrogeneous millisecond phase in both the presence and absence of the pH buffer. The bacteriorhodopsin‐induced proton release phase accounts for about 20%, and the uptake phase for about 80% of the overall potential. This is compatible with the model assuming that the proton start‐out point ‐ possibly, the protonated Schiff base connecting lysine 216 with retinal ‐ is closer to the outer rather than the inner (cytoplasmic) surface of the bacterial membrane.

Electrogenesis by bacteriorhodopsin incorporated in a planar phospholipid membrane

In 1966, Mitchell [l] put forward the idea of electrogenesis in coupling membrane systems. According to this concept, there are molecular electric generators incorporated in membranes of mitochondria, chloroplasts, photosynthetic and respiring bacteria. These generators were postulated to represent enzyme complexes charging the membranes by electron or proton transport across hydrophobic barriers. According to Mitchell, the transmembrane electric potential and pH differences generated by enzymes of respiratory or photosynthetic redox chains are utilized by a reversible ATPase to form ATP. Several years ago, Tupper and Tedeschi [2] tried to detect membrane potentials in mitochondria, using microelectrodes, but failed. The size of the mitochondrion is apparently too small to allow an electrode to be introduced into the matrix space without a sharp decrease in the very high electric resistance of the mitochondrial membrane [3]. Some light-dependent electric responses of a complex character were demonstrated by Bulychev et al. [4] with microelectrodes in isolated chloroplast. The electric phenomena in coupling membranes were also studied by other methods. To measure the membrane potential in mitochondria, Mitchell and Moyle determined K+ distribution across the membrane of energized mitochondria whose K+ permeability was increased by valinomycin [5] . Our group, has described the antiport of synthetic penetrating anions and cations across the membranes of energized mitochondria [6], submitochondrial particles [7,8], subchloroplast particles [9] , chromatophores of photo-

Generation of electric current by chromatophores of Rhodospirillum rubrum and reconstitution of electrogenic function in subchromatophore pigment-protein complexes

Lipoprotein complexes, containing (1) bacteriochlorophyll reaction centers, (2) bacteriochlorophyll light-harvesting antenna or (3) both reaction centers and antenna, have been isolated from chromatophores of non-sulphur purple bacteria Rhodospirillum rubrum by detergent treatments. The method of reconstituting the proteoliposomes containing these complexes is described. Being associtated with planas azolectin membrane, ptoteoliposomes as well as intact chromatophores were found to generate a light-dependent transmembrane electric potential difference measured by Ag/AgC1 electrodes and voltmeter. The direction of the electric field inproteoliposomes can be regulated by the addition of antenna complexes to the reconstitution mixture. The reaction center complex proteoliposomes generate an electric field of a direction opposite to that in chromatophores, whereas proteoliposomes containing reaction center complexes and a sufficient amount of antenna complexes produce a potential difference as in chromatophores. ATP and inorganic pyrophosphate, besides light, were shown to be usable as energy sources for electric generation in chromatophores associated with planar membrane.

Lipid-impregnated filters as a tool for studying the electric current-generating proteins.

Abstract Porous filters and collodion film impregnated with decane solution of phospholipids, were used for measurements of electric potential differences generated by bacteriorhodopsin, chromatophore redox chain, H+-ATPase, pyrophosphatase, and mitochondrial respiratory chain. It was shown that reconstituted proteoliposomes, containing e.g., bacteriorhodopsin or natural coupling membrane vesicles, such as Rhodospirillum rubrum chromatophores, can be attached to a filter surface by means of Ca2+ or Mg2+ ions. Addition of the respective energy source was found to result in electric potential difference being generated across the filter. This effect was measured directly by Ag AgCl electrodes immersed into electrolyte solutions on both sides of the filter. Using a phospholipid-impregnated collodion film one can measure electric responses as fast as 300 nsec. The phospholipid-impregnated filters turned out to be sensitive and reliable electrodes for measuring the concentration of synthetic penetrating ions, such as phenyldicarbaundecaborane, tetraphenylborate, tetrapentylammonium, and tetraphenylphosphonium. By measuring changes in the concentration of these ions in the suspension of proteoliposomes, chromatophores, mitochondria, or bacterial cells, one can follow the formation and dissipation of transmembrane potential differences in these systems. It is shown that the phospholipid-impregnated filters are much more reliable and handy than planar phospholipid membranes previously used for studying electrogenic activity of electric current-producing membrane proteins.

Protonation of a novel intermediate P is involved in the M → bR step of the bacteriorhodopsin photocycle

A novel intermediate (P) of the bacteriorhodopsin (bR) photocycle, appearing between M412 and bR is described. Like bR, intermediate P shows an absorption maximum at 560–570 nm. However, the extinction coefficient of P is somewhat lower than that of bR. Moreover, there are some differences in spectra of bR and P at wavelengths shorter than 450 nm. The P → bR transition correlates with the absorption of H+ from the water medium. The following conditions proved to be favourable for the detection of the new intermediate: a high salt concentration, low light intensity and low temperature (0.5°C). The P → bR transition is strongly decelerated by a small amount of Triton X‐100. Illumination of P does not produce M412 before bR is formed. It is assumed that M412 converts to P when the Schiff base is protonated by a proton transferred from a protein protolytic group which participates in the inward H+‐conductivity pathway. Reprotonation of this group results in the conversion of P to bR. No more than 1 H+ is transported per bR photocycle.

Ion permeability induced in artificial membranes by the ATP/ADP antiporter

The hypothesis on the additional function of the ATP/ADP antiporter (ANT) as uncoupling protein has been tested in proteoliposomes and planar bilayer phospholipid membranes (BLM). It is found that dissipation of the light‐induced ΔpH in the dark is very much faster in ANT‐bacteriorhodopsin proteoliposomes than in proteoliposomes containing baeteriorhodopsin as the only protein. Mersalyl treatment of ANT‐bacteriorhodopsin proteoliposomes causes further increase in the ΔpH dissipation rate due to formation of a high conductance pore. The properties of this pore are studied on ANT incorporated to BLM. They proved to be similar to those of so‐called multiple conductance channel or permeability transition pore of inner mitochondrial membrane. The conductance of the single channel is as high as 2.2 nS. The channel fails to discriminate between K+, Na+, H+ and Cl−. Thus the obtained data are consistent with the assumption that native and modified ANT might function as an H+‐specific conductor and as a permeability transition pore, respectively.

Rapid kinetics of membrane potential generation by cytochrome c oxidase with the photoactive Ru(II)‐tris‐bipyridyl derivative of cytochrome c as electron donor

Yeast iso‐1‐cytochrome c covalently modified at cysteine‐102 with (4‐bromomethyl‐4′‐methylbipyridine)[bis(bi‐pyridine)]Ru2+ (Ru‐102‐Cyt c) has been used as a photoactive electron donor to mitochondrial cytochrome c oxidase (COX) reconstituted into phospholipid vesicles. Rapid kinetics of membrane potential generation by the enzyme following flash‐induced photoreduction of Ru‐102‐Cyt c heme has been measured and compared to photovoltaic responses observed with Ru(II)(bipy‐ridyl)3 (RuBpy) as the photoreductant [D.L. Zaslavsky et al. (1993) FEBS Lett. 336, 389–393]. At low ionic strength, when Ru‐102‐Cyt c forms a tight electrostatic complex with COX, flash‐activation results in a polyphasic electrogenic response corresponding to transfer of a negative charge to the interior of the vesicles. The initial rapid phase is virtually identical to the 50 μs transient observed in the presence of RuBpy as the photoactive electron donor which originates from electrogenic reduction of heme a by CuA. CuA reduction by Ru‐102‐Cyt c turns out to be not electrogenic in agreement with the peripheral location of visible copper in the enzyme. A millisecond phase (τ ca. 4 ms) following the 50 μs initial part of the response and associated with vectorial translocation of protons linked to oxygen intermediate interconversion in the binuclear centre, can be resolved both with RuBpy and Ru‐102‐Cyt c as electron donors; however, this phase is small in the absence of added H2O2. In addition to these two transients, the flash‐induced electrogenic response in the presence of Ru‐102‐Cyt c reveals a large slow phase of Δψ generation not observed with RuBpy. This phase is completely quenched upon inclusion of 100 μM ferricyanide in the medium and originates from a second order reaction of COX with the excess Ru‐102‐Cyt c 2+ generated by the flash in a solution.

Flash-induced voltage changes in halorhodopsin from Natronobacterium pharaonis

The flash‐induced voltage response of halorhodopsin at high NaCl concentration comprises two main kinetic components. The first component with τ≈1 μs does not exceed 4% of the overall response amplitude and is probably associated with the formation of the L (hR520) intermediate. The second main component with τ≈1–2.5 ms which is independent of Cl− concentration can be ascribed to the transmembrane Cl− translocation during the L intermediate decay. The photoelectric response in the absence of Cl− has the opposite polarity and does not exceed 6% of the overall response amplitude at high NaCl concentration. A pH decrease results in substitution of the Cl−‐dependent components by the photoresponse which is similar to that in the absence of Cl−. Thus, the difference between photoresponses of chloride‐binding and chloride‐free halorhodopsin forms resembles that of bacteriorhodopsin purple neutral and blue acid forms, respectively. The photovoltage data obtained can hardly be explained within the framework of the photocycle scheme suggested by Varo et al. [Biochemistry 34 (1995), 14490–14499]. We suppose that the O‐type intermediate belongs to some form of halorhodopsin incapable of Cl− transport.

Electrogenic processes and protein conformational changes accompanying the bacteriorhodopsin photocycle

The possible mechanisms of electrogenic processes accompanying proton transport in bacteriorhodopsin are discussed on the basis of recent structural data of the protein. Apparent inconsistencies between experimental data and their interpretation are considered. Special emphasis is placed on the protein conformational changes accompanying the reprotonation of chromophore and proton uptake stage in the bacteriorhodopsin photocycle.