Figure 6. Scheme of the photocycles of SRI and NpSRII, depicting the major intermediates. Included are the proton uptake and release steps. Abbreviations: CC, cytoplasmic channel; EC, extracellular channel.
the K-like intermediate but strongly affects its relaxation pathway , Interestingly, the same volume change of 10 mL mol~' was determined for the NpSRII-transducer complex .
Early on, it was obvious that an M-like intermediate was present in the photocycles of both pigments, SRI [155,156] and HsSRII . In the case of SRI the investigation of the photocycle is hampered by the fact that two species SRI587 and SRI550 are present at physiological pH. Bogomolni and Spudich - assuming a unidirectional, unbranched reaction scheme-identified two precursor intermediates (K: S610; L: S560) which are followed by the long-lived M-like state (S373) . The latter intermediate, possessing a lifetime of about 750 ms, can absorb a second photon which short-cuts the cycle and is responsible for the UV-sensitive negative phototaxis. Apparently, the photocycle of both SRI 587 and SRI550 contain an intermediate with a deprotonated Schiff base, i.e. S373. Since in SRI5g7 the proton acceptor of the Schiff base proton (Asp76) is already protonated another group has to take over this role, which might be His 166 (see below, ).
After excitation by light S373 relaxes back to the ground state via a red-shifted intermediate (S510) with a half-life of 80 ms. These reactions have been studied in more detail by Bogomolni and co-workers [159,160]. In an interesting observation Manor et al. described the influence of the membrane potential on the photocycling rates of SRI, BR and HR , which might be important for the function of SRI under physiological conditions, which displays a proton motive force of about -250 mV [162-164].
The photochemical cycle of HsSRII is not very well characterised due to its sensitivity towards external conditions , Nevertheless, the few data available show a photocycle with a sequence of intermediates corresponding to blue-shifted L, M and O intermediates [35,96,165,166], FTIR experiments proved that Asp73, the counter-ion of the Schiff base, is the Schiff base proton acceptor . Interestingly, the O-intermediate (SII540) is characterised by an dW-trans retinal conformation, indicating that the reisomerisation from 13-cis has already occurred. More information about the photocycle is available for the homologous protein NpSRII [125,168 -171], Summarising these data, a photocycle similar to that of bacteriorhodopsin is obtained (Figure 6). However, the turnover rate of 1.2 s is much slower than that of BR. FTIR data provided evidence that, in NpSRII, Asp75 becomes protonated during the formation of the M-intermediate  which can be photoconverted into the initial state .
Interestingly, in the BR-mutant D96N the slow photocycle can be enhanced by the addition of azide which accelerates the reprotonation of the Schiff base . Similarly, in NpSRII the M-decay rate is substantially increased by the addition of azide although the overall turnover did not change [39,174], A similar acceleration of the M-decay can be obtained if residues from the cytoplasmic or extracelluar proton conducting channel are mutated to residues found in BR [39,40],
1.3.4 Proton transfer reactions of sensory rhodopsins
Investigating the light-induced proton transfer reactions in sensory rhodopsins provided important insight into the signalling mechanism of the photoreceptors. However, only recent progress in the biochemical accessibility of the SRs and the application of different electrophysiological techniques has led to a common picture of how the photoreceptors transport protons across the plasma membrane and how their transducers influence this process. For SRI and NpSRII it is evident that in the transducer-free state the sensors act as outwardly directed proton pumps and that their corresponding transducers Htrl and NpHtrll inhibit this pumping specifically [39,66,93,94,137,138, 175,176], Although the capability of HsSRII to pump protons has been questioned , the latest results indicate that NpSRII, like SRI, acts as an outward directed proton pump [93,94],
Due to their slow photocycle turnover (t ~ 1 s), SRI and HsSRII are less effective pumps than BR and HR (t ~ 10 ms) , The main differences between the ion pumps and the receptors are found in the second part of the photocycle. The molecular events in this part of the photocycle are the reprotonation of the Schiff base (M-decay), the subsequent 13-cis/a.W-trans isomerisation of the retinal and the deprotonation of the counter-ion of the Schiff base. From this comparison of the photocycles of the SRs and BR the proton release during M-formation should be as fast as in BR but the proton uptake (M-decay) should be decelerated considerably. A sequence alignment (Figure 2) supports this hypothesis since the Schiff base proton donor Asp96 in BR is replaced by aromatic residues in all SRs [36-38]. It should be noted that in Nop-1, the recently discovered eukaryotic archaeal rhodopsin, this Asp residue is conserved. Apparently, this site does not solely govern the photocycle turnover, as was also demonstrated by mutational studies on NpSRII and SRI (see below).
SRI was the first of the two receptors for which proton pumping was demonstrated. Bogomolni as well as Haupts and their co-workers analysed pH changes in suspensions of SRI-containing vesicles or H. salinarum cells and observed an acidification upon illumination [137,138]. More detailed mechanistic information was provided later on by time-resolved measurements of photocurrents using the BLM-technique with reconstituted SRI samples  and voltage-clamp recordings from Xenopus oocytes after the injection of mRNA encoding the sopl gene . The results can be summarised as follows. At neutral pH there are two species (SRI587 and SRI550, see above) which can be excited by orange light. Apparently, SRI550, the species with a deprotonated Asp76, contributes exclusively to the electrogenic photocycle, as can be concluded from the action spectrum and the pH-dependency of the voltage-clamp photocurrents. The action spectrum for the light-induced membrane potential shows a maximum at 550 nm  and the pH dependent amplitudes of the voltage-clamp signals coincide with the titration of Asp76 with a pKa of 7.2 .
The thermal M-decay can be accelerated from 1 s to 80 ms by the absorption of a second "blue" photon [33,159], resulting in an enhanced proton pumping activity [93,138,175], Consequently, under white light ("natural") conditions SRI is a two-photon driven pump. This amplifying effect of blue light is unique for SRI. For BR and NpSRII (see below) blue light has exactly the opposite effect as it quenches the photocurrent [39,179]. Apparently, the molecular switch which changes the Schiff base orientation from the extracellular side in M, to the cytoplasmic side in M2 is faster in SRI than in BR and NpSRII, thereby accumulating M2. Hence, in SRI the additional blue light would mainly excite M2, which is followed by a proton uptake from the cytoplasm but not, as was demonstrated for BR, from the extracellular side after excitation of M,. Therefore one can conclude that SRI, contrary to BR, is a two-photon driven proton pump.
For HsSRII as a proton pump the results are ambiguous. Sasaki and Spudich investigated light-induced pH changes in vesicles or open membrane sheets solutions containing HsSRII . Under continuous illumination the authors observed an increase in pH to a constant level which returned back to the initial value after switching off the light. From their results they concluded that the proton uptake and release occurs from the same (extracellular) side, resulting in a non-electrogenic photocycle. The light induced increase of pH was explained by a steady-state mixture of photo-intermediates which have picked up a proton from the bulk. Conversely, Schmies and co-workers detected from Xenopus oocytes, with more sensitive voltage-clamp recordings, a photocurrent that proves outwardly directed proton pumping by HsSRII . Although there is no explanation for this contradiction the photocycle data indicate a similar reaction pathway of HsSRII and BR, strongly supporting an electrogenic photocycle.
NpSRII-F86D + NpHtrll
Figure 7. Photocurrent traces of the NpSRII mutant F86D without and in the presence of NpHtrll (data taken from ). The currents were measured using the oocyte system which was clamped at -20 mV. The transient photocurrent is seen in the first part of the traces directly after the onset of illumination.
Recently proton transport in NpSRII has been analysed in greater detail. Schmies et al. [39,93] demonstrated an outwardly directed proton transport by NpSRII (see Figure 7). Kamo and co-workers confirmed the proton pumping by using a proton-sensitive electrochemical Sn02 cell and pH-measurements in suspensions of NpSRII-containing vesicles [94,176]. The efficiency of the light-induced photocurrent is about 100 times weaker than that of BR. It can be enhanced by replacing F86 by an Asp residue (in BR, the corresponding Asp96 serves as proton donor to the Schiff base) or by the addition of an external proton donor like azide , Interestingly, these modifications do not increase the photocycle turnover which still lasts about Is [39,174] although the addition of azide accelerates one step in the photocycle, namely the repro-tonation of the Schiff base . Since in these examples the turnover has not changed as compared to that of the wild-type the increased amplitudes of the photocurrents must originate from other sources. A proposal was put forward by Schmies et al. [39,93] who assume a two-photon process which short-cuts the photocycle thereby enhancing the pump activity. The second photon has to excite an intermediate which follows M because an excitation of M quenches pumping in a similar manner as observed for BR (see above).
The proton transfer steps resulting in a vectorial transport of protons across the membrane can be formally separated into two parts. First, during the L-M transition the proton from the Schiff base is transferred to the counter-ion Asp75 as in BR and SRI (BR-Asp85, SRI-Asp76) [127,134,180-182]. This fast charge movement towards the extracellular side is represented in the transient photocurrent (Figure 7). From this experiment it cannot directly be concluded that the proton has already been released into the bulk phase. However, Iwamoto et al.  and Sasaki et al. (for HsSRII ) have provided evidence that the proton release from the membrane occurs later during the O-decay. The second step, the reprotonation of the Schiff base from the cytoplasm is not resolved under steady state conditions, but the occurrence of a continuous photocurrent proves the net transport of protons across the membrane. These data show that both HsSRII and NpSRII are proton pumps, albeit not very efficient.
The oocyte experiments allow the measurement of the photocurrent in the presence of an applied membrane potential. At about -120 mV both the transient and the photostationary current disappear in NpSRII , For comparison, proton pumping in BR vanishes at about -250 mV [162,183], Since the membrane potential (A*F) in N. pharaonis is about -250 mV  and the proton motive force in H. salinarum is of the same order [162,163], a physiological relevance of the proton translocation can be ruled out. Nevertheless, the steep voltage-dependence of the SR's photocurrents indicates charge movements accompanied by large conformational changes. This is in line with models describing the signal transfer between photoreceptors and transducers which assume that the tilting of helix F triggers the activation of the transducers signalling domain (see 1.4.2, [21,108]).
126.96.36.199 Proton transfer reactions in the receptor!transducer complexes The observation that sensory rhodopsins can function as proton pumps led to the question whether the binding of their cognate transducers influences this property. Indeed, Bogomolni et al. reported this effect for the SRI/Htrl complex , using the same experimental approach as they applied to demonstrate proton pumping by uncomplexed SRI . In the latter experiment the illumination of sealed vesicles containing SRI resulted in an acidification of the bath medium. However, no pH-change was observed for vesicles with an incorporated SRI/Htrl-complex. This was confirmed by voltage-clamp recordings using Xenopus oocytes . Similar results were also obtained for the mutant NpSRII-F86D (which displays an enhanced photocurrent). In an experiment similar to the pH-measurements of Bogomolni et al. , Sudo et al.  verified the voltage-clamp data for NpSRII/NpHtrll. For HsSRII, Sasaki et al.  did not detect an overall proton transport in HsSRII. Consequently, the HsSRII/Htrll complex should not show vectorial proton transfer.
It is important to note that the binding of the transducer to its receptor only affects the photostationary but not the transient photo-current . Therefore, the fast proton reactions are not inhibited and neutralisation of the Schiff base^ounter-ion pair can still occur. Therefore, the proton transfer reactions, which lead to a disruption of the salt bridge between the protonated Schiff base and its counter-ion, might be important in the formation of the signalling state, as Spudich and co-workers pointed out (e.g. [130,185]).
In two different studies, the specificity of receptor/transducer interactions was demonstrated. With the oocyte system, cross co-expression of SRI with NpHtrll and vice versa, NpSRII with Htrl, does not alter the receptors' photocurrent, indicating that no complex between SRI/NpHtrll and NpSRII/ Htrl is formed, or at least no specific interaction takes place , In another approach Spudich and co-workers prepared transducer chimeras between Htrl and Htrll in which different transducer domains were combined . After expression of the chimeric signalling complexes in H. salinarium the authors analysed the phototaxis of these cells and concluded that the receptors interact specifically with their corresponding transducers. Whereas the cytoplasmic domains can be exchanged, SRI and HsSRII need the two transmembrane helices of their cognate transducers Htrl and Htrll, respectively to mediate a correct physiological response.
The question still arises, which properties of the functional complex are responsible for the inhibitory effect of the transducers and is it an important feature of the signalling mechanism? Data obtained so far indicate that these specific SR/Htr interactions are likely to be located in the cytoplasmic part of the membrane. Spudich and co-workers concluded that Htrl closes the cytoplasmic channel of SRI because the reprotonation rate of the SB (M-decay) in SRI/Htrl is insensitive to the pH, but becomes pH-sensitive after removal of the transducer [65,184,186], A closure of the cytoplasmic channel of SRI by Htrl would reduce the accessibility of protons from the cytosol which explains the inhibition of the pumping. However, an alternative explanation might also be possible. If, as discussed above, SRI550 exclusively contributes to the pumping, the shift of the SR550/SRI587 equilibrium towards almost 100% SRI587 upon Htrl binding would automatically abolish the photocurrent. Certainly, the two explanations are not exclusive and both mechanisms might occur simultanously. For NpSRII, the pKa of the SB-counter-ion Asp75 is, at 5.6, too low for a -COOH <=> -COO equilibrium under physiological conditions . It follows that the inhibition of the proton pump on binding of NpHtrll is not due to a non-pumping species.
1.3.5 Properties of the SR/Htr complex
The binding of Htrl to SRI has large kinetic effects on the SRI photocycle, a result which provided initial evidence for the formation of a functional complex [65,85], A smaller but distinct influence on the lifetime of the O-intermediate in HsSRII on binding its Htrll has been described ,
Contrary to these results no effects on the photocycle kinetics of NpSRII were observed if a shortened transducer was bound , In respect of the inhibition of the proton pump by the transducer  and a binding constant in the 100 nmol range  it was concluded that a functional complex is formed.
The observation that proton pumping is inhibited and photocycle turnover is altered in the SRI/Htrl complex has been utilised to elucidate the interaction of SRI with its transducer Htrl, including also mutational studies. A minimal structural unit comprised the receptor and the N-terminal transducer sequence (1-159) [89,91], Function-perturbing mutations in SRI and Htrl altered the rate of S373 (M) formation which was interpreted as a modulation of the electrostatic interactions of the protonated Schiff base and an optimisation of the photocycle by the transducer .
Interesting mutations in SRI presented by Spudich and co-workers involved Asp201 and Hisl66 [82,158,188], In  Olson et al. replaced Asp residues at five positions of SRI by site-specific mutagenesis. It turned out that Asp201 is most vital for the attractant signalling function, which is changed to a repellant response when this residue is substituted by the isosteric asparagine. The authors point out that this result genetically separates the attractant and repellant response of the bacteria. One proposed explanation assumes that in the dark the signalling complex is locked in an inappropriately attractant adapted state, similar to the situation of repetitive stimulation by orange light [189,190].
From mutational studies on His 166 Zhang and Spudich concluded that this residue plays a role in the proton pathway after deprotonation of the Schiff base, the modulation of SRI photoreaction kinetics by Htrl and is important in phototaxis signalling . Since under physiological conditions the protonated Asp76 is not available to accept the proton His 166 might be an alternate site. Only this reaction sequence would lead to the formation of the signalling state capable of activating Htrl. Indeed, His 166 replacements have conformational effects on the structure of Htrl at position 64 .
These results were explained by a two-conformation equilibrium model introduced by Spudich and Lanyi as a unified mechanism for ion pumping and signal transduction . According to this proposal SRI consists of an equilibrium mixture of two conformers which have similar properties to the closed and open channel conformers of BR. Orange light shifts the equilibrium towards the attractant (A) state whereas in the two-photon cycle the repellent (R) conformer is populated. In support of this model are observations concerning the "orange-light inverted" phenotype of some SRI-mutants like D201N  or HI66A  and the Htrl-mutant E56Q , Apparently, in all these mutants the equilibrium is shifted in the dark towards the A conformer, exhibiting a repellent response to both one-photon and two-photon activation. Further support comes from a suppressor mutational analysis on mutants at these two sites (D201N, H166S, H166A) as well as on the Htr mutant E56Q . Assuming that the effects of these single site mutants can be reversed to wild-type properties the authors screened for second site mutations that restored the attractant response. Fifteen such mutants were identified with three suppressor mutations at the cytoplasmic side of helix F and G of SRI and the other 12 mutations in Htrl clustering at the cytoplasmic end of TM2. These sites certainly are intimately involved in the binding surface of the receptor to the transducer.
Was this article helpful?