3.5.1 Metarhodopsin II
With the formation of Mil, the intermediate capable of catalyzing nucleotide exchange in Gt is reached . In this photoproduct, the Schiff base bond between chromophore and apoprotein is still intact but deprotonated , Blocking Schiff base deprotonation by monomethylating the active-site lysine (Lys296) abolishes light-induced formation of a Mil-like intermediate , According to FTIR analysis, Schiff base deprotonation is mechanistically coupled to the protonation of Glu113 . Based on the definition of Mil as any photoproduct that binds the chromophore via a deprotonated Schiff base, the intermediate is characterized by several special properties (for details see [7,73]) Mil is formed with a large activation energy of >150 kJ mol-1, both for the light-induced formation from the ground state (for references see ), and from MI by pressure jump . The reaction enthalpy relative to the ground state (AH = 110 kJ mol"1) was determined either directly  or was derived from the van t'Hoff equation (AH = 40 kJ mol-1 relative to its predecessor MI) , The entropy increases largely in Mil (37.9 cal K1 mol-1 was determined for the MI^MII equilibrium at pH 7) . The large entropy change and the MI—>MII reaction volume, of the order of 100 ml mol-1  (dependent on the preparation), are apparently due to an unfolding of the protein, leading to exposure of binding sites for Gt and other signal proteins. In accordance, Mil shows an enhanced susceptibility to partial digestion [84,85]. Mil formation depends on the presence of an aqueous milieu (see [86-88]). Hydroxylamine and sodium borohydride can attack the Schiff base only after formation of Mil [89-91], supporting a less constrained chromophore-protein interaction in this state. The formation of Mil from Ml is dependent on the osmotic pressure . Osmotically sensitive regions of rhodopsin, which change their hydration during the MI/MII conversion, are narrow crevices or pockets . Mil formation causes perturbations in the UV-spectrum, indicating a more hydro-philic environment of aromatic residues [93,94], changes in linear and circular dichroism, interpreted as a rotation of the chromophore relative to the plane of the disk membrane [47,95], birefringence changes [96,97] and changes in near-infrared light scattering of the disk membranes [98,99].
UV-absorption changes of Trp126 in helix III and Trp265 in helix VI can be seen in detergent solution between the ground state and Mil . Whether these changes occur in earlier photoproducts is not known. Lack of a steric interaction between Trp265 (Figure 6) and the (3-ionone ring could cause the movement of helix VI . This is in accordance with stationary linear dichroism measurements that show a Trp-residue changes its orientation ~30° during the MI-»MII transition .
Mil formation is accompanied by changes of the membrane interfacial and transmembrane potential, the latter being positive relative to the aqueous exterior of the disk. Electrostatic potentials are measured in situ as a component of the "early receptor potential" [102,103], on lipid impregnated filter materials [104,105], and on lipid bilayers with photoreceptor membranes attached at one side . The ERP can also be recorded directly in giant cells that heterologously express rhodopsin [107,108] and may, in combination with site-directed mutagenesis, offer new approaches to the relationship between electrostatics and structure of the receptor protein.
Vibrational spectroscopy indicates considerable alterations of the apoprotein structure in Mil [109-111]. Some bands in the infrared spectrum can be specifically assigned to carboxyl groups [86,112,113]. Time-resolved EPR spectra indicate small movements near the second cytoplasmic loop, kinetically correlated with Mil formation [114,115]. Many mutations in rhodopsin have been reported to affect its functional activity, but only a few have been assigned to the formation of Mil as a spectrally defined species, including the Schiff base counter-ion [116-118] and certain histidine residues [119,120]. All these data are consistent with considerable conformational changes that accompany Mil formation - necessary for interaction with Gt or arrestin which are only able to bind after Mil has formed.
3.5.2 Role of the hydrophobic environment and light-induced reorganization of disk membrane phospholipids
The native disk membrane contains approximately equimolar amounts of PC and PE phospholipids and lower proportions of PS (=14%) and phospha-tidylinositol (2.5%). The content of cholesterol in the disk membranes varies from the bottom to the top of the ROS [between 0.30 (bottom) and 0.05 (top), relative to the total lipid content] . The phospholipids consist mainly of C22-fatty acids which are polyunsaturated to a high percentage (22:6, but no 22:3 in the dominant species) [9,122,123]. In the native disk membrane environment, the Mil photoproduct is formed in milliseconds, and the MI^MII equilibrium is well on the side of MIL This equilibrium and formation of Mil is influenced by changes in the lipid composition of the disk membrane. Mil formation depends on the presence of unsaturated lipids and on the fluidity of the phospholipid hydrocarbon chains. Thus, increasing the membrane rigidity by the addition of cholesterol or removal of lipid unsaturation, e.g. by reconstitution of the receptor in egg-lecithin vesicles, shifts the MI^MII equilibrium towards MI [124-126], Similarly, in membranes with a reduced lipid content obtained by phospholipase C treatment or membranes containing short-chain, saturated lecithin, the decay of MI is retarded and yields almost no Mil but rather predominantly free retinal plus opsin [127,128], The ratio between the free energies of MI and Mil is modulated by an interfacial tension-like interaction between rhodopsin and the bilayer. However, Mil formation does not require a specific phospholipid head group , but it is enhanced in PE or PS bilayers , probably as a consequence of a surface charge effect .
In highly fluid detergent micelles (octyl glucoside or dodecyl maltoside), both an enhanced rate of Mil formation and a shift of the equilibrium to Mil are observed. In general terms, this means that the energy barrier for the MI—»-Mil conversion is lowered, and that photoproducts preceding the conversion are affected  (for the early MI380 product see ). In octyl glucoside solubilized disk membranes, the activation free energy of Mil formation is linearly dependent on the level of associated disk phospholipid , until the lipid/rhodopsin ratio of the native membrane is reached. This argues against a specific mode of interaction between rhodopsin (and/or Mil) and the lipid. However, FTIR spectroscopy has indicated that a small amount of lipid may bind tightly to the receptor ground state, which alters its interaction in the transition to the active state Mil .
In the disk membrane, lipids undergo a rapid flip-flop between outer and inner leaflet (half mean time <5 min) . In the resulting equilibrium, PS has a distinct preference for the outer leaflet, whereas PC and PE show a small, if any, asymmetry (see [135,136] and references therein). Studies on osmotically intact disk vesicles of bovine ROS have shown that light-induced formation of the active Mil state has an effect on the transbilayer redistribution of disk membrane phospholipids . Redistribution was measured by bovine serum albumin extraction of spin-labelled PC-, PE- and PS-phospholipid analogs from the outer leaflet of the membrane. Upon photolysis of rhodopsin, a change in the redistribution of PS was found as seen by a fast transient (<10 min) enhancement of spin-labelled PS extraction. This effect was augmented by a peptide stabilizing Mil, suggesting a direct release of one molecule PS per rhodopsin into the outer leaflet upon Mil formation and subsequent redistribution between the leaflets. In the case of PC and PE, more complex kinetics were observed. In both cases there was a consistent prolonged period of reduced extraction (two lipids per rhodopsin in each case). The different phases of phospholipid reorganization after illumination are likely to be related to the formation and decay of the active rhodopsin species and to the subsequent regeneration process.
To discuss the signaling state, the MI ^Mil equilibrium in the reaction scheme (Section 3.4.1) has to be extended to incorporate its pH-dependency and to do justice to the known substates of Mil:
The negative and positive enthalpies (AH) in forming MI and Mil, respectively, indicate that molecular interactions built up in MI are lost upon transition to Mil. To drive the conversion, the entropy must increase, and thus the overall disorder in the protein. In both Mil states, Mlla and Mllb , the Schiff base bond of all-ira/M-retinal is still intact but deprotonated. The first protonation switch occurs in the MI to Mlla transition, which is accompanied by translocation of the Schiff base proton to the counter-ion Glu113 (see ). At that stage the prosthetic group all-ira/w-retinal has the characteristics of a ligand agonist that facilitates Mlla formation by elevation of the free energy (AG) of MI. Although the disruption of the salt bridge shifts the conformational equilibrium towards the signaling state, there are other determinants of the active state as shown by mutagenesis of the counter-ion region , Formation of Mlla may also release inactivating constraints among H-II, H-III, H-VI and H-VII due to changes in steric interaction between opsin and alWraws-retinal.
The second step, formation of Mllb, involves the protonation of Glu134 by proton uptake from the aqueous phase [76,138,139]. This residue, a part of the highly conserved E(D)RY motif in GPCRs, forms a salt bridge with the adjacent Arg135, suggesting protonation of Glu134 as a mechanism to directly destabilize the constraints imposed by this salt bridge and to induce formation of the active cytoplasmic receptor surface. By FTIR spectroscopy, it could be shown that, in the complex with Gt, Glu134 of rhodopsin is protonated . Mutation E134Q, eliminating the negative charge, is known to evoke constitutive activity of opsin  and abolishes light-induced proton uptake . However, with bound 1 1-cis-retinal, the mutant rhodopsin is inactive but shows light-induced activity. One explanation could be that an activation mechanism, which is merely based on successive release of constraints, leads to formation of the catalytic receptor-Gt interface. This was approached by construction of mimics of the receptor surface, in which combinations of fragments corresponding to the cytoplasmic loops and/or carboxy-terminal tail of opsin were inserted onto a surface loop of thioredoxin . These mimics showed binding to Gt of varying degrees, but low catalysis of nucleotide exchange in Gt. Full Gt activation requires both the whole opsin apoprotein and the retinal ligand, which controls even the last steps of activation in the native receptor . However, conformational changes caused by release of inactivating constraints due to light-induced changes in steric interaction between opsin and all-iraws-retinal and deprotonation of the Schiff base were measured by EPR and are seen in a dominant movement of the helix VI out of the helix bundle . Similar helix movements upon receptor activation were shown for other GPCRs . A corresponding movement of helix VI cannot be seen in the ground state of the E134Q mutant; however, EPR analysis has shown part of the conformational change around helix III . Mechanistically, it would be interesting to learn whether the dominant movement of helix VI coincides with formation of Mlla or Mllb, and whether it can only occur as a consequence of the protonation changes at Glu"3 and Glu134.
To explain the sequential flow of events, we propose that, first, the N-terminal part of H-III moves into a position to induce proton transfer from the Schiff base to Glu"3 (Mlla). The tandem of glycines would then allow amplified movement of the C-terminal part of this helix, thus inducing a larger structural change at the cytoplasmic surface, which is linked to the repro-tonations and rearrangement events around the ERY tripeptide. This mechanism would require a separate conformational change, following neutralization of Glu113, in agreement with a separate Mllb state. The MIIa/MIIb two-step proton translocation scheme is independently supported by the established two-step pumping or signaling processes in the archaeal rhodopsins.
3.5.4 Mechanistic insights from archaeal rhodopsins and photor ever sal of metarhodopsin II
Archaea contain four retinal-binding proteins, two transport proteins for protons and chloride [called bacteriorhodopsin (BR) and halorhodopsin (HR), respectively], and the sensory rhodopsins (SRs) that mediate phototaxis responses (see also Chapter 1). As in rhodopsins, the chromophore, in this case all-mms-retinal, is bound to a lysine residue in H-VIL High-resolution structures of the ground states of BR, HR and SRII are known [145-147], while structural information on the active M! and M2 states of BR (the possible analogs of Mlla and Mllb in rhodopsin) is available [148,149]. In all retinal proteins, a transient movement of H-VI is found that correlates with the opening of a cytoplasmic half-channel and the relocation of water in the hydrophobic environment of the half-channel. In BR and HR, protonation of a residue (Asp96 in BR) at the cytoplasmic border of H-III occurs in this context. The SRs interact constitutively with dimers of Htr transducer proteins by lateral helix-helix contact, thus forming a receptor that can bind and activate intracellular phosphoregulatory proteins . Both SRs are proton pumps in their free state, but bound transducer proteins block the half-channel for proton transport while helix movement is still allowed to occur . Thus, in the SR-transducer photoreceptor complex, part of the free energy used for the transport mode in free SR is channeled into a long-lived signaling state to account for its sensory function . In these rhodopsins, both H+ transfer near the Schiff base and H+ uptake with movement of helix VI which might provide the trigger for transducer activation are also seen . Therefore, it might be useful to consider this part of signal transmission in rhodopsin as a SR transducer-like partial proton pump . For archaeal rhodopsins, an extended H-bonded network arises [149,152] which might also be relevant for the mechanism of signal transmission, i.e. formation of Mlla and Mllb.
Like bacterial rhodopsin ground states, the Mil intermediate carries all-trans-retinal in its retinal-binding pocket. Therefore the effect of blue light on this photoproduct was soon investigated. Among other photoproducts a measurable one absorbing at 500 nm is generated. The spectral characteristics and the accompanying proton release argued for a reversal of the activation process and photoregeneration of rhodopsin [153,154]. However, a new product with novel properties is formed rather than rhodopsin in ground state , FTIR studies and retinal extraction showed that this product has the protein conformational characteristics of Mil, which arises from Mil by thermal decay, and still carries the all-ira^-retinal isomer. The data indicates the presence of a "second switch" between active and inactive conformations that operates by photolysis but without stable isomerization around the CM=C|2 double bond. It is not known whether transient or metastable isomerizations are involved in this pathway. This emphasizes the characteristic of the rhodopsin ground state, which in contrast to invertebrate rhodopsins (see also Chapter 2) is only accessible by metabolic regeneration with 11-m-retinal and ensures by its exceedingly stable chromophore-protein interaction the low noise of the signal transduction process in rods. Under conditions of substantial bleaching, however, accumulation of the photoreverted product discussed above may influence bleaching adaptation phenomena and even may be involved in blue-light-induced retinal degeneration ,
Was this article helpful?