Shining light on rhodopsin selectivity: How do proteins decide whether to transport H 1 or Cl – ?

The versatile microbial rhodopsin family performs a variety of biological tasks using a highly conserved architecture, making it difficult to understand the mechanistic basis for different functions. Besaw et al. now report structures of a recently discovered cyanobacterial Cl 2 -pumping rhodopsin and its functionally diver-gent mutant that reveal how these transmembrane proteins create a gradient of activity with subtle changes. These insights are paral-leled by a second recent report, which in combination answers long-standing questions about rhodopsin selectivity and will facili-tate future engineering efforts.

The versatile microbial rhodopsin family performs a variety of biological tasks using a highly conserved architecture, making it difficult to understand the mechanistic basis for different functions. Besaw et al. now report structures of a recently discovered cyanobacterial Cl 2 -pumping rhodopsin and its functionally divergent mutant that reveal how these transmembrane proteins create a gradient of activity with subtle changes. These insights are paralleled by a second recent report, which in combination answers long-standing questions about rhodopsin selectivity and will facilitate future engineering efforts.
Light serves as a source of energy and information across all domains of life. Miniscule photons can be used to initiate large changes in biomolecules, particularly in the microbial rhodopsins, where photon-induced isomerization of all-trans-retinal chromophore affects the biological functions of ion pumps, ions channels, phototactic sensors, enzymes, and so on. Microbial rhodopsins are part of a large family of photoreceptive membrane proteins with a conserved structure made up of seven-transmembrane a helices (1). Because rhodopsins can perform such a wide variety of functions using this common compact structure, they have provided ideal model systems to study structure-function relationships in proteins using spectroscopy, biochemistry, and structural biology. Moreover, ion pumps and ion channels have been widely adopted as tools in optogenetics to manipulate the firing pattern of animal nerves with light, and further elucidation of the mechanism of ion-transporting rhodopsins is increasingly required to construct new ion-transporting optogenetic tools.
So far, four types of ion-pumping rhodopsins, including outward and inward H 1 pumps, an inward Cl 2 pump, and an outward Na 1 pump, have been identified. Of these, the archeal outward H 1 pump, called bacteriorhodopsin (BR), was the first to be discovered, in 1971. Then the archeal inward Cl 2 pump halorhodopsin (HR) was reported in 1977 and functionally characterized in 1982. BR has an Asp-Thr-Asp (DTD) motif in the third helix that plays a critical role during H 1 pumping; these three residues are highly conserved in closely related H 1 pumps. HR, in contrast, has a TSA motif, which is similarly critical for its chloride-specific activity. Interestingly, the mutation of the first D of the DTD motif to T in BR confers an inward Cl 2 -pumping function similar to HR (2), suggesting that this residue (Asp-85 in BR) might be responsible for determining whether an H 1 or Cl 2 is transported. However, the reverse mutation of HR, changing the T of the TSA motif to D, did not result in the conversion to an H 1 pump (3). Hence, the determinants separating H 1 and Cl 2 pump rhodopsins were not yet completely clear.
Recently, it was revealed that cyanobacteria have a new type of Cl 2 -pumping microbial rhodopsins with a TSD motif. One of them, MastR (also known as MrHR) from Mastigocladopsis repens, functions as a Cl 2 pump similar to the TSA-containing HR, but in this case, the Thr to Asp mutation (MastR T74D) was successful in creating an H 1 pump (4), thus achieving complete functional conversion between an H 1 pump (BR) and a Cl 2 pump (MastR) by swapping only one residue. Moreover, another TSD-containing HR from Synechocystis sp. PCC 7509 called SyHR was shown to be able to pump a sulfate ion (SO 22 4 ) that other HRs are unable to transport (5). These findings suggest that TSD-containing HRs have molecular mechanisms for ion transport that differ from the archeal TSA-containing HRs. To understand the mechanisms of these unique abilities, however, demands their three-dimensional structure at atomic level.
Besaw et al. now answer this call, revealing the X-ray crystallographic structure of MastR and its H 1 pumping mutant (6) in bicelles at 2.33 and 2.50 Å, respectively. In the structure of WT MastR, the substrate Cl 2 is bound to Thr-74 and Ser-78, corresponding to the T and S in the motif, and to the Schiff base part of the retinal via a water molecule. In MastR T74D, the Asp occupies the position adopted by Cl 2 in the WT protein, and strong hydrogen bonds similar to those found in BR are formed between the Asp, water molecule, and retinal (Fig. 1). This result supports the idea that a strong hydrogen-bonding network is essential for the H 1 pump, in agreement with previous spectroscopic and theoretical studies (1). Another structure of the same protein was also reported by Yun et al. (7). They further focused on the difference in an extracellular loop between MastR and SyHR. The replacement of residues in the extracellular loop in the former with the corresponding sequence from the latter resulted in MastR being able to transport SO 22 4 . A structure of the SO 22 4 -transporting mutant revealed a more positively charged entrance to the putative extracellular ion pathway compared with the WT, suggesting that this entrance determines the anion selectivity of TSD-type HRs. Besaw et al. similarly point to differences in loops between MastR and BR that seem to explain their ion selectivity, suggesting an even broader role for the extracellular entrance.
These two structural studies illuminate the long-sought functional determinants between H 1 and Cl 2 pumps and structural elements enabling divalent-anion pumping. Moreover, the new structures provide an exciting opportunity to revisit our understanding of rhodopsin evolution in general, in which the structural elements important for the function of the ancestral molecule were conserved * For correspondence: Keiichi Inoue, inoue@issp.u-tokyo.ac.jp.
in the descendants with diversified functions, as discussed by Besaw et al. There are, however, unsolved mysteries of ion transport by TSD-containing HRs remaining. For example, we do not know the structure of the SO 22 4 -bound state during SO 22 4 transport. Also, although both studies suggested possible ion transport pathways in their respective proteins, these structures represent the dark state, so the precise structure of the transiently opened pathway has not been revealed. To solve these problems, further structural studies on the SO 22 4 -bound protein and photoactivated intermediates are required. Freeze-trapping methods will help to capture intermediate structures under light illumination (8), but recently developed techniques, such as time-resolved serial millisecond crystallography (TR-SMX), available at synchrotron facilities (9), and the time-resolved serial femtosecond crystallography (TR-SFX) technique with X-ray free electron lasers (XFEL) (10), will provide not only structural insights into each photointermediate but also the dynamics of conformational change between them. It will be exciting to see how these and other explorations shed light on this fascinating protein family.