An affordable convertible: Engineering proton transfer pathways in microbial rhodopsins

Main text

Microbial rhodopsins offer seemingly endless opportunities for protein engineering, especially when it comes to proton transfers. These simple seven-helical retinal-binding membrane proteins employ the same basic scheme to perform a multitude of functions. The chromophore, which is normally present as a protonated all-trans-retinal Schiff base (SB) with a lysine side chain on the seventh helix, absorbs a photon of visible light and photoisomerizes to the 13-cis form, triggering a cascade of thermal reactions known as a photocycle. In the course of the photocycle, microbial rhodopsins experience conformational changes of retinal and protein, changes in proton affinity of various side chains, and rearrangement of internal water molecules. This often induces multiple proton transfers, which may or may not result in net proton transport (inward or outward). In the case where proton transfers do not result in proton transport, they may be coupled with sodium or chloride ion transport, ion channel opening, or photosensory function (1,2).

In either case, it appears that the key primary proton transfer is the one involving the deprotonation of retinal SB, forming an active state, which may be associated with ion transport, signaling function, or channel opening. At the very minimum, it is followed by another key proton transfer involving SB reprotonation, but usually, there are many more secondary proton transfers involved. The occurrence and rate of the light-induced SB deprotonation and reprotonation may depend on several factors, such as a suitable pKa of the SB, its proper orientation (accessibility), and the availability of proton acceptors and donors and/or water molecules conducting protons to the acceptors and from the donors. Typically, outward proton pumps, such as archaeal bacteriorhodopsin, eubacterial green proteorhodopsin, and xanthorhodopsin, and their fungal and algal homologs have two carboxylic counterions for the positively charged SB, located on the third and seventh helices on the extracellular side of retinal. The counterion on the third helix serves as a primary proton acceptor in the SB deprotonation reaction. Some sensory rhodopsins retained this geometry and the primary proton transfer to produce the signaling state. In addition, all outward proton pumps possess a strongly bound water molecule interacting with the SB (3). On the contrary, inward proton pumps, such as xenorhodopsins and schizorhodopsins, have only one counterion, and their SBs release protons toward the cytoplasmic side (4). Likewise, inward chloride and outward sodium pumps replace the carboxylic proton acceptor on the third helix with a nonprotonatable residue (usually threonine or asparagine). Most chloride pumps do not deprotonate the SB at all, while the sodium pumps use another carboxylic proton acceptor one helical turn down toward the cytoplasmic side of the third helix.

Such an amazing variety of motifs and geometries in the SB vicinity provides fertile ground for the functional interconversion of microbial rhodopsins by site-directed mutagenesis. These elegant experiments allow identifying the minimal set of amino acids required for a certain biological function by taking a protein of different functionality and strategically mutating it until the functionality switches to the desired one. For example, the outward proton-pumping function could be restored in sodium pumps and two out of the three chloride pumping families (all but archaeal) by reconstructing the carboxylic primary proton acceptor (and sometimes the cytoplasmic proton donor) on the third helix (5,6).

A new group of TAT rhodopsins (7,8) poses a particular challenge in this regard. Not only do they lack SB deprotonation, but they do not seem to show any productive photocycle at all, returning to the original state from very early intermediates in a few microseconds. Similar to the chloride pumps of archaea and cyanobacteria (halorhodopsins), TAT rhodopsins replaced their third helix proton acceptor with threonine, but this does not make them transport chloride. Instead, they seem to function as sensors of ultraviolet light and/or pH using the form with the deprotonated SB, which is present under physiological conditions due to its low pKa. The paper by Sugimoto, Katayama, and Kandori published in this issue of Biophysical Journal (9) sheds light on the origin of the unproductive photocycle of a TAT rhodopsin by studying its mutant with restored ability for SB deprotonation (10) and defines its proton transfer pathways.

The question of the origin of the unproductive fast photocycle is not trivial, as a simple replacement or protonation of the third helix counterion (the primary proton acceptor) may abolish proton transport but very rarely results in the disappearance of all photointermediates observable beyond one microsecond after the excitation (11). Still, the T82D mutant of TAT rhodopsin, in which the proton acceptor is restored, displays a slow photocycle with SB deprotonation when Asp82 is anionic and behaves like the wild-type protein when Asp82 is protonated (10). This result is a bit counterintuitive, as negatively charged Asp82 strongly increases the pKa of the SB in the parent state, which, nevertheless, does not prevent its deprotonation after excitation. In the current paper (9), Sugimoto et al. are looking for other possible reasons for the lack of proton transfers in TAT rhodopsin and identify putative proton donors and acceptors in its T82D mutant.

First, they approached the strength of retinal-protein interactions in the wild-type and the mutant using infrared difference spectroscopy. Normally, microbial rhodopsins show significant retinal distortions (polyene chain twisting) after photoisomerization, which is believed to be necessary for storing photon energy. Wild-type TAT rhodopsin does not show such distortion, in agreement with its weak hydrogen bonding of the SB (9,12), which seemingly points to the origin of the unproductive photocycle. Unexpectedly, the T82D mutant, which restores the strong hydrogen bonding of the SB, does not show any light-induced retinal distortion either. This eliminates retinal twisting as a culprit and suggests that additional sequence analysis and mutagenesis may be needed to answer this question.

Nevertheless, the T82D mutant itself shows a very interesting and unusual pattern of proton transfers, which contributes to our understanding of hydrogen-bond network diversity in microbial rhodopsins. Unsurprisingly, the SB gives up its proton to the restored acceptor, Asp82, but this does not result in the net proton transport. Instead, the authors show (9), using additional mutagenesis and infrared spectroscopy, that the reprotonation of the SB occurs from the same side (extracellular), resulting in futile proton shuttling. The newly identified proton donor, Glu54, which is also implicated in the binding of Ca²⁺ ions (9,13), may play a significant role in TAT rhodopsin photochemistry and oligomerization (8). Interestingly, its histidine homologs in many bacterial proton-pumping microbial rhodopsins are known to form complex Asp/His counterions and are implicated in proton transfer modulation and oligomer formation (14,15). Finally, a homologous glutamate was recently identified in a new subfamily of sodium pumps, where it is also involved in the regulation of oligomerization and the modulation of pKa of the proton acceptor for the SB (16). In summary, while the biological function and photochemistry of TAT rhodopsins remain somewhat mysterious, their studies continue to broaden our understanding of structural prerequisites for proton transfers in membrane proteins.

Acknowledgments

The research was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC).

Declaration of interests

The author declares no competing interests.

Editor: Ana-Nicoleta Bondar.