WO2021086219A1 - Light-gated pentameric channel - a new optogenetic tool - Google Patents

Abstract

The invention relates to newly characterized light-gated pentameric channels such as OLPVRII and their use in medicine, their utility as optogenetic tools, nucleic acid constructs encoding same, expression vectors carrying the nucleic acid construct, cells expressing said light-gated pentameric channels and their respective uses.

Description

LIGHT-GATED PENTAMERIC CHANNEL - A NEW OPTOGENETIC TOOL

The invention relates to newly characterized light-gated pentameric channels and their use in medicine, their utility as optogenetic tools, nucleic acid constructs encoding same, expression vectors carrying the nucleic acid construct, cells expressing said light-gated pentameric channels and their respective uses.

BACKGROUND OF THE INVENTION

All cells maintain a particular concentration of ions H⁺, K⁺, Na⁺ and Cl in cytoplasm, which is crucial for life. Ion gradients across cell membranes are maintained by ion transporters which are integral membrane proteins. In many archaea, bacteria, unicellular eukaryotes, these gradients are created (among other mechanisms) by light-driven microbial rhodopsins. Microbial rhodopsins that are extremely distantly related to animal visual rhodopsins comprise an expansive family of seven transmembrane proteins that contain a covalently attached cofactor, retinal. Upon absorption of a photon, retinal isomerizes triggering a series of structural transformations that correlate with functional and spectral changes and are known as the photocycle. Microbial rhodopsins are currently considered to be universal and the most abundant biological light energy transducers. Before the year 2000, only microbial rhodopsins from halophilic archaea have been known. A 2000 metagenomic study resulted in the discovery of a rhodopsin gene in marine Proteobacteria that was, accordingly, named proteorhodopsin (PR) (Beja, O. et al. Science. 289, 1902-1906 (2000)). Since 2000, thousands of microbial rhodopsins have been identified, in all three domains of life (bacteria, archaea and eukaryota) as well as in large viruses (Beja, O. & Lanyi, J. K. Proc. Natl. Acad. Sci. U. S. A. Ill, 6538-9 (2014)). The renaissance of rhodopsins as a research field has culminated in the development of optogenetics, the revolutionary method for controlling cell behavior in vivo in which microbial rhodopsins play the key role (Boyden, E. S. et al. Nat. Neurosci. 8, 1263-1268 (2005); Deisseroth, K. et al. J. Neurosci. 26, 10380— 10386 (2006)).

Several rhodopsins with unexpected functions have been discovered and characterized recently. Among the members of this family are light-driven proton, anion and cation pumps, light-gated anion and cation channels, and photoreceptors. Recently, rhodopsins that function as inward proton pumps have been discovered (Inoue, K. et al. Nat. Commun. 7, 13415 (2016)). Genomic and metagenomic studies have dramatically expanded the collection of rhodopsin sequences, some of which have been identified in unexpected organisms and habitats, for example, sodium-pumping rhodopsins (NaRs) in Flavobacteria (Kwon, S. K. et al. Genome Biol. Evol. 5, 187-199 (2013)), and the wide spread and importance of PR-based phototrophy in the marine environment have become evident (Beja, O. et al. Nature 411, 786- 789 (2001)). However, most of microbial rhodopsin diversity remains experimentally uncharacterized.

Computational analysis of proteins encoded by Nucleocytoplasmic Large DNA Viruses (NCLDV) of eukaryotes has led to the discovery of genes encoding PR homologs in Organic Lake Phycodnavirus and Phaeocystis globosa virus which belong to the extended Mimiviridae family (Yutin, N. & Koonin, E. V. Biol. Direct 7, 34 (2012)). Subsequently, a number of additional rhodopsins have been identified in other NCLDV. These viral rhodopsins show only distant (albeit statistically significant) sequence similarity to microbial rhodopsins with known functions. Phylogenetic analysis shows that viral rhodopsins are monophyletic and split into two distinct branches (Fig. 1). The structure, function, and role of the viral rhodopsins in the infection of the host protists remain unknown. Given the distant relationship between the viral rhodopsins and the rest of the microbial rhodopsin superfamily, the former could be expected to have unique properties.

By the examination of the crystal structure, functional characterization and molecular dynamics (MD) simulations of Organic Lake Phycodna virus rhodopsin II (OLPVRII), a representative of the second group of viral rhodopsins, we suggest that it is a pentameric lightgated channel that is functionally analogous to well-studied pentameric ligand-gated ion channels playing crucial roles in many cellular processes. The functional characteristics of OLPVRII make this group of proteins promising candidates for the development of next generation optogenetic tools.

It was shown, that OLPVRII forms a pentamer in the lipid membrane with the unique bottle-shaped central pore. The pore is comprised of two parts: narrow hydrophobic 'neck', covered with the ring of five positively charged amino acid residues (R29) facing the intracellular space and wide more polar part, facing the extracellular bulk. The pentamer is stabilized by a set of extremely conserved amino acids, such as E26, R36, H37, N40 and W203. The profile of the central pore is very similar to that of known channels with hydrophobic gating mechanisms, such as, for instance, pentameric ligand-gated ion channels (pLGICs). MD simulations indicated that the pore serves as an anion- selective channel (presumably for chloride). The selectivity is provided by the ring of five arginines at the cytoplasmic part of the pore.

Functional tests indicated that OLPVRII pumps protons, which is characteristic for almost all known microbial rhodopsins, as the primary reaction is the proton translocation from the Schiff base, connecting retinal and lysine residues in the middle of the protein, to the nearby residues or cluster of water molecules. Study of OLPVRII kinetics showed that the photocycle is longer than that of known proton pumps, which also supports the light-gated channel function of the protein.

The channel formed in the middle of the OLPVRII pentamer is unique among microbial rhodopsins. Indeed, it is symmetrical and it is thus convenient to rationally design the OLPVRII variants with modified selectivity. Particularly, substitution of R29 by glutamate residue switched the selectivity from anions to cations, as indicated by the MD.

Thus OLPVRII should provide photocurrents which are comparable or even exceed those of other channelrhodopsins used for optogenetics, such as channelrhodopsin-2 (ChR2), anion channelrhodopsins (GtACRs) and red-shifted cation channelrhodopsin Chrimson. The currents should be enough to hyperpolarize neuron membranes.

It is an object of the present disclosure to provide new optogenetic tools based on the use of pentameric light-gated channels. Such optogenetic tools could be valuable in the field of scientific research as well as in medicine.

SUMMARY OF THE INVENTION

Microbial rhodopsins are represented in all three domains of cellular life. Recently, rhodopsin genes have been identified in some large double-stranded DNA viruses, but the structure and functions of viral rhodopsins as well as their possible contributions to virus-host interactions remain unknown. In this disclosure we describe an Organic Lake Phycodnavirus rhodopsin II (OLPVRII), a representative of the largest group of viral rhodopsins. The protein forms a pentamer, with a symmetrical, bottle-like central channel. The narrow vestibule of the channel is positioned in the cytoplasmic part, and its entrance is covered by a ring of 5 arginine residues, whereas 5 phenylalanines form a hydrophobic barrier in the exit of the vestibule. The proton donor E42 is placed in the cytoplasmic half of transmembrane helix B. The architecture of the OLPVRII, with the central pore, is unique among the rhodopsins of known structure. The structural and functional characterization of the viral rhodopsin, together with molecular dynamics simulations, suggest that OLPVRII might be a light-gated pentameric ion channel functionally analogous to pentameric ligand-gated ion channels.

Accordingly, disclosed is a light-gated pentameric channel having at least 59% sequence similarity over the full length of SEQ ID NO: 1 (OLPVRII) for use in medicine and other optogenetic applications, as further defined in the claims. For example, the light-gated pentameric channel may comprise or consist of an amino acid sequence SEQ ID NO: 1 (OLPVRII).

Also provided is a nucleic acid construct, comprising a nucleotide sequence coding for the light-gated pentameric anionic channel as disclosed herein, wherein the nucleotide sequence is codon-optimized for expression in human cells; and an expression vector, comprising a nucleotide sequence coding for light-gated pentameric channel as disclosed herein or said nucleic acid construct, wherein the nucleotide sequence is optimized for expression in human cells.

Also contemplated is a mammalian cell expressing the light-gated pentameric channel as disclosed herein, with the proviso that the mammalian cell is not a human embryonic cell or a cell capable of modifying the genu line genetic identity of human beings; and a mammalian cell comprising the nucleic acid construct or the expression vector of the present disclosure. Furthermore, the present disclosure also provides a liposome, comprising the lightgated pentameric channel as disclosed herein.

The light-gated pentameric channel, the nucleic acid construct, the expression vector, the mammalian cell, or the liposome of the present disclosure may be advantageously used in medicine, such as for use in restoring auditory activity, recovery of vision, or for use in treating of neurological injury, brain damage, or a degenerative neurological disorder, like as Parkinson's disease and Alzheimer's disease. In addition, the present disclosure provides a non-human mammal, comprising a cell of the present disclosure, preferably wherein the cell is an endogenous cell; with the proviso that those animals are excluded, which are not likely to yield in substantial medical benefit to man or animal which will outweigh any animal suffering.

Finally, also provided is a non-therapeutic, or ex vivo, or in vitro use of a light-gated pentameric channel as disclosed herein, 1) for light-stimulation of electrically excitable cells, 2) for transporting cations or anions over a membrane, 3) or as an optogenetic tool.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

The examples herein show functional studies of the representative of the yet non- characterized viral rhodopsin OLPVRII from Organic Lake Phycodna Virus belonging to the family of microbial rhodopsins which suggest that this rhodopsin is light-gated pentameric anion channel. In particular, structural data, molecular dynamics simulations and amino acid conservancy indicate that OLPVRII is light-gated pentameric anion channel. The crystallographic structure of OLPVRII reveals the ion translocation pathway that is very different from that of the known rhodopsins. Due to its intrinsic properties, particularly its pentameric organisation, a viral rhodopsin OLPVRII is highly suitable for genetic modifications, and expected maximal currents are very high, which makes this rhodopsin an attractive alternative for the existing optogenetic tools.

Accordingly, disclosed herein is a light-gated pentameric channel having at least 59% sequence similarity over the full length of SEQ ID NO: 1 (OLPVRII) for use in medicine and other optogenetic applications. In preferred embodiments, the light-gated pentameric channel has at least 65%, more preferably at least 70%, more preferably at least 75%, more preferably at least 80%, more preferably at least 85%, more preferably at least 90%, more preferably at least 95%, more preferably at least 97%, more preferably at least 98%, and most preferably 99% sequence similarity to the full length of SEQ ID NO: 1 (OLPVRII).

Alternatively, or in addition, the light-gated pentameric channel can have at least 38%, more preferably at least 45%, more preferably at least 50%, more preferably at least 55%, more preferably at least 60%, more preferably at least 65%, more preferably at least 70%, more preferably at least 75%, more preferably at least 80%, more preferably at least 85%, more preferably at least 90%, more preferably at least 95%, more preferably at least 97%, more preferably at least 98%, and most preferably at least 99% sequence identity to the full length of SEQ ID NO: 1 (OLPVRII).

Generally, an amino acid sequence has "at least x % identity" with another amino acid sequence, e.g. SEQ ID NO: 1 above, when the sequence identity between those to aligned sequences is at least x % over the full length of said other amino acid sequence. Similarly, an amino acid sequence has "at least x % similarity" with another amino acid sequence, e.g. SEQ ID NO: 1 above, when the sequence similarity between those two aligned sequences is at least x % over the full length of said other amino acid sequence, e.g. SEQ ID NO: 1.

Such alignments can be performed using for example publicly available computer homology programs such as the "EMBOSS" program provided at the EMBL homepage at http:/ www.ebi.ac.uk/Tools/psa/emboss_needle/, using the default settings provided therein. Further methods of calculating sequence identity or sequence similarity percentages of sets of amino acid acid sequences are known in the art.

The light-gated pentameric anionic channel OLPVRII has seven transmembrane a- helices (A-G) and a co-factor retinal covalently bound to the residue 195 Lysine in SEQ ID NO: 1 via the Schiff base. The helicies A and B of one protomer contact the helicies A' and B' of the other protomer forming a pentamer.

The light-gated pentameric channel of the present disclosure is a membrane protein with at least 5 transmembrane helices, which is capable of binding a light-sensitive polyene. Transmembrane proteins with 6 or 7 transmembrane helices are preferable. Transmembrane proteins with more than 7 helices, for example 8, 9 or 10 transmembrane helices, are, however, also encompassed. Furthermore, the invention covers transmembrane proteins which in addition to the transmembrane part include N- and/or C-terminal sequences, where the terminal sequences can extend into the inside and/or outside of the lumen enclosed by the membrane or can also be arranged on the membrane surface. The length of the N- and/or C- terminal sequences is in principle subject to no restriction; however, light-gated pentameric channels with N-terminal sequences with 1 to 1000 amino acids, preferably 1 to 500, especially preferably 5 to 50 amino acids, are preferred. Independently of the length of the N- terminal sequences, the C-terminal sequences preferably comprise 1 to 1000 amino acids, preferably 1 to 500, especially preferably 5 to 50 amino acids. In a preferred embodiment, the light-gated pentameric channel is not truncated at the N and or C-terminus. The concept of the transmembrane helix is well known to the skilled person. The light-gated pentameric channel of the present disclosure is, in general, an a-helical protein, typically the helicies comprise 20 to 25 amino acids. However, depending on the nature of the membrane, which can be a natural membrane, for example a cell or plasma membrane, or also a synthetic membrane, the transmembrane segments can also be shorter or longer. For example, transmembrane segments in artificial membranes can comprise up to 30 amino acids, but on the other hand also only a few amino acids, for example 12. Most preferably, the light-gated pentameric channel has seven transmembrane a-helices (A-G) and a co-factor retinal covalently bound to 195 Lysine via the Schiff base. The helicies A and B of one protomer contact the helicies A' and B' of the other protomer forming a pentamer.

Preferably, the light-gated pentameric channel only comprises (semi)-conservative substitutions as compared to SEQ ID NO: 1. Conservative substitutions are those that take place within a family of amino acids that are related in their side chains and chemical properties. Examples of such families are amino acids with basic side chains, with acidic side chains, with non-polar aliphatic side chains, with non-polar aromatic side chains, with uncharged polar side chains, with small side chains, with large side chains etc. Typical semiconservative and conservative substitutions are:

Furthermore, the skilled person will appreciate that glycines at sterically demanding positions should not be substituted and that proline should not be introduced into parts of the protein which have an a-helical or a b-sheet structure. In particular, the light-gated pentameric channel should only comprise (semi)-conservative substitutions at the position corresponding to L24, F28, R29, E42, K195, E26, W203, R36, H37, N40 of SEQ ID NO: 1. In other words, the light-gated pentameric channel preferably comprises an "L" at position 24, an "F" at position 28, etc.

In an even more preferred embodiment, the light-gated pentameric channel comprises an amino acid sequence SEQ ID NO: 1 (OLPVRII). In a most preferred embodiment, the light-gated pentameric channel consists of an amino acid sequence SEQ ID NO: 1 (OLPVRII).

The term "pentameric" as used herein is intended to mean that when the channel is imbedded into a lipid environment, it forms a pentamer. The structural requirement of "forms a pentamer" can be tested, like illustrated in the example 1 below, using the following assay. Dissolve 20 mg of phospholipids (POPC:POPS lipids (l-palmitoyl-2-oleoyl-glycero- 3-phosphocholine and l-palmitoyl-2-oleoyl-sn-glycero-3-phospho-L-serine from Avanti Polar Lipids, USA) in 4:1 weight ratio) and dry them under a stream of N2 in a glass vial. Remove residual solvent by overnight incubation under vacuum. Then resuspend the dried lipids in 1 ml of 50 mM Na/Na-Pi pH 8.0, 100 mM NaCl buffer supplemented with 2% (w/v) sodium cholate, clarify the mixture by sonication and add a solubilized purified protein at a protein/lipid ratio of 5:100 (w/w). Remove the detergent by incubation with detergent absorbing beads (Amberlite XAD 2, Supelco). To ensure the ample removal of the detergent, add the beads to the suspension 3-4 times. The volume of the beads should be about one third of the total volume of liposome mixture. Incubate the suspension with the beads at least 2 hours stiring. The last incubation should be prolonged to 16 hours. Then place glutaraldehyde solution directly to the bottom of the well of 24-well Falcon Tissue Culture Plate (Thermo Fisher Scientific, Germany). Load the protein on the microbridges, cover the set up by cover slips and incubate it at 4°C. Vary the concentration of the protein and glutaraldehyde to find the optimal condition for incomplete cross-linking of the proteins. This allows one to observe in the denaturing conditions of SDS-PAGE the array of artificial multimers formed by several crosslinked protomers of the natural multimer. Using the molecular weight of the highest visible multimer and the number of the appered multimers, one can make a conclusion about an aggregate state of the protein in a sample. In particular, for the solubilized in DDM purified OLPVRII protein (see the Example 1) reconstituted into POPC/POPG mixture we incubated the sample for 8 hours, the concentration of the protein in liposomes was 1 mg/ml and glutaraldehyde - 10%.

In addition, pentameric form of the protein can be observed in the crystal structure with five protomers in an assymetric unit, in case the crystals are grown using in meso approach. In such crystals, the assymetric unit can contain also ten molecules per unit. The "in meso approach" is known in the art. Given the structure of the protein obtained, one can check the probability of the pentameric assembly formation or its stablity by the molecular dynamics simulations similarly to the protocol discribed in Example 2.

The terms "anionic" and "cationic" as used herein is intended to mean that when the protein is imbedded into a lipid environment, there is a possibility that it will transfer anions or cations, respectively, from one side of the membrane to the other. Given the structure of the membrane protein, the functional requirement of being either "anionic" or "cationic" channel can be tested using the following protocol that can be reproduced by a trained specialist in the art.

Carry out all-atom MD simulations of the protein using the CHARMM36m force field with the TIP3P water model. The simulation box should contain an equilibrated 1-palmitoyl- 2-oleoyl-sn-glycero-3-phosphocholine (POPC) lipid bilayer, surrounded by 200 mM NaCl aqueous solution. Model the protein using the residues from the obtained structure and embed it into the bilayer using g_membed. Assign a proton donor a protonated state, all other amino acids model in their default ionization state to reflect the most probable state at neutral pH based on pKa calculations using PROPKA 3.1. Perform simulations using GROMACS 2018 in the NPT ensemble with periodic boundary conditions and an integration time step of 2 fs. Maintain the temperature at 310 K using the velocity-rescaling thermostat; keep pressure at 1 bar using the semi-isotropic Parrinello-Rahman barostat as described in Example 3. Trancate Lennard- Jones interactions at 12 A with a force switch smoothing function from 10-12 A. Calculate electrostatic interactions using particle mesh Ewald method and a real space cutoff of 12 A. Equilibrate simulation systems with position restraints on the protein heavy atoms for 400 ns, followed by ~20 ns with backbone-only position restraints. Perform 2-3 independent production runs of 10 ps length for the protein with mutations imitating the opened pore, i.e. with a smaller amino acid substitutions known in the art. In particular, for OLPVRII F24A/L28A mutant can be used. If the channel pore is blocked by hydrocarbon or other molecule, it should be removed for the simulations. For consistency, all analyses of the simulations should be based on the time windows from at least 2000 ns in each simulation. Analyze water and ion density distributions in the molecular dynamics trajectories using GROmaps and MD Analysis and calculate electrostatic potential distributions from the charge densities in the molecular dynamics simulations using the Poisson equation as implemented in PMEPot. Analyze water, Na⁺, and CF distributions in the free molecular dynamics simulations to calculate potential of mean force profiles along the pore axis as indicated in Example 3. Calculate standard deviations of energies and pore radii using block bootstrap sampling. Spontaneous water and ion flux can be observed into the central pore from both membrane sides. However, free-energy profiles calculated for water, Na⁺, and CF permeation should show the highest barrier, which can be anion or cation selective, respectively. Reduction of the respective side chain volumes at the barier by alanine substitution should increase the hydration of the constriction, lower the energy barrier for ion permeation, and lead to an overall Cl - or Na⁺-selective ion pathway in the mutant.

In addition, an anion or cation channeling activity of the pentameric channel of present disclosure can be directly demonstrated by the experimental methods known in the art. As an instance, the light-gated pentameric channel can be further characterized electrophysiologically by using patch-clamp measurements in the whole cell configuration.

In a related aspect, the present disclosure also provides a nucleic acid construct, comprising a nucleotide sequence coding for the light-gated pentameric channel as described above. To ensure optimal expression, the coding nucleotide sequence can also be suitably modified, for example by adding suitable regulatory sequences and/or targeting sequences and/or by matching of the coding DNA sequence to the preferred codon usage of the chosen host. In a particularly preferred embodiment, the nucleotide sequence is codon-optimized for expression in human cells. For example, the nucleotide sequence may have the sequence shown in SEQ ID NO: 16. The targeting sequence may encode an N-terminal extension targeting the light-gated pentameric channel to a particular site or compartment within the cell, such as to the synapse or to a post-synaptic site, to the axon-hillock, or the endoplasmic reticulum. The nucleic acid may be combined with further elements, e.g., a promoter and a transcription start and stop signal and a translation start and stop signal and a polyadenylation signal in order to provide for expression of the sequence of the mutant light-gated pentameric channel of the present disclosure. The promoter can be inducible or constitutive, general or cell specific promoter. An example of a cell-specific promoter is the mGlu6-promotor specific for bipolar cells. In particular embodiments, the coding sequence of the light-gated pentameric channel is under the control of a neuronal cell specific human promoter, preferably the human synapsin promoter. Selection of promoters, vectors and other elements is a matter of routine design within the level of ordinary skill in the art. Many such elements are described in the literature and are available through commercial suppliers.

Also disclosed is an expression vector, comprising the nucleotide sequence coding for the mutant light-gated pentameric channel or the nucleic acid construct as disclosed herein, wherein the nucleotide sequence is optimized for expression in human cells. In a preferred embodiment, the vector is suitable for gene therapy, in particular wherein the vector is suitable for virus-mediated gene transfer, i.e. wherein the vector is a viral vector. The term "suitable for virus-mediated gene transfer" means herein that said vector can be packed in a virus and thus be delivered to the site or the cells of interest. Examples of viruses suitable for gene therapy are retroviruses, adenoviruses, adeno-associated viruses, lentiviruses, pox viruses, alphaviruses, rabies virus, semliki forest virus and herpes viruses. These viruses differ in how well they transfer genes to the cells they recognize and are able to infect, and whether they alter the cell's DNA permanently or temporarily. However, gene therapy also encompasses non-viral methods, such as application of naked DNA, lipoplexes and polyplexes, and dendrimers.

The resulting nucleic acid sequence may be introduced into cells e.g. using a virus as a carrier or by transfection including e.g. by chemical transfectants (such as Lipofectamine, Fugene, etc.), electroporation, calcium phosphate co-precipitation and direct diffusion of DNA. A methods for transfecting a cell are known in the art and may be adapted to the respective recipient cell. In general, the selection of a suitable vector or plasmid depends on the intended host cell.

Therefore, the present disclosure also pertains to a mammalian cell expressing the light-gated pentameric channel as disclosed herein, with the proviso that the mammalian cell is not a human embryonic cell or a cell capable of modifying the germ line genetic identity of human beings. Similarly, the present disclosure provides a mammalian cell comprising the nucleic acid construct or the expression vector disclosed herein.

The incorporation of the light-gated pentameric channel of the present disclosure into the membrane of cells which do not express the corresponding channels in nature can, for example, be simply carried out using known procedures of recombinant DNA technology. In preffered embodiment, the DNA coding for this pentameric channel is firstly incorporated into a suitable expression vector, e.g. a plasmid, a cosmid or a virus, the target cells are then transformed with this, and the protein is expressed in this host. Next, the cells are treated in a suitable manner, e.g. with retinal, in order to enable the linkage of a Schiffs base between protein and retinal.

The expression of the light-gated pentameric channel of the present disclosure can be advantageously effected in certain mammalian cell systems. In a preferred embodiment the mammalian cell is a neuroblastoma cell, in particular NG108-15; a HEK293 cell; a COS cell; a BHK cell; a CHO cell; a myeloma cell; or a MDCK cell. In another preferred embodiment, the mammalian cell is an electrically excitable cell. It is further preferred that the cell is a hippocampal cell, a photoreceptor cell; a retinal rod cell; a retinal cone cell; a retinal ganglion cell; a bipolar neuron; a ganglion cell; a pseudounipolar neuron; a multipolar neuron; a pyramidal neuron, a Purkinje cell; or a granule cell.

A photoreceptor cell is a specialized neuron found in the retina that is capable of phototransduction. The two classic photoreceptors are rods and cones, each contributing information used by the visual system. A retinal ganglion cell is a type of neuron located near the inner surface of the retina of the eye. These cells have dendrites and long axons projecting to the protectum (midbrain), the suprachiasmatic nucleus in the hypothalamus, and the lateral geniculate (thalamus). A small percentage contribute little or nothing to vision, but are themselves photosensitive. Their axons form the retinohypothalamic tract and contribute to circadian rhythms and pupillary light reflex, the resizing of the pupil. They receive visual information from photoreceptors via two intermediate neuron types: bipolar cells and amacrine cells. Amacrine cells are intemeurons in the retina, and responsible for 70% of input to retinal ganglion cells. Bipolar cells, which are responsible for the other 30% of input to retinal ganglia, are regulated by amacrine cells. As a part of the retina, the bipolar cell exists between photoreceptors (rod cells and cone cells) and ganglion cells. They act, directly or indirectly, to transmit signals from the photoreceptors to the ganglion cells.

The cell may be isolated (and genetically modified), maintained and cultured at an appropriate temperature and gas mixture (typically, 37°C, 5% C02), optionally in a cell incubator as known to the skilled person and as exemplified for certain cell lines or cell types in the examples. Culture conditions may vary for each cell type, and variation of conditions for a particular cell type can result in different phenotypes. Aside from temperature and gas mixture, the most commonly varied factor in cell culture systems is the growth medium. Recipes for growth media can vary in pH, glucose concentration, growth factor and the presence of other nutrient components among others. Growth media are either commercially available, or can be prepared according to compositions, which are obtainable from the American Tissue Culture Collection (ATCC). Growth factors used for supplement media are often derived from animal blood such as calf serum. Additionally, antibiotics may be added to the growth media. Amongst the common manipulations carried out on culture cells are media changes and passaging cells. Thus, the presently disclosed light-gated pentameric channel could be particularly useful as a research tool, such as in a non-therapeutic use for light-stimulation of electrically excitable cells, in particular neuron cells.

The retinal or retinal derivative is necessary for the functioning of the light-gated pentameric channel of the present disclosure. In general, it is produced by the cell to be transfected with said pentameric channel or is introduced artificially into the cultivation media. Depending on its conformation, the retinal may be all-trans retinal, 11-cis-retinal, 13- cis-retinal, or 9-cis-retinal. It is also contemplated that the light-gated pentameric channel of the present disclosure may be incorporated into vesicles, liposomes or other artificial cell membranes. Accordingly, also disclosed is a channel, comprising the light-gated pentameric channel of the present disclosure, and a retinal or retinal derivative. In most preffered embodiments, the retinal may be all-trans retinal, 11 -cis-retinal, 13-cis-retinal, or 9-cis- retinal. However, the retinal derivative could be also selected from the group consisting of 3,4-dehydroretinal, 13-ethylretinal, 9-dm-retinal, 3- hydroxyretinal, 4-hydroxyretinal, naphthylretinal; 3,7,11 -trimethyl-dodeca-2,4,6,8, 10- pentaenal; 3,7-dimethyl-deca-2,4,6,8- tetraenal; 3,7-dimethyl-octa-2,4,6-trienal; and 6-7 rotation-blocked retinals, 8-9 rotation- blocked retinals, and 10-11 rotation-blocked retinals.

Finally, there are a number of diseases in which, e.g., the natural visual cells no longer function, but all nerve connections are capable of continuing to operate. Today, attempts are being made in various research centers to implant thin films with artificial ceramic photocells on the retina. These photocells are intended to depolarise the secondary, still intact cells of the retinal and thereby to trigger a nerve impulse (bionic eyes). The deliberate expression of mutant light-gated pentameric channel according to the present disclosure in these ganglion cells, amacrine cells or bipolar cells could be a much better solution and could enable greater three-dimensional visual resolution. Therefore, the present disclosure also contemplates the light-gated pentameric channel, the nucleic acid construct, the expression vector, the mammalian cell, or the liposome according to the present disclosure for use in medicine. In particular, it is contemplated that the presently disclosed light-gated pentameric channel can be used for restoring auditory activity in deaf subjects or recovery of vision in blind subjects.

In addition, due to its estimated electrophysiological capabilities, the light-gated pentameric channel of the present disclosure is contemplated to be suitably applied in treating or alleviating neurological injury, brain damage, seizure, or a degenerative neurological disorder, such as Parkinson's disease and Alzheimer's disease. In all these treatment cases, the light-gated pentameric channel may be delivered by way of liposomes, and more preferably by way of administering the nucleic acid construct or the expression vector of the present disclosure to subject to be treated.

Further described are non-human animals which comprise a cell according to the present disclosure, i.e. a cell which functionally express the light-gated pentameric channel according to the present disclosure, e.g. in a cell such as a neuron. In preferred embodiments, the cell is an endogenous cell. The non-human animal may be any animal other than a human. In a preferred embodiment, the non-human animal is a vertebrate, preferably a mammal, more preferably a rodent, such as a mouse or a rat, or a primate.

In particular, some model organisms are preferred, such as Caenorhabditis elegans, Arbacia punctulata, Ciona intestinalis, Drosophila, usually the species Drosophila melanogaster, Euprymna scolopes, Hydra, Loligo pealei, Pristionchus pacificus, Strongylocentrotus purpuratus, Symsagittifera roscoffensis, and Tribolium castaneum. Among vertebrates, these are several rodent species such as guinea pig, hamster, mouse, and rat, as well as other species such as chicken (Gallus gallus domesticus), cat (Felis cattus), dog (Canis lupus familiaris), Lamprey, Japanese ricefish, Rhesus macaque, Sigmodon hispidus, zebra finch (Taeniopygia guttata), pufferfish (Takifugu rubripres), african clawed frog (Xenopus laevis), and zebrafish. Also preferred are non-human primates, i.e. all species of animals under the order Primates that are not a member of the genus Homo, for example rhesus macaque, chimpanzee, baboon, marmoset, and green monkey. However, these examples are not intended to limit the scope of the invention. In any case, it is noted that those animals are excluded, which are not likely to yield in substantial medical benefit to man or animal and which are therefore not subject to patentability under the respective patent law or jurisdiction. Moreover, the skilled person will take appropriate measures, as e.g. laid down in international guidelines of animal welfare, to ensure that the substantial medical benefit to man or animal will outweigh any animal suffering.

Finally, also non-therapeutic, or ex vivo, or in vitro uses of the light-gated pentameric channel of the present disclosure are contemplated. For example, the light-gated pentameric channel of the present disclosure may be advantageously applied 1 ) for light-stimulation of electrically excitable cells, 2) for transporting ions over a membrane, 3) or as an optogenetic tool. In the following, the present invention is illustrated by figures and examples which are not intended to limit the scope of the present invention.

DESCRIPTION OF THE FIGURES

Figure 1. A phylogenetic tree of microbial rhodopsins. Different microbial rhodopsins groups and their functions are indicated. Viral rhodopsins and OLPVRII are highlighted red.

Figure 2. Investigation of the OLPVRII pentamerization. A. Typical SEC profile of wild type OLPVRII is shown in blue. The elution profile of E26A/R36A/W203A is shown in red. The fractions pooled are indicated. The elution profiles clearly illustrate the distribution of the protein between monomeric and pentameric fractions. B. SDS-PAGE analysis of SEC fractions and results of crosslinking experiments. Lane 1: Monomeric fraction after gel- filtration; lane 2: Pentameric fraction after gel-filtration; lane 3: molecular weight marker with molecular weights indicated; lane 4: pentameric protein at 2 mg/ml protein, crosslinked with 10% glutaraldehyde, 8 hours incubation; lane 5: monomeric protein at 1 mg/ml reconstituted into POPC/POPG mixture crosslinked with 10% glutaraldehyde, 8 hours incubation.

Figure 3. pH titration of solubilized OLPVRII. A. Absorption spectra at pH between 7.58 and 12.93. B. Absorption spectra at pH between 7.58 and 1.93. C. Difference spectra at pH from 8.05 to 12.93. D. Difference spectra at pH from 8.05 to 1.93. E. Fitting at lower pH values of the dependence of the retinal absorbance maximum on the pH value with sigmoidal curve. The parameters of the fit gave the pKa of the proton acceptor. F. Fitting at higher pH values of dependence of the absorbance at 514 nm on the pH value with sigmoidal curve. The parameters of the fit gave the pKa of the Schiff base.

Figure 4. Functional characterization of OLPVRII. A. BLM measurements of OLPVRII in E. coli polar lipid liposomes. Red curve shows the electrical signal on BLM under illumination without addition of protonophore, while blue curve represents the signal under the same conditions, but after addition of 1799 and monensin. Illumiation onset is indicated by gray (light off) and white (light on) colors. B. Time resolved-experiment. Electrical signal (grey) was recorded after laser flash and fitted with the exponential decay curves (red). C. Schematic representation of photocycle kinetics of the OLPVRII reconstituted into soy bean liposomes. Lifetimes and absorption maxima of the intermediate states are indicated. D. Time traces of the absorption changes of OLPVRII measured at 410 nm, 525 nm, and 550 nm.

Figure 5. Intermediate states spectra of OLPVRII. The absolute spectra of 7 intermediate state determined during transient kinetics data treatments are shown with red color, while ground state spectra is represented with black line.

Figure 6. Photocycle kinetics of OLPVRII mutants. Time dependence of absorption differences at selected wavelengths of the wild type OLPVRII and its mutants E42Q and D75N colored red, blue and green, respectively.

Figure 7. Crystal packing and examples of 2Fo-Fc electron density maps A. Example of OLPVRII crystals. B. Crystal packing of OLPVRII. C. Example of 2Fo-Fc electron density maps around retinal of protomer B. Maps a contoured at the level of 1.5s. D. Simulated annealing omitted electron density map built omitting the retinal and Lysl95 residues. The map is contoured at the level of 3.0 s. E. Example of 2Fo-Fc electron density maps in the RSB region of protomer E. Maps a contoured at the level of 1.5s. F. Example of 2Fo-Fc electron density maps around Glu42 of protomer B. Maps a contoured at the level of 1.0s.

Figure 8. Architecture of OLPVRII pentamer and interprotomer contacts. A. View from the cytoplasmic side. Surface representation of the pentamer. Central pore is contoured by a red circle. B. View from the cytoplasmic side. Cartoon representation of the pentamer. Retinal cofactor is colored cyan. C. Detailed view of the main region of interprotomer contacts. Protomers C, D and E are colored gray. D. Side view of the pentamer. One protomer is hidden for clarity. Cavity inside the pentamer was calculated using HOLLOW and is colored light blue, the hydrophobic membrane core boundaries are shown with solid horizontal lines. E, F. Detailed view of the interprotomer contacts.

Figure 9. OLPVRII pentamer and lipids paving the pentamer. A. View from the cytoplasmic side. B. View from the extracellular side. Lipid fragments are colored violet. C. View from the cytoplasmic side on the cartoon representation of the pentamer with surrounding lipids. Monoolein (MO) molecules are shown with spheres. Lipidic fragments deepened between the protomers (helices A and G from one protomer and B’, C’ and D’ of the neighbouring protomer) are colored orange and shown with spheres. D. Detail view of the MO stabilization hydrogen bonds. E. Side view of the pentamer. Monoolein (MO) molecule is shown with spheres. Lipidic fragment deepened between the protomers is colored orange and shown with spheres. Hydrophobic/hydrophilic borders of the membrane are shown with gray lines. F. Detail side view of the residues comprising pocket for lipidic molecule deepened between protomers. One protomer is colored green.

Figure 10. Sequence alignment of viral rhodopsins from group II including the representative TARA metagenomic sequences. Highly conservative residues are indicated with red color, 6-letter motif residues are indicated with magenta color. Residues involved into formation of the putative channel pore are colored blue.

Figure 11. Architecture of the pentamers of microbial rhodopsins.

Figure 12. Central part of OLPVRII and other rhodopsins. A. OLPVRII. B. bR (PDB ID 1C3W). C. NsXeR (PDB ID 6EYU). D. BPR HOT75 (PDB ID 4KLY). E. BPR Medl2 (PDB ID 4JQ6). F. KR2 (PDB ID 4XTO). Cavities are shown in blue. One protomer of each oligomer is hidden for clarity.

Figure 13. Central pore inside OLPVRII pentamer. A. Overall central pore structure. One protomer is hidden for clarity. The lipid fragment is colored violet. B. View from the cytoplasmic side on the water pentagon and interaction network between water molecules and the pore-lining OLPVRII residues. C. Side view of the pore vestibule. Example of 2Fo-Fc electron density map is shown around the water pentagon and the hydrocarbon chain. The map is contoured at the level of 1.2s. The hydrocarbon chain is colored violet.

Figure 14. Hydrophobicity of the OLPVRII pentamer surface. A. View from the cytoplasmic side. B. View from the extracellular side. C. Side section view of the central pore. Residues forming the pore vestibule are shown with sticks (Arg29, Leu28 and Phe24). Lipid fragments are shown with sticks and are colored red. Water molecules are shown with red spheres. Red color of the residue indicates its hydrophobicity, while white color indicates its hydrophilicity.

Figure 15. Comparison of the inner pore of OLPVRII (left) and GLIC (right). For GLIC, PDB ID: 4HFI was used. Extracellular domain is colored green. Hydrocarbon chains are colored cyan.

Figure 16. Structural alignment of OLPVRII (yellow) and bR (magenta) protomers. A, B. Side views. C, D. View from the extracellular and cytoplasmic sides, respectively. The most notable differences are marked with black arrows.

Figure 17. Structure of OLPVRII protomer, retinal-binding pocket and extracellular part. A. Overall side view of protomer A. Helices F and G are hidden for clarity. Hydrophobic- hydrophilic boundaries of the membrane are shown with gray lines. B. Detailed view of the extracellular part. C. Detailed view of the retinal-binding pocket. Residues comprising the walls of the pocket are colored teal. D. Detailed view of the RSB region. Helices A and B are hidden for clarity. Cavities inside the protein protomer are colored pink. Lysl95 and covalently bound retinal are colored cyan.

Figure 18. Retinal binding pocket of A. OLPVRII; B. bR (PDB ID: 1C3W). Residues comprising retinal binding pocket are colored teal, retinal molecules are colored cyan.

Figure 19. Extracellular part of the Schiff base region of A. OLPVRII; B. bR (PDB ID: 1C3W). Retinal molecules are colored cyan, cavities are colored pink. Helices and C and G are annotated with capital bold letters. Cavities inside the protein protomers are colored pink.

Figure 20. Extracellular part of A. OLPVRII; B. bR (PDB ID: 1C3W), C. ChR2 (PDB ID: 6EID). Retinal molecules are colored cyan, cavities inside protein protomers are colored pink.

Figure 21. Cytoplasmic part of A. OLPVRII; B. bR (PDB ID: 1C3W). Retinal molecules are colored cyan, cavities inside protein protomers are colored pink. Helices A, B and C are annotated with capital bold letters.

Figure 22. Cytoplasmic part of the OLPVII. Gray color indicates the cytoplasmic space, which concaves inside the protein down to Leu39 side chain. Cavities inside the protein protomer are colored pink.

Figure 23. The cytoplasmic part of OLPVRII and its connection to the central pore. A. Side view. One protomer is hidden for clarity. B. View from the cytoplasmic side. The central pore is shown with the blue surface. The red circle indicates the R29 ring of the central pore. The brown arrows show the putative sequence of structural rearrangements transduced from RSB to the pore interface. Cavities inside the protein protomer are colored pink.

Figure 24. MD simulations reveal a closed anion permeation pathway. A. Averaged water densities in absence or presence of the hydrocarbon chain along the central axis from unguided MD simulations contoured at 1.2s in side view. Helices A and B of three subunits are shown in cartoon representation and pore-lining residues as sticks. B. Na+ (blue) and Cl- (red) densities for WT and F24A-L28A mutant proteins in absence of the detergent contoured at 0.2s. C. Pore radii along the pore axis (positions relative to the protein center of mass) for the conformational ensemble sampled in WT simulations without hydrocarbon chain. Side chain positions of residues Phe24, Leu28, and Arg29 are shown. D. Electrostatic potential distribution calculated from the MD simulations mapped onto the solvent-accessible surface of OLPVRII in side view. The two front subunits are hidden for clarity. E. Potential of mean force profiles for water, Na+, and Cl- permeation along the pore. F. Histograms of the number of water molecules observed within the hydrophobic gate region for different mutations.

Figure 25. Energetics of ion and water permeation in OLPVRII. A. Potential of mean force profiles for water, Na⁺, and Cl permeation along the WT OLPVRII pore with and without bound hydrocarbon chains in the pore, and for the F24A, L28A, R29E, F24/L28A, and F24/L28A/R29E mutants without the hydrocarbon chains. B. Electrostatic potential distribution calculated from the MD simulations mapped onto a slice through the pore center parallel to the pore axis for wild type OLPVRII (without hydrocarbon chains).

Figure 26. Comparison of transmembrane parts of inner pores of OLPVRII and bestrophin ligand-gated chloride channel (BEST1). As a closed and open states of BEST1 we used models with PDB IDs 6N26 and 6N28, respectively. Membrane core boundaries are shown with gray lines. Residues forming the narrowest parts of the channels are shown with sticks.

Table 1. Crystallographic data collection and refinement statistics.

DESCRIPTION OF THE SEQUENCES

SEQ ID NO: 1 (OLPVRII; UniProtID F2Y2Z0; residues of the oligomerization interface (E26, R36, H37, N40 and W203) and the central pore vestibule (F24, L28 and R29) are in bold) MSDLIE YSF YLTY AFLMTT GTITFIE ALRTKNES VRHILNLETCIS V V A AFF Y S NFIGKLEHINYEEINLNRYVDWAITTPIMLLVLVLAFRVNQTNKAMVKFSDFMIILG MN Y GMLGT GYLGDIGVIHKTMGTVLGFLFFGGLF YKLNTLRTSN ASNDLL Y GAFF VL W AL YGVF YQMEQLPRNV G YN VLDLFSKCFV GI YF WAF YAKIFTLE

SEQ ID NO: 2 (human codon-optimized OLPVRII)

AT G AGCG ACCTGATCG AGTAC AGCTT CT ACCTG ACCT ACGCCTTCCT GAT GACCACCGGCACCATCACCTTTATCGAGGCCCTGCGGACCAAGAACGAGAGCG TGCGGCACATCCTGAACCTGGAAACCTGCATCTCTGTGGTGGCCGCCTTCTTCTA CTCCAACTTCATCGGCAAGCTGGAACACATCAACTACGAGGAAATCAACCTGA ACCGCTACGTGGACTGGGCCATCACCACACCTATCATGCTGCTGGTGCTGGTCC TGGCCTTCAGAGTGAACCAGACCAACAAGGCCATGGTCAAGTTCAGCGACTTCA TGATCATCCTGGGCATGAACTACGGCATGCTCGGCACAGGCTACCTGGGCGATA TCGGCGTGATCCACAAGACCATGGGAACCGTGCTGGGCTTCCTGTTCTTCGGCG GCCTGTTCTACAAGCTGAACACCCTGAGAACCAGCAACGCCTCCAACGACCTGC TGTACGGCGCCTTCTTTGTGCTGTGGGCTCTGTATGGCGTGTTCTACCAGATGGA ACAGCTGCCCAGAAACGTGGGCTACAACGTGCTGGACCTGTTCAGCAAGTGCTT CGTGGGCATCTACTTTTGGGCCTTCTACGCCAAGATCTTCACCCTGGAA

EXAMPLES

Example 1 - Characterization of OLPVRII Expression and purification

We optimized the nucleotide sequence of OLPVRII (UniprotID F2Y2Z0) for E. coli expression using the GeneOptimizer™ software (Life Technologies, USA) and synthesized it commercially (Eurofms). We introduced the gene into the pSCodon 2.1 expression vector (Delphi Genetics) via Ndel and Xhol restriction sites and appended at the 3' terminus additional GSGIEGRSGAPHHHHHHHH* tag, which was used for metal-affinity chromatography purification and contains FXa cleavage site. The same gene was introduced into pEKT expression vector, pET vector derivative (Novagen), via Xbal and Xhol restriction sites. We expressed the protein as described previously (Bratanov, D. et al PLoS One 10, (2015)) with slight modifications. We transformed E. coli SE1 (Delphi Genetics) or C41(DE3) (Lucigen) cells with the expression plasmid and grew them at 37 °C in 2L shaking at 120 rpm baffled flasks in an autoinducing medium ZYP-5052 containing required antibiotic. At optical density OD600 of 1.0 we decreased the temperature to 20°C and supplemented the media with 10 mM all-trans-retinal. After 14h cultivation the cells were disrupted in M-110P Lab Homogenizer (Microfluidics, USA). Then we isolated total membranes by ultracentrifugation, resuspended them in 20 mM Tris-HCl pH 8.0, 100 mM NaCl and solubilized overnight in DDM (n-Dodecyl-P-D-Maltopyranoside, Anagrade, Anatrace, USA or crystallography grade, Cube-Biotech, Germany). The insoluble fraction was removed by ultracentrifugation (90000g, lh, 4°C). The supernatant was loaded on Ni- NTA column (Qiagen, Germany) and after washing the column with 5 volumes of 50 mM NaH2P04/Na2HP04 pH 8.0, 100 mM NaCl, 50 mM imidazole, 0.3% DDM buffer we eluted the protein in a buffer containing 50 mM NaH2P04/Na2HP04 pH 8.0, 100 mM NaCl, 0.5 M imidazole and 0.3% DDM. Eluted protein was concentrated using microfiltration (MW cutoff 30 kDa, Amicon Ultra, Millipore, Germany). Then we applied it to 30 ml Superdex 200i (GE Healthcare, Germany) column equilibrated with 50 mM NaH2P04/Na2HP04 pH 8.0, 100 mM NaCl, 0.2% DDM, and pooled a peak of colored functional protein. The total yield of the functional purified protein was approximately 2-3 mg per 5 liters of culture. After size exclusion chromatography, the protein was divided into two fractions, “monomers” and “pentamers”, according to the protein size appearing in the elution profile (Fig. 2). Pentamerization was confirmed by crosslinking experiments with the solubilized protein. For crystallization and functional tests, we used the monomeric fraction unless otherwise indicated. Finally, using microfiltration we concentrated homogeneous protein to 20 mg/ml, froze it in liquid nitrogen and stored it at -80°C until used for crystallization or functional measurements. Typically, we obtained the monomeric fraction of the protein with the peak ratio A280/A520 about 1.0- 1.3 and pentameric fraction with the peak ratio of about 1.3 -1.7. Generally, peak ratio of the pentameric fraction is slightly higher probably due to aggregate impurities. Low peak ratio indicates a very high purity of the samples. SE1 strain showed an instability with OLPVRII-bearing pSC plasmid and its production was discontinued, therefore we used C41 strain bearing OLPYRII gene in pEKT plasmid. Protein from C41 did not show considerable differences to SE1 -expressed protein, therefore we used it on later stages of the study.

Incorporation into lipid environment

We dissolved 20 mg of phospholipids (asolectin from soybean, Sigma-Aldrich) in CHC13 (Chloroform ultrapure, Applichem Panreac) and dried them under a stream of N2 in a glass vial. Residual solvent was removed by overnight incubation under vacuum. Then we resuspended the dried lipids in 1 ml of 0.15 M NaCl buffer supplemented with 2% (w/v) sodium cholate, clarified the mixture by sonication and added OLPVRII at a protein/lipid ratio of 5:100 (w/w). The detergent was removed by incubation with detergent-absorbing beads (Amberlite XAD 2, Supelco). To ensure the ample removal of the detergent, we added the beads to the suspension 3-4 times. The volume of the beads was about one third of the total volume of liposome mixture. We incubated the suspension with the beads at least 2 hours stiring. The last incubation wass prolonged to 16 hours. Then, the mixture was dialyzed against 0.15 M NaCl (adjusted to a desired pH) at 4°C for 1 day with several buffer changes to obtain desired pH. The described protocol leads to the formation of unilamellar vesicules with the size of 100-300 nm depending on the lipid composition that was confirmed by dynamic light scattering. We used obtained liposomes for the spectroscopic characterisation of OLPVRII.

We prepared liposomes for BLM measurements with following modifications: a) E. coli polar lipids (Anatrace, Affymetrix, USA) were used instead of asolectin as a host lipid for liposomes; b) to achieve Na⁺- and Cl -free environment lipids were resuspended in 100 mM Tris-HEPES pH 7.4 buffer supplied with 2% (w/v) cholic acid in the same buffer and liposomes were not dialised against NaCl; c) to improve the signal-to-noise ratio lipid-to- protein ratio was increased to 20:100.

The OLPVRII-containing nanodiscs were assembled using the following protocol (Ritchie, T. K. et al. Methods in Enzymology 464, 211-231 (2009)). 1,2-dimyristoyl-sn- glycero-3-phosphocholine (DMPC, Avanti Polar Lipids, USA) and an elongated in- house produced MSP1E3 version of apolipoprotein-1 were used as a lipid and scaffold protein, respectively, the molar ratio during the assembly was DMPC:MSPlE3:OLPVRII = 100:2:3. The empty nanodiscs and aggregates were removed by size exclusion chromatography.

Crosslinking For crosslinking experiments liposomes from E. coli polar lipids turned out to be not suitable, as a natural mixture of lipids contains considerable amount of NFh-groups that block the crosslinking reaction. Therefore we used an artificial mixture of POPC:POPS lipids (1- palmitoyl-2-oleoyl-glycero-3-phosphocholine and 1 -palmitoyl-2-oleoyl-sn-glycero-3- phospho-L-serine from Avanti Polar Lipids, USA) in 4:1 weight ratio. The liposomes were prepared as described above.

We performed crosslinking according to the following protocol (Fadouloglou, V. E. et al. Anal. Biochem. 373, 404-406 (2008)) with the following modifications. We placed glutaraldehude solution directly to the bottom of the well of 24-well Falcon Tissue Culture Plate (Thermo Fisher Scientific, Germany). OLPVRII was loaded on the microbridges, then the setup was covered by cover slips and incubated at 4°C. We varied the concentration of the protein and glutaraldehyde to find the optimal condition for incomplete cross-linking of the proteins. This allowed us to observe in the denaturing conditions of SDS-PAGE the array of artificial multimers formed by several crosslinked protomers of the natural multimer. Using the molecular weight of the highest visible multimer and the number of the appered multimers, we could make a conclusion about an aggregate state of the protein in a sample. We used for the experiments the following samples: solubilized in DDM multimeric fraction of OLPVRII, reconstituted into E. coli polar lipids and POPC/POPG mixture (ratio of 4:1 by weight) monomeric fraction of OLPVRII. The optimal conditions for solubilized protein were 2 mg/ml protein, 10% glutaraldehyde, 8 hours incubation. For the monomeric protein reconstituted into POPC/POPG mixture we incubated the sample for 8 hours, the concentration of the protein in liposomes was 1 mg/ml and glutaraldehyde - 10%. For the E. coli polar lipids, probably due to the presence of NFb-groups in the lipid mixture, the crosslinking was not so efficient, so the formation of the pentamers was not observed.

To prove that the observed pentameric assembly is not an artifact of crystallization, we performed crosslinking experiments with wild type protein. Using size exclusion chromatography, the protein was separated into two fractions corresponding to “monomers” and “pentamers”. Crosslinking experiments clearly confirm that multimeric fractions appearing on the size exclusion chromatography comprise of pentamers. Moreover, we reconstituted the monomers of solubilized wild type protein into liposomes, and crosslinking experiments showed that in these liposomes OLPVRII protein is in the pentameric form. Thus, the pentamers retain their oligomeric form during purification, and purified in detergent monomers, when placed into lipid environment, assemble into the pentamers again, particularly during crystallization. In addition, we expressed and purified a triple mutant E26A/R36A/W203A with three mutations of highly conserved amino acids that are located at the interface between the protomers and form hydrogen bonds between them. While the wild type protein had the pentamer:monomer ratio being 4:1 in the elution profile, the triple mutant showed an approximately 1 : 1 ratio. The fact that the triple mutant partially retained the pentameric assembly suggests that there are additional interactions between OLPVRII protomers, such as hydrogen bonding between C-D loop and C-terminus and hydrophobic cooperation. Taking into account that the key amino acids forming the pentamer are highly conserved in the viral rhodopsins of the second group, we suppose the pentameric assembly to be crucial for OLPVRII function.

Spectroscopic characterization

We recorded the absorption spectra using the Shimadzu UV-2401PC spectrophotometer. Fitting at lower pH values of the dependence of the retinal absorbance maximum on the pH value with Boltzmann sigmoidal curve gave the pKa of the proton acceptor (point of half decay ± standard error). Fitting at higher pH values of dependence of the absorbance at 514 nm on the pH value with Boltzmann sigmoidal curve gave the pKa of the Schiff base (point of half decay ± standard error). To access the photocycle kinetics of viral rhodopsin we used two different laser flash photolysis setups. The first setup was similar to that described by Chizhov and co-workers (Chemavskii, D. S. et al. Biophys. J. 71, 2329- 2345 (1996)). Surelite 11-10 Nd:YAG laser (Continuum Inc, USA) was used providing pulses of 5 ns duration at 532 nm wavelength and energy of 3 mJ/pulse. We placed the samples between two collimated mechanically coupled monochromators (1/8 m model 77250, Oriel Corp., USA), and the probing light (Xe-arc lamp, 75W, Osram, Germany) passed the first monochromator, sample and arrived after a second monochromator at a PMT detector (R3896, Hamamatzu, Japan). Two digital oscilloscopes (LeCroy 9361 and 9400A) were used to record the traces of transient transmission changes in two overlapping time windows.

The second setup was as follows. Brilliant B Nd:YAG laser (Quantel, France) was used providing pulses of 4 ns duration at 500 nm wavelength and energy near 2 mJ/pulse. Samples were placed between two collimated mechanically coupled monochromators (LSH- 150, LOT, Germany). The probing light (Xe-arc lamp, 75 W, Hamamatsu, Japan) passed the first monochromator, sample and arrived after a second monochromator at a PMT detector (R12829, Hamamatzu, Japan). Two digital oscilloscopes (Keysight DSO-X 4022A) were used to record the traces of transient transmission changes in two overlapping time windows.

On both setups we recorded transient absorption changes starting from 700 ns after the laser pulse until completion of the photocycle, at each wavelength from 330 to 730 nm in 10 nm steps 25 laser pulses were averaged to improve the signal -to-noise ratio. We analyzed the data sets using the global multi-exponential nonlinear least-squares fitting program MEXFIT as was reported earlier.

We measured the photocycle kinetics of OLPYRII at pH 7.5 and temperatures between 0 and 30 °C with 10 degree intervals. For the experiments solubilized monomeric protein expressed in C41 strain was reconstituted into soy bean lipid vesicles. Along with this experiment we conducted additional tests. For the SE1 -expressed protein we measured the photocycle of both monomers and pentamers for the protein in three different environments: solubilized in DDM, reconstituted into soy bean lipids and nanodiscs, and did not observe any significant photocycle differences. The photocycle of the monomers in detergent was shorter (35 ms comparing to 70-100 ms in the other environments), but the order and appearance of the intermediates were the same for all samples. Moreover, as we observed the deprotonation of the Schiff base at higher pH values, to ensure the robustness of the data we additionally measured the photocycles of the detergent-solubilized OLPVRII monomers at pH 8.0, 7.0, and 6.0. The protein demonstrated identical behavior at these pH values.

The UV-Vis absorption spectra of OLPVRII solubilized in DDM show maximum absorbance of bound retinal at 514 nm (Fig. 3). The position of the peak shifts depending on the buffer pH. We varied the pH of the OLPVRII sample in 6-mix buffer from 7.58 to 1.93 and from 7.58 to 12.93 for two identical samples. At alkaline pH, the maximum absorbance of the retinal changed from 514 to 367 nm indicating the RSB deprotonation. A sigmoidal fit of the dependence of the absorbance at 514 nm on pH gave the pKa of the RSB of 10.36±0.08 (point of half decay±standard error). In contrast, at acidic pH, we observed a gradual shift of the retinal absorbance maximum from 514 to 548 nm indicating protonation of the RSB counterion Asp75. We fitted the dependence of the wavelength of the absorbance maximum on pH with a sigmoidal curve and obtained the pKa of the proton acceptor of 3.34±0.11 (point of half decay±standard error). The corresponding pKa values for bR are 12.2 and 2.7 (Balashov, S. P. et al. Biophys. J. 60, 475^190 (1991); Jonas, R. & Ebrey, T. G. Proc. Natl. Acad. Sci. 88, 149-153 (1991)), whereas for PR, the pKa of the proton acceptor is 7.68 (Friedrich, T. et al. J. Mol. Biol. 321, 821-838 (2002)). These notable differences in pKa appear to stem, first, from the absence in viral rhodopsins of the histidine that is conserved in PRs and forms a hydrogen bond with the proton acceptor thus decreasing its pKa, and, second, the amino acid substitutions A215M and L93S, compared to bR, that alter the properties of the RSB.

The photocycle kinetics of monomeric OLPVRII expressed in the C41 strain and reconstituted into soy bean lipid vesicles was measured at pH 7.5. Transient absorption changes after laser illumination are shown in Fig. 4D at three characteristic wavelengths, 410, 530, and 550 nm. The dataset is best fitted by seven exponentials. The total length of the OLPVRII photocycle is about 70 ms, and the corresponding time constants were retrieved. The results of the global fit are shown in Fig. 4C.

The photocycle includes a microsecond part that is usually assigned to the release of the energized ion appearing as the RSB deprotonation, and a millisecond part of the ground state recovery by proton uptake. Like in most rhodopsins, the amount of the resolved kinetic intermediates is higher than the amount of spectrally distinct states which indicates the presence of spectrally silent transitions. The representative absolute spectra of the intermediates are shown in Fig. 5. In the microsecond part of the photocycle the RSB deprotonation occurs via a series of intermediates with characteristic times of ti=1 ps, X2=16 ps, 1₃=61 ps, and X4=340 ps. The first intermediate is composed of single K-like spectral state, while P2-P4 are mixtures of K540- and M410-like states. The appearance of the blue-shifted M4 10-like intermediate with the deprotonated RSB in the range of hundreds of microsecond is typical for microbial rhodopsins. However, in contrast to bR and many other rhodopsins, OLPVRII does not form an L-state, the same was obtained for proteorhodopsin (Friedrich, T. et al. J. Mol. Biol. 321, 821-838 (2002)). Then, in the millisecond time range, the M410-like intermediate returns to the ground state (xs=10 ms, X6=25 ms, and cg=73 ms), via a series of equilibrium states P5-P7 that consist of the mixture of M410-, N525-, and 0550-like states.

To validate the proposed mechanism of proton pumping, we analyzed E42Q and D75N mutants of solubilized OLPVRII, where the predicted negatively charged proton donor and acceptor, respectively, are replaced by polar residues. The mutant E42Q showed a photocycle closely similar to that of the wild type protein, but with a considerably longer decay of the M-state, while the D75N failed to form the M-state (Fig. 6). These experiments show that Asp75 is the proton acceptor of OLPVRII, as is the case in many other microbial rhodopsins, whereas Glu42 is the proton donor that is located in the helix B, in a sharp contrast to most known rhodopsins.

BLM measurements

The BLM setup was similar to that described by Bamberg and co-workers (Friedrich, T. et al. J. Mol. Biol. 321, 821-838 (2002)). Optically BLMs with area of ~10^-2 cm² were formed across a hole between the two compartments of a cuvette filled with an electrolyte solution (100 mM Tris-HEPES, initial pH 7.4). The membrane-forming solution consists of n-decane with addition of 1.5% (w/v) diphytanoyl-phosphatidylcholine (Avanti

Biochemicals, Birmingham, AL) and 0.025% (w/v) octadecylamine (Riedel-de-Haen, Hannover, Germany). After the addition OLPVRII-containing proteoliposomes a photosensitivity of the samples reached maximal current amplitudes after ~90 minutes. We carried out the experiments under illumination of either mercury arc lamp (Osram HBO 103 W) or excimer laser-pumped dye laser (LPX105MC, Lambda Physics, Gottingen, Germany). Afterwards, a combination of ionophores (2-3 mM 1799 ((2, 6-dihydroxy)- 1,1, 1,7, 7,7- hexafluoro-2,6-bis(trifluoromethyl)heptane-4-one) and 5 mM monensin) was added effectively permeabilizing the membrane system (final conductance 50-100 nS). We measured the photocurrents under short-circuit conditions, so that no external driving force is generated. All BLM experiments were carried out at room temperature. Signal recording and shutter/laser-triggering was carried out by Pclamp8 software via a Digidata 1200 Interface (Axon Instruments).

Solubilized OLPVRII was reconstituted into E. coli polar lipids in the buffer 100 mM Tris-HEPES pH 7.4 (pH was chosen because this is close to the natural conditions in the Organic Lake). Without an added protonophore, upon illumination of the proteoliposomes with continuous light, we observed characteristic photocurrents (Fig. 4A), where a fast upward deflection is followed by a subsequent slow decay back to zero current, whereas the stationary current is negligible. Adding ionophores 1799 and monensin had no qualitative effect on the photocurrent, but a positive stationary current was observed indicating continuous pumping (Fig. 4A). The liposomes were prepared solely in the Tris-HEPES buffer. These ions are unlikely to be transported, so that we conclude that the photocurrent is carried by protons. To investigate which steps of the transport cycle are electrogenic, we performed time- resolved BLM measurements (Fig. 4B). Upon a laser flash the fast upward deflection has two time constants, ti'=24 ps and T₂ ^,=6 ms, whereas the slow decay of the current is characterized by two constants T₃ ^,=90 ms and T₄-950 ms. Comparing these time constants to the constants obtained in the flash photolysis kinetic measurements, we can assign the time constant n ' to the K540®M410 transition, and the time constants ti' and 1₃' reflect the formation of late 0550- and N525-like intermediates. The time constant if can be correlated with the unspecific discharge of the BLM.

Taking into account the arrangement of functionally important amino acids in the protein and the results of the BLM experiments, we conclude that, at least in liposomes, OLPVRII acts as an outward proton pump. However, the charge transfer per photocycle is small compared to that of the expected channel activity. Indeed, it was shown that ChR2 also possesses outward proton pumping activity, which is negligible compared to its ion channeling (Feldbauer, K. et al. Proc. Natl. Acad. Sci. 106, 12317-12322 (2009)).

Example 2 - Crystal structure of OLPVRII

Crystallization and structure determination

The crystals were grown using the in meso approach, similar to our previous studies (Gordeliy, V. I. et al. Nature 419, 484^-87 (2002)). We mixed the solubilized protein with premelted at 42 °C monoolein (Nu-Chek Prep) to form a lipidic mesophase. Final protein concentration in the phase was 20 mg/ml. Then we spotted 150 nl drops of mesophase on a 96-well LCP glass sandwich set (Paul Marienfeld GmbH, Germany) and overlaid them with 400 nL of precipitant solution using NT8 crystallization robot (Formulatrix). The crystals grew at 22 °C temperature and appeared in 6-12 weeks. The best crystals were obtained with a protein concentration of 25 mg/ml and 0.2 M sodium malonate, pH 4.6, 15% PEG 550 MME. All crystals were harvested using micromounts (MiTeGen, USA) and were flash- cooled and stored in liquid nitrogen for further crystallographic analysis.

X-ray diffraction data were collected at P14 beamline of the PETRAIII, Hamburg at 100 K, with an EIGER 16M detector. We processed diffraction images with XDS and scaled the reflection intensities with AIMLESS from the CCP4 suite. The crystallographic data statistics are presented in Table 1. Reference model (archaerhodopsin-2, PDB 2EI4) for molecular replacement was chosen with the MoRDa pipeline. Initial phases were successfully obtained in P2i space group by an Automated Model Building and Rebuilding using Autobuild. The initial model was iteratively refined using REFMAC5, PHENIX and Coot. The cavities inside the protein were calculated using HOLLOW.

Overall OLPVRII structure

Typically for in meso crystallization, type I pyramid-shape crystals up to 50 pm in size (Fig. 7A) appeared within 3 months. The crystals diffracted up to 1.9 A, and the structure of OLPVRII was solved at 1.9 A. The space group of the crystals was determined to be P2i, with the cell lattice parameters of a=79.58 A, b=99.70 A, c=82.96 A, a=g=90° and b=116.95^°. Crystal packing and examples of the electron density maps are presented in Fig. 7B-E.

OLPVRII asymmetric unit cell contains five molecules of the protein organized in a pentamer with a bottle-like pore in the centre that has a narrow neck in the cytoplasmic part (Fig. 8A,D). Several lipid moieties with disordered polar heads are observed around the pentamer (Fig. 9). In addition, a well-ordered molecule of monoolein (MO), host molecule of the crystallization matrix, is observed between the D helix of one protomer and the G' helix of the neighbouring pentamer. MO is buried deep inside the hydrophobic groove formed by Phel97, Ile200, Tyr201, Phe205, Phe210 of one monomer and Phel03' of the other monomer (Fig. 9). The location of the MO polar head is stabilized by hydrogen bonding with the main chain of Phe210 and via Wat20 with Ala204 and lie 209 at the C-terminus.

Organization of OLPVRII pentamer

OLPVRII pentamer is stabilized by a dense net of hydrogen bonds. The most intensive contacts between protein molecules in the pentamer are located on the cytoplasmic side of the protein, where the helix A of one protomer is interposed into the cleft between helices A' and B' of other protomer (Fig. 8B). The existence of the entire network is conditioned on the presence of two highly ordered water molecules. The first one (Watl2) coordinates hydrogen bonding of the side chains of His37 and Asn40 of one protomer and Trp203' and Glu26' of the other (Fig. 8C,F). Asn40 is stabilized also by a hydrogen bond with the backbone oxygen of Ile22\ The second water molecule coordinates the interactions between Arg36 and backbone oxygens of Ala27 and Thr30 to further stabilize the position of Arg29' side chain of the neighbour protein molecule. At the center of the pentamer, oxygens of Leu28 and Arg29 side chains of all 5 protomers are also bound by strong hydrogen bonds mediated by 5 water molecules (Wat 13). Moreover, Glu26' of the second protomer strongly interacts with Arg36 from the same protomer providing a connection between the clusters of water-mediated hydrogen bonds. Thus, the following amino acids play the key role in the formation of the pentamer: Glu26, Arg36, His37, Asn40, Trp203, and these amino acids are highly conserved within group II of viral rhodopsins, but not in other rhodopsins (Fig. 10). In addition, the distorted C-terminus of one protomer interacts with the poorly ordered C-D loop of the neighboring protomer forming several hydrogen bonds between their backbone atoms and thus stabilizing the pentamer (Fig. 8B, E).

The pentameric assembly was also observed for other microbial rhodopsins, such as PRs and NaRs. Although the relative orientation of the protomers is similar for OLPVRII, KR2 and BPR HOT75 (Fig. 11), the size and profile of the central pore together with the pentamerization contacts are completely different in these proteins (Fig. 12). Oligomerization of PRs and NaRs affects the functionality of these proteins. For instance, in KR2, NaR from Krokinobacter eikastus, pentamerization is required for sodium pumping. Thus, pentamerization could also be a key determinant of the activity of group II viral rhodopsins, and a thorough analysis of the OLPVRII pentamer is required to elucidate its function.

Central pore (channel) of the pentameric structure

However, the contacts between the OLPVRII molecules described above are unusual for microbial rhodopsins. They lead to the formation of the bottle-shaped pore inside the pentamer formed by the A and B helices of the protomers (Fig. 8D and Fig. 13 A). The narrowest section (vestibule) of the pore is formed by the Phe24, Leu28 and Arg29 of helix A and has the length of 11 A and mean diameter of 6 A. Notably, Phe24 and Arg29 are highly conserved in group II viral rhodopsins, while Leu28 is interchanged with other hydrophobic residues, He and Met (Fig. 10). At the entrance of the pore, on the cytoplasmic side of the pentamer, the unique arrangement of Arg29 side chains allows the positioning of 5 neighbouring positively charged amino acids. Hereafter, we denote this configuration Arg29 ring.

Deeper into the pore, oxygen atoms of Leu28 main chain coordinate 5 water molecules forming a pentagon stabilized by hydrogen bonds with the mean distance between two neighbouring molecules of 2.6±0.2 A (mean±s.e.m., for different pentamer molecules, Fig. 13B). This pentagon weakly interacts with Arg29 side chains as water molecules are located within 3.8±0.3 A (mean±s.e.m., for different pentamer molecules) from the nearest arginine nitrogen atoms. The Leu28 and Phe24 side chains form an extremely hydrophobic part of the vestibule (Fig. 14). Surprisingly, we observed a strong electron density along the pore axis in this region (Fig. 13C). We assigned it to a short hydrocarbon chain that plugs the pore. Phe24 side chain is oriented perpendicularly to the long axis of the pore forming the narrowest section of the vestibule with the diameter of only 5.5 A. Closer to the extracellular side, Phe24 side chain forms the shoulder of the pore, where its diameter sharply increases to 26 A. The wide extracellular part of the pore is lined with the residues Thr21, Val47, Metl7, Ser54, Asn55, TyrlO and Tyrl3. The shoulder and the following part of the pore are paved with fragments of lipid molecules (Fig. 14).

The comparison of the pores inside the oligomers formed by microbial rhodopsins with known structure shows the uniqueness of the channel in OLPVRII pentamer (Fig. 12). Remarkably, however, a similar pore organization is observed in the transmembrane domains of pentameric Ligand-Gated Ion Channels (pLGICs) (Fig. 15) (Sauguet, L. et al. EMBO Journal 32, 728-741 (2013)). Indeed, the ring of 5 negatively-charged amino acids (Glu222 in each subunit) plays the role in the ion selectivity of a structurally characterized pLGIC from Gloeobacter violaceus (GLIC). Even more notably, the crystal structure of GLIC (PDB ID: 4HFI) contains similar pentagons of water molecules inside the transmembrane channel, presumably involved into hydration of the permeation ion, that are followed by the extremely hydrophobic region plugged with the carbon tails of detergent molecules. The OLPVRII pore vestibule mimics the transmembrane channel of GLIC (Fig. 15). Such an unusual narrow cytoplasmic part of the pore, its similarity to the transmembrane channels of pLGICs, and the evolutionary conservation of most residues involved in the formation of the pentamer and pore strongly suggest that the central pore of OLPVRII functions as a channel. In pLGICs, opening and closure of the transmembrane channel are triggered by conformational changes starting with ligand binding at the active center of the extracellular domain. In contrast, structural rearrangements in microbial rhodopsins are initiated by absorption of light by the retinal cofactor that resides inside the protein. Therefore, OLPVRII is likely to function as a pentameric light-gated ion channel. To identify possible mechanisms of the central channel regulation in OLPVRII, we analyzed the protomer organization.

Retinal pocket and extracellular part of OLPVRII In the OLPVRII pentamer, there are only slight differences between protomers. The root mean square deviation (RMSD) of the backbone atoms positions between different protomers varies from 0.17 to 0.20 A. Thus, below, we describe only protomer A from the OLPVRII model.

Similarly to other microbial rhodopsins, each protomer of OLPVRII is organized as a bundle of seven transmembrane helices (A-G) connected by relatively short loops. The retinal cofactor is covalently bound to Lysl95 through the Schiff base (RSB), and the electron densities indicate that the chromophore is in all-trans conformation (Fig. 7C). OLPVRII consists of only 211 amino acids and is the smallest among microbial rhodopsins with known crystal structure. Compared to other rhodopsins, the length of the protein is reduced due to extremely short N- and C-termini as well as short loops that lack defined secondary structure. Because the differences in the structures of the OLPVRII and Halobacterium salinarum bacteriorhodopsin (bR) are likely to reflect functional differences, we compare the structures of bR and OLPVRII in detail. The assembly of OLPVRII helices into a bundle differs from that in bR (PDB ID 1C3W), with RMSD of the backbone atoms positions of about 1.8 A (Fig. 16). Importantly, in OLPVRII, helix C is considerably longer, whereas helix F, conversely, is shorter than the corresponding helices of bR. The extracellular part of helix G is shifted inward for about 4 A and the tilt and positions of other helices are also slightly altered in OLPVRII compared to bR.

The walls of the retinal-binding pocket are hydrophobic and similar to those of other rhodopsins except for the regions of the b-ionone ring and RSB. However, compared to bR, there are considerable differences in the adjacent amino acids: the highly conserved Pro 186 is replaced with Glyl73, Serl41 with Glyl37 (like in PRs), and Thrl42 with Phel38 (Fig. 17C and Fig. 18). There are differences also at the cytoplasmic side of the RSB, where Leu93 of bR is replaced with Met83 and Ala215 with Seri 94. Mutation A215T is the major change converting bR to a sensory rhodopsin, whereas Leu93 and homologous residues are functionally important and responsible, for example, for the color tuning in PRs. These mutations lead to the shift of the RSB towards the extracellular part of the protein compared to bR, due to the steric conflict with Met83 and Seri 94.

Like in bR, in OLPVRII, the RSB nitrogen is hydrogen bonded to a key water molecule Watl (Wat402 in bR) that donates hydrogen bonds to two anionic residues, Asp75 and Asp 191 (Asp85 and Asp212 in bR, correspondingly). This arrangement stabilizes the positive charge of the RSB and is conserved in outward proton pumps and most other rhodopsins. In OLPVRII, structural organization of the RSB that involves Asp75, Aspl91 and Arg72 is similar to that of bR, but some differences are observed (Fig. 17A and Fig. 18). The structure and mutational analysis suggest that Asp75 is the primary proton acceptor from RSB and the deprotonation of RSB in OLPVRII occurs in a manner similar to that of bR.

The space between Arg72 and the extracellular surface of the protein is notably different in OLPVRII compared to bR (Fig. 17 and Fig. 19). Whereas in bR, Arg82 side chain is stabilized by hydrogen bonds with water molecules Wat403, Wat404, Wat405 and Wat407, and plays a key role in proton transfer, in OLPVRII, these water molecules are replaced by three Asn side chains, Asn69, Asnl84 and Asnl88. Importantly, Asn69 and Asnl84 are highly conserved and Asnl88 is completely conserved among the group II viral rhodopsins. While Arg82 was demonstrated to be a key element in the proton translocation mechanism in bR, and analogous arginine is found in most microbial rhodopsins, playing important roles in their functioning, we suggest that the strong stabilization of Arg72 in OLPVRII by three Asn side chains may affect its mobility and thus affect the function of the rhodopsin.

Strikingly different from bR, where the extracellular bulk is separated from the proton release group (Glul94 - Glu204 pair), in OLPVRII, a large cavity protrudes from the extracellular protein surface down to Arg72 (Fig. 17), similarly to Exiguobacterium sibiricum rhodopsin (ESR) and ChR2 (Fig. 20). This hydrophilic cavity is filled with water molecules and is formed by polar amino acids, including Argl83, Aspl24, Tyr73, Tyrl87, Glnl80 and Asn69 (Fig. 19). These amino acids are densely hydrogen bonded and are also linked via a hydrogen bond network to Arg72, and further, to the RSB. Thus, in OLPVRII, Arg72 side chain separates the RSB region and the extracellular space. RSB deprotonation during the photocycle triggers structural rearrangements in the extracellular part of the protein and therefore can result in opening of the gate around Arg72 similarly to the way it might occur in ChR.

Cytoplasmic part of OLPVRII and its connection to vestibule

The cytoplasmic part of OLPVRII dramatically differs from those of all rhodopsins with known structures (Fig. 21). First, there is no charged amino acid in the cytoplasmic half of the helix C (Val86 in OLPVRII at the position of Asp96 in bR) which would act as a proton donor to the RSB during the photocycle as in most of the outward proton-pumping rhodopsins. In OLPVRII, the putative proton donor Glu42 is located in helix B (which corresponds to Thr46 in bR) and thus juxtaposed to the RSB. Glu42 is fully conserved in both groups of viral rhodopsins and site-directed mutagenesis results indicate that Glu42 indeed serves as a proton donor to RSB during the OLPVRII photocycle.

The mechanism of Glu42 reprotonation remains unclear. Most likely, proton uptake proceeds from the side of Glu26, where the cytoplasmic surface is concaved inside the protomer and reaches the side chain of Leu39, which is the only residue separating Glu42 from the bulk. In this case, Leu39 might play the role of a hydrophobic gate, similar to that of Leu93 in bR (Fig. 22). An alternative mechanism of Glu42 reprotonation could involve the uptake of the proton from the part of the protomer occupied by Asn94, Asnl58, Seri 57, Tyr201 and Tyr206, which together form a small hydrophilic cavity near the cytoplasmic surface of the protein.

Moreover, a large cavity is observed between the RSB and the Glu42 side chain. This cavity is hydrophilic and filled with 5 water molecules stabilized by hydrogen bonding with Glu42, Ser36 and main chain oxygens of Seri 94 and Lysl95 in helix G. A similarly positioned cavity exists in the wild type ChR2 but, in this case, the cavity is hydrophobic, with no water molecules inside. Hydrogen bonds protrude from Glu42 via water molecules Wat7, Wat8, Wat9, WatlO, Watl l and Thr23 and Thr43 side chains to Leu39, which is located next to Asn40 from the pentamerization interface, strengthening interactions between helices A and B inside the protomer (Fig. 23). The distances between Glu42 and Wat9, as well as between Wat8 and Wat9, are extremely short, only 2.24 A and 2.20 A, respectively. The proton donor Glu42 as well as His37, Asn40 and Arg36 that form pentameric contacts in OLPVRII are all located within 2 turns of the same helix B. Therefore, in OLPVRII, the proton donor is rigidly connected to the oligomerization interface not only by the geometry and constraints of the a-helix but also by a continuous chain of strong hydrogen bonds through the water molecules in the hydrophilic cavity and polar residues of helix A. Furthermore, the dense hydrogen bonding network of the pentameric contacts protrudes directly to Arg29 ring (Fig. 23).

This distinct architecture of the cytoplasmic part, together with the conservation among group II viral rhodopsins of all amino acids responsible for the pentameric assembly of OLPVRII, strongly supports the idea that structural changes initiated by retinal isomerization upon absorption of light propagate directly from the RSB to the vestibule of the central pore. Indeed, the location of the proton donor Glu42 in the helix B, close to the RSB, its additional hydrogen bonding with helices A and G and the oligomerization interface, as well as the tight interprotomer contacts provide for an almost direct interaction between RSB and Arg29 that are located at the distance of 27 A from each other (Fig. 23). The pore is formed by helices A and B, and accordingly, its shape and characteristics are determined by the hydrogen bonding network that involves and stabilizes these helices and depends on the retinal conformation (in particular, RSB orientation). Moreover, our attempt to trap an active state of the protein in the crystals led to a dramatic destruction of the crystals (like in the case of eye rhodopsin and sensory rhodopsin) evident by a drop of structural resolution from 2 to 20A. This fact indicates that large structural rearrangements occur in OLPVRII during the photocycle and gives another hint on the channel activity of the protein, where large structural rearrangements occur upon ion channel opening like in pLGICs.

Thus, detailed examination of the OLPVRII structure not only suggests that this rhodopsin is a pentameric light-gated ion channel, but also demonstrates the possible mechanisms of the channel functioning and regulation.

Example 3 - Molecular dynamics simulations of OLPVRII

We carried out all-atom MD simulations of OLPVRII using the CHARMM36m force field with the TIP3P water model. The simulation box contained an equilibrated 1-palmitoyl- 2-oleoyl-sn-glycero-3-phosphocholine (POPC) lipid bilayer, surrounded by 200 mM NaCl aqueous solution. OLPVRII was modelled using the residue range 1-211 and embedded into the bilayer using g membed. We assigned Glu42 a protonated state, all other amino acids were modeled in their default ionization state to reflect the most probable state at neutral pH based on pKa calculations using PROPKA 3.1. Hydrocarbon chains bound to the central pore were modelled as one octane and five dodecane molecules.

We performed simulations using GROMACS 2018 in the NPT ensemble with periodic boundary conditions and an integration time step of 2 fs. Temperature was maintained at 310 K using the velocity-rescaling thermostat; pressure was kept at 1 bar using the semi-isotropic Parrinello-Rahman barostat as described recently. Lennard-Jones interactions were truncated at 12 A with a force switch smoothing function from 10-12 A. Electrostatic interactions were calculated using particle mesh Ewald method and a real space cutoff of 12 A. Simulation systems were equilibrated with position restraints on the protein heavy atoms for 400 ns, followed by ~20 ns with backbone-only position restraints. Subsequently, we performed 2-3 independent production runs of 2 ps length for OLPYRII with and without bound hydrocarbon chains in the pore, and F24A, L28A, R29E, F24/L28A, and F24/L28A/R29E mutants without the chains.

For consistency, all analyses of the simulations are based on the time windows from 1400-1800 ns in each simulation. We analyzed water and ion density distributions in the MD trajectories using GROmaps and MD Analysis and calculated electrostatic potential distributions from the charge densities in the MD simulations using the Poisson equation as implemented in PMEPot. Pore radii were calculated using FIOLE for an ensemble of conformations observed in the MD trajectories. Water, Na⁺, and Cl distributions in the free MD simulations were analyzed to calculate potential of mean force profiles along the pore axis via

where G is the free energy, T the temperature, and R the gas constant. Densities were analyzed within a cylinder placed along the pore axis with a radius of 36.23 A (resulting in a cross-sectional area identical to the area of OLPVRII in the membrane plane); n(z) is the density within a membrane slice at axial position z, and nbulk is the density in the bulk phase. We calculated standard deviations of energies and pore radii using block bootstrap sampling.

To test whether OLPVRII could function as a pentameric light-gated ion channel, we carried out unguided all-atom MD simulations of the protein embedded in a POPC lipid bilayer in a 200 mM NaCl aqueous solution. We observed spontaneous water flux into the central pore from both membrane sides except for the hydrophobic region occupied by the hydrocarbon chain (Fig. 24A). In the absence of these lipids, a continuous hydration pattern suggests a putative ion conduction pathway (Fig. 24A, B). However, closer inspection reveals a minimum in water density in the pore region lined by Phe24 and Leu28, where burst-like liquid-vapor oscillations occur on the nanosecond timescale. Calculation of pore radii shows a constriction in this region, with a minimum of 1.93 A (Fig. 24C).

Arg29 creates a positive electrostatic potential at the cytoplasmic pore entrance that is separated by the hydrophobic constriction from the wide, partially negative, extracellular entrance (Fig. 24D and Fig. 25). Free-energy profiles calculated for water, Na⁺, and Cl permeation show that the cytoplasmic side of the pore is highly anion selective, whereas the extracellular entrance is cation selective, with the highest barrier created by Phe24 and Leu28 (Fig. 24E). Reduction of the respective side chain volumes by alanine substitution increases the hydration of the constriction, lowers the energy barrier for ion permeation, and leads to an overall Cl -selective ion pathway in the double mutant. Along this pathway, we observed multiple spontaneous water and C1 permeation events (Fig. 24F and Fig. 25). Our simulations define OLPVRII as an anion channel, that is sterically open, but functionally closed by a hydrophobic gate formed by Phe24 and Leu28. Importantly, free-energy calculations demonstrate that charge reversal via the R29E mutation converts OLPVRII into a closed cation-selective channel, which is conductive in the F24A/L28A/R29E mutant, suggesting a similar gating mechanism as in the wild type protein (Fig. 25). Gating of OLPVRII thus resembles the hydrophobic gating mechanism in pLGICs and other biological ion channels. Following light-induced isomerization of the retinal, activation of OLPVRII may propagate from the RSB via the hydrogen bond network toward the helices forming the pore. It is tempting to speculate that channel opening involves rotation or tilting of helix A, which moves Phe24 and Leu28 away from the pore and thereby lowers the effective hydrophobicity in the pore, causing hydrophobic gate wetting and permitting Cl permeation. Indeed, similar rotation of the a-helix leads to significant displacement of hydrophobic residues (Phe80, Phe84 and Ile76) and the channel opening in pentameric bestrophin ligand-gated chloride channel BEST1 (Fig. 26).

Example 4 - Optogenetic approaches to the stimulation of the auditory pathway and the recovery of vision

Hernandez V. H. et al. (J Clin Invest. 124, 1114-1129 (2014)), demonstrated an approach for optogenetic stimulation of the auditory pathway in rodents, particularly the optical stimulation of neurons genetically engineered to express the light-gated ion ChR2. Confirmed by recordings of single neuron and neuronal population responses, optogenetic stimulation of spiral ganglion neurons activates the auditory pathway and restores auditory activity in deaf mice. In the followed article, the authors (Keppeler, D. et al. EMBO J 37, e99649 (2018)) demonstrated improved plasma membrane targeting and temproral resolution of fast mutant of ChR2 Chronos. Given achieved improved temporal resolution and sensitivity of optogenetic stimulation with Chronos, optical cochlear implants could provide considerable advantages over electrical cochlear implants, particularly in spectral resolution of coding. This is highly relevant clinically, as more than 450 million people suffer from a disabling hearing impairment and causal therapies for the most common form, sensorineural hearing impairment, are still lacking. The light-gated pentameric channel of the present disclosure might be employed similarly to ChR2 and its mutants in optogenetic stimulation of auditory pathway. Introducing the coding sequence such as SEQ ID NO: 2 (human codon- optimized OLPVRII) into the constructs as described by Hernandez V. H. et al. is a common practice in the art.

Mace E. et al. (Mol Ther. 23, 7-16 (2015)) described an approach for optogenetic activation of retinal neurons. Inherited retinal degenerations affect around 1 in 3,000 people, with the most common form being retinitis pigmentosa. Despite the great diversity of mutations, retinitis pigmentosa converges on a phenotype of photoreceptor cell loss in later stages of the disease. Studies of post-mortem retinas from patients have shown that a large percentage of inner retinal neurons remain present even after photoreceptor degeneration. Their optogenetic reactivation mediated by adeno-associated virus gene therapy has the potential to restore vision regardless of patient-specific mutations. The challenge for clinical translatability is to restore a vision as close to natural vision as possible, while using a surgically safe delivery route for the fragile degenerated retina. Expression of ChR2 in ON bipolar cells or retinal ganglion cells might allow greater spatial resolution and acuity than that achieved with current prosthetic devices. This has motivated a first-in-human phase 1 clinical trial that was initiated recently (ClinicalTrials.gov NCT02556736).

In Mace E. et al. (Mol Ther. 23, 7-16 (2015)) ON bipolar cells are targeted. For safe gene delivery, an engineered adeno-associated virus variant was used that can transduce the bipolar cells after injection into the eye's easily accessible vitreous humor. It is shown that adeno-associated virus encoding ChR2 under the ON bipolar cell-specific promoter mediates long-term gene delivery restricted to ON-bipolar cells after intravitreal administration. ChR2 expression in ON bipolar cells leads to restoration of ON and OFF responses at the retinal and cortical levels. Moreover, light-induced locomotory behavior is restored in treated blind mice.

In a more recent publication (Chaffiol A. et al. Mol. Ther. 25, 46-60 (2017)), the authors studied an optogenetic vision restoration in cynomolgus macaques. A more sensitive human codon-optimized Ca2+-permeable channelrhodopsin (CatCh) was used along with a newly designed promoter sequence from the regulatory region of the human gamma- synuclein gene that leads to robust and specific transgene expression of CatCh in retinal ganglion cells in both mice and in non-human primats. In the study, CatCh expression in macaque retinal ganglion cells and CatCh-mediated light responses with relatively low light intensities demonstrated optogenetic stimulation of retinal neurons.

Introducing the coding sequence for the light-gated pentameric channel of the present disclosure, such as SEQ ID NO: 2 (human codon-optimized OLPVRII), into the constructs as described, e.g. by Mace et al. or Chaffiol A. et al. is a routine practice for trained personnel. Therefore, the new light-gated pentameric channel of the present disclosure might be inserted in the cassettes for the activation of ON bipolar cells or ganglion cells in the retina for vision restoration.

LIST OF REFERENCES

Beja, O. et al. Bacterial rhodopsin: Evidence for a new type of phototrophy in the sea. Science. 289, 1902-1906 (2000).

Beja, O. & Lanyi, J. K. Nature’s toolkit for microbial rhodopsin ion pumps. Proc. Natl. Acad. Sci. U. S. A. 111, 6538-9 (2014).

Boyden, E. S., Zhang, F., Bamberg, E., Nagel, G. & Deisseroth, K. Millisecond- timescale, genetically targeted optical control of neural activity. Nat. Neurosci. 8, 1263-1268 (2005).

Deisseroth, K. et al. Next-Generation Optical Technologies for Illuminating Genetically Targeted Brain Circuits. J. Neurosci. 26, 10380-10386 (2006).

Inoue, K. et al. A natural light-driven inward proton pump. Nat. Commun. 7, 13415

(2016).

Kwon, S. K. et al. Genomic makeup of the marine flavobacterium Nonlabens (Donghaeana) dokdonensis and identification of a novel class of rhodopsins. Genome Biol. Evol. 5, 187-199 (2013).

Bratanov, D. et al. An approach to heterologous expression of membrane proteins. The case of bacteriorhodopsin. PLoS One 10, (2015). Ritchie, T. K. et al. Chapter 11 Reconstitution of Membrane Proteins in Phospholipid Bilayer Nanodiscs. Methods in Enzymology 464, 211-231 (2009).

Fadouloglou, V. E., Kokkinidis, M. & Glykos, N. M. Determination of protein oligomerization state: Two approaches based on glutaraldehyde crosslinking. Anal. Biochem. 373, 404—406 (2008).

Chemavskii, D. S., Engelhard, M., Mueller, K., Zubov, B. V & Hess, B. Spectrally Silent Transitions in the Bacteriorhodopsin Photocycle. Biophys. J 71, 2329-2345 (1996).

Balashov, S. P., Govindjee, R. & Ebrey, T. G. Redshift of the purple membrane absorption band and the deprotonation of tyrosine residues at high pH: Origin of the parallel photocycles of trans-bacteriorhodopsin. Biophys. J. 60, 475—490 (1991).

Jonas, R. & Ebrey, T. G. Binding of a single divalent cation directly correlates with the blue-to-purple transition in bacteriorhodopsin. Proc. Natl. Acad. Sci. 88, 149-153 (1991).

Friedrich, T. et al. Proteorhodopsin is a light-driven proton pump with variable vectoriality. J. Mol. Biol. 321, 821-838 (2002).

Feldbauer, K. et al. Channelrhodopsin-2 is a leaky proton pump. Proc. Natl. Acad. Sci. 106, 12317-12322 (2009).

Gordeliy, V. I. et al. Molecular basis of transmembrane signalling by sensory rhodopsin Il-transducer complex. Nature 419, 484—487 (2002).

Sauguet, L. et al. Structural basis for ion permeation mechanism in pentameric ligandgated ion channels. EMBO Journal 32, 728-741 (2013).

Hernandez, V. H. et al. Optogenetic stimulation of the auditory pathway. JClin Invest. 124, 1114-1129 (2014).

Keppeler, D. et al. Ultrafast optogenetic stimulation of the auditory pathway by targeting-optimized Chronos. EMBO J 37, e99649 (2018).

Mace E. et al. Targeting channelrhodopsin-2 to ON-bipolar cells with vitreally administered AAV Restores ON and OFF visual responses in blind mice. Mol Ther. 23, 7-16 (2015). Chaffiol A. et al. A New Promoter Allows Optogenetic Vision Restoration with Enhanced Sensitivity in Macaque Retina. Mol. Ther. 25, 46-60 (2017).

Claims

1. A light-gated pentameric channel having at least 59% sequence similarity over the full length of SEQ ID NO: 1 (OLPVRII) for use in medicine or other optogenetic application.

2. The light-gated pentameric channel of claim 1 , wherein the light-gated pentameric channel has at least 65%, more preferably at least 70%, more preferably at least 75%, more preferably at least 80%, more preferably at least 85%, more preferably at least 90%, more preferably at least 95%, more preferably at least 97%, more preferably at least 98%, and most preferably 99% sequence similarity to the full length of SEQ ID NO: 1 (OLPVRII); and/or wherein the light-gated pentameric channel has at least 38%, more preferably at least 45%, more preferably at least 50%, more preferably at least 55%, more preferably» at least

60%, more preferably at least 65%, more preferably at least 70%, more preferably at least

75%, more preferably at least 80%, more preferably at least 85%, more preferably at least

90%, more preferably at least 95%, more preferably at least 97%, more preferably at least

98%, and most preferably at least 99% sequence identity to the full length of SEQ ID NO: 1 (OLPVRII).

3. The light- gated pentameric channel of any one of claims 1 or 2, wherein the light- gated pentameric channel should only comprise (semi)-conservative substitutions at the position corresponding to L24, F28, R29, E42, K195, E26, W203, R36, H37, N40 of SEQ ID NO: 1.

4. The light-gated pentameric channel that comprises the light-gated pentameric channel of claims 1 or 2 or/and that consists of light-gated pentameric channel of claims 1 or

2.

5. The light-gated pentameric channel of claim 4 which comprises the amino acid sequence of SEQ ID NO: 1 (OLPVRII) and/or it consists of the amino acid sequence of SEQ ID NO: 1 (OLPVRII).

6. The light-gated pentameric channel of any one of claims 1-5, that forms the pentameric assembly in the lipid environment, i.e. five single molecules of claims 1-5 form a single protein cluster, when incorporated into either natural membrane in vivo or artificial membrane in vitro.

7. The light-gated pentameric channel of any one of claims 1-6, wherein the lipid environment includes E. coli polar lipid liposomes, E. coli membrane, POPC/POPG liposomes, in meso grown crystals, cell plasma membrane, etc.

8. The light-gated pentameric channel of any one of claims 1-7 for use in restoring auditory activity, recovery of vision, or for use in treating or alleviating alkalosis, neurological injury, brain damage, seizure, or a degenerative neurological disorder, such as Parkinson's disease and Alzheimer's disease.

9. A nucleic acid construct, comprising a nucleotide sequence coding for the lightgated pentameric channel as defined in any one of claims 1-8.

10. The nucleic acid construct of claim 9, wherein the nucleotide sequence is codon- optimized for expression in human cells.

11. The nucleic acid construct of any one of claims 9 and 10, wherein the nucleotide sequence has the sequence shown in SEQ ID NO: 2.

12. The nucleic acid construct of any one of claims 9 to 11 for use in medicine or any optogenetic application.

13. The nucleic acid construct of any one of claims for use in restoring auditory activity, recovery of vision, or for use in treating or alleviating alkalosis, neurological injury, brain damage, seizure, or a degenerative neurological disorder, such as Parkinson's disease and Alzheimer's disease.

14. An expression vector, comprising a nucleotide sequence coding for light-gated pentameric channel as defined in any one of claims 1-8.

15. The expression vector of claim 14, wherein the vector is a viral vector.

16. The expression vector of any one of claims 14-15, wherein the coding sequence of the light-gated pentameric channel is under the control of a neuronal cell specific human promotor.

17. The expression vector of claim 16, wherein the coding sequence of the light-gated pentameric channel is under the control of a human synapsin promotor.

18. The expression vector of any one of claims 14 to 17 for use in medicine or any optogenetic application.

19. The expression vector of any one of claims 14 to 17 for use in restoring auditory activity, recovery of vision, or for use in treating or alleviating alkalosis, neurological injury, brain damage, seizure, or a degenerative neurological disorder, such as Parkinson's disease and Alzheimer's disease.

20. A mammalian cell expressing the light-gated pentameric channel as defined in any one of claims 1-8, with the proviso that the mammalian cell is not a human embryonic cell or a cell capable of modifying the germ line genetic identity of human beings.

21. A mammalian cell comprising the nucleic acid construct according to any one of claims 9-13 or the expression vector according to any one of claims 14-19.

22. The mammalian cell of any one of claims 20 or 21, wherein the cell is a hippocampal cell; a photoreceptor cell; a retinal rod cell; a retinal cone cell; a retinal ganglion cell; a bipolar neuron; a ganglion cell; a pseudounipolar neuron; a multipolar neuron; a pyramidal neuron; a Purkinje cell; a granule cell; a neuroblastoma cell, in particular NG108- 15; a HEK293 cell; a COS cell; a BHK cell; a CHO cell; a myeloma cell; a MDCK cell; any other mammalian cell.

23. The mammalian cell of any one of claims 19 to 22 for use in medicine or any optogenetic application.

24. The mammalian cell of any one of claims 19 to 22 for use in restoring auditory activity, recovery of vision, or for use in treating or alleviating alkalosis, neurological injury, brain damage, seizure, or a degenerative neurological disorder, such as Parkinson's disease and Alzheimer's disease.

25. The non-mammalian cell expressing the light-gated pentameric channel as defined in any one of claims 1-8 or comprising the nucleic acid construct according to any one of claims 9-13 or the expression vector according to any one of claims 14-19, wherein the cell is selected from a group comprising a bacterial cell, a yeast cell, an insect cell, and a plant cell.

26. The non-mammalian cell of claim 25 for use in medicine or any optogenetic application.

27. The non-mammalian cell of claim 25 for use in restoring auditory activity, recovery of vision, or for use in treating or alleviating alkalosis, neurological injury, brain damage, seizure, or a degenerative neurological disorder, such as Parkinson's disease and Alzheimer's disease.

28. A liposome, comprising the light-gated pentameric channel as defined in any one of claims 1-8.

29. The liposome of claim 28 for use in medicine or any optogenetic application.

30. The liposome of claim 28 for use in restoring auditory activity, recovery of vision, or for use in treating or alleviating alkalosis, neurological injury, brain damage, seizure, or a degenerative neurological disorder, such as Parkinson's disease and Alzheimer's disease.

31. A non-human mammal, comprising a cell according to any one of claims 15-17, with the proviso that those animals are excluded, which are not likely to yield in substantial medical benefit to man or animal which will outweigh any animal suffering.

32. The non-human mammal of claim 31, wherein the cell is an endogenous cell.

33. A non-therapeutic, or ex vivo, or in vitro use of a light-gated pentameric channel as defined in any one of embodiments 1-8, a) for light-stimulation of electrically excitable cells, b) for transporting ions over a membrane c) or as an optogenetic tool.