
==== Front
J Phys Chem B
J Phys Chem B
jp
jpcbfk
The Journal of Physical Chemistry. B
1520-6106
1520-5207
American Chemical Society

37694950
10.1021/acs.jpcb.3c04032
Article
A Proteorhodopsin-Related Photosensor Expands the Repertoire of Structural Motifs Employed by Sensory Rhodopsins
Saliminasab Maryam †
Yamazaki Yoichi ‡
Palmateer Alyssa †
Harris Andrew †
https://orcid.org/0000-0002-3805-1319
Schubert Luiz §
Langner Pit §
https://orcid.org/0000-0001-6321-2615
Heberle Joachim §
https://orcid.org/0000-0003-2636-9773
Bondar Ana-Nicoleta *∥⊥
https://orcid.org/0000-0002-5614-8317
Brown Leonid S. *†
† Department of Physics and Biophysics Interdepartmental Group, University of Guelph, Guelph, Ontario N1G 2W1, Canada
‡ Division of Materials Science, Graduate School of Science and Technology, Nara Institute of Science and Technology, Ikoma, Nara 630-0192, Japan
§ Experimental Molecular Biophysics Group, Department of Physics, Freie Universität Berlin, D-14195 Berlin, Germany
∥ University of Bucharest, Faculty of Physics, Atomiştilor 405, Măgurele 077125, Romania
⊥ Forschungszentrum Jülich, Institute for Neuroscience and Medicine and Institute for Advanced Simulations (IAS-5/INM-9), Computational Biomedicine, Wilhelm-Johnen Straße, 52428 Jülich, Germany
* Email: lebrown@uoguelph.ca.
* Email: nbondar@fizica.unibuc.ro.
11 09 2023
21 09 2023
11 09 2024
127 37 78727886
14 06 2023
09 08 2023
© 2023 The Authors. Published by American Chemical Society
2023
The Authors
https://creativecommons.org/licenses/by-nc-nd/4.0/ Permits non-commercial access and re-use, provided that author attribution and integrity are maintained; but does not permit creation of adaptations or other derivative works (https://creativecommons.org/licenses/by-nc-nd/4.0/).

Microbial rhodopsins are light-activated retinal-binding membrane proteins that perform a variety of ion transport and photosensory functions. They display several cases of convergent evolution where the same function is present in unrelated or very distant protein groups. Here we report another possible case of such convergent evolution, describing the biophysical properties of a new group of sensory rhodopsins. The first representative of this group was identified in 2004 but none of the members had been expressed and characterized. The well-studied haloarchaeal sensory rhodopsins interacting with methyl-accepting Htr transducers are close relatives of the halobacterial proton pump bacteriorhodopsin. In contrast, the sensory rhodopsins we describe here are relatives of proteobacterial proton pumps, proteorhodopsins, but appear to interact with Htr-like transducers likewise, even though they do not conserve the residues important for the interaction of haloarchaeal sensory rhodopsins with their transducers. The new sensory rhodopsins display many unusual amino acid residues, including those around the retinal chromophore; most strikingly, a tyrosine in place of a carboxyl counterion of the retinal Schiff base on helix C. To characterize their unique sequence motifs, we augment the spectroscopy and biochemistry data by structural modeling of the wild-type and three mutants. Taken together, the experimental data, bioinformatics sequence analyses, and structural modeling suggest that the tyrosine/aspartate complex counterion contributes to a complex water-mediated hydrogen-bonding network that couples the protonated retinal Schiff base to an extracellular carboxylic dyad.

University of Guelph 10.13039/100008986 NA Deutsche Forschungsgemeinschaft 10.13039/501100001659 SFB1078 Natural Sciences and Engineering Research Council of Canada 10.13039/501100000038 NA document-id-old-9jp3c04032
document-id-new-14jp3c04032
ccc-price
==== Body
pmc1 Introduction

Microbial rhodopsins continue to display stunning diversity of functions and structures for more than two decades, owing to the rapid advances in genomics and metagenomics.1−6 These versatile retinal-binding, light-sensitive proteins modify and reuse the same seven-transmembrane helical template with a retinal chromophore attached to the seventh helix via the Schiff base. Modifications of this template are used to diversify their color, functions, ion selectivity, functional pH range, kinetics of response to light, and other characteristics. As a result, we see many examples of divergent evolution, which create various ion pumps and channels, photosensors, and light-activated enzymes, presumably from a common ancestor. In addition, we see examples of convergent evolution where the same function is created multiple times in unrelated or very distant ancestral protein groups. For example, both schizorhodopsins and xenorhodopsins are inward proton pumps,7−10 while haloarchaeal/cyanobacterial halorhodopsins and eubacterial NTQ-rhodopsins are inward chloride pumps.11−14 Both pairs are very distant in terms of sequence homology and functional motifs. Here we introduce another possible case of such convergent evolution by characterizing a new group of sensory rhodopsins, distinct from the well-known sensory rhodopsins (SRs) of haloarchaea but likely interacting with the same type of Htr transducers. While haloarchaeal SRs are close relatives of the most studied halobacterial proton pump, bacteriorhodopsin (BR),15,16 the new group appears to be related to eubacterial proton pumps, proteorhodopsins (PR).17−19 That sensory proteorhodopsins exist has been suggested before,16,20 but only for proteins with primary structures and putative transducers distinct from those of the group we characterize in this paper.

Haloarchaeal SRs have been known for more than 40 years, with the two main types, SRI and SRII, originally found in Halobacterium salinarum.21−24 An additional type of haloarchaeal SR was identified in Haloarcula marismortui,25 and two SRI-like species were found outside of haloarchaea, in eubacterium Salinibacter ruber, most likely appearing as a result of lateral gene transfer.26 The properties of SRI and SRII (learned mainly from H. salinarum and Natronobacterium pharaonis proteins) have been reviewed extensively (e.g., refs (27−35)), and we briefly summarize them here for the sake of comparison with the new group. While SRII serves as a receptor for the photophobic response to blue light and accordingly has an absorption maximum around 490 nm, SRI has a dual function of a phototactic and photophobic sensor (via its dark state absorbing around 590 nm and its M photointermediate with a deprotonated Schiff base absorbing near UV light, respectively). Both types of SR interact with dimeric Htr type transducers, HtrI and HtrII, whose genes are adjacent to the genes of SRs. The Htrs are similar to the well-known chemosensory methyl-accepting transducers, which have two transmembrane helices and a large intracellular C-terminal domain. These intracellular domains of Htrs transmit the signal received via interaction of the transmembrane helices with the cognate receptors to a histidine kinase CheA (in complex with CheW), which modulates the activity of phosphoregulatory CheY, ultimately affecting the flagellar motor. Haloarchaeal SRs retain the carboxylic primary proton acceptor of BR (homologue of Asp85) but replace the primary proton donor (homologue of Asp96) with Tyr or Phe. Accordingly, their photocycles are slow, which is thought to be beneficial for transmitting the light signal to their transducers.

In 2004, a new type of putative sensory rhodopsin with unusual amino acid sequence was identified in the genome of photosynthetic purple bacterium Thermochromatium tepidum.17 From the sequence analysis of the retinal-binding pocket, the absorption maximum of this protein was predicted to be around 480 nm, and this protein was called SRII. Even though the rhodopsin and its transducer have not been expressed or characterized, the sensory function was predicted from the proximity of its gene to the cluster of genes normally associated with SR’s signal transduction cascade, namely, Htr, CheY, CheA, and CheW. It was proposed that this rhodopsin, rather than photoactive yellow protein (PYP) mediates negative phototaxis of its host. More recently, several homologues of this unusual protein have been identified in various environmental and viral samples,18,19 but none have been characterized biophysically so far. We report on the expression and biophysical characterization of members of this new group, which we call Proteo-SRs, to stress their evolutionary relationship with proteorhodopsins. We describe their unique primary structure (with the notable tyrosine in place of carboxylic Schiff base counterion on helix C), hydrogen-bonded networks, and unusual retinal configuration and probe their interaction with the putative transducers.

2 Methods

The sample preparation procedures followed our protocols developed for the Antarctic rhodopsin9 with minor modifications, as follows.

2.1 Protein Expression

Genes encoding the wild-type (see section 3.2 for the sources) and mutant (produced by GenScript from the wild-type template) Proteo-SRs and their C-terminally truncated transducers (at the position homologous to 114 of NpHtrII) were cloned into pET21a(+) vector (EMD Millipore, Billerica, MA) by GenScript (Piscataway, NJ) using NdeI–XhoI restriction sites, which encodes a C-terminal 6× His-tag after the LE insert. Lucigen’s heat-shock transformation protocol was used to transform the Escherichia coli C41(DE3) OverExpress Chemically Competent Cells (Lucigen, Middleton, WI) with the plasmids. After thawing the E. coli C41 cells on ice for 10–15 min, 50 ng of DNA was added to 50 μL of cells, and the cells were incubated on ice for 30 min. The cells were heat shocked in a 42 °C water bath for 45 s, and then incubated on ice for 2 min. The cells were added to 950 μL of 2 × YT media (1% yeast extract, 1.6% tryptone, 1% NaCl) and incubated at 37 °C and 250 rpm shaking for 1 h. 100 μL of cell culture was spread onto 2 × YT media agar plates (2 × YT media with 1.5% agar and 0.1 mg/mL ampicillin at pH 7.0) and incubated overnight at 37 °C. During cell growth stages, 0.1 mg/mL ampicillin was always added to the 2 × YT media. For the small-scale colony screening, six isolated bacterial colonies were selected from plates to inoculate 2 mL of 2 × YT media for incubation at 37 °C and 240 rpm overnight and transferred to 25 mL 2 × YT media. When the 25 mL of culture reached an optical density at 600 nm (OD600, measured by Cary50, Varian) of ∼0.4, it was induced by 1 mM isopropyl-β-d-thiogalactoside (IPTG) and 7.5 μM all-trans-retinal. The induced cell culture was incubated for 4 h at 37 °C and 275 rpm. The cells were collected through low-speed centrifugation at 4680 × g and 4 °C. The colony with the most intense blue color and largest pellet was selected as the optimal colony for protein expression. This colony was grown in 1 L of 2 × YT media using shake flask cultures. The colony material was used to inoculate 2 mL of media and was incubated overnight at 37 °C and 240 rpm. 2 mL of cell culture was used to inoculate 25 mL of media which was then incubated overnight at 37 °C and 240 rpm. The 25 mL of cell culture was used to inoculate 1 L of media to an OD600 of 0.1 and was incubated at 37 °C and 240 rpm. Once an OD600 of ∼0.4 was reached, 1 L of cell culture was induced with a total concentration of 1 mM IPTG and 7.5 μM all-trans-retinal and incubated for 4 h at 37 °C and 275 rpm. The cells were collected with low-speed centrifugation at 4680 × g and 4 °C for 10 min. The cell pellet was resuspended in 150 mM NaCl and was centrifuged again at 4680 × g and 4 °C for 10 min. The cell pellet was resuspended in 40 mL of lysis buffer (150 mM NaCl, 0.05 M Tris base, 1 mM MgCl2, 2 μg/mL DNase I, 0.2 mg/mL lysozyme, pH 7.2). The suspended cells were left to shake (IKA VIBRAX) at 400 rpm at room temperature for 3 h and then frozen for future use.

2.2 Protein Purification and Lipid Reconstitution for Spectroscopic Studies

Spectroscopy was performed on His-tag-affinity-purified lipid-reconstituted samples. The thawed cells were sonicated (Fisher Sonic Dismembrator Model 500) to produce membrane fragments. The sonicated sample was spun down at 4680 × g and 4 °C for 10 min to get rid of unbroken cells, and the supernatant was collected. The membrane fragments were then sedimented using ultracentrifugation at 150,000 × g and 4 °C for 50 min. The membrane pellet was resuspended in 5 mM Tris base, 1% DDM, pH 7.5, and was left to stir overnight at 4 °C in the dark. The insoluble membrane debris was removed with the ultracentrifuge at 150,000 × g and 4 °C for 1 h. The supernatant was collected, and the protein yield was estimated spectrophotometrically (Cary50) using the extinction coefficient of Anabaena sensory rhodopsin (ASR),36 to determine the amount of nickel-nitrilotriacetic acid (Ni2+-NTA) resin (Qiagen) to be added. The resin was washed with ∼200 mL of 150 mM NaCl on a membrane filter (Thermo Scientific Nalgene MF75 filter, 0.8 μm pore size), resuspended in 10× binding buffer (3 M NaCl, 0.5 M Tris base, 0.05% DDM, pH 8), and added to the solubilized protein sample to stir overnight in the 4 °C chromatography fridge. A chromatographic column was used to wash the protein bound to the resin (normally, four times) with 25 mL of washing buffer (0.3 M NaCl, 0.05 M Tris, 5 mM imidazole, and 0.05% DDM, pH 8) to remove any contaminating proteins. Next, 30 mL of elution buffer (0.3 M NaCl, 0.05 M Tris, 0.5 M imidazole, 0.05% DDM, pH 8) was added to the column and was collected after 10 min of incubation. Typically, a total of 30 mL of eluted protein was collected, and this sample was syringe-filtered (0.22 μm pore size) to remove any large debris or accidentally collected resin. An Amicon Ultra 50 mL centrifugal filter tube was used to concentrate the sample down to a small volume through repeated centrifugation at 4000 × g and 4 °C for 20 min. The buffer was exchanged to reconstitution buffer (5 mM NaCl and 10 mM Tris, pH 8) through washing it with a total volume of approximately 30 mL by repeated centrifugation in the filter tube. The liposomes used for reconstitution were made of 1, 2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) and 1, 2-dimyristoyl-sn-glycero-3-phosphate (DMPA) in a 9:1 w/w ratio. The lipids were dissolved in chloroform and stirred for 90 min, and the chloroform was removed under vacuum inside a desiccator for at least 4 h. The lipid film was rehydrated with 1 mL of reconstitution buffer to form a 11.1 mg/mL suspension. The solubilized, concentrated pure protein was combined with the lipid suspension at a 2:1 protein to lipid ratio, and Triton X-100 stock (0.2 mg/μL) was added to the sample to a final concentration of 0.8 mg/mL, at which a noticeable decrease in turbidity was observed. After at least 6 h of stirring at 4 °C, 0.8 g/mL BioBeads SM-2 (Bio-Rad) was added to the sample to absorb the detergent and the sample was mixed on the Orbitron Rotator for at least 24 h. A 27 G needle (BD) was used to remove the reconstituted protein, and the BioBeads were washed with reconstitution buffer to collect most of the reconstituted protein. The sample was centrifuged at 150,000 × g and 4 °C for 1 h to collect proteoliposomes.

2.3 Sample Preparation and Experimental Setup for Time-Resolved and Static Spectroscopy in the Visible and Infrared Ranges and Raman Spectroscopy

Flash photolysis experiments were run on the proteoliposomes encased within polyacrylamide gels. The gels were prepared using 700 μL of the proteoliposome suspension, 300 μL of 33% acrylamide and 1% bis(acrylamide) solution mixture, 2.4 μL of 10% ammonium persulfate, and 3 μL of N,N,N′,N′-tetramethylethylenediamine. After solidification, the gels were washed with 1 L of distilled water for at least 4 h at room temperature and stored at 4 °C in distilled water. The pH and/or salt conditions were set by soaking the gel in a buffer (100 mM NaCl plus 50 mM buffer appropriate for the pH range–potassium acetate for 4 and 5, MES for 6, Bis-Tris-Propane for 7, TRIS for 8, CHES for 9, and CAPS for 10 and 11) for at least 90 min prior to the measurement.

Static spectroscopy was performed on a Cary50 spectrophotometer (Varian). Flash-photolysis spectroscopy was performed using a custom-built single-wavelength spectrometer described elsewhere.37 In short, the photocycle was initiated with 7 ns pulses of the second harmonic of an Nd:YAG laser at 532 nm (Continuum Minilite II). Absorption changes of the monochromatic light (provided by an Oriel QTH source and two monochromators) were observed by using an Oriel photomultiplier, an amplifier with a 350 MHz bandwidth, and a Gage AD converter (CompuScope 12100–64M). Kinetic traces were averaged (normally, 200–1000 traces) and converted into a quasi-logarithmic time scale using in-house software.

Raman spectra were collected on 5 μL of a wet paste of the proteoliposomes hydrated with desired buffer (identical to those used for the visible spectroscopy) using FRA106/s accessory to the Bruker IFS66vs spectrometer, with Nd:YAG laser excitation at 1064 nm, at a 4 cm–1 resolution, with the OPUS software, at least 1000 scans averaged per sample.

Light-induced FTIR difference spectroscopy experiments were conducted using a Bruker Vertex 80v FTIR spectrometer essentially as described previously.38 Briefly, difference spectra (under 525 nm LED illumination) were acquired at 2 cm–1 spectral resolution at a 40 kHz scanner velocity. Time-resolved difference spectra were recorded by using the rapid-scan mode at 4 cm–1 spectral resolution. The scanner velocity was set to 280 kHz and the spectral range restricted to 2257–0 cm–1 by an interference filter. Under those conditions, recording a single-sided interferogram requires ∼20 ms, yielding the first spectrum after pulsed laser excitation (Δt = 10 ns, λ = 532 nm, Eexc = 3 mJ/cm2, Minilite II, Continuum). The proteoliposome suspension in 10 mM CHES, 5 mM NaCl at pH 9 was gently sonicated at 30 °C for 15 min and dried on a BaF2 window. Rehydration was achieved by equilibration with drops of H2O/glycerol (90:10 w/w) placed next to the protein film. Deuteration of the protein film was achieved by exchanging the H2O/glycerol droplets with D2O/glycerol-d3 (90:10 w/w).

2.4 Co-Reconstitution of S22.Bin169 Proteo-SR with Its Transducer

S22.Bin169 Proteo-SR transducer (Proteo-TR) was transformed, expressed, purified, and concentrated using the protocol described above for Proteo-SRs, except that the induced cell culture was incubated for 6 h at 37 °C and 275 rpm, and no retinal was added. The S22.Bin169 Proteo-SR and its transducer were expressed and purified separately, mixed, and co-reconstituted in the liposomes including DMPC and DMPA in a 9 to 1 w/w ratio as described above. The lipid film was rehydrated with reconstitution buffer, resuspended, and added to the concentrated pure proteins (protein to lipid ratio 2:1) and Triton X-100 stock (0.2 mg/μL) and then stirred for 6 h at 4 °C. To absorb the detergent, 0.8 g/mL BioBeads SM-2 (Bio-Rad) was added to the sample, and it was mixed for 24 h on the Orbitron Rotator in the 4 °C chromatography fridge. After washing the BioBeads with reconstitution buffer (5 mM NaCl, 10 mM Tris, pH 8) and using a 27 G needle (BD) to remove the reconstituted proteins, the proteoliposomes were collected using ultracentrifugation at 150,000 × g and 4 °C for 1 h. The actual ratio of the reconstituted proteins was verified by running SDS-PAGE of proteoliposomes and quantifying the density of the respective bands.

2.5 Ion Transport Assays

Ion transport assays were performed on S13.Bin138 Proteo-SR in whole E. coli cells grown in 1 L of culture as described above and collected at 4680 × g and 4 °C for 10 min. The cells were washed three times in the unbuffered solution, 10 mM KCl, 10 mM MgSO4, and 100 μM CaCl2. One-third of the cell culture was resuspended in 25 mL of the unbuffered solution. Gently stirred cell suspensions were illuminated (Cole Parmer 9741–50 illuminator) with yellow light (>460 nm) using a glass filter and the pH changes were monitored using a glass electrode (Accumet Microprobe Extra Long Calomel Combo Electrode) and recorded using a digital oscilloscope (Agilent Technologies DSO 1052B Digital Storage Oscilloscope). 10 μM of the proton uncoupler CCCP (from ethanol stock) was added in control experiments.

2.6 Molecular Dynamics (MD) Simulations

2.6.1 Structure Modeling of S13.Bin138 Proteo-SR

Structural modeling of S13.Bin138 Proteo-SR with ColabFold39 using standard settings gave five structural models with pLDDT (predicted local distance difference test) scores between 93.9 and 95.1, suggesting high confidence in structural prediction for all models. Separately, we used Modeller 10.440,41 to generate a homology model of retinal-bound S13.Bin138 Proteo-SR based on the structure of Exiguobacterium sibiricum rhodopsin (Protein Data Bank 4hyj), which was among the ColabFold-selected templates. The Modeller-generated structure was overlapped onto one of the ColabFold structural models using Visual Molecular Dynamics, VMD.42 S13.Bin138 Proteo-SR bound to all-trans-retinal was geometry optimized using Chemistry at Harvard Molecular Mechanics (CHARMM);43,44 during geometry optimizations, the coordinates for the retinal molecule, Ser51, His55, Tyr90, Phe166, Tyr199, Asn229, and Lys230 were allowed to change, whereas all other protein coordinates were fixed. The Y90F, E134Q, and E219Q mutants were prepared using CHARMM to mutate and geometry optimize the corresponding side chain.

2.6.2 Protonation States

The protonation states of specific titratable side chains of S13.Bin138 Proteo-SR are unknown. Although, in principle, one could use software such as PROPKA45 to derive clues about likely protonation states, the fact that the predicted likely protonation state for an active-site carboxylic side chain of a microbial rhodopsin (Glu162 of the C1C2 channelrhodopsin chimera)46 was found to be inconsistent with subsequent spectroscopy47 underlines the challenges with relatively simple pKa estimations for active-site residues of retinal proteins. As a first step in our consideration of the protonation states of internal titratable side chains of S13.Bin138 Proteo-SR, here we relied on the conservation of internal titratable side chains whose protonation states have been documented by experimental data on BR.

In BR, Asp115 is protonated throughout the entire reaction cycle,48 and it is within H-bond distance from the Thr90 hydroxyl group in static crystal structures49,50 and throughout atomistic MD simulations.51 Asp115 and Thr90 are conserved in S13.Bin138 Proteo-SR as Asp123 and Thr95, respectively, and by analogy with BR, in all simulations performed, Asp123 was protonated (neutral). Likewise, in BR Asp212 is negatively charged in the BR resting state, and the corresponding Asp226 of S13.Bin138 Proteo-SR was considered negatively charged in all simulations. The BR extracellular proton release group, which includes Glu194 and Glu204, stores one proton;52 since in S13.Bin138 Proteo-SR the extracellular Glu134 and Glu219 are close not only to Arg87 (corresponding to BR Arg82) but also to Arg218, in the absence of clear spectroscopy data indicating protonation of Glu134 or Glu219, we treated both residues as negatively charged. Nevertheless, we indirectly evaluated the effect of the protonation state of either Glu side chain by studying the Glu134Q and Glu219Q mutants.

His55 is located in a key functional site near the Asp226 helix G counterion. We considered the protonation state of His55 by performing two independent simulations of the wild-type all-trans- protein: one simulation with His55 neutral and treated as −Nδ1 tautomer, and the second simulation, with His55 neutral and treated as −Nε2 tautomer. For all other titratable side chains we considered standard protonation states, i.e., Asp and Glu are negatively charged, His are neutral and −Nδ1 tautomers, and Arg and Lys are positively charged.

2.6.3 Simulation Systems and Simulation Protocol

The protein was embedded in a hydrated POPC lipid bilayer using VMD and CHARMM.44 In each simulation system, the lipid bilayer contains 365 lipids and 33473 water molecules, and chloride ions are added for charge neutrality. We used CHARMM36 for the protein, lipid molecules, and ions,53,54 the TIP3P water model, and retinal parameters from refs (55) and (56). We used a switch function between 10 and 12 Å to switch off short-range nonbonded interactions, and smooth particle mesh Ewald summation for Coulomb interactions.57,58

All MD simulations were performed using NAMD.59−61 Geometry optimization, heating to 300 K, and the first 1 ns of equilibration were performed using soft harmonic restraints of 4 kcal/molÅ–2 on the backbone heteroatoms, 2 kcal/molÅ–2 on side chains, and 1 kcal/molÅ–2 on water oxygen atoms and lipid headgroups. We continued the equilibration with 4 steps, 1 ns each, during which we reduced stepwise the magnitude of the soft harmonic restraints. The harmonic restraints were halved during the second step; in the third step, we placed harmonic restraints of 1 kcal/molÅ–2 on the backbone heteroatoms, on the retinal, solvent water oxygen atoms, and ions; we reduced all these constraints to 0.5 kcal/molÅ–2 for the fourth equilibration step, and switched off the restraints on protein and retinal atoms in the fifth equilibration step. All restraints were then switched off and the production runs were initiated. Heating was performed in the NVT ensemble (constant number of atoms N, constant volume V, and constant temperature T); equilibration and production runs were performed in the NPT ensemble (constant N, constant pressure P, constant T) using a Langevin dynamics62 scheme and Nosé–Hoover piston. During heating and the first 1 ns of equilibration we used an integration step of 1 fs; all subsequent equilibration steps and the production runs were performed using a multiple time integration scheme63 with 1 fs for the bonded forces, 2 fs for short-range nonbonded, and 4 fs for long-range nonbonded. The lengths of all covalent bonds to H atoms were fixed.64

The production runs were prolonged to 398.11 ns for the reference wild-type simulation, 361 ns for the wild-type simulation with His55 described as -Nδ1 tautomer, 332.44 ns for the Y90F mutant, 321 ns for the E134Q mutant, and 311 ns for the E219Q mutant. The total sampling time of the five production runs performed is 1.72 μs. Average values were computed based on 20,000 equally spaced coordinate snapshots from the last 200 ns of each simulation.

2.6.4 H-Bond Graph Analyses

H-Bond graphs were computed using the Bridge2 graphical user interface.65,66 H-Bonds were computed using standard distance and angle criteria: we consider that two groups are H-bonded when the distance between the H-bond donor and acceptor heteroatoms is within 3.5 Å, and the H-bond angle is within 60°. The occupancy of an H-bond is given by the percentage of the coordinate sets used in H-bond graph computations in which the H-bond criteria are met. The average number of water molecules for an edge (H-bond connection) between two nodes (H-bonding side chains) of the H-bond graph has values between 0 (direct H-bond between the two side chains) and 3 (three H-bonded water molecules bridge the two side chains). Thus, two nodes (protein side chains) connected by an edge for which the average number of water molecules in the bridge is below 1.0 connect mostly via direct H-bonding or one water molecule.

For each simulation, we computed H-bond graphs for direct H-bonds between protein side chains and water-mediated bridges between side chains; for simplicity, we considered water-mediated bridges with up to three H-bonded water molecules. H-bond graphs were computed based on 20,000 equally spaced coordinate snapshots from the last 200 ns of each simulation.

3 Results and Discussion

3.1 Identification and Sequence Analysis of the New Group of Sensory Rhodopsins

Using the sequence of the putative uncharacterized SRII previously found in the genome of Thermochromatium tepidum by Kyndt et al.17 as a BLAST query, we identified a number of its close homologues in the public genome and metagenome databases (http://www.ncbi.nlm.nih.gov/protein and http://img.jgi.doe.gov).67 The homologues form a very deep branch on the phylogenetic tree of microbial rhodopsins (Figure 1), implying a very clear separation of this subgroup in terms of sequence homology (see below), which nevertheless clusters with proteorhodopsins but not with the haloarchaeal SRs. Thus, we call this subgroup of microbial rhodopsins Proteo-SRs, following the earlier suggestion on their function17 based on the proximity of the T. tepidum SRII gene to Htr, Che A, CheW, and Che Y genes. This gene cluster also reproduces in the homologues of the putative T. tepidum SRII explored by us, in cases where the genomic context is available. In some cases, genes for retinal biosynthesis can be observed nearby as well.

Figure 1 Unrooted phylogenetic tree of selected microbial rhodopsin sequences showing the position of Proteo-SRs (marked with a star) relative to other major groups, with Proteo-SRs and haloarchaeal SRs highlighted in red. The sequences were aligned using CLUSTALO,68 and the phylogenetic tree was visualized by Dendroscope.69

Considering the very deep branching of Proteo-SRs, an additional assurance of their evolutionary relationship with PRs from a more detailed sequence analysis of conserved residues is warranted. One can see several residues known to be involved in the extensive hydrogen-bonding in PRs conserved in Proteo-SRs (Figure S1). For example, a homologue of GPR’s Glu142 (BR’s Ala126) in helix D is conserved, as are two asparagines in helix G, one of which precedes the retinal-binding lysine.70−73 On the other hand, the residues known to be important for haloarchaeal SRs signaling (homologues of Thr189, Tyr199, and Thr204 of N. pharaonis SRII,74 and Tyr210 of H. salinarum SRI)75 are not conserved in Proteo-SRs. They also lack the aromatic Tyr/Phe signature in place of the cytoplasmic proton donor (Asp96 of BR) typical for all classical SRs,15 replacing it with methionine.

Several conserved sequence regions of Proteo-SRs are shared with another deeply branched group of unusual PR-related proton pumps with a lysine proton donor but not with PRs in general (Figure S1). These are so-called DTK pumps, which include Exiguobacterium sibiricum rhodopsin (ESR) and its close homologues.73,76 Among others, the cytoplasmic end of helix C and the extracellular end of helix G display unusual shared motifs of FPLLxG and LRE, respectively. The highest similarity of Proteo-SRs to the DTK rhodopsins is supported by the BLASTP scores as well as recent phylogenetic analysis of metagenomes.19 Specifically, S13.Bin138 Proteo-SR shows 34% identity (50% similarity) to ESR, but only 24% identity (39% similarity) to Salinibacter ruber xanthorhodopsin (XR), and 22% identity (40% similarity) to EBAC31A08 GPR. This may suggest that both DTK pumps and Proteo-SRs are very evolutionary old groups, which have a common ancestor with PRs, but did not necessarily originate from them.

In addition to the residues shared with PRs and ESR, and a number of residues conserved among many microbial rhodopsins (such as homologues of BR’s R82, W86, T90, D115, W182, P186, D212, and K216), Proteo-SRs display many subgroup-specific residues (Figure S1). The carboxylic counterion and the primary proton acceptor on helix C (Asp85 in BR) is replaced by Tyr, which may create an unusual geometry of H-bonds in the Schiff base region. Together with the replacement of the carboxylic proton donor (Asp96 in BR), it creates a unique Y(T/C)M characteristic motif on helix C.77 Helix B has low homology to other microbial rhodopsins: an Asn replaces BR’s Thr46 (partner of the proton donor), and a His replaces BR’s Tyr57. Notably, a His in the preceding position, which is conserved among PR-like rhodopsins and forms the complex His/Asp counterion72,78−80 is missing (Figure S1), in agreement with its Asp partner being replaced by Tyr. There are also prominent stretches of aromatic amino acids in the extracellular side of helix B and the cytoplasmic side of helix F, which may play a role in protein–protein or protein–lipid interactions. In what follows, we use site-directed mutagenesis experiments, electronic and vibrational spectroscopy, sequence comparisons, and structural modeling to explore the interactions and putative roles of unique residues of one of the Proteo-SRs.

3.2 Expression and Basic Spectroscopic Characterization of Proteo-SRs

First, we attempted to express the original Proteo-SR from T. tepidum and its homologue from another purple sulfur bacterium (specifically, Chromatiaceae), Imhoffiella purpurea. We could not obtain their appreciable expression in C41(DE3) E. coli, either for the original sequences (C-terminally 6-His-tagged) or for the constructs with the 30 N-terminal amino acids of GPR added to the N-termini of Proteo-SRs. Next, using the same N-terminal leader addition strategy, we attempted to express a homologue from another purple sulfur bacterium, Chromatium okenii, and from Gammaproteobacteria bacterium isolate CSSed165 cm_20 (metagenome assembled genome from a Siberian soda lake). While the C. okenii rhodopsin did not show any expression, similar to the other Chromatiaceae, the metagenomic CSSed165 cm_20 homologue produced blue E. coli cells, suggesting a robust expression of a rhodopsin with the absorption maximum of >570 nm. Indeed, the absorption maximum of the purified solubilized CSSed165 cm_20 Proteo-SR (in 0.05% DDM, 300 mM NaCl, 50 mM Tris, pH 8) is at 579 nm (not shown). Encouraged by this successful expression, we found two more close homologues of the CSSed165 cm_20 Proteo-SR, from Gammaproteobacteria bacterium isolates S13.Bin138 and S22.Bin169 (metagenome assembled genomes from a Tibetan saline lake), which also showed robust expression even without the N-terminal leader of GPR (see below). The three expressed homologues share several amino acid sequence differences from the Chromatiaceae homologues, possibly responsible for their improved expression in E. coli (Figure S1), most noticeably the YCM Helix C motif, as opposed to the YTM motif for the rest of the group. From this point, we will mainly focus on biophysical characterization of the S13.Bin138 Proteo-SR (and to the lesser extent its very close homologue S22.Bin169 Proteo-SR), which were chosen for their good expression levels as well as the availability of the complete genomic context with the Htr-like transducer proteins.

The absorption maximum of purified solubilized S13.Bin138 Proteo-SR at pH 8 is 580 nm (Table S1), whereas that of the lipid-reconstituted protein ranges from 570 to 580 nm at pH 4–9 (Table S1 and Figure 2A). Most of that acid-induced blue shift of the absorption maximum occurs between pH 4 and 6, suggesting that a group with pKa close to 5 is titrated. As expected from the high sequence homology, S22.Bin169 Proteo-SR shows a very similar absorption maximum (578 nm in lipids at pH 9, Figure S2B), which suggests that at the likely physiological conditions (alkaline pH and high salt),81,82 Proteo-SRs show the absorption maxima more characteristic of SRI rather than SRII, in contrast to the 480 nm absorption maximum predicted based on the sequence analysis of the T. tepidum homologue.17 This hints that the physiological function of Proteo-SRs may be phototactic rather than photophobic, consistent with the photosynthetic capabilities of their hosts, which possess bacteriochlorophyll-based reaction centers.

Figure 2 Spectroscopic characterization of the dark state of S13.Bin138 Proteo-SR. (A) pH dependence of the absorption spectrum of the lipid-reconstituted protein. (B) FT-Raman spectrum of the lipid-reconstituted protein in H2O and D2O based buffers at pH/pD 9.

Next, we explored the configuration of the Proteo-SRs chromophore with FT-Raman spectroscopy. At the physiologically relevant pH of 9, S13.Bin138 Proteo-SR (Figure 2B) and S22.Bin169 Proteo-SR (Figure S3A) show almost identical Raman spectra with two unusual features, suggesting a unique retinal conformation. First, the ethylenic stretching band is strongly split (the main peak is at 1513 cm–1 and the secondary peak is at 1536 cm–1, almost equidistant from the ∼1525 cm–1 value expected from the well-known correlation with the visible absorption maximum).83 The secondary higher-frequency peak is strongly D2O-dependent (presumably being coupled to NH vibrations), shifting to 1527 cm–1. This behavior is reminiscent of recently observed Raman spectra of heliorhodopsins,84 where it was interpreted as originating from a linear rather than bent retinal polyene chain, but the ethylenic stretch band split is even more pronounced in Proteo-SRs. Second, there is a prominent unusual D2O-dependent HOOP (hydrogen out-of-plane) band at 980 cm–1, with a number of smaller D2O-independent bands at 959, 880, and 833 cm–1, which suggest strongly twisted retinal skeleton, especially in the vicinity of the retinal Schiff base.85 Similar (but apparently weaker) retinal twist was observed in schizorhodopsins, suggesting that it may be caused by direct hydrogen-bonding of the Schiff base nitrogen to the single counterion aspartate on helix G,86 which may also be the case for Proteo-SRs, where the other counterion carboxyl on helix C is replaced by tyrosine. The suggested linear (unbent) shape of the chromophore and its polyene chain twist (nonplanarity) are not mutually exclusive, as the former refers to bond angles and the latter to dihedral (torsion) angles of retinal. This retinal geometry is also supported by the MD simulations (see section 3.5). The Schiff base C=N stretching vibrations are observed at 1636 cm–1, shifting to 1615 cm–1 in D2O, almost identical to SRI from H. salinarum.87 The fingerprint C–C stretching region suggests predominantly all-trans-retinal (1199 and 1165 cm–1 peaks),88 even though a minor 1183 cm–1 peak indicative of 13-cis configuration is observed (Figure 2B). It should be noted that a similar minor peak in the related proton pump GPR did not seem to originate from 13-cis-retinal.89,90

To test if the spectral blue shift of 10 nm observed at low pH (Figure 2A) originates from the accumulation of the 13-cis chromophore, similar to the Antarctic schizorhodopsin AntR,9 we recorded a Raman spectrum of S13.Bin138 Proteo-SR at pH 4 (Figure S3B). While the ethylenic C=C stretch showed the expected minor upshift of ∼1 cm–1 correlated with the visible absorption shift, no pH-dependent changes in the isomeric configuration of retinal were observed. This suggests that the visible absorption change originates from a titration of some side chain in the chromophore vicinity not accompanied by retinal isomerization.

3.3 Photochemical Reactions of Proteo-SRs

S13.Bin138 and S22.Bin169 Proteo-SRs do not display the light-adaptation/dark-adaptation process typical for BR and some other microbial rhodopsins, in which the main absorption maximum shifts due to different isomeric compositions of the dark-adapted and light-adapted states.91,92 Instead, strongly blue-shifted (absorption maximum around 365 nm) long-living (characteristic decay times of ∼20 s) species with deprotonated Schiff base accumulate upon prolonged illumination (Figure S2). Such strongly blue-shifted species are well-known for various channelrhodopsins, where they are responsible for channel inactivation and likely originate from secondary photochemical processes.93−95

The photocycle of S13.Bin138 Proteo-SR at physiological pH (pH 9) is relatively slow (turnover time constant of ∼1 s) (Figure 3A), while that of S22.Bin169 is somewhat faster (turnover time constant of ∼0.25 s) (Figure S4). Both are very similar and can be described by a sequence of several spectral states, which include red-shifted and blue-shifted species, but do not show apparent Schiff base deprotonation (no M-like intermediate, in contrast to haloarchaeal sensory rhodopsins33). In agreement with that, we have not observed any light-driven proton transport in the unbuffered suspension of E. coli cells overexpressing S13.Bin138 Proteo-SR (data not shown).

Figure 3 Photocycle analysis for S13.Bin138 Proteo-SR at pH 9. (A) Laser-induced absorption difference kinetics measured every 20 nm. Solid lines are 5-exponential fits. (B) Spectra of the kinetic components extracted by the global multiexponential fit (see text for details). The spectrum labeled “initial” is a difference spectrum at the beginning of the fit, and the rest are decay-associated spectra.

Global multiexponential fitting of the single-wavelength kinetic traces (performed by FITEXP96) identifies the following photocycle transitions for S13.Bin138 Proteo-SR at pH 9 (Figure 3B). The initial spectrum represents the difference spectrum at 200 ns, reflecting the depletion of the dark state and formation of a blue-shifted (L-like) state. The following decay-associated spectra reflect transitions between various states, where positive amplitudes correspond to the decaying species and negative amplitudes to the forming species. The first kinetic component (τ ∼ 2.7 μs) shows further formation of the L-like state (with the difference maximum at ∼500 nm, which we call L1) from the mildly red-shifted K-like one. The second kinetic component (τ ∼ 26 μs) and the third minor component (τ ∼ 1.9 ms) reflect the transition from the K/L1 mixture to a more strongly blue-shifted state (with the difference maximum at ∼460 nm), which we will call L2. Finally, the fourth and the fifth kinetic components (τ ∼ 0.18 and ∼1s) show the decay of the L1/L2 mixture (with the difference maximum at ∼480 nm) to the parent dark state. The photocycle of S22.Bin169 Proteo-SR shows very similar components (Figure S4), with the respective time constants of ∼3.6 and ∼18 μs, ∼1.7 and ∼51 ms, and ∼0.25 s.

Despite the observed 10 nm blue-shift of the absorption spectrum at low pH (Figure 2A), the photocycle of S13.Bin138 Proteo-SR lacks strong pH-dependence (Figure S5). The photocycle measured at pH 4 is marginally faster than at 9, with the last two components of ∼0.1 and ∼0.7 s, but is similar otherwise. Taken together with the Raman data (Figure S3B), it suggests that the titration of a side chain affecting the color does not have a major effect on the protein photochemistry in other respects.

Next, we investigated the late intermediates of the S13.Bin138 Proteo-SR photocycle at pH 9 by steady-state (LED 525 nm illumination) and time-resolved (rapid-scan mode with laser excitation) light-induced difference FTIR spectroscopy (Figures 4 and S6). The rapid-scan FTIR data show that the laser-induced absorption difference decays in two almost identical spectral components (Figure S6) with characteristic times of ∼363 ms and ∼2.5 s, in good agreement with the two last components revealed by the visible spectroscopy (Figure 3). These two kinetic components with identical infrared and visible spectra likely reflect the mixture of the last two intermediates decaying to the dark state. We compared the FTIR spectra of these two components to the steady-state difference spectrum acquired under continuous illumination and found them to be virtually identical as well (Figure S6B). For the sake of more detailed analysis of the spectral features, we repeated the steady-state measurements in the D2O-based buffer (Figure 4).

Figure 4 Light-induced difference FTIR spectra of S13.Bin138 Proteo-SR at pH 9 acquired under continuous LED illumination at 525 nm in H2O (black) and D2O (red) based buffers.

The steady-state light-minus-dark difference spectra of S13.Bin138 Proteo-SR faithfully reproduced spectral features of the retinal skeleton seen by Raman spectroscopy (Figure 2B). Specifically, one can observe the negative double ethylenic stretch at 1534 (strongly D2O-dependent) and 1512 cm–1, negative C–C stretches corresponding to the all-trans-retinal stretch at 1198 and 1164 cm–1, and strong negative D2O-dependent HOOP at 979 cm–1 (Figure 4). Surprisingly, no positive C–C stretches are seen, giving the observed spectrum an M-like character, which is further confirmed by the putative C=N stretch of the Schiff base (negative at 1634 cm–1, positive at 1622 cm–1). This contradicts the results of the visible spectroscopy, suggesting that the last two intermediates have a maximum around ∼470 nm, which is unlikely to correspond to the deprotonated Schiff base. There is a remote possibility that a late intermediate with the deprotonated Schiff base exists but is very strongly blue-shifted and has low extinction coefficient, precluding its detection by flash-photolysis. This seems to be contradicted by the fact that we do detect M-like intermediates in some mutants (section 3.4 below). An alternative explanation is that the retinal Schiff base is protonated in those late intermediates, but the intensity of the C–C stretching vibrations is low in this unusual complex counterion environment, resulting in no positive bands.

The spectra also show perturbations of the protein backbone (Amide I negative bands at 1664 and 1657 cm–1 and positive at 1645 cm–1), lipid esters (D2O-independent positive band at ∼1740 cm–1), and some side chains. Specifically, we observed perturbation of a cysteine (negative at 2560 cm–1, positive at 2569 cm–1), for which Cys48 of helix B (see section 3.4) and Cys94 of helix C of the TCM motif (Figure S1) are both good candidates. One can also see the D2O-dependent band likely corresponding to a protonated carboxylic acid (negative at 1732 cm–1, positive at 1740 cm–1, overlapping with lipid esters), which is tempting to assign to one of the extracellular Glu/Asp residues explored by MD simulations and mutagenesis below (see sections 3.4 and 3.5) or to the internal Asp123 (homologue of BR’s Asp115). Finally, a pair of strong bands (negative at 1696 cm–1, positive at 1683 cm–1) may indicate H-bonding changes involving the C=O of an Asn side chain, for which Asn44 of helix B (see section 3.4) and two conserved Asn in helix G (Figure S1) would be the prime candidates.

3.4 Probing Selected Residues of S13.Bin138 Proteo-SR by Site-Directed Mutagenesis

Based on the conservation patterns seen in the multiple sequence alignment (Figure S1) we selected 10 residues conserved in Proteo-SRs for mutagenesis studies (Table S1). Those residues are either unique for the group or known to play important roles in BR or GPR. Specifically, we performed the following mutations in S13.Bin138 Proteo-SR: N44A (Thr46 in BR), C48A (Pro50 in BR), S51A (Ala53 in BR), H55A (Tyr57 in BR), R87Q (Arg82 in BR), Y90F (Asp85 in BR), E134Q (Glu142 in GPR), R218Q (Glu204 in BR), E219Q (Thr205 in BR), and D226N (Asp212 in BR). The mutants were characterized under the same conditions as the wild-type by static and time-resolved visible spectroscopy as well as Raman spectroscopy for selected mutants (Table S1, Figures S7–S9), except for the D226N mutant which did not express (or was severely misfolded and did not form a protonated Schiff base).

We chose Asn44 and Cys48 in helix B (Thr46 and Pro50 in BR) as they are conserved in Proteo-SRs and form an unusual motif on the cytoplasmic side, rarely seen in other rhodopsins (with the few exceptions of cyanobacterial halorhodopsins, Acetabularia proton pumps, and some putative Cryptophyta sensory rhodopsins). Although Asn44 is relatively far from retinal Schiff base (based on its correspondence to BR’s Thr46 and according to our structural model), its mutation to Ala affects the spectrum and the photocycle of S13.Bin138 Proteo-SR. First, its absorption maximum is slightly red-shifted (Table S1, Figure S7A), which does not seem to be caused by changes in the isomeric content, according to its Raman spectrum (Figure S9). The pKa of the Schiff base of both N44A and C48A is somewhat lower than in the wild-type (Table S1, Figure S7B), being in the range of 10.2–10.4 as opposed to >11. Most interestingly, the photocycle of N44A, but not of C48A, is strongly perturbed, being dramatically faster than in the wild-type, and showing modest accumulation of the M intermediate (Figure S8). These results may suggest that Asn44 resides in a conformational hotspot coupled to the retinal chromophore and may be important for signal transduction. This suggestion is supported by the light-induced difference FTIR spectra (see Figure 4 above) of the wild-type S13.Bin138 Proteo-SR, which show the bands that can be ascribed to both Asn and Cys side chains.

Ser51 from helix B corresponds to Ala53 in BR, a position known to affect ion transfers in the retinal Schiff base vicinity in several microbial rhodopsins.97−99 The photocycle and the alkaline pH spectra of the S51A mutant resemble those of the wild-type, but an acid-induced blue-shift of the spectrum with pKa ∼ 5 described above could not be observed (Table S1, Figures S7–S8), suggesting that Ser51 may interact with a residue responsible for the blue-shift.

Located just one helical turn from Ser51, His55 is an unusual replacement for the well-conserved Tyr on helix B (Tyr 57 in BR), which usually interacts with the aspartate Schiff base counterion on helix G (Asp212 in BR, Asp226 in S13.Bin138 Proteo-SR). The H55A mutant shows blue-shifted (by about 7 nm at pH 9) and somewhat broadened visible spectrum and a moderately slower photocycle, without any dramatic changes in retinal conformation as reported by Raman spectroscopy (Table S1, Figures S7–S9). Similar to S51A, the H55A mutant shows almost no acid-induced blue-shift, suggesting that it may be either His55 itself or another residue interacting with it (and Ser51), which is being titrated. In view of the potential importance of His55 for the tuning of the counterion complex of S13.Bin138 Proteo-SR, we used atomistic MD simulations to explore potential interactions of His55 (section 3.5 below).

Helix C hosts the strongly conserved Arg87 (Arg82 in BR) and, within one helical turn, Tyr90, which replaces the negatively charged counterion and primary proton acceptor of BR (Asp85). The R87Q mutant shows a modest (a few nm) blue shift of the visible spectrum, a smaller acid-induced spectral shift, and a photocycle with changed kinetics of the K/L1/L2 transitions, but it is otherwise similar to that of the wild-type protein (Table S1, Figures S7 and S8). The Y90F mutant has much stronger phenotype (Table S1, Figures S7–S9), displaying a large blue-shift of ∼20 nm, lower pKa of the Schiff base (∼10.3), and a Raman spectrum which suggests higher 13-cis content and less prominent chromophore twisting (relatively stronger 1184 cm–1 C–C stretch and weaker 982 cm–1 HOOP). The prominent double ethylenic stretch of the wild-type was unevenly affected by the mutation. While the main peak has shifted as expected from the strongly blue-shifted visible absorption maximum, the secondary peak did not, almost merging with the main peak as a result (Figure S9). All of this suggests a rather strong interaction of Tyr90 with retinal. Accordingly, the photocycle kinetics and intermediates of the Y90F mutant are quite different from those in the wild-type, in part reflecting the blue-shift of the ground state and higher contribution from the 13-cis species, and in part reflecting true changes in the kinetics of K/L1/L2 transitions of the all-trans cycle. Considering the strong phenotype of this mutant, possible interactions of Tyr90 and the effects of its mutation were further explored by structural modeling and MD simulations (see section 3.5).

At the extracellular side, helices D and G contribute a cluster of conserved charged groups: Glu134 (helix D), conserved among Proteo-SRs and PRs (Glu142 in GPR), and Arg218 and Glu219 (helix G, corresponding to, respectively, BR’s Glu204 and Thr205) conserved in Proteo-SRs and ESR (Figure S1). We found that the R218Q mutant phenotype is fairly mild as it displays a modest red-shift of the absorption spectrum and minor changes in the photocycle kinetics. In contrast, replacement of Glu134 or Glu219, aside from a modest red-shift of their absorption spectra, has rather drastic consequences (Table S1, Figures S7–S9) for the pKa of the Schiff base, both in the dark and during the photocycle. Specifically, the pH titrations of the main absorption bands at ∼580 nm reveal significant accumulation of the product absorbing at ∼400 nm at alkaline pH (presumably with the deprotonated retinal Schiff base), suggesting that the pKa of the Schiff base is ∼9.7 in E134Q and ∼8.9 in E219Q (Figure S7). In contrast to the wild-type, the M-like intermediate absorbing around ∼400 nm is seen in the photocycles of these mutants (Figure S8), but their photocycle kinetics are complex, as both protonated and deprotonated Schiff base forms of the dark state are present in equilibrium (at pH 9 used for the photocycle kinetics measurements), and some kinetics may reflect re-equilibration of these forms upon the ∼580 nm state photodepletion. Aside from the light-induced Schiff base deprotonation, the photocycles of these mutants are significantly perturbed but in the opposite ways: while the E134Q shows a much slower cycle, the E219Q displays a much faster cycle than the wild-type.

To verify that slower intermediates in the E134Q cycle are not associated with the higher 13-cis content of the dark state, we measured its Raman spectrum, which was similar to that of the wild-type (Figure S9). These spectroscopic data suggest a strong coupling of Glu134 and Glu219 to the retinal Schiff base, likely via its complex counterion. This suggestion is supported by MD simulations showing that Glu134 and Glu219 contribute to an H-bond network that connects to the retinal Schiff base (see section 3.5) and by the difference FTIR data reflecting a significant perturbation of carboxylic acid(s) in the photocycle (Figure 4).

3.5 Exploration of Hydrogen-Bonded Networks of S13.Bin138 Proteo-SR by MD Simulations

MD simulations show that wild-type S13.Bin138 Proteo-SR hosts an extended H-bond network at the extracellular half. Here, the Asp226 counterion on helix G is central to an H-bond network that effectively connects the protonated Schiff base to the extracellular side (Scheme 1, Figures 5, S10, and S11). The H-bond connections between Asp226 and its immediate neighbors (the retinal Schiff base, His55, Tyr90, Trp91, and Tyr199) tend to be either direct or mediated by one water molecule, whereas connections with the groups closer to the extracellular side, such as Arg87 and Glu219, are dynamic, less frequently visited, and mediated by 2–3 water molecules (Figure S11). Taken together, helices B, C, and G contribute not fewer than 15 H-bonding side chains to the dense H-bond network at the extracellular side of S13.Bin138 Proteo-SR; a handful of H-bond connections further bridge the extracellular network to helices A, D, and F (Scheme 1). In comparison, the cytoplasmic region hosts few interhelical H-bonds. The robust network connecting the retinal Schiff base to the extracellular side (e.g., to Glu134, Glu219, and Arg218) via its complex counterion (including, among others, Asp226, Tyr90, Ser51, His55, and Arg87) is in agreement with the expectations from the mutagenesis results. On the other hand, lack of involvement of Asn44 in the H-bond networks on the cytoplasmic side is somewhat surprising in view of its strong phenotype. Nevertheless, it is possible that Asn44 becomes part of those interhelical networks in one of the photointermediates, or that its replacement with Ala creates a water-filled cavity strongly affecting the photocycle. It should be also mentioned that Asn44 is H-bonded to the backbone carbonyl of Leu40. This side chain-backbone H-bond is within one helical turn of Y42, which is part of a network that includes several helix B side chains (Lys34, Ser35, Asn36, Tyr37, and Arg38) (Figure S10). The Asn44-Leu40 H-bond is also close to the interhelical H-bond between Trp233 and Met101.

Scheme 1 Summary of H-Bond Networks of S13.Bin138 Proteo-SR and How They Respond to Mutations

To avoid overcrowding of the scheme, we show only selected shortest-distance H-bond connections from the H-bond graphs. Gray dashed lines represent H-bonds and water-mediated bridges between protein side chains sampled in the reference simulation on the wild-type protein. Most of these H-bond connections shown with the gray dashed lines are sampled in all simulations we report here. The double-arrowed dashes indicate H-bond connections that depend on the tautomeric state of His55. Dark red, blue, and green dashed lines represent H-bonds sampled only in Y90F, E134Q, and E219Q, respectively. The red, blue, and green asterisks indicate selected shortest-distance H-bond connections absent from the Y90F, E134Q, and E219Q H-bond networks, respectively. Note that the H-bond connections shown apply to H-bond graphs computed with the criteria and occupancy thresholds as discussed in the text.

Figure 5 H-bond network of the reference simulation on wild-type S13.Bin138 Proteo-SR. (A) Selected amino acid residue side chains that H-bond during the simulation. (B) H-bond graph showing persistent H-bonds and water-mediated bridges that are sampled with occupancies of at least 70% during the last 200 ns of the simulation. We highlight yellow the nodes representing selected amino acid residue side chains whose H-bond properties were altered in independent simulations reported here. Details about the H-bond network are presented in Figures S10 and S11. (C) Close-up of the retinal Schiff base interactions. (D) Close-up of the extracellular H-bond network of E134 and E219, showing water molecules whose oxygen atoms are within 12 Å of the nitrogen atom of the retinal Schiff base. All molecular graphics were prepared by using VMD.

The tautomeric state of His55 influences the details of the H-bonding of Asp226, Ser51, Arg87, and Tyr90; however, the internal protein–water H-bond network is rather similar in simulations of the wild-type protein with His55 described as -Nε2 or -Nδ1 tautomers (Figures 5 and S10–S14). It is worth mentioning that the side chain of Arg87 switches its direct contact from Glu134 to Glu219 between the two simulations. As mentioned above, the complex Helix B His/Helix C Asp counterion of PRs and ESR is not present in Proteo-SRs, as the corresponding Asp is replaced by Tyr and His is moved down by one position. The MD simulations suggest that a different His/Asp counterion complex, Helix B His55/Helix G Asp226, may be formed in Proteo-SRs, and that the strength of this complex depends on the tautomeric state of His55 (Figures 5 and S12). This is reminiscent of the tautomeric switch proposed for the photocycle of GPR,100 which may be also at play in the novel His/Asp complex of Proteo-SRs. When His55 is a -Nδ1 tautomer, the shortest-distance H-bond connection between Arg87 and Asp226 is mediated by just two water molecules on average (Figure S14B); by contrast, in the reference simulation with His55 as −Nε2 tautomer, the shortest-distance H-bond connection between Arg87 and Asp226 passes via Tyr58 and involves two water bridges, each with 2–3 H-bonded water molecules (Figures 5 and S11B). In what follows, for simplicity, we compare the mutant simulations to the wild-type S13.Bin138 Proteo-SR with His55 treated as an −Nε2 tautomer.

Next, we explored how the extracellular H-bond network rearranges upon the mutation of several key residues. In wild-type S13.Bin138 Proteo-SR, Tyr90 samples direct H-bonds to Asp226 and Ser51 (Scheme 1, Figures 5, S10, S11). When Tyr90 is replaced by Phe, Asp226 connects to Ser51 directly or via one water molecule (the average number of water molecules in the Ser51-Asp226 bridge is 0.6) and lacks the water-mediated bridge to Tyr58 and from here to Glu219, Asp 226 connects instead to Glu219 either via Arg87, or via a relatively low-occupancy three-water bridge (Figures S15–S17). This rearrangement of H-bonding at the Asp226 site associates with altered H-bonds of Arg87, Glu134, and Glu219. Most notably, instead of a water-mediated bridge as found in the wild-type protein, Arg87 and Glu219 sample persistent, direct connections in the Y90F mutant (Figures 5, S10, S11, and S15–S17). This is similar to the above-mentioned switch of Arg87 from Glu134 to Glu219 observed upon the tautomeric switch of His55.

Reorientation of Arg87 toward Glu219 and direct H-bonding to Glu219 are also observed in the E134Q mutant (Figures S18–S20). Except for a somewhat closer interaction with Arg87, Asp226 has largely the same H-bond network in E134Q and wild-type S13.Bin138 Proteo-SR (Scheme 1, Figures 5 and S18).

Compared with the wild-type protein, the E219Q mutant lacks H-bonding between Gln219 and Tyr12. Gln219 samples water-mediated bridges not only with Arg87 and Arg218, but also with Glu134 (Figure S23); in contrast to the wild-type protein, the H-bond graph of E219Q lacks water-mediated bridging between Asp226 and Arg87, the shortest-distance connections between these two side chains being instead via Tyr58 or Gln219 (Figures S21–S23). This serious rearrangement of the counterion complex may be responsible for the lower pKa value of the Schiff base described above.

Finally, we monitored retinal conformation throughout the MD simulations to verify the predictions made by Raman spectroscopy regarding the retinal linearity and twist (see above). To estimate retinal bending, we monitored, along each of the simulations, the distance between the retinal C6 and C14 atoms. The mean values, in each of the five simulations performed, are 9.7–9.8 Å, with a standard deviation of 0.1 Å; the first quartile distance is 9.6–9.8 Å. Infrequently, the C6–C14 retinal distance can be as short as 9.2–9.3 Å. The wild-type S13.Bin138 Proteo-SR shows the mean distance of 9.8 Å; for comparison, the same distance measured from the high-resolution crystal structure of bacteriorhodopsin, PDB 7z09, is 9.54 Å (Figure S24). Thus, given that we report the SR138 retinal bending based on structural modeling and force-field simulations, we suggest that the retinal bending in our simulations is in agreement with the Raman spectroscopy prediction, being much smaller than in BR. To illustrate the retinal twist likely originating from the close interaction with Asp226, we include the five equally spaced coordinate snapshots taken from the last 50 ns of the reference wild-type simulation, which confirm both the chromophore twisting and proximity of Asp226 to the Schiff base (Figure S24A).

3.6 Interaction of Proteo-SRs with Their Transducers

Finally, we assayed the interaction of the putative Htr-like transducers of Proteo-SRs with their cognate photoreceptor. For this purpose, we used S22.Bin169 Proteo-SR with its C-terminally truncated transducer (TR), as the homologous transducer of S13.Bin138 Proteo-SR did not express well. The Colabfold modeling of the transducer showed a typical Htr-like fold, with two transmembrane helices and a well-structured loop between them bearing a putative chemosensory domain as well as a large C-terminal cytoplasmic domain (not shown). The truncation removed most of the large cytoplasmic domain of TR, leaving only about 40 amino acids after the predicted cytoplasmic end of the TM2, similar to what was done for N. pharaonis SRII.101 First, we tried to assay the interaction of S22.Bin169 Proteo-SR and TR in DDM detergent by mixing the two purified solubilized proteins at increasing molar TR/SR ratios (up to 8:1) and measuring the photocycle kinetics. Only very minor changes of spectra and kinetics have been observed (not shown), suggesting that interaction of SR and TR in DDM is weak. Thus, we changed our approach to co-reconstitution of the two proteins in lipids and assaying their interaction by following the photocycle kinetics at 580 nm (Figure 6). It appeared that the photocycle became appreciably faster with the increasing molar ratio of TR/SR, saturating at the expected 1:1 ratio, suggesting that the two proteins do interact in lipids at the given low protein:lipid ratio (2:1 w/w). When a lower protein/lipid ratio (1:5) was used for the co-reconstitution the interaction was still observed, but it appeared to be weaker (Figure 6B) suggesting that the complex of SR and TR may include lipids when they are present in sufficient amounts. We analyzed the S22.Bin169 TR/SR photocycle vs that of SR alone at the low lipid content (Figure S25) and found that it lacks the ∼2 ms component corresponding to the L1/L2 equilibration (see above). We verified that these photocycle changes are not caused by the alteration of the isomeric composition of the dark state, as the Raman spectra of both samples were identical (data not shown). It is thus feasible that the observed shift in the L1/L2 intermediate equilibrium reflects an alteration of the conformational changes in the receptor by interactions with the transducer.

Figure 6 Effect of the transducer co-reconstitution on the photocycle of S22.Bin169 Proteo-SR. (A) Photocycle kinetics (taken at 580 nm) measured at different receptor/transducer molar ratios at pH 9. The indicated ratios are based on the desired prereconstitution protein concentrations, while the actual postreconstitution ratios are two times lower, as detected by SDS-PAGE (e.g., 1:2 receptor/transducer is actually 1:1). (B) Comparison of the photocycle kinetics at 580 nm for the transducer-free and transducer-co-reconstituted receptor. Left: low lipid/protein ratio (1:2 w/w), at the 1:1 actual transducer/receptor ratio; right: high lipid/protein ratio (5:1 w/w), at the 5:1 actual transducer/receptor ratio.

4 Conclusions

In summary, we characterized a new group of sensory rhodopsins with the unique YT(C)M helix C motif, which was first discovered in 2004 but has not been studied experimentally until now. The phylogenetic analysis showed that these rhodopsins, even though they may interact with Htr-like transducers, are not related to haloarchaeal SRs. Instead, they are closer to PRs, in particular, to ESR, so we called them Proteo-SRs. Interestingly, the expressed proteins had absorption maxima around 580 nm, suggesting that they may play a phototactic (SRI-like) rather than a photophobic (SRII-like) role, contrary to the original prediction. The spectroscopic studies along with MD simulations and bioinformatic analysis revealed many unusual features of these Proteo-SRs. The retinal chromophore shows several unusual vibrational bands, suggesting that it is unbent and twisted at the same time. The extracellular side of Proteo-SRs features an extended H-bonded network with strong coupling of extracellular glutamates to the unique Schiff base counterion complex. Accordingly, FTIR spectroscopy shows significant light-induced perturbation of carboxylic acids, along with asparagine and cysteine side chains. The classical two-Asp (on helices C and G) counterion of outward proton pumps is replaced by Tyr/Asp, and wild-type Proteo-SRs do not deprotonate their Schiff base in the photocycle and do not transport protons accordingly. Moreover, the His/Asp complex of PRs formed between helices B and C is recreated between helices B and G, by moving the His one position farther. We were also able to detect interaction of Proteo-SR with its transducer upon co-reconstitution in lipids. Thus, we infer that sensory rhodopsins able to interact with Htr-type transducers appeared in the course of evolution at least twice, derived from either haloarchaeal BR or proteobacterial PR, possibly presenting another case of convergent evolution among microbial rhodopsins. The unique structural motifs of Proteo-SRs expand the repertoire of sequence signatures of sensory rhodopsins and add to the collection of functional helix C motifs of microbial rhodopsins.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jpcb.3c04032.Additional analysis, experimental, and computational details including sequence alignments, spectroscopic data, and molecular dynamics simulations visualization (PDF)

Supplementary Material

jp3c04032_si_001.pdf

The authors declare no competing financial interest.

Acknowledgments

The research was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC) and the University of Guelph grants to L.S.B., and by German research foundation (DFG) via SFB1078 to J.H. (project B3). A.-N.B. thanks Dr. Yoshitaka Moriwaki for assistance with installing ColabFold. Computations were performed using computing resources at the RWTH-Aachen University and at the JURECA-DC supercomputing center at the Forschungszentrum Jülich. A.P. was supported by NSERC USRA.

Abbreviations

BR bacteriorhodopsin

ESR Exiguobacterium sibiricum rhodopsin

FTIR Fourier-transform infrared

GPR green-absorbing proteorhodopsin

MD molecular dynamics

PR proteorhodopsin

Proteo-SR new type of sensory rhodopsins related to PR

SB Schiff base of retinal

SR sensory rhodopsin

XR xanthorhodopsin
==== Refs
References

Rozenberg A. ; Inoue K. ; Kandori H. ; Beja O. Microbial Rhodopsins: The Last Two Decades. Annu. Rev. Microbiol. 2021, 75 , 427–447. 10.1146/annurev-micro-031721-020452.34343014
Gordeliy V. ; Kovalev K. ; Bamberg E. ; Rodriguez-Valera F. ; Zinovev E. ; Zabelskii D. ; Alekseev A. ; Rosselli R. ; Gushchin I. ; Okhrimenko I. Microbial Rhodopsins. Methods in molecular biology (Clifton, N.J.) 2022, 2501 , 1–52. 10.1007/978-1-0716-2329-9_1.
Ernst O. P. ; Lodowski D. T. ; Elstner M. ; Hegemann P. ; Brown L. S. ; Kandori H. Microbial and animal rhodopsins: structures, functions, and molecular mechanisms. Chem. Rev. 2014, 114 , 126–163. 10.1021/cr4003769.24364740
de Grip W. J. ; Ganapathy S. Rhodopsins: An Excitingly Versatile Protein Species for Research, Development and Creative Engineering. Front Chem. 2022, 10 , 879609 10.3389/fchem.2022.879609.35815212
Inoue K. Diversity, Mechanism, and Optogenetic Application of Light-Driven Ion Pump Rhodopsins. Advances in experimental medicine and biology 2021, 1293 , 89–126. 10.1007/978-981-15-8763-4_6.33398809
Govorunova E. G. ; Sineshchekov O. A. ; Li H. ; Spudich J. L. Microbial Rhodopsins: Diversity, Mechanisms, and Optogenetic Applications. Annu. Rev. Biochem. 2017, 86 , 845–872. 10.1146/annurev-biochem-101910-144233.28301742
Inoue K. ; Ito S. ; Kato Y. ; Nomura Y. ; Shibata M. ; Uchihashi T. ; Tsunoda S. P. ; Kandori H. A natural light-driven inward proton pump. Nat. Commun. 2016, 7 , 13415 10.1038/ncomms13415.27853152
Inoue K. ; Tsunoda S. P. ; Singh M. ; Tomida S. ; Hososhima S. ; Konno M. ; Nakamura R. ; Watanabe H. ; Bulzu P. A. ; Banciu H. L. ; et al. Schizorhodopsins: A family of rhodopsins from Asgard archaea that function as light-driven inward H(+) pumps. Sci. Adv. 2020, 6 , eaaz2441 10.1126/sciadv.aaz2441.32300653
Harris A. ; Lazaratos M. ; Siemers M. ; Watt E. ; Hoang A. ; Tomida S. ; Schubert L. ; Saita M. ; Heberle J. ; Furutani Y. ; et al. Mechanism of Inward Proton Transport in an Antarctic Microbial Rhodopsin. J. Phys. Chem. B 2020, 124 , 4851–4872. 10.1021/acs.jpcb.0c02767.32436389
Brown L. S. Light-driven proton transfers and proton transport by microbial rhodopsins - A biophysical perspective. Biochim Biophys Acta Biomembr 2022, 1864 , 183867 10.1016/j.bbamem.2022.183867.35051382
Engelhard C. ; Chizhov I. ; Siebert F. ; Engelhard M. Microbial Halorhodopsins: Light-Driven Chloride Pumps. Chem. Rev. 2018, 118 , 10629–10645. 10.1021/acs.chemrev.7b00715.29882660
Yoshizawa S. ; Kumagai Y. ; Kim H. ; Ogura Y. ; Hayashi T. ; Iwasaki W. ; DeLong E. F. ; Kogure K. Functional characterization of flavobacteria rhodopsins reveals a unique class of light-driven chloride pump in bacteria. Proc. Natl. Acad. Sci. U. S. A. 2014, 111 , 6732–6737. 10.1073/pnas.1403051111.24706784
Schobert B. ; Lanyi J. K. Halorhodopsin is a light-driven chloride pump. J. Biol. Chem. 1982, 257 , 10306–10313. 10.1016/S0021-9258(18)34020-1.7107607
Hasemi T. ; Kikukawa T. ; Kamo N. ; Demura M. Characterization of a Cyanobacterial Chloride-pumping Rhodopsin and Its Conversion into a Proton Pump. J. Biol. Chem. 2016, 291 , 355–362. 10.1074/jbc.M115.688614.26578511
Ihara K. ; Umemura T. ; Katagiri I. ; Kitajima-Ihara T. ; Sugiyama Y. ; Kimura Y. ; Mukohata Y. Evolution of the archaeal rhodopsins: evolution rate changes by gene duplication and functional differentiation. J. Mol. Biol. 1999, 285 , 163–174. 10.1006/jmbi.1998.2286.9878396
Sharma A. K. ; Spudich J. L. ; Doolittle W. F. Microbial rhodopsins: functional versatility and genetic mobility. Trends in Microbiology 2006, 14 , 463–469. 10.1016/j.tim.2006.09.006.17008099
Kyndt J. A. ; Meyer T. E. ; Cusanovich M. A. Photoactive yellow protein, bacteriophytochrome, and sensory rhodopsin in purple phototrophic bacteria. Photochem. Photobiol. Sci. 2004, 3 , 519–530. 10.1039/b315731h.15170480
Needham D. M. ; Yoshizawa S. ; Hosaka T. ; Poirier C. ; Choi C. J. ; Hehenberger E. ; Irwin N. A. T. ; Wilken S. ; Yung C. M. ; Bachy C. ; et al. A distinct lineage of giant viruses brings a rhodopsin photosystem to unicellular marine predators. Proc. Natl. Acad. Sci. U. S. A. 2019, 116 , 20574–20583. 10.1073/pnas.1907517116.31548428
Kindler G. S. ; Wong H. L. ; Larkum A. W. D. ; Johnson M. ; MacLeod F. I. ; Burns B. P. Genome-resolved metagenomics provides insights into the functional complexity of microbial mats in Blue Holes, Shark Bay. FEMS Microbiol. Ecol. 2022, 98 , fiab158 10.1093/femsec/fiab158.34865013
Spudich J. L. The multitalented microbial sensory rhodopsins. Trends Microbiol 2006, 14 , 480–487. 10.1016/j.tim.2006.09.005.17005405
Bogomolni R. A. ; Spudich J. L. Identification of a 3Rd-Rhodopsin-Like Pigment in Phototactic Halobacterium-Halobium. Proc. Natl. Acad. Sci. U. S. A. 1982, 79 , 6250–6254. 10.1073/pnas.79.20.6250.6959114
Wolff E. K. ; Bogomolni R. A. ; Scherrer P. ; Hess B. ; Stoeckenius W. Color discrimination in halobacteria: spectroscopic characterization of a second sensory receptor covering the blue-green region of the spectrum. Proc. Natl. Acad. Sci. U. S. A. 1986, 83 , 7272–7276. 10.1073/pnas.83.19.7272.3463965
Takahashi T. ; Tomioka H. ; Kamo N. ; Kobatake Y. A photosystem other than PS370 also mediates the negative phototaxis of Halobacterium halobium. FEMS Microbiology Letters 1985, 28 , 161–164. 10.1111/j.1574-6968.1985.tb00784.x.
Spudich E. N. ; Sundberg S. A. ; Manor D. ; Spudich J. L. Properties of a second sensory receptor protein in Halobacterium halobium phototaxis. Proteins 1986, 1 , 239–246. 10.1002/prot.340010306.3449857
Fu H. Y. ; Lin Y. C. ; Chang Y. N. ; Tseng H. ; Huang C. C. ; Liu K. C. ; Huang C. S. ; Su C. W. ; Weng R. R. ; Lee Y. Y. ; et al. A novel six-rhodopsin system in a single archaeon. J. Bacteriol. 2010, 192 , 5866–5873. 10.1128/JB.00642-10.20802037
Mongodin E. F. ; Nelson K. E. ; Daugherty S. ; Deboy R. T. ; Wister J. ; Khouri H. ; Weidman J. ; Walsh D. A. ; Papke R. T. ; Sanchez Perez G. ; et al. The genome of Salinibacter ruber: convergence and gene exchange among hyperhalophilic bacteria and archaea. Proc. Natl. Acad. Sci. U. S. A. 2005, 102 , 18147–18152. 10.1073/pnas.0509073102.16330755
Hoff W. D. ; Jung K. H. ; Spudich J. L. Molecular mechanism of photosignaling by archaeal sensory rhodopsins. Annu. Rev. Biophys. Biomol. Struct. 1997, 26 , 223–258. 10.1146/annurev.biophys.26.1.223.9241419
Jung K. H. The distinct signaling mechanisms of microbial sensory rhodopsins in Archaea, Eubacteria and Eukarya. Photochem. Photobiol. 2007, 83 , 63–69. 10.1562/2006-03-20-IR-853.16968113
Sasaki J. ; Spudich J. L. Signal transfer in haloarchaeal sensory rhodopsin- transducer complexes. Photochem. Photobiol. 2008, 84 , 863–868. 10.1111/j.1751-1097.2008.00314.x.18346091
Oprian D. D. Phototaxis, chemotaxis and the missing link. Trends Biochem. Sci. 2003, 28 , 167–169. 10.1016/S0968-0004(03)00056-2.12713898
Klare J. P. ; Chizhov I. ; Engelhard M. Microbial rhodopsins: scaffolds for ion pumps, channels, and sensors. Results Probl Cell Differ 2008, 45 , 73–122. 10.1007/400_2007_041.17898961
Inoue K. ; Tsukamoto T. ; Sudo Y. Molecular and evolutionary aspects of microbial sensory rhodopsins. Biochim. Biophys. Acta 2014, 1837 , 562–577. 10.1016/j.bbabio.2013.05.005.23732219
Sasaki J. ; Spudich J. L. Proton transport by sensory rhodopsins and its modulation by transducer-binding. Biochim. Biophys. Acta 2000, 1460 , 230–239. 10.1016/S0005-2728(00)00142-0.10984603
Suzuki D. ; Irieda H. ; Homma M. ; Kawagishi I. ; Sudo Y. Phototactic and chemotactic signal transduction by transmembrane receptors and transducers in microorganisms. Sensors (Basel) 2010, 10 , 4010–4039. 10.3390/s100404010.22319339
Gushchin I. ; Gordeliy V. Transmembrane Signal Transduction in Two-Component Systems: Piston, Scissoring, or Helical Rotation?. Bioessays 2018, 40 , 1700197 10.1002/bies.201700197.
Shi L. ; Kawamura I. ; Jung K. H. ; Brown L. S. ; Ladizhansky V. Conformation of a seven-helical transmembrane photosensor in the lipid environment. Angew. Chem., Int. Ed. Engl. 2011, 50 , 1302–1305. 10.1002/anie.201004422.21290498
Waschuk S. A. ; Bezerra A. G. ; Shi L. ; Brown L. S. Leptosphaeria rhodopsin: Bacteriorhodopsin-like proton pump from a eukaryote. Proc. Natl. Acad. Sci. USA 2005, 102 , 6879–6883. 10.1073/pnas.0409659102.15860584
Radu I. ; Schleeger M. ; Bolwien C. ; Heberle J. Time-resolved methods in biophysics. 10. Time-resolved FT-IR difference spectroscopy and the application to membrane proteins. Photochem. Photobiol. Sci. 2009, 8 , 1517–1528. 10.1039/b9pp00050j.19862409
Mirdita M. ; Schütze K. ; Moriwaki Y. ; Heo L. ; Ovchinnikov S. ; Steinegger M. ColabFold: making protein folding accessible to all. Nat. Methods 2022, 19 , 679–682. 10.1038/s41592-022-01488-1.35637307
Sali A. ; Blundell T. L. Comparative protein modelling by satisfaction of spatial restraints. J. Mol. Biol. 1993, 234 , 779–815. 10.1006/jmbi.1993.1626.8254673
Marti-Renom M. A. ; Stuart A. ; Fiser A. ; Sanchez R. ; Melo F. ; Sali A. Comparative protein structure modeling of genes and genomes. Annu. Rev. Biomol. Struct. 2000, 29 , 291–325. 10.1146/annurev.biophys.29.1.291.
Humphrey W. ; Dalke W. ; Schulten K. VMD: visual molecular dynamics. J. Mol. Graph. 1996, 14 , 33–38. 10.1016/0263-7855(96)00018-5.8744570
Brooks B. R. ; Bruccoleri R. E. ; Olafson B. D. ; States D. J. ; Swaminathan S. ; Karplus M. CHARMM: a program for macromolecular energy, minimization, and dynamics calculations. J. Comput. Chem. 1983, 4 , 187–217. 10.1002/jcc.540040211.
Brooks B. R. ; Brooks C. L. I. ; MacKerell A. D. Jr. ; Nilsson L. ; Petrella R. J. ; Roux B. ; Won Y. ; Archontis G. ; Bartels C. ; Boresch S. ; et al. CHARMM: the biomolecular simulation program. J. Comput. Chem. 2009, 30 , 1545–1614. 10.1002/jcc.21287.19444816
Olsson M. H. M. ; Sondergaard C. R. ; Rostkowski M. ; Jensen J. H. PROPKA3: consistent treatment of internal and surface residues in empirical pKa predictions. J. Chem. Theory Comput. 2011, 7 , 525–537. 10.1021/ct100578z.26596171
Kato H. E. ; Zhang F. ; Yizhar O. ; Ramakrishnan C. ; Nishizawa T. ; Hirata K. ; Ito J. ; Aita Y. ; Tsukazaki T. ; Hayashi S. ; et al. Crystal structure of the channelrhodopsin light-gated cation channel. Nature 2012, 482 , 369–374. 10.1038/nature10870.22266941
Ito S. ; Kato H. E. ; Taniguchi R. ; Iwata T. ; Nureki O. ; Kandori H. Water-containing hydrogen-bonding network in the active center of channelrhodopsin. J. Am. Chem. Soc. 2014, 136 , 3475–3482. 10.1021/ja410836g.24512107
Metz G. ; Siebert F. ; Engelhardt M. Asp85 is the only internal aspartic acid that gets protonated in the M intermediate and the purple-to-blue transition of bacteriorhodopsin. A solid-state 13C CP-MAS NMR investigation. FEBS Lett. 1992, 303 , 237–241. 10.1016/0014-5793(92)80528-O.1318849
Luecke H. ; Schobert B. ; Richter H.-T. ; Cartailler J.-P. ; Lanyi J. K. Structure of bacteriorhodopsin at 1.55 Å resolution. J. Mol. Biol. 1999, 291 , 899–911. 10.1006/jmbi.1999.3027.10452895
Borshchevskiy V. ; Kovalev K. ; Round E. ; Efremov R. ; Astashkin R. ; Bourenkov G. ; Bratanov D. ; Balandin T. ; Chizhov I. ; Baeken C. ; et al. True-atomic-resolution insights into the structure and functional role of linear chains and low-barrier hydrogen bonds in proteins. Nature Struct. Mol. Biol. 2022, 29 , 440–450. 10.1038/s41594-022-00762-2.35484235
del Val C. ; Bondar L. ; Bondar A.-N. Coupling between inter-helical hydrogen bonding and water dynamics in a proton transporter. J. Struct. Biol. 2014, 186 , 95–111. 10.1016/j.jsb.2014.02.010.24576681
Garczarek F. ; Brown L. S. ; Lanyi J. K. ; Gerwert K. Proton binding within a membrane protein by a protonated water cluster. Proc. Natl. Acad. Sci. USA 2005, 102 , 3633–3638. 10.1073/pnas.0500421102.15738416
MacKerell A. D. Jr. ; Bashford D. ; Bellott M. ; Dunbrack R. L. ; Evanseck J. D. ; Field M. J. ; Fischer S. ; Gao J. ; Guo H. ; Ha S. ; et al. All-atom empirical potential for molecular modeling and dynamics studies of proteins. J. Phys. Chem. B 1998, 102 , 3586–3616. 10.1021/jp973084f.24889800
MacKerell A. D. Jr. ; Feig M. ; Brooks C. L. I. Extending the treatment of backbone energetics in protein force fields: limitations of gas-phase quantum mechanics in reproducing protein conformational distributions in molecular dynamics simulations. J. Comput. Chem. 2004, 25 , 1400–1415. 10.1002/jcc.20065.15185334
Gruia A. D. ; Bondar A.-N. ; Smith J. C. ; Fischer S. Mechanism of a molecular valve in the halorhodopsin chloride pump. Structure 2005, 13 , 617–627. 10.1016/j.str.2005.01.021.15837200
Tajkhorshid E. ; Baudry J. ; Schulten K. ; Suhai S. Molecular dynamics study of the nature and origin of retinal’s twisted structure in bacteriorhodopsin. Biophys. J. 2000, 78 , 683–693. 10.1016/S0006-3495(00)76626-4.10653781
Darden T. ; York D. ; Pedersen L. Particle mesh Ewald: an N x log(N) method for Ewald sums in large systems. J. Chem. Phys. 1993, 98 , 10089–10092. 10.1063/1.464397.
Essmann U. ; Perera L. ; Berkowitz M. L. ; Darden T. ; Lee H. ; Pedersen L. G. A smooth particle mesh Ewald method. J. Chem. Phys. 1995, 103 , 8577–8593. 10.1063/1.470117.
Kalé L. ; Skeel R. ; Bhandarkar M. ; Brunner R. ; Gursoy A. ; Krawetz N. ; Phillips J. ; Shinozaki A. ; Varadarajan K. ; Schulten K. NAMD2: greater scalability for parallel molecular dynamics. J. Comput. Phys. 1999, 151 , 283–312. 10.1006/jcph.1999.6201.
Phillips J. C. ; Braun B. ; Wang W. ; Gumbart J. ; Tajkhorshid E. ; Villa E. ; Chipot C. ; Skeel R. D. ; Kale L. ; Schulten K. Scalable molecular dynamics with NAMD. J. Comput. Chem. 2005, 26 , 1781–1802. 10.1002/jcc.20289.16222654
Phillips J. C. ; Hardy D. J. ; Maia J. D. C. ; Stone J. E. ; Ribeiro J. V. ; Bernardi R. C. ; Buch R. ; Fiorin G. ; Henin J. ; Jiang W. ; et al. Scalable molecular dynamics on CPU and GPU architectures with NAMD. J. Chem. Phys. 2020, 153 , 044130 10.1063/5.0014475.32752662
Martyna G. J. ; Tobias D. J. ; Klein M. L. Constant-pressure molecular-dynamics algorithms. J. Chem. Phys. 1994, 101 , 4177–4189. 10.1063/1.467468.
Tuckerman M. ; Berne B. J. ; Martyna G. J. Reversible multiple time scale molecular dynamics. J. Chem. Phys. 1992, 97 , 1990–2001. 10.1063/1.463137.
Ryckaert J.-P. ; Ciccotti G. ; Berendsen H. J. C. Numerical integration of the Cartesian equations of motion of a system with constraints. Molecular dynamics of n-alkanes. J. Comput. Phys. 1977, 23 , 327–341. 10.1016/0021-9991(77)90098-5.
Siemers M. ; Lazaratos M. ; Karathanou K. ; Guerra F. ; Brown L. S. ; Bondar A.-N. Bridge: A graph-based algorithm to analyze dynamic H-bond networks in membrane proteins. J. Chem. Theory Comput. 2019, 15 , 6781–6798. 10.1021/acs.jctc.9b00697.31652399
Siemers M. ; Bondar A.-N. Interactive interface for graph-based analyses of dynamic H-bond networks: application to spike protein S. J. Chem. Inf. Model. 2021, 61 , 2998–3014. 10.1021/acs.jcim.1c00306.
Chen I. A. ; Chu K. ; Palaniappan K. ; Pillay M. ; Ratner A. ; Huang J. ; Huntemann M. ; Varghese N. ; White J. R. ; Seshadri R. ; et al. IMG/M v.5.0: an integrated data management and comparative analysis system for microbial genomes and microbiomes. Nucleic Acids Res. 2019, 47 , D666–D677. 10.1093/nar/gky901.30289528
Sievers F. ; Wilm A. ; Dineen D. ; Gibson T. J. ; Karplus K. ; Li W. ; Lopez R. ; McWilliam H. ; Remmert M. ; Soding J. ; et al. Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol. Syst. Biol. 2011, 7 , 539 10.1038/msb.2011.75.21988835
Huson D. H. ; Scornavacca C. Dendroscope 3: an interactive tool for rooted phylogenetic trees and networks. Syst. Biol. 2012, 61 , 1061–1067. 10.1093/sysbio/sys062.22780991
Kralj J. M. ; Bergo V. B. ; Amsden J. J. ; Spudich E. N. ; Spudich J. L. ; Rothschild K. J. Protonation state of Glu142 differs in the green- and blue-absorbing variants of proteorhodopsin. Biochemistry-Us 2008, 47 , 3447–3453. 10.1021/bi7018964.
Bergo V. ; Amsden J. J. ; Spudich E. N. ; Spudich J. L. ; Rothschild K. J. Structural changes in the photoactive site of proteorhodopsin during the primary photoreaction. Biochemistry-Us 2004, 43 , 9075–9083. 10.1021/bi0361968.
Ran T. ; Ozorowski G. ; Gao Y. ; Sineshchekov O. A. ; Wang W. ; Spudich J. L. ; Luecke H. Cross-protomer interaction with the photoactive site in oligomeric proteorhodopsin complexes. Acta crystallographica 2013, 69 , 1965–1980. 10.1107/S0907444913017575.24100316
Gushchin I. ; Chervakov P. ; Kuzmichev P. ; Popov A. N. ; Round E. ; Borshchevskiy V. ; Ishchenko A. ; Petrovskaya L. ; Chupin V. ; Dolgikh D. A. ; et al. Structural insights into the proton pumping by unusual proteorhodopsin from nonmarine bacteria. Proc. Natl. Acad. Sci. U. S. A. 2013, 110 , 12631–12636. 10.1073/pnas.1221629110.23872846
Sudo Y. ; Spudich J. L. Three strategically placed hydrogen-bonding residues convert a proton pump into a sensory receptor. Proc. Natl. Acad. Sci. USA 2006, 103 , 16129–16134. 10.1073/pnas.0607467103.17050685
Radu I. ; Budyak I. L. ; Hoomann T. ; Kim Y. J. ; Engelhard M. ; Labahn J. ; Buldt G. ; Heberle J. ; Schlesinger R. Signal relay from sensory rhodopsin I to the cognate transducer HtrI: assessing the critical change in hydrogen-bonding between Tyr-210 and Asn-53. Biophys Chem. 2010, 150 , 23–28. 10.1016/j.bpc.2010.02.017.20303644
Petrovskaya L. E. ; Lukashev E. P. ; Chupin V. V. ; Sychev S. V. ; Lyukmanova E. N. ; Kryukova E. A. ; Ziganshin R. H. ; Spirina E. V. ; Rivkina E. M. ; Khatypov R. A. ; et al. Predicted bacteriorhodopsin from Exiguobacterium sibiricum is a functional proton pump. Febs Lett. 2010, 584 , 4193–4196. 10.1016/j.febslet.2010.09.005.20831870
Beja O. ; Lanyi J. K. Nature’s toolkit for microbial rhodopsin ion pumps. Proc. Natl. Acad. Sci. U. S. A. 2014, 111 , 6538–6539. 10.1073/pnas.1405093111.24737891
Balashov S. P. ; Petrovskaya L. E. ; Lukashev E. P. ; Imasheva E. S. ; Dioumaev A. K. ; Wang J. M. ; Sychev S. V. ; Dolgikh D. A. ; Rubin A. B. ; Kirpichnikov M. P. ; et al. Aspartate-histidine interaction in the retinal schiff base counterion of the light-driven proton pump of Exiguobacterium sibiricum. Biochemistry-Us 2012, 51 , 5748–5762. 10.1021/bi300409m.
Hempelmann F. ; Holper S. ; Verhoefen M. K. ; Woerner A. C. ; Kohler T. ; Fiedler S. A. ; Pfleger N. ; Wachtveitl J. ; Glaubitz C. His75-Asp97 cluster in green proteorhodopsin. J. Am. Chem. Soc. 2011, 133 , 4645–4654. 10.1021/ja111116a.21366243
Bergo V. B. ; Sineshchekov O. A. ; Kralj J. M. ; Partha R. ; Spudich E. N. ; Rothschild K. J. ; Spudich J. L. His-75 in Proteorhodopsin, a Novel Component in Light-driven Proton Translocation by Primary Pumps. J. Biol. Chem. 2009, 284 , 2836–2843. 10.1074/jbc.M803792200.19015272
Jiang H. ; Dong H. ; Deng S. ; Yu B. ; Huang Q. ; Wu Q. Response of Archaeal Community Structure to Environmental Changes in Lakes on the Tibetan Plateau, Northwestern China. Geomicrobiology Journal 2009, 26 , 289–297. 10.1080/01490450902892662.
Yang J. ; Jiang H. ; Dong H. ; Wang H. ; Wu G. ; Hou W. ; Liu W. ; Zhang C. ; Sun Y. ; Lai Z. amoA-encoding archaea and thaumarchaeol in the lakes on the northeastern Qinghai-Tibetan Plateau, China. Front Microbiol 2013, 4 , 329 10.3389/fmicb.2013.00329.24273535
Kakitani H. ; Kakitani T. ; Rodman H. ; Honig B. ; Callender R. Correlation of Vibrational Frequencies with Absorption Maxima in Polyenes, Rhodopsin, Bacteriorhodopsin, and Retinal Analogs. J. Phys. Chem-Us 1983, 87 , 3620–3628. 10.1021/j100242a011.
Urui T. ; Das I. ; Mizuno M. ; Sheves M. ; Mizutani Y. Origin of a Double-Band Feature in the Ethylenic C=C Stretching Modes of the Retinal Chromophore in Heliorhodopsins. J. Phys. Chem. B 2022, 126 , 8680–8688. 10.1021/acs.jpcb.2c04883.36281583
Braiman M. ; Mathies R. Resonance Raman-Spectra of Bacteriorhodopsins Primary Photoproduct - Evidence for a Distorted 13-Cis Retinal Chromophore. Proc. Natl. Acad. Sci. U. S. A. 1982, 79 , 403–407. 10.1073/pnas.79.2.403.6281770
Shionoya T. ; Singh M. ; Mizuno M. ; Kandori H. ; Mizutani Y. Strongly Hydrogen-Bonded Schiff Base and Adjoining Polyene Twisting in the Retinal Chromophore of Schizorhodopsins. Biochemistry-Us 2021, 60 , 3050–3057. 10.1021/acs.biochem.1c00529.
Haupts U. ; Eisfeld W. ; Stockburger M. ; Oesterhelt D. Sensory rhodopsin I photocycle intermediate SRI380 contains 13-cis retinal bound via an unprotonated Schiff base. Febs Lett. 1994, 356 , 25–29. 10.1016/0014-5793(94)01226-1.7988713
Smith S. O. ; Braiman M. S. ; Myers A. B. ; Pardoen J. A. ; Courtin J. M. L. ; Winkel C. ; Lugtenburg J. ; Mathies R. A. Vibrational Analysis of the All-Trans-Retinal Chromophore in Light-Adapted Bacteriorhodopsin. J. Am. Chem. Soc. 1987, 109 , 3108–3125. 10.1021/ja00244a038.
Krebs R. A. ; Dunmire D. ; Partha R. ; Braiman M. S. Resonance Raman characterization of proteorhodopsin’s chromophore environment. J. Phys. Chem. B 2003, 107 , 7877–7883. 10.1021/jp034574c.
Pfleger N. ; Lorch M. ; Woerner A. C. ; Shastri S. ; Glaubitz C. Characterisation of Schiff base and chromophore in green proteorhodopsin by solid-state NMR. J. Biomol Nmr 2008, 40 , 15–21. 10.1007/s10858-007-9203-5.17968661
Scherrer P. ; Mathew M. K. ; Sperling W. ; Stoeckenius W. Retinal isomer ratio in dark-adapted purple membrane and bacteriorhodopsin monomers. Biochemistry-Us 1989, 28 , 829–834. 10.1021/bi00428a063.
Vogeley L. ; Sineshchekov O. A. ; Trivedi V. D. ; Sasaki J. ; Spudich J. L. ; Luecke H. Anabaena sensory rhodopsin: a photochromic color sensor at 2.0 A. Science 2004, 306 , 1390–1393. 10.1126/science.1103943.15459346
Sineshchekov O.A. ; Govorunova E.G. ; Li H. ; Wang Y. ; Melkonian M. ; Wong G.K. ; Brown L.S. ; Spudich J.L. Conductance Mechanisms of Rapidly Desensitizing Cation Channelrhodopsins from Cryptophyte Algae. mBio 2020, 11 , 00657-20 10.1128/mBio.00657-20.
Oppermann J. ; Fischer P. ; Silapetere A. ; Liepe B. ; Rodriguez-Rozada S. ; Flores-Uribe J. ; Peter E. ; Keidel A. ; Vierock J. ; Kaufmann J. ; et al. MerMAIDs: a family of metagenomically discovered marine anion-conducting and intensely desensitizing channelrhodopsins. Nat. Commun. 2019, 10 , 3315 10.1038/s41467-019-11322-6.31346176
Bruun S. ; Naumann H. ; Kuhlmann U. ; Schulz C. ; Stehfest K. ; Hegemann P. ; Hildebrandt P. The chromophore structure of the long-lived intermediate of the C128T channelrhodopsin-2 variant. Febs Lett. 2011, 585 , 3998–4001. 10.1016/j.febslet.2011.11.007.22094167
Dioumaev A. K. Evaluation of intrinsic chemical kinetics and transient product spectra from time-resolved spectroscopic data. Biophys Chem. 1997, 67 , 1–25. 10.1016/S0301-4622(96)02268-5.17029887
Eisenhauer K. ; Kuhne J. ; Ritter E. ; Berndt A. ; Wolf S. ; Freier E. ; Bartl F. ; Hegemann P. ; Gerwert K. In channelrhodopsin-2 Glu-90 is crucial for ion selectivity and is deprotonated during the photocycle. J. Biol. Chem. 2012, 287 , 6904–6911. 10.1074/jbc.M111.327700.22219197
Besaw J. E. ; Ou W. L. ; Morizumi T. ; Eger B. T. ; Sanchez Vasquez J. D. ; Chu J. H. Y. ; Harris A. ; Brown L. S. ; Miller R. J. D. ; Ernst O. P. The crystal structures of a chloride-pumping microbial rhodopsin and its proton-pumping mutant illuminate proton transfer determinants. J. Biol. Chem. 2020, 295 , 14793–14804. 10.1074/jbc.RA120.014118.32703899
Govorunova E. G. ; Sineshchekov O. A. ; Brown L. S. ; Bondar A. N. ; Spudich J. L. Structural Foundations of Potassium Selectivity in Channelrhodopsins. mBio 2022, 13 , e0303922 10.1128/mbio.03039-22.36413022
Maciejko J. ; Kaur J. ; Becker-Baldus J. ; Glaubitz C. Photocycle-dependent conformational changes in the proteorhodopsin cross-protomer Asp-His-Trp triad revealed by DNP-enhanced MAS-NMR. Proc. Natl. Acad. Sci. U. S. A. 2019, 116 , 8342–8349. 10.1073/pnas.1817665116.30948633
Hippler-Mreyen S. ; Klare J. P. ; Wegener A. A. ; Seidel R. ; Herrmann C. ; Schmies G. ; Nagel G. ; Bamberg E. ; Engelhard M. Probing the sensory rhodopsin II binding domain of its cognate transducer by calorimetry and electrophysiology. J. Mol. Biol. 2003, 330 , 1203–1213. 10.1016/S0022-2836(03)00656-9.12860139
