Functional Green-Tuned Proteorhodopsin from Modern Stromatolites

The sequenced genome of the poly-extremophile Exiguobacterium sp. S17, isolated from modern stromatolites at Laguna Socompa (3,570 m), a High-Altitude Andean Lake (HAAL) in Argentinean Puna revealed a putative proteorhodopsin-encoding gene. The HAAL area is exposed to the highest UV irradiation on Earth, making the microbial community living in the stromatolites test cases for survival strategies under extreme conditions. The heterologous expressed protein E17R from Exiguobacterium (248 amino acids, 85% sequence identity to its ortholog ESR from E. sibiricum) was assembled with retinal displaying an absorbance maximum at 524 nm, which makes it a member of the green-absorbing PR-subfamily. Titration down to low pH values (eventually causing partial protein denaturation) indicated a pK value between two and three. Global fitting of data from laser flash-induced absorption changes gave evidence for an early red-shifted intermediate (its formation being below the experimental resolution) that decayed (τ1 = 3.5 μs) into another red-shifted intermediate. This species decayed in a two-step process (τ2 = 84 μs, τ3 = 11 ms), to which the initial state of E17-PR was reformed with a kinetics of 2 ms. Proton transport capability of the HAAL protein was determined by BLM measurements. Additional blue light irradiation reduced the proton current, clearly identifying a blue light absorbing, M-like intermediate. The apparent absence of this intermediate is explained by closely matching formation and decay kinetics.


Introduction
Microbial rhodopsins are considered to be the simplest energy-harvesting photoreceptors consisting of a single, membrane-embedded protein covalently binding a retinal (Vitamin-A aldehyde) chromophore [1]. The study of the microbial archetype, bacteriorhodopsin (BR) from the archaeon Halobacterium salinarum [2], led to the suggestion that BR evolved as a S34, yielded genes putatively encoding essential traits for survival under multiple environmental extreme conditions, e.g., high levels of UV radiation, elevated salinity, and the presence of poisoning arsenic concentrations [18,45,[48][49][50]. Likewise, genes putatively coding for the bacteriorhodopsin protein family were found showing sequential features indicating the presence of proteorhodopsins and xantorhodopsins [18,[45][46][47]. The former findings called for a more detailed, molecular investigation of the mechanisms involved in the resistance of these strains to extreme but common impairing factors in its original environment. We herein present the first functional characterization of E17R, a PR from Exiguobacterium sp. S17, a halotolerant, highly arsenic resistant extremophile isolated from a modern stromatolite, located at an altitude of more than 3,500 m above sea level.

Strain, media and culture conditions
The extremophile Exiguobacterium sp. S17 was previously isolated from modern stromatolites located at the shore of Lake Socompa (3,570 m; Fig 1) [50], and is currently maintained in the culture collection of Laboratory of Microbial Research on Andean Lakes (National System of Biological Data). For preculture and DNA extraction, S17 was grown in LB media or on LBagar plates (15.0 g/L).

DNA isolation
Genomic DNA from Exiguobacterium sp. S17 was purified from cells grown on LB broth for 24 h at 30°C and harvested by centrifugation (3,000 g for 10 min at 4°C). The pellets were washed twice with distilled water. Total genomic DNA was extracted with the DNeasy Blood and Tissue Kit (Qiagen) using the specific protocol for Gram-positive bacteria following the manufacture's recommendations.
Gene cloning E. coli strain XL-1 Blue (Stratagene) was used for all DNA manipulations. To amplify the putative PR gene from Exiguobacterium sp. S17, two gene-specific primers PR-S17F 5 0 GATCGAAT TCGATGGACGAAGAAGTCAATTTAC and PR-S17R 5 0 GATTCTCGAGCGTTTTAATTTGTTTG AGCGTCGCGACGCTCAAACAAATTAAAACG were designed according to the published genomic DNA sequence of S17 [48]. The gene start codon (ATG) is shown in bold-type in the forward primer. Restriction sites for EcoRI and XhoI were introduced in the forward and reverse primers (underlined). E17R -gene was PCR amplified and cloned into the pET-26b(+) vector (Novagene). The vector was transformed into Escherichia coli XL1 blue cells for further analysis.

Phylogenetic analysis of protein sequences
The sequence of the E17R protein from Exiguobacterium sp. S17 is available in GenBank (WP_016509804) together with the draft genome obtained by pyrosequencing (ASXD01). All sequences of PR-related proteins with significant homology (S1 File) were retrieved from public databases via the National Center for Biotechnology Information web site (www.ncbi.nlm. nih.gov). Multiple sequence alignments and phylogenetic analyses were carried out using MUSCLE [51] and Phylogeny.fr [52,53]. The evolutionary history was inferred using the Maximum likelihood method [54]. The tree was drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The percentages of replicate trees in which the associated taxa cluster together in the bootstrap test (1,000 replicates) are shown next to the branches. The analysis involved 22 amino acid sequences of PR-genes of Exiguobacterium strains. All positions containing gaps and missing data were eliminated.

Genomic environment analysis of proteorhodopsin
Genes for microbial rhodopsin and its genomic environment were identified using the RAST annotation platform [55] and the IMG analysis system [56]. The search and comparison was performed with all genomes available for the Exiguobacterium genus.

Protein expression and purification
Heterologous expression was performed with E. coli BL21 (DE3) cells transformed with the recombinant vector pET26-E17R. Cells from an overnight culture were used to inoculate a 200 mL culture, which then was used to inoculate 6 L of LB medium. Cells were grown with 40 μg/ mL kanamycin, until an OD 600 of 0.6 was reached. At that point, expression was induced with 0.5 mM IPTG. Cells were harvested by centrifugation after 12 h of incubation at 30°C. Cells were resuspended in 50 mM Tris, 5 mM MgCl 2 pH 8.5 added with 1 mM PMSF, 5 μM retinal (final concentrations) and DNase. Cells were then stirred for 1 h at 4°C and then lysed using a microfluidizer (Microfluidics Corporation, M110LA). The cell lysate was centrifuged for 1 h at 186,000 g at 4°C. The pellet was resuspended in 20 mM HEPES, 100 mM NaCl, pH 7.4, added with 5 μM retinal and 1% (w/v) n-dodecyl-β-maltoside (DDM) (final concentrations). Following a second centrifugation (1 h at 186,000 g, 4°C), the supernatant was filtered and proteorhodopsin was obtained by IMAC purification using a His-tag (ÄKTA, GE Healthcare).

Steady state absorbance and fluorescence spectroscopy
Steady state spectra were recorded at room temperature either on a Hitachi U3000 spectrophotometer or on a Hitachi F-4500 spectrofluorometer. The extinction coefficient of E17R was determined from the difference spectrum of a sample before and after treatment with 10 mM hydroxylamine (final concentration) and illumination (>480 nm) at room temperature. The value for E17R can be quantified from the known extinction coefficient of the retinal oxime, ε max = 33'600 M -1 cm -1 [57].

pH-titration
For pH-tritation experiments, proteorhodopsin samples were adjusted to A 540 = 0.1-0.5 in a buffer containing 20 mM HEPES, 100 mM NaCl, and 0.1% (w/v) DDM at pH 7.4. The titration was carried out in a 1 × 1 cm cuvette with continuous stirring. The pH values were adjusted with 1 M NaOH or 1 M HCl, before the spectra were recorded (Cary50, Varian, USA). Reversibility was checked at the end of the titration. Data were fitted using the Henderson-Hasselbalch equation.

Flash photolysis
Samples were prepared with identical buffer conditions as for the pH titration with pH = 7.4. Transient absorbance changes were recorded on a home-built flash-photolysis setup. The reaction was started by a 10 ns laser pulse from an excimer laser pumped dye laser (Coumarin 307, λ = 503 nm, 5-10 mJ cm -2 ). Light from 75 W XBO lamp was filtered by narrow-band interference filters and passed through the sample and a monochromator, before it was detected by a photodiode. Absorbance changes were followed over time on two oscilloscopes for the following wavelengths: 380, 400, 420, 440, 460, 480, 500, 517, 540, 562, 580, 600, 620, 645 nm. The number of data points was reduced by a logarithmic interpolation procedure before the data was analyzed by a global fitting routine. To this end, data blocks of centuries are averaged over 2 (n-1) points per n th century yielding 880 data points for the initial 45k points of the raw data for each of the oscilloscope with sampling rates of 200 ns per point and 20 μs per point, respectively. The two datasets were combined yielding 1520 point covering a range up to 1 second. 5 mM MgCl 2 , pH 8.5. The membrane-forming solutions contained 1.5% (w/v) diphytanoylphosphatidylcholine (Avanti Biochemicals, Birmingham, AL) and 0.025% (w/v) octadecylamine (Riedel-de-Haën, Hannover, Germany) in n-decane to obtain a positively charged membrane. Proteorhodopsin-containing proteoliposomes were added to the compartment under gentle stirring. A combination of ionophores (2 μM monensin and FCCP) was added, which effectively permeabilizes the compound membrane system (final conductance 50-100 nS). A 100 W Osram HBO mercury arc lamp with a combination of bandpass filters removing UV and IR irradiation (<380 nm and >750 nm) and a cut-off filter removing irradiation <495 nm was used for sample illumination. Blue light effect was generated by additional irradiation with light longer than 360 nm. The measurements were carried out at room temperature. The photocurrents were recorded with pClamp9.0 software via a MiniDigi 1A digitizer (Axon Instruments).

Results
A green proteorhodopsin from Exiguobacterium sp. S17 E17R, a gene predicted as belonging to the bacteriorhodopsin family (L479_RS10205/ WP_016509804), was found in the genome of Exiguobacterium sp. S17. It encodes a protein of 248 amino acid residues (744 bp), E17R, with features typical of a microbial rhodopsin. All essential amino acid residues of the energy transducing rhodopsins are conserved (Figs 2 and 3). E17R has 85% sequence identity with the green-PR of the psycrophilic Exiguobacterium sibiricum (ESR) for which the crystal structure was recently solved (PDB: 4HYJ) [21]. Alignment of these two proteins sequences with that of the blue-light absorbing proteorhodopsin (BPR) of uncultured Gamma-proteobacterium Hot 75m4 (http://www.uniprot.org/taxonomy/ 245185) clearly showed the instrumental difference in an amino acid position associated with spectral tuning. Both Exiguobacterium proteins contain a leucine residue at position 94 (E17R, position 93 at ESR), which fine-tunes the PR absorption peak towards green light (absorption maximum 524 nm, Fig 4). This is in accordance with the shallow nature of the stromatolite formation in the shore of L. Socompa, the original environment of S17 (Fig 1). In turn, the corresponding position in BPR (106) is occupied by a glutamine, which is known to yield an absorption maximum near 490 nm [5,32], a feature for proteorhodopsin-bearing microbes from deep-sea in coincidence with the light quality available at depths greater than 50 m.
A BLASTP search in all publicly available genomes of the Exiguobacterium genus was performed. Twenty-one genomes displayed a sequence with significant similarity (higher than 80%) to that of S17 proteorhodopsin (S1 File). E17R shares the highest sequence identity with Exiguobacterium sp. 8-11-1 (98%), Exiguobacterium sp. AB2 (95%) and Exiguobacterium sp. AT1b (94%). All PR sequences display Leu in position 106 (following the numbering of the BPR Gamma-proteobacterium Hot 75m4), assigning them to the "green-tuned" PR subfamily. A phylogenetic tree (Fig 3) constructed with the 22 related rhodopsins (S1 File) revealed two divergent clades, one including E17R (mostly form by alkaliphiles/thermophiles) and the other one mostly formed by psycrophiles/psycrotolerants including two branches, the one of Exiguobacterium acetylicum and the one of E. sibiricum ESR. Thus, in view of the phylogenetic distance from ESR, further biochemical and spectroscopic studies of the novel proteorhodopsin E17R were performed.

Biochemical and steady-state spectral properties of E17R
Gene E17R (L479_RS10205/WP_016509804) encodes for a protein with 248 amino acids (MW: 27.4 kDa) that showed typical features of a microbial rhodopsin. Secondary structure prediction identified seven transmembrane α-helices, a lysine residue serving as putative retinal binding site at the seventh helix and proton donor and acceptor residues that identify this protein as a light-driven proton pump (Fig 2). Expression and purification of E17R yielded after purification and retinal addition a protein with λ max = 524 nm (solubilized in 0.03% [w/v] DDM, Fig 4). The chromophore extinction coefficient was determined as 50,000 M -1 cm -1 from the difference spectrum after bleaching in the presence of hydroxylamine, which formed the retinal oxime (inset in Fig 4) [57]. The ratio of protein (280 nm) to chromophore (524 nm) absorbance can be taken as a purity factor and it amounts to~1.3 (taking into account increased scattering around 280 nm). The value is slightly higher than the theoretical limit of 1 as calculated from the extinction coefficients for the protein moiety and the chromophore (Fig 4).
We followed pH-induced absorbance changes in E17R between pH 2 and pH 10 in order to identify de-or reprotation of amino acids in close proximity to the chromophore (Fig 5). As reported for the closely related ESR [58], pH-titration reveals an ongoing red-shift of the absorption maximum upon acidification. Usually, this colour change is attributed to the protonation of the counterion of the protonated Schiff-base (D85 in ESR). pH-Titration experiments with E17R showed a similar absorbance change suggesting that D86 is the Schiff base counterion in E17R. pH-values lower than 2.0 could not be reached, as the protein lost stability irreversibly. In the accessible pH range, we observed a complex titration behaviour, similar to that described for ESR. For this protein, the data was interpreted as a strong interaction between the counter-ion D85 (D86 for E17R) and a conserved histidine residue (H57 in ESR, H58 in E17R) located in the second transmembrane helix (TMH B) [55]. For E17R, the fit of the absorbance changes shifting from 525 nm to 535 nm or higher gives a pK 2~3 (Fig 5). Note that this is an approximate value, as the final state at acidic pH (<2) was not reached. The second transition is observed with a pK 1~1 0 accompanied by smaller spectral changes (from 517 nm to 525 nm). The fluorescence data is in line with the argument given in Balashov et al. (2012) [58] for the assignment of the titratable groups. The major increase in fluorescence intensity observed at wavelengths longer than 620 nm can be attributed to the protonation of the proton acceptor D86 and takes place only at pH values <3 (data not shown). On the other hand, there is almost no change in fluorescence intensity at pH values >8. Therefore, the second transition might be assigned to deprotonation of the His interacting strongly with D86. It was proposed that this interaction keeps the counter-ion in its deprotonated state even at low pH values. Such behaviour is different from, e.g., the very first identified green-absorbing PR that has a pK a~7 .5 for its counter-ion D100 [58].

Flash photolysis
Flash photolysis as a time-resolved detection method for absorption changes allows following -after a short laser pulse (routinely 10 ns)-absorbance changes at a wide wavelength range on a time domain from few microseconds into the seconds range. Detection of absorbance changes, followed by fit procedures provides details on the lifetime of intermediate states formation and decay. In most cases, the global fit approach is used where all detected absorbance changes over time and wavelengths are subjected to one single fit model. This data evaluation yields lifetimes and also spectral shape of the intermediates through which the protein travels. Here, we followed the built-up and decay of photo-intermediates from time-resolved absorbance changes in the microsecond to second time range after excitation of E17R's initial dark state by a nanosecond laser flash (Fig 6A). The results allow us to deduce a photocycle with a  proteorhodopsin from Modern Stromatolites lifetime of 82 ms for the recovery of the parental dark state. Global fit analysis of the kinetic traces clearly identifies formation and decay of different intermediates and the bleaching and recovery of the ground state (Fig 6B). The first intermediate is red-shifted compared to the ground state. Its formation is below the time resolution of our system. It decays partly in a fast process (τ 1 = 3.5 μs), followed by the formation of another red-shifted intermediate. This species is built in a two-step process with small amplitudes (τ 2 = 84 μs and τ 3 = 11 ms). A lifetime of τ 4 = 82 ms corresponds to the final decay of all red-shifted intermediates and the concomitant formation of the parental 524 nm-form. These flash photolysis experiments did not reveal the presence of a deprotonated species (M-state). Such behaviour is reminiscent of the ESR that also has no noticeable transient concentration of an M-state [58][59][60]. It was argued for the ESR protein that the absence of this intermediate is solely due to kinetics parameters, i.e., caused by similarly fast formation and decay kinetics.

Light-induced proton transport currents
The protein pumping activity for the closely related ESR protein was measured in E. coli cell suspensions [58]. For E17R, we used a more direct proof employing the black lipid membrane (BLM) method [61]. E17R was reconstituted in proteoliposomes that were attached to BLM yielding capacitive coupling. Light activation generated a current of the compound membrane ( Fig 7A). Doping the membrane with the protonophore FCCP makes the membrane proton selective and turns it into a DC coupled system, now also allowing measurement of stationary transport currents as a result of the light-driven proton transport (Fig 7B). The data clearly illustrates the light-driven proton translocation by E17R, but, interestingly, this method displays a blue-light effect in the photocycle after additional irradiation with blue light (Fig 6B, illumination >380 nm), and thus gives evidence for the presence of an M-like intermediate. Its presence becomes visible by the concomitant excitation of the ground state and the M-state that absorbs in the blue-/near UV region. The excitation of the M-state provides a shunt in the photocycle abolishing proton pumping which leads to a current amplitude decrease.
Sequence comparison and phylogenetic analysis of E17R indicated that it was a green-tuned proteorhodopsin, being the case also to all proteorhodopsins found in the so far reported Exiguobacterium genomes. Accordingly, the original environments of most isolates are shallow or surficial such as microbial mats, soil, rhizosphere, glacier ice, or hot springs (S1 File). Nevertheless, PRs were also present in Exiguobacterium strains isolated from ecological niches where light is less abundant such as marine sediments, industrial waste and even animal gastrointestinal tracts (S1 File). The ecological significance of PRs for survival under stressful conditions was evident for a Vibrio sp. strain in which deletion of the PR gene reduced survival during carbon starvation [64]. Light-dependent promotion of starvation survival and enhanced PR expression was also established for the marine dinoflagellate Oxyrrhis marina [36]. Bohorquez et al., (2012) [16] found diverse PR-like genes in microorganisms from shallow, oligotrophic hot spring waters at the Colombian high-Andes indicating that rhodopsin photosystems could be advantageous in these acidic hot springs by contributing to survival in ecosystems that receive abundant sunlight and where alternative energy sources may vary or be scarce. For the Socompa stromatolite community, the use of light to counteract stressful conditions could also be an interesting strategy especially to those communities situated at the top layer where light is fully available but where UV stress is maximal [41]. In accordance, all four available genomes isolated from stromatolite indigenous strains (from the top 5 mm) contained microbial rhodopsin-likes genes even despite belonging to distant taxonomical groups (firmicutes, proteobacteria and bacterioidetes) [18,48,49].
Although isolated from completely different niches (Andean stromatolites in hypersaline lakes vs. Siberian Permafrost), E17R shows highest homology to ESR for which recently a three-dimensional structure was reported [21]. Both PRs share the essential pump amino acids and show clear differences to the blue light-absorbing BPR. Common between all three proteins compared here (Fig 2) is the proton acceptor (D86, D85 in ESR), and also the close interaction with His58 (His57 in ESR, and His76 in BPR) that keeps the proton acceptor in its deprotonated state during the photocycle. Different, however, is a lysine as putative proton donor (Lys97 in E17R, Lys96 in ESR). Here, BPR resembles bacteriorhodopsin as both proteins carry an acidic amino acid as proton donor (D96 in BR, and E109 in BPR). Apparently, the protonated side chain of lysine (better coined an ammonium group) undergoes transient deprotonation and therefore functionally complements the normally present acidic side chain of, e.g., D96 in BR. Another feature shared by both green-absorbing PRs (different from BPR and BR) is an extension of the loop between helices three and four. Seven additional, nearly fully conserved amino acids-NGGFTQL (N is exchanged for A in ESR)-are present in the green absorbing PRs.
Inspection of the sequences of E17R and ESR identify positions at which amino acid are different (vertical arrows in Fig 2). In most cases aliphatic amino acids are exchanged against each other. Alternatively, polar (Thr, Ser) and acidic amino acids (Asp, Glu) were exchanged. Overall, the exchanges found in ESR are of a more polar nature compared to the corresponding positions in E17R (numbering follows E17R, exchanges in ESR are given after the dash: T76-D, G147-S, N183-K, V224-T). A more drastic change is found in ESR that carries an ion pair (K172 and D178), located at the entrance into the sixth helix at a position, where E17R carries a serine (173) and a lysine (179).
Highest conservation between all three aligned PRs is found in helix three with the proton donor and acceptor residues, except for the exchange of the acidic proton donor in BR and BPR for a lysine. In order to accomplish its function, the pK value of this lysine has to be lowered significantly compared to aqueous solution (as that of aspartate or glutamate is remarkably increased for maintaining their protonated state). Helix seven also shows a high degree of sequence similarity, except for the exchange V220 ! F (E17R to ESR) and V224 ! T in direct vicinity to the retinal-binding lysine.
Flash photolysis of E17R showed a photocycle comprised only of long wavelengths absorbing intermediates. From the initial photoproduct on (detection of its formation being below the time resolution of the instrument), all further intermediates show continuously increasing life times (τ i = 3.5 μs, 84 μs, and 11 ms). The initial PR 524 state is then formed with τ 4 = 82 ms. At first glance the apparent absence of the M-intermediate in the transient absorbance changes at the measuring pH value of 7.4 is puzzling, as a de-and reprotonation is instrumental for a vectorial proton transport. In fact, the BLM measurements clearly demonstrate the presence of a photocurrent and blue light irradiation opens a short circuit pathway leading to reduce current. Thus, the lack of an M-intermediate at the pH-value chosen for this experiment is due to purely kinetic reasons (see also the pH-dependent measurements in [58]).

Conclusions
In this work, we have successfully cloned and heterologously overexpressed a proteorhodopsin gene from a halotolerant, UV-and arsenic resistant strain originally isolated from a high-altitude Andean Lake (ca. 3,600 m). To our knowledge, this is the first report of a functional proteorhodopsin isolated from modern stromatolites. Its presence in a stromatolite and its functional, light-driven proton transport activity identifies Exiguobacterium sp. S17 as an important member of this complex microbial community. Further studies including knock-out strategies on E17R are in progress to link light-harvesting functionality to microbial physiology. High-throughput sequencing metagenomic projects on Socompa stromatolites are also being conducted to ultimately reveal the ecological significance of these versatile novel ion pumps in the structure and function of the microbial extremophilic communities. The studies offer a high window to study ancient phototrophic systems in homologous stromatolites of their Precambrian counterparts.
Supporting Information S1 File. Isolation source and other relevant information on Exiguobacterium spp. strains in which PR -like genes were found. (DOCX)