Binding affinity of AbGS for mutated AbHeR

Next, we determined the AbHeR residues involved in AbGS binding. Considering AbGS is a soluble protein, we postulated that the binding residues in AbHeR were located on the cytoplasmic side. We identified AbHeR charged residues on the intracellular loop (ICL) using Protter [34] and amino acid alignment tools (Fig 2A and 2H). AbHeR and AbGS were predicted to be in dimer and dodecamer forms, respectively (S3A–S3D Fig). To elucidate the role of electrostatic forces in the associated PPIs, we used an electrostatic potential map and found that AbGS did not form a field with a dominant charge in a specific area (S3E Fig). Consequently, we analyzed the electrostatic potential map for AbHeR (S3F Fig). Notably, AbHeR exhibited negatively and positively charged areas on the cytoplasmic side. The positively charged area included K3, K217, K216, K220, and R226, in the front and side membrane views. In the bottom cytoplasmic side view, the negatively charged area included E150, E151, and D223. Overall, there were more positively charged areas than negatively charged fields; K3, K216, K217, K220, and R226 formed the field in the front-side-bottom view (cytoplasmic to extracellular side; S3F Fig).

Open in a separate window

Fig 2

PPIs and binding affinities of AbHeR for AbGS.

(A) Multiple sequence alignments of AbHeR and predicted heliorhodopsin in the GS group were aligned using MUSCLE. Alignments of AbHeR with heliorhodopsin in the GS group. Negatively and positively charged residues near intracellular loops of heliorhodopsin are indicated by red and blue, respectively. (B–G) Dissociation constants of AbHeR for AbGS determined using ITC analysis. The upper and lower panels represent raw data and enthalpy change per mol, respectively. AbGS was continuously added to AbHeR mutants. Nonlinear curves of lower panels were evaluated by the best-fit curve. (H) Scheme of PPIs in each amino acid position of AbHeR for AbGS, as predicted by Protter. Positively and negatively charged residues on the cytoplasmic side are indicated as red and blue circles, respectively, with arrows. Dotted arrows indicate not expressed and solubilized proteins. Fold-changes for AbHeR mutants compared to WT indicate text with black arrows. (I) Predicted model of AbHeR function associated with AbGS. AbHeR dimers are embedded in the membrane, and the N- and C-termini of AbHeR are located on the CP and EC sides, respectively. AbGS dodecamer binds to AbHeR dimers and exhibits altered enzyme activity in the presence of light. The underlying data of the ITC analysis can be found in S1 Data. CP, cytoplasmic; EC, extracellular; GS, glutamine synthetase; ITC, isothermal titration calorimetry; MUSCLE, MUltiple Sequence Comparison by Log-Expectation; PPI, protein–protein interaction; WT, wild type.

We confirmed that the charged residues of heliorhodopsins on ICL2 and ICL3 were conserved compared to in the ICL1 region and N-terminus (S2B Fig). In particular, charged residues of heliorhodopsins in the GS group were highly conserved on the ICL2 and ICL3 (Fig 2A). Hence, we constructed the following AbHeR mutants with alterations in ICL2 and ICL3 regions: E147Q, K148Q, E150Q, E151Q, K216Q, K217Q, K220Q, D223N, and E228Q. The AbHeR E147Q mutant was successfully expressed, however, was not solubilized in 1% DDM. AbHeR E150Q and E228Q mutants were not expressed. The AbHeR K148Q, E151Q, K216Q, K217Q, K220Q, and D223N mutants were successfully expressed and purified, with the absorbance maxima remaining unaltered (S1H Fig).

We performed ITC analysis for the AbHeR mutants with AbGS. The K_d values of K148Q and E151Q on ICL2 was determined to be 137 (±44) and 30 (±8) μM, respectively (Fig 2B and 2C, and Table 1). The K_d value of K148Q was similar to that of WT, indicating that the K148 position was not influenced by AbGS binding. However, the K_d value of E151Q decreased by approximately 3 times (Fig 2H and Table 1). The K_d values of K216Q, K217Q, K220Q, and D223N on ICL3 were determined to be 3,676 (±5,595), 537 (±1,751), 833 (±2,840), and 44 (±23) μM, respectively (Fig 2D–2G and Table 1). Three positively charged residues on ICL3 influenced the binding to AbGS, and the K216 position exhibited extremely weak binding when compared with WT, resulting in an approximate 37-fold increase in the K_d value of K216Q (Fig 2H and Table 1). Hence, we propose that the K216 position is critical for binding of AbGS and that the K217 and K220 residues facilitate the binding. The K_d value of D223N also decreased by approximately 2 times similar to E151Q, suggesting that the absence of negatively charged residues is required for AbGS binding (Fig 2H and Table 1). SRs can respond to various light stimuli and regulate the transducer for signal transduction in different organisms [11]. Similarly, the PPI between AbHeR and AbGS may regulate AbGS enzyme activity in the presence of light, which may influence the growth or light stimuli of Actinobacteria bacterium IMCC26103 (Fig 2I).

AbHeR regulates AbGS activity

GS is one of the key enzymes that participate in nitrogen metabolism, as it catalyzes glutamine synthesis from glutamate via the following reaction (Eq 1):

The biosynthetic reactions include conversion of L-glutamate, in the presence of ATP, ammonia, and metal ion (Me²⁺, magnesium, or manganese), to L-glutamine with adenosine diphosphate (ADP), inorganic phosphate (P_i), and a proton as by-products [35]. We performed a biosynthesis assay rather than the γ-glutamyltransferase assay as the procedures of the latter involve the use of hydroxylamine, which affects rhodopsin bleaching [36]. The biosynthesis assay involves measuring the concentration of P_i after GS catalysis and colorimetric determination of P_i with ferric sulfate and molybdate [37,38]. AbGS activity was calculated using a standard curve for P_i concentration (S4A Fig), and enzyme kinetic parameters of AbGS were determined (Table 2 and S4B and S4C Fig). k_cat and k_cat/K_m values of reported recombinant GSs for L-glutamate were 2.0 to 30.5 s^-1 and 900 to 28,000 M^-1s^-1, respectively [39–41], implying that AbGS activity is lower than those of the reported recombinant GSs.

Next, to stabilize AbHeR for heat and illumination, we prepared an inverted membrane vesicle (IMV), (inside-out (ISO) membrane vesicle) with an orientation opposite that of RSO membrane vesicles [42,43]. We predicted that the exposed cytoplasmic side of AbHeR IMV would readily bind AbGS. The absorption spectra of the prepared AbHeR IMV and empty IMV were measured; both AbHeR WT and K216Q IMV exhibited absorption maxima similar to that of purified AbHeR (S1B and S1H and S5A Figs). To check the reverse direction of IMV, a light-induced H⁺ movement assay was performed with Gloeobacter rhodopsin (GR) WT as a light-induced H⁺-pumping rhodopsin [44]. The GR WT IMV pumped H⁺ inward, compared with GR WT (S5B Fig), suggesting that our prepared IMV was successfully inverted. Moreover, AbHeR IMV exhibited thermal and light stability (S5C Fig).

As AbHeR WT binds to AbGS, we postulated that AbHeR WT may influence AbGS activity. To test this hypothesis, we employed the AbHeR K216Q mutant as a control, which was not expected to influence AbGS activity due to its lack of binding activity. To confirm PPIs between AbHeR IMV and AbGS, the associated binding affinity was analyzed using ITC analysis. The IMVs were continuously injected into the AbGS, wherein Escherichia coli C43 (DE3) and AbHeR K216Q IMVs were not observed to be bound (Fig 3A and 3C). However, AbHeR WT IMV was bound to AbGS. These results are consistent with the ITC data of purified AbHeR. The K_d value of AbHeR WT IMV was determined to be 6.06 (± 5.44; Fig 3B). Interestingly, the binding affinity of AbHeR IMV for AbGS (Fig 3B) was improved when compared with that for purified AbHeR with AbGS (Fig 1C). The only difference between the proteins was the absence or presence of detergent (DDM). DDM stabilizes membrane proteins; however, it may bind to several areas of AbGS as a soluble protein. Therefore, areas of AbGS do not interfere with the binding to AbHeR. Consequently, we suggest that AbGS suitably binds to AbHeR in the E. coli membrane and can be used for subsequent in vivo experiments. Furthermore, AbHeR may bind to AbGS in living cells.

Open in a separate window

Fig 3

AbGS activity is regulated in vitro via AbHeR binding.

The binding of AbGS to preparations of Escherichia coli C43 (DE3) (A), AbHeR WT (B), and AbHeR K216Q (C) IMVs were analyzed by ITC analysis. The upper and lower panels represent the raw data and enthalpy change per mol, respectively. AbGS was continuously added to IMVs. The nonlinear curves of the lower panels were evaluated by best-fit curve analysis. (D–F) AbGS reactions were performed with AbHeR WT and K216Q IMV in the absence (dark) and presence of light (light, 532 nm) for time traces at 30 μmol m^-2 s^-1 using a biosynthesis assay. The nonlinear and linear fits were evaluated using the Michaelis–Menten equation (E and F). (D) Biosynthesis assay of AbGS reactions with AbHeR WT IMV carried out with 100 mM L-glutamate at different times. The measurements were conducted in an independent experimental group (n = 6), and the data are expressed as mean ± standard deviation. The statistical significance between the 2 groups was analyzed using the t test; p-values are indicated by an asterisk. The underlying data of the ITC analysis and the biosynthesis assay can be found in S1 and S2 Data, respectively. IMV, inverted membrane vesicle; ITC, isothermal titration calorimetry; WT, wild type.

AbHeR IMV, is heat and light stable, binds to AbGS, and is suitable for studying the effect of AbHeR on AbGS. We designed a biosynthesis assay to investigate the effect of AbHeR in the presence of light, wherein the reactions of AbHeR IMV with AbGS were subtracted from AbHeR IMV alone because ATP hydrolysis by embedded membrane proteins in IMV can be influenced. The AbGS activity with AbHeR WT was statistically significantly increased in the presence of light, implying that AbGS was regulated by light-induced AbHeR (Fig 3D). Consistent with the ITC analysis results, AbHeR K216Q did not bind to AbGS and was not significantly altered in the presence of light (Fig 3D). In the absence of light, the amount of a product (P_i) produced by AbGS bound to AbHeR WT at 150 min was increased 1.09 times when compared with that in AbHeR K216Q. In the presence of light, the amount of the product produced by AbGS bound AbHeR WT at 150 min was increased 1.1 times when compared with that observed in the absence of light. When compared with AbGS with AbHeR K216Q, the amount of the product produced by AbGS bound to AbHeR WT at 150 min was increased 1.2 times in the presence of light. Therefore, AbGS is more productive when bound to AbHeR than in its unbound state; the effect is stronger in the presence of light.

For more accurate comparisons of AbGS activity, we quantified constant values through the enzyme kinetic parameters of AbGS with AbHeR in the absence and presence of light (Table 2 and Fig 3E and 3F). In the absence of light, K_m and k_cat values of AbGS with AbHeR WT were lower than those for AbGS alone. In the presence of light, the K_m and k_cat values were lower and greater, respectively, suggesting increased activity. Although the activity of AbGS with AbHeR WT was greater in the presence of light, the K_m and k_cat values were lower under such conditions than under AbGS alone. However, as the values differ for each condition, comparing the actual AbGS activity is challenging. It is more appropriate to compare the catalytic efficiencies based on the assessment of the specificity constant (i.e., k_cat/K_m ratio). In the absence of light, the k_cat/K_m value of AbGS in the presence of AbHeR WT was 1.52 times greater than that for AbGS alone. In the presence of light, the k_cat/K_m value of AbGS in presence of AbHeR WT was 2.24 times greater than that for AbGS alone. Moreover, the AbHeR K216Q, mutation of a critical binding residue for AbGS, did not regulate AbGS significantly when compared with AbGS alone; however, the k_cat/K_m value of AbGS in the presence of AbHeR K216Q and light was 1.27 times greater than that in the absence of light. According to the ITC results, there was weakly binding between this mutant and AbGS, resulting in an increase its activity in the presence of light. In a nonessential activation model, which is similar to reversible mixed inhibition, enzyme activity is activated, resulting in changes in both K_m and V_max [45]. Therefore, AbHeR might function as a nonessential activator for AbGS.

Since the ITC analysis is performed in the absence of light, AbGS can bind to AbHeR in the absence of light. That is, AbGS binds to the interaction residues (K216 or K216 with other residues) of AbHeR on the cytoplasmic side, followed by conversion of all-trans retinal (ATR) to 13-cis retinal in AbHeR, resulting in a conformational change in AbHeR in the presence of light. In contrast, AbGS cannot bind AbHeR in the absence of the positively charged residue (K216). Therefore, we suggest that conformational changes in AbHeR influence the interaction with, and activity of, AbGS, and that AbHeR up-regulates AbGS activity.

Growth rate test of cells with AbGS and AbHeR

We hypothesized that co-expression of AbHeR and AbGS in a cell would up-regulate AbGS activity in the presence of light. E. coli K-12 MG1655 and JW3841 strains (glnA knockout (KO) strain derived from K-12 MG1655) were used as the positive control and experimental group to study the influence of the absence of glnA encoding GS alone, respectively. Cellular growth rate was assessed in E. coli K-12 MG1655 and JW3841 strains to compare the complementation of AbGS with AbHeR.

We added L-glutamate to 3-morpholinopropane-1-sulfonic acid (MOPS) minimal medium to induce AbGS activity. The K-12 MG1655 strain was transformed with the all-in-one (AIO) plasmid containing ampicillin and kanamycin as the glnA KO strain exhibited kanamycin resistance. The glnA KO strain was transformed with pKA001 vector containing AbGS and AbHeR [46], which were regulated by P_lac and P_araBAD, respectively. The growth rate of cells was then measured in the absence or presence of L-glutamate, which is required for the catalytic activity of AbGS. Finally, we assessed the growth rate of cells upon complementation of AbGS with AbHeR.

The K-12 MG1655 strain was grown in the absence of L-glutamate and saturated for 48 h. In the presence of L-glutamate, the cells were saturated at a higher cell density, indicating that supplementary L-glutamate influenced cell density and growth rate (Fig 4A); i.e., the MOPS minimal medium in place of Luria–Bertani (LB) medium is sufficient to permit cell growth. The glnA gene in E. coli is the key enzyme involved in nitrogen fixation; glnA-deficient E. coli strain reportedly does not grow in minimal mediums [47–49]. Therefore, we postulated that if AbGS catalyzes L-glutamate to L-glutamine in the glnA KO strain in the presence of L-glutamate, it would exhibit cell growth similar to that of the K-12 MG1655 strain. As expected, the glnA KO strain containing the null vector (glnA KO + pKA001) did not grow in MOPS minimal medium. Upon the addition of L-glutamate, cell density began to decrease after 72 h (Fig 4B). Similarly, the glnA KO strain expressing AbGS (glnA KO + AbGS) did not grow in the absence of L-glutamate. Meanwhile, the cells showed slow growth and became saturated at 72 h in the presence of L-glutamate, indicating that AbGS successfully converted L-glutamate to L-glutamine (Fig 4C). However, the complementation was not recovered to the level exhibited by the K-12 MG1655 strain, suggesting that AbGS activity is lower than that of the well-known GS.

Open in a separate window

Fig 4

In vivo growth rate of E. coli glnA KO strains with AbGS and AbHeR.

Cells are grown in LB medium and washed to remove components of the LB medium. The cells are then transferred to MOPS minimal medium. The growth of transferred cells was recorded at OD₆₀₀ at different time points; growth is normalized to the initial recording. The E. coli K-12 MG1655 strain transformed with AIO vector (A), glnA KO strain transformed with pKA001 null vector (B), and glnA KO strain transformed with pKA001-GS vector (C) were incubated in the absence and presence of L-glutamate in MOPS minimal medium containing IPTG, which regulates AbGS expression. (D–G) The E. coli glnA KO strain transformed with pKA001 null or pKA001-GS vector was incubated in MOPS minimal medium containing L-glutamate and glycerol (MOPS-GG). The E. coli glnA KO strain transformed with pKA001-GS-AbHeR WT (F) and K216Q (G) with and without ATR in the absence (dark) and presence of light (light, 532 nm) conditions at 40 μmol m^-2 s^-1 were incubated in MOPS-GG medium. (A–E) The tests were performed in an independent experimental group (n = 3). (F and G) The tests were performed in an independent experimental group (n = 9). (E) Statistical significance between the 2 groups (pKA001 and AbGS) was analyzed using the t test; p-values are indicated with asterisks. (F and G) Statistical significance between the 2 groups (dark and light) was analyzed using the t test; p-values are indicated with asterisks. The data are expressed as mean ± standard deviation. Based on data (F) and (G), half-life (t_1/2) for growth rate of cells was analyzed to quantify the effect of AbGS binding in the absence and presence of light by exponential decay. Nonlinear fitted lines are shown for AbHeR WT (H) and K216Q (I). Quantified t_1/2 values are indicated by the same color as that of each rectangle, circle, and triangle. The underlying data of the graph can be found in S3 Data. AIO, all-in-one; KO, knockout; LB, Luria–Bertani; WT, wild type.

L-glucose, as a carbon source in the MOPS minimal medium, can interfere with isopropyl β-D-1-thiogalactopyranoside (IPTG), which regulates AbGS expression on the lac operon; therefore, we used glycerol as the carbon source in the MOPS minimal medium. In addition, we assumed that the medium supplemented with glycerol provides a limited source of energy for growth. The synthesis of recombinant proteins expressed by IPTG may affect cell growth rate. Therefore, we added different concentrations of IPTG to the medium. The glnA KO + AbGS showed no significant difference in cell growth across different IPTG concentrations in MOPS minimal medium containing glycerol (Fig 4D), implying that cells transferred into the medium do not require IPTG for complementation through AbGS.

Next, we conducted experiments with MOPS minimal medium containing L-glutamate and glycerol in the absence of glucose and IPTG (MOPS-GG medium). We compared glnA KO + pKA001 and AbGS to confirm complementation of AbGS in MOPS-GG medium. AbGS showed complementation in the glnA KO strain; the cells exhibited reduced growth in the absence of AbGS and complementation with AbGS statistically significantly slowed down the decay of cells (Fig 4E). We designed a strain in which AbGS and AbHeR were co-expressed in the pKA001 vector, and their expression was induced by IPTG and L-arabinose, respectively. We assumed that AbHeR without ATR cannot regulate AbGS in the absence or presence of light, whereas AbHeR with ATR regulates AbGS in the presence of light. Therefore, the growth of cells may be altered. However, the cells co-expressing AbHeR and AbGS showed decreased growth levels similar to that of glnA KO + pKA001 cells (Fig 4F). AbGS expression levels in the cells co-expressing AbHeR and AbGS were lower than those in the cells expressing AbGS after induction, indicating that a specific amount of AbGS is essential for complementation (S4D–S4F Fig). Although the cells co-expressing AbHeR and AbGS showed reduced growth rate, AbHeR may influence the growth of cells in the presence of light. Surprisingly, the growth of cells co-expressing AbGS and AbHeR with ATR in the presence of light was statistically significantly higher than that in the absence of light. In addition, the growth of cells expressing AbGS and AbHeR without ATR was not significantly different from that of cells expressing AbGS and AbHeR with ATR in the absence of light. That is, AbHeR could not regulate AbGS in the absence of light harvesting (Fig 4F). The low binding affinity of AbGS in the AbHeR K216Q mutant was measured using the same experimental method as that for AbHeR WT. AbHeR K216Q did not influence cell growth significantly (Fig 4G). We quantified the growth levels because it was difficult to compare the down decay of cells expressing AbGS as well as AbHeR WT or K216Q. Using exponential decay, the t_1/2 of growth rate of cells containing AbHeR WT in the presence of light was 1.58 times greater than that in the absence of light (Fig 4H); however, t_1/2 values of growth rates of cells containing AbHeR K216Q in presence of light were not altered when compared with those in the absence of light (Fig 4I). These t_1/2 values also suggest that AbGS can be partially regulated by AbHeR in the presence of light and that AbGS does not bind well to AbHeR K216Q.

Plasmid preparation

AbHeR (accession number: QLL25365)- and AbGS (accession number: QLL25366)-encoding genes containing NdeI and NotI restriction sites with a hexahistidine-tag were codon-optimized for protein expression in E. coli and chemically synthesized (Integrated DNA Technologies, United States of America). The hexahistidine-tags of AbHeR and AbGS were located at the N- and C-termini, respectively. The genes were cloned into the NdeI and NotI sites of the pET21b and pKA001 vector [46]. Each of the AbHeR mutants was prepared using site-directed mutagenesis, and the mutated gene sequences were cloned into the pET21b vector. For the complementation experiment, NcoI and SalI restriction enzyme sites were introduced into the AbHeR amplicon and introduced into pKA001-AbGS.

The AIO plasmid was designed using E. coli K-12 MG1655 and contains ampicillin and kanamycin resistance genes. The ccdB gene in the plasmid was removed and inserted internal ribosome entry site (IRES2) gene into sites removed the ccdB gene. The IRES2 fragment was amplified by polymerase chain reaction using the 5′-AGATCTTAACCCCCCCCCTAACGTTAC-3′ (BglII) and 5′-ATGCATGTATTATCGTGTTTTTCAAAGG-3′ (NsiI) primers containing restriction enzyme sites from the MSCV-IRES-GFP vector. The IRES2 amplicon was introduced into the All-In-One vector (Biofact, Korea), and the resultant vector was named the AIO plasmid. The GR has been studied previously [73], and the GR WT gene in the pKA001-GR WT vector was introduced into the NdeI and NotI restriction sites of the pET21b vector.

Protein expression and purification

Rhodopsin expression and purification procedures have been previously described [8]. The AbHeR WT, AbHeR mutants, and GR WT were expressed in E. coli C43 (DE3) cultured in LB medium containing 50 μg/mL ampicillin at 37°C and 200 rpm. The cells were induced by 0.84 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) and 7 μM all-trans retinal (ATR, Toronto Research Chemicals, Canada). After induction, the cells were harvested with a buffer S (150 mM NaCl and 50 mM Tris-HCl (pH 7.0)) and subsequently disrupted using a sonicator. Membrane fractions were isolated using ultracentrifugation (Beckman × L-90 ultracentrifuge, USA) at 100,000 × g and 4°C for 1 h and then solubilized with buffer S containing 1% n-dodecyl-β-D-maltopyranoside (DDM, Goldbio, USA) at 4°C for 16 h. The solubilized rhodopsins were prepared with Ni²⁺ NTA agarose (QIAGEN, USA) by rocking at 4°C for 4 h in the absence of light after ultracentrifugation at 4°C and 45,000 × g for 15 min. The protein samples were washed with buffer SD (150 mM NaCl, 50 mM Tris-HCl (pH 7.0), and 0.02% DDM) containing 25 mM imidazole and subsequently eluted with buffer SD containing 250 mM imidazole using Amicon Ultra-4 10,000 MWCO centrifugal filter units. The AbGS expression and purification procedures were slightly modified.

The E. coli BL21 (DE3) strain harboring AbGS was incubated in LB medium containing 50 μg/mL ampicillin at 37°C and 200 rpm until the optical density at 600 nm (OD₆₀₀) reached 0.4. The cells were induced with 0.84 mM IPTG and incubated at 37°C and 200 rpm for 4 h. The cells were harvested and washed with buffer S. The harvested cells were resuspended in buffer S containing 1 mM phenylmethylsulfonyl fluoride (PMSF) and subsequently disrupted using a sonicator. The cell lysates were centrifuged at 30,000 × g and 4°C for 45 min. After centrifugation, the supernatants were subjected to separation using Ni²⁺ NTA agarose by rocking at 4°C for 2 h. AbGS was washed with buffer S and then eluted with buffer S containing 50 mM imidazole.

Three-dimensional structure and protein–protein docking analysis

AbHeR and AbGS sequences were analyzed using the Swiss-Model and predicted by an available protein crystal structure HeR-48C12 (PDB code: 6su3.1.A) for AbHeR and GS (PDB code: 3qaj.1.A) for AbGS [74–76]. The predicted crystal structures were analyzed using PyMOL (The PyMOL Molecular Graphics System, Version 2.52, Schrödinger, LLC). Electrostatic potential maps of the predicted three-dimensional (3D) structures of AbHeR and AbGS were analyzed using CHARMM-GUI [77]. Protein–protein docking between the predicted crystals structure AbHeR and AbGS was conducted using ClusPro 2.0 server [78–81]. The ClusPro server performed 3 computational steps. Rigid body docking was used to sample billions of conformations using PIPER, which calculates the docked conformation energy in grids based on the fast Fourier transform (FFT) correlation approach. Root-mean-square deviation (RMSD) was generated to determine the largest clusters. Finally, the samples were refined by CHARMM minimization. The docking server resulted in 30 models of a balanced set for energy coefficients. Afterward, 1 model with a low energy value in the models was selected.

Protein–protein interaction using ITC analysis

Previously reported methods for ITC analysis of microbial rhodopsin were followed with slight modifications [8,33]. Purified AbHeR WT, AbHeR mutants, and AbGS were completely exchanged with buffer SD using Amicon Ultra-4 10,000 MWCO centrifugal filter units and Amicon Ultra-4 30,000 MWCO centrifugal filter units, respectively. The concentrations of proteins were quantified using the Bradford assay. AbGS was continuously injected into the AbHeR at 25°C using a MicroCal ITC200 (Malvern Panalytical, United Kingdom) at the Advanced Bio-Interface Core Research Facility, Korea. ITC analysis data were evaluated using Microcal LLC ITC200 (Malvern Panalytical).

Preparation of IMVs

Preparation of IMVs, which are ISO membrane vesicles, was performed in accordance with previously published methods with slight modifications [42,73]. After induction, the E. coli pellets containing empty vector (pET21b), AbHeR, and GR were washed with buffer P (5 mM MgSO₄ and 50 mM potassium phosphate (pH 7.5)) instead of buffer S. The cells were homogenized to 20% (wet weight/volume) in buffer P containing 1 mM dithiothreitol (DTT) and 1 mM PMSF, and then disrupted using an EmulsiFlex-C3 high-pressure homogenizer (Avestin, Canada) with 3 passes at 10,000 psi at the Advanced Bio-Interface Core Research Facility. Unbroken cells and large debris were removed twice by centrifugation at 4°C and 39,000 × g for 15 min. IMVs were collected by ultracentrifugation at 4°C and 150,000 × g for 1.5 h. The collected IMVs were homogenized with buffer S and 100 mM NaCl, respectively, and subsequently ultracentrifuged at 4°C and 150,000 × g for 1.5 h. The washed IMVs were homogenized with buffer S and 100 mM NaCl, respectively, and then centrifuged at 10,000 × g for 5 min. The supernatants were frozen and stored in liquid nitrogen. Before use, the IMVs were thawed rapidly in a water bath at 37°C for 2 min.

Analysis of photochemical and biophysical properties of AbHeR

The protocols for estimating pKa values of the counterion, preparation of RSO membrane vesicles, and the light-induced H⁺ movement assay, have been described [8]. Absorption spectra of purified AbHeR were measured using a UV-VIS spectrophotometer (Shimadzu UV-2550, Japan). The difference of absorbance form of the counterion of rhodopsin was fitted using Origin 9.0 (OriginLab, USA), and the pKa value of the counterion was estimated using the Henderson–Hasselbalch equation. To perform the light-induced H⁺ movement assay, membrane vesicles were prepared, resuspended, and adjusted to OD₆₀₀ of 20 (2.0 × 10¹⁰ cells/mL) with 100 mM NaCl and a mixture solution (20 mM LiCl, NaCl, KCl, CsCl, and Na₂SO₄). The GR IMV was thawed and adjusted to OD₆₀₀ of 15 (1.2 × 10¹⁰ cells/mL) with 100 mM NaCl, followed by illumination at an intensity of 100 W/m² through a short-wave 500 nm cut off filter (Sigma Koki SCF-50S-44Y, Japan). The assay was performed with and without 10 μM CCCP, and the pH value was monitored using a pH meter (Horiba pH meter F-51. Japan) with a pH electrode bar (Horiba pH electrode bar 9618S-10D, Japan).

AbGS activity with AbHeR

AbGS activity was measured by colorimetric determination of P_i concentrations using ferric sulfate and molybdate, in accordance with previous studies [36–38]. The AbHeR IMVs were adjusted to an OD₅₅₄ of 0.330. The concentration of purified AbGS, as quantified by the Bradford assay, was adjusted to 160 μM with 50 mM imidazole-HCl (pH 7.0). First, 4 μL of each AbHeR and AbGS sample was allowed to bind at 260 rpm and 25°C for 1 h. After binding, 25 μL of an additive solution (72 mM imidazole-HCl at pH 7.0, 80 mM NH₄Cl, 12.8 mM MnCl₂, and 176 mM NaCl) and 5 μL of 400 mM glutamate-NaOH (pH 7.0) were added. Bound samples were incubated at 260 rpm and 25°C for 15 min. To activate enzyme activity, 2 μL of ATP solution (400 mM adenosine triphosphate (ATP) was dissolved in 1 M imidazole-HCl (pH 7.0), 150 mM NaCl, and 50 mM Tris) was added. The reaction was carried out at 37°C in the absence and presence of light (specific green laser, 532 nm) conditions at 20 μmol m^-2 s^-1. The reaction samples were treated with 40% trichloroacetate to be a final concentration of 20%. Treated samples were cooled at 4°C and centrifuged at 4°C and 3,500 × g for 30 min. Forty microliters of the supernatants containing P_i were transferred to a 96-well microplate, and 180 μL of freshly prepared reducing agent (0.8% FeSO₄ with 0.3N H₂SO₄) was added. The color reaction was developed for 15 min by adding 15 μL of a freshly prepared developing solution (6.6% (NH₄)₆Mo7O₂₄ with 7.5N H₂SO₄). Finally, concentrations of the developed reactants were measured at an OD₆₆₀ using a 2300 EnSpire Multimode Plate Reader (PerkinElmer, USA). The P_i concentrations were calculated using the P_i standard curve, which was prepared using the abovementioned procedure (colorimetric determination of the P_i assay) with different concentrations of potassium phosphate buffer without AbHeR and AbGS.

Growth rate test of cells for AbGS complementation

E. coli JW3841 (Keio collection), a glnA-deficient strain, was used to complement AbGS in our study [49]. The pKA001 vector contains the lac promoter and araBAD promoter with the β-dioxygenase gene. AbGS and AbHeR were regulated by P_lac and P_araBAD, respectively. E. coli JW3841 was transformed with the pKA001 null vector, pKA001-AbGS, and pKA001-AbGS_AbHeR. A single colony of transformed cells was incubated in LB medium containing 50 μg/mL ampicillin and 50 μg/mL kanamycin at 37°C and 280 rpm for 16 h. The incubated cells were transferred 100:1 ratio to LB medium containing 50 μg/mL ampicillin and 50 μg/mL kanamycin at 37°C and 280 rpm until the OD₆₀₀ reached 0.4. The cells were induced by the addition of 0.84 mM IPTG, 0.2% L-arabinose, and 7 μM ATR until the OD₆₀₀ reached 3.0. The induced cells were washed 3 times, resuspended with buffer S, and subsequently transferred to 8 mL of MOPS minimal medium [82] (40 mM 3-morpholinopropane-1-sulfonic acid, 4 mM tricine, 10 μM F₂SO₄, 9.52 mM NH₄Cl, 276 μM K₂SO₄, 500 nM CaCl₂, 528 μM MgCl₂, 50 mM NaCl, 1.32 mM K₂HPO₄, 3 nM (NH₄)₆Mo₇O₂₄, 400 nM H₃BO₃, 30 nM CoCl₂, 10 nM CuSO₄, 80 nM MnCl₂, and 10 nM ZnSO₄ (pH 7.4)) containing 0.4% carbon source (D-glucose or glycerol), 10 mM L-glutamate, 50 μg/mL ampicillin, and 50 μg/mL kanamycin to reach an OD₆₀₀ equal to 0.1 (1.0 × 10⁷ cells/mL). The transferred cells were incubated at 37°C and 280 rpm in the absence and presence of light (532 nm) at 40 μmol m^-2 s^-1. Further OD₆₀₀ measurements of E. coli were obtained using the UV-VIS spectrophotometer (Shimadzu UV-2550).

E. coli K-12 MG1655 was transformed with the AIO plasmid, and further experiments were performed in the same manner as described for E. coli JW3841.

Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and western blot

The E. coli JW3841 strain after induction of AbGS and AbHeR expression was centrifuged at 13,000 × g for 30 s and subsequently resuspended in Laemmli sample buffer. The samples were vigorously vortexed for 2 min and then lysed at 1,600 rpm for 20 min in a shaking incubator (Finepcr, Korea). The cell density in the buffer was equivalent to an OD₆₀₀ value of 3.0 (3.0 × 10⁸ cells/mL). Purified GS and AbHeR were added by Laemmli sample buffer, and then boiled for 10 min and incubated at 25°C for 30 min, respectively. Prepared cell lysates and purified proteins were subjected to discontinuous SDS-PAGE that the acrylamide concentrations of stacking and separating gels are 5% and 12%, respectively. After electrophoresis, the gels were stained with Coomassie brilliant blue R-250 and developed after transferring, blocking, as well as treatment with hexahistidine-tag antibody and HRP-conjugated antibody.

We would like to thank Drs. Keon Ah Lee, Ah reum Choi, and Se-Hwan Kim for their effort, support, and helpful discussion during the study.