Heliorhodopsin-mediated light-modulation of ABC transporter

Abstract

Heliorhodopsins (HeRs) have been hypothesized to have widespread functions. Recently, the functions for few HeRs have been revealed; however, the hypothetical functions remain largely unknown. Herein, we investigate light-modulation of heterodimeric multidrug resistance ATP-binding cassette transporters (OmrDE) mediated by Omithinimicrobium cerasi HeR. In this study, we classifiy genes flanking the HeR-encoding genes and identify highly conservative residues for protein-protein interactions. Our results reveal that the interaction between OcHeR and OmrDE shows positive cooperatively sequential binding through thermodynamic parameters. Moreover, light-induced OcHeR upregulates OmrDE drug transportation. Hence, the binding may be crucial to drug resistance in O. cerasi as it survives in a drug-containing habitat. Overall, we unveil a function of HeR as regulatory rhodopsin for multidrug resistance. Our findings suggest potential applications in optogenetic technology.

Fig. 1. HeR genes flanked by neighboring genes in the same operon.

Phylogenetic trees between the HeRs (total: 448 heliorhodopsin sequences) in organisms (a) and HeRs (79 heliorhodopsin sequences) (b). a Each group of organisms is indicated by a thin-colored donut curve. HeR roots in which at least two genes (helR with adjacent genes in the same operon) are predicted to be transcribed by a single promoter and are indicated by the same color belonging to the groups. The percentages of helR with adjacent genes per total helR are labeled next to each group. b Total helR (n = 448), genes adjacent to helR (n = 173), and frequent neighboring genes (n = 79) are indicated by white, gray, and dark gray pie donut curves. helR and adjacent genes in the same operon were analyzed. The adjacent genes were classified into ten groups. Each group is indicated by different colored texts and circles. c, e The amino acids of each HeR at the residue positions in OcHeR are counted and normalized by the number of each total HeR, that is, amino acid frequency per HeR. c Difference in amino acid frequency per HeR in the classified ten groups compared to non-co-transcription group (helR alone). Residues that increased and decreased by more than 0.19 are marked as upper and lower stack-bars, respectively. e Amino acid frequency per HeR in total HeRs (n = 448). d Blue and red surfaces in the electrostatic potential distribution of OcHeR present positively and negatively charged residues, respectively. Red and blue correspond to potentials of 20 kT e⁻¹ and −20 kT e⁻¹, respectively. The structure is viewed from membrane angles, and yellow dotted rectangles indicate ICLs. ICL intracellular loop, NA not aligned position, EC extracellular side, CP cytoplasmic side.

Fig. 2. HeRs in the group of ABCT-containing operons.

a Phylogenetic tree between the microbial rhodopsins and HeRs (68 rhodopsin sequences). Each function of rhodopsins is indicated using schematic ion-pumps and -channels in microbial rhodopsins and schematic regulatory function in HeRs. The HeRs belonging to the glutamine synthetase and photolyase groups are represented as orange and dark red circles, respectively. b OcHeR and other HeRs in the group of ABCT-containing operons are indicated using pink circles and stars, respectively. The direction of transcription of genes in the operons of the organisms is indicated by arrows. The predicted promoters are indicated using black bent arrows. Nucleotide gaps between the genes are labeled for each gene. EC extracellular side, CP cytoplasmic side, PPI protein‒protein interaction, hp hypothetical protein, duf domain unknown function.

Fig. 3. Functional analysis of OmrDE.

a, b Comparison protein structures between OmrDE (predicted by TmrAB as a template, PDB: 6RAF) and TmrAB (only TMDs and NBDs are indicated). The protein structures are colored: OmrD, cyan; OmrE, yellow; TmrA, orange; TmrB, navy. b The functional important motifs in OmrDE and TmrAB are indicated by black arrows with labels. Walker A, ABC signature, and D-loop are grouped with irregular shapes indicated as dotted lines. c Purified membrane proteins subjected to SDS-PAGE. d Western blotting after the His-Tag pull-down assay. The molecular weights of the purified membrane protein containing hexahistidine (His) and trihemagglutinin (HA) tags are calculated using protein sequences: OmrD-His, 62.86 kDa; OmrE-His, 72.09 kDa; OmrE-HA, 75 kDa. c, d These experiments were performed twice, and the representative data are shown. e MIC tests for E. coli harboring plasmids treated with various drugs, as shown in Supplementary Fig. 10a‒f, Supplementary Table 1. Measurements were conducted in an independent experimental group (n = 3), in which the data were presented as mean value ± SD. Cm, chloramphenicol; Km, kanamycin; Tet, tetracycline; DAPI, 4′,6-diamidino-2-phenylindole; HO342, Hoechst 33342. f DAPI translocation in IMVs containing no protein or membrane proteins was initiated by adding Mg-ATP, followed by measuring DAPI fluorescence; measurements were conducted in an independent experimental group (n = 6). g ATP hydrolysis of the NBDs of purified membrane proteins was performed using colorimetric determination of P_i. Non-linear fits were evaluated using the Michaelis–Menten equation, and measurements were conducted in an independent experimental group (n = 3), in which the data were presented as mean value ± SD. h Suggested model of OcHeR function associated with OmrDE. Membrane proteins are embedded in the membrane. Marker, protein marker: Null, empty vector; CP cytoplasmic side, EC extracellular side.

Fig. 4. Crucial binding position of OcHeR for OmrDE.

a Binding affinities of OcHeR for OmrD^tcE^tc, OmrD, OmrE, and OmrDE from ITC analysis in Table 2 and Supplementary Figs. 12 and 13. The K_d values are estimated using a one site binding model. ITC analysis of OcHeR R233Q for OmrDE is evaluated using a sequential binding site model: R233Q-1, first binding; R233Q-2, second binding. Measurements were conducted in an independent experimental group (n = 2 to 5). Data are presented as mean value ± SEM. b, c Protein–protein docking simulation between modeled OcHeR homodimer based on PDB code: 6SU4 and modeled OmrDE heterodimer based on PDB code: 6RAF. OcHeR, OmrD, and OmrE structures are shown in pink, cyan, and yellow, respectively. b Residues at a distance of 5 Å from the interface between OcHeR and OmrDE are shown. c OcHeR homodimer is expressed in two subunits (monomer). The distance (2.0–2.2 Å) between the hydrogen bonds of the interacting residues of OcHeR and OmrDE was determined using the polar interaction tool; it is indicated by yellow dotted lines and marked by yellow arrows. The positions of residues in OcHeR and OmrDE are indicated in pink and black text, respectively. d Scheme of the crystal structure of OcHeR is shown from the top angle, with each subunit of OcHeR separated by dotted lines. Each mutation position is indicated by text with non-binding and fold-changes of the K_d values of OcHeR mutants for OmrDE compared with that of OcHeR WT for OmrDE. ICL intracellular loop, ND not detected, NB non-binding.

Fig. 5. OcHeR-mediated specific light-modulation of OmrDE.

a, b ATP hydrolysis of purified OmrDE with OcHeR using the colorimetric determination of P_i. a Non-linear fits were evaluated using the Michaelis–Menten equation, and the measurements were conducted in an independent experimental group (n = 3). b Specificity constants (k_cat/K_m) are represented as bars. c, d Drug translocation of OmrDE and OcHeR co-expressed in IMVs based on the fluorescence intensity in the absence and presence of light. c The drug translocation assay was recorded at different time points; the relative fluorescence units (RFUs) were normalized to the initial recording, performed in an independent experimental group (n = 5). The significance between the two groups (WT in the absence and presence of light) was analyzed using the two-tailed Student’s t-test. d The delta-normalized RFU means that the data in (c) at 15 min were subtracted from the initial recording, which was analyzed using a one-way repeated-measures ANOVA with the post-hoc Dunnett test. The box plots information: minima and maxima, dash marks; centre, cross mark; whisker range, standard deviation; percentile range, 25% (Q1) and 75% (Q3). e Non-liner fits were averaged from data in Supplementary Fig. 10h‒m, where the x-axis modified logarithmic number to number. Dotted and solid lines indicate the absence and presence of light, respectively. f The IC₅₀ value for DAPI of data, obtained data in Supplementary Fig. 7h‒m, is estimated using the Dose–Response equation in an independent experimental group (n = 3). Gray and colored bars indicate the absence and presence of light, respectively. g Photocycles of OcHeR without and with OmrDE were recorded at 554 [G (ground) state] and 618 nm (O state); half-life (t_1/2) values through data analysis were estimated using exponential decay equation as non-linear fitted lines of data in Supplementary Fig. 6i. h Schematic model of a non-essential activator. Data information: In (a‒d, f, g), data are presented as mean value ± SD. In (c, d), p values and no significance are indicated with asterisks and labeled “ns”, respectively. Exact p values are indicated in the asterisks.

Fig. 6. Positive cooperatively sequential binding mechanism.

a Thermodynamic parameters of OcHeR for OmrDE from ITC analysis in Table 2 and Supplementary Figs. 12 and 13. ITC analysis of OcHeR R233Q for OmrDE was performed using a sequential binding site model: R233Q-1, first binding; R233Q-2, second binding. Measurements were conducted in an independent experimental group (n = 2 to 5). data are presented as mean value ± SEM. b This mechanism is assumed by OcHeR with IF_narrow conformation of OmrDE in the absence of light, which is based on thermodynamic parameters in ITC analysis. Protein structures of OcHeR, OmrD, and OmrE are indicated in pink, cyan, and yellow, respectively. Attractants and repellents are indicated by red and blue arrows, respectively.