Fine-Tuning Modulation of Oxidation-Mediated Posttranslational Control of Bradyrhizobium diazoefficiens FixK2 Transcription Factor

Abstract

FixK₂ is a CRP/FNR-type transcription factor that plays a central role in a sophisticated regulatory network for the anoxic, microoxic and symbiotic lifestyles of the soybean endosymbiont Bradyrhizobium diazoefficiens. Aside from the balanced expression of the fixK₂ gene under microoxic conditions (induced by the two-component regulatory system FixLJ and negatively auto-repressed), FixK₂ activity is posttranslationally controlled by proteolysis, and by the oxidation of a singular cysteine residue (C183) near its DNA-binding domain. To simulate the permanent oxidation of FixK₂, we replaced C183 for aspartic acid. Purified C183D FixK₂ protein showed both low DNA binding and in vitro transcriptional activation from the promoter of the fixNOQP operon, required for respiration under symbiosis. However, in a B. diazoefficiens strain coding for C183D FixK₂, expression of a fixNOQP'-'lacZ fusion was similar to that in the wild type, when both strains were grown microoxically. The C183D FixK₂ encoding strain also showed a wild-type phenotype in symbiosis with soybeans, and increased fixK₂ gene expression levels and FixK₂ protein abundance in cells. These two latter observations, together with the global transcriptional profile of the microoxically cultured C183D FixK₂ encoding strain, suggest the existence of a finely tuned regulatory strategy to counterbalance the oxidation-mediated inactivation of FixK₂ in vivo.

Keywords: CRP/FNR proteins; in vitro transcription; microarrays; microoxia; protein–DNA interaction; rhizobia; symbiosis.

Figure 1

IVT activation from the fixNOQP promoter mediated by different FixK₂ protein derivatives. Plasmid pRJ8816 harboring the fixNOQP promoter cloned upstream of the B. diazoefficiens rrn transcriptional terminator was employed as template for multiple-round IVT activation assays with B. diazoefficiens RNAP holoenzyme. A series of concentrations of FixK₂ protein variants were added to the reactions: lane 1, no protein (-); lanes 2, 5 and 8, 0.5 µM; lanes 3, 6 and 9, 1.25 µM; lanes 4, 7 and 10, 2.5 µM. The positions of the fixNOQP transcript and the FixK₂-independent transcript (used as control for the experiments) are depicted on the right. Each panel refers to different sections of the same gel. Shown are the results of a typical experiment that was performed at least twice. nt, nucleotides.

Figure 2

Comparative SEC of native FixK₂ and C183S and C183D FixK₂ variants at different protein concentrations. Elution profiles were monitored at 280 nm following chromatography of FixK₂ loaded at 2.5, 5, 10, 20 and 30 µM (native, upper panel); C183S FixK₂ at 2.5, 5, 10, 20 and 40 µM (C183S, middle panel); and C183D FixK₂ at 2.5, 5, 10, 20 and 40 µM (C183D, bottom panel). The dashed lines show the calculated elution volume for the theoretical M_w of the monomeric (~26 kDa) and dimeric forms (~52 kDa).

Figure 3

In vitro interaction of C183S and C183D FixK₂ derivatives with the fixNOQP promoter tested by EMSA (A) and surface plasmon resonance (SPR) (B) approaches. (A) A 90-bp PCR fragment containing the FixK₂ box at 20 nM was incubated with increasing concentrations (0 to 12 µM) of FixK₂ protein variants, indicated at the top of each gel. Lower bands show free DNA, while upper bands correspond to the protein–DNA complexes. The molecular marker GeneRuler™ 1 Kb Plus DNA Ladder (Thermo Fisher Scientific, Waltham, MA, USA) is shown on the first lane. (B) A biotinylated double-stranded oligonucleotide containing the FixK₂ box from the fixNOQP promoter was immobilized on a streptavidin (SA) sensor chip by biotin–streptavidin binding. The sensorgrams with the relative resonance units (RU) of the interaction with DNA of C183S and C183D FixK₂ protein variants at 250 nM are shown. Data of the C183D FixK₂ protein did not allow us to calculate any kinetic/affinity parameters.

Figure 4

Expression data for a chromosomally integrated fixNOQP’-’lacZ fusion in different B. diazoefficiens backgrounds. Wild-type, C183D-fixK₂ and ΔfixK₂ strains were cultivated for 48 h microoxically (0.5% O₂). β-Galactosidase values are means ± standard errors of a representative experiment performed with two parallel cultures assayed in quadruples. The experiment was repeated at least twice. WT, wild type.

Figure 5

Denitrifying growth of the B. diazoefficiens C183D-fixK₂ strain (triangles). Wild type (WT, diamonds) and ΔfixK₂ (squares) were used as controls. Cells were grown anoxically with nitrate. Values ± standard errors are the mean of a representative experiment carried out with three parallel cultures. At least three replicates of the experiment were done.

Figure 6

Expression of fixK₂ at protein (A,B) and transcriptional (C) levels. Steady-state levels of FixK₂ protein in cells cultivated under microoxic free-living conditions (A) and in soybean bacteroids collected at 25 and 32 dpi (B). Immunodetection was performed with a polyclonal FixK₂ antibody [28]. (A) 60 μg of crude extract of wild-type (lane 1) and C183D-fixK₂ strains (lane 2) both cultivated microoxically (0.5% O₂). (B) 10 μL of soybean bacteroid crude extract of wild-type (lanes 1 and 3) and C183D-fixK₂ strains (lanes 2 and 4). Apparent molecular mass of FixK₂ is shown on the left. Representative results of at least three independent biological replicates are shown. (C) β-Galactosidase activity from a chromosomally integrated fixK₂′-’lacZ fusion in B. diazoefficiens wild-type, C183D-fixK₂ and ΔfixK₂ strains. Cells were cultivated for 48 h microoxically (0.5% O₂). Values are the means ± standard errors of a typical experiment performed with two parallel cultures assayed in quadruples. The experiment was repeated at least twice. WT, wild type.

Figure 7

Workflow of microarray data analyses of the C183D-fixK₂ strain. Labels of the comparisons between specific transcription profiles are depicted alongside the circles. The total number of differentially expressed genes is indicated in parentheses. Up/down arrows refer to increased and decreased gene expression. The group of genes with differential expression in the C183D-fixK₂ strain (dark grey circle, left) showed an overlap of 50 genes (light grey circle, middle) with those in ΔfixK₂ strain (white circle, right; [19]), both grown microoxically (0.5% O₂) and compared with the wild type grown in the same conditions. Within the overlap, 47 genes showed downregulated expression in both the C183D-fixK₂ and ΔfixK₂ strains, which includes 37 genes organized in mono- or polycistronic transcriptional units that harbor a putative FixK₂ box within the promoter region (26 putative transcriptional units, see Table 2).

Figure 8

Modeling of different FixK₂ protein variants with the double-stranded DNA containing the FixK₂ box present at the fixNOQP promoter. Shown are the protein–DNA distances between the negatively charged oxygens of the phosphate group of the nitrogenous bases adenine 6 and adenine 7 of the strand W of DNA [23], and cysteine (A), serine (B), aspartic acid (C), cysteine–sulfenic acid (D), cysteine–sulfinic acid (E) and cysteine–sulfonic acid (F) residues of FixK₂. The predictions of the 3D models of FixK₂ and C183D FixK₂ were obtained with the Pymol 2.2.3 program (https://pymol.org/2/; accessed on 13 October 2021), using the C183S FixK₂–DNA structure as a template (http://wwpdb.org/; code 4I2O; accessed on 7 June 2021). Visualization of molecular structures and interactions was performed using the Discovery Studio Visualizer program, version V20.1.0.19295 (BIOVIA, Waltham, MA, USA), which also allowed the modeling of sulfenic, sulfinic and sulfonic acid derivatives of FixK₂. Distances in angstroms (Å) are represented by dashed lines; adenine 6 on the left; adenine 7 on the right.