Redox sensing by a Rex-family repressor is involved in the regulation of anaerobic gene expression in Staphylococcus aureus

Abstract

An alignment of upstream regions of anaerobically induced genes in Staphylococcus aureus revealed the presence of an inverted repeat, corresponding to Rex binding sites in Streptomyces coelicolor. Gel shift experiments of selected upstream regions demonstrated that the redox-sensing regulator Rex of S. aureus binds to this inverted repeat. The binding sequence--TTGTGAAW(4)TTCACAA--is highly conserved in S. aureus. Rex binding to this sequence leads to the repression of genes located downstream. The binding activity of Rex is enhanced by NAD+ while NADH, which competes with NAD+ for Rex binding, decreases the activity of Rex. The impact of Rex on global protein synthesis and on the activity of fermentation pathways under aerobic and anaerobic conditions was analysed by using a rex-deficient strain. A direct regulatory effect of Rex on the expression of pathways that lead to anaerobic NAD+ regeneration, such as lactate, formate and ethanol formation, nitrate respiration, and ATP synthesis, is verified. Rex can be considered a central regulator of anaerobic metabolism in S. aureus. Since the activity of lactate dehydrogenase enables S. aureus to resist NO stress and thus the innate immune response, our data suggest that deactivation of Rex is a prerequisite for this phenomenon.

Fig. 1

Transcriptional organization of rex in S. aureus. Total RNA was isolated from SH1000 (w), SH1000 Δrex (r) and SH1000 ΔarcR (a) before (+O₂) and 30 min after imposition of anaerobic conditions (-O₂). Equal amounts (10 µg) of total RNA were used. The RNA was separated on formaldehyde containing agarose gels and blotted onto positively charged nylon membranes. The membranes were hybridized with the digoxygenin-labelled RNA probes of rex. Schematic representation of the gene locus based on the sequence of S. aureus NCTC8325 and the transcriptional organization of the predicted operons are shown (‘SAOUHSC_0’ was neglected in the locus tags). The gene used as a probe is marked by an asterisk (*). Methylene blue staining of the membranes is shown in Fig. S2.

Fig. 2

Verification of Rex binding sites by electrophoretic mobility shift assays (EMSAs). Binding of purified Rex to the promoter regions of different genes using EMSAs. The resulting protein–DNA complexes were separated from unbound DNA fragments using native polyacrylamide gels. The DNA fragments were visualized by ethidium bromide staining. Formation of stable Rex–DNA complexes resulted in one or more distinct shifted DNA bands. DNA fragments containing the putative Rex binding site with different Rex binding affinities are shown: (A) strong binding affinity indicated by a complete shift in the presence of 0.2 µM Rex protein; (B) low binding affinity indicated by an incomplete shift in the presence of 0.2 µM Rex protein; (C) no binding in the presence of 0.2 and 0.4 µM Rex protein. (D) As a negative control the upstream region of the anaerobically induced clpL gene which lacks a putative Rex binding site was used.

Fig. 3

Influence of NAD⁺ and NADH on the Rex affinity to several DNA motifs. A. EMSAs were performed with PCR products of the upstream regions of adhE, srrA, vicR, 5SrRNA, hemE and clpL after incubation with purified Rex protein in the presence of of NAD⁺. Formation of stable Rex–DNA complexes resulted in a distinct up-shifted DNA band. The respective inverted repeats of the DNA fragments used are shown. Inverted repeats are shown in bold capitals. Mismatches to the consensus sequence are underlined. B. EMSAs were performed with PCR products of the upstream regions of adhE, nirR, nirC and hemE incubated with 0.4 µM purified Rex protein and different concentrations of NAD⁺ and NADH. Subsequently, the PCR products were separated on native polyacrylamide gels and visualized by ethidium bromide staining.

Fig. 4

Binding affinity of Rex to NADH and NAD⁺. Isothermal titration calorimetry experiments titrating S. aureus Rex with NADH (A) and NAD⁺ (B). Raw data are shown in the top panels and in the lower panels it is fitted using a one-site model. A different scale for the y-axis is used in the inserted boxes.

Fig. 5

Characterization of the Rex consensus sequence in S. aureus. Mutants of the Rex binding site of the adhE upstream region were constructed by site-directed mutagenesis. PCR products were incubated with 0.4 or 0.1 µM purified Rex protein. PCR products were run through a native polyacrylamide gel and visualized by ethidium bromide staining. Formation of stable Rex–DNA complexes resulted in a distinct up-shifted DNA band. To illustrate the DNA shift caused by Rex binding, DNA fragments of the wild-type sequence were incubated in the presence (+Rex) and absence (-Rex) of purified Rex protein prior to separation on the gels. The wild-type sequence and the respective mutated binding sequences are shown below. Base pairs that correspond to the wild-type sequence are shown in capital letters. Mutated base pairs are shown in black boxes. Insertions and deletions are shown in grey boxes.

Fig. 6

Colocalization of Sigma A consensus sequences and verified Rex binding sites. A. Total RNA was isolated from SH1000 (w) and SH1000 Δrex (r) before and 30 min after a shift to anaerobic conditions. Equal amounts (10 µg) of total RNA were used. The transcription start sites are marked with arrows. Lanes A, C, G and T show the dideoxy sequencing ladders obtained with the equivalent primer. B. The nucleotide sequences around the respective promoter regions of Rex-regulated genes are shown. The transcriptional start points for ldh1, pflBA and adhE obtained by primer extension experiments are indicated by arrows. The potential −10 and −35 regions of SigA-dependent promoters in front of the transcriptional start points of these genes are framed.

Fig. 7

Transcriptional analysis of Rex-regulated genes. Total RNA was isolated from SH1000 (w), SH1000 Δrex (r) and SH1000 ΔarcR (a) before (+O₂) and 30 min after a shift to anaerobic conditions (-O₂). Ten micrograms of total RNA was used for Northern blot experiments of the genes indicated except for ldh1 and pflBA where 4 µg was used. The RNA was separated on formaldehyde containing agarose gels and blotted onto positively charged nylon membranes. The membranes were hybridized with digoxygenin-labelled RNA probes of the respective genes. Schematic representations of the gene loci based on the sequence of S. aureus NCTC8325 and the transcriptional organization of the predicted operons are shown (‘SAOUHSC_0’ was neglected in locus tags). The genes used as a probe are marked by an asterisk (*). Methylene blue staining of the membranes is shown in Fig. S3. Anaerobically induced genes can be divided according to their transcriptional pattern into the following classes: (A) anaerobic induction is solely controlled by Rex; (B) anaerobic induction is partially controlled by Rex; (C) anaerobic induction is not controlled by Rex.

Fig. 8

The influence of Rex on anaerobic protein synthesis in S. aureus. Cytoplasmic proteins of the wild-type and rex mutant were labelled with l-[³⁵S] methionine before and 30 and 60 min after a shift to anaerobic conditions and separated on 2D gels. The resulting 2D gel images were analysed using Delta2D software (DECODON, Greifswald). A. Fused 2D image with colour-coded Rex-regulated protein spots of S. aureus SH1000. Autoradiograms of the rex mutant and the wild-type in the presence and absence of oxygen were combined to generate a fused proteome map and thereby to illustrate the Rex modulon on the 2D image in the standard pH range of 4–7. To show the influence of Rex on the synthesis of proteins, induction ratios of the intensities of each protein spot of the mutant compared with the wild-type were calculated for aerobic as well as for anaerobic conditions. In this way, oxygen-dependent influences on the synthesis of each protein spot were determined. Based on their expression pattern and on the presence of a Rex binding site in front of the respective genes, anaerobically induced or repressed proteins regulated by Rex activity can be divided in the following classes: Class I – anaerobic induction is solely and directly controlled by Rex (red); Class II – anaerobic induction is directly but not solely controlled by Rex (yellow); Class III – synthesis of these proteins is indirectly repressed by Rex (blue); Class IV – synthesis of these proteins is indirectly activated by Rex (green). B. Details of dual channel images of selected proteins whose synthesis was influenced by Rex. The autoradiograms were normalized by using total normalization. The bar graphs on the right display relative synthesis rates (logarithm to the base 2) of the individual proteins at the different time points (red bars = increased synthesis rate, green bars = reduced synthesis rate). SrrA and Ldh1 belong to Class I, Adh1, AdhE and PflB belong to Class II, and GapA1, GapR and Pgk belong to Class IV.

Fig. 9

Impact of Rex on selected metabolic pathways. A. Selected proteomic and metabolomic data of the glycolysis and fermentation pathways. Protein synthesis rates (rex mutant/wild-type) of selected enzymes are shown in squares and ratios of selected metabolites (rex mutant/wild-type) are shown in circles for aerobic (white) and anaerobic (black) conditions. Proteins whose genes are directly Rex-regulated are underlined. Cells were grown in synthetic medium without MOPS buffer to an optical density of 0.5 and shifted to anaerobic conditions. Cytoplasmic proteins of the wild-type and rex mutant were labelled with l-[³⁵S]-methionine before and 60 min after a shift to anaerobic conditions and separated on 2D gels. The resulting 2D gel images were analysed using Delta2D software (DECODON, Greifswald). For metabolite analyses, samples were taken from the aerobically grown culture and from the anaerobically grown culture immediately before (only aerobic culture) and 1, 8 and 24 h after shifting to anaerobic conditions. Cells were separated from the supernatant by filtration and the supernatants obtained were used for further analyses. B. Extracellular metabolite analyses by ¹H-NMR. Glucose, pyruvate, formic acid, ethanol, acetate, lactate, acetoine and 2,3-butanediol were detected and quantified by ¹H-NMR. The graphs show the increase of the concentration in mM.