SosA inhibits cell division in Staphylococcus aureus in response to DNA damage

Abstract

Inhibition of cell division is critical for viability under DNA-damaging conditions. DNA damage induces the SOS response that in bacteria inhibits cell division while repairs are being made. In coccoids, such as the human pathogen, Staphylococcus aureus, this process remains poorly studied. Here, we identify SosA as the staphylococcal SOS-induced cell division inhibitor. Overproduction of SosA inhibits cell division, while sosA inactivation sensitizes cells to genotoxic stress. SosA is a small, predicted membrane protein with an extracellular C-terminal domain in which point mutation of residues that are conserved in staphylococci and major truncations abolished the inhibitory activity. In contrast, a minor truncation led to SosA accumulation and a strong cell division inhibitory activity, phenotypically similar to expression of wild-type SosA in a CtpA membrane protease mutant. This suggests that the extracellular C-terminus of SosA is required both for cell division inhibition and for turnover of the protein. Microscopy analysis revealed that SosA halts cell division and synchronizes the cell population at a point where division proteins such as FtsZ and EzrA are localized at midcell, and the septum formation is initiated but unable to progress to closure. Thus, our findings show that SosA is central in cell division regulation in staphylococci.

Figure 1

Gram‐positive SOS‐controlled cell division inhibitors. A. Schematic representation of genes encoding characterized Gram‐positive SOS‐regulated cell division inhibitors (not drawn to scale) including the uncharacterized sosA from S. aureus. Despite considerable sequence divergence, these genes are commonly chromosomally co‐localized with lexA homologous genes. The cell division inhibitors carry a single transmembrane domain (TM), and several proteins have an additional LysM domain. B. Alignment (CLUSTAL O[1.2.4]) of SosA sequences deduced from open reading frames next to lexA in S. aureus strain 8325‐4 (YP_499864) and seven Staphylococcus species: S. carnosus (CAL27889), S. simulans (AMG96201), S. arlettae (EJY94737), S. haemolyticus (YP_253482), S. lugdunensis (YP_003471776), S. epidermidis (YP_188489) and S. warneri (EEQ79882). The proteins are 77 amino acids long and are characterized by a predicted transmembrane segment at AAs 10–30 (for S. aureus SosA) and a predicted extracellular C‐terminal (TOPCONS [Tsirigos et al., 2015]) with considerable sequence conservation at the membrane‐proximal portion (‘*’ indicates fully conserved residues, ‘:’ indicates conservation of residues with highly similar properties).

Figure 2

SosA supports survival of S. aureus subjected to lethal DNA damage and is involved in bacterial swelling. A. Culture optical density at 600 nm and cell viability of S. aureus strains 8325‐4 and JE2 in comparison with their respective ΔsosA mutants upon challenge with a lethal dose of mitomycin C (MMC, 1 µg ml⁻¹) for 2 h. Error bars represent the standard deviation from three biological replicates. B. Cell size of 8325‐4 and JE2 WT and ΔsosA mutants exposed to mitomycin C estimated by flow cytometry (FSC‐A). Cells were grown exponentially prior to MMC addition at an OD₆₀₀ of 0.05. Samples were taken after 0 (black), 40 (brown), 80 (red) and 120 (blue) min of incubation with MMC. C. Effect of MMC treatment (0.04 µg ml⁻¹) on cell shape and cell number of JE2 WT and JE2ΔsosA as visualized by time‐lapse phase‐contrast microscopy. Scale bar represents 2 µm.

Figure 3

Expression of sosA alone interferes with S. aureus growth. A. The effect of a controlled expression of sosA on the ability to form colonies was assessed in S. aureus RN4220. Strains carrying either the vector control (vector) or sosA under an anhydrotetracycline (AHT)‐inducible promoter (pSosA) were grown exponentially to an OD₆₀₀ of 0.5, serially 10‐fold diluted and plated on TSA plates in the presence or absence of TSA + 300 ng ml⁻¹ of AHT. The plates were incubated overnight at 37°C and imaged. B. Evaluation of cell size distribution by flow cytometry (FSC‐A) of S. aureus RN4220 containing the control vector (black) or pSosA (red). Cells were grown exponentially prior to induction with 100 ng ml⁻¹ of AHT. At the indicated time points, the cells were collected and analyzed by flow cytometry. C. Visualization of cell size increase and reduction of cell number of JE2/pRAB12‐lacZ (control) and JE2/pRAB12‐sosA in the presence of 200 ng ml⁻¹ of AHT by time‐lapse phase‐contrast microscopy at 37°C. Scale bar represents 2 µm.

Figure 4

Evaluation of membrane localization of SosA and truncated variants SosAd10, SosAd10(44A) and SosAd40 by in frame fusion at the C‐terminus to PhoA‐LacZ in pKTop (the truncated variants are explained in Figs 5 and 6). E. coli IM08B cells carrying the constructs were streaked on a dual‐indicator plate containing LB agar plus 50 µg⁻¹ of kanamycin, 1 mM IPTG, 5‐bromo‐4‐chloro‐3‐indolyl phosphate disodium salt (80 µg ml⁻¹) and 6‐Chloro‐3‐indolyl‐β‐D‐galactopyranoside (100 µg ml⁻¹). Cytoplasmic localization of the PhoA‐LacZ chimera is indicated by red/rose color development whereas translocation across the membrane is indicated by blue color development.

Figure 5

The effect of C‐terminal truncations of SosA on the inhibitory activity of the protein. A. Schematic representation of the different truncated SosA constructs. Full‐length SosA is a 77‐amino acid peptide. SosAd10 lacks the extreme C‐terminal 10 amino acids, while SosAd40 is SosA truncated of almost its entire extracellular C‐terminal part. Indicated conserved residues (*) originate from the alignment in Fig. 1. B. Activity of the constructs was assessed in S. aureus RN4220 and compared to the vector control (vector). Cells were grown exponentially to an OD₆₀₀ of 0.5, serially 10‐fold diluted and plated on TSA (control) or TSA plus inducer (AHT) at indicated increasing concentrations followed by incubation overnight at 37°C. C. Visualization of the drastic cell size increase and the reduction in cell number of JE2/pRAB12‐lacZ (control) and JE2/pRAB12‐sosAd10 in presence of 200 ng ml⁻¹ of AHT by time‐lapse phase‐contrast microscopy at 37°C. Scale bar represents 2 µm.

Figure 6

A. Assessment of the activity of SosAd10 variants with point mutations (alanine substitutions) at conserved residues (37/38, 40/41 and 44/45) within the C‐terminal part. Activity of the constructs was assessed in S. aureus JE2 and compared to the vector control (‐), SosA and SosAd10. Cells were grown exponentially to an OD₆₀₀ of 0.5, serially 10‐fold diluted and plated on TSA (control) or TSA plus inducer (AHT) at increasing concentrations followed by incubation overnight at 37°C. B. Comparison of the activity of the point mutation protein SosA (44A) with WT SosA in S. aureus JE2. Cells were grown as above and plated on TSA or TSA plus 200 ng ml⁻¹ AHT.

Figure 7

CtpA is a possible negative regulator of SosA. A. Hypersusceptibility of an S. aureus JE2 ctpA mutant to SosA‐mediated growth inhibition. Expression of sosA (pSosA) in WT S. aureus JE2 and the corresponding JE2‐ctpA mutant (ctpA) were compared and referenced to the vector control (vector) and expression of the hyperactive SosAd10 variant (pSosAd10) in the WT. Cells were grown exponentially to an OD₆₀₀ of 0.5, serially 10‐fold diluted and plated on TSA (Control) or TSA plus inducer (AHT) at indicated concentrations. The plates were incubated overnight at 37°C and imaged. B. Assessment of mitomycin C susceptibility of a ctpA mutant by comparison of plating efficiency of S. aureus 8325‐4 and JE2 WT, ctpA and ΔsosA‐ctpA in presence of 0.5 µg ml⁻¹ MMC. Cells were grown exponentially to an OD₆₀₀ of 0.5 and serially 10‐fold diluted before plating and incubation at 37°C overnight. C. Western blot of accumulated SosA in S. aureus 8325‐4 WT and the ctpA mutant at indicated time points after the addition of 1 µg ml⁻¹ of MMC to exponentially growing cells. The full blot is displayed in Fig. S5 and may serve as a control for equal loading and transfer by comparing intensities of nonspecific high molecular weight bands.

Figure 8

SosA does not impair the localization of cell division proteins and septum formation initiation. A. Localization of FtsZ‐eYFP in SJF4694 (JE2 pSosA pCQ11‐FtsZ‐eYFP) grown in the absence and presence of 100 ng ml⁻¹ of AHT for 45 min. B. EzrA‐GFP localization in SJF4697 (JE2 pSosA ezrA‐gfp+) grown in the absence or presence of 100 ng ml⁻¹ AHT for 45 min. C. Percentages of JE2 pSosA cells exhibiting incomplete, complete, complete split or incomplete split septa (n = 573 for no AHT, n = 607 for 100 ng ml⁻¹ of AHT; organisms either isolated or co‐adhered in pairs; other represents staining in indistinct shape) after incubation with or without AHT for 45 min. D. GpsB‐GFP localization in SJF700 (JE2 pSosA gpsB‐gfp+) grown in the absence or presence of 100 ng ml⁻¹ of AHT for 45 min. All fluorescence images (A, B and D) are average intensity projections and scale bars represent 3 um. E. 3D‐SIM Z‐stack images of JE2/pSosA grown with 100 ng ml⁻¹ of AHT for 45 min and labeled with Alexa Fluor 647 NHS ester. (i) A cell with initiated septum formation and (ii) a cell splitting into daughter cells without finishing a septal disc. Scale bars represents 0.5 µm.

Figure 9

Proposed model for a regulated survival strategy for S. aureus upon DNA damage. As part of the SOS response, SosA is produced and, by a membrane‐localized activity, corrupts normal cell division activity, likely via an interaction with essential divisome components. In effect, cells are still able to initiate septum formation but are unable to complete it, leading to synchronization of the cell population. This provides a spatio‐temporal window for paralleled SOS response‐regulated DNA repair activity. At the same time, the cell size increases due to off‐septal activity of peptidoglycan‐synthesizing enzymes. As the SOS response diminishes, the cellular concentration of SosA is lowered, directly or indirectly, by the proteolytic activity of CtpA. At this stage, septa are allowed to complete and normal growth/division continues.