2.1 Bacterial strains and growth

Bacterial strains used in this study are listed in Table 1. Molecular cloning and plasmid propagation were performed using E. coli DC10B cells grown in Luria-Bertani (LB) broth (Sigma-Aldrich). Staphylococcus aureus RN4220 strains were grown in tryptic soy broth (TSB; Becton Dickinson Bacto™), minimal media (SSM9PR: 1 × M9 salts [6.8 g/L Na₂HPO₄, 3 g/L KH₂PO₄, 0.5 g/L NaCl, 1 g/L NH₄Cl], 2 mM MgSO₄, 0.1 mM CaCl₂, 1% (w/v) glucose, 1% (w/v) casamino acids, 1.5 µM thiamine, 0.05 mM nicotinamide), or on brain heart infusion (BHI) agar (Becton Dickinson Difco™). Liquid cultures were routinely grown at 37°C with aeration, unless otherwise specified.

Plasmids were introduced into chemically competent E. coli DC10B or BTH101 cells using heat shock (Sambrook & Russell, 2001). Plasmids isolated from E. coli DC10B strains were introduced into S. aureus RN4220 cells via electroporation (Löfblom, Kronqvist, Uhlén, Ståhl, & Wernérus, 2007; Monk, Shah, Xu, Tan, & Foster, 2012). Where applicable, antibiotics were used at the following concentrations: 100 µg/ml ampicillin, 50 µg/ml kanamycin, 10 µg/ml erythromycin, and 15 µg/ml neomycin. Details on the construction of strains used in this study are provided in Appendix 1.

2.2 Recombinant DNA techniques

Plasmids used in this study are listed in Table 1. Oligonucleotides were synthesized by Integrated DNA Technologies and are listed in Table A1. DNA modification enzymes were purchased from New England Biolabs.

Plasmids were constructed using standard molecular cloning techniques (Sambrook & Russell, 2001) with T4 DNA ligase or Gibson assembly (Gibson et al., 2009) using Q5 high-fidelity DNA polymerase, T5 exonuclease, and Taq DNA ligase. Plasmid DNA was isolated from E. coli DC10B cells using the QIAprep Spin Miniprep kit (Qiagen). Staphylococcus aureus genomic DNA was isolated using the Purelink Genomic DNA Mini kit (Invitrogen), with the addition of 50 µg/ml lysostaphin (Sigma-Aldrich) for cell lysis at 37°C for 30 min. DNA purification was performed using the NucleoSpin Gel and PCR Clean-up kit (Macherey-Nagel). Details on the construction of plasmids used in this study are provided in Appendix 1. Plasmid sequences were verified by DNA sequencing (Fasmac).

2.3 Bacterial growth assays

Overnight cultures of strains to be assayed were diluted 1:50 in fresh media (SSM9PR minimal media or TSB rich media) and grown at 37°C until OD_600nm ~ 0.05. Cultures were aliquoted in triplicate into separate wells of a 96-well microplate. Optical density was measured every 20 min at 595 nm using a MultiSkan Go microplate reader (Thermo Fisher Scientific) set at 37°C with moderate shaking. Experiments were performed in triplicate for at least three independent experiments.

2.4 Bacterial viability assays

Overnight cultures were diluted 1:50 in fresh media and grown to mid-exponential phase. Cultures were normalized to OD_600nm 1.0 and serially diluted in phosphate-buffered saline (PBS), after which 10 µl was spotted onto SSM9PR minimal media agar or BHI agar plates. Plates were incubated overnight at 37°C. Data were obtained from at least three independent experiments.

2.5 Bacterial two-hybrid assays

Escherichia coli adenylate cyclase, cya, mutant cells (BTH101) were cotransformed with pairs of pKNT25 and pUT18 plasmid derivatives. Five single colonies of each strain were patched onto selective LB agar plates containing 100 µg/ml 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-Gal; Sigma-Aldrich) and 0.5 mM isopropyl β-D-1-thiogalactopyranoside (IPTG; Sigma-Aldrich). Plates were incubated at 30°C for 3 days. Blue patches indicate β-galactosidase activity from interaction between T25 and T18 fusions. Data shown are representative of three independent assays.

2.6 Fluorescence microscopy

Overnight cultures of S. aureus cells were diluted 1:50 in fresh TSB with appropriate antibiotics and incubated at 37°C with aeration until mid-exponential phase (OD_600nm ~ 0.4–0.6). For fluorescence localization of Spo0J-mRFPmars and SMC-GFP, cultures were grown at 30°C until OD_600nm ~ 0.1–0.2. Expression of Spo0J-mRFPmars and SMC-GFP was induced for 2 hr with 2.5 ng/ml anhydrotetracycline and 0.1 mM IPTG, respectively. Where required, DNA was stained with 2.5 µg/ml 4′,6-diamidino-2-phenylindole (DAPI), cell membranes were stained with 5 µg/ml FM4-64 (Invitrogen), and cell walls were stained with an equal mixture of vancomycin and a vancomycin BODIPY FL conjugate (0.4 µg/ml, Invitrogen) for 2 min at room temperature. After washing with PBS, a small aliquot of cells was applied to an agarose pad containing 2% (w/v) agarose in TSB on a microscope slide, and then covered with a glass coverslip for imaging.

Super-resolution SIM imaging was performed using a Zeiss ELYRA PS.1 microscope equipped with a 100 × 1.46 NA alpha plan apochromat oil immersion objective and a pco.edge sCMOS camera. Fluorescence images were acquired sequentially using 200–300 ms exposure times per image, for a total of 15 images per SIM reconstruction. All imaging was performed at room temperature (~23°C). Raw data were reconstructed using the SIM algorithms in ZEN 2011 SP7 software (black edition, Carl Zeiss). Brightfield images were captured using widefield imaging mode. Images had a final pixel size of 25 nm × 25 nm.

Images were processed and analyzed using Fiji (Schindelin et al., 2012). Cell diameters were determined by measuring the distance between FM4-64 fluorescence peaks from line profiles drawn along the short axis of cells. Data were obtained from at least three independent experiments.

2.7 Western blot and immunodetection

Staphylococcus aureus cultures were grown using the growth and induction conditions described for fluorescence microscopy. Following induction for 2 hr, cells were harvested (1,800 g, 4°C, 10 min) and then resuspended 1:200 in lysis buffer (0.1 M NaCl, 0.1 M Tris-HCl, 0.01 M MgCl₂, pH 7.5) containing 1 × protease inhibitor cocktail (Roche) and 0.1 mg/ml DNaseI (Roche). Cells were mixed with 212–300 μm acid-washed glass beads (Sigma-Aldrich) and lysed at 50 Hz for 6 min using a TissueLyser (Qiagen) with precooled tube adapter. Lysates were cleared by centrifugation (21,100 g, 4°C, 30 min) and then incubated with reducing sample buffer at 95°C for 5 min. Samples were electrophoresed on 4%–15% Tris-glycine polyacrylamide gradient gels and then transferred to 0.45 µm nitrocellulose membranes in Towbin buffer (25 mM Tris, 192 mM glycine, 20% (v/v) methanol, pH 8.3).

Blots were blocked with 5% (w/v) EasyBlocker (GeneTex) in TBST (25 mM Tris, 0.15 M NaCl, 0.05% Tween-20). Spo0J-mRFPmars was detected with a 1:2,000 dilution of anti-RFP or anti-mCherry antibodies (Abcam), and SMC-GFP and GFP were detected with a 1:3,000 dilution of anti-GFP antibodies (Abcam). Primary antibodies were detected with a 1:3,000 dilution of goat anti-rabbit HRP-conjugated antibodies (Bio-Rad). Antibodies were diluted in blocking buffer and incubated with blots for 1 hr at room temperature. Blots were washed three times with TSBT before and after blocking and antibody incubations. Secondary antibodies were detected using Clarity Western ECL substrate (Bio-Rad). Blots were imaged using a ChemiDoc gel imager (Bio-Rad).

2.8 Immunoprecipitation

Immunoprecipitation experiments were performed on S. aureus cell lysates prepared as described for Western blotting. Cell lysates (~1.5 mg total protein) from S. aureus cells expressing Spo0J-mRFPmars (pHC052) and either GFP (pHC102) or SMC-GFP (pHC067) were incubated with pre-equilibrated GFP-Trap magnetic agarose beads (ChromoTek) for 1 hr at 4°C with gentle inversion. Beads were magnetically separated and washed twice with wash buffer (10 mM Tris-HCl, 150 mM NaCl, 0.5 mM EDTA, pH 7.5) at 4°C, before elution of bound complexes with elution buffer (50 mM Tris-HCl, 10 mM EDTA, 1% SDS, pH 8) at 65°C for 15 min. Eluates were magnetically separated from the beads and analyzed by Western blotting and immunodetection.

2.9 Statistical analyses

Analyses of statistical significance were performed using unpaired Student's t tests. p values <.05 were considered statistically significant and are indicated with asterisks.

3.1 spo0J and smc are not essential for Staphylococcus aureus growth and viability

There is much variation in the functional significance of ParB and SMC in different bacteria. For example, parB is essential in Caulobacter crescentus, while its homologue spo0J is not essential in B. subtilis nor S. pneumoniae (Ireton, Gunther, & Grossman, 1994; Minnen et al., 2011; Mohl, Easter, & Gober, 2001). Additionally, smc is not essential in S. pneumoniae and C. crescentus, but deletion causes lethality under conditions promoting fast growth in B. subtilis (Gruber et al., 2014; Jensen & Shapiro, 1999; Minnen et al., 2011). In order to determine whether spo0J and smc are essential in S. aureus, we generated separate spo0J and smc deletions in the S. aureus RN4220 parental strain. Strains were constructed using an integrative pMAD derivative followed by double-crossover homologous recombination to generate markerless deletions at the spo0J or smc loci of the S. aureus RN4220 genome (Appendix 1). Furthermore, because spo0J and smc were shown to interact in B. subtilis and S. pneumoniae, we also constructed a Δspo0J Δsmc double mutant to test whether this interaction is essential in S. aureus.

The Δspo0J (HC061), Δsmc (HC080), and Δspo0J Δsmc (HC094) mutants showed similar growth rates to the RN4220 parental strain, with no difference in viability when grown at 37°C in TSB rich medium or SSM9PR minimal medium (Figure 1a–c). As described below, we showed that the lack of significant difference in growth and viability between the strains was not a result of suppressor mutations or polar effects from the gene deletions (Figure A1).

3.2 Spo0J and SMC are involved in Staphylococcus aureus chromosome segregation

The roles of Spo0J and SMC in chromosome segregation are well-studied in B. subtilis and S. pneumoniae. Perturbations to spo0J or smc result in defects in chromosome segregation, with a spo0J smc double mutant having a synthetic lethal phenotype in B. subtilis when grown in rich medium (Britton et al., 1998). Yu et al. (2010) showed that SMC is involved in S. aureus chromosome segregation, but we wondered whether spo0J also plays a role together with SMC. To this end, we performed SIM imaging to examine the DNA content of S. aureus RN4220, Δspo0J, Δsmc, and Δspo0J Δsmc derivatives stained with DAPI.

When grown at 37°C in TSB rich medium, 0.1% of the RN4220 parental strain contained cells that were anucleate, that is, without a nucleoid and, therefore, devoid of DAPI staining (Figure 2a,b). In comparison, 0.6% of Δspo0J and 1.1% of Δsmc cells were anucleate, which are both significantly greater than the frequency observed for the parental strain (p = .0241 and 0.0021, respectively) (Figure 2a,b), suggesting that spo0J and smc are required for normal chromosome segregation in S. aureus. Notably, the Δspo0J Δsmc double mutant showed a significantly higher proportion of anucleate cells compared to the parental strain and the Δspo0J and Δsmc single mutants (3.2% anucleate cells, p = .0017, 0.0032, and 0.0070, respectively) (Figure 2a,b), indicating an increased chromosome segregation defect in the absence of both spo0J and smc.

Defects in chromosome segregation are often associated with nucleoid abnormalities such as nucleoid condensation or bisection of unsegregated nucleoids by the division septum (Minnen et al., 2011; Veiga & G Pinho, 2017; Yu et al., 2010). However, in our study, we found no difference in the occurrence of nucleoid abnormalities between the parental and mutant strains and did not observe any instances of septum formation over unsegregated chromosomes.

3.3 Spo0J is required for SMC localization in Staphylococcus aureus

Our finding that Spo0J and SMC are both required together to ensure accurate chromosome segregation led us to speculate that S. aureus Spo0J and SMC might behave similarly to their counterparts in other bacteria. A significant role of ParB/Spo0J in S. pneumoniae, B. subtilis, and C. crescentus is to recruit the condensin, SMC, to origin-proximal sites, where ParB/Spo0J binds (Gruber & Errington, 2009; Minnen et al., 2011; Sullivan et al., 2009; Tran, Laub, & Le, 2017). To determine whether this interaction might also occur in S. aureus, we constructed Spo0J-mRFPmars and SMC-GFP fusion proteins so that their localizations could be visualized by SIM imaging. Both proteins were expressed from compatible, coresident, low-copy number plasmids in S. aureus. Spo0J-mRFPmars expression was induced with 2.5 ng/ml anhydrotetracycline from the P_xyl/tetO promoter, while SMC-GFP expression was induced with 0.1 mM IPTG from the P_spac promoter. Importantly, both fluorescent fusions were functional under the induction conditions tested, since Spo0J-mRFPmars and SMC-GFP could complement the chromosome segregation defects of the Δspo0J and Δsmc mutants, respectively (Figure A1). This further verifies that the growth, viability, and chromosome segregation phenotypes of the Δspo0J and Δsmc strains were the result of gene deletions and not other mutations or polar effects.

Spo0J-mRFPmars formed fluorescent foci consistent with previously shown localization patterns and indicative of DNA-binding to origin-proximal sites (Pinho & Errington, 2004; Veiga, Jorge, & Pinho, 2011) (Figure 3a and Figure A2). SMC-GFP displayed similar localization patterns to Spo0J-mRFPmars, often with overlapping, or partially-overlapping, foci (Figure 3a). Our results suggest that S. aureus Spo0J and SMC likely localize to similar subcellular positions, consistent with potential complex formation.

If S. aureus Spo0J recruits SMC to origin-proximal sites as described in B. subtilis, S. pneumoniae, and C. crescentus (Gruber & Errington, 2009; Minnen et al., 2011; Sullivan et al., 2009; Tran et al., 2017), then we predicted that SMC-GFP might be mis-localized in the absence of Spo0J. Therefore, we endeavored to determine whether SMC-GFP localization is dependent on Spo0J. For this purpose, we expressed SMC-GFP in the S. aureus Δspo0J mutant strain and compared its localization to that in the parental RN4220 strain. Subsequent SIM imaging revealed that more than 90% of cells in the parental strain contained at least one fluorescent focus of SMC-GFP, with the majority (62%) containing two SMC-GFP foci per cell (Figure 3b,c). In contrast, SMC-GFP localization appeared more dispersed in the Δspo0J cells, with a significantly higher proportion of cells containing no SMC-GFP foci (50%, p < .0001), and an average of 0.96 foci per cell, compared to 2.1 foci in the parental strain (p < .0001, n_total = 1,345 and 1,307 cells, respectively, Figure 3b,c). Western blotting using anti-GFP antibodies showed that the loss of SMC-GFP foci in the absence of spo0J was not due to the release of free GFP by proteolysis (Figure 3d). No proteolysis was also detected for Spo0J-mRFPmars (Figure 3d). Taken together, our results suggest that staphylococcal chromosome segregation involves the formation of complexes containing Spo0J and SMC and that Spo0J is required for correct localization of SMC in S. aureus.

3.4 Spo0J and SMC form a complex in Staphylococcus aureus to mediate chromosome segregation

The increased proportion of anucleate cells in the Δspo0J Δsmc mutant and the dependence of SMC localization on Spo0J suggested potential interaction between Spo0J and SMC. Therefore, in order to verify whether Spo0J and SMC form a complex in S. aureus, we performed co-immunoprecipitation experiments on S. aureus RN4220 cells expressing Spo0J-mRFPmars and either GFP or SMC-GFP. Following nuclease digestion of genomic DNA, we used anti-GFP antibodies coupled to magnetic agarose beads to immunoprecipitate GFP or SMC-GFP and their interacting proteins. We were able to detect Spo0J-mRFPmars in the immunoprecipitated complexes isolated from S. aureus cells expressing both Spo0J-mRFPmars and SMC-GFP, but not from cells expressing Spo0J-mRFPmars and GFP (Figure 4a). Thus, Spo0J-mRFPmars was isolated specifically in the presence of SMC-GFP, indicating the formation of complexes containing Spo0J and SMC in S. aureus. Notably, the Spo0J-mRFPmars immunoprecipitated from cell lysates containing SMC-GFP was only weakly detected using anti-RFP antibodies, which may be indicative of weak or transient complex formation.

To test whether Spo0J and SMC might interact directly, we conducted bacterial two-hybrid assays using Spo0J and SMC fusions to the Bordetella pertussis adenylate cyclase (CyaA) fragments, T25 and T18 (Karimova, Pidoux, Ullmann, & Ladant, 1998). The assays revealed that in the E. coli cya mutant strain BTH101, β-galactosidase activity was detected from cells expressing the Spo0J-T25 and SMC-T18 fusions (blue patches), however, no activity was detected from cells expressing the reciprocal fusions, SMC-T25 and Spo0J-T18 (white patches) (Figure 4b). We do not exclude the possibility that an interaction between Spo0J and SMC, direct or indirect, may be dependent on Spo0J localization or DNA binding, or on the presence of other S. aureus proteins that were not present in the bacterial two-hybrid assays. The assays also showed self-interaction of Spo0J and SMC (Figure 4b), which is consistent with the fluorescent Spo0J-mRFPmars and SMC-GFP foci observed (Figure 3a).

3.5 Absence of Spo0J and SMC does not interfere with normal Staphylococcus aureus cell division

Spatiotemporal coordination of chromosome segregation and cell division is paramount for successful cell growth. Impairment of parB in C. crescentus blocks FtsZ ring assembly and subsequent cell division (Mohl et al., 2001), while chromosome segregation defects in B. subtilis correlate with elongated cells (Britton et al., 1998; Ireton et al., 1994; Lee & Grossman, 2006). In order to investigate the influence of Spo0J and SMC on S. aureus cell division, we sought to determine whether the chromosome segregation defects observed in the spo0J and smc mutants correspond to abnormalities in staphylococcal cell division.

Disruption of staphylococcal cell division, for example, by impeding FtsZ, EzrA, Noc, or DnaK function, results in cell enlargement and aberrant cell division (Bottomley et al., 2017; Jorge, Hoiczyk, Gomes, & Pinho, 2011; Lund et al., 2018; Monteiro et al., 2018; Pereira et al., 2016; Pinho & Errington, 2003; Veiga et al., 2011). Therefore, as an indicator of cell division, we measured the cell diameters of S. aureus cells to determine whether differences in cell size could be attributed to inefficient chromosome segregation. We found no significant difference in cell diameters between the parental and deletion strains (Figure 5a and Figure A3), which suggests that the absence of spo0J or smc, alone or together, does not affect S. aureus cell size.

We next investigated whether spo0J and smc play a role in cell cycle progression. We categorized the S. aureus cell cycle into three phases, characterized by cells that had no septum (Phase 1), an incomplete septum (Phase 2) or a complete septum (Phase 3) (Monteiro et al., 2015). In order to provide an indication of the relative duration of each phase of the cell cycle, we stained S. aureus cell membranes with the lipid dye FM4-64 and calculated the proportion of cells in each phase from SIM images. We observed no difference in the distribution of cells in the three cell cycle phases for the parental, Δspo0J, and Δsmc strains (Figure 5b). We did, however, observe a modest, but insignificant, increase in the proportion of cells in Phase 2 for the Δspo0J Δsmc double mutant (54 ± 4.3% compared to 42 ± 3.6% for the parental strain, mean ± SEM, p = .1032) (Figure 5b), potentially indicating slightly prolonged septum closure in these cells. Taken together, these results suggest that spo0J and smc do not significantly affect cell cycle progression and are not essential for normal cell division in S. aureus.