Staphylococcus aureus USA300, its lpl deletion mutant USA300Δlpl and the complemented mutant USA300Δlpl (pTX30-lpl) are constructed as previously described (Nguyen et al., 2015). All S. aureus cultures were performed as follows: Staphylococci were grown overnight in Brain Heart Infusion (BHI) broth, aliquots were diluted (1:50) in DMEM and cultivated at 37°C under anaerobic conditions until an OD of 0.6 at 600 nm was reached, corresponding to 10⁸ CFU/ml (CFU, colony-forming unit). Bacteria were harvested by centrifugation, washed twice with phosphate-buffered saline (PBS), and suspended in the interaction medium (DMEM). Bacterial concentrations were estimated spectrophotometrically and confirmed by plate counts.

The purification was carried out essentially as described previously (Nguyen et al., 2015). Lpl1(+LSP) was isolated from the membrane fraction of S. aureus SA113 (pTX30::lpl1-his) whereas Lpl(−LSP) was isolated from the cytoplasmic fraction of S. aureus SA113Δlgt (pTX30::lpl1(-sp)-his). Both of these Lpl1-his proteins were purified via Ni-NTA affinity chromatography. For the enhance of protein expression, the clones were first cultivated aerobically at 37°C in the absence of xylose (BO-medium) until OD_{578 nm ≈} 0.5 was reached, then following continuously cultivated for 4 h in the presence of 0.5% xylose to induce the Lpl1 expression. The bacterial cells were harvested and washed two times with Tris buffer (20 mM Tris, 100 mM HCl, pH 8.0). Then the pellet was re-suspended with Tris buffer containing protease inhibitor table (Merck, Darmstadt, Germany) and lysostaphin (30 μg/ml) and incubated at 37°C for 2 h to disrupt the cell wall. After the first ultracentrifugation (235,000 × g for 45 min at 4°C), the supernatant containing the cytoplasmic proteins were collected for the next purification step. For membrane fraction isolation, the pellet was subsequently dissolved overnight at 6°C with Tris buffer containing 2% Triton-X100. After the second ultracentrifugation step, the supernatant containing membrane fragments was collected. The purification step was carried with Ni-NTA supper flow beads (Qiagen, Germany). The Ni-NTA beads capturing Lpl1 proteins were washed four times with the first washing buffer (Tris buffer containing 0.25% TritonX-100 and 20 mM imidazole) and following 2 times with the second washing buffer (Tris buffer containing 0.25% TritonX-100 and 40 mM imidazole). Finally, the Lpl1 was eluted with the same buffer containing 500 mM imidazole. The Lpl1 were concentrated via centrifugal ultra-filter unit with a molecular mass cut-off of 10 kDa (Sartorius AG, Göttingen, Germany). The concentrated samples of Lpl1 were dialyzed overnight at 6°C with DPBS buffer (Life technologies, Darmstadt) by Dialyzer tube, MWCO 6–8 kDa (Merk, Darmstadt) and was subsequently lyophilized overnight. The lyophilized samples stored at 4°C for further experiment. Two microgram of lyophilized samples were dissolved in water and run for SDS page to check the quantity and qualification of purified protein samples.

The human cervix cancer HeLa cells were obtained from American Type Culture Collection, Manassas, VA, USA HeLa cells were cultured in DMEM (Dulbecco's modified Eagle medium, Gibco, Invitrogen Corporation, Cergy Pontoise, France) supplemented with 10% fetal calf serum (FCS) (BioWest, Paris, France), 100 U/ml penicillin and 100 μg/ml streptomycin. Trypsin/EDTA (Gibco, Saint Aubin, France) was used to release adherent cells for subculturing.

Cell cycle analysis experiments were performed by a standard method with synchronized cells when at a given time point the majority of cells are in a particular cell cycle phase. It allows us to detect even little variations in cell cycle phase distribution. In order to synchronize the cells at the G1/S border the double thymidine block (DTB) protocol was employed as described (Deplanche et al., 2015). Briefly, HeLa cells were grown in a 25-ml flask up to 30% confluence. After washing with PBS, cells were cultivated in DMEM containing 10% of FCS supplemented with 2 mM thymidine (DMEM-T) for 18 h. Thymidine was then removed by washing with PBS and cells were cultivated in fresh DMEM with 10% of FCS for 9 h to release cells. The cells were then cultivated in DMEM-T for 17 h, followed by cultivation in DMEM containing 10% of FCS.

HeLa cells were grown in 25 ml flasks (30% confluence at the beginning of DTB), then cells were infected with S. aureus at a multiplicity of infections (MOI, number of bacteria per cell at the onset of infection) of 60:1, for 16, 20, 25, and 28 h after DTB release (infection medium: DMEM). The incubation time was chosen in agreement with the recognized evaluation of the phases of HeLa cell cycle and on the basis of our preliminary experiments. There was no alteration between the phases of infected and uninfected cells before 25 h of incubation of HeLa cells with bacteria. Consequently, synchronized HeLa cells were exposed to bacteria for 20, 25, and 28 h after DTB release. HeLa cell concentrations were determined using one of the four samples. The remaining samples were used for the analysis in triplicate. The low HeLa cell density at the beginning of the experiment was used to ensure cell proliferation during the entire experiment since cells cease proliferate, when they reach confluence (Owen et al., 1989).

Bacterial concentrations were estimated spectrophotometrically and were confirmed by CFU determination. Unbound bacteria were removed 2 h post-infection by treating cells with 20 μg/ml of lysostaphin and 100 μg/ml gentamicin for 2 h in cDMEM containing 3% FCS to eliminate the extracellular bacteria (Schmaler et al., 2009). Then infected HeLa cells were cultivated in cDMEM containing 25 μg/ml of gentamicin.

In experiments assigned for the analysis of the impact of purified Lpl1-his on the cell cycle progression, HeLa cells were exposed to 2 and 5 μg/ml of Lp1 and Lpl1(−LSP) for 20, 25, and 28 h after DTB release; quite similar as described for the exposure to bacteria. We also used 2 and 5 μg/ml of Pam2C (Pam₂CSK₄) and Pam3C (Pam₃CSK₄) as controls. Pam2C and Pam3C are synthetic di- and tri-palmitoylated lipopeptides (obtained from EMC, Tübingen, Germany) mimic the acylated amino terminus of bacterial lipoproteins. In all cell cycle progression experiments with either whole bacteria or purified proteins there was no difference before 20 h exposure, however, there was strong effect after 25 h and only weak effect after 28 h exposure. Consequently, we presented the values after 25 h exposition. It is worth mentioning that the recognized duration of phases of the HeLa cell cycle is approximately 10, 8, 3, and 1 h, for G1, S, G2, and M, respectively (Comayras et al., 1997).

After 25 h post-infection cells were detached and combined with adherent cells, which were collected by Trypsin/EDTA treatment and fixed in 70% ethanol overnight. Cells were then stained with propidium iodide (PI) and analyzed with an Accuri C6 flow cytometer (Becton Dickinson, Le Pont de Claix, France) as described earlier (Alekseeva et al., 2013; Deplanche et al., 2015). Data were collected from 20,000 cells and analysis was performed with CFlow software.

Invasion assays was performed according Almeida et al. with slight modifications (Almeida et al., 1996). Briefly, 3 × 10⁶ HeLa cells were incubated in cDMEM in 12 well plates overnight at 37°C in 5% CO₂. Then cells were exposed to S. aureus strains: USA300, its deletion mutant USA300Δlpl and the complemented mutant at MOI 100:1 (with inoculum 2 × 10⁷ CFU/well) for 1 h. Afterwards, cells were washed and incubated in DMEM containing 20 μg/ml lysostaphin and 100 μg/ml gentamicin for 2 h in order to remove extracellular bacteria and to measure internalized intracellular bacteria only. After lysis with 0.05% Triton X-100 in PBS, cell lysates were plated on BHI agar, and CFU were determined after overnight incubation using a micromethod (Florence Baron et al., 2006). To compare the invasion rates of USA300 and its mutants the percentage of internalized bacteria from the inoculum (i.e., of the number of input CFU) was determined for each strain.

Different concentration of purified Lpl1(−LSP) were tested for cytotoxic activity against HaCaT cells, an immortalized human keratinocyte line by using the LDH cytotoxicity assay kit (Thermo Scientific). The experiments were performed in triplicate in a 96 well plates. First, the optimal human cell number was determined to ensure LDH signal is within the linear range. Then we selected the initial seeding number of 10⁴ cells in 100 μl media (DMEM high glucose, 10% FBS, 1% Pen/Strep) per well (4 × 10⁴/100 μl/a well). After 2 days incubation at 37°C with 5% CO₂, the HaCaT cells were incubated with 10 μl purified Lpl1(−LSP) of different concentrations for 1 and 25 h, respectively. We also used PBS buffer and PBS buffer with 0.25% of Trixton X-100 as controls. After incubation cells were centrifuged and the LDH activity determined in the supernatant by measuring the absorbance at 490 nm and 680 nm in a Tecan reader. The percent cytotoxicity was calculated following the instruction of the kit.

Staphylococcus aureus USA300 and lpl mutant strains were aerobically cultured in TSB medium for 12 h and subsequently cultured on sheep blood agar at 37°C for 18 h. After that, these plates were stored at 4°C for 12 h for a cold-shock reaction. Beta toxin is known as a hot-cold toxin, which interacts with sheep blood cells at 37°C and only lyses them at 4°C (Herbert et al., 2010). The hemolytic activity makes the clear zone on the agar around the S. aureus colonies.

The strains S.a USA 300 wild type, USA 300Δlpl, and the complemented USA 300 Δlpl (pTX30-lpl) were grown overnight in TSB at 37°C and centrifuged for 10 min at 5000 × g at 4°C. The corresponding supernatants were filtered through sterile syringe filter with 0.2 μm pore size (Sarstedt, Germany) prior to be concentrated 5x using speedvac vacuum concentrator. PSM peptides were separated from the supernatant by reversed-phase chromatography using an XBridge C8 5 μm, 4.6 × 150 mm column (Waters). A linear gradient from 0.1% TFA in water to acetonitrile containing 0.08% TFA for 15 min with additional 5 min of 100% B at a flow rate of 1 ml/min was used and a 50 μl sample volume was injected. Peaks were detected at 210 nm. The PSM peptides were eluted between 12 and 19 min. As standard PSMα, PSMβ, and delta-toxins.

At least three different assays were performed per experiment. The differences among the groups were assessed by analysis of variance (ANOVA). P < 0.05 were considered to be significant. Tukey's Honestly Significant Difference test was applied for comparison of means between groups. The values are expressed as mean ± standard deviation (±SD).

After double thymidine block (DBT) HeLa cells were synchronized at the border of G1/S phase in agreement with the evaluated phases of the HeLa cell cycle. Cell cycle experiments were carried out with S. aureus USA300 carrying the νSaα island, its Δlpl deletion mutant and the complemented mutant; the genetic organization is shown in Figures S1A,B. Exposure of HeLa cells to USA300 or its Δlpl deletion mutant for 20 h showed hardly a difference in cell cycle. However, after 25 h there was a strong difference between wild type and the mutant, which was less pronounced after 28 h. Therefore, cell cycle analysis was carried out after 25 h of cells exposed to S. aureus. The rather narrow time frame of cell cycle arrest can be explained by the onset of cellular repair systems as has been shown for example by DNA-damage (Siegl and Rudel, 2015); however, such stress situations can also trigger the destruction of the damaged cell and activation of pathways leading to programmed cell death if cellular damage cannot be precisely repaired (Nascimento et al., 2012).

The DNA content of untreated synchronized HeLa cells 25 h after DBT release was used as a control: 59 ± 5% of cells were in the G1, 17 ± 3% in the S and 24 ± 4% in the G2 phase (Figure 1A). Exposure of HeLa cells to S. aureus USA300 caused a significant delay in the G2/M phase: after 25 h post-infection only 16 ± 4% of cells were in G1, 12 ± 4% in S, and 72 ± 6% were still in the G2/M phase (Figure 1B). When HeLa cells were infected with the lpl deletion mutant, USA300Δlpl, only 49 ± 4% were in the G2/M phase; while with the complemented mutant, USA300Δlpl (pTX30-lpl), 68 ± 5% of cells were in the in G2/M phase indicating that almost wild type levels were regained (Figure 1B). The results demonstrate that in the presence of USA300 and the complemented lpl mutant the G2/M phase transition was significantly delayed compared to the lpl deletion mutant, indicating that the lpl gene cluster is responsible for this delay. The next question was whether a purified Lpl protein is also able to delay the G2/M phase transition.

To verify whether protein or lipid portion of lipoprotein is responsible for a G2/M phase transition delay, we selected Lpl1, as a prototype of the cluster, which is encoded by the first gene of the lpl cluster in USA300 (Figure S1A). Lpl1 was expressed with C-terminal His-tag in SA113 and SA113Δlgt with and without lipoprotein signal peptide (LSP). In the absence of LSP Lpl1 was not expressed in the membrane but in the cytoplasm. The plasmid encoded constructs for expression of Lpl1, carrying the lipid modification, and Lpl1(−LSP), not lipid modified, are shown in Figures S1C,D. The lipid modified Lpl1 was expressed in SA113 (pTX30::lpl1-his) with its LSP; Lpl1(+LSP) was isolated from the membrane fraction and purified by Ni-NTA chromatography and its purity verified by SDS-PAGE (Figure S1E). In S. aureus the lipid moiety of lipoproteins is tri-acylated (Kurokawa et al., 2009). As we also wanted to see whether the protein part alone, has an effect on the cell cycle, we expressed Lpl1 without its lipoprotein signal peptide in SA113Δlgt (pTX30::lpl1-(LSP)-his). This Lpl1(−LSP) was isolated from the cytoplasm and also purified by Ni-NTA chromatography (Figures S1D,E).

Analysis of the DNA content of control untreated synchronized HeLa cells 25 h after DBT release revealed that 76 ± 6% of cells were in the G1, 5 ± 2% in the S and 19 ± 3% in the G2/M phase (Figure 2A). In the presence of 2 and 5 μg/ml (68 and 171 nM) Lpl1 the percentage of HeLa cells after 25 h post incubation was increased from 19 ± 5% (control) to 26 ± 4% and 32 ± 5% in G2/M phase, respectively (Figure 2B). This result shows that Lpl1(+LSP) moderately delayed the G2/M transition phase. Next we analyzed Lpl1(−LSP) in the same way as Lpl1(+LSP) and it turned out that with Lpl1(−LSP) (70 and 176 nM) even a higher percentage of the HeLa cells (31 ± 4 and 49 ± 5%) were in the G2 phase after 25 h (Figure 2C). This result demonstrates that solely the protein portion of Lpl1 causes the G2/M transition delay.

To confirm our observation of a pivotal role of the protein portion of Lpl1 in the inhibition of the G2/M phase transition, we examined whether synthetic lipopeptides have an impact on cell cycle alteration. Therefore, synchronized HeLa cells were exposed to 2 and 5 μg/ml of Pam2 and Pam3. These are synthetic lipopeptides that mimic the acylated amino terminus of bacterial lipoproteins; the structure is shown in Table 1. Analysis of the DNA content of untreated synchronized HeLa cells 25 h after DBT release revealed that 52 ± 3% of cells were in the G1 phase, 9 ± 2% in the S phase and 39 ± 3% in the G2 phase (Figure 3A). Exposure of HeLa cells to 2 and 5 μg/ml (1574 and 3937 nM) of Pam2 resulted in an increase of cells in the G2 phase to 49 ± 4 and 61 ± 5%, respectively (Figure 3B). Interestingly, exposure HeLa cells to 2 and 5 μg/ml (1333 and 3333 nM) Pam3C showed less influence on the cell cycle. The percentage of cells in the G2 phase was increased from 39% (control) to 47% (2 μg/ml Pam3C) and to 48% (5 μg/ml Pam3C), respectively. Unlike Pam2Cys there was no dose dependent increase with Pam3C (Figure 3C). As we used a more than 20 times higher concentration of Pam2C and Pam3C compared to Lpl1 to see an effect at al, we assume that the lipopeptides have only a marginal effect on cell cycle. These results strengthen our findings that the protein portion of Lpl1 is the major player in the eukaryotic cell cycle alteration.

It has been indicated that in the presence of the lpl cluster enhances the invasion into human keratinocytes and in mouse skin (Nguyen et al., 2015). As we study here the effect on cell cycle in HeLa cells it is important to verify whether the same results were achieved in this cell line. Indeed, the internalization into epithelial cells was lowest with USA300Δlpl and was increased two-fold and even three-fold with USA300 (parent) and the complemented mutant USA300Δlpl (pTX30-lpl), respectively (Figure 4). These results clearly demonstrate that the lpl cluster contributes to HeLa cell invasion and that there might be a link between host cell invasion and G2/M cell cycle delay.

It has been shown that alpha-hemolysin and PSMs have an impact on cell cycle alteration (Haugwitz et al., 2006; Deplanche et al., 2015). In order to rule out that altered expression of these compounds have an influence on cell cycle delay we investigated whether in USA300Δlpl alpha-hemolysin and PSM was altered compared to the parent strain USA300. We found that USA300 and USA300Δlpl showed no difference in any of the two parameters. Alpha-hemolysin production, as indicated by the hemolysis zone on sheep agar plates, was unaltered in the mutant (Figure 5A). The contents of PSM-α, -β, and -γ in the supernatant of USA300 and its lpl mutant were determined by HPLC analysis; there was no difference in the production of the PSM peptides (Figure 5B). As the S. aureus strains were pre-cultivated in BHI medium for cell cycle experiments the quantification of the PSMs was also carried out in supernatant of BHI grown cells. Interestingly, when S. aureus cells were cultivated in TSB medium the HPLC resolution of PSMs was improved, but there was also no difference in the concentration between wild type and Δlpl mutant (Figure S2). These results indicate, that the observed effect of whole S. aureus cells as well of the isolated Lpl1(+LSP) and Lpl1(−LSP) on extending the G2/M transition phase are really due the lpl cluster and the Lpl proteins.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.