Staphylococcal Superantigen-like protein 11 mediates neutrophil adhesion and motility arrest, a unique bacterial toxin action

Methicillin resistant Staphylococcus aureus (MRSA) is a major human pathogen, which causes superficial to lethal clinical infections. Neutrophils are the most abundant leukocytes in the blood and are the first defense mechanism against S. aureus infections. Here we show Staphylococcal Superantigen-Like protein 11 (SSL11) from MRSA USA300_FPR3757 mediated differentiated human neutrophil-like cells (dHL60) motility arrest by inducing cell adhesion and “locking” cells in adhesion stage, without inducing oxidative burst. Pre-incubation of SSL11 with the glycan Sialyl Lewis X blocked SSL11 function and de-glycosylation of dHL60 cells by PNGase F abolished SSL11 binding, suggesting that SSL11 functions via interacting with glycans. This is the first description of a bacterial toxin inhibiting neutrophil motility by inducing adhesion and “locking” cells in an adhesion stage. Therefore, this study might provide a new target against S. aureus infections.

. SSL11 stimulates dHL60 cell adhesion. (A) 2 μg of purified SSL7, SSL11 and SSL11 3XF were separated by SDS-PAGE and stained with Coomassie Blue. (B) dHL60 cells were incubated with 80 nM of SSL7 or SSL11 in fibronectin (FN)-coated plates at 37 °C for 30 min followed by two PBS washes. Representative DIC images were shown. (C) dHL60 cells were incubated with SSL7 or SSL11 in FN-coated 96-well plates at 37 °C for 30 min followed by two PBS washes. Adherent cells were quantified by crystal violet staining and shown as adhesion arbitrary unit (AU). (D) dHL60 cells were incubated with 80 nM of SSL11 at 37 °C for 30 min and chased in fresh media without SSL11 for another 4 hours in FN-coated plates. Representative DIC images were shown. (E) dHL60 cells were treated with SSL11 as described in (D) in FN-coated 96 well plates. Adherent cells were quantified by crystal violet staining and shown as adhesion arbitrary unit (AU).
www.nature.com/scientificreports www.nature.com/scientificreports/ after SSL11 treatment, dHL60 cells were incubated with SSL11 for 30 mins, and cells were chased in media without SSL11 for another four hours. Unexpectedly, dHL60 cells remained adhesive four hours later in a dose-dependent manner, suggesting that SSL11 "locked" cells in adhesion stage (Fig. 1D,E). SSL11 is the first known member of the SSL family to induce cell adhesion.

SSL11 inhibits fMLP-mediated dHL60 cells motility. Neutrophil migration requires a well-regulated
balance between adhesion and de-adhesion, where interruption of this balance affects neutrophil motility. To test if SSL11-mediated dHL60 cell adhesion affects cell motility, the effect of SSL11 on chemotactic peptide fMLP-induced cell motility was tested. fMLP was added to the edge of a fibronectin (FN)-coated well and cell motility was recorded for 30 mins. Upon the addition of fMLP, dHL60 cells migrated an average length of 307 μm, while cells pre-incubated with SSL11 migrated an average length of 55 μm ( Fig. 2A,B). Visualization revealed that SSL11 pre-incubated dHL60 cells responded to fMLP stimulation, but did not migrate and appeared to have enlarged adhesive tails ( Fig. 2A insert, Supplementary Movies 3 and 4). Thus, SSL11 did not inhibit cell motility by directly interfering with fMLP sensing, but rather, SSL11 inhibited cell motility by increasing tail-localized cell adhesion. Moreover, SSL11-treated cells displayed minimum migration toward fMLP even four hours after SSL11 was taken out, suggesting that "locking" cells in adhesion stage by SSL11 abolished cell motility (Fig. 2C,D).
Dose-dependent SSL11 association with dHL60 cells. SSL11 association with dHL60 cells was dose-dependent and correlated with cell adhesion (Fig. 3A,B). Cell adhesion was detected at 40 nM SSL11 and reached maximum adhesion at 80 nM SSL11 (Fig. 1C). SSL7 did not bind to dHL60 cells (Fig. 3A,B) and did not www.nature.com/scientificreports www.nature.com/scientificreports/ induce cell adhesion (Fig. 1C), supporting that cell association is required for SSL11-induced cell adhesion and motility arrest. In addition, SSL11-induced cell adhesion was associated with cell spreading when the average cell diameter was quantified (Fig. 3C). Glycan-dependent SSL11 function. SSLs are divided into two subgroups depending on the presence of a Sialyl Lewis X [SLex-Neu5Aca2-3Galb1-4(Fuc1-3) GlcNAc, SLeX] binding site within the C-terminal domain. SSL2-6 and SSL11 possess this conserved carbohydrate binding motif 4,5 . To test whether SSL11-induced cell adhesion and motility arrest is glycan dependent, SSL11 was pre-incubated with exogenous SLeX before incubation with dHL60 cells. SLeX inhibited SSL11 cell association and SSL11-induced cell spreading, adhesion in a dose-dependent manner with complete inhibition at 100 μM. As a control, PMA-induced cell adhesion was not affected by SLeX, suggesting that SSL11 binds and functions via interacting with SLeX specifically (Fig. 5). SLeX is the sub-terminal disaccharide N-acetyllactosamine (LacNAc) on both N-and O-glycans. N-linked glycosylation is the most common form of covalent protein modification in human cells, especially for secretory and membrane-bound glycoproteins 16 . To further test whether SSL11 binding and function are glycan-dependent, dHL60 cells lysate were treated with PNGase F to remove N-linked glycosylation, which was confirmed by loss of Erythrina Cristagalli Lectin (ECL) binding and Silver staining ( Supplementary Fig. S2A,B). A 3XFLAG epitope tag was engineered at the N terminus of SSL11 (SSL11 3XF ) (Fig. 1A) and SSL11 3XF induced dHL60 cell adhesion, similar to non-FLAG tagged SSL11 ( Supplementary Fig. S3). SSL11 3XF bound to dHL60 cell lysate at a region of 75-150 kDa by Far Western blotting and de-glycosylation by PNGase F abolished those interactions ( Supplementary Fig. S2C). This further suggests that SSL11 binding and stimulation of neutrophil adhesion are glycan-dependent.

SSL11 does not induce oxidative burst in dHL60 cells.
Oxidative burst describes the phagocytic response of cells to produce reactive oxygen species (ROS). The uncharged and nonfluorescent Dihydrorhodamine 123 (DHR123) is a widely used ROS indicator which can passively diffuse across membranes. When oxidized, DHR123 is converted to rhodamine123 which fluoresces when excited at 488 nm. The oxidative burst of neutrophils can be triggered by PMA 17 . To test if SSL11 induced oxidative burst of dHL60 cells, dHL60 cells were pre-incubated with DHR123 followed by incubation with either PMA or SSL11. PMA induced strong ROS production shown by green fluorescence while SSL11 did not (Fig. 6), suggesting that SSL11 did not activate dHL60 cells oxidative burst although SSL11 induced cell adhesion and spreading. SSL11 does not induce cell adhesion via PSGL-1. Earlier studies reported that SSL11 blocked neutrophil rolling by interacting with PSGL-1 4 . To test if PSGL-1 is also responsible for SSL11-induced cell adhesion and www.nature.com/scientificreports www.nature.com/scientificreports/ spreading, we generated a psgl-1 knockout cell line using CRIPR-Cas9 system ( Supplementary Fig. S4). When wt, vector control or psgl−/− dHL60 cells were incubated with SSL11, SSL11 induced cell adhesion was not affected in psgl−/− cells (Fig. 7). This indicated that SSL11-induced cell adhesion was independent of PSGL-1.

Discussion
In the current study, SSL11 from USA300_FPR3757 arrested fMLP-mediated neutrophil like dHL60 cells motility by inducing adhesion and "locking" cells in adhesion stage ( Supplementary Fig. S5). This is a new function of SSL family proteins and sheds light on the understanding of the virulence and pathogenesis of S. aureus.
S. aureus is a potent pathogen as approximately 30% of the human population is continuously colonized and the vast majority of S. aureus diseases occur in immune-competent individuals. To achieve this level of pathogenicity, S. aureus deploys an arsenal of immune-evasive strategies to manipulate neutrophil function. Neutrophils, the first line of defense against invading pathogens, are the most abundant leukocytes with more than 50% of the bone marrow devoted to neutrophil production. Neutrophil recruitment into inflamed tissue is a well-regulated cascade of events: rolling, adhesion to endothelial cells, post adhesion strengthening, crawling, and finally transmigration 13,18,19 . To prevent neutrophil recruitment, S. aureus devotes various secreted factors targeting early steps of neutrophil recruitment: rolling and chemokine signaling. SSL5 and extracellular adherence protein (Eap) interfere with neutrophil rolling by binding to PSGL-1 and ICAM-1, respectively. SSL10 inhibits CXCL12-mediated responses by targeting CXCR4 20 and SSL3 blocks recognition of staphylococcal lipoproteins and peptidoglycan by binding to TLR2 6 . Chemotaxis-inhibitor protein of S. aureus (CHIPS) binds and www.nature.com/scientificreports www.nature.com/scientificreports/ inhibits formyl peptide receptor 1 (FPR1) and C5a receptor (C5aR) [21][22][23] . In the current study, SSL11 from S. aureus USA300_FPR3757 induced cell adhesion and arrested fMLP-stimulated cell motility. A previous study showed that SSL11 blocks neutrophil rolling by interacting with PSGL-1 4 . Our study showed that SSL11-induced cell adhesion was independent of PSGL-1 (Fig. 7). Thus, SSL11 has multiple functions toward neutrophil manipulation as SSL5 displays several immune evasion functions 5,[24][25][26][27] . Moreover, SSL11-induced cell adhesion was observed over different surfaces: FN-coated, ICAM-1 coated, plastic and glass surfaces (data not shown), suggesting that SSL11 might trigger multiple pathways to induce strong cell adhesion, which needs further investigation. SSL11 did not induce oxidative burst in dHL60 cells (Fig. 6) although SSL11 induced cell adhesion, suggesting that SSL11 functions as immune "suppressor" instead of immune "activator" such as Superantigens.
SSL11 provides extra mechanism of abrogating neutrophil recruitment by inducing cell adhesion and "locking" cells in adhesion stage, especially for neutrophils already rolling, ready for next step during migration. Elicitation of neutrophil adhesion by SSL11 blocks neutrophil migration to the site of infection, which is beneficial for S. aureus infection. Besides neutrophils, SSL11 also induced cell spreading for THP-1 cells, a human monocytic cell line. This suggests that SSL11 targets other leukocytes, which is beneficial for S. aureus infection. To our knowledge, this is the first description of a bacterial toxin inhibiting neutrophil motility by inducing adhesion and "locking" cells in the adhesion stage.
In the current study, SSL11-induced cell adhesion and spreading was glycan dependent, as pre-incubation of SSL11 with SLeX abolished SSL11 association and its function (Fig. 5) and de-glycosylation of cell lysate abolished SSL11 binding (Supplementary Fig. S2). Therefore, glycan SLeX serve as a functional receptor for SSL11. SLeX functions as a receptor for H. Pylori adhesion SabA, which allows the bacteria to cling to the surface of the gastric cell 28 . SLeX is an invariant ligand for P-, E-and L-selectin. The selectins bind weakly to SLeX-like glycans but with high affinity to glycoprotein with SLeX such as PSGL-1 29 . This is consistent with previous work that SSL11 binds PSGL-1 via glycans 4 . SSL11 3XF bound to dHL60 cell lysate at a region of 75-150 kDa in a glycan-dependent manner ( Supplementary Fig. S2C), suggesting glycoproteins with SLeX are receptors for SSL11. Besides PSGL-1, other receptors need to be identified. Sharing a SLeX binding pocket with SSL11, SSL5 binds to PSGL-1, GPIbα, and GPCRs via glycan binding 5,25,26 . There might be an additional receptor binding site on SSL11 which determines the binding and function specificity. Bacterial toxins sometimes employ dual-receptor binding for specificity. Botulinum neurotoxin A (BoNT/A) binds to a ganglioside and a synaptic vesicle protein SV2A/B/C, BoNT/B binds to a ganglioside and a synaptic vesicle protein Synaptotagmins I/II while Tetanus neurotoxin binds to dual-gangliosides 30,31 .
SSL11 demonstrated a new mechanism of neutrophil manipulation by bacterial toxins: inducing intense adhesion to arrest cell motility, which can prevent neutrophils from migrating to the site of infection. SSL11 might provide a new target against S. aureus infections. Plasmid construction. E. coli codon-optimized sequences of S. aureus strain USA300 _FPR3757 SSL7 (GenBank: ABD22785.1) and SSL11 (GenBank: ABD21156.1) were synthesized (IDT) and subcloned into a pET28a vector for expression. An N-terminal 3 × FLAG tag was engineered downstream of His 6 tag for construct SSL11 3XF . All constructs contain an N-terminal His 6 tag for protein purification. Constructs were confirmed by DNA sequencing.

Antibodies and reagents.
Protein expression and purification. Plasmids encoding SSL7, SSL11 and SSL11 3XF were transformed into E. coli BL21(DE3). Transformants were grown overnight on LB agar plates containing 50 µg of kanamycin/ml, which were the inoculums for liquid cultures (LB, 400 ml) containing the same antibiotic. Cells were cultured at 37 °C to an optical density at 600 nm (OD 600 ) of 0.6 when T7 promoter expression was induced with 1 mM IPTG. Cells www.nature.com/scientificreports www.nature.com/scientificreports/ were cultured overnight at 250 rpm at 16 °C. Cells were pelleted and lysed with a French press and clarified by centrifugation. His 6 -tagged proteins were purified using Ni 2+ -nitrilotriacetic acid (NTA) resin (Qiagen). Purified SSL7 and SSL11 were dialyzed into 20 mM Tris buffer (pH 7.9) with 200 mM NaCl and 40% glycerol. Aliquots were stored at −20 °C.
Protein labeling by Alexa Fluor. SSL7 and SSL11 were labeled using Alex Fluo568 Protein labeling kit (Invitrogen, A10238). SSL7 and SSL11 were dialyzed in PBS at 4 °C overnight. 50 μl of 1 M bicarbonate was added to 0.5 ml of 2 mg/ml SSL7 or SSL11 followed by transfer the protein solution to a vial of reactive dye. The reaction mixture was stirred for two hours at room temperature followed by dialysis in PBS at 4 °C overnight to eliminate free dyes. Labeled proteins were stored at 4 °C.
Cell Culture. The HL60 cell (ATCC ® CCL-240 ™ ) is from a patient with acute promyelocytic leukemia, which can be differentiated into neutrophil-like cells 14 . HL60 cells were cultured and maintained in RMPI medium with 10% FBS and 20 mM HEPES. HL60 cells were differentiated in complete medium with 1.3% DMSO as described by Fleck RA et al. 14 . Differentiated HL60 (dHL60) cells show ≥90% viability (trypan blue), ≥55% CD35 expression and ≤20% CD71 expression. Differentiation was confirmed by Flow cytometry using anti-CD11b, anti-CD35 and anti-CD71 antibodies (data not shown). dHL60 cells were used for experiments at day 6-8 after differentiation. The human monocytic cell line THP-1 is a generous gift from Dr. Thomas Zahrt at Medical College of Wisconsin. THP-1 cells were cultured and maintained in RMPI medium with 10% FBS and 20 mM HEPES. dHL60 cells adhesion assay. 96-well plates were coated with 100 μl of Fibronectin (10 μg/ml) in PBS at 4 °C overnight followed by two PBS washes. Adhesion assay was performed as previous described with modification 32 . dHL60 cells (3 × 10 5 cells/well) were incubated with SSL7 or SSL11 (20-160 nM) at 37 °C for 30 min followed by two PBS washes. Adherent cells were incubated with 0.5% crystal violet at RT for 10 min followed by four PBS washes. The plate was dried and followed by addition of ethanol to solubilize cell bound crystal violet, absorbance at 595 nm was measured to quantify crystal violet and shown as adhesion arbitrary unit (AU). For SLeX blocking experiment, indicated amount of SLeX was incubated with SSL11 or PMA (10 ng/ml) in serum free RPMI media at 4 °C for two hours before incubation with dHL60 cells. dHL60 cells motility assay. dHL60 cells (5 × 10 5 cells) were incubated with 80 nM of SSL11 at 37 °C for 30 min. 10 μM fMLP was added and cell motility was recorded by taking DIC images every 10 s for 30 min on a 37 °C heated stage. Twenty cells from each group were tracked using ImageJ "Manual tracking" and data were imported to Ibidi Chemotaxis and Migration Tool and plotted by GraphPad Prism 7.03 to show the individual cell moving tracks and total migration length (μm). SSL11 association with dHL60 cells. 24-well plates with glass cover slips were coated with Fibronectin (10 μg/ml) in PBS at 4 °C overnight followed by two PBS washes. dHL60 cells (5 × 10 5 cells/well) or THP-1 cells (2 × 10 5 cells/well) were incubated with SSL7 568 or SSL11 568 (20-160 nM) at 37 °C for 30 min. Cells were washed twice with PBS and fixed with 4% paraformaldehyde at room temperature for 15 min. Cells were incubated in blocking solution (DPBS with 10% FBS, 2.5% cold-water fish skin gelatin, 0.1% Triton-X, and 0.05% Tween 20) for one hour followed by incubation with Phalloidin 647 in incubation solution (DPBS 5% FBS, 1% cold-water fish skin gelatin, 0.1% Triton X, and 0.05% Tween 20) at RT for one hour. Cells were washed and incubated with DAPI for nuclei staining, then were fixed again with 4% paraformaldehyde at room temperature for 15 min. After wash, cover slips were mounted using ProLong Gold Antifade Mountant). For SLeX blocking experiment, indicated amount of SLeX was incubated with SSL11 or PMA (10 ng/ml) in serum free RPMI media at 4 °C for two hours before incubation with dHL60 cells. Images were captured with a Nikon TE2000 microscope using a Photometrics CoolSnap HQ2 camera. Images were captured by epifluorescence with a Sedat Quad cube (Chroma Technology Corp).
Oxidative burst by DHR123 assay. dHL60 cells (5 × 10 5 cells/well) were incubated with 25 μM DHR123 in serum free medium in glass bottom 24-well plates at 37 °C for 30 mins followed by incubation with PMA (100, 300 and 900 ng/ml) or SSL11 (20-80 nM) at 37 °C for 30 mins. Cells were washed, fixed and stained by DAPI. Dihydrorhodamine 123 is a nonfluorescent reactive oxygen species (ROS) indicator and is oxidized by ROS to exhibit green fluorescence. DHR123 fluorescence was observed under microscopy using FITC channel. Cell staining is shown: oxidized DHR123 by ROS (green) and DAPI (blue).
Far Western Blotting. dHL60 cells were lysed using RIPA buffer and the dHL60 cell lysate was quantified using BCA assay. For de-glycosylation, dHL60 cell lysate were incubated with PNGase F at 37 °C for 24 hours. dHL60 cell lysate (2 μg) with or without PNGase F treatment were separated on SDS-PAGE and were transferred (2019) 9:4211 | https://doi.org/10.1038/s41598-019-40817-x www.nature.com/scientificreports www.nature.com/scientificreports/ to PVDF membrane in duplicate. One membrane was blocked with 5% BSA, followed by incubation with biotinytlated ECL at RT for two hours and incubation with HRP-conjugated streptavidin for another one hour. The other membrane was blocked with 2% milk and followed by incubation with 10 nM SSL11 3XF and HRP-conjugated anti-FLAG M2 antibody at room temperature for 2 h. Both membranes were washed, incubated with SuperSignal, and images were taken with a Fluorchem HD2 camera system. Silver Staining. Gels from SDS-PAGE were fixed with 30% isopropanol, 10% glacial acetic acid, washed with water for 10 min for three times. Gel was incubated with 0.02% sodium thiosulfate for 90 s followed by two quick water rinses. The gel was incubated with 0.2% silver nitrate for 20 min followed by two quick water rinses. The gel was developed in develop buffer (6 g sodium carbonate, 50μl formaldehyde, and 3 ml of 0.2% sodium thiosulfate in 100 ml) as needed and was quenched using 10% acetic acid.
Data analysis and statistics. Images were generated with equal exposure times and conditions. Image intensity analysis was performed using ImageJ (NIH). Figures were compiled using Canvas X 2017 (ACD Systems). Data were shown as means with standard errors of the means (SEM). Data were analyzed by unpaired two-tailed Student's t test using GraphPad Prism 7.03 from three independent experiments. P values of <0.05 at the 95% confidence level are indicated by *, P values of <0.005 are indicated by **P values of <0.001 are indicated by *** and P values of <0.0001 are indicated by ****.