Fig 1.
SElX binds to human monocytes and neutrophils from multiple mammalian species.
Flow cytometry analysis of recombinant staphylococcal proteins binding to isolated human cells. Cell type was determined by forward (FSC-H) and sideways (SSC-H) scatter (i). SElX binding to human neutrophils (ii) monocytes (iii) and lymphocytes (iv) in addition to neutrophils isolated from mice (v), cattle (vi) and rabbits (vii), was examined. For all graphs binding was detected using mouse anti-HIS-FITC IgG binding to the 6 x HIS-tag on the recombinant proteins. Mean median fluorescence of three donors is shown ± standard error of the mean (SEM). SSl5 and SSl7 were used as positive and negative controls, respectively. The same legend is used for all graphs.
Fig 2.
SElX protein sequence encodes a conserved sialic acid-binding motif.
(i) A conserved glycan binding motif (colored red) is present in the amino acid sequences of characterised SSl-proteins and SElX. (ii) The sialic acid-binding motif is conserved in all 17 alleles of SElX. (iii) Amino acid conservation across the sialic acid-binding region of 7 staphylococcal neutrophil binding proteins (SSl2, SSl3, SSl4 SSl5, SSl6, SSl11 and SElX). The probability of residues at each position is proportional to the size of the letters. Image generated using the weblogo 3.4 program (http://weblogo.threeplusone.com/). Chemical property color scheme: green is polar, blue is basic, red is acidic, purple is neutral and black is hydrophobic. Sequences were obtained from previously published work [6, 23].
Fig 3.
SElX interaction with neutrophils is dependent upon the sialic acid binding motif.
(i) Flow cytometry analysis of SElX binding to human neutrophils and monocytes pretreated with 0.2 U/ml neuraminidase versus unteated cells. 6 x HIS-tagged SElX binding was detected using mouse anti-His FITC-labelled IgG. (ii) Hypothetical protein model of SElX indicating the location of predicted sialic acid binding residues (colored green). The motif residues are magnified in the black box to indicate their predicted atomic structure in the binding pocket. (iii) Flow cytometry analysis of site-directed alanine-replacement mutants of SElX binding to human neutrophils. 6 x HIS-tagged SElX binding was detected using Mouse anti-HIS FITC-labelled IgG.
Fig 4.
SElX binds to multiple neutrophil surface glycoproteins in a sialic acid-dependent manner.
(i) Neutrophil lysates were analysed for peptide enrichment following incubation with wild-type SElX (WT) and SElX EKQD-A. Data shown indicates the mean number of peptides detected from each enriched protein (± SD from three donors). * denotes significance difference (p-value <0.05) in peptide enrichment between the wild-type and mutant protein as determined by multiple t-test comparisons (one unpaired, two tailed test per protein), without assuming similar standard deviation. (ii) ELISAs were performed comparing the interaction between staphylococcal proteins and recombinant CD50. Two-way ANOVA statistical analyses were performed. **** denotes a significant difference between SElX and SElX EKQD-A.
Fig 5.
SElX inhibits IgG-mediated neutrophil phagocytosis and reduces killing by human neutrophils.
(i) Phagocytosis of fluorescent-labelled S. aureus opsonized with 150 μg/ml of purified human IgG or 1% (v/v) complement-inactivated human pooled serum (ΔHPS), in the presence of SElX at various concentrations. Gates indicate the populations of neutrophils that have phagocytosed bacteria or not (+/-). (ii) Inhibition by recombinant SElX protein was compared to the IgG-mediated phagocytosis inhibitor FLIPr and the neutrophil-binding deficient mutant SElX EKQD-A. Phagocytosis was calculated as the percentage of cells with fluorescent bacteria and expressed relative to buffer-treated cells with 75 μg/ml human IgG. Results shown are the means of three different human donors (error bars SE of mean). Results between SElX wild-type and SElX EKQD-A protein were tested by two-way ANOVA and found to be significantly different, *indicates at which concentration a significant difference in phagocytosis was observed. (iii) S. aureus USA300 wild-type, selx deletion and site-directed mutants were incubated with isolated human neutrophils for 60 min. Following incubation neutrophils were lysed with Triton-X 100 and surviving bacteria plated and enumerated. Killing was calculated as the difference between the no neutrophil control and the surviving CFU. Data shown are the mean % surviving bacteria incubated with cells from 5 donors ± SD. CFU data were tested by students t-test (unpaired, two-tailed) with Welches correction, (* indicates a p value < 0.05, ** indicates a p value < 0.01).
Table 1.
Bacterial strains used in this study.
Fig 6.
Neutrophil binding-deficient mutants of SElX retain mitogenic activity.
Isolated human PBMC were stimulated with recombinant SElX and sialic acid-binding mutants. After 72 h incubation, proliferation was determined by analysing the incorporation of [3H] thymidine. Results shown are the means of triplicate measurement from 3 human donors ± standard deviation of the mean.
Fig 7.
SElX does not contribute to S. aureus virulence in a mouse skin abscess infection model.
(i) Lesion size of each mouse (n = 10) was measured every 24 h post inoculation (p.i.) for 6 d. Mean lesion size ± SEM is plotted for each of the 4 USA300 LAC mutant groups. (ii) Bacteria were recovered from excised skin lesions and enumerated by serial dilutions. CFU were normalised to the weight of tissue homogenised to give the bacterial load per mg of tissue. CFU per mg are displayed for each infected animal; the horizontal line indicates the mean CFU/mg of tissue and vertical bars show the SEM for each group (n = 10) (iii) Representative images from histological examinations of skin lesions 72 h and 144 h post inoculation. Mounted sections were stained with haematoxylin and eosin. Black arrows on each image indicate the surface of the epidermis.
Fig 8.
SElX contributes to lethality in a rabbit model of necrotising pneumonia by inhibiting neutrophil function.
Kaplan-Meier curves of % survival of rabbits infected with wild-type S. aureus LAC, LACΔselx, LACΔselx rep and LAC selx EKQD-A at a dose of 6x109 CFU (i) and S. aureus LAC and LAC selx EKQD-A at a dose of 2x109 CFU (ii), p-values stated are the results of log-rank (Mantel-Cox) tests. (iii) Increase in rabbit core temperature (ΔT°C) 12 h after pulmonary infection with LAC, LACΔselx, LACΔselx rep and LAC selx EKQD-A at a dose of 6x109 CFU (data plotted are the mean of three animals ± SD). (iv) Increase in rabbit core temperature (ΔT°C) 12 h after pulmonary infection with LAC and LAC selx EKQD-A at a dose of 2x109 CFU (data plotted are the mean of four animals ± SD, ** indicates statistical significance by unpaired, two-tailed students t-test p<0.003) (v) Gross pathology of lungs from rabbits infected with S. aureus strain LAC and LAC selx EKQD-A showing representative examples of haemorrhagic lesions.
Table 2.
List of primers used in this study.