The Superantigen Toxic Shock Syndrome Toxin 1 Alters Human Aortic Endothelial Cell Function

Abstract

Staphylococcus aureus infective endocarditis (IE) is a fast-progressing and tissue-destructive infection of the cardiac endothelium. The superantigens (SAgs) toxic shock syndrome toxin 1 (TSST-1), staphylococcal enterotoxin C (SEC), and the toxins encoded by the enterotoxin gene cluster (egc) play a novel and essential role in the etiology of S. aureus IE. Recent studies indicate that SAgs act at the infection site to cause tissue pathology and promote vegetation growth. The underlying mechanism of SAg involvement has not been clearly defined. In SAg-mediated responses, immune cell priming is considered a primary triggering event leading to endothelial cell activation and altered function. Utilizing immortalized human aortic endothelial cells (iHAECs), we demonstrated that TSST-1 directly activates iHAECs, as documented by upregulation of vascular and intercellular adhesion molecules (VCAM-1 and ICAM-1). TSST-1-mediated activation results in increased monolayer permeability and defects in vascular reendothelialization. Yet stimulation of iHAECs with TSST-1 fails to induce interleukin-8 (IL-8) and IL-6 production. Furthermore, simultaneous stimulation of iHAECs with TSST-1 and lipopolysaccharide (LPS) inhibits LPS-mediated IL-8 and IL-6 secretion, even after pretreatment with either of the proinflammatory cytokines tumor necrosis factor alpha (TNF-α) and IL-1β. IL-8 suppression is not mediated by TSST-1 binding to its canonical receptor major histocompatibility complex class II (MHC-II), supporting current evidence for a nonhematopoietic interacting site on SAgs. Together, the data suggest that TSST-1 differentially regulates cell-bound and secreted markers of endothelial cell activation that may result in dysregulated innate immune responses during S. aureus IE. Endothelial changes resulting from the action of SAgs can therefore directly contribute to the aggressive nature of S. aureus IE and development of life-threatening complications.

Keywords: TSST-1; aortic endothelial cell dysfunction; human endothelial cells; infective endocarditis; superantigen.

FIG 1

Immortalized HAECs retain phenotypic and functional characteristics of primary cells. (A) (Top) Immortalized HAEC confluent monolayers stained with CD31-AF488 antibody (green) and DAPI (blue). (Bottom) Flow cytometry histogram plots of primary or immortalized HAECs labeled with CD31-AF488 antibody. Secondary antibody only (dashed line) and nonstained cells (dotted line) were used as negative controls. (B) (Top) Immortalized HAEC confluent monolayers stained with VE-cadherin-AF488 (green) antibody, Acti-Stain 555 phalloidin (red), and DAPI (blue). (Bottom) Flow cytometry histogram plots of immortalized HAECs labeled with VE-cadherin-AF488 and ZO-1-AF647 antibodies. Secondary antibody only (dashed line) was used as a negative control. (C) Tube forming assay of primary (top) and immortalized (bottom) HAECs at 0 h and 8 h after seeding in Matrigel-coated wells. (D) Acetylated low-density lipoprotein (Ac-LDL) uptake assay of primary HAECs (top), immortalized HAECs (middle), and fibroblasts (negative control; bottom) treated with fluorescently labeled Ac-LDL (red) for 2.5 h and stained with DAPI (blue). (E) IL-8 detection in culture supernatants 24 h after treatment of primary or immortalized HAECs with increasing concentrations of TNF-α, IFN-γ, or IL-6. (F) IL-8 detection in culture supernatants 24 h after treatment of immortalized HAECs with increasing concentrations of IL-2 or IL-1.

FIG 2

TSST-1 directly activates iHAECs. Shown is In-Cell Western assay detection of total protein of the cell adhesion molecules VCAM-1 (A) and ICAM-1 (B) in iHAECs treated with TSST-1 (25 μg/ml) or LPS (1 to 2 ng/ml), individually and in combination for 24 h. P values were determined by one-way ANOVA (asterisks) with Holm-Sidak's multiple-comparison test shown for TSST-1 versus medium or versus LPS plus TSST-1 (Greek letters, adjusted P value). (A) **, P = 0.003; α, P = 0.005 (TSST-1 versus medium). (B) ***, P = 0.0001; β, P = 0.0003 (TSST-1 versus medium). (A and B) TSST-1 versus LPS plus TSST-1 is not significant (ns).

FIG 3

TSST-1 disrupts iHAEC monolayer barrier function. (A and B) Leakage of fluorescent tracers across a confluent iHAEC monolayer stimulated for 24 h with TSST-1 (25 μg/ml) or LPS (1 ng/ml), individually and in combination: Lucifer yellow (low-molecular-weight tracer) (A) and BSA-Alexa Fluor 633 (high-molecular-weight tracer) (B). EDTA (50 mM) was used as a positive control. (C) In-Cell Western assay total protein detection of VE-cadherin and ZO-1 in confluent iHAEC monolayers stimulated for 24 h with TSST-1 (25 μg/ml) or LPS (1 ng/ml), individually and in combination. P values were determined by one-way ANOVA (asterisks) with Holm-Sidak's multiple-comparison test (Greek letters, adjusted P value). (A) ****, P < 0.0001; α, P = 0.001 (TSST-1 versus medium); β, P = 0.04 (TSST-1 versus LPS plus TSST-1). (B) ***, P = 0.0001; α, P = 0.001 (TSST-1 versus medium); β, P = 0.05 (TSST-1 versus LPS plus TSST-1). (C) VE-cadherin: *, P = 0.01; α, P = 0.05 (TSST-1 versus medium). ZO-1: ns, not significant (TSST-1 versus medium or versus LPS plus TSST-1).

FIG 4

TSST-1 impedes iHAEC monolayer wound closure. Shown are results of an in vitro wound healing assay with immortalized HAECs stimulated with TSST-1 and/or LPS for 19 h. (A) Representative phase-contrast images of scratched iHAEC monolayers with or without TSST-1 (12.5 μg/ml) at 0 h (top) and 19 h (bottom). (B) Quantification of the scratched area in monolayers treated with increasing concentrations of TSST-1 or LPS. P values were determined by one-way ANOVA. *, P = 0.04. (C) Quantification of the scratched area in monolayers treated with TSST-1 (12.5 μg/ml) or LPS (5 ng/ml), individually and in combination. P values were determined by one-way ANOVA (asterisks) with Holm-Sidak's multiple-comparison test (Greek letters, adjusted P value). **, P = 0.009; α, P = 0.04 (TSST-1 versus medium); ns, not significant (TSST-1 versus LPS plus TSST-1).

FIG 5

TSST-1 stimulation of iHAECs does not induce IL-8 secretion. (A) iHAECs stimulated with increasing concentrations of LPS (0.1 to 50 ng/ml) for 24 h. P value was determined by one-way ANOVA. ****, P < 0.0001. (B) iHAECs stimulated with TSST-1 (25 μg/ml) or LPS (2.5 ng/ml) for 4 to 72 h. One-way ANOVA linear trend, P < 0.0001. (C) iHAECs stimulated with increasing concentrations of TSST-1 (6 to 100 μg/ml) for 24 h. (D) iHAECs stimulated for 24 h with TSST-1 (3, 6, or 25 μg/ml) alone and in the presence of LPS (0.5 ng/ml). P values were determined by one-way ANOVA with Holm-Sidak's multiple-comparison test (Greek letters, adjusted P value). α, P < 0.0001 (LPS versus medium); β, P < 0.01 (LPS versus LPS plus TSST-1).

FIG 6

TSST-1 suppresses IL-8 responses to LPS in iHAECs. (A) iHAECs stimulated for 24 h with increasing concentrations of LPS (0.005 to 2.5 ng/ml) in the presence of TSST-1 (3 μg/ml). P values were determined by two-way ANOVA (asterisks). **, P = 0.002 (LPS versus LPS plus TSST-1 across all concentrations). (B) Percent suppression of LPS-mediated IL-8 secretion resulting from concomitant stimulation with TSST-1. P values were determined by one-way ANOVA. ****, P = 0.002.

FIG 7

TSST-1 IL-8 suppression is independent of MHC-II binding. (A) In-Cell Western assay detection of MHC-II in iHAECs treated with TSST-1 (25 μg/ml) or LPS (1 ng/ml), individually and in combination for 24 h. Data are reported as fluorescence intensity normalized to that of medium-treated cells. P values were determined by one-way ANOVA (asterisks) with Holm-Sidak's multiple-comparison test (Greek letters, adjusted P value). ***, P < 0.0003; α, P = 0.0009 (TSST-1 versus medium); β, P = 0.03 (LPS versus medium); ns, not significant (TSST-1 versus LPS plus TSST-1). (B) iHAECs treated with an MHC-II-blocking antibody or isotype control antibody (10 μg/ml) for 30 min prior to stimulation with TSST-1 (10 μg/ml) and/or LPS (0.5 ng/ml) for 24 h also in the presence or absence of antibodies. Significance was determined by one-way ANOVA. (C) iHAECs stimulated for 24 h with His tag-purified toxoid (toxoid-His; MHC-II and TCR binding site inactivated), control His tag-purified TSST-1 (TSST-1–His), and native TSST-1 at 25 μg/ml each. (D) iHAECs stimulated with toxoid-His or TSST-1–His (2.4 μg/ml each) in the presence of LPS or LPS alone (1 ng/ml). Statistical analysis was done by one-way ANOVA with Holm-Sidak's multiple-comparison test. ***, P < 0.0001.

FIG 8

TNF-α and IL-1β pretreatment does not rescue iHAECs from TSST-1-mediated IL-8 suppression during simultaneous LPS stimulation. (A) iHAECs treated for 24 h with IL-2, IFN-γ, TNF-α, or IL-1β (10 ng/ml) or IL-6 (1 ng/ml) before stimulation with TSST-1 (12.5 μg/ml) or LPS (0.5 ng/ml), individually and in combination for an additional 24 h. (B to E) iHAECs treated for 24 h with increasing concentrations of IL-2, IFN-γ, TNF-α, or IL-1β (0.01 to 10 ng/ml) or IL-6 before stimulation with TSST-1 (12.5 μg/ml) or LPS (0.5 ng/ml), individually and in combination for an additional 24 h (all concentrations of IL-6 are 1 order of magnitude lower than those of other cytokines: “0.01” on the x axis = 0.001 ng/ml of IL-6; “10” on the x axis = 1 ng/ml of IL-6). iHAECs' baseline production of IL-8 under all conditions (2 ng/ml) is indicated by a horizontal line. (B) Cytokine treatment alone (medium-only control). (C) Cytokine-pretreated iHAECs stimulated with TSST-1. (D) Cytokine-pretreated iHAECs stimulated with LPS. (E) Cytokine-pretreated iHAECs stimulated with LPS plus TSST-1. P values were determined by one-way ANOVA with Holm-Sidak's multiple-comparison test (asterisks, adjusted P value for LPS versus LPS plus TSST-1; Greek letters, adjusted P value for cytokine only versus medium alone when significant). No cytokine, *, P = 0.03; IL-6, ****, P < 0.0001; IL-2 and IFN-γ, not significant (ns); TNF-α, *, P = 0.01; IL-1β, ****, P < 0.0001. α (TNF-α), P = 0.05, β (IL-1β), P < 0.0001.