Staphylococcal Superantigens Stimulate Epithelial Cells through CD40 To Produce Chemokines

Abstract

Mucosal and skin tissues form barriers to infection by most bacterial pathogens. Staphylococcus aureus causes diseases across these barriers in part dependent on the proinflammatory properties of superantigens. We showed, through use of a CRISPR-Cas9 CD40 knockout, that the superantigens toxic shock syndrome toxin 1 (TSST-1) and staphylococcal enterotoxins (SEs) B and C stimulated chemokine production from human vaginal epithelial cells (HVECs) through human CD40. This response was enhanced by addition of antibodies against CD40 through an unknown mechanism. TSST-1 was better able to stimulate chemokine (IL-8 and MIP-3α) production by HVECs than SEB and SEC, suggesting this is the reason for TSST-1's exclusive association with menstrual TSS. A mutant of TSST-1, K121A, caused TSS in a rabbit model when administered vaginally but not intravenously, emphasizing the importance of the local vaginal environment. Collectively, our data suggested that superantigens facilitate infections by disruption of mucosal barriers through their binding to CD40, with subsequent expression of chemokines. The chemokines facilitate TSS and possibly other epithelial conditions after attraction of the adaptive immune system to the local environment.IMPORTANCE Menstrual toxic shock syndrome (TSS) is a serious infectious disease associated with vaginal colonization by Staphylococcus aureus producing the exotoxin TSS toxin 1 (TSST-1). We show that menstrual TSS occurs after TSST-1 interaction with an immune costimulatory molecule called CD40 on the surface of vaginal epithelial cells. Other related toxins, where the entire family is called the superantigen family, bind to CD40, but not with a high-enough apparent affinity to cause TSS; thus, TSST-1 is the only exotoxin superantigen associated. Once the epithelial cells become activated by TSST-1, they produce soluble molecules referred to as chemokines, which in turn facilitate TSST-1 activation of T lymphocytes and macrophages to cause the symptoms of TSS. Identification of small-molecule inhibitors of the interaction of TSST-1 with CD40 may be useful so that they may serve as additives to medical devices, such as tampons and menstrual cups, to reduce the incidence of menstrual TSS.

Keywords: CD40; Staphylococcus aureus; chemokines; superantigens; toxic shock syndrome toxin.

FIG 1

Structure as determined by PyMOL (RCSB Protein Data Bank 3TSS) of TSST-1 according to the standard front and rear views (A) and amino acid residues comprising the dodecapeptide region of SAgs (B). Residues involved in MHC II binding are shown in yellow (not all contact residues included), residues involved in Vβ2-TCR binding are shown in cyan (does not include all residues), and the location of residue Lys 121 is in magenta. The dodecapeptide regions for SEB, SEC, and TSST-1 are shown with a number representing the amino acid residue immediately preceding the dodecapeptide.

FIG 2

Lethality of TSST-1 and a K121A mutant of TSST-1, where Lys at position 121 was changed to Ala, as tested in two TSS models. The number of rabbits alive out of a total of 6 is shown after i.v. (10 or 20 µg/kg) or intravaginal (Intra-Vag; 10 µg/kg) administration. By Fisher’s exact probability test, P < 0.001 for all comparisons to rabbits administered the K121A mutant i.v.

FIG 3

Chemokine (IL-8 and MIP-3α) production by human vaginal epithelial cells after 6 h of incubation in response to various concentrations of the SAgs TSST-1, SEB, and SEC. Bars indicate SD. Chemokines quantified by ELISA. At 1 and 10 µg/ml, TSST-1 induced greater chemokine production than SEs B and C (P < 0.001).

FIG 4

Flow cytometry of (A) human vaginal epithelial cells (HVECs) stained with antibodies to CD40 (phycoerythrin [PE] labeled) or an isotype-matched irrelevant antibody, (B) CD40-deficient HVECs through CRISPR-Cas9 knockout stained with FITC-labeled antibodies to CD40, and (C) deficient HVECs reconstituted with CD40 on a plasmid as stained with FITC-labeled antibodies or isotype-matched irrelevant antibodies.

FIG 5

Chemokine (IL-8) production by human vaginal epithelial cells after 6-h exposure to TSST-1 (10 µg), TSST-1 (10 µg plus 20 µl of antibodies [Ab] against CD40), SEB (10 µg), SEB (10 µg plus 20 µl of antibodies against CD40), 20 µl of antibodies alone, or cells alone in keratinocyte serum-free medium (KSFM). Bars indicate SD. All responses in the presence of SAg with or without antibodies are significantly greater than HVECs in KSFM alone or KSFM plus antibodies (P < 0.001). Additionally, the responses of SAg alone compared to SAg plus antibodies were significantly different (P < 0.001). Finally, responses to TSST-1 with or without antibodies were greater than SEB with or without antibodies (P < 0.001).

FIG 6

Chemokine (IL-8) production by human vaginal epithelial cells (HVECs) or HVECs knocked out for CD40 after 6-h exposure to SEB (100 µg), SEB (100 µg plus 20 µl of antibodies [Ab] against CD40), 20 µl of antibodies alone, or cells alone in KSFM. Bars indicate standard SD. All responses in the presence of SAg with or without antibodies are significantly greater than HVECs in KSFM alone or KSFM plus antibodies (P < 0.001). Additionally, the responses of SAg alone compared to SAg plus antibodies were significantly different (P < 0.001). Finally, responses of HVECs to SEB with or without antibodies were greater than HVECs knocked out for CD40 with or without antibodies (P < 0.001).

FIG 7

Chemokine (IL-8) production by human vaginal epithelial cells (HVECs), HVECs knocked out for CD40, or HVECs complemented with CD40 after 6-h exposure to TSST-1 (100 µg), TSST-1 (100 µg plus 20 µl of antibodies [Ab] against CD40), 20 µl of antibodies alone, or cells alone in KSFM. Bars indicate SD. All responses in the presence of SAg with or without antibodies are significantly greater than cells in KSFM alone or KSFM plus antibodies (P < 0.001), except for HVECs knocked out for CD40. Additionally, the responses of SAg alone compared to SAg plus antibodies were significantly different (P < 0.001), except for HVECs knocked out for CD40. Finally, responses of cells to TSST-1 with or without antibodies were greater than HVECs knocked out for CD40 with or without antibodies (P < 0.001).

FIG 8

Chemokine (IL-8) production by human vaginal epithelial cells after 6-h exposure in KSFM to the K121A mutant of TSST-1 (10 and 100 µg), K121A mutant (100 µg plus 20 µl of antibodies [Ab] against CD40), TSST-1 wild type (10 and 100 µg), or TSST-1 (100 µg plus 20 µl of antibodies [Ab] against CD40). Bars indicate SD. All responses in the presence of SAg with or without antibodies when matched for doses are not significantly different.