Staphylococcus aureus Isolated from Skin from Atopic-Dermatitis Patients Produces Staphylococcal Enterotoxin Y, Which Predominantly Induces T-Cell Receptor Vα-Specific Expansion of T Cells

Abstract
While investigating the virulence traits of Staphylococcus aureus adhering to the skin of atopic-dermatitis (AD) patients, we identified a novel open reading frame (ORF) with structural similarity to a superantigen from genome sequence data of an isolate from AD skin. Concurrently, the same ORF was identified in a bovine isolate of S. aureus and designated SElY (H. K. Ono, Y. Sato’o, K. Narita, I.


Staphylococcus aureus Isolated from Skin from Atopic-Dermatitis Patients Produces Staphylococcal Enterotoxin Y, Which Predominantly Induces T-Cell Receptor Vα-Specific Expansion of T Cells
Infect Immun . 2020 Feb; 88(2): e00360-19. Published online 2020 Jan 22. Prepublished online 2019 Nov 18. doi: 10.1128/IAI.00360-19 PMCID: PMC6977126 PMID: 31740530 Fatkhanuddin Aziz , a, b Junzo Hisatsune , a, i Liansheng Yu , a, i Junko Kajimura , e Yusuke Sato’o , c Hisaya K. Ono , d Kanako Masuda , a Mika Yamaoka , e Siti Isrina Oktavia Salasia , f Akio Nakane , g Hiroki Ohge , h Yoichiro Kusunoki , e and Motoyuki Sugai a, i Fatkhanuddin Aziz a Department of Bacteriology, Graduate School of Biomedical and Health Sciences, Hiroshima University, Hiroshima, Japan b Animal Health Study Program, Department of Bioresources Technology and Veterinary, Vocational College, University of Gadjah Mada, Yogyakarta, Indonesia Find articles by Fatkhanuddin Aziz Junzo Hisatsune a Department of Bacteriology, Graduate School of Biomedical and Health Sciences, Hiroshima University, Hiroshima, Japan i Antimicrobial Resistance Research Center, National Institute of Infectious Diseases, Tokyo, Japan Find articles by Junzo Hisatsune Liansheng Yu a Department of Bacteriology, Graduate School of Biomedical and Health Sciences, Hiroshima University, Hiroshima, Japan i Antimicrobial Resistance Research Center, National Institute of Infectious Diseases, Tokyo, Japan Find articles by Liansheng Yu Junko Kajimura e Department of Molecular Biosciences, Radiation Effects Research Foundation, Hiroshima, Japan Find articles by Junko Kajimura Yusuke Sato’o c Division of Bacteriology, Department of Infection and Immunity, School of Medicine, Jichi Medical University, Shimotsuke, Tochigi, Japan Find articles by Yusuke Sato’o Hisaya K. Ono d Laboratory of Zoonoses, Kitasato University School of Veterinary Medicine, Towada, Aomori, Japan Find articles by Hisaya K. Ono Kanako Masuda a Department of Bacteriology, Graduate School of Biomedical and Health Sciences, Hiroshima University, Hiroshima, Japan Find articles by Kanako Masuda Mika Yamaoka e Department of Molecular Biosciences, Radiation Effects Research Foundation, Hiroshima, Japan Find articles by Mika Yamaoka Siti Isrina Oktavia Salasia f Department of Clinical Pathology, Faculty of Veterinary Medicine, University of Gadjah Mada, Yogyakarta, Indonesia Find articles by Siti Isrina Oktavia Salasia Akio Nakane g Department of Microbiology and Immunology, Hirosaki University, Aomori, Japan Find articles by Akio Nakane Hiroki Ohge h Department of Infectious Diseases, Hiroshima University Hospital, Hirsohima, Japan Find articles by Hiroki Ohge Yoichiro Kusunoki e Department of Molecular Biosciences, Radiation Effects Research Foundation, Hiroshima, Japan Find articles by Yoichiro Kusunoki Motoyuki Sugai a Department of Bacteriology, Graduate School of Biomedical and Health Sciences, Hiroshima University, Hiroshima, Japan i Antimicrobial Resistance Research Center, National Institute of Infectious Diseases, Tokyo, Japan Find articles by Motoyuki Sugai Victor J. Torres, Editor Victor J. Torres, New York University School of Medicine; Author information Article notes Copyright and License information Disclaimer a Department of Bacteriology, Graduate School of Biomedical and Health Sciences, Hiroshima University, Hiroshima, Japan b Animal Health Study Program, Department of Bioresources Technology and Veterinary, Vocational College, University of Gadjah Mada, Yogyakarta, Indonesia c Division of Bacteriology, Department of Infection and Immunity, School of Medicine, Jichi Medical University, Shimotsuke, Tochigi, Japan d Laboratory of Zoonoses, Kitasato University School of Veterinary Medicine, Towada, Aomori, Japan e Department of Molecular Biosciences, Radiation Effects Research Foundation, Hiroshima, Japan f Department of Clinical Pathology, Faculty of Veterinary Medicine, University of Gadjah Mada, Yogyakarta, Indonesia g Department of Microbiology and Immunology, Hirosaki University, Aomori, Japan h Department of Infectious Diseases, Hiroshima University Hospital, Hirsohima, Japan i Antimicrobial Resistance Research Center, National Institute of Infectious Diseases, Tokyo, Japan Corresponding author. Address correspondence to Motoyuki Sugai, pj.og.diin@iagus . Citation Aziz F, Hisatsune J, Yu L, Kajimura J, Sato’o Y, Ono HK, Masuda K, Yamaoka M, Salasia SIO, Nakane A, Ohge H, Kusunoki Y, Sugai M. 2020. Staphylococcus aureus isolated from skin from atopic-dermatitis patients produces staphylococcal enterotoxin Y, which predominantly induces T-cell receptor Vα-specific expansion of T cells. Infect Immun 88:e00360-19. https://doi.org/10.1128/IAI.00360-19 . Received 2019 May 7; Revisions requested 2019 Jun 23; Accepted 2019 Oct 29. Copyright © 2020 American Society for Microbiology. All Rights Reserved . This article has been cited by other articles in PMC.


Associated Data
Supplementary Materials Supplemental file 1 IAI.00360-19-s0001.pdf (279K) GUID: 424797F2-6852-4733-A0FB-99F47A32668A


ABSTRACT
While investigating the virulence traits of Staphylococcus aureus adhering to the skin of atopic-dermatitis (AD) patients, we identified a novel open reading frame (ORF) with structural similarity to a superantigen from genome sequence data of an isolate from AD skin. Concurrently, the same ORF was identified in a bovine isolate of S. aureus and designated SElY (H. K. Ono, Y. Sato’o, K. Narita, I. Naito, et al., Appl Environ Microbiol 81:7034–7040, 2015, https://doi.org/10.1128/AEM.01873-15 ). Recombinant SElY bov had superantigen activity in human peripheral blood mononuclear cells. It further demonstrated emetic activity in a primate animal model, and it was proposed that SElY be renamed SEY (H. K. Ono, S. Hirose, K. Narita, M. Sugiyama, et al., PLoS Pathog 15:e1007803, 2019, https://doi.org/10.1371/journal.ppat.1007803 ). Here, we investigated the prevalence of the sey gene in 270 human clinical isolates of various origins in Japan. Forty-two strains were positive for the sey gene, and the positive isolates were from patients with the skin diseases atopic dermatitis and impetigo/staphylococcal scalded skin syndrome (SSSS), with a detection rate of ∼17 to 22%. There were three variants of SEY (SEY 1 , SEY 2 , and SEY 3 ), and isolates producing SEY variants formed three distinct clusters corresponding to clonal complexes (CCs) 121, 59, and 20, respectively. Most sey + isolates produced SEY in broth culture. Unlike SEY bov , the three recombinant SEY variants exhibited stability against heat treatment. SEY predominantly activated human T cells with a particular T-cell receptor (TCR) Vα profile, a unique observation since most staphylococcal enterotoxins exert their superantigenic activities through activating T cells with specific TCR Vβ profiles. SEY may act to induce localized inflammation via skin-resident T-cell activation, facilitating the pathogenesis of S. aureus infection in disrupted epithelial barriers. KEYWORDS: Staphylococcus aureus , atopic dermatitis, enterotoxins, superantigens


KEYWORDS: Staphylococcus aureus , atopic dermatitis, enterotoxins, superantigens


INTRODUCTION
Staphylococcus aureus is a ubiquitous Gram-positive bacterium, known worldwide to cause a variety of infectious diseases ranging from skin infections to fatal systemic infections in humans and animals. Staphylococcal enterotoxins (SEs) are members of a protein family of more than 20 different staphylococcal exotoxins, sharing several biological activities and structural features ( 1 , 2 ). These bacterial proteins are known to have pyrogenic, superantigenic, and emetic activities. Some of these proteins are implicated in toxic shock syndrome ( 3 ) and are causative agents of staphylococcal food poisoning ( 4 , 5 ). Many studies have shown significant associations of SEs with human diseases ( 6 , 7 ). The mechanisms of action of SEs in the toxicity of staphylococcal infections and exacerbation of S. aureus pathogenicity have been documented in the context of classical SEs such as SEA-SEE, SEH, and toxic shock syndrome toxin (TSST) ( 6 , 8 , – 11 ). Most of the SEs interact with the specific variable regions of T-cell receptor β chains (TCR Vβ), leading to cytokine production and mitogenic responses in T cells that orchestrate the onset of inflammation ( 2 , 12 ). Superantigens contribute to skin disorders by skin homing of T cells and mediate the regulation of the immune response. Superantigens trigger hosts to produce specific immunoglobulin E (IgE). IgE-relevant superantigens thus stimulate mast cell and basophil degranulation, resulting in the release of mediators such as cytokines and chemokines, eventually promoting skin rashes and itch manifestation in dermatitis patients ( 7 , 13 , 14 ). While performing genomic characterization of S. aureus isolated from the skin of atopic-dermatitis (AD) patients for a unique virulence factor(s), we identified a novel open reading frame (ORF) exhibiting amino acid sequence similarity to superantigens. Concurrently, the ORF was identified by Ono et al. as a new enterotoxin-like protein from a bovine isolate of S. aureus designated SElY ( 15 ). Therefore, SElY was renamed SEY after the demonstration of the induction of vomiting activity in the common marmoset, a newly established primate model for emetic activity tests ( 16 ). Interestingly, this SEY showed very weak superantigenic activity in mouse splenocytes even at 10 μg/ml, unlike other enterotoxins, such as SEA ( 17 ) and SEB ( 18 ), that induce mouse T-cell proliferation dose dependently at nanogram concentrations. Here, we investigated the prevalence of the sey gene in human clinical isolates of diverse origins and demonstrated that ∼17 to 22% of S. aureus isolates from patients with atopic dermatitis or impetigo/SSSS possess the sey gene. Characterization of sey gene products of the isolates identified three SEY variants (SEY 1 , SEY 2 , and SEY 3 ), distinct from bovine SEY (SEY bov ) ( 15 ), that are are produced by isolates belonging to three independent clonal complexes (CCs) (CC121, CC20, and CC59). Here, we report a novel superantigenic activity of SEY.


RESULTS
Characterization of sey + S. aureus isolates of human origin. Data regarding the prevalence of sey from 270 human clinical S. aureus isolates are shown in Table 1 . Overall, 42 isolates were positive for the sey gene by PCR, and around 17 to 22% of isolates from patients with atopic dermatitis, impetigo, and staphylococcal scalded skin syndrome (SSSS) were positive for sey , whereas isolates from patients with furunculosis and sepsis were very rare or not present. Genetic characterization of sey + isolates was performed, and the data are summarized in Table 2 . Multilocus sequence typing (MLST) and analysis of clonal complexes revealed that the sey gene was primarily found in isolates of CC121, CC20, and CC59. We noted that sey + isolates of CC121 were positive for eta and ednA genes, while CC20 possessed a series of se types of the enterotoxin gene cluster, the egc island, including seg , sei , sem , sen , and seo genes. Interestingly, all sey + isolates in CC59 harbored the classical enterotoxin seb but lacked et , luk , and edn genes. table ft1 table-wrap mode="anchored" t5 TABLE 1 caption a7 Origin(s) (no. of strains) No. (%) of strains positive for the sey gene Atopy (124) 22 (17.7) Impetigo (62) 12 (19.3) SSSS (22) 5 (22) Furuncle (19) 0 Sepsis (31) 1 (3.1) Other (12) 2 (16.7) Total (270) 42 Open in a separate window Distribution of strains positive for the sey gene table ft1 table-wrap mode="anchored" t5 TABLE 2 caption a7 Origin ST CC Enterotoxin or TSST-1 gene(s) a Exfoliative toxin gene(s) Leukocidin gene(s) Other virulence factor gene(s) Presence of mecA Atopy 121 121 seg , sei , seIj , sem , sey 1 eta lukED ednA , scn , sak − Atopy 121 121 seg , sei , sen , sey 1 eta lukED ednA , scn , sak − Atopy 120 121 seg , sei , sen , sey 1 eta lukED scn , sak − Atopy 121 121 seg , sen , seq , sey 1 lukED ednA − Atopy 121 121 seg , sei , sen , sey 1 eta lukED scn , sak − Atopy 121 121 seg , sei , sey 1 eta lukED scn , sak − Atopy 59 59 seb , sek , seq , sey 2 chp , scn , sak − Atopy 59 59 seb , sek , seq , sey 2 chp , scn , sak − Atopy 59 59 seb , sek , seq , sey 2 chp , scn , sak − Atopy 59 59 seb , sek , sey 2 chp , sak − Atopy 20 20 seb , seg , sei , sem , sen , seo , sey 3 lukED chp , scn , sak − Atopy 20 20 seb , seg , sei , sem , sen , seo , sey 3 lukE chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 lukED chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 lukED chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 lukED chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 lukED chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 lukE chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 lukE chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 lukE chp , scn , sak − Atopy 20 20 seg , sei , selj , sem , seo , sey 3 lukE ednA , chp , scn , sak − Atopy 20 20 seg , sei , sen , sey 3 lukE chp , scn , sak − Impetigo 120 121 seg , sei , sem , sen , sey 1 eta lukED ednA , scn , sak − Impetigo 121 121 seg , sei , sem , sen , sey 1 eta lukED ednA , scn , sak − Impetigo 121 121 seg , sei , sem , sen , sey 1 eta lukED ednA , scn , sak − Impetigo 121 121 seg , sei , sem , sen , sey 1 eta lukED ednA , scn , sak − Impetigo 121 121 seg , sei , sen , sey 1 eta , etb ednC , scn , sak + Impetigo 121 121 seg , sei , sey 1 eta lukED scn , sak − Impetigo 121 121 seg , sei , sey 1 eta lukED ednA , scn , sak − Impetigo 121 121 seg , sei , sey 1 eta lukED ednA , scn , sak − Impetigo 121 121 seg , sei , sey 1 eta lukED scn , sak − Impetigo 121 121 seg , sei , sey 1 eta lukED scn , sak − Impetigo 20 20 seg , sei , sem , sen , seo , sey 3 lukED chp , scn , sak − Impetigo 20 20 seg , sei , sem , sen , seo , sey 3 lukED chp , scn , sak − SSSS 121 121 seg , sei , sem , sen , sey 1 eta lukED ednA , scn , sak − SSSS 121 121 seg , sei , sen , sey 1 lukED scn , sak − SSSS 121 121 seg , sei , sen , sey 1 eta lukED scn , sak − SSSS 121 121 seg , sei , sen , sey 1 eta lukED scn , sak − SSSS 121 121 seg , sei , sen , sey 1 eta lukED ednA , scn , sak − Sepsis 1358 121 seg , sei , sem , sen , sey 1 lukED scn , sak − Ulcer 121 121 seg , sei , sey 1 lukED ednA , scn , sak − Whitlow 20 20 seg , sei , sem , seo , sey 3 lukE chp , scn , sak − Open in a separate window a The sey gene is indicated in boldface type. Genetic characterization of 42 sey + S. aureus clinical isolates Comparison of deduced amino acid sequences of SEY revealed heterogeneity (see Table S1 in the supplemental material); there were at least three variants of sey gene products present in clinical isolates, and they were distinct from that of the bovine isolate ( 15 ). Therefore, we designated these gene products SEY 1 , SEY 2 , and SEY 3 , respectively, and that from bovine SEY bov . Comparison with other members of enterotoxins clearly demonstrated that they belonged to the SEY family, and the most similar relative was SET ( Fig. 1A ). This group represented a distinct group of amino acid sequence similarity from other enterotoxins. Moreover, as illustrated in Fig. 1B , there was a correlation between SEY variants and CC types. Strain JP074 was the first characterized isolate from an AD patient, and SEY 2 cloned from JP074 was selected for antiserum production, the establishment of an enzyme-linked immunosorbent assay (ELISA) system, and deep analysis of its biological activities in human peripheral blood mononuclear cells (PBMCs). fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 1 caption a7 Phylogenetic analysis of SEY variants and their association with clonal groups found in the present study. Multiple alignments of amino acid sequences were analyzed by using ClustalW. (A) Sequence database of enterotoxins assigned the indicated GenBank accession numbers. SEY variants and SEY bov are clustered into one subgroup (black dashed box) belonging to the SET/SEY group. (B) Tree showing that three SEY variants are clearly separated into three clusters, which coincide with the clonal complexes of the producing strains (colored boxes). (C) Schematic position of the sey gene on the chromosome. Three representative sey + isolates of different STs were compared to reference strain MW2. The sey gene (red arrow) is flanked by a nonmobile genetic element gene cluster. SEY variants share antigenicity and are produced in culture supernatants. To clarify the cross-reactivity among SEY variants, recombinant proteins of SEY 1 , SEY 2 , and SEY 3 were expressed in Escherichia coli BL21(DE3). They showed almost identical molecular sizes of approximately 23.5 kDa in a 12% polyacrylamide gel ( Fig. 2A , top). Western blot analysis using rabbit antisera against SEY 2 revealed that SEY 2 shared antigenicity with both SEY 1 and SEY 3 ( Fig. 2A , bottom). Subsequently, we focused on detecting SEY protein in culture supernatant samples from selected representative strains of major CCs found in this study ( Table 3 ). Signals of the same electrophoretic mobility were detected in the culture supernatants of isolates of CC121, CC59, and CC20 ( Fig. 2B ). fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 2 caption a7 Confirmation of purified recombinant proteins of the SEY subtype by SDS-PAGE and immunoblot assays. (A, top) Coomassie blue-stained SDS-PAGE gel with an eluted protein sample from Co 2+ affinity chromatography. (Bottom) Fifty nanograms of each SEY subtype per well from a 12% polyacrylamide gel was transferred onto a nitrocellulose membrane, and the signal was detected using immunized rabbit anti-SEY 2 sera. (B) SEY production in the culture supernatant (arrowhead). Representatives of S. aureus isolates from major CCs positive for the sey gene were cultured in 60 ml of BHI broth supplemented with 1% yeast extract for 24 h. Samples were treated by TCA-acetone precipitation and subjected to SDS-PAGE and Western blotting. MW, molecular weight. table ft1 table-wrap mode="anchored" t5 TABLE 3 caption a7 Strain, plasmid, or primer Characteristic(s) or sequence a Source or reference Strains E.coli DH5α Cloning strain TaKaRa BL21(DE3) Host expression for recombinant protein Novagen S. aureus JP087 sey + ; ST121, CC121, and sey 1 ; atopic dermatitis This study JP074 sey + ; ST59, CC59, and sey 2 ; atopic dermatitis This study JP096 sey + ; ST20, CC20, and sey 3 ; atopic dermatitis This study NCTC 8325 sey negative; reference strain Sugai lab Plasmids pFAY 1 Amp r ; pET22b(+) with sey 1 This study pFAY 2 Amp r ; pET22b(+) with sey 2 This study pFAY 3 Amp r ; pET22b(+) with sey 3 This study Primers sey detection sey F TCTATTGGAATAGCAGAAGTA This study sey R CAATATGTCGCCTAAATCTAT This study sey cloning sey F1 GGAATTC CATATG CACCACCACCACCACCACAAAACAACTGGATTGATTACAG This study sey R2 TTGAC GAATTC TATGTTGGAACGAC This study Open in a separate window a Restriction cut sites are in boldface type. Strains, plasmids, and primers used in this study Quantification of SEY production by a sandwich ELISA. We developed an ELISA system to quantify SEY production of S. aureus isolates harboring sey using purified polyclonal anti-SEY 2 antibodies ( Fig. 3A ). The specificity of the sandwich ELISA was confirmed by the lack of cross-reactivity with other enterotoxins ( Fig. 3B ). When cultured in brain heart infusion (BHI) broth supplemented with 1% yeast extract medium (BHIY) at 37°C for 24 h with constant agitation, most strains of major CCs were positive for SEY in vitro above the detectable level, while only one of the isolates did not produce SEY in each of the CC20 and CC121 groups. Of note, strains belonging to CC20 were detected in the range of 8 to 57 ng/ml, whereas in the CC121 group, the level of production ranged from 3 to 178 ng/ml ( Fig. 4 ). fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 3 caption a7 Establishment of the SEY ELISA. (A) Straight-line plot of SEY concentrations of 0.2 to 5.0 ng/ml for the standard curve. The detection limit for SEY by the ELISA was 0.2 ng/ml. The coefficient of determination ( R 2 ) values for linear regression are indicated. (B) Cross-reaction test of anti-SEY serum against other selected SEs using an ELISA. One hundred nanograms of purified enterotoxins (5 ng for SEY) or no SE was loaded into a microplate well. The cutoff value for the SEY ELISA was 0.3 ng/ml; values of <0.3 ng/ml were considered undetectable. (C) Growth of S. aureus isolates, starting at an OD at 660 nm of 0.01, from triplicate measurements. (D) Production of SEY from selected strains as representatives of major CCs in this study, measured by a sandwich ELISA. Each isolate or sample was assayed in three biological or technical replicates, respectively, and each assay was performed twice. Error bars represent standard errors of the means (SEM) of test results. fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 4 caption a7 SEY production by S. aureus isolates of various CCs. SEY production by sey + strains of major CCs cultured in BHIY medium was measured by a sandwich ELISA. sey -negative strain NCTC 8325 was used as a negative control. Each isolate was assayed in duplicate in each assay. Bars represent the means of the data set. ND, not determined. SEY variants from human isolates are unstable against heat treatment. SEY bov was reported to be unstable against heat and enzyme treatments ( 15 ). We investigated whether SEY variants have characteristics similar to those of SEY bov . As shown in Fig. 5A , SDS patterns after heat treatment of SEY variants were more stable than those of SEB, suggesting that SEY variants from human isolates are more stable than SEY bov . Unlike heat treatment, SEY variants were unstable against digestive enzyme treatment, like SEY bov ( 15 ). Pepsin-treated SEY variants showed signs of digestion at 30 min ( Fig. 5B ). Similarly, apparent digestion by trypsin treatment was observed in just 30 min ( Fig. 5C ). SEB was resistant to heat and pepsin treatments but susceptible to trypsin digestion ( 5 , 19 ). These results suggested that SEY variants are stable against heat treatment but susceptible to digestive enzyme treatment. fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 5 caption a7 Stability test of SEY variants against heat and digestive enzyme treatments. SEY variants, SEB, and BSA (100 μg/ml) (depicted by arrows) were heated at 100°C (A), incubated with pepsin (100 μg/ml) (B), or incubated with trypsin (50 μg/ml) (C). After the indicated times, samples were collected and analyzed by SDS-PAGE. NT, no treatment; P, pepsin; T, trypsin. T-cell proliferation and cytokine production in human PBMCs stimulated with SEY variants. In mouse splenocytes, any of the variants, SEY 1 , SEY 2 , or SEY 3 , failed to induce a significant level of splenocyte proliferation after 48 h of stimulation at any concentration ranging from 1 pg/ml to 10 μg/ml, whereas cells treated with SEA showed a dose-dependent increase in the level of proliferation ( Fig. 6 ). Next, we investigated human T-cell responses to the SEY variants by flow cytometry. Substantial cell divisions were observed in both CD4 + and CD8 + T-cell populations when PBMCs were stimulated with SEB or any of the SEY variants ( Fig. 7A ). Although the dividing cell fraction was higher in the PBMCs cultured with SEY 3 than in those cultured with SEY 1 or SEY 2 , the difference in the mean of percentages of dividing CD4 + or CD8 + T cells was not statistically significant ( Fig. 7B and ​ andC). C ). In addition, we measured the levels of tumor necrosis factor alpha (TNF-α) and gamma interferon (IFN-γ) in culture supernatants from human PBMCs stimulated with the variants ( Fig. 7D and ​ andE). E ). Similar to other SEs, all the SEY variants induced the production of TNF-α and IFN-γ as well as the proliferation of CD4 + and CD8 + T cells. Although the effects on human lymphocytes slightly differ among the SEY variants, none of the SEY variants were as potent as SEB, which is known to act as a superantigen in both mice and humans. fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 6 caption a7 Mitogenic activity of SEY variants in mouse splenocytes. Splenocytes were stimulated with several concentrations of SEA, SEY 1 , SEY 2 , and SEY 3 for 48 h. Splenocyte proliferation was assessed using CCK-8 assays. Each bar shows the mean and SEM for triplicate assays from a representative experiment. These data are representative of results from three independent experiments. fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 7 caption a7 Percentages of proliferating CD4 + and CD8 + T cells and cytokine production from human PBMCs following stimulation with SEY variants. SEB was used as the positive control. (A to C) Flow cytogram data from one representative experiment (A) and percentages of proliferating CD4 + (B) and CD8 + T (C) cells indicated as the mean values and SEM of data for three donors. (D and E) Mean TNF-α (D) and IFN-γ (E) levels and SEM of data for 4 donors. Statistical analysis among SEY variants or between SEY variants and SEB was performed using analysis of variance (ANOVA) followed by a Tukey test (*, P < 0.05; **, P < 0.01; ***, P < 0.001; NS, not significant). Human TCR expression profiles following stimulation with SEY 2 . Using a panel of monoclonal antibodies (MAbs) specific to TCR Vβ subfamilies, we investigated TCR Vβ profiles in CD4 + T cells that proliferated in response to stimulation of PBMCs with 10 ng/ml of SEY 2 or SEB. The distribution of TCR Vβ expression in CD4 + T cells from five healthy donors indicated that there was no clear enhancement in the expression of certain Vβ subfamilies ( Fig. 8A ). In addition, neither SEY 1 nor SEY 3 activated specifically different Vβ subfamilies (Fig. S1). To evaluate alterations in TCR repertoire usage following SEY 2 stimulation more extensively and precisely, we conducted deep sequencing of TCR Vα (TRAV) and Vβ (TRBV) gene transcripts in T cells cultured with the toxins. As shown in Fig. 8B , a remarkable enhancement of TCR Vα transcription in CD4 + T cells following SEY 2 stimulation was observed for TRAV 8.2 and 8.6 in three healthy individuals examined, with frequencies of 24% and 33%, respectively. On the other hand, there was no remarkable change in the frequency of any TRBV gene transcripts in SEY 2 -stimulated CD4 + T cells ( Fig. 8C ). In accord with previous reports ( 2 , 20 , 21 ) and the present flow cytometry results, Vβ subfamilies 3, 12, 14, 17, and 20 were specifically enriched by SEB stimulation ( Fig. 8A and ​ andC). C ). In addition, enhanced transcriptions of Vβ 6 and 15, which were not analyzed by flow cytometry due to the unavailability of specific antibodies to these Vβ subfamilies, were observed in SEB-stimulated CD4 + T cells ( Fig. 8A ), while there was no particular Vα enhancement in these cells ( Fig. 8B ). fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 8 caption a7 (A) TCR Vα/β profiles of CD4 + T cells after in vitro stimulation of human PBMCs with 10 ng/ml of SEY 2 or SEB. After 5 days of incubation, the cells were stained with anti-CD4 MAb and a panel of MAbs to distinct TCR Vβ subfamily segments. Data represent the mean values and SEM of data for five healthy donors. (B and C) Transcripts from TRAV (B) and TRBV (C) genes were evaluated by TCR sequencing in sorted CD4 + T cells. The mean frequencies of V gene transcripts and SEM of data for three healthy donors are indicated as bars. (D) Chord diagrams depicting V-J segment combinations in TRAV and TRBV genes in CD4 + T cells following PBS, SEY, or SEB stimulation, where clonotype distribution maps were generated by VDJtools. The relative abundance of each V-J segment combination is represented as a proportion of the colored chord. Preferential usage of a certain variable-joining (V-J) gene recombination has been shown in human viral infection ( 22 ). Using the TCR sequencing data, we therefore investigated whether there were any specific combinations between Vα and Jα or Vβ and Jβ segments in the stimulated CD4 + T cells. No certain pattern was seen in J gene segment usages among TCR Vα or Vβ repertoires in CD4 + T cells stimulated with SEY 2 or SEB ( Fig. 8D ). Overall, we demonstrate that SEY 2 acts as a superantigen that specifically stimulates human T cells bearing TCR Vα 8.2 and 8.6.


Characterization of sey + S. aureus isolates of human origin.
Data regarding the prevalence of sey from 270 human clinical S. aureus isolates are shown in Table 1 . Overall, 42 isolates were positive for the sey gene by PCR, and around 17 to 22% of isolates from patients with atopic dermatitis, impetigo, and staphylococcal scalded skin syndrome (SSSS) were positive for sey , whereas isolates from patients with furunculosis and sepsis were very rare or not present. Genetic characterization of sey + isolates was performed, and the data are summarized in Table 2 . Multilocus sequence typing (MLST) and analysis of clonal complexes revealed that the sey gene was primarily found in isolates of CC121, CC20, and CC59. We noted that sey + isolates of CC121 were positive for eta and ednA genes, while CC20 possessed a series of se types of the enterotoxin gene cluster, the egc island, including seg , sei , sem , sen , and seo genes. Interestingly, all sey + isolates in CC59 harbored the classical enterotoxin seb but lacked et , luk , and edn genes. table ft1 table-wrap mode="anchored" t5 TABLE 1 caption a7 Origin(s) (no. of strains) No. (%) of strains positive for the sey gene Atopy (124) 22 (17.7) Impetigo (62) 12 (19.3) SSSS (22) 5 (22) Furuncle (19) 0 Sepsis (31) 1 (3.1) Other (12) 2 (16.7) Total (270) 42 Open in a separate window Distribution of strains positive for the sey gene table ft1 table-wrap mode="anchored" t5 TABLE 2 caption a7 Origin ST CC Enterotoxin or TSST-1 gene(s) a Exfoliative toxin gene(s) Leukocidin gene(s) Other virulence factor gene(s) Presence of mecA Atopy 121 121 seg , sei , seIj , sem , sey 1 eta lukED ednA , scn , sak − Atopy 121 121 seg , sei , sen , sey 1 eta lukED ednA , scn , sak − Atopy 120 121 seg , sei , sen , sey 1 eta lukED scn , sak − Atopy 121 121 seg , sen , seq , sey 1 lukED ednA − Atopy 121 121 seg , sei , sen , sey 1 eta lukED scn , sak − Atopy 121 121 seg , sei , sey 1 eta lukED scn , sak − Atopy 59 59 seb , sek , seq , sey 2 chp , scn , sak − Atopy 59 59 seb , sek , seq , sey 2 chp , scn , sak − Atopy 59 59 seb , sek , seq , sey 2 chp , scn , sak − Atopy 59 59 seb , sek , sey 2 chp , sak − Atopy 20 20 seb , seg , sei , sem , sen , seo , sey 3 lukED chp , scn , sak − Atopy 20 20 seb , seg , sei , sem , sen , seo , sey 3 lukE chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 lukED chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 lukED chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 lukED chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 lukED chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 lukE chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 lukE chp , scn , sak − Atopy 20 20 seg , sei , sem , sen , seo , sey 3 lukE chp , scn , sak − Atopy 20 20 seg , sei , selj , sem , seo , sey 3 lukE ednA , chp , scn , sak − Atopy 20 20 seg , sei , sen , sey 3 lukE chp , scn , sak − Impetigo 120 121 seg , sei , sem , sen , sey 1 eta lukED ednA , scn , sak − Impetigo 121 121 seg , sei , sem , sen , sey 1 eta lukED ednA , scn , sak − Impetigo 121 121 seg , sei , sem , sen , sey 1 eta lukED ednA , scn , sak − Impetigo 121 121 seg , sei , sem , sen , sey 1 eta lukED ednA , scn , sak − Impetigo 121 121 seg , sei , sen , sey 1 eta , etb ednC , scn , sak + Impetigo 121 121 seg , sei , sey 1 eta lukED scn , sak − Impetigo 121 121 seg , sei , sey 1 eta lukED ednA , scn , sak − Impetigo 121 121 seg , sei , sey 1 eta lukED ednA , scn , sak − Impetigo 121 121 seg , sei , sey 1 eta lukED scn , sak − Impetigo 121 121 seg , sei , sey 1 eta lukED scn , sak − Impetigo 20 20 seg , sei , sem , sen , seo , sey 3 lukED chp , scn , sak − Impetigo 20 20 seg , sei , sem , sen , seo , sey 3 lukED chp , scn , sak − SSSS 121 121 seg , sei , sem , sen , sey 1 eta lukED ednA , scn , sak − SSSS 121 121 seg , sei , sen , sey 1 lukED scn , sak − SSSS 121 121 seg , sei , sen , sey 1 eta lukED scn , sak − SSSS 121 121 seg , sei , sen , sey 1 eta lukED scn , sak − SSSS 121 121 seg , sei , sen , sey 1 eta lukED ednA , scn , sak − Sepsis 1358 121 seg , sei , sem , sen , sey 1 lukED scn , sak − Ulcer 121 121 seg , sei , sey 1 lukED ednA , scn , sak − Whitlow 20 20 seg , sei , sem , seo , sey 3 lukE chp , scn , sak − Open in a separate window a The sey gene is indicated in boldface type. Genetic characterization of 42 sey + S. aureus clinical isolates Comparison of deduced amino acid sequences of SEY revealed heterogeneity (see Table S1 in the supplemental material); there were at least three variants of sey gene products present in clinical isolates, and they were distinct from that of the bovine isolate ( 15 ). Therefore, we designated these gene products SEY 1 , SEY 2 , and SEY 3 , respectively, and that from bovine SEY bov . Comparison with other members of enterotoxins clearly demonstrated that they belonged to the SEY family, and the most similar relative was SET ( Fig. 1A ). This group represented a distinct group of amino acid sequence similarity from other enterotoxins. Moreover, as illustrated in Fig. 1B , there was a correlation between SEY variants and CC types. Strain JP074 was the first characterized isolate from an AD patient, and SEY 2 cloned from JP074 was selected for antiserum production, the establishment of an enzyme-linked immunosorbent assay (ELISA) system, and deep analysis of its biological activities in human peripheral blood mononuclear cells (PBMCs). fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 1 caption a7 Phylogenetic analysis of SEY variants and their association with clonal groups found in the present study. Multiple alignments of amino acid sequences were analyzed by using ClustalW. (A) Sequence database of enterotoxins assigned the indicated GenBank accession numbers. SEY variants and SEY bov are clustered into one subgroup (black dashed box) belonging to the SET/SEY group. (B) Tree showing that three SEY variants are clearly separated into three clusters, which coincide with the clonal complexes of the producing strains (colored boxes). (C) Schematic position of the sey gene on the chromosome. Three representative sey + isolates of different STs were compared to reference strain MW2. The sey gene (red arrow) is flanked by a nonmobile genetic element gene cluster.


SEY variants share antigenicity and are produced in culture supernatants.
To clarify the cross-reactivity among SEY variants, recombinant proteins of SEY 1 , SEY 2 , and SEY 3 were expressed in Escherichia coli BL21(DE3). They showed almost identical molecular sizes of approximately 23.5 kDa in a 12% polyacrylamide gel ( Fig. 2A , top). Western blot analysis using rabbit antisera against SEY 2 revealed that SEY 2 shared antigenicity with both SEY 1 and SEY 3 ( Fig. 2A , bottom). Subsequently, we focused on detecting SEY protein in culture supernatant samples from selected representative strains of major CCs found in this study ( Table 3 ). Signals of the same electrophoretic mobility were detected in the culture supernatants of isolates of CC121, CC59, and CC20 ( Fig. 2B ). fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 2 caption a7 Confirmation of purified recombinant proteins of the SEY subtype by SDS-PAGE and immunoblot assays. (A, top) Coomassie blue-stained SDS-PAGE gel with an eluted protein sample from Co 2+ affinity chromatography. (Bottom) Fifty nanograms of each SEY subtype per well from a 12% polyacrylamide gel was transferred onto a nitrocellulose membrane, and the signal was detected using immunized rabbit anti-SEY 2 sera. (B) SEY production in the culture supernatant (arrowhead). Representatives of S. aureus isolates from major CCs positive for the sey gene were cultured in 60 ml of BHI broth supplemented with 1% yeast extract for 24 h. Samples were treated by TCA-acetone precipitation and subjected to SDS-PAGE and Western blotting. MW, molecular weight. table ft1 table-wrap mode="anchored" t5 TABLE 3 caption a7 Strain, plasmid, or primer Characteristic(s) or sequence a Source or reference Strains E.coli DH5α Cloning strain TaKaRa BL21(DE3) Host expression for recombinant protein Novagen S. aureus JP087 sey + ; ST121, CC121, and sey 1 ; atopic dermatitis This study JP074 sey + ; ST59, CC59, and sey 2 ; atopic dermatitis This study JP096 sey + ; ST20, CC20, and sey 3 ; atopic dermatitis This study NCTC 8325 sey negative; reference strain Sugai lab Plasmids pFAY 1 Amp r ; pET22b(+) with sey 1 This study pFAY 2 Amp r ; pET22b(+) with sey 2 This study pFAY 3 Amp r ; pET22b(+) with sey 3 This study Primers sey detection sey F TCTATTGGAATAGCAGAAGTA This study sey R CAATATGTCGCCTAAATCTAT This study sey cloning sey F1 GGAATTC CATATG CACCACCACCACCACCACAAAACAACTGGATTGATTACAG This study sey R2 TTGAC GAATTC TATGTTGGAACGAC This study Open in a separate window a Restriction cut sites are in boldface type. Strains, plasmids, and primers used in this study


Quantification of SEY production by a sandwich ELISA.
We developed an ELISA system to quantify SEY production of S. aureus isolates harboring sey using purified polyclonal anti-SEY 2 antibodies ( Fig. 3A ). The specificity of the sandwich ELISA was confirmed by the lack of cross-reactivity with other enterotoxins ( Fig. 3B ). When cultured in brain heart infusion (BHI) broth supplemented with 1% yeast extract medium (BHIY) at 37°C for 24 h with constant agitation, most strains of major CCs were positive for SEY in vitro above the detectable level, while only one of the isolates did not produce SEY in each of the CC20 and CC121 groups. Of note, strains belonging to CC20 were detected in the range of 8 to 57 ng/ml, whereas in the CC121 group, the level of production ranged from 3 to 178 ng/ml ( Fig. 4 ). fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 3 caption a7 Establishment of the SEY ELISA. (A) Straight-line plot of SEY concentrations of 0.2 to 5.0 ng/ml for the standard curve. The detection limit for SEY by the ELISA was 0.2 ng/ml. The coefficient of determination ( R 2 ) values for linear regression are indicated. (B) Cross-reaction test of anti-SEY serum against other selected SEs using an ELISA. One hundred nanograms of purified enterotoxins (5 ng for SEY) or no SE was loaded into a microplate well. The cutoff value for the SEY ELISA was 0.3 ng/ml; values of <0.3 ng/ml were considered undetectable. (C) Growth of S. aureus isolates, starting at an OD at 660 nm of 0.01, from triplicate measurements. (D) Production of SEY from selected strains as representatives of major CCs in this study, measured by a sandwich ELISA. Each isolate or sample was assayed in three biological or technical replicates, respectively, and each assay was performed twice. Error bars represent standard errors of the means (SEM) of test results. fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 4 caption a7 SEY production by S. aureus isolates of various CCs. SEY production by sey + strains of major CCs cultured in BHIY medium was measured by a sandwich ELISA. sey -negative strain NCTC 8325 was used as a negative control. Each isolate was assayed in duplicate in each assay. Bars represent the means of the data set. ND, not determined.


SEY variants from human isolates are unstable against heat treatment.
SEY bov was reported to be unstable against heat and enzyme treatments ( 15 ). We investigated whether SEY variants have characteristics similar to those of SEY bov . As shown in Fig. 5A , SDS patterns after heat treatment of SEY variants were more stable than those of SEB, suggesting that SEY variants from human isolates are more stable than SEY bov . Unlike heat treatment, SEY variants were unstable against digestive enzyme treatment, like SEY bov ( 15 ). Pepsin-treated SEY variants showed signs of digestion at 30 min ( Fig. 5B ). Similarly, apparent digestion by trypsin treatment was observed in just 30 min ( Fig. 5C ). SEB was resistant to heat and pepsin treatments but susceptible to trypsin digestion ( 5 , 19 ). These results suggested that SEY variants are stable against heat treatment but susceptible to digestive enzyme treatment. fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 5 caption a7 Stability test of SEY variants against heat and digestive enzyme treatments. SEY variants, SEB, and BSA (100 μg/ml) (depicted by arrows) were heated at 100°C (A), incubated with pepsin (100 μg/ml) (B), or incubated with trypsin (50 μg/ml) (C). After the indicated times, samples were collected and analyzed by SDS-PAGE. NT, no treatment; P, pepsin; T, trypsin.


T-cell proliferation and cytokine production in human PBMCs stimulated with SEY variants.
In mouse splenocytes, any of the variants, SEY 1 , SEY 2 , or SEY 3 , failed to induce a significant level of splenocyte proliferation after 48 h of stimulation at any concentration ranging from 1 pg/ml to 10 μg/ml, whereas cells treated with SEA showed a dose-dependent increase in the level of proliferation ( Fig. 6 ). Next, we investigated human T-cell responses to the SEY variants by flow cytometry. Substantial cell divisions were observed in both CD4 + and CD8 + T-cell populations when PBMCs were stimulated with SEB or any of the SEY variants ( Fig. 7A ). Although the dividing cell fraction was higher in the PBMCs cultured with SEY 3 than in those cultured with SEY 1 or SEY 2 , the difference in the mean of percentages of dividing CD4 + or CD8 + T cells was not statistically significant ( Fig. 7B and ​ andC). C ). In addition, we measured the levels of tumor necrosis factor alpha (TNF-α) and gamma interferon (IFN-γ) in culture supernatants from human PBMCs stimulated with the variants ( Fig. 7D and ​ andE). E ). Similar to other SEs, all the SEY variants induced the production of TNF-α and IFN-γ as well as the proliferation of CD4 + and CD8 + T cells. Although the effects on human lymphocytes slightly differ among the SEY variants, none of the SEY variants were as potent as SEB, which is known to act as a superantigen in both mice and humans. fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 6 caption a7 Mitogenic activity of SEY variants in mouse splenocytes. Splenocytes were stimulated with several concentrations of SEA, SEY 1 , SEY 2 , and SEY 3 for 48 h. Splenocyte proliferation was assessed using CCK-8 assays. Each bar shows the mean and SEM for triplicate assays from a representative experiment. These data are representative of results from three independent experiments. fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 7 caption a7 Percentages of proliferating CD4 + and CD8 + T cells and cytokine production from human PBMCs following stimulation with SEY variants. SEB was used as the positive control. (A to C) Flow cytogram data from one representative experiment (A) and percentages of proliferating CD4 + (B) and CD8 + T (C) cells indicated as the mean values and SEM of data for three donors. (D and E) Mean TNF-α (D) and IFN-γ (E) levels and SEM of data for 4 donors. Statistical analysis among SEY variants or between SEY variants and SEB was performed using analysis of variance (ANOVA) followed by a Tukey test (*, P < 0.05; **, P < 0.01; ***, P < 0.001; NS, not significant).


Human TCR expression profiles following stimulation with SEY 2 .
Using a panel of monoclonal antibodies (MAbs) specific to TCR Vβ subfamilies, we investigated TCR Vβ profiles in CD4 + T cells that proliferated in response to stimulation of PBMCs with 10 ng/ml of SEY 2 or SEB. The distribution of TCR Vβ expression in CD4 + T cells from five healthy donors indicated that there was no clear enhancement in the expression of certain Vβ subfamilies ( Fig. 8A ). In addition, neither SEY 1 nor SEY 3 activated specifically different Vβ subfamilies (Fig. S1). To evaluate alterations in TCR repertoire usage following SEY 2 stimulation more extensively and precisely, we conducted deep sequencing of TCR Vα (TRAV) and Vβ (TRBV) gene transcripts in T cells cultured with the toxins. As shown in Fig. 8B , a remarkable enhancement of TCR Vα transcription in CD4 + T cells following SEY 2 stimulation was observed for TRAV 8.2 and 8.6 in three healthy individuals examined, with frequencies of 24% and 33%, respectively. On the other hand, there was no remarkable change in the frequency of any TRBV gene transcripts in SEY 2 -stimulated CD4 + T cells ( Fig. 8C ). In accord with previous reports ( 2 , 20 , 21 ) and the present flow cytometry results, Vβ subfamilies 3, 12, 14, 17, and 20 were specifically enriched by SEB stimulation ( Fig. 8A and ​ andC). C ). In addition, enhanced transcriptions of Vβ 6 and 15, which were not analyzed by flow cytometry due to the unavailability of specific antibodies to these Vβ subfamilies, were observed in SEB-stimulated CD4 + T cells ( Fig. 8A ), while there was no particular Vα enhancement in these cells ( Fig. 8B ). fig ft0 fig mode=article f1 fig/graphic|fig/alternatives/graphic mode="anchored" m1 Open in a separate window FIG 8 caption a7 (A) TCR Vα/β profiles of CD4 + T cells after in vitro stimulation of human PBMCs with 10 ng/ml of SEY 2 or SEB. After 5 days of incubation, the cells were stained with anti-CD4 MAb and a panel of MAbs to distinct TCR Vβ subfamily segments. Data represent the mean values and SEM of data for five healthy donors. (B and C) Transcripts from TRAV (B) and TRBV (C) genes were evaluated by TCR sequencing in sorted CD4 + T cells. The mean frequencies of V gene transcripts and SEM of data for three healthy donors are indicated as bars. (D) Chord diagrams depicting V-J segment combinations in TRAV and TRBV genes in CD4 + T cells following PBS, SEY, or SEB stimulation, where clonotype distribution maps were generated by VDJtools. The relative abundance of each V-J segment combination is represented as a proportion of the colored chord. Preferential usage of a certain variable-joining (V-J) gene recombination has been shown in human viral infection ( 22 ). Using the TCR sequencing data, we therefore investigated whether there were any specific combinations between Vα and Jα or Vβ and Jβ segments in the stimulated CD4 + T cells. No certain pattern was seen in J gene segment usages among TCR Vα or Vβ repertoires in CD4 + T cells stimulated with SEY 2 or SEB ( Fig. 8D ). Overall, we demonstrate that SEY 2 acts as a superantigen that specifically stimulates human T cells bearing TCR Vα 8.2 and 8.6.


DISCUSSION
A molecular epidemiological study of 270 S. aureus human clinical isolates demonstrated that the sey gene was primarily found in isolates from skin infections with loss of barrier function. We show that the major sey -positive strains were distributed in CCs 121, 59, and 20 ( Table 2 ). Of note, sequence type 59 (ST59) and ST121 have been implicated in skin and soft tissue infection in many countries; distributions and prevalences were reported in Portugal ( 23 ), Japan ( 24 ), China ( 25 ), and Taiwan ( 26 ). The distribution of sey in human isolates was slightly different from that in bovine isolates, as described previously ( 27 ). Wilson et al. reported that CC151 was the main source of sey + , whereas CC20 was also identified from bovine isolates ( 27 ). Notably, CC20 was associated with human and animal infections, indicating that strains of this CC could infect both hosts ( 28 , 29 ). Most of the recently identified SEs of clinical isolates are encoded on mobile genetic elements ( 1 , 30 ). On the other hand, like SEY bov ( 15 ), genome sequence analysis of selected strains for SEY variants in this study showed that the sey genes were located on the chromosome ( Fig. 1C ). Moreover, diversity of SEY sequence existed in distinct lineages, suggesting their clonal distribution ( Fig. 1B ). S. aureus is known to produce various amounts of SEs during growth in broth culture ( 31 , – 33 ). In this study, SEYs were detected in culture supernatants using Western blot analysis as well as a sandwich ELISA. A time course experiment between 3 h and 24 h indicated that SEY was produced and accumulated gradually in culture supernatants ( Fig. 3C and ​ andD). D ). Our study shows that SEY levels varied from 3 to 178 ng/ml ( Fig. 4 ). Salgado-Pabón et al. reported that SEC was produced at up to 100 μg/ml in strain MW2 and that the toxins SEA, SEH, SEK, SEL, SEIQ, and SEIX were produced at very low concentrations (0.075 to 30 ng/ml) ( 32 ). The presence of a small amount of SE is enough to induce T-cell proliferation, triggering massive cytokine production and leading to systemic illnesses such as toxic shock syndrome and related illnesses ( 2 , 34 , 35 ). A low concentration of SEs (1 to 10 ng/ml) is frequently associated with chronic or asymptomatic mucosal infection by S. aureus ( 35 ). Thus, the potential of low-SE-producing strains to induce SE-related diseases should not be underestimated. In human PBMCs, we found that the SEY variants can induce the proliferation of both CD4 + and CD8 + T cells with slightly different degrees of efficiency ( Fig. 7B and ​ andC). C ). Such SE subtype variation has also been found in amino acid substitutes of SEB that exhibit differences in their abilities to induce the proliferation of human T-cell subsets and in causing lethality in rabbits ( 36 ). Our present results therefore suggest that the diversity in amino acid sequence at the respective residues of SEY from different CCs may affect the mitogenic response level of human T cells. Our study is the second to report that staphylococcal enterotoxin predominantly induces TCR activation through particular Vα segments. Previously, Petersson et al. demonstrated that SEH stimulates human TCR Vα (TRAV27), with no specific TCR Vβ expansion ( 37 ). Interestingly, Ono et al. ( 38 ) reported that there were no specific TCR Vα 1 to 17 by PCR or Vβ 1 to 23 by flow cytometry in human T cells that were reactive to SET, which has the closest similarity, i.e., 32% amino acid sequence identity, to SEYs among known SEs ( Fig. 1A ). We demonstrate that TCR sequencing of the whole RNA transcript is superior to TCR repertoire analyses based on flow cytometry and PCR for the characterization of specific Vα and Vβ expansions following superantigen stimulation. Further studies using TCR sequencing to elucidate the Vα or Vβ specificity of SET may uncover the structural relationship of SEY and SET with T-cell activation. SEB was reported to induce dermatitis upon application on normal skin in healthy subjects; moreover, the specific SEB-reactive TCR Vβ 12 and 17 were enriched in T cells accumulating in skin biopsy specimens of normal skin in healthy and atopic-dermatitis subjects after application of SEB ( 39 , 40 ). On the other hand, Brunner et al. ( 41 ) showed a highly polyclonal TCR pattern in 29 lesional and 19 nonlesional AD skin biopsy specimens with no specific Vβ repertoires compared to healthy controls. Since AD skin was associated with infection or colonization by SE-producing S. aureus , we demonstrate that SEY from AD isolates predominantly used specific TCR Vα for T-cell proliferation. These data suggested that TCR Vβ repertoires analyzed in T cells from the skin of AD patients may not be sufficient; thus, complete TCR Vα and Vβ repertoires should be considered. We demonstrate that sey + S. aureus human clinical isolates are common in skin pathophysiology relevant to the loss of barrier function. S. aureus is commonly adherent to the lesion, and prolonged infection may exacerbate the clinical condition. SEY in situ may contribute to the activation of skin-resident T cells via TCR Vα interactions. Analysis of TCR Vα/β repertoires at the site of inflammation in atopic-dermatitis, impetigo, and SSSS patients may provide evidence for the pathophysiological role of SEY of S. aureus .


MATERIALS AND METHODS
Bacterial strains and plasmids. Bacterial strains and plasmids used in this study are listed in Table 3 . For S. aureus clinical isolates, 270 nonrepetitively selected clinical isolates from the strain library of the Department of Bacteriology, Hiroshima University, were used in this study ( Table 1 ). Isolates were maintained and stored in a 15% glycerol stock at −80°C until use. S. aureus was routinely cultivated in tryptic soy broth (TSB; BD Microbiology System, MD, USA) at 37°C overnight with aeration in a water bath shaker or on tryptic soy agar plates prior to the experiments. Bacterial genomic DNA was prepared as previously described ( 42 ). PCR amplification. For detection, the sey gene was amplified using primers listed in Table 3 . Otherwise, multiplex PCR was performed to determine other SE (SEA, SEB, SEC, SED, SEE, SEG, SEH, SEI, SElJ, SEK, SEL, SEM, SEN, SEO, SEP, SEQ, and SER), toxic shock toxin ( tst ), exfoliative toxin ( et ), leukocidin ( luk ), epidermal cell differentiation inhibitor ( edn ), chemotaxis inhibitor ( chp ), complement inhibitor ( scn ), staphylokinase ( sak ), and methicillin-resistant ( mecA ) genes according to a method described previously ( 43 , – 45 ). Multilocus sequence typing. MLST was carried out using sequences of seven housekeeping genes according to a previously reported method ( 46 ). Clonal complexes (CCs) were deduced from the grouping of unique identifier MLST data using the eBURST v3 algorithm ( http://eburst.mlst.net/ ). Cloning of sey from major CCs. S. aureus isolates possessing sey , listed in Table 3 , were selected as DNA templates for sey gene cloning. KOD-Plus-Neo (Toyobo, Osaka, Japan) was used to amplify the DNA fragment corresponding to the predicted mature form of SEY with primer pairs including the NdeI and EcoRI sites ( Table 3 ). The signal peptide of premature SEY was predicted using online SignalP prediction software ( http://www.cbs.dtu.dk/services/SignalP ) ( 47 ). The PCR products were digested with NdeI and EcoRI before ligation into the pET 22b+ expression vector (Novagen, Madison, WI, USA). The plasmid construct was transformed and maintained in Escherichia coli strain DH5α, grown in Luria-Bertani broth (10 g NaCl, 10 g Trypticase peptone, and 5 g yeast extract per liter [pH 7.2]) containing 100 μg/ml ampicillin or on LB agar with ampicillin. The insertion was verified by DNA sequencing (ABI 3100xl; Applied Biosystems, CA, USA). The resultant plasmids containing sey 1 , sey 2 , and sey 3 were named pFAY 1 , pFAY 2 , and pFAY 3 , respectively ( Table 3 ). Expression of recombinant SEY 1 , SEY 2 , and SEY 3 . Recombinant proteins were expressed in E. coli BL21(DE3) containing the plasmid construct pFAY 1 , pFAY 2 , or pFAY 3 . Bacteria were grown with aeration at 37°C in 300 ml LB medium containing 100 μg/ml ampicillin to an optical density (OD) at 600 nm of 0.5 to 0.6, as previously described ( 42 ). Cultures were induced with 0.5 mM (final concentration) isopropyl-β- d -thiogalactopyranoside (IPTG; Nacalai Tesque, Kyoto, Japan) and further incubated with constant rotation at 30°C for another 24 h. Cells were pelleted by centrifugation (4,000 × g at 4°C for 15 min) and kept at −80°C until use. Cell pellets were thawed in lysis buffer (5 mM imidazole, 50 mM NaH 2 PO 4 , and 300 mM NaCl [pH 7.0]) and lysed by sonication on slurry ice. The cell pellet was removed by centrifugation (9,000 × g at 4°C for 20 min), and the supernatant was collected for the isolation of N-terminally 6×His-tagged SEY fusion proteins by using Co 2+ affinity chromatography (Clontech Laboratories, Inc.) according to the manufacturer’s instructions. The purity of the proteins was determined by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and proteins were stained with Coomassie brilliant blue R250 (Nacalai Tesque, Kyoto, Japan). A Bio-Rad protein assay (Bio-Rad, Hercules, CA, USA) with bovine serum albumin (BSA; Sigma-Aldrich, St. Louis, MO, USA) as the standard protein was used to measure the protein concentration. Purified proteins were dialyzed (Spectra/por dialysis membrane with a molecular weight cutoff [MWCO] of 6 to 8 kDa; Spectrum Lab, Rancho Dominguez, CA, USA) against a 1,000× volume of Dulbecco’s phosphate-buffered saline (DPBS) (2.68 mM KCl, 1.46 mM KH 2 PO 4 , 136.9 mM NaCl, 8 mM Na 2 HPO 4 ·12H 2 O [pH 7.4]) at 4°C for 24 h. For superantigen characterization, purified SEY 1 , SEY 2 , and SEY 3 were treated with EndoTrap (Hyglos, Bernried, Germany) to remove endotoxin contamination before use. Preparation of culture supernatants. Supernatant samples were prepared for Western blotting and ELISAs. A single colony of S. aureus was grown in 3 ml brain heart infusion (BHI) broth (BD Microbiology System, MD, USA) supplemented with 1% (wt/vol) yeast extract (BD Microbiology System, MD, USA) at 37°C as described previously ( 48 ). A total of 0.6 ml precultured cells was inoculated into 60 ml fresh medium and cultured with constant agitation at 37°C for 24 h. The supernatant was collected after centrifugation (15,600 × g for 20 min at 4°C) and filtered using a 0.2-μm-pore-size Minisart hydrophilic cellulose acetate membrane filter (Sartorius, Göttingen, Germany). Prior to the ELISA, the culture supernatant sample was treated with 50% (vol/vol) normal rabbit serum (Invitrogen, MD, USA) overnight at 4°C to neutralize protein A, as previously described ( 49 ). Western blotting. SEY sera were prepared by Sigma-Aldrich (Tokyo, Japan) by immunizing rabbits with purified SEY 2 . Western blotting was performed to detect SEY production in vitro from broth cultures and cross-reaction tests between SEY 1 , SEY 2 , and SEY 3 . Three S. aureus strains from major CCs listed in Table 3 were selected for toxin production in culture supernatants. Samples for Western blotting were prepared by trichloroacetic acid (TCA)-acetone precipitation from culture supernatants as described previously ( 48 ), with some modifications. A 10% volume of 100% (wt/vol) trichloroacetic acid (Nacalai Tesque, Kyoto, Japan) was added, and the mixture was incubated for 30 min on ice. The precipitated protein was collected by centrifugation (16,700 × g for 15 min at 4°C), followed by two washes with 1 ml acetone. The precipitates were dissolved in 100 μl of 1× SDS sample buffer. The sample and purified SEY 2 as a positive control were electrophoretically transferred onto a nitrocellulose membrane (Amersham Protran NC 0.45; GE Healthcare, Germany) from a 12% SDS-PAGE gel. SEY was detected by using rabbit antisera to SEY. Bound primary antibodies were detected with peroxidase-conjugated goat anti-rabbit immunoglobulin G (IgG) antibodies (Cappel Products, MP Biomedicals, LLC, OH, USA) at a 1:10,000 dilution, and chemiluminescence was developed with the Clarity Western ECL substrate (Bio-Rad, Hercules, CA, USA). Purified polyclonal antibody preparation. Antibody purification was performed as described previously by Omoe et al. ( 31 ). IgG antibodies were affinity purified from hyperimmune sera with protein G-Sepharose 4 fast-flow resin (GE Healthcare Bio-Sciences AB, Uppsala, Sweden). Purified SEY 2 -coupled N -hydroxysuccinimide (NHS)-activated Sepharose resin (GE Healthcare Bio-Sciences AB, Uppsala, Sweden) was used to purify polyclonal IgG. About 1 mg of purified polyclonal IgG was dialyzed against a 1,000× volume of DPBS (pH 7.2), conjugated to EZ-link Plus activated peroxidase (Thermo Scientific, Rockford, IL, USA), mixed with Pierce peroxidase conjugate stabilizer (Thermo Scientific, Rockford, IL, USA) according to manual instructions, and stored in aliquots at −20°C until use. SEY sandwich ELISA. A modified protocol was developed for a sandwich ELISA based on methods described previously by Omoe et al. ( 31 ). A microtiter plate (96F polysorp, white microwell SH; Nunc, Denmark) was coated with a total of 100 μl/well of a solution containing 0.5 μg/ml purified polyclonal IgG anti-SEY in DPBS, followed by blocking with blocking buffer (StartingBlock [PBS] blocking buffer; Thermo Scientific, Rockford, IL). Standards (recombinant SEY 2 ) or the sample at appropriate dilutions (10- to 1,000-fold) in immunoreaction enhancer solution 1 (Can Get Signal; Toyobo, Osaka, Japan) was added to the wells. The plate was washed 8 times (300 μl/well) with DPBS containing 0.05% (wt/vol) Tween 20 (Sigma-Aldrich, St. Louis, MO, USA). Secondary IgG-horseradish peroxidase (HRP) diluted 12,800-fold in immunoreaction enhancer solution 2 (Can Get Signal; Toyobo, Osaka, Japan) was added to all wells. After incubation and washing, the chemiluminescent substrate (SuperSignal ELISA Femto maximum-sensitivity substrate; Thermo Scientific, Rockford, IL) was added, and the reaction was allowed to develop. Relative light units (∼425 nm) were determined using a microplate reader (Varioskan Flash spectral scanning multimode reader; Thermo Fisher Scientific, Waltham, MA, USA). A standard curve was constructed by plotting luminometric values (means from duplicate wells) for each reference standard concentration. The concentration in nanograms per milliliter of each sample was determined from the corresponding standard curve. Assessment of stability of the SEY variants. We compared the stabilities of SEY variants against heat and digestive enzyme treatments as described previously ( 15 ). Briefly, SEY 1 , SEY 2 , SEY 3 , SEB, and BSA, as a protein control, were diluted to 100 μg/ml in PBS. For heat treatment stability, protein samples were heated at 100°C in a dry heat block (As One, Japan). Furthermore, the stability of the protein against digestive enzymes was assessed with pepsin (Sigma, St. Louis, MO, USA) and trypsin (Sigma, St. Louis, MO, USA). The prediluted protein samples described above were mixed with 100 μg/ml of pepsin in 0.1 M sodium acetate buffer (pH 4.5) or 50 μg/ml trypsin in 0.01 M Tris-HCl buffer (pH 8.0) and then incubated at 37°C. All samples for this experiment were collected at the desired time range from 0.5 h to 12 h and then analyzed by 12% SDS-PAGE with Coomassie brilliant blue staining. Mouse splenocyte proliferation. C57BL/6J mouse (female, 6 weeks old; Charles River Laboratories Japan, Inc., Yokohama, Japan) spleen was used as a source of splenocytes. Mitogenic assays were carried out after 48 h of incubation of 4.5 × 10 5 cells/well in a 96-well round-bottomed culture plate (Greiner Bio-One International, Kremsmünster, Austria). Different concentrations of EndoTrap-treated endotoxin-free SEY 1 , SEY 2 , SEY 3 , or SEA as a positive control were used. Splenic lymphocyte proliferation was determined by using a cell counting kit (CCK-8; Dojindo Laboratories, Kumamoto, Japan) as described previously ( 50 ). Human lymphocyte proliferation. Human peripheral blood mononuclear cells (PBMCs) were obtained from 20 ml of heparinized venous blood samples by centrifugation on a Ficoll density gradient (lymphocyte separation medium 1077; Wako Pure Chemical Industries Ltd., Osaka, Japan) for 30 min at 400 × g . The interface layer containing PBMCs was collected, washed, and resuspended in RPMI 1640 medium (Nacalai Tesque, Inc., Kyoto, Japan) supplemented with 2.5% fetal calf serum (FCS; HyClone Laboratories, Logan, UT, USA) and 1% penicillin-streptomycin (P/S; Gibco/Life Technologies, Grand Island, NY, USA) at room temperature. Cell number and viability were determined microscopically by trypan blue staining. About 40 × 10 6 PBMCs were washed and resuspended in DPBS supplemented with 1% FCS for labeling with a cell division tracker, 5 (and 6)-carboxyfluorescein diacetate succinimidyl ester (CFSE; Tonbo Biosciences, San Diego, CA, USA). Cell labeling in 10 μM CFSE was performed at 37°C for 15 min and then terminated by the addition of FCS (80% of the total volume). The CFSE-labeled PBMCs (4.0 × 10 6 cells) were stimulated with 10 ng/ml of SEY 1 , SEY 2 , SEY 3 , or SEB (Sigma-Aldrich, Tokyo, Japan) as a positive control in 2 ml of RPMI 1640 medium supplemented with 10% FCS and 1% P/S and incubated at 37°C with flushed 5% CO 2 in a 24-well plate (Corning, Inc., Corning, NY, USA) (2 ml/well). Following a 5-day incubation, the cells were washed in DPBS supplemented with 1% FCS and 0.01% sodium azide. About 0.2 × 10 6 cells were surface stained with phycoerythrin (PE)-labeled anti-CD4 (clone RPA-T4; Tonbo Biosciences, San Diego, CA, USA) and PE-Cy7-labeled anti-CD8a (clone RPA-T8; Tonbo Biosciences) antibodies for 30 min on slurry ice. After the cells were washed, cell divisions of CD4 + or CD8 + T cells were analyzed by flow cytometry (JSAN; Bay Biosciences, Kobe, Japan), and the flow cytometry data were assessed by using FlowJo software (TreeStar, Ashland, OR, USA). The percentages of proliferating cells were determined by gating on CD4 + CFSE low or CD8 + CFSE low cells. Cytokine production. Human PBMCs (2.0 × 10 6 cells) were stimulated with 10 ng/ml of SEY 1 , SEY 2 , SEY 3 , or SEB as described above. After a 5-day incubation, culture supernatants were collected, and levels of tumor necrosis factor alpha (TNF-α) and gamma interferon (IFN-γ) were measured using ELISA kits (Invitrogen, USA) according to the supplier’s instructions. Sample measurements were all performed in duplicate. TCR Vβ expression in human PBMCs. TCR Vβ expression profiles in SEY-responding T cells were assessed by flow cytometry as previously described ( 38 ). In brief, PBMCs cultured with 10 ng/ml of SEY 2 or SEB for 5 days as described above were stained with PE-Cy7-labeled anti-CD4 (clone RPA-T4; Tonbo Biosciences) and a panel of MAbs to TCR Vβ subsets (IOTest beta mark kit; Beckman Coulter, Miami, FL). The stained cells were acquired on a FACScan flow cytometer (BD Biosciences, San Jose, CA, USA), and the flow cytometry data were analyzed with FlowJo software (TreeStar, Ashland, OR). TCR sequencing. PBMCs cultured with toxins for 5 days as described above were stained with PE-Cy7-labeled anti-CD4 (clone RPA-T4; Tonbo Biosciences). CD4 cells were sorted in a flow cytometer (JSAN; Bay Biosciences, Kobe, Japan) and collected in RLT lysis buffer (Qiagen, Hilden, Germany). Total RNA was extracted using the RNeasy minikit (Qiagen, MD, USA). TCR libraries were prepared using a SMARTer human TCR a/b profiling kit (TaKaRa Bio, CA, USA) according to the manufacturer’s instructions. Purified products were sequenced on an Illumina MiSeq instrument. Analysis of TCR sequences was carried out using MiXCR software ( 51 ) for raw data processing with CDR3 extraction and gene alignment and tcR ( 52 ) and VDJtools ( 53 ) software for comparing TCR repertoires of each sample. Ethics statement. All experiments using mouse models were performed with the approval of the local ethical committee of the Kitasato University School of Veterinary Medicine, Towada, Japan (permit numbers 17-115 and 18-092). Human venous blood was obtained from healthy donors, which was approved by the Hiroshima University ethics committee. Accession number(s). The nucleotide sequences of sey 1 , sey 2 , and sey 3 were submitted to the GenBank database and assigned accession numbers {"type":"entrez-nucleotide","attrs":{"text":"LC431775","term_id":"1755080157","term_text":"LC431775"}} LC431775 , {"type":"entrez-nucleotide","attrs":{"text":"LC431776","term_id":"1755080178","term_text":"LC431776"}} LC431776 , and {"type":"entrez-nucleotide","attrs":{"text":"LC431777","term_id":"1755080207","term_text":"LC431777"}} LC431777 , respectively.


Bacterial strains and plasmids.
Bacterial strains and plasmids used in this study are listed in Table 3 . For S. aureus clinical isolates, 270 nonrepetitively selected clinical isolates from the strain library of the Department of Bacteriology, Hiroshima University, were used in this study ( Table 1 ). Isolates were maintained and stored in a 15% glycerol stock at −80°C until use. S. aureus was routinely cultivated in tryptic soy broth (TSB; BD Microbiology System, MD, USA) at 37°C overnight with aeration in a water bath shaker or on tryptic soy agar plates prior to the experiments. Bacterial genomic DNA was prepared as previously described ( 42 ).


PCR amplification.
For detection, the sey gene was amplified using primers listed in Table 3 . Otherwise, multiplex PCR was performed to determine other SE (SEA, SEB, SEC, SED, SEE, SEG, SEH, SEI, SElJ, SEK, SEL, SEM, SEN, SEO, SEP, SEQ, and SER), toxic shock toxin ( tst ), exfoliative toxin ( et ), leukocidin ( luk ), epidermal cell differentiation inhibitor ( edn ), chemotaxis inhibitor ( chp ), complement inhibitor ( scn ), staphylokinase ( sak ), and methicillin-resistant ( mecA ) genes according to a method described previously ( 43 , – 45 ).


Multilocus sequence typing.
MLST was carried out using sequences of seven housekeeping genes according to a previously reported method ( 46 ). Clonal complexes (CCs) were deduced from the grouping of unique identifier MLST data using the eBURST v3 algorithm ( http://eburst.mlst.net/ ).


Cloning of sey from major CCs.
S. aureus isolates possessing sey , listed in Table 3 , were selected as DNA templates for sey gene cloning. KOD-Plus-Neo (Toyobo, Osaka, Japan) was used to amplify the DNA fragment corresponding to the predicted mature form of SEY with primer pairs including the NdeI and EcoRI sites ( Table 3 ). The signal peptide of premature SEY was predicted using online SignalP prediction software ( http://www.cbs.dtu.dk/services/SignalP ) ( 47 ). The PCR products were digested with NdeI and EcoRI before ligation into the pET 22b+ expression vector (Novagen, Madison, WI, USA). The plasmid construct was transformed and maintained in Escherichia coli strain DH5α, grown in Luria-Bertani broth (10 g NaCl, 10 g Trypticase peptone, and 5 g yeast extract per liter [pH 7.2]) containing 100 μg/ml ampicillin or on LB agar with ampicillin. The insertion was verified by DNA sequencing (ABI 3100xl; Applied Biosystems, CA, USA). The resultant plasmids containing sey 1 , sey 2 , and sey 3 were named pFAY 1 , pFAY 2 , and pFAY 3 , respectively ( Table 3 ).


Expression of recombinant SEY 1 , SEY 2 , and SEY 3 .
Recombinant proteins were expressed in E. coli BL21(DE3) containing the plasmid construct pFAY 1 , pFAY 2 , or pFAY 3 . Bacteria were grown with aeration at 37°C in 300 ml LB medium containing 100 μg/ml ampicillin to an optical density (OD) at 600 nm of 0.5 to 0.6, as previously described ( 42 ). Cultures were induced with 0.5 mM (final concentration) isopropyl-β- d -thiogalactopyranoside (IPTG; Nacalai Tesque, Kyoto, Japan) and further incubated with constant rotation at 30°C for another 24 h. Cells were pelleted by centrifugation (4,000 × g at 4°C for 15 min) and kept at −80°C until use. Cell pellets were thawed in lysis buffer (5 mM imidazole, 50 mM NaH 2 PO 4 , and 300 mM NaCl [pH 7.0]) and lysed by sonication on slurry ice. The cell pellet was removed by centrifugation (9,000 × g at 4°C for 20 min), and the supernatant was collected for the isolation of N-terminally 6×His-tagged SEY fusion proteins by using Co 2+ affinity chromatography (Clontech Laboratories, Inc.) according to the manufacturer’s instructions. The purity of the proteins was determined by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and proteins were stained with Coomassie brilliant blue R250 (Nacalai Tesque, Kyoto, Japan). A Bio-Rad protein assay (Bio-Rad, Hercules, CA, USA) with bovine serum albumin (BSA; Sigma-Aldrich, St. Louis, MO, USA) as the standard protein was used to measure the protein concentration. Purified proteins were dialyzed (Spectra/por dialysis membrane with a molecular weight cutoff [MWCO] of 6 to 8 kDa; Spectrum Lab, Rancho Dominguez, CA, USA) against a 1,000× volume of Dulbecco’s phosphate-buffered saline (DPBS) (2.68 mM KCl, 1.46 mM KH 2 PO 4 , 136.9 mM NaCl, 8 mM Na 2 HPO 4 ·12H 2 O [pH 7.4]) at 4°C for 24 h. For superantigen characterization, purified SEY 1 , SEY 2 , and SEY 3 were treated with EndoTrap (Hyglos, Bernried, Germany) to remove endotoxin contamination before use.


Preparation of culture supernatants.
Supernatant samples were prepared for Western blotting and ELISAs. A single colony of S. aureus was grown in 3 ml brain heart infusion (BHI) broth (BD Microbiology System, MD, USA) supplemented with 1% (wt/vol) yeast extract (BD Microbiology System, MD, USA) at 37°C as described previously ( 48 ). A total of 0.6 ml precultured cells was inoculated into 60 ml fresh medium and cultured with constant agitation at 37°C for 24 h. The supernatant was collected after centrifugation (15,600 × g for 20 min at 4°C) and filtered using a 0.2-μm-pore-size Minisart hydrophilic cellulose acetate membrane filter (Sartorius, Göttingen, Germany). Prior to the ELISA, the culture supernatant sample was treated with 50% (vol/vol) normal rabbit serum (Invitrogen, MD, USA) overnight at 4°C to neutralize protein A, as previously described ( 49 ).


Western blotting.
SEY sera were prepared by Sigma-Aldrich (Tokyo, Japan) by immunizing rabbits with purified SEY 2 . Western blotting was performed to detect SEY production in vitro from broth cultures and cross-reaction tests between SEY 1 , SEY 2 , and SEY 3 . Three S. aureus strains from major CCs listed in Table 3 were selected for toxin production in culture supernatants. Samples for Western blotting were prepared by trichloroacetic acid (TCA)-acetone precipitation from culture supernatants as described previously ( 48 ), with some modifications. A 10% volume of 100% (wt/vol) trichloroacetic acid (Nacalai Tesque, Kyoto, Japan) was added, and the mixture was incubated for 30 min on ice. The precipitated protein was collected by centrifugation (16,700 × g for 15 min at 4°C), followed by two washes with 1 ml acetone. The precipitates were dissolved in 100 μl of 1× SDS sample buffer. The sample and purified SEY 2 as a positive control were electrophoretically transferred onto a nitrocellulose membrane (Amersham Protran NC 0.45; GE Healthcare, Germany) from a 12% SDS-PAGE gel. SEY was detected by using rabbit antisera to SEY. Bound primary antibodies were detected with peroxidase-conjugated goat anti-rabbit immunoglobulin G (IgG) antibodies (Cappel Products, MP Biomedicals, LLC, OH, USA) at a 1:10,000 dilution, and chemiluminescence was developed with the Clarity Western ECL substrate (Bio-Rad, Hercules, CA, USA).


Purified polyclonal antibody preparation.
Antibody purification was performed as described previously by Omoe et al. ( 31 ). IgG antibodies were affinity purified from hyperimmune sera with protein G-Sepharose 4 fast-flow resin (GE Healthcare Bio-Sciences AB, Uppsala, Sweden). Purified SEY 2 -coupled N -hydroxysuccinimide (NHS)-activated Sepharose resin (GE Healthcare Bio-Sciences AB, Uppsala, Sweden) was used to purify polyclonal IgG. About 1 mg of purified polyclonal IgG was dialyzed against a 1,000× volume of DPBS (pH 7.2), conjugated to EZ-link Plus activated peroxidase (Thermo Scientific, Rockford, IL, USA), mixed with Pierce peroxidase conjugate stabilizer (Thermo Scientific, Rockford, IL, USA) according to manual instructions, and stored in aliquots at −20°C until use.


SEY sandwich ELISA.
A modified protocol was developed for a sandwich ELISA based on methods described previously by Omoe et al. ( 31 ). A microtiter plate (96F polysorp, white microwell SH; Nunc, Denmark) was coated with a total of 100 μl/well of a solution containing 0.5 μg/ml purified polyclonal IgG anti-SEY in DPBS, followed by blocking with blocking buffer (StartingBlock [PBS] blocking buffer; Thermo Scientific, Rockford, IL). Standards (recombinant SEY 2 ) or the sample at appropriate dilutions (10- to 1,000-fold) in immunoreaction enhancer solution 1 (Can Get Signal; Toyobo, Osaka, Japan) was added to the wells. The plate was washed 8 times (300 μl/well) with DPBS containing 0.05% (wt/vol) Tween 20 (Sigma-Aldrich, St. Louis, MO, USA). Secondary IgG-horseradish peroxidase (HRP) diluted 12,800-fold in immunoreaction enhancer solution 2 (Can Get Signal; Toyobo, Osaka, Japan) was added to all wells. After incubation and washing, the chemiluminescent substrate (SuperSignal ELISA Femto maximum-sensitivity substrate; Thermo Scientific, Rockford, IL) was added, and the reaction was allowed to develop. Relative light units (∼425 nm) were determined using a microplate reader (Varioskan Flash spectral scanning multimode reader; Thermo Fisher Scientific, Waltham, MA, USA). A standard curve was constructed by plotting luminometric values (means from duplicate wells) for each reference standard concentration. The concentration in nanograms per milliliter of each sample was determined from the corresponding standard curve.


Assessment of stability of the SEY variants.
We compared the stabilities of SEY variants against heat and digestive enzyme treatments as described previously ( 15 ). Briefly, SEY 1 , SEY 2 , SEY 3 , SEB, and BSA, as a protein control, were diluted to 100 μg/ml in PBS. For heat treatment stability, protein samples were heated at 100°C in a dry heat block (As One, Japan). Furthermore, the stability of the protein against digestive enzymes was assessed with pepsin (Sigma, St. Louis, MO, USA) and trypsin (Sigma, St. Louis, MO, USA). The prediluted protein samples described above were mixed with 100 μg/ml of pepsin in 0.1 M sodium acetate buffer (pH 4.5) or 50 μg/ml trypsin in 0.01 M Tris-HCl buffer (pH 8.0) and then incubated at 37°C. All samples for this experiment were collected at the desired time range from 0.5 h to 12 h and then analyzed by 12% SDS-PAGE with Coomassie brilliant blue staining.


Mouse splenocyte proliferation.
C57BL/6J mouse (female, 6 weeks old; Charles River Laboratories Japan, Inc., Yokohama, Japan) spleen was used as a source of splenocytes. Mitogenic assays were carried out after 48 h of incubation of 4.5 × 10 5 cells/well in a 96-well round-bottomed culture plate (Greiner Bio-One International, Kremsmünster, Austria). Different concentrations of EndoTrap-treated endotoxin-free SEY 1 , SEY 2 , SEY 3 , or SEA as a positive control were used. Splenic lymphocyte proliferation was determined by using a cell counting kit (CCK-8; Dojindo Laboratories, Kumamoto, Japan) as described previously ( 50 ).


Human lymphocyte proliferation.
Human peripheral blood mononuclear cells (PBMCs) were obtained from 20 ml of heparinized venous blood samples by centrifugation on a Ficoll density gradient (lymphocyte separation medium 1077; Wako Pure Chemical Industries Ltd., Osaka, Japan) for 30 min at 400 × g . The interface layer containing PBMCs was collected, washed, and resuspended in RPMI 1640 medium (Nacalai Tesque, Inc., Kyoto, Japan) supplemented with 2.5% fetal calf serum (FCS; HyClone Laboratories, Logan, UT, USA) and 1% penicillin-streptomycin (P/S; Gibco/Life Technologies, Grand Island, NY, USA) at room temperature. Cell number and viability were determined microscopically by trypan blue staining. About 40 × 10 6 PBMCs were washed and resuspended in DPBS supplemented with 1% FCS for labeling with a cell division tracker, 5 (and 6)-carboxyfluorescein diacetate succinimidyl ester (CFSE; Tonbo Biosciences, San Diego, CA, USA). Cell labeling in 10 μM CFSE was performed at 37°C for 15 min and then terminated by the addition of FCS (80% of the total volume). The CFSE-labeled PBMCs (4.0 × 10 6 cells) were stimulated with 10 ng/ml of SEY 1 , SEY 2 , SEY 3 , or SEB (Sigma-Aldrich, Tokyo, Japan) as a positive control in 2 ml of RPMI 1640 medium supplemented with 10% FCS and 1% P/S and incubated at 37°C with flushed 5% CO 2 in a 24-well plate (Corning, Inc., Corning, NY, USA) (2 ml/well). Following a 5-day incubation, the cells were washed in DPBS supplemented with 1% FCS and 0.01% sodium azide. About 0.2 × 10 6 cells were surface stained with phycoerythrin (PE)-labeled anti-CD4 (clone RPA-T4; Tonbo Biosciences, San Diego, CA, USA) and PE-Cy7-labeled anti-CD8a (clone RPA-T8; Tonbo Biosciences) antibodies for 30 min on slurry ice. After the cells were washed, cell divisions of CD4 + or CD8 + T cells were analyzed by flow cytometry (JSAN; Bay Biosciences, Kobe, Japan), and the flow cytometry data were assessed by using FlowJo software (TreeStar, Ashland, OR, USA). The percentages of proliferating cells were determined by gating on CD4 + CFSE low or CD8 + CFSE low cells.


Cytokine production.
Human PBMCs (2.0 × 10 6 cells) were stimulated with 10 ng/ml of SEY 1 , SEY 2 , SEY 3 , or SEB as described above. After a 5-day incubation, culture supernatants were collected, and levels of tumor necrosis factor alpha (TNF-α) and gamma interferon (IFN-γ) were measured using ELISA kits (Invitrogen, USA) according to the supplier’s instructions. Sample measurements were all performed in duplicate.


TCR Vβ expression in human PBMCs.
TCR Vβ expression profiles in SEY-responding T cells were assessed by flow cytometry as previously described ( 38 ). In brief, PBMCs cultured with 10 ng/ml of SEY 2 or SEB for 5 days as described above were stained with PE-Cy7-labeled anti-CD4 (clone RPA-T4; Tonbo Biosciences) and a panel of MAbs to TCR Vβ subsets (IOTest beta mark kit; Beckman Coulter, Miami, FL). The stained cells were acquired on a FACScan flow cytometer (BD Biosciences, San Jose, CA, USA), and the flow cytometry data were analyzed with FlowJo software (TreeStar, Ashland, OR).


TCR sequencing.
PBMCs cultured with toxins for 5 days as described above were stained with PE-Cy7-labeled anti-CD4 (clone RPA-T4; Tonbo Biosciences). CD4 cells were sorted in a flow cytometer (JSAN; Bay Biosciences, Kobe, Japan) and collected in RLT lysis buffer (Qiagen, Hilden, Germany). Total RNA was extracted using the RNeasy minikit (Qiagen, MD, USA). TCR libraries were prepared using a SMARTer human TCR a/b profiling kit (TaKaRa Bio, CA, USA) according to the manufacturer’s instructions. Purified products were sequenced on an Illumina MiSeq instrument. Analysis of TCR sequences was carried out using MiXCR software ( 51 ) for raw data processing with CDR3 extraction and gene alignment and tcR ( 52 ) and VDJtools ( 53 ) software for comparing TCR repertoires of each sample.


Ethics statement.
All experiments using mouse models were performed with the approval of the local ethical committee of the Kitasato University School of Veterinary Medicine, Towada, Japan (permit numbers 17-115 and 18-092). Human venous blood was obtained from healthy donors, which was approved by the Hiroshima University ethics committee.


Accession number(s).
The nucleotide sequences of sey 1 , sey 2 , and sey 3 were submitted to the GenBank database and assigned accession numbers {"type":"entrez-nucleotide","attrs":{"text":"LC431775","term_id":"1755080157","term_text":"LC431775"}} LC431775 , {"type":"entrez-nucleotide","attrs":{"text":"LC431776","term_id":"1755080178","term_text":"LC431776"}} LC431776 , and {"type":"entrez-nucleotide","attrs":{"text":"LC431777","term_id":"1755080207","term_text":"LC431777"}} LC431777 , respectively.


ACKNOWLEDGMENTS
We thank Kazumasa Iwamoto for research suggestions and implications. We thank Keiko Tsuruda for assistance with statistical analysis. We thank Hiroki Kitagawa, Isamu Kado, Yuta Kuroo, and Le Nguyen Tra Mi for their technical assistance. We thank Editage for English language editing. The Radiation Effects Research Foundation (RERF), Hiroshima and Nagasaki, Japan, is a public interest foundation funded by the Japanese Ministry of Health, Labor, and Welfare (MHLW) and the U.S. Department of Energy (DOE). The views of the authors do not necessarily reflect those of the two governments. This study was supported in part by the Japan Agency for Medical Research and Development under grant JP18gm1010001 to M.S.


Footnotes
Supplemental material is available online only.


Metadata
Authors
Fatkhanuddin Aziz, Junzo Hisatsune, Liansheng Yu, Junko Kajimura, Yusuke Sato’o, Hisaya K. Ono, Kanako Masuda, Mika Yamaoka, Siti Isrina Oktavia Salasia, Akio Nakane, Hiroki Ohge, Yoichiro Kusunoki, Motoyuki Sugai
Journal
Infection and Immunity
Publisher
American Society for Microbiology
Date
pmc06977126
PM Id
31740530
PMC Id
6977126
Images
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8