Figures
Abstract
Pathogen and cellular by-products released during infection or trauma are critical for initiating mucosal inflammation. The localization of these factors, their bioactivity and natural countermeasures remain unclear. This concept was studied in mice undergoing pulmonary inflammation after Staphylococcal enterotoxin A (SEA) inhalation. Highly purified bronchoalveolar lavage fluid (BALF) fractions obtained by sequential chromatography were screened for bioactivity and subjected to mass spectrometry. The Inflammatory and inhibitory potentials of the identified proteins were measured using T cells assays. A potent pro-inflammatory factor was detected in BALF, and we hypothesized SEA could be recovered with its biological activity. Highly purified BALF fractions with bioactivity were subjected to mass spectrometry. SEA was the only identified protein with known inflammatory potential, and unexpectedly, it co-purified with immunosuppressive proteins. Among them was lactoferrin, which inhibited SEA and anti-CD3/-CD28 stimulation by promoting T cell death and reducing TNF synthesis. Higher doses of lactoferrin were required to inhibit effector compared to resting T cells. Inhibition relied on the continual presence of lactoferrin rather than a programming event. The data show a fraction of bioactive SEA resided in a mucosal niche within BALF even after the initiation of inflammation. These results may have clinical value in human diagnostic since traces levels of SEA can be detected using a sensitive bioassay, and may help pinpoint potential mediators of lung inflammation when molecular approaches fail.
Citation: Ménoret A, Svedova J, Behl B, Vella AT (2015) Trace Levels of Staphylococcal Enterotoxin Bioactivity Are Concealed in a Mucosal Niche during Pulmonary Inflammation. PLoS ONE 10(10): e0141548. https://doi.org/10.1371/journal.pone.0141548
Editor: Paulo Lee Ho, Instituto Butantan, BRAZIL
Received: May 5, 2015; Accepted: October 10, 2015; Published: October 28, 2015
Copyright: © 2015 Ménoret et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data are within the paper.
Funding: This work was supported by National Institutes of Health AI0142858 (http://projectreporter.nih.gov/project_info_details.cfm?aid=8810630&icde=24503495&ddparam=&ddvalue=&ddsub=&cr=4&csb=default&cs=ASC), and National Institutes of Health AI056172 (http://projectreporter.nih.gov/project_info_details.cfm?aid=8862346&icde=24503495&ddparam=&ddvalue=&ddsub=&cr=5&csb=default&cs=ASC).
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: BALF, bronchoalveolar lavage fluid; SEA, Staphylococcal enterotoxin A; DAMP, danger-associated molecular pattern; MRSA, methicillin-resistant Staphylococcus aureus; BSA, Bovine serum albumin; BFA, Brefeldin A
Introduction
There is increased recognition of S. aureus infection in the respiratory tract of patients not only as a single pathogen but also as a pivotal co-infection agent especially in combination with viruses such as influenza [1,2]. Diagnosing the root pathogenic mechanism of co-infection is challenging since complex biological processes of different infectious agents along with the host are involved. Nevertheless, there is often an ignored role for pathogen-generated toxins that can be secreted during infection. An excellent example are the enterotoxins released by S. aureus that have classically been grouped as superantigens [3]. Thus, besides the infectious agents themselves the presence of toxins such as Staphylococcal enterotoxins (SE) provides another layer of complex pathogenicity that directly impacts adaptive immunity.
Perhaps the greatest threat of SE are their potential to induce severe or even lethal shock as observed in classical cases of Toxic Shock Syndrome [4], but also documented in less well publicized instances of shock after endonasal surgery [5]. In either situation disease onset is rapid, and largely manifested by a massive cytokine storm triggered by SE-activated T cells. This stems from the oligoclonal activation of T cells by SE, which relies not on processed antigen from the pathogen but by bridging the MHC II molecule to a specific TCR Vβchain [3]. Not only do CD4 T cells become activated but CD8 T cells are also potently stimulated leading to substantial clonal expansion, effector differentiation and systemic migration throughout the body [6]. Thus, it is not surprising that SE mediate pathogenic outcomes in lung as observed in several pulmonary inflammation models of asthma and acute lung injury [7,6,8]. Nevertheless, there is growing evidence that SE may be involved in human pulmonary maladies including rhinosinitus [9] and as a co-morbid of asthma [10]. Recent evidence has also implicated the presence of SE in victims of sudden infant death syndrome [11]. Taken together, the diagnosis and rapid implementation of countermeasures against SE might aid in treating and understanding T cell associated severe pulmonary diseases including status asthmaticus and others [12,10,13].
A difficulty in determining a role for SE in human disease is that molecular diagnostics can only locate the presence of a factor after reaching a certain concentration threshold. An even more difficult challenge is that sensitive PCR diagnostics for the bacteria are not very useful in the detection of SE since these proteins often migrate from the sites of colonization [14]. For example, SE in lung mucosa may be localized away from the colonizing bacteria in the nasal polyps, making a PCR-based diagnosis of a lung specimen inconclusive. One might also consider TCR Vβ expansion as an indicator of SE, but this would not infer imminent presence and is also complicated by the loss of polyclonality of TCRs in CD8 T cells of people especially the elderly [15,16]. Also, SE by virtue of the fact that they are Proteinaceous, effective at very low concentrations, and have a high binding avidity for MHC II, makes them difficult to detect in patient samples. Thus, a major gap is how SE can be efficiently detected, but perhaps even more importantly tracking SE bioactivity which an ELISA or immunoblot cannot.
To address this idea we set out to determine if SEA could be recovered and if it retained biological activity. We tested this idea using a bioassay where cells produce IFN-γ in response to LPS or cytokines and found that BALF from intranasally-SEA treated mice contain a potent capacity to stimulate IFN-γ release. Confirming our hypothesis, we found trace levels of SEA in fractionated BALF even after inflammation was initiated. Purification of SEA from BALF showed that its presence was undetectable by immunoblot and that a T cell dependent bioactivity assay was necessary for its identification; perhaps even moreso than mass spectrometry. Secondly, SEA co-purified with host lactoferrin in fractionated BALF, which weakened T cell IL-2 and TNF synthesis. Specifically, cytokine responses were inhibited only when lactoferrin was in the continual presence of the cells. Lactoferrin did not pre-program protective or attenuated responses, instead its presence was a requirement especially for effector T cells. These results suggest that bioactive SE may be present in trace amounts in BALF of patients with severe respiratory events. A sensitive bioactivity assay may assist in identifying disease-mediating agents in complex cases of rapid shock seen in intensive care units.
Materials and Methods
Mice and immunization
C57BL/6 and TCR β/γ mice were purchased from the National Cancer Institute (Frederick, MD) or the Jackson Laboratory (Bar Harbor, ME). All mice were maintained in the central animal facility at the University of Connecticut Health (UCH) in accordance with federal guidelines. The present study was approved by the University of Connecticut Health’s Animal Care Committee. Mice were anesthetized with isoflurane (Vedco Inc., Saint Joseph, MO) and given 1 μg of SEA diluted in 50 μl of BSS, or BSS alone by intranasal (i.n.) route as previously described [17].
Reagents, antibodies and ELISA
Staphylococcus enterotoxin A (SEA) were purchased from Toxin Technology Inc. (Sarasota, FL). Lipopolysaccharide (LPS) from salmonella enteric typhimurium was purchased from Sigma (St. Louis, MO). Mouse recombinant IL-6, IL-12, IL-18, IL-33, IL-36β, IL-36β, and TNF were purchased from R&D Technologies (Kingstown, RI). Mouse recombinant IL-1α and IL-1β were purchased from BD Biosciences (San Jose, CA). IFN-α was purchased from PBL Assay Science (Piscataway, NY). Bovine serum albumin (BSA), was purchased from Gemini Bio-Products (West Sacramento, CA) and bovine lactoferrin and transferrin were purchased from Sigma (St. Louis, MO). Proteinase K and ionomycin were purchased from Life Technology (Grand Island, NY). Phorbol 12-Myristate 13-Acetate (PMA) and Brefeldin A (BFA) was purchased from EMD Millipore Corporation (Billerica, MA). Anti-Lactoferrin rabbit polyclonal antibody (ab135710) was purchased from Abcam (Cambridge, MA), anti-SEA antibody (GTX39112) was purchased from GeneTex and biotinylated using EZ-Link Sulfo-NHS-Biotin biotinylation kit from Thermo Scientific (Asheville, NC). Anti-mouse and anti-rabbit IgG secondary HRP-conjugated antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Avidin-HRP was purchased from BIORAD (Hercules, CA). Anti-CD4-V450, -TCR Vβ3-PE and -TNF-AF700 were purchased from BD biosciences (San Jose, CA). IFN-γ, IL-6, TNF and IL-2 ELISA kits were purchased from BD biosciences Pharmingen (Mountain View, CA). Mouse serum amyloid A (SAA) ELISA kit was purchased from Immunology Consultants Laboratory, Inc. (Portland, OR).
BALF processing and bioassay
BAL was collected by lavage of mouse lung with sterile PBS. BAL was centrifuged at 1,000 rpm at 4°C to separate cells from BALF and 0.2 micron filtered as described earlier [18]. The immunostimulatory activity contained in the BALF was measured using a bioassay adapted from a technique used to measure IL-18 bioactivity [19]. Briefly, RBC depleted mouse splenocytes were incubated overnight with IL-12 (2 ng/ml) in Figs 1 and 2 only, washed and incubated in tissue culture medium without IL-12. After 5 h, splenocytes were seeded in 96-well flat-bottom plates (2 x 106 per well in triplicate) and stimulated with 20 μlBALF from SEA or vehicle (alone) exposed mice for 16–18 h. Supernatants were measured for IL-2 and IFN-γ levels by ELISA.
(A) BALF obtained from mice 16 h after i.n. SEA (1 μg) diluted in BSS (BALF-SEA) or BSS alone (BALF-BSS) were tested in a bioassay as described in material and method section. Briefly, mouse splenocytes were pretreated with IL-12 (2 ng/ml) for 16–18 h, washed and rested for 5 h. Primed splenocytes (200,000/well) were incubated with BALF and controls. Culture supernatants were obtained after 16–18 h and assayed for IFN-γ by ELISA. For comparison, proinflammatory cytokines at a final concentration of 5 ng/ml, except IFN-α (1000 U/ml), and LPS (5 μg/ml) were added to the bioassay. Bar graphs show IFN-γ secretion measured in triplicate. Representative of 1 out 3 experiments is shown. (B) BALFs obtained 8 h and 18 h after i.n. SEA and BSS were compared in a bioassay. Bar graphs show IFN-γ secretion measured in triplicate. Representative of 1 out 4 experiments is shown. (C) BALFs and IL-33 (bioactive protein control) incubated with and without proteinase K at 37°C for 18 h, were either boiled for 10 min or left untreated and tested in a bioassay as described above. Bar graphs show IFN-γ secretion measured in triplicate. Representative of one out three experiments is shown. (D) Culture supernatants obtained with BALFs, cytokines and LPS applied to a bioassay as described in A were tested for IL-2. Bar graphs show IL-2 secretion measured in triplicate. Representative of 1 out 3 experiments is shown. For all panels statistical significance between groups was evaluated by two-tailed Student’s t tests. The error bars indicate the standard error of the mean between triplicates. * p<0.05, *** p<0.001.
BALFs obtained from mice 16 h after i.n. SEA and BSS were tested in a bioassay using C57BL/6 splenocytes (top panel) and TCR β/γ-/- splenocytes (bottom panel) and supernatants were assayed for IFN-γ (A) and IL-2 (B). Bar graphs show cytokines secretion measured in triplicate with n = 3. The data are representative of 3 independent experiments. Statistical significance between BALF-BSS and BALF-SEA was evaluated by two-tailed Student’s t tests. The error bars indicate the standard error of the mean between triplicates.
Chromatography
BALF samples were separated by size exclusion chromatography using Sephacryl 100-HR (Sigma, St Louis, MO) equilibrated in sterile PBS and run at 4°C at 0.5 ml/min. Absorption of the proteins was recorded at 280 nm and fractions collected every 1.5 ml. Anion exchange chromatography was performed using High Q (BIORAD, Hercules, CA) equilibrated with 20 mM phosphate (pH 7.0). The samples were applied to the column, washed at 1 ml/min and eluted with a linear gradient of 20 mM phosphate, 1 M NaCl (pH 7.0). Protein detection was recorded at 280 nm and fractions collected every 1 ml. Cation exchange chromatography was performed using High S (BIORAD, Hercules, CA) equilibrated with 20 mM MES (pH 5.5). The samples were applied to the column, washed at 1.5 ml/min, eluted with a linear gradient of 20 MES (pH 5.5) 1 M NaCl and fractions collected every 1 ml. BALF was also fractionated directly by cation exchange chromatography using the same protocol as described above except the elution gradient was performed with 20 MES (pH 5.5), 2 M NaCl. All buffers were filtered at 0.2 microns.
Mass Spectrometry
The fractions of interest were lyophilized, resuspended in trypsin digestion buffer, peptides were obtained and identified by LC MS/MS on an Orbitrap ELITE (Yale Keck facility, New Heaven, CT) using the following parameters: maximum number of missed cleavage: 1; partial oxidation of met: allow; enzyme: trypsin; monoisotopic mass threshold: 10 PPM; fragment mass tolerance 0.5 Da; monoisotopic or average: monoisotopic. All MS/MS samples were analyzed using Mascot (Matrix Science, London, UK; version Mascot). Mascot searched the SWISSPROT database for the mouse genome (Mus musculus) assuming the digestion by trypsin and was further search for the sequence of SEA (Staphylococcus aureus enterotoxin A).
SDS-PAGE and immunoblotting
BALF and proteins were resuspended in denaturing SDS sample buffer and resolved by SDS-PAGE and immunoblotted as described earlier [20]. Briefly, samples were resolved on 4–15% SDS PAGE, transferred onto 0.2 μm nitrocellulose membrane (BIORAD, Hercules, CA), probed with primary antibodies, washed, then incubated with secondary HRP-conjugated antibodies or avidin-HRP. Western blot detection was performed using ECL plus (Amersham, Arlington Heights, IL).
Flow cytometry
Following previous procedures [20], RBC-depleted spleens and lymph node cells were resuspended with staining buffer (BSS, 3% FBS and 0.1% sodium azide), followed by blocking with a mixture of normal mouse serum, anti-Fc receptor supernatant from the 2.4 G2 hybridoma [21], and human gamma globulin. Incubation with labeled primary antibodies and flow cytometry analysis was carried out as before [20]. For intracellular cytokine staining, spleen cells were cultured for 4 h with SEA or PMA + Ionomycin in the presence of brefeldin A (BFA), and then stained for CD4, Vβ3 and TNF as described before [20]. Cell viability was assessed using MitoFlow kit purchased from Cell Technology (Fremont, CA) as before [22].
Results
BALF of mice treated with SEA contain a stable T cell dependent pro-inflammatory factor
Our previous data demonstrated that SEA inhalation induced significant pulmonary inflammation coincident with the appearance of cell damage suggesting the presence of DAMPs, alarmins and/or cytokines [18]. To study this possibility, BALF from intranasal SEA-treated mice was used to induce IFN-γ release using a sensitive cell bioassay relying on IL-12-preconditioning of spleen cells [19]. BALF from SEA-treated mice (BALF-SEA) contained an activity that can also be induced by factors such as IL-12, IL-36β, and especially IL-33 (Fig 1a) an alarmin involved in SEA mediated lung inflammation [23]. This activity occurred as early as 8 h after SEA treatment, was present at 18 h, and required SEA since vehicle (BALF-BSS) was ineffective (Fig 1b). The activity was proteinase K resistant unless boiled for 10 minutes, suggesting a heat resistant protein, in contrast to IL-33 that was heat and proteinase K sensitive (Fig 1c). Like LPS, BALF-SEA stimulated IFNγ, but this did not rule out a role for adaptive immune cell cytokine production. As expected, neither LPS nor any innate cytokine induced IL-2, but BALF-SEA did (Fig 1d). To test if cytokines produced in the bioassay were derived from innate or adaptive immune cells, IFN-γ and IL-2 were measured using WT versus TCR β/γ mice spleen cells. As expected, IFN-γ was induced by BALF-SEA, IL-12, IL-33 and LPS (Fig 2a, top panel), but BALF-SEA did not induce IFN-γ in the absence of T cells even though the innate factors were able to do so (Fig 2a, bottom panel). Similarly, IL-2 was only induced by BALF-SEA when T cells were present (Fig 2b). Moreover, we looked for classical inflammatory factors secreted in the BALF after SEA inhalation and detected IL-6, serum amyloid A (SAA) but no TNF (Fig 3). IL-6 and SAA were detected in the BALF latter (38h) than the pro-inflammatory activity described in Figs 1 and 2 suggesting the presence of an earlier factor.
BALF obtained 2–88 h after i.n. SEA inhalation was tested for the presence of IL-6, TNF and SAA by ELISA. Plots show cytokine secretion, each symbol represents one mouse. BALF were collected from 10 experiments. The error bars indicate the standard error of the mean between the mice in one group. Statistical significance between the groups was evaluated by two-tailed Student’s t tests.
Purification and identification of the pro-inflammatory activity present in BALF-SEA
We hypothesized that SEA was in the BALF and responsible for this activity. To test this idea an unbiased chromatographic strategy to isolate the factor was performed using the IL-2 bioassay. A size exclusion chromatography approach was used to generate fractions containing IL-2 inducing activity that eluted as a broad peak between 15 and 45 kDa, which was based on the calibration curve (Fig 4a-inset). The fractions with the IL-2 inducing activity from BALF-SEA were pooled, concentrated, and applied for a second step of purification using anion exchange. The corresponding fractions from BALF-BSS were processed identically. Most of the proteins bound the anion exchange column and eluted in the salt gradient (Fig 4b, top panel). In contrast, the majority of the IL-2 inducing activity was detected in the unbound material (Fig 4b, bottom panel), consistent with SEA’s isoelectric point of 7.7. Fractions with the IL-2 inducing activity from Fig 4b were pooled, concentrated and then placed on cation exchange at pH 5.5 (Fig 4c, top). No protein was detected in the salt elution gradient but the IL-2 inducing activity was detected (Fig 4c, bottom). After mass spectrometry of the pooled IL-2 inducing fractions SEA was indeed recovered 16 h post inhalation at trace amounts based on percentage of peptide coverage that was only 9.3% (Table 1).
(A) BALFs were obtained from mice 16 h after i.n. SEA or BSS, concentrated using a 3 kDa centricon and loaded on a size exclusion column (Sephacryl 100-HR). Protein detection was recorded at 280 nm (top panel) and fractions from BALF-SEA and BALF-BSS were tested in the bioassay (bottom panel). BALF-SEA fractions with bioactivity were pooled and concentrated on centricon 3 kDa while the corresponding fractions from BALF-BSS were processed identically. Each sample run was a pool of BALF obtained from 3–5 mice. Representative of 1 out 5 experiments is shown. (B) Samples were concentrated, the buffer was exchanged on a PD-10 column to 20 mM Phosphate pH 7.0, applied to an anion exchange column (High Q) and an elution profile was developed with a linear gradient of 0–1 M NaCl. Protein detection was recorded at 280 nm (top panel), bound and unbound fractions were collected, tested in a bioassay (bottom panel), and pooled as described in A. Representative of 1 out 4 experiments is shown. (C) Samples were concentrated, buffer exchanged on a PD-10 column to 20 mM MES pH 5.5, applied to a cation exchange column (High S) and an elution profile was developed with a linear gradient of 0–1 M NaCl. Protein detection was recorded (top panel), bound and unbound fractions collected, tested (bottom panel) and pooled as described in A. Representative of 1 out 3 experiments is shown.
SEA co-migrates with immunosuppressive proteins
Lactoferrin was also detected in BALF-SEA which seemed to co-migrate with SEA since it was not detected by mass spectrometry in BALF-BSS (Table 1). It has been shown previously that lactoferrin can impede inflammation [24], attenuate SEB activity [25] and therefore we hypothesized it could also affect SEA activity. Immunoblotting showed that lactoferrin was present at high levels in BALF-SEA while SEA was undetectable (Fig 5a). To test if in BALF-SEA lactoferrin and SEA co-migrated, cation exchange chromatography was used and elution occurred between 1.03 M and 1.3 M NaCl (Fig 5b, top panel). Specifically, IL-2-inducing activity was found precisely where lactoferrin was detected (Fig 5b, bottom panel). Thus, lactoferrin may function as an innate factor. Lactoferrin was present in BALF-SEA and -BSS through a time course, and also detectable in BALF from naïve mice (Fig 5c). Thus, SEA fractionates at least with a portion or subspecies of lactoferrin. In sum, lactoferrin is in BALF-BSS and BALF-SEA before protein fractionation, but after fractionation a sub fraction of lactoferrin co-migrates with SEA from BALF-SEA but not in the equivalent fraction from BALF-BSS.
(A) BALF-SEA was resolved by 4–15% gradient SDS-PAGE under reducing and denaturing conditions, transferred to nitrocellulose membrane and probed with anti-lactoferrin and anti-SEA antibodies as described in the Materials and Methods. Purified Lactoferrin and SEA (30 and 10 ng) were loaded in parallel to serve as positive controls. Specificity of the antibody was assessed using isotype control. The data are representative of 2 (lactoferrin) and 3 (SEA) independent experiments. (B) 16 h BALF-SEA was directly fractionated by cation exchange chromatography. Fractions were tested in a bioassay and immunoblotted for lactoferrin (inset). Chromatogram of protein absorption at 280nM (top panel) and bar graphs of IL-2 secretion (bottom panel) are shown. Representative of 1 out 3 experiments is shown. (C) BALF from naïve mice, and BALF from mice having received a time course of i.n. SEA and BSS were immunoblotted with anti-lactoferrin antibodies and 3 experiments are shown.
Next, spleen cells were stimulated with SEA in the presence of lactoferrin to determine its impact on cytokine synthesis. Lactoferrin strongly inhibited IL-2 production while control BSA did not (Fig 6a, top panel). Transferrin was also detected by mass spectrometry in BALF fractions (Table 1), and inhibited IL-2 release. In contrast to stimulation with SEA, cells stimulated with anti-CD3/-CD28 in the presence of lactoferrin, and not transferrin, inhibited IL-2 production (Fig 6a, bottom panel). Thus, IL-2 synthesis from naïve T cells was inhibited by lactoferrin under TCR specific and mitogenic stimulation that also included PMA + Ionomycin (Fig 6b). Effector T cells generated by immunizing mice with SEA + LPS [26,27] were slightly inhibited by lactoferrin, although not to statistical significance (Fig 6b). To assess if transient lactoferrin treatment could mediate long term inhibition of IL-2 production, cells were pre-incubated with lactoferrin or BSA for 3 h and washed before overnight stimulation with SEA or PMA + Ionomycin (Fig 6c). Lactoferrin was washed away as determined immunoblot (Fig 6d). After overnight stimulation comparable IL-2 levels from lactoferrin and BSA treated cells were observed (Fig 6e), demonstrating that continuous presence of lactoferrin was required.
(A) Mouse splenocytes stimulated with SEA (10 ng/ml) (top panel) and soluble anti-CD3/-CD28 (each 1 μg/ml) (bottom panel) were co-incubated with increasing doses of BSA (negative control), lactoferrin and transferrin. Culture supernatants were assayed for IL-2 after 18 h and the bar graphs show IL-2 secretion. Data are representative of the average of 3 independent experiments with n = 3. The error bars indicate the standard error of the mean between 3 biological replicates. Statistical significance between treatments was evaluated by two-tailed Student’s t tests. * p<0.05, ** p<0.01, *** p<0.001. (B) Mouse splenocytes from either unimmunized mice (Naïve) or 3 day post immunization with 0.5 μg SEA and 10 μg LPS (Effector) were incubated with lactoferrin (white bars) or BSA (grey bars) (5 mg/ml) at 37°C with PMA (50 ng/ml) + Ionomycin (1 μg/ml). Culture supernatants were assayed for IL-2 after 18 h and the data are representative of the average of 3 independent experiments with n = 3. The error bars indicate the standard error of the mean between 3 biological replicates. Statistical significance between lactoferrin and BSA was evaluated by two-tailed Student’s t tests. (C) Cells were incubated with lactoferrin or BSA (5 mg/ml) for 3 h at 37°C, washed 4 times with tissue culture medium, and then stimulated with SEA (10 ng/ml) or PMA (50 ng/ml) + Ionomycin (1 μg/ml). Culture supernatants were assayed for IL-2 and IFN-γ after 18 h. (D) The four successive washes described above in C were analyzed by SDS-PAGE and immunoblotted with anti-lactoferrin as described in legend of Fig 5a. Representative of 1 out 3 experiments is shown. (E) IL-2 and IFN-γ secretion from the culture supernatants (as described above in C) are shown as bar graphs. Data representative of the average of 3 independent experiments are presented. The error bars indicate the standard error of the mean between triplicates.
Lactoferrin causes T cell death and blocks early cytokine synthesis
Based on these data we hypothesized that lactoferrin-mediated inhibition might be the result of decreased cell survival. To test this, the cells were stained with a marker of mitochondrial integrity after overnight stimulation as described in Fig 6a and 6b. After incubation with lactoferrin the cells had decreased mitochondrial integrity compared to BSA (Table 2), suggesting reduced cell viability. However, cells washed after 3 h incubation with lactoferrin or BSA had comparable levels of mitochondrial integrity (Table 3). Thus, overnight incubation with lactoferrin was required to reduce IL-2 production likely by inducing cell death. To address if lactoferrin could perturb early cytokine production before affecting cell viability, intracellular cytokine staining of TNF after 4 h of stimulation with PMA + Ionomycin was used. SEA-specific T cells were stained with anti-CD4, -Vβ3 and -TNF (Fig 7a). Incubation with lactoferrin inhibited TNF synthesis in naïve T cells to a much greater extent than effector T cells (Fig 7b, left panels), however, increasing lactoferrin levels did begin to block TNF production (Fig 7b, bottom panel). Similar to IL-2 production (Fig 6), lactoferrin did not inhibit TNF production when it was washed away before stimulation with the mitogen (Fig 7b, right panels), demonstrating that cells exposed to lactoferrin are not irreversibly programmed to die. Thus, lactoferrin inhibits cytokine production more effectively in naïve than in effector T cells by causing cell death and blocking early cytokine synthesis.
(A) Splenocytes from naïve mice were incubated with lactoferrin or BSA (5 mg/ml), in the presence of BFA and stimulated with or without PMA + Ionomycin for 4 h at 37°C. Cells were stained for CD4 and TCR Vβ3, fixed, permeabilized and stained for intracellular TNF. Representative dot plots show the gating strategy. The percentage of SEA-specific T cells: CD4+Vβ3+, is indicated in top right quadrant (left panels) and the percentage of those cells producing TNF is indicated as histograms in the right panels. (B) The mean percentage +/- standard error of the mean of CD4+Vβ3+ cells producing TNF from 3 independent biological replicates was analyzed as described in A is shown. Hence, naïve and effector splenocytes were obtained as described in the legend of Fig 6. One set of cells were directly incubated with lactoferrin or BSA in the presence of BFA and PMA + Ionomycin (No wash, left panels) and stained as described in A. Another set of cells (Washed, right panels) was first incubated for 3 h with lactoferrin or BSA, washed 4 times as described in legend of Fig 6c and then stimulated with or without PMA + Ionomycin for 4 h at 37°C. Data are combined from 3 experiments with n = 3 and displayed as mean +/- standard error of the mean. Statistical significance between BSA and lactoferrin was evaluated by two-tailed Student’s t tests. * p<0.05, ** p<0.01, *** p<0.001.
Discussion
Here it is shown that an exogenous pathogenic protein such as SEA can be detected outside of an active infection site using a functional bioassay. Moreover, we demonstrate a way mucosal immunity contains the powerful activity of superantigens. Firstly, trace levels of inhaled SEA were recovered from BALF as determined by mass spectrometry. This was evident even after the initiation of pulmonary inflammation and importantly the recovered SEA was bioactive showing the mucosal innate immune system could not entirely inactivate SEA even late after inhalation. Secondly, SEA fractionated with lactoferrin and transferrin demonstrating the potential of innate immune factors to shield adaptive immune cell stimulation from external agents. Lastly, lactoferrin inhibited SEA-mediated T cell cytokine production by enhancing cell death using a mechanism that relied on the continual presence of lactoferrin. Collectively, these results suggest that bioassays can add an important tool for detecting functional pathogen by-products in ways that molecular diagnostics cannot, and demonstrate that an innate barrier exists in the lung even prior to an inflammatory insult.
The bioactivity of BALF at 18 h after SEA intranasal instillation initially suggested the presence of a DAMP, alarmin, and/or cytokine with the capacity to stimulate IFN-γ release on IL-12 preconditioned cells. The activity was similar to IL-12 but less powerful than LPS or the alarmin IL-33 and resistant to boiling but sensitive to proteinase K (Fig 1a–1c). Nevertheless, the capacity to stimulate IL-2 supported the notion that SEA itself was recovered even though this was 16 h into the pulmonary inflammatory response (Fig 1d). In fact, the bioactivity of the BALF-SEA was fairly constant over a time course (data not shown), contrasting to previous data showing a 7-log reduction in viral particle recovery within 18 h of infection [28]. Several other attributes of superantigens suggested that recovery of SEA would be unlikely. The high affinity of SEA for MHC II is well documented [29]. Further, considering that alveolar macrophages and dendritic cells express MHC II [30], and SEA inhalation induces the rapid migration of monocytes to lung followed by upregulation of MHC II [17,31] it was reasoned that SEA would not be recoverable at later time points. Finally, the systemic nature of the SEA-induced response after inhalation is clear [6], which strongly suggests that SEA moves rapidly into the bloodstream. Nevertheless, BALF-SEA did not induce IFN-γ or IL-2 release in spleen cells devoid of T cells (Fig 2). Thus, this strongly suggested that SEA was in BALF and its bioactivity maintained. Using size exclusion chromatography it was demonstrated that SEA bioactivity was detected in fractions corresponding to a molecular weight between 15–45 kDa, and after subjecting these fractions to anion and cationic exchange a highly pure sample was recovered (Fig 4). After mass spectrometry it was clear that SEA was present. To our knowledge purification of bioactive SEA from BALF has not been previously reported, but interestingly only in trace amounts with about 9% peptide coverage (Table 1). In fact, immunoblotting for SEA yielded no detectable signal (Fig 5a). These results show the importance of a functional readout compared to detection alone.
While there is no doubt that detection is extremely useful in diagnosis and prediction outcomes, there is space for expanding the use of functional bioassays to better understand the effects of a presumed pathogen or factor. For instance, BALF analysis of ICU patients could include functional tests of fluid such as described here and elsewhere [32,33]. An example is the ARDS network established by the NIH which has gathered an impressive collection of BAL and respiratory fluid [34], among which some could be tested for enterotoxin bioactivity when Staphylococcus aureus is an established risk factor [35]. Although mouse cells were used here, human BALF containing enterotoxins should stimulate responses just as well since SEA is not species specific per se. Also, mouse cells devoid of T cells could be used as identity controls or cells from HLA-DR transgenic mice [36,37].
The presence of immunosuppressive factors in the BALF that co-fractionated with SEA was quite surprising given that this occurred well into the inflammatory response. One possibility is that the co-fractionation of lactoferrin, lysozyme and transferrin (Table 1) may be an innate barrier that is present even without the induction of immunity. In this study we focused on lactoferrin, an iron-binding glycoprotein with known immune effects [38,24], since it was detected with SEA (Fig 5b). Lactoferrin is found in mucosal secretions, synthesized by epithelial cells, present in neutrophilic granules [39] and in colostrum’s at concentrations up to 9 g/L [40]. Lactoferrin is a defensive factor in mice challenged with methicillin-resistant Staphylococcus aureus (MRSA) [38], and in cows infected with beta-lactamase resistant Staphylococcus aureus [41]. Moreover, lactoferrin has been shown to attenuate SEB activity [25], and clinically it has been used to prevent neonatal sepsis [42], and necrotizing enter colitis [43,40]. Thus, it is possible that these immunosuppressive factors can inhibit an environmental insult before an immune response is initiated. The data detecting lactoferrin in BALF from naïve mice is consistent with lactoferrin being part of a protective shield even in the absence of provocation. Nevertheless, since neutrophils release lactoferrin [44] it is also possible that it can be carried to sites of inflammation.
Of the two iron-binding proteins co-purified with SEA, lactoferrin had a broader spectrum of action compared to transferrin since it attenuated both anti-CD3/-CD28 and SEA stimulation while transferrin inhibited only SEA stimulation (Fig 6a). The reason for this difference is unclear, but may be related to the iron chelation capacity of transferrin [38]. Lactoferrin was more efficient at reducing IL-2 and TNF production from naïve than from effector cells. This could be linked to a differential expression of lactoferrin receptors after T cell activation [45]. Secondly, lactoferrin-mediated cytokine inhibition was in part due to a decrease of primary T cell survival (Table 2), possibly through a similar apoptotic mechanism as previously described in human T leukemia cells [46]. Lastly, a faster process occurred in the reduction of TNF at 4 h (Fig 7).
Conclusions
In summary, our work suggests that detecting SEA bioactivity in BALF can be used as a functional clinical diagnostic, especially when molecular diagnostics fail. Two such cases might involve status asthmaticus [47] and undetected co-infections [11,48]. Importantly, the copurification of immunosuppressive proteins with SEA suggests that they could be applied as countermeasures against enterotoxin-mediated lung inflammation and injury.
Acknowledgments
The authors would like to thank Drs. Francisco Sylvester (University of North Carolina, Chapel Hill, NC) and Adam Adler (Department of Immunology, UCHC, Farmington CT) for helpful comments and suggestions.
Author Contributions
Conceived and designed the experiments: AV AM. Performed the experiments: AM BB JS. Analyzed the data: AV AM BB JS. Wrote the paper: AV AM.
References
- 1. Nguyen T, Kyle UG, Jaimon N, Tcharmtchi MH, Coss-Bu JA, Lam F, et al. (2012) Coinfection with Staphylococcus aureus increases risk of severe coagulopathy in critically ill children with influenza A (H1N1) virus infection. Crit Care Med 40: 3246–50. pmid:22971587
- 2. Robinson KM, Choi SM, McHugh KJ, Mandalapu S, Enelow RI, Kolls JK, et al. (2013) Influenza A exacerbates Staphylococcus aureus pneumonia by attenuating IL-1beta production in mice. J Immunol 191: 5153–9. pmid:24089191
- 3. Krakauer T, Stiles BG (2013) The staphylococcal enterotoxin (SE) family: SEB and siblings. Virulence 4: 759–73. pmid:23959032
- 4. McCormick JK, Yarwood JM, Schlievert PM (2001) Toxic shock syndrome and bacterial superantigens: an update. Annu Rev Microbiol 55: 77–104. pmid:11544350
- 5. Younis RT, Lazar RH (1996) Delayed toxic shock syndrome after functional endonasal sinus surgery. Arch Otolaryngol Head Neck Surg 122: 83–5. pmid:8554751
- 6. Muralimohan G, Rossi RJ, Guernsey LA, Thrall RS, Vella AT (2008) Inhalation of Staphylococcus aureus enterotoxin A induces IFN-gamma and CD8 T cell-dependent airway and interstitial lung pathology in mice. J Immunol 181: 3698–705. pmid:18714046
- 7. Herz U, Ruckert R, Wollenhaupt K, Tschernig T, Neuhaus-Steinmetz U, Pabst R, et al. (1999) Airway exposure to bacterial superantigen (SEB) induces lymphocyte-dependent airway inflammation associated with increased airway responsiveness—a model for non-allergic asthma. Eur J Immunol 29: 1021–31. pmid:10092107
- 8. Rajagopalan G, Iijima K, Singh M, Kita H, Patel R, David CS (2006) Intranasal exposure to bacterial superantigens induces airway inflammation in HLA class II transgenic mice. Infect Immun 74: 1284–96. pmid:16428778
- 9. Guven M, Karabay O, Akidil O, Yilmaz MS, Yildirim M (2013) Detection of staphylococcal exotoxins in antrochoanal polyps and chronic rhinosinusitis with nasal polyps. Otolaryngol Head Neck Surg 148: 302–7. pmid:23112271
- 10. Huvenne W, Hellings PW, Bachert C (2013) Role of staphylococcal superantigens in airway disease. Int Arch Allergy Immunol 161: 304–14. pmid:23689556
- 11. Highet AR, Goldwater PN (2009) Staphylococcal enterotoxin genes are common in Staphylococcus aureus intestinal flora in Sudden Infant Death Syndrome (SIDS) and live comparison infants. FEMS Immunol Med Microbiol 57: 151–5. pmid:19702876
- 12. Corrigan CJ, Hartnell A, Kay AB (1988) T lymphocyte activation in acute severe asthma. Lancet 1: 1129–32. pmid:2896958
- 13. Tunceroglu H, Shah A, Porhomayon J, Nader ND (2013) Biomarkers of lung injury in critical care medicine: past, present, and future. Immunol Invest 42: 247–61. pmid:23473375
- 14. Mino MJ, Ortiz RT, Randad P, Moffatt LT, Jordan MH, Shupp JW (2013) Localization of superantigen virulence factors in kidney tissue of animals with Staphylococcus aureus-infected burn wounds. J Burn Care Res 34: 142–50. pmid:23292582
- 15. Degauque N, Boeffard F, Foucher Y, Ballet C, Brouard S, Soulillou JP (2011) The blood of healthy individuals exhibits CD8 T cells with a highly altered TCR Vb repertoire but with an unmodified phenotype. PLoS One 6: e21240. pmid:21738624
- 16. Posnett DN, Sinha R, Kabak S, Russo C (1994) Clonal populations of T cells in normal elderly humans: the T cell equivalent to "benign monoclonal gammapathy". J Exp Med 179: 609–18. pmid:8294871
- 17. Kumar S, Colpitts SL, Menoret A, Budelsky AL, Lefrancois L, Vella AT (2013) Rapid alphabeta T-cell responses orchestrate innate immunity in response to Staphylococcal enterotoxin A. Mucosal Immunol 6: 1006–15. pmid:23321986
- 18. Menoret A, Kumar S, Vella AT (2012) Cytochrome b5 and cytokeratin 17 are biomarkers in bronchoalveolar fluid signifying onset of acute lung injury. PLoS One 7: e40184. pmid:22792238
- 19. Lewis EC, Dinarello CA (2006) Responses of IL-18- and IL-18 receptor-deficient pancreatic islets with convergence of positive and negative signals for the IL-18 receptor. Proc Natl Acad Sci U S A 103: 16852–7. pmid:17075045
- 20. Menoret A, McAleer JP, Ngoi SM, Ray S, Eddy NA, Fenteany G, et al. (2009) The oxazolidinone derivative locostatin induces cytokine appeasement. J Immunol 183: 7489–96. pmid:19917702
- 21. Unkeless JC (1979) Characterization of a monoclonal antibody directed against mouse macrophage and lymphocyte Fc receptors. J Exp Med 150: 580–96. pmid:90108
- 22. Lee SJ, Long M, Adler AJ, Mittler RS, Vella AT (2009) The IKK-neutralizing compound Bay11 kills supereffector CD8 T cells by altering caspase-dependent activation-induced cell death. J Leukoc Biol 85: 175–85. pmid:18923104
- 23. Ngoi SM, St Rose MC, Menoret AM, Smith DE, Tovey MG, Adler AJ, et al. (2012) Presensitizing with a Toll-like receptor 3 ligand impairs CD8 T-cell effector differentiation and IL-33 responsiveness. Proc Natl Acad Sci U S A 109: 10486–91. pmid:22689946
- 24. Legrand D, Mazurier J (2010) A critical review of the roles of host lactoferrin in immunity. Biometals 23: 365–76. pmid:20143251
- 25. Hayworth JL, Kasper KJ, Leon-Ponte M, Herfst CA, Yue D, Brintnell WC, et al. (2009) Attenuation of massive cytokine response to the staphylococcal enterotoxin B superantigen by the innate immunomodulatory protein lactoferrin. Clin Exp Immunol 157: 60–70. pmid:19659771
- 26. McAleer JP, Liu B, Li Z, Ngoi SM, Dai J, Oft M, et al. (2010) Potent intestinal Th17 priming through peripheral lipopolysaccharide-based immunization. J Leukoc Biol 88: 21–31. pmid:20130220
- 27. Vella AT, McCormack JE, Linsley PS, Kappler JW, Marrack P (1995) Lipopolysaccharide interferes with the induction of peripheral T cell death. Immunity 2: 261–70. pmid:7535182
- 28. Pickett TE, Pasetti MF, Galen JE, Sztein MB, Levine MM (2000) In vivo characterization of the murine intranasal model for assessing the immunogenicity of attenuated Salmonella enterica serovar Typhi strains as live mucosal vaccines and as live vectors. Infect Immun 68: 205–13. pmid:10603389
- 29. Tiedemann RE, Urban RJ, Strominger JL, Fraser JD (1995) Isolation of HLA-DR1.(staphylococcal enterotoxin A)2 trimers in solution. Proc Natl Acad Sci U S A 92: 12156–9. pmid:8618863
- 30. Aggarwal NR, King LS, D'Alessio FR (2014) Diverse macrophage populations mediate acute lung inflammation and resolution. Am J Physiol Lung Cell Mol Physiol 306: L709–25. pmid:24508730
- 31. Muralimohan G, Rossi RJ, Vella AT (2008) Recruitment and in situ renewal regulate rapid accumulation of CD11c+ cells in the lung following intranasal superantigen challenge. Int Arch Allergy Immunol 147: 59–73. pmid:18451649
- 32. Rasooly R, Do PM (2009) In vitro cell-based assay for activity analysis of staphylococcal enterotoxin A in food. FEMS Immunol Med Microbiol 56: 172–8. pmid:19453753
- 33. Rasooly R, Hernlem BJ (2014) Quantitative analysis of staphylococcus enterotoxin A by differential expression of IFN-gamma in splenocyte and CD4(+) T-cells. Sensors (Basel) 14: 8869–76.
- 34. Agrawal A, Zhuo H, Brady S, Levitt J, Steingrub J, Siegel MD, et al. (2012) Pathogenetic and predictive value of biomarkers in patients with ALI and lower severity of illness: results from two clinical trials. Am J Physiol Lung Cell Mol Physiol 303: L634–9. pmid:22865551
- 35. Randolph AG, Vaughn F, Sullivan R, Rubinson L, Thompson BT, Yoon G, et al. (2011) Critically ill children during the 2009–2010 influenza pandemic in the United States. Pediatrics 128: e1450–8. pmid:22065262
- 36. DaSilva L, Welcher BC, Ulrich RG, Aman MJ, David CS, Bavari S (2002) Humanlike immune response of human leukocyte antigen-DR3 transgenic mice to staphylococcal enterotoxins: a novel model for superantigen vaccines. J Infect Dis 185: 1754–60. pmid:12085321
- 37. Fukui Y, Esaki Y, Kimura A, Hirokawa K, Nishimura Y, Sasazuki T (1993) T-cell repertoire in a strain of transgenic C57BL/6 mice with the HLA-DRA gene on the X-chromosome. Immunogenetics 37: 204–11. pmid:8420827
- 38. Actor JK, Hwang SA, Kruzel ML (2009) Lactoferrin as a natural immune modulator. Curr Pharm Des 15: 1956–73. pmid:19519436
- 39. Borregaard N, Sorensen OE, Theilgaard-Monch K (2007) Neutrophil granules: a library of innate immunity proteins. Trends Immunol 28: 340–5. pmid:17627888
- 40. Sherman MP, Miller MM, Sherman J, Niklas V (2014) Lactoferrin and necrotizing enterocolitis. Curr Opin Pediatr 26: 146–50. pmid:24503532
- 41. Petitclerc D, Lauzon K, Cochu A, Ster C, Diarra MS, Lacasse P (2007) Efficacy of a lactoferrin-penicillin combination to treat {beta}-lactam-resistant Staphylococcus aureus mastitis. J Dairy Sci 90: 2778–87. pmid:17517718
- 42. Turin CG, Zea-Vera A, Pezo A, Cruz K, Zegarra J, Bellomo S, et al. (2014) Lactoferrin for prevention of neonatal sepsis. Biometals 27: 1007–16. pmid:24935001
- 43. Manzoni P, Meyer M, Stolfi I, Rinaldi M, Cattani S, Pugni L, et al. (2014) Bovine lactoferrin supplementation for prevention of necrotizing enterocolitis in very-low-birth-weight neonates: a randomized clinical trial. Early Hum Dev 90 Suppl 1: S60–5. pmid:24709463
- 44. Jacobsen LC, Sorensen OE, Cowland JB, Borregaard N, Theilgaard-Monch K (2008) The secretory leukocyte protease inhibitor (SLPI) and the secondary granule protein lactoferrin are synthesized in myelocytes, colocalize in subcellular fractions of neutrophils, and are coreleased by activated neutrophils. J Leukoc Biol 83: 1155–64. pmid:18285402
- 45. Suzuki YA, Lopez V, Lonnerdal B (2005) Mammalian lactoferrin receptors: structure and function. Cell Mol Life Sci 62: 2560–75. pmid:16261254
- 46. Mader JS, Richardson A, Salsman J, Top D, de Antueno R, Duncan R, et al. (2007) Bovine lactoferricin causes apoptosis in Jurkat T-leukemia cells by sequential permeabilization of the cell membrane and targeting of mitochondria. Exp Cell Res 313: 2634–50. pmid:17570361
- 47. ten Brinke A, Sterk PJ, Masclee AA, Spinhoven P, Schmidt JT, Zwinderman AH, et al. (2005) Risk factors of frequent exacerbations in difficult-to-treat asthma. Eur Respir J 26: 812–8. pmid:16264041
- 48. Rudkjobing VB, Aanaes K, Wolff TY, von Buchwald C, Johansen HK, Thomsen TR (2014) An exploratory study of microbial diversity in sinus infections of cystic fibrosis patients by molecular methods. J Cyst Fibros 13: 645–52. pmid:24636809