Attenuation of massive cytokine response to the staphylococcal enterotoxin B superantigen by the innate immunomodulatory protein lactoferrin

Abstract

Staphylococcal enterotoxin B (SEB) is a pyrogenic exotoxin and a potent superantigen which causes massive T cell activation and cytokine secretion, leading to profound immunosuppression and morbidity. The inhibition of SEB-induced responses is thus considered a goal in the management of certain types of staphylococcal infections. Lactoferrin (LF) is a multi-functional glycoprotein with both bacteriostatic and bactericidal activities. In addition, LF is known to have potent immunomodulatory properties. Given the anti-microbial and anti-inflammatory properties of this protein, we hypothesized that LF can modulate T cell responses to SEB. Here, we report that bovine LF (bLF) was indeed able to attenuate SEB-induced proliferation, interleukin-2 production and CD25 expression by human leucocyte antigen (HLA)-DR4 transgenic mouse T cells. This inhibition was not due to bLF's iron-binding capacity, and could be mimicked by the bLF-derived peptide lactoferricin. Cytokine secretion by an engineered SEB-responsive human Jurkat T cell line and by peripheral blood mononuclear cells from healthy donors was also inhibited by bLF. These findings reveal a previously unrecognized property of LF in modulation of SEB-triggered immune activation and suggest a therapeutic potential for this naturally occurring protein during toxic shock syndrome.

Introduction

Superantigens are pathogen-derived toxins that induce massive activation of T cells resulting in systemic toxicity and profound immunosuppression with potentially detrimental consequences [1]. Unlike conventional protein antigens, that need to be degraded proteolytically into antigenic peptides and presented in the context of self major histocompatibility complex (MHC) molecules to cognate T cells, intact superantigens bind MHC class II glycoproteins outside their peptide binding groove and interact simultaneously with the Vβ region of many T cell receptors (TCRs) [2]. This distinct mode of recognition enables superantigens to stimulate a large fraction of T cells regardless of their antigen specificity and to provoke excessive cytokine secretion. These changes account for adverse systemic effects attributed to superantigens and could jeopardize the host's adaptive immunity to superantigen-harbouring infectious agents.

Staphylococcal enterotoxin B (SEB) is one of the many potent pyrogenic exotoxins produced by Staphylococcus aureus, and a prototype bacterial superantigen [3]. SEB causes a variety of clinical diseases ranging from mild and often self-limiting food poisoning to lethal non-menstrual toxic shock [3,4]. SEB is one of the major virulence factors in the context of staphylococcal infections caused by antibiotic-resistant strains [5], which represent a growing public health concern. Various preventive and therapeutic strategies have been tested in experimental animals to counter the toxicity and lethality of bacterial superantigens, including SEB. These include passive administration of neutralizing antibodies against SEB [6,7], SEB toxoid vaccines [6,8], recombinant SEB mutants with attenuated immunotoxicity [9–11] and engineered high-affinity binding inhibitors [12,13]. However, there are currently no effective treatments or vaccines approved for human use against SEB.

Lactoferrin (LF) is a multi-functional glycoprotein found abundantly in colostrum, milk and other exocrine secretions in various species, including but not restricted to bovines, humans and mice [14,15]. Potent anti-microbial and immunomodulatory properties of LF implicate this naturally occurring molecule as an attractive therapeutic candidate for infectious diseases. Bacteriostatic and bactericidal activities of LF against Gram-negative microbes are attributed to its ability to chelate iron, which is required for the growth of certain bacteria, and to destabilize the outer membrane of their cell wall [14,16,17]. LF also reportedly protects against septic shock induced by lipopolysaccharide (LPS) or Gram-negative bacterial infection in several animal models [18–21]. LF's anti-microbial activities against Gram-positive pathogens, including methicillin-resistant S. aureus, have also been reported in both in vitro and in vivo settings [22]. However, whether LF alters excessive and harmful immune responses associated with staphylococcal superantigens in general, and SEB in particular, remains essentially unexplored. This is an important question, given the anti-inflammatory characteristics of LF [23] and in light of the fact that prevention or inhibition of proinflammatory cytokine production is an attractive area of intervention for in vitro and in vivo testing of therapeutics against SEB-induced toxic shock. Given the anti-microbial and anti-inflammatory properties of LF, we hypothesized that this innate glycoprotein can modulate T cell responses to SEB. In this report, we demonstrate for the first time that bovine LF (bLF) is able to inhibit mouse and human T cell activation and cytokine secretion in response to this superantigen, and discuss how bLF may modulate SEB-induced responses.

Materials and methods

Mouse strains

C57BL/6 (B6) mice, 6–12 weeks old, were purchased from Charles River Canada (St-Constant, Québec, Canada). Human leucocyte antigen (HLA)-DR4-IE transgenic mice, referred to hereafter as DR4 Tg mice, were housed and bred under specific pathogen-free conditions in a barrier facility at the University of Western Ontario. These mice lack endogenous major histocompatibility complex (MHC) class II but express a chimeric MHC class II molecule composed of HLA-DRA-IEα and HLA-DRB1*0401-IEβ[24]. All mice were cared for in accordance with institutional regulations and the guidelines established by the Canadian Council on Animal Care. Both female and male mice were used in our experiments.

Cell lines

A Jurkat cell line expressing the Vβ17 chain of the human T cell receptor (TCR) was generated as follows. The cDNA for Vβ17·1 TCR chain (clone N17) was a kind gift from Dr Hongmin Li (Wadsworth Center, New York State Department of Health, USA). Recombinant Vβ17 was constructed and expressed in the Jurkat T cell line JRT3-T3·5 (American Tissue Culture Collection, Manassas, VA, USA), which lacks the expression of the endogenous Vβ8·1 chain present in WT Jurkat T cells [25]. The leader and transmembrane DNA sequences of human hVβ8·1 [26] were attached to the 5′ and 3′ ends of hVβ17 cDNA [27], respectively, to allow for surface expression and pairing of recombinant Vβ17 with the endogenous Jurkat TCR Vα1. Modifications to the Vβ17 cDNA were performed by sequential megaprimer polymerase chain reaction (PCR), and the Vβ17 cDNA was cloned into the unique BamHI site of pBIG2i [28]. Ten µg of linearized pBIG2i :: hVβ17 was nucleofected into 5 × 10⁶ JRT3-T3·5 cells, and stable transfectants were selected using increasing concentrations of hygromycin B. Surface expression of Vβ17 by these cells was confirmed by flow cytometry.

Human lymphoblastoid B cell line LG2 [29] or human bare lymphocyte syndrome (BLS) B cells transfected stably with DRA1*0101/DRB1*0401 to express HLA-DR4 [30], which will be referred to as BLS–DR4, were used as antigen-presenting cells (APCs) required for superantigen-induced activation of Vβ17⁺ Jurkat T cells. LG2 and BLS–DR4 cells were provided generously by Dr Eric Long (National Institutes of Health, Rockville, MD, USA) and Dr Malak Kotb (University of Cincinnati College of Medicine, USA), respectively. To ensure HLA-DR4 expression in the BLS–DR4 line, cells were selected for more than a month with 300 µg/ml hygromycin B (HyClone) and 4 mM L-histindiol (Sigma-Aldrich St Louis, MO, USA). Cell lines were maintained in RPMI-1640 medium supplemented with 10% heat-inactivated fetal bovine serum (FBS), non-essential amino acids, 2 mM L-glutamine, 1 mM sodium pyruvate, 10 mM HEPES, 100 U/ml penicillin and 100 µg/ml streptomycin (hereafter referred to as complete medium). We confirmed high expression levels of HLA-DR4 on BLS–DR4 cells by flow cytometry. Cells were grown at 37°C and humidified atmosphere containing 5% CO₂.

Reagents

Bovine lactoferrin (bLF) purified from cow's milk with a purity of 97·45% and iron saturation level of 14·09% was purchased from MJS BioLynx Inc. (Brockville, Ontario, Canada). The powder was dissolved in sterile phosphate-buffered saline (PBS) and stored at –20°C until use. N-terminus acetylated bovine lactoferricin (LfcinB) [(ACE)FKCRRWQWRMKKLGAPSITCVRRAF(OH)] was custom-synthesized by American Peptide Company Inc. (Sunnyvale, CA, USA), purified by high performance liquid chromatography (HPLC) and analysed by mass spectrometry. The purity of LfcinB was greater than 95%. Working solution of LfcinB in RPMI-1640 medium was prepared freshly from the LfcinB powder stored at –20°C, and used immediately. Desferrioxamine mesilate was from Mayne Pharma Inc. (Kirkland, Québec, Canada). The powder was dissolved in water and added to complete medium at a final concentration of 50 µg/ml to prepare media depleted of free iron, which was used immediately. Alternatively, 2,2′-dipyridyl ReagentPlus with ≥ 99% purity (Sigma-Aldrich) was used at a final concentration of 100 µM in order to chelate iron in complete medium. Bovine serum albumin (BSA, fraction V) and bovine apo-transferrin were both obtained from Sigma-Aldrich, and used at doses similar to bLF concentrations in control experiments.

SEB production and purification

Recombinant SEB was cloned from S. aureus strain COL, expressed in Escherichia coli BL21 (DE3), and purified to apparent homogeneity as we have described recently [31]. An inactive mutant of SEB was also generated by site-directed mutagenesis (SEB_N23A) for use as a negative control. This position is analogous to Asn²³ in staphylococcal enterotoxin C3, and is known to be critical for binding to mouse Vβ8·2 [32].

Endotoxin testing

Serial dilutions of various biological preparations used in our studies (e.g. SEB, SEB_N23A, BSA, apotransferrin) were prepared and tested for endotoxin contamination using a highly sensitive (sensitivity: 0·1 endotoxin unit/ml) and quantitative colorimetric assay kit (QCL-1000® Endpoint Chromogenic Limulus Amoebocyte Lysate Assay; Lonza Ltd., Walkersville, MD, USA) as per the manufacturer's instructions.

T cell proliferation

DR4 Tg mice were killed in a CO₂ chamber, and spleen cell suspensions were prepared in cold PBS using a glass tissue homogenizer. Erythrocytes were depleted by treatment with ACK lysis buffer (BioWhittaker Inc., Walkersville, MD, USA), and splenocytes were seeded at 4 × 10⁵ cells/200 µl/well of U-bottomed microtitre plates. Cells were stimulated with either 5 or 500 ng/ml of SEB in the presence or absence of the indicated doses of bLF or LfcinB, and plates were incubated for 72 or 96 h at 37°C in a humidified atmosphere containing 6% CO₂. To measure SEB-induced T cell proliferation, cells were pulsed with 0·5 µCi of tritiated thymidine ([³H]TdR) during the final 8 h of the cultures. Cultures were harvested subsequently onto glass fibre filter mats, and [³H]TdR incorporation into cellular DNA was determined by liquid scintillation counting. For proliferation and other functional assays in which the effects of bLF and LfcinB were examined, fetal bovine serum (FBS) was used at 10% and 0·5%, respectively.

Cytokine quantification

Erythrocyte-depleted splenocytes from DR4 Tg mice were seeded at 4 × 10⁵ cells/200 µl/well and stimulated with SEB in the presence or absence of bLF or LfcinB. After 24-h incubation, culture supernatants were harvested and stored at –80°C. The interleukin (IL)-2 and interferon (IFN)-γ content of these samples was quantified using mouse enzyme-linked immunosorbent assay (ELISA) kits (sensitivity: 2 pg/ml and 15 pg/ml for IL-2 and IFN-γ, respectively) purchased from eBioscience (San Diego, CA, USA).

To examine the effect of bLF on human T cell responses to SEB, human peripheral blood mononuclear cells (PBMCs) were isolated from heparinized peripheral blood of six healthy volunteers (three females and three males; age range: 21–40 years) by Ficoll-Hypaque density gradient centrifugation, according to a protocol approved by the University of Western Ontario Research Ethics Board for Health Sciences Research Involving Human Subjects. PBMCs were resuspended in complete medium containing 2 µg/ml polymyxin B, seeded at 1 × 10⁵ cells/well of a 96-well U-bottomed plate and stimulated with SEB in the presence or absence of bLF. In some experiments, Vβ17⁺ Jurkat T cells were mixed with LG-2 or BLS HLA-DR4 cells at a 5:1 T : APC ratio and stimulated with SEB in the presence or absence of bLF. Between 16 and 18 h later, culture supernatants were harvested and IL-2 concentrations in samples were determined by a human IL-2 ELISA kit (eBioscience) with a sensitivity of 4 pg/ml.

Flow cytometry

To confirm the expression of HLA-DR4 on DR4 Tg splenocytes, 1 × 10⁶ cells were placed in fluorescence activated cell sorter (FACS) tubes (BD Biosciences, San Jose, CA, USA), washed with cold FACS buffer (PBS containing 0·5% BSA) and incubated with 5 µg/ml of anti-mouse CD16/CD32 monoclonal antibody (mAb) (clone 2·4G2, Fc Block from eBioscience) for 20 min on ice. Without removing the Fc Block, cells were stained with either 1 µg (per million cells) of a fluorescein isothiocyanate (FITC)-labelled anti-human DR mAb [clone L243 (G46-6)] or FITC-conjugated mouse IgG2a isotype control (both purchased from BD Biosciences). Cells were incubated on ice for 30 min, washed twice with cold FACS buffer and once with cold PBS and fixed immediately with 1% paraformaldehyde. Samples were then acquired using a BD FACSCalibur equipped with the CellQuest acquisition software. Data were analysed using FlowJo software (Tree Star Inc., Ashland, OR, USA).

CD25 expression on conventional [(CD4⁺forkhead box P3 (FoxP3)^–] and regulatory (CD4⁺FoxP3⁺) T cells was assessed by flow cytometry using a mouse regulatory T cell (T_reg) kit purchased from eBioscience (cat. no. 88–8118), as per the manufacturer's instructions. In brief, 1 × 10⁶ splenocytes were washed and incubated on ice with 5 µg/ml of Fc block for ∼20 min followed by surface staining for 30 min on ice with 0·125 µg of FITC-labelled anti-mouse CD4 mAb (clone RM4-5) and 0·06 µg of phycoerythyrin (PE)-labelled anti-mouse CD25 mAb (clone PC61). Cells were then washed in cold FACS buffer, fixed and permeabilized for 30 min with the fixation/permeabilization solution provided in the kit, washed again and further stained with 0·5 µg of allophycocyanin-labelled anti-mouse/rat FoxP3 (clone FJK-16s). Samples were washed and analysed subsequently, as described above.

The expression of Vβ17 by transfected JRT3-T3·5 cells was verified by staining with a FITC-conjugated anti-human Vβ17 mAb (clone E17·5F3) purchased from Serotec (Raleigh, NC, USA) followed by cytofluorimetric examination of fixed cells, as described above.

Statistical analysis

For each set of data, representative results obtained from at least three independent experiments are shown unless stated otherwise. Data are expressed as mean ± standard deviation (s.d.) from replicate wells. Statistical comparisons were performed using Student's t-test, and differences with P-values less than 0·05 were considered to be statistically significant. Trend lines were fitted using polynomial regression.

Results

Inhibition of SEB-induced mouse T cell activation by bLF

The lack of optimal mouse models in which to study superantigen immunopathology and superantigen-induced illness has hampered efforts in this important area of investigation. This is due in large part to the fact that mice, unlike humans, are relatively resistant to superantigenic bacterial toxins [2,33,34]. One strategy to overcome this experimental obstacle is to use ‘humanized’ mice expressing HLA class II transgenes, in which robust T and B cell responses to superantigens are detectable [33–35]. To study the effect of bLF on SEB-induced T cell responses, we used DR4 Tg mice that are responsive to SEB [31]. We first confirmed the expression of HLA-DR4 on a substantial fraction (∼45%) of splenocytes obtained from DR4 Tg mice, but not on WT B6 splenocytes (data not shown). Next, we confirmed that SEB used at two different concentrations, 5 and 500 ng/ml, caused robust T cell proliferation (Fig. 1a) and IL-2 production (Fig. 1b) among DR4 Tg splenocytes, but not in WT controls. The SEB_N23A mutant that is unable to bind mouse Vβ8·2 [32] was used as an additional control and, as expected, failed to induce substantial T cell responses in either DR4 Tg or WT mice (Fig. 1).

Fig. 1.

Proliferative and cytokine responses of DR4 Tg splenic T cells to staphylococcal enterotoxin B (SEB). Splenocytes from wild-type (WT) and human leucocyte antigen (HLA)-DR4 transgenic (DR4 Tg) mice were left untreated or stimulated with SEB at either 5 or 500 ng/ml. Cellular proliferation was measured 96 h later by tritiated thymidine incorporation (a), as described in Materials and methods. The interleukin (IL)-2 content of culture supernatants was quantified by enzyme-linked immunosorbent assay (ELISA) 24 h post-stimulation (b). A mutant form of SEB (SEB_N23A) was used in parallel as a negative control. Error bars represent standard deviation from replicate wells (quadruplicate wells for proliferation and triplicate wells for IL-2 production) for each treatment group. Similar data were obtained in two additional experiments.

Next, we examined whether the presence of bLF in these cultures can alter the magnitude of T cell responses to this superantigen. As illustrated in Fig. 2, treatment with bLF inhibited SEB-induced T cell proliferation and IL-2 production significantly in DR4 Tg splenocyte cultures in a dose-dependent fashion. IFN-γ production following SEB stimulation was inhibited similarly by bLF (not shown). There were no detectable cytokine levels in additional control cultures containing splenocytes and bLF only (data not shown). Using trypan blue dye exclusion, we compared the viability of cultured cells at several time-points and found no marked differences between bLF-treated and -untreated groups, thus ruling out the possibility that attenuated T cell responses in the presence of bLF may have stemmed from any direct toxic effects exerted by bLF in our system (data not shown).

Fig. 2.

The effect of bovine lactoferrin (bLF) on staphylococcal enterotoxin B (SEB)-induced mouse T cell proliferation and interleukin (IL)-2 production. DR4 Tg mouse splenocytes were stimulated with SEB at either 500 ng/ml (a,c) or 5 ng/ml (b,d) in the absence or presence of the indicated doses of bLF. Cell proliferation was measured by tritiated thymidine incorporation (a,b), and the IL-2 content of culture supernatants was quantified by enzyme-linked immunosorbent assay (ELISA) (c,d), as described in Materials and methods. Error bars represent standard deviation in triplicate wells for each treatment group; *, ** and *** denote statistically significant differences with P-values less than 0·05, 0·01 and 0·001, respectively, in comparison with cultures receiving SEB only. Data are representative of three independent experiments that gave similar results.

In additional control experiments, we confirmed that SEB-induced IL-2 secretion in our system and its inhibition by bLF were not due to or affected by endotoxin contamination. First, using a sensitive colorimetric assay, the endotoxin content of our SEB and SEB_N23A mutant preparations was determined to be below the detection limit of the kit (0·1 EU/ml). Secondly, the addition of various doses (10 ng/ml, 100 ng/ml and 1 µg/ml) of E. coli 055:B5 LPS (Sigma-Aldrich) to DR4 Tg splenocytes failed to induce detectable IL-2 production by these cells. Thirdly, the presence of LPS in cultures containing bLF failed to intensify the negative immunomodulatory effect of bLF on SEB-induced IL-2 response (data not shown). Therefore, the observed effect of bLF cannot be attributed to endotoxin contamination.

The function of the IL-2/IL-2 receptor system is critical for both T cell proliferation and activation-induced cell death, which are implicated in the progression of superantigen-mediated immunosuppression [36]. The high affinity receptor for IL-2 is composed of three chains: the β chain and the common cytokine receptor γ chain that are constitutively expressed by T cells, and the α chain (CD25) whose surface expression is highly inducible in T cells after activation. This is with the exception of a small subset of T cells known as CD4⁺CD25⁺FoxP3⁺ naturally occurring T_reg (nT_reg) cells that express CD25 constitutively [37]. CD25 is thus considered a T cell activation marker for conventional T cells. We looked at the effect of bLF on both inducible and constitutive expression of CD25 on CD4⁺ T cells in our model. While DR4 Tg conventional (CD4⁺FoxP3^–) T cells did not express CD25 in their resting state, they up-regulated CD25 expression following stimulation with 500 ng/ml of SEB, as expected (Fig. 3). CD25 expression among these cells was inhibited dramatically by bLF, and this inhibition was most pronounced when bLF was used at 10 mg/ml. Similar results were obtained when DR4 Tg splenocytes were stimulated with 5 ng/ml of SEB (data not shown). Within the same experiments, we noticed no marked reduction in the percentage of CD25⁺ cells among nT_reg (CD4⁺FoxP3⁺) cells upon bLF treatment, indicating that the constitutive expression of CD25 remains largely unaffected by bLF (Fig. 3). In addition, CD25 expression on a per cell basis (mean fluorescence intensity) was reduced only moderately on nT_reg cells. Together, this data set demonstrates clearly the ability of bLF to modulate superantigen-mediated T cell activation as judged by cellular proliferation, cytokine secretion and CD25 up-regulation.

Fig. 3.

The effect of bovine lactoferrin (bLF) on CD25 expression by staphylococcal enterotoxin B (SEB)-triggered mouse T cells. DR4 Tg splenocytes were left untreated or stimulated with 500 ng/ml of SEB in the absence or presence of the indicated doses of bLF. After 24 h, the constitutive and induced expression of CD25 by CD4⁺forkhead box P3 (FoxP3)⁺ regulatory T cells and CD4⁺FoxP3^– conventional T cells, respectively, was examined by flow cytometry and after live gating on lymphocytes. Numbers indicate the percentage of CD25⁺ cells among gated subpopulations of T cells. Data are representative of three independent experiments yielding similar results.

bLF-mediated inhibition of T cell responses to SEB cannot be mimicked by other bovine proteins, and is independent of bLF's iron-binding property

In order to show that the observed effect of bLF reflects a selective immunomodulatory property of this molecule and not all bovine proteins, we stimulated DR4 Tg splenocytes with SEB in the presence of bLF, BSA or bovine apo-transferrin used in parallel cultures. We chose BSA and apo-transferrin, the iron-deficient form of transferrin, as control bovine proteins because their molecular weight is similar to that of bLF. In addition, apo-transferrin and bLF belong to the same family of proteins, i.e. the transferrin superfamily, and share structural and functional characteristics [38]. Figure 4a shows that both BSA and apo-transferrin used at similar doses to bLF concentrations were unable to prevent SEB-mediated IL-2 production, indicating that the observed effect of bLF on superantigen-induced responses is not due to a general inhibitory function exhibited by bovine proteins. Furthermore, because apo-transferrin and bLF both bind iron, yet behave differently in our model, the inhibitory effect of bLF cannot be attributed to its iron-binding capacity. This notion is supported by our finding that using desferrioxamine mesilate or 2′,2′-dipyridyl reagent to chelate free iron in complete medium did not block the observed bLF inhibitory property (Fig. 4b).

Fig. 4.

Modulation of staphylococcal enterotoxin B (SEB)-induced cytokine production by bovine lactoferrin (bLF) in comparison with other bovine proteins, or in iron-depleted cultures. DR4 Tg mouse splenocytes were cultured with 500 ng/ml of SEB in the absence or presence of the indicated doses of bovine lactoferrin (bLF), bovine serum albumin (BSA) or apo-transferrin (all shown in mg/ml). The interleukin (IL)-2 content of culture supernatants was determined by enzyme-linked immunosorbent assay (ELISA) 24 h later (a). The IL-2 response of DR4 Tg splenocytes cultured in the absence or presence of the indicated doses of bLF and stimulated with 500 ng/ml of SEB was examined in complete medium or in iron-chelated media (pretreated with 50 µg/ml of desferrioxamine mesilate or 100 µM 2,2′-dipyridyl) (b). Data are representative of two independent experiments yielding similar results.

LfcinB inhibits IL-2 production in response to SEB

LfcinB is a cationic peptide released from the N-terminal region of LF from cow's milk by acid-pepsin hydrolysis, and exerts anti-microbial properties against a wide range of pathogens, including S. aureus[15,39]. Importantly, LfcinB is believed to account, at least partially, for the iron-independent, anti-microbial action of bLF [40]. We therefore examined whether LfcinB can prevent SEB-induced IL-2 production in a similar fashion to bLF. The concentrations of LfcinB employed in these experiments were extrapolated from previous studies measuring the gastric content of LfcinB following bLF ingestion [41], as well as the dose range at which LfcinB exhibits its anti-microbial function [39]. While not markedly affecting cell viability in these experiments (data not shown), LfcinB could inhibit IL-2 production efficiently in response to SEB used at either 500 ng/ml (Fig. 5) or 5 ng/ml (data not shown). This finding demonstrates clearly that LfcinB can mimic the observed inhibitory effect of bLF on superantigen-mediated responses.

Fig. 5.

The effect of lactoferricin B on staphylococcal enterotoxin B (SEB)-induced IL-2 production. DR4 Tg mouse splenocytes were activated with 500 ng/ml of SEB in the absence or presence of several doses of lactoferricin B. Culture supernatants were harvested 24 h later and their interleukin (IL)-2 content was determined by enzyme-linked immunosorbent assay (ELISA). Error bars represent standard deviation in triplicate wells for each treatment group; *** denotes statistically significant differences with P-values less than 0·001 in comparison with cultures receiving SEB only. Data are representative of two independent experiments yielding similar results.

bLF attenuates human T cell responses to SEB

To extend our findings from a mouse model to human cells, we tested the effect of bLF on the responsiveness of Vβ17⁺ human Jurkat T cells to SEB. This cell line was generated and used because SEB is known to activate T cells expressing Vβ17 TCR β chains [42]. The surface expression of Vβ17 by these cells was confirmed by flow cytometry (Fig. 6a). We stimulated Vβ17⁺ Jurkat T cells with SEB in the presence of either LG2 or BLS–DR4 cells as APCs. The presence of bLF in these cultures led to significantly decreased IL-2 concentrations in these cultures, consistent with our findings in our mouse model (Fig. 6b). Next, we tested whether bLF can dampen the cytokine response of primary human T cells to superantigen. Figure 6c represents the inhibitory effect of bLF on SEB-induced IL-2 production by PBMCs obtained from healthy human subjects. These results demonstrate the ability of bLF to function in the capacity of an immunomodulator in the context of human T cell responsiveness to a bacterial superantigen.

Fig. 6.

The effect of bovine lactoferrin (bLF) on staphylococcal enterotoxin B (SEB)-induced interleukin (IL)-2 production by human T cells. A Jurkat human T cell line generated to express human Vβ17 (a) was co-incubated with bare lymphocyte syndrome (BLS)–DR4 antigen-presenting cells at a 5 : 1 ratio, and stimulated with 500 ng/ml SEB in the absence or presence of the indicated doses of bLF (shown in mg/ml). The IL-2 content of culture supernatants was quantified by enzyme-linked immunosorbent assay (ELISA) 16 h later (b). Error bars represent standard deviation in quadruplicate wells for each treatment group. Peripheral blood mononuclear cells from a 24-year-old healthy male donor were challenged similarly with SEB in the absence or presence of bLF followed by IL-2 quantification in culture supernatants by ELISA (c); *, ** and *** denote statistically significant differences with P-values less than 0·05, 0·01 and 0·001, respectively, in comparison with cultures receiving SEB only. Error bars represent standard deviation in quadruplicate wells for each treatment group. Similar results were obtained in independent experiments using peripheral blood mononuclear cells obtained from a total of six donors (three males and three females, age range 21–40 years).

Discussion

The therapeutic potential of naturally occurring bioactive proteins and peptides has attracted the interest of the scientific community in recent years. LF is the main whey protein in human milk, and is also present in cow's milk. While parenteral administration of bLF into mice before challenge with LPS was shown previously to lower serum levels of proinflammatory cytokine tumour necrosis factor (TNF)-α[18], our understanding of how bLF modulates T cell responses associated with toxic shock syndrome is far from complete. In this report, we demonstrate for the first time that bLF diminishes both mouse and human T cell responses to the staphylococcal superantigen SEB, a finding that is potentially of critical importance in the context of staphylococcal infections and toxic shock.

In our in vitro system, bLF prevented conventional mouse T cell activation measured by proliferation, cytokine synthesis and CD25 expression. In the same experiments, bLF did not alter the percentage of CD25⁺ cells among nT_reg (CD4⁺FoxP3⁺) cells, suggesting that the observed inhibitory effect of bLF is somewhat selective for conventional T cells and does not affect global gene expression. Of note, bLF used at one particular dose (1 mg/ml) was able to increase the overall percentage of nT_reg cells among SEB-stimulated DR4 Tg splenocytes dramatically, consistently and reproducibly. Given the profound immunosuppressive function of nT_reg cells and their potential involvement in modulation of superantigen-mediated immunopathology [43], whether bLF can influence the suppressive activity of nT_reg cells warrants further investigation.

It is unlikely that attenuated T cell responses to SEB in our model resulted from any adverse effects on APCs. In fact, bovine and human LF have been shown recently to enhance antigen-presenting and co-stimulatory functions of APCs [44–46]. It is also noteworthy that we ruled out the possibility that the observed negative modulatory effect of bLF on SEB-induced IL-2 production was a consequence of endotoxin contamination of our reagents. This is particularly important in light of a previous report that SEB and LPS act synergistically to induce cytokine secretion by human PBMCs [47]. Of note, such a synergistic effect was evident for secretion of TNF-α, IL-1β and IL-6, which are prototype proinflammatory cytokines produced by APCs (e.g. monocytes), and not for IFN-γ, a typical T cell-derived cytokine. In fact, this is consistent with the results of our control experiments in which LPS failed to augment SEB-induced activation of mouse and human T cells. Moreover, the addition of LPS in cultures containing both SEB and bLF to imitate endotoxin contamination failed to intensify the negative modulatory effect of bLF on SEB-induced responses (data not shown).

The expression patterns of certain streptococcal superantigens are reportedly influenced by LF and transferrin in a partially iron-dependent fashion [48]. It was therefore of interest to find out whether or not the observed inhibition of mouse T cell responses by bLF in our system was due to the iron-binding property of bLF. Our finding that the negative modulatory effect of bLF on SEB-induced IL-2 response could not be mimicked by bovine apo-transferrin suggested that this effect does not stem from the iron-binding capacity of bLF. Further support for this notion is provided by our experiments in which the presence of two iron-chelating reagents in cultures failed to reverse the inhibitory effect of bLF on SEB-mediated cytokine secretion.

Importantly, the inhibitory effect of bLF on responses to SEB could be recapitulated by its bioactive peptide LfcinB. This finding introduces a novel and potentially useful function for this peptide in modulation of superantigen responses. One exciting area of investigation will be to test and optimize the usage of LF-derived peptides in both in vitro and in vivo settings. These peptides can be employed either alone or in combination with LF, other naturally occurring dairy proteins or other therapeutics in an attempt to diminish adverse T cell responses. To this end, LfcinB and bLF were recently shown to act synergistically against E. coli and S. epidermidis[49].

An important aspect of our results is the similarity between mouse splenocytes and human T cells (Jurkat cells and normal PBMCs) in terms of their responsiveness to bLF in the context of the SEB challenge. It needs to be emphasized that bLF has significant homology with both human and mouse LF at the amino acid sequence level. It is also well known that bLF can act across the species barrier, and is capable of binding receptors expressed by cells from various species. Importantly, Jurkat cells also used in our studies reportedly express a bLF-binding protein [50]. Also of interest, human LF receptors were found on human T cells stimulated with the mitogen phytohaemagglutinin, but not on resting T cells [51]. Whether superantigen-activated T cells bind to bLF more efficiently than resting T cells remains to be elucidated.

The focus of the present study was the potential therapeutic benefit of bLF in the context of a superantigen-mediated cytokine storm. Although LF concentrations in cow's milk are much lower than those in human milk, bLF can be prepared in large quantities. Similarly, the production of LfcinB can be scaled-up for therapeutic use if the beneficial properties of this peptide, for instance in the context of staphylococcal toxic shock syndrome, are validated in vivo. There is no doubt that using mouse LF and human LF in their respective host models will provide invaluable additional insight into the innate immunomodulatory properties of these glycoproteins.

Here, we have demonstrated for the first time that bLF can inhibit both mouse and human responses to SEB. Future in vivo studies will have to optimize the dose and administration route of bLF before potentially protective properties of bLF can be tested in animal models of superantigen-mediated illness and toxic shock.

Disclosures

The authors have no financial conflict of interest.