Correspondence Nicolae Miron, Department of Immunology, Iuliu Hatieganu University of Medicine and Pharmacy, Croitorilor Street 19-21, Fifth Floor, 400162 Cluj Napoca, Cluj, Romania. Tel/fax: +40 264535543; mobile: +40 743013883; email: firstname.lastname@example.org
Staphylococcal enterotoxin A (SEA) is one of the bacterial products tested for modulation of unwanted immune responses. Of all the staphylococcal enterotoxins, SEA is the most potent stimulator of T cells. When administered orally, SEA acts as a superantigen (SA), producing unspecific stimulation of intra-epithelial lymphocytes (IELs) in the intestinal mucosa. This stimulation results in amplification of the normal local immunologic responses, which are mainly regulatory. This amplification is based on increased local production of IFN-γ by IELs, which acts on the nearby enterocytes. As a result, the enterocytes produce large amounts of tolerosomes, cellular corpuscles which detach themselves from the basal poles of the enterocytes and contain antigenic peptides that are conditioned to be interpreted as tolerogenic by the gut immune system. Tolerosomes are physiologically produced as a response to dietary peptides; it is already known that enterocytes posses the molecular mechanisms for processing peptides in a similar manner to lymphocytes. The fate of tolerosomes is not precisely known, but it seems that they merge with intestinal dendritic cells, conveying to them the information that orally administered peptides must be interpreted as tolerogens. SEA can stimulate this mechanism, thus favoring the development of tolerance to peptides/proteins administered subsequently via the oral route. This characteristic of SEA might be useful in therapy for regulating immune responses. The present paper reviews the current status of research regarding the impact of SEA on the enteric immune system and its potential use in the treatment of allergic and autoimmune diseases.
concentration needed for half maximum proliferation of human T cells
staphylococcal enterotoxin A
staphylococcal enterotoxin B
staphylococcal enterotoxin C
staphylococcal enterotoxin E
T cell receptor
transforming growth factor beta
tumor necrosis factor alpha
T regulatory cell
toxic shock syndrome toxin 1
Staphylococcal enterotoxin A
Staphylococcal enterotoxin A belongs to the family of staphylococcal enterotoxins, a group of molecules which have drawn the attention of researchers in the field of immunity for over 30 years. The first SE discovered was SEA, in 1966, followed by another eight (B-E, G-J). The original observations were connected with the ability of these enterotoxins to induce toxic shock when food contaminated with Staphylococcus aureus strains was ingested (1). From the beginning, it was observed that SEs are active in very small amounts (micrograms), and are very stable. Generally, foods contaminated by them retain their toxicity after boiling or freezing. Even in the digestive tract, these proteins are not degraded by local proteases and can therefore still exert their specific actions (2). In the case of SEA, at approximately 4 hr after the ingestion of less than 1 μg, symptoms such as nausea, vomiting, and abdominal cramps appear (3). This is accompanied by an inflammatory infiltrate abundant in PMNs in the lamina propria and epithelium of the intestinal wall. PMNs release large quantities of mediators such as histamine, leukotrienes, and intestinal neuropeptides including substance P, all of which contribute to the clinical picture (4). The proof for the inflammatory etiology of the symptom of emesis in toxic shock is that this symptom is reversed by the administration of antihistamines. In some animal models, it has been proved that SEA also induces secretion of monocyte chemo-attractant protein 1 (5), IL-6 and IL-8 by the intestinal myofibroblasts (6). Under the influence of SEA, the serotonin concentration increases in the intestinal wall, stimulating local vagal receptors, an absolutely necessary step in the development of the gastrointestinal symptoms (7).
In addition to their toxic activity, SEs stimulate adaptive immunity as SAs, which means that the number of T cells activated by these toxins is much greater than in the case of normal antigens. The difference is due to the peculiarity of the SAs ligation to the MHC-TCR complex (8). SAs bind the complex from the exterior in an unspecific manner, as compared to conventional specific TCR antigen binding. As a result, SAs produce undifferentiated, exaggerated activation of T lymphocytes, which generates increased production of cytokines. If SAs escape into the blood, the serum concentrations of TNF-α, IL-2 and IFN-γ produced by circulating lymphocytes rapidly reach toxic levels, which can cause death by toxic shock (9). SAs activity is evaluated by measuring P50 (h), the concentration which activates half of the human T cells. SEA has the lowest P50 (h) (0.1 pg/ml) of all SEs (10).
SEs are coded by plasmids, transposomes, prophages, and pathogenicity islands. They have a complex structure, with two important domains: one responsible for digestive toxicity and another for superantigenic activity (11). So far, it is not clear whether these two functions can be separated (12).
The enteric immune system
Apart from its effects in food-borne toxic shock, the impact of SEA on the function of the enteric immune system is connected with the immunological characteristics of the digestive tract. The intestine has an estimated mucous surface of 300 square meters and processes annually 30 kg of proteins. Daily absorption of 130–190 g of peptides occurs; these have not only a nutritive role, but also an antigenic function (13). There are approximately 1000 billion bacteria which stimulate local immunity per gram of feces, and as many lymphocytes per meter of intestine (14). Thus, there is more lymphoid tissue in the whole digestive tract than in the whole of the rest of the human body (15). This lymphoid tissue is distributed between the intestinal epithelium and the lamina propria, the sub-epithelial connective tissue of the mucosa. In the epithelial layer, lymphocytes are located in the spaces between the latero-basal sides of normal enterocytes. It is estimated that there are 20 intraepithelial lymphocytes for every 100 enterocytes (13). In the lamina propria, the lymphoid tissue is organized in the form of solitary lymph nodes or classical Peyer's patches, which are veritable secondary lymphoid organs.
IELs are relatively difficult to classify according to the classical criteria used for T cells. The majority of IELs express αEβ7-integrin (which binds the E-cadherin expressed on enterocytes) and belong to the CD8+ type; however the CD8 molecule is heterodimeric, as is true in the general circulation, in only 50% of cases (16). Some of the homodimeric CD8+ IELs are autoreactive, and these are functionally more similar to γδTCR T cells than to αβTCR T cells (17). Likewise, some of the CD8+ IELs with αα-homodimeric CD8 are MHC-II restricted, and not MHC-I restricted (18). IELs are the result of intestinal migration of lymphocytes, which begins in the neonatal period, sometimes after antigenic stimulation in secondary lymphoid organs. However, once they have entered the intestinal wall after stimulation by local antigens, CD8+ cells have a tendency to produce immunoregulatory cytokines, such as TGF-β, and to refrain from cytotoxic activity (19).
In addition to CD8+ IELs, the gut also hosts γδTCR T cells, NKT cells, and classical CD4+ T cells with αβTCR. The exact immune function of all these cells is unknown. The general tendency of these lymphocytes is to generate a tolerogenic immune response to antigens encountered in the gut lumen (20, 21). Other cellular types also participate in mounting an immune response. The most important for promoting oral tolerance are dendritic cells in the lamina propria, which infiltrate the area between the latero-basal sides of the enterocytes and reach into the intestinal lumen with their projections, taking up antigens which are afterwards processed and presented into the mesenteric lymph nodes (22). Another important cell is the so-called M cell, placed as a hood over the luminal region of the PP. These M cells are in contact with the gut content at their upper pole, allowing them to capture antigens and pass them over to the immune milieu of the PP, where they are processed by other dendritic cells and then presented to lymphocytes in the local lymph nodes (23). It has been proved that a large proportion of intestinal dendritic cells express an enzyme called retinal dehydrogenase, (responsible for vitamin A metabolism), which produces a shift toward a tolerogenic phenotype in the case of the T helper cells that interact with these dendritic cells (24, 25). All these particularities of the enteric immune system result in generation, at the intestinal level, of Th regulatory cells, also known as iTreg, Tr1, Th3 and Th2 (26). Although intestinal T regulatory cells are classical CD4+CD25+FoxP3+ regulatory cells, they appear in the intestine, and not in the thymus (27). Tr1 (CD4+ IL-10+ FoxP3-) are regulatory cells which exert their function especially through the synthesis of IL-10, while Th3 (CD4+ TGF-β+ FoxP3+) rely on the release of TGF-β for the down regulation of immune responses. These regulatory subpopulations present numerous interconnections in vivo, probably leading to the existence of intermediate cellular types (28). All these characteristics make the gut a predominantly tolerogenic immune environment.
Oral tolerance to dietary antigens
The oral administration of any peptide can have three consequences: the secretion of anti-peptide IgA; a systemic immune response with the appearance of serum antibodies and cell-mediated immunity; or a state of anergy, local and/or general tolerance, which prevents an unwanted immune response when re-encountering an innocuous antigen. The first two situations are encountered in the case of pathogens with invasive potential, while the third possibility applies to commensal bacteria and dietary antigens, which do not cause local injuries or systemic immune responses (29).
Oral tolerance is classically defined as the specific suppression of cellular and humoral immune responses to an antigen which has entered the human body via the digestive route. Oral tolerance likely evolved as an analog of self tolerance, in order to prevent hypersensitivity reactions to foods and commensal bacteria. Oral tolerance is a continuously developing immunological process, stimulated by exogenous antigens which enter the gut. Due to their preferential access to the internal medium, antigens entering via the gut represent a special category of antigens, at the border between self and non-self. Dietary tolerance thus becomes a form of peripheral tolerance, a process by which food antigens and commensal microorganisms are considered a future part of the self (30).
There are two main pathways for inducing oral tolerance: stimulation of the development of Tregs to an antigen which has been eaten, and clonal anergy of effector cells which might react to a particular antigen (31). The most important factor determining what kind of tolerance will develop is the antigen dose (32). Small doses of oral antigen favor the development of Tregs, while larger doses lead to deletion of active clones. Small doses lead to antigen presentation through dendritic cells belonging to the gut-associated lymphoid tissue, with consequent increased synthesis of regulatory cytokines, such as IL-10, TGF-β and IL-4 (33). Afterwards, these dendritic cells migrate to local lymph nodes, where they suppress immune responses by inhibiting effector cells through regulatory cytokines. These cytokines act not only on effector cells which recognize the antigen presented by the tolerogenic dendritic cells, but also on effector cells from the immediate proximity, inside the lymph node (bystander suppression) (34).
The impact of staphylococcal enterotoxin A on the enteric immune system and its influence on oral tolerance
As previously shown by Lonnqvist et al., treatment of neonatal mice with orally administered SEA promotes the development of oral tolerance to OVA when it is fed to adult mice (Fig. 1) (35). SEA, one of the strongest known T-cell mitogens, does not reverse, but rather augments, the tolerogenic type of intestinal immune responses. SEA binds to the TCR of IELs and to the MHC-II of the dendritic cells which cross the epithelium to take up samples from the intestinal lumen. The result is excessive stimulation of IELs, with increased local IFN-γ production, probably through a MyD88-dependent mechanism (36). IFN-γ stimulates normal enterocytes to process peptides rapidly for presentation through MHC-II (37). Although enterocytes are not professional antigen presenting cells, it has been found that they participate in the development of oral tolerance by production of MHC-II-associated peptides (38). Such production occurs, not only when stimulated by SEA or other inflammatory stimuli, but also physiologically, in which case it is at a lower rate (39). MHC-II-associated peptides can be presented directly to CD4+ lymphocytes (40) or packed in the form of corpuscles, or small cellular fragments, which detach from the basal poles of enterocytes. These corpuscles belong to the exosome group, and are called tolerosomes because it has been proved that they also participate in creation of a tolerogenic intestinal environment, synergic with the Tregs described above (41). The exact composition of tolerosomes is not known, but it is thought that they may contain other co-stimulatory molecules, which may induce tolerance to the MHC-associated peptide (42).
The discovery of tolerosomes is relatively recent, having occurred less than 10 years ago. It has been known since 1983 that, in order for oral tolerance to develop, an intact portal circulation is needed, and that oral tolerance is transferrable through serum. These cell fragments, the so-called tolerosomes, first discovered by electron microscopy in 2001, were found in the insoluble fraction produced by ultracentrifugation from the serum of animals which had been subjected to induction of oral tolerance. The soluble fraction, serum without tolerosomes, was no longer able to mediate the transfer of oral tolerance (41). This proved that intercellular communication occurs through exosomes during development of oral tolerance. The fate of tolerosomes after their production has not yet been fully elucidated. It is supposed that they bind to local or distant antigen presenting cells (43, 44), conveying the necessary information for mounting tolerance to food antigens. In any case, the fact that the portal circulation is involved in this process has lead to the speculation that tolerosomes can be directed to the liver, another recognized tolerogenic site (45, 46).
Oral tolerance in therapeutics and the rationale for staphylococcal enterotoxin A utilization
Oral tolerance has been exploited for therapeutic purposes to inhibit all forms of unwanted immune responses, from the secretion of different antibody classes, to type IV hypersensitivity reactions. It is to be noted that Th1-type responses are much easier to inhibit than Th2 responses. In order to suppress a Th2 immune response, it is necessary to administer greater antigen quantities, or to increase the frequency of administration (47). An exception to this rule is that of IgE-mediated Th2 immune responses associated with increased production of IL-4, such as allergies, which respond very well to oral tolerization schemes (48).
The idea of using SEA in order to augment oral tolerance to different peptides arose from epidemiologic studies (49). Staphylococcus aureus is now a common commensal in the gut in the occidental population (50, 51). It has been demonstrated that Western infants with a greater degree of colonization with SEA-producing S. aureus strains are protected against food allergy (52, 53). Toxigenic S. aureus residing in the gut induce greater concentrations of IgA in children's serum and protect from eczema (54). Animal models of allergic diseases suggest that neonatal oral administration of SEA followed by feeding the sensitizing protein OVA in adulthood prevents the development of airway allergy when the mice are re-exposed to intranasal OVA (35). Oral administration of SEA in the neonatal period with the aim of augmenting Tregs generation is favored because of the propensity of lymphocytes to home into the gut versus the skin in this period of ontogenesis (55). Other animal studies have indicated that parenteral inoculation of SEA promotes the generation and function of regulatory lymphocytes (56, 57). SEA is less well absorbed from the gut lumen through facilitated transcytosis than are other staphylococcal SAs such as SEB and TSST-1 (58), and is probably less prone to produce systemic effects when orally administered.). SEA is less well absorbed from the gut lumen through facilitated transcytosis than are other staphylococcal SAs such as SEB and TSST-1 (58), and is probably less prone to produce systemic effects when orally administered.[T1] Also, SEA seems to be more efficient at induction of regulatory-type immune responses than TSST-1 (59). For these reasons, SEA might be a better choice for therapeutic studies of oral tolerance.
The impact of staphylococcal enterotoxins on achievement of tolerance to antigenic proteins in experimental autoimmune encephalomyelitis
Three main molecules are affected by autoimmunity in multiple sclerosis, the disease mimicked by EAE: myelin basic protein, proteolipid protein, and myelin oligodendrocyte glycoprotein. There have been attempts at inducing oral tolerance to these proteins in animal models of EAE (60–64) and also in humans (65–67).
The history of the use of staphylococcal enterotoxins in EAE has some aspects in common with oral administration of antigenic myelin proteins. Experiments on animals were first conducted with SEB, and only later with SEA, although SEA is more potent in regard to its effects on T cells. So far, there are no studies of SEA or SEB administration in humans with MS. Also, there are no studies in humans or animals of associations between SEA and any of the myelin antigenic proteins, MBP, PLP or MOG.
In general, previous studies using SEA or SEB in animals were focused on parenteral (intravenous or intraperitoneal) administration. The reason for this is connected to the discovery that in mice which develop EAE, especially the PL/J species, which were massively used in the 1990s, there is TCR restriction of the myelin-reactive cells (68). A significant proportion of these lymphocytes have a TCR that contains the Vβ8 chain (69). SEB is a molecule with tropism for this chain (70). With high doses, lymphocyte stimulation by SAs leads to their deletion (71). The first experiments with SEB on mice actually tried to produce deletion of autoreactive lymphocytes. When given before immunization with MBP, SEB has a protective effect to the development of EAE, because those T cells which might have become autoreactive are eliminated. When SEB is given after immunization, EAE aggravates, because there is supplementary stimulation of the effector cells by the SA (72). Unlike MBP, PLP is not recognized by Vβ8+ T cells (73), accordingly PLP-induced EAE is differently influenced by administration of SEB. EAE induced by PLP is worsened if SEB is given prior to immunization because the remaining lymphocytes are stimulated by the cytokine production which results from the activation/destruction of Vβ8+ T cells (74).
SEA possesses a different tropism for the Vβ chain of the TCR, preferentially binding to the Vβ1, 3, 10, 11 and 12 types (75). Intraperitoneal administration of SEA can reactivate MBP-induced EAE after one month of clinical remission (76). Soos et al. have shown that SEA produces new episodes of EAE when given in mice which have previously been immunized with MBP after depletion of Vβ8 cells by SEB pretreatment. As previously mentioned, the explanation relies on the types of lymphocytes that remain in place to be stimulated by SEA. This experiment revealed that it is not only Vβ8 cells that can participate in EAE pathogenesis, as was previously believed (77). To our knowledge, there has been no study of oral administration of SEA in EAE.
In any case, the variable behavior seen after administration of SEB/SEA can be explained by the affinity for certain T cells, different TCR restrictions for effector lymphocytes in different species, and differing routes of administration. When administered parenterally, SEA acts as a major stimulant of the systemic lymphocyte compartment. Thus, staphylococcal enterotoxins have the opportunity to reactivate EAE, even in animals which have entered a remission period (78).
Staphylococcal enterotoxins in type 1 diabetes mellitus
Insulin is now recognized as the major auto-antigen in type 1 diabetes (79). As a consequence, a number of clinical trials have tested the possibility of producing oral tolerance to insulin, in the hope of preventing or delaying the onset of the disease in non-diabetic relatives at high risk of diabetes. The Diabetes Prevention Trial–Type 1 showed that 7.5 mg of oral insulin daily did not confer a benefit when compared to placebo. In a subgroup of this trial which included only those relatives who had tested positive on two occasions for anti-insulin autoantibodies, orally administered insulin proved to be useful in preventing the onset of diabetes, compared with placebo (80). Currently, the Pre-POINT (Primary Oral/intranasal Insulin Trial) is addressing the group of children who are at high risk of developing type 1 diabetes and who have not yet developed anti-insular autoantibodies. This trial is ongoing (81).
There has been no trial in humans or animals that has tested the efficacy of SEA as an adjuvant for augmenting oral tolerance to insulin or any other peptides that function as autoantigens in type 1 diabetes.
In animal models the results of SEA usage appear to be in conflict. Kawamura et al. have shown that staphylococcal enterotoxins (SEA, SEC1, SEC2, or SEC3), when injected iv into non-obese diabetic female mice at 4 and 10 weeks of age, significantly reduce the incidence of diabetes at 32 weeks compared with a saline treated group (82). The explanation, according to the authors, originates in the fact that SAs are able to stimulate a CD4+ fraction of T lymphocytes which is capable of immunoregulatory activity.
Ellerman et al. succeeded in transferring the disease from BB/Wor diabetes-prone rats to healthy counterparts using staphylococcal enterotoxin-activated spleen cells (83). In this experiment the donor animals were first depleted of RT6.1 T-cells, which are the Tregs of this rat strain. Thus, in the absence of the regulatory arm, SAs activated only the effector arm of the immune system in these animals. The diabetogenic T cells were strongly activated by SEA, SEC3, and SEE, whereas SEB and SEC2 were less effective in the adoptive transfer of diabetes. The results of this experiment, considered together with those of Kawamura's, strongly suggest that SAs have a nonspecific action on both effector and regulatory lymphocytes. Preservation of the regulatory arm of the immune system might be of special importance in the case of BB rats because their effector autoimmune lymphocytes present specific resistance to apoptosis when challenged with normal or high doses of SAs (84).
Staphylococcal enterotoxins in allergic diseases
It is clear that, when present in their skin lesions, SEA can aggravate the condition of atopic dermatitis patients (85, 86). SEA also seems to have implications in the pathogenesis of atopic keratoconjunctivitis (87), psoriasis, erythroderma (88), and chronic urticaria (89). In all these diseases, SEA acts topically, at the surface of the external epithelia.
The effects of attempting to produce tolerance by sequential oral administration of SEA and an allergenic protein are currently under investigation in animal models of allergic diseases. The formula of neonatal treatment with oral SEA followed by oral administration of OVA in adulthood has proven useful in preventing the development of induced allergic asthma in mice (35). As we have said before, tolerization is better achieved in the neonatal period, due to the fact that most neonatal lymphocytes home to the gut, where they are educated towards a regulatory phenotype, the gut being a medium which predisposes to this type of immune response. The combination of α4β7 integrin and MAdCAM-1, which is expressed only on high-endothelial venules in gut-associated lymphoid tissues and post capillary venules in the gut (90), ensures a major flow of lymphocytes towards the gut wall in early infancy, a phenomenon that is lost in adult life. It seems that, at the beginning of ontogenesis, regulatory responses are easier to elicit (91).
Results from similar studies are different in adult life. Oral co-administration of SEB with a food allergen – ovalbumin or whole peanut extract – to mice aged 4 to 8 weeks resulted in highly Th2 polarized immune responses to the antigen (92). Subsequent oral challenge with antigen led to anaphylaxis, and local and systemic mast cell degranulation. SEB-induced sensitization triggered eosinophilia in the blood and intestinal tissues. SEB impaired tolerance specifically by limiting the expression of TGF-β and regulatory T cells, and tolerance was regained with high-dose antigen.
When SEA acts on the respiratory mucosa of adult animals, an experimental setup that does not involve oral tolerance, the results are detrimental to the course of allergic response. Exposure to SEA 4 hr prior to OVA sensitization triggers an increased accumulation of eosinophils in bronchoalveolar lavage fluid, bone marrow, and lung tissue at 24 hr after OVA re-challenge (93).
Our intention was to present the current status of knowledge regarding the use of SEA as a tool for increasing immune tolerance to proteins that function as allergens or autoantigens in different diseases. Current studies are still trying to determine the exact route of administration that could provide a benefit in human or animal therapy. In our opinion, the oral route and the sequence of SEA followed by the incriminated peptide or protein can provide a solution to augmenting the immune regulatory responses. Still, some difficulties remain to be solved. So far, only administration of SEA in the neonatal period has proven to be successful. For humans, it would be of great interest to also improve oral tolerance in adult life. It is reasonable to foresee difficulties in establishing the appropriate dose of this potentially toxic molecule in human therapy, both in adults and, even more so, in neonates. On the other hand, research regarding SEA could open a window to other approaches to boosting physiological ways of gaining tolerance to molecules that enter the digestive tract.
This work was funded by the Romanian National Council of Scientific Research in Higher Education – CNCSIS (PD_477).