Superantigens and Microbial Pathogenesis

The results of a randomized, controlled trial of endoscopic ligation of esophageal varices compared with sclerotherapy by Laine and colleagues, in this issue of Annals, is the second report of a relatively new endoscopic technique. The two techniques appear to be approximately comparable, except that the complication rate is less with ligation. Whether ligation will replace sclerotherapy remains to be seen.

Recent analysis of various bacterial and viral products has revealed that microbial proteins may be not only responsible for microbial infection and disease but also connected to the onset of autoimmunity. These microbial products are called superantigens because of their strong effect on the proliferation of certain T-lymphocyte populations. Superantigens elicit various biological activities, including lymphocyte mitogenesis, pyrogenicity, depressed antibody production, and shock [1, 2]. We describe the effects of microbial superantigens on immune function and their possible association with various acute and chronic inflammatory diseases.

T cells mediate protection against infectious agents by producing lymphokines after activation by specific antigens. Lymphokines affect the function of other cell types, including phagocytic cells, involved in antimicrobial immunity. Most mature T cells express receptors (T-cell receptors) composed of and chains, which are necessary for the specific recognition of antigens [3]. Both and chains consist of constant and variable (V) regions. Many V and V families of genes can encode the V regions of T-cell receptors; however, the antigen receptor of a single T cell is encoded by only a single V and V family member.

T-cell receptors only recognize antigens when the antigens are bound in the groove of major histocompatibility complex class II molecules. These class II molecules are expressed on the surface of antigen-presenting cells such as macrophages. The class II-antigen peptide complex on antigen-presenting cells is recognized by that portion of the T-cell receptor formed by interaction between the V and V chains (in much the same way that antigen binds to the variable regions of light and heavy chains of antibody). This binding process results in the antigen-specific activation of only the small number of T cells bearing the specific receptor for that particular class II-antigen peptide complex.

In contrast to antigens, superantigens bind only to the V chain and induce activation of the large number of cells expressing specific V chains. The T-cell response to superantigens is therefore much larger than antigen-specific responses. Like antigens, superantigens require antigen-presenting cells expressing class II molecules to induce T-cell proliferation; however, superantigens do not bind to class II molecules through the groove structure [1, 2, 4, 5]. Like superantigens, mitogens such as concanavalin A and phytohemagglutinin stimulate large numbers of T cells, but, unlike superantigens, mitogens do so without regard for the particular V chain expressed by the T cells and do not have to be bound to class II molecules to induce T-cell activation.

There is increasing evidence of a fundamental difference between the effect of in vitro and in vivo exposure of T cells to superantigens. In vitro, superantigens cause a dramatic proliferation of T cells bearing specific V molecules. Various superantigens have been reported to stimulate the production of several cytokines, including interleukin-2, 4,and 6,tumor necrosis factor- and , and interferon- [6-9]. On the other hand, exposure to superantigen in vivo can abrogate the T-cell response to subsequent stimulation with the same superantigen in vitro [10]. Activation by superantigen in vivo, therefore, can result in either the establishment of unresponsiveness or the deletion of T cells.

Staphylococcus aureus produces a family of protein enterotoxins that possess superantigen properties [11]. These enterotoxins belong to a larger family of related gram-positive bacterial pyrogenic toxins including toxic shock syndrome toxin 1 (TSST-1). A Mycoplasma arthritidis-secreted protein [12] and streptococcal M protein [13] also possess superantigen properties.

Several toxic syndromes are associated with staphylococcal enterotoxins and other bacterial superantigens. Although staphylococcal enterotoxins are generally associated with food-borne intoxication, they have also been implicated in the toxic shock syndrome [14]. The superantigens TSST-1 and staphylococcal enterotoxins induce a toxic shock syndrome-like disease in a rabbit model of menstrual toxic shock syndrome [15]. Although TSST-1 has been strongly associated with menstrual toxic shock syndrome, other staphylococcal toxins, primarily staphylococcal enterotoxin B, play a pathogenetic role in cases not associated with menstruation. Recent studies have shown that patients with the toxic shock syndrome have a predominance of peripheral T cells expressing V 2 chains, a subset of the V family [16]. These results suggest that superantigen-induced activation of T cells expressing specific V chains may result in massive lymphokine production that might, in turn, be responsible for the clinical syndrome observed. From these studies, it is clear that not all in vivo exposure to superantigen leads to T-cell unresponsiveness as described above. The circumstances under which unresponsiveness or disease is induced by superantigens are being investigated.

Other bacterial superantigens have also been implicated in the development of various disease states. The identification of M protein of group A streptococci as a superantigen and its interaction with human T cells may be of central importance in the pathogenesis of poststreptococcal autoimmune diseases, acute rheumatic fever, and acute glomerulonephritis [13].

Mycoplasma arthritidis, predominantly a rodent pathogen, can induce a chronic inflammatory arthritis in rats and has been cultured from the bone marrow of patients with systemic lupus erythematosus [17, 18]. It has been postulated that the M. arthritidis superantigen-induced T-cell activation may contribute to the polyclonal B-cell activation seen in patients with systemic lupus erythematosus. In addition, the finding that V () 14-positive T cells are significantly elevated in the synovial fluid of affected joints in patients with rheumatoid arthritis when compared with peripheral blood from such patients suggests that the development of disease may initially involve the superantigen-induced activation of V 14-bearing T cells [19].

Superantigens have also been found to be produced by retroviruses [20, 21]. The in vivo expression of one of these retroviruses, mouse mammary tumor virus, results in the selective depletion of V 14-bearing T cells [22]. A selective depletion of T cells that express specific V chains also occurs in patients infected with human immunodeficiency virus (HIV), suggesting that an HIV superantigen exists [23]. The role of this superantigen in the pathogenesis of HIV-related disease is being investigated.

As described above, activation by superantigen in vivo can, under certain circumstances, either establish unresponsiveness in T cells expressing specific V chains or cause their deletion. Although the mechanism of this unresponsiveness remains poorly understood, it may be possible to selectively disable T cells bearing certain V molecules that are involved in autoimmune disorders. Thus, superantigens may have a potential use in prophylaxis or treatment of certain autoimmune disorders. This possibility is shown by the fact that the in vivo administration of a superantigen, Staphylococcus enterotoxin E, but not other Staphylococcus enterotoxins, ablated the induction of experimental allergic encephalomyelitis, a disease resembling multiple sclerosis [24]. Thus, superantigen-mediated induction of T-cell nonresponsiveness may have relevance not only for the role of these molecules in microbial pathogenesis but also for therapeutic considerations in autoimmune diseases involving specific V-bearing T cells in their pathogenesis. However, the systemic administration of these products would not be without risk; the intravenous administration of these toxins in mice resulted in several toxic side effects. Molecular alterations of superantigens into less toxic forms that retain their activity on T cells in vivo may result in effective agents for treatment of autoimmune disorders.

Bacterial superantigens may have a role in the induction and development of autoimmunity as well as acute and chronic inflammatory disease states. The polyclonal expansion of superantigen-activated T cells may result in the production of large quantities of lymphokines and the activation of autoreactive lymphocyte populations. Superantigens have been proposed to be relevant to the induction of rheumatoid arthritis and multiple sclerosis because of the expansion of particular autoreactive V-bearing T cells in these patients. However, the role of these expanded T-cell populations in disease developments needs further evaluation. In addition, the nature of these bacterial products and their inhibitory effects on specific T cells expressing particular V chains suggest that these proteins may have potential as therapeutic agents in the treatment of T-cell-mediated autoimmune and inflammatory disease induction and progression.

• Copyright 2004 by the American College of Physicians