Atopy, autoimmunity, and the TH1/TH2 balance

The functional classification of murine CD4⁺ T helper cell clones into distinct subtypes, designated T_H1 and T_H2, based primarily on the pattern of their cytokine production, has greatly enhanced our understanding of T-cell biology.¹ The existence of T_H1 and T_H2 subsets has since been reported in humans as well.²

See related article, p. 470.

The concept of a type 1 or type 2 response has now been broadened to include other cells such as cytotoxic (CD8⁺) T cells. These two subsets of the T cells are characterized by differences in the cytokines produced, transcription factors involved, the nature of the immune responses mediated, and their disease associations.³ T_H1 cells are characterized by the production of interleukin-2, interferon-, and tumor necrosis factor-. In the absence of IL-4, IL-12, produced by activated macrophages and dendritic cells, is the principal T_H1-inducing cytokine; its actions are mediated by the transcription factor Stat4. T_H1 cells are critical in the defense against intracellular pathogens, eliciting phagocyte-mediated responses. T_H1-dominant responses are associated with inflammation and injury and are exemplified by the delayed hypersensitivity reaction. Thus T_H1 cells are believed to play a crucial role in cell-mediated immunity. Many autoimmune diseases including rheumatoid arthritis, juvenile rheumatoid arthritis, insulin-dependent diabetes mellitus, and multiple sclerosis are associated with a T_H1 phenotype. The signature cytokines of the T_H2 cells are IL-4 and IL-5, but T_H2 cells also produce IL-6 and IL-13. IL-10, although thought to be predominantly a T_H2 cytokine, can be secreted by T_H1 cells as well.³ IL-4 is the principal cytokine that induces naive T-cell precursors to differentiate into T_H2 cells, mediated by the transcription factor Stat6. IL-4 is involved in B-cell switching to produce immunoglobulin E, and IL-5 is important for eosinophil activation. T_H2 cells play an important role in humoral immunity. T_H2 cells and their cytokines are associated with atopic and allergic disorders. In addition, several cytokines produced by T_H2 cells such as IL-4, IL-10, and IL-13 have anti-inflammatory actions.

It is now believed that the same T-cell precursor (designated T_H0), a mature, naive CD4⁺ T lymphocyte, is capable of developing either a T_H1 or T_H2 phenotype.³ Once a T-cell immune response begins to develop along one pathway, it tends to be progressively polarized in that direction. Ouyang et al⁴ showed that the transcription factor GATA-3 directly repressed T_H1 development, in an IL-4–independent mechanism, which may involve repression of IL-12. This suggests that the expression of GATA-3 and IL-12 signaling are mutually antagonistic, facilitating dominance of one pathway over the other during early development of the T helper cell. Besides cytokines, other factors that determine the balance between T_H1 and T_H2 subsets in immune responses include the type of antigen-presenting cell, the antigenic dose, and the strength of stimulation and co-stimulation.

The T_H1 and T_H2 pathways each control a unique set of immune responses, promoting the development of cells of the same subset while suppressing the expansion and effector cells of the other subset. Factors that alter this balance between these two subsets may drive the immune response toward one pathway or the other. Although this description is simplistic, the T_H1/T_H2 paradigm is a useful one. As stated earlier, many autoimmune diseases are associated with a T_H1 phenotype, and allergic/atopic diseases are associated with a T_H2 phenotype. Given the polarization of the developing T-cell response into one of these two types, it is reasonable to speculate that autoimmune and allergic diseases represent two ends of the same spectrum of immune responses and that these two types of disorders would be mutually exclusive. In this issue of The Journal, the EURODIAB Substudy 2 Study Group⁵ reports results that aimed to determine whether occurrence of atopic diseases early in life is associated with a reduced risk of development of type 1 IDDM. In a multicenter, population-based, case-control study, the authors collected data on atopic diseases (asthma, eczema, and allergic conjunctivitis) from children with type 1 IDDM and age-matched control subjects. They found a decreased prevalence of atopic diseases, in particular asthma, in children with IDDM. The risk reductions associated with the atopic diseases were marked in children in the 10- to 14-year age group (odds ratio 0.59; 95% CI 0.42-0.83). A negative association between atopy and IDDM has been previously described. In 1971 Hermansson et al⁶ reported a lower frequency of atopy in children with IDDM, and their siblings as well, compared with control subjects. Douek et al⁷ found that a significantly lower proportion of children with IDDM had symptoms of asthma, compared with siblings or control subjects. The frequency and the severity of the symptoms were also significantly lower among the children with IDDM. However, the differences seen were for self-reported symptoms of asthma, but not an actual diagnosis of asthma, which was similar in all 3 groups.

In addition to these epidemiologic studies, there is evidence from the laboratory that supports the T_H1/T_H2 paradigm. Rapoport et al⁸ investigated the secretory patterns of interferon-, IL-2, IL-4, and IL-10 from stimulated peripheral blood mononuclear cells from patients with IDDM and control subjects. Patients with IDDM had an early decreased secretion of the T_H2 cytokines IL-4 and IL-10 and a late increased secretion of the T_H1 cytokines IL-2 and interferon-, providing evidence for impairment of the T_H1 and T_H2 cytokine secretory pattern in patients with IDDM.

Patients with RA, another T_H1-mediated disease, have also been investigated for the prevalence of T_H2-mediated diseases. Verhoef et al⁹ demonstrated that the prevalence of hay fever in patients with RA was significantly lower than that in patients without RA (4% vs 8%). Those patients with RA who had hay fever had less severe disease compared with control patients with RA (without hay fever) as determined by the erythrocyte sedimentation rate, C-reactive protein level, joint score, and radiographic joint damage score. Allanore et al¹⁰ reported that the incidence and point prevalence of atopy was lower among patients with RA than control subjects. These studies support the notion that T_H1 and T_H2 diseases are mutually exclusive and that the occurrence of a T_H2-mediated disease, such as atopy, might be protective against the development of a T_H1-mediated autoimmune disease.

However, not all studies support this view. Stromberg et al,¹¹ in a study of 81 children with type 1 IDDM and 72 control subjects, did not find a significant difference in prevalence of atopic disease as defined by history, clinical features, skin prick test results, serum IgE, or circulating IgE antibodies to allergens.¹¹ Another study showed that children with juvenile chronic arthritis, another T_H1-mediated disease, had no abnormalities of their responses to inhalational allergens or IgE production compared with control subjects.¹² O’Driscoll et al¹³ compared 40 patients with RA (9 of whom were receiving steroids) and 40 age- and sex-matched control subjects and failed to detect a significant difference between the two groups with respect to prevalence of positive skin prick test results, atopic diseases, blood eosinophil counts, or total serum IgE.¹³ Although these negative studies are hospital-based and are limited by small sample sizes, they emphasize the complexity of the interactions between various components of the immune system. Measurement of cytokines from peripheral blood cells or the serum may not be representative of the true nature of a localized inflammatory response, with its particular cytokine and cellular environment. In addition, cytokines are pleiotropic, being produced by more than one type of cell and having many targets with different effects. Thus attempts to dichotomize complex diseases in terms of just T_H1 or T_H2 may be an oversimplification. In fact, T_H2 responses mediating IDDM have been reported, and this disease could be mediated by both T_H1 and T_H2 cells.¹⁴ JRA provides another example of the complexity of the T_H1/T_H2 balance. There is evidence that JRA is a T_H1-mediated disease.¹⁵ Murray et al¹⁶ demonstrated that synovial fluid from patients with pauciarticular disease significantly overexpressed IL-4 messenger RNA compared with synovial fluid from patients with polyarticular JRA and RA. Thus it is of interest that the T_H2 cytokine IL-4 is characteristic of pauciarticular JRA, which has a better overall articular prognosis than polyarticular JRA.

Natural infection, mediated by IL-12, is believed to promote the differentiation of T cells into a T_H1 phenotype. Immunizations, on the other hand, promote the development of a T_H2 phenotype.¹⁷ The decreasing incidence of natural infections in the Westernized world (hygiene hypothesis),¹⁸ in combination with the increasing success of the immunization programs, would be expected to tilt the balance toward T_H2, depriving the immune system of signals that promote T_H1 development. This raises the question of whether this tilt toward T_H2 will alter the prevalence of T_H1-mediated diseases.¹⁹ Interestingly, evidence for such a trend has been presented by Spector et al²⁰ who reported that there was a secular decline in the prevalence of RA and rheumatoid factor in women.

In conclusion, establishing that T_H1 and T_H2 diseases are mutually antagonistic would lend further support for attempts to promote one type of T-cell response selectively, in order to alter or control the other and to preferentially manipulate the cytokine network. This is part of the rationale behind trials of oral tolerance, in which antigen is administered orally to preferentially inhibit T_H1 responses in the periphery.²¹ The EURODIAB Substudy 2 Study Group⁵ has provided further evidence in support of the T_H1/T_H2 paradigm.

I thank Gurjit K. Khurana Hershey, MD, PhD, and David N. Glass, MD, for critically reviewing the manuscript.

References

TOP

1. Mosmann TR, Cherwinski H, Bond MW, Giedlin MA, Coffman RL. Two types of murine helper T cell clone. I. Definition according to profiles of lymphokine activities and secreted proteins. J Immunol 1986;136:2348-57.