Main

Growth and development, which are critical in children, are also relevant for immunopathology in relation to allergy. Important developments in various compartments of the immune system occur in young children that are different from the adult situation (Table 1). These differences could contribute to the widely held view that during the first years of life there is an enhanced risk of allergic sensitization leading to development of allergic diseases in childhood and later on in life(1). It is generally accepted that the prevalence of both asthma and hay fever is increasing in children worldwide(2). This increase could be the result of better diagnostic capabilities and better recognition of childhood asthma, air pollution and environmental risk factors, and socioeconomic factors.

Table 1 Differences of the immune system of newborns and infants as compared with adults in relation to allergy
Full size table

Here we discuss the interaction between the development of the immune system and genetic and environmental factors considered to be involved in the pathogenesis of allergic diseases. Risk factors for the development of allergic diseases, cytokine patterns in allergic children, and cell surface markers thought to be involved in allergy are discussed. Finally, we present possible mechanisms of allergy development and new approaches for therapy.

PREDICTIVE FACTORS FOR ALLERGY DEVELOPMENT

Genetic Factors

In newborns, a positive family history has been shown to be an important risk factor for the development of atopic disease(3, 4). If a child has one allergic parent, the risk to become allergic is approximately 20%, which is increased to 60% with two allergic parents, whereas without allergic parents the risk is around 10%(5). Family history could be very useful as an early marker for allergic disease, with an acceptable sensitivity (61%) and specificity (83%)(6). Allergic diseases are presently considered as at least a two-locus inheritance, with expression being influenced by environmental variables, resulting in heterogeneous expression. Moreover, an individual's likelihood of developing allergic disease is not constant over time, even if the general propensity is genetically determined.

Recently, evidence of linkage of high levels of total serum IgE to loci on chromosome 5q31.1, including the cytokine gene cluster, was reported(7, 8). It was also demonstrated that elevated serum IgE is coinherited with a trait for bronchial hyperresponsiveness and that a gene governing bronchial hyperresponsiveness is located near a major locus that regulates serum IgE levels on chromosome 5q(9). Further support for genetic influences comes from several studies reporting significantly higher cord blood IgE concentrations in male than in female subjects(10–13), although conflicting information exists(14, 15). The gender influence on cord blood IgE can hypothetically be due to differential expression of the gene encoding for atopy in male subjects(13). An increased number of atopic boys compared with girls is in line with this observation(16). Cookson et al.(17) found a strong linkage between a marker on chromosome 11, localized at 11q13, with atopic IgE responses in large pedigrees. This finding was confirmed in 60 nuclear families(18). Abnormalities in the chromosome 11q region account for 60% of allergy in British families. It was also noticed that this linkage was detected only with maternally derived alleles(19). A genetic mechanism, known as genomic imprinting, in which a paternal“atopy gene” is be suppressed, may explain this(20). A maternal inheritance will mean that unaffected mothers may still carry the trait if they have inherited an abnormal gene from their father. One specific site on chromosome 11q accounts for 20% of the 11q-associated allergy. This site represents a single amino acid change in theβ-chain of the high affinity IgE receptor (FcεRI)(21). Furthermore, genetic linkage was found between a gene (or genes) in the T cell receptor α region, encoded on chromosome 14, and specific IgE responses(22).

Neonatal IgE Levels

The principal feature so far that distinguishes atopic from nonatopic individuals is their predisposition to develop antigen-specific IgE antibodies upon repeated exposure to low doses of foreign proteins (i.e. pollen, animal products, house dust mite, and food)(23). This results in higher serum IgE levels in atopic than in nonatopic subjects. However, twin studies strongly suggest that, although the tendency to IgE production is genetically determined, the antigenic specificity is governed most likely by environmental influences(24). These results exemplify our current understanding of allergy development, in which we consider the interaction between genetic makeup (genotype) and environmental factors as determining the development of disease. Thus it is important to be able to discern children at risk for allergy development as early as possible and to determine those environmental factors, which subsequently should be avoided.

Another prediction factor is measurement of umbilical cord blood IgE levels(16, 25–27). Elevated cord blood IgE concentrations may be genetically determined (genetic high IgE responders)(28) and/or be due to intrauterine sensitization(25, 27). The latter alternative seems to be only rarely the case. The IgE antibodies present in cord blood seem to be more the consequence of nonspecific spontaneous IgE production, perhaps lack of suppression, than of antigenic stimulation. However, as discussed later (see“Th Cell Subsets”) some antigen-specific T cell responses were recorded in cord blood.

Furthermore, the mother's, more than the father's, atopic history had a significant correlation with cord blood IgE levels(13, 29). Presumably, cord blood IgE levels are, in addition to possible placental transfer of maternal IgE, mainly determined by the fetal immunologic reaction to intrauterine allergens and other factors(such as smoke), based upon genetic determinants.

Several considerations support the use of cord blood IgE levels, because a fetus can make IgE from the 11th week of gestation(30), IgE cannot cross the placental barrier(31), and a well established correlation exists between a high IgE responder and the clinical tendency to develop atopic disease for adults as well as for children(32).

Although the specificity of cord blood total IgE levels in predicting allergy is relatively high (94%), the sensitivity is quite low (26%)(3). The predictive value is not very high, hence relevance of elevated cord blood total IgE levels as the sole marker for predicting atopic disease has been questioned(33–37). Furthermore, a rapid decrease of IgE levels during the first weeks occurred in half of the infants born into allergic families(33). Hattevig et al.(38) reported a transient IgE response to food proteins, which occurred relatively frequently, even in infants who did not develop atopic diseases. Thus, young children may show temporarily positive results, developing IgE antibodies against food allergens. Persistently high IgE concentrations, however, were more strongly related to the development of atopic disease(38). This transient IgE production is an integral part of the normal immune response, because at the time of initial contact(s) with an allergen not previously encountered, the immune system initially “recognizes” the allergen and responds to it. In immunologically normal (nonatopic) children and animals, these initial responses are self-limiting and spontaneously switch off, after a few weeks or months, despite continuing allergen exposure. The initial response is replaced by a state of lifelong allergen-specific immunologic tolerance(unresponsiveness). This is supported by animal studies showing that in healthy (nonatopic) animals the natural response of the respiratory mucosal immune system to inhaled allergens involves initial recognition, accompained by transient low level IgE production due to selective suppression, resulting in a state of tolerance(39–41). Tolerance induction functions poorly in the preweaning period(42), presumably due to delayed postnatal maturation of one or more key elements of the mucosal immune function that are rate-limiting in inducing tolerance(1). Therefore allergen exposure in the very early phase of infancy primes for subsequent T cell reactivity(43, 44). This results in a persisting IgE response, leading to a state of clinical hypersensitivity. High total IgE levels at 2 and 6 mo after birth showed no better predictive values than cord blood IgE(45, 46). However, cord blood IgE levels in conjunction with a positive family history (especially in the mother) can be useful for identifying infants at risk for early development of atopic disease.

Allergenic Reactivity

Measurement of allergen reactivity can add to the prediction of allergy(47). A positive skin reaction to egg in young children resulted in a specificity of 54% and a sensitivity of 54% in the prediction of atopy at a later age. Recently, Sigurs et al.(48) confirmed this finding by showing that atopic disease appeared before the age of 4 y in 80% of the 40 children who had IgE antibodies to egg white at 9 mo of age. A similar sensitivity to that with skin testing was obtained by Kobayashi et al.(49) by measuring in vitro proliferative responses of cord blood lymphocytes to food allergens (egg and cow's milk). Combining the proliferative response values with cord blood total IgE concentrations results in a rise in sensitivity to 79%.

Hattevig et al.(50) showed that specific IgE antibodies to inhalant allergens appeared in increasing frequency starting at the age of 2 y, later than to food allergens. This agrees with the difference in clinical and immunologic expression of the atopic status in infancy and childhood. In infancy food allergen sensitization and atopic dermatitis predominate. Beyond that age range, inhaled allergen sensitization and respiratory tract symptoms become important(10).

New specific immune factors need to be considered. Possibly, differential cytokine expression may be relevant in this respect (see“Cytokines”).

THE ROLE OF ENVIRONMENTAL FACTORS IN ALLERGIC SENSITIZATION

The identified risk factors are presented in Table 2. Smoking during pregnancy has been shown to be associated with increased cord blood IgE levels(51) and the subsequent development of asthma(52). Other studies have failed to demonstrate this(12, 14). Paternal smoking did neither influence cord blood IgE levels nor infant allergy(51). After birth exposure to concentrations of pollutants indoors when people are smoking may be far in excess of outdoor concentrations of pollutants that are found regularly(53, 54).

Table 2 Potential risk factors for the development of allergy
Full size table

Outdoor pollution, by combustion by-products (CO2, NOx, SO2, CO), solid particles, and ozone, is potentially able to enhance allergic sensitization. The actual contribution of pollutants acting as adjuvants in allergy development is, however, controversial. The prevalence of allergic diseases in the more polluted former East Germany was found to be less than that in West Germany, against expectations(55).

The month of birth was studied by de Groot et al.(56) in a group of 45 000 patients in relation to the presence of IgE antibodies to inhalant allergens. Children born between December and February had a slight but significantly higher chance to become sensitized to grass pollen. On the other hand, children born in autumn(August-November) had a higher chance to become sensitized to indoor allergens, such as house dust mite and animal danders(56). This was in agreement with earlier studies of Björksten and Suonemi(57) and Businco et al.(58). Such information suggests that, in the first 3 mo after birth, an increased sensitivity for allergic sensitization is present.

Low birth weight has been found a highly significant risk factor for allergic disorders probably due to immaturity of the immune system that renders infants more liable to allergic sensitization(52). The effect was more pronounced than that of an atopic family history. These children were also at risk to develop house dust mite allergy by the age of 2 y(52) (see below,“Immunopathology of Allergy Development in Childhood”). Recently, an association has been suggested between impaired or disproportionate fetal growth and raised serum IgE levels as an adult(59). Other studies have not seen any indication for prematurity as a risk factor for atopic sensitization(5). In addition, Martinez et al.(60) established a relationship between impaired neonatal lung function and development of wheezing respiratory illness in infants; however, the allergic state was not taken into account.

Viral Infections

Viral infections in childhood are thought to have a modulating role on the immune system, which can be stimulatory, as well as inhibitory. Viral infections may cause epithelial injury, inflammation, enhanced mediator release, and sensitization of IgE(61). Variations over time in an individual's susceptibility to sensitization might be due to respiratory tract infection, thereby occurring more easily in early life. RSV and parainfluenza virus are the predominant viruses isolated from infants and children less than 6 y of age, whereas Mycoplasma pneumoniae and rhinovirus predominate in older children(62). The occurrence of recurrent wheezing and asthma after RSV bronchiolitis in infancy is frequently reported(63, 64). Thus, respiratory infections are an important cause of wheezing, and RSV is the predominantly associated organism. In subjects with existing allergic disease, RSV infections promote the development of inflammatory response in the lungs,i.e. the late allergic response.

In infants with confirmed RSV infections during the first 6 mo of life, the frequency of persistent wheezing up to 7-8 y was directly related to the level of RSV-specific IgE in nasopharyngeal secretions during the initial period(65). Moreover, allergy in childhood often occurs shortly after an RSV infection(66). Only 28% of infants with undetectable titers of RSV-IgE had subsequent episodes of wheezing. In contrast, 70% of infants with high RSV-IgE titers experienced wheezing(67). Recently, a link has been shown between viral infections and the Th1 and Th2 profile. RSV contains an attachment protein called glycoprotein G, which could skew the immune response toward a Th2 phenotype(68). In this way, RSV infections could induce or enhance the development of allergy in a hypersensitive period of life. However, PBMC from 22 infants previously infected with RSV usually had RSV-specific increases in Th1 cytokine-specific mRNA (IFN-γ and IL-2). This was not observed in PBMC from RSV antibody-negative children, indicating that naturally acquired RSV normally induces a Th1 memory response(69). This is in accordance with the hypothesis formulated by Martinez(70) that recurrent early life infections render a preferential selection of Th1 clones with subsequent inhibition of allergic sensitization. Furthermore, long-time persistence of RSV genome and protein in an animal model has recently been shown(71), which may result in latent and persistent viral infections. Another possible mechanism is the effect of virus infections upon adhesion molecules which regulate eosinophilic and neutrophilic migration. ICAM-1 permits eosinophilic and neutrophilic migration, and it has been demonstrated that it functions as the major human rhinovirus receptor(72). However, the precise relationship between respiratory virus infections and the development of allergic diseases may well depend upon host factors, the period in life, and virus characteristics, which should be further studied(73, 74).

IMMUNOPATHOLOGY OF ALLERGY DEVELOPMENT IN CHILDHOOD

T h Cell Subsets

Based on their cytokine production profile, CD4+ Th cells can be divided into a Th1 (e.g. IL-2 and IFN-γ), a Th2 subset (e.g. IL-4, IL-5, and IL-10), and a naive Th0 subset (producing all of these cytokines)(75, 76). The intricate cytokine network provides the most appropriate and convincing basis for intercellular communications and control. Several factors, most of them cytokines themselves, may specifically promote Th1 or Th2 outgrowth. IFN-γ and IL-12 strongly promote the generation of Th1-like cells, whereas IL-4, IL-10, and PGE2 promote the generation of Th2-like cells(77–79) (Fig. 1). Because IL-12 and PGE2 are produced by APC, eventually the balance between different APC populations may determine the selective outgrowth of Th1 and Th2 cells. On the other hand, T cell-derived cytokines, such as IL-4 and IFN-γ, may be important for the maintenance of Th1 and Th2 responses. Naive T cells require for their activation inductive signals from dendritic cells, whereas sensitized memory T cells can respond to processed antigen presented by virtually any cell type expressing class II MHC molecules on the surface. Although monocytes can act as APC for memory T cells, resident alveolar macrophages strongly suppress the APC function of dendritic cells. Hereby, alveolar macrophages might induce a state of tolerance reflected in decreased (local) memory T cell expression and suppression of IgE formation(80). The immunologic outcome of initial T cell activation events after inhaled allergens can be detected as T cell anergy or tolerance(80). In most individuals, repeated exposure to an allergen results in the induction of such tolerance. This is less likely to occur in atopic persons. The eventual outcome of individual encounters with inhaled antigens depends to an important extent on the functional capacity of the APC, which initially removes the antigen. The balance between the potential immunoregulatory APC populations is a key factor in hypersensitivity and the pathogenesis of immunoinflammatory diseases(80).

Figure 1
figure 1

Schematic representation of T-B cell interaction. B cells acting as APC present peptides of the allergen to the specific T cell receptor, expressed on CD4+ Th2 cells. Subsequently B7 expression is up-regulated on the APC and interacts with CD28 expressed on the T cell. Signals through the T cell receptor and the CD28 costimulatory molecule are obligatory for the activation of the T cell. Upon activation the T cell transiently expresses CD40L and secretes IL-4. Binding of IL-4 to the IL-4 receptor on B cells provides signal 1, whereas interaction between CD40L and CD40 expressed on the B cell provides signal 2. These two signals result in B cell activation and subsequent isotype switching to the synthesis of IgE. This process of isotype switching can be modulated by several cytokines interacting with IL-4-directed processes.

Full size image

An antigen also can determine to some extent the preferential outgrowth of Th1 versus Th2 type cells. Very low doses of peptide, presented by dendritic cells, lead to selective priming of Th2-like cells that produce IL-4 and not IFN-γ on restimulation in vitro. Priming with much higher doses of antigen causes selective outgrowth of IFN-γ-producing Th1 cells. Low doses of allergens can, therefore, induce Th2 responses that lead to IgE production, whereas higher doses, used for desensitization, could convert this response to inhibitory Th1-like cells(81).

Induction of IgE Synthesis

Cytokines play a crucial role in the regulation of IgE synthesis (Fig. 1). IL-4 is a key cytokine in the production of IgE by naive B cells(82). This is based on its ability to induce expression of ε-germline transcripts, indicating that IgE synthesis reflects Ig isotype switching(83). Another T cell-derived cytokine, IL-13, can also induce germline-ε expression in B cells, but it is 2-5-fold less potent than IL-4 in inducing IgE synthesis(84). For this activation by IL-4 a physical contact between activated CD4+ T cells and B cells is required, called cognate interaction(82) (Fig. 2). Antigen-activated T-B cell interactions require binding of the T cell receptor to MHC class II-peptide complexes on antigen-presenting B cells, which results in T cell activation and cytokine production. Once Th cells are activated and express the CD40L, they interact with B cells that constitutively express CD40 and induce proliferation and differentiation in an antigen-independent fashion(85). The costimulatory signal provided by CD4+ T cells can be replaced by anti-CD40 MAb(86), infection of B cells with EBV(87), or the addition of hydrocortisone to T cell-depleted PBMC cultures(88). CD40-CD40L interaction causes the induction of B7 on B cells. The membrane antigen B7.1, expressed on activated B cells, monocytes, and dendritic cells, interacts with CD28 on T cells. This interaction leads to higher IL-4 production, resulting in higher IgE synthesis(89, 90). CTLA-4 can also act as a ligand for B7.2 and is expressed on activated T lymphocytes(91, 92).

Figure 2
figure 2

Th2-dependent IgE and IgG4 formation. Allergen exposure in atopic individuals leads to the activation of naive CD4+ T cells. Depending on different types of APC and the local microenvironment, these T cells differentiate into functional Th1 or Th2 effector cells. This microenvironment dictates the selective outgrowth of Th1 or Th2 cells, primarily by the presence of IL-12 (Th1) and PGE2 (Th2) produced by monocytes. IL-12 can either directly induce Th1 development or indirectly via up-regulating the IFN-γ production by NK cells. Activated Th2 cells induce in activated B cells, through the process of cognate T-B cells interaction (see Fig. 1), proliferation, up-regulation of CD23, and isotype switching to IgE and IgG4. IFN-γ produced by Th1 cells inhibits these processes, whereas other cytokines (such as IL-5, IL-6, and IL-13) further enhance the IL-4 activity. The ratio of Th1 vs Th2 cells therefore determines the final outcome of the B cell response. Cross-regulation between Th1 and Th2 cells by IFN-γ and IL-10 further modifies this ratio.

Full size image

Another important interaction in IgE regulation takes place between CD23(low affinity receptor for IgE (FcεRII) expressed on many hemopoietic cells and CD21 (receptor for EBV and complement receptor-2) on B cells, T cells, and follicular dendritic cells(93, 94). Triggering of CD21 specifically increased IL-4-induced IgE production from PBMC(93). Thus, compared with other interactions, the CD21-CD23 pairing controls IgE synthesis in an isotype-specific manner.

The activating effect of IL-4 is promoted by IL-5 and IL-6, but inhibited by IFN-α and -γ, TGF-β, and PGE2(95–98). Moreover, PGE2, IFN-α, and IFN-γ also inhibit IL-4-induced expression of CD23 on B cells, indicating that there is an association between CD23 expression and IL-4-induced IgE production(99, 100). Furthermore, IL-10 has a down-modulating effect on the production of IFN-γ(101).

Cell Surface Markers

CD4+/CD8+T cells. White blood cell and lymphocyte counts are at their highest values at birth and decline with age. Proportions of lymphocyte subsets vary, but the percentage of CD4+ Th cells and CD8+ cytotoxic T cells progressively increases over time(102).

The number of CD4 and CD8 T lymphocytes in the peripheral blood of adult asthmatic patients did not differ from those in the normal control subjects(103). Gemou-Engesaeth et al.(104) studied a group of atopic asthmatic children (7-16 y), compared with age-matched atopic nonasthmatic children, in which also no differences were found in the percentages of CD4+ and CD8+ T lymphocytes.

CD45RA+/CD45RO+T cells. Neonates are immunologically naive, because exposure to microorganisms and foreign antigens has been prevented to some extent by the placental barrier. Indeed, most (91%) neonatal CD4+ T cells express CD45RA, a well accepted marker of naive T cells. This CD45RA+ population of neonatal T cells is phenotypically identical to the CD45RA+ naive T cells in adult peripheral blood(105). These cells are gradually replaced by CD45RO+ (marker of activated or memory T cells) cells, presumably as a result of repeated antigen exposure(102). A plateau phase of CD45RO+ cells is reached during puberty(106). The switch from CD45RA+ to the CD45RO+ phenotype appears to reflect an age-related accumulation of memory cells in the circulation. These phenotypic changes are associated with functional changes, such as an increased proliferative response to IL-2 and an increased production of IFN-γ(106). In vitro CD45RO+ T lymphocytes show enhanced help for IgE synthesis(107).

CD4+CD45RA+ cells produced high levels of IL-2 mRNA upon polyclonal stimulation, but they expressed trace quantities of mRNA for IL-4 and IFN-γ. In contrast, CD4+CD45RO+ cells produced high levels of mRNA for IL-4 and IFN-γ, and less for IL-2(108). CD45 is now considered to be an important regulatory protein that controls activation of T and B cells after ligation of the antigen receptor(109).

In adult asthmatic patients the percentage of CD4+CD45RO+ T cells was significantly elevated and that of CD4+CD45RA+ T cells significantly reduced compared with healthy controls(103). In a study of asthmatic children (1-17 y), who were largely above 10 y of age, it was shown that the percentage of CD4+CD45RA+ T cells in peripheral blood was not significantly increased compared with the percentage in atopic nonasthmatic children. Three different patient groups (asthmatic patients, atopic nonasthmatic patients, and patients with atopic dermatitis) did, however, show an increased expression of CD45RA as compared with nonatopic children(110). This suggests that at least in children, elevated percentages of peripheral CD4+ T lymphocytes expressing CD45RA were associated with the atopic state rather than specifically with asthma. This may, however, be caused by accumulation of CD45RO+ T cells in the target organs.

CD25+ T cells. The IL-2 receptor(CD25) is generally expressed on activated T cells, B cells, NK cells, monocytes, and macrophages. The IL-2 receptor is useful as a marker of activation, but not for discrimination between virgin and memory cells, as the expression is lost when cells return to the resting state(111).

In adult patients with acute severe asthma, it was demonstrated that the percentage of CD4+ T cells expressing the IL-2 receptor was elevated(112). Interestingly, percentages of peripheral blood CD4+ T lymphocytes expressing the activation markers CD25 and HLA-DR, and of CD8+ T lymphocytes expressing CD25, were elevated in asthmatic children compared with atopic nonasthmatic children(104). This suggests that T lymphocyte activation may be physiologically associated with asthma and not simply with the atopic state per se.

CD40/CD40L. Recently, two comparable studies were published, both suggesting an association between ineffective expression of CD40L on cord blood T cells and the low Ig production in the newborn(113, 114). Cord blood MNC have no(113) or decreased(114) expression of CD40L compared with adult MNC. This defect can be largely attributed to the lack of antigenic exposure in the neonate, because in vivo primed cord blood lymphocytes could express substantial amounts of CD40L upon appropriate stimulation (priming and restimulation with PMA and ionomycin)(113). This ineffective CD40L expression by cord blood CD4+ T cells together with their failure to produce a number of costimulatory cytokines, might be regarded as the primary cause for the poor Ig production in the neonate(114).

The CD40 molecule may also play a role in allergic disease. In adult patients with atopic dermatitis, the mean fluorescence intensity of CD40 expression on B cells was increased when compared with expression on PBMC from nonatopic donors(115).

CD23+ B cells. CD23, the low affinity receptor for the Fc portion of IgE, is also a presumed activation antigen. The number of B cells that express CD23 nearly doubles between infancy and adulthood(102). Allergic young children(0-2 y) showed a significantly higher percentage of FcεRII-positive cells than nonallergic young children. FcεRII-positive lymphocytes may play a regulatory role in IgE synthesis in allergic young children(116).

Cytokines

Neonatal cytokine production differs from the adult situation and may play a role in the increased sensitivity in the neonatal phase to allergic sensitization. Because cloning procedures potentially distort the pattern of in vivo cytokine production, cytokine mRNA expression and production should be studied in freshly isolated and short-term cultured T cells obtained from peripheral blood(117). In this way the intrinsic capacity of T cells to produce cytokines can now be studied precisely, because possible influences, such as cytokine production by other cell populations, are prevented(118).

IFN- γ. In neonatal T cells, IFN-γ production is strongly decreased to as low as 10% of the adult IFN-γ production, whereas production of IL-2 and expression of IL-2 receptor are comparable to those in adults(119). Several causes of this low IFN-γ production have been suggested: an increased sensitivity of infant T cells to the inhibitory effect of PGE2(120), an intrinsic deficiency in the capacity of infant T cells to synthesize IFN-γ(119), or an inefficient accessory cell activity on the part of infant macrophages(121). This reduced capacity to produce IFN-γ coincides with the time when the baby's gastrointestinal and respiratory tract is first exposed to environmental antigens, and this may be important in the development of specific allergic disease(122).

After polyclonal stimulation, IFN-γ production by cord blood MNC was found to be reduced in newborns with a positive atopic family history, compared with those without familial risk. This was independent of cord blood IgE levels. IL-4 production was mostly below the detection limits of the assay and could not be analyzed(123). Production of other Th2-derived cytokines exerting enhancing effects on IgE synthesis(IL-5, IL-6) did not differ significantly, neither between adults and newborns, nor between newborns in different familial risk groups(123).

In a study using stimulated CD4+ T cell clones from infants genetically “at risk” for atopy, both the IFN-γ- and the IL-4-producing capacity of clones from babies born into allergic families were significantly reduced relative to clones from normal infants(124). This suggests that, in children who develop allergy, initial immune responses to environmental allergens in early childhood occur against a background of maturational CD4+ T cell dysfunction, resulting in abnormal immunologic responsiveness.

IL-4. IL-4 production is also strongly decreased in neonatal T cells(125). In adults all of the IL-4 and most of the IFN-γ are produced by CD45RO+ T cells, whereas IL-2 is produced by both CD45RO- and CD45RA-positive cells. Therefore, the acquisition of memory T cells may be a major determinant of maturation permitting both IL-4 and IFN-γ production(126). However, given the appropriate stimulation, in vitro neonatal T cells were able to express the lymphokine gene repertoire characteristic for adult T cells (IL-3, IL-4, IL-6, IFN-γ, and granulocyte/macrophage colony-stimulating factor were measured). This suggests that T cells generated from neonatal blood by a primary stimulation in vitro are functionally indistinguishable from the T cells in adult blood(127). If previous in vivo activation is a major determinant for the capacity of T cells to produce IL-4 and IFN-γ, then the capacity of T cells which presumably have had a minimal exposure to exogenous antigens, will exhibit a profile of lymphokine production similar to that of adult virgin T cells. Therefore, the cytokine production profile of neonatal T cells may reflect their antigenically naive status(125). These differences comprise a diminished capacity to provide help for, or actual suppression of, Ig production by B cells, a diminished generation and activity of cytotoxic T cells, and a decreased capacity to activate macrophages(125).

Tang and Kemp(128) studied the production of IL-4 in PHA-stimulated PBMC cultures in healthy neonates, children, and adults. In vitro IL-4 production was found to be significantly reduced in neonates and children under 10 y of age as compared with adults, and to increase progressively with age. These age-dependent variations in IL-4 production may reflect differences in naive and memory T cell populations. Interestingly, the addition of plasma from neonates to PBMC cultures from adults at a final concentration of 10%, resulted in approximately 50% inhibition of IL-4 production. This points into the direction of inhibitory factors present in the plasma of neonates, the nature of which is as yet unknown.

A large body of evidence suggests that allergen-specific Th2-like cells are expanded in atopic subjects, particularly at the level of target tissues, such as skin and lungs(129). Already at birth, allergen-specific Th2 cells can be expanded, suggesting that aeroallergen sensitization can occur during fetal life (intrauterine sensitization)(130). Apart from IL-4, overexpression of other genes of the IL-4 gene family, such as IL-3, IL-5, and granulocyte/macrophage colony-stimulating factor, also appears to be present in atopic subjects. Several genetic alterations could be present at the level of signal transduction pathways or complexes formed by transcription factors and their corresponding promotor regulatory elements. Alternatively, deficient production or regulatory activity of the cytokines responsible for the inhibition of Th2 cell development (e.g. IFN-γ) might induce such Th2 cytokine overexpression(131).

Recently, Borres et al.(132) demonstrated that IL-4 levels in serum are associated with development of allergic disease in infancy. Elevated levels were recorded before onset of clinical symptoms, suggesting that atopic disease is associated with a primary disorder in T cell function.

Allergen-induced cytokine profiles. To establish the relationship between cytokine production, exposure to environmental allergens, and development of atopic disease, it is interesting to study in vitro allergen-induced cytokine profiles. Warner et al.(133) showed that the development of atopic eczema, with positive cow's milk skin prick tests at 1 y of age, was associated with increased proliferative responses and defective IFN-γ production to stimulation with β-lactoglobulin at birth. The deficient IFN-γ response was specific for certain allergens, including a generalized maturational delay in T cell function.

However, data on IFN-γ production show conflicting results. The group of Tang et al.(134–137) found in children with atopic dermatitis (mean age 3.5 y) a higher percentage of IFN-γ-producing cells by intracellular staining in unstimulated PBMC cultures compared with those of healthy control children. This indicates in vivo activation of T cells in the atopic group. IFN-γ secretion after stimulation, however, was significantly reduced in children with atopic dermatitis, compared with controls, whereas the percentage of IFN-γ-producing cells was not different(134). Thus the reduced ability of atopic children to secrete IFN-γ in vitro does not relate to a lack of IFN-γ-producing cells, but to a difference in the regulation of IFN-γ production beyond the stage of gene transcription and translation. This was confirmed by analyzing IFN-γ mRNA expression in children with atopic dermatitis in whom a constitutive expression of IFN-γ mRNA was found. In controls this could be detected only after stimulation(135). In the same group of children spontaneous IL-4 mRNA expression was detected in four out of eight severe atopic dermatitis patients, whereas this was absent in healthy children(136). This suggests in vivo activation of T cells in children with atopic dermatitis. Additional evidence is provided by reports of increased serum levels of soluble IL-2 receptors(138). As T cells from patients with atopic dermatitis have been shown to have an intrinsic defect of IFN-γ secretion, the imbalance of IL-4 and IFN-γ secretion documented in atopic dermatitis could reflect general activation of T cells in the presence of an intrinsically defective IFN-γ secretion(136).

Adhesion Molecules

Leukocyte-endothelial adhesion molecules are thought to be involved in the initial stage of selective recruitment and migration of inflammatory cells from the circulation to the sites of inflammation. Several cytokines, including IL-4 and IFN-γ, promote induction and up-regulation of adhesion molecules, both on the endothelium and on leukocyte surfaces(139). At the sites of allergic inflammation, increased expression of ICAM-1 and E-selectin (endothelial cell adhesion molecule-1) on vascular endothelium have been demonstrated(140, 141). The selective accumulation of eosinophils and lymphocytes in allergic responses has been hypothesized to be generated by a VLA-4/VCAM-1 pathway induced by the release of IL-4 and IL-5 from Th2 lymphocytes(142). This is supported by the marked expression of VLA-4 on basophils, eosinophils and lymphocytes, but not on neutrophils and minimally on monocytes(143, 144). Furthermore, it was identified that IL-4 selectively up-regulates VCAM-1 expression on vascular endothelium(144, 145). However, it has been demonstrated that, although the initial attachment of eosinophils and lymphocytes to endothelial cells is mediated by LFA-1/ICAM-1 and VLA-4/VCAM-1, the subsequent transendothelial migration process relies heavily on ICAM-1 and LFA-1(146, 147). IFN-γ is the strongest cytokine enhancer of epithelial ICAM-1 expression, with IL-1β, TNF-α, and IL-4 also being effective(148).

In allergic asthmatic patients, adhesion molecule expression on endothelium was correlated with eosinophil and total leukocyte infiltrate(149). This suggests that adhesion molecule expression in the vessels of the bronchi may be involved in leukocyte and particularly in eosinophil infiltration.

Adhesion molecules can also be present in a circulating form(150). In adult allergic asthmatic patients, increased serum levels of soluble ICAM-1 and E-selectin during asthma attacks have been demonstrated(151). This may reflect the upregulation of ICAM-1 and E-selectin expression in allergic inflammation. The soluble form of these adhesion molecules may be a useful marker for the presence of allergic inflammation(151). However, in a study in asthmatic children who underwent allergen provocation, no differences were observed in soluble ICAM-1 expression(152). These soluble adhesion molecules might slow down the allergic response by binding to ligands on leukocyte surfaces and, thereby, blocking adherence to tissue ligands and limiting the inflammatory process(139). Because little is known about the role of adhesion molecules in the pathophysiology of allergy in children, more research in this field is necessary.

Other Cell Types

Neonatal B lymphocytes secrete minimal amounts of Ig in response to stimulation with anti-CD3, in contrast to adult cells. This deficiency could be corrected by the addition of IL-2, IL-4, or IL-6, resulting in secretion of all isotypes(153, 154). There were no differences in the distribution of Ig isotypes secreted in response to the cytokines alone or in combination. Deficient production of these cytokines is likely to contribute to the decreased capacity of neonatal lymphocytes to generate an Ig response. Maturation of B lymphocytes into antibody-producing plasma cells occurs gradually during the first weeks of life(155).

An abnormal monocyte/macrophage function may contribute at least partly to deficient neonatal immune responses. Specifically, the newborn infant's monocytes/macrophages, when studied in vitro, are permissive for intracellular multiplication of herpes simplex virus and have impaired phagocytosis and chemotaxis, compared with monocytes/macrophages of adults(156). The decreased ability of neonatal cells to produce IFN-γ appears to be primarily related to the immature function of the neonatal macrophage(157). Cord blood macrophages cultured with adult T cells show only a minimal production of IFN-γ. Adult macrophages cultured with cord blood T cells, however, give rise to high IFN-γ production(158). This points to an important limiting role of macrophages in young infants.

NK cells in cord blood are less active than those in adults, for example lysis of bound targets and the production of NK cytotoxic factor and IFN-ε are both reduced(159–161). This may reflect their immaturity or the absence of cytokines necessary for full NK cell activation. A deficient NK cell function could play an additional role in the development of allergy, because they produce IFN-γ among other cytokines.

Thus developmental aspects in non-T cells, namely macrophages and NK cells, may contribute to the generation of atopy by comprimised IFN-γ production, although specific abnormalities related to allergy are not yet documented.

THERAPEUTIC APPROACHES

The progress in our understanding of allergic mechanisms in children gives rise to new therapeutic options. Interference can take place on different levels: first, primary prevention of sensitization (allergen avoidance); second, at the level of factors regulating the disease process (several forms of immunotherapy); and third, suppression of inflammatory mechanisms(corticosteroids and others).

Allergen Avoidance

A variety of studies have investigated the possible beneficial role of allergen avoidance. The concept of intrauterine sensitization led to studies evaluating the effects of maternal diets. It has decisively been demonstrated that maternal allergen avoidance diets (cow milk's or egg) during pregnancy fail to affect the frequency of atopic disease, serum IgE levels, and food sensitization, from birth up to 5 y of age in offspring at high risk for atopy(162, 163). Elimination diet during lactation was beneficial for high risk infants and reduced the development of atopic eczema(164–166). These studies, however, need to be proven, because no double-blind placebo-controlled food challenges were performed. However, breast-feeding, with or without diet restriction, has given a favorable outcome in many studies(167, 168), but without effect in others(169, 170). The recently published results of a 17-y follow-up study(171) show that the prevalence of manifest atopy throughout follow-up was highest in the group who had little or no breast-feeding. The original groups, however, were not selected by randomization. In addition, the studied population was not selected for a positive family history of atopy. The combination of exclusive breast-feeding with maternal lactation and infant elimination diets has given the most beneficial effect(164, 172, 173). The effect was limited to the 1st or 2nd y of life and the reduction of eczema, whereas the incidence of respiratory symptoms and allergic manifestations appearing later was not affected. These results are not surprising due to the antigen specificity of the allergen avoidance protocol that will not protect against sensitization to inhaled allergens or food allergens not avoided. In an European Society of Pediatric Allergy and Clinical Immunology position paper, extensively hydrolyzed formula is recommended for avoidance of cow's milk in cow's milk-allergic infants to prevent cow's milk allergy and associated atopic dermatitis in high risk infants, but not for prevention of respiratory symptoms. Soy is not beneficial in this respect, considering its immunogenicity and allergenicity(174). Late introduction of solid food besides breast-feeding seems to have long lasting benefits on the development of atopic diseases in high risk children(175–177). This needs to be confirmed in a randomized study. The addition of mite reduction to the food avoidance results in a reduction of respiratory allergy(178). The Isle of Wight study showed that the dual approach of avoidance of food antigens and a reduction of exposure to house dust mite can reduce the manifestations of atopy in the first 2 y of life (13% instead of 40%)(179). Sanda et al.(180) demonstrated a similar beneficial effect of house dust mite avoidance in children with atopic dermatitis.

Immunotherapy

Peptides that represent epitopes on allergens can down-regulate T cells as suggested in cell culture and animal studies(181). In this context it is important not to rely on a single synthetic peptide, but to focus on the identified immunodominant regions of the allergens(182). It has been reported that peptide analogues based on known immunogenic T cell epitopes, which maintain the necessary MHC binding sites but have altered T cell receptor contact residues, may alter the effector response of T cells both in vitro and in vivo(183). In adult allergic patients clinical trials with such altered peptide ligands inducing a state of long-term nonresponsiveness are currently under way.

Recent adult trials, as reviewed by Bousquet an Michel(184), suggest a beneficial effect of traditional s.c. immunotherapy in allergic rhinitis, conjunctivitis, and more recently in allergic asthma. The study of Creticos et al.(185) did not show a practical benefit of immunotherapy for allergic asthma. A limited number of double blind placebo controlled trials of specific oral immunotherapy in children have been reported. Studies using birch pollen (rhinoconjunctivitis)(186) have shown clearly demonstrable effects, whereas studies with grass pollen (pollen asthma and/or rhinitis)(187) and Dermatophagoides (perennial asthma and/or rhinitis)(188) have failed to demonstrate efficacy. A major challenge for successful immunotherapy is to convert an existing Th2 response into a Th1 response that is beneficial, or at least not pathologic. Allergen immunotherapy of adult allergic patients caused a remarkable reduction in the quantity of IL-4 production by their allergen-specific CD4+ T cells(189). Furthermore, repeated parenteral challenge with tolerizing doses of grass pollen allergen in atopic adults selectively stimulates allergen-specific Th1-like T cell reactivity(190). This may be even more effective in children who do not have an established T cell memory(191). Therefore, vaccination with a mixture of relevant inhalant allergens at the appropriate time in childhood, especially in combination with an Th1-selective adjuvant, may provide a method to strengthen a population of appropriate Th1 cells(191). Further research is needed to investigate whether this approach is useful, as well as the time sequence and profile of children who would benefit from it.

Cytokine Therapy

When considering cytokine therapy, IL-4 is a highly attractive target, because of its critical role in regulating both IgE isotype switch and the commitment of T cells to a Th2 cytokine pattern. Key regulatory units in the IL-4 gene promotor structure have been identified, but most of the nuclear factors that bind to these regulator sequences are common to many related genes (such as IL-5). However, there are indications of specific regulatory proteins, offering the possibility for selective inhibitors(192). Inhibition of the IL-5 response might prevent terminal differentiation of eosinophils and abolish the eosinophilic component of inflammation in asthma and allergy(193). The removal of specific cytokines in animal models often had more profound and informative effects than cytokine addition(192). This suggests that cytokine antagonists may often be more valuable in patients than the addition of cytokines.

Recombinant IFN-γ therapy over a 12-wk period in adult patients with severe atopic dermatitis turned out to be safe, well accepted, and effective in reducing inflammation, clinical symptoms, and eosinophilia(194). Despite clinical improvement, no change in serum IgE levels was found, although spontaneous in vitro IgE synthesis decreased. This was not unexpected, because serum IgE levels reflect IgE synthesis by terminally differentiated B cells, which may no longer be responsive to rIFN-γ. Two 4- and 5-y-old children suffering from refractory atopic dermatitis were treated with rIFN-γ, which was well tolerated. Clinically, some benefit was observed in one child, whereas no beneficial effect was seen in the other, although marked immunologic changes were noted(195). Further studies are needed before IFN-γ can be recommended for therapy of pediatric atopic eczema. Considering the pleiotropy and the redundancy in the cytokine network, an approach targeting combinations of several involved cytokines seems appropriate.

Glucocorticosteroids

Glucocorticosteroids are widely used as the most effective agents currently available in the treatment of atopic diseases. The anti-inflammatory actions of topical corticosteroids may depend partly on down-regulation of cytokine synthesis. In vitro it was demonstrated that corticosteroids inhibit the production of IL-4 and also down-regulate the transcription of IL-4 mRNA in PBMC cultures(196, 197). Furthermore, dexamethasone-suppressed IL-5 production in atopic human PBMCs through an inhibitory action on its gene expression(198). This suppression of IL-5 gene expression is one of the most important mechanisms by which glucocorticosteroids inhibit eosinophil functions in the treatment of atopic diseases.

Anti-IgE

Anti-IgE forms a prophylactic/treatment agent designed to decrease the amount of IgE by targeting and subsequent lysis of IgE-secreting B cells. Furthermore, anti-IgE treatment could decrease the biologic activity of IgE by initiating IgE clearance and by preventing IgE binding to receptors on effector cells. For anti-IgE to be safe, a MAb was selected that did not bind to receptor-bound IgE (either CD23 or FcεRI). This antibody was humanized and is being currently evaluated in trials with adult atopic subjects, but not yet in young allergic children(199).

Disodium cromoglycate also has an established role in the prophylactic treatment of allergic disease. The possible mode of action of cromolyn sodium was thought to be stabilization of mast cells and subsequent prevention of mediator release after antigen challenge(200). More recently, it was shown that disodium cromoglycate inhibits T cell-driven IgE synthesis by human B cells, by inhibiting deletional switch recombination fromμ to ε(201). It was also demonstrated that disodium cromoglycte inhibited the recruitment of inflammatory cells, particularly eosinophils, possibly by affecting the generation of cytokines and the expression of leukocyte-specific adhesion molecules(202). These results suggest novel potential mechanisms for the prevention of allergic disease by disodium cromoglycate.

CONCLUDING REMARKS

In the understanding of the development of allergic diseases in childhood, progress is rapidly evolving with respect to the contribution of genetic, environmental, and immunopathologic factors. The key factors currently considered relevant in the development of allergic diseases at a young age are: 1) With respect to genetic markers the linkage of genes located on chromosome 5q31.1, encoding the cytokine gene cluster, to total serum IgE levels, is now the subject of many investigations. The linkage and relationship of genetic markers on chromosome 14 (T cell receptorsα/δ) to allergen-specific IgE levels, and the linkage between genes on chromosome 11q (FcεR1) and allergy, however, still remain to be further unraveled. This genetic propensity can be further modified by environmental factors, such as allergen exposure, pollution, and viral infections, whose implications remain controversial and in need of further study. 2) Concerning cytokines, a reduced allergen-induced IFN-γ production, as well as elevated IL-4 serum concentrations, before the onset of allergic symptoms in infants point to a developmentally early skewing toward a Th2 rather than a Th1 response. 3) Transiently elevated IgE levels in healthy infants could be followed by development of immunologic tolerance, whereas persistently elevated IgE levels in atopic children could be associated with failure in immunologic tolerance. This might be due to the presence or absence of suppressive factors, possibly IFN-γ. Further research to establish the role of other factors involved is needed. In addition Th2-derived stimulatory factors, like IL-4 or IL-13 are likely to be present in higher levels in atopic individuals. 4) Failure of tolerance induction based on or accompanied by immaturity of the immune system, may create increased sensitivity to allergic sensitization in early infancy. T cells in combination with APC most likely play an important role in allergy development. This role is probably largely due to genetic factors. 5) Abnormalities at the level of B cells, manifested by an increased CD40 expression, resulting in an enhanced responsiveness to CD40, may also contribute to disease. It remains to be established whether elevated IgE levels in atopics are based on increased precursor frequencies of IgE secreting B cells, differences in activation requirements of B cells or differential activity of regulatory T cells (either cytokines or membrane-bound molecules). 6) The contribution of the various immunopathologic factors and their possible interaction in the clinical heterogeneity of allergic diseases, like atopic dermatitis, food allergy, allergic asthma, and allergic rhinitis, awaits further elucidation. Currently, we and other groups are investigating the role that these factors may play in the immunopathologic changes in young potentially atopic children, which probably will result in detection of disease markers and new therapeutic approaches.