The Role of Costimulatory Molecules in Allergic Disease and Asthma

The prevalence of allergy is increasing worldwide, especially in Western countries [1,2]. This trend may be due to increased hygiene which prevents early exposure to microorganisms in infancy and increases the risk of developing an allergic disease in the future [3]. Air pollution may also favor the development of allergies [4]. Among allergic diseases, asthma is of particular concern as it can evolve into a life-threatening chronic pulmonary disease. It is now recognized that inhalation of aeroallergens by sensitized patients initiates a massive T helper 2 (TH2) cell response resulting in the production of interleukin-4 (IL-4), IL-5 and IL-13 by CD4+ T cells [5]. This triggers the development of typical signs of bronchial asthma including airway hyperresponsiveness, bronchoconstriction, mucus secretion and airway remodeling. The recruitment, differentiation and activation of TH2 cells play a cardinal role in the development of allergic disease. This TH2 polarization is mainly controlled by antigen-presenting cells (APCs) that provide signals capable of supporting the differentiation of naïve T cells into TH2 cells.

Initially, Lafferty et al. [6] suggested that T lymphocytes require two coordinate signals to be fully activated. The first signal confers specificity to the immune response and plays an important role in the recognition of the major histocompatibility class-peptide complex at the surface of APCs by T lymphocytes. However, this primary signal is not sufficient to completely activate these T cells. To become fully effective, a second, nonspecific costimulatory signal is often required by T cells. These signals are provided by interactions between APCs and T cells through surface molecules expressed on T lymphocytes. APCs express costimulatory molecules such as CD80 (B7-1) and CD86 (B7-2) which belong to the B7 family. They provide modulation of T-cell function by ligation to their receptors, CD28 or cytotoxic T lymphocyte-associated antigen 4 (CTLA-4). Engagement of CD28 or CTLA-4 has been observed to have opposite effects. CD28 promotes T-cell activation and survival while CTLA-4 inhibits T-cell responses and regulates peripheral T-cell tolerance [7]. This demonstration provides clear evidence that costimulatory molecules are involved in the fine tuning of the T-cell response by mediating both stimulatory as well as inhibitory signals. In many recent studies, numerous new costimulatory molecules have been described leading to the recognition that costimulation pathways are more complex than the classical two-signal model. In general, costimulatory molecules are divided into 2 main families: molecules from the B7:CD28 family, such as CTLA-4 or programmed death (PD)-L1, and from the tumor necrosis factor receptor (TNFR) superfamily such as OX40 or CD27 [8]. All these costimulatory molecules have particular effects on T-cell activation, function and survival (table 1) and are implicated in nearly all inflammatory diseases. Studies to better characterize the specific role of these molecules in allergy and asthma are still ongoing. In this review, we summarize current knowledge concerning the role of costimulatory molecules in allergy and analyze the potential functions of the emerging new subsets of costimulatory molecules recently described.

The first costimulatory molecules described were the ligands of CD28: CD80 (B7-1) and CD86 (B7-2). The CD28 costimulation pathway is an important factor for the promotion of an effective antigen-specific immune response. However, CD28-deficient mice are still able to develop allergic airway inflammation showing that CD28 cannot be solely responsible for the development of an allergic response [9]. The expression of CD28 ligands (CD80 and CD86) has been extensively studied in clinical samples from asthmatic patients. Hofer et al. [10] reported that B lymphocytes from asthmatic patients exposed to allergens express higher levels of surface CD86, but not CD80, compared with those from asthmatic patients not exposed to allergen or with those from healthy individuals (table 1). Another study demonstrated that CD80 and, to a lesser extent, CD86 were upregulated at the surface of alveolar macrophages from allergic patients compared with those from pulmonary sarcoidosis, extrinsic allergic alveolitis patients or from normal subjects [11]. In contrast, Burastero et al. [12] observed that in allergic individuals, CD80, but not CD86, is highly expressed by alveolar macrophages. CD86 can also be expressed in a soluble form, where the transmembrane domain is deleted. This form is mostly produced by circulating monocytes and, like membrane-bound CD86, cross-links CD28 or CTLA-4 and activates T lymphocytes [13]. In patients with acute asthma, the level of soluble CD86 has been shown to be increased relative to that in patients with stable asthma or in healthy individuals [14]. Monocytes from allergic patients produce more soluble CD86 compared with those of healthy individuals [14]. It was also observed that the level of soluble CD86 is correlated with the severity of the airway hyperreactivity (AHR). These results are consistent with studies which show that the concentrations of soluble CD80 and soluble CD86 are elevated in asthmatic patients. Interestingly, the administration of a glucocorticoid (e.g., prednisolone), used to reduce airway inflammation in allergic patients, reduces the level of circulating CD86 [15]. Recently, Ritprajak et al. [16] demonstrated that the topical administration of a silencer RNA specific to the CD86 gene reduced local inflammation in a mouse model of atopic dermatitis by decreasing the recruitment of dendritic cells (DCs) into the skin, production of antigen-specific IL-4 and induction of serum immunoglobulin E (IgE) and IgG1. It has also been reported that pulmonary tolerogenic DCs stimulated by allergen exposure express high levels of both CD80 and CD86 [17]. Taken together, these studies suggest that overexpression of CD80 and/or CD86 is correlated with the development of allergic disease and asthma.

Within the B7 family, CD28 and inducible costimulator (ICOS) are most homologous with respect to structure and function. Both are type I transmembrane receptors expressed as homodimers, with an extracellular (Ig)V-like domain, a hallmark of receptors of the B7-related family [18]. ICOS expression is induced in vitrowithin 24–48 h of activation on all T helper-primed cells [18,19]. ICOS has been shown to regulate the production of TH2 cytokines [20] and plays a critical role in lung mucosal inflammatory responses [21]. Furthermore, this costimulatory molecule has been suggested not only to intensify some of the functions of CD28 during an already established immune response but also to induce additional T-effector cell functions [22]. Recent studies suggest that ICOS-mediated costimulation may regulate TH2-effector cell function without affecting TH2 differentiation [23]. Another study showed that transfer of ICOS-enriched T cells followed by allergen airway challenge induced infiltration of recipient T and B cells as well as local production of allergen-specific IgE by intrapulmonary plasma cells [24]. In contrast, transfer of the ICOS-depleted T-cell fraction resulted in the recruitment of significantly lower numbers of B cells with no local IgE production. These data indicate that expression of ICOS defines a subset of T effector cells that are required for B-cell infiltration and local IgE production in lung tissue. According to Tesciuba et al. [23], ICOS stimulation increases the migration of lymphocytes into draining lymph nodes by augmenting the expression of attractant chemokines CCL21 and CXCL13. In other reports, the increased production of IL-5, which is a main factor for the differentiation, maturation and recruitment of eosinophils, is attributed to ICOS+ cells [24]. ICOS+ T cells also promote the differentiation of B cells and IgE-producing plasma cells through the enhanced production of IL-4 and IL-10 [25]. ICOS-deficient mice are unable to induce high IgE responses demonstrating their role in the induction of IgE production [26]. Some studies also suggest that intermediate ICOS expression is associated with high production of TH2 cytokines, whereas high levels of ICOS predominantly translate into high IL-10 production [27]. Surprisingly, ICOS/ICOS-L interaction not only promotes the development of TH2-driven inflammation but also mediates mucosal tolerance, as studies have indicated that pulmonary DCs in the bronchial lymph nodes of mice exposed to respiratory allergen induced the costimulation of regulatory T cells (T_reg cells) via the ICOS/ICOS-L pathway [28]. These T_reg cells produce IL-10, show inhibitory activity and, when adoptively transferred into sensitized mice, have the ability to inhibit the development of AHR. These reports suggest that both the development and inhibitory function of T_reg cells are depending upon the presence of IL-10 and ICOS/ICOS-L interaction. Studies in ICOS-deficient patients supported the findings in mice by demonstrating that CD4+ T cells in those patients cannot be skewed towards suppressive or anergic phenotype [29]. Interestingly, Taylor et al. [30] demonstrated that IL-10 suppresses CD28 and ICOS-mediated T-cell costimulation by a mechanism dependent on Src homology 2 domain-containing protein tyrosine phosphatase 1 (SHP-1). Specifically, IL-10 ligation to its receptor triggers Tyk2 activation followed by SHP-1 phosphorylation. This active form of SHP-1 dephosphorylates the costimulatory molecules CD28 and ICOS. Thus, signal transduction downstream of engagement of either molecule is abolished and T-cell activation through these costimulatory molecules is impaired. Beside IL-10, ICOS is involved in expression and differentiation of IL-17-producing helper T cells [31]. A recent study by Bauquet et al. [31] suggests that there are significantly lower levels of IL-17 from T cells in ICOS-deficient mice. The authors suggest that engagement of ICOS induces the expression of c-Maf, which regulates IL-21 production and controls the expansion of TH17 cells. Beyond the role of ICOS on T cells, two recent studies have demonstrated the importance of ICOS in iNKT cell function in the development of AHR and in the homeostasis and survival of CD4+ iNKT cells. ICOS expression is upregulated on iNKT cells upon α-galactosylceramide stimulation. Moreover, blockade of the ICOS pathway using a specific antibody or gene knock-out strategy abrogates cytotoxicity and cytokine production triggered by α-galactosylceramide treatment [32]. The number of CD4+ iNKT cells were greatly reduced in the periphery but not in the thymus of both ICOS–/– and ICOS-L–/– mice compared with wild-type mice, indicating that ICOS/ICOS-L interactions are critical in homeostatic survival of CD4 iNKT cells [19]. In contrast, the number of iNKT cells and the level of ICOS expression in CD28–/– mice is comparable with that in wild-type mice, suggesting that signaling via ICOS but not via CD28 plays a unique role in regulation of CD4 iNKT cell homeostatic survival. Thus, the iNKT cells expressing ICOS contribute significantly to the development of AHR.

CTLA-4 has been described to be an important regulator of T-cell activation. CTLA-4 is constitutively and exclusively expressed by T lymphocytes in both mice and humans [7]. CTLA-4 expression confers to T-lymphocyte regulatory functions [33]. Indeed, the contribution of CTLA-4 in the regulation of the immune system is demonstrated by the development of multiple organ autoimmune pathologies and lymphoproliferative disease in CTLA-4-deficient mice [34,35]. Blockade of CTLA-4 activity abolishes the suppressive function of CD4+ CD25+ T cells [33]. In a mouse model of inflammatory bowel disease, the effect of transfer of a population of CD4+ CD45RB^low T_reg cells in decreasing intestinal inflammation is abrogated by the coadministration of a blocking anti-CTLA-4 antibody (table 1) [36]. These studies suggest that the engagement of CTLA-4 at the surface of T_reg cells by its ligands CD80 or CD86 contributes to the regulation of suppressive functions of T_reg cells. Polymorphism in CTLA-4 gene is also considered a risk factor for allergy and asthma. Howard et al. [37] characterized four single nucleotide polymorphisms which were related to allergic and asthma phenotypes. They demonstrated that these specific polymorphisms alone or in combinations are correlated with an elevated IgE titer or bronchial hyperresponsiveness in patients with asthma. Interestingly, in the same study, no correlation between allergic phenotype and single nucleotide polymorphisms for CD28 was observed. In a similar report, Lee et al. [38] studied the impact of two polymorphisms in the CTLA-4 promoter (–318 C/T) and gene (+49 C/G) [39]. They demonstrated that a polymorphism at the level of the promoter was correlated to asthma severity while the +49 C/G polymorphism is associated with airway hyperresponsiveness. These findings confirm that CTLA-4 is indeed involved in the course of allergic diseases. Furthermore, CTLA-4 is often considered as a marker of T_reg cells. In that regard, Meiler et al. [40] demonstrated that, in response to multiple bee stings, beekeepers develop a protective immune response by the development of antigen-specific type 1 regulatory T cells. The authors observed that in peripheral blood mononuclear cells isolated from beekeepers, the suppressive activity of antigen-specific IL-10-secreting cells is blocked by an anti-CTLA-4 and an anti-PD-1 antibody. The authors confirmed the role of CTLA-4 in the suppressive activity of T_reg cells in humans and suggest that the engagement of CTLA-4 on T_reg cells leads to reduced T-cell-receptor-derived signaling which is required for the induction of the suppressive activity [40]. Similarly, analysis of Bet v 1-specific CD4+ T cells from healthy individuals at the single-cell level using major histocompatibility class II peptide tetramer revealed that a fraction of these cells expressed CTLA-4 and Foxp3, suggesting that they represent a population of T_reg cells [41]. Interestingly, CTLA-4 seems to play a more important role in the sensitization phase than in established allergy. In a model of mice sensitized with grass pollen, the administration of a blocking anti-CTLA-4 or a blocking anti-CD154 (anti-CD40L) antibody during the sensitization phase prevents the production of allergen-specific antibody. In contrast, in sensitized mice, anti-CTLA-4 or anti-CD154 antibodies failed to decrease the level of allergen-specific IgE [42].

The PD-1 receptor and its ligands PD-L1 (B7-H1) and PD-L2 (B7-DC) belong to the B7:CD28 family of receptors. The PD-1 receptor was initially discovered in T cells undergoing cell death [43]. The inhibitory signal provided by engagement of PD-1 was demonstrated by the development of autoimmune diseases in PD-1-deficient mice [44]. Several groups have subsequently reported that engagement of PD-1 by PD-L1 or PD-L2 results in inhibition of proliferation and polarized or altered cytokine production [45,46,47]. Studies are just beginning to elucidate PD-L1 and PD-L2 function in allergy and asthma. In a mouse model of asthma, Matsumoto et al. [48,49] demonstrated that PD-L2 is highly expressed on pulmonary DCs and macrophages of sensitized mice. Moreover, administration of blocking antibody against PD-L2, but not PD-1 or PD-L1, during challenge enhances the airway hyperresponsiveness and production of TH2 cytokines (table 1) [49]. This effect is mediated by interferon (IFN)-γ, given that no improvement is observed in IFN-γ-deficient mice following treatment with anti-PD-L2 [48]. In addition, administration of the sHIgM12 antibody (an antibody inducing reverse signaling through PD-L2) in a mouse model of allergic asthma blocks the development of AHR [50]. An additional study demonstrates that administration of PD-L2-Fc in a mouse model of allergic asthma resulted in elevated levels of serum IgE and increased eosinophilic and lymphocytic infiltration into the bronchoalveolar lavage fluid [51]. These studies emphasize the pivotal role of PD-L2 in the development of allergic asthma. In addition, in a model of experimental allergic conjunctivitis, treatment with anti-PD-L2 blocking antibody during the effector phase enhanced infiltration of eosinophils into the conjunctiva without change in the systemic response [52]. Finally, using PD-L2-deficient mice, it was reported that this molecule is dispensable for TH2 differentiation and required for the induction of mucosal tolerance [53]. Taken together, these reports strongly suggest that PD-L2 is involved in the downregulation of TH2-allergic immune response. In a mouse model of hapten-induced contact hypersensitivity, Kim et al. [54] demonstrated that blockade of PD-L1 enhanced the activity of hapten-specific T cells and the administration of hapten-carrying PD-L1 on DC-induced tolerance in animals sensitized by hapten challenge. Similarly, Tsushima et al. [55] demonstrated that anti-PD-L1 blocking antibody, but not anti-PD-L2 blocking antibody, enhanced contact hypersensitivity reaction, possibly by increasing the proliferative response of T cells in response to hapten-pulsed APCs. This suggests a unique role of PD-L1 in the regulation of inflammatory responses (table 1). Piconi et al. [56] had reported earlier that during allergen-specific immunotherapy, the expression of PD-L1 on both monocytes and B lymphocytes is increased relative to that in the untreated control group. The authors proposed that PD-L1 could be used as a marker to monitor the effect of allergen-specific immunotherapy and could also be targeted to enhance immunosuppression. A subset of PD-L1-positive tolerogenic Langerhans cells was described by Allam et al. [57] in the sublingual mucosa in humans. The authors observed that upon stimulation by a Toll-like receptor 4 (TLR4) ligand, these cells release a higher level of IL-10 compared with untreated control cells. These Langerhans cells have decreased capacity to stimulate T cells and are able to support the differentiation of T_reg cells expressing Foxp3, producing IL-10 and transforming growth factor-β [58]. These sublingual tolerogenic Langerhans cells stimulated via their TLR4 expressed higher levels of the coinhibitory molecules PD-L1 and B7-H3, while CD86 expression is lowered. Consequently, the expression of these molecules by APCs seems to be linked with tolerogenic properties. These studies represent a body of evidence which suggests that PD-L1 is involved in the maintenance of the peripheral tolerance and may contribute to the induction of allergy. Clearly, further work is required to understand the role of PD-1 and its ligands in allergic diseases and asthma.

The precise role of the other ligands of the B7 family, such as B7-H3, is far less clear. The ligand of B7-H3 is still unknown but seems to be present at the surface of activated T cells [40]. A recent study proposed that B7-H3 supports the differentiation of TH2 cells during sensitization in animal models of asthma, as blocking B7-H3 with an antibody results in a decreased production of TH2 cytokines in draining lymph nodes, reduced infiltration of eosinophils in lungs and in reduced AHR [59]. In contrast, administration of blocking B7-H3 antibody during the sensitization phase increases the severity of allergic conjunctivitis in mice, possibly by inducing IL-5 production in the spleen [60]. These studies indicate that B7-H3 is potentially involved in the regulation of allergic disease and asthma. However, many questions pertaining to B7-H3 will still need to be addressed in future mechanistic and prospective studies.

Members of the TNFR superfamily can have distinctive cytoplasmic death domains which are involved in apoptotic signaling. Other members of the superfamily lack such a domain, with no apparent homology in the cytoplasmic tail. This latter group of receptors is involved in gene activation and antiapoptotic signaling. The role in allergy of TNFR family members such as OX-40, 4-1BB, CD30, Fas, CD27 and CD40 has been recently studied and reported by several investigators.

OX40 and its ligands play an important role in costimulation of allergen-specific lymphocytes. Activated CD4+T cells express OX40, whereas OX40L is mainly expressed by APCs [61,62]. It was observed that ligation of OX40 increases IL-4 production by naïve cells and promotes their development into effector cells producing high levels of the TH2 cytokines IL-4, IL-5 and IL-13 [62]. OX40-OX40L interaction also plays an important role in deciding the fate of CD4+ T cells during allergic inflammation. It was reported that OX40L expressed by thymic stromal lymphopoietin-activated DCs enables these APCs to trigger allergic inflammatory TH2 responses [63]. In a similar study, Ito et al. [63] showed that blocking of OX40-OX40L interaction inhibits the production of TNF-α and TH2 cytokines and enhances the production of IL-10. In allergen-induced models of asthma, OX40- or OX40L-deficient mice exhibit markedly impaired reactivation of TH2 memory cells and TH2 responses as well as diminished lung inflammation [64,65]. OX40-deficient mice developed a weak TH2 response and airway inflammation after sensitization to ovalbumin indicating the cardinal role of this molecule in the initiation of allergic immune responses [66]. Additionally, Duan et al. [67] reported that intranasal exposure to lipopolysaccharide/endotoxin leads to the interaction between OX40 and OX40L, which, with other inflammatory effects of TLR4 signaling, alters the balance between Foxp3+ T_reg cells and effector T cells and influences the susceptibility to allergic inflammatory disease. Moreover, it was reported that OX40 inhibits the development of adaptive Foxp3+ T_reg cells from naïve CD4+ T-cell populations in response to transforming growth factor-β [68,69]. These studies suggest that preventing OX40-OX40L interaction might be beneficial in improving the effectiveness of allergen immunotherapy as those interactions might induce mucosal tolerance through development of regulatory cells. Interestingly, OX40L can also be present at the surface of mast cells, which play a pivotal role in allergy. T_reg cells can abrogate mast cell degranulation following FcεRI engagement by IgE, via OX40-OX40L interaction [70].

Similarly to ICOS, OX40-OX40L interaction can modulate the function of iNKT cells which also play an important role in allergic disease and asthma. A recent report suggests that iNKT cells interact with plasmacytoid DCs via OX40-OX40L interaction, downregulate the CD8+ immune response and prevent tissue damage [71], while iNKT stimulated with OX40L-expressing DCs produces more IFN-γ and CD69 [72]. However, the phenotypical difference between OX40-deficient and OX40L-deficient mice suggests that OX40 might have more than 1 ligand; thus, it is difficult to draw a definite conclusion regarding OX40L.

Several clinical studies have revealed a link between the upregulation of CD30 and allergic diseases and asthma. Initially, it was reported that the concentration of the soluble form of CD30 is higher in patients with asthma or atopic dermatitis than in healthy controls and correlates directly with the severity of the disease [73,74]. In atopic dermatitis patients, CD30 expression is increased at the surface of Langerhans cells, CD4+ and CD8+ T cells [74,75]. Rojas-Ramos et al. [75] demonstrated that the level of CD30 expression at the surface of CD4+ T cells was correlated with the production of IL-4 after restimulation of CD4+ T cells isolated from allergic patients. These data suggest that the expression of CD30 at the surface of different immune cells or in a soluble form in the serum could be linked to a TH2 polarization. A recent study demonstrating a reduction of the level of soluble CD30 in venom-specific immunotherapy also suggests a relationship between TH2 activity and CD30 expression [76].

Another member of the TNFR superfamily is 4-1BB (CD137), which has been suggested to suppress antigen-specific helper T cells and B cell-dependent humoral immune response [77]. Additionally, 4-1BB is specifically expressed by eosinophils from atopic patients with IgE-mediated dermatitis or asthma [78]. In a mouse model of allergic asthma, 4-1BB blocking antibody decreased airway hyperresponsiveness and reduced the level of allergen-specific IgE in sera of sensitized mice [79]. Also, pulmonary T lymphocytes of anti-4-1BB-treated mice showed a decreased proliferation and produced less IL-5 in response to ovalbumin. However, IL-4 and IL-5 levels in BAL fluid were only marginally reduced. 4-1BB is also expressed by mast cells and acts as a costimulatory molecule when mast cells are stimulated through their FcεRI. Agonistic anti-4-1BB antibody enhances mast cell cytokine production after engagement of FcεRI [80]. Polte et al. [81] demonstrated that in a mouse model of asthma, prophylactic administration of a blocking anti-4-1BB is capable of preventing the establishment of airway hyperresponsiveness, eosinophil infiltration and production of allergen-specific IgE and reduces the production of TH2 cytokines while enhancing secretion of TH1 cytokines. Depletion of CD8+ cells or blockade of IFN-γ abolished the protective effect of 4-1BB blocking antibody, which suggests that the effect of 4-1BB is dependent on IFN-γ-producing CD8+ T cells [82]. Interestingly, after sensitization, injection of anti-4-1BB blocking antibody totally reverses the allergic phenotype in mice, which suggests that intervention within the 4-1BB pathway might offer a novel therapeutic approach in patients with asthma. Finally, in a similar approach using a model of atopic conjunctivitis, treatment with an agonistic anti-4-1BB before or after sensitization abolished the development of allergic conjunctivitis [83].

The CD95 (Fas or APO-1) antigen is a 40- to 50-kDa transmembrane glycoprotein that also belongs to the TNF superfamily. This cell surface molecule mediates apoptosis (programmed cell death) and is strongly upregulated on activated T cells, B cells, natural killer cells and thymocytes. In the field of allergy, it is important to note that Fas is also expressed by eosinophils in sensitized mice [84], and stimulation of this receptor with an anti-Fas antibody triggers apoptosis of eosinophils. In vivostudies have also shown that lung eosinophilia is reduced by administration of anti-Fas antibody. Moreover, failure of induction of eosinophil apoptosis by Fas/FasL interaction could explain the chronic eosinophilic airway inflammation found in patients with asthma. Fas-deficient mice sensitized to ovalbumin show a delayed AHR resolution compared with wild-type mice [85], an observation that could be explained by a decreased apoptosis of eosinophils and effector T cells. In this regard, a study by Tong et al. [86] proposed that this delayed resolution of eosinophilia in Fas-deficient mice is due to the absence of Fas on T cells but not on eosinophils. Consistent with those studies is the observation that transfer of DCs genetically engineered to express FasL significantly reduces induction of AHR in sensitized recipients [87]. On the other hand, Uller et al. [88] showed that anti-Fas antibody exacerbates established airway inflammation since stimulation of Fas receptor by eosinophils triggers cytolysis and causes secondary necrosis of apoptotic eosinophils. Therefore, time kinetics and location of Fas/FasL expression play a major role in the regulation of immune responses in allergy and asthma.

CD27 is a type I glycoprotein expressed on T and B cells and also belongs to the TNFR family. The ligand for CD27 is CD70, another member of the TNF family [89], which is expressed not only on activated B cells but also on T cells, particularly on activated CD4+ CD45RO+ T cells [90]. It has been demonstrated that the CD27/CD70 interaction is involved in the differentiation of B cells into plasma cells [17,18]. The roles of CD27 and CD70 as costimulatory molecules in allergic diseases are less clear. One report suggests that B cells transfected with a plasmid expressing CD70 significantly augments IgE production by enhancing B-cell proliferation and differentiation into plasma cells. However, another recent report shows that CD27 and CD70 do not play a role in the development of experimental allergic conjunctivitis in mice [91]. Although it seems that the effects of CD27/CD70 interaction on B-cell and Ig synthesis in the murine and human systems are somewhat different, more experiments in both animal models and humans need to be done to further clarify their role in allergy and asthma.

Another member of the TNFR superfamily, CD40, is a costimulatory protein found on APCs and is required for their activation. This receptor has been found to be essential in mediating a broad variety of immune and inflammatory responses including T cell-dependent immunoglobulin class switching, memory B-cell development, germinal center formation, B- and T-lymphocyte activation and regulation [92]. During allergen sensitization, cooperation between T and B lymphocytes through CD40-CD40L interaction is a fundamental signal to trigger isotype class switching towards IgE [93]. CD40 is also expressed at the surface of airway epithelial cells. Its engagement increases the production of inflammatory mediators suggesting that CD40 ligation favors airway inflammation [94,95]. Recently, Suzuki et al. [96] showed that the inhibition of CD40 expression using small interfering RNA in ovalbumin-sensitized mice results in a decreased production of TH2 cytokines and increases the number of T_reg cells.

CD2 is a member of the immunoglobulin superfamily and is expressed on all peripheral blood T cells. It is one of the earliest T-cell markers, being present on more than 95% of thymocytes, and is also found on some natural killer cells but not on B lymphocytes. CD2 interacts with lymphocyte function-associated antigen-3 to mediate adhesion between T cells and other cell types [97]. CD2 is also considered as a costimulatory molecule on T and natural killer cells. Engagement of this receptor by its ligand lymphocyte function-associated antigen-3 induces T-cell proliferation and cytokine production [98]. CD2 engagement seems to inhibit TH1 activity while favoring TH2 development [99]. In a model of mercury chloride-induced autoimmune disease in mice, administration of a specific CD2 antibody increased the production of TH2 cytokines such as IL-4 and increased the serum level of IgE and IgG1 autoantibodies. Interestingly, the CD2 costimulation signal can be counteracted by regulatory cytokines such as IL-10 [100]. Since CD2 is also expressed by monocytes, the high CD2 expression on monocytes defines a population with elevated FcεRI expression [101]. In asthmatic patients, CD2 expression at the surface of these monocytes is correlated with the level of plasma IgE. In addition, administration of an anti-IgE (omalizumab) decreases the expression of FcεRI at the surface of CD2^high monocytes. Therefore, it has been proposed that monocytes expressing a high level of CD2 represent an attractive target for the treatment of allergic disease and asthma.

Families of costimulatory molecules are involved in the regulation of most inflammatory diseases by finely controlling the intensity of the immune response. Costimulatory molecules are implicated in the development and control of allergic inflammation characterized by the establishment of an acute TH2 polarization. Elucidation of the role of costimulation pathways in the development of new subsets of T helper cells has just begun, and most of the mechanisms underlying the regulation of atopic diseases by costimulatory molecules are unknown and require further investigation. With an increased understanding of these immunological mechanisms, new therapeutic strategies in the treatment of allergic airway diseases can be created by analyzing the role of costimulatory molecules which are critically involved in the induction and maintenance of allergen-induced airway diseases. Taken together, recent studies have begun to provide insight into the role of costimulatory molecules and give us new clues to design more efficient therapies to fight the increasing public health problem that allergies represent.

O.A. was supported by grant NIH R01 AI066020.