- Email: [email protected]
IgE synthesis
IgE synthesis jl G. Delespesse, University
of Montreal,
Notre-Dame
Canada and *CIBA-GEIGY, Current
M. Sarfati
Opinion
and C. Heusser*
Hospital,
Immunology in Immunology
Introduction The lirst demonstration that the production of IKE antibody is T-cell dependent and that its regulation is isotypespecilic was reported in the early 1970s (Okumura et al, JZmmuno11971,106:101~1025). Tada and co-workers have shown that neonatally thymectomized rats are incapable of developing an IgE antibody response, but that after adult thymectomy the IgE but not the 1gG antibody response is increased and sustained for longer. The IgE-potentiating effect of adult thymectomy is abolished by reconstitution with syngeneic T lymphocytes. Ishizaka (Annu Rev Zmmunol 1988, 6:513-535) has extensively documented one of the mechanisms by which rodent T cells mediate the isotype-specific but not antigen-spe cilic regulation of IgE synthesis. Because this work has recently been raiewed in detail (Ishizaka, 1988), it will only be summarized very brieily here. According to this model, a subset of T helper (TH) cells [capable of expressing the low-affinity receptor for IgE, the FCE receptor (FGR)II] may release &E-binding factors that may either suppress or increase the secretion of IgE by activated I@-bearing B cells. The effect of these factors is isotype-specific and their suppressive or potentiating activity is dependent upon their N-linked glycosylation which, in turn, is regu lated by T cells. A complementary DNA (cDNA) coding for rodent IgE-binding factor has been cloned and functionally expressed (Martens et al, Proc NatlAcud Sci USA 1985,82:2460-2464); the predicted amino-acid sequence has no homology with that of the human (Liidin et al, EMBOJ1987,6:10!+116) or the mouse [l] Fc&RII,which is also known as CD23. The rodent IgE-binding factors are, therefore, not related to the soluble fragments of human Fc&RII, also called &E-binding factors because they bind to human IKE. In the human (but not in the mouse) model, Fc&RII and its soluble fragments have been shown to regulate IgE synthesis. The availability of several recombinant lymphokines and of the corresponding neutralizing antibodies has permitted a major advance in our understanding of the isotypespecific regulation of the immune response. It is now
Research Center, Montreal,
Division,
Basel, Switzerland
1990, 2:506-512
clear that selection of IgE isotype is dictated by the preferential production of the T-cell-derived lymphokine interleukin (IL)-4, whereas it is suppressed by interferon (IFN)-y. In the mouse, T, lymphocytes may differentiate in two distinct subsets with diiferent lymphokine production profiles: Tul cells secrete IL-2, IFNy and lymphotoxin (LT), but not IL-4 or IL-5, and TH2 cells produce IL-4 and IL-5, but not IL-2, IFNy or LT (Mosmann et al, J Zmmunol 1986, 136:234%2357). Hence, the development of an IgE response appears to depend upon the preferential differentiation of T, cells into Tn2 lymphocytes. Here, I will focus on the role of lymphokines, T cells and &E-binding factors (soluble CD23) in the regulation of IgE synthesis.
The role of interleukin-4 In vivo studies in the mouse, using neutmlizing antiIL-4 or anti-IL-4-receptor antibodies, have unequivocally demonstrated that the primary, and most of the secondary, antigen-specific or polyclonal IgE (but not IgG) antibody responses depend upon the production of IL-4 [L&3]. However, another form of T-cell help, which acts at a later stage than IL-4 and which can be inhibited by anti-CD4 (but not by anti-IL-4), is also required for the induction of an IgE response. Most of the secondary and of the established ongoing IgE response is suppressed by the administration of saturating doses of anti-IL-4. The small part of the secondary response that is resistant to anti-IL-4 treatment is thought to be mediated by IgE memory B cells whose activation should be IL-4independent. Another conclusion from these in vivo studies is that the role of IL-4 in the persistent IgE response of parasite-infected animals is to recruit new IgE B cells from their IgM precursors rather than to promote the activity or to expand the pool of &E-secreting cells (Finkelman et al, Annu Rev Zmmunol 1990, in press). This view is consistent with recent observations that the spontaneous in
Abbreviations B-Cll~hronic lymphocytic leukemia B cells; cDNA-complementary DNA; CH
506
@
Current
Science
Ltd ISSN
0952-7915
IgE synthesis Delespesse, Sarfati and Heusser
vitro synthesis of IgE by lymphocytes from atopic patients is IL-4-independent, in as much as it is not inhibited by a neutralizing anti-IL-4 antibody (Delespesse et al., 7th Int Cong Zmmunol1989, A75-A77). In contrast to anti-IL4, anti-CD4 injected at the peak of a secondary response has no effect on the decline in serum IgE antibody levels. This suggests that in the course of an ongoing IgE response, II-4 may be produced by a CD4- cell; mast cells might have such a role because they may release IL-~ after &E-dependent stimulation (Plaut et al, Nature 1989, 3396467). The mode of action of IL-4 at the cellular and molecular levels has been analysed in vitro. The first conclusion of these studies is that IL-4 is necessary but not sufficient to transform resting IgM precursor B cells into IgE-producing cells. Hence, the addition of IL-4 to cultures of punlied mouse or human B cells does not induce IgE syn thesis unless additional signals are provided, for example lipopolysaccharide (LPS) to mouse B cells (Coffman et al., J Immunoll986,I36:4538-454I) or Epstein-Barr &US (EBV) to human B cells [4]. Several groups have demonstrated that normal resting B cells may be induced to synthesize IKE after cognate irteraction with selected helper T-cell clones (P&e et al, JZmmunol19@3,141:121~1224). In this model, two signals are required in addition to IL-4 to obtain an IgE response. The first signal is provided by direct cellular contact between T and B cells [5] which primes the B cell to respond to soluble lymphokines. The formation of T/B-cell conjugates with synaptic-like connections explains why, at least in some cases, the IgE response is not blocked by neutralizing antibodies to IL-4 (Stevens et al., Nature 1988, 334:255-257). The second, additional, signal was recently identified as the low-molecular-weight B-cell growth factor (LMWBCGF) [6]. This was disclosed by the observation that one IL-4producing akoreactive TB cell clone fails to induce IgE synthesis during cognate interaction with B cells unless LMW-BCGF is added to the cultures. It is worth noting that this particular T-cell clone does not produce IFNy or IL-2, and that in the presence of LMW-BCGF it induces the synthesis of IgE but not of the other inmunoglobulin classes. However, further addition of IL-2 to the cultures not only increases the synthesis of IgE but also results in the production of IgM and IgG. Stimula tion of the purilied B cells with a cocktail containing optimal concentrations of II-4, IL-2 and LMWBCGF does not induce immunoglobulin synthesis unless cyclosporine Apretreated T cells are added to the B cells. The only function of such T cells is to provide cellular contact with the B cells, because cyclosporine A treatment has completely suppressed their lymphokine production. In most of these experiments using T-cell clones, the B/T-cell interactions are major histocompatibility complex (MIX)restricted. Bystander B cells (expressing MIX class II molecules that are not recognized by the T cells) are not induced to secrete IgE unless they have been preactivated. Such in vivo preactivated B cells have been found in the blood of atopic patients (Umetsu et al, J EqDMed 1985, 162:202-214). MHC restriction is not required in a
similar model using selected human T-cell clones grown in the presence of phytohaemagglutinin (Del Prete et al., Eur J Zmmunoll986, 16:1509-1514). In this case, the T cells are first stimulated with mitogen and then coculhued with B cells from unselected normal donors; direct cellular contacts between T and B cells are also required. Further insight into the mode of action of IL-4 was obtained by dissecting the mechanisms whereby the addition of IL-4 to human peripheral blood mononuclear cells (PBMC) induces the synthesis of IgE. IL-4 triggers cellular interactions between B/T lymphocytes and monocytes and it significantly changes their endogenous production of cytokines. In this model, IL-4 also induces d.rect and MI-E-restricted cellular interactions between T and B cells. Indeed, the IgE response is blocked by antibodies either to the T-cell receptor for antigen or to MHC class II molecules, as well as by antibodies to adhesion molecules such as CD2, CD4 and lymphocyte function-associated antigen @A)-1 [ 51. Monocytes are required for an optimal IgE response and they cannot be replaced by a cocktail of IL-1 and IL-~ (Vercelli et al, Eur J Zmmunol 1989, 19:14191424) nor by the culture supernatant of IL-4-stimulated monocytes. The IL-4-induced synthesis of IgE by PBMC also depends upon a cascade of cytokine interactions triggered by IL-4. This lymphokine increases the endogenous production of IL-~ (Smeland et al, J Exp Med 1989, 170:1463-1468) which was shown to have a necessary role in as much as the IgE response is completely inhibited by anti-IL6 neutralizing antibodies (Vercelli, 1989). Other effects of IL-4 in this system include the inhibition of the production of IL-la, IL-lj3, tumour necrosis factor (TNF)a, prostaglandin (PG)E~ [7] and IFNr [8]. The suppressive effect of IL-4 on the secretion of PGE, and IFNy is relevant because these agents are potent inhibitors of the IgE response. Finally, IL-4 increases the expression of FceEII (CD23) and the release of its soluble fragments, whose function will be discussed below. The addition of IL-5 to PBMC cultures stimulated with suboptimal concentrations of IL-4 signilicantly potentiates the synthesis of IgE (P&e et al, Cell B&hem 1989,39:253-264) but the role of IL-~ in the regulation of IgE synthesis is still unclear. Treatment of parasite-infected mice with neutralizing anti-IL-5 antibody does not alfect the serum IgE level whereas it completely suppresses eosinophilia (Coffman et al., Science 1989, 245:30%311). Using a neutralizing anti-IL-4 monoclonal antibody, it has recently been found that IL-4 only needs to be present during the first 4-6 days of the culture to induce IgE synthesis by either PBMC or by monoclonal surface IgM leukemic B cells co-stimulated with IL-4 and hydrocortisone (Delespesse et al., 1989) [9]. These observations are in agreement with several studies that have indicated that the major role of IL-4 in the regulation of the IgE response is in directing the switching of IgM precursor B cells to IKE. This was unequivocally demonstrated by the observation that single-cell cultures of antigen-specilic B cells can be induced by T cells to generate a large proportion of clones producing IgE and IgGr antibodies [lo]. IL-4 alone is not capable of inducing the expression of
507
508
Atopic allergy
and other
hypersensitivities
IgE by resting IgM precursor B cells and in physiological conditions the other activation signals are provided by the direct interactions between these B cells and T cells. Hence, the addition of IL-4 to purified B cells induces the accumulation of yl but not of E messenger RNA (mRNA) transcripts, unless LPS is added as a co-stimulant [ 1 l] Interestingly, the yl and the E mRNA are first transcribed in the germline conliguration; this was taken to suggest that IL-4 directs the class switching by altering me accessibility of the E and yl switch regions to a common recombinase (Berton et al., Proc Nat1Acud Sci USA 1989,86:2829-2833). Whether the expression of the s gene leading to IgE secretion involves either gene rearrangement (with deletion of the intervening sequences between the assembled VDJ genes and the E gene) or differential RNA processing is still debated. Clearly, terminally differentiated U266 IgE myeloma cells have rearranged their immunoglobulin gene whereas EBV-transformed monoclonal IgEsecreting cell lines have not [ 121. Moreover, in the latter case, the same monoclonal B cells were shown to secrete IgE, IgM and IgD. The monoclonal IgA and IgG B-cell lines generated in the same experiments only secrete one isotype and display the appropriate deletions in their genes. These data, together with the recent Iinding that cloned B cells freshly isolated from the blood of allergic donors also secrete the same three isotypes, led the authors to conclude that: (1) gene rearrangement is not necessary for the secretion of IgE; and (2) the IgE precursor B cells, also named ‘IgE-committed cells’ may belong to a distinct subset of B cells that is capable of co-secreting IgM/IgD and IgE. AS indicated above, in vivo studies have suggested that the activation of 1gE memoty cells is IL-4-independent. This view is now supported by the in vitro observation that purified human B cells may be driven to secrete IgE when co-stimulated with protein-A-bearing Stupkyhxccus aweus and IL-2 [ 131. Paradoxically, the IgE response is inhibited by the addition of IL-4 which inhibits the IL2-supported proliferation of the B-cell blasts.
The role of interferons IFNs can suppress the in vivo synthesis of IgE in mouse and, to some extent, in humans. The polyclonal IgE (and IgGl) response following the injection of goat anti-mouse IgD is inhibited by daily administration of high doses of IFNy. Most importantly, anti-IFNy neutralizing antibody has the opposite effect, indicating that endogenous IFNy down-regulates the IgE response [14]. The same experiments also show that IFNy up-regulates IgG2, production. Similar results were obtained more recently with IFNa, which appeared to be much less toxic (Finkelman, 1990). One pilot clinical study (Souillet et al, Luncet 1989, i:1384) clearly shows that IFNa is most active in reducing the in vivo production of IgE and in clearing the symptoms of a patient with severe atopic dermatitis. Administration of IFNy to patients with the hyperIgE syndrome reduces the serum IgE level in some cases (King et al, Proc Nat1Acud Sci USA,in press). However,
similar treatment of atopic dermatitis patients has no ef5 feet (Bogumiewicz et al., J Allergy Clin Immunol 1989, 83AlOO). In each group of patients, the spontaneous in vitro synthesis of IgE by circulating lymphocytes is inhibited by treatment with IFNy. The mode of action of IFNa has not been analysed and that of IFNy deserves further study. Like the other lymphokines, IFNy has multiple effects on various cell types and, moreover, it might have opposite effects on the same cells at different stages of differentiation or activation, as has been shown for IL-2 (Tigges et al, Science 1989, 243:781-786). The ability of IFNy to block the IgE response of highly purified mouse or human B cells costimulated with LPS or EBV, respectively, clearly indicates that IFNy may act directly on B cells [ 3,4] (Snapper et al, Science 1987, 236944-946). Moreover, the observation that IFNy prevents the IL-4-induced transcription of the germline yl mRNA in mouse B cells suggests that it may also inhibit transcription of the E gene [ 151. Hence, IFNy may act very early in isotype selection as an antagonist of IL-4. This view is in keeping with recent observations of the suppressive effects of IFNy on IL-4-induced IgE synthesis by human PBMC (Pelemann et al, submitted). Hence, IFNy suppresses this response only when added at the same time as IL-4 or shortly thereafter. Moreover, parallel time-course inhibitions are obtained by adding either anti-IL-4 monoclonal antibodies or IFNy to the IL4stimulated cultures. After 4-6 days, when the anti-IL-4 antibody has no more effect, the cells may be washed and recultured in the absence of IL-4. IgE secretion in these secondary cultures is not inhibited by anti-IL-4 or by IFNy. Additional evidence suggests that the selection of the IgE isotype is decided very early by the order in which the IL-4 and IFNy genes are activated at the site of the immune response. As mentioned, IL-4 is capable of suppressing the production of IFNy by human lymphocytes [8], but further analysis indicates that IL-4 prevents the induction of IFNy production, probably by inhibiting the transcription of the IFNy gene (Wu and Delespesse, personal observations) but that it does not suppress the ongoing production of IFNy by activated T cells or Tcell clones [ 161. The addition of IL-~ at the initiation of a mixed lymphocyte culture has two effects: (1) it inhibits the expression of IFNy mRNA and protein; and (2) it induces IgE synthesis. If IL-4 is added after 16 h or more, IFNy is produced and there is no IgE response. These observations lead us to imagine two scenarios regarding the induction of an IgE response at the site of the interactions between antigen-presenting cells and T/B lymphocytes. In one, IL-4 is produced first, resulting in IgE expression and inhibition of IFNy production. In the . other, IFNy is produced first with the following consequences: (1) the B cells are unable to express IgE upon subsequent exposure to IL-4 [the production of which is not inhibited by IFNy (Gajewski et 61, J Immunoll988, 140:4245-4252)]; and (2) IFNy preferentially induces the differentiation of TH lymphocytes into Trill cells (see below). In addition to exerting a direct effect on B cells and selecting the expansion of Tnl cells, IFNy may also act at
IgE synthesis Delespesse,
a much later stage of the IgE response. For example, it has been shown to suppress the SpOntaneOUS in Vitro synthesis of IgE by PBMC cells of atopic donors. As indicated, in this model the production of IgE is due to in vivo pre-activated B cells and is IL-4independent. All the available evidence indicates that IFNy is a potent suppressor of the IgE response in vitro and that it is likely to exert the same effect in vivo under physiological conditions. However, two unexplained observations should be considered to avoid coming to a premature conclusion, First, some IL-4producing T-cell clones are capable of inducing a strong IgE response in spite of their concomitant production of high amounts of IFNy [6]. Set ond, the addition of IFNy to IL-4-stimulated PBMC cell cultures may, occasionally, increase rather than suppress the IgE response. This paradoxical effect of IFNy , which is observed more frequently on neonatal (umbilical cord blood) than on adult PBMC, is at present unexplained (Peleman et al, submitted). Finally, it is worth noting that transforming growth factor p, which selectively increases the production of &A by LPS-stimulated mouse B cells [17], is a very good inhibitor of the in vitro synthesis of human IgE (Delespesse, unpublished observation). It is, therefore, possible that this cytokine also plays a role in the physiological regulation of IgE synthesis.
The role of T lymphocytes As discussed above, the production of IgE appears to depend upon the balance between the production of IL4 and IFNy, which are mainly produced by T cells. In other words, the &type selection is dictated by the na ture of the interactions between antigen-presenting cells and T cells. Long-term mouse T-cell clones can be di vided into two distinct subsets, TH1 and TH2, which dilfer in their patterns of lymphokine production (Mosmann et al, 1986). Trill produce IL-2, IFNy and LT but not IL-4 or IL-5, whereas Tn2 produce IL-4, IL-5, IL-~ but not IL-2 or IFNy; other lymphokines (IL-3 granulocyte/macrophage colony-stimulating factor, etc.) are produced by both subsets. Human Tn cells cannot be divided into TH1 and Tn2 subsets because most of the IL-4-producing T-cell clones also secrete either IFNy or IL-2, or both (Maggi et al., Eur J ZmmunoZ1988, 18:1045-1050). More recent studies [ 16,181 in the mouse have shown that short-term Tn cell clones resemble their human counterparts, in as much as they secrete both T,l and TH2 types of cy tokines. These T cells, named T,O, are thought to be at an Intermediate stage of differentiation: they are derived from the antigen stimulation of a precursor primary T, cell which is capable of producing only IL-2, and upon repeated antigenic stimulation, THO may differentiate into stable TH1 and T,2 cells. The functions of Trill and Tn2 cells are clearly distinct and as in the case of IgE regulation, they are often antagonistic. Moreover, Trill and Tn2 cross-inhibit each other, as shown by the observations that, firstly, IFNy inhibits the proliferation of Tn2 cells (without affecting their lymphokine production) and, set
Sarfati
and Heuser
ondly, that activated Tn2 cells inhibit lymphokine production by Tnl cells via the release of a soluble factor called cytokine synthesis-inhibiting factor (Mosmann et al, FRSEBJ 1989,3~1269). Because Tnl and Tn2 are fully differentiated cells gen erated by repetitive and prolonged antigen stimulations, they cannot be involved in isotype selection during a pnmary response. They probably act as memoty T cells and are involved in the maintenance of ongoing and prolonged immune responses. The hallmark of the IgE an tibody response of atopic patients to airborne allergens is that it is prolonged and that it may be boostered, as shown by the post-seasonal rise in IgE anti-pollen antibody. This probably accounts for the presence in these patients of memory T+like cells, but this possibility remains to be tested. Advances in our understanding of the mechanisms determining the production of IgE antibodies in the allergic patient will come from the analysis of the factors goveming the preferential different&ion of primary Tn cells into either T,2 or T,l lymphocytes. The most informative result in this respect is the recent linding that antigen-specific TH clones, freshly derived in the presence of antigen and IL-2, are mainly of the Tn2 phenotype, whereas Trill cells predominate if the cloning is performed in the presence of IFNy. Hence, the presence of IFNy in the early course of an immune response may result in the preferential expansion of T,l cells [ 161. Other factors that may critically influence the differentiation pathway of either the primary or the T,O cells include: the nature of the antigen-presenting cell and the density of antigen on it, the presence of bystander CD8+ T cells or natural killer cells secreting IFNy and, probably, the microenvironment in which the interaction between T cells and antigen-presenting cells takes place.
The role of Fc&RII (CD231 and its soluble fragments Cellular distribution, structure and regulation of FcsRll expression
In contrast with the high-alfiity receptor for IgE (FcsRI; Kinet, this issue, pp 499-505), FcsRII may be expressed on several haematopoietic cells as well as on some epithelial cells (Delespesse et al, Zmmunol Today 1989, 10:159-164; Billaud et al., J Viral 1989, 63:41214128). Hence, FcsRII may be found on B/T lymphocytes, monocytes, eosinophils, platelets, follicular dendritic cells and epidermal Iangerhans’ cells (Delespesse et al, 1989). On B lymphocytes, FcERII is selectively expressed on surface &M/surface IgDbearing cells and it is lost after isotype switching (Kikutani et LA, J Eqb Med 1986, 164:1455-1569). The structure of FcsRII is well known and has recently been reviewed in detail (Delespesse et al, Cbem Zmmunoll989, 47:78105). Brie& FcERII is a 45kD sialoglycoprotein containing one N-linked carbohydrate chain of the complex type. Human and mouse FcsRII cDNAs have been cloned and functionally expressed (Liidin et al, 1987) [ll. They display significant
509
510
Atopic allergy
and other
hypersensitivities
homology and both have an inverted membrane orientation, with the amino terminal inside and the carboxy terminal outside the cell membrane. The &E-binding domain has been mapped to a region that also displays a striking homology with several animal lectins [19] and with a newly discovered family of cell adhesion molecules (Mary et al, Science 1989, 243:1144). Unexpectedly, the binding of IgE to Fc&RII does not seem to involve carbohydrate interactions, as unglycosylated native or recombinant IgE or &I-derived peptides are capable of binding to FcERII [ 201. In humans, the Fc&RII gene may encode two tierent molecules (named a and b) which have different initiation sites and alternative splicing (Yokota et al., Cell 1988, 55:611-618). FcERII a and b dilfer only by a few amino acids in their cytoplasmic tail; the extracellular and transmembrane domains are identical. Type a is found on freshly isolated and unstimulated normal B cells while type b is expressed on IL-4-stimulated B cells, monocytes, eosinophils and T cells. FcERII is cleaved into soluble fragments, some of which
retain the ability of binding IgE (&E-binding factors).
FcERII is first cleaved into unstable 37 and 33 kD IgE
binding factors that are subsequently transfomred into more stable 25kD molecules (Letellier et al, Mol Zmmunol 1989, 26: 1105-l 112). Recent evidence indicates that this cleavage is an autoproteolytic process and that the 37kD &E-binding factors are endowed with the enzyme activity (Letellier et al., submitted). The rate of cieavage of FcsRII is reduced by IgE and increased by treatment of the cells with inhibitors of N-linked glycosylation. The expression of FcsRII and the concomittant production of IgEbinding factors are tightly regulated. They are induced by IL-4 and suppressed by IFNy; IFNy may in hibit FcERII expression by B cells but not by monocytes, platelets and Iangerhans’ cells (Delespesse et al., 1989) [21]. FcERII b is overexpressed in patients with increased production of IgE, for example those with severe atopy or some parasitic diseases (Watanabe et al., Appl Zmmunoll989, 89:103-107). Abnormally high expression of this molecule is also found in patients with B-cell chronic lymphocytic leukemia, a condition associated with hypogammaglobulinemia and very low serum IgE levels (Sarfati et al, Blood 1988, 71:94-101). On B cells, FcERII may also be up-regulated by direct interaction with T cells (in the absence of IL-4) and by various agents including leukotriene B4 and IL-2 (Yamaoka et al, J Zmmunoll989,143:1996-2000; Crow et al, J Eap Med
1986, 164:1760-1772). FcERII is a multifunctional
molecule
Membrane-bound FcERII may exert different functions according to the cell types on which it is expressed. On eosinophils, platelets and monocytes, Fc&RII b mediates &E-dependent cellular cytotoxicity against parasites as well as &E-dependent release of inllammatory mediators (Capron et al., Zmmunol Today 1986,7:15-21). On B cells, FcERII focuses the &E-dependent presentation of antigen to T cells [22]. In addition, FcERII may also play the role of a cell adhesion molecule (Kishimoto, per-
Sonal observation), probably by virtue of its lectin activity, which remains to be demonstrated. Soluble FcERU fragments are reported to have several activities in vitro including: (1) regulation of the synthesis of human IgE; (2) promotion of B-cell and Tcell proliferation (Swendeman et al, EMBO J 1987, 6~1637-1642; An&age et al., Eur J Zmmunol 1989, 10:31-35); and (3) most interestingly, induction, in synergy with IL-I, of the maturation of early thymocytes (CD2-, CD3_CD4/CD8-) unresponsive to IL-2, into CD2+, CD3 + , CD4/8+ cells that are responsive to IL-2 (Mossalayi et al, J Eq Med 1990, in press).
Role of FcsRll/lgE-binding
factors
in IgE regulation
role of these molecules has been documented in four models: (1) the stimulation of unfractionated PBMC with IL-4; (2) the established, ongoing synthesis of IgE by IL-4-prestimulated normal PBMC; (3) the spontaneous synthesis of IgE by PBMC of atopic patients, and (4), more recently, IgE synthesis induced by the co-stimulation of monoclonal chronic lymphocytic leukemia B cells (B-CLLs) with IL-4 and hydrocortisone [9]. In each case, the IgE synthesis can be markedly suppressed by some monoclonal antibodies to FcERII (FILMIER), cross-reacting with &E-binding factors. The suppression is dose-dependent and, in the case of B-CLLs, complete inhibition is observed with as little as 0.1 clg/ml of tdb~R. The effect of mAbsR or of its F(ab’)2 fragments is isotype-specific (Sarfati et al, J Zmmunol1988, 141:2195-219, Pene et al, Proc Nat1 Acad Sci USA 1988, 856880-6884). Two observations indicate that IIIA~ER do not suppress IgE synthesis by delivering a negative signal to &E-secreting cells. First, these cells have lost their FcERII (Kikutani et al., J Exp Med 1986,164:1455-1469). Second, not all the mAb&Rs are equally active and only those antibodies preventing the cleavage of FcERII are capable of suppressing IgE synthesis. The
Taken together, the inhibition studies using mAbsR suggest that the soluble, rather than the membrane-bound, FcsRII up-regulates the synthesis of IgE. This view is supported by the observations that &E-binding factors augment the production of 1gE in the above models. The ability of these molecules signiftcantly to enhance ongoing IgE synthesis by unstimulated lymphocytes from atopic donors was described several years ago in work which used afhnity-purilied &E-binding factors isolated from the culture supernatant of FcsRIIbearing B-cell lines (P&e et al., Eur J Zmmunol 1988, 18:929-935). These results have now been conIirmed (P&e et al., 1989) and extended using recombinant IgEbinding factors. These molecules signilicantly increase the ongoing and IL-4-independent synthesis of IgE by either in viva or in vitro preactivated B cells. Native or recombinant IgE binding factors may also augment the 1gE response of PBMC stimulated with either suboptimal concentrations of IL-4 (P&e et al, Eur J Zmmunoll988, 18:92+935) or after partial depletion of monocytes. Further studies have clearly indicated that the IgE-potentiating activity is mainly mediated by the unstable 3337kD &E-binding factors. This was shown by coculturing target lympho-
IgE synthesis
qtes with various Chinese hamster ovaly (CHO) transfectants expressing either the complete FcsRII molecule (CHO l-7), a secreted form of 37 kD IgE-binding factors (CHO-37) or a secreted form of 25kD &E-binding factors (CHO-25). A pronounced increase in IgE synthesis is only observed with CHO 1-7 and CHO-37 cells; under these conditions, where the unstable 37kD binding factor is constantly produced, a 2-20-fold enhancement of the IgE response is obtained with very low concentrations of &E-binding factors, i.e. 0.1 nmol/litre (Delespesse et al, submitted). It may therefore be concluded that IgEbinding factors (soluble CD23) potentiate the synthesis of IgE by in& en&g the activity of the IL-4-independent &E-activated B cells. Current studies are aimed at determining whether this effect is direct or indirect, via T cells or monocytes.
Acknowledgements Dr G. Delespesse is an MRC associate; Dr M. Sarfati is the recipient of an MRC scholarship; this work is supported by the MRC of Canada. The authors gratefully acknowledge Drs F. Fielman and T. Nutman for providing access to unpublished manuscripts. The excellent secretarial assistance of Mrs Norma Del Bosco is most appreciated.
Annotated
references
and recommended
reading
Delespesse,
Sarfati
and Heusser
postrongylus brasiliensis or by injection of goat anti-mouse I’D antibody, and (2) the secondaty IgE anti-dinltrophenol response, Anti-IL-4 has no effect on the IgG, responses elicited under the same conditions. When given at the peak of the secondary IgE anti-dinitrophenol response or of the N. brasiliensis infection, anti-IL-4 accelerates the decline in total and specific anti-dinitrophenol IgE, respectively, but the suppression is not complete. This paper unequivocally demonstrates that IL-4 is required for an in vivo primary IgE response and that it also plays a major role in the secondary and ongoing Igl? response. 4. 0
THYF?RONITISG, T~OKO~GC, JUNE CH, IEV~NEAD, EINKELMAN FD: I@ secretion by Epstein-Barr virus-infected purified human B lymphocytes is stimulated by interleukin 4 and suppresed by interferon y. Pm Nat1 Acad Sci USA 1989, 86:558@5584. Co-stimulation of highly purified human B cells with EBV and IL-4 induces a signiIicant production of IgE after H weeks of CUlNre. Both stimulants are required to induce IgE synthesis; in contrast, EBV alone induces the secretion of other immunoglobulin isotypes. IFNr at 10 U/ml specifically suppresses the IgE response. This Study demonstrates that: (1) as in the mouse, highly purihed human B cells may be induced to synthesize IgE, provided a second activation signal is given, and (2) IFNy may suppress the IgE response by a direct effect on human B cells. VERCELLI D, JABARAHU, ARAIKI, GEHAR Induction of human IgE synthesis requires interleukin 4 and T/B cell interactions involving the T cell receptor/CD3 complex and MHC class II antigens. J &?I$ Med 1989, 169:129>1307. The induction of human IgE synthesis by IL-4-stimulated PBMC requires cognate interaction betwen autologous T/B cells; it is blocked by antibodies against the T-cell receptor /CD23 complex or against MHC class II antigens on B cells. Antibodies to adhesion molecules (CD2, CD4, LFA-1) also Inhibit the IgE response, indicating that cell-cell contact is required. By means of selected alloreactive T-cell clones it is also shown that bystander B cells (whose MHC class II molecules are not recognized by T cells) are not induced to synthesize IgE. 5. 0
6.
DE KRLIYFF RH, TURNERT, ABRAMS JS, PAUADINO JR MA, U~TSU
??e
0 00
Of interest Of outstanding Interest
1. a
BETIXER B, HOFETETIXRM, RASM, YLIKOYOUKA WM, KILCHHERR F, CONRADDH: Molecular structure and expression of the murine lymphocyte low-atTin@ receptor for IgE (FceRII). Pm Nat1 Acad Sci USA 1989, 86~7567570. The cDNA coding for mouse Fc&RII was cloned and functionally expressed in COS cells. The predicted amino acid sequence demonstrates: (1) 57% identity to human FcsRII; (2) the presence of an additional repetitive sequence of 21 amino acids giving 4 repeats compared with 3 in the human sequence; and (3) the carboxy terminus is truncated and does not contain the Asp-GlyArg sequence that is expressed in human FcERI~and that may be involved in the binding to integrin. 2.
FINKELMAN FD, HOIMESJ, URBANJR JF, PAULWE, KATONAIM: T help requirements for the generation of an in viva IgE response: a late acting form of T cell help other than IL-4 is required for IgE but not for IgGl production. J Immunol 1989, 142:403-$08. The Injection of mice with goat antibody to mouse IgD induces a pronounced polyclonal IgE and IgGl response. Both responses are blocked by anti-CD4 antibody whereas anti-IL-4 only blocks the IgE rep spouse. When the injection of anti-CD4 or anti-IL-4 antibodies is delayed to day 5 after the Injection of goat anti-mouse IgD, anti-IL-4 has little effect on the IgE response whereas anti-CD4 strongly inhibits the IgE response and slightly reduces IgGl production. This paper demonstrates that: (1) the mechanisms regulating the in uivo production of IgGl and IgE are different; and (2) a late-acting form of T-cell help, other than K-4, is required for IgE but not for IgGl production. ??.
3. a.
FINKEU~AN ED, KATONA IM, URBANJF, HOLMESJ, OHARA J, TUNG AS, SAMPIXJV, PAULWE: IL-4 Is required to generate and sustain in vivo IgF, responses. / Immunoll988, 141:23352341. The injection of anti-IL-4 neutralizing antibody to mice suppresses: (1) the polyclonal IgE responses induced either by infection with N@
DT: Induction of human IgE synthesis by CD4+ T ceU clones-requirement for interleukin 4 and low molecular weight B cell growth factor. J @ Med 1989, 170:1477-1493. WeU characterized alloreactive human CD4+ T-cell clones were tested for their ability to induce IgE synthesis under cognate interaction with resting B cells. One clone co-producing substantial amounts of IFNy and II-4 is a good Inducer of IgE; another clone producing only IL4 (and not IL-2 or IFNr) fails to induce IgE unless the CUlNreS are supplemented with the LMWBCGF. Under these conditions, only IgE is induced; the production of the other isotypes requires the addition of K-2. This paper demonstrates that 3 sets of signals are required for the induction of IgE synthesis by alloreactive T-cell clones: (1) those derived from the direct T/F-cell contact; (2) LMWBCGF; and (3) IL-4. HART PH, Vrrn GF, BURGESS DR, WHITIY GA, PICCOLIDS, anti-inflammatory effects of interleuldn 4: Suppression of human monocyte tumor necrosis factor, interleukin 1, and prostaglandin E> Proc Natl Acud Sci USA 1989, 86:380>3807. The production of IL-I, TNFa and PGE, by human monocytes stimulated with LPS or with LPS and IFNy is inhibited by small concentrations of IL-4 (0.1-0.5 U/ml). II-4 has the same effect as 0.1 nmol/litre dexamethasone. 7. 0
HAMIUON JA: Potential
PELEMAN R, Wu J, EARGENEC, DELESPESSE G: Recombinant lnterleukin 4 suppresses the production of interferon y by human mononuclear cells. J k$!) Med 1989, 170:1751-1757. IL-4 inhibits the production of IFNy by human PBMC stimulated with mitogen or allogeneic cells. The suppression is dose and time-dependent; this is observed at the mRNA and at the protein level. The data further suggest that the balance between the production of IL-4 and IFNy, governing the synthesis of IgE, may be signiscantly iniluenced by the chronological order of activation of the IL-4/IFNy genes.
8. 0
SAWATIM, Luo H, DEIESPESSEG: IgE synthesis by chronic 9. 0 lymphocytic leukemia cells. J Exp Med 1989, 170:17751781. Unfractionated lymphocytes from patients with B-cell chronic lymphocytic leukemia are induced to secrete monoclonal IgE upon co-stimula
511
512
Atopic allergy
and other
hypemensitivities
tion with IL-4 and hydrocortisone. Hydrocottisone is required to induce of normal lymphocytes it IgE synthesis by B-CLIs, whereas in CdNre increases the IgE response to IL-4. 10. ??e
COFFMAN RL Interleuldn 4 causes isotype~ switching to IgE in T cell-stimulated clonal B ceU cultures. J E~J Med 1988, 168853-862. Clonal CdNres of B cells stimulated with a Tu2 ceU line speciIic for rabbit IgG were established. In this system, the majority of B cells undergo clonal expansion in the presence of rabbit anti-mouse IgM The addition of IL-4 or of anti-IL-4 indicates that: (1) II-4 increases the proportion of IgE-secreting clones, has no effect on the frequency of IgGrproducing clones and reduces the proportion of @-producing clones; and (2) IL-4 increases the quantity of IgE produced by single clones. This study demonstrates that IL-4 promotes IgE synthesis by directing the isotype switching to IgE and not by selectively supporting the mats uration of precommltted precursor cells. LEBMAN DA,
11.
ROTHMAN P, LUTZKER S, COOKW, COFFMAN R, ALT W Mitogen plus lnterleukln 4 induction of CE transcripts in B lymphoid cells. J ELI Med 1988, 168:23852389. Murine splenic B cells and Abelson murine leukemia virus-transfomred cells were cultured with LPS and IL-4 and assayed for E and yl transcripts. Germline yl and E transcripts appear first, and are followed by the accumulation of normal E mRNA. The resuks suggest that IL-4 can direct class switching by increasing recombinase accessibility to yl and E switch regions. ??
12.
T, DCKH HM: Clonal and molecular characteristics of the human &E-committed B cell subset. J Exp Med 1989, 169:407-430. Monoclonal EBV~transformed B-cell lines were obtained by ceU soning followed by limiting dilution. It was observed that: (1) individual IgEsecreting cells co-secrete IgM and IgD; (2) these IgEsecreting cells have not rearranged their lmmunoglobulin gene; (3) u and E mRNA are of normal and comparable size; and (4) U266 IgE myeloma cells have rearranged their immunoglobulin gene. This study demonstrates that rearrangement of the E gene (with deletion of the intervening sequences between assembled VDJ and E) is not always required for IgE production. It is postulated that B cells capable of secreting IgE belong to a unique B-cell sub-lineage termed ‘IgEcommiued cells’. MACKENZIE
??
SPIAWSKIJR, JELINCKDF, LIPSKYPF: Immunomodulatory role of IL-4 on the secretion of Ig by human B cells. J Immunol 1989, 142:15691575. Highly purified human B-cell blasts induced by stimulation with S. aureus are capable of producing immunoglobulins of all isotypes (including IgE) when cultured in the presence of IL-2 but not IL-4. Addition of II-4 to the IL-2-containing secondaty CUlNreS inhibits immunoglobulin production (including IgE) and cellular proliferation. The IL-4-mediated suppression is partially reversed by IFNy and LMWBCGF. Additional data document the complex immunoregulatory action of IL-4 on human B cells. 13. ??
TR, COFFMAN RI: IFNy FINKELMAN ID, KATONA IM, MOSMANN regulates the isotypes of Tu secreted during in uivo humoral Immune responses. J Immunol 1988, 140:1022-1027. Multiple injections of IFNy signiIicantly inhibit the IgE and IgGt response of mice to goat anti-mouse IgGD. Anti-IFNy neutralizing antibody has the reverse effect, The data also show that IgGaa production is enhanced by IFNy. Immunization of mice with killed and tixed Brucelkz abortus, known to induce IFNY production, increases the production of IgG2a and suppresses that of IgE and IgGl. Injection of B. &or&s-immunized animals with anti-IFNy antibody inhibits the increase of IgG2a and enhances the production of IgGt but does not reverse the suppresslve effect of B. ubortus on IgE production. Hence, the B. &or&&n duced IgE suppression is not mediated by IFNy. 14. ??
BERTONMT, UHRJW, VITET~AES: Synthesis of germ-line yl ImmunoglobuRn heavy-chain transcripts In resting B cells: Induction by Interleukin 4 and inhibition by Interferon. Pnx Nat1 Acad Sci USA 86:28292833. By means of an RNase protection assay, it was observed that IL-4 induces expression of germline yl transcripts in B cells after 4 h of cul15. ??
Nre; the induction is inhibited by IFNy The accumulation of yl mRNA * may be induced in resting B cells stimulated with IL-4 alone but it is increased by co-stimulation with LPS. This study strongly suggests that IL-4 directs the class switching to IgG, by opening the Syl switch region and making it accessible to switch recombinase. 16. ee
GAJEXVXITF, SCHE~LSR, NAU G, F~CH FW: Regulation of TceU activation: Differences among T-cell subsets. Zmmunol Ra, 1989, 79111. This is a timely review of the mechanisms governing the differentiation of cells into T,l or TH2 ceU types. The roles of IFNy and antigen-presenting cells are emphasized as weU as the different requirements and signalling mechanisms for the antigen-induced activation of the subsets. 17.
COFFMAN R, &BMAN DA, SARADERB: Transforming growth fattor fi specifically enhances IgA production by lipopolysaccharlde-stimulated murine B lymphocytes. J Exp Med 1989, 170:10391044. The addition of transforming growth factor g to purified mouse B cells stimulated with LPS induces pronounced production of Ig.4 and suppresses the other isotypes. The IgA response is further increased by IL-2 and IL-5. Transforming growth factor j3 induces a substantial increase in IgA production by surface IgA cells whereas it suppresses IgA production by B surface IgA+ cells. This paper strongly suggests that transforming growth factor j3 acts as an isotype-specific switch factor for lg.& ??
18.
MOSMANN TR, COFFMANRL Heterogeneity of cytoldne secretion patterns and functions of helper T cells. Adv Immunol 1989, 46:111-147. This is a most up-to-date review of the author’s work on TH1 and TH2 cells with regard to their phenotypic and functional characteristics, their role in the immune response, and their cross-inhibition. New concepts regarding the diUerentiatlon of precursor TH cells into Tul and TH2 types are presented. ??
site for IgE of the human FCE receptor II (FcsRII/CD23) is cormned to the domain homologous with animal lectins. Proc Nat1 Acud Sci USA 1989, 86:7118-7123. FcERII mutants were expressed in mammalian ceUs with a view to mapping the IgEbindlng site of the receptor. This is localized to a domain of 123 amino acid residues which is also homologous to the carbohydratebinding domain of animal lectins. The epitopes of several monoclonal antibodies to CD23, known to inhibit the binding of IgE to FcERII, are clustered in the same region. This is also the case with the MHM6 monoclonal antibody, which is known to mimic the action of LMWBCGF. 19.
BE’ITLER B, MAIEXRR, RUEGG D, HOFSTETIER H: Binding
??
20.
VERCELLI D, HELM B, MARSH P, PADIAN E, GEHA RS, GOULD
H: The B-cell binding site on human lmmunoglobulin E. Nature 1989, 338649651. By means of recombinant E-chain fragments synthesized in Escherichia coli, the CD23-binding site of IgE has been mapped to a sequence of the ~3 domain. Glycosylation of IgD is not required for binding to FcERII 0 or FcERI. ??
21. ??
BIEBER T, RIEGERA, NEUCHK~T C, PRINZJC, RICHER EP, BOLT& Nwnscu G, %IEINJ?R 0, KRAFT D, RING J, STINGLG: Induc-
tion of FcsR2/CD23 on human epidermal Langerhans cells by human recombinant interleukin 4 and interferon. J EQ Med 1989, 170:309-314. Langerhans’ cells, freshly isolated from the skin of normal donors, do not express CD23/FcaRU These molecules are induced after 24 h incubation in the presence of either IL-4 or IFNy; neither phorbol-12. my&ate-13.actetate (PMA) nor IL-2 have any effect; a synergistic effect is observed between IL-4 and IFNy. 22. ??e
KHERYMR, YAMASHITA LC: Low-atlkity IgE receptor (CD23) function on mouse B cells: role in &E-dependent antigen focusing. Proc Nat1 Acad Sci 1989, 86:75567560. ti Mouse B cells treated with IgE monoclonal antibodies to trinitrophenol are 100.fold more effective than untreated B cells in presenting low concentrations of ttinitrophenol antigen to T cells. This is blocked by This paper preventing IgE binding to the B cells by means of IIIA~ER documents one function of FcERII on B cells, i.e. the &E-dependent internalization of antigen resulting in efficient presentation to T cells.
of Montreal,
Notre-Dame
Canada and *CIBA-GEIGY, Current
M. Sarfati
Opinion
and C. Heusser*
Hospital,
Immunology in Immunology
Introduction The lirst demonstration that the production of IKE antibody is T-cell dependent and that its regulation is isotypespecilic was reported in the early 1970s (Okumura et al, JZmmuno11971,106:101~1025). Tada and co-workers have shown that neonatally thymectomized rats are incapable of developing an IgE antibody response, but that after adult thymectomy the IgE but not the 1gG antibody response is increased and sustained for longer. The IgE-potentiating effect of adult thymectomy is abolished by reconstitution with syngeneic T lymphocytes. Ishizaka (Annu Rev Zmmunol 1988, 6:513-535) has extensively documented one of the mechanisms by which rodent T cells mediate the isotype-specific but not antigen-spe cilic regulation of IgE synthesis. Because this work has recently been raiewed in detail (Ishizaka, 1988), it will only be summarized very brieily here. According to this model, a subset of T helper (TH) cells [capable of expressing the low-affinity receptor for IgE, the FCE receptor (FGR)II] may release &E-binding factors that may either suppress or increase the secretion of IgE by activated I@-bearing B cells. The effect of these factors is isotype-specific and their suppressive or potentiating activity is dependent upon their N-linked glycosylation which, in turn, is regu lated by T cells. A complementary DNA (cDNA) coding for rodent IgE-binding factor has been cloned and functionally expressed (Martens et al, Proc NatlAcud Sci USA 1985,82:2460-2464); the predicted amino-acid sequence has no homology with that of the human (Liidin et al, EMBOJ1987,6:10!+116) or the mouse [l] Fc&RII,which is also known as CD23. The rodent IgE-binding factors are, therefore, not related to the soluble fragments of human Fc&RII, also called &E-binding factors because they bind to human IKE. In the human (but not in the mouse) model, Fc&RII and its soluble fragments have been shown to regulate IgE synthesis. The availability of several recombinant lymphokines and of the corresponding neutralizing antibodies has permitted a major advance in our understanding of the isotypespecific regulation of the immune response. It is now
Research Center, Montreal,
Division,
Basel, Switzerland
1990, 2:506-512
clear that selection of IgE isotype is dictated by the preferential production of the T-cell-derived lymphokine interleukin (IL)-4, whereas it is suppressed by interferon (IFN)-y. In the mouse, T, lymphocytes may differentiate in two distinct subsets with diiferent lymphokine production profiles: Tul cells secrete IL-2, IFNy and lymphotoxin (LT), but not IL-4 or IL-5, and TH2 cells produce IL-4 and IL-5, but not IL-2, IFNy or LT (Mosmann et al, J Zmmunol 1986, 136:234%2357). Hence, the development of an IgE response appears to depend upon the preferential differentiation of T, cells into Tn2 lymphocytes. Here, I will focus on the role of lymphokines, T cells and &E-binding factors (soluble CD23) in the regulation of IgE synthesis.
The role of interleukin-4 In vivo studies in the mouse, using neutmlizing antiIL-4 or anti-IL-4-receptor antibodies, have unequivocally demonstrated that the primary, and most of the secondary, antigen-specific or polyclonal IgE (but not IgG) antibody responses depend upon the production of IL-4 [L&3]. However, another form of T-cell help, which acts at a later stage than IL-4 and which can be inhibited by anti-CD4 (but not by anti-IL-4), is also required for the induction of an IgE response. Most of the secondary and of the established ongoing IgE response is suppressed by the administration of saturating doses of anti-IL-4. The small part of the secondary response that is resistant to anti-IL-4 treatment is thought to be mediated by IgE memory B cells whose activation should be IL-4independent. Another conclusion from these in vivo studies is that the role of IL-4 in the persistent IgE response of parasite-infected animals is to recruit new IgE B cells from their IgM precursors rather than to promote the activity or to expand the pool of &E-secreting cells (Finkelman et al, Annu Rev Zmmunol 1990, in press). This view is consistent with recent observations that the spontaneous in
Abbreviations B-Cll~hronic lymphocytic leukemia B cells; cDNA-complementary DNA; CH
506
@
Current
Science
Ltd ISSN
0952-7915
IgE synthesis Delespesse, Sarfati and Heusser
vitro synthesis of IgE by lymphocytes from atopic patients is IL-4-independent, in as much as it is not inhibited by a neutralizing anti-IL-4 antibody (Delespesse et al., 7th Int Cong Zmmunol1989, A75-A77). In contrast to anti-IL4, anti-CD4 injected at the peak of a secondary response has no effect on the decline in serum IgE antibody levels. This suggests that in the course of an ongoing IgE response, II-4 may be produced by a CD4- cell; mast cells might have such a role because they may release IL-~ after &E-dependent stimulation (Plaut et al, Nature 1989, 3396467). The mode of action of IL-4 at the cellular and molecular levels has been analysed in vitro. The first conclusion of these studies is that IL-4 is necessary but not sufficient to transform resting IgM precursor B cells into IgE-producing cells. Hence, the addition of IL-4 to cultures of punlied mouse or human B cells does not induce IgE syn thesis unless additional signals are provided, for example lipopolysaccharide (LPS) to mouse B cells (Coffman et al., J Immunoll986,I36:4538-454I) or Epstein-Barr &US (EBV) to human B cells [4]. Several groups have demonstrated that normal resting B cells may be induced to synthesize IKE after cognate irteraction with selected helper T-cell clones (P&e et al, JZmmunol19@3,141:121~1224). In this model, two signals are required in addition to IL-4 to obtain an IgE response. The first signal is provided by direct cellular contact between T and B cells [5] which primes the B cell to respond to soluble lymphokines. The formation of T/B-cell conjugates with synaptic-like connections explains why, at least in some cases, the IgE response is not blocked by neutralizing antibodies to IL-4 (Stevens et al., Nature 1988, 334:255-257). The second, additional, signal was recently identified as the low-molecular-weight B-cell growth factor (LMWBCGF) [6]. This was disclosed by the observation that one IL-4producing akoreactive TB cell clone fails to induce IgE synthesis during cognate interaction with B cells unless LMW-BCGF is added to the cultures. It is worth noting that this particular T-cell clone does not produce IFNy or IL-2, and that in the presence of LMW-BCGF it induces the synthesis of IgE but not of the other inmunoglobulin classes. However, further addition of IL-2 to the cultures not only increases the synthesis of IgE but also results in the production of IgM and IgG. Stimula tion of the purilied B cells with a cocktail containing optimal concentrations of II-4, IL-2 and LMWBCGF does not induce immunoglobulin synthesis unless cyclosporine Apretreated T cells are added to the B cells. The only function of such T cells is to provide cellular contact with the B cells, because cyclosporine A treatment has completely suppressed their lymphokine production. In most of these experiments using T-cell clones, the B/T-cell interactions are major histocompatibility complex (MIX)restricted. Bystander B cells (expressing MIX class II molecules that are not recognized by the T cells) are not induced to secrete IgE unless they have been preactivated. Such in vivo preactivated B cells have been found in the blood of atopic patients (Umetsu et al, J EqDMed 1985, 162:202-214). MHC restriction is not required in a
similar model using selected human T-cell clones grown in the presence of phytohaemagglutinin (Del Prete et al., Eur J Zmmunoll986, 16:1509-1514). In this case, the T cells are first stimulated with mitogen and then coculhued with B cells from unselected normal donors; direct cellular contacts between T and B cells are also required. Further insight into the mode of action of IL-4 was obtained by dissecting the mechanisms whereby the addition of IL-4 to human peripheral blood mononuclear cells (PBMC) induces the synthesis of IgE. IL-4 triggers cellular interactions between B/T lymphocytes and monocytes and it significantly changes their endogenous production of cytokines. In this model, IL-4 also induces d.rect and MI-E-restricted cellular interactions between T and B cells. Indeed, the IgE response is blocked by antibodies either to the T-cell receptor for antigen or to MHC class II molecules, as well as by antibodies to adhesion molecules such as CD2, CD4 and lymphocyte function-associated antigen @A)-1 [ 51. Monocytes are required for an optimal IgE response and they cannot be replaced by a cocktail of IL-1 and IL-~ (Vercelli et al, Eur J Zmmunol 1989, 19:14191424) nor by the culture supernatant of IL-4-stimulated monocytes. The IL-4-induced synthesis of IgE by PBMC also depends upon a cascade of cytokine interactions triggered by IL-4. This lymphokine increases the endogenous production of IL-~ (Smeland et al, J Exp Med 1989, 170:1463-1468) which was shown to have a necessary role in as much as the IgE response is completely inhibited by anti-IL6 neutralizing antibodies (Vercelli, 1989). Other effects of IL-4 in this system include the inhibition of the production of IL-la, IL-lj3, tumour necrosis factor (TNF)a, prostaglandin (PG)E~ [7] and IFNr [8]. The suppressive effect of IL-4 on the secretion of PGE, and IFNy is relevant because these agents are potent inhibitors of the IgE response. Finally, IL-4 increases the expression of FceEII (CD23) and the release of its soluble fragments, whose function will be discussed below. The addition of IL-5 to PBMC cultures stimulated with suboptimal concentrations of IL-4 signilicantly potentiates the synthesis of IgE (P&e et al, Cell B&hem 1989,39:253-264) but the role of IL-~ in the regulation of IgE synthesis is still unclear. Treatment of parasite-infected mice with neutralizing anti-IL-5 antibody does not alfect the serum IgE level whereas it completely suppresses eosinophilia (Coffman et al., Science 1989, 245:30%311). Using a neutralizing anti-IL-4 monoclonal antibody, it has recently been found that IL-4 only needs to be present during the first 4-6 days of the culture to induce IgE synthesis by either PBMC or by monoclonal surface IgM leukemic B cells co-stimulated with IL-4 and hydrocortisone (Delespesse et al., 1989) [9]. These observations are in agreement with several studies that have indicated that the major role of IL-4 in the regulation of the IgE response is in directing the switching of IgM precursor B cells to IKE. This was unequivocally demonstrated by the observation that single-cell cultures of antigen-specilic B cells can be induced by T cells to generate a large proportion of clones producing IgE and IgGr antibodies [lo]. IL-4 alone is not capable of inducing the expression of
507
508
Atopic allergy
and other
hypersensitivities
IgE by resting IgM precursor B cells and in physiological conditions the other activation signals are provided by the direct interactions between these B cells and T cells. Hence, the addition of IL-4 to purified B cells induces the accumulation of yl but not of E messenger RNA (mRNA) transcripts, unless LPS is added as a co-stimulant [ 1 l] Interestingly, the yl and the E mRNA are first transcribed in the germline conliguration; this was taken to suggest that IL-4 directs the class switching by altering me accessibility of the E and yl switch regions to a common recombinase (Berton et al., Proc Nat1Acud Sci USA 1989,86:2829-2833). Whether the expression of the s gene leading to IgE secretion involves either gene rearrangement (with deletion of the intervening sequences between the assembled VDJ genes and the E gene) or differential RNA processing is still debated. Clearly, terminally differentiated U266 IgE myeloma cells have rearranged their immunoglobulin gene whereas EBV-transformed monoclonal IgEsecreting cell lines have not [ 121. Moreover, in the latter case, the same monoclonal B cells were shown to secrete IgE, IgM and IgD. The monoclonal IgA and IgG B-cell lines generated in the same experiments only secrete one isotype and display the appropriate deletions in their genes. These data, together with the recent Iinding that cloned B cells freshly isolated from the blood of allergic donors also secrete the same three isotypes, led the authors to conclude that: (1) gene rearrangement is not necessary for the secretion of IgE; and (2) the IgE precursor B cells, also named ‘IgE-committed cells’ may belong to a distinct subset of B cells that is capable of co-secreting IgM/IgD and IgE. AS indicated above, in vivo studies have suggested that the activation of 1gE memoty cells is IL-4-independent. This view is now supported by the in vitro observation that purified human B cells may be driven to secrete IgE when co-stimulated with protein-A-bearing Stupkyhxccus aweus and IL-2 [ 131. Paradoxically, the IgE response is inhibited by the addition of IL-4 which inhibits the IL2-supported proliferation of the B-cell blasts.
The role of interferons IFNs can suppress the in vivo synthesis of IgE in mouse and, to some extent, in humans. The polyclonal IgE (and IgGl) response following the injection of goat anti-mouse IgD is inhibited by daily administration of high doses of IFNy. Most importantly, anti-IFNy neutralizing antibody has the opposite effect, indicating that endogenous IFNy down-regulates the IgE response [14]. The same experiments also show that IFNy up-regulates IgG2, production. Similar results were obtained more recently with IFNa, which appeared to be much less toxic (Finkelman, 1990). One pilot clinical study (Souillet et al, Luncet 1989, i:1384) clearly shows that IFNa is most active in reducing the in vivo production of IgE and in clearing the symptoms of a patient with severe atopic dermatitis. Administration of IFNy to patients with the hyperIgE syndrome reduces the serum IgE level in some cases (King et al, Proc Nat1Acud Sci USA,in press). However,
similar treatment of atopic dermatitis patients has no ef5 feet (Bogumiewicz et al., J Allergy Clin Immunol 1989, 83AlOO). In each group of patients, the spontaneous in vitro synthesis of IgE by circulating lymphocytes is inhibited by treatment with IFNy. The mode of action of IFNa has not been analysed and that of IFNy deserves further study. Like the other lymphokines, IFNy has multiple effects on various cell types and, moreover, it might have opposite effects on the same cells at different stages of differentiation or activation, as has been shown for IL-2 (Tigges et al, Science 1989, 243:781-786). The ability of IFNy to block the IgE response of highly purified mouse or human B cells costimulated with LPS or EBV, respectively, clearly indicates that IFNy may act directly on B cells [ 3,4] (Snapper et al, Science 1987, 236944-946). Moreover, the observation that IFNy prevents the IL-4-induced transcription of the germline yl mRNA in mouse B cells suggests that it may also inhibit transcription of the E gene [ 151. Hence, IFNy may act very early in isotype selection as an antagonist of IL-4. This view is in keeping with recent observations of the suppressive effects of IFNy on IL-4-induced IgE synthesis by human PBMC (Pelemann et al, submitted). Hence, IFNy suppresses this response only when added at the same time as IL-4 or shortly thereafter. Moreover, parallel time-course inhibitions are obtained by adding either anti-IL-4 monoclonal antibodies or IFNy to the IL4stimulated cultures. After 4-6 days, when the anti-IL-4 antibody has no more effect, the cells may be washed and recultured in the absence of IL-4. IgE secretion in these secondary cultures is not inhibited by anti-IL-4 or by IFNy. Additional evidence suggests that the selection of the IgE isotype is decided very early by the order in which the IL-4 and IFNy genes are activated at the site of the immune response. As mentioned, IL-4 is capable of suppressing the production of IFNy by human lymphocytes [8], but further analysis indicates that IL-4 prevents the induction of IFNy production, probably by inhibiting the transcription of the IFNy gene (Wu and Delespesse, personal observations) but that it does not suppress the ongoing production of IFNy by activated T cells or Tcell clones [ 161. The addition of IL-~ at the initiation of a mixed lymphocyte culture has two effects: (1) it inhibits the expression of IFNy mRNA and protein; and (2) it induces IgE synthesis. If IL-4 is added after 16 h or more, IFNy is produced and there is no IgE response. These observations lead us to imagine two scenarios regarding the induction of an IgE response at the site of the interactions between antigen-presenting cells and T/B lymphocytes. In one, IL-4 is produced first, resulting in IgE expression and inhibition of IFNy production. In the . other, IFNy is produced first with the following consequences: (1) the B cells are unable to express IgE upon subsequent exposure to IL-4 [the production of which is not inhibited by IFNy (Gajewski et 61, J Immunoll988, 140:4245-4252)]; and (2) IFNy preferentially induces the differentiation of TH lymphocytes into Trill cells (see below). In addition to exerting a direct effect on B cells and selecting the expansion of Tnl cells, IFNy may also act at
IgE synthesis Delespesse,
a much later stage of the IgE response. For example, it has been shown to suppress the SpOntaneOUS in Vitro synthesis of IgE by PBMC cells of atopic donors. As indicated, in this model the production of IgE is due to in vivo pre-activated B cells and is IL-4independent. All the available evidence indicates that IFNy is a potent suppressor of the IgE response in vitro and that it is likely to exert the same effect in vivo under physiological conditions. However, two unexplained observations should be considered to avoid coming to a premature conclusion, First, some IL-4producing T-cell clones are capable of inducing a strong IgE response in spite of their concomitant production of high amounts of IFNy [6]. Set ond, the addition of IFNy to IL-4-stimulated PBMC cell cultures may, occasionally, increase rather than suppress the IgE response. This paradoxical effect of IFNy , which is observed more frequently on neonatal (umbilical cord blood) than on adult PBMC, is at present unexplained (Peleman et al, submitted). Finally, it is worth noting that transforming growth factor p, which selectively increases the production of &A by LPS-stimulated mouse B cells [17], is a very good inhibitor of the in vitro synthesis of human IgE (Delespesse, unpublished observation). It is, therefore, possible that this cytokine also plays a role in the physiological regulation of IgE synthesis.
The role of T lymphocytes As discussed above, the production of IgE appears to depend upon the balance between the production of IL4 and IFNy, which are mainly produced by T cells. In other words, the &type selection is dictated by the na ture of the interactions between antigen-presenting cells and T cells. Long-term mouse T-cell clones can be di vided into two distinct subsets, TH1 and TH2, which dilfer in their patterns of lymphokine production (Mosmann et al, 1986). Trill produce IL-2, IFNy and LT but not IL-4 or IL-5, whereas Tn2 produce IL-4, IL-5, IL-~ but not IL-2 or IFNy; other lymphokines (IL-3 granulocyte/macrophage colony-stimulating factor, etc.) are produced by both subsets. Human Tn cells cannot be divided into TH1 and Tn2 subsets because most of the IL-4-producing T-cell clones also secrete either IFNy or IL-2, or both (Maggi et al., Eur J ZmmunoZ1988, 18:1045-1050). More recent studies [ 16,181 in the mouse have shown that short-term Tn cell clones resemble their human counterparts, in as much as they secrete both T,l and TH2 types of cy tokines. These T cells, named T,O, are thought to be at an Intermediate stage of differentiation: they are derived from the antigen stimulation of a precursor primary T, cell which is capable of producing only IL-2, and upon repeated antigenic stimulation, THO may differentiate into stable TH1 and T,2 cells. The functions of Trill and Tn2 cells are clearly distinct and as in the case of IgE regulation, they are often antagonistic. Moreover, Trill and Tn2 cross-inhibit each other, as shown by the observations that, firstly, IFNy inhibits the proliferation of Tn2 cells (without affecting their lymphokine production) and, set
Sarfati
and Heuser
ondly, that activated Tn2 cells inhibit lymphokine production by Tnl cells via the release of a soluble factor called cytokine synthesis-inhibiting factor (Mosmann et al, FRSEBJ 1989,3~1269). Because Tnl and Tn2 are fully differentiated cells gen erated by repetitive and prolonged antigen stimulations, they cannot be involved in isotype selection during a pnmary response. They probably act as memoty T cells and are involved in the maintenance of ongoing and prolonged immune responses. The hallmark of the IgE an tibody response of atopic patients to airborne allergens is that it is prolonged and that it may be boostered, as shown by the post-seasonal rise in IgE anti-pollen antibody. This probably accounts for the presence in these patients of memory T+like cells, but this possibility remains to be tested. Advances in our understanding of the mechanisms determining the production of IgE antibodies in the allergic patient will come from the analysis of the factors goveming the preferential different&ion of primary Tn cells into either T,2 or T,l lymphocytes. The most informative result in this respect is the recent linding that antigen-specific TH clones, freshly derived in the presence of antigen and IL-2, are mainly of the Tn2 phenotype, whereas Trill cells predominate if the cloning is performed in the presence of IFNy. Hence, the presence of IFNy in the early course of an immune response may result in the preferential expansion of T,l cells [ 161. Other factors that may critically influence the differentiation pathway of either the primary or the T,O cells include: the nature of the antigen-presenting cell and the density of antigen on it, the presence of bystander CD8+ T cells or natural killer cells secreting IFNy and, probably, the microenvironment in which the interaction between T cells and antigen-presenting cells takes place.
The role of Fc&RII (CD231 and its soluble fragments Cellular distribution, structure and regulation of FcsRll expression
In contrast with the high-alfiity receptor for IgE (FcsRI; Kinet, this issue, pp 499-505), FcsRII may be expressed on several haematopoietic cells as well as on some epithelial cells (Delespesse et al, Zmmunol Today 1989, 10:159-164; Billaud et al., J Viral 1989, 63:41214128). Hence, FcsRII may be found on B/T lymphocytes, monocytes, eosinophils, platelets, follicular dendritic cells and epidermal Iangerhans’ cells (Delespesse et al, 1989). On B lymphocytes, FcERII is selectively expressed on surface &M/surface IgDbearing cells and it is lost after isotype switching (Kikutani et LA, J Eqb Med 1986, 164:1455-1569). The structure of FcsRII is well known and has recently been reviewed in detail (Delespesse et al, Cbem Zmmunoll989, 47:78105). Brie& FcERII is a 45kD sialoglycoprotein containing one N-linked carbohydrate chain of the complex type. Human and mouse FcsRII cDNAs have been cloned and functionally expressed (Liidin et al, 1987) [ll. They display significant
509
510
Atopic allergy
and other
hypersensitivities
homology and both have an inverted membrane orientation, with the amino terminal inside and the carboxy terminal outside the cell membrane. The &E-binding domain has been mapped to a region that also displays a striking homology with several animal lectins [19] and with a newly discovered family of cell adhesion molecules (Mary et al, Science 1989, 243:1144). Unexpectedly, the binding of IgE to Fc&RII does not seem to involve carbohydrate interactions, as unglycosylated native or recombinant IgE or &I-derived peptides are capable of binding to FcERII [ 201. In humans, the Fc&RII gene may encode two tierent molecules (named a and b) which have different initiation sites and alternative splicing (Yokota et al., Cell 1988, 55:611-618). FcERII a and b dilfer only by a few amino acids in their cytoplasmic tail; the extracellular and transmembrane domains are identical. Type a is found on freshly isolated and unstimulated normal B cells while type b is expressed on IL-4-stimulated B cells, monocytes, eosinophils and T cells. FcERII is cleaved into soluble fragments, some of which
retain the ability of binding IgE (&E-binding factors).
FcERII is first cleaved into unstable 37 and 33 kD IgE
binding factors that are subsequently transfomred into more stable 25kD molecules (Letellier et al, Mol Zmmunol 1989, 26: 1105-l 112). Recent evidence indicates that this cleavage is an autoproteolytic process and that the 37kD &E-binding factors are endowed with the enzyme activity (Letellier et al., submitted). The rate of cieavage of FcsRII is reduced by IgE and increased by treatment of the cells with inhibitors of N-linked glycosylation. The expression of FcsRII and the concomittant production of IgEbinding factors are tightly regulated. They are induced by IL-4 and suppressed by IFNy; IFNy may in hibit FcERII expression by B cells but not by monocytes, platelets and Iangerhans’ cells (Delespesse et al., 1989) [21]. FcERII b is overexpressed in patients with increased production of IgE, for example those with severe atopy or some parasitic diseases (Watanabe et al., Appl Zmmunoll989, 89:103-107). Abnormally high expression of this molecule is also found in patients with B-cell chronic lymphocytic leukemia, a condition associated with hypogammaglobulinemia and very low serum IgE levels (Sarfati et al, Blood 1988, 71:94-101). On B cells, FcERII may also be up-regulated by direct interaction with T cells (in the absence of IL-4) and by various agents including leukotriene B4 and IL-2 (Yamaoka et al, J Zmmunoll989,143:1996-2000; Crow et al, J Eap Med
1986, 164:1760-1772). FcERII is a multifunctional
molecule
Membrane-bound FcERII may exert different functions according to the cell types on which it is expressed. On eosinophils, platelets and monocytes, Fc&RII b mediates &E-dependent cellular cytotoxicity against parasites as well as &E-dependent release of inllammatory mediators (Capron et al., Zmmunol Today 1986,7:15-21). On B cells, FcERII focuses the &E-dependent presentation of antigen to T cells [22]. In addition, FcERII may also play the role of a cell adhesion molecule (Kishimoto, per-
Sonal observation), probably by virtue of its lectin activity, which remains to be demonstrated. Soluble FcERU fragments are reported to have several activities in vitro including: (1) regulation of the synthesis of human IgE; (2) promotion of B-cell and Tcell proliferation (Swendeman et al, EMBO J 1987, 6~1637-1642; An&age et al., Eur J Zmmunol 1989, 10:31-35); and (3) most interestingly, induction, in synergy with IL-I, of the maturation of early thymocytes (CD2-, CD3_CD4/CD8-) unresponsive to IL-2, into CD2+, CD3 + , CD4/8+ cells that are responsive to IL-2 (Mossalayi et al, J Eq Med 1990, in press).
Role of FcsRll/lgE-binding
factors
in IgE regulation
role of these molecules has been documented in four models: (1) the stimulation of unfractionated PBMC with IL-4; (2) the established, ongoing synthesis of IgE by IL-4-prestimulated normal PBMC; (3) the spontaneous synthesis of IgE by PBMC of atopic patients, and (4), more recently, IgE synthesis induced by the co-stimulation of monoclonal chronic lymphocytic leukemia B cells (B-CLLs) with IL-4 and hydrocortisone [9]. In each case, the IgE synthesis can be markedly suppressed by some monoclonal antibodies to FcERII (FILMIER), cross-reacting with &E-binding factors. The suppression is dose-dependent and, in the case of B-CLLs, complete inhibition is observed with as little as 0.1 clg/ml of tdb~R. The effect of mAbsR or of its F(ab’)2 fragments is isotype-specific (Sarfati et al, J Zmmunol1988, 141:2195-219, Pene et al, Proc Nat1 Acad Sci USA 1988, 856880-6884). Two observations indicate that IIIA~ER do not suppress IgE synthesis by delivering a negative signal to &E-secreting cells. First, these cells have lost their FcERII (Kikutani et al., J Exp Med 1986,164:1455-1469). Second, not all the mAb&Rs are equally active and only those antibodies preventing the cleavage of FcERII are capable of suppressing IgE synthesis. The
Taken together, the inhibition studies using mAbsR suggest that the soluble, rather than the membrane-bound, FcsRII up-regulates the synthesis of IgE. This view is supported by the observations that &E-binding factors augment the production of 1gE in the above models. The ability of these molecules signiftcantly to enhance ongoing IgE synthesis by unstimulated lymphocytes from atopic donors was described several years ago in work which used afhnity-purilied &E-binding factors isolated from the culture supernatant of FcsRIIbearing B-cell lines (P&e et al., Eur J Zmmunol 1988, 18:929-935). These results have now been conIirmed (P&e et al., 1989) and extended using recombinant IgEbinding factors. These molecules signilicantly increase the ongoing and IL-4-independent synthesis of IgE by either in viva or in vitro preactivated B cells. Native or recombinant IgE binding factors may also augment the 1gE response of PBMC stimulated with either suboptimal concentrations of IL-4 (P&e et al, Eur J Zmmunoll988, 18:92+935) or after partial depletion of monocytes. Further studies have clearly indicated that the IgE-potentiating activity is mainly mediated by the unstable 3337kD &E-binding factors. This was shown by coculturing target lympho-
IgE synthesis
qtes with various Chinese hamster ovaly (CHO) transfectants expressing either the complete FcsRII molecule (CHO l-7), a secreted form of 37 kD IgE-binding factors (CHO-37) or a secreted form of 25kD &E-binding factors (CHO-25). A pronounced increase in IgE synthesis is only observed with CHO 1-7 and CHO-37 cells; under these conditions, where the unstable 37kD binding factor is constantly produced, a 2-20-fold enhancement of the IgE response is obtained with very low concentrations of &E-binding factors, i.e. 0.1 nmol/litre (Delespesse et al, submitted). It may therefore be concluded that IgEbinding factors (soluble CD23) potentiate the synthesis of IgE by in& en&g the activity of the IL-4-independent &E-activated B cells. Current studies are aimed at determining whether this effect is direct or indirect, via T cells or monocytes.
Acknowledgements Dr G. Delespesse is an MRC associate; Dr M. Sarfati is the recipient of an MRC scholarship; this work is supported by the MRC of Canada. The authors gratefully acknowledge Drs F. Fielman and T. Nutman for providing access to unpublished manuscripts. The excellent secretarial assistance of Mrs Norma Del Bosco is most appreciated.
Annotated
references
and recommended
reading
Delespesse,
Sarfati
and Heusser
postrongylus brasiliensis or by injection of goat anti-mouse I’D antibody, and (2) the secondaty IgE anti-dinltrophenol response, Anti-IL-4 has no effect on the IgG, responses elicited under the same conditions. When given at the peak of the secondary IgE anti-dinitrophenol response or of the N. brasiliensis infection, anti-IL-4 accelerates the decline in total and specific anti-dinitrophenol IgE, respectively, but the suppression is not complete. This paper unequivocally demonstrates that IL-4 is required for an in vivo primary IgE response and that it also plays a major role in the secondary and ongoing Igl? response. 4. 0
THYF?RONITISG, T~OKO~GC, JUNE CH, IEV~NEAD, EINKELMAN FD: I@ secretion by Epstein-Barr virus-infected purified human B lymphocytes is stimulated by interleukin 4 and suppresed by interferon y. Pm Nat1 Acad Sci USA 1989, 86:558@5584. Co-stimulation of highly purified human B cells with EBV and IL-4 induces a signiIicant production of IgE after H weeks of CUlNre. Both stimulants are required to induce IgE synthesis; in contrast, EBV alone induces the secretion of other immunoglobulin isotypes. IFNr at 10 U/ml specifically suppresses the IgE response. This Study demonstrates that: (1) as in the mouse, highly purihed human B cells may be induced to synthesize IgE, provided a second activation signal is given, and (2) IFNy may suppress the IgE response by a direct effect on human B cells. VERCELLI D, JABARAHU, ARAIKI, GEHAR Induction of human IgE synthesis requires interleukin 4 and T/B cell interactions involving the T cell receptor/CD3 complex and MHC class II antigens. J &?I$ Med 1989, 169:129>1307. The induction of human IgE synthesis by IL-4-stimulated PBMC requires cognate interaction betwen autologous T/B cells; it is blocked by antibodies against the T-cell receptor /CD23 complex or against MHC class II antigens on B cells. Antibodies to adhesion molecules (CD2, CD4, LFA-1) also Inhibit the IgE response, indicating that cell-cell contact is required. By means of selected alloreactive T-cell clones it is also shown that bystander B cells (whose MHC class II molecules are not recognized by T cells) are not induced to synthesize IgE. 5. 0
6.
DE KRLIYFF RH, TURNERT, ABRAMS JS, PAUADINO JR MA, U~TSU
??e
0 00
Of interest Of outstanding Interest
1. a
BETIXER B, HOFETETIXRM, RASM, YLIKOYOUKA WM, KILCHHERR F, CONRADDH: Molecular structure and expression of the murine lymphocyte low-atTin@ receptor for IgE (FceRII). Pm Nat1 Acad Sci USA 1989, 86~7567570. The cDNA coding for mouse Fc&RII was cloned and functionally expressed in COS cells. The predicted amino acid sequence demonstrates: (1) 57% identity to human FcsRII; (2) the presence of an additional repetitive sequence of 21 amino acids giving 4 repeats compared with 3 in the human sequence; and (3) the carboxy terminus is truncated and does not contain the Asp-GlyArg sequence that is expressed in human FcERI~and that may be involved in the binding to integrin. 2.
FINKELMAN FD, HOIMESJ, URBANJR JF, PAULWE, KATONAIM: T help requirements for the generation of an in viva IgE response: a late acting form of T cell help other than IL-4 is required for IgE but not for IgGl production. J Immunol 1989, 142:403-$08. The Injection of mice with goat antibody to mouse IgD induces a pronounced polyclonal IgE and IgGl response. Both responses are blocked by anti-CD4 antibody whereas anti-IL-4 only blocks the IgE rep spouse. When the injection of anti-CD4 or anti-IL-4 antibodies is delayed to day 5 after the Injection of goat anti-mouse IgD, anti-IL-4 has little effect on the IgE response whereas anti-CD4 strongly inhibits the IgE response and slightly reduces IgGl production. This paper demonstrates that: (1) the mechanisms regulating the in uivo production of IgGl and IgE are different; and (2) a late-acting form of T-cell help, other than K-4, is required for IgE but not for IgGl production. ??.
3. a.
FINKEU~AN ED, KATONA IM, URBANJF, HOLMESJ, OHARA J, TUNG AS, SAMPIXJV, PAULWE: IL-4 Is required to generate and sustain in vivo IgF, responses. / Immunoll988, 141:23352341. The injection of anti-IL-4 neutralizing antibody to mice suppresses: (1) the polyclonal IgE responses induced either by infection with N@
DT: Induction of human IgE synthesis by CD4+ T ceU clones-requirement for interleukin 4 and low molecular weight B cell growth factor. J @ Med 1989, 170:1477-1493. WeU characterized alloreactive human CD4+ T-cell clones were tested for their ability to induce IgE synthesis under cognate interaction with resting B cells. One clone co-producing substantial amounts of IFNy and II-4 is a good Inducer of IgE; another clone producing only IL4 (and not IL-2 or IFNr) fails to induce IgE unless the CUlNreS are supplemented with the LMWBCGF. Under these conditions, only IgE is induced; the production of the other isotypes requires the addition of K-2. This paper demonstrates that 3 sets of signals are required for the induction of IgE synthesis by alloreactive T-cell clones: (1) those derived from the direct T/F-cell contact; (2) LMWBCGF; and (3) IL-4. HART PH, Vrrn GF, BURGESS DR, WHITIY GA, PICCOLIDS, anti-inflammatory effects of interleuldn 4: Suppression of human monocyte tumor necrosis factor, interleukin 1, and prostaglandin E> Proc Natl Acud Sci USA 1989, 86:380>3807. The production of IL-I, TNFa and PGE, by human monocytes stimulated with LPS or with LPS and IFNy is inhibited by small concentrations of IL-4 (0.1-0.5 U/ml). II-4 has the same effect as 0.1 nmol/litre dexamethasone. 7. 0
HAMIUON JA: Potential
PELEMAN R, Wu J, EARGENEC, DELESPESSE G: Recombinant lnterleukin 4 suppresses the production of interferon y by human mononuclear cells. J k$!) Med 1989, 170:1751-1757. IL-4 inhibits the production of IFNy by human PBMC stimulated with mitogen or allogeneic cells. The suppression is dose and time-dependent; this is observed at the mRNA and at the protein level. The data further suggest that the balance between the production of IL-4 and IFNy, governing the synthesis of IgE, may be signiscantly iniluenced by the chronological order of activation of the IL-4/IFNy genes.
8. 0
SAWATIM, Luo H, DEIESPESSEG: IgE synthesis by chronic 9. 0 lymphocytic leukemia cells. J Exp Med 1989, 170:17751781. Unfractionated lymphocytes from patients with B-cell chronic lymphocytic leukemia are induced to secrete monoclonal IgE upon co-stimula
511
512
Atopic allergy
and other
hypemensitivities
tion with IL-4 and hydrocortisone. Hydrocottisone is required to induce of normal lymphocytes it IgE synthesis by B-CLIs, whereas in CdNre increases the IgE response to IL-4. 10. ??e
COFFMAN RL Interleuldn 4 causes isotype~ switching to IgE in T cell-stimulated clonal B ceU cultures. J E~J Med 1988, 168853-862. Clonal CdNres of B cells stimulated with a Tu2 ceU line speciIic for rabbit IgG were established. In this system, the majority of B cells undergo clonal expansion in the presence of rabbit anti-mouse IgM The addition of IL-4 or of anti-IL-4 indicates that: (1) II-4 increases the proportion of IgE-secreting clones, has no effect on the frequency of IgGrproducing clones and reduces the proportion of @-producing clones; and (2) IL-4 increases the quantity of IgE produced by single clones. This study demonstrates that IL-4 promotes IgE synthesis by directing the isotype switching to IgE and not by selectively supporting the mats uration of precommltted precursor cells. LEBMAN DA,
11.
ROTHMAN P, LUTZKER S, COOKW, COFFMAN R, ALT W Mitogen plus lnterleukln 4 induction of CE transcripts in B lymphoid cells. J ELI Med 1988, 168:23852389. Murine splenic B cells and Abelson murine leukemia virus-transfomred cells were cultured with LPS and IL-4 and assayed for E and yl transcripts. Germline yl and E transcripts appear first, and are followed by the accumulation of normal E mRNA. The resuks suggest that IL-4 can direct class switching by increasing recombinase accessibility to yl and E switch regions. ??
12.
T, DCKH HM: Clonal and molecular characteristics of the human &E-committed B cell subset. J Exp Med 1989, 169:407-430. Monoclonal EBV~transformed B-cell lines were obtained by ceU soning followed by limiting dilution. It was observed that: (1) individual IgEsecreting cells co-secrete IgM and IgD; (2) these IgEsecreting cells have not rearranged their lmmunoglobulin gene; (3) u and E mRNA are of normal and comparable size; and (4) U266 IgE myeloma cells have rearranged their immunoglobulin gene. This study demonstrates that rearrangement of the E gene (with deletion of the intervening sequences between assembled VDJ and E) is not always required for IgE production. It is postulated that B cells capable of secreting IgE belong to a unique B-cell sub-lineage termed ‘IgEcommiued cells’. MACKENZIE
??
SPIAWSKIJR, JELINCKDF, LIPSKYPF: Immunomodulatory role of IL-4 on the secretion of Ig by human B cells. J Immunol 1989, 142:15691575. Highly purified human B-cell blasts induced by stimulation with S. aureus are capable of producing immunoglobulins of all isotypes (including IgE) when cultured in the presence of IL-2 but not IL-4. Addition of II-4 to the IL-2-containing secondaty CUlNreS inhibits immunoglobulin production (including IgE) and cellular proliferation. The IL-4-mediated suppression is partially reversed by IFNy and LMWBCGF. Additional data document the complex immunoregulatory action of IL-4 on human B cells. 13. ??
TR, COFFMAN RI: IFNy FINKELMAN ID, KATONA IM, MOSMANN regulates the isotypes of Tu secreted during in uivo humoral Immune responses. J Immunol 1988, 140:1022-1027. Multiple injections of IFNy signiIicantly inhibit the IgE and IgGt response of mice to goat anti-mouse IgGD. Anti-IFNy neutralizing antibody has the reverse effect, The data also show that IgGaa production is enhanced by IFNy. Immunization of mice with killed and tixed Brucelkz abortus, known to induce IFNY production, increases the production of IgG2a and suppresses that of IgE and IgGl. Injection of B. &or&s-immunized animals with anti-IFNy antibody inhibits the increase of IgG2a and enhances the production of IgGt but does not reverse the suppresslve effect of B. ubortus on IgE production. Hence, the B. &or&&n duced IgE suppression is not mediated by IFNy. 14. ??
BERTONMT, UHRJW, VITET~AES: Synthesis of germ-line yl ImmunoglobuRn heavy-chain transcripts In resting B cells: Induction by Interleukin 4 and inhibition by Interferon. Pnx Nat1 Acad Sci USA 86:28292833. By means of an RNase protection assay, it was observed that IL-4 induces expression of germline yl transcripts in B cells after 4 h of cul15. ??
Nre; the induction is inhibited by IFNy The accumulation of yl mRNA * may be induced in resting B cells stimulated with IL-4 alone but it is increased by co-stimulation with LPS. This study strongly suggests that IL-4 directs the class switching to IgG, by opening the Syl switch region and making it accessible to switch recombinase. 16. ee
GAJEXVXITF, SCHE~LSR, NAU G, F~CH FW: Regulation of TceU activation: Differences among T-cell subsets. Zmmunol Ra, 1989, 79111. This is a timely review of the mechanisms governing the differentiation of cells into T,l or TH2 ceU types. The roles of IFNy and antigen-presenting cells are emphasized as weU as the different requirements and signalling mechanisms for the antigen-induced activation of the subsets. 17.
COFFMAN R, &BMAN DA, SARADERB: Transforming growth fattor fi specifically enhances IgA production by lipopolysaccharlde-stimulated murine B lymphocytes. J Exp Med 1989, 170:10391044. The addition of transforming growth factor g to purified mouse B cells stimulated with LPS induces pronounced production of Ig.4 and suppresses the other isotypes. The IgA response is further increased by IL-2 and IL-5. Transforming growth factor j3 induces a substantial increase in IgA production by surface IgA cells whereas it suppresses IgA production by B surface IgA+ cells. This paper strongly suggests that transforming growth factor j3 acts as an isotype-specific switch factor for lg.& ??
18.
MOSMANN TR, COFFMANRL Heterogeneity of cytoldne secretion patterns and functions of helper T cells. Adv Immunol 1989, 46:111-147. This is a most up-to-date review of the author’s work on TH1 and TH2 cells with regard to their phenotypic and functional characteristics, their role in the immune response, and their cross-inhibition. New concepts regarding the diUerentiatlon of precursor TH cells into Tul and TH2 types are presented. ??
site for IgE of the human FCE receptor II (FcsRII/CD23) is cormned to the domain homologous with animal lectins. Proc Nat1 Acud Sci USA 1989, 86:7118-7123. FcERII mutants were expressed in mammalian ceUs with a view to mapping the IgEbindlng site of the receptor. This is localized to a domain of 123 amino acid residues which is also homologous to the carbohydratebinding domain of animal lectins. The epitopes of several monoclonal antibodies to CD23, known to inhibit the binding of IgE to FcERII, are clustered in the same region. This is also the case with the MHM6 monoclonal antibody, which is known to mimic the action of LMWBCGF. 19.
BE’ITLER B, MAIEXRR, RUEGG D, HOFSTETIER H: Binding
??
20.
VERCELLI D, HELM B, MARSH P, PADIAN E, GEHA RS, GOULD
H: The B-cell binding site on human lmmunoglobulin E. Nature 1989, 338649651. By means of recombinant E-chain fragments synthesized in Escherichia coli, the CD23-binding site of IgE has been mapped to a sequence of the ~3 domain. Glycosylation of IgD is not required for binding to FcERII 0 or FcERI. ??
21. ??
BIEBER T, RIEGERA, NEUCHK~T C, PRINZJC, RICHER EP, BOLT& Nwnscu G, %IEINJ?R 0, KRAFT D, RING J, STINGLG: Induc-
tion of FcsR2/CD23 on human epidermal Langerhans cells by human recombinant interleukin 4 and interferon. J EQ Med 1989, 170:309-314. Langerhans’ cells, freshly isolated from the skin of normal donors, do not express CD23/FcaRU These molecules are induced after 24 h incubation in the presence of either IL-4 or IFNy; neither phorbol-12. my&ate-13.actetate (PMA) nor IL-2 have any effect; a synergistic effect is observed between IL-4 and IFNy. 22. ??e
KHERYMR, YAMASHITA LC: Low-atlkity IgE receptor (CD23) function on mouse B cells: role in &E-dependent antigen focusing. Proc Nat1 Acad Sci 1989, 86:75567560. ti Mouse B cells treated with IgE monoclonal antibodies to trinitrophenol are 100.fold more effective than untreated B cells in presenting low concentrations of ttinitrophenol antigen to T cells. This is blocked by This paper preventing IgE binding to the B cells by means of IIIA~ER documents one function of FcERII on B cells, i.e. the &E-dependent internalization of antigen resulting in efficient presentation to T cells.