Investigation of inflammatory and allergic responses to common mold species: Results from in vitro experiments, from a mouse model of asthma, and from a group of asthmatic patients.

Most studies on molds focus on Alternaria alternata and Aspergillus fumigatus. Here, we report on inflammatory and allergenic properties of more typical indoor species Aspergillus versicolor, P. chrysogenum, C. cladosporioïdes, and C. sphaerospermum that were compared to A. alternata and A. fumigatus. In a mouse model, after intranasal instillation, A. alternaria, A. versicolor, and C. sphaerospermum induced the early recruitment of neutrophils and the strong expression of inflammatory markers in the bronchoalveolar lavages fluids. A. fumigatus also induced the early accumulation of neutrophils but with lower levels of inflammatory markers. Chronic treatment induced variable response according to species: P. chrysogenum and A. fumigatus appeared strong pro-allergenic inducers compared to A. alternata and C. sphaerospermum while A. versicolor and C. cladosporioides induced a mixed pro-allergenic/pro-inflammatory response. In mold-sensitized asthmatics, mold-specific Immunoglobulin E (IgE) were detected with an in-house dot-blot assay. A. fumigatus and A. alternata were the most frequent sensitizers. Altogether, P. chrysogenum, P. brevicompactum, C. sphaerospermum, and C. cladosporïoides were the "major sensitizer" (defined as the strongest response against a single mold species) for almost 30% of the asthmatics. These results show that, not only A. alternata and A. fumigatus, but also indoor species have strong inflammatory and allergic properties and a harmful potency.

Maintenance of B cells during chronic murine Trypanosoma brucei gambiense infection.

African trypanosomosis is a debilitating parasitic disease occurring in large parts of sub-Saharan Africa. Trypanosoma brucei gambiense accounts for 98% of the reported HAT infections and causes a chronic, gradually progressing disease. Multiple experimental murine models for trypanosomosis have demonstrated inflammation-dependent apoptosis of splenic follicular B (FoB) cells and the destruction of B-cell memory against previously encountered pathogens. Here, we report that during murine infection with a chronic T. b. gambiense field isolate, FoB cells are retained. This coincided with reduced levels of IFN-γ and TNF-α during the acute phase of the infection. This result suggests that in chronic infections with low virulent parasites, less inflammation is elicited and consequently no FoB cell destruction occurs.

In vivo characterization of two additional Leishmania donovani strains using the murine and hamster model.

Leishmania donovani is a protozoan parasite causing the neglected tropical disease visceral leishmaniasis. One difficulty to study the immunopathology upon L. donovani infection is the limited adaptability of the strains to experimental mammalian hosts. Our knowledge about L. donovani infections relies on a restricted number of East African strains (LV9, 1S). Isolated from patients in the 1960s, these strains were described extensively in mice and Syrian hamsters and have consequently become 'reference' laboratory strains. L. donovani strains from the Indian continent display distinct clinical features compared to East African strains. Some reports describing the in vivo immunopathology of strains from the Indian continent exist. This study comprises a comprehensive immunopathological characterization upon infection with two additional strains, the Ethiopian L. donovani L82 strain and the Nepalese L. donovani BPK282 strain in both Syrian hamsters and C57BL/6 mice. Parameters that include parasitaemia levels, weight loss, hepatosplenomegaly and alterations in cellular composition of the spleen and liver, showed that the L82 strain generated an overall more virulent infection compared to the BPK282 strain. Altogether, both L. donovani strains are suitable and interesting for subsequent in vivo investigation of visceral leishmaniasis in the Syrian hamster and the C57BL/6 mouse model.

Immunomodulatory properties of Lactobacillus plantarum and its use as a recombinant vaccine against mite allergy.

BACKGROUND: Selected lactic acid bacteria were reported to prevent atopic dermatitis and experimental asthma but the mechanisms of their immunomodulatory effects are not fully elucidated. In this study, the signaling pathways triggered by Lactobacillus plantarum NCIMB8826 were investigated and the potential use of this strain producing a variant of the mite allergen Der p 1 as live vaccine vehicle was evaluated. METHODS: Mouse bone marrow-derived dendritic cells were stimulated with wild-type or a L. plantarum teichoic acid mutant to evaluate the secretion of cytokines. A recombinant L. plantarum expressing Der p 1 was engineered, its in vitro immunomodulatory properties were characterized and its prophylactic potential was evaluated in a Der p 1-sensitization murine model. RESULTS: Mouse dendritic cells stimulated by L. plantarum triggered the release of interleukin-10 (IL-10), IL-12 p40, IL-12 p70 and tumor necrosis factor-alpha (TNF-alpha). IL-12 p40 secretion was dependent on nuclear factor-kappaB (NF-kappaB), mitogen-activated protein (MAP) kinases, Toll-like receptor 2 (TLR2), TLR9 and on the bacterial teichoic acid composition. Recombinant L. plantarum producing Der p 1 exhibited similar immunostimulatory properties as wild-type. Prophylactic intranasal pretreatment of mice with this recombinant strain prevented the development of the typical Th2-biased allergic response by a drastic reduction of specific IgE and the induction of protective allergen-specific IgG2a antibodies. Moreover, both wild-type or recombinant L. plantarum reduced airway eosinophilia following aerosolized allergen exposure and IL-5 secretion upon allergen restimulation. CONCLUSION: By combining both Th1-type immunostimulatory properties and an efficient allergen delivery capacity, recombinant L. plantarum producing Der p 1 represents a promising vaccine against house dust mite allergy.

The SH2 domain-containing 5-phosphatase SHIP2 is expressed in the germinal layers of embryo and adult mouse brain: increased expression in N-CAM-deficient mice.

The germinative ventricular zone of embryonic brain contains neural lineage progenitor cells that give rise to neurons, astrocytes and oligodendrocytes. The ability to generate neurons persists at adulthood in restricted brain areas. During development, many growth factors exert their effects by interacting with tyrosine kinase receptors and activate the phosphatidylinositol 3-kinase and the Ras/MAP kinase pathways. By its ability to modulate these pathways, the recently identified Src homology 2 domain-containing inositol polyphosphate 5-phosphatase 2, SHIP2, has the potential to regulate neuronal development. Using in situ hybridization technique with multiple synthetic oligonucleotides, we demonstrated that SHIP2 mRNA was highly expressed in the ventricular zone at early embryonic stages and subventricular zones at latter stages of brain and spinal cord and in the sympathetic chain. No significant expression was seen in differentiated fields. This restricted expression was maintained from embryonic day 11.5 to birth. In the periphery, large expression was detected in muscle and kidney and moderate expression in thyroid, pituitary gland, digestive system and bone. In the adult brain, SHIP2 was mainly restricted in structures containing neural stem cells such as the anterior subventricular zone, the rostral migratory stream and the olfactory tubercle. SHIP2 was also detected in the choroid plexuses and the granular layer of the cerebellum. The specificity of SHIP2 expression in neural stem cells was further demonstrated by (i) the dramatic increase in SHIP2 mRNA signal in neural cell adhesion molecule (N-CAM)-deficient mice, which present an accumulation of progenitor cells in the anterior subventricular zone and the rostral migratory stream, (ii) the abundant expression of 160-kDa SHIP2 by western blotting in proliferating neurospheres in culture and its downregulation in non-proliferating differentiated neurospheres. In conclusion, the close correlation between the pattern of SHIP2 expression in the brain and the proliferative and early differentiative events suggests that the phosphatase SHIP2 may have important roles in neural development.

The SH2 domain containing inositol 5-phosphatase SHIP2 associates to the immunoreceptor tyrosine-based inhibition motif of Fc gammaRIIB in B cells under negative signaling.

Fc gammaRIIB are single-chain low-affinity receptors for IgG that bear an immunoreceptor tyrosine-based inhibition motif (ITIM) in their intracytoplasmic domain and that negatively regulate immunoreceptor tyrosine-based activation motif (ITAM)-dependent cell activation. In B cells, coaggregation of the B cell receptor (BCR) and Fc gammaRIIB leads to an inhibition of B cell activation. Inhibitory properties of Fc gammaRIIB have been related to the recruitment of SHIP, an SH2 domain-containing inositol 5-phosphatase (referred to as SHIP1), via ITIM phosphorylated Fc gammaRIIB. Here, we demonstrate that the second SH2 domain-containing inositol 5-phosphatase SHIP2 could also bind to the Fc gammaRIIB ITIM. As a model, a Fc gammaRIIB deficient B cell line (IIA1.6), transfected with a cDNA encoding either w.t. Fc gammaRIIB1' or Fc gammaRIIB1' whose ITIM tyrosine was mutated has been used. SHIP2 tyrosine phosphorylation and association to the adaptator protein Shc were only found in transfectants expressing w.t. Fc gammaRIIB1'. SHIP2 was also found to bind to a phosphopeptide corresponding to the ITIM sequence of Fc gammaRIIB. There was no binding to the nonphosphorylated peptide. Finally, both SHIP2 and SHIP1 were coprecipitated with Fc gammaRIIB1' upon coaggregation with BCR in IIA1.6 transfectants.

Downregulation of antigen-presenting cell functions after administration of mitogenic anti-CD3 monoclonal antibodies in mice.

Antibodies against CD3epsilon are widely used as immunosuppressive agents. Although it is generally assumed that these reagents exert their immunomodulatory properties by inducing T-cell deletion and/or inactivation, their precise mechanism of action remains to be elucidated. Using a murine model, we demonstrate in this report that administration of anti-CD3epsilon antibodies causes the migration and maturation of dendritic cells (DC) in vivo, as determined by immunohistochemical analysis. This maturation/migration process was followed by selective loss of splenic DC, which resulted in a selective inhibition of antigen-presenting cell (APC) functions in vitro. Spleen cells from anti-CD3epsilon-treated animals were unable to productively stimulate naive alloreactive T cells and Th1-like clones in response to antigen, while retaining the ability to present antigen to a T-cell hybridoma and Th2 clones. Anti-CD3epsilon treatment was found to induce a selective deficiency in the ability of spleen cells to produce bioactive interleukin-12 in response to CD40 stimulation. APC dysfunction was not observed when nonmitogenic forms of anti-CD3epsilon antibodies were used, suggesting that splenic DC loss was a consequence of in vivo T-cell activation. Nonmitogenic anti-CD3epsilon monoclonal antibodies were found to be less immunosuppressive in vivo, raising the possibility that APC dysfunction contributes to anti-CD3epsilon-induced immunomodulation. Collectively, these data suggest a novel mechanism by which mitogenic anti-CD3epsilon antibodies downregulate immune responses.

Distribution of the src-homology-2-domain-containing inositol 5-phosphatase SHIP-2 in both non-haemopoietic and haemopoietic cells and possible involvement of SHIP-2 in negative signalling of B-cells.

The termination of activation signals is a critical step in the control of the immune response; perturbation of inhibitory feedback pathways results in profound immune defects culminating in autoimmunity and overwhelming inflammation. FcgammaRIIB receptor is a well described inhibitory receptor. The ligation of B-cell receptor (BCR) and FcgammaRIIB leads to the inhibition of B-cell activation. Numerous studies have demonstrated that the SH2-domain-containing inositol 5-phosphatase SHIP (referred hereto as SHIP-1) is essential in this process. The cDNA encoding a second SH2-domain-containing inositol 5-phosphatase, SHIP-2, has been cloned [Pesesse, Deleu, De Smedt, Drayer and Erneux (1997) Biochem. Biophys. Res. Commun. 239, 697-700]. Here we report the distribution of SHIP-2 in mouse tissues: a Western blot analysis of mouse tissues reveals that SHIP-2 is expressed in both haemopoietic and non-haemopoietic cells. In addition to T-cell and B-cell lines, spleen, thymus and lung are shown to coexpress SHIP-1 and SHIP-2. Moreover, SHIP-2 is detected in fibroblasts, heart and different brain areas. SHIP-2 shows a maximal tyrosine phosphorylation and association to Shc after ligation of BCR to FcgammaRIIB but not after stimulation of BCR alone. Our results therefore suggest a possible role for SHIP-2 in the negative regulation of immunocompetent cells.

Role and regulation of IL-12 in the in vivo response to staphylococcal enterotoxin B.

Injection of a staphylococcal superantigen (SAg) such as staphylococcal enterotoxin B (SEB) in adult mice results in cytokine production and cell proliferation which can lead to septic shock. The aim of the present work was to identify the cytokines and co-stimulatory molecules regulating the in vivo systemic release of IFN-gamma, a cytokine known to play an important role in the pathophysiology associated with bacterial infections. We demonstrate in this study that (i) in contrast to lipopolysaccharide (LPS), SEB administration induces high levels of the p70, bioactive form, of IL-12; (ii) IL-12 production in response to SEB requires both CD40-dependent signals and IFN-gamma secretion; (iii) the early systemic release of IFN-gamma (3 h post-treatment) in response to SEB is IL-12 independent, while the sustained, late response (6-9 h post-treatment) requires endogenous IL-12 production; (iv) IL-12 produced during the primary SEB response (day 0) is responsible for priming cells in vivo to high IFN-gamma production upon secondary challenge (day 2); (v) the priming effect of IL-12 is TCR unrelated, as SEB-primed animals secrete high levels of IFN-gamma in response to both staphylococcal enterotoxin A and LPS administered 48 h later. The ability of bacterial SAg to induce septic shock and to modulate the immune response to unrelated antigens may therefore be related to their unique capacity to induce systemic IL-12 production in vivo. These observations also help to explain why SEB-primed animals, known to express an anergic phenotype 48 h post-treatment (as judged by defective IL-2 production and proliferation), nevertheless display an increased capacity to secrete the inflammatory cytokine IFN-gamma.

Carbohydrate-bearing cell surface receptors involved in innate immunity: interleukin-12 induction by mitogenic and nonmitogenic lectins.

Based on the observation that pathogen-derived lectins play an important role in cell adhesion and invasion, we examined the possible role of host carbohydrate-bearing molecules in inducing the secretion of IL-12, a crucial proinflammatory cytokine. The ability of 12 plant lectins to recognize and stimulate naive murine mononuclear cells in vitro has been characterized in this study. Mitogenic lectins (comprising Con A, PHA, PSA, and LCA) were found to induce the secretion of multiple cytokines in vitro, including IL-2, interferon (IFN)-gamma, and IL-12. Of interest, WGA, a nonmitogenic lectin unable to promote IL-2 secretion, was found to induce IL-12 and IFN-gamma production in a T and B cell-independent fashion. The functional properties of WGA were inhibited by N-acetylneuraminic acid and N,N'-diacetylchitobiose. WGA therefore represents a potentially useful tool for the study of membrane glycoproteins involved in the early proinflammatory response characteristic of innate immunity.

Revisiting the Th1/Th2 paradigm.

The identification of subsets of CD4+ helper cells producing distinct pattern of cytokines has provided a valuable framework for understanding how different effector populations of immune cells can be recruited in vivo during infection. In the view of most investigators, Th1 and Th2 cells produce factors that serve as their own autocrine factors and cytokines exerting suppressive activities on each other's development and activity. This concept intuitively explains the natural tendency of immune responses to become progressively polarized. However, several experimental observations appear difficult to rationalize with a simple, 'symmetrical' Th1/Th2 paradigm including those that Th1 cells do not produce their own growth factor; that both Th1 and Th2 cells can promote inflammatory responses; that interleukin-10 (IL-10) inhibits inflammatory responses in a Th1/Th2-independent fashion; that IL-10 promotes the development of Th1-type effector cells; and that IL-12 can amplify pre-established Th2 responses. The purpose of the present analysis is to provide a revised model for better understanding how cytokines regulate immune responses in vivo.

Staphylococcal enterotoxin B induces an early and transient state of immunosuppression characterized by V beta-unrestricted T cell unresponsiveness and defective antigen-presenting cell functions.

Staphylococcal enterotoxin B (SEB) is a bacterial enterotoxin able to simultaneously bind to class II molecules on APCs and to selected V beta regions (including V beta 8) of the TCR complex. Administration of SEB to adult BALB/c mice results in clonal activation of T cells bearing V beta 8 receptors, leading to an excessive release of proinflammatory cytokines. This initial immune response is followed by a long-lasting state of V beta 8-specific unresponsiveness, thought to benefit both the host (as it contributes to the down-regulation of the inflammatory response) and the bacterium (through ligand-specific T cell anergy). However, it is not clear how this type of restricted unresponsiveness can effectively impair the generation of an antibacterial response. To gain insight into the mechanism by which Gram-positive bacteria subvert the host immune response, we have investigated the immune competence of SEB-treated mice 48 h following SEB administration. We demonstrate in this report that in vivo, SEB induces a transient but profound state of unresponsiveness affecting both T and Ag-presenting cell functions. Although in vivo activation by SEB appears to be V beta-restricted under our experimental conditions, SEB-treated mice displayed an early (lasting 48 to 72 h postinjection) and V beta-unrestricted unresponsive state characterized by the inability to produce IL-2 in response to polyclonal TCR mitogens including third party bacterial superantigens (staphylococcal enterotoxin A and toxic shock syndrome toxin 1, SEA and TSST-1, respectively), Abs to non-SEB reactive V beta regions (V beta 6), anti-CD3 epsilon Abs, and a lectin (Con A). Spleen cell populations from SEB-treated mice also displayed defective APC functions, possibly related to a selective decrease in splenic dendritic cells numbers. Taken together, these observations indicate that SEB induces an early and transient state of immunodeficiency in vivo, representing a potential mechanism for escaping host immune surveillance.

Toxicity and neuroendocrine regulation of the immune response: a model analysis.

Various models have been proposed for the regulation of the primary immune response. Most of the models focus on the ability of the immune system to control a multiplying pathogen, and take into account the cross-regulations between different immune components. In the present study, we integrate the immune system in the general physiology of the host and consider the interaction between the immune and neuroendocrine systems. In addition to pathogen growth and toxicity, our four-variable model takes into account the toxic consequences for the organism of the immune response itself, as well as a neuro-hormonal retro-control of this immune response. Formally, the dynamics of the model is first explored on the basis of a discrete caricature, with special emphasis on the role of the constitutive feedback loops for determining the essential dynamical behavior of the system. This logical analysis is then completed by a classical continuous approach using differential equations. From a biological point of view, our model accounts for four stable regimes which can be described as "pathogen elimination/organism healthy", "pathogen elimination/ organism death", "pathogen growth/organism death" and "chronic infection". The size of the basins of attraction of these different regimes varies as a function of some crucial parameters. Our model allows moreover to interpret the interplay between pathogen immunogenicity and neuro-hormonal feedback, the effects of stress on immunity and the toxic shock syndrome, in terms of transitions among the steady states.

Regulation of dendritic cell numbers and maturation by lipopolysaccharide in vivo.

Dendritic cells (DC) are described as "nature's adjuvant," since they have the capacity to sensitize T cells in vivo upon first encounter with the antigen. The potent accessory properties of DC appear to develop sequentially. In particular, the ability to process antigens and to sensitize native T cells develops in sequence, a process termed "maturation" that is well described in vitro. Here, we obtain evidence for maturation in vivo in response to the bacterial product lipopolysaccharide (LPS). Before LPS treatment, many DC are found at the margin between the red and white pulp. These cells lack the M342 and DEC-205 markers, but process soluble proteins effectively. 6 h after LPS, DC with the M342 and DEC-205 markers are found in increased numbers in the T cell areas. These cells have a reduced capacity to process proteins, but show increases in the B7 costimulator and T cell stimulatory capacity. 48 h after LPS, the number of DC in the spleen is reduced markedly. We interpret these findings to mean that LPS can cause DC in the marginal zone to mature and to migrate into and then out of the T cell areas.

Costimulation regulates the kinetics of interleukin-2 response to bacterial superantigens.

The aim of this study was to investigate the mechanisms by which B7-related costimulatory molecules (CD80, CD86) increase T-cell responsiveness to extracellular ligands. As a model study, the in vitro response of purified splenic CD4+ T cells to a bacterial superantigen, SEB, was characterized. Previous analysis of this experimental model led us to conclude that expression of B7-related molecules is strictly required in order to activate CD4+ T cells in the presence of bacterial superantigens. In the present report, we demonstrate that antigen-presenting cell-derived costimulatory signals regulate the kinetics of interleukin-2 (IL-2) production by SEB-activated splenic CD4+ T cells. Indeed, experiments performed with purified subpopulations of antigen-presenting cells and using B7-transfected cell lines indicated that increased levels of CD80 and/or CD86 cell surface expression is associated with a faster kinetics of IL-2 production in response to SEB. Accordingly, blocking of CD80 or CD86-derived signals by specific monoclonal antibodies led to a slower kinetics of IL-2 production in response to SEB. Thus these data demonstrate that similar strength of signal through the T-cell receptor can lead to immune responses displaying distinct kinetics depending on the level of costimulatory ligands on APC.

Glucocorticoids down-regulate dendritic cell function in vitro and in vivo.

Exogenous glucocorticoid hormones are widely used as therapeutical agents, whereas endogenous glucocorticoids may act as physiological immunosuppressants involved in the control of immune and inflammatory responses. The optimal activation of T lymphocytes requires two distinct signals: the major histocompatibility complex-restricted presentation of the antigen and an additional co-stimulatory signal provided by the antigen-presenting cells. There is ample evidence that, among the cells able to present the antigen, the dendritic cells (DC) have the unique property to activate antigen-specific, naive T cells in vitro and in vivo, and are therefore required for the induction of primary immune responses. In this work, we tested whether glucocorticoids affected the capacity of DC to sensitize naive T cells. Our data show that, in vitro, the steroid hormone analog dexamethasone (Dex) affects the viability of DC, selectively down-regulates the expression of co-stimulatory molecules on viable DC, and strongly reduces their immunostimulatory properties. In vivo, a single injection of Dex results in impaired antigen presenting function, a finding which correlates with reduced numbers of splenic DC. These results show that glucocorticoids regulate DC maturation and immune function in vitro and in vivo and suggest that this mechanism may play a role in preventing overstimulation of the immune system.

B7.2 provides co-stimulatory functions in vivo in response to staphylococcal enterotoxin B.

Excessive T cell activation induced by bacterial superantigens plays an important role in the pathology associated with Gram-positive bacteremia. To gain insight into the early phases of T cell activation by bacterial enterotoxins in vivo, we investigated the ability of antibodies to well-defined co-stimulatory molecules to inhibit T cell activation and the subsequent toxic shock syndrome induced in BALB/c mice following the injection of staphylococcal enterotoxin B (SEB). We demonstrate here that a single dose of anti-B7.2 antibodies, but not anti-B7.1 antibodies, significantly inhibits T cell activation, as judged by lower systemic IL-2 release, blastogenesis and IL-2 receptor expression, and reduces the lethal effect of SEB in D-galactosamine-sensitized mice. These results demonstrate that co-stimulation through the B7.2 molecule plays an important role in the activation of T cells in response to SEB in vivo and suggest alternative therapies for septic shock caused by bacterial enterotoxins based on blocking antibodies to co-stimulatory molecules.

Activation of murine T cells by bacterial superantigens requires B7-mediated costimulation.

Staphylococcus enterotoxins bind class II MHC molecules on antigen-presenting cells (APC) and stimulate T cells expressing appropriate V beta gene products. Although the role of non-TcR-associated costimulatory receptors during antigen-specific T cell stimulation has been clearly established, the involvement of costimulatory activity in T cell activation by superantigens (SAgs) has been the matter of controversy. The aim of this study was to evaluate the role of the costimulatory-receptor ligand molecules CD28/B7 on bacterial SAg-mediated activation of naive murine T cells. We demonstrate in this report that a combination of monoclonal antibodies to murine B7.1 and B7.2 molecules inhibits the in vitro response of naive T cells to SAgs SEA, SEB, and TSST-1. The inhibition of T cell responses required simultaneous blocking of B7.1 and B7.2, suggesting that either B7.1 or B7.2 is sufficient to provide costimulatory signals to naive T cells in response to bacterial exotoxins. Inhibition of T cell activation by antibodies to B7-related molecules can be overcome by antibodies to CD28, a finding in agreement with the hypothesis that CD28-mediated signals participate in T cell activation by bacterial SAgs. These observations suggest that, as demonstrated for conventional antigen, T cell activation by SAgs requires the coordinated participation of TcR- and CD28-derived signals.

Co-stimulation lowers the threshold for activation of naive T cells by bacterial superantigens.

Staphylococcus enterotoxins bind class II MHC molecules on antigen presenting cells (APC) and stimulate T cells expressing appropriate V beta gene products. Although the role of non-TCR associated co-stimulatory receptors during antigen-specific T cell stimulation has been clearly established, the involvement of co-stimulatory activity in T cell activation by superantigens has been the matter of controversy. In this report, we examine the role of co-stimulation provided by selected APC populations in the response to bacterial exotoxins (staphylococcal enterotoxin A, staphylococcal enterotoxin B and toxic shock syndrome type 1). We demonstrate that the APC population able to activate naive T cells to IL-2 production is heterogeneous, comprising both adherent (presumably dendritic) and non-adherent (mostly B lymphocytes) cells. By stimulating naive T cells in the presence of graded doses of superantigens, we have observed that half-maximal IL-2 production was achieved at lower doses of superantigens in the presence of dendritic cells. Similarly, addition of antibodies to CD28 or B7.1-transfected cell lines increased the sensitivity of naive T cells to lower doses of superantigens. These observations indicate therefore that superantigens can be presented to naive T cells by APC displaying distinct levels of co-stimulatory activity, although with different efficacy. Thus, naive T cells are sensitive to CD28-mediated co-stimulation during superantigen-mediated responses but IL-2 production can be induced by high doses of superantigens in the presence of APC expressing weak co-stimulatory activity. These observations are compatible with the hypothesis that CD28-mediated signals participate in T cell activation by lowering T cell sensitivity to TCR ligands.