Homepage bandeau_genéral

  Director : Guy BORDENAVE (gbordena@pasteur.fr)



Cells of the immune system talk to each other by means of cell surface contacts. Cell surface receptors involved in these interactions are able to transduce informations inside the cells and to provoke the appropriate reaction. We designed systems (essentially with T and B lymphocytes) which allow to mimic some of these situations and to study the biochemical and genetical mechanisms underlying these relationships.



Activation of T-cells in presence of CSA induces the expression of a novel genetic program (Laurent Mascarell & Paolo Truffa-Bachi).

The discovery of the immunosuppressive property of CSA and its successful use in organ transplantation were important hallmarks in health care. CSA/cyclophilin complexes bind to calcineurin and inhibit the phosphatase activity of this enzyme. The cytosolic component of the nuclear factor of activated T-lymphocytes (NFAT) is not anymore dephosphorylated by calcineurin, preventing its translocation into the nucleus. Thus, the transcription of a set of genes possessing NFAT binding regulatory regions, such as IL2 and other cytokines participating in T-cell clonal expansion and differentiation, does not occur. CSA does not block the generation of AP-1, a transactivator that couples extracellular signals to gene-activation events in a developmental and tissue-specific manner. We hypothesized that the immunosuppressor may not interfere with the expression of genes regulated by AP-1 and we have asked whether activation of T lymphocytes in presence of CSA could modify the transcription of a set of inducible genes.

We used the proteomic approach, labeling new synthesized peptide with 35S-methionine. This study was a collaborative project between our group and the group of Dr. Lefkovits at the Roche Center for Medical Genomics, formerly the Basel Institute for Immunology. The proteomic technique is not restricted to the analysis of one or few cellular products but allows for qualitative and a quantitative measurement of several hundreds of proteins synthesized during a given interval and thus reflects the transcriptional and translational potential of the cells under various conditions. The comparison of the 2D gels from T-cells activated by ConA in presence or absence of CSA at different times showed that some polypeptides were made only in ConA-activated cells and that were common to both groups of cells. Strikingly, when activation was carried out in presence of CSA several radiolabeled polypeptides (>200), not found in ConA-activated T cells nor in non-activated cells, were detected supporting our hypothesis that CSA modifies the genetic program of LT activation.

The inhibitory effects of CSA on T cell activation processes are well documented; nevertheless, the outcome of CSA removal has received little attention despite the fact that discontinuation of the immunosuppressive treatment results in a rapid graft rejection. We have studied the effect of CSA removal by proteomic analyses. Around 80 polypeptides were synthesized de novo following CSA removal confirming that T cells, triggered in presence of CSA, have acquired a new program of gene expression.

In conclusion, we have shown that, in addition to its well-documented inhibitory effect on genes encoding for cytokines, CSA profoundly affects the genetic program of T-cell activation. T-cells triggered in presence of the immunosuppressor do not proliferate; nevertheless, they differentiate into cells expressing surface markers important in cell cooperation and migration. CSA promotes the synthesis of numerous proteins that are not found in resting or activated T-cells indicating that the immunosuppressor must act at the level of a class of transactivators that regulate the expression of the corresponding genes. Whether the changes in polypeptide representation indicate a spontaneous reshuffling of cellular differentiation functions or just a general shift towards a new cellular homeostasis remains to be assessed.

A molecule of vegetal origin, urtica dioica agglutinin, acts as a superantigen on mouse T-cells (Paula Rovira & Paolo Truffa-Bachi).

Although the physiological function of vegetal lectins is poorly understood, these proteins have been powerful tools for the analysis of T-cell activation. During the analysis of various lectins, we found that one of these molecules, Urtica dioica agglutinin (UDA) was a superantigen that triggers only the sub-population of T cells carrying the Vß8.3 domains of ß-chain of the TCR. Since the disclosing of UDA superantigenic activity, we have progressed in the understanding of its mode of action. The capacity of being presented by both class I and class II MHC molecules made of UDA a unique superantigen. In addition, we demonstrated that binding to MHC and TCR Vß chain was afforded by the recognition of oligosaccharides on these molecules. The tertiary structure of UDA was carried as a collaborative project with the group of Graham Bentley in our department. UDA crystals and their structures were obtained in free and oligosaccharide-complexes. These structures gave a molecular explanation of how a lectin, devoid of particular specificity towards two molecules mandatory for the immune response, could act as a superantigen. Vb8.3 chains possess a glycan residue, not found on other Vß, located in a position that allows UDA to bind simultaneously the TCR and class I (or class II) MHC molecules. The relative positioning of the MHC and Vb8.3 TCR molecules is indispensable to ensure T-cell activation and proliferation.

Studies on the interactions of LPSs from bacteria of different genera, and growing in different biotopes, with cells of the immune system. (group conducted by Robert Girard).


The knowledge of the first steps of the interaction of LPS with cells of the immune system, and of other tissues (pulmonary, intestinal) represents a major step for understanding the various and pleiotropic effects (septic shock being the most dramatic of them) triggered by these amphiphilic molecules. We have shown that LPSs from different bacterial strains (including Salmonella enterica subsp. enterica and Bordetella spp.) can induce, at the surface of mouse or human bone marrow granulocytes, the expression of receptors with high affinity for Lipid A. These receptors, identified as CD14, exhibit a marked specificity for various bacterial constituents such as LPSs, peptidoglycans, and lipophosphoglycans, and are distinct from constitutive receptors. This induction phenomenon is under the control of the lps locus present on mouse chromosome 4. Recently, Beutler's group described the sequence of a gene in this locus, coding for a transmembrane protein termed TLR4, belonging to the family of Toll-like receptors (TLRs).In line with our initial observation, we continued to study the constitutive receptor(s) of LPS, and the signal transduction pathways involved in the expression of CD14 at the surface of mouse bone marrow granulocytes.

Constitutive LPS receptor(s).

We have shown previously that the interaction of LPS with receptors expressed on the surface of bone marrow granulocytes was inhibited by the synthetic glycolipid PPDm2. This specific interaction is reversible, saturable, and time- and temperature-dependent. Using the novel compound PPDm2-B, a synthetic analogue of lipid A prepared by the group of R. CHABY, which specifically inhibits the induction of CD14, we have been able to prepare a new photoactivatable compound, radiolabeled with 125I. With this reagent, we carried out experiments in which cytoplasts of bone marrow cells expressing the Ly6-G marker were radiolabeled. Analyses by SDS-PAGE, and 2 D gels, indicated that several proteins (about 70, 30, and 20 kDa) were radiolabeled, and that labeling could be partially inhibited in the presence of LPS. By centrifugation through a saccharose gradient, the 30 kDa protein was enriched in a fraction of density 1.12. The 30 kDa protein was also found in membranes of cytoplasts from LPS-unresponsive mice. It is noteworthy that the molecular mass of this protein is similar to that of the recently described MD-2 molecule which, in association with TLR-4, may represent the LPS sensor in some macrophage cell lines. Definite identification of these radiolabeled molecules should allow us to define the constitution of the LPS-receptor present on the surface of mouse bone marrow granulocytes, and to consider to carry out a comparative study on receptors present on more differentiated cell types, or on different cell types (B- and pre-B lymphocytes).

Our experimental results indicated that LPSs from Rhizobium spp. can induce the expression of CD14 in bone marrow granulocytes of LPS-unresponsive mouse strains. This observation indicates that the lps locus is not involved, and that other TLRs, different from TLR4, can take part in the detection of structurally atypical lipid A molecules.

2. Transduction signals involved during the induction of CD14.

We have shown previously that the LPS-induced expression of CD14 requires the participation of protein tyrosine kinase, and the MAP kinase p38, but not of protein kinase C. However, other serine/thréonine protein kinases are certainly involved, as shown by the decreased response of the granulocytes to LPS in the presence K252a and KT-5823, which are known inhibitors of this family of kinases. We found that pre-treatment of the bone marrow granulocytes with phorbol myristate acetate (PMA) completely blocks the induction of CD14 by LPS. The unmodified binding of radiolabeled LPS to cells pre-treated with PMA indicated that the phorbol ester triggers a biochemical event which follows the interaction of LPS with its constitutive receptor. The observation that a specific activator of the PKC-a isoform mimics the inhibitor effect of PMA, whereas a specific inhibitor of this isoform antagonizes this effect, suggests a regulatory role of PKC-a in the signaling pathway triggered by LPS.

Cellular and molecular mechanisms of the T-cell-induced suppression of mouse IgG2ab expression and of its T-cell- immune tolerance (group conducted by Guy Bordenave).


We previously observed that, in the absence of any preliminary stimulation or immunization, T splenocytes from normal Igha mice with different genetic backgrounds — BALB/c (H-2d), BC8 (H-2b) strains — possessed a constitutive inhibitory activity against the expression of IgG2ab (IgG2a from mice of the Ighb haplotype, prototype C57Bl/6 strain). This intrinsic anti-IgG2ab TL activity of normal Igha mice can be enhanced in vivo by one or two i.v. injections of, for instance, B splenocytes from an Ighb congenic strain. Such an operation is termed "sensitization" and the resulting mice and T lymphocytes (L) are called sensitized mice and T1sens or T2sens respectively, according to the number of sensitizations. With Tsens injected to Igha/b or Ighb/b newborns (prepared by mating Igh congenic strains, which thus are histocompatible) we are able to induce a total, specific and chronic (but experimentally reversible) suppression of IgG2ab expression in 100 % of the recipients. As the suppression is strictly restricted to the IgG2ab allotype, we can say that it operates downstream the allelic exclusion and the isotypic switch leading to IgG2ab producing plasmocytes. The suppression is transferable into histocompatible Igha/b newborns by means of T splenocytes from mice subjected to the suppression.

One of the advantages of our model is that it deals with an unambiguous case of T-cell-mediated downregulation of immunoglobulin production. With this in vivo experimental system we are thus able to interrupt or provoke, at will, the production of large amounts of an immunoglobulin. Consequently it allowed us and continues to allow us a fine dissection of the partners, LT, LB taking part in a negative regulation of the immune system. More recently it offers a way to study problems pertaining to the immune tolerance.

For instance, we established that the suppression induction, in the Tsens newborn recipients, required the cooperation between CD4+ and CD8+ TL which had to possess the anti-IgG2ab specificity. Nevertheless, CD8+ TL alone were essential for suppression maintenance. During the sensitization phase, in Igha mice, CD4+ helper TL were indispensable to obtain a full recruitment of anti-IgG2ab cytotoxic TL.

Role of the CD40 molecule in the suppression of IgG2ab expression.

A new model called "sequential cellular interactions" recently emerges to try to explain the CD4+-dependent generation of CD8+ cytotoxic TL (CTL). The CD4+ TL help for CD8+ CTL responses was shown to be mediated by the intermediary of an antigen presenting cell (in fact dendritic cells : DC) carrying both MHC class I and MHC class II components able to present antigen derived peptides to CD8+ TL and CD4+ TL, respectively. The involvement of CD40 molecule on DC by CD40 ligand on activated CD4+ TL would trigger the DC which, in turn, would acquire the capacity to stimulate CD8+ TL directed against the same antigenic determinant and to lead them to the state of cytotoxic effectors. For instance, in the absence of CD4+ TL, their help for CD8+ TL could be reconstituted by the action of agonistic anti-CD40 antibodies. As the suppression of IgG2ab expression provides a good example of CD4+ TL-dependent generation of CD8+ CTL, we wondered whether CD4+ TL-depleted Igha mice, subjected to an in vivo agonistic anti-CD40 treatment, and of course sensitized against Ighb congenic BL, were able to recruit regulatory CD8+ TL with potential to inhibit IgG2ab production in Igha/b recipients. We observed that it was the case compared to CD4+ TL-depleted Igha mice which did not suffer the anti-CD40 injections. Nevertheless, despite the significant ability to negatively regulate IgG2ab production, the anti-IgG2ab CD8+ CTL recruited by CD40 stimulation in CD4+ TL-depleted Igha mice, failed to induce full suppression in all Igha/b recipients as if they were less effective than anti-IgG2ab CD8+ CTL normally obtained in the presence of CD4+ TL. This incomplete CD4+ TL help replacement by in vivo CD40 triggering could reflect the need for a parallel CD40-independent DC stimulation or direct CD4+-CD8+ TL communication for the priming of CD8+ CTL. Nevertheless, our data provide evidence that agonistic anti-CD40 treatment helps in recruiting CD8+ CTL implicated in immune regulatory functions.

We formerly showed that IgG2ab+ target BL were lysed by anti- IgG2ab TL which, alternatively, or concomitantly, used perforin or Fas cell-death pathway. By employing pore-forming protein (perforin) "knockout" Igha mice (Pfp-/-) we established that this suppression could be fully induced by only the Fas mediated cytotoxic mechanism. As Fas is not expressed on the surface of mature BL, its up-regulation which can occur particularly via CD40 triggering, would constitute a prerequisite for induction and maintenance of IgG2ab suppression by means of the Fas pathway. Thus our model of suppression of IgG2ab expression also furnishes the opportunity to investigate the possible involvement of CD40 expression on Ighb BL when the Igha CD8+ TL effectors operate exclusively by the Fas pathway. We demonstrated by utilizing Pfp-/- Igha mice (exclusively operating through the Fas pathway) and CD40 "knockout" (CD40-/-) Ighb mice that Pfp-/- Igha TL failed to establish full suppression against CD40-/- Ighb BL while they were able to totally induce it against CD40+/+ BL. One can imagine the existence of two kinds of IgG2ab-producing BL : i) a highly represented population, readily susceptible to Fas induced suppression, which up-regulates Fas through a CD40 independent mechanism, and ii) a weakly represented population, chronically escaping Fas-induced suppression in the absence of CD40-mediated Fas up-regulation. There is obviously a direct involvement of BL CD40 expression in in vivo negative control of an immunoglobulin production.

From another point of view we also observed that transfer of CD40+/+ and mainly CD40-/- mature peripheral Ighb BL into RAG2 "knockout" mice (RAG2-/-), the RAG2 gene governs the recombination of the genetic elements of antibodies and TCR and its absence results in mice without T and B cell compartments, led into these LB to an enhanced switch recombination to IgG2ab. For CD40-/- Ighb BL, the order of magnitude of frequency and absolute number of IgG2ab-producing cells was increased around 500 times when transferred into a RAG2-/- environment.


puce Publications of the unit on Pasteur's references database


  Office staff Researchers Scientific trainees Other personnel

Madame THEPAUT Sylvana, I.P.

CHABY Richard DR2 CNRS (à temps très partiel)

GIRARD Robert M.C. University Paris VII



MASCARELL Laurent PhD student




PEDRON Thierry I. I.P.



Page Top research Institut Pasteur homepage

If you have problems with this Web page, please write to rescom@pasteur.fr.