HLA class II signaling mediates cellular activation and programmed cell death.

The class II human leucocyte antigens (HLA class II) are principally peptide presentation molecules. Signal transduction by these molecules has also been shown to transmit activation signals in both B and T lymphocytes by a pathway including protein tyrosine kinase activation, an intracellular calcium flux, and both the activation and transcriptional regulation of protein kinase C (PKC) isoforms. Apoptosis can also result from human leukocyte antigen class II stimulation. Inhibitors of gene transcription were used to inhibit activation and, therefore, to distinguish the signal transduction pathways important for apoptosis. This approach provided evidence that cellular activation and apoptosis undertook separate signaling pathways, and that PKC and intracellular calcium were shared between the two pathways, while tyrosine kinase activity was essential for cell activation. Further studies using cord blood B cells showed that these cells were incapable of generating a calcium flux after HLA class II ligation and were not subject to cell death. The importance of sustained levels of calcium for programmed cell death (PCD) was underlined since the restoration of a calcium flux enabled PCD of cord blood B cells via HLA class II. These results demonstrate that HLA class II stimulation initiates two distinct signal transduction pathways.

Role of the CD1a molecule in the superantigen-induced activation of MHC class II negative human thymocytes.

Bacterial superantigens (Sag) are potent activators of T cells. This T-cell activation has been described as an MHC class II dependent phenomenon. We have observed that human thymocytes depleted of MHC class II positive cells are still able to proliferate in response to the staphylococcal enterotoxin A (SEA). This proliferation was clearly inhibited by the addition of monoclonal antibodies directed against the CD1a molecule. In contrast, monoclonal antibodies directed against the CD1b and CD1c molecules have no effect on the Sag-induced activation of the CD2 (+) MHC class II (-) thymocytes. We next examined the ability of the CD1a molecule to transmit transmembrane signals. Results obtained indicate that CD1a ligation on these thymocytes induced tyrosine phosphorylation of the p56(lck) tyrosine kinase. Signal transduction via CD1a is further confirmed by the observation of a significant intracellular calcium flux (Ca(i)(++)) in thymocytes following CD1a engagement. These data demonstrate that CD1a ligation induces a signal transduction pathway which has a potential role in the bacterial superantigen-induced activation of human CD2 (+) MHC class II (-) thymocytes.

An in vitro model of allogeneic stimulation of cord blood: induction of Fas independent apoptosis.

Cord blood is increasingly used in transplantation as it is a readily available source of progenitor cells and is reputed to generate less severe graft-versus-host disease (GVHD) than adult bone marrow. We have compared apoptosis of cord blood lymphocytes (CB) and adult lymphocytes (PBMC) after stimulation via HLA class I, HLA class II or CD3 in order to reproduce in vitro some of the stimuli occurring after allotransplantation. CB spontaneously apoptose more than PBMC ex vivo, stimulation via HLA class I dramatically increased CB apoptosis without altering viability of PBMC. Expression of Fas was markedly lower on CB than on PBMC and this difference was maintained even after activation. Fas ligand was expressed in CB and in PBMC. CB were activated via either HLA class I or class II molecules although proliferation was not observed. Only phorbol ester pre-activation allowed Fas to subsequently induce a death signal. Proliferation of PBMC via CD3 led to enhanced Fas signals. CB therefore differ from PBMC with regard to both spontaneous and activation induced apoptosis and either allo- or CD3 mediated stimulation. Finally, the apoptosis of CB via HLA-class I could have an important role in the moderation of graft-versus-host disease.

Characterisation of the roles of CD95 and CD95 ligand in cord blood.

The increasing use of umbilical cord blood in transplantation has led to a renewed interest in the immunological characterisation of this material. This study addresses the question of whether the CD95 molecule and its ligand are expressed and are functional in mediating cell death in cord blood mononuclear cells. These molecules have a crucial role in the homeostasis of haematopoietic cell populations in the adult and also contribute to graft-versus-host disease. CD95 is the most well studied receptor mediating a signal for cell death by apoptosis and its inducible ligand has been demonstrated to mediate cell death of multiple types of CD95 expressing cells. The object of this study was to examine whether cord blood mononuclear cells could behave either as targets for CD95-mediated cell death or as mediators of cell death due to the expression of CD95L. The results of this study lead us to suggest that cord blood mononuclear cells enjoy some immunological privilege due to the relatively low level of expression of CD95 (in comparison with adult peripheral blood lymphocytes) and the expression of the CD95 ligand.

T-lymphocyte control of HLA-DR blood monocyte differentiation into neo-fibroblasts. Further evidence of pluripotential secreting functions of HLA-DR monocytes, involving not only collagen but also uromodulin, amyloid-beta peptide, alpha-fetoprotein and carcinoembryonic antigen.

Previous studies led us to demonstrate in pathological situations that the fibroblast, not the macrophage, was the terminal maturation step of the HLA-DR monocyte and that the entire process came under T-lymphocyte control. Fibrosis which developed under immunosuppressive treatment (cyclosporin) after organ transplantation is an illustration of these in vitro observations. The present in vitro study was undertaken in order to investigate whether or not this transformation process takes place under physiological conditions and if so, the nature of the T-lymphocyte control. We report that normal HLA-DR monocytes/macrophages are able to secrete type 1 collagen and to differentiate into neo-fibroblasts. However, contrarily to what happened in pathology, only a few neo-fibroblasts developed transiently. The addition of conditioned medium (CM) from activated T-lymphocytes greatly enhanced the transformation process. Counteracting this CM effect, cell-to-cell contact between neo-fibroblasts and T-cells resulted in the loss of fibroblastic shape. The 'end-result' macrophage engulfed numerous lymphocytes giving rise to a multinucleated cell. This giant cell no longer adhered to the slide and died. The question is raised as to whether the process observed in vitro is involved in vivo in tissue repair. We also report that HLA-DR monocytes and the neo-fibroblasts which derive from them are able to secrete, in addition to type 1 collagen, a variety of proteins such as uromodulin, amyloid-beta peptide, alpha-fetoprotein and carcinoembryonic antigen. In cystic fibrosis we previously reported a high level of uromodulin production by HLA-DR monocytes differentiating towards the fibroblastic phenotype. Pathologies characterized by excessive production of either alpha-feto-protein, carcinoembryonic antigen, beta-amyloid protein (Alzheimer's disease) should be investigated, taking into account the involvement of HLA-DR monocytes and their possible uncontrolled differentiation into neo-fibroblasts.

Possible monocytic origin of chondrosarcoma: in vitro transdifferentiation of HLA-DR blood monocyte-like cells from a patient with chondrosarcoma, into neo-fibroblasts and chondrocyte-like cells.

Nodules and multilayered areas composed of fibroblasts and chondrocyte-like cells embedded in an abundant extracellular matrix appeared spontaneously in in vitro culture of mononucleated blood cells taken from a patient with chondrosarcoma. Using specific antibodies it was demonstrated that the neo-fibroblasts which developed in the culture resulted from a direct transdifferentiation of monocytes expressing HLA-DR specificity. The experiment was carried out twice, once before surgery and then two years later. In both cases the spontaneous transdifferentiation of HLA-DR monocytes into neo-fibroblasts was observed. Previously it was shown that normal monocytes were also able to give rise in vitro to neo-fibroblasts. However, the latter are normally rapidly destroyed by cell-cell contact with T-cells. Normal T-cells adhere to normal neo-fibroblasts by which they are finally engulfed. As a result, the neo-fibroblasts lose their fibroblastic shape, no longer adhere to their support and die. Therefore the abnormal proliferation and persistence of neo-fibroblasts in pathological situations such as the present case may result either from an intrinsic defect in monocytes, T-cells or both. The question is whether or not this transdifferentiation process observed in vitro accounts for the development of chondrosarcoma in vivo. The present results suggest that in vivo chondrosarcoma may start in a necrotic zone (resulting for instance from trauma) and attract HLA-DR monocytes, where they accumulate and transdifferentiate into neo-fibroblasts and chondrocyte-like cells. The uncontrolled transdifferentiation of these HLA-DR monocytes resulting from a dysregulation of the immune system is probably linked to the malignant process which may have a retroviral origin. The question is raised regarding the embryologic origin of this special sub-population of blood monocytes in which pluripotential capabilities are retained; its origin may differ from that of the other circulating monocytes.

Cystic fibrosis: production of high levels of uromodulin-like protein by HLA-DR blood monocytes differentiating towards a fibroblastic phenotype.

We report here the spontaneous in vitro transformation of blood monocytes into fibroblasts, in a patient suffering from cystic fibrosis (CF). The blood monocytes with this capacity express HLA-DR specificity. Monocytes were identified by non-specific esterase activity and by immunofluorescence using monoclonal antibodies against monocytes/macrophages antigens. Neo-fibroblasts were identified by electron microscopy and immunofluorescence using monoclonal antibodies against a cytoplasmic enzyme specifically involved in the synthesis of collagen. The secretion of collagen was evidenced using antibodies against type I collagen. Both monocytes/macrophages and neo-fibroblasts express the monocytic and the fibroblastic markers and synthesize type I collagen. This transformation observed in vitro might mimick the process of fibrosis development which takes place in vivo, particularly in pancreatic acini, lungs and intestine of patients with CF. Interestingly, the whole process in vitro is inhibited when T lymphocytes are properly stimulated by IL2. In addition, both monocytes and neo-fibroblasts secrete high quantities of uromodulin-like glycoprotein. The significance of this finding is discussed in relation to the thick mucus secretion which characterizes the disease. In addition, from a fundamental point of view, it confirmed in a large series of patients that this observation may have significant implications, since CF mutation impairs the gene coding for cAMP-regulated Cl- channel and that it has been proposed that uromodulin might be implicated in Cl- transport. Therefore the question of the relationships between uromodulin and the cAMP-regulated Cl- channel arises.

HLA class II molecule signal transduction leads to either apoptosis or activation via two different pathways.

The class II human leukocyte antigens (HLA class II) are constitutively expressed on antigen presenting cells (APC) and are essential for peptide presentation to helper T lymphocytes. Signal transduction by HLA class II molecules on B lymphocytes has been described and has been shown in many cases to induce cellular proliferation. However, since signalling via HLA class II can also lead to apoptosis, it has not been clear how the outcome of the signals is determined. We have distinguished two separate HLA class II-initiated pathways leading to either proliferation or apoptosis of primary human B lymphocytes. Proliferation requires new gene transcription and activation of src family tyrosine kinases. In contrast, apoptosis is significantly increased in the absence of transcription/translation. It is dependent on serine/threonine phosphatases and cytoskeletal mobility. An extracellular source of calcium was essential for apoptosis, suggesting the need for sustained high level of intracellular calcium. Activation of iso-enzymes of the protein kinase C family was needed for both pathways. We therefore conclude that HLA class II molecules can initiate two distinct signalling pathways leading to either proliferation or apoptosis of APC.

HLA class II-mediated death is induced via Fas/Fas ligand interactions in human splenic B lymphocytes.

HLA class II molecules, expressed on the surface of antigen-presenting cells, are responsible for the presentation of antigen-derived peptides to CD4+ helper T lymphocytes. Signaling via these molecules initiates the generation of second messengers leading to programed cell death (PCD) of activated B lymphocytes. The present study examined the mechanism of HLA class II-mediated apoptosis and describes the essential role of the molecule Fas and its ligand (FasL). FasL was expressed in B lymphocytes after stimulation via HLA class II or with phorbol esters. Expression of FasL protein was significantly increased in 50% of B lymphocytes after stimulation via HLA class II, and the level of FasL mRNA was also increased either by activation with phorbol esters and ionomycin or by signaling via HLA class II. Although HLA class II signaling did not change the expression of the Fas molecule, it did lead to increased sensitivity to Fas-mediated apoptosis. The crucial role of Fas/FasL interactions was confirmed by the absence of cell death via HLA class II in B cells lacking Fas expression, and by the significant inhibition of HLA class II-mediated apoptosis in the presence of either an antagonistic anti-Fas or anti-FasL antibody. These data demonstrate FasL expression on activated human B lymphocytes and support the idea that antigen presentation could contribute to the regulation of lymphocyte populations via Fas and FasL interactions.