Homepage bandeau_genéral

  Director : CUMANO Ana (cumano@pasteur.fr)



Abstract: The Unit for Lymphocyte Development aims at the understanding of the events leading to the establishment of a functional immune system. We carry on three main topics of research: 1. The establishment of the hematopoietic system during embryogenesis; 2. The development and physiology of a subset of T lymphocytes, gd T cells; 3. Ontogenesis, repertoire and physiology of regulatory T cells, important in the control of the homeostasis of the immune system and involved in the protection from pathological autoimmunity.



The establishment of embryonic hematopoiesis, in the mouse embryo. A. Cumano.

Hematopoietic stem cells.

Hematopoietic stem cells (HSCs) are the key elements responsible for the maintenance of blood cell formation throughout life. It has long been accepted that HCSs or their immediate precursors are not generated in the hematopoietic organs (fetal liver, bone marrow, thymus and more transiently the spleen) but have an exogenous origin.

In mouse embryonic development, the yolk sac (YS) comprises mesoderm and primitive endoderm, while the embryo proper is composed of the three layers formed during gastrulation. The aorta first develops from the intraembryonic mesoderm called Splanchnopleura (Sp), primordiums of the mesonephros, mesentery and gonads soon become apparent, this region has thus been named AGM (Aorta, Gonads, Mesonephros).

The first hematopoietic cells identifiable during mouse development appear within the YS-blood islands. Most of these cells belong to the primitive erythroid lineage. We have identified the site of origin of adult hematopoietic precursors, in the mouse embryo. We showed that hematopoietic intraembryonic precursors could be detected in the Sp/AGM of embryos isolated from the YS before the establishment of circulation; these precursors are thus generated in situ. Interestingly, in contrast with Sp/AGM, YS was consistently unable to give rise to a lymphoid progeny.

These experiments indicate that hematopoietic cells originate in two independent locations, in the mouse embryo. The first to develop is the YS apparently oriented to a fast production of erythrocytes. The second set of hematopoietic precursors originates in the mesenchyme around the aorta. Similar intra-embryonic hemogenic sites have been identified in other species, namely Amphibia, Birds, Mice and Humans.

While Sp/AGM generates de novo hematopoietic stem cells between days 9.5-12.5 of gestation, we show that it does not support hematopoietic differentiation. Progenitors within this region thus constitute a pool of undifferentiated hematopoietic cells readily accessible for characterization.

We have previously reported that, before the onset of circulation, the mesoderm-derived precursors in the Sp/AGM or in the YS were unable to reconstitute the hematopoietic system of irradiated recipients. We used a comparative cell transfer scheme in which recipient mice are either deprived of T and B cells (Rag2-/-) or deficient also in NK cells (Rag2gc-/-). Six to eight months post-transplantation, donor derived multilineage reconstitution was observed in Rag2gc-/- mice injected with Sp/AGM explants. Mice injected with YS cells only showed short-term erythromyeloid reconstitution.

This result establishes, for the first time, that early intra-embryonic precursors may engraft in an adult environment and provide long-term multilineage reconstitution of the hematopoietic system. Hematopoietic precursors derived from the intra-embryonic hemogenic site, explanted before the circulation is established, thus qualified as HSC. In contrast, YS-derived cells appear devoid of such potential when tested in the same conditions, a result that shed doubts on the hypothesis of a second generation of LTR-endowed HSC by the YS.

T-cell development.

T cell generation occurs in the thymus from precursors present in fetal liver and adult bone marrow. The role of the thymic microenvironment as a unique site for the induction of commitment to the T-cell lineage, remains a matter of debate. HSCs as well as common lymphoid progenitors (CLP) had been purified from bone marrow. However, little is known about the intermediates that define irreversible commitment to the T cell lineage as such progenitors have not been identified in the major hematopoietic organs, fetal liver and bone marrow. Populations of precursors restricted to the T and NK cell lineages have however been reported in the fetal thymus and fetal blood. Initial studies indicated that increasing numbers of T-cell precursors exit the fetal liver between days 11 and 15 of gestation. We isolated a population, corresponding to 0.2% of fetal liver cells, that includes around 70% of T cell precursors in this organ. The presence of these cells in fetal liver from athymic embryos indicates their pre-thymic origin. Clonal analysis revealed that this population comprises T/NK bipotent cells. This population can reconstitute transiently not only TCRab+ and NK peripheral compartment but also the TCRgd and the TCRabCD8aa compartments in the intestinal epithelium of recipient mice. These two latter cell types are considered to be of extra-thymic origin. PCR and RT-PCR analysis of sorted cells show that they do not have D-Jb rearrangements and that they do not express detectable levels of Rag2 or pTa transcripts. However, they express GATA-3, a transcription factor associated with T cell development. This population designated as a common T/NK cell progenitor (C-TNKP), differs from previously described bipotent T/NK precursors present in fetal blood, spleen and thymus, both by surface markers and by gene expression analysis. We conclude that commitment to the T/NK pathway of differentiation occurs in central hematopoietic organs (fetal liver and bone marrow) and we proposed that the precursors identified here, represent the immediate developmental step before thymic immigration.

Physiology of gd T-cells. P. Pereira

1. The Thy-1dull gd cell population.

We have previously characterized the function, phenotype, ontogenic development and TCR repertoire of a population of gd thymocytes which exhibit properties equivalent to those of NK TCRab T cells. Initially defined by expressing low levels of Thy-1, this gd population produces high levels of several cytokines, including interferon- (IFN)-g, IL-4, IL-10 and IL-3, whereas Thy-1bright gd thymocytes produce only IFN-g. Virtually all Thy-1dull gd thymocytes express high levels of CD44 and low levels of the heat stable antigen (HSA) and of CD62 ligand (CD62L), whereas about half of them express the NK1.1 marker. Thy-1dull gd thymocytes are barely detectable in newborn animals, and their representation increases considerably during the first two weeks of postnatal life. Thy-1dull gd thymocytes from DBA/2 mice express majoritarily the Vg1 gene and a newly characterized Vd6 gene named Vd6.4 with highly restricted Vd-D-Jd junctions, and more diverse Vg-Jg junctions. Further experiments indicated that the homogeneous TCRd repertoire expressed by the Thy-1dull gd thymocytes results from cellular selection rather than by molecular mechanisms, suggesting the existence of a limited set of self-ligands.

In search of putative polymorphism in the self-ligand(s) selecting the homogeneous repertoire of TCRs expressed by Thy-1dull gd cells, we characterized this gd T cell population in a large number of laboratory mouse strains. Thy-1dull gd thymocytes were present in every strain tested, albeit at different frequencies. Moreover, the repertoire of TCRs expressed by Thy-1dull gd thymocytes varies in different strain of mice, from oligoclonal (i.e. DBA/2 mice) to polyclonal (i.e. FVB/N mice). (FVB/N x DBA/2)F1 animals showed an intermediate phenotype between the two parental strains with about half of their Thy-1dull gd thymocytes expressing VDJd and Vg1Jg4 junctions identical to those find in DBA/2 mice and the other half expressing a polyclonal repertoire of TCRs. These two subsets of cells may also be distinguished by their differential expression of cell surface markers. Thus, oligoclonal Thy-1dull gd thymocytes express higher levels of TCR, may express CD4 and do not express NK markers like Ly49 or DX5, whereas the polyclonal Thy-1dull gd thymocytes express higher levels of TCR, may express different NK markers and are CD4-negative.

The existence of two cell subsets that differ in the levels of diversity of their TCRs is indicative of strain-specific differences in the selection of the expressed TCRs. Such differences may be found in their TCRg and/or d chains or in the putative self-ligands selecting for the repertoire of TCRs of Thy-1dull gd thymocytes. Cloning and sequencing of PCR amplified Vg1 and Vd6 chains expressed by oligoclonal Thy-1dull gd thymocytes from (FVB/N x DBA/2)F1 animals showed roughly identical frequency of both parental alleles, strongly suggesting that polymorphism at the structural genes coding for the g and d chains are not responsible for the repertoire differences observed among these mouse strains. We, therefore, decided to identify the genetic elements determining the selection of the homogeneous TCR expressed by Thy-1dull gd thymocytes in a FVB/N x (FVB/N x DBA/2)F1 backcross. About 150 individual backcrossed animals were analyzed for the frequency of Thy-1dull gd cells among gd thymocytes, the g and d chains they used to form their TCRs and their frequency of expression of the CD4 marker. A map of the genome of each animal using SSLP markers is underway. From these experiments we hope to gain insights into the nature of the endogenous ligand(s) selecting the restricted repertoire of Thy-1dull gd thymocytes.

2. Selection of the repertoire of TCRs expressed by gd cells located in the intestinal epithelium of the small intestine: Allele-specific negative selection of Vg1/Vd4 i-IELs.

In the intestinal epithelium of the small intestine, gd cells constitute the largest population of peripheral gd lymphocytes and represent about half of the intraepithelial lymphocytes (i-IEL). Isolation of i-IELs is easy and, thus, gd i-IELs constitute an optimal population to analyze TCRgd-ligand interactions in vivo in non-immunized mice. We have been interested in the identification of genetic elements influencing the representation of different gd i-IEL subsets in normal mice. Our initial results analyzing V-gene usage by gd i-IELs in a panel of recombinant inbreed strains of B6 and DBA/2 origin (BXD RI), showed a complex genetic control with marked influences by genes closely linked to the structural genes for TCRg and TCRd and to the MHC locus

C57BL/10 (B10) and B6 are related strains that share about 90% of their genomic composition. In particular, they share the same TCRd and MHC haplotypes but differ in their TCRg haplotype. Analyses of gd i-IEL subsets in B6, B10, (B10 x B6)F1 hybrids and in a cohort of (B10 x B6)F2 animals showed that the representation of the Vg1/Vd4 subset was influence by gene(s) closely linked to the TCRg locus, as it was previously suggested by our analyses of BXD RI strains. (B10 x B6)F1 hybrid mice showed an intermediate phenotype indicative of a non-recessive model of action.

Formal evidence for an allele-specific selective event came from two experiments. First, introduction of a rearranged Vg1-Jg4 chain of B6 origin in B6 or in (B10 x B6)F1 mice results in a B6-like phenotype (i.e. low expression of Vd4+ cells among Vg1+ cells). Second, analyses of Vd4+ cells in Vg1+ i-IELs from (B10 x B6)F1 mice separated on the basis of the Vg1-Cg4 allele they express showed a B6-like phenotype in cells carrying the B6 allele and a B10-like phenotype in cells carrying the B10 allele. This experiments indicate that an allele-specific cellular selection is responsible for the different representation of Vg1+Vd4+ cells among gd i-IELs in B6 and B10 mice. Such allele-specific mehcanism could be an expansion (positive selection) of Vg1+Vd4+ cells in B10 mice or a deletion (negative selection) of Vg1+Vd4+ cells in B6 mice. To distinguish among these possibilities we studied the junctional diversity of the g and d chains expressed by Vg1+Vd4+ i-IELs in B10 and B6 mice. Our results showed a reduced junctional diversity of Vg1-Jg4 chains expressed by Vg1+Vd4+ i-IELs in B6 mice when compared to that present in B10 mice, strongly suggesting a deletion of most Vg1+Vd4+ i-IELs in B6 mice. Furthermore, such deletion appears to involve the Fas-FasL pathway as indicated by the recovery of normal numbers of Vg1+Vd4+ i-IELs in B6 mice deficient in the Fas molecule.

1. Ontogenesis and physiology of natural regulatory CD4 T lymphocytes. A. Bandeira

Regulatory CD4 T cells protect the integrity of tissues and organs against the eventual destructive activity of other T cells. This has been demonstrated in natural peripheral tolerance, in the regulation of immune responses against exogenous antigens and in transplantation tolerance.

All normal individuals contain potentially aggressive self-reactive T cells. Autoimmune diseases may result from a failure of the regulatory T cell system, thus explaining the frequent association of autoimmune pathology with certain forms of immunodeficiencies. Studies in an experimental system of autoimmune disease associated with chronic lymphopenia, following neonatal thymectomy of normal mice, revealed that normal CD4 T cells expressing the CD25 marker display protective activity.

Regulatory T cells also control immune responses against exogenous antigens. Transfers of naive CD4 T cells into T and B cell-deficient, syngeneic recipients lead to the development of inflammatory bowel disease (IBD), caused by massive Th1-mediated immune responses in the gut promoted by local antigens. The disease can be prevented by co-transferring naturally activated/memory, CD45RBlow CD4 T cells.

Natural regulatory T cells have not been isolated to the homogeneity. No specific marker for this functional population is yet available. They seemed to have poor capacity of expansion, maybe the result of its own functional activity. Little is known on the origin, repertoire and physiology of this population.

We investigated the link between the regulatory events implicated in peripheral tolerance and in mucosal immune responses, with the hypothesis that such events are at the core of the systemic homeostatic mechanisms that control total T cell (and B cell) numbers. Our results show that 1) the same pool of natural regulatory T cells (CD25+ CD4+) that controls peripheral tolerance, also regulates immune responses at the level of the mucosa and the systemic homeostasis of peripheral CD4 T cell pools, via a mechanism involving the production of IL-10; 2) a small fraction of these cells has a very high potential of expansion; 3) the expression of the CD25 molecule in vivo is not stable and requires the presence of other CD4 T cells; 4) CD25- CD45RBlow CD4 T cell pool also contains regulatory T cells capable of IBD protection but they do not efficiently regulate peripheral expansion of CD4 T cells; 5) regulatory CD4 T cells are present in germ-free mice, thus indicating that at least a fraction of them is self-reactive.

Thymic epithelium (TE) is capable of inducing full tolerance to a variety of tissues and organs, via a mechanism dependent on regulatory CD4 T cells. These studies revealed a major event in T cell development, namely the thymic origin of the first regulatory T cells, selected and activated into this effector function as the result of high avidity TCR/ligand interactions exclusively with TE. CD4 T cells expressing the CD25 marker are present in the thymus of normal mice. We showed that these T cells are capable of long-term protection of CD4 T cell-induced IBD in Rag-2° recipients. As their peripheral equivalents, they are capable of inhibiting peripheral expansion of CD4 T cells. Thus, at least part of the CD25+ CD4 T cells acquires their functional properties already in the thymus. Our studies also suggest that the major part of the peripheral compartment of the CD25+ CD4 T cells is dependent on thymic output, while peripheral expansion or induction of new regulatory T cells have a minor contribution to the maintenance of the pool. Moreover, the rate of production of these cells is similar between the newborn and the adult. We find thus no evidence for the existence of a special developmental program for the generation of these cells.

TCR repertoire analyses of thymic and peripheral CD25+ CD4 T cells in C57BL/6 mice revealed no particular bias in the TCR repertoire (with the exception of the Vb5 family) and showed a level of diversity equivalent to the one of naive CD4 T cells. This suggests that 1) the clonal size within this population is in general small; 2) these cells recognize a diverse set of natural ligands. These results reinforce the notion that regulatory T cells do not belong to a lineage distinct from the one of conventional ab T cells.

In cooperation with T. Laufer (University of Pennsylvania, Philadelphia, USA), we could show that 1) the development of CD25+ CD4 T cells with regulatory activity requires Class II molecules and occurs when MHC expression is restricted to the cortical TE. Also, the CD25+ CD4 T cell population of normal mice is purged of cells reactive to self-antigens presented by HCs, a conclusion also obtained from our own studies on natural deletion involving endogenous superantigens.

Two other aspects of the repertoire of the CD25+ CD4 T cells were establishment : 1) a fraction of these cells expands in vivo in conditions that imply the recognition of class I molécules. This reinforces the notion that their regulatory activity is also extended to the control of the activity of CD8 T cells ; 2) a fraction of these cells is capable to expand in allogeneic recipients without any obvious deleterious effect. This observation opens quite interesting perspectives in transplantation of MHC-mismatched tissues and organs.

2. Grafts of embryonic thymic epithelium (TE) naturally deprived of " professional " antigen-presenting cells are acutely rejected in adult allogeneic recipients.

According to the concept of the " passenger leukocyte ", HC APCs present in epithelial or parenchyma tissues trigger graft rejection. Grafts of tissues submitted to physical or chemical treatments resulting in the elimination of HC, survive longer or are not destroyed in allogeneic recipients. This was the case for example of grafts of embryonic thymic lobes of 14 days treated with 2-deoxi-guonosine (2-dGuo). The lack of co-receptors necessary to T cell activation reinforced the notion that epithelia and parenchyma per se are not immunogenic. Responding T cells either " ignore " them or are rendered anergic. Thus, grafts of embryonic tissues, not yet colonized by HCs, should not be rejected.

This is in fact not the case. Grafts of the third branchial pouches of day 10 embryos, which contain the thymic epithelium anlagen not yet colonized by HCs, are the target of T cell-mediated, acute rejection in adult allogeneic recipients. The heart primordium of E8 embryos (at this stage, circulation is not yet established) is also rejected. We could also show that the vast majority of the cells infiltrating the grafts were derived from the pool of activated/memory T cells, strongly suggesting that these cells are the major effectors the acute rejection of tissues deprived of professional APCs.

These observations are not readily accommodated in the " passenger leukocyte " concept, and raise concerns on the rational of tissue depletion of HC as a benefit in transplantation.

Furthermore, they imply that treatments that confer to tissues protection against rejection, as mentioned above must also directly affect the parenchyma. Re-evaluating the protective effect of 2-Guo on E14 thymic grafts against rejection, we showed that this drug is active on pure TE and is specific for this tissues since treated E10 and E14 hearts, as well as E14 intestines are always rejected. In contrast to previous conclusions, we could also show that 2-dGuo-treated E14 thymic lobes are immunogenic but trigger a non-destructive T cells immune response.


puce Publications of the unit on Pasteur's references database


  Office staff Researchers Scientific trainees Other personnel

VOUGNY Marie-Christine (mcvougny@pasteur.fr)

CUMANO Ana (DR Inserm) cumano@pasteur.fr

BANDEIRA Antonio (CR CNRS) bandeira@pasteur.fr

PEREIRA Pablo (Chef Labo IP) ppereira@pasteur.fr

VIEIRA Paulo (chercheur Associé CNRS, arrivé 2/2001 pveira@pasteur.fr

ANDRE Isabelle (post doc, partie 5/2000)

ANNACKER Oliver (PhD student, parti 12/2000)

ARAUJO Ricardo (PhD student) raraujo@pasteur.fr

BARRETO Vasco (PhD student, parti 12/2000)

BERTRAND Julien (DEA) jbertran@pasteur.fr

BRAUN Deborah (PhD student)

DIAS Sheila (PhD student) sdias@pasteur.fr

DOUAGI Iyadh (PhD student, parti avril 2001)

GOLUB Rachel (post-doc) rgolub@pasteur.fr

GRIGORIADOU Kalliopi (PhD student) kalliopi@pasteur.fr

BURLEN-DEFRANOUX Odile (Ingeneer IP) oburlen@pasteur.fr

BOUCONTET Laurent (Tech Sup Labo IP) bouconte@pasteur.fr

VOUGNY Marie-Christine (Sec Dir IP) mcvougny@pasteur.fr


Page Top research Institut Pasteur homepage

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