Molecular cloning and characterization of murine ICOS and identification of B7h as ICOS ligand.

Human ICOS (huICOS) is a T cell-specific molecule structurally related to CD28 and CTLA-4 with potent co-stimulatory activities on T cell proliferation, cytokine induction and T cell help for B cells. We have now cloned and characterized murine ICOS (muICOS). muICOS mRNA of 1.5 kb and 3.3 kb encodes a protein with a deduced molecular mass of 20.3 kDa, which is 71.7 % identical to huICOS. On the cell surface, muICOS is expressed as a disulfide-linked, glycosylated homodimer of 47-57 kDa, with subunits of approximately 26 kDa. With a panel of monoclonal antibodies we have determined the expression of muICOS in vitro and in vivo. Following activation of splenic T cells via CD3, muICOS became detectable at 12 h and reached a maximum of expression at around 48 h, thus exhibiting expression kinetics similar to huICOS. In vivo, muICOS was found to be substantially expressed in the thymic medulla and in the germinal centers and T cell zones of lymph nodes and Peyer's patches. Non-lymphoid tissue was ICOS negative. The muICOS gene was mapped to a region of chromosome 1 also harboring the CD28 and CTLA-4 genes. Using recombinant chimeric muICOS-Ig we determined that B7h, a recently cloned B7-like molecule, is a ligand for muICOS.

Induction, binding specificity and function of human ICOS.

Recently, we have identified the inducible co-stimulator (ICOS), an activation-dependent, T cell-specific cell surface molecule related to CD28 and CTLA-4. Detailed analysis of human ICOS presented here shows that it is a 55-60-kDa homodimer with differently N-glycosylated subunits of 27 and 29 kDa. ICOS requires both phorbol 12-myristate 13-acetate and ionomycin for full induction, and is sensitive to Cyclosporin A. ICOS is up-regulated early on all T cells, including the CD28- subset, and continues to be expressed into later phases of T cell activation. On stimulation of T cells by antigen-presenting cells, the CD28/B7, but not the CD40 ligand/CD40 pathway is critically involved in the induction of ICOS. ICOS does not bind to B7-1 or B7-2, and CD28 does not bind to ICOS ligand; thus the CD28 and ICOS pathways do not cross-interact on the cell surface. In vivo, ICOS is expressed in the medulla of the fetal and newborn thymus, in the T cell zones of tonsils and lymph nodes, and in the apical light zones of germinal centers (predominant expression). Functionally, ICOS co-induces a variety of cytokines including IL-4, IL-5, IL-6, IFN-gamma, TNF-alpha, GM-CSF, but not IL-2, and superinduces IL-10. Furthermore, ICOS co-stimulation prevents the apoptosis of pre-activated T cells. The human ICOS gene maps to chromosome 2q33 - 34.

Sequential involvement of NFAT and Egr transcription factors in FasL regulation.

The critical function of NFAT proteins in maintaining lymphoid homeostasis was revealed in mice lacking both NFATp and NFAT4 (DKO). DKO mice exhibit increased lymphoproliferation, decreased activation-induced cell death, and impaired induction of FasL. The transcription factors Egr2 and Egr3 are potent activators of FasL expression. Here we find that Egr2 and Egr3 are NFAT target genes. Activation of FasL occurs via the NFAT-dependent induction of Egr3, as demonstrated by the ability of exogenously provided NFATp to restore Egr-dependent FasL promoter activity in DKO lymph node cells. Further, Egr3 expression is enriched in Th1 cells, suggesting a molecular basis for the known preferential expression of FasL in the Th1 versus Th2 subset.

Utilization of an NF-ATp binding promoter element for EGR3 expression in T cells but not fibroblasts provides a molecular model for the lymphoid cell-specific effect of cyclosporin A.

Cyclosporin A (CsA) mainly exerts its immunosuppressive action by selectively inhibiting Ca2+/calcineurin-dependent gene transcription in lymphoid cells. A model explaining the tissue-specific effect of this drug on gene expression has not been established to date, since none of the known intracellular targets of CsA (e.g., cyclophilins, calcineurin, and NF-AT) is lymphoid cell specific. To investigate this issue, we performed a detailed comparative analysis of the promoter regulating the two-signal-dependent (Ca2+ ionophore plus phorbol myristate acetate [PMA]), CsA-sensitive expression of EGR3 in T cells and the one-signal-dependent (PMA), CsA-insensitive expression of EGR3 in fibroblasts. As a result, we identified a 27-bp promoter element functionally interacting with transcription factors NF-ATp and NF-ATc that is crucial for the CsA-sensitive expression of the EGR3 gene in T cells. In contrast, the same element was without function in fibroblasts, and other, CsA-insensitive promoter regions were found to be responsible for EGR3 gene expression in these cells. The inactivity of the 27-bp element in fibroblasts was apparently due to insufficient expression levels of NF-ATp, since overexpression of NF-ATp, but not NF-ATc, restored the two-signal phenotype and CsA sensitivity of EGR3 promoter induction in these cells. The differential usage of an NF-AT binding site explains the selective effect of CsA on EGR3 gene expression in T cells versus fibroblasts and may represent one of the basic mechanisms underlying the tissue specificity of CsA.

Cloning of ATAC, an activation-induced, chemokine-related molecule exclusively expressed in CD8+ T lymphocytes.

A cDNA clone, designated ATAC, was isolated from a collection of human T cell activation genes. Analysis of tissue distribution determined that ATAC mRNA of approximately 0.9 kb is exclusively expressed in activated CD8+ T cells. Induction of the ATAC gene requires stimulation by both phorbol 12-myristate 13-acetate and Ca2+ ionophore A23187 ("two-signal gene") and is fully abrogated by the immunosuppressive agent cyclosporin A. Upon stimulation, ATAC mRNA is detectable within 30 min, maximal expression is seen after 4 h. The polypeptide encoded by the open reading frame of ATAC mRNA is 114 amino acids long with a calculated M(r) of 12.52 kDa. The structural features predict the cleavage and secretion of a mature ATAC protein of approximately 10 kDa from the 12.52-kDa precursor. ATAC is highly similar to a very recently identified murine molecule designated lymphotactin both at the cDNA (73.8% identity) and the protein (61.4% identity) levels, and related to members of the C-C and C-X-C chemokine families. Two variants of the ATAC protein were expressed and tested for chemotaxis and Ca2+ release on a variety of target cells. The ATAC gene was located to chromosome 1q23.

NOT, a human immediate-early response gene closely related to the steroid/thyroid hormone receptor NAK1/TR3.

By analyzing the early genetic response of human T cells following mitogenic activation we have identified NOT, a member of the steroid/thyroid hormone family of receptors. NOT has all structural features of steroid/thyroid hormone receptors (C2C2 zinc-finger domain, ligand binding domain), but is rapidly and only very transiently expressed after cell activation, which is clearly at variance with classical steroid receptors such as glucocorticoid or estrogen receptors. NOT gene induction is independent of de novo protein synthesis, defining NOT as an immediate-early response gene. Short-lived NOT mRNA (4.2 kilobases) expression could be observed in vitro in a greater number of tissue types following activation by a variety of distinct stimuli. In vivo, NOT mRNA expression was detected exclusively in the brain, where a very strong signal was observed. By immunoblot analysis of human T cell lysates with NOT specific antisera two activation-dependent protein bands (66 and 59 kilodaltons) could be detected. NOT gene was localized to human chromosome 2q22-q23. Sequence comparison revealed that NOT is the human homolog of the murine NURR1 and rat RNR-1. Moreover NOT is closely related to NAK1/TR3, a previously identified human orphan steroid receptor. Several lines of evidence indicate that NOT and NAK1/TR3 form a distinct and exclusive subgroup of orphan steroid receptors, whose expression characteristics in vitro and in vivo resemble the expression of nonsteroid immediate-early transcription factors such as jun and fos. NOT and NAK1/TR3 thus may function as general coactivators of gene transcription rather than participate in the induction of specific target genes, as is the case with classical steroid receptors.

Defective expression of CD40 ligand on T cells causes "X-linked immunodeficiency with hyper-IgM (HIGM1)".

X-linked immunodeficiency with hyper-IgM (HIGM1) is a rare disorder, characterized by recurrent infections associated with very low or absent IgG and IgA, and normal to increased IgM serum levels. The disease has been earlier mapped to the q26-27 region of the X-chromosome. We have identified a novel molecule expressed on the surface of activated T cells, which was designated TRAP (Tumor necrosis factor Related Activation Protein), and could demonstrate that TRAP is a ligand for the CD40 receptor expressed on B cells. Our mapping of the TRAP gene to the Xq26.3-27.1 region suggested a causal relationship to HIGM1. Further work revealed that various mutations of the TRAP/CD40 ligand (CD40L) gene may lead to a defective expression of the TRAP/CD40L molecule on the T-cell surface in HIGM1 patients. A combination of structural and functional analyses finally demonstrated that the failure of TRAP/CD40L on T cells to interact with CD40 on B cells is responsible for the inefficient T-cell help for B cells observed in HIGM1. The observations made in HIGM1 allowed us to conclude that TRAP/CD40L is not required for IgM synthesis. In contrast, functional expression of TRAP is a prerequisite for effective immunoglobulin isotype switching and subsequent production of IgG, IgA and IgE by B cells in vivo. The interaction of TRAP/CD40L with CD40 thus provides a very critical link between the cellular and the humoral part of the immune system. The knowledge of TRAP/CD40L cDNA sequence, the availability of various reagents for the testing of expression and function of TRAP/CD40L, and our recent elucidation of the exon-intron structure of the TRAP/CD40L gene now provide all necessary tools for early diagnosis of affected patients and the detection of female carriers of HIGM1. The available information will also provide a basis for future attempts at gene therapy in this disease.

Defective expression of T-cell CD40 ligand causes X-linked immunodeficiency with hyper-IgM.

X chromosome-linked immunodeficiency with hyper-IgM (HIGM1, MIM number 308230) is a rare disorder characterized by recurrent bacterial infections, very low or absent IgG, IgA and IgE, and normal to increased IgM and IgD serum levels. HIGM1 has been suggested to result from ineffective T-cell help for B cells. We and others have identified a novel, TNF-related activation protein (TRAP) that is exclusively expressed on the surface of stimulated T cells. TRAP, a type II transmembrane protein of M(r) 33,000, is the physiological ligand for CD40 (refs 5-8). Crosslinking of CD40 on B cells induces, in the presence of lymphokines, immunoglobulin class switching from IgM to IgG, IgA or IgE. Mapping of the TRAP gene to the X-chromosomal location q26.3-q27.1 (ref. 6) suggested a causal relationship to HIGM1, which had previously been assigned to Xq26 (refs 12-14). Here we present evidence that point mutations in the TRAP gene give rise to nonfunctional or defective expression of TRAP on the surface of T cells in patients with HIGM1. The resultant failure of TRAP to interact with CD40 on functionally intact B cells is responsible for the observed immunoglobulin isotype defect in HIGM1.

Expression of PILOT, a putative transcription factor, requires two signals and is cyclosporin A sensitive in T cells.

Few known genes (IL-2, members of the IL-8 family, interferon-gamma) are induced in T cells only through the combined effect of phorbol myristic acetate (PMA) and a Ca(2+)-ionophore, and expression of only these genes can be fully suppressed by Cyclosporin A (CyA). We have identified a putative transcription factor, designated PILOT, with an identical dual signal requirement for expression. Induction of the PILOT gene is detectable in human T cells 20 min following activation in the presence of cycloheximide and is fully suppressed by CyA. The PILOT protein has a calculated M(r) of 42.6 kDa and contains three zinc fingers of the C2H2-type at the carboxyl-terminus which are highly homologous to the zinc finger regions of the transcription factors EGR1, EGR2, and pAT 133. In contrast to T cells, in fibroblasts PILOT gene expression requires only one signal (PMA) and is not affected by CyA. This observation directly demonstrates the existence of a Ca2+ signal-dependent regulatory element obligatory for expression of some genes in T cells but not in fibroblasts. This differential expression model will be valuable in the dissection of the dual signal pathway in T cells and the effects of CyA upon it.

Cloning of TRAP, a ligand for CD40 on human T cells.

A cDNA clone, designated TRAP (TNF-related activation protein) was isolated from a collection of T cell activation genes. The polypeptide encoded by a mRNA of approx. 2.3 kb is 261 amino acids long with a calculated M(r) of 29.3 kDa. The structural features predict a type II transmembrane protein, but are also compatible with a secreted form. TRAP is highly similar to an identified murine CD40 ligand both at the cDNA (82.8% identity) and the protein (77.4% identity) levels, and related to tumor necrosis factor/lymphotoxin. Expressed in a murine myeloma, TRAP was identified as a ligand for CD40 by binding to a soluble CD40 construct. TRAP mRNA is expressed in a T cell-specific fashion with a maximum at 8 h after stimulation. The TRAP gene is located in the q26.3-q27.1 region of the X chromosome.

Oligoclonal T cells in rheumatoid arthritis: identification strategy and molecular characterization of a clonal T-cell receptor.

Immunodominant antigens in rheumatoid arthritis (RA) should induce an expansion of T cells bearing a corresponding T-cell receptor (TCR). We therefore analysed the TCR repertoire at the site of inflammation using two fundamentally different strategies. The total TCR repertoire was examined by generating 'representative' T-cell clone panels, which were subsequently tested for clonality by restriction mapping of the TCR beta gene locus. No clonality was detected in large T-cell clone panels generated with cells from three patients. However, when we selectively analysed the TCR repertoire of in vivo pre-activated, interleukin-2 (IL-2)-responsive T cells, significant T-cell/TCR clonality was found in 2 out of 4 patients. The clonal T cells represented a minority of the total T-cell population with an estimated frequency of 1 in 300 to 1 in 1000 cells. Molecular characterization of a clonal TCR and the use of a specific TCR V beta MoAb ruled out an over-representation of T cells bearing the same V beta element in the total T-cell population, rendering the involvement of super-antigens in the induction of T-cell clonality in this case unlikely.

The sexual inducer of Volvox carteri. Primary structure deduced from cDNA sequence.

The primary structure of the sexual inducer of Volvox carteri f. nagariensis has been deduced by cloning and sequence analysis of cDNA. The sexual inducer contains 208 amino acids including a signal sequence. A total of six potential N-glycosylation sites are found within the polypeptide chain. At the genomic level, the sexual inducer protein is encoded in five exons.