Cd1d-restricted cellular lysis by peripheral blood lymphocytes: relevance to the inflammatory bowel diseases.

The CD1d molecule has been implicated to play a role in inflammatory bowel diseases (IBD), possibly through its presentation of an intestinal antigen trigger. To understand the role of the CD1d class I-like protein in IBD, we investigated the molecule's expression in diseased intestinal tissue and determined its potential to undergo specific recognition by intraepithelial and peripheral blood lymphocytes (PBLs) derived from IBD patients. We have observed an increase in precipitable CD1d in inflamed tissues, which suggests CD1d up-regulation in IBD; this was not accompanied by the occurrence of CD1d-specific cytotoxicity by lymphocytes isolated from the same tissue sites. In contrast, we have observed CD1d-specific cytotoxicity by PBLs from both patients and normal controls mediated by a possibly unique type of lymphocytic cell. These observations support a model in which intestinal inflammation may be initiated by circulating PBLs following the tissue-specific upregulation of CD1d. These activated PBLs may then be the source of the extraintestinal manifestations observed with IBD. We therefore propose that the cells responsible for this activity may play a role in regulating immune responses through the specific recognition of CD1d-specific antigen(s).

Absence of GNAI2 codon 179 oncogene mutations in inflammatory bowel disease.

The human GNAI2 gene coding for G protein, Galphai2, is located on chromosome 3p21 in proximity to the region where an inflammatory bowel disease (IBD) locus has been suggested. Galphai2-deficient mice develop a lethal diffuse colitis that resembles human ulcerative colitis (UC) and frequently progresses to colon adenocarcinoma. Furthermore, the human GNAI2 gene is subject to point mutations at certain positions, including three at codon 179, all of which have been reported in human endocrine tumors. In order to evaluate the possible involvement of this gene in IBD pathogenesis, we have examined GNAI2 codon 179 sequences in 28 familial IBD patients, including 13 UC, 15 Crohn's disease (CD), and 7 patients with colon cancer/dysplasia, from 12 multiplex IBD families. The wildtype codon 179, CGC for arginine, plus the first G of the codon 180 engender a sequence recognizable by the enzyme BstUI. Mutations, therefore, can result in the abrogation of BstUI digestion of polymerase chain reaction (PCR) products containing the codon 179. Using the PCR-restriction fragment length polymorphism technique, all 28 IBD patients, including those with colon cancer, and 14 non-IBD family members show a BstUI-cleavable PCR-banding pattern indicating the presence of wildtype codon 179. We conclude that, in the familial IBD and colon cancer/dysplasia patients studied, there is no detectable mutation in the codon 179 of the GNAI2 gene.

Gamma delta intraepithelial lymphocytes drive tumor necrosis factor-alpha responsiveness to intestinal iron challenge: relevance to hemochromatosis.

The dependence of intestinal epithelial cell (IEC) growth and differentiation on intraepithelial lymphocytes (IELs) expressing the gamma/delta (gamma delta) T-cell receptor (TCR), suggested a potential role for gamma delta + IELs in the regulation of iron absorption. We therefore examined the levels of hepatic iron and the IEL cytokine responses in C57BL/6J control and class I and TCR knockout lines (placed on a C57BL/6J genetic background) following the administration of supplemental dietary iron. The highest level of liver iron was found in the beta 2-microglobulin knockout (beta 2m-/-) mice followed by the TCR-delta knockout (TCR delta-/-) animals. TCR-alpha knockout (TCR alpha-/-) and control animals did not differ in their iron levels. Liver iron loading correlated inversely with the ability of the mice to generate an IEL tumor necrosis factor (TNF)-alpha response. These observations suggest a model in which IEC iron loading is communicated to IELs via the HFE class I protein. The result of this communication is the initiation of TNF-alpha release by gamma delta + IELs (sustained by macrophages and dendritic cells) contributing to the upregulation of ferritin expression and possibly to the normal maintenance of the IEC apoptotic pathway.

Conservation of ancestral haplotypes telomeric of HLA-A.

Genes that predispose to haemochromatosis are though to be located within the several megabases telomeric of HLA-A. Further recombinant mapping has been used previously to map susceptibility genes for diseases such as insulin-dependent diabetes mellitus, myasthenia gravis and cystic fibrosis, and should be useful in relation to haemochromatosis. However, this method requires the recognition of ancestral haplotypes within the susceptibility region. Using a panel of six microsatellite markers from this region (MOG A, MOG B, MOG C, D6S464, D6S306 and D6S105), we show that ancestral haplotypes extend telomeric of HLA-A, at least as far as D6S105. Nine of 14 haplotypes carrying HLA-B7 and HLA-A3 shared the same microsatellite alleles between HLA-A and at least D6S105. Similarly, nine of 10 haplotypes sharing HLA-B8 and HLA-A1 shared the same microsatellite alleles, although a different set to those with HLA-B7 and HLA-A3. Haplotypes representing historical recombination events were also identified. These two findings demonstrate that recombinant mapping may be applicable to the mapping of disease genes in this region.

An increased gamma delta T cell population in the intestine of thymus-leukemia antigen transgenic mice.

In an approach to study thymic leukemia antigen's (TL's) function, we have developed transgenic mice that express T18d on virtually all somatic cells; in such mice, we initially observed changes in T cells within the thymus and lymph nodes as well as the ability of TL to undergo recognition by splenic T cells. As phase II of our study, we now present the results on the composition of gut T cell populations which may be a better measure of TL's true function. We have demonstrated an increase in the number of gamma delta T cells as well as the increase in gamma delta T cells expressing the V gamma 2 chain. These cells appear to be both CD4 and CD8 negative. This suggests that TL may select for a subset of gamma delta T cells within the gut and bolsters earlier reports implicating an H-2T regional gene product as the major histocompatibility complex ligand for gamma delta T cells.

Transgenic mice expressing the thymus leukemia antigen fail to control cutaneous herpes simplex virus infection.

TL-transgenic mice expressing the thymus leukemia antigen demonstrate a lack of viral clearance following cutaneous HSV infection of the footpad. In this study, both uninfected and HSV-infected TL-transgenic mice demonstrate increased concentrations of IL-4 as well as decreased concentrations of IFN-gamma which may possibly underlie the impairment of viral clearance. Furthermore, lymphocytes from HSV-infected nontransgenic mice, adoptively transferred into HSV-infected TL-transgenic mice, promoted viral clearance and led to an increase in IFN-gamma production. Transgenic mice which were subcutaneously injected with IFN-gamma in the right footpad were also capable of clearing the viral challenge; however, clearance was restricted solely to the right footpad. These studies support the possibility of perturbations in the immune system of TL-transgenic mice and effectively demonstrate the utility of this model system in the study of HSV clearance, persistence, and potential spontaneous reactivation. Moreover, the TL-transgenic animals may provide a useful model system for additional studies requiring a host system skewed toward a Th2 phenotype.

Expansion of natural (NK1+) T cells that express alpha beta T cell receptors in transporters associated with antigen presentation-1 null and thymus leukemia antigen positive mice.

Thymic selection of natural killer-1+ natural T cells that express alpha beta T cell receptors requires a conserved beta 2-microglobulin-associated molecule, presumably CD1d, displayed by CD4+8+ thymocytes. Here we demonstrate that positive selection of natural T cells occurs independent of transporters associated with antigen presentation-1 (TAP-1) function. Moreover, natural T cells in TAP-1o/o mice are numerically expanded. Several H-2 class Ib molecules function in a TAP-independent manner, suggesting that if expressed in TAP-1o/o thymocytes, they could play a role in natural T cell development. Of these class Ib molecules, H-2TL is expressed by TAP-1o/o thymocytes. Moreover, we find that thymi of TL+ mice congenic or transgenic for H-2T18 also have a numerically expanded natural T cell repertoire compared with TL- mice. This expansion, as in TAP-1o/o thymi, is evident in each of the limited T cell receptor V beta chains expressed by natural T cells, suggesting that TL and CD1d impact similar repertoires. Thus TL, in addition to CD1d, plays a role in natural T cell development.

Thymus-leukemia antigen interacts with T cells and self-peptides.

The functional role of the class Ib thymus-leukemia (TL) Ag expressed within the thymic cortex and intestinal mucosa of the mouse remains unknown. In an approach to elucidate the potential functionality of TL, we developed transgenic mice that ectopically express the H-2T18d gene product on essentially all nucleated cells through the control of a heterologous H-2Kb gene promoter. Transgenic mice demonstrated an increase in the number of CD4+ lymphocytes within the thymus and lymph nodes; these cells displayed an altered T cell receptor repertoire possibly suggesting a role for the ectopically expressed TL protein. The TL protein additionally displayed the characteristics of a bona fide transplantation Ag, because skin grafts from transgenic animals onto MHC- and minor histocompatibility Ag-matched nontransgenic recipient mice resulted in a rapid and vigorous immunologic rejection of the allograft. In MLR studies, transgenic stimulator cells induced the proliferation of responders to a level intermediate between genetically identical and H-2-disparate responder-stimulator combinations. The TL protein was also capable of stimulating cytotoxic T lymphocytes, thereby resulting in specific lysis of TL+ target cells. Further data demonstrated that the TL protein assembles with peptides that are modified at the amino terminus, and that TL retains these molecules at the cell surface. Together, these data suggest that H-2T18d is capable of interacting with T cells via a bound peptide. These data further support the possibility that TL may subserve a specialized function within the immunologic system.

Clinical and molecular analyses of deletion 3p25-pter syndrome.

Hemizygous deletion of 3p25-pter is associated with a phenotype of profound growth failure, microcephaly, characteristic facial changes, and mental retardation. Since the severity may be quite variable, we have studied 3 cases of del 3p25-pter to define the clinical manifestations and the critical chromosome region for phenotypic expression. The patient we now report died at age 6 months and provided an opportunity for a detailed necropsy analysis for only the second time in a del(3p) patient. He had marked hypoplasia of all organs, hypomyelination of white matter, and multiple renal cortical microcysts. Ordered genomic markers from the distal regions of chromosome 3p aided in determining the parent of origin of each deletion and in defining the boundaries of the deleted chromosomal segments. The deleted markers distal to the RAF1 oncogene in 2 of the 3 patients were consistently hemizygous. One patient had an interstitial deletion based on evidence of diploid inheritance of one of the most distal loci (D3S17). Available genetic linkage maps suggest that the deletion spans at least 19 centimorgans (cM).

Characterization of the thymus leukemia (TL) product encoded by the BALB/c T3c gene by DNA-mediated gene transfer. Comparison to the T13c product and BALB/c leukemia TL.

Using DNA-mediated gene transfer, we have studied the TL protein products encoded by both the T3c and T13c BALB/c genes. Biochemically, the proteins differed in their m.w. and pI points; serologically, although both molecules were recognized by TL alloantiserum, only the T13c protein was recognized by monoclonal TL antibodies. Interestingly, both proteins were serologically and immunochemically recognized by leukemia-specific TL.4 antiserum. The quantity of cell surface T13 was significantly greater than T3 possibly due to the less efficient splicing of T3 transcripts in the L cell nucleus; both genes directed the synthesis of cytoplasmic RNA containing an unspliced intron 3 as assessed by S1 analysis. In toto, the results suggest that T3c is similar or perhaps identical with the novel TL product previously identified on the surface of certain x-ray-induced BALB/c leukemias.

Additional products of the Tla locus of the mouse.

Thymocytes and leukemia cells of some mouse strains yield TL proteins, precipitable by anti-TL antiserum and by anti-TL monoclonal antibodies, that include not only the familiar heavy (H) chain of 45-50 kDa but also products of higher molecular mass. Production of a 53-kDa TL form by Tlad thymocytes was studied in detail. A cross was made between B10.M (Tlad) mice, which produce the 53-kDa TL, and mice of the A strain (Tlaa), which make only the usual H chain. Hemi-expression of apparently unaltered 53-kDa TL was observed in thymocytes of the Tlad/Tlaa heterozygous F1 progeny. Thus, there was no indication of positive or negative trans interaction with respect to production of the 53-kDa TL form associated with Tlad. We conclude that production of 53-kDa TL is governed intrachromosomally. Two-dimensional chymotryptic peptide maps of the TL H chain and the 53-kDa TL of Tlad thymocytes differed only by added features found in the map of the 53-kDa TL. With the exception of Tlaa, all Tla alleles (Tlab-f) yielded TL products of higher molecular weight than the accompanying H chain, although in the case of Tlab this was evident only in TL+ leukemia cells because Tlab thymocytes are TL-. For H-2, representing other class I genes, no products other than the familiar H chain were demonstrable under similar conditions.

Structural characteristics of Tla products.

Biochemical study of thymus leukemia antigen (TL) from thymocytes of various Tla genotypes and from leukemia cells revealed features that, given present evidence, are peculiar to TL among class I products of the H-2:Qa:Tla region of chromosome 17. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) of TL from thymocytes of all TL+ mouse strains, precipitated by anti-TL antiserum or monoclonal antibodies, showed two closely migrating bands of equal intensity in the heavy (H) chain position (45-50,000 mol wt). Comparison of these two bands by two-dimensional isoelectric focusing (2D IEF)-SDS-PAGE and 2D chymotryptic peptide mapping showed no differences indicative of protein dissimilarity. Thus, the two components of the H chain doublet may differ only in a feature of glycosylation that does not affect charge. The two leukemias studied gave only a single band in the H chain position. On 2D peptide mapping and 2D IEF-SDS-PAGE, the patterns for TL of Tlaa and Tlae thymocytes, which are closely related serologically, were broadly similar, but clearly different from the pattern typical of Tlac and Tlad thymocytes. 2D peptide maps of TL from Tlaa thymocytes and Tlaa leukemia cells did not differ. Leukemia cells of Tlab origin (thymocytes TL-) gave 2D peptide and 2D IEF-SDS-PAGE patterns of a third type. With the exception of Tlaa, thymocytes of TL+ mice yielded additional TL products of higher molecular weight than the TL H chain.

Biochemical characterization of a p43,12 complex: comparison with human and murine class I molecules.

A monoclonal antibody designated as C21 reacting with a p43,12 complex was developed against human thymocytes. It stained predominantly the early hematopoietic cells of the lymphoid lineage and also thymocytes, peripheral B-cells and activated T- and B-cells similarly to OKT10. The heavy chain of this antigen was a glycoprotein of Mr 43,000 (p43). Sequential immunoprecipitation with C21 and OKT10 antibodies indicated that they both reacted with an identical heavy-chain molecule. This observation was further documented by two-dimensional analysis. Monoclonal antibody C21 was used to probe a p43,12 complex further. Structural polymorphism of the p43 heavy chain isolated from T- and B-cells of different individuals was not detected by chymotryptic peptide mapping, although molecules from these cell types possessed a different charge on two-dimensional gels. An unusual observation was made regarding this complex on MOLT4 cells. The light chain co-precipitated from these cells was 12,000 daltons and had a pI distinct from that of beta 2-microglobulin but similar to the pI of the beta t molecule. Comparison between chymotryptic peptide maps of the p43 heavy chain and those of the human and murine class I molecules such as HLA, T6, H-2K, Qa-2 and TL revealed no apparent homology. We have shown, however, that the peptide backbone of p43, as studied by both tunicamycin treatment of cells and endoglycosidase F digestion of immunoprecipitates, was identical in size to that of murine Qa-1. These results suggest that the p43 antigen may be homologous to murine Qa-1 or another class I antigen encoded in the murine TL:Qa region.

Biochemical characterization of the human T6 antigen: a comparison between T6 and murine TL.

The human T6 antigen was studied by two monoclonal antibodies: OKT6 and Leu-6. A third monoclonal antibody, C56 (developed in our laboratory), was found to have similar properties to those of OKT6. On SDS-PAGE, all three antibodies precipitated a 48,000-12,000-dalton heterodimer. Two-dimensional gel electrophoresis and chymotryptic peptide map analysis revealed that these antibodies precipitated in identical 48,000-dalton heavy chain which was distinguishable from the HLA-A,B,C heavy chains. The single 12,000-dalton light chain precipitated with OKT6 antibody was shown to be distinct from beta 2-microglobulin by its pI. The two light chains precipitated with Leu-6 antibody were resolved by charge into beta 2-microglobulin and the more basic 12,000-dalton peptide identical to that precipitated with OKT6. In addition to beta 2-microglobulin, the latter component (presumably beta t) was also found in the light-chain fraction precipitated from the thymocytes with a monoclonal antibody recognizing the framework of HLA-A,B,C heavy chains. Using chymotryptic peptide mapping, no polymorphism was detected among the heavy chains of the T6 antigen isolated from thymocytes of four individuals. All three monoclonal antibodies failed to precipitate murine TL from ASL1 leukemia cell lysates. Similarly, none of the six monoclonal and two conventional anti-TL antibodies reacted with T6. Although a high degree of homology was found by peptide map analysis among the TL molecules encoded by the Tlaa, Tlad and Tlae alleles, a comparison between their peptide maps and that of T6 revealed no similarity. Despite previous suggestions that T6 is homologous to murine TL, the present biochemical studies do not support this hypothesis.

Further polymorphism of the Tla locus defined by monoclonal TL antibodies.

Six new monoclonal TL antibodies are described. At least one new TL antigen is defined (TL.7), and at least one more Tla allele, bringing the total number of known Tla alleles to six. Five of the monoclonal antibodies, and probably all six, identify distinct TL antigenic specificities. Four of these antigens conform in strain distribution and expression on leukemia cells to antigens defined by conventional antisera. The data contain a hint that monoclonal TL antibodies like TL.m6 may serve to identify a region of the Tla gene, which determines whether or not prothymocytes will respond to physiological induction by expressing TL, and thus may provide a means to study the regulatory mechanism that determines whether mouse strains are phenotypically TL+ or TL-.