Three classes of mouse H-2 messenger RNA distinguished by analysis of cDNA clones

The class I MHC gene family in mice

Molecular genetic analysis of the murine MHC reveals that the class I gene family is far larger than expected from immunogenetic analysis of class I polypeptides. Of approximately 30 class I genes present in mouse DNA most are located telomeric to the H-2 regions in the adjacent Qa and Tla regions. Although it is not yet known how many class I genes from the Qa and Tla regions are expressed to produce class I polypeptides, evidence is accumulating that at least some of these genes encode polypeptides whose structure, and presumably function, is very different from the H-2K, D and L class I polypeptides. Here, Andrew Mellor reviews the structural data which reveals the unexpeted complexity of the class I gene family, particularly in the Qa and Tla regions. He discusses how molecular cloning and analysis of these genes are providing evidence that class I genes in these regions encode polypeptides which are highly homologous to the classical H-2 antigens but which are structurally and functionally quite distinct from them.

Inhibition of IFN-gamma induction of class II MHC genes by cAMP and prostaglandins

Triggering of the cyclic AMP (cAMP) signal transduction pathway inhibits the the interferon gamma (IFN-gamma)-mediated induction of class II major histocompatibility (MHC) genes. We have investigated the mechanism of the inhibition of IFN-gamma induction of the murine A alpha class II MHC gene by cAMP and E series prostaglandins (PGEs). 151 base pairs of the A alpha promoter were sufficient to confer positive regulation by IFN-gamma and negative regulation by cAMP which accurately mirrored the regulation of the endogenous A alpha gene. cAMP also inhibited the IFN-gamma activation of the Fc gamma receptor I (Fc gamma RI) gene promoter, an "early" promoter which is activated immediately after treatment of cells with IFN-gamma. PGEs, which cause an elevation in intracellular cAMP, inhibited the induction of the A alpha promoter, and inhibition was greater in the presence of tumor necrosis factor alpha (TNF alpha). A mutational analysis of the A alpha promoter showed that all four conserved class II promoter elements, the S, X1, X2, and Y boxes, play a role in mediating A alpha promoter activation by IFN-gamma. Mutations in these elements did not diminish the cAMP inhibition of promoter activation by IFN-gamma. Thus, conserved class II promoter sequences which mediate most known examples of positive and negative regulation, including cAMP inhibition of constitutive class II expression, do not mediate cAMP inhibition of IFN-gamma activation of the A alpha promoter. We suggest that this inhibition may be mediated by a novel class II promoter element or by disruption of an early step in the IFN-gamma signal transduction pathway.

A soluble form of the adhesion receptor CD58 (LFA-3) is present in human body fluids

The human adhesion receptor CD58 (LFA-3) is expressed on most human cell types. Here we report on a soluble form of CD58 (sCD58) in human serum, human urine, and culture supernatants of several cell lines. sCD58 partially purified from human serum, from supernatant of the Hodgkin cell line L428, and purified sCD58 from human urine were found to have a molecular mass of 40-70 kDa under denaturating conditions (sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blotting). However, gel filtration of sCD58 purified from human urine gave a molecular mass of 118-166 kDa, suggesting a noncovalent homotrimer conformation or its association with other molecules. Using an enzyme-linked immunosorbent assay specific for CD58 we found that sera from patients suffering from different forms of hepatitis contained elevated sCD58 levels (n = 108). Accordingly, there was a fivefold increase of supernatant sCD58 when the hepatocellular carcinoma cell line Hep G2 was incubated with 25 ng/ml recombinant tumor necrosis factor-alpha in vitro. In contrast, sCD58 serum levels of 337 additional patients suffering from various other immunological disorders were not found to be raised. At high concentrations sCD58 binds to CD2-positive cells and inhibits rosette formation of human T cells to human erythrocytes. Thus, local release of large quantities of naturally occurring sCD58 may interfere with intercellular adhesion in vivo.

Soluble interleukin-6 receptor is released from receptor-bearing cell lines in vitro

Soluble interleukin-6 receptor (sIL-6R) was found to be spontaneously released from human myeloma cell line U266 cells into culture supernatant, and was quantitatively measured with a fluorescence sandwich enzyme-linked immunosorbent assay employing antibodies specific to IL-6R. The supernatant IL-6R was generated only from IL-6R-positive cell lines; myeloma cell lines RPMI8226 and PRMI1788, and myelomonocytic cell lines U937, THP-1, and HL-60. In contrast, it was not released from the IL-6R-negative cells; T cell line Molt-4 and Burkitt lymphoma cell line Raji. SDS-PAGE analysis of the soluble IL-6R from U266 cells suggested a molecular weight of approximately 50-55 kDa, 25-30 kDa smaller than the mature cell surface receptor. These results suggest that the generation of soluble IL-6R may be a maker of myeloma cells and myelomonocytic cells.

Hormonal regulation of major histocompatibility complex class I genes in rat thyroid FRTL-5 cells: thyroid-stimulating hormone induces a cAMP-mediated decrease in class I expression

Thyrocytes normally express major histocompatibility complex (MHC) class I, but not class II, cell surface antigens. A rat thyrocyte cell line, FRTL-5, also expresses MHC class I antigens, in addition to a variety of thyroid-specific genes. Treatment of FRTL-5 thyrocytes with physiological concentrations of thyroid-stimulating hormone (TSH) has been shown to induce increased expressed of thyroglobulin and thyroid peroxidase but to simultaneously decrease expression of the TSH receptor. The reduction in TSH receptor expression by TSH is cAMP mediated. In the present study, it is demonstrated that, in thyrocytes treated with TSH, MHC class I expression decreases concomitant with the decrease in TSH receptor expression. This decreased expression is evidenced by reduced cell surface levels of MHC class I antigens, by reduced steady-state RNA levels, and by reduced transcription of the class I genes. TSH-mediated reduction of MHC class I gene transcription in FRTL-5 cells was mapped to a region within 135 base pairs of the promoter.

New evidence for trans-species evolution of the H-2 class I polymorphism

A serological survey using alloantisera specific for the H-2 class I antigens in Japanese wild mice. Mus musculus molossinus, revealed a high frequency of the H-2Kf antigen. This antigen has also been found in European wild mice, M. m. domesticus and M. m. musculus. In this survey, the H-2Kf antigen was characterized through the use of ten newly isolated monoclonal antibodies raised against cells of a Japanese wild mouse, and by Southern blot analysis using an H-2K locus-specific probe which hybridizes with the 3' end of the gene. The serologically identified H-2Kf antigens revealed several minor variations in reactivities to the monoclonal antibodies. However, all the antigens examined could be clearly separated into two types with respect to the restriction fragment length polymorphism (RFLP) pattern. The first type, found together with a single, characteristic RFLP pattern, was always associated with the presence of reactivity to one particular monoclonal antibody MS54. The second type, found to represent different RFLP patterns, is associated with the absence of reactivity to MS54. This concordance between the presence of an antigenic determinant and a particular RFLP was observed not only within Mus musculus subspecies but also in a different species: M. spretus, carrying the same antigenic determinant, gave an identical RFLP to that of the other MS54-positive Mus musculus subspecies. The data suggest that the antigenic determinant specific for MS54 is an ancient polymorphic structure which has survived the long period of diversification of Mus species (approximately 2-3 million years) without alteration, and is associated with a stable DNA structure at the 3' end of the H-2K gene.

Ki-1 (CD30) antigen is released by Ki-1-positive tumor cells in vitro and in vivo. I. Partial characterization of soluble Ki-1 antigen and detection of the antigen in cell culture supernatants and in serum by an enzyme-linked immunosorbent assay

An enzyme-linked immunosorbent assay (ELISA) has been developed that allows the quantitative determination of the Ki-1 (CD30) antigen in soluble form. Similar levels of sensitivity of this new Ki-1 ELISA and the ELISA previously described for measuring the soluble 55-kDa chain of the interleukin 2 receptor were seen. As assessed with this ELISA, the investigated Ki-1+ permanent cell lines released the Ki-1 antigen into the culture supernatant. In culture supernatants of concanavalin A-activated human peripheral blood lymphocytes, however, this antigen could not be detected. The released Ki-1 antigen has an apparent molecular weight (Mr) of 85,000, whereas the cell-associated Ki-1 antigen has an Mr of 105,000. We investigated sera from 30 normal donors, 15 patients with systemic infections, and 63 patients suffering from lymphomas for soluble Ki-1 antigen. In all sera from normal donors and patients with systemic infectious diseases, soluble Ki-1 antigen was below the detection limit (i.e., less than 70 pg). In contrast, high amounts of the soluble Ki-1 antigen were found in sera from 18 malignant lymphomas containing Ki-1+ tumor cells. This finding demonstrates that the release of the Ki-1 antigen takes place not only in vitro, but in vivo as well. Moreover, these results imply that the Ki-1 antigen may be used as a serum tumor marker.

Expression and secretion of type beta transforming growth factor by activated human macrophages

Alveolar macrophages activated with concanavalin A and peripheral blood monocytes activated with lipopolysaccharide secrete type beta transforming growth factor (TGF-beta). There is minimal TGF-beta secretion in unactivated monocytes, even though TGF-beta mRNA is expressed in these cells at a level similar to that in activated, lipopolysaccharide-treated cultures. U937 lymphoma cells, which have monocytic characteristics, also express mRNA for TGF-beta. Freshly isolated monocytes, both control and lipopolysaccharide-treated, secrete an acid-labile binding protein that inhibits TGF-beta action. We conclude the following: (i) that expression of TGF-beta mRNA is unrelated to monocyte activation, (ii) that secretion of TGF-beta is induced by monocyte activation, and (iii) that cosecretion of TGF-beta and its monocyte/macrophage-derived binding protein may modulate growth factor action. In contrast, monocytic expression of other growth factor genes, such as the B chain of platelet-derived growth factor, is not constitutive and requires activation.

Regulated expression of a murine class I gene in transgenic mice

The major histocompatibility complex class I genes play an essential role in the immune presentation of aberrant cells. To gain further insight into the regulation of the expression of these class I genes and to better define the functions of their protein products, we made use of the technique of gene transfer into the germ line of inbred mice. With the use of locus-specific DNA probes, we observed that a transgenic class I gene was expressed in a tissue-dependent fashion analogous to that of an endogenous class I gene. In addition, the level of expression of the transgenic gene was substantially higher that that of the endogenous gene. The availability of transgenic mice properly expressing a foreign murine class I gene provides a unique system to further define the role of the class I antigens in the maturation of the immune response and in determining the malignant and metastatic phenotypes of tumor cells.

Developmental and tissue-specific regulation of the Q10 class I gene by DNA methylation

The H-2 class I genes encode cell-surface glycoproteins that play a critical role in the immune presentation of aberrant cells. The Q10 class I gene, however, encodes a secreted glycoprotein that is highly homologous to the membrane-bound molecules. While the H-2 genes are activated in all tissue types, the expression of the Q10 gene is restricted to only the liver. Analysis of DNA from different tissues revealed a unique methylation profile for the Q10 gene in liver. Developmental activation of this gene in newborn mice is also reflected by a coordinated temporal change in DNA methylation. By comparing the methylation profiles between congenic mice, which differed in their levels of expression of the Q10 gene, it is observed that methylation at the 3'-flanking region correlates with expression. Methylations were at both CG and CC sequences. Since treatment of newborns with 5-azacytidine, which led to inhibition of methylation, resulted in the suppression of Q10, we conclude that hypermethylation in the 3'-flanking region is responsible, at least in part if not in full, for the activation of the Q10 gene in the liver.

A deletion in a rat major histocompatibility complex class I gene is linked to the absence of beta 2-microglobulin-containing serum molecules

Class I major histocompatibility antigens are composed of a heavy chain that is noncovalently associated with beta 2-microglobulin (beta 2m). Most class I molecules are membrane bound, but mouse and rat cDNA clones and genes without a functional code for the transmembrane amino acids have been identified. The membrane-associated class I molecules are important in the control of cell-mediated cytotoxicity, while the function of the soluble molecules remains unclear. Previous studies have shown that beta 2m circulates in rat serum in three different molecular weight classes. The first is free beta 2m (Mr, 12,000), the second is about Mr 70,000, and the third is roughly Mr 200,000. In an inbred subline of immunodeficient, diabetes-prone BioBreeding rats (BioBreeding/Hagedorn), previous work detected two restriction fragment polymorphisms in class I major histocompatibility complex genes, one of them a gene deletion on a 7-kilobase BamHI fragment and the other on a 2-kilobase BamHI fragment. In these rats we have found that the third serum beta 2m-binding size class is absent. Analysis of F1 and F2 individuals following cross-breeding between BioBreeding/Hagedorn rats and genetically related (nondiabetic) control BioBreeding w-subline rats demonstrated that the large-size serum peak of beta 2m was associated with the presence of the class I restriction fragments.

Molecular analysis of membrane immunoglobulin-negative variants

The mouse B-cell lymphoma WEHI 279.1 is a tumor which synthesizes both membrane and secreted immunoglobulin M (IgM). We have immunoselected variants which fail to express the membrane form (mIgM-); the most frequently isolated phenotype is a complete loss of both membrane expression and synthesis of the mu heavy chain within the cells. We have chosen four of these mIgM- mutants for detailed molecular investigation. One of these has suffered a large deletion which covers the region of chromosome 12 containing the expressed mu gene, but three have no detectable changes in the DNA arrangement of the mu gene. All of the mutants, including the deletion mutant, synthesize 10-30% of the wild-type level of cytoplasmic mu RNA; however, none is the appropriate size for membrane mu (mu m) or secreted mu (mu s) message. Based on our studies of the deletion mutant, which retains its nonproductively arranged allele, at least some of these RNAs may be 'sterile' transcripts from the nonproductively arranged allele. However, if all of these mRNAs derive from the other allele, they represent a substantial elevation of these sterile messages relative to the wild-type level. Furthermore, the three nondeletion mutants transcribe mu RNA at a level indistinguishable from the wild type. It is likely that their defects lie in the stability, processing, or transport of the mu RNA within the nucleus. Somatic cell hybrids between P3X and the IgM- variants produced mostly mIgM- hybrids. However, a few mIgM+ hybrids were produced, suggesting that the mu- defects may be partly complemented by the P3X fusion partner.

Regulated expression of a murine class I gene in transgenic mice

The major histocompatibility complex class I genes play an essential role in the immune presentation of aberrant cells. To gain further insight into the regulation of the expression of these class I genes and to better define the functions of their protein products, we made use of the technique of gene transfer into the germ line of inbred mice. With the use of locus-specific DNA probes, we observed that a transgenic class I gene was expressed in a tissue-dependent fashion analogous to that of an endogenous class I gene. In addition, the level of expression of the transgenic gene was substantially higher that that of the endogenous gene. The availability of transgenic mice properly expressing a foreign murine class I gene provides a unique system to further define the role of the class I antigens in the maturation of the immune response and in determining the malignant and metastatic phenotypes of tumor cells.

Stable transfer and restricted expression of a cloned class I gene encoding a secreted transplantation-like antigen

The identification of a unique major histocompatibility complex class I gene, designated Q10, which encodes a secreted rather than a cell surface antigen has led to questions regarding its potential role in regulating immunological functions. Since the Q10 gene is specifically activated only in the liver, we sought to define the molecular mechanisms which control its expression in a tissue-specific fashion. Results obtained by transfection of the cloned Q10 gene, either in the absence or presence of a heterologous transcriptional enhancer, into a variety of cell types of different tissue derivations are consistent with the Q10 gene being regulated at two levels. The first is by a cis-dependent mechanism which appears to involve site-specific DNA methylation. The second is by a trans-acting mechanism which would include the possibility of an enhancer binding factor. The ability to efficiently express the Q10 gene in certain transfected cell lines offers an opportunity to obtain this secreted class I antigen in quantities sufficient for functional studies; this should also make it possible to define regulatory sequences which may be responsible for the tissue-specific expression of Q10.

Expression of hybrid class I genes of the major histocompatibility complex in mouse L cells

The class I genes of the major histocompatibility complex of the mouse can be divided into two categories: those encoding the transplantation antigens and those encoding the Qa and Tla antigens. The inbred BALB/c mouse has 28 potential Qa/Tla genes. The sites of tissue expression, developmental regulation, and functions of these genes are virtually unknown. We have used the technique of exon shuffling to construct hybrid genes between each of three Qa region genes (Q5, Q7, and Q8) and two other class I genes (H-2Ld and Q6). The hybrid genes have been transfected into mouse L cells, in which intact transplantation antigen genes generally are expressed and in which intact Qa genes generally are not expressed. Analysis of expression of the hybrid gene constructs indicates that the 5' half of two of the Qa genes (Q5 and Q8) can readily be expressed in the context of a hybrid molecule, whereas the 3' half prevents cell-surface expression. The exon shuffling approach described here will be useful in characterizing Qa/Tla genes and in identifying or producing new reagents to study the Qa/Tla gene products, their tissue distribution, their developmental stages of expression, and, ultimately, their functions.

Expression in L cells of transfected class I genes from the mouse major histocompatibility complex

One of the major surprises of the molecular analysis of major histocompatibility complex (MHC) genes is the large number of class I (K/D)-related sequences in the genome. Both restriction fragment length polymorphisms and cosmid cloning experiments showed them all to be closely linked to the MHC. Until now little information was available concerning either their expression or recognition by the immune system. Here we report that these non-K/D genes can provoke antibody responses and be recognized by cytolytic T cells. Immunization of C3H mice with L cells transfected with class I genomic clones resulted in antisera that reacted preferentially with cells from strain B10.P (the gene donor). Thus, these genes can be expressed by L cells. These products were recognized by cytolytic T cells produced by mixed lymphocyte culture with B10.P stimulators. One gene, represented in clone lambda 3a, was chosen for further analysis. A restriction fragment length polymorphism, detected between B10.P (KpDp) and B10.F(14R) (KbDp) and between B10 (KbDb) and B10.F(13R) (KpDb), has enabled us to map the lambda 3a sequence to the D or Tla region. Restriction endonuclease mapping of the lambda 3a clone shows that the gene is intact and that, although many restriction sites are conserved, the gene in lambda 3a differs from other class I genes. When the lambda 3a clone was transfected into mouse L cells, a new product was expressed. Cells expressing this product (designated L3a cells) were killed by primary D-end-reactive, allospecific cytolytic T lymphocytes. The L3a cells were unreactive with monoclonal antibodies specific for the Kp,Dp,Qa-2, Tla.3, and Tla.5 molecules.

Stable transfer and restricted expression of a cloned class I gene encoding a secreted transplantation-like antigen

The identification of a unique major histocompatibility complex class I gene, designated Q10, which encodes a secreted rather than a cell surface antigen has led to questions regarding its potential role in regulating immunological functions. Since the Q10 gene is specifically activated only in the liver, we sought to define the molecular mechanisms which control its expression in a tissue-specific fashion. Results obtained by transfection of the cloned Q10 gene, either in the absence or presence of a heterologous transcriptional enhancer, into a variety of cell types of different tissue derivations are consistent with the Q10 gene being regulated at two levels. The first is by a cis-dependent mechanism which appears to involve site-specific DNA methylation. The second is by a trans-acting mechanism which would include the possibility of an enhancer binding factor. The ability to efficiently express the Q10 gene in certain transfected cell lines offers an opportunity to obtain this secreted class I antigen in quantities sufficient for functional studies; this should also make it possible to define regulatory sequences which may be responsible for the tissue-specific expression of Q10.

Secretion of a soluble class I molecule encoded by the Q10 gene of the C57BL/10 mouse

The DNA sequence of the Q10 genes appears to be highly conserved amongst strains of mice and has only been found to be transcribed in the liver. An examination of the nucleotide sequence of the exon that normally encodes the transmembrane domain of class I molecules suggested that the Q10 gene encodes a secreted protein. We have established this by showing that L cells transformed with an expression vector containing the Q10 gene secrete a class I molecule which was identified with an antiserum raised against a peptide predicted by the Q10 transmembrane exon. Both the L cell-derived Q10 molecule and a class I protein immunoprecipitated from serum with this anti-peptide antiserum have mol. wts. of approximately 38 000; the Q10 molecule secreted by L cells is heterogeneous in mol. wt. This heterogeneity was drastically reduced after endoglycosidase F treatment, suggesting that Q10 molecules secreted into the serum by the liver may be glycosylated differently from those secreted by L cells. Endoglycosidase F treatment of both the L cell and serum forms of the soluble molecule yielded two products with mol. wts. of approximately 32 000 and 35 000; this is consistent with the observation that the predicted Q10 protein sequence has two potential glycosylation sites. In contrast to previous published results, the Q10 molecule reacted with rabbit anti-H-2 antisera which is consistent with its greater than 80% homology to the classical transplantation antigens.

Characterization of two class I major histocompatibility rat cDNA clones, one of which contains a premature termination codon

A cDNA clone that shares considerable homology with mRNA for class-I major-histocompatibility (MHC) antigens has been isolated from a cDNA library constructed from an insulin-secreting tumor of the NEDH rat (RIN). Although there is very strong homology between the third extracellular domain of this gene, and other class-I antigen genes, a stop codon was noticed at the junction between the external and transmembrane regions of its protein. Thus, this gene codes for a truncated class-I antigen that lacks both transmembrane and cytoplasmic domains but has hydrophilic amino acids towards its C-terminus and therefore resembles a soluble protein. This gene was found to be transcribed at significant levels in the rat testis and an insulin-secreting tumor, but transcription was negligible in liver and kidney. The partial nucleotide sequence of this gene in a Wistar rat genomic clone indicates that the cDNA clone is represented in the rat genome complete with premature termination codon, thus eliminating the possibility of artifactual origins for the clone.

Occurrence of a unique MHC class I gene in distantly related members of the genus Mus

There is unequivocal evidence that a relatively nonpolymorphic class I gene (designated Q10) from the Qa region of inbred mice encodes a secreted class I molecule. We have used a cDNA probe specific for this gene and an antiserum specific for its secreted protein product to investigate the occurrence and expression of this gene in different species of wild mice broadly representing the entire genus Mus. Evidence is presented that a Q10-like gene has been conserved and is transcribed and translated throughout the genus, suggesting that it serves an important function. However, the data also show that some differences have appeared in this gene over the period of evolutionary time covered by this sampling of wild mice. These results indicate that a specific class I DNA probe isolated from inbred mice can be used to discriminate a particular gene among the multiple class I genes present in other species.

Functional insertion of an Alu type 2 (B2 SINE) repetitive sequence in murine class I genes

The regulation of expression of the family of MHC (major histocompatibility complex) class I genes is complex. Sequence analysis has revealed that class I genes from the H-2D subregion of the MHC (which includes the D and L genes) differ from the class I gene from the H-2K subregion (the K gene) by the insertion of a type 2 Alu-like repetitive element (the murine B2 sequence) within the 3' noncoding region of the D and L genes. The consequence of this insertion in the D and L genes is the introduction of a novel polyadenylation signal, which is preferentially used over the more distal signal, the analog of that found in the K gene. The insertion of the type 2 Alu-like sequence results in a change in the preferred site for endonucleolytic cleavage which is necessary for generating a correct 3' terminus for polyadenylation. The data demonstrate that the type 2 Alu-like sequence has a function; the data also suggest a possible regulatory role of this sequence in the expression of class I genes.

Organization and evolution of the class I gene family in the major histocompatibility complex of the C57BL/10 mouse

The major histocompatibility complex (MHC) encodes several classes of protein vital to the regulation of the immune response. We have isolated 26 class I genes that map to this region in the C57BL/10 mouse and linked these into three gene clusters. The number of genes differs from the number found in the BALB/c strain and comparison of the organization of the class I genes in these two strains shows conserved regions and polymorphic regions which probably result from deletions, insertions and translocations within the MHC.

Detection of a secreted form of the murine H-2 class I antigen with an antibody against its predicted carboxyl terminus

Analysis of H-2 class I-specific cDNA clones has suggested the synthesis by the liver of a class I molecule that is secreted rather than membrane bound. To detect this putative class I-related molecule, we have predicted a unique region of amino acid sequence located toward the carboxyl terminus of the molecule that is not expected to be shared with any of the classical H-2 class I antigens, and we have generated specific antibodies to a synthetic peptide corresponding to this sequence. Indirect immunoprecipitation with this antibody led to the identification of a Mr 40,000 polypeptide in association with beta 2-microglobulin in the serum of mice of five different H-2 haplotypes. This class I molecule is also detected in the liver together with lower molecular weight components, which are presumably underglycosylated precursors. Synthesis of this molecule is not detected in thymus, spleen, kidney, or testis. This class I serum component has no detectable reactivity with either a broad-specificity alloantiserum against H-2b or a xenoantiserum against purified H-2a class I molecules. The availability of a specific antibody against the secreted class I molecule offers a means to purify this protein for structural and functional studies.

Alien antigens on tumour cells revisited

New findings on the detection of alien histocompatibility antigens (AHA) on tumour cells are reviewed. Three points are discussed, namely: (1) isolation and molecular characterization of AHA; (2) frequency of expression of AHA by tumour cells, and (3) mechanisms of expression of AHA. Mechanisms of expressions of AHA are particularly discussed in the light of the recent advances in the understanding of the molecular genetic organization of H-2 antigens. Gene conversion involving coding or non-coding sequences might be responsible for the appearance of unexpected H-2 determinants on cancer cells.

Inter-locus and intra-allelic polymorphisms of HLA class I antigen gene mRNA

We have constructed cDNA clone libraries from two lymphoblastoid cell lines, JY (HLA-A2, B7, C untypeable) and LB (HLA-A28, B40, Cw3), and isolated clones encoding class I HLA antigens. We have characterized short oligonucleotide probes derived from the coding region of the HLA class I antigens which are specific for the HLA-A and -B loci. These probes have been used to subdivide the class I cDNA clones into subclasses. DNA sequencing of several HLA-A and -B related clones has allowed us to extend the primary structural characterization of these cell-surface antigens. This analysis has also detected a sequence polymorphism at the HLA-A locus, indicating that the previously considered homozygous typing cell line LB expresses two alleles of similar, although not identical, serological specificity.

Alternative RNA splicing in expression of the H-2K gene

A major role of the classical transplantation antigens (designated class I antigens) is the presentation of virus-infected cells to cytotoxic T cells, a process that leads to the destruction of the cell displaying the viral antigen. Consistent with this function is the finding that these transplantation antigens (encoded by the H-2K, H-2D and H-2L genes in mice) are cell-surface glycoproteins with their amino-termini protruding extracellularly and their carboxy-termini located inside the cell. While the external domain is expected to provide biological specificity required for the associative presentation of viral antigens, the role of the cytoplasmic domain remains obscure. The recent observation that this latter region of the molecule is encoded by three separate DNA exons has suggested a complex role for this portion of the polypeptide chain. We have now obtained evidence for the use of alternative acceptor splice sites in the H-2K gene, resulting in two RNA transcripts that would encode H-2K antigens differing in their carboxy-termini. This is the first demonstration of the use of alternative splice acceptor sites in the same class I gene, and indicates the existence of different functional subsets of antigens encoded by the same gene.

Developmental activation of the H-2K gene is correlated with an increase in DNA methylation

Embryonal carcinoma cell lines present strong developmental analogies with early embryonic cells. Previous studies have shown that treatment of F9 teratocarcinoma cells with retinoic acid induces the expression of the classical transplantation antigens which are indispensable for effective interactions between cells. In contrast to several genes that were analyzed, all of which were highly and heterogeneously methylated in F9 cells, the H-2K gene was poorly methylated if at all. Activation of the H-2K gene upon differentiation of F9 cells was accompanied by an increase in DNA methylation. While increased methylation of the H-2K gene in one of the two homologous chromosomes was correlated with a low level of expression, increased methylation in both chromosomes was associated with a high level of expression. Treatment of differentiated F9 cells with 5-azacytidine resulted in inhibition of DNA methylation and a concomitant repression of the H-2K gene expression. Thus, in contrast to the many examples of an inverse correlation between the level of gene methylation and its transcriptional activity, expression of the H-2K gene is directly correlated with the extent of methylation. This finding offers an explanation whereby the hypomethylation observed in tumor cells may be responsible for the establishment and maintenance of a malignant state of growth.

Specific association of repetitive DNA sequences with major histocompatibility genes

The DNA sequence organization of a 17.8-kilobase segment of porcine DNA, containing a functional major histocompatibility (MHC) gene, has been studied. The DNA flanking the MHC gene contains at least 10 distinct repetitive DNA sequence elements, each of which occurs only once within the 17.8-kilobase DNA segment. Their reiteration frequencies in the genome range from 10(2) to 10(4). The genomic organization of seven of these sequence elements has been examined; all are interspersed with other, unrelated DNA sequences. These seven repeated sequences are not generally associated in the genome. However, they appear to be nonrandomly linked in MHC-associated regions of the genome: at least two additional DNA segments containing MHC-homologous DNA also contain sequences homologous to DNA fragments bearing the seven different repeats. Of the seven sequences, four can be detected in splenic total RNA. These results suggest that these repeated elements are specifically associated with the MHC locus.

A major rearrangement in the H-2 complex of mouse t haplotypes

A proportion of wild mice carry a chromosome 17 of which a large part is very different from the standard mouse chromosome 17. The affected region is called the t complex, and the anomalous chromosomal types are the t haplotypes. In combination with various other chromosomes 17, t haplotypes can produce crossover suppression, taillessness, transmission distortion, male sterility and lethality early in development. The various t haplotypes also carry H-2 specificities which are different from those of other mice. This, together with the fact that the lethality genes map to both sides of H-2, suggests that the major histocompatibility complex is contained within the t complex. The lack of recombination between t haplotypes and standard chromosomes 17 may be due to large-scale rearrangements. Genetic data support this idea, in that the tufted gene, the H-2 complex and a group of H-2-related genes appear to be in inverted order in t haplotypes relative to the standard chromosome 17. The mapping of several t-lethal factors close to the H-2-related genes in t haplotypes suggests that breakpoint(s) may be found here. We have now investigated the major histocompatibility complex of t haplotypes by Southern blots using a variety of cloned DNA probes, and find a major rearrangement, specific to the t haplotypes, in the Qa-2,3 region of the complex. This involves the loss of several large homology units, probably including several class I H-2-related genes, and the creation of two possible breakpoints.

Secretion of a transplantation-related antigen

Analysis of mouse cDNA clones has led to the identification of a class I (H-2)-related gene that encodes a truncated transplantation-like antigen. Unlike the products of the class I genes (H-2K, H-2D, and H2-L), which are synthesized and displayed on the surface of all cells, the class I-related gene product is expressed only in liver cells and is secreted. The region of the secreted molecule corresponding to the extracellular domain of the membrane-bound class I antigens shows unusual amino acid substitutions at positions otherwise invariably conserved. There is also loss of a glycosylation site that is used in all class I antigens. Within the region corresponding to the transmembrane domain are multiple nonconservative substitutions of hydrophobic residues, alterations that render the encoded protein incapable of inserting into the plasma membrane. Toward the end of the same domain, the polypeptide chain terminates abruptly and thus lacks the intracellular domain present on all class I antigens. A candidate for this secreted molecule, detected using various heteroantisera against class I antigens, has been identified. A potential role for this serum protein in mediating active tolerance is discussed.

Specific association of repetitive DNA sequences with major histocompatibility genes

The DNA sequence organization of a 17.8-kilobase segment of porcine DNA, containing a functional major histocompatibility (MHC) gene, has been studied. The DNA flanking the MHC gene contains at least 10 distinct repetitive DNA sequence elements, each of which occurs only once within the 17.8-kilobase DNA segment. Their reiteration frequencies in the genome range from 10(2) to 10(4). The genomic organization of seven of these sequence elements has been examined; all are interspersed with other, unrelated DNA sequences. These seven repeated sequences are not generally associated in the genome. However, they appear to be nonrandomly linked in MHC-associated regions of the genome: at least two additional DNA segments containing MHC-homologous DNA also contain sequences homologous to DNA fragments bearing the seven different repeats. Of the seven sequences, four can be detected in splenic total RNA. These results suggest that these repeated elements are specifically associated with the MHC locus.

A cDNA clone containing the entire coding sequence of a mouse H-2Kd histocompatibility antigen

We have isolated a cDNA clone carrying a 1560 bp long insert which contains the entire coding and 3' untranslated regions of an H-2K(d) mouse histocompatibility antigen. Its sequence and overal features are described. They point to the existence of unique properties of DNA sequences associated with the H-2K(d) antigen.

Identification of a second class I antigen controlled by the K end of the H-2 complex and its selective cellular expression

Immunoprecipitates obtained from [35S]methionine-labeled spleen cells by using monoclonal antibodies specific for H-2Kd and H-2Dd have been separated by two-dimensional polyacrylamide gel electrophoresis. Analysis of these gel patterns revealed the presence of an additional product of the K end of the H-2d complex, designated here as H-2K'. To determine whether H-2K' is a unique protein or a differentially glycosylated form of the previously characterized H-2Kd histocompatibility antigen, nonglycosylated molecules labeled in the presence of tunicamycin were examined. The results showed that both H-2K and H-2K' have distinct nonglycosylated polypeptide precursor forms. The approximate molecular weight differences between the fully glycosylated and nonglycosylated molecules also indicated the presence of three oligosaccharide side chains on H-2K', as is the case with H-2Kd, whereas H-2Dd has only two oligosaccharide units. The cellular expression of H-2K' was also investigated. Comparison of H-2 antigens immunoprecipitated from normal spleen cells and from thioglycollate-induced adherent peritoneal exudate cells cultured in the presence or absence of supernatant fluids from concanavalin A-stimulated spleen cells revealed that H-2K' was not expressed on the adherent peritoneal cells. This indicates that H-2K' is expressed in a tissue-specific manner, unlike the classical histocompatibility antigens H-2K and H-2D.

Amino acid sequence of the cytoplasmic segment of the h-2Kd (H2.31) murine major histocompatibility antigen

The primary structure of the carboxy-terminal portion of the H-2Kd murine major histocompatibility antigen has been determined using radiosequencing methodology. The two peptides encompassing the entire cytoplasmic portion of the H-2Kd molecule were isolated from cyanogen bromide digests of the detergent solubilized molecule. These two peptides are not present in CNBr digests of papain-solubilized H-2Kd. Alignment of the two CNBr peptides was deduced from tryptic overlap peptides derived from the whole molecule. Alignment with the corresponding region of the H-2Kb antigen shows 90% homology and supports the assignment of this segment of H-2Kd as the C-terminal. The sequence obtained in this study is (Met)-Arg-Arg-Asn-Thr-Gly-Gly-Lys-Gly-Val-Asn-Tyr-Ala-Leu-Ala-Pro-Gly-Ser-Gln- Thr-Ser-Asp-Leu-Ser-Leu-Pro-Asp-Pro-Lys-Val-Met-Val-His-Asp-Pro-His-Ser-Leu- Ala. These data allow extensive comparisons with the protein sequences deduced from the 3' ends of H-2d haplotype cDNA and genomic clones as well as with the homologous regions of H-2Kb and H-2Db.

Mapping of class I DNA sequences within the human major histocompatibility complex

Mutants that had lost expression of alleles of one or more HLA loci were isolated with immunoselection after gamma-irradiation of a human lymphoblastoid cell line LCL 721. DNAs from the mutants were digested with restriction endonucleases and analyzed by Southern blotting using probes for class I HLA genes. Eight polymorphic cut sites for HindIII and PvuII were discovered in class I-associated sequences of LCL 721. Losses of specific fragments generated by restriction enzymes could be associated with losses of specific antigenic expressions and it was possible in this way to assign HLA-A1, HLA-A2, and HLA-B8 to specific DNA fragments. Patterns of gamma-ray-induced segregations of DNA fragments permitted rough linkage alignment of about 30% of the fragments generated by PvuII. The resultant map showed that there are class I HLA genes on the telomeric side of the HLA-A locus. Restriction enzyme site polymorphisms were also examined in a panel of DNAs isolated from peripheral blood lymphocytes (PBLs) of HLA-typed individuals. This panel of PBL DNA complemented the analysis using the HLA deletion mutants.