Three Ia species with different structures and alloantigenic determinants in an HLA-homozygous cell line.

Three distinct molecular subsets with different structures and alloantigenic determinants were identified in human Ia antigens from cells of an HLA-Dw7 homozygous cell line. The subsets carried DR7 specificity, BR4X7 supertypic specificity and MB2 supertypic specificity, respectively, and were immunospecifically separated by the use of operationally monospecific alloantisera. These specificities showed HLA-linked segregation in families and they were distributed in the population according to different but partially overlapping patterns. On peptide mapping analysis, the three subsets showed marked differences in the beta-chains. The alpha-chains of DR7 and BR4X7 subsets were very similar to each other, whereas the alpha-chains of MB2 subset were distinctive from those of DR7 and BR4X7. These data indicate the presence of a minimum of three HLA-linked loci; DR locus, a locus that encodes BR4X7, and a locus that encodes MB2, and substantiate the three-loci concept for the genetic control of human I antigens.

Immunological dissection of human Ia molecules.

The immunochemical analysis of Daudi Ia molecules by a variety of alloantisera led to the recognition of at least three molecular species carrying different antigenic determinants: DRw6, DC-1, and DC-2. Genetic as well as structural evidence indicates that DRw6 and DC-1 molecules are controlled by separate, HLA-linked loci, rather than by alleles at the same locus. The alloantigenic determinants appear to be expressed on the small Ia subunit. DC-1 and DC-2 determinants discussed had not been defined by serological analysis at the population level, but were demonstrated to be present by immunochemical analysis at the molecular level.

Two brc-abl junction peptides bind HLA-A3 molecules and allow specific induction of human cytotoxic T lymphocytes.

Intracellular processing of the products of the bcr-abl junction region in CML Philadelphia chromosomes would generate novel peptides which, if they are capable of binding to HLA-class I molecules, would be potential targets of a cytotoxic T cell response. The 18 nonamers corresponding to the b2-a2 and b3-a2 fusions and differing from the parental bcr and abl sequences for at least one amino acid have been synthesized and tested for binding with HLA class I alpha chain preparations from HLA-homozygous B lymphoblastoid cell lines. Two peptides derived from the b3-a2 junction bound to HLA-A3 and elicited detectable specific CTL responses in vitro. The binding affinity of one of the two peptides could be increased by appropriate substitutions of the anchor residues with those of the known HLA-A3 anchor motifs. More important, the modified peptide had increased capacity to prime a specific CTL response in vitro. The interaction with HLA-A3 of these two peptides and their substitution derivatives seems to be promising for target trials aimed at eliciting a specific CD8 T cell response against CML.

Influence of ionic strength on the antibody binding of MHC products.

Ionic strength has a strong influence on the binding of human Ia and HLA(A,B,C) molecules by the corresponding antibodies. A control system, BSA-anti-BSA, was influenced only minimally. The effect on the binding of anti-beta 2-microglobulin was very small when isolated chains were used, but more pronounced when using whole HLA(A,B,C) molecules. An inverse relationship exists between degree of inhibition and antibody concentration.

Microfingerprinting analysis of human Ia molecules favours a three loci model.

Alpha subunits from DC1 Ia molecules, when compared with DR alpha subunits, are shown to possess distinctive features revealed by differences in microfingerprinting patterns after peptic digestion. Alpha chains from BR4X7 molecules differ from DC1 alpha chains and are more similar to DR alpha chains. Since DC1 and BR4X7 beta chains (which carry the HLA-controlled alloantigenic determinants) associate with different alpha subunits, it is considered unlikely that they are controlled by alleles at the same locus. The proposed model implies the existence of three tightly linked HLA loci controlling the beta subunits of DR, DC and BR molecules respectively.

The importance of secondary anchor residue motifs of HLA class I proteins: a chemometric approach.

In this paper we report a chemometric approach to Quantitative Structure-Activity Relationship (QSAR) analysis applied to a study of the binding of peptides to Major Histocompatibility Complex (MHC) class I proteins. Peptides which possess the known primary anchor residue motif for HLA-B27 binding do not necessarily bind to HLA-B27 proteins. Secondary anchor residues are also involved, but it is not yet clear which amino acids are required or in which positions. A classic approach to this problem would be to synthesize multiple peptides each varying by a single amino acid from a starting peptide, and test them for their binding properties. Not only is this approach inefficient, but it is essentially unable to provide information about possible mutual interactions of amino acid residues in different positions. Using a statistical design to select the most informative compounds to use in the QSAR study, it was possible to analyse the effects on HLA-B27 peptide binding of different amino acids in four positions by means of only nine peptides. The relative binding activity of these peptides could then be modeled mathematically to provide information about the relative contribution of each of the four positions and to suggest a new peptide with high binding affinity. Our results demonstrate the usefulness of the chemometric strategy for studying peptides of interest in molecular immunology.

Papin-solubilized Ag-B antigens. II. Characterization of small sized Ag-B molecules.

Papain solubilization of rat Ag-B histocompatibility antigens produces Ag-B molecules of about 59,000 daltons which have been shown to contain two fragments bound noncovalently: one fragment about 37,000 daltons carrying Ag-B allospecificity, and another about 11,000 daltons, an apparent rat homologue of human beta2-microglobulin. Beside the 59,000-dalton Ag-B molecules, papain digests of liver cell membranes of ACI strain rats were found to contain Ag-B molecules of about 25,000 and 35,000 daltons. These smaller Ag-B molecules carried Ag-B private specificity of the rat strain (i.e., Ag-B4), as did the 59,000-dalton Ag-B molecules, and accounted for 40% of the solubilized Ag-B alloantigenic activity. The smaller Ag-B molecules were tested for the antigenic specificities that are characteristic of each of the two fragments of the 59,000-dalton molecules and detected by rabbit antiserum against rat cell membranes. The 35,000-dalton Ag-B molecules were found to contain the Ag-B 11,000-dalton fragment (i.e., rat beta2-microglobulin homologue) and to differ from the 59,000-dalton Ag-B molecules only in absence of a part of the 37,000-dalton fragment portion. The 25,000-dalton Ag-B molecules did not contain the rat beta2-microglobulin homologue and contained only a single component that is similar to the alloantigenic fragment portion of the 35,000-dalton Ag-B molecules. Similar 25,000-dalton Ag-B molecules (carrying Ag-B1 private specificity) of a single component were found in Fischer rat material. They accounted for 10% of the solubilized Ag-B alloantigenic activity.

Papain-solubilized Ag-B antigens. I. Isolated and characterization of two components composing Ag-B antigens.

Ag-B antigen molecules of about 59,000 daltons were partially purified from papain digests of liver cell membranes of Fischer and ACI rats. These preparations were radioiodinated and the labeled Ag-B antigen molecultes were isolated as specific immune complexes with alloantibodies directed to Ag-B1 or Ag-B4. These specifically purified Ag-B antigen molecules were found to give two fragments of 37,000 and 11,000 daltons on sodium sulfate-acrylamide gel electrophoresis. The two fragments (or very similar ones) were isolated from the radioiodinated partially purified Ag-B antigen preparations by acid dissociation and subsequent gel filtration. The 37,000-dalton fragment retained the same Ag-B alloantigenic specificity as the parental 59,000-dalton Ag-B antigen molecules, whereas the 11,000-dalton fragment did not carry any detectable Ag-B alloantigenic activity. In the reaction with rabbit antisera raised against rat cell membranes, each fragment was shown to be antigenically distinctive.

Absence of reaction of a xenogenic anti-H-2 serum with mouse embryonal carcinoma cells.

A rabbit antiserum raised against papain-solubilized H-2 antigens has been used to investigate the eventual expression of H-2 antigens and related molecules on embryonal carcinoma cells and on other types of mouse cells. No material reacting with this serum could be detected on cells carrying the F9 antigen. It is concluded that no H-2 antigen or cross-reacting material is expressed on these cell types.

HLA-DR typing by radioimmunoassay.

A radioimmunoassay procedure is described by which peripheral blood lymphocytes can be typed for HLA-DR specificities. The major advantages of this method are the following: simple and reproducible procedure, no need for B lymphocyte separation, no need for optimal viability, and no need for preabsorption of antisera with platelets. This method will find an application in the genetic and biochemical analysis of the HLA complex, and in the clinical tests of Ia antigens for diagnostic or prognostic purposes and in retrospective transplant studies.

Absence of a serologically detectable association of murine beta2-microglobulin with the embryonic F9 antigen.

The association of murine beta2-microglobulin to the early embryonic F9 antigen has been investigated by indirect immunofluorescence and by radioimmunoassay. Although some cell lines carry both types of molecules, the beta2-microglobulin was not found expressed on primitive teratocarcinoma cells, where F9 antigen was primarily detected. It is concluded that the low m.w. (12,000 daltons) subunit that was reported to be associated to the F9 antigen is not the adult beta2-microglobulin.

Serological and functional analysis of the epitope clusters on Leu3/T4 antigen.

Monoclonal antibodies (MoAbs) were generated against cells or cell membrane glycoproteins of a human T-cell line, HPB-ALL. Five, designated MCN 1, 3, 12, 19, and 29, were found to be specific to helper/inducer T-cells; they gave a positive membrane staining to approximately 48% of peripheral blood lymphocytes and 84% of thymocytes and these proportions did not change upon costaining with Leu 3a, a known anti-helper/inducer T-cell MoAb. Furthermore, their reaction pattern with a panel of human lymphoid cell lines was identical to that of Leu 3a. A reciprocal binding blocking test showed that the epitopes reactive with the MCN MoAbs are divided into three separate clusters. The MCN 3- and Leu 3a-reactive epitopes formed a cluster and they appeared to be the same epitope. This cluster was well separated from that represented by the MCN 1-reactive epitope. The MCN 12-, 19-, and 29-reactive epitopes could be assigned to a third cluster. MCN 12 and 19 were probably toward the same epitope. A sequential binding test indicated that the three epitope clusters reside on the molecule carrying Leu 3a-defined epitope, i.e., the Leu3/T4 antigen. On the functional analysis, MCN 3 gave a profound inhibitory effect on T-cell proliferative response to MHC class II antigens, whereas other MCN MoAbs did not show any modifying effect on the T-cell function.

Polystyrene beads coated with antibodies directed to HLA class I intracytoplasmic domain: the use in quantitative measurement of peptide-HLA class I binding by flow cytometry.

Protein-reactive, conformation-independent anti-peptide antibodies were raised in rabbits against a C-terminal sequence SDSAQGSDVSLA, common to most HLA-A and -B locus products. Antibodies were coupled to 4.5-microm polystyrene beads through the Fc portion by the use of protein A. The antibody-coupled beads showed a high capacity to bind HLA-A and -B proteins as well as their alpha chains by the intracytoplasmic domain, keeping the extracellular domains solvent exposed. The density of HLA class I proteins bound on the beads was approximately the same as that on cultured B cells. The antibody beads made it possible to quantitate peptide-HLA class I binding, i.e., in vitro HLA class I assembly by flow cytometry. The assembly rate determined by the provisionally called flow cytometric HLA class I assay was 15%-19% for the reassembly of dissociated HLA class I proteins with the released selfpeptides. With single synthetic peptides, the highest rate so far obtained was 6.5%. The assay specificity and reproducibility were satisfactory.

A current view of the human Ia system based on serological and immunochemical analysis.

We have approached the analysis of the human Ia system by localizing antigenic determinants on subsets of Class II molecules. For this purpose two classes of reagents have been used: alloantisera and mouse anti human monoclonal antibodies (moabs). We summarize here the most recent data obtained by these two approaches.

Unfolded HLA class I alpha chains and their use in an assay of HLA class-I-peptide binding.

Unfolded HLA class I alpha chains were isolated from B-cell lysates by alkaline denaturation and subsequent gel filtration and used for the detection of HLA class-I-peptide binding. Binding to specific peptides in the presence of excess beta 2-microglobulin induced the unfolded alpha chains to refold and acquire a conformation that is specific to folded alpha chains. This conformational change was measured by a specific RIA that involves inhibition of the binding of 125I-labeled HLA-A2 alpha/beta dimers and rabbit anti-HLA-B7 serum absorbed with beta 2-microglobulin. This assay procedure does not require labeling of either test peptides or test class I proteins and does not seem to have specificity degeneracy. It is applicable to the detection of peptide binding by all HLA class I allelic proteins. Evaluation of the assay conditions and HLA allelic specificity of the peptide binding defined by the use of synthetic peptides are described here, including the technical details, specificity, and reproducibility.

A new allodeterminant on HLA-DQ molecules carrying the DQw3 specificity.

The la subset that reacts with alloantiserum HON known to possess a strong anti-DRw53 activity was isolated from a 125I-labeled Ia preparation obtained from cells of RPMI 8057 cell line (DR1,4) and was found on peptide mapping to be lacking in the pattern characteristic of DR-like molecules carrying the DRw53 specificity and to display the structural features of DQ molecules, particularly those carrying the DQw3 specificity. Distribution analysis on a panel of selected la-positive cell lines indicated that the specificity involved is associated only with DR4 and DRw9, differing from the known DRw53 pattern (DR4, 7, and w9) and also from the known DQw3 pattern (DR4 and 5). Reciprocal sequential binding experiments demonstrated that the HON-defined specificity resides along with DQw3 specificity on the same molecules. Thus, HON alloantiserum possesses two different antibody activities; one directed to DRw53 specificity and another directed to a new DR4- and w9-associated DQ specificity.

Further characterization of two "DR supertypic" specificities. Additional evidence that they reside on molecules different from those carrying HLA-DR locus specificities.

Peripheral lymphocytes from a panel of individuals who had been assayed for DR specificities by the conventional cytotoxicity assay were typed for DR "supertypic" specificities, DC1 and BR4 x 7, by the radioimmunoassay. A positive and a negative population were clearly distinguished for both specificities and the strong association of the DC1 specificity with DR1, 2, and w6 was confirmed as well as the BR4 x 7 specificity with DR4 and 7. Family study also supported this strong association. Appropriate papain digestion separated molecules carrying DC1 determinant from those carrying DR2 as well as from those carrying DRw6, and separated molecules carrying BR4 x 7 from those carrying DR4. Specificity analysis of the 8th Workshop antisera by use of these separated antigen preparations showed that some anti-DC1 antisera do not possess appreciable anti-DR1, 2, or w6 activity and vice versa. The same was found for DR4 x 7 in its relationship with DR4 and 7. The existing evidence could be explained most economically by assuming a genetic model of two loci in linkage disequilibrium each coding for analogous but distinct forms of the small (beta) subunits of Ia molecules.

Allele- and temperature-dependency of in vitro HLA class I assembly.

Allelic variations of in vitro HLA class I assembly have been investigated in both the absence and the presence of binding peptides by flow cytometry using human leukocyte antigen (HLA) class I alpha chains isolated by alkali treatment from cultured HLA homozygous B cells and polystyrene beads coated with anti-HLA class I alpha chain antibodies specific to the C-terminal segment (anti-HLA class I beads). The specificity of assembly was temperature dependent, while the stability of the assembled complex depended on the bound peptide. The efficiency of assembly was allele dependent and primarily ruled by the binding affinity of alpha chains with beta(2)m. Thus, an allele hierarchy could be defined for the binding of HLA-B alpha chain with beta(2)-microglobulin: B7, B18 > B35, B62 > B27, B51. Allele and temperature dependency was found in HLA class I reassembly on acid treated B cells. The HLA class I proteins, reassembled with specific single peptides, could be efficiently transferred to anti-HLA class I beads. These findings would be used to produce microspheres coupled at high surface density with oriented single-peptide loaded HLA class I molecules and also to improve the preparation efficiency of HLA class I tetramers by the use of site-specific biotinylation.

DR5-associated DC molecules carry three different allospecificities.

Antiserum 8w670 which had been classified as anti-DR5 in the Workshop analysis was found by the direct binding test to react with about 40% of Ia molecules in an Ia preparation from LG38 cells (DR5,5) that was deleted of DR molecules and DR-like molecules we call BR and enriched in DC molecules by pretreatment with a rabbit antiserum raised against alpha-subunits of DR and BR molecules. The specificity involved in the binding reaction was shown by the binding inhibition assay to be present in 100% of DR5-positive cases (53 out of 53) and 8w13-positive cases (2 out of 2) and in 23% of DR4-positive cases (4 out of 17). This association pattern did not correspond to any of the known supertypic specificities including the DC beta 4 and DC alpha 3, both of which had been found on DR5-associated DC molecules. Yet sequential binding analysis revealed that the 8w670-defined specificity was indeed present on the same molecules carrying DC beta 4 and DC alpha 3 specificities. This specificity was designated DC5.

The peptide-binding specificity of HLA-B27 subtype (B*2705) analyzed by the use of polyalanine model peptides.

Model peptides have been used to quantitate the effect on HLA-B*2705 binding of the spacing between primary anchor residues, the type of amino acid accepted in the P9 anchor position, and the type of amino acid accepted in the "secondary anchor positions" P3 and P7. Peptide binding was measured by the HLA class I alpha-chain-refolding assay. The results obtained show that (a) Among the model peptides differing in the spacing between anchor residues, the nonamer with Arg in P2 and Lys in P9 (R2, K9) has the maximum binding with B*2705 molecule. The decamer, with an extra Ala inserted between Arg and Lys (R2, K10), has much lower binding, and still lower binding is observed for the octamer, where an Ala is removed (R2, K8). (b) Besides the "classic" Lys and Arg, several other aminoacids such as Tyr, Leu, Ala, and Gln can be accepted in P9, but with significant differences in binding affinity. (c) Different amino acids in P3 have an influence on peptide binding. Trp and Phe have a favorable influence, whereas Lys and Val appear to hinder the binding. Some variations are seen also for different amino acids in P7.