A biochemical function for attractin in agouti-induced pigmentation and obesity.

Agouti protein, a paracrine signaling molecule normally limited to skin, is ectopically expressed in lethal yellow (A(y)) mice, and causes obesity by mimicking agouti-related protein (Agrp), found primarily in the hypothalamus. Mouse attractin (Atrn) is a widely expressed transmembrane protein whose loss of function in mahogany (Atrn(mg-3J)/ Atrn(mg-3J)) mutant mice blocks the pleiotropic effects of A(y). Here we demonstrate in transgenic, biochemical and genetic-interaction experiments that attractin is a low-affinity receptor for agouti protein, but not Agrp, in vitro and in vivo. Additional histopathologic abnormalities in Atrn(mg-3J)/Atrn(mg-3J) mice and cross-species genomic comparisons indicate that Atrn has multiple functions distinct from both a physiologic and an evolutionary perspective.

Immunochemical studies on the specificity of cellular hypersensitivity. The in vitro inhibition of peritoneal exudate cell migration by cehmically defined antigens.

The immunochemical specificity of antigen-induced inhibition of peritoneal exudate cell migration was studied in animals sensitized to chemically defined alpha,DNP(Lys)(18) peptides. It was shown that sensitized peritoneal exudate cells could discriminate between various DNP-oligolysines. Only immunogenic members of the homologous series of alpha,DNP-L-lysines equal to or larger in size than the heptamer inhibited the migration of specifically sensitized peritoneal exudate cells. In contrast, nonimmunogenic alpha,DNP-L-lysines, a D-lysine containing stereoisomer of alpha,DNP L(Lys)(9) (alpha,DNP-L(4)DL(4)) and (Lys)(9)epsilon, DNP were not inhibitory to the migration of peritoneal exudate cells derived from animals immunized to alpha,DNP(Lys)(18). The exquisite specificity of the in vitro reaction of sensitized cells with antigen contrasts with the previously observed in vivo or in vitro specificity of anti-alpha,-DNP(Lys)(n) antibody, but parallels the specificity of the in vivo delayed or anamnestic response. These results suggest the presence of a still undefined but highly specific binding site, which functions as the cellular receptor for antigen on the sensitized lymphoid cell or on some "processing" cell.

Antigen recognition and antibody specificity. Carrier specificity and genetic control of anti-dinitrophenyl-oligolysine antibody.

The exact specifiicity of anti-DNP antibody produced by Hartley guinea pigs immunized with a series of defined alpha,DNP and epsilon,DNP-oligolysines was studied by fluorescence quenching. All responder animals made anti-DNP antibody which recognized the precise chain length, +/- 1 lysyl residue, of the DNP-oligolysines used to induce the immune response as measured by an increase in binding energy (-DeltaF degrees ) for that antigen. The ability of the immune system to detect the smallest possible change in oligolysine chain length suggests that the anti-hapten antibody-forming cell possesses a highly specific recognition system for carrier conformation. When DNP-oligolysines are incorporated in an adjuvant containing M. tuberculosis H37Rv, both responder and nonresponder produce anti-DNP antibody, but only the responder develops delayed skin sensitivity. In addition to their failure to develop delayed hypersensitivity, nonresponders produced anti-DNP oligolysine antibody which did not show the increase in -DeltaF degrees for the immunizing antigen characteristic of responder antibody. These observations support a local environment hypothesis for antigen recognition at the level of the anti-hapten antibody-forming cell and suggest that the polylysine gene exerts its control at the same cell.

T and B cell in hapten-specific carrier-determined tolerance.

BDF1 mice were made tolerant by a single i.v. injection of 1 mg of DNAP-gamma1 or by weekly i.v. injections of 0.2 mg of DNP-gamma1 given for a month. In both instances, spleen cells of tolerant animals were fractionated to obtain pure populations of T cells (nonimmunoglobulin-bearing cells), referred to as tolerant T cells, and B cells (immunoglobulin-bearing cells) referred to as tolerant B cells (immunoglobulin-bearing cells) referred to as tolerant B cells. The control cells were similarly fractionated to obtain normal T and B cells. Mixtures of tolerant T cells and normal B cells, or conversely, normal T cells and tolerant B cells were used to repopulate lethally irradiated recipients. These recipients were then immunized with dinitrophenyl-keyhole limpet haemocyanin and in certain instances with other antigen horse red blood cells. The immune response to both antigens was measured using the direct hemolytic plaque assay. It was found that both T and B cells were tolerant and that tolerance was hapten specific at both T- and B-cell levels. While B-cell tolerance was demonstrated at a 1/1 T/B ratio, a 4/1 T/B ratio was necessary to show T-cell tolerance. Thus, the hapten-specific carrier-determined tolerance involves not only B cells but also T cells. The implication of this finding for the cellular mechanism of tolerance in an experimental model closely related to self tolerance is discussed.

Antigen recognition: in vitro studies on the specificity of the cellular immune response.

Studies of the immunochemical specificity of antigen-induced thymidine-2-(14)C incorporation in lymph node cells obtained from animals immunized to a series of closely related alpha-DNP-oligolysines, epsilon-DNP-oligolysines, and oligolysines have shown that the sensitized cell exhibits an extraordinary degree of specificity for antigen. The sensitized cell is maximally stimulated by the homologous immunizing antigen and can discriminate among compounds which differ from one another only in the position of a dinitrophenyl group or D-lysine residue on an identical oligolysine backbone. These studies support the view that the immunogen is not degraded prior to the induction of the immune response, and that the majority of cells produced as a consequence of immunization have stereospecific antigen receptors for the DNP-oligolysine used to induce the response; a smaller and more variably sized population of cells is produced with receptors specific for the oligolysine portion of the immunizing antigen. When specifically sensitized lymph node cell cultures are stimulated in vitro by heterologous DNP-oligolysines, the oligolysine- and not the DNP-oligolysine-sensitive population of cells appears to play a crucial role in the specificity of such cross-reactions. It is concluded from these studies that the antigen receptor on the sensitized lymph node cell differs in both kind and degree from conventional antibody. The chemical nature of the receptor and the means by which this receptor reacts with antigen to initiate the biosynthetic or proliferative cellular immune response still remain undefined.

Identification of the sequence required for expression of the 2H4 epitope on the human leukocyte common antigens.

The anti-2H4 antibody, which subdivides the T4+ population of human T lymphocytes into T4+, 2H4+ suppressor-inducer cells and T4+, 2H4- helper cells, recognizes an epitope on a subset of the human leukocyte common antigens (LCAs). LCAs are a family of cell surface glycoproteins generated from a single gene by the differential usage of three exons near the NH2-terminus. Using cDNA clones corresponding to four of the different forms of LCA molecules, extracellular domains of the LCA molecules were synthesized in vitro. Immunoprecipitation of these molecules with the anti-2H4 antibody demonstrated that exon A is required for the expression of the 2H4 epitope.

The subdivision of the T4 (CD4) subset on the basis of the differential expression of L-C/T200 antigens.

The T4 (CD4) subset of T lymphocytes has been subdivided into two major subsets, a suppressor/inducer subset (T4+,2H4+) and a helper subset (T4+,2H4-) on the basis of the differential expression of the L-C/T200 (CD45) antigens. The 2H4 antigen itself comprises at least three distinct polypeptides at 125,200, and 220 X 10(3) Mr, of which the 200 and 220 X 10(3) Mr polypeptides constitute the highest Mr isoforms of a pool of five distinct L-C/T200 antigens. The T4+,2H4+ subset expresses at least four of these isoforms at 180, 190, 200, and 220 X 10(3) on the cell surface, while the T4+,2H4- subset expresses only the 180 and 190 X 10(3) Mr forms. Pulse-chase analysis and endoglycosidase treatment revealed that the 125 X 10(3) Mr chain of the 2H4 antigen is nonglycosylated, while the 200 and 220 X 10(3) polypeptides are structurally related and derived by N- and O-linked glycosylation from two nascent subunits at 150 and 160 X 10(3) Mr. The function of the T4+,2H4+ subset could be blocked only by an antibody reactive with the L-C/T200 isoforms enriched with O-linked oligosaccharides at 200 and 220 X 10(3) Mr.

Identification of amino acids at the junction of exons 3 and 7 that are used for the generation of glycosylation-related human CD45RO and CD45RO-like antigen specificities.

The CD45 transmembrane protein tyrosine phosphatase plays an essential role in lymphocyte activation. In humans, CD45 is composed of five isoforms that are generated by alternative splicing of three exons of a common precursor mRNA. Expression of the smallest molecular mass 180-kD CD45 isoform (CD45-O) results from splicing out of exons 4(A), 5(B), and 6(C), which encode peptide regions near the NH2 terminus, and is regulated during T cell maturation and activation. Two monoclonal antibodies (mAb), UCHL1 (anti-CD45RO) and A6 (anti-CD45RO-like), were studied that selectively bind to murine transfectant cells expressing the human CD45-O isoform. The anti-CD45RO-like A6 mAb, but not the anti-CD45RO UCHL1 mAb, also weakly reacted with transfectant cells expressing the human CD45 isoforms that contained exons 4 and 5(AB), or exon 5(B) encoded sequences. The structural basis of the antigen specificities of these two different human anti-CD45RO mAbs was investigated at the molecular level by using potential glycosylation-defective CD45-O isoform variants containing amino acid substitutions at the junction of exons 3 and 7. Replacement of the threonine residue at position 8 (last amino acid encoded in exon 3 and a putative O-linked carbohydrate anchorage site) by an alanine, completely abrogated the reactivity of the UCHL1 mAb, but did not affect that of the A6 mAb. Conversely, replacement of either the asparagine at position 174 or the serine at position 176 (the first two putative carbohydrate anchorage sites in exon 7) by alanine, abrogated the reactivity of the A6 mAb, but not that of the UCHL1 mAb. Both the UCHL1 and A6 epitopes were dependent on the presence of O-linked carbohydrates; and the UCHL1, but not the A6 epitope, was dependent on the presence of sialic acid. These results demonstrate a pivotal role for the amino acids encoded at the junction of exons 3 and 7 for the generation of glycosylation-related CD45RO epitopes that are expressed in a cell lineage- and activation-regulated fashion.

Immunochemical studies on the antigenic determinants required to elicit delayed and immediate hypersensitivity reactions.

Hartley and strain 2 guinea pigs were sensitized to chemically defined alpha-DNP(Lys)(n)-BuAm, alpha-DNP(Lys)(n)-, and (Lys)(n)-BuAm peptides and skin tested with individual members of these homologous series, related peptides and hapten-substituted proteins. The immediate skin response (Arthus) could be elicited with hapten-substituted tetra-, penta-, or hexamers, whereas both immediate and delayed skin responses could be provoked by the octamer or nonamer. The hapten is an integral part of the determinant for both immediate and delayed skin reactivity, since poly-L-lysine was unable to elicit either delayed or immediate reactions in sensitized animals. Arthus type cross-reactions occurred only when the sensitizing and test antigen shared a common haptenic determinant. In contrast to this, in this system, delayed type cross-reactions occurred only when the test antigen and the sensitizing antigen contained both a large oligo-L-lysine carrier as well as the same haptenic determinant. These observations imply that the mediation of the delayed response requires a larger determinant than is necessary to elicit the immediate response. The role of high affinity antibody as the mediator of the delayed response is discussed in terms of the size of the antigenic determinants required to elicit this response. It was found that the ability to elicit the delayed response paralleled the immunogenic capacity of these peptides, whereas the immediate response could be elicited by nonimmunogenic peptides. This finding suggests that the delayed response may require the continued biosynthesis of antibody and may be analogous to a local in vivo secondary response.

Antibody-dependent cellular cytotoxicity by allosensitized human T cells.

Peripheral human T cells, isolated by sheep erythrocyte-rosette formation and density centrifugation, were highly cytotoxic to both Ab-coated autologous lymphocytes and antibody (Ab)-coated chicken erythrocytes when stimulated in mixed lymphocyte culture, but were not lytic when freshly purified, or when unstimulated in 6-day culture. Allosensitized T cells were shown to effect this activity by a specific effector-target cell interaction dependent on Ab, as indicated by: (a) induction of killing by Ab to target cells not lysed in the absence of Ab. (b) inhibition of Ab-dependent killing by aggregated Ig. The mechanism by which allosensitized T cells effect antibody-dependent cellular cytotoxicity is discussed.

Delineation of an immunoregulatory amplifier population recognizing autologous Ia molecules. Analysis with human T cell clones.

Autoreactive T lymphocytes were generated by culturing human peripheral blood mononuclear cells with an antigen-specific major histocompatibility complex (MHC)-restricted autologous inducer T cell, termed RW17C and subsequently cloned in soft agar. The majority of such clones (AC1-13) expressed the T3+T4+T8-T11+Ia+ phenotype and were directed at autologous class II MHC gene products found on B cells, macrophages, and B lymphoblastoid cells as judged by their proliferative response to the latter. For this recognition, the clones employed a T3-Ti molecular complex and a T4 structure analogous to those found on allospecific T cells. Perhaps more importantly, it was observed that the same AC1-13 autoreactive clones (AC) induced autologous B cells to produce high levels of immunoglobulin in the absence of exogenous antigen and could synergize with the RW17C clone to effect maximal B cell Ig production. These results support the notion that such autoreactive cells can function in a physiologic amplifier role by facilitating induction via an internal set of signals (i.e. autologous MHC).

Human CD4 helper T cell activation: functional involvement of two distinct collagen receptors, 1F7 and VLA integrin family.

In the present study, we showed that activation of human CD4 T cells can be induced by anti-CD3 and collagen in a serum-free system. This activation was inhibited by the addition of peptides containing the RGD or Gly-Pro-X sequences. Significantly, we demonstrated that both the 1F7 (CD26) structure and the VLA integrin family, particularly the VLA-3 complex, contribute to the functional interaction between collagen and CD4 cells since anti-1F7 and anti-VLA-3 specifically inhibited this collagen-induced CD4 cell activation. Biochemical studies showed that the 1F7 structure is not a member of the VLA integrin family. These results thus indicated that two different families of antigens serve as functional collagen receptors for CD4 T cell activation.

A subset of natural killer cells in peripheral blood displays a mature T cell phenotype.

Normal human PBMC were analyzed for the presence of cells expressing both T3 and NKH1 antigens, using direct two-color immunofluorescence. In six individuals, NKH1+T3+ cells were found to represent 2.5% of PBMC and 24% of the total number of NKH1+ cells. Purified NKH1+T3+ cells were shown to have the typical morphology of large granular lymphocytes (LGL). NKH1+T3+ cells also exhibited spontaneous cytotoxicity against K562 target cells and this lytic activity could be inhibited by anti-T3 mAb. Similar results were obtained with NKH1+T3+ cells cultured in vitro in lymphocyte-conditioned medium. Taken together, these results indicate that NKH1+T3+ cells represent a unique population of NK-active cells in normal peripheral blood. Although these cells exhibit LGL morphology and NK activity, this appears to be mediated through a functional T cell-like receptor for target antigen.

Differential usage of three exons generates at least five different mRNAs encoding human leukocyte common antigens.

Leukocyte common antigens (LCAs, also known as T200 and CD 45) are integral membrane proteins expressed exclusively on hematopoietic cells. These molecules exhibit varying molecular masses and epitopes when expressed in different cell types. To determine the genetic bases for the generation of this diversity, three classes of human LCA cDNA clones that are different near their 5' ends have been isolated. These differences arose as a result of differential usage of three exons as determined from an analysis of a genomic DNA clone. Furthermore, Northern blot analysis with LCA exon-specific probes demonstrates the existence of at least two more LCA mRNA forms that are generated by differential splicing. A comparison of the human and mouse LCA protein sequences revealed a marked difference only in the extracellular domain.

Ligation of VLA-4 on T cells stimulates tyrosine phosphorylation of a 105-kD protein.

The VLA/integrins are a family of heterodimeric adhesion receptors shown to be involved in cell-to-cell and cell-to-extracellular matrix (ECM) interactions. Given recent evidence that VLA molecules can synergize with the CD3/T cell receptor (TCR) pathway to activate T cells, it is important to identify biochemical event(s) generated by these molecules. Here, we report that the engagement of VLA-4 on T cells with specific antibodies or its ligand activates protein-tyrosine kinase (PTK) activity as detected by antiphosphotyrosine immunoblotting. The crosslinking of VLA-beta 1 (CD29) with a specific monoclonal antibody (mAb) (anti-4B4) plus anti-mouse immunoglobulin resulted in the rapid tyrosine phosphorylation of a 105-kD protein (pp105) in the human T cell line H9, as well as in peripheral resting T cells. The increase in tyrosine phosphorylation of pp105 was specifically mediated by VLA-4, since mAbs against alpha 4, but not against other VLA alpha chains, could induce this phosphorylation. In addition, the binding of T cells with the CS1 alternatively spliced segment of fibronectin (the binding site recognized by VLA-4) induced pp105 tyrosine phosphorylation. Crosslinking the CD3 complex or VLA-4 molecules with mAbs demonstrated that each of these molecules stimulated the tyrosine phosphorylation of unique sets of proteins with different kinetics, suggesting that these two receptor systems are coupled to distinct PTKs. Since tyrosine phosphorylation of cellular proteins has been shown to be a crucial biochemical event in cell growth, our findings suggest that the induction of pp105 tyrosine phosphorylation via VLA-4 may play a role in the transduction of activation signals through this molecule.

A new member of the immunoglobulin superfamily that has a cytoplasmic region homologous to the leukocyte common antigen.

A human gene (LAR) that hybridizes to mouse leukocyte common antigen cDNA under relaxed hybridization conditions was isolated. The LAR gene is expressed in a broad range of cells, including T lymphocytes, kidney, and prostate cells. The structure of the protein encoded by the LAR gene was deduced by determining the nucleotide sequences of a 7.7-kb LAR cDNA. The putative LAR protein is composed of a 1,234 amino acid extracellular region, a 24 amino acid transmembrane segment, and a 623 amino acid cytoplasmic region. The cytoplasmic region contains two homologous domains that have extensive sequence similarity to the cytoplasmic region of the leukocyte common antigens. The NH2-terminal region of the extracellular segment of the LAR protein contains three tandem Ig-like domains and nine non-Ig-like domains. Among the known Ig-like proteins, the LAR protein has the highest degree of similarity to neural-cell adhesion molecule. The non-Ig-like domains of the LAR protein are also similar to the non-Ig-like domains of neural-cell adhesion molecule.

Peptide variability exists within alpha and beta subunits of the T cell receptor for antigen.

The T cell receptor for antigen (Ti) has recently been identified as a 90-kdalton T3-associated clonotypic structure composed of one 49-51-kdalton alpha and one 43-kdalton beta subunit, which are disulfide linked. Here, Ti molecules from two alloreactive CTL clones derived from the same donor but of differing specificities (CT8III and CT4II) are directly compared following isolation with anticlonotypic monoclonal antibodies. Isoelectric focusing shows that the alpha subunits (pI 4.4-4.7) are more acidic than the beta subunits (pI 6.0-6.2) but that each glycoprotein species is distinctive. More importantly, two-dimensional peptide maps of 125I-labeled surface receptors indicate that the beta chains of Ti1 and Ti2 appear unique and share only two peptides in common. In contrast, peptide maps of Ti1 and Ti2 alpha chains are more related although not identical. These results suggest that the human T cell receptor is composed of constant as well as variable regions and that at least one of the latter is located within the beta subunit.

Location of T cell and major histocompatibility complex antigens in the human thymus.

A series of monoclonal antibodies were used to study the intrathymic distribution of T cell-specific antigens, Ia antigens, and beta 2-microglobulin in frozen sections of human thymus by immunofluorescence and immunoperoxidase techniques. Most of the cortical thymocytes reacted with anti-T4, anti-T5, anti-T6, anti-T8, and anti-T10 antibodies, thus indicating coexpression of multiple antigens on cortical lymphocytes. The staining of cells in the medulla was most satisfactorily judged in sections stained with the immunoperoxidase technique. Many medullary cells reacted with anti-T4--and a smaller fraction with anti-T5, anti-T6, anti-T8, and anti-T10 antibodies. In addition, T1 and T3 antibodies, which react with all peripheral T cells, stained a majority of medullary cells. The medullary cells were also more intensely stained with antibodies directed against beta 2-microglobulin than the majority of cortical cells. Hence, the staining profile of medulla approximates the staining pattern of peripheral T cells, with large numbers of cells bearing T1+, T3+, and T4+ antigens (helper/inducer cells) and a small number of cells bearing T1+, T3+, and T5+/T8+ antigens (suppressor/cytotoxic cells). This supports the conclusion that mature cells present in the medulla are derived from immature cells in the cortex. However, a small number of cells scattered throughout the cortex stained with T1 and T3 antibodies, which suggests that maturation of thymocytes can also occur in the cortex. Antibody directed against Ia antigens resulted in a characteristic patchy pattern of staining in the cortex and in diffuse staining in the medulla, which was interpreted as resulting from staining of epithelial reticulum. The majority of thymocytes did not stain. The staining pattern suggests a close relationship between epithelial cells and thymocytes.

Inhibition of antibody-dependent cellular cytotoxicity and immunoglobulin synthesis by an antiserum prepared against a human B-cell Ia-like molecule.

Rabbit antisera to the human B-cell-specific antigen complex, p23,30, was used to define further the functional heterogeneity of isolated human lymphocyte subpopulations. Specific depletion of p23,30-bearing cells from Ig-negative cell populations and Ig-negative, E rosette-negative (Null) populations by either complement-mediated lysis or by physical separation on goat antirabbit Fab immunoabsorbent columns, eliminates the antibody-dependent cellular cytotoxic (ADCC) function. Furthermore, binding of anti-p23,30 serum to the effector cell surface inhibits ADCC but does not interfere with EA rosette formation. Apparently p23,30 represents a cell surface site which is distinct from the Fc receptor but which is important in the triggering of ADCC. In addition, depletion of p23,30-bearing cells from unfractionated cell populations, Ig-positive B-cell populations and Ig-negative, E rosette-negative (Null) populations eliminates the capacity of these populations to secrete immunoglobulin during subsequent culturing. Thus both the Ig-secreting cells and the ADCC effector cells within the Ig-negative, E rosette-negative (Null) population reside in the same population of cells which bears the p23,30 antigen.

The compartmentalization of antigen-reactive lymphocytes in desensitized guinea pigs.

A single injection of epsilon,DNP-Lys(7-10) can render previously sensitized guinea pigs specifically unreactive to subsequent intradermal challenge with that antigen. Antigen-reactive lymphocytes, as assayed by macrophage-migration in-inhibition or thymidine incorporation, were depleted from the peritoneal exudates of those animals. In contrast, it was intriguing to find that lymph node lymphocytes from such animals responded normally in the antigen-induced thymidine incorporation assay. These studies demonstrate a compartmentalization of antigen-reactive lymphocytes in desensitized animals which may account for the short-lived nature of this phenomenon.