Peter J. Thurlow

Extracellular ATP causes loss of L-selectin from human lymphocytes via occupancy of P2Z purinoceptors

Lymphocytes from normal subjects or patients with chronic lymphocytic leukemia are known to possess receptors for extracellular ATP termed P2Z purinoceptors whose physiological role is undefined. Addition of extracellular ATP (50–500 μM) to both normal and leukemic lymphocytes caused loss of binding of monoclonal antibodies to L‐selectin (CD62L) on the cell surface. UTP, ADP, and adenosine (all at 500 μM) had no effect on L‐selectin expression. Several features of the ATP‐induced loss of L selectin indicate that this effect is mediated by lymphocyte P2Z purinoceptors. First the loss was attenuated in isotonic NaCl medium compared to 150 mM KCl medium. Second the loss of L‐selectin was immediately halted by addition of Mg2+ ions in molar excess of ATP. The most potent nucleotide causing L‐selectin loss was benzoylbenzoic ATP (>10 μM) which is also the most potent agonist for the P2Z purinoceptor. Finally preincubation of lymphocytes with oxidized ATP, an irreversible inhibitor of P2Z purinoceptors, also inhibited ATP induced loss of L‐selectin. Extracellular ATP is known to open an ion channel associated with the P2Z purinoceptor on B‐lymphocytes which allows influx of Ca2+. However, ATP‐induced loss of L‐selectin did not require extracellular Ca2+. Moreover addition of the calcium ionophore, ionomycin, had minimal effect on L‐selectin expression. Staurosporine (500 nM), an inhibitor of protein kinase C, inhibited only 10% of ATP induced loss of L‐selectin but completely inhibited the loss of L‐selectin caused by 50 nM PMA. Thus extracellular ATP interacts with lymphocyte P2Z purinoceptors which leads to shedding of L‐selectin via a pathway which requires neither Ca2+ influx nor activation of protein kinase C. ATP may have a physiological role in the loss of L‐selectin which occurs during the interactions of lymphocytes with other cells.

A monoclonal anti-pan-T-cell antibody. In vitro and in vivo studies.

A murine monoclonal antihuman T cell antibody is described that is pan-T in that it reacts with all T cells, including those from thymus, but not with B cells or other cells, except for a small subpopulation of null cells. Biochemical analysis demonstrated that the antigen detected is a glycoprotein, Mr 50,000 daltons, in which sialic acid, alpha-fucose and alpha-mannose are the important terminal or subterminal carbohydrates. Further analysis indicated that the antigen is the E-rosette-forming-cell (ERFC) receptor or a closely associated structure. Three patients with renal allograft rejection were treated with the pan-T cell monoclonal antibody and, although the antibody was in excess and bound to circulating T cells, no lasting effect on T cell numbers or graft rejection was noted.

Characterization of activated lymphocyte-tumor cell adhesion.

This study demonstrates the variable expression of ICAM‐1 and leukocyte function antigen‐3 (LFA‐3) on four tumor cell lines (COLO526, K562, Daudi, and HT‐29). In addition, phorbol ester (PMA) activation of lymphocytes modulated LFA‐1 from a uniform to a clustered surface distribution; whereas after treatment with high levels of Mg2+ ions, the unique epitope for high‐affinity LFA‐1 was identified using clone Mab24. Using a flow cytometric adhesion assay it was demonstrated that PMA‐activated lymphocytes formed conjugates with COLO526 and Daudi, and that these conjugates were inhibited by anti‐CD2 with varying inhibition by LFA‐1 clones MHM24 and 25.3.1. When lymphocytes were induced to express the high‐affinity form of LFA‐1, conjugates were identified with COLO526, K562, and Daudi and these conjugates were sensitive to the presence of both CD2 and LFA‐1 antibodies. Further studies using confocal microscopy confirmed significant adhesion between peripheral blood lymphocytes pre‐treated with either PMA or high levels of Mg2+ and the adherent cell line COLO526. In conclusion, this unique study has demonstrated for the first time the important role of the active form of LFA‐1 on the lymphocyte cell surface for conjugate formation with an ICAM‐1‐expressing tumor cell; also, two pathways of cell signaling were identified for conjugate formation to occur. J. Leukoc. Biol. 67: 847–855; 2000.

Acquired protein S deficiency in a patient with systemic lupus erythematosus causing central retinal vein thrombosis.

A 16 year old girl with systemic lupus erythematosus (SLE) developed the rare complication of central retinal vein occlusion. Although classically a disease of older patients, it has been recognised in association with SLE but only in the presence of the lupus anticoagulant or antiphospholipid antibodies. The thrombosis occurred when free protein S concentrations were transiently reduced and there was no family history or other known causes of reduced protein S concentrations. No other prothrombotic risk factors were present.

Lymphocyte-facilitated tumour cell adhesion to endothelial cells: the role of high affinity leucocyte integrins

Aim: Lymphocytes transiently express an active form of the β_2 integrin LFA‐1 (LFA‐1Af) which has conformational changes in extracellular domains enabling higher affinity binding to the ligand ICAM‐1. In this study, we investigated the role of lymphocytes bearing LFA‐1Af as potential mediators of binding of ICAM‐1‐positive tumour cells to endothelium. Methods: LFA‐1 expression on 51Cr‐PBLs was modulated in order to express high affinity LFA‐1Af and conjugates were formed with 35S‐labelled COLO526. The binding of the conjugates to resting or IL‐1β‐stimulated human umbilical vein endothelial cells (HUVECs) was then assessed via a modified radioactive HUVEC binding assay. In addition, the binding of PBL‐COLO526 conjugates to HUVECs was demonstrated by confocal microscopy. Results: The binding of COLO526 to endothelial cells did not change significantly between unstimulated and stimulated HUVECs. In addition, pre‐incubating the COLO526 with fresh PBLs did not significantly alter the binding of COLO526 to resting or activated HUVECs; whereas, in the presence of PBLs with LFA‐1Af, the COLO526 conjugate binding dramatically increased from basal levels to 41% on resting HUVECs and 81% on stimulated HUVECs. COLO526‐PBL(LFA‐1Af) conjugate adhesion to stimulated HUVECs was inhibited by blocking antibody to LFA‐1 (50%), VLA‐4 (32%) or L‐selectin (40%). Antibodies to the HUVEC adhesion molecules ICAM‐1, VCAM‐1 and E‐selectin also inhibited COLO526‐PBL(LFA‐1Af) conjugate binding to activated HUVECs by 79, 60 and 73%, respectively. Conclusions: PBLs bearing LFA‐1Af can enhance COLO526 adhesion to both resting and activated HUVECs. Furthermore, blocking studies demonstrate that a range of pathways are involved in this phenomenon (LFA‐1/ICAM‐1, VLA4/VCAM‐1, L‐selectin/E‐selectin). These studies have identified a novel alternative pathway for lymphocyte‐facilitated tumour cell adhesion to endothelial cells.

Detection of glycoprotein IIb and IIIa by monoclonal antibodies

Murine monoclonal antibodies were produced against human platelet membranes and screened on platelets by a 125I protein A radioimmunoassay. Several clones produced platelet specific antibodies as they showed no reaction with peripheral blood lymphocytes, neutrophils, bone marrow (excluding megakaryocytes) or several cell lines. Two antibodies (designated anti‐HuPl‐mla and anti‐HuPl‐mlb) were of particular interest in that although platelet specific they were non‐reactive with platelets from a thrombasthenic patient. In functional assays these two antibodies could specifically inhibit ADP and collagen induced aggregation of platelets and release of ATP, retard platelet aggregation and ATP release induced by epinephrine, and inhibit ADP induced platelet fibrinogen binding. These two antibodies appear to recognize glycoproteins IIb and IIIa as analysis by SDS‐PAGE using radiolabelled membranes revealed a two chain structure of molecular weight 112 000 and 122 000 daltons when run after reduction and 87 000 and 140 000 daltons non‐reduced.

A monoclonal antibody detecting a new human T cell antigen, HuLy-M2

HuLy-m21 is a new T cell antigen detected by a monoclonal antibody. The antibody (anti-HuLy-m2) reacts with 85% of human T cells, 20% B cells, and 50% null cells in peripheral blood. The antibody is cytotoxic, binds protein A, and is of the IgG2a subclass. HuLy-m2 antigen consists of a glycoprotein dimer Mr 37,000 subunit structure, in which sialic acid and carbohydrate play some role in antigenicity. Capping studies showed the HuLy-m2 antigen to be distinct from the previously described OKT-3, 4, 8, HuLy-m1 (OKT11), and HuLym3 antigens.: HuLy-m2 is a new T cell antigen detected by a monoclonal antibody. The antibody (anti-HuLy-m2) reacts with 85% of human T cells, 20% B cells, and 50% null cells in peripheral blood. The antibody is cytotoxic, binds protein A, and is of the IgG2a subclass. HuLy-m2 antigen consists of a glycoprotein dimer Mr 37,000 subunit structure, in which sialic acid and carbohydrate play some role in antigenicity. Capping studies showed the HuLy-m2 antigen to be distinct from the previously described OKT-3, 4, 8, HuLy-m1 (OKT11), and HuLy-m3 antigens.

HLA (class I) antigens on platelets are involved in platelet function.

HLA (Class I) antigens are ubiquitous in their cellular distribution and, while their function in major histocompatability complex (MHC)‐restricted phenomena are clear, their function on other cells, such as platelets, is not so obvious. We now report that several anti‐HLA monoclonal antibodies (including an anti‐β2 microglobulin antibody) selectively affect platelet function in that three different anti‐HLA monoclonal antibodies caused not only the aggregation of human platelets, but also caused the release of 14C‐serotonin. In addition, the anti‐HLA antibodies could selectively block the binding of several platelet agonists such as collagen, adrenalin, ADP, but not the binding of others such as thrombin and arachidonic acid. In blocking studies there also appeared to be an association between platelet glycoprotein IIb‐IIIa and HLA Class I antigens. We propose that both heavy and light chains of Class I HLA antigens on platelets may be involved in platelet aggregation and release and suggest an additional role for HLA antigens on platelets.

Antithyroid and antiadrenal autoantibodies in antiglomerular basement membrane disease, thin basement membrane disease and Alport syndrome

Summary The basement membranes of the glomerulus, thyroid and adrenal all contain the Goodpasture antigen, the target of autoantibodies in antiglomerular basement membrane (GBM) disease. Antithyroid antibodies can be associated with antiGBM disease, and there have been occasional reports of antithyroid antibodies in Alport syndrome, an inherited kidney disease where the GBM lacks the Goodpasture antigen. The aim of this study was to determine how often antithyroid and antiadrenal autoantibodies occurred in antiGBM disease, Alport syndrome and a related condition, thin basement membrane disease (TBMD).. Sera from patients with antiGBM disease (n = 19), Alport syndrome (n = 5) or TBMD (n = 13) were tested for antithyroglobulin, antithyroid microsomal and antiadrenal antibodies. Five of the patients with antiGBM disease (5/19, 26%, P NS) had antimicrosomal, and one had antithyroglobulin, antibodies (1/19, 5%, P NS). No patient with Alport syndrome had antithyroid antibodies. One with TBMD (1/13, 8%, P NS) had antithyroglobulin and antimicrosomal antibodies at titres of 1/400 and 1/25,600, respectively. Both patients with antithyroglobulin antibodies had previously been diagnosed with hypothyroidism. No one with antiGBM disease, Alport syndrome or TBMD had antiadrenal antibodies.. Antithyroid microsomal antibodies do not occur significantly more often in patients with antiGBM disease than in normals, and antithyroid and antiadrenal antibodies are not associated with Alport syndrome or TBMD.Abbreviations: GBM, glomerular basement membrane; TBMD, thin basement membrane disease.

INVESTIGATION OF T AND NATURAL KILLER CELL FUNCTION WITH MONOCLONAL ANTIBODIES

The relationship of T cell and natural killer (NK) cell antigens was examined using two monoclonal anti-T cell antibodies. The first--anti-HuLy-m1 (identical to OKTlla), reactive with all E-RFC+ cells and 50% null cells including NK cells--could partially block (greater than 50%) NK activity. The second--anti-HuLy-m2, reactive with T cells, some B cells, and null cells--could stimulate and not block NK activity. The enhancement of NK cell function by anti-HuLy-m2 was not related to interferon stimulation, and neither antibody altered effector:target cell binding. When used together neither augmentation nor blocking effects were observed, indicating different sites of reaction with NK cells. To further explore a possible association between T cell and NK antigens, anti-HuLy-m1 and -m2 were used to investigate certain T cell functions. Neither antibody was mitogenic for peripheral blood lymphocytes (PBLs) or blocked PHA stimulation of PBLs. In addition anti-HuLy-m1 did not inhibit a mixed lymphocyte culture (MLC) but could inhibit cytotoxic T lymphocytes (CTLs) almost as effectively as OKT3. However, anti-HuLy-m2 had no effect in an MLC and showed only minimal inhibition of a CTL assay. These studies demonstrate a relationship of the membrane structure of T cells and NK cells at an antigenic and functional level.