Complete prevention of the clinical expression of adjuvant-induced arthritis in rats by cyclosporin-A and lobenzarit: the regulation of lymph node cell populations and cytokine production.

A single dose of either cyclosporin-A (CsA) or lobenzarit (CCA) given with an arthrogenic adjuvant completely prevented expression of experimental adjuvant arthritis in rats. The aim of this study was to understand how these drugs prevented the arthritis expression by studying the popliteal lymph nodes draining the arthritic joints at various times after adjuvant injection. Neither drug affected the proliferation in popliteal lymph nodes at the time arthritis was normally expressed, however, there was a marked change in the types of cells present. Immunofluorescence assays showed a reduction in the proportion of CD4+ cells, while the proportion of B-lymphocytes was almost doubled. This coincided with a marked elevation in the ability of these cells to produce interleukin (IL)-6. At the same time production of other cytokines (IL-2, tumour necrosis factor (TNF) and interferon (IFN)-gamma) was not greatly affected. However, one day after adjuvant injection IL-2 and IFN-gamma production was reduced. In vitro experiments showed that IL-6 production by lymphoid cells was relatively unaffected by CsA and CCA but IL-2, TNF and IFN-gamma were suppressed by CsA. The results indicate that CsA and CCA may modify the response to the arthritic adjuvant by specifically inhibiting IL-2, TNF and IFN-gamma production at the time of adjuvant injection. The lack of inhibition of IL-6 by these drugs reveals it may not play a key role in the initiation of this model of chronic inflammation.

Signal transduction via Fc gamma R and Mac-1 alpha-chain in monocytes and polymorphonuclear leucocytes.

Some (VIM12, Leu-15, 5A4.C5), but not all, Mac-1-specific monoclonal antibodies (mAb) induced a clear respiratory burst in unprimed monocytes but not in unprimed polymorphonuclear leucocytes (PMN). We showed that this monocyte stimulation occurred via formation of Mac-1 mAb-Fc gamma RI or Mac-1 mAb-Fc gamma RII complexes, as human monomeric IgG1 could completely block the respiratory burst induced by the murine IgG2a subclass anti-Mac-1 mAb Leu-15 and the Fc gamma RII-specific mAb IV.3 inhibited respiratory burst formation by IgG1 subclass anti-Mac-1 mAb VIM12 and 5A4.C5, respectively. F(ab')2 fragments of mAb VIM12 did not stimulate. This association between Mac-1 and Fc gamma RII may be due to a near spatial association between these molecules in monocytes, as we observed partial inhibition of FITC-labelled anti-Fc gamma RII mAb IV.3 binding after prior incubation with mAb VIM12. If monocytes were preincubated with mAb IV.3 or aggregated IgG, there was partial inhibition of mAb VIM12 binding. The non-stimulating anti-Mac-1 mAb (JML.H11,44, OKM1, LM2/1, Mo1) did not show any significant competition with mAb IV.3 binding to Fc gamma RII. Both non-stimulating CD18-specific mAb, however, showed strong competition with mAb IV.3 binding to Fc gamma RII. On unprimed PMN, the situation was different. No Mac-1-specific mAb induced a respiratory burst and there was no competitive inhibition between anti-Mac-1 mAb and antibodies binding to Fc gamma RII. In interferon-gamma (IFN-gamma)-primed PMN, however, we observed a functional association between Mac-1 and Fc gamma RI as IgG2a subclass mAb Leu-15 induced a respiratory burst which could be inhibited by monomeric human IgG1, as observed in monocytes. However, no other anti-Mac-1 mAb was able to induce a respiratory burst in IFN-gamma-primed PMN. Therefore, a similar signal transducing capability may exist between Mac-1 and Fc gamma RI on both monocytes and PMN, despite a different relationship between Mac-1 and Fc gamma RII on these cell populations. As no Mac-1 beta-chain-specific (CD18)mAb were able to induce a respiratory burst in monocytes, despite being able to interact with Fc gamma R via their Fc regions, as detected by competition with mAb IV.3 for binding to Fc gamma RII, we conclude that intracellular signalling via Mac-1 mAb-Fc gamma RII complexes requires the alpha-chain.

Is aspirin a prodrug for antioxidant and cytokine-modulating oxymetabolites?

Aspirin and salicylate are transformed by stimulated human polymorphonuclear leucocytes (PMN), likely to be found at inflammatory sites, into both 2,3- and 2,5-dihydroxybenzoates (DHB). These DHB inhibit both the production of hydrogen peroxide by stimulated human PMN and prostaglandin (PG) E2 by activated rat macrophages. In contrast, DHB stimulated production of interleukin (IL)-1 and tumour necrosis factor (TNF) but inhibited IL-6 production by rat macrophages. These effects were probably a consequence of PGE2 inhibition. Gentisate (2,5-DHB) and homogentisate (a tyrosine metabolite) inhibited the lymphoproliferative action of IL-1. Some related phenols, e.g. 5-aminosalicylate, inhibited H2O2 production but had little effect on PGE2 production. These findings suggest that the local synthesis of DHB may contribute to the overall anti-inflammatory activity of salicylate, which (unlike aspirin) has little direct effect on PG production.

Comparison of techniques for the assessment of polymorphonuclear leukocyte polarisation in suspension.

Polarisation of polymorphonuclear leukocytes (PMN) in suspension was assessed using three techniques: 1) visual classification; 2) computerised morphometry; and 3) flow cytometry. While visual classification detected the formation, polarisation and type of cytoplasmic extensions produced by PMN, morphometry and flow cytometry detected only the formation of extensions. The area, perimeter and ellipticity were, in general, statistically different for each subtype of PMN-shape identified by visual classification. Furthermore, the magnitude and direction of changes detected by flow cytometry were affected by the use of erythrocyte lysis (during isolation of the cells) and the fixative used prior to analysis. The findings of this study demonstrate that visual classification is a more sensitive, reliable and appropriate assay of PMN polarisation than current morphometric and flow cytometric methods.

Phenotypic analysis of functionally associated molecules on peripheral blood and synovial fluid monocytes from arthritis patients.

Surface expression of 16 different membrane molecules was analyzed in peripheral blood and synovial fluid monocytes from patients with rheumatoid arthritis and reactive arthritis compared to controls. The most significant findings were modulated expression of function-associated FcRI, CR1, CR3, MHC class II and activation-associated CD31, M5, and M6 molecules in arthritis patients compared to controls. Of these molecules, only upregulated expression of MHC class II has previously been reported in synovial fluid monocytes of patients with rheumatoid arthritis.

Molecular characterization and functional analysis of the leukocyte surface protein CD31.

The CD31 Ag is a surface glycoprotein of 130 kDa with a broad cellular distribution. We show that among peripheral human blood cells, it is expressed on monocytes, granulocytes, platelets, and a subpopulation of lymphocytes. Activation of granulocytes leads to down-regulation of CD31 molecule expression. Sequence analysis and quantitative measurements of the relatedness of the CD31 molecule to other known proteins demonstrate that it consists of six Ig constant domains and that each domain bears substantial similarity to Fc gamma R domains. We find, however, that the CD31 molecule does not bind Ig Fc domains. On human monocytes we demonstrate that CD31 mAb recognizing certain epitopes of the CD31 molecule induce the generation of reactive oxygen metabolites. No such effect was seen with human granulocytes. By using two CD31 mAb, termed 1B5 and 7E4, we analyzed the requirements for activation of the monocyte respiratory burst via CD31 Ag in more detail. We show that signal transduction occurs via formation of a CD31 Ag-mAb-Fc gamma R complex involving either Fc gamma RI (CD64) or Fc gamma RII (CDw32) molecules.

M5, a phosphoinositol-linked human myelomonocytic activation-associated antigen.

When investigating the previously described monoclonal antibody (MoAb) VIM-5, raised against THP1 cells and binding to human monocytes and granulocytes, we found that the antigen detected by this antibody, designated M5, becomes very strongly expressed on monocytes after overnight culture with phorbol myristate acetate (PMA) or lipopolysaccharide (LPS) but not with recombinant human interferon-gamma (rhIFN-gamma). Granulocytes stimulated with formyl-methionyl-leucyl-phenylalanine (FMLP) become negative for binding VIM-5. Immature granulocytes from bone marrow do not express M5, thus its expression on granulocytes is differentiation linked. The antigen bound by VIM-5 is sensitive to hydrolysis by phosphoinositol-specific phospholipase C (PI-PLC). The immunoprecipitated M5 antigen on monocytes is a broad band, with a peak of 50 kD (unreduced) and two bands of 53 kD and 44 kD (reduced). We have therefore detected an antigen that is upregulated on stimulated monocytes but, conversely, down-regulated on FMLP-stimulated granulocytes.

Expression of a 150-kD cell surface antigen identified by monoclonal antibody YB5.B8 is associated with poor prognosis in acute non-lymphoblastic leukaemia.

Peripheral blood specimens, obtained from 71 patients with newly-diagnosed acute non-lymphoblastic leukaemia (ANLL) prior to the initiation of therapy, were assayed for the presence of a myeloid leukaemia-associated cell surface antigen identified by monoclonal antibody YB5.B8. The antibody bound to cells from 22 patients, and these patients had a poorer overall survival rate than those whose cells failed to bind the antibody (p less than 0.025). Fifty patients were treated with daunorubicin/cytosine arabinoside/6-thioguanine (DAT) according to a standard protocol and survived at least to the end of the induction phase (7 days). Of the 34 patients whose cells were YB5.B8 negative, 28 obtained a complete remission. In contrast, only four of the 16 patients whose cells expressed YB5.B8 antigen obtained complete remission (p less than 0.001). Expression of the YB5.B8 antigen in ANLL appears to be a strong prognostic indicator which is independent of other known prognostic factors such as patient age, leucocyte count and pre-existing hematopoietic abnormality.

Specificity of a mouse monoclonal antibody raised against acute myeloid leukaemia cells for mast cells in human mucosal and connective tissues.

A mouse monoclonal antibody raised against acute myeloid leukaemia cells (YB5.B8 monoclonal antibody; Gadd, S. J. and Ashman, L. K. (1985): Leukaemia Res. 9, 1329-1336) has been found by an indirect immunoperoxidase technique to bind to scattered cells in frozen sections from a number of human tissues. They have been identified as mast cells in fixed sections of skin, tonsil and duodenum by simultaneous staining of glycosaminoglycan with Alcian blue in 0.7 N HCl. The antibody does not distinguish mast cells in mucosal tissues from those in connective tissue, although the level of expression by cells at both sites appears to be heterogeneous. With the exception of low affinity binding to B lymphocytes, no other bone marrow-derived cells were found to bind the antibody. In particular, basophils and eosinophils were not stained, suggesting that they are not related closely to mast cells and that the antigen detected by YB5.B8 monoclonal antibody is not an IgE Fc receptor. Therefore, among all mature haemopoietic lineages, the antibody is specific for mast cells.

Specificity of the passive antibody-induced suppression of the humoral immune response of mice to surface antigens on human cells.

The effect of passively administered antibody on the humoral immune response of BALB/c mice to antigenic determinants on human cells has been examined. Antiserum raised by immunizing mice with the human leukaemic cell line K562, which lacks HLA-A,B,C antigens, was administered to mice, together with the HLA-A,B,C-positive cell line, BALM-1. The antibody response to the unique antigen was assessed by measuring the ability of the resultant antiserum to inhibit the binding to BALM-1 cells of a labelled monoclonal antibody, 7B6, which is specific for a monomorphic HLA-A,B,C determinant. As an indication of the immune response to antigens common to K562 and BALM-1, the ability of the same antiserum to inhibit the binding of monoclonal antibody 6B1, which detects an epitope common to both cell lines, was measured. Passive antibody to K562 blocked the immune response of mice to the common antigen on BALM-1 cells. However, the response to the antigen not recognized by the passive antibody was unaffected, even though the two antigens were present on the same immunizing cell. Thus, the effect of passive antibody was 'determinant specific'. Similar results were obtained, irrespective of whether the i.v. or i.p. route of immunization was used, and whether the passive antibody was adsorbed onto the immunizing cells prior to injection, or administered separately. The blocking of the immune response did not depend on simple masking of the antigenic determinants by the passive antibody, since non-saturating amounts of antibody were effective. In addition, blocking activity was dependent on antibody class and on an intact Fc region. The latter considerations also imply that the outcome of passive antibody administration in this system was determined by factors other than the ability of the antigen-antibody complexes to interact directly with B cells, and indicate the importance of antigen processing and/or a mechanism such as antigen-reactive cell opsonization.

A murine monoclonal antibody specific for a cell-surface antigen expressed by a subgroup of human myeloid leukaemias.

Murine monoclonal antibody YB5.B8 was raised against leukaemic blasts from a patient with M1-type acute non-lymphocytic leukaemia (ANLL). The antibody, which is of IgG1 class, bound to the majority of leukaemic blasts in the immunizing population, but not to cells of an autologous EBV-transformed B cell line. The antigen was not detected on normal blood or bone marrow cells, or on any of the eleven haemopoietic cell lines tested. It was present on some cells in peripheral blood specimens from 7/37 patients with ANLL and 1/5 patients with chronic myelomonocytic leukaemia and one patient with myelofibrosis with blastic change. In contrast, the antigen was not detected on cells in any of the 32 lymphoid leukaemic specimens tested, or on cells from four patients with chronic granulocytic leukaemia in accelerated chronic phase. In the ANLL group, expression of the antigen usually occurred on cells from types M1 or M2 according to the F.A.B. classification, and appeared to be associated with an unfavourable response to chemotherapy. The antigen was removed from the cell surface by digestion with pronase, and was re-expressed after 24 h in culture. Re-expression was prevented by the protein synthesis inhibitor, cycloheximide, but not by tunicamycin which inhibits glycosylation. Therefore, it seems likely that YB5.B8 binds to a peptide antigenic determinant.

Binding of mouse monoclonal antibodies to human leukaemic cells via the Fc receptor: a possible source of 'false positive' reactions in specificity screening.

Mouse monoclonal antibodies (MoAbs) of different classes and subclasses directed against antigens not expected to be present on human cells have been screened by indirect immunofluorescence using flow cytometry for binding to human non-lymphocytic leukaemic cells and normal peripheral blood leucocytes. Antibodies of the IgG2a and IgG3 subclass, but not of the IgG1, IgG2b or IgM class bound to the blast cell and monocyte populations in a peripheral blood mononuclear cell preparation from a patient with acute myelomonocytic leukaemia. F(ab')2 fragments of an anti-salmonella antibody of IgG2a subclass failed to bind, indicating that the results were not due to cross-reactivity with antigens of the cell membrane, thus implicating the Fc region in binding. Furthermore, binding was partly blocked by inclusion of 10% heat-inactivated normal rabbit serum in the assays. IgG2a bound to varying degrees to the leukaemic cell populations in seven of the nine non-lymphocytic leukaemic specimens tested, but no binding to normal peripheral blood mononuclear cells or granulocytes was detected. The results emphasize the importance of including appropriate controls when screening MoAbs for binding to various types of human cells.