Isoforms of c-KIT differ in activation of signalling pathways and transformation of NIH3T3 fibroblasts.

Alternate splicing of mRNA encoding c-KIT results in isoforms which differ in the presence or absence of four amino acids (GNNK) in the juxtamembrane region of the extracellular domain of the receptor. In this study we show that these isoforms of human c-KIT, expressed at similar levels in NIH3T3 cells, display differential effects on various attributes of transformation. The GNNK- isoform strongly promoted anchorage independent growth (colony formation in semi-solid medium), loss of contact inhibition (focus formation), and led to tumorigenicity in nude mice. In contrast, the GNNK+ isoform elicited colony formation but relatively poor focus formation and no tumorigenicity. Saturation binding analysis indicated that the isoforms do not differ significantly in their affinity for the KIT ligand, Steel Factor (SLF). Negligible ligand-independent receptor phosphorylation was observed in either case but, after ligand stimulation, the GNNK- isoform displayed more rapid and extensive tyrosine autophosphorylation and faster internalization. Both isoforms recruited the p85 subunit of phosphatidylinositol 3-kinase and led to similar phosphorylation of its downstream effector c-Akt, but the GNNK- isoform gave rise to more MAP kinase phosphorylation. Thus the c-KIT isoforms display different signalling characteristics and have different transforming activity in NIH3T3 cells.

Transformation of NIH3T3 fibroblasts by the c-Kit receptor tyrosine kinase: effect of receptor density and ligand-requirement.

Ectopic expression of the normal murine receptor tyrosine kinase, c-Kit, in NIH3T3 cells induced many phenotypic changes characteristic of transformation including anchorage-independent growth, focus formation and tumorigenicity in nude mice. Although transformation was largely dependent on the presence of recombinant murine Steel Factor (SLF), the ligand to the c-Kit receptor, anchorage independent growth did occur at a low frequency in the absence of added factor, and this could not be inhibited by neutralising antibodies or by SLF anti-sense mRNA. Clones from factor-independent colonies in semi-solid agar displayed a narrow range of c-Kit surface protein levels (4.3-6.4 x 10(4) receptors/cell) which was relatively high compared with the pool from which they were derived. Analysis of a larger series of random clones derived from adherent cultures expressing different levels of c-Kit demonstrated a positive correlation between SLF-dependent, anchorage-independent growth and c-Kit protein and mRNA expression levels (respectively, Rs = 0.58, P < 0.01; and Rs = 0.53, P < 0.01) with consistent colony formation observed with clones having > 2.5 x 10(4) receptors/cell. Interestingly, two of the three clones expressing the highest levels of c-Kit protein and mRNA produced few or no colonies in the presence or absence of SLF. Sequential overexpression of human c-KIT in NIH3T3 cells using a dihydrofolate reductase (DHFR)-encoding vector and gene co-amplification through methotrexate selection, which resulted in pools expressing up to 1.5 x 10(5) receptors/cell, confirmed that high receptor densities resulted in a decrease in colony numbers. Thus, analysis of clonal and selected populations has indicated that an optimal level of c-Kit is required for transformation of NIH3T3 cells in the presence of SLF, and that some ligand-independent transformation occurs.

Increased expression of c-Kit or its ligand Steel Factor is not a common feature of adult acute myeloid leukaemia.

Cell surface levels of the receptor tyrosine kinase P145(c-kit), the product of the c-kit proto-oncogens, in a panel of 80 primary adult acute myeloid leukaemia (AML) specimens collected at presentation were quantitated by immunofluorescence and flow cytometry, and compared with levels on CD34+ bone marrow cells from normal donors. Receptor levels on AML blast cells were extremely variable and were similar to, or less than, those on normal stem and progenitor cells. In general P145(c-kit) expression was higher on cells of immature phenotype (FAB M1 and M2). c-kit mRNA was quantitated by ribonuclease protection assay (RPA) and was shown to be correlated with cell surface protein expression (r=0.76; P<0.001). This indicates that ligand-mediated receptor internalisation or other mechanisms of increased protein turnover are not responsible for variations in the level of P145(c-kit) in AML specimens. Quantitative Southern blotting was used to examine c-kit gene copy number in 25 of these specimens and was found to be normal in all but one. Thus we have found little evidence of over-expression of c-kit in adult AML. mRNA for the c-kit ligand, Steel Factor (SLF) was also quantitated by RPA in these specimens. While SLF message was detectable (limit of detection approximately 10(4) copies per 10 microgram total RNA; equivalent to 1 copy per 100 cells) in 19% of cases, these specimens in general contained low levels of c-kit mRNA. Thus, an autocrine cycle involving c-kit and SLF does not appear to be a common feature of AML.

A monoclonal antibody that inhibits the action of GM-CSF on normal but not leukaemic progenitors.

Monoclonal antibody YB5.B8 was previously shown to inhibit haemopoietic colony formation in response to a complex growth factor supplement in vitro (Cambareri A. C., Ashman L. K., Cole S. R. & Lyons A. B. (1988), Leukemia Res. 12, 929). We now report studies of the effect of the antibody on colony formation by normal human bone marrow cells in response to recombinant human colony-stimulating factors GM-CSF, G-CSF and IL-3. MAb YB5.B8 significantly reduced the yield of colonies of all types examined (granulocyte-macrophage, granulocyte, macrophage and eosinophil) in response to GM-CSF but not to IL-3 or G-CSF. However, MAb YB5.B8 failed to influence the proliferation of the myelomonocytic leukaemia cell line RC-2A in response to GM-CSF, G-CSF or IL-3. Direct binding studies demonstrated the presence of low numbers of receptors for GM-CSF on RC-2A cells, however, the binding of this cytokine was not influenced by co- and/or pre-incubation with MAb YB5.B8. Therefore the antigen identified by YB5.B8 is probably not a receptor for GM-CSF and may indirectly influence the response of normal haemopoietic progenitors to this cytokine.

A monoclonal antibody to a human mast cell/myeloid leukaemia-specific antigen binds to normal haemopoietic progenitor cells and inhibits colony formation in vitro.

An antigen identified by murine monoclonal antibody YB5.B8 has previously been detected only on acute non-lymphoblastic leukaemia (ANLL) cells and tissue mast cells. We now report that the YB5.B8 antigen is present on a minor population (up to 3%) of normal bone marrow mononuclear cells which overlaps the set of progenitor cells capable of forming haemopoietic colonies in vitro. The results indicate that the antigen is a normal haemopoietic progenitor cell marker which is selectively retained on mast cells during maturation, and that leukaemias which express the antigen are not necessarily committed to the mast cell lineage. Furthermore, the antibody was capable of partially inhibiting the formation of haemopoietic colonies in vitro, indicating an important functional role for the antigen. This is consistent with the observation, reported in the accompanying paper, that expression of the YB5.B8 antigen is strongly correlated with poor response to therapy in patients with ANLL.

Effects of mutant c-kit in early myeloid cells.

Activating mutations in c-Kit, the receptor for Stem Cell Factor (SCF), have been identified in dysplasias and leukaemias of the mast cell lineage and have been shown to contribute to transformation in model systems. Early myeloid cells also normally express c-Kit and their survival, proliferation and differentiation is promoted by SCF. It might therefore be expected that c-Kit mutations could also be involved in some acute and/or chronic myeloid leukaemias. We have found that mutant c-Kit (and normal c-Kit in the presence of SCF) provides a strong differentiation stimulus in normal and immortalised murine early myeloid cells. Since maturation of haemopoietic cells, with the exception of mast cells, results in down-regulation of c-Kit expression, the transforming effects of mutant receptor may be self-limiting in most lineages. This is consistent with the observation that multipotential progenitor cells from some patients with systemic mastocytosis express mutant c-Kit. However, c-Kit mutations have been observed in a few cases of myelodysplastic syndromes or AML without mast cell features. Oncogenesis involves multiple genetic changes and the phenotype of malignant haemopoietic cells expressing mutant c-Kit may be influenced by co-oncogenic events. For example mutations blocking the differentiative effect of mutant c-Kit might result in AML rather than mastocytosis. Thus the extent to which c-Kit mutations contribute to malignancies of early myeloid phenotype remains unknown, and resolution of this issue is complicated by the heterogeneity of this family of diseases.

Effects of mutant c-Kit in early myeloid cells.

Activating mutations in c-Kit, the receptor for Stem Cell Factor (SCF), have been identified in dysplasias and leukaemias of the mast cell lineage and have been shown to contribute to transformation in model systems. Early myeloid cells also normally express c-Kit and their survival, proliferation and differentiation is promoted by SCE It might therefore be expected that c-Kit mutations could also be involved in some acute and/or chronic myeloid leukaemias. We have found that mutant c-Kit (and normal c-Kit in the presence of SCF) provides a strong differentiation stimulus in normal and immortalised murine early myeloid cells. Since maturation of haemopoietic cells, with the exception of mast cells, results in down-regulation of c-Kit expression, the transforming effects of mutant receptor may be self-limiting in most lineages. This is consistent with the observation that multipotential progenitor cells from some patients with systemic mastocytosis express mutant c-Kit. However, c-Kit mutations have been observed in a few cases of myelodysplastic syndromes or AML without mast cell features. Oncogenesis involves multiple genetic changes and the phenotype of malignant haemopoietic cells expressing mutant c-Kit may be influenced by co-oncogenic events. For example mutations blocking the differentiative effect of mutant c-Kit might result in AML rather than mastocytosis. Thus the extent to which c-Kit mutations contribute to malignancies of early myeloid phenotype remains unknown, and resolution of this issue is complicated by the heterogeneity of this family of diseases.

Differential tissue expression of epitopes of the tetraspanin CD151 recognised by monoclonal antibodies.

CD151, a member of the tetraspanin family of cell membrane proteins, is widely expressed in epithelial, endothelial and muscle cells as well as platelets and megakaryocytes. Several monoclonal antibodies recognising CD151 in transfected cells and immunoprecipitating typical bands of 28 and 32 kDa from cell lysates have been produced. Surprisingly, these antibodies show different patterns of staining on tissue sections and on haemopoietic cells. Here we show that these differences are at least in part due to masking of certain epitopes in integrin/CD151 complexes. These data have important implications for the use of monoclonal antibodies in studies of the distribution and function of CD151. Of six monoclonal antibodies from four laboratories, 11B1 was found to be the most reliable for detection of CD151 in different cell and tissue contexts.

Different epitopes of the CD31 antigen identified by monoclonal antibodies: cell type-specific patterns of expression.

3 mAb-5A2.G5, B2B1 and 2BD4-all of IgG1 isotype were identified as belonging to the CD31 cluster by their binding to transfected murine cell lines expressing the CD31 antigen and by sequential immunoprecipitation experiments. Competitive binding experiments were carried out using the human myelomonocytic cell line RC-2A. mAb B2B1 and 2BD4 did not cross block. mAb 5A2.G5 partly inhibited binding of 2BD4 but not B2B1. Thus the epitopes identified by 5A2.G5 and 2BD4 appear to overlap but to be quite separate from that recognized by B2B1. All 3 antibodies bound to a 130 kDa species on Western blots after nonreducing polyacrylamide gel electrophoresis of platelet lysates. Culture of RC-2A cells in the presence of tunicamycin (after removal of surface antigens by pronase) blocked re-expression of the epitopes recognised by all 3 mAb suggesting that they involve N-linked glycosylation. Furthermore, treatment of platelet lysates with Endoglycosidase F prior to electrophoresis and Western blotting abolished the binding of the mAb but not a rabbit polyclonal antiserum to CD31. Nevertheless, neuraminidase treatment of RC-2A cells failed to affect mAb binding. The antibodies displayed typical properties of CD31 antibodies in that all 3 precipitated at 130 kDa cell surface protein and bound strongly to platelets, monocytes, neutrophils and vascular endothelium. All 3 antibodies were positive on hemopoietic progenitor cells which give rise to colonies in the CFU-C assay. However, differences in binding to peripheral blood lymphocytes and to certain human leukemic cell lines were noted. In particular, mAb 5A2.G5 bound weakly to lymphocytes and to the lymphoid cell line HSB-2 compared with the other 2 antibodies.

Expression of the YB5.B8 antigen (c-kit proto-oncogene product) in normal human bone marrow.

The c-kit proto-oncogene product is a member of the family of growth factor receptors with intrinsic tyrosine kinase activity. In the mouse c-kit maps to the W locus, which is known to be of central importance in hematopoiesis. Monoclonal antibody (MoAb) YB5.B8, which was raised against peripheral blood blast cells from a patient with acute myeloid leukemia (AML), was recently shown to bind to the extracellular domain of the c-kit product. This antibody does not bind detectably to normal peripheral blood cells and identifies a sub-group of AML patients with poor prognosis. We have used MoAb YB5.B8 to study the expression of c-kit by normal human bone marrow cells by immunofluorescence and flow cytometry, and to isolate multipotential and erythroid colony-forming cells. In a series of 11 normal adult bone marrow specimens, MoAb YB5.B8 bound to 4.0% +/- 1.8% of the cells in the low-density fraction. Dual-labeling experiments were performed with YB5.B8, and CD33, CD34, and CD10 MoAbs. Three populations of cells binding YB5.B8 could be identified based on their pattern of coexpression of the other markers; ie, YB5.B8+/CD34+/CD33-, YB5.B8+/CD34+/CD33+ and YB5.B8+/CD34+/CD33+. These populations had distinctive two-dimensional light scatter characteristics and are likely to correspond to precursor colony-forming cells, colony-forming cells, and maturing mast cells, respectively. No cells binding both YB5.B8 and an MoAb to the early lymphoid marker CD10 were found, implying that most early lymphoid cells do not express c-kit. MoAbs to the c-kit protein should prove valuable in multimarker studies of human hematopoietic stem and progenitor cells. Definition of a reference range of c-kit expression in normal human bone marrow will provide a sound basis for further studies of this marker in diagnosis and prognosis in AML.

The murine monoclonal antibody, 14A2.H1, identifies a novel platelet surface antigen.

A murine monoclonal antibody 14A2.H1, raised against acute myeloid leukaemia cells, identifies a previously undescribed 27 kDa platelet surface glycoprotein which is expressed at low copy number (10(3)/platelet). MAb 14A2.H1 caused aggregation of platelets which was dependent on Fc gamma RII. Binding of the antibody to platelets was not altered by activation by thrombin or phorbol ester. In haemopoietic cell populations the antibody bound to megakaryocytes, monocytes (weakly), several myeloid leukaemic cell lines and fresh myeloid leukaemic blasts from some patients. Lymphocytes, lymphoid cell lines, neutrophils and haemopoietic progenitor cells were negative. Expression of the antigen was not restricted to haemopoietic cells as epithelial cells in tonsillar crypts and endothelial cells were positive.