Synergistic interaction between differentiation inducers and DNA synthesis inhibitors: a new approach to differentiation induction in myelodysplasia and acute myeloid leukaemia.

Numerous agents induce differentiation and maturation of neoplastic and dysplastic myeloid cells in vitro and some of these agents have been used with limited success in the treatment of patients with myelodysplastic syndromes (MDS) and myeloid leukaemias. We recently proposed that physiological and pharmacological agents which enhance differentiation and maturation in vitro act by two fundamentally different routes: (1) by hastening the progression through various differentiation/maturation steps; (2) by slowing proliferation (usually by inhibition of DNA synthesis). In order to test this thesis we looked for synergistic effects on differentiation using pairs of agents from the two groups in cultures of cells from myelodysplastic and acute myeloid leukaemia (AML) patients and from normal marrow donors. The results with three MDS, two AML and three normal samples show that combinations of differentiation inducing agents (retinoic acid, N-methylformamide) with DNA synthesis inhibitors (6-mercaptopurine, cytosine arabinoside and aphidicolin) produce a differentiation inducing effect equivalent to that of 10-100, or even 1000 fold higher concentrations of single agents. Myelotoxic effects in vitro were not synergistic. The use of these synergistic combinations should greatly enhance the usefulness of differentiation inducers in the therapy of MDS and myeloid leukaemia.

Differentiation of human acute myeloid leukaemia cells in response to exogenous and endogenous stimuli: different patterns of response suggested by a bioassay system.

The object of this study was to devise quantitative bioassay systems suitable for the analysis of differentiation in acute myeloid leukaemia (AML) in response to both endogenous (bone marrow) and exogenous stimuli. Dose response analyses of the relationship between exogenous differentiation stimulus (supplied by peripheral blood mononuclear leucocytes) and clone cell maturity in double layer semi-solid agar cultures, were used to identify a measure of response with a linear relationship to differentiation stimulus on which bioassay systems could be based. All 7 AML samples tested at presentation and 1 sample from a patient during the regenerative phase after therapy, exhibited some response to exogenous differentiation stimulus. Only in the latter was there no significant difference in the response of test and reference cells. The other 7 showed marked variation in the pattern of the response in vitro. There was no close correlation between differentiation capacity in vitro and extent of differentiation observed in vivo. This discrepancy might be related to varying availability of endogenous differentiation stimulus in different patients. The technique for assaying differentiation response was adapted to demonstrate and quantify the endogenous differentiation stimulus (the stimulus provided by the leukaemic patient's marrow cells). The results suggest that in some patients the level of this stimulus may be a major determinant in the level of differentiation achieved in vivo.

DNA strand breakage and ADP-ribosyl transferase mediated DNA ligation during stimulation of human bone marrow cells by granulocyte-macrophage colony stimulating activity.

ADP-ribosyl transferase (ADP-RT) is a chromatin-bound nuclear enzyme catalysing the transfer of ADP-ribose from NAD+ to chromatin proteins. The enzyme is activated by DNA strand breaks and has been suggested to have roles in both DNA repair (via its effect on DNA ligase II) and in differentiation. We recently demonstrated that specific inhibitors of ADP-RT preferentially inhibit differentiation of human granulocyte-macrophage progenitor cells to the macrophage lineage and that the specific proliferation/differentiation stimulus granulocyte-macrophage colony stimulating activity (GM-CSA) activates ADP-RT in human marrow cells within 3 h of exposure. The purpose of this study was to investigate the role of ADP-RT in monocyte-macrophage differentiation. By altering the time of addition of ADP-RT inhibitor it was demonstrated that maximal inhibition of macrophage differentiation only occurs when the inhibitor is added within the first 24 h of culture. This suggests that it is an early event during the induced differentiation of granulocyte-macrophage progenitor cells which requires ADP-RT. Fluorometric assay of the level of DNA strand breaks showed that GM-CSA induces DNA strand breaks which are rapidly ligated only if ADP-RT is available. These data and those of our earlier studies suggest that DNA rearrangement may be involved in differentiation of granulocyte-macrophage progenitors to the monocyte-macrophage pathway. Such a DNA rearrangement could provide a molecular basis for commitment of multipotent progenitors to a single lineage.

Role of ADP-ribosyl transferase in differentiation of human granulocyte-macrophage progenitors to the macrophage lineage.

Nuclear adenosine diphosphate-ribosyl (ADP-ribosyl) transferase is a chromatin-bound enzyme catalyzing the transfer of ADP-ribose from NAD+ to chromatin proteins. The physiologic function of this covalent modification of chromatin has not been fully established, but roles in both DNA repair and in differentiation have been proposed. We demonstrate that three specific inhibitors of ADP-ribosyl transferase (5-methylnicotinamide, 3-methoxybenzamide, 3-aminobenzamide) inhibit differentiation of human granulocyte-macrophage progenitor cells to the macrophage lineage. Differentiation to the neutrophil-granulocyte lineage is much less affected. The inhibition of macrophage differentiation seems to relate to the ability of these compounds to inhibit ADP-ribosyl transferase. A structural analogue (3-methoxybenzoic acid), which is not inhibitory for the enzyme, did not inhibit macrophage differentiation. Additional evidence for a role of ADP-ribosyl transferase in the differentiation of granulocyte-macrophage progenitors was obtained from experiments in which enzyme activity levels were measured in permeabilized marrow cells. Marrow cell ADP-ribosyl transferase activity increased after 3-hr stimulation by the differentiation/proliferation stimulus--granulocyte-macrophage colony-stimulating activity (GM-CSA). Unstimulated marrow cells showed low or undetectable levels of enzyme activity.

CFUGEMM, CFUGM, CFUMK: analysis by equilibrium density centrifugation.

A modification of the technique of Fauser and Messner was used for the culture of human multipotent hematopoietic progenitors. This modified technique shows a linear relationship between mixed granulocytic-erythrocytic colonies and the number of cells plated, even at extremely low cell doses (10(4) cells/dish) and is therefore a more suitable assay system for cell separation studies than the original non-linear method. The buoyant density of CFUGEMM (colony forming unit granulocytic-erythrocytic-megakaryocytic-macrophage) was determined using equilibrium density centrifugation. CFUGEMM were of lower buoyant density than the majority of nucleated marrow cells. Substantially enriched populations of CFUGEMM could be obtained with a single density separation procedure. The density distribution profile for CFUGEMM was also distinct from the density distribution of granulocyte-macrophage colony forming cells (CFUGM), the latter being of somewhat greater buoyant density than the former. Cells which formed clones containing only megakaryocytes in culture (CFUMK) had an intermediate density between that of CFUGEMM and CFUGM. The morphological characteristics of these progenitor cells were studied using correlation analysis. Results suggested that the CFUGEMM correspond to transitional cells without granules, the CFUGM to transitional cells with 1-4 granules and the heterogeneous group of GM-cluster forming cells to a broad category including myeloblasts, promyelocytes, myelocytes and metamyelocytes.

Use of bone-marrow culture in prediction of acute leukaemic transformation in preleukaemia.

Bone-marrow granulocyte-macrophage progenitor cell proliferation and regulatory factor (colony-stimulating activity; CSA) production were assessed at presentation and, if possible, subsequently in twenty-one patients with dysmyelopoiesis and less than 5% bone-marrow blasts. Seven patients underwent acute leukaemic transformation 1-35 months after the first marrow culture. Assay of bone-marrow endogenous CSA proved the most useful prognostic test. The rate of transformation in the seven patients with raised CSA at presentation was significantly greater (five transformed and one died at or before 4 months) than that in the fourteen patients with normal, low, or undetectable CSA (two transformed at 27 and 35 months). In one of the latter an increase in bone-marrow CSA occurred 6 weeks before transformation (serial marrow samples were not available in the other case). No other marrow culture feature measured, including the presence or absence of granulocyte-macrophage colony-forming cells and total clone numbers (colonies and smaller clusters) was as useful.

Brief report: qualitative differences in the biological activity of two sources of colony stimulating factor: GCT conditioned medium and leucocyte feeder layers.

A most stringent test for the non-identity of biologically active materials is provided by the techniques used for parallel line type bioassay of these materials. We have studied commercially available GCT conditioned medium using a parallel line type biological assay for granulocyte-macrophage colony stimulating factors (GM-CSF) and mononuclear leucocytes as a reference source of GM-CSF. Only test and reference GM-CSF sources which have active constituents with identical dose response characteristics can produce parallel line displacement in this assay system. GCT conditioned medium failed to produce parallel displacements in five different assays. We must therefore conclude that one or more active factors from the test and reference sources of GM-CSF (GCT conditioned medium and mononuclear leucocytes) are not identical. This indicates a need for caution when substituting GCT conditioned medium for leucocyte feeder layers in human bone marrow cultures.

The use of image analysis for estimation of growth in bone marrow cultures.

An image analysis system was used to assess growth in agar bone marrow cultures. The technique has advantages over visual counting techniques; it saves time, increases objectivity, reduces operator errors, and may be used to estimate total cell growth or mean clone size. Either measurement may be used to determine activity levels of factors that affect growth such as colony stimulating factors. The advantages and limitations of the technique will be discussed.

Distinct T lymphocyte subsets affect granulo-monocytic differentiation and proliferation.

Bone marrow culture techniques using semi-solid support-media provide an opportunity to observe granulocyte-macrophage progenitor cells proliferating and differentiating in vitro. Cells from the bone marrow and peripheral blood produce specific growth factors which stimulate both the proliferation and differentiation processes (collectively termed colony-stimulating factors or activity, CSF/CSA). Even highly purified CSA induces proliferation and concomitant differentiation suggesting that the two processes are inseparable. We here present evidence to suggest that the two processes are, at least in part, individually regulated. We observed delayed differentiation by morphological, cytochemical and functional criteria in granulocyte-macrophage clones formed in cultures in which the growth stimulus was supplied by marrow or blood cells which have been depleted of T lymphocytes bearing the OKT3/MBG6 antigen(s) by complement lysis. The proliferation stimulus was unaffected. T cell depletion using another monoclonal antibody, OKT11a (which removed an overlapping but not identical population of T cells), did not produce the same effect but did increase the level of CSA produced by the remaining marrow cells. Selective replacement experiments using OKT4+ or OKT8+ cell preparations suggested that removal of an MBG6+, OKT3+, OKT8+, OKT4-lymphocyte subset is responsible for the effect on differentiation.

Relative and absolute deficiency of bone marrow endogenous colony stimulating activity in acute myeloid leukemia: relation to response to chemotherapy.

The purpose of this study was to determine possible mechanisms for the recently observed association between insensitivity of acute myeloid leukemia (AML) clonogenic cells to colony stimulating activity (CSA) and poor response to induction chemotherapy. The bone marrow endogenous CSA was determined using semi-solid agar cultures by measuring the response of the AML patients' own clonogenic cells to endogenous CSA. The results show that whereas 31% (5/16) of patients at presentation have deficient bone marrow endogenous CSA production, over 50% (11/21) have relative deficiency of endogenous CSA, due to insensitivity of the patients' clonogenic cells to CSA. Although there is an association between relative deficiency of endogenous CSA and a poor response to therapy, the relationship is not close enough to explain the previously observed highly significant correlation between insensitivity to CSA and poor response to therapy. The CSA-insensitive phenotype and poor response to therapy, one via the tendency to relative endogenous CSA deficiency in the CSA-insensitive group and another via some additional feature of these poor response AML phenotypes which is independent of the presence or absence of endogenous CSA deficiency.

Reactivity of monoclonal antibodies with human myeloid precursor cells.

Five monoclonal antibodies have been tested for their ability to bind to myeloid precursor cells in normal human bone marrow. Indirect immunofluorescence and the fluorescence activated cell sorter was used to separate cells according to their reactivity for trial culture in vitro in order to grow granulocyte-macrophage colony forming cells (CFUc). Two antibodies (OKT3 and OKT11) which react strongly with bone marrow T lymphocytes were found to be unreactive with CFUc. YD1/23 reacts very strongly with both T and B lymphocytes but is only weakly reactive with CFUc. In contrast, OKT10 and YE2/36 did react with CFUc. The consequences of these findings and the potential clinical use of these antibodies in bone marrow transplantation are discussed.

Bone marrow endogenous colony stimulating factor(s): relation to granulopoiesis in vivo.

Granulocyte-macrophage colony stimulating factors (CSF) are produced both in peripheral tissues and within the bone marrow. Stimulators from either source might regulate granulocyte and monocyte production in vivo. The purpose of this study was to devise an assay for bone marrow endogenous CSF so that its role in regulation of granulopoiesis might be assessed. There was a significant positive correlation between the endogenous CSF level and the subsequent trend in peripheral blood neutrophil count in both normal and infected patients, suggestive of a regulatory role. In addition, granulopoietic recovery after neutropenia was associated with increased endogenous CSF levels. There are two candidate regulators of endogenous CSF production, a stimulatory humoral factor and a neutrophil derived inhibitor which inhibits endogenous CSF production in vitro. Our results suggest that the marrow neutrophil level had a much more significant inhibitory effect than the neutrophil level in the culture.

Clone size potential and sensitivity to colony-stimulating activity: differentiation-linked properties of granulocyte-macrophage progenitor cells.

Previous studies using semi-solid agar cultures of acute myeloid leukaemia (AML) cells have shown that proliferative capacity (clone size potential) and the degree of sensitivity of clone-forming cells to the specific granulocyte-macrophage regulatory colony-stimulating activity (CSA) are closely linked AML cell properties. The purpose of this study was to determine whether this association is confined to AML cells or whether the linking of these two properties in AML represents retention of an association occurring in normal granulocyte-monocyte progenitor cell populations. Bone marrow cells from normal donors were studied using four independent techniques to enrich for clonogenic cells with different clone size potentials (equilibrium density centrifugation, adherence to microcarrier beads, osmotic lysis and fluorescence-activated cell sorting). It was shown that mean clone size was directly related to the mean CSA threshold (amount of CSA required to stimulate 50% of the cells to clone formation). Further studies (including analysis of the kinetics of production of clonogenic cells with different clone size potentials in suspension culture) suggested that these two properties were linked to differentiation, with proliferative capacity decreasing and sensitivity to CSA increasing as cells differentiate down the granulocyte-monocyte pathway.