Production of hematopoietic growth factors by human B lymphocytes: mechanisms and possible implications

In this study we have investigated the ability of human B lymphocytes to produce granulocyte-macrophage colony stimulating factor (GM-CSF) and, in preliminary experiments, granulocyte CSF (G-CSF). The sources of human B cells were surgically removed tonsils from normal individuals and peripheral blood from patients with B cell chronic lymphocytic leukemia (B-CLL). Tonsil B lymphocytes were purified by E rosetting and complement-mediated cytotoxicity with selected monoclonal antibodies and subsequently fractionated by a Percoll density gradient into in vivo activated and resting cells. The latter cell fractions were subsequently cultured with or without stimuli. GM-CSF was detected by a bioassay, G-CSF by an enzyme-linked immunoassay. In vivo and in vitro activated B cells produced GM-CSF, whereas in vivo activated, but not in vitro activated, B lymphocytes produced G-CSF. These results were confirmed by Northern blot experiments with cDNA probes specific for GM-CSF and G-CSF genes. Many B cell suspensions from B-CLL patients produced GM-CSF or G-CSF only following Staphylococcus Aureus Cowan I (SAC) stimulation; in some cases, a spontaneous production or no production at all of the two cytokines was detected. The possible implications of these results for B cell physiology and for the pathogenesis of immunologically mediated diseases will be discussed.

Pentasomy 21 in leukemia complicating Diamond-Blackfan anemia

We present the cytogenetic pattern of a leukemic infant with Diamond-Blackfan anemia (DBA). The karyotype was characterized by clonal evolution involving consecutive gains of chromosome 21 up to pentasomy. No chromosomal changes were present in normal lymphocytes. Such a karyotype evolution has been described in some cases of acute leukemia associated with Down Syndrome, but rarely in non-Down cases.

Production of granulocyte-macrophage colony-stimulating factor but not IL-3 by normal and neoplastic human B lymphocytes

The ability of human B cells to produce granulocyte-macrophage (GM)-CSF and IL-3 was investigated. B cells, isolated from tonsils or from the peripheral blood of patients with B cell chronic lymphocytic leukemia using mAb and immune rosettes, were cultured with or without Staphylococcus aureus Cowan strain I. GM-CSF and IL-3 were measured in the culture supernatants using a bioassay based on the selective proliferative response of the MO7e megakaryoblastic cell line to IL-3 or GM-CSF. S. aureus Cowan I-stimulated normal B cells released measurable amounts of GM-CSF but not of IL-3 as determined in neutralization assays with specific mAb in the MO7e cell line test. Some of the unstimulated normal B suspensions also produced GM-CSF, albeit in lower quantities. When normal B cells were fractionated into small (resting) and large (activated) B cells by Percoll density gradients, spontaneous GM-CSF production was detected only in the large cell fractions, but small cells were induced to produce GM-CSF upon S. aureus Cowan I stimulation. On a per cell basis, tonsillar B cells were found capable of releasing more GM-CSF than activated peripheral blood monocytes. The amount of GM-CSF produced by B cells was always inferior to that released by stimulated peripheral blood T cells or NK cells. The purified B cell suspensions from all 14 B cell chronic lymphocytic leukemia patients studied released GM-CSF but not IL-3 in the culture supernatants, sometimes even in the absence of stimuli. Northern blot analysis on total or poly(A)+ RNA confirmed the presence of GM-CSF, but not of IL-3, mRNA transcripts in both normal and malignant B cells. The results of these studies support the notion that activated human B lymphocytes release sufficient GM-CSF to play a role in the control of both hematopoiesis and the inflammatory process.

Production of granulocyte-macrophage colony-stimulating factor but not IL-3 by normal and neoplastic human B lymphocytes

The ability of human B cells to produce granulocyte-macrophage (GM)-CSF and IL-3 was investigated. B cells, isolated from tonsils or from the peripheral blood of patients with B cell chronic lymphocytic leukemia using mAb and immune rosettes, were cultured with or without Staphylococcus aureus Cowan strain I. GM-CSF and IL-3 were measured in the culture supernatants using a bioassay based on the selective proliferative response of the MO7e megakaryoblastic cell line to IL-3 or GM-CSF. S. aureus Cowan I-stimulated normal B cells released measurable amounts of GM-CSF but not of IL-3 as determined in neutralization assays with specific mAb in the MO7e cell line test. Some of the unstimulated normal B suspensions also produced GM-CSF, albeit in lower quantities. When normal B cells were fractionated into small (resting) and large (activated) B cells by Percoll density gradients, spontaneous GM-CSF production was detected only in the large cell fractions, but small cells were induced to produce GM-CSF upon S. aureus Cowan I stimulation. On a per cell basis, tonsillar B cells were found capable of releasing more GM-CSF than activated peripheral blood monocytes. The amount of GM-CSF produced by B cells was always inferior to that released by stimulated peripheral blood T cells or NK cells. The purified B cell suspensions from all 14 B cell chronic lymphocytic leukemia patients studied released GM-CSF but not IL-3 in the culture supernatants, sometimes even in the absence of stimuli. Northern blot analysis on total or poly(A)+ RNA confirmed the presence of GM-CSF, but not of IL-3, mRNA transcripts in both normal and malignant B cells. The results of these studies support the notion that activated human B lymphocytes release sufficient GM-CSF to play a role in the control of both hematopoiesis and the inflammatory process.

Chronic childhood neutropenia: studies on the in vitro clonogenicity of bone marrow myeloid progenitors

Bone marrow mononuclear cells (MNC) from 6 pediatric patients with chronic neutropenia were tested for myeloid colony formation upon stimulation with the supernatant of the 5637 cell line or with recombinant granulocyte-macrophage colony-stimulating factor or interleukin 3. Heterogeneous patterns of response of myeloid progenitors were observed in the individual patients, with no colony growth in 2 cases and abnormalities of colony formation or composition in two additional cases. Morphologic and surface marker analyses showed that the bone marrow of some patients contained an excess of lymphocytes with an altered subset distribution. In order to investigate whether or not there was a relationship between the latter abnormality and the observed clonogenic defects, marrow MNC were tested for myeloid colony formation before and after lymphocyte depletion. No evidence for a cell-mediated suppression of colony growth was obtained; likewise, patient sera failed to inhibit colony formation by normal bone marrow myeloid progenitors. Taken together, these data make it unlikely that, in our cases, immunologic mechanisms were responsible for the pathogenesis of chronic neutropenia.

Accelerated chemotherapy with or without GM-CSF for small cell lung cancer: a non-randomised pilot study

Two series of five consecutive patients with small cell lung cancer were treated with an "accelerated" chemotherapy regimen of cyclophosphamide-doxorubicin-vincristine (CAV) and cisplatin-etoposide (PE) alternated possibly every week. In the first group of patients (median age 49 years, range 46-52) recombinant granulocyte-macrophage colony-stimulating factor (GM-CSF) was given as soon as grade IV leukopenia occurred, while in the second group (median age 59 years, 55-68) no growth factor was administered. The mean interval between chemotherapy courses and the mean duration of chemotherapy were 10 and 57 days, respectively, in the patients supported with GM-CSF compared with 13 and 72 days in the control group. One GM-CSF treated patient was withdrawn after the third cycle because of severe toxicity. The mean white blood cell and platelet nadirs were 600 and 46,000/microliters in the first group vs. 840 and 105,000/microliters in the controls. Overall chemotherapy dose-intensity was increased by two fold in the patients given GM-CSF compared with a 1.5 fold increase in the control patients. In all cases, irrespective of their treatment, there was an impaired colony forming capacity of circulating and marrow haemopoietic progenitor cells when grade IV leukopenia occurred, with recovery after the end of leukopenia. This pilot study suggests that accelerated CAV/PE chemotherapy is feasible both with and without GM-CSF. Different GM-CSF schedules as well as combinations of different haemopoietic growth factors may further improve dose-intensity.

Promotion and inhibition of haemopoiesis by NK cells: a model for immune-mediated haemopoietic suppression

Human NK cells, besides mediating cytotoxic functions, can exert a number of regulatory activities on the proliferation and maturation of hematopoietic stem cells through the release of soluble factors. A preliminary characterization of these factors demonstrates that NK cell-derived colony stimulating activities (CSA) are represented by GM-CSF and IL3, whereas the colony inhibiting activity (CIA) produced by NK cells is identifiable as TNF. CSA is mainly released by resting NK cells, while, following appropriate activation, the latter cells produce large amounts of CIA. NK cells also influence B cell activation by releasing a 50 kd BCGF and a 26 kd BCDF. The latter factor, which is active in vitro only in the presence of IL2, is identifiable as IL6. The production of regulatory cytokines by NK cells may have some bearing on the pathogenesis of a number of autoimmune disorders.

Production of colony-stimulating activity by human natural killer cells: analysis of the conditions that influence the release and detection of colony-stimulating activity

Highly purified natural killer (NK) cell suspensions were tested for their capacity to release colony-stimulating activity (CSA) in vitro. NK cell suspensions comprised primarily CD16+ cells and were devoid of CD3+ T cells, CD15+ monocytes, and of B cells. CSA was detected in the NK cell supernatants and sustained the growth of myeloid colonies from both normal peripheral blood and bone marrow. CSA could be in part inhibited by pretreating NK cell culture supernatants with a specific goat anti-granulocyte-macrophage colony-stimulating factor (GM-CSF) antiserum. The inhibition, however, was never complete, a finding that suggests that additional factors were responsible for CSA. Incubation of NK cells with K562 cells (an NK-sensitive target) or with normal bone marrow cells resulted in the appearance of a strong colony-inhibiting activity (CIA) in the culture supernatants. Such CIA was demonstrable in an experimental system where bone marrow or peripheral blood progenitors were induced to form myeloid colonies in the presence of conditioned medium by CSA-producing giant cell tumor (GCT) cells. Stimulation of NK cells with NK-insensitive targets failed to induce CIA production. Neutralizing antitumor necrosis factor (TNF) monoclonal antibodies (MoAbs) were found capable of inhibiting CIA present in the supernatants of NK cells stimulated with K562 cells. Following treatment with anti-TNF antibodies, CSA was again detectable in the same supernatants. This finding indicates that induction of TNF production did not concomitantly switch off CSA production by NK cells. Pretreatment of NK cells with recombinant interleukin-2 (rIL-2) or gamma interferon (r gamma IFN) did not change the amount of CSA released. However, treatment with rIL-2 caused the appearance of a factor in the NK cell supernatants capable of sustaining the formation of colonies of a larger size.

Anti-lymphocyte globulin stimulates normal human T cells to proliferate and to release lymphokines in vitro. A study at the clonal level

Human peripheral blood mononuclear cells (PBMC) were stimulated in vitro with anti-lymphocyte globulin (ALG), and the phenotypic and functional properties of the blasts obtained were investigated. When stained with monoclonal antibodies (MoAbs), all of the blasts were identified as T cells that expressed predominantly the CD4 phenotype (70% of the cells). The remaining blasts were CD8+. These findings demonstrate that ALG stimulates both helper-inducer and cytotoxic-suppressor cells at random since the CD4 to CD8 ratio in the stimulated blasts was the same as in resting PBMC. This ratio is different from that observed in short-term cultures of T cells stimulated with phytohemagglutinin (PHA) under the same conditions (CD4 to CD8 ratio less than 1). ALG-stimulated T cells were cloned by limiting dilution in the presence of recombinant Interleukin-2 (rIL-2). The clones obtained were expanded and maintained in long term cultures with rIL-2. Thirty-two clones were tested for their capacity of producing colony stimulating activity (CSA) or burst promoting activity (BPA). Twenty-eight of them produced CSA and 12 produced BPA. No correlation was found between the surface phenotype and the ability of the clones to produce CSA or BPA (ie, both the CD4+ and CD8+ clones released the cytokines). When 16 of the same clones were tested for II-2 and gamma interferon (gamma IFN) production, 12 were found to be gamma INF and IL-2 producers. All of the gamma IFN producers also released IL-2, whereas in the single clones no correlation was found with the capacity of releasing BPA and CSA. Supernatants from selected T-cell clones were also tested for hematopoietic growth factor activities in the presence of neutralizing antisera to human granulocyte-macrophage colony stimulating factor (GM-CSF) or to Interleukin-3 (IL-3). It was found that most CSA was attributable to GM-CSF, whereas BPA was mainly related to the presence of IL-3.