Cytotoxicity of an 125I-labeled DNA-binding compound that induces double-stranded DNA breaks.

[125I]Iodorivanol (6,9-diamino-2-ethoxy-5-[125]iodoacridine) has been prepared by direct iodination of rivanol (6.9-diamino-2-ethoxyacridine). In vitro binding of [125I]iodorivanol to PM2 DNA resulted in induction of double-stranded DNA breaks following decay of the 125I atom, presumably in the same way as decay of 125I atoms in 125I-labeled DNA causes double-stranded DNA breaks. Treatment of mouse L-cell cultures with [125I]iodorivanol resulted in a cell kill, the extent of which was dependent on the 125I specific activity and the duration of exposure. A clonogenic assay was used to quantitate cell kill. It was concluded that at least some of the [125I]iodorivanol in the culture medium was taken up by the cells, transported to the nucleus, and bound to DNA and that subsequent decay of the 125I atoms induced double-stranded DNA breaks in the genome, with consequent loss of viability. 125I-labeled DNA-binding compounds are suggested as a novel class of cytotoxic agents.

Inhibition of DNA synthesis and cell death.

The association between DNA synthesis inhibition and cell death in mouse L-cells was investigated using the drug hydroxyurea. This drug produces a preferential labelling of low molecular weight DNA and dose-response studies revealed a correlation between this effect and cytoxicity. Investigation of the reassociation kinetics of DNA labelled during hydroxyurea inhibition showed an over-replication of middle repetitive sequences, but the concentration dependence of this effect was quite different to that of cytotoxicity.

In vivo effects of interleukin-1 alpha on regenerating mouse bone marrow myeloid colony-forming cells after treatment with 5-fluorouracil.

Injection of a single dose of recombinant human interleukin-1 alpha (r-hu-IL-1 alpha) into mice 24 hr after 5-fluorouracil (FU) treatment resulted in an increased rate of recovery of three types of colony-forming cells (CFCs) in the bone marrow. Myeloid progenitors with high proliferative potential (responsive to CSF-1 + IL-3 + IL-1 alpha), low proliferative potential (responsive to CSF-1), megakaryocyte progenitors, and total nucleated cells per femur increased up to 5-fold, 7-fold, 3-fold, and 3-fold, respectively, in a dose related fashion compared with the control FU treated marrows. The kinetics of FU kill and recovery of these CFCs are shown.

Haemopoietic recovery in spleen and marrow after transplantation of bone marrow from either normal or hydroxyurea treated mice.

Haemopoietic regeneration was studied following x-irradiation and transplantation of bone marrow from either normal or hydroxyurea-treated donor mice, to ascertain the contribution of proliferating progenitor cells to regeneration. With transplantation of equivalent numbers of CFU-S, total DNA and 3HTdR uptake into DNA in spleen and femoral bone marrow and the erythroid, granulocytic and mononuclear cell populations were not significantly different between normal (NBM) and hydroxyurea-treated (HUBM) marrow. The response of hypertransfused x-irradiated mice to erythropoietin (EPO) administration was also not significantly different in spleens of mice receiving normal or hydroxyurea-treated marrow.

In vivo kinetic status of hematopoietic stem and progenitor cells as inferred from labeling with bromodeoxyuridine.

The extent of incorporation of bromodeoxyuridine (BrdUrd) into DNA of different types of hematopoietic stem cells assayed in vivo and of progenitor cells assayed in vitro was determined after continuous infusion of BrdUrd into mice for either four or seven days. Cells surviving subsequent exposure to 320 nm ultraviolet light (UV320) were considered not to have incorporated BrdUrd. Assays of stem cells were carried out in 8-Gy-irradiated Balb/c-specific pathogen-free (SPF) mice by measuring the ability of injected marrow: (a) to form spleen colonies at either seven, ten, or 13 days (the units giving rise to such colonies were named CFU-S-7, CFU-S-10, and CFU-S-13), (b) to increase the marrow content at 13 days of colony-forming cells (CFC) responsive in vitro to pregnant-mouse-uterus extract (P) (this ability was named marrow-repopulating ability P-CFC [MRA-P-CFC]) and, (c) to increase blood platelet counts at day 13 (this was named platelet-repopulating ability [PRA]). The in vitro assays carried out on marrow from BrdUrd-infused mice were measurements of the content of CFC responsive to P and to P plus human spleen-conditioned medium (H). The percentage survival after exposure to UV320 in marrows obtained after four and seven days of infusion of BrdUrd, respectively, was: MRA-P-CFC, 100% and 100%; PRA, 80% and 50%; CFU-S-13, 65% and 25%; CFU-S-10, 11% and 3%; CFU-S-7, 8% and 2%; P + H CFC, 20% and 12%; and P-CFC, 6% and 6%. These results are in agreement with predictions from previous experiments that studied the effects of 5-fluorouracil (5-FU) on bone marrow using the same assays and are compatible with a model that considers marrow to be organized as a concatenated series of compartments in which turnover rate increases as maturity increases.

Multiparameter analysis of transplantable hemopoietic stem cells. II. Stem cells of long-term bone marrow-reconstituted recipients.

Marrow obtained from mice (referred to as [X + BM] mice) 3 months after gamma-irradiation (9 Gy) and bone marrow inoculation (0.1 femur equivalents) showed a reduced capacity to reconstitute hemopoiesis of irradiated mice and an increased sensitivity to 5-fluorouracil. Sorting of marrow from (X + BM) mice on the basis of low angle and 90 degrees scatter, and low rhodamine 123 fluorescence, showed that the set of cells that in normal mice is enriched for cells efficient at hemopoietic reconstitution manifested the greatest reduction in hemopoietic reconstituting ability. In spite of this reduction this fraction contained as many 13-day spleen colony-forming units (CFU-S13) and high proliferative potential colony-forming cells (HPP-CFC) as the equivalent fraction from normal littermate mice. This could be explained by postulating that neither CFU-S13 nor HPP-CFC are responsible for hemopoietic reconstitution, but that this is dependent on an earlier, pre-CFU-S13 cell. Alternatively only a subset of either CFU-S13 or HPP-CFC is responsible for long-term hemopoietic reconstitution after lethal irradiation. It would appear that at present there is no adequate method of predicting the hemopoietic reconstituting ability of a given marrow, other than to test it by injection into lethally irradiated hosts.

Pretreatment with cytosine arabinoside does not protect against the delayed effects of irradiation on thrombopoiesis and erythropoiesis.

We have examined the capacity of the thrombopoietic and erythropoietic systems to respond to challenge with the cytotoxic drug 5-fluorouracil (5-FU) following sublethal doses of irradiation. Normal adult mice respond to 5-FU with a mild thrombocytopenia followed by a marked rebound thrombocytosis. One month after 2 Gy whole-body radiation, platelet counts took longer than normal to reach a nadir after 5-FU, and the rebound thrombocytosis was reduced. A normal response was seen when the interval after radiation was extended to 4 months. Increasing the radiation dose to 6.5 Gy resulted in a much smaller rebound thrombocytosis when 5-FU was given 1 month later. Extending the interval before the drug was given resulted in a normal response being regained between 4 and 7 months. Erythrocyte levels were temporarily depressed after 5-FU in all mice, and it took longer for recovery to occur if they had been irradiated. In another series of experiments, we investigated the effect of priming the mice before irradiation to see if this resulted in radioprotection. An injection of cytosine arabinoside (Ara-C) 2 days before a dose of 6.5 Gy resulted in the expected earlier recovery in platelet counts. To see if cells of the megakaryocyte lineage were protected from the delayed effects of irradiation by this treatment, mice were given 5-FU 1 month after Ara-C plus irradiation. The period of thrombocytopenia was followed by only a small rebound thrombocytosis, and platelet counts were indistinguishable from those found for mice not primed with Ara-C before irradiation. These experiments revealed a delayed effect of irradiation on the thrombopoietic and erythropoietic systems, which was long-lasting but not permanent at the doses used. The effects were not eliminated by priming mice with Ara-C before irradiation.

The concentration and resolution of primitive hemopoietic cells from normal mouse bone marrow by negative selection using monoclonal antibodies and Dynabead monodisperse magnetic microspheres.

High proliferative potential colony-forming cells (HPP-CFC) detected in clonal agar culture in the presence of the combined stimulus of colony-stimulating factor 1 (CSF-1) + interleukin 3 (IL-3) + interleukin 1 alpha (IL-1 alpha) are closely related to developmentally early progenitor cells capable of reconstituting the hemopoietic system of lethally irradiated mice following transplantation. Flow cytometric analysis and sorting of normal, unperturbed bone marrow has shown that HPP-CFC are B220- and 7/4-, whereas the committed progenitors of the macrophage lineage responsive to CSF-1 alone (CSFCSF-1) are B220- and 7/4+. Negative immunomagnetic selection using an anti-7/4, anti-B220 antibody cocktail and second-antibody-coupled Dynabead microspheres to replace flow cytometry results in the highly reproducible and specific enrichment of HPP-CFC, and simultaneous resolution of HPP-CFC from CFCCSF-1. The tenfold enrichment of HPP-CFC compared with unfractionated bone marrow cell suspensions was comparable to that obtained by fluorescence-activated cell sorting. Enrichment was achieved with negligible loss of HPP-CFC at the immunomagnetic bead selection step, and 65% of HPP-CFC were recovered. The method is rapid, highly reproducible, and efficient, and has wide application to the separation of rare hemopoietic cells from normal bone marrow.

Multiparameter analysis of transplantable hemopoietic stem cells: I. The separation and enrichment of stem cells homing to marrow and spleen on the basis of rhodamine-123 fluorescence.

A multiparameter cell separative procedure is described that enables normal transplantable hemopoietic stem cells that preferentially home to the marrow of lethally irradiated mice to be enriched and separated from the majority of spleen colony-forming cells that are assayed 13 days after transplantation (CFU-S13). First, bone marrow cells are centrifuged in a discontinuous bovine serum albumin gradient. Low-density cells (1.060-1.068 g/cm3) are harvested and labeled with the supravital cationic fluorochrome rhodamine 123 (Rh123). Labeled cells are analyzed using a fluorescence-activated cell sorter, and cells are sorted on the basis of relative Rh123 fluorescence within a predetermined forward versus 90 degrees red light scatter window that has been optimized for the recovery and enrichment of cells with marrow repopulating ability (MRA). Cells with MRA were characterized by relatively low Rh123 fluorescence and could be separated from a fraction that fluoresced more intensely and contained the majority of CFU-S13 but low MRA. The ratios of low-fluorescent:high-fluorescent MRA and CFU-S13 were 20:1 and 0.5:1, respectively, and the resultant ratio of MRA:CFU-S13 was 556:1 in the low-fluorescent fraction, and 48:1 in the high-fluorescent fraction. Cells with platelet repopulating ability cofractionate with MRA whereas cells with erythroid repopulating ability remain associated with CFU-S13. High-proliferative-potential macrophage colony-forming cells that form colonies in agar in the presence of the combined stimulus of pregnant mouse uterus extract plus human-spleen-conditioned medium were present in both fractions. Between 20% and 30% of transplantable hemopoietic stem cells has been recovered in approximately 1% of total nucleated cells from the starting preparation, and enrichments of 25- to 30-fold have been achieved permitting hemopoietic reconstitution of lethally irradiated host mice with as few as 400 donor cells.

Characterization and enrichment of macrophage progenitor cells from normal and 5-fluorouracil treated mouse bone marrow by unit gravity sedimentation.

Unit gravity sedimentation has been used to characterize and enrich cells from normal, post-5-fluorouracil (FU) and post 5-fluorouracil plus endotoxin (FUEt) regenerating mouse bone marrow with respect to two classes of macrophage progenitor cells. The two classes of progenitor cells assayed were (1) those responsive to the combined stimulus of pregnant mouse uterus extract (PMUE) plus human spleen conditioned medium (HUSPCM), and (2) those responsive to PMUE alone. In contrast to the bimodal nucleated cell distribution of normal marrow, 7 day post-FU marrow exhibited a unimodal nucleated cell profile. In marrow from 7 day post-FUEt treated mice, in ;which marrow regeneration was accelerated, there was a reemergence of a second nucleated cell peak. In addition, in 7 day post-FU and 7 day post-FUEt marrow there was a shift in the modal sedimentation velocities of both PMUE responsive and PMUE + HUSPCM responsive populations to higher values. The combined effect of these changes resulted in a marked increase in plating efficiencies of the peak enrichment fractions, reaching 12.5% in 7 day post-FUEt marrow. However, the highest yield of progenitor cells responsive to the combined stimulus of PMUE + HUSPCM compared to those responsive to PMUE alone were obtained in 7 day post-FU marrow.

The resolution, enrichment, and organization of normal bone marrow high proliferative potential colony-forming cell subsets on the basis of rhodamine-123 fluorescence.

Cell sorting on the basis of rhodamine-123 (Rh123) fluorescence has been used in conjunction with negative immunomagnetic selection to analyze the high proliferative potential colony-forming cell (HPP-CFC) compartment of normal murine bone marrow and to resolve and enrich HPP-CFC subpopulations responsive to different combinations of the hemopoietic growth factors interleukin 1 alpha (IL-1 alpha), interleukin-3 (IL-3), and colony-stimulating factor 1 (CSF-1). HPP-CFC with a specific requirement for IL-1 alpha plus IL-3 plus CSF-1 in order to proliferate were resolved and enriched on the basis of their low Rh123 retention (Rh-dull), whereas HPP-CFC that grew in the presence of IL-3 plus CSF-1, IL-3 alone, or CSF-1 alone were Rh-bright. Further addition of IL-1 alpha to IL-3 plus CSF-1 stimulated few additional HPP-CFC in the Rh-bright fraction. Our data confirm the value of Rh123 as a probe for the dissection and analysis of the primitive hemopoietic stem cell (PHSC) compartment. These data also show that the Rh123 staining characteristics of IL-1 alpha plus IL-3 plus CSF-1-responsive HPP-CFC are consistent with the hypothesis that these HPP-CFC are closely related to PHSC with long-term reconstituting capacity in vivo and that they are among the most primitive progenitors yet detected in clonal agar culture.

The organization of hemopoietic tissue as inferred from the effects of 5-fluorouracil.

Mouse bone marrow obtained one day after injection of 5-fluorouracil (FU) had a markedly diminished content of spleen colony forming units (CFUs) but retained its capacity to repopulate the marrow granulocyte-macrophage colony forming cell (GM-CFC) and CFUs compartments of 850 R irradiated hosts and had only a slightly reduced platelet repopulating ability (PRA). A significant correlation (r = 0.94, P less than 0.001) was observed between the content of high proliferative potential granulocyte-macrophage progenitor (HPP-GM-CFC) and the platelet and marrow GM-CFC repopulating abilities of bone marrow cell suspensions. Spleens of irradiated mice, injected with marrow from donors treated with FU between 1 and 7 days before showed an increase in colony numbers with time of sampling between 8-13 days after transplantation. In contrast, the colony counts observed in mice injected with normal bone marrow remained constant over that time interval. The colonies derived from bone marrow of FU treated mice grew faster than those from bone marrow of normal mice. Spleens obtained from irradiated mice, 10 days after injection of bone marrow derived from donors treated with FU 1 or 3 days before, showed only a few macroscopic surface colonies but when sectioned were found to contain large numbers of microscopic colonies, 80% of which were megakaryocytic. The results are interpreted on the basis of a clonal succession model of hemopoiesis with stem cells of varying proliferative potential and proliferation rates increasing as capacity for cell production decreases.

Enhanced megakaryocyte repopulating ability of stem cells surviving 5-fluorouracil treatment.

The effect of 5-fluorouracil treatment of donor mice on the capacity of transplanted bone marrow to produce megakaryocytes in the spleens of lethally irradiated recipients has been examined. At both 10 and 13 days after transplantation, the spleens of recipients of 5-fluorouracil treated bone marrow had significantly more megakarocytes per unit area of spleen section than recipients injected with an equivalent number of spleen colony forming units from normal bone marrow. It is suggested that such treatment may provide a sensitive in vivo system for the investigation of endogenous factors influencing megakaryocyte progenitor proliferation. The results are consistent with the concept of stem cells being heterogeneous with respect to self-renewal capacity.

Subpopulations of mouse bone marrow high-proliferative-potential colony-forming cells.

Bone marrow cells taken from mice treated eight days previously with 5-fluorouracil, formed colonies consisting of 10-100 cells after four days of incubation in methylcellulose cultures containing only 500 cells/dish, in the presence of partially purified synergistic factor from human placental-conditioned medium (SFHPlac) and macrophage colony-stimulating factor (CSF-1). Replating of these colonies revealed a high incidence (27%) of another class of high-proliferative-potential colony-forming cells (HPP-CFC) responsive only to the synergistic factor in WEHI-3B-conditioned medium (SFW, which appears to be identical to interleukin 3) plus CSF-1. These colonies contained no HPP-CFC responsive to SFHPlac plus CSF-1, although primary cultures incubated for 14 days in the presence of SFHPlac plus CSF-1 formed large colonies (diameter greater than 0.5 mm), indicating the presence of HPP-CFC responsive to SFHPlac plus CSF-1 in the starting marrow. Primary cultures containing SFW alone, or purified interleukin 3 alone, also gave rise to colonies consisting of 10-100 cells after four days of incubation in methylcellulose cultures; however, the cells from these colonies were unable to form large colonies on replating in the presence of either CSF-1 plus SFHPlac or CSF-1 plus SFW. These results suggest that two distinct populations of HPP-CFC exist and that the population of HPP-CFC stimulated by CSF-1 plus SFHPlac differentiates to form HPP-CFC that respond to CSF-1 plus SFW.

Detection of murine hemopoietin-1 in media conditioned by EMT6 cells.

We have previously reported replating experiments which demonstrated the existence of subpopulations of murine high-proliferative-potential colony-forming cells (HPP-CFC). One population of HPP-CFC, termed HPP-CFC-1, is stimulated by the combination of macrophage colony-stimulating factor (CSF-1) plus hemopoietin-1 (H-1), and actively generate a second population of HPP-CFC, termed HPP-CFC-2, which is responsive to CSF-1 plus interleukin-3 (IL-3). These reclonal experiments represent an assay system that discriminates between the two types of synergistic factors, namely H-1 and IL-3. To date H-1 has only been detected in medium conditioned by human cells. In this paper we have utilized these recloning experiments to study the synergistic factor(s) present in media conditioned by the murine mammary carcinoma cell line EMT6. Colony formation in secondary cultures containing cells picked up from primary cultures incubated in CSF-1 plus EMT6-conditioned medium was identical to that seen in secondary cultures containing cells picked up from primary cultures incubated in CSF-1 plus a source of H-1. Both sets of cultures demonstrated the generation of HPP-CFC-2 in the primary cultures, indicating the presence of a molecule in EMT6-conditioned medium that is the murine equivalent of H-1.

Identification of the "factor" in erythrocyte lysates which enhances colony growth in agar cultures.

The activity in human erythrocyte lysates which enhances colony growth of mouse bone marrow (BM) and other cell types in agar culture, could not be separated from hemoglobin (Hb). This conclusion was reached after various procedures, including purification of Hb in human hemolysates by crystallisation, separation of Hb into its major (A0) and minor (A1 and A2) components by DEAE-Sephadex chromatography and separation of a hemolysate into a Hb fraction and a non-Hb protein fraction by DEAE-cellulose chromatography; all resulted in the enchancement activity remaining with the Hb fraction. Separation of globins from rat or human lysates by an acid acetone precipitation, resulted in an acetone powder (AP) which retained the enhancement activity towards both mouse BM and tumour cell lines. The AP was separated into alpha and beta globins by chromatography on Sephadex G100 in 20% formic acid followed by CM-cellulose chromatography in a 8 M urea system. Since the enhancement activity is associated with both the alpha and beta globin peaks even under these dissociating conditions, it has been concluded that the enhancement factor in erythrocyte lysates is Hb itself. The enhancement activity of an AP is abolished by treatment with N-ethylmaleimide, suggesting that sulfhydryl groups in Hb are required for the activity.

Sources of murine macrophage colony-stimulating factor that also contain growth factors for primitive macrophage progenitor cells.

A new growth factor (synergistic factor, SF) has recently been described, which, in combination with a macrophage CSF source, is able to stimulate the proliferation of primitive high-proliferative-potential macrophage-progenitor cells (HPP-CFC) in mouse bone marrow in culture. It has been found that the addition of supraoptimal amounts of the crude CSF sources, extracts of pregnant mouse uterus and embryo, media conditioned by L cells or the mouse mammary tumor cell line (EMT6), stimulated the proliferation of HPP-CFC in the absence of any added SF. These preparations thus appear to contain a factor with similar biological properties to those of SF. This conclusion is supported by the results obtained from Sephacryl S200 chromatography of these three CSF sources, which indicate that in all three cases fractions with apparent molecular weight (MW) greater than 68,000 contained the major portion of the CSF-1 activity, whereas fractions with MW less than 68,000 contained the SF-like activity, together with minor amounts of CSF-1 activity.

Progenitor cells in murine bone marrow stimulated by growth factors produced by the AF1-19T rat cell line.

The AF1-19T rat cell line has been found to produce an activity that acts synergistically with colony-stimulating factor 1 (CSF-1) to stimulate primitive high proliferative potential colony-forming cells (HPP-CFC) in mouse bone marrow (BM) that appear to be the same as those stimulated by the combination of 5637-cell-conditioned medium (CM) plus CSF-1 or recombinant human (rh) interleukin 1 (IL-1) plus recombinant murine (rm) interleukin 3 (IL-3) plus CSF-1. AF1-19T also produced granulocyte-macrophage colony-stimulating factor (GM-CSF), which could be separated from this synergistic activity by gel filtration followed by hydroxylapatite chromatography. Results obtained from the mouse thymocyte costimulation assay for IL-1, the hybridoma growth factor assay for interleukin 6 (IL-6), the ability to stimulate HPP-CFC, and the ability to block this stimulation with an antibody to murine IL-1 alpha suggest that the synergistic activity in AF1-19T-CM is probably a mixture of IL-1 activity and IL-6 or an IL-6-like activity. Other workers have described a progenitor cell population in mouse BM (CFU-A) that forms large colonies in response to AF1-19T-CM plus CSF-1 or GM-CSF plus CSF-1. Experiments involving the kinetics of recovery after 5-fluorouracil treatment and generation of progenitors suggest that the GM-CSF-plus-CSF-1-responsive progenitors, and hence CFU-A, are a more mature cell type than the more primitive HPP-CFC, responsive to 5637-cell-CM plus CSF-1 or rhIL-1 plus rmIL-3 plus CSF-1.

Generation of murine hematopoietic precursor cells from macrophage high-proliferative-potential colony-forming cells.

High-proliferative-potential colony-forming cells (HPP-CFC) have been described as primitive murine macrophage progenitors. We have previously demonstrated the existence of two populations of HPP-CFC: one population, termed HPP-CFC-1, is stimulated by the combination of macrophage colony-stimulating factor (CSF-1) plus haemopoietin-1 (H-1) and actively generates a second population of HPP-CFC, termed HPP-CFC-2. HPP-CFC-2 are stimulated by CSF-1 plus interleukin-3 and generate macrophage CFC that differentiate to form mature macrophages. In this study, we have demonstrated that HPP-CFC-1, when stimulated by CSF-1 plus H-1, generate colony-forming cells (CFC) for the megakaryocyte and granulocyte lineages in addition to HPP-CFC-2 and M-CFC. No CFC were detected with erythroid potential. In addition, HPP-CFC-1 generated cells that formed day-13 spleen colonies, cells that repopulated the bone marrow, cells with platelet-repopulating ability, and cells with erythroid-repopulating ability in lethally irradiated mice. These data support previous data that the HPP-CFC-1 represent a primitive hemopoietic cell population and demonstrate the multipotentiality but not totipotentiality of these cells.

Interleukin 1 plus interleukin 3 plus colony-stimulating factor 1 are essential for clonal proliferation of primitive myeloid bone marrow cells.

The clonal growth in nutrient agar at low cell densities of high-proliferative potential colony-forming cells (HPP-CFC) of bone marrow obtained from mice treated 2 days earlier with 5-fluorouracil (FU) (FU2dBM) has been shown to require a combination of three growth factors, interleukin 1 (IL-1), interleukin 3 (IL-3), and macrophage colony-stimulating factor (CSF-1). These HPP-CFC have been enriched 140-fold from FU2dBM by fluorescence-activated cell sorting of 7/4-, B220-, and L3T4-negative cells. The mean of the plating efficiencies of these enriched populations was 4.4% and no growth was observed when the factors were used singly. Similarly, enrichments of 16-fold were obtained from FU2dBM using immunomagnetic Dynabeads with anti-7/4 plus anti-B220 (meaning plating efficiency 0.5%). The further additions of human granulocyte CSF or mouse granulocyte-macrophage CSF or both to IL-1 plus IL-3 plus CSF-1 did not increase HPP-CFC colony formation, but both augmented the small colony formation with IL-1 plus IL-3, IL-3 plus CSF-1, or IL-1 plus CSF-1.