Friend virus-infected long-term bone marrow cultures produce colony stimulating factor dependent and independent granulocyte-macrophage progenitor cells for over four years in vitro

C3H/HeJ mouse long-term bone marrow cultures infected at initiation with a cloned polycythemic strain of Friend spleen focus forming virus in a cloned N-tropic murine leukemia virus helper virus coat, persistently produced: colony-forming unit spleen (CFUs) for 55 weeks that formed macroscopic spleen colonies in syngeneic or allogeneic C57B10.Br/J mice; and L-cell or WEHI-3 cell conditioned medium-dependent granulocyte-macrophage colony forming unit culture (GM-CFUc); and morphologically normal granulocytes for over 245 weeks. Colony stimulating factor (CSF)-independent colony forming progenitor cells were first detectably produced in vitro at 75 weeks, and when subcultured generated karyotypically distinct permanent factor-independent tumorigenic cell lines. Nonadherent cells removed from long-term marrow cultures at 19 but not at 77 weeks reconstituted donor origin hematopoiesis in C57B10.Br/J mice as measured by B-cell lineage surface immunoglobin allotype. Nonadherent cells removed at 77 weeks produced lethal splenomegaly and marrow infiltration with culture origin cells in C57B10.Br/J mice. Despite generation of clonal malignant cell lines, L-cell DSF (CSF-1, M-CSF) responsive GM-CFUc that were simultaneously produced over 4 years in the same long-term marrow cultures, grew to 7 day colonies in semisolid medium and terminally differentiated. Thus, adherent stromal cells in Friend virus-infected long-term bone marrow cultures simultaneously support CSF-responsive and malignant CSF-independent hematopoietic progenitor cells.     

In vitro quantitation of lethal and physiologic effects of total body irradiation on stromal and hematopoietic stem cells in continuous bone marrow cultures from Rf mice

The role of stromal-supportive cells in hematopoietic stem cell responses to irradiation is poorly understood. The effects of in vivo total body irradiation (TBI) and interval from TBI to explant of marrow on: stromal cell proliferation in vitro; stromal cell support of hematopoiesis in continuous bone marrow culture; and generation of WEHI-3 growth factor (GF)-dependent lines of hematopoietic progenitor cells were evaluated. Continuous marrow cultures from non-irradiated control RfM/UN, C57BL/6J, C3H/HeJ, and N:NIH (Swiss) mice generated pluripotential hematopoietic stem cells (CFUs) and committed granulocyte-macrophage progenitor cells (GM-CFUc) for over 20 weeks. Explant of marrow at 2, 4, 5, or 6 months after single fraction TBI (300-800 rad) was associated with decreased longevity of hemopoiesis (2-12 weeks), and a decrease in the proliferative capacity of fibroblastic adherent-stromal colony forming cells (CFUf) as measured by colony size at 14 days and number of colonies per 10(6) cells plated. In contrast, explant of marrow 8 to 24 months after TBI produced cultures with longevity that was indistinguishable from age-matched control cultures (19-24 weeks). Marrow from irradiated first and second generation recipients of serially transferred marrow demonstrated a similar 7-month in vivo recovery period; however, the plateau maximum duration of hemopoiesis did not return to control levels. Purified stromal cell cultures were prepared by corticosteroid-deprivation of explanted marrow for 28 days and were then engrafted in vitro with marrow from C57BL/6J or RfM/UN mice that had been irradiated 1 month previously. Hemopoiesis in these cultures was restored, and they produced GM-CFUc and granulocytes for 15-24 weeks. Thus, healthy stroma supported growth of recently irradiated hemopoietic cells in vitro. Nonadherent cells removed from the above continuous marrow cultures generated clonal non-leukemogenic WEHI-3 GF-dependent hemopoietic progenitor cell lines with a frequency concordant with radiation effects on culture longevity, and this was increased by the presence of purified healthy stromal cultures. Indirect effects of x-irradiation on hemopoietic stem cells through damage and repair in the stromal cell compartment can be effectively studied with the present bone marrow culture system.     

Effects of monoclonal antibody and complement treatment of human marrow on hematopoiesis in continuous bone marrow culture

Long-term bone marrow cultures were established from single-cell suspensions of human bone marrow that had been treated with monoclonal antibodies and complement. Each treated cell suspension was evaluated for production of hematopoietic stem cells over 20 weeks. Treatment with antibody to HLA-DR (Ia), B1, J2, or J5 did not remove adherent cells including those differentiating to adipocytes in 17-hydroxycorticosteroid. In contrast, treatment with monoclonal antibody directed against human beta 2-microglobulin reduced adipocyte numbers by 100-fold and reduced the total adherent cell density over 70%. Cumulative total nonadherent cell and granulocyte-macrophage colony-forming units (GM-CFUc) production over 20 weeks was not significantly altered by one cycle of anti-Ia plus complement or up to three cycles of treatment with complement and anti-J2, -J5, or -B1. However, one cycle of treatment with anti-beta 2-micro-globulin depressed production of both GM-CFUc and nonadherent cells by over 100-fold compared to other treatment groups. While one cycle of treatment of anti-Ia and complement killed all detectable cells forming CFU-erythroid-granulocyte-megakaryocyte-macrophage, blast-forming units (erythroid), and GM-CFUc, GM cluster-forming cells survived. Treatment of marrow with three cycles of anti-Ia and complement removed all detectable GM colony- and GM cluster-forming cells; however, this marrow produced fewer cumulative Ia-positive GM-CFUc. Long-term bone marrow cultures may prove to be an interesting system for in vitro analysis of the effects of new immunotherapeutic agents including other monoclonal antibodies prior to clinical use.     

Phenotypically distinct target cells for murine sarcoma virus and murine leukemia virus marrow transformation in vitro.

An in vitro hematopoietic microenvironment was established from explained fragments of bone marrow from adult noninbred NIH Swiss mice with the use of corticosteroid-reconstituted horse serum. Infection with Kirsten murine sarcoma virus (Ki-MuSV) with either a Rauscher murine leukemia virus (R-MuLV) or Balb:virus-1 helper virus coat reduced proliferation of granulocytic and pluripotent hematopoietic stem cells and produced neoplastic transformation of both macrophages and preadipocytes in the adherent cell population within a 4-week period. Ki-MuSV-transformed, virus-releasing macrophages formed clusters of 4-49 cells in 0.8% methylcellulose-containing medium in the absence of added colony-stimulating factor (CSF), synthesized lysozyme, ASD-chloroacetate substrate-specific esterase-M, and CSF, and produced tumors following inoculation iv into adult NIH Swiss mice or ip into newborn NIH Swiss mice. In cultures infected with helper leukemia viruses R-MuLV or Balb:virus-1, gradual transformation of a distinct cell phenotype was observed over a 9-week period with generation of increasing numbers of atypical myeloblasts and promyelocytes which showed dyssynchronous nuclear-cytoplasmic maturation, basophilic granulation, cytoplasmic vacuolation, and formation of incompletely maturing CSF-dependent granulocyte-macrophage colonies in vitro and small spleen colonies in vivo. These data demonstrated that rapid biologic expression of the murine sarcoma virus genome in specific adherent "stromal" marrow cells prevents detection of a more subtle helper-virus-induced dysmyelopoiesis in a distinct nonadherent cell population.     

Presence of colony promoting activity (CPA) in the supernatant of the long-term bone marrow cultures

Colony promoting activity (CPA) was detected in the supernatants of the bone marrow cell cultures. CPA itself had no granulocyte-macrophage colony stimulating activity (CSF); but did increase the number and the size of colonies induced by the optimal dose of CSF. This activity was dose dependent and not specific to CSF sources. CPA-responsive cells (target cells) were found in the bone marrow cell cultures as well as in the freshly isolated bone marrow.     

Pool size of pluripotential hematopoietic stem cells increased in continuous bone marrow culture by Friend spleen focus-forming virus

Continuous mouse bone marrow cultures were infected with Friend murine leukemia virus. Production of nonadherent (NA) and adherent cells, granulocyte-macrophage colony-forming unit(s) of progenitor cells (GM-CFUc), pluripotential hematopoietic stem cells (CFUs), the self-renewal potential (Rs) of CFUs, and generation of factor-dependent (FD) multipotential and committed permanent stem cell cloned lines were measured. Uninfected marrow cultures from C57BL/6J, C57BL/6JUt, B6.S, C3H/HeJ, (C57BL/6J x DBA/2J)F1, CD- 1 Swiss, or N:NIH(S) mice generated NA cells, GM-CFUc, and CFUs for 20-41 weeks; cultures infected with Rauscher or other helper viruses generated them for 35-45 weeks. GM-CFUc and CFUs production in SFFV-positive cultures persisted for over 65 weeks and exceeded control levels by twentyfold to fiftyfold. The Rs of CFUs in SFFV-positive cultures was not detectably increased above control cultures. Multipotential (erythroid-neutrophil-mast cell-basophil-eosinophil) permanent FD cell clones were derived from control and SFFV-positive cultures. Thus SFFV amplifies the stem cell pool in vitro without detectably increasing the Rs capacity of CFUs.     

Characteristics of Granulocyte-Macrophage Colony Formation in Chronic Myelomonocytic Leukemia: A Comparative Study with Other Myelodysplastic and Myeloproliferative Disorders

We studied granulocyte-macrophage (GM) colony formation in chronic myelomonocytic leukemia (CMML, 6 cases), as compared with that in myelodysplastic syndromes (MDS, 6 cases) and myeloproliferative disorders (MPD, 12 cases). GM colony formation of bone marrow cells by colony-stimulating factor (CSF) was normal in CMML and MPD patients, but was decreased in MDS patients. Circulating granulocyte-macrophage progenitors (CFU-GM) were detected in CMML and MPD patients, but not in MDS patients. GM colony formation without CSF was observed in CMML patients, but not in MDS or MPD patients. These endogenous colonies decreased markedly after adherent cell (AdC) depletion, but AdC did not form endogenous colonies in sufficient numbers to explain their marked decrease after AdC depletion. In CMML patients, the numbers of circulating CFU-GM and endogenous colonies correlated with leukocyte and monocyte counts, respectively. The cellular composition of GM colonies was normal in MDS and MPD patients, whereas granulocytic colonies predominated in all CMML patients but one. The CSF-producing capacity of peripheral blood cells was also studied and was found to be increased in CMML patients. This capacity was markedly decreased by AdC depletion; and AdC could produce CSF only in CMML patients. CSF produced by CMML patients supported granulocytic colonies to a greater extent than CSF produced by MDS or MPD patients. These results suggest that enhanced granulopoiesis in CMML patients is closely associated with the possible hyperproduction of granulocytic CSF by their adherent monocytes.     

Characteristics of granulocyte-macrophage colony formation in chronic myelomonocytic leukemia: a comparative study with other myelodysplastic and myeloproliferative disorders

We studied granulocyte-macrophage (GM) colony formation in chronic myelomonocytic leukemia (CMML, 6 cases), as compared with that in myelodysplastic syndromes (MDS, 6 cases) and myeloproliferative disorders (MPD, 12 cases). GM colony formation of bone marrow cells by colony-stimulating factor (CSF) was normal in CMML and MPD patients, but was decreased in MDS patients. Circulating granulocyte-macrophage progenitors (CFU-GM) were detected in CMML and MPD patients, but not in MDS patients. GM colony formation without CSF was observed in CMML patients, but not in MDS or MPD patients. These endogenous colonies decreased markedly after adherent cell (AdC) depletion, but AdC did not form endogenous colonies in sufficient numbers to explain their marked decrease after AdC depletion. In CMML patients, the numbers of circulating CFU-GM and endogenous colonies correlated with leukocyte and monocyte counts, respectively. The cellular composition of GM colonies was normal in MDS and MPD patients, whereas granulocytic colonies predominated in all CMML patients but one. The CSF-producing capacity of peripheral blood cells was also studied and was found to be increased in CMML patients. This capacity was markedly decreased by AdC depletion; and AdC could produce CSF only in CMML patients. CSF produced by CMML patients supported granulocytic colonies to a greater extent than CSF produced by MDS or MPD patients. These results suggest that enhanced granulopoiesis in CMML patients is closely associated with the possible hyperproduction of granulocytic CSF by their adherent monocytes.     

Regulation of granulocyte and macrophage colony-stimulating factor production by phytohaemagglutinin-treated blood mononuclear cells

The colony-stimulating activity (CSA) of medium conditioned by phytohaemagglutinin (PHA)-stimulated blood mononuclear cells was tested using granulocyte-macrophage colony-forming cells (GM-CFC) from normal bone marrow. Low concentrations of the conditioned medium stimulated granulocytic colony-forming cells (CFC) which formed colonies by the seventh day of incubation; higher concentrations stimulated the formation of macrophage colonies which were not seen until the end of the second week in culture.The colony-stimulating activities could not be demonstrated in adherent cell-depleted bone marrow cultures. This dependence of activity on adherent cells was confirmed by incubating different concentrations of conditioned medium with isolated adherent cells and then testing for colony-stimulating activity in cultures of non-adherent bone marrow cells. The activities of conditioned media following exposure to adherent cells corresponded to the results seen when the conditioned medium from PHA-stimulated mononuclear cell cultures was used to stimulate agar cultures of unseparated marrow. The results suggest that PHA-responsive mononuclear cells (probably T lymphocytes) may modulate the regulation of colony-stimulating factor (CSF) production by adherent colony-stimulating cells (CSC).     

Clonal analysis of the action of GM-CSF on the proliferation and differentiation of myelomonocytic leukemic cells.

With 214 subclones of the BALB/c myelomonocytic leukemia WEHI-3B, the granulocyte-macrophage colony-stimulating factor (GM-CSF) in impure or purified form, consistently increased the proportion of colonies exhibiting partial or complete differentiation in agar cultures. GM-CSF also increased colony size and content of daughter colony-forming cells. Serial recloning of WEHI-3B colonies in the presence of GM-CSF showed that when colonies differentiated completely, self-replication of the colony-forming cell was suppressed (clonal extinction). However, WEHI-3B cells exhibited clonal instability and even in the continuous presence of GM-CSF many colony-forming cells still generated cells able to form undifferentiated colonies. It appears unlikely that GM-CSF can completely suppress the progressive proliferation of a myeloid leukemic population of the WEHI-3B type.     

Role of stromal and hematopoietic stem cells in Friend spleen focus forming virus effects in continuous bone marrow culture

Replication of the Friend spleen focus forming virus (SFFV) in C3H/HeJ or C57BL/6J mouse continuous bone marrow cultures is associated with an increased cumulative production of pluripotential/hematopoietic stem cells (CFUs), granulocyte-macrophage progenitor cells (GM-CFUc), and total granulocytes, compared to uninfected or helper virus infected control cultures. The site(s) of action of the virus are not known. To determine whether viral effect(s) occurred in adherent stromal and/or non-adherent hematopoietic stem cells, purified cell populations, comprised exclusively of cells from each compartment, were separated from C57BL/6J (Fv-1^bb) or C3H/HeJ (Fv-1ⁿⁿ) marrow cultures and were left uninfected or were infected with host range replication restricted B-tropic or N-tropic helper pseudotype viruses of SFFV respectively. Cell populations were then mixed to establish continuous hematopoiesis in allogeneic or syngeneic combinations. Virus host range restriction between compartments was maintained in allogeneic reconstituted cultures which showed active hematopoiesis for 16–17 weeks and no significant differences between: (1) virus infected and uninfected groups or (2) groups comprised of stromal cells from one or the other strain. Thus, these data indicate that the prolongation of hematopoiesis in undisturbed long-term marrow cultures by SFFV occurs through an interaction of adherent hematopoietic stem cells with the marrow stroma.     

Retroviral src gene expression in continuous marrow culture increases the self-renewal capacity of multilineage hematopoietic stem cells

To define the action of the retroviral src gene on hematopoietic stem cells, C57BL/6 x DBA/2 (B6D2F1) mouse long-term marrow cultures were infected at initiation with Moloney murine leukemia virus (MuLV) pseudotypes of src-recombinant retroviruses with the src gene inserted in the env region of an amphotropic MuLV (src-Ampho), or in the gag region of Moloney MuLV (src-Mo). Other cultures were infected with Friend spleen focus-forming virus polycythemia-inducing strain (SFFVp), Moloney MuLV, or amphotropic MuLV, or were uninfected controls. Harvested nonadherent cells were tested weekly for multilineage, granulocyte-erythroid-megakaryocyte macrophage (CFU-GEMM) colony formation in vitro in recombinant murine IL-3 and erythropoietin, and individual colonies were removed, split 1:2, with half of each replated for in vitro self-renewal and the other half examined morphologically for number of hematopoietic cellular lineages, or tested for release of MuLV and src virus. Cultures infected with src-Ampho, src-Mo, or SFFVp demonstrated a significant increase in cumulative nonadherent cell and CFU-GEMM production. There was prolonged self-renewal over seven serial transfers of individual CFU-GEMM from src virus-infected cultures over seven serial transfers, and five of 61 individual colonies from the second or third generations contained detectable v-src gene sequences, but none released detectable src virus. Self-renewal of CFU-GEMM was similar to that with permanent IL-3-dependent cell line B6SUtA. In contrast, MuLV-infected or control uninfected cultures produced fewer cells, and self-renewal of CFU-GEMM did not exceed three generations. IL-3-dependent clonal hematopoietic progenitor cell lines, derived from each culture group, formed no detectable tumors in vivo; however, each released the original helper and/or transforming virus. Adherent cell lines, derived from src-Ampho-infected cultures released src virus and formed fibro-sarcomas in vivo. The data support the conclusion that src-recombinant virus expression in long-term marrow cultures increases the self-renewal capacity of multilineage hematopoietic stem cells.     

Cytogenetic analysis of human acute and chronic myeloid leukemic cells cloned in agar culture

Karyotypic analysis was performed on agar cultures of blood or bone marrow from 12 patients with acute or chronic myeloid or myelomonocytic leukemia in whom karyotypic markers were present. The granulocytic colonies and clusters which developed on culture were shown to be derived from representative cells of the leukemic population. In two patients with acute leukemia in remission, normal colonies with a normal karyotype were grown from marrow cells but in two patients with chronic myeloid leukemia in remission the Ph¹ abnormality persisted in colony cells. The agar culture technique appears to be ideal for following the emergence and disappearance of leukemic and normal granulopoietic populations in patients with these types of leukemia.     

In vitro erythroid colony formation in acute myelogenous leukemia in the rat.

Erythroid colonies were grown in vitro in plasma clot cultures. Normal adult rat bone marrow responded to exogenous erythropoietin with the formation of an average of 2 colonies/10(3) cells plated. No erythroid colonies were observed in cultured normal spleen preparations. Shay chloro-leukemia cells administered iv induced an acute myelogenous leukemia. During the progressive stages of the disease, the numbers of erythrocyte colony forming units (CFU-E) in the marrow decreased; concomitantly, these progenitors appeared in leukemic spleen cultures. Paralleling changes in CFU-E, the numbers of nucleated red blood cells in the marrow declined but increased in the leukemic spleen. However, compensatory spleen erythropoiesis was transient, due to continued leukemia cell colonization. The loss of erythroid progenitor cells from the bone marrow played a significant role in the anemia associated with this leukemia.     

Clonal analysis of the response of HL60 human myeloid leukemia cells to biological regulators

Human myeloid leukemia (HL60) cells formed colonies in semi-solid agar cultures with a cloning efficiency of 20–90%. Addition of unfractionated human placental conditioned medium (HPCM) or the two semi-purified granulocyte-macrophage colony stimulating factors from HPCM (GM-CSFα and β) increased colony size and the frequency of colonies exhibiting differentiation of colony cells. Differentiation induction using GM-CSFα or β did not reduce the total number of clonogenic cells per colony but did reduce the proportion of clonogenic cells within colonies. Sera from some patients with acute infections exhibited an elevated capacity to induce differentiation both in HL60 colonies and in colonies of the mouse myelomonocytic leukemia cell line, WEHI-3B. A low percentage of HL60 colonies contained maturing eosinophils but the frequency was not influenced by human-active eosinophil colony stimulating factors or by sequential recloning of HL60 colonies.The studies suggest that, as is true for mouse myeloid leukemia cell lines, the granulocyte-macrophage colony stimulating factors are able to induce significant differentiation in human HL60 myeloid leukemia cells.     

Vaccination with leukemia cells expressing cell-surface-associated GM-CSF blocks leukemia induction in immunocompetent mice

The fundamental basis for immunotherapy of leukemia is that leukemic cells express specific antigens that are not expressed by normal hematopoietic cells. However, the host immune system appears to be tolerant to leukemia cells. To overcome this tolerance, we vaccinated immunocompetent mice with murine leukemia cells (WEHI-3B and BCR-ABL+ 32D cells) transduced with a specifically constructed transmembrane form of granulocyte-macrophage colony-stimulating factor (tmGM-CSF). The transduced cells expressed tmGM-CSF on the cell-surface. To determine whether tmGM-CSF-expressing WEHI-3B leukemia cells would prevent leukemia formation as a vaccine, immunocompetent mice (BALB/c and C3H/HEJ) were immunized with lethally irradiated murine leukemia cells expressing cell-surface tmGM-CSF before challenging mice with murine leukemia cells. Two immunocompetent mouse models were investigated, either WEHI-3B cells in BALB/c mice or BCR-ABL+ 32D cells in C3H/HEJ mouse. The results showed that 100% of WEHI-3B/tmGM-CSF-vaccinated BALB/c mice and about 65% of 32D+ BCR-ABL/tmGM-CSF-vaccinated C3H/HEJ mice were protected from leukemia after leukemia cell challenge, whereas all non-vaccinated mice succumbed to leukemia. Spleen and marrow cell suspensions from vaccinated mice challenged with WEHI-3B cells lacked detectable GFP+ WEHI-3B cells at 82 days post-challenge. A significant delay of death was observed in C3H/HEJ mice challenged with the very aggressive DA-1 cell line expressing BCR-ABL. Vaccination of mice with WEHI-3B/CD40L cells protected 80% of the mice from the WEHI-3B challenge. Notably, 60% of the WEHI-3B/BALB/c mice were also protected from leukemia when WEHI-3B/tmGM-CSF vaccination was carried out after the leukemia challenge. In order to determine whether cellular immunity is involved in this vaccine-mediated protection, either CD4+ or CD8+ T cells were depleted from mice after the WEHI-3B/tmGM-CSF vaccination. The results indicate that CD8+ T-cells mediated the protective immune response provided by the irradiated tmGM-CSF-expressing WEHI-3B cells. In addition, vaccination of nude mice did not provide protection from WEHI-3B leukemia induction. Importantly, 80% of non-vaccinated mice were also protected from a WEHI-3B cell challenge after receiving spleen cells from vaccinated mice 1 day before challenge with leukemia cells. These results indicate that overexpression of tmGM-CSF on the leukemia cell-surface can enhance the recognition of leukemic cells by CD8+ T cells, and can either prevent or significantly delay leukemia induction. These findings suggest that injection of irradiated leukemia cells expressing cell-surface-bound GM-CSF has the potential as an immunological approach to treat leukemia.     

Trophoblast cell line conditioned medium for in vitro culture and antigenic characterization of acute myeloid leukemia clonogenic cells

The colony-stimulating factor-containing supernatant of the human trophoblast cell line TPA-30-1 was used to stimulate in vitro growth of acute myeloid leukemia clonogenic cells from 54 patients. Prevalent colony growth (10-greater than 1000) was observed in 63% of cases. In 31% clusters and a few colonies (1-9/1 x 10(5) plated cells) were scored. Neither colonies nor clusters could be detected in the remaining 6%. The best growth was observed in subtype M5 (8 of 9 cases, 89%). Morphological examination and recloning tests suggested that the colonies originated from leukemic progenitors. TPA-30-1 supernatant stimulation can therefore be compared with that of phytohemagglutinin or phytohemagglutinin-leukocyte-conditioned medium. In addition it does not require T-lymphocyte removal and batch screening. Extension of the culture for antigenic characterization of acute myeloid leukemia clonogenic cells to more patients than in a previous study confirmed the existence of a subgroup (39%) of patients whose acute myeloid leukemia clonogenic cells constantly expressed late myeloid differentiation antigens recognized by the monoclonal antibody S4-7. Since S4-7 spares early normal hemopoietic progenitors, this subgroup (54% M2) can be considered as candidates for autologous bone marrow transplantation after in vitro purging with S4-7 monoclonal antibody and complement.     

Multipotent and committed CD34+ cells in bone marrow transplantation.

In order to study the role of CD34+ cells in hematological recovery following bone marrow transplantation (BMT), bone marrow cells stained with HPCA-1 (CD34) and MY-9 (CD33) monoclonal antibodies were analyzed by using a fluorescence-activated cell sorter on or about days 14 and 28, as well as at later times, following BMT in 6 recipients. Single cell cultures of CD34+ cells were also performed to evaluate their in vitro hematopoietic function. CD34+ cells were detectable in bone marrow cells on day 14. More than 80% of CD34+ cells co-expressed the CD33 antigen, and macrophage (Mac) colony-forming cells predominated among total colony-forming cells of CD34+ cells. In normal bone marrow cells, CD34+, CD33+ cells amounted to about 40% of CD34+ cells, and the incidences of erythroid bursts, granulocyte/macrophage (GM) colonies, and Mac colonies were similar to each other. After more than 10 weeks, CD34+, CD33- cells gradually recovered, as erythroid burst colony-forming cells increased following GM colony-forming cells. This phenomenon was well-correlated with the time course of peripheral blood cell recovery. CD34+, CD33+ cells as committed progenitors and CD34+, CD33- cells as multipotent stem cells have distinctive biological behaviors in BMT.     

Heterogeneity of Ia antigen expression by proliferating nonlymphocytic leukemia blast cells

In vitro systems were used to detect Ia-like antigens on proliferating normal myeloid and acute nonlymphocytic leukemia (ANLL) blast cells. Incubation of normal bone marrow cells with a monoclonal anti-Ia antibody and complement resulted in toxicity for both granulocyte/macrophage progenitors (CFU-GM) (toxicity 79%-100%) and cells proliferating in liquid culture in response to placenta-conditioned medium colony-stimulating factor (CSF) or medium conditioned by normal, phytohemagglutinin (PHA)-stimulated mononuclear cells. In contrast, effects of anti-Ia antibody and complement on blast colony-forming cells and 3H-TdR incorporation in liquid culture from eight patients with ANLL were variable. Colony growth with CSF after treatment was 0% to 91% of control growth and did not correlate with display of Ia-like antigens. Survival of ANLL cells growing in liquid cultures was even more variable after anti-Ia+ complement treatment (28%-227% of control). The presence of Ia-like antigens did not distinguish ANLL cells responding to PHA-conditioned medium from those responding to CSF in either colony or liquid culture. Dose-response curves for ANLL cells in liquid culture were similar before and after treatment with anti-Ia+ complement. In contrast to normal myeloid precursor cells, which show uniform display of Ia-like antigens, display of Ia antigen by proliferating leukemia cells is highly variable from patient to patient. Anti-Ia reagents such as this one would not be effective in treating ANLL marrow for autologous transplantation.     

In myelodysplastic syndromes progression to leukemia is directly related to PHA dependency for colony formation and independent of in vitro maturation capacity

In an agar-liquid double-layer colony assay in which myeloid leukemia colony-forming cells require the presence of both the lectin PHA and CSF for in vitro proliferation, colony formation of bone marrow cells derived from patients with a myelodysplastic syndrome (MDS) was studied. In five of 14 MDS and all five leukemic transformed MDS cases, colony formation was found to require both PHA and CSF. Three of these five PHA-dependent MDS cases progressed to overt leukemia within 1 year, one progressed from RA to RAEB, and one patient received AML chemotherapy. PHA-dependent colony formation was associated with higher bone marrow blast counts, but not directly to FAB type or cytogenetic abnormalities. In nine other MDS cases only CSF was required for colony formation. In these PHA-independent cases the course of the disease was stable during the observation time (5-17 months). Two types of colonies were observed in this in vitro system: colonies adherent and colonies nonadherent to the agar underlayer. The former consisted of terminally differentiated myeloid cells, and the latter comprised immature cells. This suggests that the percentage of adherent colonies formed in vitro may be used as a measure for the maturation defect in MDS. However, no correlation was found between the percentage of adherent colonies and progression to leukemia of the MDS cases. Our findings suggest that the dependency on PHA for in vitro colony formation of colony-forming cells in MDS is predictive for the progression to leukemia. However, the in vitro differentiation capacity has no apparent prognostic significance.