AC133, a novel marker for human hematopoietic stem and progenitor cells.

AC133 is one of a new panel of murine hybridoma lines producing monoclonal IgG antibodies (mAbs) to a novel stem cell glycoprotein antigen with a molecular weight of 120 kD. AC133 antigen is selectively expressed on CD34(bright) hematopoietic stem and progenitor cells (progenitors) derived from human fetal liver and bone marrow, and blood. It is not detectable on other blood cells, cultured human umbilical vein endothelial cells (HUVECs), fibroblast cell lines, or the myeloid leukemia cell line KG1a by standard flow cytometric procedures. All of the noncommitted CD34(+) cell population, as well as the majority of CD34(+) cells committed to the granulocytic/monocytic pathway, are stained with AC133 antibody. In vitro clonogenicity assays have demonstrated that the CD34(+)AC133(+) double-positive population from adult bone marrow contains the majority of the CFU-GM, a proportion of the CFU-Mix, and a minor population of BFU-E. The CD34(dim) and AC133(-) population has been shown to contain the remaining progenitor cells. AC133-selected cells engraft successfully in a fetal sheep transplantation model, and human cells harvested from chimeric fetal sheep bone marrow have been shown to successfully engraft secondary recipients, providing evidence for the long-term repopulating potential of AC133(+) cells. A cDNA coding for AC133 antigen has been isolated, which codes for a polypeptide consisting of 865 amino acids (aa) with a predicted size of 97 kD. This antigen is modeled as a 5-transmembrane molecule, a structure that is novel among known cell surface structures. AC133 antibody provides an alternative to CD34 for the selection and characterization of cells necessary for both short- and long-term engraftment, in transplant situations, for studies of ex vivo expansion strategies, and for gene therapy.

Characterization of c-kit expression by primitive hematopoietic progenitors in umbilical cord blood.

Human umbilical cord blood (CB) appears to be an exciting new source of transplantable stem cells for a variety of clinical conditions. In this study, we have attempted to further characterize the primitive progenitors in CB. First we analyzed the effects of early-acting growth factors on blast cell colony formation from CD34+ progenitors. Addition of Steel factor (SF), interleukin-6 (IL-6), or granulocyte colony-stimulating factor (G-CSF) to cultures containing interleukin-3 enhanced blast cell colony formation. These results indicated that cell cycle-dormant progenitors are present in CB. Next, based on results obtained in the murine system, we tested whether c-kit expression could separate the CB progenitors into cycle-dormant vs. cycle-active progenitors. Cells were separated into CD34+ c-kit-, c-kitlow, and c-kithigh. The results suggested that the c-kitlow population contains the majority of cycle-dormant progenitors and the c-kithigh population contains most of the forming cells were in the c-kitlow population, while the opposite is true for other colony-forming cells. Expression of c-kit may be useful in identifying CB progenitors with long-term engraftment capability.

Human bone marrow and umbilical cord blood cells generate CD4+ and CD8+ single-positive T cells in murine fetal thymus organ culture.

Murine fetal thymus lobes isolated from both normal and scid/scid mice can be colonized by donor cells from either human bone marrow or human umbilical cord blood in vitro. Subsequent organ culture results in a transient production of a few CD4+ CD8+ (double-positive) cells and then the accumulation of CD4+ or CD8+ (single-positive) T cells. A significant number of immature T-cell intermediates (e.g., CD8low, CD3-/low cells) were present in early organ cultures, suggesting that these were progenitors of the mature CD3+/high single-positive T cells that dominated late cultures. Depletion of mature T cells from the donor-cell populations did not affect their ability to colonize thymus lobes. However, colonization depended on the presence of CD7+ progenitor T cells. Limiting dilution experiments using mature T-cell populations (human peripheral blood leukocytes, human bone marrow cells, and human umbilical cord blood cells) suggested that thymic organ culture supports the growth of progenitor T cells but does not support the growth of mature human T cells. Each of these donor populations produced single-positive populations with different CD4/CD8 ratios, suggesting that precursor cells from different sources differ qualitatively in their capacity to differentiate into T cells.

Growth factor requirements for survival in G0 and entry into the cell cycle of primitive human hemopoietic progenitors.

In this study we have isolated populations of dormant human hemopoietic progenitors by two different approaches. First CD34+ cells isolated by panning were further separated on the basis of absence of HLA-DR expression by using fluorescence-activated cell sorting. Second, CD34+ HLA-DR- cells were isolated by nonadherence to soybean agglutinin, negative immunomagnetic bead selection with lineage-specific antibodies, and two-color cell sorting. Progenitors in either cell population were unable to form colonies in the presence of interleukin (IL)-3 alone but yielded a substantial number of colonies, including multilineage colonies, in the presence of combinations of IL-3 and IL-6. Similarly, IL-3 plus any one of the other synergistic factors, including granulocyte colony-stimulating factor, IL-11, leukemia inhibitory factor, and steel factor, effectively supported colony formation from CD34+ HLA-DR- progenitors. Sequential observation of colony formation from single CD34+ HLA-DR- cells provided definitive evidence that the synergistic factors trigger cell divisions of dormant cells. Studies with delayed addition of factors to the cultures provided evidence that this population of cells also requires IL-3 or granulocyte/macrophage colony-stimulating factor (GM-CSF) to survive even while dormant. In contrast, none of the synergistic factors were able to replace IL-3 or GM-CSF in this function. These findings confirm and extend the model that multiple factors with overlapping functions operate both independently and in combination to regulate early stages of hemopoiesis.

Blast colony-forming cells and precursors of colony-forming cells detectable in long-term marrow culture express the same phenotype (CD33- CD34+).

We have previously shown that CD33- CD34+ human marrow cells are capable of giving rise to colony-forming cells (CFC) in long-term marrow culture (LTMC) but are mainly depleted of progenitors that directly form colonies in semi-solid media. In contrast, the CD33+ CD34+ population contains most of the CFC but not the precursors of these cells detectable in LTMC. The purpose of the present study was to determine if a form of CFC with self-renewal potential, the blast CFC, is contained within the CD33+ CD34+ or CD33- CD34+ population. The results demonstrate that blast CFC segregate within the CD33- population, representing a distinct progenitor population that may be analogous to progenitors for CFC detectable in LTMC.

Leukemia inhibitory factor differentiation-inhibiting activity/human interleukin for DA cells augments proliferation of human hematopoietic stem cells.

Leukemia inhibitory factor (LIF)/differentiation-inhibiting activity (DIA)/human interleukin for DA cells (HILDA) is a cytokine with biologic activities involving a variety of different types of target cells. Here we have tested LIF/DIA for possible effects on the growth and differentiation of normal human hematopoietic cells in culture. As a single agent, LIF/DIA had no effect on colony formation by CD34-positive human bone marrow cells. However, LIF/DIA was as effective as either interleukin-6 (IL-6) or granulocyte colony-stimulating factor (G-CSF) in the enhancement of IL-3-dependent colony formation of very primitive blast colony-forming cells. Studies using neutralizing antibodies against IL-6 or G-CSF demonstrated that this was not due to induction in culture of either of the other known synergistic factors for blast cell colony formation. A 1-day delay in the time course of appearance of blast cell colonies grown in the presence of LIF/DIA relative to those grown in the presence of IL-6 suggests that the different synergistic factors may operate through different mechanisms, although we cannot rule out that high doses of LIF/DIA might yield accelerated blast cell colony formation. Our findings provide evidence that LIF/DIA may play an important role, along with IL-6 and G-CSF, in the regulation of early hematopoietic stem cells.

Survival of hemopoietic progenitors in the G0 period of the cell cycle does not require early hemopoietic regulators.

Although it is generally held that hemopoietic stem cells in steady-state marrow are dormant in the cell cycle, the direct proof for this concept has been lacking. In the present study, we have documented the development of human multipotential blast cell colonies from single cells by daily observation of the growth of candidate progenitors. The results clearly demonstrated that early hemopoietic progenitors may remain as single cells for more than 2 weeks of incubation. Once the progenitors began proliferation, the subsequent growth was characterized by steady cell doubling. Next, we tested the survival of blast cell colony progenitors in the presence of neutralizing antibodies prepared against early acting hemopoietic factors including interleukin (IL) 1 alpha, IL-1 beta, IL-3, IL-6, and granulocyte colony-stimulating factor. Cultures were initiated with individual antibodies, and, on day 14, IL-3 and the corresponding growth factor in concentrations that neutralize the antibodies were added. On days 18-27 of culture, blast cell colonies containing 25 or more cells were identified and replated for analysis of their ability to form secondary colonies. The cumulative frequency of the blast cell colonies in cultures containing antibody did not differ significantly from that of the control group containing rabbit IgG. A combination of anti-IL-1 alpha, anti-IL-1 beta, anti-IL-6, and anti-granulocyte colony-stimulating factor did not affect the survival of dormant blast cell colony-forming cells. These results indicate that survival of hemopoietic stem cells in the G0 period of the cell cycle is independent of early hemopoietic regulators.

Synergism between interleukin-6 and interleukin-3 in supporting proliferation of human hematopoietic stem cells: comparison with interleukin-1 alpha.

Currently available evidence suggests that in the steady state, the majority of hematopoietic stem cells are dormant in cell cycle and reside in the so-called G0 period. Studies in our laboratory indicated that once a stem cell leaves G0, its subsequent proliferation requires the presence of interleukin-3 (IL-3). Recently it was reported that interleukin-1 (IL-1) may stimulate stem cells to become sensitive to IL-3. In a separate study, we observed that interleukin-6 (IL-6, also known as B cell stimulatory factor-2/interferon beta 2) possesses synergism with IL-3, shortening the G0 period of murine hematopoietic stem cells. We report here that human IL-6 and IL-3 act synergistically in support of the proliferation of progenitors for human blast cell colonies and that IL-1 alpha reveals no synergism with IL-3 when tested against purified human marrow progenitors. Panned My-10+ human marrow cells were plated in culture and on day 14 of incubation, either IL-3, IL-6, IL-1 alpha or a combination of these factors was added to the cultures. Blast cell colony formation was analyzed daily between days 18 and 32 of culture. IL-6 or IL-1 alpha alone failed to support blast cell colony formation. In the presence of IL-3 alone, blast cell colonies continued to emerge between days 21 and 27. When a combination of IL-3 and IL-6 was added, blast cell colonies developed earlier than in cultures with IL-3 alone and twice as many blast cell colonies were identified. IL-1 alpha failed to augment IL-3-dependent blast cell colony formation. Replating studies of the individual blast cell colonies revealed various types of single as well as multilineage colonies. These observations suggest that IL-6 shortens the G0 period of human hematopoietic stem cells and that the reported synergistic activities of IL-1 on primitive hematopoietic cells may be indirect.

Recombinant gibbon interleukin 3 supports formation of human multilineage colonies and blast cell colonies in culture: comparison with recombinant human granulocyte-macrophage colony-stimulating factor.

The genetic sequences encoding the gibbon and human interleukin 3 (IL 3) proteins were molecularly cloned. The amino acid sequence of the mature gibbon IL 3 protein proved to share 93% homology with the corresponding human protein. We examined the effects of biosynthetic (recombinant) gibbon IL 3 on the proliferation and differentiation of an enriched population of human hematopoietic progenitors and compared the results with the effects of recombinant human granulocyte-macrophage colony-stimulating factor (GM-CSF). Gibbon IL 3 supported the formation of various types of single lineage as well as multilineage colonies by My-10+ bone marrow cells in the presence of human erythropoietin (Ep). In contrast, recombinant human GM-CSF supported the formation of single-lineage colonies and only a small number of multilineage colonies. Both IL 3 and GM-CSF had significant erythroid burst-promoting activity (BPA). Delayed addition of gibbon IL 3 to low serum culture of My-10+ marrow cells supported the formation of blast cell colonies with variable but high replating capability. Human GM-CSF was less effective than IL 3 in support of multipotential blast cell colonies. These results are analogous to the effects of murine IL 3 and GM-CSF on murine progenitors and support the notion that the primary factor for multipotential progenitors is IL 3.

Blast cell colony assay for umbilical cord blood and adult bone marrow progenitors.

We previously described candidate human blast cell colonies in culture of umbilical cord blood cells. However, their replating efficiencies were low, and we were unable to grow colonies from adult marrow cells. We report here a consistent method of growth and identification of human blast cell colonies that are supported by low serum culture and by delayed addition of medium conditioned by a T lymphoblast cell line, C5MJ. Nonadherent mononuclear cord blood and bone marrow cells were prepared by use of Ficoll-Paque and overnight adherence to plastic. Bone marrow cells were further enriched for progenitors by panning with monoclonal anti-My-10 antibody. Cells were plated in methylcellulose culture containing 2% fetal calf serum and supplemented with bovine serum albumin, lecithin, cholesterol, and transferrin. On day 14 of culture, concentrated C5MJ-conditioned medium was carefully added to each dish. Blast cell colonies consisting of 18 to 100 cells were detected on days 21 to 28. Forty percent to 75% of the blast cell colonies in individual samples yielded secondary colonies upon replating (positive colonies). The replating efficiency of the positive colonies ranged from 3% to 100%. The largest secondary colony contained 7,800 cells. In addition to single-lineage colonies, multilineage colonies revealing two to five lineage combinations were seen. These results suggest that human primitive progenitors are dormant in cell cycle and that they survive in the absence of colony-stimulating factors. Human blast cell colonies may provide a unique population of progenitors for studies of the early process of human hemopoiesis.

Disparate differentiation in hemopoietic colonies derived from human paired progenitors.

We analyzed the differentiation of hemopoietic colonies derived from human paired daughter cells. Candidate progenitor cells were isolated by use of a micromanipulation technique from cultures of My-10 antigen-positive cord blood cells. Then nine to 36 hours later, the paired daughter cells were separated with a micromanipulator and allowed to form colonies in methylcellulose medium containing erythropoietin, phytohemagglutinin leukocyte-conditioned medium, and platelet-poor plasma. The cellular composition of the colonies was determined by differentiating all of the cells of the May-Grünwald-Giemsa-stained preparation. Of a total of 75 evaluable pairs of colonies, 35 consisted of 28 types of disparate pairs revealing nonhomologous lineage combinations. Forty pairs were homologous in lineage expression. However, the proportions of the individual cell lineages were significantly different in the members of some of the homologous pairs. Some pairs revealed significant differences in colony size. These observations are similar to those reported for murine paired progenitors and are consistent with the stochastic model of human stem cell differentiation.

Single cell origin of multilineage colonies in culture. Evidence that differentiation of multipotent progenitors and restriction of proliferative potential of monopotent progenitors are stochastic processes.

In this paper, we report analysis of differentiation in human hemopoietic colonies derived from a single cell. Cord blood mononulear cells and panned My-10 antigen-positive bone marrow and cord blood cells were plated in methylcellulose medium containing erythropoietin and conditioned medium. Initially, we performed mapping studies to identify candidate colony-forming cells. Subsequently, using a micromanipulator, we transferred single cells individually to 35-mm dishes for analysis of colony formation. Cellular composition of the colony was determined by identifying all of the cells in the May-Grunwald-Giemsa stained preparation. Of 150 single candidate cells replated, 63 produced colonies. The incidences of single lineage colonies included 19 erythroid, 17 monocyte-macrophage, and 9 eosinophil colonies. There were 18 mixed hemopoietic colonies consisting of cells in two, three, four, and five lineages in varying combinations. In some instances, we noted the predominance of one lineage and the presence of very small populations of cells in a second or third lineage. These results provide evidence for the single-cell origin of human multilineage hemopoietic colonies, and are consistent with the stochastic model of stem cell differentiation in man. They also indicate that restriction of the proliferative potential of committed progenitors is a stochastic process.

Circulating eosinophil colony-forming cells in pure eosinophil aplasia.

Using a methylcellulose clonal culture method, we found eosinophil colony-forming cells in the blood of two patients: a 46-year-old white woman with a long history of urticaria and allergic rhinitis and complete absence of eosinophils secondary to IgG-mediated and complement-mediated destruction of her eosinophils; and a white woman with a history of allergic rhinitis and asthma and a complete absence of eosinophils in her peripheral blood. Addition of the patients' plasma or IgG to cultures of peripheral blood mononuclear cells derived from normal persons or the patients inhibited formation of eosinophil colonies without influencing colony formation in other hemopoietic cell lineages. In contrast, normal plasma or IgG had no effect on the formation of eosinophil colonies. That the IgG-mediated cytotoxic activity of eosinophils prevents the expression of an apparently normal eosinophilopoietic potential suggests an autoimmune basis for the eosinophil aplasia.

Identification of pure and mixed basophil colonies in culture of human peripheral blood and marrow cells.

We present a colony assay system that allows in situ identification of human basophil/mast cell (basophil) colonies. In methylcellulose culture, in the presence of phytohemagglutinin-leukocyte conditioned media (PHA-LCM), human peripheral blood and bone marrow cells form colonies that can be distinguished by their unique morphological characteristics. Pure basophil colonies are diffuse, small colonies containing small, round, highly refractile cells. These characteristics of the constituent cells led us to the observation that a significant number of basophils are found in combination with eosinophils. The mixed eosinophil/basophil colonies have the distinctive elements of pure eosinophil and pure basophil colonies. Usually, these are diffuse colonies with compact clusters of slightly larger, darker-appearing cells. We also found colonies that contained basophils and neutrophils/monocytes, but this type could not be consistently identified by in situ morphology. Cytochemical analysis confirmed the metachromatic nature of the granules in the basophils. The presence of IgE receptors on the cells was documented by indirect immunofluorescent staining after passive sensitization with purified human IgE. Peripheral blood cells from six healthy volunteers formed 5.7 +/- 1.0 (mean +/- SEM) pure colonies in 2 X 10(5) cells. Cultures of bone marrow cells from patients with various types of anemia had 9.0 +/- 1.5 colonies in 10(5) cells. This is the first description of a colony assay system for in situ identification of a pure population of basophilic granulocytes.

Suspension culture of human mast cells/basophils from umbilical cord blood mononuclear cells.

Selective growth of human mast cells/basophils was obtained in suspension cultures of mononuclear cells from umbilical cord blood. A fraction of culture supernatant of phytohemagglutinin-stimulated T cells, which lacked interleukin 2, was required for the selective growth of mast cells. When the mononuclear cells were cultured for 2-4 wk in the presence of the fraction, 50-90% of the total cells in the cultures contained metachromatic granules. Under the optimal culture conditions, the number of mast cells/basophils recovered from the cultures was 30-60% of the number of mononuclear cells plated. Cultured mast cells/basophils bear 1.2-3.83 X 10(5) IgE receptors per cell and contained 0.48-1.6 micrograms of histamine per 10(6) cells. The average forward rate constant, k1, and dissociation constant, k-1, for the binding of human IgE to IgE receptors on the cells were 1.9 X 10(5) M-1 sec-1 and 6.9 X 10(-5) sec-1, respectively (average equilibrium constant = 2.75 X 10(9) M-1). Specific binding of human IgE with high affinity indicates that the cells recovered in the suspension culture are human mast cells/basophils. Cultured cells sensitized with human IgE released a substantial amount of histamine upon exposure to anti-IgE. The results indicate that human mast cells/basophils obtained in the culture are functionally mature.

Quantitation of hemoglobin biosynthesis with agarose isoelectric focusing.

Isoelectric focusing in polyacrylamide gels has been widely used to separate fetal (HbF) and adult (HbA) hemoglobins, and the relative synthesis of HbA and HbF has been estimated by fluorography or autography of dried gels. In order to measure the absolute synthesis of hemoglobins, we developed a system that utilizes isoelectric focusing in agarose and quantitates the total radioactivity of separated hemoglobins. After isoelectric focusing, the protein bands are individually eluted from the agarose and the radioactivity measured by liquid scintillation counting. We used this technique to study the synthetic capabilities of erythroid precursors at sequential times in culture. As previously reported, the relative ratio of HbF decreased over time in culture. However, our results clearly revealed that the absolute synthesis of HbF did not decrease until there was a parallel decrease in hemoglobin A, and that changes in the relative ratio occur because of disproportionate increases in HbA. This methodology, allowing independent evaluation of the radioactivity in synthesized hemoglobin, enabled us to gain new insight into the biosynthetic capabilities of erythroid precursors in clonal cell culture.

Enhancement of the growth of human early erythroid progenitors by bone marrow conditioned media.

The inclusion of condition media (CM) obtained from cultures of human bone marrow cells increased the number of erythroid bursts observed after a 14-day culture period. In contrast, CM had no effect on the number of erythroid colonies observed at five days. The effect of CM on burst growth was more pronounced at limiting growth conditions--low serum or low erythropoietin (EPO) concentrations. Burst-forming units (BFUe) survived in culture and were able to respond to CM by increased colony growth and hemoglobin synthesis as late as eight days following the initiation of cultures. CM prevented the decline in the number of burst observed in the absence of EPO during the first four days of culture; however, when EPO was added at later times, a decrease in burst formation was seen. Increased sensitivity of burst-forming cells to low concentrations of EPO and a reduced serum requirement were observed in cultures containing CM. The incorporation of 59Fe into heme was also increased by CM, giving a more sensitive test of CM activity than burst number. CM obtained from bone marrow cultures, therefore, probably contains burst-promoting activity (BPA), one of the factors required for the proliferation of early erythroid progenitors.

Cell-mediated immune responses to Pseudomonas aeruginosa.

Of six healthy laboratory workers with no history of illness caused by Pseudomonas aeruginosa, four exhibited positive skin responses at 24 hours, strong inhibition of leukocyte migration, and marked stimulation of blastogenesis of peripheral lymphocytes when exposed to a "cocktail" of three phenol-killed strains of P aeruginosa. There was no correlation between the extent of these responses with serum hemagglutinin titers. It is postulated that cell-mediated immunity against P aeruginosa may have a biologic role in host defense.

The effect of antibiotics on cell-mediated immunity.

Normal human lymphocytes in culture were stimulated by the addition of phytohemagglutinin, and deoxyribonucleic acid (DNA) synthesis was measured by the incorporation of tritiated thymidine. The effect of 11 commonly used antibiotics on DNA synthesis then was measured by adding each antibiotic to the culture in concentrations ranging from 2 to 64 microng/ml, a range which covers the plasma levels obtained during customary clinical therapy. Severe dose-dependent suppression of DNA synthesis was found in the presence of two preparations of minocycline, oxytetracycline, and the ascorbic acid salt of tetracycline. Less severe but stil significant suppression was found in the presence of chloramphenicol, clindamycin, tetracycline, and ascorbic acid alone. No effect was noted when penicillin, carbenicillin, or cephalothin was added, and slight stimulation was found in the presence of gentamicin. It is postulated that these findings may help to explain fungal and viral superinfection following antibiotic therapy and that they may play a role in the failure of antibiotic therapy to eliminate some infections.