D Bonnet 1,2, M Bhatia 1,3, JCY Wang 1, U Kapp 1 and JE Dick 1. Summary:

Transcription

1 Bone Marrow Transplantation, (1999) 23, Stockton Press All rights reserved /99 $ Cytokine treatment or accessory cells are required to initiate engraftment of purified primitive human hematopoietic cells transplanted at limiting doses into NOD/SCID mice D Bonnet 1,2, M Bhatia 1,3, JCY Wang 1, U Kapp 1 and JE Dick 1 1 Department of Genetics, Research Institute, Hospital for Sick Children and Department of Molecular and Medical Genetics, University of Toronto, Toronto, Ontario, Canada Summary: Little is known about the cell types or mechanisms that underlie the engraftment process. Here, we have examined parameters affecting the engraftment of purified human Lin CD34 CD38 normal and AML cells transplanted at limiting doses into NOD/SCID recipients. Mice transplanted with 500 to 0 Lin CD34 CD38 cord blood (CB) or AML cells required the co-transplantation of accessory cells (ACs) or short-term in vivo cytokine treatment for engraftment, whereas transplantation of higher doses ( 5000 Lin CD34 CD38 cells) did not show these requirements suggesting that ACs are effective for both normal and leukemic stem cell engraftment in this model. Mature Lin CD34 and primitive Lin CD34 CD38 cells were capable of acting as ACs even though no repopulating cells are present. Cytokine treatment of NOD/SCID mice could partially replace the requirement for co-transplantation of AC. Furthermore, no difference was seen between the percentage of engrafted mice treated with cytokines for only the first 10 days after transplant compared to those receiving cytokines for the entire time of repopulation. Surprisingly, no engraftment was detected in mice when cytokine treatment was delayed until 10 days posttransplant. Together, these studies suggest that the engraftment process requires pluripotent stem cells plus accessory cells or cytokine treatment which act early after transplantation. The NOD/SCID xenotransplant system provides the means to further clarify the processes underlying human stem cell engraftment. Keywords: stem cells; xenotransplant; NOD/SCID mice Correspondence: Dr JE Dick, Department of Genetics, Research Institute, Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada, M5G 1X8 Present addresses: 2 Coriell Institute for Medical Research, 401 Haddon Avenue, Camden, NJ, USA; 3 The John P Robarts Research Institute, Gene Therapy and Molecular Virology, Perth Drive, London, Ontario, N6A 5K8 and Department of Microbiology and Immunology, University of Western Ontario The first two authors contributed equally to this work Received 29 May 1998; accepted 7 September 1998 The most conclusive way to identify stem cells is to follow repopulation after transplantation. 1,2 Clinically, it has been difficult to identify the factors and cell types that play a role in the engraftment process because of the absence of assays for human repopulating cells. Recently we have developed an in vivo assay for primitive human hematopoietic cells based on their ability to repopulate the bone marrow (BM) of NOD/SCID mice following intravenous injection. 3 The engrafting cell was operationally termed the SCID-repopulating cell (SRC). In this in vivo system, the engraftment of human cells after transplantation infers that the SRC must navigate to the mouse BM, survive in this microenvironment and then proliferate and differentiate extensively. Accumulated data from gene marking, cell purification, and kinetic analysis indicated that SRC are biologically distinct from and more primitive than most progenitors that can be assayed in vitro. 46 Similar approaches were taken to identify acute myeloid leukemia (AML) stem cells, termed SCID-leukemia initiating cells (SL-IC), based on their ability to establish leukemic proliferation in SCID and NOD/SCID mice. 7,8 Although this xenotransplant system has allowed for the detection and purification of engrafting human stem cells that are undoubtedly relevant to clinical bone marrow transplantation, relatively little is known about the mechanism that underlies the engraftment process itself. Available clinical data suggest that depletion of mature cells, such as T cells, from the transplanted BM decreases the rate and extent of donor cell engraftment. 9 In addition, autotransplants of BM which were combined with accessory cell populations such as those in peripheral blood demonstrated that reconstitution occurred earlier than if BM alone was used. This impairment in hematological recovery after depletion of cell fractions has lead to the hypothesis that a facilitating mechanism provided by a subpopulation of cells is required early in the engraftment process especially for allogeneic transplantation. 10 It has been difficult to conclusively establish the requirement and nature of these ACs in murine studies that involve syngeneic transplantation of highly purified murine stem cell populations into lethally irradiated recipients, since short-term repopulating cells must be co-transplanted to enable survival of the recipient. 1,11 Highly purified subfractions found within CD34 populations are important clinically in autotransplants for multiple myeloma and breast cancer since recent studies indicate that cells enriched solely on the basis of CD34

2 204 a Light density cord blood or bone marrow cells S T E M S E P Lin Lin F A C S CD3438 CD3438 CD34 b c d CD % CD % CD % CD34 CD34 CD34 Figure 1 Purification of Lin CD34 CD38, Lin CD34 CD38, and Lin CD34 cell subsets. (a) A schematic diagram depicting the strategy used to obtain highly purified cells. Light density mononuclear cells from whole CB or BM were separated into lineage negative (Lin ) or positive (Lin ) fractions by staining with monoclonal antibodies against lineage antigens, followed by purification on an immunomagnetic column (StemSep). Lin cells were stained with monoclonal antibodies specific for CD34 and CD38 and sorted on a FACStar Plus cell sorter into Lin CD34 CD38 and Lin CD34 CD38 populations. Lin cells were stained with anti-cd34 and sorted for the CD34 population. Expression of CD34 and CD38 on mononuclear cord blood cells analyzed by flow cytometry and re-analysis of Lin CD34 CD38 (b), Lin CD34 CD38 (c), and Lin CD34 (d) sorted cell fractions are shown. Purities are given as a percentage of the lymphocyte/blast window. expression may still be contaminated with tumor cells. 12,13 In order to begin to better understand the processes involved in the engraftment of purified human stem cells, we have studied the roles of cytokine treatment and the cotransplantation of non-repopulating cell fractions of CB or BM on the engraftment potential of limiting cell doses of highly purified human hematopoietic cells. We report that in order for less than 0 highly purified Lin CD34 CD38 cells from CB to engraft, the cotransplantation of accessory cells (ACs) or short-term treatment with cytokines were necessary. Facilitation of engraftment by AC in this xenotransplant model is not limited to normal stem cells, but is also required for the engraftment of purified CD34 CD38 cells from AML patients. Co-administration of ACs and treatment with cytokines had no synergistic effect on the frequency of engraftment, and cytokine treatment alone was only required during the first 10 days after transplant. This study demonstrates that other components extrinsic to stem cells are required for successful engraftment and that this model will be a useful tool to identify the mechanism by which cytokines and ACs facilitate engraftment of human repopulating cells in patients. Materials and methods Human cells Human BM cells were obtained from harvests of normal donors for allogeneic transplantation in accordance with procedures approved by the Human Experimentation Committee at the Ontario Cancer Institute (OCI), Toronto, Ontario. Samples of CB were obtained from discarded placental and umbilical tissues. BM and CB samples were diluted 1 to 3 in Iscove s modified Dulbecco s medium (IMDM; Gibco BRL, Burlington, Ontario, Canada) and enriched for mononuclear cells by centrifugation on Ficollpaque (Pharmacia, Uppsala, Sweden). Leukemic peripheral blood specimens from patients with AML were obtained after informed consent according to procedures approved by the Human Experimentation Committee. AML samples were processed as previously described. 7 Purification of cell populations As depicted in Figure 1, mononuclear cells were stained with a mixture of lineage specific antibodies (CD2, CD3, CD14, CD16, CD19, CD24, CD56, CD66b, CD41, glycophorin A) provided by the manufacturer (StemCell Technologies, Vancouver, BC, Canada), followed by addition of secondary antibody conjugated to metal colloid. Cells were then eluted through a magnetized column in order to enrich for cells not expressing lineage markers (Lin ); the cells remaining on the column (Lin ) were then washed off the demagnetized column using phosphate-buffered saline (PBS) containing 5% FCS. As described previously, cell fractions were then stained with anti-hu CD34-fluorescein isothiocyanate (FITC) and anti-hu CD38-phycoerythrin (PE), and analyzed and sorted on a FACStar Plus (Becton Dickinson, San Jose, CA, USA) equipped with 5 W argon and 30 mw helium neon lasers. 6,7 Fluorescence of FITC

3 and PE excited at 482 nm (0.38 W) and 633 nm (30 mw), respectively, as well as known forward and side scatter properties of normal live human hematopoietic cells were used to establish sorting gates. Data acquisition and analysis were performed using LYSIS II software (Becton Dickinson). Transplantation of purified cells into NOD/SCID mice Eight-week-old NOD/LtSz-scid/scid (NOD/SCID) mice used in this study were bred from breeding pairs originally obtained from L Shultz (Jackson Laboratory, Bar Harbor, ME, USA), and maintained in the defined flora animal facility located at the OCI. All animals were handled under sterile conditions and maintained in micro-isolators. Purified cell populations, at the indicated dose, were transplanted by tail vein injection into sublethally irradiated mice (375 cgy using a 137 Cs -irradiator) according to our standard protocol as previously described. 14,15 Mice transplanted with purified Lin CD34 CD38 cells were cotransplanted with non-repopulating cells (ACs) and/or received alternate day intraperitoneal (i.p.) injections of human cytokines (huscf 10 g, huil-3 and hugm-csf 6 g, from Amgen, Thousand Oaks, CA, USA). AC populations used were CD34 cells derived from the Lin fraction, CD34 CD38 cells derived from the Lin fraction, or mononuclear cells from CB and BM which had been irradiated (1500 cgy using a 137 Cs -irradiator). Mice were killed 79 weeks after transplantation, and the BM from the femurs, tibiae and iliac crests of each mouse was flushed into IMDM containing 10% FCS. Analysis of human cell engraftment in transplanted mice Genomic DNA was isolated from the BM of transplanted mice by standard extraction protocols. 14,15 EcoRI digested DNA was separated by agarose gel electrophoresis, transferred on to a positively charged nylon membrane and probed with a labeled human chromosome 17-specific satellite probe (p17h8). The level of human cell engraftment was determined by comparing the characteristic 2.7 kb band with those of human:mouse DNA mixtures as controls. Multiple exposures are performed to ensure sensitivity down to the limit of detection of 0.05% human DNA. Results Engraftment of low doses of Lin CD34 CD38 cells requires cytokine treatment and/or co-transplantation of accessory cells To isolate populations enriched for SRC, CB cells were sorted based on CD34 and CD38 antigen expression as previously reported. 6 The SRC is found exclusively in the Lin CD34 CD38 fraction, while Lin CD34 CD38 and Lin CD34 cells have no repopulating capacity. Limiting dilution analysis demonstrated that the frequency of SRC was 1 in 617 Lin CD34 CD38 cells. 6 A stringent twostep strategy involving depletion of lineage-positive cells followed by fluorescence-activated cell sorting was performed in order to sort purified populations of Lin CD34 CD38, Lin CD34 CD38, and Lin CD34 cells (Figure 1a). Re-analysis of sorted cell fractions demonstrated high levels of purity ( 98%) of the cell populations obtained in this experiment. Purity ranges for the 16 blood samples used in this study were 84% for Lin CD34 CD38 (Figure 1b), 94% for Lin CD34 CD38 (Figure 1c) and 99% for Lin CD34 (Figure 1d) populations. Using this strategy, the Lin CD34 CD38 and Lin CD34 CD38 populations did not include contaminating CD38 or CD38 cells, respectively. Using the NOD/SCID xenotransplant model, we have previously shown that the injection of cytokines was not needed to obtain engraftment of unseparated BM or CB even at limiting cell doses. 14,16 However, in five experiments without cytokine administration, the purified Lin CD34 CD38 cells from CB or BM did not engraft NOD/SCID mice (Figure 2). Mice transplanted with limited numbers of Lin CD34 CD38 cells required i.p. injection of cytokines (SCF, GM-CSF, IL-3) every other day, throughout the experimental period in order to engraft. Although the level of human cells in the murine BM was low, the phenotype of the engrafted human cells in these experiments was similar to that shown previously, and included both myeloid and lymphoid progenitors and mature cells. 6,17 In a second series of experiments, we evaluated whether co-transplantation of ACs ( Lin CD34 or Lin CD34 CD38 cells), together with 5000 Lin CD34 CD38 cells (one to two SRC) could replace cytokine treatment. A representative experiment is shown in Figure 2a. The mice were sacrificed 8 weeks post-transplant and the presence of human cell engraftment was evaluated by Southern blot. Mice injected with 500 Lin CD34 CD38 cells alone did not contain human cells (detection limit 0.05%), whereas the same number of Lin CD34 CD38 cells did engraft NOD/SCID mice when cytokines were administered and/or Lin CD34 or Lin CD34 CD38 cells were co-transplanted (Figure 2a). As reported previously, up to Lin CD34 and Lin CD34 CD38 cells injected alone did not engraft mice, confirming the absence of contaminating SRC in these two populations. 6 The results from 11 experiments with the four different treatment groups are summarized in Figure 2b. Similar to initial results, with the exception of one mouse, mice not treated with cytokines or co-transplanted with ACs were not engrafted with 5000 Lin CD34 CD38 cells. A lower percentage of engrafted mice (48%) was obtained with cytokine treatment alone vs co-transplantion with ACs (80%). However, the effects of cytokine treatment and co-transplantation ACs were not additive since the frequency of engraftment in mice that received both treatments was not different from that observed in mice that were co-transplanted with ACs alone (73% vs 80%). There was no significant difference in the level of engraftment or phenotype of human cells found in mice from the various treatment groups. 205

4 206 a GFs Accessory CD34 CD34 Human mouse CD38 mixtures (%) b Level of engraftment (% human) GFs Accessory Frequency of engraftment (%) Figure 2 Effect of cytokine treatment and accessory cell co-transplantation on the engraftment of NOD/SCID mice transplanted with limiting numbers of Lin CD34 CD38 cells. (a) Mice were transplanted with 500 Lin CD34 CD38 cells purified from a human CB sample with or without co-transplanted CD34 cells. These two groups of mice were further separated into two groups in which one group received alternate day intraperitoneal injections of human cytokines (GF) until sacrifice. DNA from the BM of mice from these four groups was digested and probed for the presence of human engraftment. (b) Summary of the level of human cell engraftment from 82 mice from 11 experiments in which mice were transplanted with 5000 Lin CD34 CD38 cells purified from BM and CB with or without co-transplantation of Lin CD34 cells ( to ) and were treated with or without cytokines until sacrifice. Treatment groups are indicated along the horizontal axis. Accessory cells are required for the engraftment of highly purified AML leukemic stem cells transplanted into NOD/SCID mice To determine whether the requirement of ACs was common to both the normal and leukemic transplantation models, we transplanted limiting numbers of purified CD34 CD38 cells from AML patients. We have previously reported that the cells in AML which are capable of engrafting SCID and NOD/SCID mice, termed SCID leukemia-initiating cells (SL-IC), were restricted to the CD34 CD38 fraction and that this engraftment required administration of cytokines. 7 As shown in Figure 3, in two different samples from patients with myelomonocytic (FAB M4) AML, both CD34 CD38 co-injection and cytokine treatment were required to obtain engraftment of 2000 or 0 CD34 CD38 cells. However, in contrast to normal SRC, low doses of transplanted CD34 CD38 cells from AML patients did not engraft mice that were treated with cytokines in the absence of co-transplanted ACs. These experiments indicate that irrespective of the source of repopulating cells, ACs are required for human cell engraftment in NOD/SCID mice and that their facilitating effect is obligatory when transplanting low doses of highly purified AML cells even in the presence of cytokines. Mature Lin CD34 or primitive Lin CD34 CD38 function as accessory cells In order to identify the population of cells which are most efficient in facilitating engraftment, Lin CD34, Lin CD34 CD38 or human irradiated mononuclear cells (MNC) from CB or BM were compared as the source of ACs. Figure 4 summarizes the frequency of engraftment in mice co-transplanted with these 4 different AC populations.

5 CD34CD38 CD34CD CD34 CD34 CD34 CD34 CD38 CD38 CD38 CD Human:mouse mixtures (%) Patient 1 Patient 2 Figure 3 Role of co-transplanted accessory cells on the engraftment of NOD/SCID mice with limiting numbers of CD34 CD38 AML cells. Mice were transplanted with purified CD34 CD38 cells from two AML patient samples at various doses with or without co-transplanted CD34 CD38 cells as ACs. Control mice were transplanted with CD34 CD38 cells. Table 1 Effect of cytokine treatment on the frequency of SRC engraftment in the absence of co-transplanted accessory cells Period of cytokine Engrafted mice Total Percentage of administration engrafted mice Day 0 to Day 0 to Day 10 to Mice were transplanted with 0 Lin CD34 CD38 CB cells on day 0 (n 5). In one group of mice, cytokine treatment began the following day and then every other day after transplant until mice were sacrificed at day 56. The other group was treated with cytokines for 10 days while the last group was treated starting at 10 days after transplant and continued to day 56. There was a trend towards co-transplanted Lin CD34 and Lin CD34 CD38 cells resulting in a higher level of engraftment than irradiated MNC from CB or BM. The frequency of engrafted mice was similar between all groups. Furthermore, irradiation of Lin CD34 and Lin CD34 CD38 cells did not affect the facilitating function of these populations (Figure 4, open circles), demonstrating that the irradiation procedure does not affect the ability of any of these populations to act as ACs. The effect of cytokines to facilitate engraftment is mediated only during the initial period following transplant We have found that AC transplanted alone cannot be detected by DNA analysis (sensitivity of detection 0.05%) in the BM or other organs of transplanted mice when evaluated 10 days after the transplant suggesting that the facilitating effect occurs shortly after the transplant (data not shown). To determine whether another facilitating compo- Level of engraftment (% human) CD34 CD34CD38 Irradiated BM Accessory cell population Irradiated CB Figure 4 Effect of different accessory cell populations on the engraftment of NOD/SCID mice transplanted with limiting numbers of Lin CD34 CD38 CB cells. Mice were transplanted with 5000 purified Lin CD34 CD38 cells from CB and co-transplanted ACs populations as indicated on the horizontal axis. The frequency and level of engraftment is shown for each group of mice. Control mice transplanted with five-fold greater doses of each AC did not engraft when transplanted under identical conditions (n 4, data not shown). nent such as cytokine treatment is also important only for the short term, we compared the frequency of engraftment when cytokines were given only for the first 10 days posttransplantation, for the entire period, or when cytokine treatment was delayed for 10 days after transplantation (Table 1). In all cases the mice were transplanted with 800 Lin CD34 CD38 cells in the absence of ACs. The percentage of engrafted mice treated with cytokines for the

6 208 first 10 days was similar to mice treated with cytokines for the entire experimental period (810 weeks), 54% and 48%, respectively (Table 1). However, mice which did not receive cytokines for the first 10 days post-transplantation did not engraft even though cytokines were administered for the remainder of the experimental period (Table 1). These data clearly demonstrate that cytokine treatment is only required for the first 10 days after transplantation of low numbers of purified human repopulating cells. Discussion The data reported here demonstrate that short-term cytokine treatment or co-transplantation of AC are required for engraftment of highly purified SRC. Similar treatments are also required for engraftment of SL-IC when highly purified AML cells are transplanted, suggesting that the mechanisms governing engraftment are not restricted to normal stem cells. Several different cell types can act as AC including Lin CD34 cells, Lin CD34 CD38 cells, and irradiated mononuclear BM or CB cells. While the precise mechanism of action of ACs is not known, the fact that cotransplantation of ACs can replace cytokines suggests that cytokine secretion from AC may be important. The fact that cytokines were required only during the first 10 days after transplant, concordant with the short-term presence of AC post-transplant supports the idea that cytokine treatment and AC co-transplantation may act via similar mechanisms. Furthermore, during hematopoietic recovery, mononuclear cells secrete more cytokines than those harvested during steady-state hematopoiesis and recent studies have demonstrated that cytokine levels are elevated following transplant, and coincide with hematological recovery. 18 Since mononuclear cells are known to be a source of cytokines during transplantation, it seems reasonable to ascribe a complementary role of ACs to the engraftment process by releasing cytokines in vivo. Precisely how the cytokines and co-transplantation of ACs impact on SRC engraftment is unknown but could be both direct and indirect. Since the effect of both treatments was only short term, it appears that the action must target the SRC during the early phases of the transplant. We speculate that the cytokines could be inducing adhesion molecules on the SRC enabling efficient homing to the murine microenvironment. 19 Future studies should include detailed kinetic experiments where cytokines are given for different times and doses starting from before the transplant. Similarly the AC could be acting to prevent nonspecific homing. Only secondary transplantation of mice transplanted with the BM of donor mice which were recently transplanted with human stem cells would allow one to address whether cytokines or ACs affect homing of SRC. These experiments however are limited by our current difficulty in performing serial transplantation of human cells in NOD/SCID mice. Another mechanism by which cytokines and ACs may act is in promoting the survival of the SRC during early stages of the transplant. In this scenario, cytokine or AC function would be required only until enough human cells were produced that could produce selfsustaining amounts of the required cytokines. This hypothesis could explain why ACs are not required when high doses of SRC are transplanted. We have shown previously that continuous cytokine treatment was required to generate high level human cell engraftment when human BM was transplanted into SCID mice. 15 In these studies, cytokines seemed to be stimulating the engrafting cells to proliferate and differentiate. However, it is difficult to compare the relationship between this previous work to the present study since these earlier studies involved transplantation of large cell doses which included a large number of Lin CD34 CD38 and Lin CD34 cells, cell types that we show here have facilitating function. Using the NOD/SCID recipients, high levels of engraftment could be achieved without cytokine treatment probably because these mice are more immunedeficient than SCID mice and their BM stroma environment is more permissive to human cell engraftment. 5,20,21 Since cytokines are required for less than 10 days post-transplant in NOD/SCID mice when highly purified cells are injected, we favor the idea that cytokine treatment promotes the survival and/or homing of the SRC in the early phase of the engraftment process. The fact that cytokine treatment is less effective than co-administration of ACs may be due to the presence of less than optimal conditions (concentration or cytokine combination) required to stimulate highly purified human stem cells. In addition, it is possible that cellcell contact between Lin CD34 CD38 cells and ACs alone or in combination with mouse stroma is crucial for facilitating engraftment. ACs may have an advantage over cytokine treatment in that they are able to bypass these limitations. Our studies indicate that both Lin CD34 and Lin CD34 CD38 cells have a comparable capacity to act as ACs. Clinical trials have found that T cell-depleted transplants were associated with a significant increase in the failure of engraftment. 22 These clinical data, as well as rodent models, suggested that T cells are required for engraftment of the BM stem cell in an allogeneic environment. 15,23,24 Transplantation of highly enriched or purified stem cell populations alone have been shown to suffer from the same limitations as the T cell depletion, ie failure of BM engraftment. This failure has been traditionally attributed to rejection by the host microenvironment. However, an alternative hypothesis to these observations is that a nonstem cell component in the donor BM may be required to facilitate stem cell engraftment in allogeneic recipients. Indeed, Ildstad s group 10 has reported the characterization of a bone marrow-derived cell that facilitates engraftment of allogeneic BM stem cells. This facilitating cell fraction was found to reside in the characteristic lymphoid gate and phenotypically defined to be CD8 CD3 TCR CD45R class II dim. Like Lin CD34 CD38 or Lin CD34 cell fractions, as few as of these cells were sufficient to facilitate engraftment. However, the fact that Lin CD34 CD38 are derived from a population of cells depleted of lineage markers (including T cell markers) seems to indicate that the facilitating cell characterized by Ildstad s group differs from the AC populations characterized in our study. Recently, a parallel study from Verstegen et al 25 has demonstrated that SCID mice depleted of macrophages and transplanted with relatively high doses of CD34 CD38 cells produced 10-fold fewer CD34 cells

7 in vivo compared to mice transplanted with equivalent numbers of unfractioned or purified CD34 cells suggesting that CD34 CD38 cells promote the expansion of the engrafting stem cells. 25 The co-transplantation of nonrepopulating CD34 CD38 cells with repopulating CD34 CD38 cells was capable of restoring the graft size. Although the readout was different from our study (ie production of CD34 cells and not frequency of engrafted mice as we measured), these data support the idea that SRC, enriched in the CD34 CD38 subset, require ACs that are found in the CD34 CD38 population. In conclusion, the results described here demonstrate the usefulness of the SRC assay to address issues regarding the complex process of human hematopoietic repopulation, and to elucidate the mechanism of action of accessory cells and/or cytokine treatment for the homing or survival of human stem cells. Acknowledgements We thank I McNiece at Amgen for cytokines, L McWhirter for providing cord blood specimens, M Minden for providing AML samples and members of the laboratory for critically reviewing the manuscript. This work was supported by grants to JED from the Medical Research Council of Canada (MRC), the National Cancer Institute of Canada (NCIC) with funds from the Canadian Cancer Society, the Canadian Genetic Diseases Network of the National Centers of Excellence, an MRC Scientist award (JED), postdoctoral fellowships from the NCIC (MB), the Leukemia Research Fund of Canada and the MRC (JCYW), the Deutsche Krebshilfe (UK), and the Human Frontier Science Organization Program and the French Cancer Research Association (DB). References 1 Ogawa M. Differentiation and proliferation of hematopoietic stem cells. Blood 1993; 81: Phillips R. Hematopoietic stem cells: concepts, assays, and controversies. Semin Immunol 1991; 3: Dick JE. Normal and leukemic human stem cells assayed in SCID mice. Semin Immunol 1996; 8: Larochelle A, Vormoor J, Hanenberg H et al. Identification of primitive human hematopoietic cells capable of repopulating NOD/SCID mouse bone marrow: implications for gene therapy. Nature Med 1996; 2: Cashman JD, Lapidot T, Wang JC et al. Kinetic evidence of the regeneration of multilineage hematopoiesis from primitive cells in normal human bone marrow transplanted into immunodeficient mice. Blood 1997; 89: Bhatia M, Wang JCY, Kapp U et al. Purification of primitive human hematopoietic cells capable of repopulating immunedeficient mice. Proc Natl Acad Sci USA 1997; 94: Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nature Med 1997; 3: Lapidot T, Sirard C, Vormoor J et al. A cell initiating human acute myeloid leukaemia after transplantation into SCID mice. Nature 1994; 367: Sykes M, Sheard M, Sachs DH. Effects of T cell depletion in radiation bone marrow chimeras. I. Evidence for a donor cell population which increases allogeneic chimerism but which lacks the potential to produce GVHD. J Immunol 1988; 141: Kaufman CL, Colson YL, Wren SM et al. Phenotypic characterization of a novel bone marrow-derived cell that facilitates engraftment of allogeneic bone marrow stem cells. Blood 1994; 84: Morrison SJ, Uchida N, Weissman IL. The biology of hematopoietic stem cells. Ann Rev Cell Dev Biol 1995; 11: Berenson RJ, Bensinger WI, Hill RS et al. Engraftment after infusion of CD34 marrow cells in patients with breast cancer or neuroblastoma. Blood 1991; 77: Gazitt Y, Reading CC, Hoffman R et al. Purified CD34 Lin - Thy stem cells do not contain clonal myeloma cells. Blood 1995; 86: Vormoor J, Lapidot T, Pflumio F et al. Immature human cord blood progenitors engraft and proliferate to high levels in immune-deficient SCID mice. Blood 1994; 83: Lapidot T, Pflumio F, Doedens M et al. Cytokine stimulation of multilineage hematopoiesis from immature human cells engrafted in scid mice. Science 1992; 255: Wang JC, Doedens M, Dick JE. Primitive human hematopoietic cells are enriched in cord blood compared with adult bone marrow or mobilized peripheral blood as measured by the quantitative in vivo SCID-repopulating cell assay. Blood 1997; 89: Bhatia M, Bonnet D, Kapp U et al. Quantitative analysis reveals expansion of human hematopoietic repopulating cells after short-term ex vivo culture. J Exp Med 1997; 186: Takamatsu Y, Akashi K, Harada M et al. Cytokine production by peripheral blood monocytes and T cells during hematopoietic recovery after intensive chemotherapy. Br J Haematol 1993; 83: Hardy CL. The homing of hematopoietic stem cells to the bone marrow. Am J Med Sci 1995; 309: Shultz L, Schweitzer P, Christianson S et al. Multiple defects in innate and adaptive immunological function in NOD/LtSzscid mice. J Immunol 1995; 154: Gan OI, Murdoch B, Larochelle A, Dick JE. Differential maintenance of primitive human SCID-repopulating cells, clonogenic progenitors, and long-term culture-initiating cells after incubation on human bone marrow stromal cells. Blood 1997; 90: O Reilly RJ. Allogenic bone marrow transplantation: current status and future directions. Blood 1983; 62: Vallera DA, Soderling CC, Kersey JH. Bone marrow transplantation across major histocompatibility barriers in mice. III. Treatment of donor grafts with monoclonal antibodies directed against Lyt determinants. J Immunol 1982; 128: Vallera DA, Soderling CC, Carlson GJ, Kersey JH. Bone marrow transplantation across major histocompatibility barriers in mice. Effect of elimination of T cells from donor grafts by treatment with monoclonal Thy-1.2 plus complement or antibody alone. Transplantation 1981; 31: Verstegen MM, van Hennik PB, Terpstra W et al. Transplantation of human umbilical cord blood cells in macrophagedepleted SCID mice: evidence for accessory cell involvement in expansion of immature CD34 CD38 cells. Blood 1998; 91: