CD34' cell fractions, cells in the small Fr 25/29 CD34+ cell population possessed .... leukocyte conditioned medium and erythropoietin (Stem Cell Tech- nologies .... 90 + 7. 1 2 1. 13 + 7. 3 2 2. 33 2. 10 39. 2 10. Fr R 0 CD34+ 0.4. + 0.2. 77 + 12.
Isolation of Small, Primitive Human Hematopoietic Stem Cells: Distribution of Cell Surface Cytokine Receptors and Growth in SCID-Hu Mice By John E. Wagner, Daniel Collins, Shawn Fuller, Lisa R. Schain, Amy E. Berson, Camillo Almici, Melinda A. Hall, Karen E. Chen, Thomas B. Okarma, and Jane S. Lebkowski Human CD34' cells were subfractionated into three size classes using counterflowcentrifugal elutriation followedby immunoadsorption t o polystyrene cell separation devices. The three CD34+ cell fractions (Fr), Fr 25/29, Fr 33/37, and Fr RO, had mean sizes of 8.5,9.3 and 13.5 pm, respectively. The majority of cells in the large Fr R 0 CD34+ cell population expressed the committedstage antigens CD33, CD19, CD38, or HLA-DR and contained the majority granulocyte-macroof phage colony-forming units (CFU-GM), burst-forming unitserythroid (BFU-E), and CFU-mixed lineage (GEMM). In contrast, the smallFr 25/29 CD34+cells were devoid of committed cell surface antigens and lacked colony-forming activity. When seeded t o allogeneic stroma, Fr R 0 CD34' cells produced few CFU-GM at week 5, whereas cells from theFr 25/29 CD34+ cell population showeda 30- t o 55-fold expansion ofmyeloid progenitors at this same time point. Furthermore, CD34' cells from each size fraction supported ontogeny of T cells in humanthymus/liver grafts in severe
combinedimmunodeficient W I D ) mice. Upon cell cycle analyses, greater than 97% of the Fr 25/29 CD34' cells were in Go/GI phase, whereas greater proportions of the two larger CD34+ cell fractions were in active cell cycle. Binding of the cytokines interleukin (lL)-la, IL-3, IL-6, stem cell factor (SCF), macrophage inhibitory protein (MIP)-la, granulocyte colony-stimulating factor (G-CSF), and granulocyte-macrophage (GM)-CSF t o these CD34' cell populations was also analyzed by flow cytometry. As compared with the larger CD34' cell fractions, cells in the small Fr 25/29 CD34+ cell population possessed the highest numbers of receptorsfor SCF, MIPIa, and IL-la. Collectively, these results indicate that the Fr 25/29 CD34' cell is a very primitive, quiescent progenitor cell population possessing a high number of receptors for SCF and MIPla and capable of yielding both myeloid and lymphoid lineages when placed in appropriate in vitro or in vivo cultureconditions. 0 1995 by The American Societyof Hematology.
T
purified CD34+ cells." Other cell surface antigens have been used to subfractionate the CD34' population into immature and committed subsets. Specifically, antibodies against the transferrin re~eptor,'~"'CD33,14,15CD38,16 HLA-DR,I2." CD45 7B9,20 c-kit,2',22and combinations of lineage-specific antigens23s24 have been used to separate immature cells from committed progenitors. More recently, positive expression of both the Thy1 and CD34 antigens has beenusedto isolate very primitive human hematopoietic p r e ~ u r s o r s .Approximately ~~,~~ 25% of human bone marrow CD34' cells express Thy 1, and these cells contain the vast majority of long-term bone marrow culture-initiating activity.26 Counterflow centrifugal elutriation (CCE) has been used to separate murine pluripotent hematopoietic precursors from committed progenitor^.^"" Jones et al" used CCE to subfractionate murine bone marrow into four discrete size and density populations. The largest fraction, highly enriched for granulocyte-macrophage colony-forming units (CFUGM), produced only short-term transient engraftment of lethally irradiated mice. The smallest fraction, which was depleted of CFU-GM and day- 12 spleen colony-forming units (CFU-S), was necessary to achieve long-term reconstitution, suggesting the physical separation of committed and reconstitutive stem cells. In further studies, Yoder et a132used CCE to separate six different subsets of murine high proliferative potential colony-forming cells (HPP-CFC). The lowest density HPP-CFC with lymphocytic morphology appeared the most primitive, requiring numerous growth factors in vitro for maximal proliferation. Using antibody selection after CCE, Orlic et al" isolated Lin-, c-kit' cells from each of the elutriated fractions. Each Lin-, c-kit+ fraction possessed pluripotent activity, yet only the small cell fraction lacked committed progenitor activity. In this study, we used CCE to separate CD34' cells from humanbonemarrow into committed andmore primitive populations. We demonstrate that the smallest subset, less
HE HUMAN PLURIPOTENT hematopoietic stem cell has been postulated to respond to a variety of internal and environmental cues, giving rise to both identical copies of itself and to differentiated cells of the hematopoietic system. Several studies have focused on the isolation and characterization of these rare cells. In the mouse, pluripotent hematopoietic stem cells are Thyl'",Sca+,Lin-, and as few as 100 isolated cells were able to reconstitute both the myeloid and lymphoid compartments of lethally irradiated recipients.'.' Further separation of the Thyll",Sca+,Lin- population using the mitochondrial membrane dye rhodamine 123 subfractionates the stem cells into resting and activated subsets that differ in their ability to proliferate and reconstitute lethally irradiated primary and secondary recipient^.^.^ In humans, the CD34 antigen is expressed by both committed andprimitive hematopoietic stem cells,6-' andpurified CD34+ cell populations engraft lethally irradiated autologous and allogeneic Moreover, hematopoiesis has been restored in humans after high-dose chemotherapy using From the Department of Pediatrics, Universityof Minnesota, Minneapolis, MN; Applied Immune Sciences, Santa Clara,CA; R&D Systems, Minneapolis, MN; and Cattedra Ematologia, Parma, Italy. Submitted November 10, 1994; accepted February 28, 1995. Supported in partbygrantsfrom CNR (Progetto Finalizzato ACRO), Associazione Italiana per la Ricerca sul Cancro (AIRC), American Cancer Society Grant No. 92-277, and Applied Immune Sciences, Inc. J . E. W. is the recipient of the American Cancer Society Career Development Award. Address reprint requests to Jane S. Lebkowski, PhD,Applied Immune Sciences, 5301 Patrick Henry Dr, Santa Clara, CA 950541114.
The publication costsof this article were defrayedin part by page chargepayment. This article must therefore be hereby marked "advertisement" in accordance with I8 U.S.C. section 1734 solely to indicate this fact. 0 1995 by The American Society of Hematology. 0006-4971/95/8602-0020$3.00/0 512
Blood, VOI 86, NO 2 (July 15), 1995: pp 512-523
ISOLATION OF HUMAN HEMATOPOIETIC STEM CELLS
than 9.5 pm in diameter, is substantially devoid of lineage markers and colony-forming cells and is in the GdG, phase of the cell cycle. However, after incubation with allogeneic stroma, the cells proliferate and produce CFU-GM for at least 5 weeks. Furthermore, cells from the small fraction also repopulate fetal thymus liver grafts; producing both CD4+ and CD8+ T cells. Lastly, using biotinylated cytokines, we characterize the distribution of receptors for a variety of different hematopoietic growth factors on all size fractions of CD34+ cells. The small CD34+ fraction expresses receptors for interleukin-l (L-l),IL-3, L-6, granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), stem cell factor (SCF), and macrophage inhibitory protein (MP)-la. Compared with the larger CD34+ subsets, the density of receptors for IL- la, SCF, and MIP-la on these small CD34+ cells is high, potentially priming them for response to these positive and negative regulators of hematopoiesis. MATERIALS AND METHODS
Human bone marrow samples. Human bone marrow samples were aspirated from the posterior iliac crest of healthy volunteers using standard procedures. Informed consent was obtained following the guidelines approved by the Committee on the Use of Human Subjects at the University of Minnesota (Minneapolis, MN). Appropriate institutional approvals were obtained. CCE. Low-density bone marrow mononuclear cells were isolated using Histopaque-l077 (Sigma Chemical CO, St Louis, MO) and density gradient centrifugation at 400g for 30 minutes at room temperature. The mononuclear cells were washed twice and resuspended in 10 mL sterile elutriation medium (Hank‘s buffered saline containing 0.5% human serum albumin [HSA] and 0.3 mmol/L EDTA, pH 7.2). The cells were injected via a sampling site coupler into the inlet stream of a Beckman J6M/E centrifuge (Spenco Division, Beckman Instruments, Palo Alto, CA) equipped with a JE-5.0 rotor and standard chamber. One peristaltic pump (Masterflex; Cole Palmer Instruments, Chicago, K) provided continuous flow of elutriation media. Cells were delivered at a total flow rate of 15 mL/min (Fr 15), at a rotor speed of 3,000 rpm and a temperature of 25°C. After 100 mL of eluate was collected, the flow rate was increased to 25 mL/min (Fr 25). With the rotor speed held constant, the flow rates were sequentially increased to29 mL/min (Fr 29), 33 min (Fr 33), and 37 mL/min (Fr 37), collecting 200 mL with each increment. The cells that remained in the chamber were captured by turning the rotor off (Fr RO) and flushing the chamber with 100 mL of elutriation media. The cell fractions were washed and centrifuged at 300g for 10 minutes. Fractions 25 and 29 (Fr 25/29) and fractions 33 and 37 (Fr 33/37) were pooled for these experiments. Viability was determined by trypan blue exclusion, and cell numbers were determined by either hemacytometer counts or by a ZBT Coulter Counter (Coulter Electronics, Hialeah, FL). Isolation of CD34+ cells. Cells obtained after the elutriation procedure were collected by centrifugation, washed with elutriation medium, and resuspended in Dulbecco’s phosphate-buffered saline Ca/Mg-free containing 1 mmol/L EDTA (DPBSE), and 0.5% Gamimune (Cutter Biological, Miles, Inc, Berkeley, CA) for CD34 isolation using the method of Lebkowski et al.33The cells were incubated for 20 minutes at room temperature and loaded into AIS MicroCELLectors SBA (Applied Immune Sciences, Inc, Santa Clara, CA), which contain covalently immobilized soybean agglutinin (SBA). Up to 2 X lo7 cells were loaded per flask and incubated for 1 hour at room temperature. The nonadherent (SBA-) cells were removed and subsequently loaded into AIS MicroCELLectors CD34. Up to
513
2 x lo7 cells were loaded onto these devices, which contain the covalently immobilized anti-CD34 monoclonal antibody, ICH3.’ After incubation for l hour at room temperature, the nonadherent (CD343 cells were removed, and the adherent (CD347 cells were collected after physical agitation in DPBSE containing 10% fetal bovine serum (FBS) or 0.5% HSA. The released cells were withdrawn, the devices were rinsed with DPBSE, and the cells were collected after centrifugation in tubes that had been precoated with 5% HSA. Light microscopy of CD34’ cells. Cytospin slides were separately prepared for each of the three elutriated CD34+ cell fractions. The cells were stained with Wright’s stain and visualized by light microscopy. Photographs were taken at 450X magnification. Light scatter and cell size determination. The various CD34+ cells subsets were fixed with 1% paraformaldehyde, run on an Epics Profile I1 (Coulter Electronics) or FACScan (Becton Dickinson, Milpitas, CA) and analyzed for forward and 90” (side) scatter. A standard size curve was constructed by analyzing defined diameter beads (Molecular Probes, Eugene, OR) and measuring forward light scatter channel signals. Keeping all instrument settings constant, the mean channel forward scatter of the various CD34+ cell fractions was obtained and plotted on the standard curve to obtain a mean cell diameter for the population. Immunophenotypic analysis. Cells for immunophenotyping were preincubated in0.5% Gamimmune and washed with Dulbecco’s phosphate-buffered saline (DPBS) containing 0.1%sodiumazide. The cells were dispensed for immunophenotyping in DPBS containing 1% FBS. Fluorescein- and phycoerythrin-conjugated CD3, CD19, CD33, CD38, HLA-DR, and CD34 (HPCA-2) antibodies (Becton Dickinson, Milpitas, CA) were added to the cells and incubated for 30 minutes at 4°C. The cells were washed twice with DPBS-azide and finally resuspended in 1% paraformaldehyde. Cells were analyzed on a FACStar Plus or FACScan (Becton Dickinson) and analyzed using vendorsupplied software. Positive staining was defined as fluorescence beyond that of cells stained with irrelevant isotype controls. Methylcellulose cultures for progenitors. Cells were cultured in 0.9% methylcellulose containing phytohemagglutinin-stimulated leukocyte conditioned medium and erythropoietin (Stem Cell Technologies, Vancouver, BC). Cultures were seeded in triplicate, 1 mL culture in 35 mm dishes, and maintained in a humidifiedatmosphere at 37°C in 5% CO,. The plates were scored at 14 to 18 days using an inverted microscope for CFU-GM, burst-forming unit-erythroid (BFU-E), and mixed lineage (CFU-GEMM) colonies. Long-term bone marrow culture. Normal allogeneic stromal layers were established from bone marrow mononuclear cells in T-25 flasks (Corning Gear Works, Corning, NY) in long-termbone marrow culture medium, McCoy’s5A medium supplemented with12.5% FBS (HyClone Labs, Logan, UT), 12.5% horse serum (HyClone Labs), 2 mmoVL L-glutamine, 0.7% sodium bicarbonate, 1% (vol/vol) minimum essential medium (MEM) nonessential amino acids (Life Technologies,GrandIsland, NY), 1%(voUvol) MEM vitaminsolution (Life Technologies), 1% (vol/vol) sodium pyruvate (Life Technologies), 100 U/mL penicillin, 100 &mL streptomycin, and 10” mol/ L hydrocortisone (A-Hydrocort; Abbott Laboratories, Chicago, K). Stromal layers were cultured 21 to 28 days, d a t e d at 1,500 cGy, and used to support long-term bone marrowculture using a modification of the method of Gartner and Kaplan.” Immediately after isolation,1 X lo4 CD34+ cells from each of the size fractions were seeded onto the stromal layers to initiate the cultures. Cultures were maintained in a humidifiedatmosphere at 37°C in 5% CO,. At weeks 2 and 5 of culture, the nonadherent layer was removed andthe adherent layer wascollected after trypsinization. Cells from the nonadherent and adherent layers were enumerated, and 1 X lo4 cells from each fraction were culturedin methylcellulose progenitor culture (see above). Colonies were quantitated 14 to 18 days after culture.
514
WAGNER ET AL
Cell cycle analysis. For cell cycle analysis, 0.5 X 10' to 1.0 x IO6 CD34' cells and its subsets were mixed and incubated in 300 to 500 mL of DNA fragmentation buffer (0.5 m o m EDTA, Triton X, 1 mg/mL propidium iodide, 1% RNase A inDPBS) for 10 minutes at 4°C in the dark. Chick erythrocytes were prepared as a control. DNA content as measured by fluorescence intensity was determined using a FACScan flow cytometer (Becton Dickinson) and analyzed using the manufacturer's program Cell Fit. Characterization of growth factor receptor distribution. CD34' cell subsets were assayed for the presence of receptors for SCF (ckit ligand), GM-CSF, G-CSF, IL-l, L-3, E-6, and MIP-la. For such analyses, 1 X 10' cells were washed twice in HEPES buffered saline (150 mmoVL NaCI, 25mmoVL HEPES pH 7.4)and incubated with 100 ng ofbiotinylated growth factor (R&D Systems, Minneapolis, MN) for 60 minutes on ice. After incubation, the cells were washed twice in 2 mL of RDFI wash buffer (R&D Systems) and subsequently incubated with 100 ng avidin-carboxyfluroescein (R&D Systems) for 30 minutes on ice in the dark. The cells were again washed with 2 mL of RDFI buffer and resuspended in 300 pL of HEPES buffered saline. Stained cells were analyzed for fluorescence intensity on either an EPICS Profile I1 or Epics 752 flow cytometer (Coulter Electronics). The percent fluorescence positivity and mean channel fluorescence were measured. Positive samples were stained with avidin-carboxyfluorescence alone. Specificity of binding of the biotinylated cytokine was confirmed by inhibition of binding in the presence of a 100-molar excess of unlabeled growth factor. Establishment of CD34+ cell subsets inthymush'iver grajis in SCID-Hu mice. SCID-Hu micewere established using modified proCB cedures ofNamikawa et al,35 McCune et al?6 andRowley et 17 SCID mice (Taconic, Germantown, NY) were received at8 weeks of ageandhousedin standard microisolator cages. Allprocedures were approved by the Applied Immune Sciences Institutional Animal Care and Use Committee. Five days before surgery, theanimals were given Ciprofloxacin(Cipro; Miles, Inc) in theirdrinking water, which wascontinueduntil 10 days postsurgery. Human fetal thymusand liver from 18- to 22-week gestationfetuses were obtained after curettages after informed consent. Pieces of both fetal thymus and liver were placed in direct contact under the kidney capsules of the SCID mice. Pentobarbital anesthesia was used for all surgical procedures. At 10 weekspostimplantation ofthe thymudliver graft, asecond surgical procedure was performed to seed 10,OOO to 50,000 of the CD34' cell subsets. For this procedure, the CD34' cells were directly injected into the thymusfliver graft in a volume of 1 to 2 pL. For these analyses, the thymudiver graft and cells from the donor of the CD34' cells were HLA-typedfor various class I markers.The CD34' cells were seeded into thymudiver grafts with at least one disparate class I marker, so that the seeded cells could be tracked after implantation. The following antibodies wereused for HLAtyping:HLAA2,clone BB7.2; HLA-A3, clone GAPA3; HLA-A11/A24,clone A11.1M; HLA-BS, clone 4D12; HLA-B7, clone BB7.2 (American Type Culture Collection, Rockville, MD). Fluorescentanalysis of stained cells wasperformed as described above. Tenweeks after CD34' cell seeding, the animals were killed, and the graft tissue was removed. The tissue was dispersed, and the mononuclear cells were isolated using Histopaque 1077 (Sigma, St Louis, MO). The origin of the cells in the graft was determined by two-color fluorescence analysis using monoclonal antibodies specific for the human CD45 antigen (clone 2D1;Becton Dickinson) andtheappropriateHLA class I marker. The lineage of the cells was further determined using appropriate major histocompatibility complex (MHC) classI and lineage-specific monoclonal antibodies. RESULTS
For these studies, CCE was used to subfractionate adult bone marrow CD34' cells. To start the process, bone marrow
mononuclear cells (BMMC) were separated by CCE into three fractions, Fr 25/29, Fr 33/37, and Fr RO, based on the flow rate of the elutriation buffer and rotor speed. Fr 25/29, Fr 33/37, and Fr R 0 contained, on average, 33.9%, 25.6%, and 31.9% of the starting BMMC population, respectively (Table I). CD34' cells were separately isolated from each of the three elutriation fractions. Of the CD34' cells that were isolated by these procedures, IO%, 35%, and 60% were recovered in the Fr 25/29, Fr 33/37, and Fr R 0 fractions, respectively. A mean 0.06%, 0.2%, and 0.4% of the BMMC were found in the final CD34+ cell populations from Fr 25/ 29, Fr 33/37, and Fr R 0 samples, respectively (Table 1). CD34+ cells from each of the elutriation fractions were visualized by light microscopy (Fig l). Cells in Fr 25/29, Fr 33/37, and Fr R 0 CD34' were progressively larger in size with distinct morphologies. The Fr 25/29 CD34' cells were lymphocyte-like with relatively scant cytoplasm, large homogeneous nuclei, and rare nucleoli. The Fr 33/37 CD34' cells were lymphoblast-like with more cytoplasm, large nuclei, and one to three nucleoli. CD34+ cells of the Fr R 0 were myeloblast-like with large amounts of cytoplasm-containing granules, large granular nuclei, and multiple nucleoli. The relative sizes of the three CD34' cell fractions were confirmed by light scatter analysis. Figure 2 shows the forward versus side scatter of the input BMMC and the three CD34+ cell fractions. Allof the cell samples showed relatively low side scatter, indicating low levels of granularity in all of the populations. The Fr 25/29 (Fig 2B), Fr 33/37 (Fig 2C), and Fr R 0 (Fig 2D) CD34' cell fractions show progressively increasing forward scatter, confirming their increasing cell size. The mean forward scatter was determined for each CD34+ cell fraction and plotted on a standard forward scatter curve generated using beads of known size. Using such analysis, the average sizes of the Fr 25/29, Fr 33/37, and Fr R 0 CD34' cells were 8.5 ? 1.5, 9.3 t 1.4, and 13.5 ? 2.0 pm, respectively. The BMMC-elutriated fractions and the purified CD34' cells were immunophenotyped for a variety of antigens (Table l). The inputBMMC fraction had,on average, 3.3% CD34' cells. The percentage of CD34' cells remained relatively constant in the elutriated fraction, with 2.0%, 7.6%, and 5.4% of the Fr 25/29, Fr 33/37, and Fr R 0 cells expressing the CD34+ antigen, respectively. The final CD34' cell fractions were substantially enriched in CD34' cells with, on average, 50.7%, 90.1%, and 76.8% purities from the Fr 25/29, Fr 33/37, and Fr R 0 populations, respectively. The lower purities in the Fr 25/29 CD34' cells most likely are a function of their lower frequency in the starting material. The different purified CD34' cell fractions varied widely in their expression of a variety of lineage-specific antigens. The majority of the large Fr R 0 CD34' cells expressed differentiation markers, with 62%, 83%, and80% of the cells possessing the CD33, CD38, and HLA-DR antigens, respectively. Zero to 4% of the Fr R 0 CD34' cells expressed the lymphoid markers CD3 or CD19. The Fr 33/37 CD34' cells expressed intermediate levels of these differentiation antigens, with13% being CD19+, and 3% to 39%having the CD33, CD38, or HLA-DR surface markers. In contrast, CD34' cells from the Fr 25/29 population were essentially
ISOLATION OF HUMAN HEMATOPOIETICSTEM CELLS
515
Table l.Cel Numbers and Surface Antigen Expressionof Fractionated Bone Marrow Cells and CD34+ Cells % Cells With Designated Surface Antigen BMMC in
Cells Unseparated 33.9 Fr 25/29 25.6Fr 33/37 Fr R 0 0.06 CD34+ Fr 25/29 0.2 Fr CD34+ 33/37 Fr R 00.4 CD34+
Fraction (%l t 11.5 4.4 31.9 t 9.2 f 0.04 0.1 0.2
+
+
+
CD34
9 + 13 + 1 2 2 2 221 8 f 3 4 24 1+5 12 2 23 51 25 90 7 77 12
+ +
+
CD19
CD3
28 f 77 2 5 37 13 15 6
+ +
+
13 121 24 2+ 11
723 321 1 f8O2 1
Of0 13
+7
CD33
Of0 21 2 0 2 32 62 83
+
4 0 2 11
CD38
2 r 14 + 1 17 3 39 2 14 9+2 33 392 10 80 2
+
+
HLA-DR
5+2 15 2 3 18 f 4 2 10 23
BMMC, the elutriated fractions(Fr 25/29, Fr 33/37, and FR RO), and their various CD34+ cell fractions (Fr 25/29 CD34+, Fr 33/37 CD34+, FR R 0 CD349 were isolated as described in Materials and Methods. The number of cells in each fraction were enumerated, and the proportion of BMMC (%) in each population was calculated. The immunophenotyping data were collected using the procedures detailed in Materials and Methods. All data were gathered using gating procedures in which all livecells were analyzed. The results represent the mean f SD from seven separate experiments.
devoid of lineage markers. Less than 23% of the Fr 25/29 CD34+cells expressed any of the tested antigens, suggesting that these purifiedCD34+ cells represented a less committed subfraction. To test the short-term hematopoietic function of the various cell fractions,CFU-GM, BFU-E, and CFU-GEMM progenitor colonies were enumerated. The results of seven such experiments are shown in Table 2. When the unseparated
BMMC were fractionatedby elutriation, thebulk of the progenitor activity tracked into the larger RFr0 and intermediate Fr 33/31 populations. In contrast, very few progenitors were found in the small cell Fr 25/29 population. Similar results were observed whenCD34+ cells were isolated from theelutriatedfractions. All three CD34+ cellfractions showed substantial enrichment of colony activity over their elutriated unpurified counterparts, with 10- to 60-fold enrich-
WAGNER ET AL
516
C Fr 33/37 CD34+
, .... . . .
0
t
...
11:1I
I l~l..l I I l
...
l-BOO
.
:i 0
.
0
. ,..
'. .>. . . . .... .. . . . . .. . . ... &..&... . . .
.:.. .
:;.i.;
..:::.
.
. .
.
...
'.
.. . ..
.. . . ....
. . . .. . .. .
..
...
: :
,. .
..
I
I ,
I
,
600
. . . . .... .. ... .. . .. ..
.
.
.
800
10100
Fr R0 CD34+
.
:.
...
.'
.
. ..
.
.
0
.
:
1
..
.... .. . .
400
200
!:
.
I I I
0
q
1
I
.
.
0 N
0
0
T
1 1 1 ~ 1 1 1 1 ~ 1 1 1 1 ~ 1 1 1 1 ~ 1 1 1 1 200
400
600
11DO0
800
ssc +
Fig 2. Forward andside scatter of theCD34+ cell subsets. Forward andside scatter were analyzed as described in Materials and Methods using theCD34' cells isolated from the three elutriated fractions. Forward scatter (FSC) is plotted on the Y axis; side scatter (SSC) is plotted on theX axis. (A) The unseparated BMMC fraction; (B) the Fr 25/29 CD34' cell fraction; (C) the Fr 33/37 CD34' cell fraction; and (D) the Fr R 0 CD34' cell fraction.
ment of progenitor activity after CD34' cell isolation. However, compared with the other two fractions, the Fr 25/29 CD34' cells were relatively devoid of progenitor activity, consistent with their immature phenotypic profile. The CD34' cell populations were tested for their ability to initiate
long-term bonemarrow cultures. In these experiments, 10,000 cells from the Fr 25/29, Fr 33/37, or Fr R 0 CD34' fractions were seeded in duplicate onto established irradiated allogeneic stromal layers. At 2 and 5 weeks of culture, the nonadherent and adherent layers were harvested, counted,
ISOLATION
OF
HUMAN HEMATOPOIETICSTEMCELLS
517
Table 2. Frequency and Enrichment of Hematopoietic Progenitors No. Coloniesi5 x 10' Cells
CFU-GM
Fraction
105.2 1.6 120.0 275.4 62.1 1,216.4 3,072.9
Unseparated Fr 25/29 Fr 33/37 Fr R 0 Fr 25/29 CD34* Fr 33/37 CQ34' Fr R 0 CD34'
BFU-E
2 53.7 75.9
2 1.7
t 100.7 i: 111.5 2 47.3 2 446.3 2 968.1
3.1 146.7 163.4 86.8 2,628.9 1,791.7
CFU-GEMM
2 52.8 5.1 i: 5.0 -c 99.3 i: 97.6 2 53.7 2 1,380.3 t 936.6
0.1 50.4 13.7 1.4 94.1 57.5
t 4.5 2 0.3 2 112.0 2 11.3 2 3.8 -c 107.8 2 52.9
Progenitor colonies were cultured as detailed in Materials and Methods. Only colonies containing greater than 50 cells were counted. In allcases, each sample was cultured in triplicate. The results represent the mean -c SD from seven separate experiments.
andplated in methylcellulose culture, andCFU-GM were enumerated. Figure 3A and B shows the number of colonies observed at weeks 0, 2, and 5 in these long-term bone marrow cultures in two experiments. Figure 4 shows the foldincrease in CFU-GM at 5 weeks of stromal culture of each of the CD34' cell fractions. In all experiments performed to
"-y
- ""
0
-
- _" l
""_
" "
2
3
- --__4
6
m k s In cullum
Fig 3. Long-term bone marrow culture ofCD34' cell subsets. The isolated CD34' cells from thethree elutriated fractions wereplaced in long-term bone marrow culture as described in Materials and Methods. Ten thousand CD34' cells were seeded per T-25 flask for all of the fractions. Duplicate flasks were made for each sample. At 2 and 5 weeks of culture, the nonadherent and adherent layers of one of the cultures were harvested, enumerated, and plated in methylcellulose t o determine CFU-GM content. The figure shows the CFUGM of theweek 0,2, and 5 cultures: A, Fr RO; 0, Fr 33/37; m, Fr 25/ 29. (A and B) Shown are the results of independent experiments, representative of four experiments.
date, thevastmajorityof CFU-GM activity was observed in the adherent layers of the stromal based cultures (data not shown). In both experiments shown in Fig 3 and in all cultures tested, the cultures seeded with the Fr R 0 CD34' cells contained the greatest number of initial CFU-GM. Despite this large number of progenitors, the number of CFU-GM increased by a maximum of 25% or decreased in number by week 2 of culture. In all Fr R 0 cultures, the number of CFUGM decreased by 60% to 80% by week 5 of culture. CD34' cells from the Fr 33/37-elutriated fraction producedmoderate increases in CFU-GM number after long-term stromal culture (Figs 3A and B and 4). Peak CFU-GM production was observed at 2 weeks, with approximately fourfold increases in colonies at this time point. After further culture to 5 weeks, colony production from the Fr 33/37 CD34' cells decreased to 80% to 300% of that at the initial starting point. In contrast, the Fr 25/29 CD34" cells increased in CFU-GM number at both 2 weeks and 5 weeks of stromal culture (Figs 3A and B and 4). The CFU-GM from the Fr 25/29 CD34' cells increased 4- to 40-fold at 2 weeks, and 25- to 55-fold at 5 weeks of culture (Figs 3 and 4). The cell cycle status of thethreeCD34'cell fractions was analyzed. Using propidium iodide staining (see Materials and Methods), cells containing normal2Nand greater amounts of DNA were quantitated. Figure 5 shows the fluorescence traces of the various CD34' populations, and the percentage of cells in the different phases of the cell cycle. In these analyses, the purity of cells in each of the three CD34' cell fractions was greater than 85%. In unfractionated bone marrow, approximately 90% of the cells were in the G&, stage, with 10% in either S phase or active division. The three CD34'cell populations showedvery divergent activity. The Fr R 0 CD34' cell fraction showed the greatest proportion of cells in active cell cycle. In this case, greater than 43% of the cells were in either S , G?, or M phase. The Fr 33/37 CD34' cells had intermediate numbers of cells in active cycle, with approximately 10% of the cells in either S , G?. or M phase. Unlike the Fr R 0 CD34' cell sample, CD34' cells from the Fr 25/29 were relatively quiescent, with greater than 97% of the cells in the GJG, stage.
WAGNERETAL
518
C
A
-L R 33/37
GO/G1 = 90.0 S = 9.1 G2/M = 0.9
06/01 = 09.6 S = 0.0 G2/M = 1.6
.
i
B =
.n
... ... ...
3.-
D
U
GO/G1 = 56.1 S = 37.1 G2/M = 6.6
GOlGl = 97.2 S = 2.5 G2/M = 0.3
.
1..
. I.e
...
surface receptors for a variety of hematopoietic growth factors were analyzed using flow cytometry. In these studies, the cells were incubated with biotinylated cytokines and counterstained with fluorescein isothiocyanate (F1TC)-conjugated avidin. For these analyses, only the CD34' cells were analyzed using two-color flow cytometry. The cytokine fluorescence profile and the mean fluorescence were determined byflow cytometry. Figure 7 shows representative fluorescence histograms from one experiment where the Fr 25/29, Fr 33/37, andFr R 0 CD34' cells weretested for receptors for seven different growth factors: IL-la, IL-3, IL-6, G-CSF, GM-CSF, SCF, and MIP-la. In most cases, fluorescence with the bound growth factor appeared as a single, sometimes broad, peak. As a control, minimal to no binding was observed in the CD34- fractions (data not shown). The fluorescent growth factor binding couldbe quenched in the presence of a 100-fold excess of the nonbiotinylated growth factor (Fig g), indicating the specificity of binding. The level of growth factor binding was examined for each of the three CD34' cell fractions. Only G-CSF showed less than 25% increases in fluorescence over negative controls in all three CD34' cell fractions. With the other six growth factors, there was a consistent trend in which the smaller CD34' cell fractions possessed higher meanfluorescence levels and consequent growth factor binding. In greater than 80% of the determinations for each growth factor, the fluorescence intensity of cytokine binding in the Fr25/29 CD34' cells was greater than that observed for the Fr R 0 CD34' cells. When statistically compared, P values of less than . l were observed for IL-la ( P = .027), MIP-la ( P = .006), and SCF ( P = .081), suggesting significantly higher fluorescence binding of these growth factors to the smaller Fr 25/29 CD34' cells.
I...
. ... .M
e.
fluorescense Fig 5. Cell cycle analysis from (A) the input BMMC, (B) R 25/29 CD34' cells, (C) Fr 33/37 CD34+ cells, and (Dl Fr R 0 CD34' cells. The X axis depicts propidium iodide fluorescence (DNA content), and the Y axis indicates cell number.
To further analyze the hematopoietic activity of the three CD34' cell fractions, 25,000 to 50,000 of these cells from adult bone marrow were injected into human fetal thymus/ liver grafts that had previously been established in CB 17 SCID mice. For these experiments the purity of the seeded Fr 25/29, Fr 33/37, and Fr R 0 CD34' cells exceeded 50%, 94%, and 92%, with less than 25%, 3%, and 1% contaminating CD34' T cells, respectively. In these studies, the thymus/ liver graft and the donor of the CD34' cells had been HLA class l-typed. To track the donor CD34' cells after injection, only graft-donor CD34' cell pairs with at least one identified HLA mismatch were used. Ten weeks after seeding of the thymusfliver grafts, the grafts were harvested and phenotyped for the presence of human CD45' cells of the donor CD34' cell HLA type. Table 3 shows the results from three independent experiments. When no donor CD34' cells were administered, none of the grafts showed evidence of donortype human cells. In contrast, 23%, 30%, and 40% of the thymusfliver grafts seeded with Fr 25/29, Fr 33/37, and Fr R 0 CD34' cells, respectively, showed evidence of donorderived cells I O weeks after seeding. Approximately 35% to 48% of the cells were of donor origin. These results are approximately equivalent to our previous results with unfractionated CD34' cells, where approximately 40% of the seeded grafts possessed donor-derived cells (data not shown). The cells from the SCID-Hu grafts were further phenotyped. The CD3 T-cell marker was present on 79% to 99% of the donor-derived cells (Fig 6), with 35% to 83% of the cells being CD4'and 55% to 78% being CD8'. CD4'CDS' cells were also observed. The results indicate that CD34' cells from the three CD34' populations not only produce myeloid cells but yield cells of the T-cell lineage when provided the appropriate microenvironment. To further characterize the three CD34' cell populations,
DISCUSSION
In this report, we isolate and characterize subpopulations of human CD34' cells that differ in size and .density. The individual size fractions, having mean diameters of 8.5, 9.3 and 13.5 pm, have distinct phenotypic profiles and prolifera-
Table 3. Reconstitution of Thymus/Liver Grafts in SCID-HU Mice With CD34' Cell Subsets CD34* Cell Subset
No. Grafts Donor-Positive/ No. Animals Tested
% Human Cells in Graft Donor-Positive
No donor cells Fr 25/29 Fr 33/37 Fr R 0
0111 3113 3110 4110
43.7 -c 12.9 47.6 -c 20.7 34.8 z 18.1
ND
The reconstitution of thymuslliver grafts in SCID-HU mice was performed as described inMaterials and Methods. In these experiments, the cells of 0 3 4 ' cell donor origin were discriminated using monoclonal antibodies against the HLA AllIA24 marker and the human CD45 antigen. All original thymus/liver grafts used for these experiments were HLA AllIA24-negative. The results were pooledfrom three separate experiments. CD34' donor cell positivity was observed in the grafts in all three experiments. Abbreviation: ND, not determined.
ISOLATION
OF HUMAN HEMATOPOIETIC STEM CELLS
519
Fr 25/29 CD%+
Fig 6. Cell phenotype of progeny of donor CD34' cells in thymus/liver grafts from SCID-HU mice. Cells from thymus/liver grafts were isolated and phenotyped usingFITC-conjugated anti-HLA-A11/A24 (x axis) and phycoerythrin-conjugatedanti-CD3 (y-axis) monoclonalantibodies.Cells in the upper right quadrants representthe progeny of the seeded CD34' cell subsets: (A) Fr 25/29 CD34' cells, (B) Fr 33/37 CD34' cells, (C) Fr R 0 CD34' cells.
tive capacities. The largest, Fr R 0 CD34' cell fraction, has the highest percentage of cells with lineage antigens, and the bulk of the colony-forming unit-cells (CFU-C) activity. The medium-sized Fr 33/37 fraction showed intermediate levels of differentiation antigen expression and activity in hematopoietic assays. This fraction contains both primitive and committed cells, along with cells at intermediate steps of differentiation. In contrast, the small Fr 25/29 CD34' cells appear to be a quiescent primitive cell population that is almost devoid of the tested surface antigens. In fact, all differentiation antigens detected in the population could be attributed to contaminating T cells in some of the preparations. The actual CD34' cells didnot coexpress the CD3,CD19,CD33, CD38, or HLA-DR antigens. The Fr 25/29 CD34' cells do not proliferate in methylcellulose culture yet produce greater than 100-fold increases in CFU-GM after long-term culture on allogeneic stroma. It is the only one of the three CD34' cell fractions that continues to proliferate and increase CFUGM production after 5 weeks in long-term bone marrow culture. Moreover, the Fr 25/29 CD34' cells appear especially sensitive to culture conditions; these cells do not proliferate in normal progenitor colony assays, and our attempt to grow these cells in any stroma-free system using single growth factors or combinations of the growth factors IL3, IL-l, IL-6, SCF, G-CSF, GM-CSF, erythropoietin, basic fibroblast growth factor, and IL-l1 uniformly did not reproducibly support growth. To date, we have only observed extensive in vitro proliferation of these cells in the adherent layers of stromal cultures supplemented with 100 ng/mL of SCF. Very few CFU-GM were observed in the nonadherent layers in these same cultures at up to 5 weeks. These findings are consistent with those of Rowley et al,"7 who showed that primitive 4-hydroperoxycyclophosphamidetreated CD34TD38- cells could only be maximally stimulated whenboth growth factors andmarrow stroma were used. CD34' cells from each of the three size fractions reconstituted fetal thymusfliver grafts implanted in SCID-Hu mice. At I O weeks after implantation, CD3' T cells of donor origin were observed in 30% to 40% of the grafts. In these studies, the donor T cells observed in the grafts most likely derived
Fr 33/37CD34+
Fr R 0 CD=+
HIA-A11/A24
from the seeded CD34' cells. In all cases, the seeded CD34' cells were at least 50% pure, and the Fr 33/37 and Fr R 0 CD34' cells were greater than 92% pure. In the Fr 33/37 and Fr R 0 populations, less than 3% of the seeded cells were contaminating T cells. Moreover, in all cases examined, double-positive donor CD4'CDS' T cells could also be observed, suggesting the development and maturation ofT cells from an earlier precursor. The data confirm that cells from adult bone marrow can be established and expanded 10- to 100-fold in such mouse models. The results further demonstrate that the Fr 25/29 CD34' cells have multipotential activity leading to both the production of myeloid cells in vitro and lymphoid cells in SCID-HUmice. Notably, CD34' cells from the large R 0 fraction were also found to produce T cells in this mouse model at the same frequency and level as the small Fr 25/29 CD34' cells. These data suggest the possible presence of multipotent cells in all size fractions, and that CD34' cells, which do not express T-cell lineage markers or which express certain myeloid lineage markers, can produce T lymphocytes after culture on adherent thymic stromal cells. This latter conclusion is consistent with recent results using thymic organ cultures.3*~39 Other investigators have used CCE as an initial step in the isolation of primitive human stem Cells isolated using this method have also beenshown to engraft xenogeneic sheep.'" In these isolation procedures, fractions at flow rates of 12 to 14 mUmin are collected at a rotor speed of 1,950 rpm before CD34' cell selection and subsequent DR' cell depletion. Cells collected using these conditions range from 9.5 to 1I pm and most closely correspond in size to those cells found in the Fr 33/37 population. This correspondence is further suggested by the frequency of day14 CFU-GM or BFU-E progenitors in Fr 12/14 CD34' cells and Fr 33/37CD34' cells, whichwas higher than those presented here for the fraction 25/29 CD34' cells. Therefore, the small Fr 25/29 cells represent a smaller sized population of primitive stem cells separated from more committed progenitors that have been discarded in other isolation procedures. Importantly, these cells can be isolatedwithout extensive lineage-specific cell depletion. Receptors for IL-l, IL-3, IL-6, G-CSF, GM-CSF, SCF, and MIP-la were examined in all three CD34' cell fractions
WAGNER ET AL
520
Fr R 0
Fr 33/37 a
I
I
( I
l
e
.
( I
L
e
( I
-a14
II
( I
Fig 7. Growth factor receptor distribution on CD34+cell subsets. The dotted lines represent the control fluorescence of avidin carboxyfluorescein alone; the solid lines represent fluorescence using the indicated biotinylated growth factor conjugate. The X axis plots fluorescence intensity; the Y axis, relative cell number. The growth factors tested and CD34' cell fractions are indicated.
ISOLATION OF HUMAN HEMATOPOIETIC STEM CELLS
ILla
=l
c
521
=l
G-CSF
V
GM-CSF
c
=l 1
IL6
c
l
IQ
1mO
SCF c
Fig 8. Blocking oflabeled growth factor binding by unlabeled biotinylated growth factor. Biotinylated growth factor binding was measured in the presence (dotted lines) or absence (solid lines) of a 100-fold excess of unbiotinylated growth factor. Data in the figure show only the results when Fr 25/29 cells were used; however, similar results were observed for the Fr 33/37 and Fr R 0 cells.
by examining direct cytokine binding to cells. Although the methods used here cannot be used to determine the actual number of receptors in such cells, it can be used tocharacterize the relative numbers of receptors on cells from different populations. The specificity of these methods was confirmed by a number of control experiments. The biologic activity of the biotinylated growth factor was confirmed in growth assays using growth factor-dependent lines. The biotinylated cytokines were also compared with native cytokines in Scatchard analyses for their ability to displace "'I-labeled cytokines on reference cell lines known to express specific receptors. Lastly, the specificity of the reactivity of the biotinylated cytokines was further confirmed by the inability of nonmatching cytokines to block the binding of the biotinylated reagents. In these studies, binding sites were found for IL-I, IL-3, IL-6, G-CSF, GM-CSF, SCF, and MIP-la on each of the three CD34' cell populations. Significant cytokine binding levels were not observed on the CD34- cells. After analysis of the mean channel fluorescence of growth factor binding to the CD34' cells, trends were observed suggesting that the density of receptors for all of these growth factors on the Fr
25/29 CD34' cells was much higher than on the large CD34' cell fraction. Statistical significance was achieved for SCF, IL-la, and MIP-la. This increased density of receptors on the small cell fraction becomes especially significant when one considers the relative sizes of the cells andthemean fluorescence per square micrometer of cell surface area. This type of analysis of receptor bindinghasbeenpreviously described by Shabtai et al:' The proportion of these growth factor binding sites that are functional is unknown. Moreover, the exact function for the high density of receptors is also unknown. These receptors might help sensitize these cells to respond to these agonists, and the fractional occupancy of these receptors may help modulate their reactivity in the marrow microenvironment. Receptors for SCF have been previously characterized on hematopoietic progenitors using specific monoclonal antibodies to c-kit. In the mouse, differing levels of c-kit receptors have been demonstrated on stem cells. Ikuta and Weissman4' showed that 70% to 80% of Thy1 ',Sca',Lin- cells express c-kit, and when this population was fractionated into c-kit+ and c-kit- cells, only the c-kit' cells led to the production of donor-derived cells in irradiated mice. Katayama et
522
WAGNER ETAL
alMsuggested that the most primitive stem cells as found in 5-fluorouracil-treated animals were c-kit", which eventually increased in intensity after cycling. In contrast, Oriic et a130 found the primitive repopulating stem cells in untreated mice to be c-kit bright and rhodamine dull. Similar studies have been reported for the human stem cell population. Briddell et al" showed that only the c-kit+ fraction of CD34+, HLADR- cells generated progenitors after 10 weeks in culture. Moreover, using stromal cultures, Gunji et a12' suggested that the majority of long-term bone marrow culture-initiating cells were c-kit'", whereas colony-forming cells were c-kithi. Again, c-kit- cells showed little hematopoietic activity. These results are in agreement with our results that the primitive CD34+ cells do express receptors for SCF. Any difference in the perceived levels of receptors in these studies may reflect the different assays used to detect receptors. In our studies, direct growth factor binding is measured, whereas in these other studies, specific antibodies are used to indirectly detect receptors. In such assays, measurement of alternative receptors may be missed. The primitive Fr 25/29 CD34' cells were found to have a high density of binding sites for both L - 3 and MIP-la. This finding is complementary to the recent results of Verfaillie et al?' who showed that MIP-la together with IL-3 could lead to the maintenance of long-term bone marrow culture-initiating cells in noncontact stromal cultures. Results indicated that these factors exerted their effects only by direct interaction with the CD34+, HLA-DR- cells. These studies with human stem cells are in agreement with findings using CCE to study murine hematopoietic progenitor~.~"~' As in the murine system, progenitors capable of long-term hematopoietic activity can be found in all fractions. However, the small Fr 25/29 CD34+ cell fraction is devoid of progenitor activity, yet is most enriched in cells that initiate long-term bone marrow cultures. Although direct evidence for the long-term repopulating activity of this cell in humans is not available, several clinical observations support this hypothesis. In clinical studies where CCE has been used for T-cell depletion in allogeneic bone marrow transplantation, cells of this size range are eliminated from the final graft due to the high concentration of T lymphocytes in these fractions. Patients transplanted with such manipulated grafts, although showing early engraftment with reduced graft-versus-host disease, have a higher rate of late graft failure and mixed chimerism as compared with patients receiving unseparated bone marrow?6 Clinical studies are now underway to evaluate whether the purification and infusion of these CD34' cells from the normally discarded fractions will improve engraftment after CCE-manipulated grafts. The Fr 25/29 CD34' cells could be a useful starting population to study the interaction of soluble factors and elements of the stromal microenvironment that act to control proliferation and differentiation of hematopoietic stem cells. Most importantly, successful expansion and manipulation of these cells will lead to broad clinical applications in transfusion, transplantation, and gene therapy. ACKNOWLEDGMENT We thank Sonia Jain, Diane Rood, Leanne Barber, and Ken Dinwiddie for their contributions to the experiments performed in this
study. We also thank Sohel Talib, Dewey Moody, Ramila Philip, Maureen McNally, Susan Alters, Kevin Leiby, Karla Knobel, Kimvan Tran, and Lydia Kilinski for valuable discussions and technical assistance during the course of this work. Special thanks go toElaine Andrews for typing the manuscript. REFERENCES 1. Spangrude RJ, Heimfeld S, Weissman IL: Purification and characterization of mouse hematopoietic stem cells. Science 241 :58, 1988 2. Chung LL, Johnson GR: Long-term hemopoietic repopulation by Thy-l'", Lin-, Ly6A/E' cells. Exp Hematol 20:1309, 1992 3. Uchida N, Weissman IL: Searching for hematopoietic stem cells: Evidence that Thy-l .l'"Lin-Sca-l+ cells are the only stem cells in C57BWKa-Thy-1.1 bone marrow. J Exp Med 175:175, 1992 4. Spangrude GJ, Johnson GR: Resting and activated subsets of mouse multipotent hematopoietic stem cells. ProcNatlAcadSci USA 87:7433, 1990 5. Li CL, Johnson GR: Rhodamine123 reveals heterogeneity within murine Lin-,Sca-l' hemapoietic stem cells. J Exp Med 175:1443, 1992 6. Civin C1, Strauss LC,Brovall C, Fackler MJ, Schwartz JF, Shaper JH: A hematopoietic progenitor cell surface antigen defined by a monoclonal antibody raised against KG-la cells. J Immunol 133:157, 1984 7. Watt S, KarhiK, Gatter K, Furley AJ,Katz FE, HealyLE, Altass LJ, Bradley NJ, Sutherland DR. Levinsky R, Greaves MF: Distribution and epitope analysis of the cell membrane glycoprotein (HPCA-1) associated withhuman hematopoietic progenitor cells. Leukemia I :4 17, I986 8. Andrews RG, Singer JW, Bernstein ID: Monoclonal antibody 12-8 recognizes a 115 kd molecule present on both unipotent and multipotent hematopoietic differentiation antigens. Blood 67342, 1986 9. Berenson RJ, Andrews RG, Bensigner WI, Kalamasz D, Knitter G: Antigen CD34' marrow cells engraft lethally irradiated baboons. J Clin Invest 81:951, 1988 IO. Andrews RG, Bryant EM, Barteimez SH, Muirhead DY, Knitter GH, Bensinger W, Strong DM, Bernstein ID: CD34'marrow cells, devoid of T and B lymphocytes, reconstitute stable lymphopoiesis and myelopoiesis in lethally irradiated allogeneic baboons. Blood 80:1693, 1992 11. Shpall ET. Jose RB, Bearman SI, Franklin WA, Archer PG. Curie1 T, Bitter M, Claman HN, Stemmer SM, Purdy M, Myers SE, Hami L, Taffs S, Heimfeld S, Hallagan J, Berenson RJ: Transplantation ofenriched CD34-positive autologous marrow into breast cancer patients following high-dose chemotherapy: Influence of CD34-positive peripheral-blood progenitors and growth factors on engraftment. J Clin Oncol 12:28, 1994 12. Brandt J, Srour EF, van Besien K, Briddell RA, Hoffman R: Cytokine-dependent long-term culture of highly enriched precursors of hematopoietic progenitor cells from human bone marrow. J Clin Invest 86:932, 1990 13. Lansdorp PM, Dragowska W: Long-term erythropoiesis from constant numbers of CD34+ cells in serum-free cultures initiated with highly purified progenitor cells from human bone marrow. J Exp Med 175:1501, 1992 14. Andrews RG, Singer JW, Bernstein ID: Precursors of colonyforming cells in humans can be distinguished from colony-forming cells expression of the CD33 and CD34 antigens and light scatter properties. J Exp Med 169:1721, 1989 15. Buhring W,Asenbauer B, Katrilaka K,Hummel G, Busch F W : Sequential expression of CD34 and CD33 antigens on myeloid colony-forming cells. Eur J Haematol 42:143, 1989 16. Terstappen WMM, Huang S, Safford M, Lansdorp PM, Loken
ISOLATION OF HUMAN HEMATOPOIETICSTEM CELLS
MR: Sequential generations of hematopoietic colonies derived from single nonlineage-committed CD34+ CD38- progenitor cells. Blood 177:1218, 1991 17. Verfaillie C, Blakolmer K, McGlave P Purified primitive human hematopoietic progenitor cells with long-term in vitro repopulating capacity adhere selectively to irradiated bone marrow stroma. J Exp Med 172:509, 1990 18. Lansdorp PM, Sutherland HJ, Eaves CJ: Selective expression of CD45 isoforms on functional subpopulations of CD34' hemopoietic cells from human bone marrow. J Exp Med 172:363, 1990 19. Fritsch G, Buchinger P, Printz D, Fink FM, Mann G, Peters C, Wagner T, Adler A, Gadner H: Rapid discrimination of early CD34+ myeloid progenitors using CD34-RA analysis. Blood 81:2301, 1993 20. Litzow MR, Brashem-Stein C, Andrews RC, Bernstein ID: Proliferative responses to interleukin-3 and granulocyte colony-stimulating factor distinguish a minor subpopulation of CD4-positive marrow progenitors that do not express CD33 and a novel antigen, 7B9. Blood 77:2354, 1991 21. Briddell RA, BroudyVC, Bruno E, Brandt JE, Srour EF, Hoffman R Further phenotypic characterization and isolation of human hematopoietic progenitor cells using a monoclonal antibody to the c-kit receptor. Blood 79:3159, 1992 22. Gunji Y, Nakamura M, Osawa H, Nagayoshi K, Nakauchi H, Miura Y, Yanagisawa M, Suda T Human primitive hematopoietic progenitor cells are more enriched in KIT'"" cells than in KIThlgh cells. Blood 82:3283, 1993 23. Andrews RC, Singer JW, Bernstein ID: Human hematopoietic precursors in long-term culture: Single CD34+ cells that lack detectable T cell, B cell, and myeloid cell antigens produce multiple colony-forming cells when cultured with marrow stromal cells. J Exp Med 172:355, 1990 24. Smith C, Gasparetto C, Collins N, Gillio A, Muench MO, O'Reilly RI, Moore MAS: Purification and partial characterization of a human hematopoietic precursor population. Blood 77:2122, 1991 25. Baum CM, Weissman IL, Tsukamoto AS, Buckle AM, Peault B: Isolation of a candidate human hematopoietic stem-cell population. Proc Natl Acad Sci USA 89:2804, 1991 26. Craig W, KayR, Cutler RL, Lansdorp PM: Expression of Thy-l on human hematopoietic progenitor cells. J Exp Med 177:1331, 1993 27. Schwartz GN, MacVittie TJ, Monroy RL, Vigneulle RM: Countercurrent centrifugal elutriation (CCE) recovery profiles of hematopoietic stem cells in marrow from normal and 5-W-treated mice. Exp Hematol 14:963, 1986 28. Ploemacher RE, Brons RHC: Separation of CFU-S from primitive cells responsible for reconstitution of the bone marrow hematopoietic stem cell compartment following irradiation: Evidence for a pre-CFU-S cell. Exp Hematol 17:263, 1989 29. Orlic D, Bodine DM: Pluripotent hematopoietic stem cells of lowand high density can repopulate W / W mice. Exp Hematol 20:1291, 1992 30. Orlic D, Fischer R, Nishikawa SI, Nienhuis AW, Bodine DM: Purification and characterization of heterogeneous pluripotent hematopoietic stem cell populations expressing high levels of c-kit receptor. Blood 82:762, 1993
523
3 1. Jones RJ, Wagner JE, Celano P, Zicha MS, Sharkis SJ: Preparation of pluripotent haematopoietic stem cells from spleen colonyforming cells. Nature 347:188, 1990 32. Yoder MC, Du SS, Williams DA: High proliferative potential colony-forming cell heterogeneity identified using counterflow centrifugal elutriation. Blood 82:385, 1993 33. Lebkowski JS, Schain LR, Okrongly D, Levinsky R, Harvey MJ, Okarma TB: Rapid isolation ofhuman CD34 hematopoietic stem cells-Purgingof human tumor cells. Transplant 53:1011, 1992 34. Gartner S, Kaplan HS: Long-term culture of human bone marrow cells. Proc Natl Acad Sci USA 77:4756, 1980 35. Namikawa R, Weilbaecher KN, Kaneshima H, Yee El, McCune JM: Long-term human hematopoiesis in the SCID-hu mouse. J Exp Med 172:1055, 1990 36. McCune JM, Namikawa R, Kaneshima H, Shultz LD, Lieberman M, Weissman IL: The SCID-hu mouse, murine model for the analysis of human hematolymphoid differentiation and function. Science 241:1632, 1994 37. Rowley SD, Brashem-Stein C, Andrews R, Bernstein ID: Hematopoietic precursors resistant to treatment with 4-hydroperoxycyclophosphamide: Requirement for an interaction with marrow stroma in addition to hematopoietic growth factors for maximal generation of colony-forming activity. Blood 8260, 1993 38. Tjennfjord GE, Veiby OP, Steen R, Egeland T: T lymphocyte differentiation in vitro from adult human prethymic CD34+ bone marrow cells. J Exp Med 177:1531, 1993 39. Egeland T, Tj@nnfjordGE, Veiby OP, Steen R Expression of CD33, CD2, CD7, CDlO, or CD19 on CD34' bone marrow progenitor cells indicates multilineage potential andnot lineagerestricted commitment. J Cell Biochem 66, 1994 (abstr G 130, suppl 18B) 40. Brandt J, Baird N, Lu L, Srour E, Hoffman R: Characterization of a human hematopoietic progenitor cell capable of forming blast cell containing colonies in vifro.J Clin Invest 82:1017, 1988 41. Srour EF, Brandt E,Briddell RA, Grigsby S, Leemhuis T, Hoffman R: Long-term generation and expansion of humanprimitive hematopoietic progenitor cells in vitro. Blood 81:661, 1992 42. Shabtai M, Malinowski K, Waltzer WC, Pullis C, Raisbeck AP, Rapaport W:Quantitative analysis of surface marker densities after exposure of T-cells to concanavalin A (Con A): A sensitive early index of cellular activation. Cell Immunol 133:519, 1991 43. Ikuta K, Weissman L:Evidence that hematopoietic stem cells express mouse c-kit but do not depend on steel factor for their generation. Proc Natl Acad Sci USA 89:1502, 1991 44. Katayama N. Shih JP, Nishikawa SI, Kina T, Clark SC, Ogawa M: Stage-specific expression of c-kit protein by murine hematopoietic progenitors. Blood 82:2353, 1993 45. Verfaillie CM, Catanzarro PM, Li WN: Macrophage inflammatory protein la, interleukin 3 and diffusible marrow stromal factors maintain human hematopoietic stem cells for at least eight weeks in vitro. J Exp Med 179643, 1994 46. Wagner JE, Zaturak M, Piantados S, Geller RB, Vogelsant GP, Wingard JR, Sara1 R, Griffin C, Shah N, Zehnbauer BA: Bone marrow transplantation of chronic myelogenous leukemia in chronic phase: Evaluation of risks and benefits. J Clin Oncol 10:779, 1992
