FLT3/FLK2 ligand promotes the growth of murine stem cells and... expansion of colony-forming cells and spleen colony-forming units

From www.bloodjournal.org by guest on November 10, 2014. For personal use only.
1995 85: 2747-2755
FLT3/FLK2 ligand promotes the growth of murine stem cells and the
expansion of colony-forming cells and spleen colony-forming units
S Hudak, B Hunte, J Culpepper, S Menon, C Hannum, L Thompson-Snipes and D Rennick
Updated information and services can be found at:
http://www.bloodjournal.org/content/85/10/2747.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.
From www.bloodjournal.org by guest on November 10, 2014. For personal use only.
FLT3/FLK2 Ligand Promotes the Growth of Murine Stem Cells and the
Expansion of Colony-Forming Cells and Spleen Colony-Forming Units
By Susan Hudak, Brisdell Hunte, Janice Culpepper, Satish Menon, Charles Hannum, LuAnn Thompson-Snipes,
and Donna Rennick
The effect of FLT3/FLK2 ligand (FL) on the growthof primitive hematopoietic cells was investigated using Thy'"Sca1'
stem cells. FL was observed t o interact with a variety of
factors t o initiate colony formation by stem
cells. When stem
cells were stimulated in liquid culture with FL plus interleukin (IL)-3,IL-6, granulocyte colony-stimulating factor (GCSF), or stem cellfactor (SCF), cells capable of formingcolonies in secondary methylcellulosecultures (CFU-c) were
produced in high numbers. However, only FL plus IL-6 supported an increase in the number ofcells capable of forming
colonies in the spleens of irradiated mice (CFU-S). Experiments with accessory cell-depleted bone marrow(Lin- BM)
showed that FL alone lacks significant colony-stimulating
activity for progenitor cells. Nevertheless, FL enhanced the
growth of granulocyte-macrophage progenitors (CFU-GM)
in cultures containing SCF,G-CSF, IL-6, or IL-11. In these
assays, FLincreased the number of CFU-GM initiating colony
formation (recruitment), as well as the number of cells per
colony (synergy). Many of the colonies were macroscopic
and contained greater than 2 x lo4granulocytes and macrophages.Therefore,
FL appears t o function as a potent
costimulus for primitive cells of high proliferative potential
(HPP). FL was also observed t o costimulate the expansion
of CFU-GM in liquid cultures of Lin- BM. In contrast, FL had
no growth-promoting affects on progenitors committed to
the erythrocyte, megakaryocyte, eosinophil, or mast cell lineages.
0 1995 by The American Societyof Hematology.
T
have focused on the influence of FL on early events in stem
cell development leading to the production and expansion
of primitive clonogenic cells. We have also investigated
whether FL can stimulate directly the growth of murine myeloid progenitors and enhance their growth by amplifying
the actions of other growth factors, as has been shown with
human progenitor cells.'33Our results with FL suggest that
there are several striking differences between the murine and
human systems.
HE FLT3FLK2 LIGAND (FL) has been cloned from
both stromal cells' and T
Previous studies have
provided indirect evidence that FL may play a role in the
regulation of hematopoiesis. First, the F L T 3 m K 2 gene was
shown to encode a receptor of the tyrosine kinase receptor
family type III."lo The ligands for several of these receptors
were known to induce the growth and differentiation of hematopoietic cells."-15 Second, FLT3FLK2 receptors appeared to be selectively expressed by highly enriched stem/
progenitor cell populations.16Finally, antisense oligonucleotides of the human homologue (STK-1) of the murine FLT3/
FLK2 receptor inhibited colony-forming activity by CD34'
human bone marrow (BM) cells."
The hematopoietic activities of FL have now been directly
assessed in murine and human systems. In the murine
system, FL induced 'H-thymidine incorporation by
AA4.1+Scal +Lid"fetal liver cells2This population contains
totipotent stem cells and myeloid-erythroid progenitors.I8
Furthermore, the response of AA4.1+ cells was increased
synergistically by combining FL with stem cell factor (SCF).
Similar results were obtained with c-kit+ stem cells.2In other
studies, it was found that FL alone was insufficient to support
colony formation by murine Thy'"Sca1' stem cells.' Nevertheless, FL in combination with interleukin (1L)-3 or IL-6
potentiated their clonal growth. Based on the sum of these
studies, FL appears to possess stem cell-stimulating activities, although these activities have not been thoroughly characterized.
Studies of human hematopoietic cells have determined
the effects of FL on the growth of more mature progenitor
populations. It was reported that FL alone could stimulate
the proliferation of humanCD34' progenitor cells in 3Hthymidine incorporation and colony-forming assay^."^ In addition, FL enhanced myeloid colony formation by human
CD34+ cells in the presence of either granulocyte-macrophage colony-stimulating factor (GM-CSF) or IL-3.l.' When
the activities of FL and SCF were compared, FL was less
effective' and, unlike SCF, did not exhibit any erythroidpromoting
The murine studies presented herein extend the evaluation
of the stem cell-stimulating activities of FL. These studies
Blood, Vol 85, No 10 (May 15). 1995: pp 2747-2755
MATERIALS AND METHODS
Animals. C57BI/Ka Thyl.1 mice were bred and maintained at
Simonsen Laboratory (Gilroy, CA). CBA/J mice were purchased
from Simonsen Laboratory.
Growth factors. Purified recombinant human erythropoietin
(epo) [specific activity (spec act), greater than IO4 U/mg] and mouse
stem cell factor (SCF; spec act, IO5 U/mg) were purchased from
R & D Systems (Minneapolis, MN) and Genzyme (Cambridge, MA),
respectively. Purified recombinant murine GM-CSF (spec act, 1.3
X 10' Ulmg) was a gift of Schering Plough Research Institute (Kenilworth, NJ). Purified human G-CSF (spec act, 3 X IO' Ulmg),
murine IL-3 (spec act, 2 X 10' Ulmg) and murine IL-6 (spec act, 4
X IO' Ulmg) were provided by Drs G. Zurawski, A. Miyajima, and
S. Menon, respectively (DNAX, Palo Alto, CA). Supernatants of
cos-7 cells transfected with cDNA encoding murine IL-I 1 were
provided by Dr F. Lee (DNAX). One unit of cos-7 cell-expressed
IL-l1 was defined as the amount of factor that stimulates half-
From the Departments of Immunology and Molecular Biology,
DNAX Research Institute of Molecular and Cellular Biology, Palo
Alto, CA.
Submitted September I, 1994; accepted January 4, 1995.
DNAX Research Institute is supported by Schering-Plough Corp,
Kenilworth, NJ.
Address reprint requests to Donna M. Rennick, PhD, DNAX Research Institute of Molecular and Cellular Biology, 901 California
Ave, Palo Alto, CA 94304.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
"advertisement" in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 I995 by The American Society of Hematology.
0006-4971/95/8510-0001$3.00/0
2747
From www.bloodjournal.org by guest on November 10, 2014. For personal use only.
2748
maximal 3H-thymidineincorporation by a factor-dependent cell line
platedat 5 X lo4 cells per mL. Purified recombinant murine FL
(spec act, 2 X lo5 U/mg) was produced at DNAX. Briefly, a mouse
FL fragment, encoding amino acid residues 28 to 162, was isolated
by polymerase chain reaction (PCR) using the TI 18 cDNA clone'
as a template. This fragment was inserted into the expression vector
pET3a. Inclusion bodies were isolated from transformed Escherichia
coli-carrying FL-pET3a and solubilized in Tris buffer, pH 8.5, containing 6 mom guanidine HCI and 10 mmol/L dithiothreitol (DTT).
The solubilized inclusion bodies were renatured by dilution in 50
mmol/L Tris, pH 8.5, containing 2.5 mmol/L reduced glutathione,
0.5 m o l / L oxidized glutathione, and 0.15 mol/L NaCI. The renatured protein was purified by sequential chromatography on an anion
exchange column (POROS-Q; PerSeptive Biosystems, Cambridge,
MA) at pH 7.5 and on a cation exchange column (Poros S; PerSeptive Biosystems) at pH 3.0. Protein fractions were assayed for activity using the BaF3 cell line expressing murine FLK2/FLT3,' as
described below. Active fractions were pooled, lyophilized, and
stored at 4°C.Pyrogen levels were determined by the Limulus
Amebocyte Lysate method (Whittaker Bioproducts, Walkersville,
MD) and were found to be approximately 2 EU/mg of protein.
Antibodies. Monoclonal antibodies specific for murine IL-6
(MP2-20F3). M-CSF (5A1), and GM-CSF (22E5) were a gift of Dr
J. Abrams (DNAX). Each of these antibodies wasusedat 15 pg/
mL. This amount has been shown to neutralize more than 200 U/
mL of a specific factor in a proliferation assay of a factor-dependent
cell line. A rabbit antiserum specific for murine G-CSF was provided
by Dr N. Shigekazu (Osaka Bioscience Institute, Osaka, Japan) and
was used at a final dilution of 1:500.
Bioassay. Baflt cells, a stable transformant of Ba/F3 cells expressing the FLT3/FLK2 receptor, wereusedto
quantify FL, as
previously described.' Briefly,Baflt cells were plated at 6 X 10'
cells per well with varying concentrations of purified recombinant
FL, and 3-(4,5-dimethylthiazol-2-yI)-2,5-diphenyltetrazolium bromide (MTT) colorimetric assays were performed. One unit of FL
activity is defined as the amount that stimulates half-maximal MTT
conversion.
Lineage-depleted bone marrow cells (Lin- BM). BM cells were
isolated from femurs and tibias of 6- to 8-week-old mice, then overlaid on Lymphopaque (Accurate Chemicals, Westbury, NY), and
centrifuged at 1,OOOg for 20 minutes. Cells were removed from the
interface, washed,and incubated with unmodified rat monoclonal
antibodies specific for CD4 (GK1.5), CD8 (53.6.7). B220 (RA36B2), Macl (M1/70), GRI (RB6-8C5), and erythrocytes (Ter-119).
Lineage-positive cells were depleted using magnetic particles coated
with goat antirat antibodies (PerSeptive Diagnostics, Cambridge,
MA) in two successive rounds of treatment.
Purification of Thy'"Scal+ Lin- (Thy'"ScaI+)cells. The protocol
used is a modification of that described previously by Spangrude et
al." Briefly, BM cells from C57BVKa Thyl.1 mice were prepared
by isolation of the interface from Lymphopaque followed by incubation with rat monoclonal antibodies specific for CD4, CD8, B220,
Macl, GR1, and erythrocytes. Lineage-positive cells were removed
by two rounds of depletion with antirat coated magnetic particles.
The remaining cells were stained in succession with phycoerythrin
(PE)-goat antirat antibodies (Biomeda, Foster City, CA), fluorescein
isothiocyanate (FITC)-conjugated anti-Thy 1.1 (19XE5), biotinylated anti-Scal (E13 161.7), and Texas Red-conjugated streptavidin
(Biomeda). Cell separation was performed on a dual laser
FACStarP'"5(Becton Dickinson, Milpitas, CA). An initial sort gate
was set to select for cells with intermediate forward light scatter and
low-to-negative staining with PE/propidium iodide (lineage marker
negative, viable cells designated Lin-). Secondary sorting criteria
were intermediate levels of fluorescein staining (Thy'") and high
levels of Texas Red staining (Scal +).
HUDAK ET AL
Colony-forming assays. Either 1 X IO5 nonadherent BM cells,
5 X IO' Lin- BM cells, or 1.5 X IOz sorted Thy'"ScaI+ cells were
seeded in 35-mm culture dishes containing I mL modilied Iscove's
medium (GIBCO, Grand Island, NY), 20% fetal calf serum (FCS;
GIBCO), 50 pmol/L 2-mercaptoethanol, and 0.8% (wt/vol) methylcellulose. All cultures were supplemented with saturating concentrations of FL, various growth factors, or a combination of these, as
indicated in Results. Plates were incubated at 37°C in a humidified
atmosphere flushed with 5 % CO,. After 7 or 14 days of culture, the
number and size of colonies were analyzed. Cell morphologies were
determined after sequentially isolated colonies were applied to glass
slides and stained with Wright-Giemsa (Sigma, St Louis, MO). For
megakaryocyte and eosinophil colony formation, agar (0.3% wt/vol)
cultures were used. After 7 days of incubation, the agar cultures were
fixedin 2.5% glutaraldehyde and stained for acetylcholinesterase
(megakaryocytes) or with Luxol blue (eosinophils) and counterstained with hematoxylin. For mast cells, methylcellulose cultures
were incubated for 21 days. Sequentially isolated colonies were
stained with toluidine blue, and mast cells were identified by their
metachromatic granules.
Liquid culture. Thy'"Scal+ cells (400) or 5 X 10' Lin BM cells
were cultured in 1.5-mL microcentrifuge tubes in a total volume of
315 pL of modified Iscove's medium, 20% FCS (vol/vol), 50 pmol/
L 2-mercaptoethano1,and various growth factors. After 7 days in
culture, cells were harvested, washed, and counted. Cells were resuspended in medium and cultured in methylcellulose to detect colonyforming cells (CFU-c). A combination of hematopoietic growth factors (SCF + IL-3 + IL-6 + epo) wasused in the colony-forming
assays to support the development of all cell lineages. The net increase in CFU-c was calculated based on the number of colonies
formed by the original Lin- BM population and the number of
colonies observed in the secondary cultures.
Spleen colony-forming unit (CFU-S) assay. The CFU-s assay
was performed by injecting various concentrations of cells into lethally irradiated recipients (six miceper group). Spleens were removed 12 days after transplantation andfixedin Tellycsniczky's
fixative (70% ethano1:acetic acid:formalin at 20:l:l). The number
of macroscopic colonies per spleen was determined, and the mean
number of CFU-s detected per six recipients was used to calculate
the total number of CFU-s generated per culture.
Statisrical analysis. Levels of significance for comparisons between samples were determined by Student's t test. When no significance levels are given, the results were not statistically different
from control values.
~
RESULTS
Dose response study. Units of FL activity were established by measuring the survival of the pro-B cell line B d
F3 transfected with a cDNA clone encoding mouse FLT3/
FLK2.' The dose response of the stable transformants called
Baflt cells is shown in Fig 1A. This assay has been used to
standardize all purifiedpreparations of recombinant FL used
in this study. In our previous study, 25 U/mL of native FL
was used to costimulate colony formation byThy'"Sca1'
stem cells in the presence of IL-3.' To determine the amount
of recombinant FL required to stimulate optimal growth of
Thy'"Scal+ cells, varying concentrations of FL were added
to stem cell cultures containing 300 U/mL of IL-3 (Fig IB).
Two different preparations of FL stimulated maximum colony numbers when used at 30 U/mL or more as defined by
the Baflt cell assay. To ensure that all colony assays contained saturating concentrations of F L , 100 U/mL (500 ng/
mL) was used in all subsequent experiments.
From www.bloodjournal.org by guest on November 10, 2014. For personal use only.
HEMATOPOIETICEFFECTS OF FLT3/FLK2LIGAND
2749
FL Units per mL
Fig 1. FL dose response curves. (A) Varying concentrations ofFL
were used t o stimulate the murine Baflt cell line(6 x l o 3 cells per
well1 in a 24-hour growth assay. Data are reported as mean values
2 SEM of triplicate wells. (B1 Varying concentrations of FL were
added t o Thy"'Scal+ cultures supplemented with IL-3 (300 UlmL).
Data reported are mean values ? SEM of triplicate plates (150 cells
per culture1 from two independent experiments In = 6 ) .
FL interacts with a selected set of factors to stimulate
colonv formation by Thy'"Scal+stem cells. The data presented in Fig 2A confirm that FL alone does not initiate the
clonal growth of Thy'"Scal+ cells, whereas it enhances their
growth when combined with IL-3 or IL-6. FL was also found
to promote colony formation by Thy'"Sca I cells when combined with SCF, GM-CSF, G-CSF, or IL-11. No colonies
were observed when FL was combined with IL- I , IL- 10, or
M-CSF (data not shown). When the costimulatory actions
of FL and SCF were compared, FL always supported lower
colony numbers than SCF, regardless of the second factor
present. Furthermore, FL did not significantly increase colony numbers in cultures that already contained SCF plus
another factor. The one exception was the higher number of
colonies observed with FL plus SCF and IL-11.
The data presented in Fig 2B show the number of stem
cell colonies that achieved a diameter of greater than 0.5
mm and contained greater than 2 X IO' cells. It was found
+
that FL, like SCF, was capable of interacting synergistically
with GM-CSF, G-CSF, IL-3, or IL-6 to generate colonies
containing large numbers of cells. Interestingly, only small
colonies were observed when FL and SCF were combined
in the absence of another factor. This result suggests that FL
and SCF can synergize with other factors, but not with each
other, to support the continuous proliferation of stem cell
progeny. However, this didnot appear tobethe case, as
combining FL and SCF with a third factor always resulted
in greater numbers of cells per colonies than could be supported by any two-factor combination containing either FL
or SCF. Indeed, both FL and SCF were required to generate
large colonies in the presence of IL- 1 I .
Cellular composition of stem cell colonies. The colonies
costimulated by FL (shown in Fig 2) were sequentially isolated and analyzed for their cellular composition after staining with Wright-Giemsa. All colonies supported by FL plus
GM-CSF, G-CSF, SCF, IL-6, or IL-I 1 contained only granulocytes and macrophages. Most colonies stimulated by FL
plus IL-3 were also found to consist of granulocytes and
macrophages. FL plus IL-3 did not stimulate a higher proportion of mixed colonies than were stimulated by IL-3 alone
(2%). Similarly, FL plus IL-3 and SCF did not increase the
incidence of mixed colonies above that support by IL-3 plus
SCF (approximately 12%). Based on these results, we have
concluded that FL can enhance cell production. However,
the types of cells that arise in these stem cell colonies are
determined by the actions of the other factors present.
During our morphologic analyses, large numbers of undifferentiated cells (blasts) were observed in colonies grown
for I O to 14 days in the presence of FL plus IL-3, IL-6, GCSF, or SCF. In contrast, few blasts were detected in colonies grown in FL plus GM-CSF. These results suggested
that FLmay interact with some but not all factors to expand a
primitive population of cells in the absence of differentiation.
Additional studies to investigate the significance of this finding are presented below.
FL costimulates the expansion of CFU-c and CFU-S in
stem cell cultures. FL was tested for its ability to expand
A
Fig 2.FL
enhances colony formation by Thy'"Seal' cells. (AI Methylcellulose cultures containing
150 sorted cells were stimulated with SCF (50 U/
mL), GM-CSF (200 UlmL), G-CSF (100 UlmL), 11-11
(100 UlmLl, IL-3 (300UlmLI, IL-6 (250 UlmLI, and FL
(100 UlmLI as indicated. To promote the development oferythrocytes, epo (1U/mL) wasadded t o all
cultures. Arrows indicate theabsence of colonies in
response t o single or multiple factors. (B) Macroscopic colonies (greater than 0.5 mm in diameter)
were scored on day 21 of culture. Data are reported
as mean values t SD of triplicate plates from three
independent experiments In = 9 ) .* P < .05compared
with groups stimulated by a singlefactor; * * P < .05
two factors; t P
compared with groups stimulated by
c .05 compared with thegroup costimulated byFL.
Factors added
none
FL
GM-CSF
G-CSF
IL-l1
IL-3
IL-6
0
LO
Colonies
per
40
60
0
150 Thylo Scal+ Colonies
cells
10
20
>0.5rnm
30
40
From www.bloodjournal.org by guest on November 10, 2014. For personal use only.
HUDAK ET AL
2750
L
Input c 400
FL
IL-3
SCF
IL-6 I
G-CSF
FL + IL-3
FLrSCFk *
FL + IL-6
1''
SCF + IL-6
FL + GCSF
SCF+G-CSF
t
'*
m
'
.
h
,
.
'
- L
*l
'*§
,
+
,
,
B'.
,
.
,
.
,
Fig 3. Clonogenic cells recovered from liquid cultures of Thy'"Sca1' cells. Four hundred cells per well were stimulated with FL in the
presence and absence of other factors for 7 days. The concentrationof each factor used isthe same as that indicated in the Fig 2 legend. The
total numbers of cells (A), CFU-c (B), and day-l2 CFU-S(C) recovered per culture are shown. Arrows indicate the absence of colonies. IA and
B) Data represent means 2 SEM from triplicates culturesfrom two independentexperiments (n = 6). IC) Data represent mean 2 SEM obtained
with six mice per treatment group from two independent experiments (n = 12). ' P < .05 compared with the input values obtained with
Thy'"Sca1' cells not cultured before CFU-c assay. t P < .05 compared with value obtained with SCF
IL-6. §f < .05 compared with value
obtained with SCF + G-CSF.
+
clonogenic cells in 7-day liquid cultures of Thy'"ScaI+ cells.
We also determined the total number of cells produced in
these cultures. It was found that FL plus IL-3, 1L-6, or GCSF supported a small increase in cell number over the input
value of 400 and greatly enhanced cell production (Fig 3A).
Similarly, enhanced production was observed whenstem
cells were precultured in SCF plus IL-6 or G-CSF. Only a
modest increase in cell numbers was stimulated byFL plus
SCF. This outcome was not unexpected based on the small
size of the stem cell-derived colonies supported by these two
factors in our primary methylcellulose cultures.
None of the individual factors were able to support the
expansion of CFU-c in liquid cultures of Thy'"Sca1' cells
(Fig 3B). However, high numbers of CFU-c were generated
when FL was combined with other factors. The most dramatic expansion was obtained with FL plus IL-6 (30-fold)
and with FL plus G-CSF (greater than 20-fold). The precultured cells were assayed for CFU-c in secondary methylcellulose cultures containing a combination of factors (IL-3,
IL-6, SCF, and epo) known to support the growth of many
hematopoietic cell lineages. The sizes of the colonies were
variable, ranging from a few hundred to thousands of cells.
Approximately 30% of the colonies were large and multicentric. Although the majority of the colonies consisted of granulocytes and/or macrophages, there was a small number (3%
to 5%) of large, mixed colonies. Despite the inability of FL to
directly enhance the outgrowth of mixed colonies in primary
cultures, it was capable of costimulating the proliferation of
primitive cells from which multipotential progenitors were
derived.
Figure 3C shows that day-l2 CFU-s were increased fivefold when stem cells were precultured in FL plus IL-6. Interactions between FL and the other factors (IL-3. SCF, or GCSF) did not result in the expansion of CFU-s. Instead, these
factor combinations supported CFU-S numbers equivalent to
or below the input number. A significant but less impressive
expansion of CFU-s occurred in the presence of SCF plus
IL-6 when compared with that obtained with FL plus IL-6.
FL does not support the growth of CFU-c in Lin- BM
cdtures. Experiments with unseparated BM cells showed
that FL alone could stimulate only a small number of colonies as compared with GM-CSF (Fig 4). In contrast, FL was
unable to stimulate colony formation of Lin- BM cells above
background levels. These results suggested that factors produced by accessory cells in unseparated BM cultures may
have contributed to the colony formation observed with FL.
This was confirmed by showing that the number of colonies
induced by FL was diminished in unseparated BM cultures
containing anti-CSF antibodies (Fig 4). Therefore, it appears
that FL does not possess a strong colony-stimulating activity
but can serve as a cofactor.
FL enhances the growth of granulocyte-macrophage colony-forming rrnits (CFU-GM)of low and high proliferative
potential ( H P P ) . Although FL alone didnot support significant colony formation by Lin- BM cells, it markedly
increased the number ofGM colonies present in cultures
containing SCF, G-CSF, IL-6, or IL-I1 (Fig 5). The most
striking finding was theability of FL to interact withGCSF, IL-6, or IL-l I to generate macroscopic colonies (Fig
5. hatched bars) containing greater than 2 X IO4 cells. Such
colonies comprised more than 30% of all colonies formed
in the presence of FL plus IL-6 or IL-I 1. Only a few macroscopic colonies were observed in cultures stimulated with
FL plus SCF. In contrast with these results, FL had no affect
on the number or size of the colonies stimulated by M-CSF
or GM-CSF (Fig 5). Furthermore, FL did not increase the
total number of large and small colonies stimulated by IL3. However, there were twice as many cells in colonies
measuring greater than 0.5 mm in diameter when FL was
used as a cofactor with 1L-3 (Fig 5).
Colonies were sequentially isolated from each treatment
group to verify their cellular composition. A small number
of mixed colonies (3%) were present in cultures containing
IL-3 but their frequency was not significantly changed by
costimulation with FL. The colonies from all other groups
consisted entirely of neutrophilic granulocytes and macrophages. This was also true of the macroscopic colonies, although some differences between the treatment groups were
From www.bloodjournal.org by guest on November 10, 2014. For personal use only.
HEMATOPOIETICEFFECTSOFFLT3/FLK2
2751
LIGAND
Colonies
per
0
100
1 ~ 1 0 5 cells
200
300
I
Unseparated BM
FL
GM-CSF
FL
+
anti-CSFantibodies
FL + isotypecontrol
Fig 4. FL does not support colony formation in
accessory cell-depleted bone marrow cultures. Unseparated (1 x lo5 cells per plate) and Lin- BM cells GM-CSF + anti-CSF
(5 x lo3 cells per plate) were cultured with FL (100
UlmLI or GM-CSF (200 UlmL).
cultures
were
GM-CSF + isotype
control
.
.Some
also supplemented with a mixture of neutralizing antibodies specific for GM-CSF (15 pg/mL), M-CSF (15
FL
pg/mL), G-CSF (15 pg/mL), and IL-6 (15 pg/mL) or
with isotype control antibodies (60 pg/mL) as indiGM-CSF
cated. Data are reported as mean values f SD of
triplicate plates from three independent experiments
(n = 9). "P < .05 compared with groups not treated
0
with antibodies or groups treatedwith isotype control antibodies.
noted. The macroscopic colonies supported by FL plus 1L1 1 contained predominately macrophages, whereas those
supported by FL plus SCF, IL-6. or G-CSF contained predominately granulocytes. Furthermore, the macroscopic colonies generated in the presence of FL contained large numbers of undifferentiated cells, suggesting that primitive cells
were expanded in the absence of differentiation. Figure 6
shows the appearance of cells grown in IL-6 as compared
with those grown in IL-6 plus FL.
Effects of FL on the expansion of CFU-GM in liquid cultures of Lin- BM cells. Lin- BM cells were cultured in
liquid medium containing FL in the presence or absence of
other factors. After 7 days, the precultured cells were assessed for CFU-c activity in secondary methylcellulose
cultures. When factors were present individually in the precultures, only FL was found to expand CFU-c numbers (17fold) over the input value of 118 (Fig 7). An even greater
expansion of CFU-c (greater than 40-fold) was observed
when FL was combined with SCF, IL-6, or G-CSF (Fig 7).
The colonies formed after plating of the precultured cells
Colonies
per
h€-
5x103 Lin- BM cells
Lin- BM
l no
200
300
Colonies Der 5x103 cells
were relatively small (containing 100 to 400 cells) and were
comprised of granulocytes and/or macrophages.
FL does not promote the growth of etythroid burstforming units (BFU-e), mast cellcolony-forming
units
(CFU-mast), eosinophil colonv-jiorming units (CFU-eo), or
megakanvcyte colonyforming units (CFU-meg). We have
investigated the possibility that FL combined with appropriate lineage-specific growth factors may enhance colony
formation by different types of progenitor cells. Our results
show that FL lacks erythroid-promoting activities in epodependent BFU-e assays (Fig 8). A number of factors appear
to regulate megakaryocytopoiesis (ie, IL-3, IL-6, IL-IO, and
IL- 1 ) . m 5 We observed that FL alone did not support the
*
growth of megakaryocyte progenitors or augment the generation of megakaryocyte colonies in the presence of IL-3 (Fig
8).
Wetestedthe
ability of FL to promote the growth of
eosinophil progenitors. In these studies, FL had no detectable
activity when combined with IL-3, GM-CSF, or IL-5 (Fig
8). In earlier studies with Lin- BM cells, FL appeared to
cells/colony
measuring >O.Smm
+
SCF
G-CSF
IL-6
IL-l1
GM-CSF
IL-3
M-CSF
-
+
25950
-
+
27900
+
27150
-
+
47400
+
4300
+
15300
5850
63W
-
+
Fig 5. FL enhances colony formation byCFU-GM.
Methylcellulose cultures containing5 x lo' Lin- BM
cells were supplemented with FL and other growth
factors as indicated. The concentration ofeach factor
used is the
same as that indicatedin the Fig 2 legend.
Data are reported as mean values f SD of triplicate
cultures from three independent experiments (n =
9). The hatched portions of the bars indicate the
number of GMcolonies measuring greater than 0.5
m m i n diameter and containing more than 2 x 10'
cells. Average numbers of cells per colony were determined after large colonies (greater than 0.5 mm)
were pooled from
one plate per treatment group. *P
< .05 compared with groups not supplemented with
FL.
From www.bloodjournal.org by guest on November 10, 2014. For personal use only.
HUDAK ET AL
2752
A
IL-6
C
I
IL-6 + FL 1
-1
Fig 6. Photomicrographs of colonies and harvested cells after 14 days of culture. Cells were harvested from cultures supplemented with
11-6 or IL-6 plus FL and stained with Wright-Giemsafor morphologicexamination. The cells harvestedfrom the microscopic colonies supported
by IL-6 (A) contained primarily mature myeloid cells IBI, whereas the cells harvestedfrom the macroscopic colonies supported byIL-6 plus FL
(C) contained large numbers of blasts (D).A and C, original magnification x 2; B and D, original magnification x 1,200.
have no effect on the IL-3-dependent growth of mast cell
progenitors. Because the strong mast cell-stimulating activities of IL-3 may have masked any weaker activities of FL,
we tested FL with IL-4 and IL-IO. Although IL-4 and ILI O are unable to support the growth of mast cell progenitors
individually.theycan
augment the actions of other factors.*"." Data presented in Fig 8 demonstrate that the combination of FL plus 1L-4 and IL-IO was noninductive, whereas
SCF plus IL-4 and IL-IO induced the formation of colonies
containing many mast cells.
DISCUSSION
Our initial studies showed that FL is incapable of supporting colony formation by Thy'"Sca1.' stem cells.' The fa1' I ure
of FL to induce the clonal growth of these primitive cells
was not surprising due to their requirement for signaling by
multiple factors.'X Significant colony formation was observed when FL was combined with either IL-3 or IL-6.'
Herein, it is shown that FL also promotes stem cell growth
when combined with SCF, G-CSF, GM-CSF, or IL- 1 1. In
contrast, FL was ineffective when combined with IL- I , ILIO, or M-CSF. These results cannot be attributed tothe
absence of any stem cell-stimulating activities by these latter
factors'3.*5.?R
and may simply indicate that not all factor interactions lead to enhanced responses.
Because FL and SCF signal through different but related
tyrosine kinase
we compared their actions in
stem cell assays. In the presence of other factors, FL was
found to be less effective than SCF in recruiting Thy'"Sca1'
cells to form colonies. Furthermore, the recruiting activity
of FL was redundant with that of SCF, as FL usually did not
cause additional colony formation when SCF was present. In
these same cultures, however, FL and SCF exhibited synergistic actions with respect to the total number of
maturing
cells that could bederived from a single stem cell. Therefore,
combining both FL and SCF with a third factor (ie, IL-3,
1L-l I , G-CSF, or GM-CSF) invariably supported the generation of larger colonies. The mechanism responsible for this
From www.bloodjournal.org by guest on November 10, 2014. For personal use only.
HEMATOPOIETICEFFECTS
OF
FLT3/FLK2LIGAND
2753
Input
FL
SCF c-
+-
IL-6
G-CSF t
FL + SCF
FL + IL-6
FL + G-CSF
a
5+*
I
0
20
40
60
80
100
CFU-C per culture
(x 102)
Fig 7. CFU-c recovered from liquid cultures of Lin- BMcells. Five
thousand Lin- BMcells were culturedfor 7 days in FL in the presence
and absence of otherfactors. The total numbersof CFU-c per culture
are shown. Data are presented as the mean r SEM of triplicate cultures from two independent experiments (n = 6). * P
.05 as compared with the input
number obtainedwith Lin- BMcells not cultured
before CFU-c assay.
cultures. We have also combined FL with a variety of factors
known to induce colony formation and found that FL possesses a strong potentiating effect onGM progenitors. No
effect on the growth of other types of progenitor cells was
detected. Similar results havebeen obtained withhuman
cells.',' In our murine assays, FL dramatically augmented
the number and size of GM colonies if combined with GCSF, IL-6, or IL-l I . Little or no potentiation was observed
when FL was combined withIL-3, GM-CSF, or M-CSF.
These latter results were also in contrast with findings in the
human system, as strong synergies between FL and IL-3 or
GM-CSF were responsible for the enhanced growth observed
with human CD34' cells.',' It is possible that the discrepancies observed between human and murineassays may reflect
fundamental differences between the species. Distinguishing
between these possibilities will require further investigation.
In Lin- BM cultures costimulated with FL, some of the
-=
FL t
BFU-e
epo c
type of synergy has yettobe defined. It has been argued
that increased cell production occurs when the cofactors involved (in this case, FLand SCF) provide different but
complimentary signals or stimulate different progeny based
on the differential expression of cytokine receptor^.'^
One important goal of our studies was to determine
whether the actions of FL on Thy'"ScaI+ stem cells resulted
in the gcneration of primitive cells that retained clonogenic
properties. After Thy'"Sca1' cells were precultured for 7
days in FL plus IL-3, IL-6, G-CSF, or SCF, the numbers
of CFU-c recovered were greatly increased over the input
number. Although all of our factor combinations supported
CFU-c production, only FL plus IL-6 stimulated the expansion of day-l2 CFU-S. We also found that SCF plus IL-6
supported an increase in both CFU-c and CFU-S numbers,
albeit to a lesser extent thanFL plus IL-6. The ability of
SCF to interact selectively withIL-6to expand these two
clonogenic populations in stem cell cultures hasbeenreported byothers.'".''
These investigators also showed that
SCF plus IL-6 could stimulate the production of cells capable
of in vivo reconstitution of lymphoid and myeloid lineages
and capable of protecting micefromlethalirradiation.'"
Studies are in progress to determine whether FL can enhance
the generation of cells with similar repopulating and survival
capabilities.
Our studies with murine progenitor cells have shown that
FL alone does not support colony formation in accessory
cell-depleted cultures. This observation is in contrast with
that found in the human system, where FL stimulated significant GM colony formation by CD34' progenitor cells.'.3
The reason for this difference is not known and cannot be
explained by the secondary effects of accessory cells because
the human CD34' cells used in these experiments were also
devoid of accessory cells. It is possible that the use of serum
containing small amounts of colony-stimulating factors may
account for the stimulatory activity of FL in thehuman
+ epo
FL
1
CFU-meg
+
11-3
+
IL-3
FL
IL-l1
-
11-5
FL
F1 + IL-5
GM-CSF
FL
+
GM-CSF
IL-5
+
GM-CSF
FL
SCF
IL-10
+ IL-4 +
11-10
+
11-4
CFU-eos
I
'm
4-i
+
11-4
L
CFU-mast
+ IL-10 -~
50
0
25
75
Colonies per Culture
Fig 8. FL has no effect on colony formation byBFU-e, CFU-meg,
CFU-eo, or CFU-mast. The ability of FL t o either support or enhance
the clonal growth ofvarious types of lineage-committed progenitor
cells was evaluated. See Materials and Methods for details about
specific assays and the identification of cells comprising colonies
generated in these cultures. IL-4 and IL-l0 were each added at 100
UlmL. All other factors were
present at the same concentrations as
indicated in the Fig 2 legend. Data are reported as the mean values
? SD of triplicate cultures from threeindependent experiments (n =
9).
From www.bloodjournal.org by guest on November 10, 2014. For personal use only.
2754
GM colonies were enormous, containing greater than 2 X
lo4 cells. Such colonies are known to be formed by a subset
of progenitor cells of HPP. The growth requirements of
CFU-HPP are complex, as stimulation by two or more factors is needed to initiate their growth and to produce optimal
numbers of maturing cell^.^*"^ CFU-HPP have been shown
to be the precursors of C m - c of lower proliferative potential.36-38 Our results indicate that FL in conjunction with GCSF, IL-6, or IL-11 supports the growth of CFU-HPP. These
results agree with those obtained with human CD34+ cells,
where the growth of CFU-HPP was elicited by FL plus GMCSF or IL-3.' The colonies formed by CFU-HPP in our
murine cell cultures contained large numbers of blasts as
well as maturing granulocytes and macrophages. Based on
this observation, we tested the possibility that substantial
numbers of clonogenic cells had been generated. In the absence of other factors, we found that FL alone supported a
17-fold expansion of CFU-c in suspension cultures of LinBM cells. It is conceivable that the actions of FL weredependent on cosignals provided by interactions between Lin- BM
cells, or that mature accessory cells were generated rapidly
when some of the Lin- BM cells attached to plastic. In
liquid cultures, unlike semisolid cultures, these events are
unavoidable. Regardless of whether FL alone is sufficient to
support the proliferation of pre-CFU, the production of CFUc was clearly augmented if FL were combined with SCF,
G-CSF, or IL-6.
The growth characteristics of the CFU-c derived from
Thy'"Scal+ or from Lin- BM precultures were slightly different, although the same factor combinations were used for
their generation. Specifically, the CFU-c from Thy'"Scal+
precultures formed large, multicentric colonies containing
mixtures of granulocytes and macrophages. A few colonies
(3% to 5%) contained additional cell types and blasts, suggesting they were formed by multipotential CFU-c. In contrast, the CFU-c from Lin- BM precultures formed smaller
colonies and contained only mature granulocytes and/or
macrophages. Apparently, the CFU-c generated in Lin- BM
cultures were late GM-committed progenitors with relatively
low proliferative potential. Therefore, it is likely that they
were derived from an early population equivalent to GMcommitted CFU-HPP. Because considerable expansion of
CFU-c occurred in both the Thy'"Scal+ and Lin- BM precultures, FL seems to be very effective in regulating the sequential development of early and late CFU-c from ancestral
cells.
Stromal cells are known to play an essential role in supporting hematopoiesis in the bone marrow microenvironment. The local production of growth factors by stroma is
believed to provide most of the signals required for normal
stem cell and progenitor cell development. The isolation of
FL from a bone marrow stromal cell line' suggests that this
factor may contribute to steady-state hematopoiesis. Therefore, we have studied the activities of FL in the presence of
factors that are derived mostly, if not exclusively, from stromal cells. Presently, the role that FL plays in the de novo
generation of pluripotential stem cells is unknown. However,
we have shown that FL interacts with certain stromal-derived
factors to initiate stem cell proliferation, resulting in the
HUDAK ET AL
production and expansion of primitive decendents that are
believed to comprise reserve progenitor cell pools (ie, CFUS and CFU-c). Furthermore, interactions between FL and
specific stromal factors appear to favor the generation of
CFU-GM and to enhance the subsequent proliferation of
CFU-GM. Based on the results of these and previous studies,'-3FL appears to optimize not only the growth of early
hematopoietic populations but to skew bone marrow-dependent myelopoiesis toward the preferential production of
granulocytes and monocytes.
ACKNOWLEDGMENT
We thank Dr G. Zurawski (DNAX), Dr J. Abrams (DNAX), and
Schering-Plough Research (Kenilworth, NJ) for generous gifts of
reagents and Dr G. Holland for critical reading of the manuscript.
REFERENCES
1. Hannum C, Culpepper J, Campbell D, McClanahan T, Zuraw-
ski S , Bazan JF, Kastelein R, Hudak S , Wagner J, Mattson J, Luh
J, Duda G, Martina N, Peterson D, Menon S , Shanafelt A, Muench
M, Kelner G, Namikawa R, Rennick D, Roncarolo M-G, Zlotnik A,
Rosnet 0, Dubreuil P, Birnbaum D, Lee F: Ligand for FLT3FLK2
receptor tyrosine kinase regulates growth of haematopoietic stem
cells and is encoded by variant RNAs. Nature 368:643, 1994
2. Lyman SD, James L, Vanden Bos T, de Vries P, BraselK,
Gliniak B, Hollingsworth LT, Picha KS, McKenna HJ, Splett RR,
Fletcher FA, Maraskovsky E, Farrah T, Foxworthe D, Williams DE,
Beckmann MP: Molecular cloning of a ligand for the flt3/flk2 tyrosine kinase receptor: A proliferative factor for primitive hematopoietic cells. Cell 75: 1157, 1993
3. Lyman SD, James L, Johnson L, Brasel K, de Vries P, Escobar
S S , Downey H, Splett RR, Beckmann MP, McKenna W :Cloning
of the human homologue of the murine flt3 ligand A growth factor
for early hematopoietic progenitor cells. Blood 83:2795, 1994
4. Rosnet 0, Marchetto S , deLapeyriere 0, Birnbaum D: Murine
Flt3, a gene encoding a novel tyrosine receptor of the PDGFW
CSFlR family. Oncogene 6:1641, 1991
5 . Shibuya M, Yamaguchi S, Yamane A, Ikeda T, Tojo A, Matsushime H, Sat0 M: Nucleotide sequence and expression of a novel
human receptor-type tyrosine kinase gene @ t ) closely related to the
fms family. Oncogene 5519, 1990
6. Rosnet 0, Schiff C, Pebusque M-J, Marchetto S , Tonnelle C,
Toiron Y, Birg F, Bimbaum D: Human FLT3FLK2 gene: cDNA
cloning and expression in hematopoietic cells. Blood 82:1110, 1993
7. Bazan J F Genetic and structural homology of stem cell factor
and macrophage colony-stimulating factors. Cell 65:9, 1991
8. Matthews W, Jordan CT, Gavin M, Jenkins NA, Copeland
NG, Lemischka IR: A receptor tyrosine kinase cDNA isolated from
a population of enriched primitive hematopoietic cells and exhibiting
close genetic linkage toc-kit. Proc NatlAcad Sci USA 88:9026,
1991
9. Geissler E, Ryan M, Housman D: The dominant-white spotting
(W) locus of the mouse encodes the c-kit proto-oncogene. Cell
55:185, 1988
10. Sherr CJ, Rettenmier CW, Sacca R, Roussel MF, Look AT,
Stanley ER: The c-fms proto-oncogene product is related to the
receptor for the mononuclear phagocyte growth factor, CSF- 1. Cell
41:665, 1985
11. Stanley ER, Chen DM, Lin H-S: Induction of macrophage
production and proliferation by a purified colony stimulating factor.
Nature 274:168, 1978
12. Stanley ER, Guilbert LJ, Tushinski RI, Bartelmez SH: CSF1-A mononuclear phagocyte lineage-specific hemopoietic growth
factor. J Cell Biochem 21:151, 1983
From www.bloodjournal.org by guest on November 10, 2014. For personal use only.
HEMATOPOIETIC EFFECTSOF
FLT3FLK2LIGAND
13. Zsebo KM, Wypych J, McNiece IK, Lu HS, Smith KA, Karkare SB, Sachdev RK, Yuschenkoff VN, Birkett NC, Williams LR,
Satyagal V N , Tung W, Bosselman RA, Mendiaz EA, Langley KI:
Identification, purification, and biological characterization of hematopoietic stem cell factor from buffalo rat liver-conditioned medium.
Cell 63:195, 1990
14. Metcalf D, Nicola NA: Direct proliferative actions of stem
cell factor on murine bone marrow cells in vitro: Effects of combinationwith colony-stimulating factors. Proc Natl Acad Sci USA
88:6239, 1991
15. Huang E, Nocka K, Beier DR, Chu T-Y, Buck J, Lahm
H-W, Wellner D, Leder P, Besmer P: The hematopoietic growth
factor KL is encoded by the SI locus and is the ligand of the c-kit
receptor, the gene product of the W locus. Cell 63:225, 1990
16. Matthew W, Jordan CT, Wiegand GW, Pardoll D, Lemischka
IR: A receptor tyrosine kinase specific to hematopoietic stem and
progenitor cell-enriched populations. Cell 65:1143, 1991
17. Small D, Levenstein M, Kim E. Carow C, Amin S , Rockwell
P, Witte L, Burrow C, Ratajczak M Z , Gewirtz AM, Civin CI: STK1, the human homolog of Flk-2/Flt-3, is selectively expressed in
CD34' human bone marrow cells and is involved in the proliferation
of early progenitor/stem cells. Proc Natl Acad Sci USA 91:459,
1994
18. Jordan CT, McKearn JP, Lemischka IR: Cellular and developmental properties of fetal hematopoietic stem cells. Cell 61:953,
1990
19. Spangrude GJ, Heimfeld S , Weissman IL: Purification and
characterization of mouse hematopoietic stem cells. Science 24158,
1988
20. Rennick D, Jackson J, Yang G, Wideman J, Lee F, Hudak
S : Interleukin-6 interacts with interleukin-4 and other hematopoietic
growth factors to selectively enhance the growth of megakaryocytic,
erythroid, myeloid, and multipotential progenitor cells. Blood
73:1828, 1989
21. Bruno E, Hoffman R: Effects of interleukin-6 on in vitro
human megakaryocytopoiesis: Its interaction with other cytokines.
Exp Hematol 17:1038, 1989
22. Williams N, De Giorgio T, Banu N, Withy R, Hirano T,
Kishimoto T:Recombinant interleukin 6 stimulates immature murine
megakaryocytes. Exp Hematol 18:69, 1990
23. Paul SR, Bennett F, Calvetti JA, Kelleher K,Wood CR,
O'Hara R M , Leary AC, Sibley B, Clark SC, Williams DA, Yang
YC: Molecular cloning of a cDNA encoding interleukin 11, a stromal
cell-derived lymphopoietic and hematopoietic cytokine. Roc Natl
Acad Sci USA 87:7512, 1990
24. Bruno E, Briddell RA, Cooper RJ, Hoffman R: Effects of
recombinant interleukin 11 on human megakaryocyte progenitor
cells. Exp Hematol 19:378, 1991
2755
25. Rennick D, Hunte B, Dang W, Thompson-Snipes L, Hudak
S : Interleukin-l0 promotes the growth of megakaryocyte, mast cell
and multilineage colonies: Analysis with committed progenitors and
Thyl'"Scal+ stem cells. Exp Hematol 22:136, 1994
26. Thompson-Snipes L, Dhar V, Bond MW, Mosmann TR,
Moore KW, Rennick D: Interleukin-IO: A novel stimulatory factor
for mast cells and their progenitors. J Exp Med 173:507, 1991
27. Rennick D, Hunte B, Holland G, Thompson-Snipes L: Cofactors are essential for stem cell factor-dependent growth and maturation of mast cell progenitors in assessory cell-depleted cultures.
Comparative effects of IL-3, IL-4, IL-IO, and fibroblasts on SCFdependent mast cell progenitor development. Blood 8557, 1995
28. Heimfeld S , Hudak S , Weissman I, Rennick D: The in vitro
response of phenotypically defined mouse stem cells and myeloerythroid progenitors to single or multiple growth factors. Proc Natl
Acad Sci USA 88:9902, 1991
29. Metcalf D: Hematopoietic regulators: Redundancy or subtlety? Blood 82:3515, 1993
30. Miura N, Okada S , Zsebo KM, Miura Y, Suda T: Rat stem
cell factor and IL-6 preferentially support the proliferation of c-kitpositive murine hemopoietic cells rather than their differentiation.
Exp Hematol 21:143, 1993
31. Bodine D, Orlic D, Birkett N, Seidel N. Zsebo K: Stem cell
factor increases colony-forming unit-spleen number in vitro in synergy with interleukin-6, and in vivo in SUSld mice as a single factor.
Blood 79:913, 1992
32. Bradley TR, Hodgson GS: Detection of primitive macrophage
progenitor cells in mouse bone marrow. Blood 54:1446, 1979
33. McNiece IK, Robinson BE, Quesenbeny PJ: Stimulation of
murine colony-forming cells with high proliferative potential by the
combination of GM-CSF and CSF-I. Blood 72:191, 1988
34. Muench MO, Schneider JG, Moore MAS: Interactions among
colony-stimulating factors, IL-Ip, IL-6, and kit-ligand in the regulation of primitive murine hematopoietic cells. Exp Hematol 20:339,
1992
35. Metcalf D, Nicola NA: The clonal proliferation of normal
mouse hemopoietic cells: Enhancement and suppression by CSF
combinations. Blood 79:2861, 1992
36. McNiece IK, Bradley TR, Kriegler AB, Hodgson GS: Subpopulations of mouse bone marrow high-proliferative-potential colony-forming cells. Exp Hematol 142456, 1988
37. McNiece IK, Williams NT, Johnson G, Kriegler AB, Bradley
TR, Hodgson GS: Generation of murine hematopoietic precursor
cells from macrophage high-proliferative-potential colony-forming
cells. Exp Hematol 15:972, 1987
38. Stanley ER, Bartocci A, Patinkin D, Rosendaal M, Bradley
TR: Regulation of very primitive, multipotential, hemopoietic cells
by hemopoietin-l. Cell 45:667, 1986
`