RAPID COMMUNICATION
Direct Contact Between Human Primitive Hematopoietic Progenitors and Bone
Marrow Stroma Is Not Required for Long-term In Vitro Hematopoiesis
By Catherine M . Verfaillie
Long-term bone marrow cultures support both differentiation and conservation of primitive human hematopoietic
progenitors in the absence of exogenous cytokines. It is
believed that hematopoiesis in such cultures requires direct
contact between hematopoietic progenitors and stroma. In
the present study, we demonstrate that primitive progenitors physically separatedfrom the stromal layer by a 0.45-pm
microporous membrane continue to generate differentiated
progenitorsfor at least 8 weeks. Moreover, primitive progenitors are conserved to a greater extent under these conditions, as when cultured in direct contact with the stroma.
However, excessive production of granulocyte-macrophage
progenitors occurs when primitive progenitors are not allowed to interact directly with the stroma. Thus, direct
contact between hematopoietic and stromal cells is not
required for either differentiation or conservation of primitive hematopoietic progenitors but is essential for the regulated production of mature blood elements. These findings
can now be used to define the role of diffusible factors and
cell-cell or cell-extracellular matrix adhesion events in the
regulation of conservation, proliferation, and differentiation
of primitive human hematopoietic progenitors in vitro.
Q 1992 by The American Society of Hematology.
S
killer (NK) cells by sheep elythrocyte rosetting as previously
described.’ Further depletion of committed lymphoid and myeloid/
monocytic cells was obtained by negative immunomagnetic depletion of cells expressing CD2, CD3, CDllb, CD19, CD56, CD71,
MY8, and glycophorin-A antigens using previously described
methods.2
The resultant lineage negative cells were labeled with anti-CD34
and anti-HLA-DR antibodies as previously described.* Cells were
sorted on a FACS-Star-Plus laser flow cytometry system (BectonDickinson, Mountain View, CA) equipped with a Consort 40
computer (Becton-Dickinson).Cells were initially selected for low
vertical and very low/low horizontal light scatter properties. Cells
selected in the first window expressing high numbers of CD34
antigens and lacking HLA-DR antigens expression were then
sorted (DR- cells)? The latter windows were chosen on the basis
of the fluorescence pattern of control samples labeled with mouse
IgGl-phycoerythrin (PE) and mouse IgG2a-fluoresceine (FITC)
antibodies.
TROMA-DEPENDENT long-term bone marrow cultures support proliferation and differentiation of primitive human hematopoietic progenitor cells for at least 12
weeks in the absence of exogenously added cytokines.1,2
Although it is well established that stromal cells in these
cultures are an absolute requirement for the maintenance
of hematopoiesis: it is not clear if proliferation and
differentiation of hematopoietic cells in stroma-dependent
cultures is the result of direct interactions between hematopoietic elements and stromal cells or their extracellular
n ~ a t r b &or
~ is the result of stroma-derived soluble cytokines. An alternative long-term culture system has recently
been described3in which highly purified human hematopoietic progenitors induced hematopoiesis for up to 8 weeks in
the absence of an adherent stromal layer but in the
presence of repeatedly added cytokines, suggesting that an
important role of the stromal cells in stroma-dependent
cultures is to provide cytokines that promote differentiation
of primitive hematopoietic progenitors. In the present
study we examined if close cell-cell contact between stromal
cells and purified human primitive progenitors is necessary
for the initiation and maintenance of hematopoiesis in
stroma-dependent cultures.
MATERIALS AND METHODS
Purijcation of DR- Cells
Bone marrow was obtained from 22 healthy young volunteers
after informed consent by aspiration from the posterior iliac crest
in preservative-free heparin. Bone marrow mononuclear cells
(BMMNC) were obtained after Ficoll-Hypaque separation (Sigma
Diagnostics, St Louis, MO) (specificgravity 1.077).
BMMNC were purified further in an initial counterflow elutriation step using previously described methods, with some modificatiom6 In short, BMMNC were resuspended in phosphate-buffered
saline supplemented with 0.3% bovine serum albumin (Sigma) and
0.01% EDTA (Sigma). The cells were injected into an elutriator
system with standard separation chamber (Beckman Instruments,
Inc, Palo Alto, CA) primed with Iscove’s modified Dulbecco’s
medium with fetal calf serum and EDTA (IMDM+FCS+EDTA).
Rotor speed and temperature were maintained at 1,950 rpm and
10°C. After loading, 200 mL of effluent was collected at a flow rate
of 14 mL/min. The rotor was then stopped and the remaining
BMMNC flushed from the separation chamber. Cells collected in
fraction 14 were then depleted from T lymphocytes and natural
Blood, Vol79, No 11 (June 1). 1992: pp 2821-2826
Phenotype
Cells collected from the upper wells of “stroma-non-contact”
cultures were analyzed at week 5 of culture for the presence of
From the Department of Medicine, University of Minnesota, Minneapolis
Submitted February 5,1992; accepted March 13,1992.
Supported in part by National Institutes of Health Grants ROI-CA4581401, POI-CA-21737, and ROI-CA-43924.Also supported by the
Gambel-Skogmo Foundation, the Paul Christiansen Foundation, the
University of Minnesota Bone Marrow TransplantResearch Fund, the
Minnesota Medical Foundation, the Leukemia Task Force, the
Children’s Cancer Research Fund, the Graduate School of the
University of Minnesota and the American Cancer Society.
C.M.K is a Special Fellow of the Leukemid Society ofAmerica and
a Special Fellow of the “Fundacionlntenzacional Jose Carreras Para
La Lucha Contra La Leucemia.”
Address reprint requests to Catherine M. Veqaillie, MD, Department
of Medicine, Division of Hematology, UMHC Box 480, Harvard St at
E River Rd, Minneapolis, MN 55455.
The publication costs of this article were defrayed in part by page
charge payment. nhis article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. section 1734 solely to
indicate thisfact.
8 1992 by TheAmerican Society of Hematology.
0006-4971I921 7911-OO39$3.00/0
282 1
CATHERINE M. VERFAlLLlE
2822
CD34+/HLA-DR+ and CD34+/HLA-DR- cells. Cells were labeled with anti-CD34-PE antibody (Becton-Dickinson) and antiHLA-DR-FITC antibody (Becton-Dickinson). Cells were analyzed
for the expression of these antigens on a FACS-Star-Plus flow
cytometry system, equipped with a Consort computer. PE and
FITC coupled isotype matched mouse Igs were used as control.
Short-term Methylcellulose Assay
Sorted DR- cells or cells recovered from long-term in vitro
cultures initiated with DR- cells were plated in clonogenic
methylcellulose assay supplemented with 3 IU recombinant erythropoietin (Epoetin) (Amgen, Thousand Oaks, CA) and 10%
conditioned media from the bladder carcinoma cell line 5637 as
previously described? Cultures were incubated in a humidified
atmosphere at 37°C and 5% COz for 18 to 21 days. The cultures
were assessed at days 18 to 21 of culture for the presence of
CFU-MIX, CFU-GM, and BFU-E as previously described?
Long-tem In Etro Cultures
“Stroma-fiee” cultures. DR- cells, 2 to 8 x l@/mLwere plated
in complete media in wells of 24- (1 mL) or 6-well plates (4 mL)
(Costar, Cambridge, MA) (Fig 1). No stromal layers were established. No cytokines were added to the complete media. The
culture media consisted of IMDM with 12.5% fetal calf serum
(HyClone Laboratories, Logan, UT), 12.5% horse serum (HyClone
Laboratories), 2 mmol/L L-glutamine (GIBCO Laboratories),
penicillin 1,OOO U/mL and streptomycin 100 U/mL (GIBCO), and
mol/L hydrocortisone (A-Hydrocort) (Abbott Laboratories,
North Chicago, IL).
“Sfroma-contact”cu1tures. Irradiated stromal cells were subcultured in 6-well (2 x lo6 cells suspended in 4 mL) or 24-well
(0.5 x 106 cells suspended in 1 mL) plates. DR- cells (2 to
8 x 103/mL) were then plated onto the irradiated allogeneic
stromal layers as described* (Fig 1).
STROMA-FREE
“Stroma-non-contact” cultures. “Stroma-non-contact” cultures
consisted of allogeneic irradiated stromal cells derived from the
same donors as the stromal cells used in the “stroma-contact”
cultures subcultured in the bottom well of 6- (2 x 106 cells
suspended in 3 mL) or 24- (0.5 x 106 cells suspended in 0.8 mL)
well plates. A collagen-treated Transwell insert (0.45 *m microporous filter) (Costar) was then placed on top of the stromal layer,
and sorted DR- cells placed in the upper wells (2 to 8 x lo3 cells in
0.2 mL complete media for 24-well plates, or 4 to 32 x lo3 DRcells in 1mL complete media for 6-well plates) (Fig 1).
Maintenance of cultures. All cultures were maintained in a
humidified atmosphere at 37°C and 5% COz. At weekly intervals
“stroma-contact” and “stroma-free” cultures were fed by removing
half of the cell-free supernatant and replacing it with fresh
complete media. For “stroma-non-contact” cultures, half the
media from the bottom wells only was removed and replaced by
fresh complete media.
Evaluation of long-term cultures. Nonadherent and adherent
cells recovered from selected “stroma-contact” cultures after
treatment with 0.15% collagenase (Boehringer Mannheim) were
assayed at different timepoints in short-term methylcellulose assay
for the presence of committed progenitors. Likewise, cells from
selected “stroma-free” cultures or present in the upper wells of
selected “stroma-non-contact” cultures were collected at different
timepoints, enumerated under hemocytometer, examined for their
morphology and phenotype, and assayed for the presence of
committed or primitive progenitors.
Limiting DilutionAssays (LDA)
At day 0, DR- cells (24 replicates per concentration) (experiments 1 through 3: 1,000,333,111, or 33; experiment 4: 500,200,
100, or 20; experiments 5 and 6: 400,150,50,15 DR- cells/well)
were plated onto 3 x 104 irradiated allogeneic stromal cells,
subcultured in 96-well plates (Costar) (day 0 LDA).*x9Likewise,
cells recovered after 5 weeks from collagenase-treated “stroma-
STROP4A-CONTACT
STROUA-NON-CONTACT
TRnrrvell
€Ei3
1-
Bonomvell
Uicropomns membrane
24.5 mm diameter (6 pall phte)
6.5 mm diameter (24 vell plate)
+
PROGENITORS
MEDIA
ALLOGENEIC IRRADIATED
STROMA
ALLOGENEIC IRRADIATED
STROMA
UEDIA
MEDIA
+
A
PROGENITORS
1CELL-STROHA CONTACTI
+
A
PROGENITORS IN
TRANSWELL-INSERT
NO CELL-STROYA CONTACT
Fig 1. DR- cells were suspended in fetal calf serum, horse serum, and hydrocortisonecontaining media but without exogenouscytokines. Cell
suspensionswere plated either without stromal layer (“stroma-free“) directly onto allogeneic irradiated stromal layers (“stroma-contact“) or in
Transwell-inserts that separated DR- cells from the stroma by a 0.45-pm microporous collagen-coated membrane allowing free passage of
diffusible factors but preventing cell-cell contact (“stroma-non-contact“). Transwellinserts are porous bottom dishes that can be placed with and
overhanging lip over wells of 24-well or 6-well cluster plates (6.5 mm diameter for 24-well plates and 24.5 mm diameter for 6-well plates). When
placed correctly, the Transwell membrane will separate cells present in the insert by 1 mm from the stromal layer attached to the bottom well. In
our experiments, Transwell-Colinserts were used. The membrane of Transwell-Col inserts has been treated with an equimolar mixture of types I
and 111 collagen derived from bovine placenta. The result is a collagen-treatedmembrane that retains its microporosity but is transparentwhen
wet (informationprovidedby manufacturer).
2823
HEMATOPOIESISWITHOUT STROMAL CONTACT
“stroma-free’’ cultures nor in the transwell inserts of
“stroma-non-contact”cultures.
As expected? when DR- cells were plated in the absence
of a stromal layer (“stroma-free”), we observed a progressive decrease in cell number (Fig 2). Virtually all cells were
monocytes at week 2. In contrast, serial evaluation of
“stroma-non-contact”cultures showed that, after an initial
decrease in cell number at week 1, the cell number in the
Transwell-inserts increased steadily (Fig 2). At week 1,
more than 55% of cells were blasts admixed with promyelocytes (Fig 3). Over the next 4 weeks, the percentage of
blasts declined; the percentage of promyelocytes remained
constant and a gradual increase in mature myeloid elements was seen. At week 8, blasts and myeloid precursors
decreased further with a reciprocal increase in monocytes.
FACS analysis of cells present in “stroma-non-contact”
cultures at week 5 demonstrated that 4.1% & 1.2% of cells
were CD34+/HLA-DR+ (n = 6) associated with more
differentiated hematopoietic progenitors,*p8 and 1.1% 2
0.16% of cells remained CD34+/HL,A-DR- (n = 6). Taking into account that the total cell number was 8 3.8-fold
higher at week 5 compared with day 0, these studies
demonstrate that for each DR- cell used to initiate the
cultures 19 2 5 CD34+/HLA-DR+ cells were generated
and approximately 6% of DR- cells could be conserved for
a minimum of 5 weeks.
These observations suggested to us that, although stroma
is important for in vitro hematopoiesis, direct contact
between primitive hematopoietic progenitors and the stromal layer is necessary neither for the differentiation of such
progenitors in more differentiated 34+/DR+ cells and
mature myeloid cells nor for the conservation of a fraction
of primitive 34+/DR- progenitors. To test this hypothesis
further, we plated cells recovered from “stroma-free,”
“stroma-contact,’’ and “stroma-non-contact” cultures in
methylcellulose progenitor assay to evaluate the production
of clonogenic cells. Very few clonogenic cells were present
in “stroma-free’’ cultures during the first 3 weeks while
none were present in such cultures at week 5 and 8 (Table
1). “Stroma-contact”cultures yielded an increasing number
of clonogenic cells over the first 5 weeks with a decrease in
committed progenitor recovery by week 8, consistent with
contact” cultures or Transwell-inserts of “stroma-non-contact”
cultures initiated at day 0 with 35,488 (experiments 1 through 3),
19,680 (experiments 4 through 6), or 14,760 (experiments 5 and 6)
DR- cells were plated in LDA (cell number = the equivalent of
1,000,333,111, and 33 (experiments 1 through 3), 500,200,100, or
20 (experiment 4) or 400, 150,50, 15 (experiments 5 and 6) DRcells at day 0; 23 2 1 replicates per concentration). Stromal layers
used to perform LDA at day 0 and at day 35 after initial culture in
“stroma-contact” or “stroma-non-contact” cultures were derived
from bone marrow samples from the same allogeneic donor.
Cultures were maintained in a humidified atmosphere, at 37°C and
5% C02, and fed weekly with 75 p,L fresh media. At week 5,
nonadherent and adherent cells were collected2 and evaluated for
the presence of committed progenitors. The absolute number of
long-term bone marrow culture initiating cells (LTBMC-IC)
present in the different cell populations was calculated as the
reciprocal of the concentration of test cells that gave 37% negative
cultures using the Poisson statistics10 and the weighted mean
method.”
StatisticalAnalysis
Results of experimental points obtained from multiple experiments were reported as the mean f 1 SEM. Significance levels were
determined by two-sided Student’st-test analysis.
RESULTS AND DISCUSSION
Primitive CD34+ hematopoietic progenitors lacking
HLA-DR antigens and antigens associated with myeloid or
lymphoid lineage commitment termed DR- cells2 were
selected using FACS (Lin-CD34+DR- or DR- cells). DRcells were suspended in fetal calf serum, horse serum, and
hydrocortisone containing media but without exogenous
cytokines. Cell suspensions were plated either without
stromal layer (“stroma-free”) directly onto allogeneic irradiated stromal layers (“stroma-contact”) or in Transwellinserts that separated DR- cells from the stroma by a
0.45-+m microporous collagen-coated membrane allowing
free passage of diffusible factors but preventing cell-cell
contact (“stroma-non-contact”) (Fig 1). These translucent
Transwell inserts were placed 1 mm above the stromal
layer, which was adherent to the bottom well but remained
completely separated from the Transwell inserts throughout the culture period. Repeated visual inspection demonstrated that no adherent stromal layer was formed in
Fig 2. DR- cells suspended in IMDM supplemented with fetal calf serum, horse serum, and
hydrocortisone4were plated in either ”stroma-free“
(A) or ”stroma-non-contact” (B) cultures. Cells from
selected ”stroma-non-contact” cultures or “stromafree” cultures were collected and enumerated. Data
represent the mean f SEM cell number present in
cultures initiated with 5,000 DR- cells. Numbers in
parentheses on the x-axis represent the number of
experiments.
“STROMA-FREE ‘I
”STROMA-NON-CONTACT”
A
B
WEEKS IN CULTURE
T
CATHERINE M. VERFAlLLlE
2824
..
I
-
I.-
Fig 3. Representative example of cells recovered
at day 0 (A), week 1(B), week 3 (C), and week 5 (D) of
"stroma-non-contact" cultures initiated with FACS
sorted DR- cells. Cells were collected from the
Transwell insert of selected "stroma-non-contact"
cultures. Cytospin preparationswere Wright-Giemsastained for morphologic examination.
Table 1. Recovery of Committed Progenitors From Primitive DRCells Culture in "Stroma-free," "Stroma-contact," and
"Stroma-non-contact" Cultures
No. of Colonies
5,000 DR- Cells
Culture
CFU-GM
BFU-E
Week(n=)
CFC
Sorted DR- cells
0 (5)
66.4f 13.9
25.5 f 1.65 40.9 f 12.6
"Stroma-free"
1 (3)
2 (4)
3(2)
5 (4)
8 (2)
12.2 f 5.1
4.2 f 3.2
13.3 f 0
Of0
Of0
8.8 f 1.8
1.8 f 1.2
13.3 f 0
Of0
Of0
"Stroma-contact"
1 (4)
2 (5)
3 (2)
5 (12)
8 (2)
81.7 f 10
39.3 8.1
85.8 f 17.7 53.7 f 12.1
96.3 f 23.3 58.3 15
150 f 16.7 108 f 23
47.8 f 1.8 34.95 f 1.65
"Stromanon-contact"
1 (4) 107.5 f 15'
2 (5) 126.8 f 30'
3 (2) 165.7 f 79
1
'
t
5 (12) 333 f 4
8 (2)
99.6 f 19.6'
3.3 f 3.3
2.5 f 2.5
Of0
Of0
Of0
45.3 f 9.9
35.7 f 7.7
38.3 f 8.3
44.1 f 9.8
13.3 f 3.3
88.8 f 8.8' 25.4 f 4.2
86.4 f 28.8* 31.8 f 9.7
154 f 74
9.5 f 2.5
273 f 37*t 28.3 f 9.8,
98 f 1
8
'
t
1.6 f 1.6
DR- cells were plated in "stroma-free," "stroma-contact," or "stromanon-contact" cultures as described in Fig l. Freshly sorted DR- cells
(week 0) or cells recovered from the wells of "stroma-free" cultures,
from nonadherent and adherent fractions of "stroma-contact" cultures
after collagenase treatment, or from Transwell-inserts of "stroma-noncontact" cultures were plated in methylcellulose progenitor assays4
Colonies were enumerated at days 14 through 19 (CFC, colony-forming
cells; CFU-GM, granulocyte-macrophage colony-forming unit; BFU-E,
erythroid burst-forming unit.' Results are the mean number f SEM of
colonies obtained from 5,000 DR- cells. Values in parentheses represent the number of experiments.
' P 5 .01: Comparison between "stroma-free" and "stroma-noncontact cultures; Student's t-test.
tP 5 .01:Comparison between "stroma-contact" and "stroma-noncontact" cultures; Student's t-test.
previous studies.2When DR- cells were plated in "stromanon-contact" culture, we observed a similar increase in the
generation of committed progenitors during the first 5
weeks that decreased by week 8. These studies demonstrated that bone marrow-derived stromal cells are required for the proliferation and differentiation of primitive
hematopoietic progenitors when no exogenous cytokines
are provided. However, induction of differentiation does
not require direct contact between the hematopoietic
progenitors and the stromal layer, suggesting that diffusible
factors are released from the stromal environment that
induce differentiation of primitive progenitors placed in a
Transwell-insert 1 mm above the stromal layer. "Stromanon-contact" cultures differed from "stroma-contact" cultures in that a significantly greater number of CFU-GM
were recovered from "stroma-non-contact" cultures at
weeks 5 and 8 compared with "stroma-contact'' cultures
(Table 1). This may indicate that, in contrast to differentiation-inducing factors, negative regulators of primitive progenitorsI2J3are either not released or are present in lower
concentrations in the Transwell-insert of "stroma-noncontact" cultures. Alternatively, direct hematopoietic cellstroma interaction may be required to convey differentiationinhibiting signals. In contrast, generation of BFU-E was
slightly lower when DR- cells were separated from the
stromal layer by a microporous membrane. The mechanisms underlying this divergent effect are not clear. It has
been postulated that BFU-E proliferate in long-term bone
marrow cultures in close contact with nurturing stromaderived macrophages.' Absence of such monocytes/macrophages in the Transwell inserts, at least during the first 2
to 3 weeks of culture, may therefore be responsible for the
less-pronounced proliferation of BFU-E compared with
CFU-GM.
It has been postulated that one of the major roles of
stromal tissue may be to maintain the most primitive
progenitor compartment.14Therefore, removal of the close
cell-cell interactions between hematopoietic and stromal
cells could induce differentiation only and result in an
accelerated exhaustion of the more primitive cell pool.15To
HEMATOPOIESIS WITHOUT STROMAL CONTACT
test this possibility, in six separate experiments we compared the absolute number of primitive progenitors capable
of initiating long-term bone marrow cultures (LTBMCIC)*J6still present after culture of DR- cells for 5 weeks in
“stroma-contact” or “stroma-non-contact” cultures with
the absolute number of LTBMC-IC present in the FACSsorted DR- p o p ~ l a t i o n .As
~ ~ ~shown in Table 2, 1
LTBMC-IC per 123 2 22 sorted DR- cells was present at
day 0. When DR- cells were cultured for 5 weeks in either
“stroma-contact” culture (1 LTBMC-IC per 524 3. 89 initially sorted DR- cells; P = .001) or “stroma-non-contact”
culture (1 LTBMC-IC per 239 f 28 initially sorted DRcells; P = .009) (Table 2) and then assessed for their
primitive progenitor content, a decrease in absolute number of LTBMC-IC was observed. However, the decrease in
primitive progenitors capable of initiating long-term in vitro
hematopoiesis was significantlygreater in “stroma-contact’’
than in “stroma-non-contact” cultures (P= .009). This
indicates that direct hematopoietic cell-stroma contact is
not required to maintain a fraction of primitive progenitor
cells capable of initiating in vitro hematopoiesis. Further
studies will be needed to determine if the increased number
of LTBMC-IC recovered after 5 weeks of culture of DRcells in “stroma-non-contact” cultures is the result of a
decreased apoptosis of primitive progenitors in these cultures compared with “stroma-contact” cultures or because
of an increased proliferation of DR- cells when direct
contact with stromal layers is not allowed.
We conclude that stromal cells are required for the
conservation of primitive hematopoietic progenitors in vitro
and for their differentiation into more mature myeloid
progeny. However, direct cell-cell contact between stromal
cells and hematopoietic cells is not necessary. Several
reports have indicated that defined cytokines cannot be
detected in significant amounts in unconcentrated stromaconditioned media.17-19However, high levels of cytokines
are concentrated by glycosaminoglycans present in the
extracellular matrix (ECM) surrounding the stromal
cell^.^^^^^ Therefore, growth of hematopoietic progenitors
cultured in Transwell-inserts separated from the stroma by
a 0.45-c~mmicroporous membrane may be the result of
diffusible stroma-derived growth factors concentrated by
the ECM and likely present at high concentrations 1 mm
above the stroma. Alternatively, hematopoietic progenitors
located in the Transwell inserts of “stroma-non-contact”
cultures may interact with collagen present on the membrane or more likely with serum- or fibroblast-derived
fibronectinzl bound to the collagen-treated membrane.2z
This interaction per se may have an instructive or permissive effect on the proliferation and differentiation of human
2825
Table 2. Primitive Progenitors Are Conserved Equally Well When
Primitive DR- Cells Are Cultured in ”Stroma-contact” and
“Stroma-non-contact” Cultures
Absolute No. of LTBMC-IC
Sorted DR- Cells
”Stromanon-contact”
Experiment
Sorted DRCells
1
2
3
4
5
6
1I73
11204
11132
11102
1I68
11168
11415
1I825
11480
11303
1I600
I
11180
11251
11283
11168
11208
11349
Mean f SEM
11123 -t 22’
11524 f 89t
11239 2 28
“Stroma-contact”
We compared the absolute number of LTBMC-IC4.l0present in freshly
sorted DR- cells with the number of LTBMC-IC still present in cells
recovered after 5 weeks from “stroma-contact” and ”stroma-noncontact” cultures1l,l2 by LDA. The absolute number of LTBMC-IC
present in the different cell populations was calculated as the reciprocal
of the concentration of test cells that gave 37% negative cultures using
the Poisson statisticsz7and the weighted mean method.z8
*P = ,001 and P = ,009:Comparison between day 0 LDA and
“stroma-contact” and ”stroma-non-contact“ cultures, respectively.
t P = .009: Comparison between “stroma-contact” and ”stroma-noncontact“ cultures.
primitive hematopoietic progenitor^.^^,^^,^ In addition, glycosaminoglycans released by stroma119,20,25,26
and/or hematopoietic cellsz7may interact with the collagen-fibronectincoated membrane” and serve as a reservoir of growthpromoting factors for the hematopoietic progenitors
colocalizing on the same substratum. Surprisingly, culture
of normal primitive progenitors separated from the adherent stromal layer results in an increased generation of
committed granulocyte-macrophage progenitors and conserves primitive progenitors with long-term in vitro repopulating capacity better than culture of such progenitors in
direct contact with the stromal layer. This suggests that,
although proliferation and differentiation of hematopoietic
progenitors does not require cell-cell contact between
stromal and hematopoietic cells, the close association of
these progenitors with the stroma in the “stem-cell-nichemodel”% may be crucial for the regulation and ordered
progression of normal human hematopoiesis.
ACKNOWLEDGMENT
The author thanks K. Boland, L. Burch, J. Burroughs, M.
Hupke, and K. Myhre for their excellent technical help, and Philip
B. McGlave, MD, for his critical review of this manuscript.
REFERENCES
1. Dexter TM, Allen TD, Lajtha LG: Conditions controlling the
3. Brandt J, Srour EF, Van Besien K, Bride11 RA,Hoffman R:
proliferation of hemopoietic stem cells in vitro. J Cell Physiol
Cytokine-dependent long-term culture of highly enriched precur91:344,1976
sors of hematopoietic progenitor cells from human bone marrow. J
2. Verfaillie C, Blakholmer K, McGlave P: Purified primitive
Clin Invest 86:932,1990
human hematopoietic progenitors with long term in vitro repopulat4. Bentley SA: Close range cell-cell interactions required for
ing capacity adhere selectively to irradiated bone marrow stroma. J
stem cell maintenance in continuous bone marrow culture. Exp
Exp Med 172509,1990
Hematol9:308, 1981
2826
5. Coloumbel L, Eaves AC, Eaves CJ:Enzymatic treatment of
long-term marrow cultures reveals the preferential location of
primitive hematopoietic progenitors in the adherent layer. Blood
62291,1983
6. Verfaillie CM, Miller WJ, Boylan K, McGlave PB: Selection
of benign primitive hematopoietic progenitors in chronic myelogenous leukemia on the basis of HLA-DR antigen expression. Blood
79:1003,1992
7. Verfaillie CM, McGlave PB: Leukemia inhibitory factor/
human interleukin for DA cells: A growth factor that stimulates
the in vitro development of multipotential human hematopoietic
progenitors. Blood 77263,1991
8. Sutherland HJ, Eaves CJ,Landsdorp PM, Thacker JD, Hogge
D E Differential regulation of primitive human hematopoietic cells
in long-term cultures maintained on genetically engineered murine
stromal cells. Blood 78:666,1991
9. Sutherland HJ, Landsdorp PM, Henkelman DH, Eaves AC,
Eaves CJ: Functional characterization of individual human hematopoietic stem cells cultured at limiting dilution on supportive
marrow stromal layers. Proc Natl Acad Sci USA 87:2584,1990
10. Porter EH, Berry RJ: The efficient design of transplantable
tumor assays. J Cancer 17583,1963
11. Taswell C: Limiting dilution assays for the determination of
immunocompetent cell frequencies: I Data analysis. J Immunol
126:1614,1981
12. Cashman JD, Eaves AC, Raines EW, Ross R, Eaves CJ:
Mechanisms that regulate the cell cycle status of very primitive
hematopoietic cells in long term human marrow cultures. I.
Stimulatory role of a variety of mesenchymal cell activators and
inhibitory role of TGF-b. Blood 75:89,1990
13. Graham GJ, Wright EG, Hewick R, Wolpe SD, Wilkie NM,
Donaldson D, Lorimore S, Pragnell IB: Identification and characterization of an inhibitor of hematopoietic stem cell proliferation.
Nature 344442,1990
14. Zipori D, Sasson T: Adherent cells from mouse bone
marrow inhibit the formation of colony stimulating factor (CSF)
induced myeloid colonies. Exp Hematol6:816,1980
15. Siminovich L, Till JE, McCulloch E A Decline in colonyforming ability of marrow cells subjected to serial transplantation
into irradiated mice. J Cell Comp Physiol64:23, 1964
16. Sutherland HJ, Eaves CJ, Eaves AC, Dragowska W, Lands-
CATHERINE M. VERFAlLLlE
dorp PM: Characterization and partial purification of human
marrow cells capable of initiating long-term hematopoiesis in vitro.
Blood 74:1563,1989
17. Toogood IRG, Dexter TM, Allen TD, Suda T, Lajtha LG:
The development of a liquid culture system for the growth of
human bone marrow. Leuk Res 4449,1980
18. Caldwell J, Palsson BO, h e y B, Emerson SG: Culture
perfusion schedules influence the metabolic activity and growth
factor production rates of human bone marrow stromal cells. J Cell
Physiol147344,1991
19. Gordon MY, Riley GO, Watt SM, Greaves M F Compartmentalization of a hematopoietic growth factor (GM-CSF) by
glycosaminoglycansin the bone marrow microenvironment. Nature
326:403,1987
20. Roberts R, Gallagher J, Spooncer E, Allen TD, Bloomfield
F, Dexter TM: Heparan sulphate bound growth factors: A mechanism for stromal cell mediated hemopoiesis. Nature 322376, 1988
21. Verfaillie CM, McCarthy JB, McGlave PB: Differentiation
of primitive human hematopoietic progenitors into dingle lineage
clonogenic progenitors is accompanied by alterations in their
interaction with fibronectin. J Exp Med 174:693,1991
22. Ruoslahti E Fibronectin and its receptors. Annu Rev
Biochem 57:375,1988
23. Dike LE, Farmer SF: Cell adhesion induces expression of
growth-associated genes in suspension arrested fibroblasts. Proc
Natl Acad Sci USA 58:6792,1988
24. Weinstein R, Riordan MA, Wemc K, Kreczko S, Shou M,
Daniak N Dual role of fibronectin in hematopoietic differentiation. Blood 73:111,1989
25. Gallagher JT, Spooncer E, Dexter TM: Role of the cellular
matrix in hematopoiesis. I. Synthesis of glycosaminoglycans by
mouse bone marrow cell cultures. J Cell Sci 63:155,1983
26. Wight TN, Kinsella MG, Keating A, Singer JW: Proteoglycans in human long-term bone marrow cultures: Biochemical and
ultrastructural analysis. Blood 67:1333,1986
27. Morris AJ, Dexter TM, Gallagher J T Metabolic properties
of a homogeneous proteoglycan of a hematopoietic stem cell line,
FDCP-MIX. Biochem J 260479,1989
28. Shofield R: The relationship between the spleen colonyforming cell and the hematopoietic stem cell: A hypothesis. Blood
Cells 4:7, 1978