Calcif Tissue Int (1992) 51:195-201

Calcified Tissue International 9 1992 Springer-Vedag New York Inc.

Laboratory Investigations Hemopoietic Functions of Marrow-Derived Osteogenic Cells D. Benayahu, 1 M. Horowitz, 2 D. Zipori, 3 and S. Wientroub 1'4 1Department of Histology and Cell Biology, Sackler Faculty of Medicine, Tel-Aviv University, Ramat-Aviv 69978, Israel; 2Department of Orthopaedics, School of Medicine, Yale University, New Haven, Connecticut, USA; 3Department of Cell Biology, Weizmann Institute of Science, Rehovot, Israel; and 4Department of Pediatric Orthopaedics, Tel-Aviv Medical Center, Tel-Aviv, Israel Received December 24, 1991, and in revised form April 6, 1992

Summary. Osteoblasts, members of the marrow stromal cellular network, may play an active role in the hemopoietic microenvironment as well as in bone remodeling. In this study, we examined the extent to which marrow-derived osteogenic cells (MBA-15) possess various stromal functions. This marrow stromal-derived cell line was shown by us to exhibit osteoblastic characteristics in culture and to form bone in vivo. These cells are shown here to constitutively produce and secrete cytokines identified as M-CSF, GMCSF, and IL-6. MBA-15 cells modulate growth of normal and malignant myeloid and lymphoid cells as well as leukemia cell lines in vitro. Cell-cell interactions were studied in co-cultures with adherent MBA-15 cells and the target hemopoietic cells. Growth inhibition effects, observed under various experimental conditions, can be attributed to the presence of different soluble and membrane-bound inhibitory activities produced by MBA-15 cells. Thus, MBA-15 cells spontaneously produce both stimulators and inhibitors that can affect myeloid and lymphoid cell growth. Marrow osteogenic cells may therefore participate in the stromal regulation of hemopoiesis.

Key words: Marrow stroma - Stromal cell lines - Osteoblasts - Cytokines - MBA-15.

Osteoblasts are known to be a major contributor to the process of bone formation. They are localized at specific sites and make complex physical associations with cells and extracellular matrix components at the bone/bone marrow interface. The in vivo structure-function relationship in the bone marrow is still not completely understood, due to the large number of concomitant cell interactions. Ultrastructure studies revealed close physical association of stromal cells with hemopoietic stem and progenitor cells [1-3]. The stromal cell system is composed of a heterogenous population which differs in morphology and function. Members of the marrow stromal cellular network, including osteoblasts [4, 5], may play an active role in the regulation of hemopoiesis. Cultured stromal cells are capable of forming an in vitro milieu that resembles the organized hemopoietic Offprint requests to: S. Wientroub at the Sackler Faculty of Medicine

structures in many of its functional aspects [6-12]. Stromadependent hemopoiesis [13] is characterized by several unique features: induction of stem cell renewal, restriction of the activity of differentiation-inducing cytokines and the consequent reduction in the yield of mature cells, and secretion of stimulators and inhibitors of hemopoiesis and expression of cell surface-associated restricting molecules. This study was conducted to investigate the possible relationship between stroma-derived osteoblasts and the hemopoietic compartment of the marrow. The hemopoietic functions of bone-forming cells were investigated using a unique marrow stromal-derived osteoblastic cell line, MBA15 r1 l]. This is the first osteogenic cell line of marrow origin that possesses osteoblastic phenotype in vitro and forms bone when implanted in diffusion chambers in vivo. The osteoblastic markers exhibited by the MBA-15 cell line include production of type I collagen solely, alkaline phosphatase activity enhanced when cells are exposed to dexamethasone, responsiveness to parathyroid hormone (PTH) and prostaglandin Ez (PGE2), and ability to mineralize extracellular matrix in vitro [11, 12].

Materials and Methods Stromal Cell Lines

MBA-15 is an osteoblastic marrow stromal cell line of SJL/J mouse strain [9-12]. 14Fl.1 is a cloned adipocyte stromal cell line of BALB/C mouse strain ~9, 10]. Cells were grown as adherent monolayer and maintained by weekly passage in Dulbecco's modified Eagle's medium (DMEM) (Beth Haemek, Israel) supplemented with 10% fetal calf serum (FCS) (Bio-Lab, Jerusalem, Israel). Factor-dependent Cell Lines

The experiments aimed to detect interleukin-like activity were performed using the following: 14M1.4 marrow-derived clonal macrophage cell line tested with mouse and human granulocyte macrophage-colony stimulating factor (GM-CSF), mouse IL-3, and mouse G-CSF and were found to proliferate only in response to L-cell CM or purified M-CSF. This activity can be neutralized by antibodies [9], and (D. Zipori, unpublished data); BAC cell line which proliferates in response to M-CSF [t4]; 32D cells which respond to G-CSF and IL-3 [15]; LBRM-33-1A5 stimulated in the presence of IL-1 [16]; HT-2 T-cells known to respond to IL-2, IL-4, and GMCSF; CTLL-2 T-cells which respond to IL-2 and a high concentration of IL-4; CTLD-20 T-cells known to respond to IL-2 [17]; MC9 mast cells which proliferate in the presence of IL-3 [18] and

196

D. Benayahu et al.: Marrow Osteoblasts and Hemopoiesis

Medl~ Removad

)|

|

Liquid Medium Culture

Semisolid Culture

I

f2 Methylcellulose (M.C.)

fl

Methyloellulose (M.C.) + agar

12

Short-term 2,4,6,8 days

B.M. in M C . . . ~

Long-term 1,2,3,4 weeks

i3 Cell Lines

B.M. in Medium

agar 8 days ~'t'

i

Methyleeltulose ~ + CSF I['zl,n'J' z .,~ /

ii 8 days

Colonies Counted

i.pyqlr162

, Fig. 1. Experimental design.

Colonies Counted

B13.29(B9); and cloned B-cells which respond specifically to IL-6 [19].

Leukemia Cell Lines These cell lines include ABLS-8.1, a pre-B lymphoma, WEHI-231 B-cell lymphoma, MPC-11 plasmacytoma, and ST 1.3 murine leukemia virus-induced T-lymphoma. WEHI-3 is a myeloid tumor, P815 a mastocytoma, and Friend 745 cloned cell line is an erythroleukemia. Most of the established cell lines were cultured in DMEM supplemented with 10% FCS, incubated at 37~ in a humidified atmosphere of 10% CO2/90% air and serially diluted weekly at 103104 cells/ml [20].

Antibodies The presence of IL-6 in CM was confirmed by the addition of neutralizing rat antimouse monoclonal antibodies to the B9 cell assay (a gift from Dr. J. Feyen, Sandoz, Switzerland) [21] used at a final concentration of 10 Ixg/ml. GM-CSF was neutralized with goat antimouse GM-CSF (1:1000) added to the HT-2 cell assay.

Conditioned Medium (CM) MBA-15 cells were seeded at low density (1.5 x 105 cells) into 60-mm tissue culture dishes (Nunc, Copenhagen, Denmark) in DMEM with 10% FCS and grown to confluence. At indicated intervals the medium was removed and fresh medium was added to the culture dishes. The collected CM was centrifuged, millipore filtered (0.45 I~m), and stored in small aliquots at -20~ until use. In some experiments, CM was collected in the presence of lipopolysaccharide (LPS) (Difco, Detroit, MI) at a final concentration of 1 ~g/ml for 3 days.

Assays for Cell Growth and Viability Single cell suspensions were prepared from thymus of BALB/C mice and tested by measuring their proliferative response to MBA15 CM. Factor-dependent cell lines were harvested from mainte-

nance cultures and washed three times in phosphate-buffered saline (PBS), diluted to appropriate cell density, and plated in 96-well plates (Nunc). Cell suspension and tested samples of CM at various dilutions were mixed in wells and incubated in triplicate cultures for 48--96 hours at 37~ in a 95% air/5% CO 2 incubator. Growth of cells was determined using the colorimetric MTT assay [22] or cultures pulsed with [3H]thymidine (1 IxCi/well; specific activity (SA), 6--8 Ci/mmol; New England Nuclear, Boston, MA) harvested with an automatic sample harvester (Cambridge Technologies, Cambridge, MA). Radioactivity was measured with a Beckman scintillation spectrometer (Fullerton, CA). All experiments were repeated at least three times and the results were expressed as the mean of triplicate cultures. In all experiments, positive controls, including the required cytokine, were used.

Granulocyte/Macrophage (G/M) Colonies Bone marrow cell (BMC) suspensions were adjusted to a concentration of 105 cells/ml in DMEM supplemented with 20% FCS. MBA-15 CM or L-cells CM [9] were used as a source of CSF, as indicated. Cells were seeded in 0.8% (W/V) methylcellulose or 0.5% (W/V) Bacto agar (Difco, Detroit, MI). Aliquots (1 ml) of these suspensions were inoculated into 30 mm culture dishes (triplicate cultures) and then incubated for 8 days. Myeloid colonies were counted by an inverted microscope. G/M colonies were distinguished by their colony morphology appearance, cell size, and nuclear shape, using an orcein staining technique.

Co-cultures Formation of G/M Colonies in the Presence of MBA-15 Adherent Layer. BMC (105/ml) were seeded as described above ("methylcellulose cultures") into 30 mm culture dishes (Fig. 1A). The same number of cells was also seeded in 1.0 ml of this medium onto a preformed agar layer composed of DMEM supplemented with 10% FCS, CM as indicated, and 0.5% (W/V) Bacto agar ("agar cultures", Fig. 1Az). In both cases, seeding was performed onto confluent layers of MBA-15 cells or in empty control plates. The accumulation of myeloid progenitor cells (CFU-GM) in these cultures was determined on day 8 [9].

D. Benayahu et al.: Marrow Osteoblasts and Hemopoiesis Table 1. Myelopoietic activities of MBA-15 cells detected in coculture system No. of myeloid colonies

% CM

(1) Semisolid cultures

(2) Methylcellulose cultures

(3) Agar cultures

0 5 15 30

0 63 --- 7 41 -+ 4a 31 --- 26

0 0 0 0

38,5 -+ 4.0 6.7 -+ 1.0a 3.5 -+ 0.3a 0

(1) Colonies formed in semisolid culture in presence of MBA-15 CM only (2) BMC seeded directly onto MBA-15 feeder layer with or without CM, as indicated (methylcellulose cultures) (3) BMC in methylcellulose cultures plated over a thin agar layer covering the MBA-15 feeder layer (agar cultures). For more details see Materials and Methods For significant difference from control (P < 0.005). Values were obtained with the unpaired Student's t-test

Recovery of CFU-GMfrom Short and Long-term Liquid Bone Marrow Cultures. BMCs (105/ml) suspended in DMEM supplemented with 20% FCS were seeded on top of preformed adherent cell layers (1 ml/plate) or in control cultures without adherent cells. Following various incubation intervals as indicated (Fig. 1B1,2), 0.8-ml aliquots were collected from cultures, diluted in a final volume of semisolid medium comprised of DMEM, supplemented with 20% FCS, 30% L-cell CM, and 0.8% (W/V) methylcellulose; 1.0 ml of the suspensions were then inoculated into 30-mm cultures dishes. G/M colonies were counted on day 8. In Vitro Growth of Tumor Cells (Fig. 1B3). The growth medium was removed from the preformed adherent layer and replaced by DMEM, supplemented with 10% FCS and 50 I~M 2-mercaptoethanol (if required by the cell line) containing the tumor cell suspension. Growth of tumor cells was measured by cell counting using a Coulter Counter. Those lower than 0.5 • 104 were not included because they were below the threshold at which accuracy could be obtained by the counting methods used. Average cell counts of 2-4 cultures are shown.

Results

Studies of hemopoietic regulation revealed two major systems. One involved C S F and other cytokines, and the other presumably cell-bound activities [13]. Thus, we explored the participation of marrow stromal-derived osteoblasts (MBA15) in these systems [4, 5, 11, 12].

Modulation of Myelopoiesis by Co-culture of BM with MBA-15 Cells MBA-15 cells were examined for their ability to support differentiation of myeloid progenitors. An inoculum of 105 BMC in methylcellulose medium was seeded in the presence or absence of the adherent layer of MBA-15 cells (Fig. 1A). This adherent cell layer did not support colony formation (Table 1). MBA-15 CM added to these cultures was also incapable of inducing G/M colonies (Table 1). Induction was observed only when a thin agar layer separated the adherent and the target marrow cells (Table 1). When the agar layer contained various concentrations of CSF, formation of G/M

197 colonies was lower in the presence of MBA-15 (Table 1) compared with culture without stroma (Table 1). This inhibitory effect on G/M colony formation requires the adherent MBA-15 cells to be in close vicinity to the C F U - G M (Table 1). Colony-stimulating activity of MBA-15 cells seemed to be present in a soluble form, at least partially, as it could pass thin agar barriers. Similar activity was demonstrated also in the CM, as depicted in Table 1. Short-term bone marrow cultures result in rapid death of progenitor cells. Liquid culture experiments were designed to determine the ability of MBA-15 adherent layer to affect the in vitro survival of myeloid progenitors. This function was previously described for other stromal cell type (Fig. 1B1). Incubation of marrow cells on top of MBA-15 adherent layer did not result in survival of myeloid progenitors, regardless of the addition of CSF (L-cells CM). Long-term bone marrow co-culture is the most common method used in the study of hemopoiesis and in vitro stem cell renewal. We examined the ability of MBA-15 cells to support the continuous proliferation of myeloid progenitors using this assay (Fig. 1B2). The endothelial-adipocytic clonal cell line 14Fl.1 known for its ability to support stem cells, myeloid progenitors, as well as pre-B and pre-T lymphocytes, was used in the control cultures [20]. BMC were plated at a density of 2.5 • 105/culture. Two and 3 weeks later, the average number of cells in liquid phase of culture was determined. MBA-15 and 14Fl.1 cultures had similar numbers of cells. CFU-GM assays performed using these cells revealed myeloid progenitors among the cell population maintained by 14Fl.1 cells, but not by MBA-15 osteoblasts (data not shown). It can be concluded that MBA-15 does not support long-term hemopoiesis.

Modulation of Tumor Cell Line Proliferation by MBA-15 Cells This part of the study involves a series of tumor cell lines belonging to specific lineages and stages of hemopoietic differentiation. The tumor cells were co-cultured with a layer of MBA-15 cells, or with 14FI.1 (as control) in liquid culture (Fig. 1B3). The effects on tumor cell growth are quantified and summarized in Figure 2. The results indicate that some of the cell lines (Friend, W E H I , ST 1.3) were strongly inhibited by MBA-15, whereas others were only partly affected. No inhibition was noted in 14F1.1 cultures.

In Vitro Growth Modulation of Myeloid and Lymphoid Cells by MBA-15 CM The following experiments were designed to test whether the activities of MBA-15 are secreted by the cells. CM from MBA-15 cells induced the formation of G/M colonies. The maximal induction was observed after a conditioning period of 6 days. A longer conditioning period caused a decrease in colony-stimulating activity (Fig. 3). The highest number of G/M colonies was noted when 5% CM was added to the growth medium. A higher concentration of CM was inhibitory and the number of colonies d e c r e a s e d in a dosedependent manner. Based on morphological criteria, the colonies formed in agar cultures were 52.5% macrophages, 19.6% g r a n u l o c y t e s , and 27.9% m i x e d g r a n u l o c y t e / macrophage colonies. Thus, the ability of MBA-15 CM to support G/M colony formation suggests that MBA-15 cells produce CSF and other cytokines.

D. Benayahu et al.: Marrow Osteoblasts and Hemopoiesis

198 ]-

14 F I.I

r~-

MBA ~ 15

:i~ii! ;.:.3 ii!~ !il);; iiilii i:?!: !~!!ii

I00 -- ,......, I . . . . . . . . I??i:iI

i!i!)

!iliii

I:::,.::;~ I:::::::1

::::: I iilil

II ;ii

o

iiiii

i:!:i:

i:?!: ?ilil

::::::

ililii iii!

1::::::

i::i::i::

~)i:

iiili

!ili!i

;2

!ii!ii:i|

iii:

o Fr*end Erythroleukemia

P-815 Moltor

WENt-3

ST4b ST 13 L - T - Lymphomo-J

Myeloid

MPC-II Plalmocytomo

100 "

0

o ~o60 6 I

40

20

D a y s

0

B Lymphomo

Discussion ---q

~

ABLS~8 WEHI-231 B Lymphoma prl-

Fig. 2. A variety of leukemia cell lines were co-cultured in the presence of MBA-15 or 14Fl.1 (control) cells. Their growth was determined following 7 days incubation using MTT assay. Data of 1/3 representative experiments are shown. Reading difference did not exceed 10%. Each column represents the average of triplicate cultures.

(IL-2), CTLL-2 (IL-2, IL-4), and MC-9 (IL-3) for detection of IL-1, IL-2, IL-3, and IL-4 in the CM harvested from MBA-15, with or without LPS activation.

8O _o

iiiili

:i:~: :::::: !:i:i: :i:i: :::::: i;i;i:

(.9 O--

ili!::i

I~j2S

.

,

5

9

T

"

,

'

I

15

"

I

"

",

30

C . M . in Culture (%)

Fig. 3. Myeloid colony formation induced by MBA-15 CM. The CM was harvested at different time intervals after confluence (days). Data of 1/10 representative experiments are shown. Results are mean of 3-5 methylcellulose cultures per point.

Cytokine Analysis in MBA-15 CM The MBA-15 CM was tested for the ability to induce proliferation of hemopoietic cells; BMC (Fig. 4a), and thymocyte proliferation (Fig. 5) were stimulated in a dose-dependent manner. Further studies enable us to define and characterize the activities as M-CSF by BAC cells (Fig. 4b) and 14M1.4 cell line (Fig. 6). MBA-15 CM contain IL-6 activity, as detected by B9 assay (Figs. 4c, 7a). This activity was neutralized by antibodies to IL-6, and B9 cell proliferation was reduced to background levels (Fig. 7a). MBA-15 cells secrete to the CM low levels of GM-CSF, whereas activation with LPS caused a significant increase, as detected by proliferation of HT-2 cells (Figs. 4d, 7b). This activity was neutralized by anti-GM-CSF in the HT-2 assay (Fig. 7b). Negative results were obtained using the following cell lines: 32-D (G-CSF, IL-3), LBRM-33-1A5 (IL-1), CTLD-20

Studies of hemopoietic organs reveal cell arrangement in defined compartments delineated by stromal cells. Stem cells and progenitors are mostly found in discrete sites, tightly packed in an organized tissue. Stromal cells physically separate hemopoietic cells from the bloodstream, and intimate cell-cell interactions occur between various stromal components and hemopoietic cells. In the present study, we examined the extent to which marrow-derived osteoblastic cells (MBA-15) share various stromal functions, and directly or indirectly mediate control of hemopoiesis. Our data show that these cells can affect the in vitro growth of normal and malignant hemopoietic cell lines. CM collected from MBA15 cells contains mitogenic activity for marrow and thymus cells. MBA-15 cells constitutively produce and release cytokines that induce various types of myeloid colonies. These cytokines were identified as M-CSF, IL-6 and, GM-CSF. The latter was mainly detected following LPS stimulation. Horowitz et al. [23] and Felix et al. [24] have shown that calvarial osteoblasts do not secrete GM-CSF constitutively and that PTH or LPS stimulation is required. CM from unstimulated ROS 17/2.8 cells [25] and from calvaria cells [26] enhanced proliferation of HT-2 cell line. This mitogenic activity was indistinguishable from GM-CSF. M-CSF was the only cytokine released constitutively from isolated calvarial cells and MC3T3-E1 cell-line and it was significantly increased in the presence of LPS or 1,25(OH)2D3 [27], rlL-lct, tumor necrosis factor (rTNFa), or by concurrent presence of transforming growth factor-13 (TGFI3) and epidermal growth factor (EGF) [28]. In the same cell line, GM-CSF was secreted after LPS stimulation and to a lesser extent PTH [29]. 1,25(OH)2D3 failed to induce these cells to secrete GM-CSF [29]. In our study, marrow-derived MBA-15 cells were shown

D. Benayahu et al.: Marrow Osteoblasts and Hemopoiesis

199

BM

75000'

(cylO~ines)

180000'

BAC (M-CSF)

140000

50000

1~ 0 0

6OOO0

E

~'0

25000

O.

,'~

2'0

;5

20OO0

4

6

8

10

b

a tO 0 0 o C

rr "o F--r

8000o

Bg (IL-6) 40000'

HT-2 {GM-CFS)

6oooo

L

30000 84

40000

20000

20000

10000

/ 0

o!,

oi~

'

110

"

'

~2 $

25

d

C

Conditioned Medium (%) Fig. 4. Growth of factor-dependent cells with MBA-15 CM. The CM harvested from confluent MBA-15 (9 and from LPS-treated (0) cultures; examined for the ability to stimulate the proliferation of different cell populations. The CM used in concentrations indicated. Results are the mean -+SD of triple dishes for each point. The upper left comer indicates the cell line used and the factor on which its growth is dependent. (a) MBA-15 CM (6-25%) was incubated with 105 BM cells/well for 72 hours. Proliferation was measured by the incorporation of 3H-thymidine added during the last 16 hours. (b) BAC 1.2F5 cells respond specifically to M-CSF. Co-culture of MBA15 CM (2.5-10%) with 10 4 BAC1.2F5 cells/well for 72 hours. Prolif-

eration was measured by the incorporation of 3H-thymidine added during the last 16 hours. (c) B9 cells respond specifically to IL-6. Dilutions of MBA-15 CM (0.01-1%) were co-cultured with 5 • 103 B9 cells/well for 72 hours. Proliferation was measured by the incorporation of 3H-thymidine added during the last 4 hours. (d) HT-2 cells respond specifically to GM-CSF. This activity was assayed by co-culture of MBA-15 (6--25%) with 2 x 104 HT-2 cells/well for 48 hours. Proliferation was measured by the incorporation of 3Hthymidine added during the last 4 hours. LPS induced significantly more activity.

to secrete M - C S F constitutively and to increase GM-CSF secretion following LPS activation. These results are in contrast to those of primary calvarial cells [23]. The data indicate that osteoblasts do not respond uniformly. There is a differential response to osteotropic agents as measured by cytokine secretion. With regard to M-CSF secretion, it appears that osteoblasts from different sources can respond differentially to the same stimulation/activation signal and that a single osteoblast population can respond differentially to independent regulatory molecules. This may imply that in vivo cells with osteoblastic phenotype do not secrete CSFs continually but rather in response to specific signals. The latter differ at the various anatomic locations and/or at different stages of differentiation. Marrow stromal osteoblasts (MBA-15) constitutively produce and release large amounts of IL-6 to the medium.

IL-6 is produced by various types of lymphoid and nonlymphoid cells, and promotes the growth of phytohemagglutinin (PHA)-stimulated thymocytes and peripheral T-cells [30, 31] similar to the results obtained with MBA-15 CM. It is well known that IL-6 and IL-3 act synergistically to activate hematopoietic stem cells. IL-6 acts on the multipotent progenitors but not on the more mature progenitors [32, 33]. The result may suggest a direct effect of marrow stromal IL-6 secretion by the osteoblasts on hematopoiesis. The possible involvement of IL-6 in bone metabolism was examined by exposing MC3T3-EI and primary mouse calvarial osteoblast cells to local bone-resorbing agents [34]. In agreement with the expression of IL-6 mRNA, biologically active IL-6 was produced by MC3T3-E1 cells in response to treatment with ILlet, T N F a , and LPS. IL-6 stimulated 45Ca release from fetal mouse calvaria, and resulted in

D. Benayahu et al.: Marrow Osteoblasts and Hemopoiesis

200 0,3"

0.3

Thymus cells conA

1 4 M 1.4

M- CSF

PHA E

E

c O tO '-..O C~ LI3 tt3

O LO

0.2

's O

Lt3 Lt3

d O. 1

0.1

:iii::

i~

Fig. 5. Thymus cells induced by Con A and PHA to proliferate in the presence of CM-harvested MBA-15 (5-20%) were measured using MTT colorimetric assay. Results of 1/3 representative experiments are shown.

~iill

d d

O

iii!!;

20

10

5

20

10

25 1 2 5 6 2 ~ 3 1 5

5

6

Conditioned Medium (%)

10000-

B9

)

~ 80O0"

~6

Conditioned Medium (%)

Fig. 6. 14M1.4 cells respond specifically to M-CSF. Various concentrations of MBA-15 CM were co-cultured 103 cell/well 14M1.4 for 96 hours. Proliferation was measured by MTT colorimetric assay. Results of 1/3 representative experiments are shown.

250O0-

NT-2

200O0 I/I/// //I///

g

/1"I//~ III/I/ I///// ////IJ

15000-

8 c

IIIII/

I

4000-

2 I

2000

co

0

. . . . . . Medium

CM

LPS CM

v////,

10000

I, I

II//I/ IIIIII I l l l l t Iit11/

5000 -

I

II CM

0

LPS CM + anti IL 6

,'///// Medium

LPS CM

CM I

I

a

CM

+ anti

LPS CM GM-CSF I

b

Fig. 7. Cytokines present in MBA-15 CM. (a) The presence of IL-6 in MBA-15 CM was confirmed by the addition of monoclonal neutralizing anti-IL-6 antibody (anti:IL-6) in a final concentration of 10 p.l/ml to B9 cell assay. Experiments were performed with 0.3% CM harvested with or without LPS induction. (b) The presence of GM-

CSF in MBA-15 CM was confirmed by addition of neutralizing polyclonal anti-GM-CSF (1:1000) antibodies (anti-GM-CSF) to the HT-2 cell assay. Experiments were performed with 25% CM harvested with or without LPS induction.

three times more osteoclasts in bone calvaria cultured with IL-6. MBA-15 secrete IL-6 constitutively whereas other osteoblast-like ceils secrete the cytokine in response to local bone resorbing agents, indicating their possible role in inducing bone resorption. The typographical association between hemopoietic tissue and b o n e i m p l i e s m u t u a l d e p e n d e n c e and crossregulation. In addition to the existence of secretory cytokine activity by MBA-15 cells, cell-cell interactions were studied in co-culture systems. When sparse BMC suspensions were seeded in semisolid medium supplemented with or without MBA-15 CM onto confluent MBA-15 cell layer, no myeloid colonies were formed. When agar barrier separated between the adherent MBA-15 cells and the target myeloid cells, the inhibiting activities were significantly reduced. This differentiation-restraining activity of marrow-derived osteoblasts was also observed in experimental systems when high concentrations of CM were used. Therefore, we may conclude that MBA-15 stromal cells antagonize the activity of cyto-

kines and other inducers in the medium and create conditions favorable for limited differentiation only. The inhibition effects observed under various sets of experimental conditions can be at least partially attributed to the presence of different inhibitory molecules in each setting. Some of these molecules are soluble and secreted to the medium and others may be membrane-bound and found on the cell surfaces. In vitro studies have described inhibitory activity of prostaglandin E, known to be produced by osteoblasts, on the proliferation of G/M progenitor cells [35, 36]. In our study, indomethacin did not affect the inhibitory activity of MBA-15 CM (Benayahu and Wientroub, unpublished observation). While studying the interactions between malignant cell lines and marrow stromal-derived osteoblasts, we noticed inhibited growth of various tumor cell lines. These tumor cells represent a variety of hemopoietic lineages and are responsive to signaling molecules from tissues. All tumoral cell lines tested showed sensitivity and reduced growth. Thus, both stimulators and inhibitors involved in myelopoietic and

D. Benayahu et al.: Marrow Osteoblasts and Hemopoiesis lymphopoietic modulation are spontaneously and constitutively secreted by marrow stromal osteoblasts.

Acknowledgments. This work was supported in part by grants to S.W. from the Basic Research Foundation, the Israel Academy of Science and Humanities, the Israel Cancer Association, Kessler Foundation, and the Moise and Frida Eskenasy Institute for Cancer Research, Tel-Aviv University. This study is part of the Ph.D. thesis of D.B. carried out at the Department of Histology and Cell Biology, Sackler Faculty of Medicine, Tel-Aviv University.

References 1. Allen TD, Dexter TM (1983) Long-term bone marrow cultures: an ultrastructural review. Scan Electron Microscop 4:1851-1866 2. Lambertsen RH, Weiss L (1984) A model ofintramedullary haemopoietic microenvironments based on sterologic study of the distribution of endoclonal marrow colonies. Blood 63:287-297 3. Marshall A, Lichtman MD (1981) The ultrastructure of the hemopoietic environment of the marrow: a review. Exp Hematol 9:391-410 4. Owen ME, Friedenstein AJ (1988) Stromal stem cells: marrowderived osteogenic precursors. In: Chichester (ed) Cell and molecular biology of vertebrate hard tissues. Ciba Foundation Symposium, 136 Wiley, pp 42--60 5. Owen ME (1985) Lineage of osteogenic cells and their relationship to the stromal system. In: Peck WA (ed) Bone and mineral research, vol 3. Elsevier, Amsterdam, New York, Oxford, pp 1-25 6. Metcalf D (1989) The molecular control of cell division, differentiation, commitment and maturation in haemopoietic cells. Nature 339:27-30 7. Zipori D (1989) Stromal cells from the bone marrow. Evidence for a restrictive role in regulation of hemopoiesis. Eur J Haematol 42:225-232 8. Dexter TM, Testa NG, Allen TD, Rutherford T, Scolnick E (1981) Molecular and cell biology aspects of erythropoiesis in long-term bone marrow cultures. Blood 58:699-707 9. Zipori D, Friedman A, Tamir M, Silverberg D, Malik Z (1984) Cultured mouse marrow cell lines: interactions between fibroblastoid cells and monocytes. J Cell Physiol 118:143-152 10. Zipori D, Duksin D, Tamir M, Argaman A, Toledo J, Malik Z (1985) Cultured mouse marrow stromal cell lines II: distinct subtypes differing in morphology, collagen types, myelopoietic factors and leukemic cell growth modulation activities. J Cell Physiol 122:81-90 11. Benayahu D, Kletter J, Zipori D, Wientroub S (1989) Bone marrow-derived stromal cell line expressing osteoblastic phenotype in vitro and osteogenic capacity in vivo. J Cell Physiol 140:1-7 12. Benayahu D, Fried A, Zipori D, Wientroub S (1991) Subpopulations of marrow stromal cells share a variety of osteoblastic markers. Calcif Tissue Int 49:202-207 13. Zipori D (1990) Regulation of hemopoiesis by cytokines that restrict options for growth and differentiation. Cancer Cells 2(7):205-211 14. Morgan C, Pollard JW, Stanley ER (1987) Isolation and characterization of a cloned growth factor-dependent macrophage cell line BAC1.2F5. J Cell Physiol 130:420--427 15. Greenberger JS, Sakakeeny MA, Humphries RK, Eaves CJ, Eckner RJ (1983) Demonstration of permanent factordependent multipotential (erythroid/neutrophil/basophil) hematopoietic progenitor cell lines. Proc Natl Acad Sci 80:2931-2935 16. Conlon PJ (1983) A rapid biological assay for the detection of interleukin 1. J Immunol 131(3):1280-1282 17. Watson J (1979) Continuous proliferation of murine antigen specific helper T lymphocytes in culture. J Exp Med 150:1510-1519 18. Nabel G, Galli SJ, Dvorak AM, Dvorak HF, Cantor H (1981)

201 Inducer T lymphocytes synthesize a factor that stimulates proliferation of cloned mast cells. Nature 291:332-334 19. Aarden LA, DeGroot ER, Schaap OL, Lansdorp PM (1987) Production of hybridoma growth factor by human monocytes. Eur J Immunol 17:1411-1416 20. Zipori D, Tamir M, Toledo J, Oren T (1986) Differentiation stage and lineage-specific inhibitor from stroma of mouse bone marrow that restricts lymphoma cell growth. Proc Natl Sci 83: 4547-4551 21. Feyen JHM, Elford P, Di Padova FE, Trechsel U (1989) Interleukin-6 is produced by bone and modulated by parathyroid hormone. J Bone Miner Res 4(4):633--638 22. Mosmann T (1983) Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Method 65:55--63 23. Horowitz MC, Einhorn TA, Philbrick W, Jilka RL (1989a) Functional and molecular changes in colony-stimulating factor secretion by osteoblasts. Connect Tissue Res 20:159-168 24. Felix R, Elford PR, Stoerckle C, Cecchini M, Wetterwald A, Trechsel U, Fleisch H, Stadler BM (1988) Production of hemopoietic growth factors by bone tissue and bone cells in culture. J Bone Miner Res 3(1):27-36 25. Weir EC, Insogna KL, Horowitz MC (1989) Osteoblast-like cells secrete granulocyte-macrophage colony-stimulating factor in response to parathyroid hormone and lipopolysaccharide. Endocrinology 124(2):899-904 26. Horowitz MC,Coleman DL, Flood PM, Kupper TS, Jilka RL (1989) Parathyroid hormone and lipopolysaccharide induce murine osteoblast-like cells to secrete a cytokine indistinguishable from granulocyte-macrophage colony-stimulating factor. J Clin Invest 83:14%157 27. Elford PR, Felix R, Cecchini M, Trechsel U, Fleiscb H (1987) Murine osteoblast-like cells and the osteogenic cell MC3T3-E1 release a macrophage colony-stimulating activity in culture. Calcif Tissue Int 41:151-156 28. Felix R, Fleisch H, Elford PR (1989) Bone-resorbing cytokines enhance release of macrophage colony-stimulating activity by the osteoblastic cell MC3T3-E1. Calcif Tissue Int 44:356-360 29. Horowitz MC, Coleman DL, Ryaby JT, Einhorn T (1989b) Osteotropic agents induce the differential secretion of granulocytemacrophage colony-stimulating factor by the osteoblast cell line MC3T3-E1. J Bone Miner Res 4(6):911-919 30. Lotz M, Jirik F, Kabouridis P, Tsoukas C, Hirano T, Kishimoto T, Carson DA (1988) B cell-stimulating factor 2/interleukine 6 is a costimulant for human thymocytes and T lymphocytes. J Exp Med 167:1253-1258 31. Uyttenhove C, Coulie PG, Van Snick J (1988) T cell growth and differentiation induced by interleukin-HP1/IL-6, the murine hybridomaJplasmacytoma growth factor. J Exp Med 167:14171427 32. Ikebuchi K, Wong GG, Clark SC, Ihle JN, Hirai Y, Ogawa M (1987) Interleukin-6 enhancement of interleukin-3-dependent proliferation of multipotential hemopoietic progenitors. Proc Natl Acad Sci USA, 84;9035 33. Koike K, Nakahata T, Takagi M, Kobayashi T, Ishiguro A, Tsuji K, Naganuma K, Okano A, Akiyama Y, Akabane T (1988) Synergism of BSF-2/interleukin 6 and interleukin 3 on development of multipotential hemopoietic progenitors in serum-free culture. J Exp Med 168:879-890 34. Ishimi Y, Miyaura C, He Jim C, Akatsu T, Abe E, Nakamura Y, Yamaguchi A, Yoshiki S, Matsuda T, Hirano T, Kishimoto T, Suda T (1990) IL-6 produced by osteoblasts and induces bone resorption. J Immunol 145(10):3297-3303 35. Kurland J, Moore MAS (1977) Modulation of hemopoiesis by prostaglandins. Exp Hematol 5:357-373 36. Kurland JI, Bockman RS, Broxmeyer HE, Moore MAS (1978) Limitation of excessive myelopoiesis by the intrinsic modulation of macrophage-derived prostaglandin E. Science 199:552555