- Email: [email protected]
Expression of hematopoietic growth factor RNAs in human mesenchymal cells from various organs
Leukemia Research Vol. 15, No. 7, pp. 551-558, 1991.
0145-2126/91 $3.00 + .00 Pergamon Press pie
Printed in Great Britain.
EXPRESSION OF HEMATOPOIETIC GROWTH FACTOR RNAs IN H U M A N M E S E N C H Y M A L CELLS FROM V A R I O U S O R G A N S KENJI YAMATO,*t ZIAD EL-HAJJOUI* and H. PHILLIPKOEFFLER* *Department of Medicine, UCLA School of Medicine, Cedars-Sinai Medical Centre, 8700 Beverly Blvd, Los Angeles, CA 90024, U.S.A. and tDepartment of Medicine, Kochi Medical School, Okocho, Nankoku, Kochi 783, Japan
(Received 19 September 1990. Revision accepted 18 December 1990) Abstract--Experiments were undertaken to study expression of hematopoietic growth factor RNAs in mesenchymal cells from a variety of organs including bone marrow, foreskin, gingiva, and lung. Cells from each organ had negligible expression of RNAs coding for granulocyte (G), macrophage (M), and granulocyte-macrophage (GM) colony stimulating factor (CSF), interleukin lfl (IL-lfl), and IL-6. Fibroblasts from each tissue had a comparable ability to express the same cytokine RNAs. Surprisingly, the stimuli for expression of G-CSF RNA was disparate from the stimuli for expression of the other cytokine RNAs. While IL-1/5 enhanced accumulation of G-CSF RNA, tumor necrosis factor oc(TNF) and 12-O-tetradecanoylphorbol-13-acetate (TPA) did not. In contrast, IL-lfl, TNF, and TPA equally stimulated increased levels of M-CSF, GM-CSF, IL-lfl and IL-6 RNAs.
Key words: Cytokine, RNA, fibroblasts, hematopoiesis.
INTRODUCTION
organs before and after their stimulation. Regardless of their origins, the various mesenchymal cells were equally capable of increasing their levels of hematopoietic growth factor RNAs. Levels of G-CSF RNA markedly increased after cells were cultured with IL-lfl, but not when they were cultured with either tumor necrosis factor a: (TNF) or 12-O-tetradecanoylphorbol-13-acetate (TPA). In contrast, all three stimuli (IL-lfl, TNF, TPA) increased expression of GM-CSF, M-CSF, IL-lfl, and IL-6 RNAs, suggesting that signals for regulation of expression of G-CSF differ from those for several other hematopoietic growth factor genes.
BONE MARROWconsists of hematopoietic progenitor cells and stromal cells including fibroblasts, endothelial cells, adipocytes, and macrophages. These stromal cells provide a hematopoietic microenvironment, which is essential for the survival, proliferation and differentiation of blood stem cells [1, 2]. Stromal cells exert these actions through cellto-cell contact with hematopoietic stem cells and by their elaboration of growth factors [3, 4]. Studies showed that mesenchymal cells from other organs also expressed hematopoietic growth factors in response to extraceltular stimulation including human umbilical cord vascular endothelial cells [510] as well as skin and lung fibroblasts [11-16]. This study was undertaken to investigate RNA expression of hematopoietic growth factor genes including granulocyte (G), macrophage (M), and granylocytemacrophage (GM) colony stimulating factor (CSF, reviewed in [17]), interleukin lfl (IL-lfl, [18, 19]) and IL-6 [18, 19] in fibroblasts from several different
MATERIALS AND METHODS
Cell culture Bone marrow fibroblastoid cells were obtained from a long-term culture of normal bone marrow cells. Mononuclear cells were separated by a Ficoll-Paque gradient from whole bone marrow aspirated from healthy volunteers, each of whom was provided with informed consent. Adherent cells were subcultured after 2 months from the start of cultivation. They were morphologically homogenous and had fibroblastoid appearance. Gingival diploid fibroblasts were kindly provided by Dr R. Saglie (UCLA, Los Angeles, CA) and foreskin diploid fibroblasts were provided by Dr R. Sparks (UCLA). Human embryonic lung fibroblasts (MRC 5 cells) were purchased from American Type Tissue Culture (Rockville, MD). Cells were cultured with o; MEM medium (Sigma, St. Louis, MO) supplemented with 10% fetal calf serum (Flow) and
Abbreviations: CSF, colony stimulating factor; G, granulocyte; M, macrophage; GM, granulocyte-macrophage; IL, interleukin; TNF, tumor necrosis factor o~; TPA, 12O-tetradecanoylphorbol- 13-acetate. Correspondence to: H. Phillip Koeffler, M.D., UCLA Department of Medicine, Division of Hematology/ Oncology, 11-240 Factor Building, 10833 Le Conte Avenue, Los Angeles, CA 90024-1678, U.S.A. 551
K. YAMATOet al.
552
1% penicillin/streptomycin (Sigma) at a density of 5 x 1055 x 106/ml in 175 cm2 flasks. Cells from the third to the tenth passage were used for the experiments. All the experiments were performed using confluent cultures. TPA was purchased from Sigma. Recombinant human TNF (4 × 10 7 U/mg) was generously provided by Dr M. Shepard (Genentech, Inc., San Francisco, CA) and IL-lfl (1 × 108U/mg) was a kind gift from Dr S. Gilles (Immunex, Seattle, WA).
Northern-blotting Cytoplasmic RNA was extracted by the method as described elsewhere [20]. RNA was separated by a 1% agarose gel and transferred to a nylon-membrane filter. Blots were probed with random hexamer primed 32p-labeled cDNAs. Hybridization was performed at 42°C for 15 h in 50% formamide, 2x SSC (1× = 150mM sodium chloride, 15 mM sodium citrate), 5 x Denhardt's solution, 0.1% sodium dodecyl sulfate, 10% dextran sulfate and 100 gg/ml salmon sperm DNA. Filters were washed at a stringency of 0.1 × SSC at 65°C and exposed to Kodak XAR films. Autoradiograms were developed at different exposures and representative pictures were presented. All the experiments were done, at least, in duplicate and consistent results were obtained for each experiment. GCSF cDNA (Xho I, 1,800 bp) was obtained from pGCSF2 (a generous gift from S. Clark, Genetics Institute, Cambridge, MA), M-CSF cDNA (Xho I/EcoR I, 2000 bp) was from p3ACSFRI [21], GM-CSF (EcoR I, 800 bp) was from pCSF-1 [22], IL-lfl (Pst I, 900 bp) was from pA-26 [23], IL-6 cDNA (Xho, I, 1,200 bp) was from pXM309 (a generous gift from S. Clark, Genetics Institute, Cambridge, MA), fl-actin cDNA (EcoR I/BamH I, 700 bp) was from pHFflA-3'-ut [24]. RESULTS
Expression of hematopoietic growth factor RNAs in bone marrow fibroblastoid cells Expression of RNAs coding for various cytokines in human mesenchymal cells was studied using the Northern-blot technique. Bone marrow fibroblastoid cells were cultured with either TNF (50 ng/ml), TPA (50 nM) or IL-1/3 (500 pg/ml) for 4 h. Prior experiments showed that TNF, TPA and IL-lfl had the maximal effect on accumulation of cytokine RNAs at the concentration used in our experiments [25, 26]. As shown in Fig. 1, unstimulated bone marrow fibroblastoid cells constitutively expressed negligible levels of RNAs coding for G, GM-CSF, IL-1/3 and IL-6. Exposure to IL-lfl increased levels of all these growth factor RNAs. TNF or TPA increased levels of GM-CSF, IL-1/3 and IL-6 RNAs but failed to accumulate G-CSF RNA. Hybridization of the same blot with fl-actin cDNA showed that equivalent amounts of RNA were applied to each lane (Fig. 1).
Time response of accumulation of growth factor RNAs In order to confirm that the time point employed in
the first experiments were appropriate, time-course experiments were undertaken. TNF (50ng/ml) increased levels of M-CSF, IL-lfl and IL-6 RNAs were detected within 2 h of culture (Fig. 2A). Levels of expression of these cytokine RNAs stayed nearly the same at least for 16 h. Hybridization of the same blot with G-CSF cDNA showed undetectable levels of G-CSF RNA (data not shown). Expression of MCSF RNA was detected at 4 h after exposure of the bone marrow fibroblastoid cells to TPA, and levels of this RNA increased slightly at 16 h (Fig. 2B). Maximal levels of IL-6 RNA were observed 2 h after exposure to TPA; levels returned to baseline by 8 h (Fig. 2B). Rehybridization of the blot showed that G-CSF RNA was not detected in TPA-stimulated cells over the 16 h duration of culture (data not shown). Levels of fl-actin RNA in TPA-treated cells decreased with time (Fig. 2B). Ethidium bromide staining showed that each lane contained equivalent amounts of ribosomal RNA, suggesting that 50 nM TPA may have an inhibitory effect on levels of flactin RNA (data not shown). IL-lfl co-ordinately increased levels of all the cytokine RNAs including G-CSF in the bone marrow fibroblastoid cells, with the maximal levels at 4 h (Fig. 2C).
Expression of hematopoietic growth factor RNAs in fibroblasts from various organs Expression of growth factor RNAs was also studied in foreskin (Fig. 3A), gingival (B) and lung fibroblasts (C) using the Northern-blot technique. TNF failed to increase levels of G-CSF RNA in all these mesenchymal cells even though it stimulated them to increase levels of RNAs coding for other growth factors (IL-6, M-CSF). On the other hand, IL-lfl caused the accumulation of all these growth factor RNAs including G-CSF RNA. These results suggested that signals causing the expression of G-CSF RNA may be different from those of other hematopoietic growth factor RNAs. Induction of expression of GM-CSF RNA slightly differed among gingival and lung fibroblasts: in gingival fibroblasts, IL-1/3 was a more potent inducer of expression of GM-CSF RNA than TNF; while, in lung fibroblasts, TNF was more potent than IL-lfl. The same findings were obtained in another separate experiment. DISCUSSION Expression of hematopoietic growth factor RNAs were studied in bone marrow fibroblastoid cells and fibroblasts from other organs. The cells constitutively expressed negligible levels of M, GM-CSF, IL-lfl and IL-6 RNAs. Each of these RNAs accumulated in a coordinated fashion when stimulated by IL-lfl
IL-6
-
L-I/3
-
/~-actin
-
)
FIG. 1. Expression of hematopoietic growth factor RNAs in bone marrow fibroblastoid cells. Bone marrow fibroblastoid cells were cultured with either TNF (50 ng/ml), TPA (50 nM) or IL-lfl (500 pg/ml) for 4 h. Cytoplasmic RNA was extracted and analysed by Northern-blotting as described in "Materials and Methods". The same blot was sequentially hybridized with 32p-labeled GM-CSF, G-CSF, IL-6, IL-lfl and fl-actin cDNAs. Lane 1, unstimulated bone marrow fibroblastoid cells; lane 2, cells stimulated with TNF; lane 3, TPA; lane 4, IL-lfl.
553
.&. '~q
¢=0
M-CSF"
IL-6
tL-1B B-~tin
B
FIG. 2. Time-response increase of levels of hematopoietic growth factor RNAs in bone marrow fibroblastoid cells. Bone marrow fibroblastoid cells were cultured with either TNF (50 ng/ml) (A), TPA (50 nm) (B), IL-1/~ (500 pg/ml) (C); cytoplasmic RNA was harvested at different time points (h) and was analysed by Northern-blotting.
554
A §kb 3kb lkb
B
C O
I-"
s
--
GM-CSF
0.8kb
M-CSF
4.0kb
G-CSF
1.6kb
IL-6
.3kb
B-actin
.lkb FIG. 3. Expression of hematopoietic growth factor RNAs in fibroblasts from foreskin, gingiva and lung. Fibroblasts cells from foreskin (A), gingiva (B), lung (C) were stimulated with either TNF (50 ng/ml), TPA (50 nM), or IL-lfl (500 pg/ml) for 4 h. Cytoplasmic RNA was analysed for cytokine RNAs by Northern-blotting. The same blots were sequentially rehybridized with 32P-labeled G-CSF, M-CSF, GM-CSF, IL-lfl, IL-6 and/~-actin. In panel A, less RNA was applied on lane 3 (IL-lfl) as compared with lane 1 (control) and 2 (TNF).
555
Hematopoietic growth factor RNAs in mesenchymal cells and TNF. In contrast, the accumulation of G-CSF R N A was inducible by IL-lfl but not by either TNF or TPA. These observations suggested that regulatory mechanisms leading to accumulation of G-CSF R N A in mesenchymal cells were different from those leading to accumulation of GM-CSF, IL-lfl and IL-6 RNAs. Previous studies showed that T N F stimulated human foreskin fibroblasts to increase levels of GCSF R N A [27]. The T N F used in our experiments is probably as potent as that used in other experiments since our recombinant T N F stimulated bone marrow stromal cells to increase their expression of other growth factor R N A s to levels equal, or even greater than IL-lfl was able to stimulate these cells to express growth factor RNAs. And, moreover, the same results were repeatedly obtained using fibroblasts from foreskin, gingiva and lung. Our method for detecting G-CSF R N A theoretically may be less sensitive than those employed in the other investigation, although we easily detected G-CSF R N A after ILlfl stimulation. We previously showed that an embryonic lung fibroblast cell line known as WI-38 could be stimulated to express increased levels of G-CSF m R N A after exposure of the cells to either T N F or T P A [13]. The WI-13 cells are unusual in so far as they also constitutively express a number of cytokines (Ross H., Sato N., Koeffler H. P., unpublished observations) and, therefore, are different from the fibroblasts used in this study which are not immortalized cells. Our present study clearly showed that, at the very least, T N F as compared with IL-1/3 was a much less potent inducer of accumulation of G-CSF R N A even though T N F and IL-lfl equally induced expression of R N A s coding for other growth factors. Constitutive accumulation of M-CSF R N A was not detected in mesenchymal cells; expression of this R N A was inducible by TNF, T P A and IL-lfl. These results were consistent with several other reports [9, 28]. However, constitutive expression of M-CSF R N A was reported in fetal liver fibroblastoid cells and vascular endothelial cells in vitro [7, 16]. Possibly the liver fibroblasts and endothelial cells elaborated an unidentified factor that induced the expression of M-CSF RNA. We observe only minor differences in expression of hematopoietic growth factor RNAs in bone marrow fibroblastoid cells as compared with fibroblasts from other organs. At least, superficially, therefore, our results do not explain why hematopoiesis is prominent in bone marrow as compared with other sites. Analysis of sites of extramedullary hematopoiesis in several diseases may be informative. In addition, G o r d o n et al. reported that bone marrow stromal cells trap GM-CSF protein on their cellular surface and facilitate exposure of blood progenitor cells to
557
this multipotential hematopoietic factor at a high local concentration [29]. Perhaps fibroblasts from other sites are unable to present cytokines in a similar fashion.
Acknowledgements--Supported in part by USPHS Grants CA26038, CA3396, the Weisz Family Foundation and the 4E Leukemia Fund in memory of Marilyn Levine, Erwin Epstein and the Clinical Nutritional Research Unit. Dr. H. Phillip Koeffler is a member of the Jonsson Comprehensive Cancer Center. We would like to thank Elisa Weiss, Marge Goldberg and Elaine Epstein for their secretarial assistance.
REFERENCES 1. Dexter T. M., Allen T. D. & Lajtha L. G. (1977) Conditions controlling the proliferation of hematopoietic stem cells in vitro. J. cell. Physiol. 91, 335. 2. Whitlock C. A. & Whitte O. N. (1982) Long-term culture of B lymphocytes and their precursors from murine bone marrow. Proc. natn. Acad. Sci. U.S.A. 79, 3608. 3. Coulombel L., Eaves A. C. & Eaves C. J. (1983) Enzymatic treatment of long-term human marrow culture reveals the preferential location of primitive hematopoietic progenitors in the adherent layer. Blood 62, 291. 4. Hunt P., Robertson D., Weiss D., Rennick D., Lee F. & Witte O. N. (1987) A single bone marrow derived stromal cell type support the in vitro growth of early lymphoid and myeloid cells. Cell 48, 997. 5. Sieff C. A., Tsai S. &Faller D. V. (1987) Interleukin 1 induces cultured human endothelial cell production of granulocyte-macrophage colony stimulating factor. J. clin. Invest. 80, 818. 6. Bagby G. C., Dinarello C. A., Wallace P., Wagner C., Hefeneider S. & McCall E. (1986) Interleukin 1 stimulates granulocyte macrophage colony stimulating activity release by vascular endothelial cells. J. clin. Invest. 78, 1316. 7. Broudy V. C., Kaushansky K., Segal G. M., Harlan J. M. & Adamson J. W. (1987) Tumor necrosis factor type alpha stimulates human endothelial cells to produce granulocyte/macrophage colony stimulating factor. Proc. hath. Aead. Sci. U.S.A. 83, 7467. 8. Nawroth P. P., Bank I., Handley D., Cassimeris J., Chess L. & Stern D. (1986) Tumor necrosis factor/ cachectin interacts with endothelial cell receptors to induce release of interleukin 1. J. expl Med. 163, 1363. 9. Seelentag W. K., Mermod J.-J., Montesano R. & Vassali P. (1987) Additive effect of interleukin 1 and tumor necrosis factor-alpha on the accumulation of the three granulocyte and macrophage colony stimulating factor mRNA in human endothelial cells. E M B O J. 6, 2261. 10. Sironi M., Breviario F., Proserpio P., Biondi A., Vecchi A., Van Damme J., Dejanna E. & Montovani A. (1989) IL-1 stimulates IL-6 production in endothelial cells. J. Immun. 142, 549. 11. Zucali J. R., Dinarello C. A., Oblon D. J., Gross M. A., Anderson L. & Weiner R. S. (1986) Interleukin 1
558
K. YAMATOet al.
stimulates fibroblasts to produce granulocyte macrophage colony stimulating activity and prostaglandin E2. J. clin. Invest. 78, 316. 12. Munker R., Gasson J., Ogawa M. & Koeffler H. P. (1986) Recombinant human TNF induces production of granylocyte macrophage colony stimulating factor. Nature 323, 79. 13. Koeffler H. P., Gasson J., Ranyard J., Souza L., Shepard M. & Munker R. (1987) Recombinant human TNF alpha stimulates production of granulocyte colony stimulating factor. Blood 70, 55. 14. Kaushansky K., Lin N. & Adamson J. W. (1988) Interleukin 1 stimulates fibroblasts to synthesize granulocyte-macrophage and granulocyte colony stimulating factors. J. clin. Invest. gl, 92. 15. Kohase M., Hendriksen-DeStefano D., May L. T., Vilcek J. & Sehgal P. B. (1986) Induction of beta-2interferon by tumor necrosis factor: A homeostatic mechanism in the control of cell proliferation. Cell 45, 659. 16. Yang Y.-C., Tsai S., Wong G. G. & Clark S. C. (1988) Interleukin 1 regulation of hematopoietic growth factor production by human fibroblasts. J. cell. Physiol. 134, 292. 17. Metcalf D. (1988) The Molecular Control of Blood Cells. Harvard University Press, Cambridge, MA, U.S.A. 18. Stanley E. R., Bartocci A., Patinkin D., Rosendaal M. & Bradley T. R. (1986) Regulation of very primitive, multipotent, hematopoietic cells by hematopoietin-1. Cell 45, 667. 19. Ikeuchi K., Wong G. G., Clark S. C., Ihle J. N., Hirai Y. & Ogawa M. (1987) Interleukin 6 enhancement of interleukin 3-dependent proliferation of multipotent hematopoietic progenitors. Proc. natn. Acad. Sci. U.S.A. 84, 9035. 20. Berk A. J. & Sharp P. A. (1977) Sizing and mapping of early adenovirus mRNAs by gel electrophoresis and S1 endonuclease-digested hybrid. Cell 12, 721. 21. Wong G. G., Temple P. A., Leary A. C., Wite-Giannotti J. S., Yang Y. C., Ciarletta A. B., Chung M., Murths P., Kriz R., Kaufman R. J., Fernz C. R., Sibley
22.
23.
24.
25. 26.
27.
28.
29.
B. S., Turner K. J., Hewick R. M., Clark S. C., Yanai N., Yokota H., Yamada M., Saito M., Motoyoshi K. & Takaku H. (1987) Human CSF-I: Molecular cloning and expression of 4-kd cDNA encoding the human urinary protein. Science 235, 1504. Wong G. G., Witek J. S., Temple P. A., Wilkens K. M., Leary A. C., Luxenburg D. P., Jones S. S., Brown E. L., Kay R. M., Orr E. C., Shoemaker C., Golde D. W., Kaufman R. J., Hewik R. M., Wang E. A. & Clark S. C. (1985) Human GM-CSF: Molecular cloning of the complementary DNA and purification of the natural and recombinant proteins. Science 228, 810. Auron P. E., Webb A. C., Rosenwasser L. J., Mucci S. F., Rich A., Wolf S. M. & Dinarello C. A. (1984) Nucleotide sequence of human monocyte interleukin 1 precursor cDNA. Proc. natn. Acad. Sci. U.S.A. 81, 7907. Ponte P., Gunning P., Blau H. & Kedes L. (1983) Human actin genes are single copy for tr-skeltal and a~cardiac actin but multicopy for fl-, and y-cytoskeltal gene: 3' untranslated regions are isotype specific but are conserved in evolution. Mol. Cell Biol. 3, 1783. Yamato K., EI-Hajjaoui Z. & Koeffler H. P. (1989) Regulation of levels of IL-1 mRNA in human fibroblasts. J. cell. Physiol. 139, 610. Yamato K., EI-Hajjoui Z., Kuo J. F. & Koeffler H. P. (1989) Granulocyte macrophage colony stimulating factor: Signals for its mRNA accumulation. Blood 74, 314. Seelentag W., Mermad J.-J. & Vassali P. (1989) Interleukin 1 and tumor necrosis factor-o: additively increase the levels of granulocyte-macrophage and granulocyte colony stimulating factor (CSF) mRNA in human fibroblasts. Eur. J. Immun. 19, 209. Fibbe W. E., Van Damme J., Billiau A., Goselink H. M., Voogt P. J., Van Eeden G., Altrak B. W. & Falkenburg J. H. P. (1988) Interleukin 1 induces human bone marrow stromal cells in long-term culture to produce G-CSF and M-CSF. Blood 71, 430. Gordon M. Y., Riley G. P., Watt S. M. & Greeves M. F. (1987) Compartmentalization of a hematopoietic growth factor (GM-CSF) by glycosaminoglycans in the bone marrow microenvironment. Nature 326, 403.
0145-2126/91 $3.00 + .00 Pergamon Press pie
Printed in Great Britain.
EXPRESSION OF HEMATOPOIETIC GROWTH FACTOR RNAs IN H U M A N M E S E N C H Y M A L CELLS FROM V A R I O U S O R G A N S KENJI YAMATO,*t ZIAD EL-HAJJOUI* and H. PHILLIPKOEFFLER* *Department of Medicine, UCLA School of Medicine, Cedars-Sinai Medical Centre, 8700 Beverly Blvd, Los Angeles, CA 90024, U.S.A. and tDepartment of Medicine, Kochi Medical School, Okocho, Nankoku, Kochi 783, Japan
(Received 19 September 1990. Revision accepted 18 December 1990) Abstract--Experiments were undertaken to study expression of hematopoietic growth factor RNAs in mesenchymal cells from a variety of organs including bone marrow, foreskin, gingiva, and lung. Cells from each organ had negligible expression of RNAs coding for granulocyte (G), macrophage (M), and granulocyte-macrophage (GM) colony stimulating factor (CSF), interleukin lfl (IL-lfl), and IL-6. Fibroblasts from each tissue had a comparable ability to express the same cytokine RNAs. Surprisingly, the stimuli for expression of G-CSF RNA was disparate from the stimuli for expression of the other cytokine RNAs. While IL-1/5 enhanced accumulation of G-CSF RNA, tumor necrosis factor oc(TNF) and 12-O-tetradecanoylphorbol-13-acetate (TPA) did not. In contrast, IL-lfl, TNF, and TPA equally stimulated increased levels of M-CSF, GM-CSF, IL-lfl and IL-6 RNAs.
Key words: Cytokine, RNA, fibroblasts, hematopoiesis.
INTRODUCTION
organs before and after their stimulation. Regardless of their origins, the various mesenchymal cells were equally capable of increasing their levels of hematopoietic growth factor RNAs. Levels of G-CSF RNA markedly increased after cells were cultured with IL-lfl, but not when they were cultured with either tumor necrosis factor a: (TNF) or 12-O-tetradecanoylphorbol-13-acetate (TPA). In contrast, all three stimuli (IL-lfl, TNF, TPA) increased expression of GM-CSF, M-CSF, IL-lfl, and IL-6 RNAs, suggesting that signals for regulation of expression of G-CSF differ from those for several other hematopoietic growth factor genes.
BONE MARROWconsists of hematopoietic progenitor cells and stromal cells including fibroblasts, endothelial cells, adipocytes, and macrophages. These stromal cells provide a hematopoietic microenvironment, which is essential for the survival, proliferation and differentiation of blood stem cells [1, 2]. Stromal cells exert these actions through cellto-cell contact with hematopoietic stem cells and by their elaboration of growth factors [3, 4]. Studies showed that mesenchymal cells from other organs also expressed hematopoietic growth factors in response to extraceltular stimulation including human umbilical cord vascular endothelial cells [510] as well as skin and lung fibroblasts [11-16]. This study was undertaken to investigate RNA expression of hematopoietic growth factor genes including granulocyte (G), macrophage (M), and granylocytemacrophage (GM) colony stimulating factor (CSF, reviewed in [17]), interleukin lfl (IL-lfl, [18, 19]) and IL-6 [18, 19] in fibroblasts from several different
MATERIALS AND METHODS
Cell culture Bone marrow fibroblastoid cells were obtained from a long-term culture of normal bone marrow cells. Mononuclear cells were separated by a Ficoll-Paque gradient from whole bone marrow aspirated from healthy volunteers, each of whom was provided with informed consent. Adherent cells were subcultured after 2 months from the start of cultivation. They were morphologically homogenous and had fibroblastoid appearance. Gingival diploid fibroblasts were kindly provided by Dr R. Saglie (UCLA, Los Angeles, CA) and foreskin diploid fibroblasts were provided by Dr R. Sparks (UCLA). Human embryonic lung fibroblasts (MRC 5 cells) were purchased from American Type Tissue Culture (Rockville, MD). Cells were cultured with o; MEM medium (Sigma, St. Louis, MO) supplemented with 10% fetal calf serum (Flow) and
Abbreviations: CSF, colony stimulating factor; G, granulocyte; M, macrophage; GM, granulocyte-macrophage; IL, interleukin; TNF, tumor necrosis factor o~; TPA, 12O-tetradecanoylphorbol- 13-acetate. Correspondence to: H. Phillip Koeffler, M.D., UCLA Department of Medicine, Division of Hematology/ Oncology, 11-240 Factor Building, 10833 Le Conte Avenue, Los Angeles, CA 90024-1678, U.S.A. 551
K. YAMATOet al.
552
1% penicillin/streptomycin (Sigma) at a density of 5 x 1055 x 106/ml in 175 cm2 flasks. Cells from the third to the tenth passage were used for the experiments. All the experiments were performed using confluent cultures. TPA was purchased from Sigma. Recombinant human TNF (4 × 10 7 U/mg) was generously provided by Dr M. Shepard (Genentech, Inc., San Francisco, CA) and IL-lfl (1 × 108U/mg) was a kind gift from Dr S. Gilles (Immunex, Seattle, WA).
Northern-blotting Cytoplasmic RNA was extracted by the method as described elsewhere [20]. RNA was separated by a 1% agarose gel and transferred to a nylon-membrane filter. Blots were probed with random hexamer primed 32p-labeled cDNAs. Hybridization was performed at 42°C for 15 h in 50% formamide, 2x SSC (1× = 150mM sodium chloride, 15 mM sodium citrate), 5 x Denhardt's solution, 0.1% sodium dodecyl sulfate, 10% dextran sulfate and 100 gg/ml salmon sperm DNA. Filters were washed at a stringency of 0.1 × SSC at 65°C and exposed to Kodak XAR films. Autoradiograms were developed at different exposures and representative pictures were presented. All the experiments were done, at least, in duplicate and consistent results were obtained for each experiment. GCSF cDNA (Xho I, 1,800 bp) was obtained from pGCSF2 (a generous gift from S. Clark, Genetics Institute, Cambridge, MA), M-CSF cDNA (Xho I/EcoR I, 2000 bp) was from p3ACSFRI [21], GM-CSF (EcoR I, 800 bp) was from pCSF-1 [22], IL-lfl (Pst I, 900 bp) was from pA-26 [23], IL-6 cDNA (Xho, I, 1,200 bp) was from pXM309 (a generous gift from S. Clark, Genetics Institute, Cambridge, MA), fl-actin cDNA (EcoR I/BamH I, 700 bp) was from pHFflA-3'-ut [24]. RESULTS
Expression of hematopoietic growth factor RNAs in bone marrow fibroblastoid cells Expression of RNAs coding for various cytokines in human mesenchymal cells was studied using the Northern-blot technique. Bone marrow fibroblastoid cells were cultured with either TNF (50 ng/ml), TPA (50 nM) or IL-1/3 (500 pg/ml) for 4 h. Prior experiments showed that TNF, TPA and IL-lfl had the maximal effect on accumulation of cytokine RNAs at the concentration used in our experiments [25, 26]. As shown in Fig. 1, unstimulated bone marrow fibroblastoid cells constitutively expressed negligible levels of RNAs coding for G, GM-CSF, IL-1/3 and IL-6. Exposure to IL-lfl increased levels of all these growth factor RNAs. TNF or TPA increased levels of GM-CSF, IL-1/3 and IL-6 RNAs but failed to accumulate G-CSF RNA. Hybridization of the same blot with fl-actin cDNA showed that equivalent amounts of RNA were applied to each lane (Fig. 1).
Time response of accumulation of growth factor RNAs In order to confirm that the time point employed in
the first experiments were appropriate, time-course experiments were undertaken. TNF (50ng/ml) increased levels of M-CSF, IL-lfl and IL-6 RNAs were detected within 2 h of culture (Fig. 2A). Levels of expression of these cytokine RNAs stayed nearly the same at least for 16 h. Hybridization of the same blot with G-CSF cDNA showed undetectable levels of G-CSF RNA (data not shown). Expression of MCSF RNA was detected at 4 h after exposure of the bone marrow fibroblastoid cells to TPA, and levels of this RNA increased slightly at 16 h (Fig. 2B). Maximal levels of IL-6 RNA were observed 2 h after exposure to TPA; levels returned to baseline by 8 h (Fig. 2B). Rehybridization of the blot showed that G-CSF RNA was not detected in TPA-stimulated cells over the 16 h duration of culture (data not shown). Levels of fl-actin RNA in TPA-treated cells decreased with time (Fig. 2B). Ethidium bromide staining showed that each lane contained equivalent amounts of ribosomal RNA, suggesting that 50 nM TPA may have an inhibitory effect on levels of flactin RNA (data not shown). IL-lfl co-ordinately increased levels of all the cytokine RNAs including G-CSF in the bone marrow fibroblastoid cells, with the maximal levels at 4 h (Fig. 2C).
Expression of hematopoietic growth factor RNAs in fibroblasts from various organs Expression of growth factor RNAs was also studied in foreskin (Fig. 3A), gingival (B) and lung fibroblasts (C) using the Northern-blot technique. TNF failed to increase levels of G-CSF RNA in all these mesenchymal cells even though it stimulated them to increase levels of RNAs coding for other growth factors (IL-6, M-CSF). On the other hand, IL-lfl caused the accumulation of all these growth factor RNAs including G-CSF RNA. These results suggested that signals causing the expression of G-CSF RNA may be different from those of other hematopoietic growth factor RNAs. Induction of expression of GM-CSF RNA slightly differed among gingival and lung fibroblasts: in gingival fibroblasts, IL-1/3 was a more potent inducer of expression of GM-CSF RNA than TNF; while, in lung fibroblasts, TNF was more potent than IL-lfl. The same findings were obtained in another separate experiment. DISCUSSION Expression of hematopoietic growth factor RNAs were studied in bone marrow fibroblastoid cells and fibroblasts from other organs. The cells constitutively expressed negligible levels of M, GM-CSF, IL-lfl and IL-6 RNAs. Each of these RNAs accumulated in a coordinated fashion when stimulated by IL-lfl
IL-6
-
L-I/3
-
/~-actin
-
)
FIG. 1. Expression of hematopoietic growth factor RNAs in bone marrow fibroblastoid cells. Bone marrow fibroblastoid cells were cultured with either TNF (50 ng/ml), TPA (50 nM) or IL-lfl (500 pg/ml) for 4 h. Cytoplasmic RNA was extracted and analysed by Northern-blotting as described in "Materials and Methods". The same blot was sequentially hybridized with 32p-labeled GM-CSF, G-CSF, IL-6, IL-lfl and fl-actin cDNAs. Lane 1, unstimulated bone marrow fibroblastoid cells; lane 2, cells stimulated with TNF; lane 3, TPA; lane 4, IL-lfl.
553
.&. '~q
¢=0
M-CSF"
IL-6
tL-1B B-~tin
B
FIG. 2. Time-response increase of levels of hematopoietic growth factor RNAs in bone marrow fibroblastoid cells. Bone marrow fibroblastoid cells were cultured with either TNF (50 ng/ml) (A), TPA (50 nm) (B), IL-1/~ (500 pg/ml) (C); cytoplasmic RNA was harvested at different time points (h) and was analysed by Northern-blotting.
554
A §kb 3kb lkb
B
C O
I-"
s
--
GM-CSF
0.8kb
M-CSF
4.0kb
G-CSF
1.6kb
IL-6
.3kb
B-actin
.lkb FIG. 3. Expression of hematopoietic growth factor RNAs in fibroblasts from foreskin, gingiva and lung. Fibroblasts cells from foreskin (A), gingiva (B), lung (C) were stimulated with either TNF (50 ng/ml), TPA (50 nM), or IL-lfl (500 pg/ml) for 4 h. Cytoplasmic RNA was analysed for cytokine RNAs by Northern-blotting. The same blots were sequentially rehybridized with 32P-labeled G-CSF, M-CSF, GM-CSF, IL-lfl, IL-6 and/~-actin. In panel A, less RNA was applied on lane 3 (IL-lfl) as compared with lane 1 (control) and 2 (TNF).
555
Hematopoietic growth factor RNAs in mesenchymal cells and TNF. In contrast, the accumulation of G-CSF R N A was inducible by IL-lfl but not by either TNF or TPA. These observations suggested that regulatory mechanisms leading to accumulation of G-CSF R N A in mesenchymal cells were different from those leading to accumulation of GM-CSF, IL-lfl and IL-6 RNAs. Previous studies showed that T N F stimulated human foreskin fibroblasts to increase levels of GCSF R N A [27]. The T N F used in our experiments is probably as potent as that used in other experiments since our recombinant T N F stimulated bone marrow stromal cells to increase their expression of other growth factor R N A s to levels equal, or even greater than IL-lfl was able to stimulate these cells to express growth factor RNAs. And, moreover, the same results were repeatedly obtained using fibroblasts from foreskin, gingiva and lung. Our method for detecting G-CSF R N A theoretically may be less sensitive than those employed in the other investigation, although we easily detected G-CSF R N A after ILlfl stimulation. We previously showed that an embryonic lung fibroblast cell line known as WI-38 could be stimulated to express increased levels of G-CSF m R N A after exposure of the cells to either T N F or T P A [13]. The WI-13 cells are unusual in so far as they also constitutively express a number of cytokines (Ross H., Sato N., Koeffler H. P., unpublished observations) and, therefore, are different from the fibroblasts used in this study which are not immortalized cells. Our present study clearly showed that, at the very least, T N F as compared with IL-1/3 was a much less potent inducer of accumulation of G-CSF R N A even though T N F and IL-lfl equally induced expression of R N A s coding for other growth factors. Constitutive accumulation of M-CSF R N A was not detected in mesenchymal cells; expression of this R N A was inducible by TNF, T P A and IL-lfl. These results were consistent with several other reports [9, 28]. However, constitutive expression of M-CSF R N A was reported in fetal liver fibroblastoid cells and vascular endothelial cells in vitro [7, 16]. Possibly the liver fibroblasts and endothelial cells elaborated an unidentified factor that induced the expression of M-CSF RNA. We observe only minor differences in expression of hematopoietic growth factor RNAs in bone marrow fibroblastoid cells as compared with fibroblasts from other organs. At least, superficially, therefore, our results do not explain why hematopoiesis is prominent in bone marrow as compared with other sites. Analysis of sites of extramedullary hematopoiesis in several diseases may be informative. In addition, G o r d o n et al. reported that bone marrow stromal cells trap GM-CSF protein on their cellular surface and facilitate exposure of blood progenitor cells to
557
this multipotential hematopoietic factor at a high local concentration [29]. Perhaps fibroblasts from other sites are unable to present cytokines in a similar fashion.
Acknowledgements--Supported in part by USPHS Grants CA26038, CA3396, the Weisz Family Foundation and the 4E Leukemia Fund in memory of Marilyn Levine, Erwin Epstein and the Clinical Nutritional Research Unit. Dr. H. Phillip Koeffler is a member of the Jonsson Comprehensive Cancer Center. We would like to thank Elisa Weiss, Marge Goldberg and Elaine Epstein for their secretarial assistance.
REFERENCES 1. Dexter T. M., Allen T. D. & Lajtha L. G. (1977) Conditions controlling the proliferation of hematopoietic stem cells in vitro. J. cell. Physiol. 91, 335. 2. Whitlock C. A. & Whitte O. N. (1982) Long-term culture of B lymphocytes and their precursors from murine bone marrow. Proc. natn. Acad. Sci. U.S.A. 79, 3608. 3. Coulombel L., Eaves A. C. & Eaves C. J. (1983) Enzymatic treatment of long-term human marrow culture reveals the preferential location of primitive hematopoietic progenitors in the adherent layer. Blood 62, 291. 4. Hunt P., Robertson D., Weiss D., Rennick D., Lee F. & Witte O. N. (1987) A single bone marrow derived stromal cell type support the in vitro growth of early lymphoid and myeloid cells. Cell 48, 997. 5. Sieff C. A., Tsai S. &Faller D. V. (1987) Interleukin 1 induces cultured human endothelial cell production of granulocyte-macrophage colony stimulating factor. J. clin. Invest. 80, 818. 6. Bagby G. C., Dinarello C. A., Wallace P., Wagner C., Hefeneider S. & McCall E. (1986) Interleukin 1 stimulates granulocyte macrophage colony stimulating activity release by vascular endothelial cells. J. clin. Invest. 78, 1316. 7. Broudy V. C., Kaushansky K., Segal G. M., Harlan J. M. & Adamson J. W. (1987) Tumor necrosis factor type alpha stimulates human endothelial cells to produce granulocyte/macrophage colony stimulating factor. Proc. hath. Aead. Sci. U.S.A. 83, 7467. 8. Nawroth P. P., Bank I., Handley D., Cassimeris J., Chess L. & Stern D. (1986) Tumor necrosis factor/ cachectin interacts with endothelial cell receptors to induce release of interleukin 1. J. expl Med. 163, 1363. 9. Seelentag W. K., Mermod J.-J., Montesano R. & Vassali P. (1987) Additive effect of interleukin 1 and tumor necrosis factor-alpha on the accumulation of the three granulocyte and macrophage colony stimulating factor mRNA in human endothelial cells. E M B O J. 6, 2261. 10. Sironi M., Breviario F., Proserpio P., Biondi A., Vecchi A., Van Damme J., Dejanna E. & Montovani A. (1989) IL-1 stimulates IL-6 production in endothelial cells. J. Immun. 142, 549. 11. Zucali J. R., Dinarello C. A., Oblon D. J., Gross M. A., Anderson L. & Weiner R. S. (1986) Interleukin 1
558
K. YAMATOet al.
stimulates fibroblasts to produce granulocyte macrophage colony stimulating activity and prostaglandin E2. J. clin. Invest. 78, 316. 12. Munker R., Gasson J., Ogawa M. & Koeffler H. P. (1986) Recombinant human TNF induces production of granylocyte macrophage colony stimulating factor. Nature 323, 79. 13. Koeffler H. P., Gasson J., Ranyard J., Souza L., Shepard M. & Munker R. (1987) Recombinant human TNF alpha stimulates production of granulocyte colony stimulating factor. Blood 70, 55. 14. Kaushansky K., Lin N. & Adamson J. W. (1988) Interleukin 1 stimulates fibroblasts to synthesize granulocyte-macrophage and granulocyte colony stimulating factors. J. clin. Invest. gl, 92. 15. Kohase M., Hendriksen-DeStefano D., May L. T., Vilcek J. & Sehgal P. B. (1986) Induction of beta-2interferon by tumor necrosis factor: A homeostatic mechanism in the control of cell proliferation. Cell 45, 659. 16. Yang Y.-C., Tsai S., Wong G. G. & Clark S. C. (1988) Interleukin 1 regulation of hematopoietic growth factor production by human fibroblasts. J. cell. Physiol. 134, 292. 17. Metcalf D. (1988) The Molecular Control of Blood Cells. Harvard University Press, Cambridge, MA, U.S.A. 18. Stanley E. R., Bartocci A., Patinkin D., Rosendaal M. & Bradley T. R. (1986) Regulation of very primitive, multipotent, hematopoietic cells by hematopoietin-1. Cell 45, 667. 19. Ikeuchi K., Wong G. G., Clark S. C., Ihle J. N., Hirai Y. & Ogawa M. (1987) Interleukin 6 enhancement of interleukin 3-dependent proliferation of multipotent hematopoietic progenitors. Proc. natn. Acad. Sci. U.S.A. 84, 9035. 20. Berk A. J. & Sharp P. A. (1977) Sizing and mapping of early adenovirus mRNAs by gel electrophoresis and S1 endonuclease-digested hybrid. Cell 12, 721. 21. Wong G. G., Temple P. A., Leary A. C., Wite-Giannotti J. S., Yang Y. C., Ciarletta A. B., Chung M., Murths P., Kriz R., Kaufman R. J., Fernz C. R., Sibley
22.
23.
24.
25. 26.
27.
28.
29.
B. S., Turner K. J., Hewick R. M., Clark S. C., Yanai N., Yokota H., Yamada M., Saito M., Motoyoshi K. & Takaku H. (1987) Human CSF-I: Molecular cloning and expression of 4-kd cDNA encoding the human urinary protein. Science 235, 1504. Wong G. G., Witek J. S., Temple P. A., Wilkens K. M., Leary A. C., Luxenburg D. P., Jones S. S., Brown E. L., Kay R. M., Orr E. C., Shoemaker C., Golde D. W., Kaufman R. J., Hewik R. M., Wang E. A. & Clark S. C. (1985) Human GM-CSF: Molecular cloning of the complementary DNA and purification of the natural and recombinant proteins. Science 228, 810. Auron P. E., Webb A. C., Rosenwasser L. J., Mucci S. F., Rich A., Wolf S. M. & Dinarello C. A. (1984) Nucleotide sequence of human monocyte interleukin 1 precursor cDNA. Proc. natn. Acad. Sci. U.S.A. 81, 7907. Ponte P., Gunning P., Blau H. & Kedes L. (1983) Human actin genes are single copy for tr-skeltal and a~cardiac actin but multicopy for fl-, and y-cytoskeltal gene: 3' untranslated regions are isotype specific but are conserved in evolution. Mol. Cell Biol. 3, 1783. Yamato K., EI-Hajjaoui Z. & Koeffler H. P. (1989) Regulation of levels of IL-1 mRNA in human fibroblasts. J. cell. Physiol. 139, 610. Yamato K., EI-Hajjoui Z., Kuo J. F. & Koeffler H. P. (1989) Granulocyte macrophage colony stimulating factor: Signals for its mRNA accumulation. Blood 74, 314. Seelentag W., Mermad J.-J. & Vassali P. (1989) Interleukin 1 and tumor necrosis factor-o: additively increase the levels of granulocyte-macrophage and granulocyte colony stimulating factor (CSF) mRNA in human fibroblasts. Eur. J. Immun. 19, 209. Fibbe W. E., Van Damme J., Billiau A., Goselink H. M., Voogt P. J., Van Eeden G., Altrak B. W. & Falkenburg J. H. P. (1988) Interleukin 1 induces human bone marrow stromal cells in long-term culture to produce G-CSF and M-CSF. Blood 71, 430. Gordon M. Y., Riley G. P., Watt S. M. & Greeves M. F. (1987) Compartmentalization of a hematopoietic growth factor (GM-CSF) by glycosaminoglycans in the bone marrow microenvironment. Nature 326, 403.