- Email: [email protected]
Leukemia inhibitory factor is mitogenic to osteoblasts
Bone Vol. 21, No. 3 September1997:243-247 ELSEVIER
Leukemia Inhibitory Factor Is Mitogenic to Osteoblasts J. C O R N I S H , 1 K. E. C A L L O N , 1 S. G. E D G A R , 2 and I. R. R E I D l Departments of 1 Medicine and 2 Pathology, University of Auckland, Auckland, New Zealand
that LIF is an inducible cytokine that is produced by a wide variety of cells and tissues including osteoblasts, z'6"27 We have previously demonstrated that LIF increases DNA synthesis in bone, and there is evidence of increases in the activity or number of both osteoclasts and osteoblasts. 12'13 In in vitro studies, we have shown that LIF stimulates bone resorption, and this is associated with a near doubling of the number of active osteoclasts in neonatal mouse calvariae. 22 At low concentrations, however, LIF predominantly stimulates DNA synthesis, 13 indicating that it may play a role in the physiological regulation of bone cell function. The LIF-stimulated increase in bone resorption is prostaglandin dependent, whereas the stimulation of DNA synthesis is independent of prostaglandin production. The effects of LIF on osteoblast cell cultures are complex, with cell growth and the synthesis of enzymes and bone matrix proteins being regulated. 2"6'8'2°'2s Furthermore, some effects of LIF are inconsistent between different models. In a fetal rat long bone organ culture system, LIF inhibits bone resorption, and the rate of DNA synthesis is not changed, 11 suggesting that the species used, the time course of experiments, and the state of osteoblast differentiation may affect the cellular response. In an in vivo study in which mice were engrafted with a hematopoietic cell line modified to produce high levels of LIF, Metcalf and Gearing demonstrated an increase in bone mass with the formation of irregularly shaped bone trabeculae in the bone marrow at the ends of the long bones, as well as some evidence of cortical bone resorption. 17 In another in vivo study 19 involving repeated intraperitoneal injections of LIF, no specific bone changes were noted. We have demonstrated that LIF accelerates bone turnover locally in a prostaglandin-independent manner in normal mice studied in vivo. 5 In light of our earlier studies in which there were increases in thymidine incorporation as well as in indices of both osteoblast and osteoclast activity following exposure to LIF in vitro, the present study set out to determine in which cell population(s) the increase in DNA synthesis occurs. We used autoradiography to localise [3H]thymidine incorporation and enzyme histochemistry to identify mature osteoblasts and osteoclasts. These techniques also allow a detailed description of the effects of LIF on the activity and distribution of these cell types in intact bone.
Leukemia inhibitory factor (LIF) regulates cell growth and is produced by a variety of tissues, including bone. Previously we have shown that recombinant human LIF induced an increase in osteoclast number, bone formation, and DNA synthesis. In the present study, we have defined the cells in intact bone at which the proliferative effects of LIF occur, using simultaneous enzyme histochemistry and autoradiographic techniques. The area of alkaline phosphatase-positive staining was increased twofold (p = 0.0008) and the number of [3H]thym~dine-positive cells was increased twofold (p = 0.0024) in I IF-treated bones. The radiolabeled cells either colocalized with alkaline phosphatase or were in the osteoprogenitor regi,~n. They were not found in the acid phosphatase-positive staining osteoclasts. These results indicate that cells which have a mitogenic response to LIF are bone-forming rather than bone-resorbing cells. (Bone 21: 243-247; 1997) © 1997 by Elsevier Science Inc. All rights reserved. Key Words: Leukemia inhibitory factor; Calvaria; Histochemical; Alkaline phosphal~se; Acid phosphatase; Autoradiography.
Introduction Growth factors and cytokines are important in the localized regulation of bone cell function, in an autocrine or paracrine manner. Leukemia inl~Jbitory factor (LIF), a single-chain glycoprotein (molecular weight of 58,000), is a pleiotrophic cytokine which acts on blood cells, embryonic stem cells, hepatocytes, and neuronal and adipose tissue, as well as bone. I° LIF is characterized by its ability to induce differentiation of a murine myeloid leukemic cell line; however, the molecule can promote both suppression and stimulation of proliferation and/or differentiation of different leukemic cell lines. 25 Transcription of the LIF gene has been detected at low levels in normal adult t i s s u e s . 4 L1F levels are elevated in septic shock and in various inflamrnatory body fluids, including the synovial fluid from subjects wil:h inflammatory arthritis. 27 LIF appears to be primarily a locally acting molecule, with its circulating concentrations being maintained at low levels. LIF or its transcripts have been detected in vitro in many cell types, including fibroblasts, T lymphocytes, bone marrow stromal cells, and osteoblasts. LIF production can be induced by various stimuli including lipopolysaccharide, interleukin-1, transforming growth factor-13, and tumour necrosis factor. 16 These findings suggest
Materials and Methods Materials
Recombinant murine LIF was kindly provided by Dr. Nicola and Dr. Gough, Walter and Eliza Institute (Melbourne, Australia). LIF was produced in Escherichia coli, as previously described. 7 LIF activity of 50 U/mL is defined as the concentration that induces 50% of M1 cells to exhibit differentiation. 18
Address for correspondence and reprints: Dr. J. Cornish, Department of Medicine, University of Auckland, Private Bag 92-019, Auckland, New Zealand. E-mail: [email protected]
© 1997 by ElsevierScienceInc. All rights reserved.
243
8756-3282/97/$17.00 PII $8756-3282(97)00144-0
244
J. Cornish et al. LIF is mitogenic to osteoblasts
Experimental Design Bone organ culture. The hemicalvaria of 6-day-old mice from a single litter were dissected out as described previously. 22 Calvariae were preincubated for 24 h in minimal essential medium with 1% charcoal-stripped, heat-inactivated serum, then changed to fresh medium containing LIF (300 U/mL) or vehicle. Incubation was continued for a further 24 or 48 h. There were ten calvariae in each group.
[3H]thymidine pulse label for autoradiography, The calvariae were pulse-labeled with [3H]thymidine (TRK 565; Amersham, UK) for the last hour of the incubation period, at a dose of 5 p~Ci/mL. In further experiments, the calvariae were pulselabeled with thymidine for the first 6 h of the incubation period to determine whether LIF affects proliferation of osteoblast and osteoclast precursor ceils. Histology. At the end of the incubation period, the specimens were washed in media with cold thymidine. The hemicalvariae were bisected through the parasagittal plane and used for the following enzyme histochemical studies. The tissue was fixed in cold 2% glutaraldehyde in 0.1 mol/L cacodylate buffer for 30 min at 4°C. Following fixation, the bones were washed in further cacodylate buffer and dehydrated at 4°C in glycol methacrylate in water (1 h in each of 50%, 70%, 95%, and 100%), followed by three 1-h changes in LR white resin (London Resin Company Ltd, Hants, UK). Polymerization of the resin was carried out on ice to minimize depression of enzyme activity by this exothermic reaction. Serial 1.5 Ixm sections were cut on a Reichert-Jung microtome (Austria). The sections were mounted on vectabondcoated slides and air-dried before further processing.
Staining for alkaline phosphatase. A simultaneous coupling azo-dye method was employed for demonstration of alkaline phosphatase activity. The substrate used was 5-bromo-4-chloro3-indolyl phosphate (BCIP) and the diazotized coupler employed was Nitro blue tetrazolium (NBT). Combined BCIP/NBT tablets were used (Sigma Diagnostics, St. Louis, MO). Incubations of sections were carried out for 30 min in the dark at 37°C. Because the prepared staining solution deteriorates rapidly, it was filtered and used immediately. This staining process, using new reagents, was repeated to obtain the desired intensity of staining. The sections were thoroughly washed in distilled water before proceeding to the acid phosphatase staining. Staining for acid phosphatase. A azo-dye method comparable to that described for alkaline phosphatase was employed to demonstrate acid phosphatase activity. The substrate used was naphthol AS-BI phosphate (Sigma Diagnostics) in N,N-dimethyl formamide (Sigma), and the diazotized coupler employed was Fast garnet GBC (Sigma) in acetate-tartrate buffer (pH 5.0). The sections were incubated for 1 h at 37°C. This staining process was repeated using new reagents to obtain the desired intensity of staining. Finally the sections were washed in distilled water and air-dried before processing for autoradiography. Autoradiography. Slides that had been histochemically stained were subsequently dipped in Kodak NTB2 emulsion (diluted 1:2 with distilled water) heated to 42°C. After coating, the slides were air-dried at room temperature, then stored in light-tight boxes and left to expose at 4°C for 2 weeks. The slides were developed in Kodak D19 (1:2) at 20°C for 2 min, fixed in Hypam (Ilford Australia Ltd., Sydney, Australia) rapid fixer (1:4) for 2 min, washed in distilled water for 5 min, and air-dried. Sections were counterstained with 1% Fast green.
Bone Vol. 21, No. 3 September 1997:243-247
Histomorphometry and data analysis. The slides were examined under an Olympus BX 50 microscope (Olympus Optical Co. Ltd., Tokyo, Japan) which was attached to an Osteomeasure Image Analyzer (Osteometrics, Inc., Atlanta, GA). Histomorphometric analyses were made of four adjacent fields (using a X20 objective) of parasaggital sections of hemicalvariae, enabling an examination of all levels of the calvariae from the frontal bone through the coronal suture and parietal bone to the lambdoidal suture. The following parameters per millimeter calvaria were measured: the area of alkaline phosphatase staining, the number of positive-staining acid phosphatase cells, and the number of radiolabeled cells. Data were compared between control and LIF-treated calvariae and the differences were analyzed using Student's t-test. Results Both alkaline phosphatase and acid phosphatase activities as well as tritiated-thymidine-labeled nuclei were demonstrated simultaneously in sections of neonatal mouse calvariae, as illustrated in Figure 1. The alkaline phosphatase activity stained blue to purple in a highly localized pericellular pattern at bone sites immediately adjacent to the mineralized bone (i.e., in osteoblasts and possibly some of their progenitor cells). Fibroblasts, their progenitor cells, and other cells in the peripheral regions of the periosteum were not stained. The acid phosphatase activity stained red and was confined mainly to the cell cytoplasm of the osteoclasts with occasional staining of the bone matrix adjacent to osteoclasts, presumably where enzymes had been released. The osteoclasts were located both at the marrow-bone interface and at the periosteal-bone surface. The autoradiographic silver grain deposits of the radiolabeled thymidine were highly localized over cell nuclei and there was very little nonspecific background grain deposition. These thymidine-labeled nuclei colocalized with alkaline phosphatase-positive cells or cells in the osteoprogenitor zone (defined as the cells located in the periosteum between the osteoblast layer and the fibroblastic cells) but not with acid phosphatase-positive cells in both LIF-treated and control bones incubated for 24 or 48 h. In bones which were pulse-labeled for an extended period of time (6 h) at the beginning of the culture, there were still no cells exhibiting thymidinelabeled nuclei which were also acid phosphatase-positive. Consequently, although this increased time of pulse-labeling at the beginning of the culture would have enabled any proliferating preosteoclasts to become labeled with [3H]thymidine and then later become incorporated into mature acid phosphatase-positive cells, this did not occur. Leukemia inhibitory factor-treated bones exhibited significant morphological changes compared to control. There was a significant increase in the area of alkaline phosphatase staining in the bone tissue incubated with LIF, compared to control tissue (Figure 2). The staining appeared to be less intense in the control bone tissue compared to the LIF-treated tissue. LIF also significantly increased the number of acid phosphatase-positive osteoclasts (Figure 3). When [3H]thymidine was incorporated for the last 4 h of the incubation period, there was a significant increase in the number of radiolabeled nuclei which colocalized with the alkaline phosphatase-positive ceils in the LIF-treated bones (Figure 4), but no significant change in the number of these labeled nuclei colocalizing with cells in the osteoprogenitor zone in the periosteum (Figure 5).
Discussion We have used a combination of enzyme histochemistry and autoradiography to demonstrate that LIF is mitogenic to osteo-
Bone Vol. 21, No. 3 September 1997:243-247
J. Cornish et al. LIF is mitogenic to osteoblasts
245
Figure 1. Photomicrograph of a calvaria treated with LIF, simultaneously demonstrating acid phosphatase-positive cells (a), alkaline phosphatase staining (b), and tritiated thymidine-positive nuclei (c). Original magnification ×900. blasts in intact bone. The findings indicate that LIF stimulates proliferation of mature alkaline phosphatase-positive osteoblasts rather than preosteoblasts located in the osteoprogenitor region. The results of this study suggest that there is no proliferation of preosteoclasts or acid phosphatase-positive osteoclasts in this neonatal mouse organ culture system. LIF significantly increased the number of acid phosphatase-positive osteoclasts; however, there were no thymidine-labeled nuclei that colocalized with these cells. Consequently, LIF may stimulate differentiation of existing osteoclasts, but the present findings indicate that LIF does not stimulate proliferation in osteoclasts or osteoclast precursor cell populations. The results of this study are consistent with our previous in
vitro and in vivo studies, in which we found increases in cell proliferation in LIF-treated osteoblast-like cells isolated from fetal rat calvariae, increases in osteoclast number and bone resorption, and increases in DNA synthesis in neonatal mouse calvariae exposed to LIF and an overall increase in bone turnover associated with increases in osteoblast and osteoclast numbers in adult mice where LIF is injected locally above the bone. Others have identified LIF as an osteoclast-stimulating factor ] which induces an increase in bone resorption in cultures of murine calvariae 9 but inhibits resorption in organ cultures of mouse fetal long bones, 11,26 models in which resorption is mostly dependent on osteoclast precursor differentiation. An in vivo study in which mice were engrafted with nonleukemic hemopoietic cells that 4-
0:0075e,m
.__."~
-~ ~
m~ ~
T: -
0.0025-
' ~ oto. 0.0000 - - - -
3-
"-~a
0oo5o-
2
//
oeo,. 0
Control
LIF
Figure 2. Effect of LIF on alkaline phosphatase activity. Data are the mean +- SEM and are expressed per millimeter length of calvaria. * Significant difference from control (p = 0.0008).
-
-
Control
LIF
Figure 3. Effect of LIF on numbers of acid phosphatase-positive cells. Data are the mean _+ SEM and are expressed per millimeter length of calvaria. * Significant difference from control (p = 0.018).
246
J. Cornish et al. LIF is mitogenic to osteoblasts
Bone Vol. 21, No. 3 September 1997:243-247
7.5-
.=_'~ 5.0-
~ 0 ~
~o z
®.;
~ e" ,13
2.5-
"N -~ n 0.0 - -
/ Control
LIF
Figure 4. Effect of LIF on number of thymidine-positive cells colocalizing with alkaline phosphatase. Thymidine incorporation was measured over the last 4 h of the incubation period. Data are the mean + SEM and are expressed per millimeter length of calvaria. * Significant difference from control (p = 0.0024).
produced high levels of LIF resulted in an increase in the deposition of new bone and a possible increase in osteoblast numbers. 17 Osteoblastogenesis is also increased in bone marrow cultures treated with LIF. 24 The effects of LIF on osteoblast proliferation in cell line cultures are variable, with a decrease being detected in murine osteogenic sarcoma UMR-106 cells,12 whereas in the murine MC3T3-E1 cells, LIF has been found to either inhibit2° or increase cell growth, 3 or exert no effect. 8 A recent study has also demonstrated that LIF is mitogenic to isolated human osteoblast-like cells. 6 Our present findings, demonstrating that LIF causes an increase in the area of alkaline phosphatase staining, confirm some previous studies. LIF increases the alkaline phosphatase activity induced by retinoic acid in preosteoblastic cell lines 23 and UMR-201. 2 In human osteoblasts, there was generally no effect of LIF on basal and 1,25-dihydroxyvitamin D3-induced alkaline phosphatase activity; yet, one culture in which the 1,25-dihydroxyvitamin D3-induced alkaline phosphatase levels were l o w - - a potentiation of the 1,25-dihydroxyvitamin D 3 effect by L I F - - w a s observed. 6 In two recent studies, LIF increased alkaline phosphate activity in undifferentiated mesenchymal progenitors of the bone marrow 24 and in MC3T3-E1 cells. 3 This is in contrast to an earlier study in the same cell line, in which LIF was found to decrease the basal alkaline phosphatase levels. 2° Similarly, LIF decreased alkaline phosphatase in rat calvarial cell cultures grown to quantify bone nodule production. 14 Since total alkaline phosphatase activity is dependent on both cell differentiation and cell number, these variable effects of LIF on this
1.00
e- ~
0.75
"~_~ ~ °.5°
ZO
a., 0.00 Control
LIF
Figure 5. Effect of LIF on number of thymidine-positive cells in the osteoprogenitor zone of the periosteum. Thymidine incorporation was measured over the last 4 h of the incubation period. Data are the mean +_ SEM and are expressed per millimeter length of calvaria.
index may just reflect the model-to-model differences in its action on cell proliferation, referred to earlier. Osteoblasts are target cells for LIF in bone, and it is likely that this is a direct effect, with receptors for LIF having been identified on mature osteoblasts and their precursors. 2 On the other hand, LIF is likely to stimulate osteoclasts indirectly to increase bone resorption via the osteoblast. Highly purified isolates of osteoclasts have not been found to respond to L I F ) 5 Leukemia inhibitory factor is synthesized by the osteoblast and its production is influenced by other cytokines; for example, it is increased by tumor necrosis factor. 2 LIF has been found to be produced both as a diffusible protein and in an immobilized form associated with the extracellular matrix in fibroblast cultures. 2] It is possible, then, that LIF made by the osteoblast is stored in the bone matrix and further LIF is released with the increased bone resorption. LIF may also interact with and influence the production of other local growth factors in bone and act as an autocrine/paracrine regulator of osteoblast and osteoclast function and differentiation, t5 In conclusion, the present data in intact bone confirm that the osteoblast is the principal bone cell showing a mitogenic response to LIF. These data, along with previous in vitro and in vivo findings, support the role of LIF as a significant cytokine involved in the local regulation of normal bone cell function.
Acknowledgments: This work was supported by the Health Research Council of New Zealand and the Auckland Medical Research Foundation, New Zealand. The authors are grateful to Lesley Scott for technical assistance.
References 1. Abe, E., Tanaka, H., Ishimi, Y., Miyaura, C., Hayashi, T., Nagasawa, H., Tomida, M., Yamaguchi, Y., Hozumi, M., and Suda, T. Differentiationinducing factor purified from conditioned medium of mitogen-treated spleen cell cultures stimulates bone resorption. Proc Natl Acad Sci USA 83:5958-5962; 1986. 2. Allan, E. H., Hilton, D. J., Brown, M. A., Evely, R. S., Yumita, S., Metcalf, D., Gough, N. M., Ng, K. W., Nicola, N. A., and Martin, T. J. Osteoblasts display receptors for response to leukemia-inhibitory factor. J Cell Physiol 145:110119; 1990. 3. Belido, T., Borba, V. Z. C., Stnhl, N., Yancopoulos, G., Jika, R. L., and Manolagas, S. C. Cilary neurotrophic factor (CNTF) and leukemia inhibitory factor (LIF) stimulate cell proliferation and increase alkaline phosphatase and interleukin-6 production by osteoblastic cells. J Bone Miner Res 10(Suppl 1):$319; 1995. 4. Bhatt, H., Brunet, L. J., and Stewart, C. L. Uterine expression of leukemia inhibitory factor coincides with the onset of blastocyst implantation. Proc Natl Acad Sci USA 88:11408-11412; 1991. 5. Cornish, J., Callon, K., King, A., Edgar, S., and Reid, I. R. The effect of leukemia inhibitory factor on bone in vivo. Endocrinology 132:1359-1366; 1993. 6. Evans, D. B., Gerber, B., and Feyen, J. H. M. Recombinant human leukemia inhibitory factor is mitogenic for human bone-derived osteoblast-like cells. Biochem Biophys Res Commun 199:220-226; 1994. 7. Gough, N. M., Gearing, D. P., King, J. A., Willson, T. A., Hilton, D. J., Nicola, N. A., and Metcalf, D. Molecular cloning and expression of the human homologue of the murine gene encoding myeloid leukemia-inhibitory factor. Proc Natl Acad Sci USA 85:2623-2627; 1988. 8. Hakeda, Y., Sudo, T., Ishizuka, S., Tanaka, K., Higashino, K., Kusuda, M., Kodama, M., and Kumegawa, M. Murine recombinant leukemia inhibitory factor modulates inhibitory effect of 1,25 dihydroxyvitamin D3 on alkaline phosphatase activity in MC3T3-E1 cells. Biochem Biophys Res Commun 175:577-582; 1991. 9. Ishimi, Y., Abe, E., He Jing, C., Miyanra, C., Hong, M. H., Oshida, M., Kurosawa, H., Yamagushi, Y., Hozumi, M., and Soda, T. Leukemia inhibitory factor/differentiation-stimulating factor (LIF/D-factor): Regulation of its production and possible roles in bone metabolism. J Cell Physiol 152:71-78; 1992.
B o n e Vol. 21, No. 3 September 1997:243-24'7
10. Kurzrock, R., Estrov, Z., Wetzler, M., Gutterman, J. U., and Talpaz, M. LIF: Not just a leukemia inhibitory factor. Endocrine Rev 12:208-217; 1991. 11. Lorenzo, J. A., Sousa, S. L., and Leahy, C. L. Leukemia inhibitory factor (LIF) inhibits basal bone resorption in fetal rat long bone cultures. Cytokine 2:1-5; 1990. 12. Lowe, C., Cornish, J., Callon, K., Martin, T. J., and Reid, I. R. Regulation of osteoblast proliferation by leukemia inhibitory factor. J Bone Miner Res 6:1277-1283; 1991. 13. Lowe, C., Cornish, J., Martin, T. J., and Reid, I. R. Effects of leukemia inhibitory factor on bone resorption and DNA synthesis in neonatal mouse calvaria. Calcif Tissue lint 49:394-397; 1991. 14. Malaval, L., Gupta, A. K., and Aubin, J. E. Leukemia inhibitory factor inhibits osteogenic differentiatiJn in rat calvafia cell cultures. Endocrinology 136: 1411-1418; 1995. 15. Martin, T. J., Allan, E. H., Evely, R. S., and Reid, I. R. Leukemia inhibitory factor and bone cell ~unction. In: Block, G. R. and Widdows, K., Eds. Polyfunctional Cytokines: IL-6 and LIF. Chichester, UK: Wiley; 1992; 141,155. 16. Metcaif, D. The leukenTia inhibitory factor (LIF). Int J Cell Cloning 9:95-108; 1991. 17. Metcalf, D. and Gearing, D. P. A fatal syndrome in mice engrafted with cells producing high levels cf the leukemia inhibitory factor (LIF). Proc Natl Acad Sci 86:5948-5952; 1989. 18. Metcalf, D., Hilton, D. L, and Nicola, N. A. Clonal analysis of the action of the murine leukemia inhibilory factor on leukemic and normal murine hemopoietic cells. Leukemia 2:216-.221; 1988. 19. Metcalf, D., Nicola, N. A., and Gearing, D. P. Effects of injected leukemia inhibitory factor on he~nopoietic and other tissues in mice. Blood 76:50-56; 1990. 20. Noda, T., Vogel, R. I_,., Hasson, D. M., Nicola, N. A., and Rodan, G. A. Leukemia inhibitory factor suppresses proliferation, alkaline phosphatase activity and type I collagen mRNA level and enhances osteopontin mRNA levels
J. C o r n i s h et al. LIF is mitogenic to osteoblasts
21.
22.
23.
24.
25.
26.
27.
247
in murine osteoblast-like (MC3T3-E1) cells. Endocrinology 127:185-190; 1990. Rathjen, P. D., Toth, S., Willis, A., Heath, J. K., and Smith, A. G. Differentiation inhibiting activity is produced in matrix-associated and diffusible forms that are generated by alternate promoter usage. Cell 62:1105-1114; 1990. Reid, I. R., Lowe, C., Cornish, J., Skinner, S. J. M., Hilton, D. J., Willson, T. A., Gearing, D. P., and Martin, T. J. Leukemia inhibitory factor: A novel bone-active cytokine. Endocrinology 126:1416-1420; 1990. Rodan, S. B., Wesolowski, G., Hilton, D. J., Nicola, N. A., and Rodan, G. A. Leukemia inhibitory factor binds with high affinity to preosteoblastic RCT-I cells and potentiates the retinoic acid induction of alkaline phosphatase. Endocrinology 127:1602-1608; 1990. Takahashi, K., Manolagas, S. C., and Jika, R. L. Leukemia inhibitory factor (LIF) stimulates osteoblastogenesis from undifferentiated mesenchymal progenitors of the bone marrow. J Bone Miner Res 10(Suppl 1):$317; 1995. Tomida, M., Yamamoto-Yamaguchi, Y., and Hozumi, M. Purification of a factor inducing differentiation of mouse myeloid leukemic MI cells from conditioned medium of mouse fibroblast L929 cells. J Biol Chem 259:1097810982; 1984. Van Beek, E., Van Der Wee-Pals, L., Van de Ruit, M., Nijweide, P., Papopoulos, S., and Lowik, C. Leukemia inhibitory factor inhibits osteoclastic resorption, growth, mineralization, and alkaline phosphatase activity in fetal mouse metacarpal bones in culture. J Bone Miner Res 8:191-199; 1993. Waring, P., Wycherley, K., Cary, D., Nicola, N., and Metcalf, D. Leukemia inhibitory factor levels are elevated in septic shock and various inflammatory body fluids. J Clin Invest 90:2031-2037; 1992.
Date Received: January 7, 1997 Date Revised: May 22, 1997 Date Accepted: May 27, 1997
Leukemia Inhibitory Factor Is Mitogenic to Osteoblasts J. C O R N I S H , 1 K. E. C A L L O N , 1 S. G. E D G A R , 2 and I. R. R E I D l Departments of 1 Medicine and 2 Pathology, University of Auckland, Auckland, New Zealand
that LIF is an inducible cytokine that is produced by a wide variety of cells and tissues including osteoblasts, z'6"27 We have previously demonstrated that LIF increases DNA synthesis in bone, and there is evidence of increases in the activity or number of both osteoclasts and osteoblasts. 12'13 In in vitro studies, we have shown that LIF stimulates bone resorption, and this is associated with a near doubling of the number of active osteoclasts in neonatal mouse calvariae. 22 At low concentrations, however, LIF predominantly stimulates DNA synthesis, 13 indicating that it may play a role in the physiological regulation of bone cell function. The LIF-stimulated increase in bone resorption is prostaglandin dependent, whereas the stimulation of DNA synthesis is independent of prostaglandin production. The effects of LIF on osteoblast cell cultures are complex, with cell growth and the synthesis of enzymes and bone matrix proteins being regulated. 2"6'8'2°'2s Furthermore, some effects of LIF are inconsistent between different models. In a fetal rat long bone organ culture system, LIF inhibits bone resorption, and the rate of DNA synthesis is not changed, 11 suggesting that the species used, the time course of experiments, and the state of osteoblast differentiation may affect the cellular response. In an in vivo study in which mice were engrafted with a hematopoietic cell line modified to produce high levels of LIF, Metcalf and Gearing demonstrated an increase in bone mass with the formation of irregularly shaped bone trabeculae in the bone marrow at the ends of the long bones, as well as some evidence of cortical bone resorption. 17 In another in vivo study 19 involving repeated intraperitoneal injections of LIF, no specific bone changes were noted. We have demonstrated that LIF accelerates bone turnover locally in a prostaglandin-independent manner in normal mice studied in vivo. 5 In light of our earlier studies in which there were increases in thymidine incorporation as well as in indices of both osteoblast and osteoclast activity following exposure to LIF in vitro, the present study set out to determine in which cell population(s) the increase in DNA synthesis occurs. We used autoradiography to localise [3H]thymidine incorporation and enzyme histochemistry to identify mature osteoblasts and osteoclasts. These techniques also allow a detailed description of the effects of LIF on the activity and distribution of these cell types in intact bone.
Leukemia inhibitory factor (LIF) regulates cell growth and is produced by a variety of tissues, including bone. Previously we have shown that recombinant human LIF induced an increase in osteoclast number, bone formation, and DNA synthesis. In the present study, we have defined the cells in intact bone at which the proliferative effects of LIF occur, using simultaneous enzyme histochemistry and autoradiographic techniques. The area of alkaline phosphatase-positive staining was increased twofold (p = 0.0008) and the number of [3H]thym~dine-positive cells was increased twofold (p = 0.0024) in I IF-treated bones. The radiolabeled cells either colocalized with alkaline phosphatase or were in the osteoprogenitor regi,~n. They were not found in the acid phosphatase-positive staining osteoclasts. These results indicate that cells which have a mitogenic response to LIF are bone-forming rather than bone-resorbing cells. (Bone 21: 243-247; 1997) © 1997 by Elsevier Science Inc. All rights reserved. Key Words: Leukemia inhibitory factor; Calvaria; Histochemical; Alkaline phosphal~se; Acid phosphatase; Autoradiography.
Introduction Growth factors and cytokines are important in the localized regulation of bone cell function, in an autocrine or paracrine manner. Leukemia inl~Jbitory factor (LIF), a single-chain glycoprotein (molecular weight of 58,000), is a pleiotrophic cytokine which acts on blood cells, embryonic stem cells, hepatocytes, and neuronal and adipose tissue, as well as bone. I° LIF is characterized by its ability to induce differentiation of a murine myeloid leukemic cell line; however, the molecule can promote both suppression and stimulation of proliferation and/or differentiation of different leukemic cell lines. 25 Transcription of the LIF gene has been detected at low levels in normal adult t i s s u e s . 4 L1F levels are elevated in septic shock and in various inflamrnatory body fluids, including the synovial fluid from subjects wil:h inflammatory arthritis. 27 LIF appears to be primarily a locally acting molecule, with its circulating concentrations being maintained at low levels. LIF or its transcripts have been detected in vitro in many cell types, including fibroblasts, T lymphocytes, bone marrow stromal cells, and osteoblasts. LIF production can be induced by various stimuli including lipopolysaccharide, interleukin-1, transforming growth factor-13, and tumour necrosis factor. 16 These findings suggest
Materials and Methods Materials
Recombinant murine LIF was kindly provided by Dr. Nicola and Dr. Gough, Walter and Eliza Institute (Melbourne, Australia). LIF was produced in Escherichia coli, as previously described. 7 LIF activity of 50 U/mL is defined as the concentration that induces 50% of M1 cells to exhibit differentiation. 18
Address for correspondence and reprints: Dr. J. Cornish, Department of Medicine, University of Auckland, Private Bag 92-019, Auckland, New Zealand. E-mail: [email protected]
© 1997 by ElsevierScienceInc. All rights reserved.
243
8756-3282/97/$17.00 PII $8756-3282(97)00144-0
244
J. Cornish et al. LIF is mitogenic to osteoblasts
Experimental Design Bone organ culture. The hemicalvaria of 6-day-old mice from a single litter were dissected out as described previously. 22 Calvariae were preincubated for 24 h in minimal essential medium with 1% charcoal-stripped, heat-inactivated serum, then changed to fresh medium containing LIF (300 U/mL) or vehicle. Incubation was continued for a further 24 or 48 h. There were ten calvariae in each group.
[3H]thymidine pulse label for autoradiography, The calvariae were pulse-labeled with [3H]thymidine (TRK 565; Amersham, UK) for the last hour of the incubation period, at a dose of 5 p~Ci/mL. In further experiments, the calvariae were pulselabeled with thymidine for the first 6 h of the incubation period to determine whether LIF affects proliferation of osteoblast and osteoclast precursor ceils. Histology. At the end of the incubation period, the specimens were washed in media with cold thymidine. The hemicalvariae were bisected through the parasagittal plane and used for the following enzyme histochemical studies. The tissue was fixed in cold 2% glutaraldehyde in 0.1 mol/L cacodylate buffer for 30 min at 4°C. Following fixation, the bones were washed in further cacodylate buffer and dehydrated at 4°C in glycol methacrylate in water (1 h in each of 50%, 70%, 95%, and 100%), followed by three 1-h changes in LR white resin (London Resin Company Ltd, Hants, UK). Polymerization of the resin was carried out on ice to minimize depression of enzyme activity by this exothermic reaction. Serial 1.5 Ixm sections were cut on a Reichert-Jung microtome (Austria). The sections were mounted on vectabondcoated slides and air-dried before further processing.
Staining for alkaline phosphatase. A simultaneous coupling azo-dye method was employed for demonstration of alkaline phosphatase activity. The substrate used was 5-bromo-4-chloro3-indolyl phosphate (BCIP) and the diazotized coupler employed was Nitro blue tetrazolium (NBT). Combined BCIP/NBT tablets were used (Sigma Diagnostics, St. Louis, MO). Incubations of sections were carried out for 30 min in the dark at 37°C. Because the prepared staining solution deteriorates rapidly, it was filtered and used immediately. This staining process, using new reagents, was repeated to obtain the desired intensity of staining. The sections were thoroughly washed in distilled water before proceeding to the acid phosphatase staining. Staining for acid phosphatase. A azo-dye method comparable to that described for alkaline phosphatase was employed to demonstrate acid phosphatase activity. The substrate used was naphthol AS-BI phosphate (Sigma Diagnostics) in N,N-dimethyl formamide (Sigma), and the diazotized coupler employed was Fast garnet GBC (Sigma) in acetate-tartrate buffer (pH 5.0). The sections were incubated for 1 h at 37°C. This staining process was repeated using new reagents to obtain the desired intensity of staining. Finally the sections were washed in distilled water and air-dried before processing for autoradiography. Autoradiography. Slides that had been histochemically stained were subsequently dipped in Kodak NTB2 emulsion (diluted 1:2 with distilled water) heated to 42°C. After coating, the slides were air-dried at room temperature, then stored in light-tight boxes and left to expose at 4°C for 2 weeks. The slides were developed in Kodak D19 (1:2) at 20°C for 2 min, fixed in Hypam (Ilford Australia Ltd., Sydney, Australia) rapid fixer (1:4) for 2 min, washed in distilled water for 5 min, and air-dried. Sections were counterstained with 1% Fast green.
Bone Vol. 21, No. 3 September 1997:243-247
Histomorphometry and data analysis. The slides were examined under an Olympus BX 50 microscope (Olympus Optical Co. Ltd., Tokyo, Japan) which was attached to an Osteomeasure Image Analyzer (Osteometrics, Inc., Atlanta, GA). Histomorphometric analyses were made of four adjacent fields (using a X20 objective) of parasaggital sections of hemicalvariae, enabling an examination of all levels of the calvariae from the frontal bone through the coronal suture and parietal bone to the lambdoidal suture. The following parameters per millimeter calvaria were measured: the area of alkaline phosphatase staining, the number of positive-staining acid phosphatase cells, and the number of radiolabeled cells. Data were compared between control and LIF-treated calvariae and the differences were analyzed using Student's t-test. Results Both alkaline phosphatase and acid phosphatase activities as well as tritiated-thymidine-labeled nuclei were demonstrated simultaneously in sections of neonatal mouse calvariae, as illustrated in Figure 1. The alkaline phosphatase activity stained blue to purple in a highly localized pericellular pattern at bone sites immediately adjacent to the mineralized bone (i.e., in osteoblasts and possibly some of their progenitor cells). Fibroblasts, their progenitor cells, and other cells in the peripheral regions of the periosteum were not stained. The acid phosphatase activity stained red and was confined mainly to the cell cytoplasm of the osteoclasts with occasional staining of the bone matrix adjacent to osteoclasts, presumably where enzymes had been released. The osteoclasts were located both at the marrow-bone interface and at the periosteal-bone surface. The autoradiographic silver grain deposits of the radiolabeled thymidine were highly localized over cell nuclei and there was very little nonspecific background grain deposition. These thymidine-labeled nuclei colocalized with alkaline phosphatase-positive cells or cells in the osteoprogenitor zone (defined as the cells located in the periosteum between the osteoblast layer and the fibroblastic cells) but not with acid phosphatase-positive cells in both LIF-treated and control bones incubated for 24 or 48 h. In bones which were pulse-labeled for an extended period of time (6 h) at the beginning of the culture, there were still no cells exhibiting thymidinelabeled nuclei which were also acid phosphatase-positive. Consequently, although this increased time of pulse-labeling at the beginning of the culture would have enabled any proliferating preosteoclasts to become labeled with [3H]thymidine and then later become incorporated into mature acid phosphatase-positive cells, this did not occur. Leukemia inhibitory factor-treated bones exhibited significant morphological changes compared to control. There was a significant increase in the area of alkaline phosphatase staining in the bone tissue incubated with LIF, compared to control tissue (Figure 2). The staining appeared to be less intense in the control bone tissue compared to the LIF-treated tissue. LIF also significantly increased the number of acid phosphatase-positive osteoclasts (Figure 3). When [3H]thymidine was incorporated for the last 4 h of the incubation period, there was a significant increase in the number of radiolabeled nuclei which colocalized with the alkaline phosphatase-positive ceils in the LIF-treated bones (Figure 4), but no significant change in the number of these labeled nuclei colocalizing with cells in the osteoprogenitor zone in the periosteum (Figure 5).
Discussion We have used a combination of enzyme histochemistry and autoradiography to demonstrate that LIF is mitogenic to osteo-
Bone Vol. 21, No. 3 September 1997:243-247
J. Cornish et al. LIF is mitogenic to osteoblasts
245
Figure 1. Photomicrograph of a calvaria treated with LIF, simultaneously demonstrating acid phosphatase-positive cells (a), alkaline phosphatase staining (b), and tritiated thymidine-positive nuclei (c). Original magnification ×900. blasts in intact bone. The findings indicate that LIF stimulates proliferation of mature alkaline phosphatase-positive osteoblasts rather than preosteoblasts located in the osteoprogenitor region. The results of this study suggest that there is no proliferation of preosteoclasts or acid phosphatase-positive osteoclasts in this neonatal mouse organ culture system. LIF significantly increased the number of acid phosphatase-positive osteoclasts; however, there were no thymidine-labeled nuclei that colocalized with these cells. Consequently, LIF may stimulate differentiation of existing osteoclasts, but the present findings indicate that LIF does not stimulate proliferation in osteoclasts or osteoclast precursor cell populations. The results of this study are consistent with our previous in
vitro and in vivo studies, in which we found increases in cell proliferation in LIF-treated osteoblast-like cells isolated from fetal rat calvariae, increases in osteoclast number and bone resorption, and increases in DNA synthesis in neonatal mouse calvariae exposed to LIF and an overall increase in bone turnover associated with increases in osteoblast and osteoclast numbers in adult mice where LIF is injected locally above the bone. Others have identified LIF as an osteoclast-stimulating factor ] which induces an increase in bone resorption in cultures of murine calvariae 9 but inhibits resorption in organ cultures of mouse fetal long bones, 11,26 models in which resorption is mostly dependent on osteoclast precursor differentiation. An in vivo study in which mice were engrafted with nonleukemic hemopoietic cells that 4-
0:0075e,m
.__."~
-~ ~
m~ ~
T: -
0.0025-
' ~ oto. 0.0000 - - - -
3-
"-~a
0oo5o-
2
//
oeo,. 0
Control
LIF
Figure 2. Effect of LIF on alkaline phosphatase activity. Data are the mean +- SEM and are expressed per millimeter length of calvaria. * Significant difference from control (p = 0.0008).
-
-
Control
LIF
Figure 3. Effect of LIF on numbers of acid phosphatase-positive cells. Data are the mean _+ SEM and are expressed per millimeter length of calvaria. * Significant difference from control (p = 0.018).
246
J. Cornish et al. LIF is mitogenic to osteoblasts
Bone Vol. 21, No. 3 September 1997:243-247
7.5-
.=_'~ 5.0-
~ 0 ~
~o z
®.;
~ e" ,13
2.5-
"N -~ n 0.0 - -
/ Control
LIF
Figure 4. Effect of LIF on number of thymidine-positive cells colocalizing with alkaline phosphatase. Thymidine incorporation was measured over the last 4 h of the incubation period. Data are the mean + SEM and are expressed per millimeter length of calvaria. * Significant difference from control (p = 0.0024).
produced high levels of LIF resulted in an increase in the deposition of new bone and a possible increase in osteoblast numbers. 17 Osteoblastogenesis is also increased in bone marrow cultures treated with LIF. 24 The effects of LIF on osteoblast proliferation in cell line cultures are variable, with a decrease being detected in murine osteogenic sarcoma UMR-106 cells,12 whereas in the murine MC3T3-E1 cells, LIF has been found to either inhibit2° or increase cell growth, 3 or exert no effect. 8 A recent study has also demonstrated that LIF is mitogenic to isolated human osteoblast-like cells. 6 Our present findings, demonstrating that LIF causes an increase in the area of alkaline phosphatase staining, confirm some previous studies. LIF increases the alkaline phosphatase activity induced by retinoic acid in preosteoblastic cell lines 23 and UMR-201. 2 In human osteoblasts, there was generally no effect of LIF on basal and 1,25-dihydroxyvitamin D3-induced alkaline phosphatase activity; yet, one culture in which the 1,25-dihydroxyvitamin D3-induced alkaline phosphatase levels were l o w - - a potentiation of the 1,25-dihydroxyvitamin D 3 effect by L I F - - w a s observed. 6 In two recent studies, LIF increased alkaline phosphate activity in undifferentiated mesenchymal progenitors of the bone marrow 24 and in MC3T3-E1 cells. 3 This is in contrast to an earlier study in the same cell line, in which LIF was found to decrease the basal alkaline phosphatase levels. 2° Similarly, LIF decreased alkaline phosphatase in rat calvarial cell cultures grown to quantify bone nodule production. 14 Since total alkaline phosphatase activity is dependent on both cell differentiation and cell number, these variable effects of LIF on this
1.00
e- ~
0.75
"~_~ ~ °.5°
ZO
a., 0.00 Control
LIF
Figure 5. Effect of LIF on number of thymidine-positive cells in the osteoprogenitor zone of the periosteum. Thymidine incorporation was measured over the last 4 h of the incubation period. Data are the mean +_ SEM and are expressed per millimeter length of calvaria.
index may just reflect the model-to-model differences in its action on cell proliferation, referred to earlier. Osteoblasts are target cells for LIF in bone, and it is likely that this is a direct effect, with receptors for LIF having been identified on mature osteoblasts and their precursors. 2 On the other hand, LIF is likely to stimulate osteoclasts indirectly to increase bone resorption via the osteoblast. Highly purified isolates of osteoclasts have not been found to respond to L I F ) 5 Leukemia inhibitory factor is synthesized by the osteoblast and its production is influenced by other cytokines; for example, it is increased by tumor necrosis factor. 2 LIF has been found to be produced both as a diffusible protein and in an immobilized form associated with the extracellular matrix in fibroblast cultures. 2] It is possible, then, that LIF made by the osteoblast is stored in the bone matrix and further LIF is released with the increased bone resorption. LIF may also interact with and influence the production of other local growth factors in bone and act as an autocrine/paracrine regulator of osteoblast and osteoclast function and differentiation, t5 In conclusion, the present data in intact bone confirm that the osteoblast is the principal bone cell showing a mitogenic response to LIF. These data, along with previous in vitro and in vivo findings, support the role of LIF as a significant cytokine involved in the local regulation of normal bone cell function.
Acknowledgments: This work was supported by the Health Research Council of New Zealand and the Auckland Medical Research Foundation, New Zealand. The authors are grateful to Lesley Scott for technical assistance.
References 1. Abe, E., Tanaka, H., Ishimi, Y., Miyaura, C., Hayashi, T., Nagasawa, H., Tomida, M., Yamaguchi, Y., Hozumi, M., and Suda, T. Differentiationinducing factor purified from conditioned medium of mitogen-treated spleen cell cultures stimulates bone resorption. Proc Natl Acad Sci USA 83:5958-5962; 1986. 2. Allan, E. H., Hilton, D. J., Brown, M. A., Evely, R. S., Yumita, S., Metcalf, D., Gough, N. M., Ng, K. W., Nicola, N. A., and Martin, T. J. Osteoblasts display receptors for response to leukemia-inhibitory factor. J Cell Physiol 145:110119; 1990. 3. Belido, T., Borba, V. Z. C., Stnhl, N., Yancopoulos, G., Jika, R. L., and Manolagas, S. C. Cilary neurotrophic factor (CNTF) and leukemia inhibitory factor (LIF) stimulate cell proliferation and increase alkaline phosphatase and interleukin-6 production by osteoblastic cells. J Bone Miner Res 10(Suppl 1):$319; 1995. 4. Bhatt, H., Brunet, L. J., and Stewart, C. L. Uterine expression of leukemia inhibitory factor coincides with the onset of blastocyst implantation. Proc Natl Acad Sci USA 88:11408-11412; 1991. 5. Cornish, J., Callon, K., King, A., Edgar, S., and Reid, I. R. The effect of leukemia inhibitory factor on bone in vivo. Endocrinology 132:1359-1366; 1993. 6. Evans, D. B., Gerber, B., and Feyen, J. H. M. Recombinant human leukemia inhibitory factor is mitogenic for human bone-derived osteoblast-like cells. Biochem Biophys Res Commun 199:220-226; 1994. 7. Gough, N. M., Gearing, D. P., King, J. A., Willson, T. A., Hilton, D. J., Nicola, N. A., and Metcalf, D. Molecular cloning and expression of the human homologue of the murine gene encoding myeloid leukemia-inhibitory factor. Proc Natl Acad Sci USA 85:2623-2627; 1988. 8. Hakeda, Y., Sudo, T., Ishizuka, S., Tanaka, K., Higashino, K., Kusuda, M., Kodama, M., and Kumegawa, M. Murine recombinant leukemia inhibitory factor modulates inhibitory effect of 1,25 dihydroxyvitamin D3 on alkaline phosphatase activity in MC3T3-E1 cells. Biochem Biophys Res Commun 175:577-582; 1991. 9. Ishimi, Y., Abe, E., He Jing, C., Miyanra, C., Hong, M. H., Oshida, M., Kurosawa, H., Yamagushi, Y., Hozumi, M., and Soda, T. Leukemia inhibitory factor/differentiation-stimulating factor (LIF/D-factor): Regulation of its production and possible roles in bone metabolism. J Cell Physiol 152:71-78; 1992.
B o n e Vol. 21, No. 3 September 1997:243-24'7
10. Kurzrock, R., Estrov, Z., Wetzler, M., Gutterman, J. U., and Talpaz, M. LIF: Not just a leukemia inhibitory factor. Endocrine Rev 12:208-217; 1991. 11. Lorenzo, J. A., Sousa, S. L., and Leahy, C. L. Leukemia inhibitory factor (LIF) inhibits basal bone resorption in fetal rat long bone cultures. Cytokine 2:1-5; 1990. 12. Lowe, C., Cornish, J., Callon, K., Martin, T. J., and Reid, I. R. Regulation of osteoblast proliferation by leukemia inhibitory factor. J Bone Miner Res 6:1277-1283; 1991. 13. Lowe, C., Cornish, J., Martin, T. J., and Reid, I. R. Effects of leukemia inhibitory factor on bone resorption and DNA synthesis in neonatal mouse calvaria. Calcif Tissue lint 49:394-397; 1991. 14. Malaval, L., Gupta, A. K., and Aubin, J. E. Leukemia inhibitory factor inhibits osteogenic differentiatiJn in rat calvafia cell cultures. Endocrinology 136: 1411-1418; 1995. 15. Martin, T. J., Allan, E. H., Evely, R. S., and Reid, I. R. Leukemia inhibitory factor and bone cell ~unction. In: Block, G. R. and Widdows, K., Eds. Polyfunctional Cytokines: IL-6 and LIF. Chichester, UK: Wiley; 1992; 141,155. 16. Metcaif, D. The leukenTia inhibitory factor (LIF). Int J Cell Cloning 9:95-108; 1991. 17. Metcalf, D. and Gearing, D. P. A fatal syndrome in mice engrafted with cells producing high levels cf the leukemia inhibitory factor (LIF). Proc Natl Acad Sci 86:5948-5952; 1989. 18. Metcalf, D., Hilton, D. L, and Nicola, N. A. Clonal analysis of the action of the murine leukemia inhibilory factor on leukemic and normal murine hemopoietic cells. Leukemia 2:216-.221; 1988. 19. Metcalf, D., Nicola, N. A., and Gearing, D. P. Effects of injected leukemia inhibitory factor on he~nopoietic and other tissues in mice. Blood 76:50-56; 1990. 20. Noda, T., Vogel, R. I_,., Hasson, D. M., Nicola, N. A., and Rodan, G. A. Leukemia inhibitory factor suppresses proliferation, alkaline phosphatase activity and type I collagen mRNA level and enhances osteopontin mRNA levels
J. C o r n i s h et al. LIF is mitogenic to osteoblasts
21.
22.
23.
24.
25.
26.
27.
247
in murine osteoblast-like (MC3T3-E1) cells. Endocrinology 127:185-190; 1990. Rathjen, P. D., Toth, S., Willis, A., Heath, J. K., and Smith, A. G. Differentiation inhibiting activity is produced in matrix-associated and diffusible forms that are generated by alternate promoter usage. Cell 62:1105-1114; 1990. Reid, I. R., Lowe, C., Cornish, J., Skinner, S. J. M., Hilton, D. J., Willson, T. A., Gearing, D. P., and Martin, T. J. Leukemia inhibitory factor: A novel bone-active cytokine. Endocrinology 126:1416-1420; 1990. Rodan, S. B., Wesolowski, G., Hilton, D. J., Nicola, N. A., and Rodan, G. A. Leukemia inhibitory factor binds with high affinity to preosteoblastic RCT-I cells and potentiates the retinoic acid induction of alkaline phosphatase. Endocrinology 127:1602-1608; 1990. Takahashi, K., Manolagas, S. C., and Jika, R. L. Leukemia inhibitory factor (LIF) stimulates osteoblastogenesis from undifferentiated mesenchymal progenitors of the bone marrow. J Bone Miner Res 10(Suppl 1):$317; 1995. Tomida, M., Yamamoto-Yamaguchi, Y., and Hozumi, M. Purification of a factor inducing differentiation of mouse myeloid leukemic MI cells from conditioned medium of mouse fibroblast L929 cells. J Biol Chem 259:1097810982; 1984. Van Beek, E., Van Der Wee-Pals, L., Van de Ruit, M., Nijweide, P., Papopoulos, S., and Lowik, C. Leukemia inhibitory factor inhibits osteoclastic resorption, growth, mineralization, and alkaline phosphatase activity in fetal mouse metacarpal bones in culture. J Bone Miner Res 8:191-199; 1993. Waring, P., Wycherley, K., Cary, D., Nicola, N., and Metcalf, D. Leukemia inhibitory factor levels are elevated in septic shock and various inflammatory body fluids. J Clin Invest 90:2031-2037; 1992.
Date Received: January 7, 1997 Date Revised: May 22, 1997 Date Accepted: May 27, 1997