LEUKEMIA INHIBITORY FACTOR (LIF) INHIBITS BASAL BONE RESORPTION IN FETAL RAT LONG BONE CULTURES Joseph A. Lorenzo,”
Sandra L. Sousa,
Christina
L. Leahy
Leukemia inhibitory factor (LIF) has a wide variety of biologic actions. In vivo, its net effect on bone is to increase new bone formation. Recently, the sequence of human LIF was found to differ by only a single amino acid from that of human differentiation-inducing factor (D-factor). The effects of LIF on bone appear to be complex since purified murine D-factor and recombinant LIF stimulate bone resorption in cultured newborn mouse calvaria. To examine further the responses of bone to LIF, we studied the effects of recombinant human LIF (glycosylated and nonglycosylated) and recombinant human D-factor (non-glycosylated) on resorption in another in vitro organ culture model, fetal rat long bones. Both LIF preparations and D-factor inhibited basal bone resorption rates by 25% to 44% in these cultures. Resorption rates in maximally inhibited LIF-treated cultures were similar to those in devitalized bones. Inhibitory effects typically occurred at concentrations of ~10 ng/mL (0.5 nM) for the non-glycosylated LIF and D-factor and 1000 U/mL for glycosylated LIF. Neither LIF nor D-factor blocked the resorptive response to interleukin 1 (IL 1) or parathyroid hormone (PTH) nor did they alter total DNA synthesis. Hence, their inhibitory effects appeared to be specific for the mechanisms regulating basal resorptive activity. These results demonstrate that LIF has potent inhibitory actions on basal resorption rates in these cultures. These effects may be important for the anabolic responses that LIF has on bone in vivo. In addition, they may also be involved in the interactions between inflammatory or tumor cells and bone. o 1990 by W.B. Saunders Company.
Leukemia inhibitory factor (LIF) is a recently purified human glycoprotein with multiple biologic activities.le3 LIF stimulates the differentiation of the murine myeloid leukemia cell line M 1 and is identical to the cytokine human interleukin for DA cells (HILDA),4 which supports the proliferation of the murine leukemic cell line DA-la. LIF also prevents the differentiation of pluripotential embryonic stem cells.536The activities of LIF are similar to those of differentiation-inducing factor (D-factor),3,7-9 a cytokine from murine and human cells that has been reported to simulate bone resorption in some model systems as well as Ml cell differentiation. Furthermore, the sequence of human
The
Departments of Medicine, Veterans Administration Medical Center, Newington, CT 06111 and The University of Connecticut Health Center, Farmington, CT 06032. *To whom reprint requests should be addressed at Veterans Administration Medical Center, 555 Willard Avenue, Newington, CT 06111. 0 1990 by W.B. Saunders Company. 1043-4666/90/0204-0007$05.00/O KEY
WORDS:
Bone/LIF/Resorption
LIF differs from that of human D-factor by only one amino acid. lo Hence, D-factor app ears to be a form of LIF. Metcalf and Gearing recently reported that LIF had potent effects on bone in viva.” They found that when they transplanted cells that had been genetically engineered to express high levels of LIF into irradiated mice, the bones of the mice developed marked changes. These included irregularities in the width of the cortical bone and an increase in new bone formation. They also believed that they found evidence of increased bone resorption. However, curiously, they reported finding only occasional osteoclasts. This data implied that rather complex pathways were involved in the actions of LIF on bone metabolism since osteoclasts are the principal cells that mediate bone resorption. In order to further explore the effects of LIF on bone, we studied its actions on fetal rat long bone cultures, an in vitro bone resorption model.‘* Three forms of LIF were examined, recombinant human LIF (glycosylated and non-glycosylated) and recombinant human D-factor (non-glycosylated).
CYTOKINE,
Vol. 2, No. 4 (July),
1990: pp 266-271
LIF
and bone resorption
/ 267
RESULTS At 10 and 100 ng/mL (0.5 and 5 nM) nonglycosylated LIF inhibited the 120-hr resorptive response of the cultures by 17% and 25% (Fig. 1A). D-factor at 10 and 100 ng/mL (0.5 and 5 nM) had similar effects and inhibited basal resorption by 28% and 37% (Fig. 1B). There were no significant effects of either peptide on the basal resorptive responses of 48-hr cultures. As with the purified non-glycosylated human LIF, glycosylated human recombinant LIF had only inhibitory effects on resorption in 120-hr cultures (25%
A
=
o-o
120hr
.-a
48hr
Control
Devitalized
LIF + Devitalized
Figure 2. Comparison of the effects of human (non-glycosylated) and three cycles of freeze-thawing on bone resorption in fetal rat long bone cultures.
recombinant (devitalized
For all groups, p < .Ol.
Control
0.1
1
10
100
LIF (ng/mL)
\\
IS,
\\ Control
120hr
0-m
48hr
.
.
0.1
O--O
1
10
100
D-factor(ng/mL)
Dilution of COS medium
Figure 1. Effects of (A) human recombinant LIF (B) human recombinant D-factor (non-glycosylated) recombinant LIF (glycosylated) on bone resorption bone cultures.
(non-glycosylated), and (C) human in fetal rat long
Bones were cultured for the indicated times in BGJ/BSA medium. The ordinates in A and B have been truncated. The number of experiments is 4 to 12 for all groups. *, significantly different from respective control value, p < .05. **, significantly different from respective control value, p < .O 1.
LIF IOOngimL
n = 6. **, significantly
different
from
control
LIF bone)
value,
inhibition at l/100 dilution, 1,000 U/mL, Fig. 1C). In addition, it had no effects on 48-hr cultures (data not shown). The 120-hr inhibitory effects on resorption of human recombinant LIF (non-glycosylated, 30 ng/mL) in live bones were similar to those seen in bones that were devitalized by three cycles of freeze-thawing (Fig. 2). This agent did not further inhibit basal resorptive effects in devitalized bones; its inhibitory effects on resorption occurred only in live bones. Neither purified LIF (non-glycosylated, 30 ng/mL) nor D-factor (30 ng/mL) significantly altered the resorptive response of the cultures to interleukin 1 (IL 1) (Fig. 3A). Similar responses were seen with the glycosylated LIF (l/ 1,000 dilution, 100 U/mL) (Fig. 3B). Glycosylated LIF and D-factor also failed to prevent the development of a resorptive response to parathyroid hormone (PTH) (Fig. 4 A and B). Variations in the lowest concentration of PTH that stimulates resorption in fetal rat long bones can occur. Typically, we find this threshold concentration to be between 1 and 10 ng/mL of bovine PTH (bPTH). For this reason, we examined the response of the bones to LIF and D-factor over this range of PTH doses. As shown in Fig. 4A, a dose of 10 ng/mL of PTH was needed to stimulate a resorptive response. D-factor (30 ng/mL) decreased the response to 10 ng/mL PTH (86 * 4% in controls, 57 + 10% with D-factor, p < .Ol). However, in both cases the ratio of the response to that of its respective non-PTH-treated control were similar (2.4 rt 0.1 in controls and 2.5 + 0.4 with D-factor). In contrast, in another experiment (Fig. 4B) glycosylated LIF (l/ 1,000 dilution, 100 U/mL) blocked the response of the culture to a threshold concentration of PTH (3 ng/mL) but not to higher doses. Hence, LIF and D-factor appear to have variable inhibitory effects on threshold resorptive responses to PTH. The reasons why different threshold concentrations
268 / Lorenzo, Sousa. Leahy
100~ 100
A
';;‘ 0‘ i% 2
1
CYTOKINE,
0-O 0-0
Control
O-O
D-Factor
A---A
LIF 30 ng/mL
30 ng/mL
** **
80-
f** 1
60-
?
% G
40-
? ::
20\\
0 Control
\\
’ 1
3
10
IL 1 (ng/mL)
100
B
1 O-O 0-0
2. iz : L 7~ 5 2 6 _ N
Mock LIF
**
111,000 1 i 1,000
0
r** P 6080-
/ /
1990: 266-271)
long bone cultures. Hence, glycosylation of the peptide does not appear to alter the absolute effects of this cytokine on these cultures nor does there appear to be a difference between the response to human recombinant LIF (non-glycosylated) and the response to D-factor. However, it was impossible for us to determine if glycosylation aItered the relative potency of the protein because we only examined a crude preparation of glycosylated LIF. Our finding that 120-hr basal resorption rates in bones treated with LIF were similar to those of dead bones suggests that this agent has a powerful effect on the mechanisms regulating basal resorption in these cultures. In the experiment depicted in Fig. 2, control levels of resorption were higher than in the experiments shown in Figs. 1, 3 and 4. Basal resorption rates of between 35% and 50% after 120 hr are typical for the feta1 rat long bone assay in our hands. However, basal rates of 60% to 70% are not uncommon. The reasons for these differences and the mechanisms regulating basal resorption rates are unknown. We consider the high (67%) 120-hr basal resorption rate in this
A
100
20-
00
Vol. 2, No. 4 (July
O-O 1 0-O
Control D-Factor
** (30 na/mL)
61 **
Control
1
3
10
T
IL 1 (ng/mL) 3
&--” 0-Y-i-l
Figure 3. Effects of (A) human recombinant D-factor and human recombinant LIF (non-glycosylated) and (II) human recombinant LIF (glycosylated) on the resorptive response of fetal rat long bone cultures to interleukin 1 (IL 1).
+
Ei
T/ -f-T
/
i
1 /
.-++-•----a # #
20
-
1
O.
For all groups, II = 6. **, significantly different from respective control value, p < .Ol. ##, significant inhibition by LIF or D-factor compared to respective control value, p < .Ol.
of PTH stimulate resorption in different experiments are unknown. However, within an experiment all dilutions were done in parallel from a single dilution of the PTH stock, Hence, we think it unlikely that errors in dilution or variations in the age of the PTH were responsible for the differences we observed in the effects that LIF and D-factor have on resorption at threshold concentrations of PTH. Neither D-factor (3 to 100 ng/mL) nor LIF (glycosylated, 1,000 U/mL) had any effect on DNA synthesis as measured by [3H] thymidine incorporation into the cultures (Table 1).
-iv
1
3
10
PTH (ng/mL)
B
100
2
O-O
Mock
0-O
LIF 1 / 1,000
80 i
00
I/ 1,000
**
Control
3
10
30
PTH (ng/mL)
DISCUSSION Our results demonstrate that glycosylated and nonglycosylated human recombinant LIF and non-glycosylated human recombinant D-factor have similar inhibitory effects on basal bone resorption rates in fetal rat
Figure 4. Effects of (A) murine recombinant D-factor and (B) human recombinant LIF (glycosylated) on the resorptive response of fetal rat long bone cultures to parathyroid hormone (PTH). For all groups, n = 6. **, significantly different from respective control value, p < .Ol. #, significant inhibition by LIF or D-factor compared to respective control value, p < .05.
LIF
TABLE 1. Effects of D-Factor and LIF (glycosylated) DNA synthesis in fetal rat long bone cultures.
on
Values are mean t standard error of the mean for 4 to 6 determinations per group. In experiment 2 mock conditioned medium was used at an equal concentration to that of LIF (1000 U/mL). Treatment
Experiment 1 Control D-factor (3 ng/mL) D-factor (10 ng/mL) D-factor (30 ng/mL) D-factor (100 ng/mL) Experiment 2 Control LIF (glycosylated, 1000 U/mL) Mock
cold acid insoluble total Q3
[3H]thymidine
counts
counts x 10-l
8.5 8.5 9.7 10.8 6.8
2 f i f i
2.2 1.4 2.1 1.3 0.7
2.4 + 0.4
3.0 f 0.5 3.5 f 0.3
experiment fortuitous since it better demonstrated the marked inhibitory effects that LIF had on this parameter. The inhibitory effects that LIF and D-factor have on bone resorption do not appear to result from nonspecific toxic actions since they did not prevent the resorptive response to either IL 1 or PTH and total DNA synthesis rates in the cultures were not altered. We have previously determined that basal resorption rates in fetal rat long bone cultures are decreased in bones that are incubated with generalized inhibitors of DNA synthesis such as hydroxyurea.” Our finding that LIF and D-factor do not inhibit total DNA synthesis in the cultures suggests that nonspecific DNA synthesis inhibition is not the mechanism by which they inhibit resorption. However, we cannot determine from these experiments whether LIF or D-factor has specific effects on DNA synthesis in individual populations of cells in the cultures. Crude and partially purified D-factor and recombinant murine and human LIF have been reported to stimulate bone resorption in cultured newborn mouse calvaria.8~13,‘4 Reid et a1.14 found that 2,000 to 7,500 U/mL of recombinant human and murine LIF stimulated resorption in newborn mouse calvaria cultures by a mechanism that was dependent on prostaglandin synthesis. These concentrations of LIF are similar to those used in the current experiments (10 to 100 ng/mL) since the specific activities of their peptides were 1 x IO* to 2 x 10’ U/mg. In contrast to their results, we have found that purified recombinant LIF and D-factor have only inhibitory effects on resorption in fetal rat long bone cultures. The local concentrations that LIF may achieve in vivo are unknown. However, the finding that LIF and D-factor inhibit bone resorption in fetal rat long bone cultures at concentrations as low as 0.5 nM
and bone resorption
/ 269
suggests that these agents have potent effects on bone cells. Differences between the resorptive responses of newborn mouse calvaria and fetal rat long bone cultures have been reported with other cytokines including transforming growth factor (TGF)-a15-17 and TGF-/3153’8and tumor necrosis factor (TNF)-01.‘9,20 TGF-a and TNF-a stimulate bone resorption in newborn mouse calvaria by mechanisms that are dependent on prostaglandin synthesis. However, in fetal rat long bone cultures, inhibitors of prostaglandin synthesis have little or no effect on the resorptive responses to either agent. TGF-/3 has even more divergent effects in these two bone culture systems. In cultured newborn mouse calvaria it stimulates bone resorption but in cultured fetal rat long bones it inhibits resorption. The reasons for these differences are unknown but may result from differences in the production of local cytokines by these cultures.*l It is unlikely that the inhibitory effects of LIF or D-factor on basal bone resorption resulted from a local increase in prostaglandin synthesis because prostaglandins have only stimulatory effects on resorption in 120-hr fetal rat long bone cultures.** Both LIF and D-factor induce the differentiation of the Ml leukemia cell line.1-3S7-9Hence, these factors have maturational effects on hematopoietic cells. Osteoclasts, the principal cells responsible for bone resorption, are believed to originate from hematopoietic precursors. 22,23However, it is unlikely that the inhibitory resorptive responses to LIF and D-factor result from a decrease in the rate at which osteoclasts are generated since LIF does not alter this parameter in vitro.23 LIF and D-factor may influence osteoclast function at the level of the osteoblast since osteoblasts appear to regulate the resorptive response of osteoclasts to a variety of stimuli. 24-27This effect may be mediated by an additional cytokine24 or by changes in the shape of osteoblasts lining the bone.*’ Recently, LIF was shown to have direct effects on osteoblast function. These responses occurred at concentrations that were similar to those used in our studies. In addition, LIF receptors were found on an osteoblast-like cell line but not on osteoclasts.29 Hence, it is unlikely that the effects of LIF on resorption are mediated by the direct responses of osteoclasts. The recent findings that LIF can stimulate new bone growth in vivo” suggest that the net effect of this protein on bone is anabolic. Our finding that LIF inhibits bone resorption is consistent with this hypothesis since bone resorption rates must be less than bone formation rates for an anabolic response to occur. Both LIF and D-factor are made by tumor cells.‘-337S9In addition, D-factor may also be the product of activated spleen cells.* Hence, production of LIF or D-factor may
270
/ Lorenzo,
Sousa,
CYTOKINE,
Leahy
be important in the interactions system or tumor cells and bone.
MATERIALS
between
the immune
AND METHODS
Bone Cultures Bone organ culture was performed as previously described.‘* Nineteen-day-old fetal rat forelimb bones labeled in utero with 45Ca (Amersham, Arlington Heights, IL) were dissected free of surrounding muscle, cartilage, and fibrous tissue. Bones were cultured in 0.5 mL of BGJ medium (Gibco, Grand Island N.Y.) that was supplemented with 1 mg/mL of bovine serum albumin. Cultures were incubated in 95% air, 5% CO, at 37OC. We utilized a 24-hr preculture in medium alone to remove readily exchangeable 45Ca and then cultured the bones in experimental medium for 120 hr. Bones were transferred to fresh experimental medium after 48 hr. Experiments that examined only bone resorption were terminated by placing the bones in 0.2 mL of 5% trichloroacetic acid (TCA) for 1 hr. Aliquots (0.1 mL) of medium and the TCA extract of the bones were counted for 45Ca by liquid scintillation in Polyfluor scintillation fluid (Packard, Downers Grove, IL). Bone resorption was assessed as the percentage of total 45Ca that was released into the medium. In some experiments bones were devitalized by three cycles of freeze-thawing during the preculture period to evaluate the component of 45Ca release that was not mediated by cellular mechanisms.
Vol. 2, No. 4 (July
1990: 266-271)
Research, Melbourne, Australia). It was identical to a preparation that had been previously described.’ The specific activity of the purified peptide was 1 x 10’ to 2 x 10’ U of LIF activity (in the Ml assay) per mL. Purified recombinant human D-factor was obtained from Genentech, South San Francisco, CA.” It was a gift from P. Koeffler, UCLA School of Medicine, Los Angeles, CA. Recombinant human LIF/ HILDA (glycosylated) was obtained from S. Clark, Genetics Institute, Cambridge, MA.4 Glycosylated LIF was tested as the crude conditioned medium of COS-1 transfected cells (specific activity, 1 x lo5 U/mL). One unit is the dilution of COS conditioned medium that half maximally stimulated the growth of DA- 1a cells.4 Control CM was obtained from COS cells that were transfected with a mock (empty) vector. Human recombinant IL lol, purified to homogeneity, was a gift from P. Lomedico, Hoffman-LaRoche Inc., Nutley, NJ. Bovine parathyroid hormone (PTH l-34) was from Bachem, Torrence, CA. All other reagents were from Sigma, St. Louis, MO.
Statistics Significant by ANOVA.
differences
between groups were determined
Acknowledgements Supported by grants #AR-31263 and AR-21707 from the N.I.H. and funds from the Veterans Administration.
DNA Synthesis DNA synthesis in 45Ca-labeled bones was assessed as the rate at which [3H]thymidine was incorporated into the cold acid insoluble fraction of the bones using previously described techniques.‘* Two hours prior to the end of an experiment, 1 PCi [methyl-3H]thymidine (specific activity 5 Ci/mM; Amersham, Arlington Heights, IL) was added to each culture. In these experiments, no additional cold thymidine was present in the medium. Experiments were terminated by washing the bones in saline, blotting them on filter paper and placing them in a counting vial with 0.2 mL of 5% TCA at 4°C. After 1 hr, the TCA was removed to another vial and the bones were washed with a second 0.2 mL of cold TCA. The two TCA samples were then pooled and the bones were rinsed with 1 mL of 70% ethanol, air dried, and dissolved in 0.4 mL of NCS tissue solubilizer (Amersham) at room temperature for at least 8 hr. The medium, TCA extract and the NCS digest were counted by liquid scintillation in ACS for ‘H and 45Ca. Total [3H]thymidine counts in the cold TCA-insoluble fraction of the bones were normalized for variations in bone size by dividing by the total “Ca counts in both the medium and the bones. Experiments utilized bones from a single litter or equal numbers of bones from two litters to further minimize bone size variability.
Reagents Recombinant human homogeneity, was obtained Dohme, West Point, PA) Nicola (The Walter and
LIF (non-glycosylated), purified to from G. Rodan (Merck Sharp and who originally received it from N. Eliza Hall Institute of Medical
REFERENCES 1. Gough NM, Gearing DP, King JA, Wilson TA, Hilton DJ, Nicola NA, Metcalf D (1988) Molecular cloning and expression of the human homologue of the murine gene encoding myeloid leukemiainhibitory factor. Proc Nat1 Acad Sci USA 85:2623-2627. 2. Gearing DP, Gough NM, King JA, Hilton DJ, Nicola NA, Simpson RA, Nice EC, Kelso A, Metcalf D (1987) Molecular cloning and expression of cDNA encoding a murine myeloid leukaemia inhibitory factor (LIF). EMBO J 6:3995-4002. 3. Tomida M, Yamamoto-Yamaguchi Y, Hozumi M (1984) Purification of a factor inducing differation of mouse myeloid leukemic Ml cells from conditioned medium of mouse fibroblast L929 cells. J Biol Chem 259:10978-10982. 4. Moreau J, Donaldson DD, Bennett F, Witek-Giannotti J, Clark SC, Wong GG (1988) Leukaemia inhibitory factor is identical to the myeloid growth factor human interleukin for DA cells. Nature 336:690-692. 5. Williams RL, Hilton DJ, Pease S, Willson TA, Stewart CL, Gearing DP, Wagner EF. Metcalf D. Nicola NA. Gouah NM (1988) Myeloii leukaemia inhibitory factor maintains thl development potential of embryonic stem cells. Nature 336:684-687. 6. Smith AG, Heath JK, Donaldson DD, Wong GG, Moreau J, Stahl M, Rogers D (1988) Inhibition of pluripotential embryonic stem cell differentiation by purified polypeptides. Nature 336:688-690. 7. Tomida M, Yamamoto-Yamaguchi Y, Hozumi M (1984) Characterization of a factor inducing the differentiation of mouse myeloid leukemic cells purified from conditioned medium of mouse Ehrlich ascities tumor cells. FEBS Lett 178:291-296. 8. Abe E, Tanaka H, Ishimi Y, Miyaura C, Hayashi T, Nagasawa H, Tomida M. Yamaauchi Y. Hozumi M. Suda T (1986) Differentiation-inducing factor purified from conditioned medium oi mitogen-treated spleen cell cultures stimulates bone resorption. Proc Nat1 Acad Sci USA 83:5958-5962.
LIF
and bone resorption
/ 271
9. Shiina-Ishimi Y, Abe E, Tanaka H, Suda T (1986) Synthesis of colony-stimulating factor (CSF) and differentiation inducing factor (D-factor) by osteoblastic cells, clone MC3T3-El. Biochem Biophys Res Comm 134:400-406.
19. Tashjian AH Jr, Voelkel EF, Lazzaro M, Goad D, Bosma T, Levine L (1987) Tumor necrosis factor-alpha (cachectin) stimulates bone resorption in mouse calvaria via a prostaglandin-mediated mechanism. Endocrinology 120:2029-2036.
10. Lowe DG, Nunes W, Bombara M, McCabe S, Ranges GE, Henzel W, Tomida M, Yamamoto-Yamaguchi Y, Hozumi M, Goedde1 DV (1989) Genomic cloning and heterologous expression of human differentiation-stimulating factor. DNA 8:351-359.
(1987) factor, 1468.
11. Metcalf D, Gearing DP (1989) Fatal syndrome engrafted with cells producing high levels of leukemia factor. Proc Nat1 Acad Sci USA 86:5948-5952.
21. Lorenzo JA, Sousa SL, Centrella M (1988) Interleukin-1 combination with transforming growth factor-alpha produces hanced bone resorption in vitro. Endocrinology 123:2194-2200.
12. Lorenzo JA, Raisz not necessary for osteoclastic cultured fetal rat long bones.
LG, Hock JM (1983) DNA responses to parathyroid J Clin Invest 72:1924-1929.
in mice inhibitory synthesis hormone
is in
13. Abe E, Ishimi Y, Takahashi N, Akatsu T, Ozawa H, Yamana H, Yoshiki S, Suda T (1988) A differentiation-inducing factor produced by the osteoblastic cell line MC3T3-El stimulates bone resorption by promoting osteoclast formation. J Bone Min Res 31635-645. 14. Reid IR, Lowe C, Cornish J, Skinner SJM, Hilton DJ, Willson TA, Gearing DP, Martin TJ (1990) Leukemia inhibitory factor: a novel bone-active cytokine. Endocrinology 126:1416-1420. 15. Tashjian AH, Voelkel EF, Lazzaro M, Singer JR, Roberts AB, Derynck R, Winkler ME, Levine L (1985) Alpha and beta human transforming growth factors stimulate prostaglandin production and bone resorotion in cultured mouse calvaria. Proc Nat1 Acad Sci USA 82:4535-4538. 16. Ibbotson KJ, Harrod J, Gowen M, D’Souza S, Smith DD, Winkler ME, Derynck R, Mundy GR (1986) Human recombinant transforming growth factor alpha stimulates bone resorption and inhibits formation in vitro. Proc Nat1 Acad Sci USA 2228-2232.
20. Stashenko P, Dewhirst FE, Peros WJ, Kent RL, Ago JM Synergistic interactions between interleukin 1, tumor necrosis and lymphotoxin in bone resorption. J Immunol 138:1464in en-
22. Dietrich JW, Goodson JM, Raisz LG (1975) Stimulation of bone resorption by various prostaglandins in organ culture. Prostaglandins 10:231-240. 23. Shinar DM, Sato M, Rodan GA (1990) The hemopoietic growth factors on the generation of osteoclast-like mouse bone marrow cultures. Endocrinology 126:1728-1735.
effect of cells in
24. McSheehy PMJ, Chambers TJ (1986) Osteoblast-like cells in the presence of parathyroid hormone release a soluble factor that stimulates osteoclastic bone resorption. Endocrinology 119:16541659. 25. Thomson BM, Saklatvala J, Chambers TJ (1986) blasts mediate interleukin 1 stimulation of bone resorption osteoclasts. JExp Med 164:104-112. 26. Thomson BM, Mundy GR, necrosis factors alpha and beta induce osteoclastic bone resorption. J Immunol
Osteoby rat
Chambers TJ (1987) Tumor osteoblastic cells to stimulate 138:775-779.
27. McSheehy PM, Chambers TJ (1987) 1,25 dihydroxyvitamin D3 stimulates rat osteoblastic cells to release a soluble factor that increases osteoclastic bone resorption. J Clin Invest 80:425-429.
17. Stern PH, Krieger NS, Nissenson Winkler ME, Derynck R, Strewler GJ (1985) growth factor-alpha stimulates bone resorption 76:2016-2019.
RA, Williams RD, Human transforming in vitro. J Clin Invest
28. Rodan GA, Martin TJ (1981) Role of osteoblasts in hormonal control of bone resorption-a hypothesis. Calcif Tissue Int 33:349-351.
18. Pfeilschifter J, Seyedin SM, Mundy ing growth factor beta inhibits bone resorption cultures. J Clin Invest 82:680-685.
GR (1988) Transformin fetal rat long bone
29. Allen EH, Hilton DJ, Brown MA, Evely RS, Yumita S, Metcalf D, Gough NM, Ng KW, Nicola NA, Martin TJ (in press) Osteoblasts display receptors for and responses to leukemia inhibitory factor. J Cell Physiol.
Sandra L. Sousa,
Christina
L. Leahy
Leukemia inhibitory factor (LIF) has a wide variety of biologic actions. In vivo, its net effect on bone is to increase new bone formation. Recently, the sequence of human LIF was found to differ by only a single amino acid from that of human differentiation-inducing factor (D-factor). The effects of LIF on bone appear to be complex since purified murine D-factor and recombinant LIF stimulate bone resorption in cultured newborn mouse calvaria. To examine further the responses of bone to LIF, we studied the effects of recombinant human LIF (glycosylated and nonglycosylated) and recombinant human D-factor (non-glycosylated) on resorption in another in vitro organ culture model, fetal rat long bones. Both LIF preparations and D-factor inhibited basal bone resorption rates by 25% to 44% in these cultures. Resorption rates in maximally inhibited LIF-treated cultures were similar to those in devitalized bones. Inhibitory effects typically occurred at concentrations of ~10 ng/mL (0.5 nM) for the non-glycosylated LIF and D-factor and 1000 U/mL for glycosylated LIF. Neither LIF nor D-factor blocked the resorptive response to interleukin 1 (IL 1) or parathyroid hormone (PTH) nor did they alter total DNA synthesis. Hence, their inhibitory effects appeared to be specific for the mechanisms regulating basal resorptive activity. These results demonstrate that LIF has potent inhibitory actions on basal resorption rates in these cultures. These effects may be important for the anabolic responses that LIF has on bone in vivo. In addition, they may also be involved in the interactions between inflammatory or tumor cells and bone. o 1990 by W.B. Saunders Company.
Leukemia inhibitory factor (LIF) is a recently purified human glycoprotein with multiple biologic activities.le3 LIF stimulates the differentiation of the murine myeloid leukemia cell line M 1 and is identical to the cytokine human interleukin for DA cells (HILDA),4 which supports the proliferation of the murine leukemic cell line DA-la. LIF also prevents the differentiation of pluripotential embryonic stem cells.536The activities of LIF are similar to those of differentiation-inducing factor (D-factor),3,7-9 a cytokine from murine and human cells that has been reported to simulate bone resorption in some model systems as well as Ml cell differentiation. Furthermore, the sequence of human
The
Departments of Medicine, Veterans Administration Medical Center, Newington, CT 06111 and The University of Connecticut Health Center, Farmington, CT 06032. *To whom reprint requests should be addressed at Veterans Administration Medical Center, 555 Willard Avenue, Newington, CT 06111. 0 1990 by W.B. Saunders Company. 1043-4666/90/0204-0007$05.00/O KEY
WORDS:
Bone/LIF/Resorption
LIF differs from that of human D-factor by only one amino acid. lo Hence, D-factor app ears to be a form of LIF. Metcalf and Gearing recently reported that LIF had potent effects on bone in viva.” They found that when they transplanted cells that had been genetically engineered to express high levels of LIF into irradiated mice, the bones of the mice developed marked changes. These included irregularities in the width of the cortical bone and an increase in new bone formation. They also believed that they found evidence of increased bone resorption. However, curiously, they reported finding only occasional osteoclasts. This data implied that rather complex pathways were involved in the actions of LIF on bone metabolism since osteoclasts are the principal cells that mediate bone resorption. In order to further explore the effects of LIF on bone, we studied its actions on fetal rat long bone cultures, an in vitro bone resorption model.‘* Three forms of LIF were examined, recombinant human LIF (glycosylated and non-glycosylated) and recombinant human D-factor (non-glycosylated).
CYTOKINE,
Vol. 2, No. 4 (July),
1990: pp 266-271
LIF
and bone resorption
/ 267
RESULTS At 10 and 100 ng/mL (0.5 and 5 nM) nonglycosylated LIF inhibited the 120-hr resorptive response of the cultures by 17% and 25% (Fig. 1A). D-factor at 10 and 100 ng/mL (0.5 and 5 nM) had similar effects and inhibited basal resorption by 28% and 37% (Fig. 1B). There were no significant effects of either peptide on the basal resorptive responses of 48-hr cultures. As with the purified non-glycosylated human LIF, glycosylated human recombinant LIF had only inhibitory effects on resorption in 120-hr cultures (25%
A
=
o-o
120hr
.-a
48hr
Control
Devitalized
LIF + Devitalized
Figure 2. Comparison of the effects of human (non-glycosylated) and three cycles of freeze-thawing on bone resorption in fetal rat long bone cultures.
recombinant (devitalized
For all groups, p < .Ol.
Control
0.1
1
10
100
LIF (ng/mL)
\\
IS,
\\ Control
120hr
0-m
48hr
.
.
0.1
O--O
1
10
100
D-factor(ng/mL)
Dilution of COS medium
Figure 1. Effects of (A) human recombinant LIF (B) human recombinant D-factor (non-glycosylated) recombinant LIF (glycosylated) on bone resorption bone cultures.
(non-glycosylated), and (C) human in fetal rat long
Bones were cultured for the indicated times in BGJ/BSA medium. The ordinates in A and B have been truncated. The number of experiments is 4 to 12 for all groups. *, significantly different from respective control value, p < .05. **, significantly different from respective control value, p < .O 1.
LIF IOOngimL
n = 6. **, significantly
different
from
control
LIF bone)
value,
inhibition at l/100 dilution, 1,000 U/mL, Fig. 1C). In addition, it had no effects on 48-hr cultures (data not shown). The 120-hr inhibitory effects on resorption of human recombinant LIF (non-glycosylated, 30 ng/mL) in live bones were similar to those seen in bones that were devitalized by three cycles of freeze-thawing (Fig. 2). This agent did not further inhibit basal resorptive effects in devitalized bones; its inhibitory effects on resorption occurred only in live bones. Neither purified LIF (non-glycosylated, 30 ng/mL) nor D-factor (30 ng/mL) significantly altered the resorptive response of the cultures to interleukin 1 (IL 1) (Fig. 3A). Similar responses were seen with the glycosylated LIF (l/ 1,000 dilution, 100 U/mL) (Fig. 3B). Glycosylated LIF and D-factor also failed to prevent the development of a resorptive response to parathyroid hormone (PTH) (Fig. 4 A and B). Variations in the lowest concentration of PTH that stimulates resorption in fetal rat long bones can occur. Typically, we find this threshold concentration to be between 1 and 10 ng/mL of bovine PTH (bPTH). For this reason, we examined the response of the bones to LIF and D-factor over this range of PTH doses. As shown in Fig. 4A, a dose of 10 ng/mL of PTH was needed to stimulate a resorptive response. D-factor (30 ng/mL) decreased the response to 10 ng/mL PTH (86 * 4% in controls, 57 + 10% with D-factor, p < .Ol). However, in both cases the ratio of the response to that of its respective non-PTH-treated control were similar (2.4 rt 0.1 in controls and 2.5 + 0.4 with D-factor). In contrast, in another experiment (Fig. 4B) glycosylated LIF (l/ 1,000 dilution, 100 U/mL) blocked the response of the culture to a threshold concentration of PTH (3 ng/mL) but not to higher doses. Hence, LIF and D-factor appear to have variable inhibitory effects on threshold resorptive responses to PTH. The reasons why different threshold concentrations
268 / Lorenzo, Sousa. Leahy
100~ 100
A
';;‘ 0‘ i% 2
1
CYTOKINE,
0-O 0-0
Control
O-O
D-Factor
A---A
LIF 30 ng/mL
30 ng/mL
** **
80-
f** 1
60-
?
% G
40-
? ::
20\\
0 Control
\\
’ 1
3
10
IL 1 (ng/mL)
100
B
1 O-O 0-0
2. iz : L 7~ 5 2 6 _ N
Mock LIF
**
111,000 1 i 1,000
0
r** P 6080-
/ /
1990: 266-271)
long bone cultures. Hence, glycosylation of the peptide does not appear to alter the absolute effects of this cytokine on these cultures nor does there appear to be a difference between the response to human recombinant LIF (non-glycosylated) and the response to D-factor. However, it was impossible for us to determine if glycosylation aItered the relative potency of the protein because we only examined a crude preparation of glycosylated LIF. Our finding that 120-hr basal resorption rates in bones treated with LIF were similar to those of dead bones suggests that this agent has a powerful effect on the mechanisms regulating basal resorption in these cultures. In the experiment depicted in Fig. 2, control levels of resorption were higher than in the experiments shown in Figs. 1, 3 and 4. Basal resorption rates of between 35% and 50% after 120 hr are typical for the feta1 rat long bone assay in our hands. However, basal rates of 60% to 70% are not uncommon. The reasons for these differences and the mechanisms regulating basal resorption rates are unknown. We consider the high (67%) 120-hr basal resorption rate in this
A
100
20-
00
Vol. 2, No. 4 (July
O-O 1 0-O
Control D-Factor
** (30 na/mL)
61 **
Control
1
3
10
T
IL 1 (ng/mL) 3
&--” 0-Y-i-l
Figure 3. Effects of (A) human recombinant D-factor and human recombinant LIF (non-glycosylated) and (II) human recombinant LIF (glycosylated) on the resorptive response of fetal rat long bone cultures to interleukin 1 (IL 1).
+
Ei
T/ -f-T
/
i
1 /
.-++-•----a # #
20
-
1
O.
For all groups, II = 6. **, significantly different from respective control value, p < .Ol. ##, significant inhibition by LIF or D-factor compared to respective control value, p < .Ol.
of PTH stimulate resorption in different experiments are unknown. However, within an experiment all dilutions were done in parallel from a single dilution of the PTH stock, Hence, we think it unlikely that errors in dilution or variations in the age of the PTH were responsible for the differences we observed in the effects that LIF and D-factor have on resorption at threshold concentrations of PTH. Neither D-factor (3 to 100 ng/mL) nor LIF (glycosylated, 1,000 U/mL) had any effect on DNA synthesis as measured by [3H] thymidine incorporation into the cultures (Table 1).
-iv
1
3
10
PTH (ng/mL)
B
100
2
O-O
Mock
0-O
LIF 1 / 1,000
80 i
00
I/ 1,000
**
Control
3
10
30
PTH (ng/mL)
DISCUSSION Our results demonstrate that glycosylated and nonglycosylated human recombinant LIF and non-glycosylated human recombinant D-factor have similar inhibitory effects on basal bone resorption rates in fetal rat
Figure 4. Effects of (A) murine recombinant D-factor and (B) human recombinant LIF (glycosylated) on the resorptive response of fetal rat long bone cultures to parathyroid hormone (PTH). For all groups, n = 6. **, significantly different from respective control value, p < .Ol. #, significant inhibition by LIF or D-factor compared to respective control value, p < .05.
LIF
TABLE 1. Effects of D-Factor and LIF (glycosylated) DNA synthesis in fetal rat long bone cultures.
on
Values are mean t standard error of the mean for 4 to 6 determinations per group. In experiment 2 mock conditioned medium was used at an equal concentration to that of LIF (1000 U/mL). Treatment
Experiment 1 Control D-factor (3 ng/mL) D-factor (10 ng/mL) D-factor (30 ng/mL) D-factor (100 ng/mL) Experiment 2 Control LIF (glycosylated, 1000 U/mL) Mock
cold acid insoluble total Q3
[3H]thymidine
counts
counts x 10-l
8.5 8.5 9.7 10.8 6.8
2 f i f i
2.2 1.4 2.1 1.3 0.7
2.4 + 0.4
3.0 f 0.5 3.5 f 0.3
experiment fortuitous since it better demonstrated the marked inhibitory effects that LIF had on this parameter. The inhibitory effects that LIF and D-factor have on bone resorption do not appear to result from nonspecific toxic actions since they did not prevent the resorptive response to either IL 1 or PTH and total DNA synthesis rates in the cultures were not altered. We have previously determined that basal resorption rates in fetal rat long bone cultures are decreased in bones that are incubated with generalized inhibitors of DNA synthesis such as hydroxyurea.” Our finding that LIF and D-factor do not inhibit total DNA synthesis in the cultures suggests that nonspecific DNA synthesis inhibition is not the mechanism by which they inhibit resorption. However, we cannot determine from these experiments whether LIF or D-factor has specific effects on DNA synthesis in individual populations of cells in the cultures. Crude and partially purified D-factor and recombinant murine and human LIF have been reported to stimulate bone resorption in cultured newborn mouse calvaria.8~13,‘4 Reid et a1.14 found that 2,000 to 7,500 U/mL of recombinant human and murine LIF stimulated resorption in newborn mouse calvaria cultures by a mechanism that was dependent on prostaglandin synthesis. These concentrations of LIF are similar to those used in the current experiments (10 to 100 ng/mL) since the specific activities of their peptides were 1 x IO* to 2 x 10’ U/mg. In contrast to their results, we have found that purified recombinant LIF and D-factor have only inhibitory effects on resorption in fetal rat long bone cultures. The local concentrations that LIF may achieve in vivo are unknown. However, the finding that LIF and D-factor inhibit bone resorption in fetal rat long bone cultures at concentrations as low as 0.5 nM
and bone resorption
/ 269
suggests that these agents have potent effects on bone cells. Differences between the resorptive responses of newborn mouse calvaria and fetal rat long bone cultures have been reported with other cytokines including transforming growth factor (TGF)-a15-17 and TGF-/3153’8and tumor necrosis factor (TNF)-01.‘9,20 TGF-a and TNF-a stimulate bone resorption in newborn mouse calvaria by mechanisms that are dependent on prostaglandin synthesis. However, in fetal rat long bone cultures, inhibitors of prostaglandin synthesis have little or no effect on the resorptive responses to either agent. TGF-/3 has even more divergent effects in these two bone culture systems. In cultured newborn mouse calvaria it stimulates bone resorption but in cultured fetal rat long bones it inhibits resorption. The reasons for these differences are unknown but may result from differences in the production of local cytokines by these cultures.*l It is unlikely that the inhibitory effects of LIF or D-factor on basal bone resorption resulted from a local increase in prostaglandin synthesis because prostaglandins have only stimulatory effects on resorption in 120-hr fetal rat long bone cultures.** Both LIF and D-factor induce the differentiation of the Ml leukemia cell line.1-3S7-9Hence, these factors have maturational effects on hematopoietic cells. Osteoclasts, the principal cells responsible for bone resorption, are believed to originate from hematopoietic precursors. 22,23However, it is unlikely that the inhibitory resorptive responses to LIF and D-factor result from a decrease in the rate at which osteoclasts are generated since LIF does not alter this parameter in vitro.23 LIF and D-factor may influence osteoclast function at the level of the osteoblast since osteoblasts appear to regulate the resorptive response of osteoclasts to a variety of stimuli. 24-27This effect may be mediated by an additional cytokine24 or by changes in the shape of osteoblasts lining the bone.*’ Recently, LIF was shown to have direct effects on osteoblast function. These responses occurred at concentrations that were similar to those used in our studies. In addition, LIF receptors were found on an osteoblast-like cell line but not on osteoclasts.29 Hence, it is unlikely that the effects of LIF on resorption are mediated by the direct responses of osteoclasts. The recent findings that LIF can stimulate new bone growth in vivo” suggest that the net effect of this protein on bone is anabolic. Our finding that LIF inhibits bone resorption is consistent with this hypothesis since bone resorption rates must be less than bone formation rates for an anabolic response to occur. Both LIF and D-factor are made by tumor cells.‘-337S9In addition, D-factor may also be the product of activated spleen cells.* Hence, production of LIF or D-factor may
270
/ Lorenzo,
Sousa,
CYTOKINE,
Leahy
be important in the interactions system or tumor cells and bone.
MATERIALS
between
the immune
AND METHODS
Bone Cultures Bone organ culture was performed as previously described.‘* Nineteen-day-old fetal rat forelimb bones labeled in utero with 45Ca (Amersham, Arlington Heights, IL) were dissected free of surrounding muscle, cartilage, and fibrous tissue. Bones were cultured in 0.5 mL of BGJ medium (Gibco, Grand Island N.Y.) that was supplemented with 1 mg/mL of bovine serum albumin. Cultures were incubated in 95% air, 5% CO, at 37OC. We utilized a 24-hr preculture in medium alone to remove readily exchangeable 45Ca and then cultured the bones in experimental medium for 120 hr. Bones were transferred to fresh experimental medium after 48 hr. Experiments that examined only bone resorption were terminated by placing the bones in 0.2 mL of 5% trichloroacetic acid (TCA) for 1 hr. Aliquots (0.1 mL) of medium and the TCA extract of the bones were counted for 45Ca by liquid scintillation in Polyfluor scintillation fluid (Packard, Downers Grove, IL). Bone resorption was assessed as the percentage of total 45Ca that was released into the medium. In some experiments bones were devitalized by three cycles of freeze-thawing during the preculture period to evaluate the component of 45Ca release that was not mediated by cellular mechanisms.
Vol. 2, No. 4 (July
1990: 266-271)
Research, Melbourne, Australia). It was identical to a preparation that had been previously described.’ The specific activity of the purified peptide was 1 x 10’ to 2 x 10’ U of LIF activity (in the Ml assay) per mL. Purified recombinant human D-factor was obtained from Genentech, South San Francisco, CA.” It was a gift from P. Koeffler, UCLA School of Medicine, Los Angeles, CA. Recombinant human LIF/ HILDA (glycosylated) was obtained from S. Clark, Genetics Institute, Cambridge, MA.4 Glycosylated LIF was tested as the crude conditioned medium of COS-1 transfected cells (specific activity, 1 x lo5 U/mL). One unit is the dilution of COS conditioned medium that half maximally stimulated the growth of DA- 1a cells.4 Control CM was obtained from COS cells that were transfected with a mock (empty) vector. Human recombinant IL lol, purified to homogeneity, was a gift from P. Lomedico, Hoffman-LaRoche Inc., Nutley, NJ. Bovine parathyroid hormone (PTH l-34) was from Bachem, Torrence, CA. All other reagents were from Sigma, St. Louis, MO.
Statistics Significant by ANOVA.
differences
between groups were determined
Acknowledgements Supported by grants #AR-31263 and AR-21707 from the N.I.H. and funds from the Veterans Administration.
DNA Synthesis DNA synthesis in 45Ca-labeled bones was assessed as the rate at which [3H]thymidine was incorporated into the cold acid insoluble fraction of the bones using previously described techniques.‘* Two hours prior to the end of an experiment, 1 PCi [methyl-3H]thymidine (specific activity 5 Ci/mM; Amersham, Arlington Heights, IL) was added to each culture. In these experiments, no additional cold thymidine was present in the medium. Experiments were terminated by washing the bones in saline, blotting them on filter paper and placing them in a counting vial with 0.2 mL of 5% TCA at 4°C. After 1 hr, the TCA was removed to another vial and the bones were washed with a second 0.2 mL of cold TCA. The two TCA samples were then pooled and the bones were rinsed with 1 mL of 70% ethanol, air dried, and dissolved in 0.4 mL of NCS tissue solubilizer (Amersham) at room temperature for at least 8 hr. The medium, TCA extract and the NCS digest were counted by liquid scintillation in ACS for ‘H and 45Ca. Total [3H]thymidine counts in the cold TCA-insoluble fraction of the bones were normalized for variations in bone size by dividing by the total “Ca counts in both the medium and the bones. Experiments utilized bones from a single litter or equal numbers of bones from two litters to further minimize bone size variability.
Reagents Recombinant human homogeneity, was obtained Dohme, West Point, PA) Nicola (The Walter and
LIF (non-glycosylated), purified to from G. Rodan (Merck Sharp and who originally received it from N. Eliza Hall Institute of Medical
REFERENCES 1. Gough NM, Gearing DP, King JA, Wilson TA, Hilton DJ, Nicola NA, Metcalf D (1988) Molecular cloning and expression of the human homologue of the murine gene encoding myeloid leukemiainhibitory factor. Proc Nat1 Acad Sci USA 85:2623-2627. 2. Gearing DP, Gough NM, King JA, Hilton DJ, Nicola NA, Simpson RA, Nice EC, Kelso A, Metcalf D (1987) Molecular cloning and expression of cDNA encoding a murine myeloid leukaemia inhibitory factor (LIF). EMBO J 6:3995-4002. 3. Tomida M, Yamamoto-Yamaguchi Y, Hozumi M (1984) Purification of a factor inducing differation of mouse myeloid leukemic Ml cells from conditioned medium of mouse fibroblast L929 cells. J Biol Chem 259:10978-10982. 4. Moreau J, Donaldson DD, Bennett F, Witek-Giannotti J, Clark SC, Wong GG (1988) Leukaemia inhibitory factor is identical to the myeloid growth factor human interleukin for DA cells. Nature 336:690-692. 5. Williams RL, Hilton DJ, Pease S, Willson TA, Stewart CL, Gearing DP, Wagner EF. Metcalf D. Nicola NA. Gouah NM (1988) Myeloii leukaemia inhibitory factor maintains thl development potential of embryonic stem cells. Nature 336:684-687. 6. Smith AG, Heath JK, Donaldson DD, Wong GG, Moreau J, Stahl M, Rogers D (1988) Inhibition of pluripotential embryonic stem cell differentiation by purified polypeptides. Nature 336:688-690. 7. Tomida M, Yamamoto-Yamaguchi Y, Hozumi M (1984) Characterization of a factor inducing the differentiation of mouse myeloid leukemic cells purified from conditioned medium of mouse Ehrlich ascities tumor cells. FEBS Lett 178:291-296. 8. Abe E, Tanaka H, Ishimi Y, Miyaura C, Hayashi T, Nagasawa H, Tomida M. Yamaauchi Y. Hozumi M. Suda T (1986) Differentiation-inducing factor purified from conditioned medium oi mitogen-treated spleen cell cultures stimulates bone resorption. Proc Nat1 Acad Sci USA 83:5958-5962.
LIF
and bone resorption
/ 271
9. Shiina-Ishimi Y, Abe E, Tanaka H, Suda T (1986) Synthesis of colony-stimulating factor (CSF) and differentiation inducing factor (D-factor) by osteoblastic cells, clone MC3T3-El. Biochem Biophys Res Comm 134:400-406.
19. Tashjian AH Jr, Voelkel EF, Lazzaro M, Goad D, Bosma T, Levine L (1987) Tumor necrosis factor-alpha (cachectin) stimulates bone resorption in mouse calvaria via a prostaglandin-mediated mechanism. Endocrinology 120:2029-2036.
10. Lowe DG, Nunes W, Bombara M, McCabe S, Ranges GE, Henzel W, Tomida M, Yamamoto-Yamaguchi Y, Hozumi M, Goedde1 DV (1989) Genomic cloning and heterologous expression of human differentiation-stimulating factor. DNA 8:351-359.
(1987) factor, 1468.
11. Metcalf D, Gearing DP (1989) Fatal syndrome engrafted with cells producing high levels of leukemia factor. Proc Nat1 Acad Sci USA 86:5948-5952.
21. Lorenzo JA, Sousa SL, Centrella M (1988) Interleukin-1 combination with transforming growth factor-alpha produces hanced bone resorption in vitro. Endocrinology 123:2194-2200.
12. Lorenzo JA, Raisz not necessary for osteoclastic cultured fetal rat long bones.
LG, Hock JM (1983) DNA responses to parathyroid J Clin Invest 72:1924-1929.
in mice inhibitory synthesis hormone
is in
13. Abe E, Ishimi Y, Takahashi N, Akatsu T, Ozawa H, Yamana H, Yoshiki S, Suda T (1988) A differentiation-inducing factor produced by the osteoblastic cell line MC3T3-El stimulates bone resorption by promoting osteoclast formation. J Bone Min Res 31635-645. 14. Reid IR, Lowe C, Cornish J, Skinner SJM, Hilton DJ, Willson TA, Gearing DP, Martin TJ (1990) Leukemia inhibitory factor: a novel bone-active cytokine. Endocrinology 126:1416-1420. 15. Tashjian AH, Voelkel EF, Lazzaro M, Singer JR, Roberts AB, Derynck R, Winkler ME, Levine L (1985) Alpha and beta human transforming growth factors stimulate prostaglandin production and bone resorotion in cultured mouse calvaria. Proc Nat1 Acad Sci USA 82:4535-4538. 16. Ibbotson KJ, Harrod J, Gowen M, D’Souza S, Smith DD, Winkler ME, Derynck R, Mundy GR (1986) Human recombinant transforming growth factor alpha stimulates bone resorption and inhibits formation in vitro. Proc Nat1 Acad Sci USA 2228-2232.
20. Stashenko P, Dewhirst FE, Peros WJ, Kent RL, Ago JM Synergistic interactions between interleukin 1, tumor necrosis and lymphotoxin in bone resorption. J Immunol 138:1464in en-
22. Dietrich JW, Goodson JM, Raisz LG (1975) Stimulation of bone resorption by various prostaglandins in organ culture. Prostaglandins 10:231-240. 23. Shinar DM, Sato M, Rodan GA (1990) The hemopoietic growth factors on the generation of osteoclast-like mouse bone marrow cultures. Endocrinology 126:1728-1735.
effect of cells in
24. McSheehy PMJ, Chambers TJ (1986) Osteoblast-like cells in the presence of parathyroid hormone release a soluble factor that stimulates osteoclastic bone resorption. Endocrinology 119:16541659. 25. Thomson BM, Saklatvala J, Chambers TJ (1986) blasts mediate interleukin 1 stimulation of bone resorption osteoclasts. JExp Med 164:104-112. 26. Thomson BM, Mundy GR, necrosis factors alpha and beta induce osteoclastic bone resorption. J Immunol
Osteoby rat
Chambers TJ (1987) Tumor osteoblastic cells to stimulate 138:775-779.
27. McSheehy PM, Chambers TJ (1987) 1,25 dihydroxyvitamin D3 stimulates rat osteoblastic cells to release a soluble factor that increases osteoclastic bone resorption. J Clin Invest 80:425-429.
17. Stern PH, Krieger NS, Nissenson Winkler ME, Derynck R, Strewler GJ (1985) growth factor-alpha stimulates bone resorption 76:2016-2019.
RA, Williams RD, Human transforming in vitro. J Clin Invest
28. Rodan GA, Martin TJ (1981) Role of osteoblasts in hormonal control of bone resorption-a hypothesis. Calcif Tissue Int 33:349-351.
18. Pfeilschifter J, Seyedin SM, Mundy ing growth factor beta inhibits bone resorption cultures. J Clin Invest 82:680-685.
GR (1988) Transformin fetal rat long bone
29. Allen EH, Hilton DJ, Brown MA, Evely RS, Yumita S, Metcalf D, Gough NM, Ng KW, Nicola NA, Martin TJ (in press) Osteoblasts display receptors for and responses to leukemia inhibitory factor. J Cell Physiol.
