JOURNAL OF BONE AND MINERAL RESEARCH Volume 5, Number 10, 1990 Mary Ann Liebert, lnc., Publishers
Bone-Resorbing Activity and Prostaglandin E Produced by Human Periodontal Ligament Cells In Vitro SHIGERU SAITO,' THOMAS J. ROSOL,* MAYUMI SAITO,' PETER W. NGAN,3 JOSEPH SHANFELD,4 and ZEEV DAVIDOVITCH'
ABSTRACT Human periodontal ligament (PDL) cells were derived from healthy premolars extracted for orthodontic treatment and were utilized for in vitro experiments in passages 4-6. Human PDL cells were seeded in tissue culture tubes and incubated with interleukin-la (IL-la), IL-10, tumor necrosis factor-a (TNF-a), interferon-y (IFN-y), indomethacin, parathyroid hormone (PTH), or their combinations, for 1 h. The medium was then replaced with serum-free BGJb medium and incubated for 24 h without further additions. Prostaglandin E (PGE) concentrations in the conditioned media (CM) were measured by radioimmunoassay, and bone-resorbing activity was measured using 45Ca-labeledneonatal mouse calvariae. The results of this study indicated that (1)unstimulated cultured PDL cells produced PGE, and PDL CM stimulated bone resorption; (2) cytokine-treated (IL-la, IL-10, and TNF-a) PDL cells had increased production of PGE and bone-resorbing activity compared to unstimulated PDL cells; (3) indomethacin completely inhibited PGE production from unstimulated PDL cells but only partially inhibited bone-resorbing activity, indicating that PDL cells produced nonprostaglandin bone-resorbing factor(s); (4) IFN-7 did not change PGE or bone-resorbing activity production by cytokine-stimulated PDL cells; and ( 5 ) PTH treatment of PDL cells in addition to cytokines (IL-la, IL-IP, and TNF-a) had additive effects on the production of bone-resorbing activity and synergistic effects on PGE production compared to cytokine treatment alone.
INTRODUCTION
T
in response to the application of mechanical force is associated with apposition and resorption of alveolar bone. Although the effector cell of bone resorption is the osteoclast, studies have shown that preosteoblasts and osteoblasts are the cells that exhibit receptors for parathyroid hormone (PTH), vitamin D metabolites, and prostaglandins (PGs) and that these cells are responsible for osteoclast recruitment.' I . * ) Periodontal ligament (PDL), the soft connective tissue interface between tooth and bone, is a source of osteoblasts in response to mechanical loading during physiologic and orthodontic tooth movement. PDL cells have been shown OOTH MOVEMENT
'Department of 'Department of 3Department of 'Department of
to respond to orthodontic force by increased proliferation as well as by differentiation into o s t e ~ b l a s t s . 'PDL ~ ~ cells are likely important in the control of bone remodeling that occurs during orthodontic tooth movement. The process of bone remodeling is affected by both systemic factors,(4ias well as by such locally produced factors as interleukin-1 (IL-l),(5,61 tumor necrosis factors (TNF),"' epidermal growth factor,(sitransforming growth f a c t o r ~ , ' and ~ ) certain PGs.''"' Davidovitch et al.(") have shown specific immunohistochemical staining for IL-la and IL-16 in PDL cells around teeth undergoing orthodontic movement in cats. In addition, the application of IL-16 to PDL cells in vitro was found to increase the synthesis of PGE and CAMP."*' Thus, cytokines or PGE synthesized
Orthodontics, School of Dentistry, Showa University, Tokyo, Japan. Veterinary Pathobiology, Ohio State University, Columbus, OH. Orthodontics, Ohio State University, Columbus, OH. Oral Biology, Ohio State University, Columbus, OH.
1013
SAITO ET AL.
1014
by PDL cells may affect the local process of bone remodeling by interacting with neighboring bone cells. In the present study, conditioned media (CM) derived from primary cultures of human PDL cells were used to characterize the effect of IL-la, IL-l& and TNF-a on the levels of PGE and bone resorption in vitro. The addition of IFN-7, indomethacin, and PTH to this response was also studied in an attempt to better understand potentially important interactions among locally produced cytokines, systemic hormones, and inhibitors of PG synthesis.
MATERIALS AND METHODS
Collection and culture of PDL cells PDL cells were prepared according to the method of Ragnarsson et al.,(131with slight modifications. Healthy premolars extracted in the course of orthodontic treatment were obtained fresh from the Oral Surgery Department, College of Dentistry, Ohio State University. The teeth were washed in Hank's balanced salt solution (HBSS), and the gingival attachments were carefully removed using a sharp scalpel. The crowns were dipped in a 5.25% sodium hypochlorite solution for 2 minutes to kill bacteria and any remaining gingival cells. The teeth were rinsed in three changes of HBSS. Each tooth was placed in a sterile 15 ml centrifuge tube with 5 ml of 0.1% collagenase (Cooper Biomedical, West Chester, PA) and incubated at 37°C for 6 h. After incubation, the tube was centrifuged at lo00 x g for 10 minutes, the tooth was removed, and the cell pellet was collected. The pellet was washed in complete medium consisting of Dulbecco's modified Eagle's medium (DMEM, GIBCO, Grand Island, NY) containing 10% bovine calf serum (Hyclone Laboratories, Logan, UT), penicillin G (100 U/ml), and streptomycin (100 pg/ml, GIBCO) and plated into 60 mm culture dishes. The dishes were incubated in a humid environment of 95% air and 5% CO, at 37°C for 48 h to permit attachment to occur; the dishes were then washed free of debris and refed daily with complete medium. To separate fibroblastic cells from epithelial cells, subculturing of the cells was performed by differential trypsinization. The dishes were washed with HBSS, and the fibroblastic cells were removed by incubation in 0.15% trypsin (GIBCO) for 7 minutes at 37°C. The removal of fibroblastic cells was closely monitored by phase-contrast microscopy. Trypsin removes the fibroblastic cells more rapidly than the epithelial cells. ( I 4 ) PDL fibroblastic cells from four donors (A through D) in passages 4-6 were used for the experiments.
Incubation and assay for PGE PDL cells (4 x los) were seeded in 12 x 75 mm plastic tissue culture tubes and allowed to attach for 6-12 h. The cytokines and PTH were obtained and used in dosages as follows: recombinant human IL-10 (kind gift from Dr. S. Gillis; Immunex, Seattle, WA), 1.0 ng/ml(S.8 x lo-" M); recombinant human IL-la (kind gift from Dr. S. Gillis),
3.0 ng/ml (1.8 x lo-" M); recombinant human TNF-a (Genentech, San Francisco, CA), 1.0 nM; recombinant human IFN-7 (Genzyme, Boston, MA), 100 U/ml (2.3 x M), and synthetic bovine PTH [bPTH-(1-34), M). Bachem, Inc., Torrence, CAI, 1.0 U/ml (3.7 x Indomethacin (kind gift from Dr. C.A. Stone, Merck Sharp and Dohme Research Laboratories, West Point, PA), when used, was added to a final concentration of 1.O pM. These agents were added to PDL culture tubes with 0.5% radioimmunoassay (RIA)-grade bovine serum albumin (BSA, Sigma Chemical Co., St. Louis, MO) for a 1 h preincubation; the media were then replaced with BGJb (Fitton-Jackson modification, GIBCO) supplemented with 0.1070 BSA, and incubation continued without any further additions for 24 h. At the end of the incubation, the PGE content in the conditioned media was measured in duplicate by radioimmunoassay as described
Bone resorption assay Bone-resorbing activity was measured using the CM produced from the incubation of PDL cells. This assay was performed as described with minor modifications.(16) Briefly, timed-pregnant mice (ND-4, Harlan Sprague-Dawley) were injected with 50 pCi 45Caat day 18 of pregnancy. Calvarial havles from neonatal mice (1 day old) were cultured with BGJb medium (0.7 ml) containing 0.1% BSA in 24-well plates for 24 h in a humid atmosphere of 95% air and 5% CO, at 37°C. The media were then discarded and replaced with either fresh BGJb medium or CM (0.7 ml) for a further 48 h incubation. Five paired calvarial halves served as treatment and control bones in each experimental group, and CM were used for the treatment groups. Quantification of bone resorption was determined as the percentage 4sCareleased into the culture medium during the second 48 h period compared to the total radioactivity in the medium and half-calvaria. Control calvarial cultures released 4-6% of the total 45Caduring the second 48 h culture period. Data are expressed as the treatment/control (T/C) ratio of percentage 45Carelease from the paired calvarial halves. PGE, (Upjohn Diagnostics, Kalamazoo, MI) was used to a final concentration of 1 x lo-' M as a positive control in all assays, which typically induced a twofold increase in 4sCa release from the mouse calvariae.
Statistical analysis of data Data are expressed as mean + standard error of the mean and were analyzed using analysis of variance (ANOVA). In all the analyses there were at least four samples in each experimental design group. The variability of PGE responses required transformations to natural logarithms. Degrees of freedom for the effects of agents and their interactions in the ANOVA were adjusted by the method of Geisser and Greenhouse.(L71 All agent comparisons were made with reference to control values using Dunnett's t-test and the appropriate error term from the ANOVA.
1015
BONE RESORPTION BY PDL CELLS
RESULTS Unstimulated PDL cells from four donors (A through D) produced PGE after 24 h of incubation that ranged from 980 to 2650 pg per lo4 cells (2.7 x 10” to 7.4 x M). Treatment of the PDL cells with IL-10 further increased the amount of PGE production, which was significantly diferent from that of unstimulated PDL cells @ < 0.01, Fig. 1). PGE production by IL-10-treated cells at 24 h was 5000-20,000 pg per lo4 cells (1.4 x lo-* to 5.6 x M). Unstimulated CM from four donors contained significant (p < 0.01) bone-resorbing activity after 24 h of incubation, and treatment of PDL cells with IL-10 significantly (p < 0.01) increased the level of bone-resorbing activity in the CM. The effect of conditioned medium on bone resorption was related to the amount of CM added to the calvarial
cultures. Conditioned medium produced by PDL cells from donor D was diluted 1 :2, 1 :4, and 1 :8 with BGJb medium and stimulated bone resorption @ < 0.01) as follows: 2.37 * 0.18, 1.51 f 0.04, and 1.28 + 0.08, respectively. Figure 2 shows the effects of IL-10 (1 ng/ml), IFN-y (100 U/ml), and their combination on PGE production and bone-resorbing activity produced by PDL cells. Treatment of PDL cells with IFN-y and IL-10 did not significantly change the amount of PGE or bone-resorbing activity production compared to PDL cells treated with It-10 alone (Fig. 2). Similar results were obtained from experiments performed using IL-10 instead of 1L-10 (data not shown) . The experiment shown in Fig. 3 demonstrates the effect of indomethacin on PGE production and bone-resorbing activity in CM from unstimulated and IL-lp-treated PDL
4‘0 3.5
I
1 T
..
0UNSTIMULATED a l L - 1 P TREATMENT
2
. I
T
v
3.0
+.
2.5
< 2.0
0 Y)
1.5 1 .o
DONOR
A
B
C
..
D
DONOR
A
e
~ ( 0 . 0 1SIGNIFICANT INCREASE IN BONE RESORPTION p(0.01 SIGNIFICANT INCREASE COMPARED TO UNSTIMULATED CELLS
‘p(0.01 DIFFERENT FROM UNSTIMUUTED CELLS
FIG. 1. Prostaglandin E and bone-resorbing activity produced by unstimulated and IL-lbstimulated ( 1 ng/ml) human periodontal ligament cells after 24 h of culture (mean + SEM).
15T
3.5
i
--e
UNSTlYUlATED
IL-1@
L
1
-t
IFN-7
UNSTIMULATED IL-1@
IL-16
+ IFN-7
p(O.01 DIFFERENT FROM U N S T I M U U T E D AND IFN-7-TREATED
CELLS
T
-
t-
IFN-1
IL- 1@
+
IFN-Y
~ ( 0 . 0 1DIFFERENT FROM UNSTIMUUTED C E U S
FIG. 2. Prostaglandin E and bone-resorbing activity produced by unstimulated, IL-I0 (1 nglml), IFN-7 (100 U/ ml), or IL-10 and IFN-y-treated human periodontal ligament cells (donor B).
SAITO ET AL.
1016
cells. Indomethacin inhibited PGE production from unstimulated PDL cells and completely abolished the increase in PGE production induced by IL-IP treatment of PDL cells at 24 h. However, treatment of PDL cells with indomethacin caused only a partial inhibition of bone-resorbing activity production at 24 h in 1L-lp-treated PDL cells (Fig. 3). Similar results were obtained from experiments performed using IL-la instead of IL-I@(data not shown). The interactive effects of cytokines and PTH on PGE and bone-resorbing activity production by PDL cells is shown in Fig. 4. Treatment of PDL cells with PTH and IL-la, IL-16, or TNF-a caused a synergistic increase in PGE production by PDL cells. PTH treatment increased the production of bone-resorbing activity by PDL cells. The three cytokines (IL-la, IL-10, and TNF-a) significantly increased bone-resorbing activity produced by PDL
cells at 24 h compared to unstimulated PDL cells. Combined treatment of PDL cells with PTH and the cytokines (IL-Ip and TNF-a) further increased the bone-resorbing activity of PDL CM.
DISCUSSION This investigation demonstrated that unstimulated human PDL cells synthesize nonprostaglandin bone-resorbing factors and PGE. In addition, the PDL cells responded to the administration of cytokines (IL-I@, IL-la, and TNF-a) with further increases in the synthesis of PGE and bone-resorbing activity compared to unstimulated PDL cells.
3.0
i
T
i
.. + 4 UNSTIMULATED
IL- 18
INDO
UNSTIMULATED
INDO
IL-lg
IL- 1 /3 INDO
'p(0.01 DIFFERENT FROM UNSTIUUUTEO CELLS
IL-lg
+
INDO *p(0.01 DIFFERENT FROM UNSTIMULATED AND INDO-TREATED CELLS p(0.05 DIFFERENT FROM INDO OR IL-l@-TREATED CELLS
0.
FIG. 3. Prostaglandin E and bone-resorbing activity produced by unstimulated, IL-IP (1 ng/ml), indomethacin (1.0 rM), or IL-lP and indomethacin (IND0)-treated human periodontal ligament cells (donor D).
A
T
..
*p(O.Ol DIFFERENT FROM UNSTIUUUTEO CELLS ~ ( 0 . 0 1DIFFERENT FROM CYTOKINE-TREATED CELLS
M
*p(0.01 and **p(0.05 DIFFERENT FROM UNSTIMUUTED hPDL CELLS ' ~ ( 0 . 0 5 and MP(O.Dl
DIFFERENT FROM CYTOKINE-TREATED CELLS
FIG. 4. Prostaglandin E and bone-resorbing activity produced by unstimulated, IL-la (3 ng/ml), IL-10 (1 ng/ml), TNF-a (1.0 nM), and PTH-treated (1.0 U/ml) human periodontal ligament cells (donor D).
1017
BONE RESORPTION BY PDL CELLS The addition of indomethacin completely inhibited PGE production and partially inhibited the production of boneresorbing activity by PDL cells. The increase of PGE in the PDL cells was due to de novo production since it was inhibited by the addition of indomethacin to the medium. Meghji et al.(18)reported that unstimulated CM derived from mouse osteoblastic cells caused bone resorption that was partially inhibited by indomethacin, dexamethasone, and nordihydroguaiaretic acid. In addition, Lerner and Hanstrom(I9) reported that human gingival fibroblasts produce nonprostaglandin bone-resorbing factors in vitro. Their findings and ours suggest that PGE is not the only factor present in PDL, osteoblast, or fibroblast C M that is responsible for stimulating bone resorption. In fact, the PGE concentration of unstimulated CM at 24 h of incubation was 2.7 x lo-' to 7.4 x 10'' M. Despite these low levels of PGE production, the stimulation of bone resorption (1.8- to 2.7-fold increase in " T a release) caused by these CM were almost the same or even higher than those caused by addition of PGE, M), which was used as a positive control (data not shown). Other factors, such as cytokines, growth factors, and leukotrienes that are synthesized by PDL cells, may contribute to the enhancement of bone resorption by interacting with PGE. Mohammed et reported on the interaction between prostaglandins and leukotrienes on orthodontic tooth movement. They found a significant inhibition of tooth movement in rats injected with indomethacin. However, since a previous has shown that inhibition of the cyclooxygenase pathway potentiates the conversion of arachidonic acid to leukotrienes, it is conceivable that the addition of indomethacin may have potentiated the synthesis of leukotrienes. The administration of IFN-7 did not change unstimulated or IL-1-stimulated PGE production or bone-resorbing activity from PDL cells. These results correspond with the finding by Friteau et al.,'zz) since they reported that IFN-y did not suppress IL- I@-stimulated PGE production from human fibroblasts. However, IFN-y did suppress ILI@-stimulated PCiE production in human monocytesiZz) and endogenous prostaglandin synthesis in mouse calvariae,lz3)suggesting that the effect of IFN-7 on PGE production is different depending on the particular cell type. IFN-7 has also been shown to have variable effects on PGE production in human osteoblasts."") PTH is a potent stimulator of bone resorption that does not depend on PGE production for its a ~ t i v i t y . ~How~~] ever, a recent study'z6' has shown that the increase in inosito1 triphosphate by P T H is abolished by the addition of indomethacin, suggesting that PGE mediates the effects of P T H on inositol phosphate production. In studies with bone explants and isolated bone cells, it was found that CAMP alone was not the exclusive second messenger for the action of PTH on bone resorption and that calcium plays an important role in this p r o c e s ~ . In ( ~ our ~ ~ study a slight increase in PGE production from PDL cells was observed following the administration of PTH. It is possible that PGE increases in response to PTH administration are transient and that the level of PGE synthesized and secreted into the medium is relatively low. I n contrast, the
addition of P T H synergistically stimulated the cytokine-induced PGE production. This result corresponds with that of Tatakis et al.,'z8)who found synergistic PGE, elevation between IL-la and PTH by mouse osteoblastic cells. It is possible that P T H can influence the stability of surface membrane receptors, as a calcium ionophore, and make them more available to the binding of cytokine molecules. Therefore, an additive effect on bone-resorbing activity caused by P T H and cytokines may be due to synergistic PGE synthesis. However, P T H increased the production of nonprostaglandin bone-resorbing factors from unstimulated PDL cells since the increase in PGE production in PTH-treated cells was not great enough to account for the increase in bone-resorbing activity. In conclusion, human PDL cells produced bone-resorbing activity and PGE in vitro that was increased after treatment with cytokines (IL-la, IL-10, and TNF-a). Indomethacin completely inhibited PGE production and partially inhibited bone-resorbing activity. IFN-y did not alter PGE production or bone-resorbing activity produced by unstimulated PDL cells. This indicated that PDL cells produced bone-resorbing activity that was due to both nonprostaglandin factors and PGE. The nonprostaglandin factors and their mechanisms of induced bone resorption may differ depending on the time period of collection of conditioned medium from the PDL cells. Future investigations are necessary to identify the nonprostaglandin factors and the time course of their secretion from human PDL cells.
ACKNOWLEDGMENTS This work was supported by the National Institutes of Health research grant DE-08428.
REFERENCES 1. Rodan GA, Martin T J 1981 Role of osteoblasts in hormonal
2.
3.
4.
5.
6.
7.
control of bone resorption. A hypothesis. Calcif Tissue Int 33~349-351. Rouleau MF, Mitchell J, Goltzman D 1988 In vivo distribution of parathyroid hormone receptors in bone: Evidence that a predominant osseous target cell is not the mature osteoblast. Endocrinology 123:187-191. Roberts WE, Chase DC 1981 Kinetics of cell proliferation and migration associated with orthodontically-induced osteogenesis. J Dent Res 60:174-181. Aubauch GD, Marx SJ, Speigel AM 1981 Parathyroid hormone, calcitonin, and calciferols. In: Williams R H (ed.) Textbook of Endocrinology, 6th ed. W.B.Saunders, Philadelphia, pp. 932-1031. Gowen M, Wood DD, lhrie EG, McGuire MKB, Russell RGG 1983 An interleukin I like factor stimulates bone resorption in vitro. Nature 306:378-380. Dewhirst FE, Stashenko P , Mole JE, Tsurumachi T 1985 Purification and partial sequence of human osteoclast-activating factor: Identity with interleukin I-beta. J lmmunol 135~2562-2568. Bertolini DR, Nedwin GE, Bringmam TS, Smith DD, Mundy GR 1986 Stimulation of bone resorption and inhibition of
1018
SAITO ET AL.
bone formation in vitro by human tumor necrosis factors. stimulate bone resorption in vitro. J Periodont Res 22:284Nature 3195 16-5 18. 289. 8. Raisz LG, Simmons HA, Sandberg AL, Canalis E 1980 20. Mohammed AH, Tatakis DN, Dziak R 1989 Leukotrienes in Direct stimulation of bone resorption by epidermal growth orthodontic tooth movement. Am J Orthodont Dentofac factor. Endocrinology 107:270-273. Orthop 99231-237. 9. Tashjian AH Jr, Voekel EF, Lazzaro M, Singer FR, Roberts 21. Raisz LG, Martin TJ 1983 Prostaglandins in bone and minAB, Derynck R, Winkler ME, Levine L 1985 Alpha and beta eral metabolism. In: William AP (ed.) Bone and Mineral Rehuman transforming growth factors stimulate prostaglandin search Annual 2. Elsevier, Amsterdam, pp. 286-310. production and bone resorption in cultured mouse calvaria. 22. Friteau L, Francesconi E, Land D, Dugas B, Damais C 1988 Proc Natl Acad Sci USA 82:4535-4538. Opposite effect of interferon-gamma on PGE, release from 10. Dietrich JW, Goodson JM, Raisz LG 1975 Stimulation of interleukin-1-stimulated human monocytes or fibroblasts. bone resorption by various prostaglandins in organ culture. Biochem Biophys Res Commun 157:1197-1204. Prostaglandins 10:231-238. 23. Peterlik M, Hoffman 0, Swetly P, Klaushofer K, Koller K 11 Davidovitch Z, Nicolay 0, Ngan PW, Shanfeld JL 1988 1985 Recombinant y-interferon inhibits prostaglandin-mediNeurotransmitters, cytokines and the control of alveolar ated and parathyroid hormone induced bone resorption in bone remodeling in orthodontics. Dent Clin North Am 32: cultured neonatal mouse calvaria. FEBS Lett 185:287-290. 41 1-435. 24. Gowen M, MacDonald BR, Russell RGG 1988 Actions of re12 Ngan PW, Zadeh YZ, Shanfeld J, Davidovitch 2 1988 The combinant human y-interferon and tumor necrosis factor CY effect of interleukin-lp and parathyroid hormone on cyclic on the proliferation and osteoblastic characteristics of hunucleotide and prostaglandin levels in human periodontal man trabecular bone cells in vitro. Arthritis Rheum 31:1500ligament fibroblasts in vitro. In: Davidovitch Z (ed.) The 1507. Biological Mechanisms of Tooth Eruption and Root Resorp- 25. Dewhirst FE, Ago JM, Peros WJ, Stashenko P 1987 Synertion. EBSCO Media, Birmingham, AL, pp. 261-267. gism between parathyroid hormone and interleukin I in stim13. Ragnarsson B, Carr G, Daniel J C 1985 Isolation and growth ulating bone resorption in organ culture. J Bone Min Res 2: of human periodontal ligament cells in vitro. J Dent Res 64: 127-1 34. 1026-1030. 26. Sandy JR, Meghji S, Farndale RM, Meikle MC 1989 Dual 14. Owens RB 1974 Glandular epithelial cells from mice: A elevation of cyclic AMP and inositol phosphates in response method for selective cultivation. JNCI 52:1375-1378. to mechanical deformation of murine osteoblasts. Biochim 15. Ngan PW, Crock B, Varghese J, Lanese R, Shanfeld J, Biophys Acta 1010:265-269. Davidovitch Z 1988 Immunohistochemical assessment of the 27. Herrmann-Erlee MPM, Van der Meer JM, Lowik CWGM, effect of chemical and mechanical stimuli on CAMP and Van Leeuwen JPTM, Boonekamp PM 1988 Different roles prostaglandin E levels in human gingival fibroblasts in vitro. for calcium and cyclic AMP in the action of PTH: Studies in Arch Oral Biol 33:163-174. bone explants and isolated bone cells. Bone 993-100. 16. Rosol TJ, Capen CC 1988 Inhibition of in vitro bone resorp- 28. Tatakis DN, Schneeberger G, Dziak R 1988 Recombinant intion by a parathyroid hormone receptor antagonist in the terleukin 1 stimulates prostaglandin E, production by osteocanine adenocarcinoma model of humoral hypercalcemia of blastic cells: Synergy with parathyroid hormone. Calcif Tismalignancy. Endocrinology 122:2098-2102. sue lnt 42:358-362. 17. Geisser S, Greenhouse SW 1958 An extension of box’s results on the use of the F distribution in multivariate analysis. Ann Address reprint requests to: Math Stat 29:885-891. Dr. Thomas J . Rosol 18. Meghji S, Sandy JR, Scutt AM, Harvey W, Harris M 1988 Department of Veterinary Pathobiology Heterogeneity of bone resorbing factors produced by unstimThe Ohio State University ulated murine osteoblasts in vitro and in response to stimula1925 Coffey Road, Columbus, OH 43210 tion by parathyroid hormone and mononuclear cell factors. Archs Oral Biol 33:773-778. 19. Lerner U, Hanstrom L 1987 Human gingival fibroblasts se- Received for publication August 31, 1989; in revised form May crete nondialyzable, prostanoid-independent products which 17, 1990; accepted May 18, 1990.
Bone-Resorbing Activity and Prostaglandin E Produced by Human Periodontal Ligament Cells In Vitro SHIGERU SAITO,' THOMAS J. ROSOL,* MAYUMI SAITO,' PETER W. NGAN,3 JOSEPH SHANFELD,4 and ZEEV DAVIDOVITCH'
ABSTRACT Human periodontal ligament (PDL) cells were derived from healthy premolars extracted for orthodontic treatment and were utilized for in vitro experiments in passages 4-6. Human PDL cells were seeded in tissue culture tubes and incubated with interleukin-la (IL-la), IL-10, tumor necrosis factor-a (TNF-a), interferon-y (IFN-y), indomethacin, parathyroid hormone (PTH), or their combinations, for 1 h. The medium was then replaced with serum-free BGJb medium and incubated for 24 h without further additions. Prostaglandin E (PGE) concentrations in the conditioned media (CM) were measured by radioimmunoassay, and bone-resorbing activity was measured using 45Ca-labeledneonatal mouse calvariae. The results of this study indicated that (1)unstimulated cultured PDL cells produced PGE, and PDL CM stimulated bone resorption; (2) cytokine-treated (IL-la, IL-10, and TNF-a) PDL cells had increased production of PGE and bone-resorbing activity compared to unstimulated PDL cells; (3) indomethacin completely inhibited PGE production from unstimulated PDL cells but only partially inhibited bone-resorbing activity, indicating that PDL cells produced nonprostaglandin bone-resorbing factor(s); (4) IFN-7 did not change PGE or bone-resorbing activity production by cytokine-stimulated PDL cells; and ( 5 ) PTH treatment of PDL cells in addition to cytokines (IL-la, IL-IP, and TNF-a) had additive effects on the production of bone-resorbing activity and synergistic effects on PGE production compared to cytokine treatment alone.
INTRODUCTION
T
in response to the application of mechanical force is associated with apposition and resorption of alveolar bone. Although the effector cell of bone resorption is the osteoclast, studies have shown that preosteoblasts and osteoblasts are the cells that exhibit receptors for parathyroid hormone (PTH), vitamin D metabolites, and prostaglandins (PGs) and that these cells are responsible for osteoclast recruitment.' I . * ) Periodontal ligament (PDL), the soft connective tissue interface between tooth and bone, is a source of osteoblasts in response to mechanical loading during physiologic and orthodontic tooth movement. PDL cells have been shown OOTH MOVEMENT
'Department of 'Department of 3Department of 'Department of
to respond to orthodontic force by increased proliferation as well as by differentiation into o s t e ~ b l a s t s . 'PDL ~ ~ cells are likely important in the control of bone remodeling that occurs during orthodontic tooth movement. The process of bone remodeling is affected by both systemic factors,(4ias well as by such locally produced factors as interleukin-1 (IL-l),(5,61 tumor necrosis factors (TNF),"' epidermal growth factor,(sitransforming growth f a c t o r ~ , ' and ~ ) certain PGs.''"' Davidovitch et al.(") have shown specific immunohistochemical staining for IL-la and IL-16 in PDL cells around teeth undergoing orthodontic movement in cats. In addition, the application of IL-16 to PDL cells in vitro was found to increase the synthesis of PGE and CAMP."*' Thus, cytokines or PGE synthesized
Orthodontics, School of Dentistry, Showa University, Tokyo, Japan. Veterinary Pathobiology, Ohio State University, Columbus, OH. Orthodontics, Ohio State University, Columbus, OH. Oral Biology, Ohio State University, Columbus, OH.
1013
SAITO ET AL.
1014
by PDL cells may affect the local process of bone remodeling by interacting with neighboring bone cells. In the present study, conditioned media (CM) derived from primary cultures of human PDL cells were used to characterize the effect of IL-la, IL-l& and TNF-a on the levels of PGE and bone resorption in vitro. The addition of IFN-7, indomethacin, and PTH to this response was also studied in an attempt to better understand potentially important interactions among locally produced cytokines, systemic hormones, and inhibitors of PG synthesis.
MATERIALS AND METHODS
Collection and culture of PDL cells PDL cells were prepared according to the method of Ragnarsson et al.,(131with slight modifications. Healthy premolars extracted in the course of orthodontic treatment were obtained fresh from the Oral Surgery Department, College of Dentistry, Ohio State University. The teeth were washed in Hank's balanced salt solution (HBSS), and the gingival attachments were carefully removed using a sharp scalpel. The crowns were dipped in a 5.25% sodium hypochlorite solution for 2 minutes to kill bacteria and any remaining gingival cells. The teeth were rinsed in three changes of HBSS. Each tooth was placed in a sterile 15 ml centrifuge tube with 5 ml of 0.1% collagenase (Cooper Biomedical, West Chester, PA) and incubated at 37°C for 6 h. After incubation, the tube was centrifuged at lo00 x g for 10 minutes, the tooth was removed, and the cell pellet was collected. The pellet was washed in complete medium consisting of Dulbecco's modified Eagle's medium (DMEM, GIBCO, Grand Island, NY) containing 10% bovine calf serum (Hyclone Laboratories, Logan, UT), penicillin G (100 U/ml), and streptomycin (100 pg/ml, GIBCO) and plated into 60 mm culture dishes. The dishes were incubated in a humid environment of 95% air and 5% CO, at 37°C for 48 h to permit attachment to occur; the dishes were then washed free of debris and refed daily with complete medium. To separate fibroblastic cells from epithelial cells, subculturing of the cells was performed by differential trypsinization. The dishes were washed with HBSS, and the fibroblastic cells were removed by incubation in 0.15% trypsin (GIBCO) for 7 minutes at 37°C. The removal of fibroblastic cells was closely monitored by phase-contrast microscopy. Trypsin removes the fibroblastic cells more rapidly than the epithelial cells. ( I 4 ) PDL fibroblastic cells from four donors (A through D) in passages 4-6 were used for the experiments.
Incubation and assay for PGE PDL cells (4 x los) were seeded in 12 x 75 mm plastic tissue culture tubes and allowed to attach for 6-12 h. The cytokines and PTH were obtained and used in dosages as follows: recombinant human IL-10 (kind gift from Dr. S. Gillis; Immunex, Seattle, WA), 1.0 ng/ml(S.8 x lo-" M); recombinant human IL-la (kind gift from Dr. S. Gillis),
3.0 ng/ml (1.8 x lo-" M); recombinant human TNF-a (Genentech, San Francisco, CA), 1.0 nM; recombinant human IFN-7 (Genzyme, Boston, MA), 100 U/ml (2.3 x M), and synthetic bovine PTH [bPTH-(1-34), M). Bachem, Inc., Torrence, CAI, 1.0 U/ml (3.7 x Indomethacin (kind gift from Dr. C.A. Stone, Merck Sharp and Dohme Research Laboratories, West Point, PA), when used, was added to a final concentration of 1.O pM. These agents were added to PDL culture tubes with 0.5% radioimmunoassay (RIA)-grade bovine serum albumin (BSA, Sigma Chemical Co., St. Louis, MO) for a 1 h preincubation; the media were then replaced with BGJb (Fitton-Jackson modification, GIBCO) supplemented with 0.1070 BSA, and incubation continued without any further additions for 24 h. At the end of the incubation, the PGE content in the conditioned media was measured in duplicate by radioimmunoassay as described
Bone resorption assay Bone-resorbing activity was measured using the CM produced from the incubation of PDL cells. This assay was performed as described with minor modifications.(16) Briefly, timed-pregnant mice (ND-4, Harlan Sprague-Dawley) were injected with 50 pCi 45Caat day 18 of pregnancy. Calvarial havles from neonatal mice (1 day old) were cultured with BGJb medium (0.7 ml) containing 0.1% BSA in 24-well plates for 24 h in a humid atmosphere of 95% air and 5% CO, at 37°C. The media were then discarded and replaced with either fresh BGJb medium or CM (0.7 ml) for a further 48 h incubation. Five paired calvarial halves served as treatment and control bones in each experimental group, and CM were used for the treatment groups. Quantification of bone resorption was determined as the percentage 4sCareleased into the culture medium during the second 48 h period compared to the total radioactivity in the medium and half-calvaria. Control calvarial cultures released 4-6% of the total 45Caduring the second 48 h culture period. Data are expressed as the treatment/control (T/C) ratio of percentage 45Carelease from the paired calvarial halves. PGE, (Upjohn Diagnostics, Kalamazoo, MI) was used to a final concentration of 1 x lo-' M as a positive control in all assays, which typically induced a twofold increase in 4sCa release from the mouse calvariae.
Statistical analysis of data Data are expressed as mean + standard error of the mean and were analyzed using analysis of variance (ANOVA). In all the analyses there were at least four samples in each experimental design group. The variability of PGE responses required transformations to natural logarithms. Degrees of freedom for the effects of agents and their interactions in the ANOVA were adjusted by the method of Geisser and Greenhouse.(L71 All agent comparisons were made with reference to control values using Dunnett's t-test and the appropriate error term from the ANOVA.
1015
BONE RESORPTION BY PDL CELLS
RESULTS Unstimulated PDL cells from four donors (A through D) produced PGE after 24 h of incubation that ranged from 980 to 2650 pg per lo4 cells (2.7 x 10” to 7.4 x M). Treatment of the PDL cells with IL-10 further increased the amount of PGE production, which was significantly diferent from that of unstimulated PDL cells @ < 0.01, Fig. 1). PGE production by IL-10-treated cells at 24 h was 5000-20,000 pg per lo4 cells (1.4 x lo-* to 5.6 x M). Unstimulated CM from four donors contained significant (p < 0.01) bone-resorbing activity after 24 h of incubation, and treatment of PDL cells with IL-10 significantly (p < 0.01) increased the level of bone-resorbing activity in the CM. The effect of conditioned medium on bone resorption was related to the amount of CM added to the calvarial
cultures. Conditioned medium produced by PDL cells from donor D was diluted 1 :2, 1 :4, and 1 :8 with BGJb medium and stimulated bone resorption @ < 0.01) as follows: 2.37 * 0.18, 1.51 f 0.04, and 1.28 + 0.08, respectively. Figure 2 shows the effects of IL-10 (1 ng/ml), IFN-y (100 U/ml), and their combination on PGE production and bone-resorbing activity produced by PDL cells. Treatment of PDL cells with IFN-y and IL-10 did not significantly change the amount of PGE or bone-resorbing activity production compared to PDL cells treated with It-10 alone (Fig. 2). Similar results were obtained from experiments performed using IL-10 instead of 1L-10 (data not shown) . The experiment shown in Fig. 3 demonstrates the effect of indomethacin on PGE production and bone-resorbing activity in CM from unstimulated and IL-lp-treated PDL
4‘0 3.5
I
1 T
..
0UNSTIMULATED a l L - 1 P TREATMENT
2
. I
T
v
3.0
+.
2.5
< 2.0
0 Y)
1.5 1 .o
DONOR
A
B
C
..
D
DONOR
A
e
~ ( 0 . 0 1SIGNIFICANT INCREASE IN BONE RESORPTION p(0.01 SIGNIFICANT INCREASE COMPARED TO UNSTIMULATED CELLS
‘p(0.01 DIFFERENT FROM UNSTIMUUTED CELLS
FIG. 1. Prostaglandin E and bone-resorbing activity produced by unstimulated and IL-lbstimulated ( 1 ng/ml) human periodontal ligament cells after 24 h of culture (mean + SEM).
15T
3.5
i
--e
UNSTlYUlATED
IL-1@
L
1
-t
IFN-7
UNSTIMULATED IL-1@
IL-16
+ IFN-7
p(O.01 DIFFERENT FROM U N S T I M U U T E D AND IFN-7-TREATED
CELLS
T
-
t-
IFN-1
IL- 1@
+
IFN-Y
~ ( 0 . 0 1DIFFERENT FROM UNSTIMUUTED C E U S
FIG. 2. Prostaglandin E and bone-resorbing activity produced by unstimulated, IL-I0 (1 nglml), IFN-7 (100 U/ ml), or IL-10 and IFN-y-treated human periodontal ligament cells (donor B).
SAITO ET AL.
1016
cells. Indomethacin inhibited PGE production from unstimulated PDL cells and completely abolished the increase in PGE production induced by IL-IP treatment of PDL cells at 24 h. However, treatment of PDL cells with indomethacin caused only a partial inhibition of bone-resorbing activity production at 24 h in 1L-lp-treated PDL cells (Fig. 3). Similar results were obtained from experiments performed using IL-la instead of IL-I@(data not shown). The interactive effects of cytokines and PTH on PGE and bone-resorbing activity production by PDL cells is shown in Fig. 4. Treatment of PDL cells with PTH and IL-la, IL-16, or TNF-a caused a synergistic increase in PGE production by PDL cells. PTH treatment increased the production of bone-resorbing activity by PDL cells. The three cytokines (IL-la, IL-10, and TNF-a) significantly increased bone-resorbing activity produced by PDL
cells at 24 h compared to unstimulated PDL cells. Combined treatment of PDL cells with PTH and the cytokines (IL-Ip and TNF-a) further increased the bone-resorbing activity of PDL CM.
DISCUSSION This investigation demonstrated that unstimulated human PDL cells synthesize nonprostaglandin bone-resorbing factors and PGE. In addition, the PDL cells responded to the administration of cytokines (IL-I@, IL-la, and TNF-a) with further increases in the synthesis of PGE and bone-resorbing activity compared to unstimulated PDL cells.
3.0
i
T
i
.. + 4 UNSTIMULATED
IL- 18
INDO
UNSTIMULATED
INDO
IL-lg
IL- 1 /3 INDO
'p(0.01 DIFFERENT FROM UNSTIUUUTEO CELLS
IL-lg
+
INDO *p(0.01 DIFFERENT FROM UNSTIMULATED AND INDO-TREATED CELLS p(0.05 DIFFERENT FROM INDO OR IL-l@-TREATED CELLS
0.
FIG. 3. Prostaglandin E and bone-resorbing activity produced by unstimulated, IL-IP (1 ng/ml), indomethacin (1.0 rM), or IL-lP and indomethacin (IND0)-treated human periodontal ligament cells (donor D).
A
T
..
*p(O.Ol DIFFERENT FROM UNSTIUUUTEO CELLS ~ ( 0 . 0 1DIFFERENT FROM CYTOKINE-TREATED CELLS
M
*p(0.01 and **p(0.05 DIFFERENT FROM UNSTIMUUTED hPDL CELLS ' ~ ( 0 . 0 5 and MP(O.Dl
DIFFERENT FROM CYTOKINE-TREATED CELLS
FIG. 4. Prostaglandin E and bone-resorbing activity produced by unstimulated, IL-la (3 ng/ml), IL-10 (1 ng/ml), TNF-a (1.0 nM), and PTH-treated (1.0 U/ml) human periodontal ligament cells (donor D).
1017
BONE RESORPTION BY PDL CELLS The addition of indomethacin completely inhibited PGE production and partially inhibited the production of boneresorbing activity by PDL cells. The increase of PGE in the PDL cells was due to de novo production since it was inhibited by the addition of indomethacin to the medium. Meghji et al.(18)reported that unstimulated CM derived from mouse osteoblastic cells caused bone resorption that was partially inhibited by indomethacin, dexamethasone, and nordihydroguaiaretic acid. In addition, Lerner and Hanstrom(I9) reported that human gingival fibroblasts produce nonprostaglandin bone-resorbing factors in vitro. Their findings and ours suggest that PGE is not the only factor present in PDL, osteoblast, or fibroblast C M that is responsible for stimulating bone resorption. In fact, the PGE concentration of unstimulated CM at 24 h of incubation was 2.7 x lo-' to 7.4 x 10'' M. Despite these low levels of PGE production, the stimulation of bone resorption (1.8- to 2.7-fold increase in " T a release) caused by these CM were almost the same or even higher than those caused by addition of PGE, M), which was used as a positive control (data not shown). Other factors, such as cytokines, growth factors, and leukotrienes that are synthesized by PDL cells, may contribute to the enhancement of bone resorption by interacting with PGE. Mohammed et reported on the interaction between prostaglandins and leukotrienes on orthodontic tooth movement. They found a significant inhibition of tooth movement in rats injected with indomethacin. However, since a previous has shown that inhibition of the cyclooxygenase pathway potentiates the conversion of arachidonic acid to leukotrienes, it is conceivable that the addition of indomethacin may have potentiated the synthesis of leukotrienes. The administration of IFN-7 did not change unstimulated or IL-1-stimulated PGE production or bone-resorbing activity from PDL cells. These results correspond with the finding by Friteau et al.,'zz) since they reported that IFN-y did not suppress IL- I@-stimulated PGE production from human fibroblasts. However, IFN-y did suppress ILI@-stimulated PCiE production in human monocytesiZz) and endogenous prostaglandin synthesis in mouse calvariae,lz3)suggesting that the effect of IFN-7 on PGE production is different depending on the particular cell type. IFN-7 has also been shown to have variable effects on PGE production in human osteoblasts."") PTH is a potent stimulator of bone resorption that does not depend on PGE production for its a ~ t i v i t y . ~How~~] ever, a recent study'z6' has shown that the increase in inosito1 triphosphate by P T H is abolished by the addition of indomethacin, suggesting that PGE mediates the effects of P T H on inositol phosphate production. In studies with bone explants and isolated bone cells, it was found that CAMP alone was not the exclusive second messenger for the action of PTH on bone resorption and that calcium plays an important role in this p r o c e s ~ . In ( ~ our ~ ~ study a slight increase in PGE production from PDL cells was observed following the administration of PTH. It is possible that PGE increases in response to PTH administration are transient and that the level of PGE synthesized and secreted into the medium is relatively low. I n contrast, the
addition of P T H synergistically stimulated the cytokine-induced PGE production. This result corresponds with that of Tatakis et al.,'z8)who found synergistic PGE, elevation between IL-la and PTH by mouse osteoblastic cells. It is possible that P T H can influence the stability of surface membrane receptors, as a calcium ionophore, and make them more available to the binding of cytokine molecules. Therefore, an additive effect on bone-resorbing activity caused by P T H and cytokines may be due to synergistic PGE synthesis. However, P T H increased the production of nonprostaglandin bone-resorbing factors from unstimulated PDL cells since the increase in PGE production in PTH-treated cells was not great enough to account for the increase in bone-resorbing activity. In conclusion, human PDL cells produced bone-resorbing activity and PGE in vitro that was increased after treatment with cytokines (IL-la, IL-10, and TNF-a). Indomethacin completely inhibited PGE production and partially inhibited bone-resorbing activity. IFN-y did not alter PGE production or bone-resorbing activity produced by unstimulated PDL cells. This indicated that PDL cells produced bone-resorbing activity that was due to both nonprostaglandin factors and PGE. The nonprostaglandin factors and their mechanisms of induced bone resorption may differ depending on the time period of collection of conditioned medium from the PDL cells. Future investigations are necessary to identify the nonprostaglandin factors and the time course of their secretion from human PDL cells.
ACKNOWLEDGMENTS This work was supported by the National Institutes of Health research grant DE-08428.
REFERENCES 1. Rodan GA, Martin T J 1981 Role of osteoblasts in hormonal
2.
3.
4.
5.
6.
7.
control of bone resorption. A hypothesis. Calcif Tissue Int 33~349-351. Rouleau MF, Mitchell J, Goltzman D 1988 In vivo distribution of parathyroid hormone receptors in bone: Evidence that a predominant osseous target cell is not the mature osteoblast. Endocrinology 123:187-191. Roberts WE, Chase DC 1981 Kinetics of cell proliferation and migration associated with orthodontically-induced osteogenesis. J Dent Res 60:174-181. Aubauch GD, Marx SJ, Speigel AM 1981 Parathyroid hormone, calcitonin, and calciferols. In: Williams R H (ed.) Textbook of Endocrinology, 6th ed. W.B.Saunders, Philadelphia, pp. 932-1031. Gowen M, Wood DD, lhrie EG, McGuire MKB, Russell RGG 1983 An interleukin I like factor stimulates bone resorption in vitro. Nature 306:378-380. Dewhirst FE, Stashenko P , Mole JE, Tsurumachi T 1985 Purification and partial sequence of human osteoclast-activating factor: Identity with interleukin I-beta. J lmmunol 135~2562-2568. Bertolini DR, Nedwin GE, Bringmam TS, Smith DD, Mundy GR 1986 Stimulation of bone resorption and inhibition of
1018
SAITO ET AL.
bone formation in vitro by human tumor necrosis factors. stimulate bone resorption in vitro. J Periodont Res 22:284Nature 3195 16-5 18. 289. 8. Raisz LG, Simmons HA, Sandberg AL, Canalis E 1980 20. Mohammed AH, Tatakis DN, Dziak R 1989 Leukotrienes in Direct stimulation of bone resorption by epidermal growth orthodontic tooth movement. Am J Orthodont Dentofac factor. Endocrinology 107:270-273. Orthop 99231-237. 9. Tashjian AH Jr, Voekel EF, Lazzaro M, Singer FR, Roberts 21. Raisz LG, Martin TJ 1983 Prostaglandins in bone and minAB, Derynck R, Winkler ME, Levine L 1985 Alpha and beta eral metabolism. In: William AP (ed.) Bone and Mineral Rehuman transforming growth factors stimulate prostaglandin search Annual 2. Elsevier, Amsterdam, pp. 286-310. production and bone resorption in cultured mouse calvaria. 22. Friteau L, Francesconi E, Land D, Dugas B, Damais C 1988 Proc Natl Acad Sci USA 82:4535-4538. Opposite effect of interferon-gamma on PGE, release from 10. Dietrich JW, Goodson JM, Raisz LG 1975 Stimulation of interleukin-1-stimulated human monocytes or fibroblasts. bone resorption by various prostaglandins in organ culture. Biochem Biophys Res Commun 157:1197-1204. Prostaglandins 10:231-238. 23. Peterlik M, Hoffman 0, Swetly P, Klaushofer K, Koller K 11 Davidovitch Z, Nicolay 0, Ngan PW, Shanfeld JL 1988 1985 Recombinant y-interferon inhibits prostaglandin-mediNeurotransmitters, cytokines and the control of alveolar ated and parathyroid hormone induced bone resorption in bone remodeling in orthodontics. Dent Clin North Am 32: cultured neonatal mouse calvaria. FEBS Lett 185:287-290. 41 1-435. 24. Gowen M, MacDonald BR, Russell RGG 1988 Actions of re12 Ngan PW, Zadeh YZ, Shanfeld J, Davidovitch 2 1988 The combinant human y-interferon and tumor necrosis factor CY effect of interleukin-lp and parathyroid hormone on cyclic on the proliferation and osteoblastic characteristics of hunucleotide and prostaglandin levels in human periodontal man trabecular bone cells in vitro. Arthritis Rheum 31:1500ligament fibroblasts in vitro. In: Davidovitch Z (ed.) The 1507. Biological Mechanisms of Tooth Eruption and Root Resorp- 25. Dewhirst FE, Ago JM, Peros WJ, Stashenko P 1987 Synertion. EBSCO Media, Birmingham, AL, pp. 261-267. gism between parathyroid hormone and interleukin I in stim13. Ragnarsson B, Carr G, Daniel J C 1985 Isolation and growth ulating bone resorption in organ culture. J Bone Min Res 2: of human periodontal ligament cells in vitro. J Dent Res 64: 127-1 34. 1026-1030. 26. Sandy JR, Meghji S, Farndale RM, Meikle MC 1989 Dual 14. Owens RB 1974 Glandular epithelial cells from mice: A elevation of cyclic AMP and inositol phosphates in response method for selective cultivation. JNCI 52:1375-1378. to mechanical deformation of murine osteoblasts. Biochim 15. Ngan PW, Crock B, Varghese J, Lanese R, Shanfeld J, Biophys Acta 1010:265-269. Davidovitch Z 1988 Immunohistochemical assessment of the 27. Herrmann-Erlee MPM, Van der Meer JM, Lowik CWGM, effect of chemical and mechanical stimuli on CAMP and Van Leeuwen JPTM, Boonekamp PM 1988 Different roles prostaglandin E levels in human gingival fibroblasts in vitro. for calcium and cyclic AMP in the action of PTH: Studies in Arch Oral Biol 33:163-174. bone explants and isolated bone cells. Bone 993-100. 16. Rosol TJ, Capen CC 1988 Inhibition of in vitro bone resorp- 28. Tatakis DN, Schneeberger G, Dziak R 1988 Recombinant intion by a parathyroid hormone receptor antagonist in the terleukin 1 stimulates prostaglandin E, production by osteocanine adenocarcinoma model of humoral hypercalcemia of blastic cells: Synergy with parathyroid hormone. Calcif Tismalignancy. Endocrinology 122:2098-2102. sue lnt 42:358-362. 17. Geisser S, Greenhouse SW 1958 An extension of box’s results on the use of the F distribution in multivariate analysis. Ann Address reprint requests to: Math Stat 29:885-891. Dr. Thomas J . Rosol 18. Meghji S, Sandy JR, Scutt AM, Harvey W, Harris M 1988 Department of Veterinary Pathobiology Heterogeneity of bone resorbing factors produced by unstimThe Ohio State University ulated murine osteoblasts in vitro and in response to stimula1925 Coffey Road, Columbus, OH 43210 tion by parathyroid hormone and mononuclear cell factors. Archs Oral Biol 33:773-778. 19. Lerner U, Hanstrom L 1987 Human gingival fibroblasts se- Received for publication August 31, 1989; in revised form May crete nondialyzable, prostanoid-independent products which 17, 1990; accepted May 18, 1990.
