JOURNAL OF BONE A N D MINERAL RESEARCH Volume 7, Number 4, 1992 Mary Ann Liebert, Inc., Publishers
Effects of Phorbol Myristate Acetate on Rat and Chick Osteoclasts RICHARD J . MURRILLS,’ LINDA S. STEIN,’ WENDY R. HORBERT,’ and DAVID W. DEMPSTER’.’
ABSTRACT The role of protein kinase C in the regulation of osteoclast function is not known. We therefore compared the effect of phorbol myristate acetate (PMA), which activates protein kinase C, on the resorptive function, motility, and morphology of osteoclasts from rat and chick. PMA caused a significant reduction in resorption pit number in both species; rat osteoclasts were more sensitive, being significantly inhibited at doses of 10-9-10-6M compared with 10-7-10-6M for chick osteoclasts. The inactive analog PMA-(r was without significant effect, and inhibition was not blocked by M indomethacin. In time course experiments, inhibition at 24 h was similar to or greater than inhibition at 6 h, indicating a persistent or progressive effect on bone resorption. Removal of PMA after 6 h prompted partial recovery of bone-resorptive ability in chick M osteoclasts but not rat, at least over a 48 h incubation. In time-lapse video studies of rat osteoclasts, PMA produced an immediate but transient cessation of motility and retraction of the cell margin into prominent filopodia. Motility resumed within 2.5 h after addition, but the osteoclasts remained partially contracted. Chick osteoclasts behaved similarly but showed no formation of filopodia at the cell periphery and a more rapid recovery of motility than rat osteoclasts; chick osteoclasts also underwent a transient vacuolation following PMA exposure, whereas rat osteoclasts did not. Despite differences in the sensitivity of rat and chick osteoclasts to PMA, these results suggest a fundamental role for protein kinase C in the inhibition of osteoclasts from both species.
INTRODUCTION H E R E IS NOW strong evidence that both mammalian and and avian osteoclasts possess an inhibitory second messenger pathway mediated by cyclic AMP. However, in addition to cyclic AMP, most cells also possess intracellular signal transduction pathways dependent on the hydrolysis of membrane phospholipids.(’’ In one such pathway, binding of a hormone or growth factor to the cell surface causes breakdown of membrane phospholipids generating diacylglycerol (DAG), which in turn binds to and activates protein kinase C (PKC). This apparently ubiquitous enzyme then phosphorylates proteins involved in cell differentiation, division, or function, thus effecting a change in the cell’s behavior. PKC exists in multiple
T
forms that vary in their tissue distribution, substrate specificity, and regulation.(8) Its role in osteoclasts is unknown. Phorbol esters are structurally similar to the DAG molecule and activate protein kinase C without requiring binding of a hormone or growth factor to the cell surface. To investigate the role of protein kinase C in osteoclasts, we tested the effects of phorbol 12-myristate 13-acetate (PMA) on bone resorption and the motility and morphology of isolated rat and chick osteoclasts.
MATERIALS AND METHODS
Phorbol myristate acetates and indomethacin Phorbol 12-myristate 13-acetate was obtained from Sigma Chemical Co. (St. Louis, MO) and 4-a-phorbol 12-
‘Regional Bone Center, Helen Hayes Hospital, West Haverstraw, New York. zDepartment of Pathology, Columbia University, New York, New York.
415
416
MURRILLS ET AL.
myristate 13-acetate (PMA-a) from Biomol Research Laboratories, Inc. (Plymouth Meeting, PA). Each was dissolved in dimethylsulfoxide (DMSO) at 20 mM and stored -80°C. PMA and PMA-a were then added to 199+Earle’s (see later) and used at 10~’o-10~6 M. Indomethacin was dissolved in absolute ethanol at lo-* M, M. Appropriate stored at -8O”C, and used at amounts of vehicles were added to ensure that the concentration of vehicles in culture media was the same in all treatments within an experiment.
Rat Osteoclasts
Bone resorption was assayed using the disaggregated osteoclast resorption assay originated by Boyde et al.C9)and Chambers et aI.l1O1and performed as described by Dempster et al.C”lBriefly, osteoclasts were recovered by mincing the bones of neonatal rats or embryonic chicks in 3 ml medium 199 containing Hank’s salts (199+ Hank’s) and settling 25-100 pl of the resultant cell suspension for 30 minutes (rat) or 90 minutes (chick) onto devitalized 4.4 x 4.4 mm bovine cortical bone slices prewetted with 100 p1 of 199+ Hank’s. The slices were then rinsed to remove nonadherent cells and transferred to tissue culture medium (medium 199 containing Earle’s salts, 0.7 g/liter of sodium bicarbonate, and 10% heat-inactivated fetal bovine serum, FBS) containing vehicle alone or specified concentrations of phorbol esters and/or indomethacin. After a 24 h incubation in a 5 % CO, humidified incubator, the slices were fixed in 2% glutaraldehyde and stained with either toluidine blue (TB) or for tartrate-resistant acid phosphatase (TRAP) using Sigma kit 386-A (Sigma Chemical Company, St. Louis, MO). The number of multinucleate osteoclasts ( 2 three nuclei for TB-stained slices and 2 two nuclei for TRAP-stained slices) were counted on one to four replicate slices per treatment per animal. Cells were stripped from the slices by sonication in 0.25 M NH,OH (two 15 s bursts), rinsed in distilled water, and air dried from acetone. The number of resorption pits, which we have shown to be significantly correlated to the total plan area of bone resorbed,I6)was counted in the light microscope on three to four replicates per treatment per animal after staining in toluidine blue for a total of 4 minutes. The experimental protocols were approved by the Institutional Animal Care and Use Committee (IACUC) of Helen Hayes Hospital.
Statistical analysis Slices were allocated to treatment groups using a randomized block design.‘”’ Osteoclasts recovered from an individual rat pup or chick embryo were settled onto four replicate slices per treatment, and typically five animals, each constituting one “block,” were used per experiment. The data were subjected to two-way analysis of variance (ANOVA) to determine the standard error of the difference between treatment means and differences from control tested for significance using the method of least significant difference. When variances between treatments were unacceptably different, data were transformed by square root or logarithmic transformation. A multiple-range test
**
20
Bone resorption assay
T
**
**
T
T
10
Chick Osteoclasts
Y
I
I
FIG. 1. Effect of PMA on rat and chick osteoclasts. Values are means + SEM, n = 5 ; **p < 0.001; * p < 0.05.
was employed for multiple comparisons. In two instances, indicated in the legend to Fig. 4 and in Table 2, unacceptably different variances could not be rectified by data transformation, thus preventing analysis of the experiment as a whole by a multiple-range test; when this occurred, a limited number of selected treatments were compared using a t-test.
RESULTS
Bone resorption and osteoclast counts PMA caused a significant dose-responsive reduction in the number of pits excavated by both neonatal rat and em-
417
EFFECT OF PHORBOL ESTER ON OSTEOCLASTS
Rat Osteoclasts
I% v) u
30
T
iz
L
0 L
0,
20
n
E, 10
0 Control
1WRM PMA
1O-RM 1W6 M P M A alpha
Chick Osteoclasts
T
I
& A
*
Cell morphology and motility
T
Control
10.’ M PMA
deed, P M A ( M) significantly increased osteoclast numbers over 24 h in five of nine chick experiments. In these experiments, the increase in osteoclast number tended to offset the inhibitory effects of P M A on the number of pits per slice, but significant inhibition was clearly present if resorption was expressed as pits per osteoclast (Table I). Significant osteoclast formation was seen more consistently (three of four experiments) with dense preparations of chick embryo bone cells (100 pI plating volume). At the lower cell densities (25 pl plating volume), significant osteoclast formation was seen in two of five experiments. Time course studies indicated that between 6 and 24 h there was no statistically significant change in the degree of inhibition, as determined by test to control ratios, in either the rat or the chick. However, inhibition tended to be greater at 24 h than at 6 h (Table 2), and significant inhibition was noted in the rat after 48 h exposure to a low dose but none was demonstrable at 6 h (Fig. 4). In recovery experiments in which slices exposed to PMA were transferred to control media for subsequent incubation, the chick osteoclasts partially recovered from the effects of P M A on bone resorption, whereas the rat osteoclasts did not, at least over 48 h at the doses tested (Fig. 4 and Table 3). The effect on rat osteoclasts was “inducible”: exposure for only the first 6 h of the experiment resulted in almost exactly the same degree of inhibition after 48 h as osteoclasts exposed continuously. I t should be noted that no loss of osteoclasts was observed in either species in the incubation periods examined (up to 48 h).
10.‘ M 10.7~ P M A alpha
On rat and chick F1c* 2’ Specificity Of effect Of osteoclasts. Values are means f SEM, n = 5 (rat) or 10 (chick); * p < 0.05.
bryonic chick osteoclasts over 24 h. Rat osteoclasts were more sensitive, significant inhibition being noted at doses of 10-9-10-6M compared with lo-’ and lo+’ M in the chick (Fig. 1 and Table 1). Doses higher than M were not tested because of the unacceptably high concentration of vehicle necessary. Inhibition was specific, the inactive analog of PMA, PMA-a, having no significant effect (Fig. 2). Inhibition by P M A was not blocked by M indomethacin, suggesting that inhibition was not due to prostaglandin production (Fig. 3). Reduction in pit number was not due to loss of osteoclasts in either the rat or the chick experiments, suggesting that inhibition was not due to cytotoxicity (Table 1). In-
Neonatal rat osteoclasts underwent a rapid change in morphology and motility in response to lo-‘ M PMA, which was most marked soon after addition (Fig. 5). Before addition, osteoclasts constantly extended and then retracted phase-dense lamellipodia from apparently random locations at the cell periphery (lamellipodial activity). Within 5 minutes of adding PMA, lamellipodial activity ceased and the cell periphery withdrew, forming prominent filopodia, such that in some osteoclasts the cell periphery consisted almost entirely of filopodia after as little as 10 minutes. Lamellipodial activity returned at least partially within 2.5 h after addition, the timing and extent varying from osteoclast to osteoclast. The inactive analog of PMA, P M A - a , had no apparent effect on the motility or morM. phology of rat osteoclasts at The motility and morphology of chick osteoclasts was M PMA, the effects being most also influenced by marked soon after addition (Fig. 6). Like rat osteoclasts, lamellipodial activity was immediately eliminated or greatly reduced and the cell periphery appeared to withdraw, the strength of the response varying from osteoclast to osteoclast. Unlike rat osteoclasts, prominent filopodia were not noted and lamellipodial activity returned after a shorter period than in rat osteoclasts. Time course studies of the morphology of cells on bone slices revealed a tranM sient vacuolation of chick osteoclasts exposed to PMA, vacuoles being visible (at x 160 magnification) in 42% of the TRAP-positive multinuclear cells after 6 h in-
MURRILLS ET AL.
418
Rat Osteoclasts
Rat Osteoclasts
I
I ‘ l i m e (hrs.)
4.f’l 4.5,
C h ic k
( )st cod a s t s
3.5
Control
PMA
Indo
PMA+Indo
Chick Osteoclasts
’l’imc. ( h n . )
FIG. 4. Recovery of chick, but not rat osteoclasts following 6 h exposure to PMA. Doses used are M (chick) SEM for four to five and 10+’ M (rat). Values are means animals, two replicates per treatment per animal. Significant differences ( P < 0.05) were detected using a multiplerange test. Because of unacceptably different variances, the 48 h group was analyzed by a multiple-range test separately from the 6 h group (which was analyzed using a 1test). Hence no comparison between these groups is shown. asignificant difference from control 6 h; bSignificant difference from PMA 6 h; CSignificant difference from control 48 h; dSignificant difference from PMA removed at 6 h; esignificant difference from PMA 48 h.
*
I
W L
-
B. Y)
h
10
L
0 L W
n
z$
5
0
(’ontrol
mit\
Indo
PhlA+lndo
FIG. 3. Effect of indomethacin M) on the response of rat and chick osteoclasts to PMA. PMA was used at lo-* M (rat) and M (chick). Values are means SEM, n = 5 (rat) or 10 (chick); * * p < 0.001; *p < 0.05.
*
vacuoles. Vacuole formation was never observed in rat osteoclasts in association with PMA exposure. The M PMA-(Y had no apparent effect on the motility or morphology of chick osteoclasts.
DISCUSSION cubation in PMA (compared with none in controls), the proportion falling to 13% after 24 h (compared with 6% in controls; Table 4). Importantly, it should be noted that no significant reduction in osteoclast numbers was noted at either of these times (Table 4) and at 24 h the total number of chick osteoclasts was increased in the presence of PMA. Video studies confirmed the PMA-induced vacuolation (Fig. 61, which was sometimes extensive, and also revealed the ability of the osteoclasts to rid themselves of vacuoles by either a rapid “deflation,” suggestive of a possible expulsion from the cell, or a more gradual shrinkage of the
These results show that both neonatal rat and embryonic chick osteoclasts were inhibited by the phorbol ester, PMA, a known activator of protein kinase C. The effect was specific, not occurring with the inactive analog PMA(Y and was not due to loss of osteoclasts, even at the highest doses tested. PMA also caused characteristic and reproducible changes in osteoclast motility and morphology. This strongly implicates protein kinase C in the inhibition of osteoclasts and has important ramifications for understanding the signal transduction mechanisms of antiresorptive agents. The finding of a common inhibitory effect of PMA in both neonatal rat and embryonic chick osteoclasts
419
EFFECT OF PHORBOL ESTER ON OSTEOCLASTS TABLE1. EFFECTOF PMA Treatment
ON
BONERESORPTION BY RATOR CHICKOSTEOCLASTS~
Pits per slice
Rat osteoclasts Control M PMA 10-9 M PMA M PMA Chick osteoclasts Control M PMA 10-7 M PMA M PMA
15.9 14.7 10.2 7.4
f f f f
3.4 2.3 1.9b 1.4~
52.2 f 17.8 35.2 k 7.7 43.2 f 9.8 30.9 f 7.7
Osteoclasts per slice
Pits per osteoclast
20.6 21.0 22.6 18.4
f f f f
4.5 1.1 3.0 1.7
0.81 f 0.15 0.73 f 0.10 0.47 f 0.09b 0.30 k 0.03d
22.9 25.4 28.7 35.6
f f f f
5.6 7.8 3.7 4.1b
2.31 k 0.63 1.62 f 0.41 1.47 k 0.29b 0.85 f 0.17d
aMultinucleate osteoclasts were counted and pits per osteoclast calculated on four (chick) or one (rat) replicate slices following tartrate-resistant acid phosphatase staining. Plating volume was 100 pl. Values are means f SEM, n = 5 animals (four replicate bone slices per treatment per animal); significant difference from control is indicated. hp < 0.05. Lp < 0.01. dp < 0.02.
TABLE 2 . TIMECOURSEOF INHIBITION OF RATAND CHICKOSTEOCLASTS BY
M PMAa T / C Ratio
Treatment
Pits per sfiee
Rat osteoclasts Control 6 h PMA 6 h Control 24 h PMA 24 h Chick osteoclasts Control 6 h PMA 6 h Control 24 h PMA 24 h
OCs per slice
12.9 7.0 27.2 10.0
f f f f
2.9 1.4b 4.7c-e 2.6b
24.4 f 7.3 27.0 k 2.9 29.2 k 4.4 27.2 f 5 . 1
4.9 4.0 46.0 33.5
f f f f
1.6b3e 1.3b.e 14.7C3d 7.6Cvd
19.8 43.5 39.7 76.4
f f f f
3.8b,",e 5.lc3e 5.7cve 7.2b-d
Pits per OC 0.34 f 0.12 0.08 f 0.02NS 0.67 f 0.05 0.25 k 0.09f
0.15 -t 0.05b.e 0.08 f 0.02b.e 1.09 k 0.16c-e 0.44 f 0.09b-d
Pits per sfiee
Piis/OC
0.21
0.56 f 0.12
0.48
0.36 f 0.05
0.39 f 0.14
0.90 f 0.17
0.69 f 0.29
0.83 f 0.15
0.41 f 0.06
k
aosteoclasts (OCs) were counted following TRAP staining; chick osteoclast counts include vacuolated cells recognizable as osteoclasts. SEM for five animals, one to four replicate slices per treatment per animal. Significant differences (p < 0.05) were Values are means detected using a multiple-range test, except for the Pits per OC column in the rat where, for reasons detailed in Materials and Methods, I-tests were used to compare PMA-treated groups to their respective controls (NS = not significant). hsignificant difference from control 24 h. CSignificant difference from control 6 h. dsignificant difference from PMA 6 h. eSignificant difference from PMA 24 h. (p < 0.005.
*
echoes our finding of a common inhibitory effect of stimulators of intracellular cAMPl6) and points to fundamental similarities of osteoclasts from these two sources, despite differences in the numbers of their calcitonin receptor^."^) In organ culture, phorbol esters have been found to have both stimulatory and inhibitory effects. In neonatal mouse calvariae, basal resorption was stimulated by phorbol esters by a process that was dependent on prostaglandin ~ y n t h e s i s . ' ~ Parathyroid ~,~~) hormone (PTH)-stimulated resorption, however, was inhibited by phorbol ester.i16)In fetal rat long bone organ culture, phorbol esters stimulated resorption'L'.18)independently of prosta-
glandin synthesis but required DNA synthesis. (I8) Our results show that the primary effect of PMA on mature osteoclasts was a powerful inhibition, in both the rat and the chick, which was independent of prostaglandin synthesis. In the chick osteoclast assay there was also evidence of significant increases in osteoclast numbers with PMA. This dual effect of inhibiting the action of mature osteoclasts while stimulating the formation of new osteoclasts is reminiscent of the actions of prostaglandins(l91 and agents that stimulate cyclic AMP(6) and may explain the ability of phorbol esters to have both stimulatory and inhibitory effects in organ culture.
420
MURRILLS ET AL.
FIG. 5 . Time-lapse micrographs of the response of rat osteoclasts to M PMA (A) immediately before the addition of PMA, (B) 100 minutes after, and ( C )200 minutes after addition.
FIG. 6. Time-lapse micrographs of the response of chick osteoclasts to M PMA (A) immediately before the addition of PMA; (B) 5 minutes after the addition of PMA; and (C) 60 minutes after addition. Note the beginnings of vacuolation in C.
Activation of protein kinase C caused changes in the morphology and motility of both rat and chick osteoclasts. These have some similarity with the changes occurring after exposure of sensitive osteoclasts to calcitonin‘ 2 0 , 2 1 ) or prostaglandins‘1.2z’and raise the possibility that protein kinase C may play a role in the transduction of these changes in morphology and motility. I t should be noted, however, that for logistical reasons the dose of PMA in the present study was insufficient to eliminate bone resorption. This should be considered when comparing the transient and incomplete effects of PMA on motility and mor-
phology with the full-blown contraction and adoption of the “stellate” morphology achieved by doses of calcitonin (CT) or stimulators of CAMP that completely inhibit resorption. (6.L1.10.21)In addition to the changes in motility and morphology usually associated with inhibitory agents, chick osteoclasts also underwent a transient vacuolation. The significance of this vacuolation is not known: vacuolation has been noted in hepatoma cells exposed to the PKC agonist t e l e o ~ i d i n , ‘ ~ ~ ) and PMA was also shown to stimulate the secretion of insulin in pancreatic islet cells,124’ raising the possibility that PKC plays a role in exocytosis or endocytosis in chick osteoclasts.
42 1
EFFECT OF PHORBOL ESTER ON OSTEOCLASTS N i s h i z ~ k a 'listed ~ ~ ) a large number of substrates reported to be targets for phosphorylation by PKC; of possible relevance to the morphologic and functional changes r e p o r t e d here a r e vinculin, ( N a + , H + ) - A T P a s e , (Na+,K+)-ATPase,and the pp60src protein, each of which is thought to play a prominent role in osteoclast activity.
TABLE 3. PIT AND OSTEOCLAST COUNTS FOR EXPERIMENTS IN FIG. 4a
Treatment
Pits per slice
Rat osteoclasts Control 6 h PMA 6 h Control 48 h PMA 0-6 h then control 6-48 h PMA 48 h Chick osteoclasts Control 6 h PMA 6 h Control 48 h PMA 0-6 h then control 6-48 h PMA 48 h
THE
OCs per slice
16.0 f 4.0b 14.6 f l . l b 54.5 f 8.lC-f
31.6 f 4.4 35.1 f 5.6 32.6 f 4.0
24.9 T 3 . P 23.8 f 4.3h
35.6 k 3.9 33.9 f 5.8 33.2 f 10.3e.f 53.7 f 14.8f 52.3 f 15.3f 83.1 f 2 3 . 4 ~ 93.2 + 18.Ob-d
M and aNote that the dose of PMA used for the chick was for the rat M; 100 pl plating volume; osteoclasts counted following TRAP staining. Values are means f SEM for four to five animals, two replicates per animal. Significant differences (p < 0.05) determined by a multiple-range test. bsignificant difference from control 48 h. CSignificant difference from control 6 h. dSignificant difference from PMA 6 h. eSignificant difference from PMA 0-6 h then control 6-48 h. ?Significant difference from PMA 48 h.
Classically, exposure to PMA initially activates and then downregulates PKC as a result of a n increased rate of degradation of the enzyme,(26)the timing and extent of downregulation being dependent on the cell type. (27) Consistent with this, transient effects were seen on the motility and morphology of both rat and chick osteoclasts, the initial dramatic changes being almost completely reversed within a few hours. However, studies on bone resorption indicated that the resorptive activity of both chick and rat osteoclasts was inhibited to a constant or increasing degree. These results suggest that although some downregulation of PKC occurred, this was not sufficient for reinstatement of bone resorption. It should be borne in mind that changes in morphology and motility may reflect only a part of the multifactorial process of bone resorption and that the restoration of several other systems (which may have different sensitivities or kinetics) may be necessary for bone resorption to recommence. Bone resorption may therefore be more sensitive to inhibitory agents than changes in morphology or motility, an observation we have already made with respect to C T and stimulators of intracellular cyclic AMP.(6." **' Inhibition of osteoclast activity could theoretically be due to a cytotoxic effect on the osteoclast. We tested this possibility using cell counts conducted at the end of the incubation period and "recovery" experiments in which slices were removed from PMA after a brief exposure and transferred to control media. Chick osteoclasts showed an increase in osteoclast number over time and were clearly capable of recovering from an inhibitory dose of PMA; rat osteoclasts, however, did not recover from the effects of PMA, a short (6 h) exposure being sufficient to elicit all the inhibition seen at 48 h with continuous exposure. A cytotoxic effect on the rat osteoclast cannot therefore be ruled out, but we think this unlikely for the following reasons. First, no loss of osteoclasts was noted in cell counts
TABLE4. EFFECTOF PMA M) ON VACUOLATION OF CHICK OSTEOCLASTS (100 p~ PLATINGVOLUME)^
Treatment
Non vacuolated OCs per slice
Control 6 h PMA 6 h Control 24 h PMA 24 h
19.7 3.7 24.7 f 2.5 37.6 k 5.9bSd3f 66.2 ? 5.4b,c,f
*
Vacuolated OCs per slice 0.1 f O.lb-d 18.8 f 3.4CJ 2.1 f 0.7b,d,f 10.2 1.8C,f
Total OCs per slice 19.8 43.5 39.7 76.4
f f f f
3.8b-d 5.ld3f 5.74f 7.2b.C,f
Ratio of vacuolated to total OCs per slice
0 f oe 0.42 0.04C.d 0.06 f 0.02b 0.13 f 0.01b
=TRAP-stained osteoclasts were deemed vacuolated if at least one cytoplasmic vacuole, however small, was visible at a magnification of x 160. Greatly distended TRAP-positive vacuolated cells in which nuclei could not be seen were excluded. Values are means k SEM for five chicks, two to four replicate slices per treatment per animal. Significance (p < 0.05) was determined using a multiple-range test. bSignificant difference from PMA 6 h. CSignificant difference from control 24 h. dSignificant difference from PMA 24 h. eExcluded from statistical analysis. fSignificant difference from control 6 h.
MURRILLS ET AL.
422
conducted at the end of our experiments, even when incubation was extended to 48 h. Second, organ culture studies using higher doses than those used here showed a longterm stimulation of resorption (presumably due to osteoclast formation), an observation inconsistent with cytotoxicity. Third, video studies showed recovery of motility in mature osteoclasts. Fourth, the inactive analog, PMA-a, showed no inhibitory effects. On balance, we therefore interpret the rat osteoclast’s relative inability to recover from an inhibitory dose of PMA as being due either to its extreme sensitivity or to the long-term nature of PMA-induced activation of PKC,‘”’ not to a cytotoxic effect. In conclusion, we have shown that phorbol myristate acetate had potent inhibitory effects on bone resorption by both neonatal rat and embryonic chick osteoclasts. Exposure of osteoclasts to PMA caused a significant reduction in their ability to resorb bone and was associated with marked but transient changes in cell morphology. These observations suggest that protein kinase C plays a fundamental role in the transduction of inhibitory signals in both mammalian and avian osteoclasts and raise the possibility, for which there is already some preliminary evidence,(**’that protein kinase C acts, either directly or indirectly, on the proteins of the osteoclast cytoskeleton.
ACKNOWLEDGMENTS Parts of this study were presented at the Washington University Continuing Edueat ion Symposium, Molecular Basis of Bone Cell Physiology: Transcellular Signalling, St. Louis, MO, April 1991, and at the Thirteenth Annual Meeting of the ASBMR, San Diego, CA, August 1991. This work was supported in part by NIH Grant AR39191. We thank Ms. S. Roberts, Mr. B. Yarborough, and Mr. L . Ferguson for their expertise in graphics and photography.
REFERENCES I . Chambers T J , Dunn CJ 1982 The effect of parathyroid hormone, 1,25-dihydroxycholecalciferoland prostaglandins on the cytoplasmic activity of isolated osteoclasts. J Pathol 137: 193-203. 2. Nicholson GC, Livesey SA, Moseley JM, Martin TJ 1986 Actions of calcitonin, parathyroid hormone, and prostaglandin E1 on cyclic AMP formation in chicken and rat osteoclasts. J Cell Biochem 31:229-241. 3. Rifkin BR, Auszmann J M , Kleckner AP, Vernillo AT, Fine AS 1988 Calcitonin stimulates CAMP accumulation in chicken osteoclasts. Life Sci 42:799-804. 4. Eliam MC, Basle M, Bouizar Z , Bielakoff J , Moukhtar M, de Vernejoul MC 1988 Influence of blood calcium on calcitonin receptors in isolated chick osteoclasts. J Endocrinol 119:243-248. 5. Ransjo M, Lerner UH. Heersche JNM 1988 Calcitonin-like effects of forskolin and choleratoxin o n surface area and motility of isolated rabbit osteoclasts. J Bone Miner Res 3:611619. 6. Murrills RJ, Dempster DW 1990 The effects of stimulators of intracellular cyclic AMP on rat and chick osteoclasts in vitro: Validation of a simplified light microscope assay of bone resorption. Bone 11:333-344.
7. Nishizuka Y 1984 The role of protein kinase C in cell surface signal transduction and tumour promotion. Nature 308:693698. 8. Exton J H 19% Signalling through phosphatidylcholine breakdown. J Biol Chem 265: 1-4. 9. Boyde A, Ali NN, Jones SJ 1984 Resorption of dentine by isolated osteoclasts in vitro. Br Dent J 156:210-220. 10. Chambers T J , McSheehy PMJ, Thomsom BM, Fuller K 1985 The effect of calcium-regulating hormones and prostaglandins on bone resorption by osteoclasts disaggregated from neonatal rabbit hones. Endocrinology 116:234-239. 11. Dempster DW. Murrills RJ, Horhert WR, Arnett TR 1987 Biological activity of chicken calcitonin: Effects on neonatal rat and embryonic chick osteoclasts. J Bone Miner Res 2: 443-448. 12. Parker RE 1979 Introductory Statistics for Biology, 2nd ed. Studies in Biology, No. 43. Institute of Biology, London. 13. Nicholson GC, Moseley JM. Sexton PM, Martin TJ 1987 Chicken osteoclasts d o not possess calcitonin receptors. J Bone Miner Res 2:53-59. 14. Tashjian A H Jr, lvey JL, Delclos B, Levine L 1978 Stimulation of prostaglandin production in hone by phorbol diesters and melittin. Prostaglandins 16:221-232. 15. Ransjo M. Lerner U H 1990 Calcitonin causes a sustained inhibition of protein kinase C-stimulated bone resorption in contrast 10 the transient inhibition of parathyroid hormoneinduced inhibition. Acta Endocrinol (Copenh) 123:251-256. 16. Ransjo M, Lerner UH 1990 12-O-tetradecanoylphorbol-13acetate, a phorhol ester stimulating protein kinase C , inhibits bone resorption in vitro induced by parathyroid hormone and parathyroid-hormone-related peptide of malignancy. Acta Physiol Scand 139:249-250. 17. Abraham DC, Wadkins CL. Conaway H H 1988 Enhancement of fetal rat limb bone resorption by phorbol ester and ionophore A-23187.Calcif Tissue Int 42:191-195. 18. Lorenzo JA, Sousa S 1988 Phorbol esters stimulate hone resorption in fetal rat long-bone cultures by mechanisms independent of prostaglandin synthesis. J Bone Miner Res 3:6367. 19. Collins DA, Chambers TJ 1991 Effect of prostaglandins E , , E,, and FZnon osteoclast formation in mouse bone marrow cultures. J Bone Miner Res 6:157-164. 20. Arnett TR, Dempster DW 1987 A comparative study of disaggregated chick and rat osteoclasts in vitro: Effects of calcitonin and prostaglandins. Endocrinology 120:602-608. 21. Chambers T J , Moore A 1983 The sensitivity of isolated osteoclasts to morphological transformation by calcitonin. J Clin Endocrinol Metab 572319-824. 22. Chambers TJ, Ali NN 1983 Inhibition of osteoclastic motility by prostaglandins 12, E , , E, and 6-oxoE,. J Pathol 139:383397. 23. Kaneko Y 1991 Vacuole formation and cytokeratin rearrangement of hepatoma cells induced by teleocidin are not associated with down-regulation of protein kinase C . Cancer Res 51:2677-2682. 24. Virji MAG, Steffes MW, Estensen RD 1978 Phorbol myristate acetate: Effect of a tumor promoter on insulin release from isolated rat islets of Langerhans. Endocrinology 102: 706-711. 25. Nishizuka Y 1986 Studies and perspectives of protein kinase C. Science 233:305-312. 26. Young S, Parker P J , Ullrich A, Stabel S 1987 Down-regulation of protein kinase C is due to an increased rate of degradation. Biochem J 244:775-779. 27. Stabel S, Rodriguez-Pena A, Young S, Rozengurt E, Parker PJ 1987 Quantitation of protein kinase C by immunoblot -
EFFECT OF PHORBOL ESTER ON OSTEOCLASTS expression in different cell lines and response to phorbol esters. J Cell Physiol 13O:lll-117. 28. Murrills RJ, Shane E, Lindsay R, Dempster DW 1989 Bone resorption by isolated human osteoclasts in vitro: Effects of calcitonin. J Bone Miner Res 4:259-268. 29. Teti A, Colucci S, Grano M, Argentino L, Zambonin Zallone A 1990 Protein kinase C regulates the organization of the cytoskeleton in osteoclasts (abstract). J Bone Miner Res S(Suppl 2):S213.
Address reprint requests to: Dr. D. W. Dempster Regional Bone Center Helen Hayes Hospital Route 9 W West Haverstraw, N Y 10993 Received for publication March 28, 1991; in revised form October 21, 1991; accepted October 22, 1991.
423
Effects of Phorbol Myristate Acetate on Rat and Chick Osteoclasts RICHARD J . MURRILLS,’ LINDA S. STEIN,’ WENDY R. HORBERT,’ and DAVID W. DEMPSTER’.’
ABSTRACT The role of protein kinase C in the regulation of osteoclast function is not known. We therefore compared the effect of phorbol myristate acetate (PMA), which activates protein kinase C, on the resorptive function, motility, and morphology of osteoclasts from rat and chick. PMA caused a significant reduction in resorption pit number in both species; rat osteoclasts were more sensitive, being significantly inhibited at doses of 10-9-10-6M compared with 10-7-10-6M for chick osteoclasts. The inactive analog PMA-(r was without significant effect, and inhibition was not blocked by M indomethacin. In time course experiments, inhibition at 24 h was similar to or greater than inhibition at 6 h, indicating a persistent or progressive effect on bone resorption. Removal of PMA after 6 h prompted partial recovery of bone-resorptive ability in chick M osteoclasts but not rat, at least over a 48 h incubation. In time-lapse video studies of rat osteoclasts, PMA produced an immediate but transient cessation of motility and retraction of the cell margin into prominent filopodia. Motility resumed within 2.5 h after addition, but the osteoclasts remained partially contracted. Chick osteoclasts behaved similarly but showed no formation of filopodia at the cell periphery and a more rapid recovery of motility than rat osteoclasts; chick osteoclasts also underwent a transient vacuolation following PMA exposure, whereas rat osteoclasts did not. Despite differences in the sensitivity of rat and chick osteoclasts to PMA, these results suggest a fundamental role for protein kinase C in the inhibition of osteoclasts from both species.
INTRODUCTION H E R E IS NOW strong evidence that both mammalian and and avian osteoclasts possess an inhibitory second messenger pathway mediated by cyclic AMP. However, in addition to cyclic AMP, most cells also possess intracellular signal transduction pathways dependent on the hydrolysis of membrane phospholipids.(’’ In one such pathway, binding of a hormone or growth factor to the cell surface causes breakdown of membrane phospholipids generating diacylglycerol (DAG), which in turn binds to and activates protein kinase C (PKC). This apparently ubiquitous enzyme then phosphorylates proteins involved in cell differentiation, division, or function, thus effecting a change in the cell’s behavior. PKC exists in multiple
T
forms that vary in their tissue distribution, substrate specificity, and regulation.(8) Its role in osteoclasts is unknown. Phorbol esters are structurally similar to the DAG molecule and activate protein kinase C without requiring binding of a hormone or growth factor to the cell surface. To investigate the role of protein kinase C in osteoclasts, we tested the effects of phorbol 12-myristate 13-acetate (PMA) on bone resorption and the motility and morphology of isolated rat and chick osteoclasts.
MATERIALS AND METHODS
Phorbol myristate acetates and indomethacin Phorbol 12-myristate 13-acetate was obtained from Sigma Chemical Co. (St. Louis, MO) and 4-a-phorbol 12-
‘Regional Bone Center, Helen Hayes Hospital, West Haverstraw, New York. zDepartment of Pathology, Columbia University, New York, New York.
415
416
MURRILLS ET AL.
myristate 13-acetate (PMA-a) from Biomol Research Laboratories, Inc. (Plymouth Meeting, PA). Each was dissolved in dimethylsulfoxide (DMSO) at 20 mM and stored -80°C. PMA and PMA-a were then added to 199+Earle’s (see later) and used at 10~’o-10~6 M. Indomethacin was dissolved in absolute ethanol at lo-* M, M. Appropriate stored at -8O”C, and used at amounts of vehicles were added to ensure that the concentration of vehicles in culture media was the same in all treatments within an experiment.
Rat Osteoclasts
Bone resorption was assayed using the disaggregated osteoclast resorption assay originated by Boyde et al.C9)and Chambers et aI.l1O1and performed as described by Dempster et al.C”lBriefly, osteoclasts were recovered by mincing the bones of neonatal rats or embryonic chicks in 3 ml medium 199 containing Hank’s salts (199+ Hank’s) and settling 25-100 pl of the resultant cell suspension for 30 minutes (rat) or 90 minutes (chick) onto devitalized 4.4 x 4.4 mm bovine cortical bone slices prewetted with 100 p1 of 199+ Hank’s. The slices were then rinsed to remove nonadherent cells and transferred to tissue culture medium (medium 199 containing Earle’s salts, 0.7 g/liter of sodium bicarbonate, and 10% heat-inactivated fetal bovine serum, FBS) containing vehicle alone or specified concentrations of phorbol esters and/or indomethacin. After a 24 h incubation in a 5 % CO, humidified incubator, the slices were fixed in 2% glutaraldehyde and stained with either toluidine blue (TB) or for tartrate-resistant acid phosphatase (TRAP) using Sigma kit 386-A (Sigma Chemical Company, St. Louis, MO). The number of multinucleate osteoclasts ( 2 three nuclei for TB-stained slices and 2 two nuclei for TRAP-stained slices) were counted on one to four replicate slices per treatment per animal. Cells were stripped from the slices by sonication in 0.25 M NH,OH (two 15 s bursts), rinsed in distilled water, and air dried from acetone. The number of resorption pits, which we have shown to be significantly correlated to the total plan area of bone resorbed,I6)was counted in the light microscope on three to four replicates per treatment per animal after staining in toluidine blue for a total of 4 minutes. The experimental protocols were approved by the Institutional Animal Care and Use Committee (IACUC) of Helen Hayes Hospital.
Statistical analysis Slices were allocated to treatment groups using a randomized block design.‘”’ Osteoclasts recovered from an individual rat pup or chick embryo were settled onto four replicate slices per treatment, and typically five animals, each constituting one “block,” were used per experiment. The data were subjected to two-way analysis of variance (ANOVA) to determine the standard error of the difference between treatment means and differences from control tested for significance using the method of least significant difference. When variances between treatments were unacceptably different, data were transformed by square root or logarithmic transformation. A multiple-range test
**
20
Bone resorption assay
T
**
**
T
T
10
Chick Osteoclasts
Y
I
I
FIG. 1. Effect of PMA on rat and chick osteoclasts. Values are means + SEM, n = 5 ; **p < 0.001; * p < 0.05.
was employed for multiple comparisons. In two instances, indicated in the legend to Fig. 4 and in Table 2, unacceptably different variances could not be rectified by data transformation, thus preventing analysis of the experiment as a whole by a multiple-range test; when this occurred, a limited number of selected treatments were compared using a t-test.
RESULTS
Bone resorption and osteoclast counts PMA caused a significant dose-responsive reduction in the number of pits excavated by both neonatal rat and em-
417
EFFECT OF PHORBOL ESTER ON OSTEOCLASTS
Rat Osteoclasts
I% v) u
30
T
iz
L
0 L
0,
20
n
E, 10
0 Control
1WRM PMA
1O-RM 1W6 M P M A alpha
Chick Osteoclasts
T
I
& A
*
Cell morphology and motility
T
Control
10.’ M PMA
deed, P M A ( M) significantly increased osteoclast numbers over 24 h in five of nine chick experiments. In these experiments, the increase in osteoclast number tended to offset the inhibitory effects of P M A on the number of pits per slice, but significant inhibition was clearly present if resorption was expressed as pits per osteoclast (Table I). Significant osteoclast formation was seen more consistently (three of four experiments) with dense preparations of chick embryo bone cells (100 pI plating volume). At the lower cell densities (25 pl plating volume), significant osteoclast formation was seen in two of five experiments. Time course studies indicated that between 6 and 24 h there was no statistically significant change in the degree of inhibition, as determined by test to control ratios, in either the rat or the chick. However, inhibition tended to be greater at 24 h than at 6 h (Table 2), and significant inhibition was noted in the rat after 48 h exposure to a low dose but none was demonstrable at 6 h (Fig. 4). In recovery experiments in which slices exposed to PMA were transferred to control media for subsequent incubation, the chick osteoclasts partially recovered from the effects of P M A on bone resorption, whereas the rat osteoclasts did not, at least over 48 h at the doses tested (Fig. 4 and Table 3). The effect on rat osteoclasts was “inducible”: exposure for only the first 6 h of the experiment resulted in almost exactly the same degree of inhibition after 48 h as osteoclasts exposed continuously. I t should be noted that no loss of osteoclasts was observed in either species in the incubation periods examined (up to 48 h).
10.‘ M 10.7~ P M A alpha
On rat and chick F1c* 2’ Specificity Of effect Of osteoclasts. Values are means f SEM, n = 5 (rat) or 10 (chick); * p < 0.05.
bryonic chick osteoclasts over 24 h. Rat osteoclasts were more sensitive, significant inhibition being noted at doses of 10-9-10-6M compared with lo-’ and lo+’ M in the chick (Fig. 1 and Table 1). Doses higher than M were not tested because of the unacceptably high concentration of vehicle necessary. Inhibition was specific, the inactive analog of PMA, PMA-a, having no significant effect (Fig. 2). Inhibition by P M A was not blocked by M indomethacin, suggesting that inhibition was not due to prostaglandin production (Fig. 3). Reduction in pit number was not due to loss of osteoclasts in either the rat or the chick experiments, suggesting that inhibition was not due to cytotoxicity (Table 1). In-
Neonatal rat osteoclasts underwent a rapid change in morphology and motility in response to lo-‘ M PMA, which was most marked soon after addition (Fig. 5). Before addition, osteoclasts constantly extended and then retracted phase-dense lamellipodia from apparently random locations at the cell periphery (lamellipodial activity). Within 5 minutes of adding PMA, lamellipodial activity ceased and the cell periphery withdrew, forming prominent filopodia, such that in some osteoclasts the cell periphery consisted almost entirely of filopodia after as little as 10 minutes. Lamellipodial activity returned at least partially within 2.5 h after addition, the timing and extent varying from osteoclast to osteoclast. The inactive analog of PMA, P M A - a , had no apparent effect on the motility or morM. phology of rat osteoclasts at The motility and morphology of chick osteoclasts was M PMA, the effects being most also influenced by marked soon after addition (Fig. 6). Like rat osteoclasts, lamellipodial activity was immediately eliminated or greatly reduced and the cell periphery appeared to withdraw, the strength of the response varying from osteoclast to osteoclast. Unlike rat osteoclasts, prominent filopodia were not noted and lamellipodial activity returned after a shorter period than in rat osteoclasts. Time course studies of the morphology of cells on bone slices revealed a tranM sient vacuolation of chick osteoclasts exposed to PMA, vacuoles being visible (at x 160 magnification) in 42% of the TRAP-positive multinuclear cells after 6 h in-
MURRILLS ET AL.
418
Rat Osteoclasts
Rat Osteoclasts
I
I ‘ l i m e (hrs.)
4.f’l 4.5,
C h ic k
( )st cod a s t s
3.5
Control
PMA
Indo
PMA+Indo
Chick Osteoclasts
’l’imc. ( h n . )
FIG. 4. Recovery of chick, but not rat osteoclasts following 6 h exposure to PMA. Doses used are M (chick) SEM for four to five and 10+’ M (rat). Values are means animals, two replicates per treatment per animal. Significant differences ( P < 0.05) were detected using a multiplerange test. Because of unacceptably different variances, the 48 h group was analyzed by a multiple-range test separately from the 6 h group (which was analyzed using a 1test). Hence no comparison between these groups is shown. asignificant difference from control 6 h; bSignificant difference from PMA 6 h; CSignificant difference from control 48 h; dSignificant difference from PMA removed at 6 h; esignificant difference from PMA 48 h.
*
I
W L
-
B. Y)
h
10
L
0 L W
n
z$
5
0
(’ontrol
mit\
Indo
PhlA+lndo
FIG. 3. Effect of indomethacin M) on the response of rat and chick osteoclasts to PMA. PMA was used at lo-* M (rat) and M (chick). Values are means SEM, n = 5 (rat) or 10 (chick); * * p < 0.001; *p < 0.05.
*
vacuoles. Vacuole formation was never observed in rat osteoclasts in association with PMA exposure. The M PMA-(Y had no apparent effect on the motility or morphology of chick osteoclasts.
DISCUSSION cubation in PMA (compared with none in controls), the proportion falling to 13% after 24 h (compared with 6% in controls; Table 4). Importantly, it should be noted that no significant reduction in osteoclast numbers was noted at either of these times (Table 4) and at 24 h the total number of chick osteoclasts was increased in the presence of PMA. Video studies confirmed the PMA-induced vacuolation (Fig. 61, which was sometimes extensive, and also revealed the ability of the osteoclasts to rid themselves of vacuoles by either a rapid “deflation,” suggestive of a possible expulsion from the cell, or a more gradual shrinkage of the
These results show that both neonatal rat and embryonic chick osteoclasts were inhibited by the phorbol ester, PMA, a known activator of protein kinase C. The effect was specific, not occurring with the inactive analog PMA(Y and was not due to loss of osteoclasts, even at the highest doses tested. PMA also caused characteristic and reproducible changes in osteoclast motility and morphology. This strongly implicates protein kinase C in the inhibition of osteoclasts and has important ramifications for understanding the signal transduction mechanisms of antiresorptive agents. The finding of a common inhibitory effect of PMA in both neonatal rat and embryonic chick osteoclasts
419
EFFECT OF PHORBOL ESTER ON OSTEOCLASTS TABLE1. EFFECTOF PMA Treatment
ON
BONERESORPTION BY RATOR CHICKOSTEOCLASTS~
Pits per slice
Rat osteoclasts Control M PMA 10-9 M PMA M PMA Chick osteoclasts Control M PMA 10-7 M PMA M PMA
15.9 14.7 10.2 7.4
f f f f
3.4 2.3 1.9b 1.4~
52.2 f 17.8 35.2 k 7.7 43.2 f 9.8 30.9 f 7.7
Osteoclasts per slice
Pits per osteoclast
20.6 21.0 22.6 18.4
f f f f
4.5 1.1 3.0 1.7
0.81 f 0.15 0.73 f 0.10 0.47 f 0.09b 0.30 k 0.03d
22.9 25.4 28.7 35.6
f f f f
5.6 7.8 3.7 4.1b
2.31 k 0.63 1.62 f 0.41 1.47 k 0.29b 0.85 f 0.17d
aMultinucleate osteoclasts were counted and pits per osteoclast calculated on four (chick) or one (rat) replicate slices following tartrate-resistant acid phosphatase staining. Plating volume was 100 pl. Values are means f SEM, n = 5 animals (four replicate bone slices per treatment per animal); significant difference from control is indicated. hp < 0.05. Lp < 0.01. dp < 0.02.
TABLE 2 . TIMECOURSEOF INHIBITION OF RATAND CHICKOSTEOCLASTS BY
M PMAa T / C Ratio
Treatment
Pits per sfiee
Rat osteoclasts Control 6 h PMA 6 h Control 24 h PMA 24 h Chick osteoclasts Control 6 h PMA 6 h Control 24 h PMA 24 h
OCs per slice
12.9 7.0 27.2 10.0
f f f f
2.9 1.4b 4.7c-e 2.6b
24.4 f 7.3 27.0 k 2.9 29.2 k 4.4 27.2 f 5 . 1
4.9 4.0 46.0 33.5
f f f f
1.6b3e 1.3b.e 14.7C3d 7.6Cvd
19.8 43.5 39.7 76.4
f f f f
3.8b,",e 5.lc3e 5.7cve 7.2b-d
Pits per OC 0.34 f 0.12 0.08 f 0.02NS 0.67 f 0.05 0.25 k 0.09f
0.15 -t 0.05b.e 0.08 f 0.02b.e 1.09 k 0.16c-e 0.44 f 0.09b-d
Pits per sfiee
Piis/OC
0.21
0.56 f 0.12
0.48
0.36 f 0.05
0.39 f 0.14
0.90 f 0.17
0.69 f 0.29
0.83 f 0.15
0.41 f 0.06
k
aosteoclasts (OCs) were counted following TRAP staining; chick osteoclast counts include vacuolated cells recognizable as osteoclasts. SEM for five animals, one to four replicate slices per treatment per animal. Significant differences (p < 0.05) were Values are means detected using a multiple-range test, except for the Pits per OC column in the rat where, for reasons detailed in Materials and Methods, I-tests were used to compare PMA-treated groups to their respective controls (NS = not significant). hsignificant difference from control 24 h. CSignificant difference from control 6 h. dsignificant difference from PMA 6 h. eSignificant difference from PMA 24 h. (p < 0.005.
*
echoes our finding of a common inhibitory effect of stimulators of intracellular cAMPl6) and points to fundamental similarities of osteoclasts from these two sources, despite differences in the numbers of their calcitonin receptor^."^) In organ culture, phorbol esters have been found to have both stimulatory and inhibitory effects. In neonatal mouse calvariae, basal resorption was stimulated by phorbol esters by a process that was dependent on prostaglandin ~ y n t h e s i s . ' ~ Parathyroid ~,~~) hormone (PTH)-stimulated resorption, however, was inhibited by phorbol ester.i16)In fetal rat long bone organ culture, phorbol esters stimulated resorption'L'.18)independently of prosta-
glandin synthesis but required DNA synthesis. (I8) Our results show that the primary effect of PMA on mature osteoclasts was a powerful inhibition, in both the rat and the chick, which was independent of prostaglandin synthesis. In the chick osteoclast assay there was also evidence of significant increases in osteoclast numbers with PMA. This dual effect of inhibiting the action of mature osteoclasts while stimulating the formation of new osteoclasts is reminiscent of the actions of prostaglandins(l91 and agents that stimulate cyclic AMP(6) and may explain the ability of phorbol esters to have both stimulatory and inhibitory effects in organ culture.
420
MURRILLS ET AL.
FIG. 5 . Time-lapse micrographs of the response of rat osteoclasts to M PMA (A) immediately before the addition of PMA, (B) 100 minutes after, and ( C )200 minutes after addition.
FIG. 6. Time-lapse micrographs of the response of chick osteoclasts to M PMA (A) immediately before the addition of PMA; (B) 5 minutes after the addition of PMA; and (C) 60 minutes after addition. Note the beginnings of vacuolation in C.
Activation of protein kinase C caused changes in the morphology and motility of both rat and chick osteoclasts. These have some similarity with the changes occurring after exposure of sensitive osteoclasts to calcitonin‘ 2 0 , 2 1 ) or prostaglandins‘1.2z’and raise the possibility that protein kinase C may play a role in the transduction of these changes in morphology and motility. I t should be noted, however, that for logistical reasons the dose of PMA in the present study was insufficient to eliminate bone resorption. This should be considered when comparing the transient and incomplete effects of PMA on motility and mor-
phology with the full-blown contraction and adoption of the “stellate” morphology achieved by doses of calcitonin (CT) or stimulators of CAMP that completely inhibit resorption. (6.L1.10.21)In addition to the changes in motility and morphology usually associated with inhibitory agents, chick osteoclasts also underwent a transient vacuolation. The significance of this vacuolation is not known: vacuolation has been noted in hepatoma cells exposed to the PKC agonist t e l e o ~ i d i n , ‘ ~ ~ ) and PMA was also shown to stimulate the secretion of insulin in pancreatic islet cells,124’ raising the possibility that PKC plays a role in exocytosis or endocytosis in chick osteoclasts.
42 1
EFFECT OF PHORBOL ESTER ON OSTEOCLASTS N i s h i z ~ k a 'listed ~ ~ ) a large number of substrates reported to be targets for phosphorylation by PKC; of possible relevance to the morphologic and functional changes r e p o r t e d here a r e vinculin, ( N a + , H + ) - A T P a s e , (Na+,K+)-ATPase,and the pp60src protein, each of which is thought to play a prominent role in osteoclast activity.
TABLE 3. PIT AND OSTEOCLAST COUNTS FOR EXPERIMENTS IN FIG. 4a
Treatment
Pits per slice
Rat osteoclasts Control 6 h PMA 6 h Control 48 h PMA 0-6 h then control 6-48 h PMA 48 h Chick osteoclasts Control 6 h PMA 6 h Control 48 h PMA 0-6 h then control 6-48 h PMA 48 h
THE
OCs per slice
16.0 f 4.0b 14.6 f l . l b 54.5 f 8.lC-f
31.6 f 4.4 35.1 f 5.6 32.6 f 4.0
24.9 T 3 . P 23.8 f 4.3h
35.6 k 3.9 33.9 f 5.8 33.2 f 10.3e.f 53.7 f 14.8f 52.3 f 15.3f 83.1 f 2 3 . 4 ~ 93.2 + 18.Ob-d
M and aNote that the dose of PMA used for the chick was for the rat M; 100 pl plating volume; osteoclasts counted following TRAP staining. Values are means f SEM for four to five animals, two replicates per animal. Significant differences (p < 0.05) determined by a multiple-range test. bsignificant difference from control 48 h. CSignificant difference from control 6 h. dSignificant difference from PMA 6 h. eSignificant difference from PMA 0-6 h then control 6-48 h. ?Significant difference from PMA 48 h.
Classically, exposure to PMA initially activates and then downregulates PKC as a result of a n increased rate of degradation of the enzyme,(26)the timing and extent of downregulation being dependent on the cell type. (27) Consistent with this, transient effects were seen on the motility and morphology of both rat and chick osteoclasts, the initial dramatic changes being almost completely reversed within a few hours. However, studies on bone resorption indicated that the resorptive activity of both chick and rat osteoclasts was inhibited to a constant or increasing degree. These results suggest that although some downregulation of PKC occurred, this was not sufficient for reinstatement of bone resorption. It should be borne in mind that changes in morphology and motility may reflect only a part of the multifactorial process of bone resorption and that the restoration of several other systems (which may have different sensitivities or kinetics) may be necessary for bone resorption to recommence. Bone resorption may therefore be more sensitive to inhibitory agents than changes in morphology or motility, an observation we have already made with respect to C T and stimulators of intracellular cyclic AMP.(6." **' Inhibition of osteoclast activity could theoretically be due to a cytotoxic effect on the osteoclast. We tested this possibility using cell counts conducted at the end of the incubation period and "recovery" experiments in which slices were removed from PMA after a brief exposure and transferred to control media. Chick osteoclasts showed an increase in osteoclast number over time and were clearly capable of recovering from an inhibitory dose of PMA; rat osteoclasts, however, did not recover from the effects of PMA, a short (6 h) exposure being sufficient to elicit all the inhibition seen at 48 h with continuous exposure. A cytotoxic effect on the rat osteoclast cannot therefore be ruled out, but we think this unlikely for the following reasons. First, no loss of osteoclasts was noted in cell counts
TABLE4. EFFECTOF PMA M) ON VACUOLATION OF CHICK OSTEOCLASTS (100 p~ PLATINGVOLUME)^
Treatment
Non vacuolated OCs per slice
Control 6 h PMA 6 h Control 24 h PMA 24 h
19.7 3.7 24.7 f 2.5 37.6 k 5.9bSd3f 66.2 ? 5.4b,c,f
*
Vacuolated OCs per slice 0.1 f O.lb-d 18.8 f 3.4CJ 2.1 f 0.7b,d,f 10.2 1.8C,f
Total OCs per slice 19.8 43.5 39.7 76.4
f f f f
3.8b-d 5.ld3f 5.74f 7.2b.C,f
Ratio of vacuolated to total OCs per slice
0 f oe 0.42 0.04C.d 0.06 f 0.02b 0.13 f 0.01b
=TRAP-stained osteoclasts were deemed vacuolated if at least one cytoplasmic vacuole, however small, was visible at a magnification of x 160. Greatly distended TRAP-positive vacuolated cells in which nuclei could not be seen were excluded. Values are means k SEM for five chicks, two to four replicate slices per treatment per animal. Significance (p < 0.05) was determined using a multiple-range test. bSignificant difference from PMA 6 h. CSignificant difference from control 24 h. dSignificant difference from PMA 24 h. eExcluded from statistical analysis. fSignificant difference from control 6 h.
MURRILLS ET AL.
422
conducted at the end of our experiments, even when incubation was extended to 48 h. Second, organ culture studies using higher doses than those used here showed a longterm stimulation of resorption (presumably due to osteoclast formation), an observation inconsistent with cytotoxicity. Third, video studies showed recovery of motility in mature osteoclasts. Fourth, the inactive analog, PMA-a, showed no inhibitory effects. On balance, we therefore interpret the rat osteoclast’s relative inability to recover from an inhibitory dose of PMA as being due either to its extreme sensitivity or to the long-term nature of PMA-induced activation of PKC,‘”’ not to a cytotoxic effect. In conclusion, we have shown that phorbol myristate acetate had potent inhibitory effects on bone resorption by both neonatal rat and embryonic chick osteoclasts. Exposure of osteoclasts to PMA caused a significant reduction in their ability to resorb bone and was associated with marked but transient changes in cell morphology. These observations suggest that protein kinase C plays a fundamental role in the transduction of inhibitory signals in both mammalian and avian osteoclasts and raise the possibility, for which there is already some preliminary evidence,(**’that protein kinase C acts, either directly or indirectly, on the proteins of the osteoclast cytoskeleton.
ACKNOWLEDGMENTS Parts of this study were presented at the Washington University Continuing Edueat ion Symposium, Molecular Basis of Bone Cell Physiology: Transcellular Signalling, St. Louis, MO, April 1991, and at the Thirteenth Annual Meeting of the ASBMR, San Diego, CA, August 1991. This work was supported in part by NIH Grant AR39191. We thank Ms. S. Roberts, Mr. B. Yarborough, and Mr. L . Ferguson for their expertise in graphics and photography.
REFERENCES I . Chambers T J , Dunn CJ 1982 The effect of parathyroid hormone, 1,25-dihydroxycholecalciferoland prostaglandins on the cytoplasmic activity of isolated osteoclasts. J Pathol 137: 193-203. 2. Nicholson GC, Livesey SA, Moseley JM, Martin TJ 1986 Actions of calcitonin, parathyroid hormone, and prostaglandin E1 on cyclic AMP formation in chicken and rat osteoclasts. J Cell Biochem 31:229-241. 3. Rifkin BR, Auszmann J M , Kleckner AP, Vernillo AT, Fine AS 1988 Calcitonin stimulates CAMP accumulation in chicken osteoclasts. Life Sci 42:799-804. 4. Eliam MC, Basle M, Bouizar Z , Bielakoff J , Moukhtar M, de Vernejoul MC 1988 Influence of blood calcium on calcitonin receptors in isolated chick osteoclasts. J Endocrinol 119:243-248. 5. Ransjo M, Lerner UH. Heersche JNM 1988 Calcitonin-like effects of forskolin and choleratoxin o n surface area and motility of isolated rabbit osteoclasts. J Bone Miner Res 3:611619. 6. Murrills RJ, Dempster DW 1990 The effects of stimulators of intracellular cyclic AMP on rat and chick osteoclasts in vitro: Validation of a simplified light microscope assay of bone resorption. Bone 11:333-344.
7. Nishizuka Y 1984 The role of protein kinase C in cell surface signal transduction and tumour promotion. Nature 308:693698. 8. Exton J H 19% Signalling through phosphatidylcholine breakdown. J Biol Chem 265: 1-4. 9. Boyde A, Ali NN, Jones SJ 1984 Resorption of dentine by isolated osteoclasts in vitro. Br Dent J 156:210-220. 10. Chambers T J , McSheehy PMJ, Thomsom BM, Fuller K 1985 The effect of calcium-regulating hormones and prostaglandins on bone resorption by osteoclasts disaggregated from neonatal rabbit hones. Endocrinology 116:234-239. 11. Dempster DW. Murrills RJ, Horhert WR, Arnett TR 1987 Biological activity of chicken calcitonin: Effects on neonatal rat and embryonic chick osteoclasts. J Bone Miner Res 2: 443-448. 12. Parker RE 1979 Introductory Statistics for Biology, 2nd ed. Studies in Biology, No. 43. Institute of Biology, London. 13. Nicholson GC, Moseley JM. Sexton PM, Martin TJ 1987 Chicken osteoclasts d o not possess calcitonin receptors. J Bone Miner Res 2:53-59. 14. Tashjian A H Jr, lvey JL, Delclos B, Levine L 1978 Stimulation of prostaglandin production in hone by phorbol diesters and melittin. Prostaglandins 16:221-232. 15. Ransjo M. Lerner U H 1990 Calcitonin causes a sustained inhibition of protein kinase C-stimulated bone resorption in contrast 10 the transient inhibition of parathyroid hormoneinduced inhibition. Acta Endocrinol (Copenh) 123:251-256. 16. Ransjo M, Lerner UH 1990 12-O-tetradecanoylphorbol-13acetate, a phorhol ester stimulating protein kinase C , inhibits bone resorption in vitro induced by parathyroid hormone and parathyroid-hormone-related peptide of malignancy. Acta Physiol Scand 139:249-250. 17. Abraham DC, Wadkins CL. Conaway H H 1988 Enhancement of fetal rat limb bone resorption by phorbol ester and ionophore A-23187.Calcif Tissue Int 42:191-195. 18. Lorenzo JA, Sousa S 1988 Phorbol esters stimulate hone resorption in fetal rat long-bone cultures by mechanisms independent of prostaglandin synthesis. J Bone Miner Res 3:6367. 19. Collins DA, Chambers TJ 1991 Effect of prostaglandins E , , E,, and FZnon osteoclast formation in mouse bone marrow cultures. J Bone Miner Res 6:157-164. 20. Arnett TR, Dempster DW 1987 A comparative study of disaggregated chick and rat osteoclasts in vitro: Effects of calcitonin and prostaglandins. Endocrinology 120:602-608. 21. Chambers T J , Moore A 1983 The sensitivity of isolated osteoclasts to morphological transformation by calcitonin. J Clin Endocrinol Metab 572319-824. 22. Chambers TJ, Ali NN 1983 Inhibition of osteoclastic motility by prostaglandins 12, E , , E, and 6-oxoE,. J Pathol 139:383397. 23. Kaneko Y 1991 Vacuole formation and cytokeratin rearrangement of hepatoma cells induced by teleocidin are not associated with down-regulation of protein kinase C . Cancer Res 51:2677-2682. 24. Virji MAG, Steffes MW, Estensen RD 1978 Phorbol myristate acetate: Effect of a tumor promoter on insulin release from isolated rat islets of Langerhans. Endocrinology 102: 706-711. 25. Nishizuka Y 1986 Studies and perspectives of protein kinase C. Science 233:305-312. 26. Young S, Parker P J , Ullrich A, Stabel S 1987 Down-regulation of protein kinase C is due to an increased rate of degradation. Biochem J 244:775-779. 27. Stabel S, Rodriguez-Pena A, Young S, Rozengurt E, Parker PJ 1987 Quantitation of protein kinase C by immunoblot -
EFFECT OF PHORBOL ESTER ON OSTEOCLASTS expression in different cell lines and response to phorbol esters. J Cell Physiol 13O:lll-117. 28. Murrills RJ, Shane E, Lindsay R, Dempster DW 1989 Bone resorption by isolated human osteoclasts in vitro: Effects of calcitonin. J Bone Miner Res 4:259-268. 29. Teti A, Colucci S, Grano M, Argentino L, Zambonin Zallone A 1990 Protein kinase C regulates the organization of the cytoskeleton in osteoclasts (abstract). J Bone Miner Res S(Suppl 2):S213.
Address reprint requests to: Dr. D. W. Dempster Regional Bone Center Helen Hayes Hospital Route 9 W West Haverstraw, N Y 10993 Received for publication March 28, 1991; in revised form October 21, 1991; accepted October 22, 1991.
423
