JOURNAL OF HONE AND MINERAL RESEARCH Volume I I , Number 4, I996 IHackwell Science, Inc. ((') I996 American Society for Bone and Mineral Research

Effects of Continuous Calcitonin Treatment on Osteoclast-like Cell Development and Calcitonin

Receptor Expression in Mouse Bone Marrow Cultures

MIKA IKEGAME. MARYANN RAKOPOULOS, T. JOHN MARTIN, JANE M. MOSELEY and DAVID M. FINDLAY

ABSTRACT

Continuous treatment with calcitonin (CT) to inhibit osteoclastic bone resorption results in acquired resistance. The mechanisms of this "escape" phenomenon are not yet established. The aim of this study was to examine the effects of continuous treatment with CT on the generation of osteoclasts and calcitonin receptor (CTR) expression in mouse bone marrow cultures. This was done by daily CT treatment of mouse bone marrow cultures from day 0, when only undifferentiated mononuclear precursors of osteoclast-like cells were present, or commencing from day 6, when differentiated osteoclast-like cells were abundant. The response to CT treatment was determined by quantitation of cells positive for tartrate-resistant acid phosphatase (TRAP) and binding of 12slI-salmon CT. Calcitonin receptor and TRAP mRNA levels were determined using semi-quantitative reverse transcription/ polymerase chain reaction. When cultures were treated with CT from day 0, TRAP-positive multinucleated cells appeared. These cells expressed only very low levels of CTR or CTR mRNA and were morphologically indistin- guishable from osteoclast-like cells formed in control cultures. They also displayed the ability to resorb bone. Continuous CT treatment of cultures from day 6 rapidly reduced the CTR mRNA levels, with a t1,* of 6 to 12 h, and these levels remained low thereafter. '*'I-salmon CT binding capacity, as determined by autoradiography, was lost in parallel. These effects were specific for the CTR since there was no consistent effect on TRAP mRNA levels. Based on these data, we suggest that the "escape" phenomenon may result from a prolonged CT-induced loss of CT responsiveness due, at least in part, both to reduced synthesis of CTR, and to the appearance in bone of CTR-deficient osteoclasts. (J Bone Miner Res 19961 1:456-465)

INTRODUCTION

sit oci A ~ I ' S , which arc multinuclcated cells dcrived 0 from hemopoictic progcnitors,' ' are rcsponsible for physiological bone resorption. They ;ire characterized by cxpression o f tartrate-resistant acid phosphatase (TRAP),'" and nbundant calcitonin rcceptors (CTRS).'~.''

The pcptide hormonc. calcitonin (CT), directly acts o n ostcoclasts and potently inhibits osteoclastic bone resorp-

Prcscntcd in part at the 12th International Conference on CnI- cium Regulating Hormones, Mclhourne, Australia. February, 1995.

tion. Howcver, in the face o f continuous administration of exogenous CT, bone becomes resistant t o the hormone. Termed "escape", this phenomenon was first demonstrated in bone organ culture."^" Although the cellular and nio- lccular mechanisms arc not yet understood, i t has been suggested that one of the mechanisms of the phenomenon is reduced osteoclastic responsiveness to CT,'X' which may be related to CT-induced down-regulation of CTR.'"' Ho- mologous down-regulation of the CTR has been described in non-osteoclastic cells,'"'' where it is accompanied by desensitization o f the cells to CT challenge." '."' Calcito- nin-induced desensitization has also been observed in short term experiments with freshly isolated rat osteoclasts.'-3'

St. Vincent'\ Institute of Medical Rcsc;lrch, Melhournc. Victoria. Australia.

456

CALCITONIN EFFECTS ON OSTEOCLAST DEVELOPMENT AND CTR mKNA 457

Additionally, i t has been suggested that the generation o f new ostcoclasts might contribute t o "escape"""~'J' with the iniplication that osteoclasts formed in the presence of CT would have reduced responsiveness t o CT. However, the effects o f C'T o n osteoclast development and osteoclast C'TR expression remain unclear.

The aim o f the experiments described here was to exam- ine the effect of continuous exposure to C T in mouse bone ni;irrow cells in culture. This system enabled us to distin- guish between effects o f CT o n expression of the CTR and effects on development o f ostcoclast-like cells in culture. The results ;ire consistent with a persistent down-regulation o f the CTR i n ostcoclast-like cells generated in culture. However. i t is significant that in the continuous presence of CT. CTK-deficient osteochst-like cells were formed that were functionnl i n terms o f bone resorbing ability.

MATERIALS AND METHODS Mirriiic hoiic imtrow cirltirrc

Uone marrow cells harvested from the tibiae. femora and humcri of 3- to 0-week-old C57BliOJ mice (Austin Hospital Animal House. Heidelberg. Victoria, Australia), a s previ- ously dcscrihed,''"' were cultured in 0 cm diameter pctri- dishes (Nunc, Inc.. Napervillc, I L , U.S.A.). The marrow ccHs were plated at ;I concentration o f I x 10" ceIIsicm' and incubated in alpha modified Eagle's medium (aMEM) ( I C " Bioniedicals Inc.. CA. U.S.A.) containing 12.5% fetal bovine serum (Cytosystems Pty. Ltd., Castlehill, N.S.W., Australia) and antibiotics (5 IUirnl penicillin and 5 nigiml streptomycin) at 37°C in a humidified atmosphere o f 5 % CO, and O S ' i , air. Osteoclast-like cell formation was in- duced by I tu.25-dihydroxyvit~imin D, ( I0 nM) (generous gift from Dr M. Uskokovic, Hoffman-LaRoche Inc., Nutlcy, NJ, U.S.A.), added to the cultures on the first day and every third day thereafter, when half the medium was replaced with fresh medium. In parallel cultures, the bone marrow cells wcrc grown on 13 mni diameter Thermanox coverslips (Nunc, Inc.) f o r assessment of cell number, and in some experiments. four devitalized bovine bone slices (4 mm X 4 nim) were added t o the cultures to examine the bone rcsorptivc activity o f the cells.

E.q)crit?icw tcil .schcduIcs

C'T trc~ittnivit ji-oin r l q 0 of'ciiltiirc: To focus on the effects o f continuous CT treatment on ostcoclast development, treatment with 0. I nM salmon C T (sCT) (Bachem, Tor- rance, C'A. U.S.A.) was begun on day 0 of culture (Fig. I ) , when on ly precursor cells were present. and supplemented daily. This concentration o f CT was chosen following pre- liminary dose-response experiments, which showed that this was the lowest (and therefore the most closely physiologi- cal) concentration that rcproducihly resulted in a large (y 8 0 5 ) change in CTR mRNA levels (as assessed by RTIPCR) when administered continuously to cultures. On day 2. 4, 0, and 8 of culture both CT-treated and non- treated control cultures were harvested for RNA extraction, and osteoclast-like cells present on coverslips from parallel

C'T treatment from dav 0 of culture $(.'I

C'T treatment from day 6 of culture

roltiirc i L 1 \Vl

euperinient linw point

FIG. 1. Schematic diagram of the experimental schcdulc. Treatment with 0. I nM sCT was begun on day 0 or day h o f culture. In each experimental protocol. the same conccn- tration of fresh sCT (0.1 nM) was addcd daily. The culture4 were harvested on 2, 4, 6 and 8 days following the start o f treatment on day 0, or a t I . 3, 0, 12. 24, 48 o r 72 h following the start of treatment on day h. At each experimental time point, CT non-treated cultures were harvested a s controls.

cultures, identified by TRAP histochemistry and 1251-1;~- beled sCT ( '251-sCT) autoradiography, were quantitated. On day 8 of culture resorption pit formation on bone slices from parallel cultures was examined by scanning clcctron microscopy following removal o f the cclls.

CT trcutnierit,frotn duy 6 o fu i l t i iw : To examine the effects o f C T on differentiated ostcoclasts. treatment with s('T (0 . I nM) was begun on day 6 of culture. when niultinucleatccl osteoclast-like cells were abundant, and supplemented daily. At 1,3, 6 , 12, 24,48, and 77 ti following the treatment, both treated and control cultures wcrc examined ;IS de- scribed above.

To examine the number o f cells expressing ostcoclast markers, the cells grown on coverslips were used for detec- tion o f TRAP by histochemistry and C T binding by auto- radiography. For '"I-sCT autoradiogr~iphy, the coverslips were first placed in I ml ice-cold 0. IS M NaCT0.OS M glycinc buffer, pH 2.5, for I minute'"" to remove any sCI' ligand already bound t o cell surface CT receptors ("acid- washed'' cells). The cells were then washed twice with uMEM. (In control experiments, thc effects of "acid wash- ing" on the osteoclast-like cclls were investigatcd by first treating thc cells with 0.3 p M sCT briefly for 5 minutes, and then either washing with aMEM alone or "acid-washing". Calcitonin-treated "acid-washed" cells displayed the siimc density of silver granules after '"I-sC'T autoradiography a s non-treated cells, whereas CT-trcatcd cells washed in aMEM alone bound n o '251-~C'T. Hence, from these con- trol experiments, "acid-washing" was observed to effcc- tivcly remove receptor-bound ligand without any damaging effects to the cells). Specific binding 21s well ;is non-specific binding was assessed, followed by TRAP histochemistry ;IS previously described.' I") Briefly. after the specific binding with '"I-sCT (2 X 10' cpminil; 4 . 2 nM) and non-specific binding with excess unlabeled sC'T ( 1 pgiml) in competition

458 IKEGAME ET AL.

with the labeled pcptide for I h, the cells were fixed for 25 minutes in 4%) formaldehyde and then a modified version of the TRAP histochemistry by Janckila et al."') was per- formed. The cells o n coverslips were incubated in 50 mM sodium acetate and 40 mM potassium sodium tartrate buffer (pH S ) , containing naphthol AS-BI phosphate (10 mglml; Sigma Chemical CO., St. Louis, MO, U.S.A.) as substrate and fast garnet GBC (Sigma) for 25 minutes at 37°C. Then, the coverslips were dipped in LM-I nuclear emulsion for autoradiography (Amersham International, Buckinghamshirc, U.K.) and exposed in the dark for 10 days at 4°C. The number of cells exhihiting either TRAP staining alone, or both TRAP staining and CTR positivity per coverslip in 20 random fields of view (covering approx- imately 5% of total coverslip area; x200 magnification) for four coverslips per time point, were counted under light microscopy. In addition, the number of nuclei from 40 randomly selected TRAP( +) multinucleated osteoclast-like cells (4 coverslips; 10 multinucleated cells/coverslip), con- taining two or more nuclei, for each time point were also counted.

Scanning electron microscopy

heim). The amplification was employed after denaturing at 95°C for 5 minutes with a melting temperature o f 92°C for 1 minute, annealing temperatures of 55°C or 65°C for 30 seconds for CTR and TRAP, respectively, and polymcriza- tion at 72°C for 30 seconds, in a Perkin Elmer Cetus 480 thermal cycler. The final polymerization step of the PCR was extended to 10 minutes. Optimal cycle numbers of 28 cycles for CTR and 22 cycles for TRAP were determined from saturation curves, and cycle numbers representing the ascending portion of the amplification curves were chosen.

To ensure equal starting quantities of RNA for the mouse bone marrow time course experiments, the reverse- transcribed RNA samples were also amplified using oligo- nucleotide primers specific for rat glyceraldehyde 3-phos- phate dehydrogenase (GAPDH)."") These were 5' primer, GAPDH-4 (5'-CATGGAGAAGGCTGGGGCTC-3') and 3' primer, GAPDH-5 (5'-AACGGATACATTGGGGGTA G-3'), representing nucleotides 306-325 and 701-720. re- spectively. The amplification was performed with annealing temperature of 55°C and cycled for 20 times, as was dcter- mined from saturation curves.

Southern blotting On day 8 of culture with C T treatment from day 0, cells

o n the bone slices were removed by sonication. The bone slices were then dehydrated in a series of graded ethanols (70 t o 100%) and dried in a 37°C oven overnight. They were then sputter-coated with gold (Dynavac SCl00-M) and ob- served by scanning electron microscopy (Philips SEM515).

Twelve microliters of each PCR product were electro- phoresed in a 1.2% agarose gel, transferred to nylon filters (Hybond-N) (Amersham, Arlington Heights, IL, U.S.A.) and probed separately with ["PI-yATP end-labeled oligo- nucleotides using T4 polynucleotide kinase (BioLabs, Bev- erly, MA, U.S.A.). The CTR oligonucleotide probe used was PS-5 14 (5'-ACCAAGATGAGGCAAACC-3'), com- plementary to a common sequence of the rat C l a and C lb CTR isoforms,'") and which could also detect CTR iso-

Polymerase chain reaction amplification of reverse-transcribed m R N A

Expression of CTR and TRAP reverse-transcribed mRNA from mouse bone marrow cultures was examined by polymerase chain reaction (PCR) amplification and South- ern blot analysis. Total cellular RNA from the bone marrow cells cultured in petri dishes was extracted using guanidine thiocyanate.' I x ) First strand cDNA was synthesized from 2.5 pg o f total RNA by first priming with 0.5 pg random hexamer (Promega) at 70°C for 5 minutes, and then incu- bating for at least I h at 42°C with 12 U o f AMV reverse transcriptasc (Promega, Madison, WI, U.S.A.), 10 mM DTT, I mM each dATP, dCTP, dCTP and dTTP, and 25 U o f RNase inhibitor (Boehringer Mannheim, Mannheim, Germany) in AMV RT reaction buffer (Promega). Two microliters of 25 pI of this reaction mixture were submitted to PCR. A region of the CTR was amplified as previously described,'"') using a 3' primer (5'-ACAAACTGGA(T/ C)(T/G)CCCAGCAGGGGCAC-3') and a 5' primer (5 ' - AAGAACATGTT(C/T)CT(C/GIT)ACTTA-3') based on the rat CTR sequence. For TRAP cDNA amplification, a 3' primer (5'-TCTCTAGATCAGGGTCTGGGTCTCCTTG (3-3') and a 5' primer (5'-GAATTCACAGCCCCCCACT CCCACCCT-3') based o n the mouse TRAP sequence"") were used. Fifty pmol of each primer was added to 20 pl of reaction mixture containing I U Taq DNA polymerase (Boehringer Mannheim), 0.1 mM each dATP, dCTP, dGTP and dTTP, in PCR reaction buffer (Boehringer Mann-

forms found in mouse bone marrow culture-derived osteo- clast-like cells."") Positive controls were PCR products am- plified from C l a or C lb CTR cDNA. The probes used for TRAP and GAPDH hybridization were 5'-GCCACACAG CATCACTGTGTCCAGCAT-3' (representing nucleotides 21 78-2204) and S'-GCTGTGGGCAAGGTCATCCC-3' (representing nucleotides 640-659), according t o the mouse TRAP(''') and rat G A P D H ' ~ ~ ) ) sequences, respcc- tively.

The filters were hybridized with thesc probes for more than 5 h at 42°C and then washed with 2X SSC/O.I%l SDS for 30 minutes at 42°C. Blots were analyzed using it Phos- phorImager (Molecular Dynamics, Sunnyvale, CA, U.S.A.).

Statistical analysis

Graphical representations of cell numbers arc expressed as the mean 5 SEM (standard error of the mean) of four coverslips, the mean values representing the number o f cells counted in 20 random views under light microscopy for each coverslip. Graphical representations of nucleus number are expressed as the mean 5 SEM of 40 TRAP(+) multinuclc- ated cells per time point. Statistical significance was iitiii- lyzed between control and treated cells of the same time point by use of the unpaired Student's t-test. Statistical significance was achieved when p < 0.05.

CALCITONIN EFFECTS ON OSTEOCLAST DEVELOPMENT AND CTR mRNA 459

TRAP(+) TRAP( +)/CTR(. d.

c .

Culture t ime (day)

FIG. 2. Effects of continuous CT treatment from day 0 o n ostcoclast-like cell formation. TRAP( +) ( a ) total (mono- nuc1c;itcd and niultinuclcated). (b) mononucleated. and (c) multinuclc;itcd osteoclast-like cell formation. TRAP(+)/ CTK( +) ( J ) total , ( c ) niononuclcated, and (f) multinuclc- atcd osteoclast-like cell number. CTR( +) cells were dc- fined ;IS ccl ls exhibiting ;I greater concentration of granules compared to background. and midtinucleated cells were defined ;IS those containing two or more nuclei. Results itre expressed ;IS the nican f SEM of the number o f positive cel ls counted in 20 random light microscopy views on each o f four coverslips per point. Open circles, control cultures; closed circles, CT-treated cultures. * / I < 0.OS; "17 < 0.01 compared to control cultures.

RESULTS

E f l i ~ ~ s of C'T trrtrttnctlt ,fi.oni r i r y 0

O.vtc~oc~lrr.st-liI\(. coI1 pwrrrtioti Jkorii prcciir.sot:s: One of the ;tinis o f these experiments was to dctcrniinc the effects o f continuotis C T treatment upon the generation in culture o f ostcoclast-like cells from mousc marrow precursors. CT was therefore iiddcd daily from the beginning o f the culture (sce schcdulc. Fig. I ) . In control cultures of CS7B116J mousc marrow cells treated with I a.25 dihydroxyvitamin D.,, iiiononuclc~itcd and multinucleated cells characterized by the ostcoclast markers TRAP and CTR, typically developed from day 4 of culture. its wc have previously described."") Thcsc cells increased in numher up t o day 8 (Fig. 221-2f). I here was great variability in the intensity o f TRAP staining among the cclls: the very large multinucleated cells tended to show weak staining. Sonic o f the TRAP-positive and C'TK-positive (TKAP( + ) /CrR( +)) cells had it shrunken appearance, which was prohably duc to exposure for 1 h t o '2iI-sCT ( --0.2 nM) iiscd for uutor~idiography, because multiniicleatcd cells stained for TRAP alone were n o t rc- tracted. I n the continuous presence o f CT, similar numbers of hoth monontiele~ited and niultinuclcated TRAP( +) cells developed ;it each time point when compared t o control

?.

cultures (Fig. 2a-2c). These cells were morphologically in- distinguishable from those in control cultures (Fig. 3a. 3h). However, the average number of nuclei possessed by TRAP( +) multinucleatcd cells in CT-treated cultures was found to be significantly less than in controls at id1 time points examined, suggesting an action o f C T i n delaying fusion o f mononucleated precursors (Fig. 4). As deter- mined by ;iutoradiography, the number of cells expressing detectable CTR in these cultures was dramatically reduced by CT treatment from day 0; most of the TRAP( +) cells in these cultures displayed essentially only background silver grains (cornpare Fig. 3a and 31)). tlcncc, we observed very few TRAP( +)/CTR( +) cells in the cultures continuously treated with CT (Fig. 2d-2f). These decreases in i'51-sCT binding were not simply due to prolonged receptor occu- pancy by poorly dissociablc CT, since cel ls were acid- washed to remove cell-surface bound <'T prior t o ",I-sCT binding. Thus, these results indicate tha t the continuous presence of 0.1 iiM sCT in these cultures greatly suppresses CTR expression, but does not alter the number o f TRAP( t ) multinucleated cells formed.

Boric rcsorptir'e uhilir); of ostc~oclrrst-liki, w I I , s fi)nilcil ill tlic prcsorcc of C71 To further authenticate the ostcoclast-like cells that fornied in control and CT-treated cultures. we examined the bone rcsorptive ahility o f the cultures, by investigating pit formation on bone slices. Scanning clec- tron microscopy o f the bone slices showed t h a t resorption pits were formed in control cultures by day H (Fig. S a ) . I n parallel cultures continuously treated with CT from day 0, we found that pits were also formed (Fig. Sh) even though acute CT treatment very potently inhibits ostcoclast pit formation in this assay. with a n ED,,, of 10 I" M.'") Sub- jectively, the pits fornicd in t he continuously treated cul- tures were indistinguishable from control pits in number and size. variability in ostcoclast colony numbcr on the bone slices in these experiments made quantitntion d in these experiments, but there wits no obvious difference i n pit formation between control and CT-treated cultures. Hence, we conclude tha t the cells that arise in cultures con t in uousl y treated with CT ii re ;I ti t he n t ic o s t eocl ;is t s de - spite thc ;ihsence of CTR because these cells express TRAP, are morphologically similar t o multinuclear cells fornied in control cultures, and can resorb bone.

f3/)wssiori of CTR riiRNA iiritl 'I'KAP r?fKN,4: In the control cultures, CTR mRNA levels dctermincd by the reverse transcription PCR (RT-PC'K) method increased in parallel to the increase in number o f TRAP( +)/C'TR( + ) cells, a s described above. These results were consistent with our previous studies.'"" In CT-treated cultures, the in- crease in CTR RT-PCR products was almost totally abro- gated, although low levels were obtained from clay 0 o f culture (Fig. ha, hb). TRAP mKNA, asscsscd by KT-PCR, was expressed at approximately the sanic level a s i n control cultures (Fig. ha, he). These results were again consistent with the data shown in Figs. 2 and 3 which showed t h a t c"I' treatment did not affect the development o f TRAP( +) cells, although CT blocked expression o f CTK in those cells.

Wc have previously reported two C T R isoforms, C I ii and Clb, which differ by the presence in C l b o f a 37 amino acid insert in the second extracellular domain of the receptor.''"

460 IKEGAME ET AL.

. ,

. . '

The C lb isoform displays different pharmacology and dif- ferent tissue localization compared with the C l a isoform. C l a has a wider expression whereas Clb is predominantly expressed in the central nervous system, with very low expression in the periphery,(21) including mouse and rat osteoclasts.(") In the present experiments, only a single band, corresponding to Cla , was seen since the level of expression of the C l b isoform was too low to be detected by 28 cycles of amplification. However, by using 40 cycles of amplification, we observed down regulation of C l b mRNA in cultures continuously treated with CT, in parallel with C l a mRNA down regulation (results not shown).

Effects of continuous CT treatment from day 6

Osteoclust-like cell numbers and CTR expression: To ex- amine the effect of continuous C T treatment upon expres- sion of the CTR in mature osteoclast-like cells, treatment was begun from day 6 of culture, at which time abundant multinucleated osteoclast-like cells were present. Calcito- nin treatment had no consistent effect on the number of TRAP( +) cells in these cultures (Fig. 7a-7c). The binding of '"I-sCT, represented by cell-associated silver grains on individual TRAP( +) cells, appeared to be reduced by 6 h of CT-treatment (compare Fig. 3c and 3d) and remained low at 24 h (Fig. 3e, 3f) and up to 72 h (not shown). This was not necessarily reflected in the number of CTR( +) cells since

FIG. 3. Effects of continuous CT treatment on osteoclast-like cell mor- phology, CTR expression and TRAP activity. After culture, cells on the coverslips were "acid-washed'' to re- move ligand occupying cell surface receptors and "'I-sCT autoradiogra- phy and TRAP staining were per- formed. CTRs are visualised as silver granules surrounding or on the cell. The cells shown here are representa- tive of many cells examined. Calcito- nin treatment from day 0: osteoclast- like multinucleated cells on day 8 from (a) control culture; (b) CT- treated culture, showing very few sil- ver granules. Calcitonin treatment from day 6: osteoclast-like multinu- cleated cell from (c) control culture at 6 h of experimental time; (d) CT- treated culture at 6 h following treat- ment, showing shrunken cytoplasm and relatively few silver granules; (e) control culture at 24 h; (f) CT-treated culture at 24 h following treatment, displaying much less CTR density when compared to control cells, al- though the cytoplasm is spread simi- lar to control. Bars: 25 pm.

Culture time (days)

FIG. 4. Effects of continuous C T treatment on osteoclast- like cell multinucleation. The number of nuclei from 40 randomly selected TRAP( +) osteoclast-like cells (4 cover- slips; 10 multinucleated cells/coverslip), containing two or more nuclei, for each time point were counted. Results are expressed as the mean ? SEM number of nuclei/multinu- cleated cell at each time point from experiment of C T treatment from day 0. Open circles, control cultures; closed circles, CT-treated cultures. *p < 0.05 compared to control cultures.

cells expressing grain density at any level above back ground were scored as CTR(+). However, by 72 h there was a significant decrease in CTR( +) cell number in the cultures (Fig. 7d-7f). Morphologically, CT treatment resulted in

CALCITONIN EFFECTS ON OSTEOCLAST DEVELOPMENT AND CTR mRNA 46 1

FIG. 5. Bone resorptive activity of osteoclast-like cells developed under continuous C T treatment. Scanning elec- tron micrograph of devitalized bovine bone slices on day 8 from (a) control and (b) CT-treated cultures from day 0. Cells were removed from the bonc slices. Both photographs hhow abundant resorption pit formation on the bone sur- face. Bars: 25 pm.

contraction o f the cytoplasm o f TRAP( +) multinucleated cells which was most obvious between 3 and 6 h (for exam- ple, Fig. 3d). The cells showed a return to pretreatment morphology after 12 to 24 h (for example, Fig. 3f), suggest- ing resistance o f the osteoclast-like cells to the inhibitory effects of CT.

Ejil’rcwioii of C‘TR mRNA wid TRAP niRNA: When C T treatment was bcgun from day 6 of culture, RT-PCR prod- ucts corresponding t o CTR mRNA rapidly fell to low levels, with a t , , ? of between 6 and 12 h (Fig. 8a, 8b). Levels were maximally reduced at 24 h and remained low for the course o f thc experiment. In contrast. C T treatment from day 6 had n o consistent effect o n TRAP mRNA levels, again detected by RT-PCR (Fig. 8a, Xc). The decrease at 1 h and increase at 6 h were not consistent in data from other 2 experiments.

DISCUSSION

In this study we have shown that continuous CT treat- ment o f mouse bone marrow cultures. which represent an ostcoclast developmental model system, resulted in pro- longed suppression o f the expression of CTR mRNA and CTR. An important feature o f this system is that it allowed us to study CTR mRNA lcvelh in parallel with observations

2 4 6 8 (day) a.

’ - CT CTR

TRAP

I. 30 I U.

C.

FIG. 6. Effects of continuous CT treatment from day 0 on CTR and TRAP mRNA levels determined by RT-PCR. (a) Southern blot data showing CTR RT-PCR products (up- per), TRAP RT-PCR products (middle), and GAPDH RT- PCR products (lower panel) amplified from control ( - ) and CT-treated (+) cultures on days indicated. Graphical representations of CTR and TRAP RT-PCR products shown in (a) relative to GAPDH RT-PCR products on culture day indicated are shown in (b) and (c), respectively. Each data point is shown as a ratio compared with the value of the control culture on day 2 (arbitrarily assigned the value 1). Open bars, control cultures; closed bars, CT- treated cultures.

on expression of CTRs, bone resorptive capacity, and TRAP mRNA levels and protein expression, during osteo- clast development. The most notable finding was that sup- pression of CTR by CT can result from both a decrease of CTR mRNA in existing osteoclast-like cells and the devel- opment of newly formed osteoclast-like cells that have very low levels of CTR mRNA.

In experiments in which CT was added daily from day 0, TRAP( +) mononucleated and multinucleated cells in- creased in number to a similar extent as in control cultures. However, CTR mRNA expression was dramatically sup- pressed throughout these experiments. This suppression was somewhat specific, in that TRAP mRNA was not ap- preciably affected. Almost no CTR expression was detect- able by autoradiography on TRAP( +) cells in CT-treated cultures. However, the multinucleated cells that appeared in these cultures were indistinguishable from those in con- trol cultures on the basis of morphology and TRAP stain- ing. Importantly, when bone marrow cells were grown on devitalized bone slices, cells were formed which were capa- ble of bone resorption despite continuous C T treatment. There is therefore strong evidence that these cells which developed in the presence of CT exhibited authentic osteo- clast behaviour.

462 IKEGAME ET AL.

TRAP(+) TRAP(+)/CTR(i

d .

e .

6

f .

'l'rciitmrnt time (h)

FIG. 7. Effects o f continuous C T treatment from day h o n osteoclast-like cell formation. TRAP( +) (a) total (mono- nucleated and multinucleated), (b) mononucleated, and (c) multinucle:itcd osteoclast-like cell formation. TRAP(+)/ CTR( + ) ( d ) total, ( c ) mononucleated, and (f) multinucle- iited osteoclast-like cell number. CTR( +) cells were de- fined a s cells exhibiting a greater concentration o f granules compared to background, and multinucleated cells were defined iis those containing two or more nuclei. Results arc exprcsscd a s thc mean ? SEM o f the number o f positive cells counted in 20 random light microscopy views on each o f lour coverslips per point. Open circles, control cultures; closed circles. CT-treated cultures. "1) < 0 . 0 S ; " p < 0.01 compared t o control cultures.

In cultures containing abundant differentiated osteoclast- like cells, C T markedly down-rcgulated CTR mRNA from 0 h and up to at least 72 h o f the treatment, but had no consistent effect o n TRAP mRNA. Down-regulation o f the ('TR was obvious in the decreased density o f autoradio- graphic silver grains in ostcoclast-like cells after 6 h of CT treatment. One possible mechanism o f CTR down-rcgula- tion in osteoclasts is internalization o f the CT-CTR com- plex. for which we have obtained evidence using electron microscopic ~iutoradiography.".~) Here we suggest that the prolonged down-regulation of C'TK mRNA, and therefore decreased CTR synthesis. is another important component i n CTR down-regulation. We have also reported homolo- gous down-regulation of CTR and CTR mRNA in a differ- ent moclel system"" in which osteoclast-like cells were generated by mouse bone marrow cells o n a type-l collagen

I n those experiments. using cultures contain- ing ;ibund;int multinucleated osteoclast-likc cells, con- ducted under conditions where n o further maturation of precursors was possible. even brief C T treatment produced prolonged down-regulation o f CTR mRNA. In addition, a recent report showed that sCT, albeit at ii high concentra- tion (I00 ngjnil; about 3.2 x 10 ' M), induced down-

regulation of CTR mRNA in human bone marrow and osteoclastoma cultures, as assessed by RT-PCR.(2") Consis- tent with the well known effects of CT on osteoclast mor- phology, we also saw the characteristic CT-induced shrink- age of multinucleated osteoclast-like cells in the bonc marrow cultures. These cells showed recovery of morphol- ogy in terms of cytoplasmic spreading between 12 and 24 h after the start of CT treatment, perhaps suggesting an escape of these cells from the inhibitory effect of CT. This recovery is likely to stem from the reduced CTR and con- sequent loss of sensitivity to CT. However, there is ;I pos- sibility that CT resistance may be due to additional factors as well as down-regulation of the CTR in osteoclasts. Low levels of CTR mRNA were detected in CT-treated cells in both our experimental protocols, implying that some C'TR expression persisted in the presence o f CT. Tashjian et itl.(")

suggested that sufficient CT binding sites for ii biological response to CT remained in cultured bone tissues which had been rendered insensitive to CT by C T treatment. Ransjii et al."7' reported that forskolin and cholera toxin, which are adenylate cyclase activators and increase intra- cellular CAMP. inhibited bone resorption in organ cultures and, as is the case with CT, this effect was also transient. Several such studies suggest that desensitization o f CT- induced intracellular signal transduction is also important. We have also obtained evidence that intraccllular signalling is necessary, rather than simple receptor occupancy, since the 8-32 truncated analogue of sCT had n o effect on CTR mRNA levels (results not shown). This analogue binds to murine CTRs with high affinity but is inactive in transduc- ing intracellular signals."" Further studies focusing on the relationship between signal transduction and CTR mRNA down-regulation are required.

I t was n o t possible t o demonstrate a significant difference in the numbers of TRAP( +) mononucleated and multinu- cleated cells between control and CT-treated cultures in repeated experiments, in CT-treated cultures from day 0. However, the average number of nuclei in multinuclcated cells was significantly less than that in control cultures on each culture day. There have been several reports o f the inhibitory effects o f C T on osteoclast multinucleation in bone marrow cultures derived from ;I number of spe-

I t has been claimed that CT decreases dose- cies~'15.2~l-.32) dependently the formation o f multinucleated cells i n hu- man,I3") baboon,(32' and mouse(15) bone marrow cultures. However, the concentrations o f C T used i n those studies were greater than in the present experiments, and it is likely that the choice of CT concentration for expcrinicnts o f this type has important implications. In fact the development o f "escape" in vitro has been shown to depend both o n the concentration of CT,'7.77' and the agent used to stimulate bonc resorption."' The concentration o f CT we used (0.1 nM) greatly suppressed CTR mRNA levels and caused ii

decreased number of nuclei in multinucleated cells. but was not enough to affect the number o f multinucleated osteo- clast-like cells which dcvcloped. I n bone organ culture it was also shown that "escape" was prevented by irradiation that will inhibit proliferation o f ostcoclast precursors' ")

and that "escape" was accompanied by the appearance o f newly formed, small, active osteocIasts.(l3' The implication

CALCITONIN EFFECTS ON OSTEOCLAST DEVELOPMENT AND CTR mRNA 463

E l I

FIG. 8. Effects of continuous CT treatment from day 6 on CTR and TRAP mRNA levels determined by RT-PCR. (a) Southern blot data showing CTR RT-PCR products (upper), TRAP RT-PCR products (middle), and GAPDH RT-PCR products (lower panel) amplified from control (-) and treated (+) cultures at the times indicated following treatment from day 6. Graphical representations of CTR and TRAP RT-PCR products shown in (a) relative to GAPDH RT-PCR products at the times indicated following sCT treatment are shown in (b) and (c), respectively. Each data point is shown a s a ratio compared with the value of the control culture at 1 h after CT treatment (arbitrarily assigned the value I ) . Open bars, control cultures; closed bars, CT-treated cultures.

o f these experiments was that the newly formed cells had reduced responsiveness to CT, and this is supported by the present data showing development of CTR deficient osteo- clast-like cells in the face of CT treatment.

We have therefore shown that continuous treatment of bone marrow cultures with 0.1 nM sCT causes prolonged down-regulation of CTR in osteoclasts, and formation of CTR-deficient osteoclasts, by reducing CTR mRNA ex- pression. These data provide strong evidence at the cell and molecular level that desensitization of osteoclasts to CT and the development of new osteoclast populations that are unresponsive t o C T are both likely to play a role in the “escape” phenomenon.

We found no clear evidence of C T effects on the number o f TRAP(+) cells or on TRAP mRNA expression. How- ever. we did not measure directly the TRAP activity of individual cclls. Using the more sensitive approach of scan- ning microdensitometry in individual isolated osteoclasts, Yumita ct al.‘.3J) reported that C T reduced TRAP activity after 6 to Y h of C T treatment. Roodman et al.‘”) found a decrease o f TRAP activity in baboon marrow cultures but this was after 3 weeks of high dose (100 ngiml) sCT treat- ment. We attempted t o score the TRAP activity on individ- ual osteoclast-like cclls according to their method.‘3’’

TRAP scores at 24 h after C T trcatnient tended t o be less than in control in two repeated cxpcriments, although the differences were not significant (data not shown).

In summary, the present study clearly shows that long- term treatment with 0.1 nM sCT can dramatically and continuously inhibit CTR mRNA expression in ostcoclast- like cells. However, it had no discernible effect o n TRAP expression and did not greatly affect osteoclast dcvelop- ment. Extrapolating from this bone marrow culture system, the results suggest that long-term administration of C T may result in the formation of CTR down-regulated osteoclast populations which consist of existing osteoclasts and ostc- oclasts formed during the treatment. We suggest that to- gether, these populations may be responsible for cmcrgencc of resistance to calcitonin treatment.

ACKNOWLEDGMENTS

We are grateful to Wayne Holloway, Jason Hodgc (Dept. of Medicine, Geelong Hospital) and Ann Friedhyber (Dept. of Pathology, University 0 1 Melbournc) for assis- tance with electron microscopy. We also would like t o thank Dr. Patrick M. Sexton and Dr. Seiki Wada (St. Vincent’s

464 IKEGAME ET AL.

Institute o f Medical Research) for helpful discussion. This work was supported by a program grant from the National Health and Medical Research Council of Australia. J.M.M. and D.M.F. are Senior Research Fellows of the NH & MRC of Australia. M.R. is an N H & MRC Biomedical Research Scholar.

REFERENCES

I . Nijweide PJ, Burger EH, Feycn JHM 1986 Cells of hone: proliferation, differentiation, and hormonal regulation. Physiol Rev 66855-886.

2. Hammarstrom LE, Hanker JS, Toverud SU 1971 Cellular differences in acid phosphatasc isoenzymes in bone and teeth. Clin Orthop Re1 Res 78151-167.

3. Nicholson GC, Moseley JM, Yates AJP, Martin TJ 1987 Con- trol o f cyclic adenosine 3’,5’-monophosphate production in osteoclasts: calcitonin-induced persistent activation and ho- mologous desensitization of adenylate cyclase. Endocrinology 120: 1902-1908.

4. Warshawsky H, Goltzman D, Rouleau MF, Bergeron JJM 1980 Direct in vivo demonstration by radioautography of spe- cific binding sites for calcitonin in skeletal and renal tissues o f the rat. J Cell Biol 85:682-694.

5. Holtrop ME, Raisz LG, Simmons HA 1974 The effects of parathyroid hormone, colchicine, and calcitonin on the ultra- structure and the activity o f osteoclasts in organ culture. J Cell Biol 60346-355.

6. Raisz LG, Au WYW, Friedman J, Niemann I I067 Thyrocal- citonin and bone resorption. Studies employing a tissue culture bioassay. Am J Med 43:684-690.

7. Wcner JA, Gorton SJ, Raisz LG 1972 Escape from inhibition o f resorption in cultures o f fetal bone treated with calcitonin and parathyroid hormone. Endocrinology 90752-759.

8. Kanehisa J I980 Time course o f “escape” from cnlcitonin- induced inhibition o f motility and resorption of disaggregated osteoclasts. Bone 10125-120.

0. Tashjian AH Jr., Wright DR, Ivey JL. Pont A 1978 Calcitonin binding sites in bone: relationships to biological response and “escape”. Recent Prog Horm Res 34285-334.

10. Findlay DM, Martin TJ 1984 Relationship between internal- ization and calcitonin-induced receptor loss in T47D cells. Endocrinology 11578-83.

I I . Lamp SJ, Findlay DM. Moselcy JM, Martin TJ 1981 Calcitonin induction of a persistent activated state of adenylxte cyclase in human breast cancer cells (T47D). J Biol Chem 256312269- 12274.

12. Michelangeli VP. Findlay DM, Moseley JM, Martin TJ 1983 Mechanisms of calcitonin induction o f prolonged activation o f adenylate cyclase in human cancer cells. J Cyclic Nucleotide Protein Phosphor Res 9129-141.

13. Klaushofer K. HOandner H, Hoffmann 0, Czenvenka E, KOnig U, Koller K, Peterlik M 19x9 Interferon y and calcitonin induce differential changes in cellular kinetics and morphology o f osteoclasts in cultured neonatal mouse calvaria. J Bone Miner Res 4585-606.

14. Krieger NS, Feldman RS, Tashjian AH Jr. 1982 Parathyroid hormone and calcitonin interactions in bone: irradiation- induced inhibition o f escape in vitro. Calcif Tissue Int 34197- 203.

15. Takahashi N, Yamana H, Yoshiki S, Roodman GD, Mundy G R , Jones SJ, Boyde A, Suda T 1988 Osteoclast-like cell

formation and its regulation by osteotropic hormones in niousc hone marrow cultures. Endocrinology 122: 1373-1382.

16. Ikegame M, Rakopoulos M, Zhou H, Houssanii S, Martin TJ. Moseley JM, Findlay DM 1095 C‘alcitonin rcccptor isofornis in mouse and rat osteoclasts. J Bone Miner Rcs 1050-05.

17. Janckila AJ, Li C-Y, Lam K-W, Yam LT 1978 The cytocheni- istry of tartrate-resistant acid phosphatasc. Am J Clin Pathol 7045-55.

18. Chomczynski P, Sacchi N I087 Single-step method o f RNA isolation by acid guanidinium tliioeyanate-phenol-chloroform extraction. Anal Biochem 162:150-159.

19. Cassady AI, King AG, Cross NCP, Hume DA 1993 Isolation and characterization of the genes encoding mouse and humiin type-5 acid phosphatasc. Gene l30201-207.

20. Tso JY, Sun X-H, Kao T-h, Rccce KS. Wu R 10x5 Isolation and characterization o f rat and human glyccraldchyde-3- phosphate dehydrogenase cDNAs: genomic complexity and molecular evolution of the gene. Nucleic Acid Res 13:2485- 2502.

21. Sexton PM, Houssanii S, Hilton JM, O’Keeffe LM. (’enter RJ. Gillespie MT, Dnrcy P, Findlay DM 1903 Identification o f brain isoforms of the rat calcitonin receptor. Mol Endocrinol 7:815-X2 I .

22. Tamura T, Takahashi N. Akntsu T. Snsaki T. Udagawii N . Tanaka S. Suda T 1993 New resorption ;issay with nioiisc osteoclast-like multinuclcated cells formed i n vitro. J Bone Miner Rcs 8:053-960.

23. lkegame M, Ejiri S, Ozawa H 1004 Histochemical and autora- diographic studies on elcatonin internalization and intrx-cllu- lar movement in osteo

24. Wada S. Martin TJ, Findlay DM I905 Homologous regulation of the calcitonin receptor in mouse osteoclnst-like cells and human breast cancer T47D cells. Endocrinology 136:201 I - 2621.

25. Akatsu T, Tamura T, Takahashi N. Udagawn N. Tanaka S, Sasaki T, Yamaguchi A, Nagate N, Sudn T 1002 Preparation and characterization of a mouse osteoclnst-like multinuclcated cell population. J Bone Miner Res 7:1207-1300.

26. Takahashi S, Goldring S, Katz M. Hilsenheck S, Williams R. Roodman G D I 0% Downregulation of calcitonin receptor mRNA expression by calcitonin during human osteoclast-like cel l differentiation. J Clin Invest 95:167-171.

27. Ransjii M, Lerner UH, Hecrsche JNM I988 Calcitonin-like effects o f forskolin and choleratoxin on surface area and nio- tility of isolated rabbit osteoclasts. J Bone Miner Res 3:hl I- 610.

28. Houssami S, Findlay DM. Brady CL. Martin TJ, Epand RM. Moore EE, Murayama E, Tamura T, Orlowski RC. Sexton PM 1004 Divergent structural requirements exist for calcitonin receptor binding specificity and adenylate cyclase activation. Mol Pharmacol 47:798-809.

20. Kurihara N, Civin C, Roodman G D I991 Ohtcotropic factor responsiveness of highly purified populations o f early and late precursors for human multinucleatcd cells expressing the o h -

teoclast phenotype. J Bone Miner Res 6257-261. 30. MacDonald BR. Takahashi N, McManus LM, Holahan J .

Mundy GR, Roodman G D 1987 Formation of multinucleated cells that respond to osteotropic horniones in long term human bone marrow cultures. Endocrinology 1202326-2333.

31. Roodman GD, lbhotson KJ, MacDonald BR. Kuehl TJ. Mundy G R 1985 1.25-dihydroxyvitaniin D3 ciiuses formation of multinucleated cells with several ostcoclast characteristics in cultures of primate marrow. Proc Natl Acnd Sci USA 82:8-713- 8217.

32. Takahashi N, Mundy G R , Kuehl TJ. Roodman G D I087 O s - teoclast-like cell formation in fetal and newborn long-term

‘ts. J Bone Miner Rcs 9 9 - 3 7 .

CALCITONIN EFFECTS ON OSTEOCLAST DEVELOPMENT AND CTR mRNA 465

luhoon n1;irrow culture\ is more sensitive t o I ,25-dihydroxy- vitaiiiin I ) , than ;idtilt long-term marrow cultures. J Bone Miner Kcs 2 3 1 1-317.

33. I lccrschc JNM I002 <’;ilcitonin cflccts on osleoclnstic rcsorp- tion: thc ‘csc;ipe phcnonicnon’ revisited. Bonc Miner 16: 174- 177.

34. Yumita S. Nicholson M‘. Rowc DJ. Kent GN, Martin TJ 1901 13iph;isic cflcct of calcitonin on tartrate-resistant acid phos- phatasc activity i n isolated rat ostcoclasts. J Hone Miner Rcs 6:50 1-507.

Address reprint requests to: Dr. I). iM. Fiti(1lu.y

St. Vit i coii t Ys l i ist itir tc, of McdicuI licscurch 41 Victoria f’urudc

I.i’tzroy. .YO65 Victoria A iistruliu

Rcccivcd in original form Junc 12, 1095; i n revised lorm Novcnihcr 10. 1995; accepted Deccmhcr 4. IWS.