0013-7227/9l/1283-1638$03.
00/0
Endocrinology
Copyright 1991 by The Endocrine Society
Cyclosporin-A in Vitro Decreases Bone Resorption,
Osteoclast Formation, and the Fusion of Cells
of the Monocyte-Macrophage Lineage
PHILIPPE ORCEL, MARY ANNICK DENNE,
AND MARIE CHRISTINE DE VERNEJOUL
INSERM Unite 18, F 75010 Paris, France
ABSTRACT. We studied the in vitro effect of cyclosporin-A
(CyA) on bone resorption using a fetal rat long bone-resorbing
assay. CyA inhibited both PTH-stimulated and unstimulated
bone resorption. The inhibitory effect of CyA on basal resorption
was dose dependent, and it was more pronounced during the
second period (<0.1 Mg/ml) of culture (days 5-7) than during
the first period (days 2-4). A cytotoxic effect was ruled out by
the absence of decrease in [3H]thymidine incorporation into
bones up to a concentration of 5 ng/m\ CyA. Histomorphometry
performed after 4 and 7 days of culture showed that CyA (1 jig/
ml) decreased the number of osteoclasts per bone section after
7 days of culture (23.5 4.0 vs. 41.7 2.9 osteoclasts/bone
section; P < 0.05), but not after 4 days (25.6 3.3 vs. 23.0
2.5). These data suggested an effect of CyA on osteoclastic
differentiation rather than on the function of mature osteoclasts.
We further assessed the mechanisms of the inhibitory effect
of CyA on osteoclastic differentiation in order to determine 1)
the level of this action (proliferation and/or fusion of osteoclast
precursors), and 2) if this action is direct or indirect. Autoradi-
ographic studies were performed on bone sections after incuba-
tion of bones with [:iH]thymidine for the last 48 h of culture.
CyA decreased slightly but significantly the percentage of labeled
B ONE resorption is physiologically achieved by
multinucleated osteoclasts. The mechanisms by
which osteoclasts are recruited and activated remain
poorly understood as yet. However, experimental evi-
dence suggests that cells of the immune system, present
close to the bone surface, play a role in the local regula-
tion of bone resorption (1). Several cytokines, secreted
by lymphocytes or monocytes, have been demonstrated
infuence the differentiation and the activity of osteo-
clasts (2-5).
Cyclosporin-A (CyA). an undecapeptide of fungal ori-
gin, is widely used in the prevention of graft rejection
after organ transplantation (6). CyA exerts its immuno-
suppressive effect by selectively interfering with the
function of T-lymphocytes and with the lymphokine-
Received September 4,1990.
Address all correspondence and requests for reprints to: Dr. Philippe
Orcel, INSERM Unite, 6 rue Guy Patin, F 75010 Paris, France.
1638
Vol. 128, No. 3
Printed in U.S.A.
nuclei per osteoclast and the number of osteoclasts containing
at least one labeled nucleus (20.2 0.7 us. 33.2 3.5; P < 0.02).
Moreover the number of nuclei per osteoclast was decreased
after 7 days in CyA-treated bones (2.4 0.05 vs. 3.0 0.1; P <
0.02). Taken together these results demonstrate that CyA
slightly decreased the proliferation of osteoclast precursors, but
markedly decreased their fusion.
Similar effects were observed in cultures of rat marrow mac-
rophages. CyA (1 Mg/ml) inhibited the fusion of macrophages
into multinucleated cells elicited by 1 nM 1,25-dihydroxyvitamin
D3, but had only a slight effect on the proliferation of these cells,
as assessed by autoradiography. CyA also inhibited the forma-
tion of multinucleated cells and the fusion index in long term
cultures of human cord blood monocytes, a cellular model for
osteoclastic differentiation. By contrast, CyA had no effect on
the formation of myotubes by fusion of cultured mononucleated
rat myoblasts.
Our results demonstrate that the inhibitory effect of CyA on
unstimulated bone resorption in vitro is related to a direct
negative action on the formation of osteoclasts, predominantly
by decreasing the fusion of osteoclast precursors. (Endocrinology
128: 1638-1646, 1991)
monokine cascade (7, 8). CyA has been shown to dra-
matically inhibit bone resorption elicited by calcemic
hormones (PTH and 1,25-dihydroxyvitamin D3 [1,25-
(OH)2D3]) or cytokines (interleukin-1, osteoclast-acti-
vating factor, and prostaglandin E2) in fetal rat long bone
and neonatal mouse calvaria culture system (9, 10).
However, neither research group, using short periods of
observation (48-96 h), showed any effect of CyA in
control cultures that were not stimulated by bone-re-
sorbing agents. On the other hand, we recently reported
that CyA, given orally to normal growing rats, inhibited
bone resorption, as assessed by biochemical markers and
histomorphometric osteoclastic parameters on vertebral
trabecular bone (11). These data remain controversial,
since Movsowitz and co-workers latter found inverse
results, an increase in histological parameters of bone
resorption in appendicular bone (12, 13). These discrep-
ancies may result from different compositions of tibial
and vertebral marrow as a source of osteoclasts (14, 15),
 CyA DECREASES OSTEOCLAST FORMATION IN VITRO
thus explaning a distinct action of CyA at each site.
From all of these results the mechanism of action of CyA
on osteoclast activity and/or formation is unclear.
We thus decided to evaluate the effect of CyA on
unstimulated osteoclastic bone resorption using the fetal
rat long bone culture system. We found that CyA de-
creased resorption during the late period of culture (4-7
days) and decreased the recruitment and the fusion or
multinucleation of osteoclasts. Conversely, CyA inhib-
ited the multinucleation of rat marrow macrophages and
the formation of multinucleated cells in human cord
blood monocyte cultures, but did not influence the mul-
tinucleation of cultured rat myoblasts. Taken together
our results suggest that the inhibitory action of CyA on
bone resorption could result from a direct effect on the
fusion of osteoclast precursors, and that this antifusion
effect would be specific for cells of the monocyte-mac-
rophage lineage.
Materials and Methods
Culture treatments
CyA (1 Mg/ml stock solution in ethanol) was a generous gift
from Laboratoires Sandoz (Rueil Malmaison, France). Final
concentrations of CyA in treated cultures were 0.1, 0.5, 1, 2.5,
and 5 Mg/ml. PTH (synthetic human fragment 1-34), purchased
from Peninsula Laboratories (Belmont, CA), was diluted in
0.05 M acetic acid and added to the cultures at a concentration
of 400 ng/ml. 1,25-(OH)2D3 was a gift from Dr. Husic (Hoff-
man-LaRoche, Basel, Switzerland). 1,25-(OH)2D3 was dissolved
in ethanol and added to the cultures after further dilution in
the culture medium. The final ethanol concentration in the
cultures was less than 0.10%.
Bone cultures
Fetal rat long bone cultures were performed as previously
reported in detail by Stern et al. (16) with some adaptations.
Briefly, bones were labeled with 45Ca (calcium chloride in
aqueous solution; 2.27 mCi/ml; Amersham International, Ay-
lesbury, United Kingdom) by injecting 0.5 mCi to the pregnant
mothers 1 day before removal of the fetuses. The shafts of radii
and ulnae of 19-day-old fetal rats were dissected in PBS, pH
7.4, with removal of cartilage and fibrous tissue. Bones were
then precultured for 24 h in serum-free Biggers-Gwatkin-Judah
(BGJ) medium (Gibco, Grand Island, NY) to remove exchange-
able 45Ca. After preculture the bones were transferred into
stationary culture plates (Multiwell, Falcon, Oxnard, CA) con-
taining BGJ medium supplemented with 1 mg/ml BSA (grade
V; Sigma Chemical Co., St. Louis, MO). Bones were then
cultured for 3 days (days 2-4) in fresh medium, and after a
second medium change the incubation was continued for an-
other 3 days (days 5-7). Bones were cultured in the presence
or absence of CyA, PTH, or both or with the vehicle alone.
Media were collected on days 4 and 7. Bones were removed
from the culture after day 7, placed in 0.5% formic acid over-
night at room temperature. Aliquots of media and formic acid
extracts were diluted in 4 ml scintillant, and 45Ca was counted
1639
in a /3-liquid scintillation counter. The mobilization of 45Ca
from bone was expressed as the percentage of total 45Ca that
was released into the medium for each period. The total 45Ca
was the amount of 45Ca in the medium harvested after each
culture period and of the 45Ca from formic acid extracts.
DNA synthesis
DNA synthesis was assessed by the incorporation of [3H]
thymidine into the acid-insoluble fraction of the bones, as
described by Lorenzo et al. (17) with modifications described
by Pfeilschifter et al. (18). After 4 or 7 days in culture, bones
were pulsed for 2 h with 1 /iCi [6-3H]thymidine (SA, 25 Ci/mM;
CEA, Gif-sur-Yvette, France). Bones were then washed in PBS,
blotted, placed in 200 ml 5% trichloroacetic acid at 4 C for 1 h,
rinsed in ethanol, air dried, and dissolved in 1 ml Nuclear-
Chicago solvent (NCS, Amersham) tissue solubilizer overnight
at 50 C. Trichloroacetic acid extracts and NCS digests were
counted for 3H in 4 ml scintillant in a /3-liquid scintillation
counter. The background was measured by the incorporation
of 3H into freeze-thaw killed bones.
Histology and autoradiography
For all histological studies bones were fixed in cold ethanol
and embedded undecalcified in glycol-methacrylate. The blocks
were cut longitudinally, and three 5-mm sections located, ex-
actly at the midregion of the radii were collected.
For histomorphometry 1-, 4-, and 7-day cultured bones were
processed as described above. Sections were stained for tartrate-
resistant acid phosphatase (TRAP) using naphtol ASTR phos-
phate (Sigma) as a substrate. They were not counterstained
because of the weak purple staining after acid phosphatase
histochemistry. Quantitative measurement of the total number
of TRAP-stained cells per bone section was performed in the
blind on coded sections, but the number of nuclei per cell was
not countable in these bones.
For autoradiography bones were incubated with 1 juCi/ml
[3H]thymidine for the last 48 h of culture. After sectioning, the
slides were dipped in NTB-2 photographic emulsion (Eastman
Kodak Co., Rochester, NY) and developed for 14 days. Slides
were counterstained with toluidine blue and random ordered.
Osteoclasts were identified by their multinucleation (two or
more nuclei), proximity to bone surface, and characteristic
cytoplasm (19). Labeled nuclei were defined as having five or
more overlying silver grains.
Cell cultures
Cultures of rat marrow macrophages. Rat marrow macrophages
were isolated and cultured according to the method of Lee and
Wong (20) with some modifications. Long bones of 2-month-
old rats were dissected out and flushed in Minimum Essential
Medium (MEM). Bone marrow cells were washed in MEM and
plated at a density of 106 cells/well (2 X 106/ml or 2.5 X 106/
cm2) in plastic eight-well chamber slides (Lab Tek, Miles,
Naperville, IL). After an overnight incubation the medium was
changed, and nonadherent cells were removed. Cells were fur-
ther cultured for 6 days, in the presence or absence of CyA (1
jug/ml), in MEM supplemented with 100 U/ml penicillin, 100
jug/ml streptomycin, 2 mM glutamine, 4 mM HEPES buffer,
1640 CyA DECREASES OSTEOCLAST FORMATION IN VITRO
20% heat-inactivated horse serum, and 1 nM 1,25-(OH)2D3 in
a humidified atmosphere of 5 CO2-air at 37 C. After 6 days cells
were fixed in cold ethanol and stained for TRAP (substrate-
naphtol AS-TR phosphate; tartrate, 50 mM) or for alkaline
phosphatase (substrate-AB-BI phosphate; Sigma). The slides
were then counterstained with toluidine blue. Quantitative
measurements were performed on coded sections. All cells of a
central low power field were counted, and cells containing three
or more nuclei were scored as multinucleated cells. In one
culture the cells were pulsed with 1 jtCi/ml [3H]thymidine for
the first 24 h of culture. After 6 days the slides were processed
for autoradiography, as described above for bone sections.
Slides were developed for 7 days, counterstained with toluidine
blue, and randomly ordered. The same quantitative measure-
ments were performed, and labeled nuclei were counted and
defined as having five or more overlaying silver grains.
Cultures of human cord blood monocytes. Human cord blood
monocytes were isolated and cultured as previously reported
(21). Briefly, mononuclear leukocyte suspensions were isolated
from umbilical cord blood by density gradient centrifugation,
washed in MEM, and plated at a density of 2.5 x 106 cells/cm2
in eight-well plastic chamber slides. After an overnight incu-
bation nonadherent cells were removed by three repeated
washes, and the medium was changed. Cord monocytes were
further cultured for 3 weeks in MEM supplemented with 20%
heat-inactivated horse serum and 1 nM 1,25-(OH)2D3, in the
presence or absence of CyA, in a humidified atmosphere of 5%
CO2-air at 37 C. After 21 days the cells were washed, ethanol
fixed, and stained with toluidine blue. Quantitative measure-
ments were performed to determine the number of multinucle-
ated cells as a percentage of total cells and the fusion index,
defined by the percentage of the total number of nuclei within
multinucleated cells (22).
Cultures of rat skeletal myoblasts. Myoblast cultures were pre-
pared according to the method of O'Neill and Stockdale (23)
with some modifications. Briefly, limb muscles of 2-day-old
rats were dissected out in Ca2+-Mg2+-free buffered saline,
minced, and digested with 0.25% trypsin (Eurobio, Paris,
France) in Dulbecco's Modified Eagle's Medium for 30 min at
37 C. Cells were dispersed by repeated pipetting through a glass
pipette, and a second digestion was performed in the same
manner. After repeated pipetting the cells were collected and
suspended in Dulbecco's Modified Eagle's Medium supple-
mented with 15% fetal calf serum (Flow Laboratories, Rock-
ville, MD). The cell suspension was then filtered through silk
to remove debris and undissociated cells. Cells were collected
by centrifugation, suspended in culture medium, counted, and
plated at a density of 5 x 104 cells/ml in Falcon 35-mm tissue
culture dishes (not collagen coated). Medium was changed after
24 h and then every 3 days. Cultures were placed in a humidified
atmosphere of 5 CO2-air at 37 C. CyA (1, 5, or 10 Mg/ml) was
added in some cultures 24 h after plating. After 7 days, cultured
cells were washed three times in PBS, fixed in 2% glutaralde-
hyde at room temperature, washed in PBS, stained with Groat
hematoxylin, and rinsed thoroughly with tap water. The nuclei
were counted in four randomly determined high power fields
for each culture, and the fusion index was calculated as the
Endo 1991
Voll28No3
ratio of the number of nuclei within myotubes containing three
or more nuclei to the total number of nuclei.
Statistics
Data are presented as the mean SE. Statistical significance
was determined by unpaired Student's t test.
Results
We first examined the effect of CyA on PTH-stimu-
lated and unstimulated bone resorption in fetal rat long
bone cultures in order to confirm the findings of other
researchers (9,10) in our modified organ culture system.
As shown in Fig. 1, PTH enhanced bone resorption
during both the first and the second halves of the 7-day
incubation period, and CyA inhibited this effect of PTH.
Moreover, in these experiments 1 /ug/ml CyA decreased
basal bone resorption during the second half of the
incubation period, but no effect was observed during the
first period. However, in another experiment a slight but
significant inhibitory effect of CyA was observed in the
first period for concentrations of 1 ixg/ml or more (Fig.
2). During the second period, concentrations as low as
0.1 fig/ml CyA decreased basal resorption, and the effect
of CyA was dose dependent (Fig. 2).
We then assessed the effect of CyA on DNA synthesis
by the incorporation of [3H]thymidine into cultured
bones. After 4 days in culture, [3H]thymidine incorpo-
ration was 826 150 cpm/bone in control bones, 656
82 in 1 jug/ml CyA-treated bones, and 46 5 in killed
bones. The difference between control and CyA-treated
bones was not significant. Figure 3 shows that 0.5 or 1
CyA CyA
CyA CyA
FIG. 1. Effect of CyA on PTH-stimulated and unstimulated bone
resorption in fetal rat long bone culture. 45Ca-prelabeled fetal rat long
bones were cultured as described in Materials and Methods. The effect
of 400 ng/ml PTH (PTH bars), 1 fig/ml CyA (CyA bars), or both
(PTH+CyA bars) was assessed by the release of 45Ca into the culture
medium and expressed as the percentage of the initial radioactivity
released into the medium after 4 and 7 days in culture. C bars are
control cultures. Each bar represents the mean SEM of four cultures.
Stars denote statistical significance at P < 0.001 vs. control bones (*)
or P < 0.05 vs. PTH-treated bones (*).
 CyA DECREASES OSTEOCLAST FORMATION IN VITRO 1641

itrols) 110- TABLE 1. Effect of CyA on the number of TRAP-positive cells in fetal
rat long bone culture
5

I
100-

90-
k
k Treatment
Time in
culture
TRAP-positive
cells
(days) (no./bone section)
released

80.
Njr ... r" Vehicle 1 16.4 2.6

| 70- ^ i ^ * ^ T ... Vehicle 4 23.0 2.5

60.
r^* CyA (1 Mg/ml) 4 25.6 3.3

Vehicle 7 41.7 2.9

oX.
0.5 2.5 5
CyA
FIG. 2. Dose effect of CyA on unstimulated bone resorption in fetal
rat long bone culture. 45Ca-prelabeled fetal rat long bones were cultured
as described in Materials and Methods. The effect of CyA was assessed
by the release of 45Ca into the culture medium and expressed as the
percentage of the initial radioactivity released into the medium for
each period of culture. Circles represent the mean SEM of four cultures
assessed after the early period (2-4 days). Triangles represent the mean
SEM of four cultures assessed after the late period (5-7 days). Stars
denote statistical significance us. control bones, at P < 0.05 {one star),
P < 0.02 {two stars), or P < 0.01 {three stars).
T
3000-
2000-
X, T
1000-
X precursors. Autoradiographs of bones cultured with [3H]
contrail killed
bonei
CyA. ng/ml
FIG. 3. Effect of CyA on the incorporation of [3H]thymidine into fetal
rat long bone cultured for 7 days. Unlabeled fetal rat long bones were
cultured, as described in Materials and Methods, in the absence or
presence of CyA (0.5-5 Mg/ml). Bones were pulsed with 1 juCi [3H]
thymidine for the last 2 h of culture, and the radioactivity was counted
in dissolved bones. Each bar represents the mean SEM of four bones.
One star denotes a statistical significance at P < 0.01 us. control bones.
jug/ml CyA did not affect [3H]thymidine incorporation
after 7 days in culture, whereas a higher concentration
(5 Mg/ml) decreased up to 2-fold the incorporation. These
results suggest that low or mild concentrations of CyA
were not cytotoxic to bone cells and their precursors, and
that such doses did not strongly inhibit the proliferation.
CyA (1 Mg/ml) significantly decreased the number of
TRAP-positive osteoclasts per bone section (Table 1).
This effect was observed only after 7 days of culture,
whereas in bones treated for 4 days no difference was
noted in the total number of TRAP-positive osteoclasts
per bone section (Table 1). These data suggest that CyA
affected osteoclast generation rather than the activity of
CyA (1 Mg/ml) 7 23.5 4.0
Fetal rat long bones were cultured in the absence or presence of 1
Mg/ml CyA, as described in Materials and Methods, for 1 day (n = 4),
4 days (n = 7), or 7 days (n = 7). After their respective culture period,
bones were processed for histomorphometry. Histochemistry for TRAP
was performed, and all TRAP-positive cells in endosteal and periosteal
bone surfaces were counted.
0
Significantly different from control bones, P < 0.05.
preexisting mature osteoclasts.
The formation of osteoclasts is a multistep process,
including proliferation, fusion, and differentiation proc-
esses. Our further purpose was, thus, to study the effect
of CyA on the proliferation and fusion of osteoclast
precursors, using autoradiographic studies and histolog-
ical measurements of osteoclastic parameters in cultured
feral rat long bones. The number of nuclei per osteoclast
was significantly decreased in CyA-treated bones (Table
2), suggesting an effect of CyA on the fusion of osteoclast
thymidine for the last 48 h (Table 2) showed a decreased
number of labeled nuclei and a smaller percentage of
osteoclasts that contained at least one labeled nucleus in
CyA-treated cultures. In the same series of experiments
we observed that in 7-interferon-treated bones none of
the nuclei present in the osteoclasts were labeled (data
not shown). Since 7-interferon is a potent inhibitor of
the proliferation of osteoclast precursors (24) and of bone
resorption in tissue culture (25), these data demonstrate
a weak antiproliferative effect of CyA on osteoclast pre-
cursors, in contrast to a predominant antifusion effect.
Osteoclasts are believed to derive from a precursor cell
of the monocyte-macrophage lineage. We thus decided
to evaluate the effect of CyA on these cells. For this
purpose we used two different models of monocyte-mac-
rophage cultures. First, we isolated bone marrow mac-
rophages from rat long bone and cultured them on plastic
chamber slides. After 2 days of culture, more than 90%
of the cells were stained with acid phosphatase, but this
staining was completely abolished when 50 mM tartrate
was added during the processing for acid phosphatase
cytochemistry. No alkaline phosphatase cell was ob-
served in these culture. This shows that marrow cells
cultured in these experiments were almost all macro-
phages. In these cultures multinucleated cells were ob-
1642 CyA DECREASES OSTEOCLAST FORMATION IN VITRO Endo'1991
Voll28No3

TABLE 2. Effect of CyA on the multinucleation and the incorporation of recently replicated nuclei into osteoclasts

Osteoclasts Nuclei
Treatment (no./bone section) (no./osteoclast)
Vehicle 20.5 1.7 3.0 0.1
CyA (1 Mg/ml) 12.4 1.3" 2.4 0.05

% Labeled nuclei/ Osteoclasts

labeled containing >1
osteoclast labeled nucleus

1.10 0.34 33.2 3.5

0.41 0.62" 20.2 0.7
Fetal rat long bones were cultured in the presence (n = 4) or absence (n = 4) of 1 ng/ml CyA, as described in Materials and Methods, for 7 days.
Bones were incubated with [3H]thymidine for the last 48 h of culture. Bone sections were processed for autoradiography and then stained with
toluidine blue. Osteoclast (multinucleated-facing-bone-cells), unlabeled nuclei, and labeled nuclei (>5 overlaying silver grains) were counted.
0
Significantly different from vehicle-treated bones, P < 0.02.

served on and after the third day of culture, and their
number rose with time, until days 5-6. Multinucleated
cells formed in these cultures by fusion of mononuclear
macrophages, and the number of nuclei per multinucle-
ated cell increased with time in culture (10 1 on day 4,
42 3 on day 6, and 78 5 on day 8). The process of
fusion was markedly enhanced by 1,25-(OH)2D3; the
number of nuclei per multinucleated cell was 25 1 in
the absence of 1,25-(OH)2D3 and 39 1 in the presence
of 1 nM 1,25-(OH)2D3 (P < 0.001). In the presence of 1
nM 1,25-(OH)2D3, CyA (1 /*g/ml) decreased the fusion of
mononuclear macrophage (Fig. 4). The number of multi-
nucleated cells formed after 5 days of culture was un-
changed, but the number of nuclei per multinucleated
cell was significantly decreased (Table 3). In one exper-
iment autoradiography was performed after 5 days of
culture of macrophages pulsed with [3H]thymidine for
controls
\ rise to multinucleated myotubes. The fusion index (0.096
1 nM 1,25-(OH)2D3 < controls) and the number of nuclei per myotube (6.91
lOnM 1,25-(OH)2D3
1 nM 1,25-(OH)2D3
+ lug/ml CyA H ft
10 20 30 40 SO
number of nuclei per MNC
FIG. 4. Effects of 1,25-(OH)2D3 and CyA on the multinucleation of
cultured rat marrow macrophages. Rat marrow macrophages were
isolated and cultured, as described in Materials and Methods, in the
presence of 1,25-(OH)2D3 and or CyA (1 Mg/ml). The number of nuclei
per multinucleated cell (MNC) was counted in six cultures for each
treatment (bars represent the mean SEM). Stars denote statistical
significance vs. control cultures (*) at P < 0.02 or vs. 1,25-(OH)2D3
alone cultures (*) at P < 0.01.
the first 24 h of culture. CyA induced a slight but signif-
icant decrease in the percentage of labeled nuclei counted
in the multinucleated cells (Table 3), indicating an in-
hibitory action on the early proliferation of mononuclear
macrophages.
We then tested the effect of CyA on cultures of human
cord blood monocytes. We have previously shown that
these cells give rise to multinucleated cells in long term
culture in the presence of 1,25-(OH)2D3, some of them
expressing characteristic features as osteoclast precur-
sors (21). In these cultures CyA at a concentration as
low as 0.1 ixg/m\ inhibited the formation of multinucle-
ated cells (16.1 1.2% of total cells vs. 23.2 2.9% in
control cultures; P < 0.05; Fig. 5) in the presence of 1
nM 1,25-(OH)2D3. Moreover, CyA decreased the fusion
of mononucleated precursors, as assessed by the fusion
index (37.7 2.2% vs. 51.7 5.7%; P < 0.05).
Finally, we assessed the effect of CyA on rat myoblast
cultures. After 7 days of culture, these myoblasts gave
0.007 in CyA-treated cultures vs. 0.097 0.005 in
0.47 vs. 7.16 0.22) were not different in CyA (1 ng/m\)-
treated cultures compared to control cultures. In one
experiment we tested higher concentrations of CyA. A
slight but not significant decrease in both the fusion
index and the number of nuclei per myotube was noted
when myoblasts were cultured in the presence of 10 ng/
ml (0.076 0.003 vs. 0.098 0.010 and 5.85 0.69 vs.
7.07 0.50, respectively).
Discussion
We report in this paper that CyA decreases bone
resorption in unstimulated fetal rat limb bone culture.
This effect was the result of an inhibitory action on
osteoclast formation, rather than on the activity of preex-
isting osteoclasts. Moreover, the results obtained from
 CyA DECREASES OSTEOCLAST FORMATION IN VITRO 1643

TABLE 3. Effect of CyA on the formation of multinucleated cells in cultures of rat marrow macrophages

Exp No. of No. of % of labeled

no. MNC nuclei/MNC nuclei
1 Vehicle 71 6 26 3
CyA (1 Mg/ml) 73 9 31

2 Vehicle 83 5 35 3 53 2
CyA (1 Mg/ml) 113 8b 21 3 41 2
Rat marrow macrophages were isolated and cultured, as described in Materials and Methods, in the presence of 20% horse serum, 1 nM 1,25-
(OH)2D3 and, in some cultures, 1 tig/nA CyA. After 6 days of culture the number of multinucleated cells (MNC), and the number of nuclei per
MNC were counted in six cultures for each treatment. In the second experiment the cells were pulsed with [3H]thymidine for the first 24 h, and
the slides were processed for autoradiography after 6 days of culture, allowing calculation of the percentage of labeled nuclei (>5 overlaying silver
grains) into MNC.
Significantly different from vehicle-treated cells, P < 0.01.
6
Significantly different from vehicle-treated cells, P < 0.05.

blasts, suggesting that its effect on cellular fusion is

10
CyA, no/ml
FIG. 5. Effect of CyA on the number of multinucleated cells (A) and
the fusion index (B) in long term cultured human cord blood monocytes.
Cord blood monocytes were isolated and cultured for 3 weeks, as
described in Materials and Methods, in the presence or absence of CyA
(0.1 and 1 fig/ml). The cells and nuclei were counted. The percentage
of multinucleated cells (MNC) and the percentage of the total number
of nuclei that are in MNC (fusion index) were calculated. Bars represent
the mean SEM of four cultures for each treatment. Stars denote
statistical significance at P < 0.05 vs. control cells.
the [3H]thymidine incorporation assay and autoradi-
ographic and histomorphometric studies suggest that
CyA had a slight inhibitory effect on the proliferation of
osteoclast precursors and strongly decreased the fusion
of these precursors, resulting in a decreased osteoclast
number. Cultures of two different cell types of the mon-
ocyte/macrophage lineage allowed us to demonstrate
that CyA directly affects the fusion of these cells. In
contrast, CyA did not influence the fusion of rat myo-
restricted to cells of the monocyte-macrophage lineage.
We confirmed the data previously reported by Stewart
et al. (9) and Klaushofer et al. (10) showing an in vitro
inhibitory effect of CyA on bone resorption elicited by
PTH. We also found, as described in these two studies,
that CyA did not or only minimally influenced unstim-
ulated bone resorption after a short exposure to CyA
(24-96 h). However, we observed that CyA inhibited
markedly and in a dose-dependent manner the basal
unstimulated bone resorption during the second half of
the culture period (96-168 h). A direct inhibitory or
cytotoxic effect of CyA on mature osteoclasts cannot be
ruled out, but our results do not suggest such an action.
Indeed, DNA synthesis, assessed by [3H]thymidine in-
corporation into bones, was not dramatically affected by
CyA. The number of osteoclasts progressively rose in
untreated cultures, although this increase was abolished
in CyA-treated bone during the late period of culture.
These data strongly suggest that the effect of CyA on
bone resorption is, rather, due to a decrease in the
formation of new osteoclasts during the culture. Similar
findings have been reported with transforming growth
factor-/? (18), which inhibits bone resorption in feral rat
long bone cultures only during the second period of
culture (days 4-6). Interestingly, transforming growth
factor-/? has been recently shown to decrease the prolif-
eration and fusion of osteoclasts precursors in two
models of osteoclast-like cell formation (5, 26).
A slight effect of CyA on DNA synthesis could result,
if not from cytotoxicity, from an action on bone cell
proliferation. The absence of cytotoxicity observed by
Stewart et al. (9) and Klaushofer et al. (10) with CyA
concentrations as high as 10 and 30 Mg/ml in short term
(48-h) cultures suggest that the effect we observed after
prolonged exposure to CyA (7 days) might be attributed
to a slight inhibition of the proliferation of bone cells.
An inhibition of the proliferation of osteoclast precursors
is demonstrated by our autoradiographic findings, show-
1644 CyA DECREASES OSTEOCLAST FORMATION IN VITRO
ing a decrease in the number of osteoclasts containing at
least one labeled nucleus in CyA-treated bones. More-
over, fewer [3H]thymidine-labeled nuclei were counted
in osteoclasts from CyA-treated bones. However, in the
same culture model we observed that 7-interferon, a
potent inhibitor of both osteoclast precursor prolifera-
tion (24) and bone resorption (25), completely abolished
cell replication, so that no labeled nuclei could be ob-
served in the osteoclasts at the end of the culture (data
not shown). The decrease in the number of osteoclasts
and in the number of nuclei per osteoclast in CyA-treated
bone can, therefore, be interpreted as the result of an
inhibitory effect of CyA predominating on the fusion of
the osteoclastic precursors.
Osteoclasts are believed to originate from early pre-
cursors located in the hemopoietic bone marrow and
belonging to the monocyte macrophage lineage (27, 28).
Multinucleated osteoclasts derive from these precursors
through a complex multistep process that comprises,
successively, proliferation and fusion of precursor cells
(28, 29). In the past few years several attempts have been
made to study osteoclastic differentiation in vitro, and
cellular models have been developed. Long term cultures
of nonadherent bone marrow cells have been shown to
give rise to multinucleated cells with several morpholog-
ical and some fonctional characteristics of osteoclasts
(review in Ref. 29). In our hands the culture of adherent
cells derived from rat bone marrow formed multinuclea-
ted cells containing the tartrate-labile isoenzyme of acid
phosphatase. Such a phenotype is characteristic of mac-
rophages, but not of osteoclast precursors (28). In these
cultures 1,25-(OH)2D3 stimulated the time-dependent
increase in the formation of multinucleated cells by
increasing the fusion of macrophages. This effect of 1,25-
(OH)2D3 on the fusion of cells of the monocyte family
has been reported by several researchers in many other
culture systems (30,31). We observed that CyA inhibited,
in the presence of 1 nM 1,25-(OH)2D3, the fusion of rat
marrow macrophages. Autoradiographic study performed
in one culture showed only a slight decrease in the
percentage of labeled nuclei per multinucleated cell in
CyA-treated cells, suggesting that the effect of CyA was
more marked on the fusion than on the proliferation of
macrophages (32). Taken together these results further
demonstrate that CyA has a direct effect on the fusion
of marrow macrophages, since the macrophage cultures
were more than 95% pure.
Long term cultures of human cord blood monocytes
allow the development of multinucleated cells with some
phenotypic characteristics of osteoclast precursors: ap-
propriate reponse to calciotropic hormones, presence of
TRAP, and expression of the vitronectin receptor (21).
We have previously observed that the fusion of cord
blood monocytes increases with the time in culture and
Endo1991
Voll28No3
in the presence of 1,25-(OH)2D3 (21). In this system CyA
decreased the formation of multinucleated cells and the
fusion index. Again, the effect of CyA on these cells is
likely to be a direct one.
A direct action of CyA on cells of the monocyte lineage
is controversial. The main target cell of this potent
immunosuppressant is the T-lymphocyte. CyA has been
shown to inhibit the function of T-lymphocytes and the
secretion of several lymphokines (8, 33). Thus, the sup-
pressive action of CyA on monocyte-macrophage and on
the production of monokines (among which is interleu-
kin-1, a potent resorbing agent in vitro and in vivo) is
commonly thought to be lymphocyte mediated (34). In-
deed, Stewart and Stern (35) have observed that the
actions of different cyclosporine analogs with different
immunosuppressive potencies on PTH- or interleukin-
1-stimulated bone resorption are tightly correlated with
their activities as immunosuppressive agents. Moreover,
CyA seems to have only a small direct effect, if any, on
the main immunological functions of monocyte-macro-
phages (6, 7). It is, therefore, possible to assume that the
inhibition of bone resorption and osteoclast formation
by CyA is mediated by its effects on the production or
action of cytokines on bone.
However, on the basis of our observations of the action
of CyA on rat marrow macrophages and on cord blood
monocytes, a possible direct effect on these cells needs
to be considered. In this regard some recently reported
data are of importance. Benson and Ziegler (36) reported
that pretreatment of macrophages with CyA could inhibit
antigen presentation and macrophage-mediated antigen-
specific T-cell activation (interleukin-2 production) in
macrophage-T-cell cocultures. In these experiments CyA
demonstrated saturable binding to macrophages, sug-
gesting the existence of CyA-binding sites (36). The
effects of CyA on cellular membranes are complex. CyA
has been shown to alter membrane fluidity (37), to cause
membrane depolarization (38), and to modulate mem-
brane calcium fluxes (39-41). In peritoneal rat macro-
phages, CyA could interfere with membrane arachidon-
ate metabolism, stimulation of phospholipase activity,
and partial blockage of thromboxane synthetase (42). All
or some of these actions of CyA on the cellular membrane
might be relevant in the process of fusion of monocyte-
macrophages. CyA might modulate the synthesis of spe-
cific proteins implicated in the fusion of these cells.
Indeed, Vignery (43) reported very recently that macro-
phage multinucleation is accompanied by the expression
of new soluble and membrane antigen in mice.
By contrast to the negative effect on the fusion of cells
of the monocyte-macrophage lineage, we observed that
CyA did not influence the fusion of rat myoblasts. In
skeletal muscles the mature muscle fibers, or myotubes,
are multinucleated and form by fusion of mononucleated
 CyA DECREASES OSTEOCLAST FORMATION IN VITRO
myoblasts. The fusion of myoblasts involves surface ele-
ments, such as proteins, and fibronectin has been shown
to enhance this process of fusion (44). Law et al. also
found that cloned myoblasts survived and fused at CyA
concentrations of 25 jug/ml or less. It is, therefore, pos-
sible that CyA induces a specific membrane modification
in monocytes, leading to decreasing fusion, although a
different process may be involved in myoblast fusion.
In the light of these data, our results suggest that CyA
inhibits by a direct mechanism the fusion of cells of the
monocyte lineage and latter inhibits the fusion of osteo-
clast precursors. This could lead to a decreased differ-
entiation of mature bone-resorbing cells and, thus, ex-
plain the decreased bone resorption induced by this drug
in vitro.
Acknowledgments
We are indebted to Caroline Morieux for expert technical
assistance, to Halah Allameddine (INSERM, Unite 153, Prof.
Fardeau) for helpful advice in the preparation of myoblast
cultures, and to Brigitte Gouin for preparing the manuscript.
References
1. Baron R, Vignery A, Horowitz M 1984 Lymphocytes, macrophages
and the regulation of bone remodeling. In: Peck WA (ed) Bone and
Mineral Research, annual 2. Elsevier, Amsterdam, pp 175-243
2. de Vernejoul MC, Horowitz M, Demignon J, Neff L, Baron R 1988
Bone resorption by isolated chick osteoclasts in culture is stimu-
lated by murine spleen cell surpernatant fluids (osteoclast-activat-
ing factor) and inhibited by calcitonin and prostaglandin E2. J
Bone Mineral Res 3:69-79
3. Gowen M, Wood DD, Ihrie EJ, McGuire MKB, Russell RGG 1983
An interleukin-1 like factor stimulates bone resorption in vitro.
Nature 306:378-380
4. Pfeilschifter J, Chenu C, Bird A, Mundy GR, Roodman GD 1989
Interleukin-1 and tumor necrosis factor stimulate the formation of
human osteoclast like cells in vitro. J Bone Mineral Res 4:113
5. Chenu C, Pfeilschifter J, Mundy GR, Roodman GD 1988 Trans-
forming growth factor beta inhibits the formation of osteoclast-
like cells in long-term human marrow cultures. Proc Natl Acad Sci
USA 85:5683-5687
6. Kahan BD 1989 Cyclosporine. N Engl J Med 321:1725-1738
7. Bunjes D, Hard C, Rollinghoff M, Wagner H 1981 Cyclosporin A
mediates immunosuppression of primary cytotoxic T cell responses
by impairing the release of interleukin 1 and interleukin 2. Eur J
Immunol 11:657-661
8. Thompson AW, Moon DK, Gecey CL, Nelson DS 1983 Cyclosporin
A inhibits lymphokine production but not the responses of mac-
rophages to lymphokines. Immunology 48:291-299
9. Stewart PJ, Green OC, Stern PH 1986 Cyclosporin A inhibits
calcemic hormone-induced bone resorption in vitro. J Bone Mineral
Res 1:285-291
10. Klaushofer K, Hoffmann O, Stewart PJ, Czerwenka E, Roller K,
Peterlik M, Stern PH 1987 Cyclosporin A inhibits bone resorption
in cultured neonatal mouse calvaria. J Pharmacol Exp Ther
243:584-590
11. Orcel Ph, Bielakoff J, Modrowski D, Miravet L, de Vernejoul MC
1989 Cyclosporin A induces in vivo inhibition of resorption and
stimulation of formation in rat bone. J Bone Mineral Res 4:387-
391
12. Movsowitz C, Epstein S, Fallon M, Ismail F, Thomas S 1988
Cyclosporin-A in vivo produces severe osteopenia in the rat: effect
of dose and duration of administration. Endocrinology 123:2571-
1645
2577
13. Schlosberg M, Movsowitz C, Epstein S, Ismail F, Fallon MD,
Thomas S 1989 The effect of cyclosporin A administration and its
withdrawal on bone mineral metabolism in the rat. Endocrinology
124:2179-2184
14. del Pozo E, Graeber M, Elford P, Payne T 1990 Regression of bone
and cartilage loss in adjuvant arthritic rats after treatment with
cyclosporin A. Arthritis Rheum 33:247-252
15. Klein L, Li QX, Donovan CA, Powell AE 1989 Variation of resorp-
tion rates in vivo of various bones in immature rats. Bone Mineral
8:169-175
16. Stern PH, Philips TE, Mavreas T 1980 Bioassay of 1,25-dihydroxy-
vitamin D in human plasma purified by partition, alkaline extrac-
tion and high-pressure chromatography. Anal Biochem 102:22
17. Lorenzo JA, Hock JM, Raisz LG 1983 DNA synthesis is not
necessary for osteoclastic response to parathyroid hormone in
cultured fetal rat long bones. J Clin Invest 72:1924-1929
18. Pfeilschifter J, Seyedin SM, Mundy GR 1988 Transforming growth
factor beta inhibits bone resorption in fetal rat long bone cultures.
J Clin Invest 82:680-685
19. King GJ, Holtrop ME, Raisz LG 1978 The relation of ultrastruc-
tural changes in osteoclasts to resorption in bone cultures stimu-
lated with parathyroid hormone. Metab Bone Dis Rel Res 1:67-74
20. Lee KC, Wong M 1980 Functional heterogeneity of culture-grown
bone marrow-macrophages. I. Antigen presenting function. J Im-
munol 125:86-91
21. Orcel Ph, Bielakoff J, de Vernejoul MC 1990 Formation of multi-
nucleated cells with osteoclast precursor features in human cord
monocytes cultures. Anat Rec 226:1-9
22. Weinberg JB, Hobbs MM, Misukonis MA 1984 Recombinant hu-
man 7-interferon induces human monocyte polykaryon formation.
Proc Natl Acad Sci USA 81:4554-4557
23. O'Neill MC, Stockdale FE 1972 A kinetic analysis of myogenesis
in vitro. J Cell Biol 52:52-65
24. Klaushofer K, Horandner H, Hoffmann O, Czerwenka E, Konig
U, Koller K, Peterlik M 1989 Interferon-g and calcitonin induce
differential changes in cellular kinetics and morphology of osteo-
clasts in cultured neonatal mouse calvaria. J Bone Mineral Res
4:585-606
25. Hoffmann O, Klaushofer K, Gleispach H, Leis HJ, Luger T, Koller
K, Peterlik M 1987 Gamma interferon inhibits basal and interleu-
kin 1-induced prostaglandin production and bone resorption in
neonatal mouse calvaria. Biochem Biophys Res Commun 143:38-
43
26. Orcel Ph, Bielakoff J, de Vernejoul MC 1990 Effects of transform-
ing growth factor-/3 on long-term human cord blood monocyte
cultures. J Cell Physiol 142:293-298
27. Ash P, Loutit JF, Townsend KMS 1980 Osteoclasts derived from
hematopoietic stem cells. Nature 283:669-671
28. Nijweide PJ, Burger EH, Feyen JHM 1986 Cells of bone: prolifer-
ation, differentiation, and hormonal regulation. Physiol Rev
66:855-886
29. Mundy GR, Roodman GD 1987 Osteoclast ontogeny and function.
In: Peck WA (ed) Bone and Mineral Research, annual 5. Elsevier,
Amsterdam, pp 209-279
30. Abe E, Miyaura C, Tanaka H, Shuna Y, Kuribayashi T, Suda SO,
Nishii Y, DeLuca HF, Suda T 1983 la,25-Dihydroxyvitamin D
promotes fusion of mouse alveolar macrophages both by a direct
mechanism and by a spleen cell-mediated indirect mechanism.
Proc Natl Acad Sci USA 80:5583-5587
31. Bar-Shavit Z, Teitelbaum SL, Reitsma P, Hall A, Pegg LE, Trial
J, Kahn AJ 1983 Induction of monocytic differentiation and bone
resorption by 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA
80:5907-5911
32. Postlethwaite AE, Jackson BK, Beachey EH, Kang AH 1982
Formation of multinucleated giant cells from human monocyte
precursors, mediation by a soluble protein from antigen- and
mitogen-stimulated lymphocytes. J Exp Med 155:168-178
33. Espevik T, Figari IS, Shalaby MR, Lackides GA, Lewis GD, She-
pard HM, Palladino Jr MA 1987 Inhibition of cytokine production
by cyclosporin A and transforming growth factor /3. J Exp Med
166:571-576
1646 CyA DECREASES OSTEOCLAST FORMATION IN VITRO
34. Skjodt H, Grawford A, Elford PR, Ihrie E, Wood DD, Russell RGG
1985 Cyclosporin A modulates interleukin-1 activity on bone in
vitro. Br J Rheumatol [Suppl 1] 24:165-169
35. Stewart PJ, Stern PH 1989 Cyclosporines: correlation of immu-
nosuppressive activity and inhibition of bone resorption. Calcif
Tissue Int 45:222-226
36. Benson A, Ziegler HK 1989 Macrophages as targets for inhibition
by cyclosporine. Transplantation 47:696-703
37. O'Leary TJ, Ross PD, Lieber MR, Levin IW 1986 Effects of
cyclosporine A on biomembranes. Vibrational spectroscopic, calor-
imetric and hemolysis studies. Biophys J 49:795-801
38. Matyus L, Balazs M, Aszalos A, Mulhern S, Damjonovich S 1986
Cyclosporin A depolarizes cytoplasmic membrane potential and
interacts with Ca2+ ionophores. Biochim Biophys Acta 886:353-
360
39. Nicchitta C, Kamoun M, Williamson JR 1985 Cyclosporine aug-
ments receptor-mediated cellular Ca2+ fluxes in isolated hepato-
cytes. J Biol Chem 260:13613-13618
40. Crompton M, Ellinger H, Costi A 1988 Inhibition by cyclosporin
Endo 1991
Voll28No3
A of a Ca2+-dependent pore in heart mitochondria activated by
inorganic phosphate and oxidative stress. Biochem J 255:357-357
41. Draznin B, Metz SA, Sussman KE, Leitner JW 1988 Cyclosporin-
induced inhibition of insulin release. Possible role of voltage-
dependent calcium transport channels. Biochem Pharmacol
37:3941-3945
42. Sraer J, Bens M, Ardaillou R 1989 Dual effects of cyclosporine A
on arachidonate metabolism by peritoneal macrophages. Phospho-
lipase activation and partial thromboxane-synthase blockage.
Biochem Pharmacol 38:1947-1954
43. Vignery A 1989 Macrophage multinucleation is accompanied by
the expression of new soluble and membrane antigens in mice. Am
J Pathol 135:565-569
44. Chung CY, Kang MS 1990 Correlation between fibronectin and its
receptor in chick myoblast differentiation. J Cell Physiol 142:392-
400
45. Law PH, Goodwin TG, Li HJ 1988 Histoincompatible myoblast
injection improves muscle structure and function of dystrophic
mice. Transplant Proc 20:1114-1119