Full text of DOI:10.1359/jbmr.2002.17.3.434

JOURNAL OF BONE AND MINERAL RESEARCH
Volume 17, Number 3, 2002
©
2002 American Society for Bone and Mineral Research
Evidence That Peroxynitrite Affects Human Osteoblast
Proliferation and Differentiation
´
FRANCISCO AIRTON CASTRO DA ROCHA and ARTUR JOSE DE BRUM-FERNANDES
ABSTRACT
⅐
⅐
Peroxynitrite (PN), a nitric oxide (NO )–derived anion, has been associated with NO damage in various cell types.
We examined the effects of adding PN to cultured human osteoblast-like (hOB) cells obtained after hip arthro-
3
plasty. Exposure to PN (0.1–0.4 mM) decreased both hOB proliferation and differentiation, measured by [ H]thy-
midine uptake and alkaline phosphatase production, respectively. Incubation with 3-morpholinosydnonimine
⅐
؊
2
(
SIN-1; 0.25–1 mM), an NO and O donor that leads to PN release, also reduced both hOB proliferation and
differentiation. Coincubation with both superoxide dismutase (SOD; 100 U/ml) and catalase (CAT; 50 U/ml),
⅐
rendering SIN-1 a pure NO donor, reversed its effects on hOB proliferation and differentiation. However,
⅐
SIN-1–induced NO production, measured by nitrite release to the hOB medium, was not altered by cotreatment
with SOD and CAT. Expression of nitrotyrosine by hOB, a marker of PN action, was significantly increased after
SIN-1 addition, as compared with untreated cells, as revealed by Western blot analysis. Interleukin-1␣ (IL-1␣)
and interferon ␥ (IFN-␥) but not tumor necrosis factor ␣ (TNF-␣) also significantly increased nitrotyrosine
expression in these cells. These data show that PN is at least partially responsible for osteoblast derangement
⅐
by NO and that cytokines released during inflammatory arthropathies can induce PN production in hOB cells.
(
J Bone Miner Res 2002;17:434–442)
Key words: nitric oxide, peroxynitrite, osteoblasts, osteoporosis, cytokines
⅐
INTRODUCTION
release of large amounts of NO , would lead to increased
(6)
bone resorption in certain disease states.
Deleterious effects attributed to NO may be at least
⅐
⅐
ITRIC OXIDE (NO ), a gaseous free radical, has been
N
linked to a diverse array of physiological phenomena partially linked to the reactive nitrogen species generated
(1)
(7)
⅐
including vascular tone regulation and neurotransmission
as well as to cytotoxic activities such as inhibition of mito-
after NO release. If compared with other free radicals,
⅐
NO is relatively stable and unreactive toward biological
(2)
(3)
chondrial respiration and DNA damage. The identifica-
tion of the expression of both constitutive and inducible
nitric oxide synthase (cNOS and iNOS, respectively) iso-
forms in both osteoblasts and osteoclasts suggests that NO
is involved in bone metabolism.
els of NO via cNOS has been assigned a physiological role
in bone function, whereas activation of iNOS, mainly under
the influence of proinflammatory cytokines, by provoking
⅐
molecules. However, NO has great reactivity with other
(8)
free radicals. When produced in large amounts in a mi-
Ϫ
croenvironment rich in superoxide (O2 ), which happens
⅐
⅐
after macrophage triggering in an inflammatory milieu, NO
(4,5)
Production of low lev-
Ϫ
and O2 rapidly combine to produce a strong oxidant, the
peroxynitrite (PN) anion (ONOO ). Although the rate con-
⅐
Ϫ
9
Ϫ1 Ϫ1
Ϫ
⅐
stant (Ͼ10 M
s
) of this reaction (O ϩ NO 3
2
Ϫ
⅐
ONOO ) is similar to the reaction of NO with other free
radicals, the former is assumed to have more pathophysio-
logical relevance, because many cell types are able to re-
The authors have no conflict of interest.
Department of Medicine, Rheumatic Diseases Unit, Faculty of Medicine, Universit e´ de Sherbrooke, Sherbrooke, Canada.
4
34

PEROXYNITRITE AFFECTS HUMAN OSTEOBLASTS
435
⅐
Ϫ
(9)
lease both NO and O after being activated. Hence, a Amersham Pharmacia Biotech, Inc. (Piscataway, NJ, USA).
2
⅐
Ϫ
potential beneficial effect of NO to scavenge O , which All other reagents used were of high-grade quality.
2
⅐
might happen in low NO concentrations, is outweighed
Ϫ
during inflammation by the favored kinetics of ONOO
Cell culture
formation, thus leading to the release of a much powerful
(7)
and long-lived oxidant.
Specimens of human trabecular bone were obtained from
PN has been shown to affect a diverse array of molecules patients undergoing total hip arthroplasty. In conformity
including proteins, lipids, and nucleic acids. Nitrosylation with Qu e´ bec’s civil law, each patient signed an informed
of tyrosine residues in proteins by PN can significantly consent. HOBs were cultured from these tissues following a
(20)
affect their function. Formation of 3-nitrotyrosine from PN previously reported method with minor modifications.
activity in proteins has been assumed to be a biological Briefly, thin slices of trabecular bone were treated with 2.5
(9)
marker for PN formation in vivo. In vitro studies have mg/ml of trypsin in phosphate-buffered saline (PBS) for 1 h
reported cytotoxic activity of PN. Cultures of HL-60 and at 37°C, washed extensively in PBS, and cut into fragments
Ϫ
2
PC-12 tumor cell lines in the presence of ONOO have of approximately 2 mm . Fragments were seeded in
been shown to induce cell apoptosis.
(10,11)
Nitrosylation of 100-mm culture dishes containing 5 ml MEM supplemented
tyrosine residues and subsequent inactivation of surface recep- with 10% FBS, 50 U/ml of penicillin, 150 ␮g/ml of strep-
tors in both endothelial and epithelial cells have been reported tomycin, 2.5 ␮g/ml of amphotericin, 2.2 mg/ml of sodium
(12–15)
as evidence of PN derangement of cell function.
More bicarbonate, and 50 ␮g/ml of L-ascorbate. The plates were
recently, early loss of P450 activity of cultured rat hepato- incubated at 37°C in a 5% CO atmosphere. The culture
2
Ϫ
cytes was associated with ONOO production via cNOS medium was replaced at 3- to 4-day intervals. After reach-
(
16)
activation.
Detection of nitrotyrosine in atherosclerotic plaques,
ing confluence (4–5 weeks), cells were trypsinized and
subcultured at a ratio of 20,000 cells/cm in 100-mm culture
(17)
2
in the serum and synovial fluid of rheumatoid arthritis dishes. First-passage cells were grown to confluence, incu-
(
18)
(19)
patients and in rejected human kidney allografts, have bated 24 h in MEM containing 0.2% FBS, and used in the
been described as evidence of both presence and linkage of experiments. These cells have been shown previously to
PN formation to human disease states. However, appropri- exhibit phenotypic characteristics of osteoblasts such as
Ϫ
ate studies focusing on ONOO role in bone metabolism response to parathyroid hormone (PTH) and production of
(20)
and pathologies are still lacking. Considering that produc- alkaline phosphatase and osteocalcin.
⅐
Ϫ
tion of NO and O in the bone microenvironment occurs
2
in both normal and disease states, we aimed to investigate
the effect of PN addition to cultured human osteoblasts
PN, SIN-1, SOD, CAT, and cytokine treatments
(
hOBs). Because stimulation of bone cells with cytokines
The original light yellow solution of PN was stored for no
⅐
has been linked to increased NO release, we also investi- more than 1 week on arrival at Ϫ80°C in the buffer (NaOH,
gated whether cytokine stimulation of hOBs could be asso- 0.3N), as supplied by the manufacturer. Immediately before
ciated with PN formation.
use, PN concentration was checked by measuring absor-
Ϫ1
bance at 302 nm (extinction coefficient; ⑀ ϭ 1670 M
Ϫ1
cm ) using 0.3N of NaOH as blank. Working solutions
were made in 5-␮l aliquots of 0.3N of NaOH. PN incuba-
tions were made in FBS-free ␣-MEM, during 10 minutes.
To maximize the number of cells exposed to ONOO , the
MATERIALS AND METHODS
Ϫ
Materials
culture dish was inclined and the PN solution was added to
Eagle’s modified essential medium (MEM), ␣-MEM, and the side of the dish followed by a rapid swirl. The aliquots
fetal bovine serum (FBS) were purchased from Gibco BRL were adjusted using 0.3N of NaOH to achieve the final
(
(
Rockville, MD, USA). PN, 3-morpholinosydnonimine concentration so that all cells were exposed to the same pH.
SIN-1), nitrotyrosine, mouse antinitrotyrosine monoclonal After these 10-minute incubations, the medium was aspi-
antibody, nitrotyrosine-bovine serum albumin nitrotyrosine- rated and the cells were washed with PBS and reincubated
bovine serum albumine (nitrotyrosine-BSA), L-NMMA, with ␣-MEM supplemented with 10% FBS, as mentioned
and a nitrite/nitrate detection kit were all purchased from previously. Controls for PN included incubation with the
Cayman Chemical Co. (Hornby, Ontario, Canada). 0.3N of NaOH buffer, as well as inactive PN, that consisted
Ϫ
Interleukin-1␣ (IL-1␣), tumor necrosis factor ␣ (TNF-␣), of a solution of ONOO allowed to decompose in buffer
penicillin/streptomycin, L-ascorbate, amphotericin, trypsin, before addition (reverse-order addition) so that the influence
Ϫ
p-nitrophenylphosphate, and 1,25-dihydroxyvitamin D3 of nitrates and nitrites present in the ONOO solution could
[
1,25(OH) D ] were purchased from Sigma (St. Louis, MO, be excluded. The absorbance of this inactive PN solution
2 3
USA). Complete (Mini protease inhibitor cocktail), human was checked at 302 nm before addition to the culture
interferon ␥ (IFN-␥) human superoxide dismutase (SOD) medium.
and catalase (CAT) were purchased from Boehringer In-
gelheim (Mannheim, Germany). [ H]Thymidine, hybond C radicals such as NO and O , ONOO has a short half-life
Although it can be considered stable if compared with
3
⅐
Ϫ Ϫ
2
extra nitrocellulose membranes, horseradish peroxidase– of Ͻ1 s in physiological solutions. Attempts to overcome
linked rabbit immunoglobulin G (IgG), and ECL-plus en- the obvious difficulties in interpreting data when using
hanced chemiluminescence system were purchased from authentic PN with both in vivo and in vitro systems include

4
36
DA ROCHA AND DE BRUM-FERNANDES
Ϫ
Ϫ
the use of ONOO donors. SIN-1 is considered an ONOO
donor because its unique ability to chemically release both
Nitrotyrosine detection
⅐
Ϫ
Ϫ
(21)
Nitrotyrosine detection was used as a marker of PN effect
on cell proteins. For the Western blots, cell suspensions of
NO and O leads to ONOO generation.
2
The compounds SIN-1, IL-1␣ (100 U/ml), TNF-␣ (100
U/ml), or IFN-␥ (10 U/ml) were added directly to the
culture medium. Cells were aerated gently after SIN-1 ad-
dition to the medium to avoid hypoxia. SOD (50 U/ml),
dissolved in PBS, and CAT (100 U/ml) were added to the
medium before incubation to give the desired final concen-
tration. Both enzymes always were added before SIN-1.
1
00-mm culture dishes obtained after different treatments
were lysed in the presence of protease inhibitors (Complete,
Mini) with Tris-NaCl and NP-40 followed by sonication
and centrifugation. The protein concentration on the super-
natant was assessed using the BioRad method and compared
with a standard curve using the lysis buffer as reference at
6
20 nm. Samples were run on 9% sodium dodecyl sulfate–
polyacrylamide gel electrophoresis (SDS-PAGE; 50 ␮g of
protein/lane). After migration, proteins were electrophoreti-
cally transferred to nitrocellulose membranes. Nonspecific
binding sites were blocked overnight in 0.1% blocking
buffer Tris-buffered saline (TBS)/Tween containing 5%
BSA. Membranes were then probed with 1.5 ␮g/ml of
mouse antinitrotyrosine monoclonal antibody for 1 h at
room temperature in 0.1% TBS/Tween 0.1% containing 1%
BSA. After five washings with 0.1% TBS/Tween containing
0.1% BSA the membranes were incubated with a horserad-
ish peroxidase–linked anti-mouse polyclonal antibody
raised in sheep (1:2000). After five washings with 0.1%
TBS/Tween containing 0.1% BSA, the membranes were
developed using the ECL detection system. Controls for the
antinitrotyrosine antibody specificity included preincuba-
tion of the first antibody with nitrotyrosine followed by pH
adjustment as well as incubation with just the second anti-
body. Nitrotyrosine staining was quantitated by scanning
densitometry of the Western blot membranes. The means Ϯ
SEM of the band intensities of three experiments obtained
with the various stimuli were compared with nonstimulated
hOB.
Cell proliferation
Osteoblast proliferation was studied by measuring
H]thymidine (2 ␮Ci/ml) incorporation into trichloroacetic
3
[
acid (TCA)–insoluble material. Briefly, cells were plated at
a density of 3 ϫ 10 cells/well in 24-well plates and cultured
for 48 h in ␣-MEM culture medium with antibiotics, sup-
plemented with 10% FBS, in the presence of [ H]thymidine.
4
3
4
8 h later, the plates were washed with PBS and a 10% TCA
3
solution was added to the wells. Incorporated [ H]thymidine
was released through washing with 0.2N of NaOH and
radioactivity was measured using a ␤-scintillation counter.
Results are expressed as counts per minute of triplicates of
three different experiments obtained from different donors.
Measurement of alkaline phosphatase activity
Alkaline phosphatase was used as a marker of osteoblast
differentiation and was measured in disrupted cells using
p-nitrophenyl phosphate as substrate, as described else-
(
22)
4
where. Briefly, cells were seeded at a density of 3 ϫ 10
cells/well in 24-well plates and cultured in the presence of
0 nM of 1,25(OH) D for 72 h with varying treatments. A
1
2
3
baseline control for this alkaline phosphatase production Statistics
consisted of cells cultured in the absence of vitamin D3.
Data are presented as means Ϯ SEM of triplicate values
After the incubation period, cells were harvested in 0.4 M of
Tris buffer and sonicated, and the alkaline phosphatase
activity in the supernatant was measured spectrophotometri-
cally at 405 nm. Values obtained were compared with a
standard curve using p-nitrophenol as standard. Data were
expressed as micromoles of alkaline phosphatase activity
per milligram of protein. Protein concentration in the su-
pernatants was assessed using the BioRad method (BioRad,
Mississauga, Ontario, Canada).
of at least three experiments, using different donors. To
compare means we used analysis of variance (ANOVA),
followed by Tukey’s test. The value of p Ͻ 0.05 was
considered significant.
RESULTS
PN decreases hOB proliferation and differentiation
Authentic PN addition to hOB culture decreased both
Nitrite assay
basal proliferation (Fig. 1A) and differentiation (Fig. 1B)
3
⅐
Total NO (nitrate ϩ nitrite) production was assessed measured by [ H]thymidine incorporation and vitamin D
–
3
indirectly using the release of nitrite to the medium based on stimulated alkaline phosphatase production, respectively. At
(
23)
the Griess reaction, as described previously,
commercially available kit. Briefly, cells were seeded at a trypan blue exclusion was not affected (Ͼ95% viable).
using a the doses and conditions used, cell viability assessed by
4
density of 3 ϫ 10 cells/well in 24-well plates and cultured Addition of vitamin D
significantly increased alkaline
3
in the presence of 10 nM of 1,25(OH) D for 72 h, with phosphatase production when compared with baseline val-
2
3
varying treatments. A baseline control for this experiment ues (Fig. 1B).
consisted of cells cultured in the absence of vitamin D3.
After the incubation period, cells were harvested in 0.4 M
SIN-1 decreases hOB proliferation
Tris buffer and sonicated, and total nitrate/nitrite was as-
sessed using the Griess reaction. Data were expressed as
micromoles per liter of nitrite.
⅐
Ϫ
Addition of SIN-1, a NO and O donor, to the hOB
2
culture significantly and dose-dependently decreased prolif-

PEROXYNITRITE AFFECTS HUMAN OSTEOBLASTS
437
FIG. 2. Effect of addition of different concentrations of SIN-1 iso-
lated or SIN-1 (1 mM) combined with SOD (50 U/ml; SIN-1 ϩ SOD),
CAT (100 U/ml; SIN-1 ϩ CAT), or SIN-1 (1 mM) ϩ SOD (50 U/ml)
ϩ CAT (100 U/ml; SIN-1 ϩ SOD ϩ CAT) on the proliferation of
3
cultured hOBs. hOB proliferation was assessed by [ H]thymidine in-
corporation 48 h after exposure to SIN-1. Data are expressed as the
3
mean Ϯ SEM of [ H]thymidine counts of three experiments performed
with three different donors (*p Ͻ 0.05 compared with nonstimulated
#
hOB [CTL] using ANOVA, followed by Tukey’s test; p Ͻ 0.05
compared with SIN-1 [1 mM], using ANOVA, followed by Tukey’s
test).
SOD ϩ CAT without SIN-1 did not significantly affect hOB
proliferation.
FIG. 1. Effect of addition of PN on the (A) proliferation and (B) 10
nM of vitamin D –induced alkaline phosphatase production by cultured
hOBs. hOB proliferation was assessed by [ H]thymidine incorporation
3
3
SIN-1 decreases hOB differentiation
4
8 h after exposure to PN at increasing concentrations. Alkaline
Addition of SIN-1 to the hOB culture significantly and
phophatase was assessed in the cell lysate using p-nitrophenyl phos-
phate as substrate. A solution of PN allowed to decompose completely
before addition to the medium (reverse order addition) was used as
dose-dependently decreased vitamin D –induced alkaline
3
phosphatase production when compared with control (Fig.
inactive PN to exclude the influence of contaminants (nitrate, nitrite, 3). SIN-1 (1 mM) decreased hOB alkaline phosphatase
etc.) present in the original PN solution (see text for details). Data are production to 64% of the control level (p Ͻ 0.01), whereas
3
expressed as the mean Ϯ SEM of either [ H]thymidine counts or as SIN-1 at the concentrations of either 0.5 mM or 0.25 mM
micromoles of alkaline phosphatase activity per milligram of protein of
three experiments performed with three different donors. Protein con-
centration was measured in the cell lysates using the Bradford method
was ineffective (p Ͼ 0.05). A combined treatment with
SIN-1 (1 mM) ϩ SOD (50 U/ml) ϩ CAT (100 U/ml) but
not with either of the enzymes isolated completely reversed
SIN-1 effect on hOB differentiation (Fig. 3). Incubation
with a combination of SOD ϩ CAT without SIN-1 did not
significantly affect the parameter studied (Fig. 3).
(
panel A, *p Ͻ 0.05 compared with nonstimulated hOB (CTL), using
ANOVA followed by Tukey’s test; panel B, *p Ͻ 0.05 compared with
vitamin D –stimulated cells, using ANOVA followed by Tukey’s test).
3
eration when compared with nonstimulated cells (Fig. 2).
SIN-1 (0.5 mM or 1 mM) decreased hOB proliferation to
Cytokine treatment effect on hOB proliferation
3
7% and 63.6% of the untreated cells, respectively, whereas
Figure 4 shows that incubation of hOB with IL-1␣,
SIN-1 at a concentration of 0.25 mM was ineffective (p Ͼ TNF-␣, or IFN-␥ isolated did not significantly alter prolif-
.05). At the doses and conditions used, cell viability as- eration. However, when all three cytokines were combined,
0
sessed by trypan blue exclusion was not affected (Ͼ95% a reduction of 36% in hOB proliferation was observed, as
viable cells). Combined treatment with SIN-1 (1 mM) ϩ compared with untreated cells. Addition of 100 ␮M of
SOD (50 U/ml) ϩ CAT (100 U/ml), rendering SIN-1 a pure L-NMMA, an NOS inhibitor, to the hOB culture arrested
⅐
NO donor, partially but significantly reversed SIN-1 effect proliferation regardless of cytokine presence, suggesting
⅐
on hOB proliferation (Fig. 2). Incubation of SIN-1 with that a certain NO basal level is needed in order for osteo-
either of the enzymes separately or a combination of blasts to proliferate. Cell viability (Ͼ95%) assessed by

4
38
DA ROCHA AND DE BRUM-FERNANDES
FIG. 3. Effect of addition of different concentrations of SIN-1 iso-
lated or SIN-1 (1 mM) combined with SOD (50 U/ml; SIN-1 ϩ SOD),
CAT (100 U/ml; SIN-1 ϩ CAT), or SIN-1 (1 mM) ϩ SOD (50 U/ml)
ϩ CAT (100 U/ml; SIN-1 ϩ SOD ϩ CAT) on the 10-nM vitamin
FIG. 4. Effect of addition of IL-1␣ (100 U/ml), TNF-␣ (100 U/ml),
IFN-␥ (10 U/ml), a combination of the three cytokines (IL-1 [100
U/ml] ϩ TNF [100 U/ml] ϩ IFN [10 U/ml]), L-NMMA (LN; 100 ␮M)
or a combination of LN (100 ␮M) ϩ IL-1 (100 U/ml) ϩ TNF (100U/
ml) ϩ IFN (10 U/ml) on the proliferation of cultured hOBs. hOB
D –induced alkaline phosphatase production of cultured hOBs. Alka-
3
line phosphatase was assessed in the cell lysate using p-nitrophenyl
phosphate as substrate. Data are expressed as micromoles of alkaline
phosphatase activity per milligram of protein of three experiments
performed with three different donors. Protein concentration was mea-
sured in the cell lysates using the Bradford method (*p Ͻ 0.05 com-
pared with nonstimulated hOB [CTL] using ANOVA followed by
Tukey’s test).
3
proliferation was assessed by [ H]thymidine incorporation 48 h after
exposure to cytokines. Data are expressed as the mean Ϯ SEM of
3
[
H]thymidine counts of three experiments performed with three dif-
ferent donors (*p Ͻ 0.05 compared with nonstimulated hOB [CTL]
using ANOVA followed by Tukey’s test).
SIN-1 (1 mM) ϩ SOD (50 U/ml) ϩ CAT (100 U/ml)
produced 122.1 Ϯ 3.4 ␮M compared with 4.1 Ϯ 0.2 ␮M for
trypan blue exclusion was not affected either by cytokine or control (nonstimulated) hOBs (data not shown).
L-NMMA addition to the culture medium.
SIN-1 induces tyrosine nitration in hOBs
Cytokine treatment effect on hOB differentiation
We performed a Western blot analysis in hOB-extracted
proteins using an antinitrotyrosine monoclonal antibody.
Figure 7A shows the result of a typical experiment. Each
treatment was repeated three times using hOB from differ-
ent donors. SIN-1 addition to hOB culture medium in-
creased tyrosine nitration, a marker of PN action, of hOB
proteins, as compared with control (nonstimulated) cells.
Exposure of the cells to a combination of SIN-1 (1 mM) ϩ
SOD (50 U/ml) ϩ CAT (100 U/ml) reversed the increased
tyrosine nitration induced by SIN-1 and thus implied that
generation of PN was significantly inhibited by SOD ϩ
CAT. Tyrosine nitration by SIN-1 also was inhibited by
addition of CAT (100 U/ml), whereas isolated addition of
SOD (50 U/ml) to SIN-1 did not alter SIN-1–induced ni-
trotyrosine formation (Fig. 7B). IFN-␥ treatment to hOB
induced higher tyrosine nitration as compared with control
whereas treatment with a combination of the three cytokines
Figure 5 shows that incubation of hOB in the presence of
either IL-1␣ or TNF-␣ alone did not alter alkaline phospha-
tase production. However, addition of IFN-␥ alone or as a
combined treatment with IL-1␣ ϩ TNF-␣ greatly enhanced
alkaline phosphatase production. Addition of 100 ␮M of
L-NMMA arrested alkaline phosphatase production, regard-
⅐
less of cytokine presence, suggesting that a certain NO
basal level is needed in order for osteoblasts to differentiate.
Nitrite production
Nitrite release to the culture medium was used to evaluate
⅐
NO release. Incubation of hOB in the presence of vitamin
D3 slightly but significantly increased nitrite release as
compared with control (nonstimulated) cells (Fig. 6). PN,
IL-1␣, or TNF-␣ did not alter nitrite release by hOB
whereas IFN-␥ increased (36%) nitrite release as compared
with control (nonstimulated) hOB. Addition of 100 ␮M of
L-NMMA reduced (50%) nitrite levels as compared with
nonstimulated cells, regardless of cytokine presence (p Ͻ
(
IL-1␣ ϩ TNF-␣ ϩ IFN-␥) did not alter tyrosine nitration.
Addition of L-NMMA significantly decreased nitrotyrosine
formation when used alone or when combined with cyto-
kines (Fig. 7B).
0
.05). SIN-1 (1 mM) addition to the culture markedly
(
around 30 times control level) increased nitrite release, an
DISCUSSION
effect that was not altered by a combined treatment with
SIN-1 (1mM) ϩ SOD (50 U/ml) ϩ CAT (100 U/ml). Nitrite
This study provides evidence, for the first time, that PN
production by 1 mM of SIN-1 was 120.1 Ϯ 2.8 ␮M and addition to hOB culture inhibits proliferation and differen-

PEROXYNITRITE AFFECTS HUMAN OSTEOBLASTS
439
FIG. 5. Effect of addition of IL-1␣ (100 U/ml), TNF-␣ (100 U/ml),
IFN-␥ (10 U/ml), a combination of the three cytokines (IL-1 [100
U/ml] ϩ TNF [100 U/ml] ϩ IFN [10 U/ml]), L-NMMA (LN; 100 ␮M),
or a combination of LN (100 ␮M) ϩ IL-1 (100 U/ml) ϩ TNF (100
FIG. 6. Effect of addition of 0.4 mM PN, inactive PN, IL-1␣ 100
(U/ml), TNF-␣ (100 U/ml), IFN-␥ (10 U/ml), a combination of the
three cytokines (IL-1 [100 U/ml] ϩ TNF [100 U/ml] ϩ IFN [10 U/ml]),
L-NMMA (LN; 100 ␮M), or a combination of LN (100 ␮M) ϩ IL-1
U/ml) ϩ IFN (10 U/ml) on the 10 nM of vitamin D –induced alkaline
3
phosphatase production by cultured hOBs. Alkaline phosphatase was (100 U/ml) ϩ TNF (100 U/ml) ϩ IFN (10 U/ml) or LN (100 ␮M) ϩ
assessed in the cell lysate using p-nitrophenyl phosphate as substrate. IFN (10 U/ml) on the nitrite production of cultured hOBs. Nitrite
Data are expressed as micromoles of alkaline phosphatase activity per
milligram of protein of three experiments performed with three differ-
ent donors. Protein concentration was measured in the cell lysates using
production was assessed in the cell supernatant using the Griess reac-
tion. Data are expressed as micromoles of three experiments performed
with three different donors (*p Ͻ 0.05 compared with nonstimulated
the Bradford method (*p Ͻ 0.05 compared with vitamin D –stimulated hOB [CTL] using ANOVA followed by Tukey’s test).
3
cells, using ANOVA followed by Tukey’s test).
previous studies focusing on PN effects on osteoblasts.
tiation of these cells. It also shows that IFN-␥ addition to Further, most studies have used transformed cell lines and
cultured hOB leads to 3-nitrotyrosine formation, suggesting PN was added to alkaline buffers, a strategy that we could
that these cells are able, under cytokine stimulation, to not use with hOB.
⅐
generate PN.
Pharmacologic manipulation using NO donors that can
As mentioned previously, technical difficulties in work- generate PN is an alternative to bypass the difficulties when
ing with PN pose a serious problem in exploring the mech- working with authentic PN. SIN-1 has the unique ability to
⅐
Ϫ
anisms of action of this anion. Because of the short half-life chemically generate both NO and O and, thus, leads to
2
of PN, especially in physiological solutions, addition of PN production. In this study, incubation of hOB in the
authentic PN to cell culture medium leads to irregular presence of SIN-1 dose-dependently decreased both prolif-
exposition of the cells. Incubation in alkaline solutions by eration and differentiation. The SIN-1 effect on hOB pro-
increasing the half-life of PN is a strategy to homogenize liferation was more pronounced because either 0.5-mM or
this exposition. However, in our system, use of alkaline 1-mM solutions were equally effective in inhibiting prolif-
buffers severely affected cell viability. Our alternative was eration, whereas differentiation was inhibited only with the
to add PN to the ␣-MEM FBS-free medium in an attempt to 1-mM SIN-1 solution. These were reproducible and sus-
increase PN availability to the cells.
tained effects that did not affect cell viability. Coincubation
We found that addition of 0.2–0.8 mM of authentic PN to with SOD and CAT but not with either of these scavengers
hOB significantly affected both proliferation and differen- alone did significantly reverse the SIN-1 effect on hOB
tiation of the cells. However, a typical dose-response curve proliferation. Moreover, vitamin D –induced alkaline phos-
3
was not obtained. Although 0.2 mM of PN arrested hOB phatase production was restored completely when cells
proliferation, a concentration of PN as low as 0.1 mM was were cotreated with SIN-1 ϩ SOD ϩ CAT, thus confirming
enough to significantly decrease vitamin D –induced alka- that the effect observed was caused by the action of PN and
3
⅐
line phosphatase production by these cells. At the concen- not by NO . This strategy has been used already in a
trations studied, hOB viability was not affected by PN previous study showing that incubation of normal human
treatment. The fact that a decomposed PN solution was keratinocytes in the presence of 1 mM of SIN-1 inhibited
ineffective argues against the possibility that contaminants their growth, an effect that was reversed by a combined
(
25)
(
hydrogen peroxide, nitrite, etc.) present in the PN solution treatment with SIN-1 ϩ SOD ϩ CAT. It has been shown
(24)
were involved in the response attributed to PN.
that adding only SOD to SIN-1 does not inhibit PN gener-
Although the PN concentrations that we used were sim- ation, as could be expected. In addition to the unfavored
Ϫ
ilar to previous studies using other cell types, comparison of kinetics of O scavenging by SOD as compared with the
2
⅐
Ϫ
our data with the literature is difficult because there are no faster NO and O reaction, the presence of SOD progres-
2

4
40
DA ROCHA AND DE BRUM-FERNANDES
FIG. 7. (A) Results of a typical experiment to
analyze nitrotyrosine presence in cultured hOBs
after exposure to different treatments, assessed
in hOB lysates. hOBs were cultured in ␣-MEM
supplemented with 0.2% FBS and either of the
following treatments was added to the medium:
PN (400, 200, or 100 ␮M), SIN-1 (1 mM) with
or without SOD (50 U/ml; SIN-1 ϩ SOD) or
CAT (100 U/ml; SIN-1 ϩ CAT), IL-1␣ (100
U/ml), TNF-␣ (100 U/ml), IFN-␥ (10 U/ml), a
combination of the three cytokines (IL-1 [100
U/ml] ϩ TNF [100 U/ml] ϩ IFN [10 U/ml]),
L-NMMA (LN; 100 ␮M), or a combination of
LN (100 ␮M) ϩ IL-1 (100 U/ml) ϩ TNF (100
U/ml) ϩ IFN (10 U/ml) or LN (100 ␮M) ϩ IFN
(
10 U/ml). Positive control (CTLϩ) means
nitrotyrosine-BSA and negative control (CTLϪ)
means nonstimulated hOB. The PN and SIN-1
incubations were done in ␣-MEM FBS-free me-
dium. Nitrotyrosine expression in hOB proteins
obtained after cell lysis was analyzed by West-
ern blot (see Material and Methods section for
details). (B) Quantitative analysis of nitroty-
rosine staining after scanning the membranes
and analyzing the mean Ϯ SEM of the band
intensities. Results are expressed as percentage
of the nonstimulated hOB. One hundred percent
was set as the intensity of the negative control
(
CTLϪ) cells (*p Ͻ 0.05 using ANOVA fol-
lowed by Tukey’s test).
sively increases NO concentration that will outcompete creased both hOB proliferation and differentiation, the
Ϫ
⅐
SOD for O2 . Also, the combination of hydrogen peroxide SIN-1 effects in that study were attributed to NO generation
and PN can be more harmful to the cells than either of them without mention to the possible role of other reactive spe-
(
26)
(28)
alone.
Further, SOD leads to H O formation, which, cies as final mediators of the SIN-1 effects.
coupled to SOD, has been shown recently to be able to study, treatment with two different NO donors, S-nitroso-
generate PN probably through increasing NO break- acetyl-penicillamine (SNAP) and S-nitroso-glutathione
In another
2
2
⅐
⅐
(
27)
down.
These facts explain why SIN-1 effects on hOBs (GSNO) at concentrations ranging from 0.1 to 1 mM in-
were reversed only when cells were cotreated with SIN-1 ϩ duced apoptosis of mouse osteoblasts, an effect that was
⅐
(29)
SOD ϩ CAT. As expected, SIN-1 greatly increased nitrite associated with increased NO levels.
Our data add to
release (30 times the level in nonstimulated cells), and that these previous studies by providing evidence that SIN-1
effect was not altered by cotreatment with SOD ϩ CAT. It effects on normal hOB proliferation and differentiation are
⅐
⅐
should be pointed out that, in this case, NO release was not caused by PN rather than NO release.
caused by hOB stimulation. Rather, it was a chemical re-
action provoked by SIN-1 addition to the cells.
Addition of 100 ␮M of L-NMMA arrested hOB prolifer-
ation and differentiation and significantly reduced nitrite
To our knowledge, there are no previous data exploring levels, as compared with nonstimulated cells. This is in
the effects of SIN-1, as a source of PN, on hOB. Although accordance with recent data showing that endothelial
a previous study reported that SIN-1 (0.1–10 mM) de- (e)NOS knockout mice display bone abnormalities and also

PEROXYNITRITE AFFECTS HUMAN OSTEOBLASTS
441
that osteoblasts derived from these animals display reduced
Finally, we aimed to show PN release after stimulation of
growth rate in vitro and are less differentiated, as compared hOB with cytokines and SIN-1. Immune identification of
(
30,31)
with wild-type mice.
These data suggest an important nitrotyrosine in the cell extracts was chosen as a fingerprint
⅐
anabolic effect of NO produced via cNOS in hOB metab- of PN presence. Authentic PN addition had at most a slight
olism.
effect on increased nitrotyrosine expression. In addition, 0.2
Although our results with PN and SIN-1 pointed to an mM of PN induced higher nitrotyrosine levels compared
important role for this oxidant (or other reactive nitrating with 0.4 mM of PN, thus reinforcing the possibility that
species) in hOB metabolism, the pathophysiological rele- difficulties in handling PN accounting for irregular cell
vance of these data are not straightforward. It has been exposition might explain these data. As expected, SIN-1
suggested that increased bone resorption during inflamma- greatly increased (70%) nitrotyrosine expression compared
tory arthropathies may be associated with cytokine-induced with nonstimulated cells. This increase was inhibited when
⅐
NO release by osteoblasts. Hence, we sought for a possible SIN-1 was coadministered with SOD ϩ CAT. These data
⅐
association between NO and/or PN production by hOB confirm our results obtained on hOB proliferation and dif-
under cytokine stimulation and cytokine effects on prolif- ferentiation showing that the effect of SIN-1 was associated
eration and differentiation of these cells.
with PN release, as shown by tyrosine nitrosylation. The
Our data have shown that isolated addition of IL-1␣, fact that coincubation with SOD ϩ CAT inhibited SIN-1–
TNF-␣, or IFN-␥ did not modify hOB proliferation whereas induced nitrotyrosine increase suggests that PN release
a combined treatment with the three cytokines significantly rather than other reactive nitrating species was responsible
decreased it. A previous study showed that treatment with for SIN-1 effects on hOB proliferation and differentiation.
IFN-␥ significantly decreased hOB proliferation, whereas L-NMMA significantly decreased nitrotyrosine formation
(32)
treatment with TNF-␣ increased it.
However, culture when used alone or when combined with cytokines. IL-1␣
conditions in that study were not specified, thus limiting or TNF-␣ did not modify nitrotyrosine expression whereas
comparison to our results. In a more recent study using IFN-␥ increased it up to 44% compared with nonstimulated
hOB, addition of either IL-1␤ or IFN-␥ but not of TNF-␣ to cells, thus showing that this proinflammatory cytokine is
the culture medium slightly decreased osteoblast prolifera- able to induce PN production by hOB.
tion, whereas treatment with a combination of these three
Curiously, we observed that nitration of tyrosine residues
(4)
cytokines arrested it. The discrepancy between these re- in hOB appeared to be restricted to a specific polypeptide.
(
4)
sults and ours could be caused by methodological reasons Apparently, PN displays specificity regarding protein nitra-
5
(33)
such as the density of the cells plated (1 ϫ 10 cells/well in tion in the cells.
For instance, it was shown that specific
4
9
6-well plates in the mentioned study and 3 ϫ 10 cells/well manganese SOD nitration by PN is involved in the mecha-
(19)
in 24-well plates in our study). Additionally, our osteoblasts nisms of chronic rejection of human renal allografts.
were obtained from patients with severe osteoarthritis, Presently, we are trying to characterize the polypeptide that
(4)
whereas normal osteoblasts were used in that study.
Our data also showed that neither IL-1 nor TNF-␣ altered
was targeted at nitration in hOB after exposure to PN.
In summary, this work shows that either authentic PN
vitamin D –induced alkaline phosphatase production by addition to cultured hOB cells or its indirect release through
3
hOB, as compared with nonstimulated cells. Incubation SIN-1 affects both hOB proliferation and differentiation.
Ϫ
⅐
with IFN-␥ greatly increased it, whereas a combination of Blockade of ONOO release in spite of persistent NO
Ϫ
the three cytokines did not add to the IFN-␥ effect alone. generation can efficiently inhibit ONOO effect on hOB.
Few studies have investigated the effects of cytokines on Thus, these data provide evidence that PN may be a final
osteoblast functions. Using human cells, a previous study mediator of NO -induced deleterious effects on hOB metab-
⅐
obtained data similar to ours, showing a great increase in olism. It also shows that hOBs are able to produce PN after
alkaline phosphatase production by IFN-␥ whereas TNF-␣ IFN-␥ stimulation.
(
32)
decreased it.
⅐
The effect of cytokines on NO release by hOB was
evaluated by measuring the amount of nitrite released into
the culture medium. Nitrite levels were significantly higher
ACKNOWLEDGMENTS
in vitamin D –stimulated cells compared with nonstimu-
This work was supported by grants from the Medical
3
lated cells. Incubation with either IL-1␣ or TNF-␣ increased Research Council of Canada and from the Arthritis Society.
ϳ22% nitrite levels that did not reach statistical signifi- Francisco Airton Castro da Rocha is a recipient of a post-
cance. Addition of IFN-␥ significantly increased up to doctoral scholarship from CNPq–Brazil.
3
2.2% nitrite release compared with nonstimulated cells.
Addition of a combination of the three cytokines did not
⅐
alter NO production. Thus, the decrease in hOB prolifera-
REFERENCES
tion observed with the combined cytokine treatment as well
as the great enhancement of alkaline phosphatase produc-
tion induced by IFN-␥ could not be ascribed directly to
1
2
. Moncada S, Higgs A 1993 The L-arginine nitric oxide path-
way. N Engl J Med 329:2002–2012.
. Cassina A, Radi R 1996 Differential inhibitory action of nitric
oxide and peroxynitrite on mitochondrial electron transport.
Arch Biochem Biophys 328:309–316.
⅐
alteration of NO production by these cytokines. As a whole,
although these data indicate that cytokines affect hOB me-
tabolism, the exact mechanism implicated in this effect still
demands investigation.
3. Nguyen T, Brunson D, Crespi CL, Penman BW, Wishnok JS,
Tannenbaum SR 1992 DNA damage and mutation in human

4
42
DA ROCHA AND DE BRUM-FERNANDES
cells exposed to nitric oxide in vitro. Proc Natl Acad Sci USA 21. Crow JP, Ischiropoulos H 1996 Detection and quantitation of
8
9:3030–3034.
nitrotyrosine residues in proteins: In vivo marker of peroxyni-
trite. Methods Enzymol 269:185–194.
4
5
. Ralston SH, Todd D, Helfrich M, Benjamin N, Grabowski PS
1
994 Human osteoblast like cells produce nitric oxide and 22. Beresford JN, Gallagher JA, Russell RGG 1986 1,25-
express nitric oxide synthase. Endocrinology 135:330–336.
. Brandi ML, Hukkanen M, Umeda T, Moradi-Bidheni N, Bi-
anchi S, Gross SS, Polak JM, MacIntyre I 1995 Bidirectional
dihydroxyvitamin D3 and human bone-derived cells in vitro:
Effects on alkaline phosphatase, type I collagen synthesis and
proliferation. Endocrinology 119:1776–1785.
regulation of osteoclast function by nitric oxide synthase iso- 23. Green LC, Wagner DA, Glogowski J, Skipper PL, Wishnock
forms. Proc Natl Acad Sci USA 92:2954–2958.
. Ralston SH, Ho L-P, Helfrich Mp, Grabowski PS, Johnston
PW, Benjamin N 1995 Nitric oxide: A cytokine-induced reg- 24. Brito C, Naviliat M, Tiscornia AC, Vuillier F, Gualco G,
ulator of bone resorption. J Bone Miner Res 10:1040–1049.
. Pryor WA, Squadrito GL 1995 The chemistry of peroxynitrite:
A product from the reaction of nitric oxide and superoxide.
Am J Physiol 268:L699–L722.
. Halliwell B, Zhao K, Whiteman M 1999 Nitric oxide and 25. Vallette G, Tenaud I, Branka J-E, Jarry A, Sainte-Marie I,
peroxynitrite. The ugly, the uglier and the not so good. Free
Rad Res 31:651–669.
JS, Tannenbaum SR 1982 Analysis of nitrate, nitrite and (15N)
nitrate in biological fluids. Anal Biochem 126:131–135.
6
7
8
9
Dighiero G, Radi R, Cayota AM 1999 Peroxynitrite inhibits T
lymphocyte activation and proliferation by promoting impair-
ment of tyrosine phosphorylation and peroxynitrite-driven ap-
optotic death. J Immunol 162:3356–3366.
Dreno B, Laboisse CL 1998 Control of growth and differen-
tiation of normal human epithelial cells through the manipu-
lation of reactive nitrogen species. Biochem J 331:713–717.
. van der Vliet A, Eiserich JP, Kaur H, Cross CE, Halliwell B
1
996 Nitrotyrosine as a biomarker for reactive nitrogen spe- 26. Est e´ vez AG, Spear N, Pelluffo H, Kamaid A, Barbeito L,
cies. Methods Enzymol 269:175–184.
0. Lin K-T, Xue J-Y, Nomen M, Spur B, Wong PY-K 1995
Peroxynitrite-induced apoptosis in HL-60 cells. J Biol Chem
Beckman JS 1999 Examining apoptosis in cultured cells after
exposure to nitric oxide and peroxynitrite. Methods Enzymol
301:393–402.
1
1
2
70:16487–16490.
27. McBride AG, Borutait e´ V, Brown GC 1999 Superoxide dis-
mutase and hydrogen peroxide cause rapid nitric oxide break-
down, peroxynitrite production and subsequent cell death.
Biochim Biophys Acta 1454:275–288.
1. Est e´ vez AG, Radi R, Barbeito L, Shin JT, Thompson JA,
Beckman JS 1995 Peroxynitrite-induced cytotoxicity in PC12
cells: Evidence for an apoptotic mechanism differentially
modulated by neurotrophic factors. J Neurochem 65:1543– 28. Danziger RS, Zuckerbraun BS, Pensler JM 1997 Role of nitric
1
550.
oxide in the regulation of osteoblast metabolism. Plast Recon-
1
1
2. Gow AJ, Duran D, Malcolm S, Ischiropoulos H 1996 Effects
str Surg 100:670–673.
of peroxynitrite-induced protein modifications on tyrosine 29. Damoulis PD, Hauschka PV 1997 Nitric oxide acts in con-
phosphorylation and degradation. FEBS Lett 385:63–66.
3. Boota A, Zar H, Kim Y-M, Johnson B, Pitt B, Davies P 1996
junction with proinflammatory cytokines to promote cell death
in osteoblasts. J Bone Miner Res 12:412–421.
IL-1␤ stimulates superoxide and delayed peroxynitrite produc- 30. Armour KE, Armour KJ, Gallagher ME, Godecke A, Helfrich
tion by pulmonary vascular smooth muscle cells. Am J Physiol
71:L932–L938.
4. Estrada C, G o´ mez C, Mart ´ı n-Nieto J, De Frutos T, Jim e´ nez A,
MH, Reid DM, Ralston SH 2001 Defective bone formation
and anabolic responses to exogenous estrogen in mice with
targeted disruption of endothelial nitric oxide synthase. Endo-
crinology 142:760–766.
2
1
1
1
1
1
1
2
Villalobo A 1997 Nitric oxide reversibly inhibits the epidermal
growth factor receptor tyrosine kinase. Biochem J 326:369– 31. Aguirre J, Buttery L, O’Shaughnessy M, Azfal F, Fernandez
3
76.
de Marticorena I, Hukkanen M, Huang P, MacIntyre I, Polak
J 2001 Endothelial nitric oxide synthase gene-deficient mice
demonstrate marked retardation in postnatal bone formation,
reduced bone volume, defects in osteoblast maturation activ-
ity. Am J Pathol 158:247–257.
5. van der Vliet A, Hristova M, Cross CE, Eiserich JP, Goldkorn
T 1998 Peroxynitrite induces covalent dimerization of epider-
mal growth factor receptors in A431 epidermoid carcinoma
cells. J Biol Chem 273:31860–31866.
6. L o´ pez-Garcia MP, Sans-Gonz a´ lez SM 2000 Peroxynitrite gen- 32. Gowen M, MacDonald BR, Russell RGR 1988 Actions of
erated from constitutive nitric oxide synthase mediates the
early biochemical injury in short-term cultured hepatocytes.
recombinant human ␥-interferon and tumor necrosis factor ␣
on the proliferation and osteoblastic characteristics of human
trabecular bone cells in vitro. Arthritis Rheum 31:1500–1507.
4
66:187–191.
7. Beckman JS, Ye YZ, Anderson PG, Chen J, Accavitti MA, 33. Ye YZ, Strong MQ, Huang Z-Q, Beckman JS 1996 Autoan-
Tarpey MM, White CR 1994 Extensive nitration of protein
tyrosines in human atherosclerosis detected by immunohisto-
chemistry. Biol Chem Hoppe-Seyler 375:81–88.
tibodies that recognize nitrotyrosine. Methods Enzymol 269:
201–209.
8. Kaur H, Halliwell B 1994 Evidence for nitric oxide-mediated
oxidative damage in chronic inflammation. Nitrotyrosine in
serum and synovial fluid from rheumatoid patients. FEBS Lett
Address reprint requests to:
Artur J. de Brum-Fernandes, M.D., Ph.D.
Rheumatic Diseases Unit
3
50:9–12.
9. MacMillan-Crow LA, Crow JP, Kerby JD, Beckman JS 1996
Nitration and inactivation of manganese superoxide dismutase
in chronic rejection of human renal allografts. Proc Natl Acad
Sci USA 93:11853–11858.
0. de Brum-Fernandes AJ, Laporte S, Heroux M, Lora M, Patry
C, M e´ nard H-A, Dumais R, Leduc R 1994 Expression of
prostaglandin endoperoxide synthase-1 and prostaglandin en-
Faculty of Medicine
Universit e´ de Sherbrooke
3
001 Twelth Avenue North
Sherbrooke, Qu e´ bec J1H 5N4, Canada
doperoxide synthase-2 in human osteoblasts. Biochem Bio- Received in original form February 27, 2001; in revised form
phys Res Commun 198:955–960. September 26, 2001; accepted October 29, 2001.