JOURNAL OF BONE AND MINERAL RESEARCH Volume 7, Number 8, 1992 Mary Ann Licbert, Inc., Publishers

Nitroblue Tetrazolium Reduction and Bone Resorption by Osteoclasts In Vitro Inhibited by a Manganese-Based Superoxide Dismutase Mimic WILLIAM L. RIES, L. LYNDON KEY, JR., and RAMONA M. RODRIGUIZ

ABSTRACT Oxygen-derived free radicals are produced by osteoclasts. Oxygen radical formation occurs at the osteoclast/ bone surface interface. This location next to bone implies that oxygen radicals, including but not limited to superoxide, are needed for bone resorption. Compounds that scavenge superoxide are being developed as pharmaceutical agents to inhibit the damaging effects of oxygen radical formation on tissues. One such scavenger is the Desferal-manganese complex (DMnC). DMnC reduced the amount of formazan staining produced by the interaction of oxygen radicals with nitroblue tetrazolium (NBT) in both individual mouse cnlvarial osteoclasts in tissue explants and isolated osteoclasts. As a result of the reduced concentrations of oxygen radicals, DMnC inhibited bone resorption by calvarial explants and isolated osteoclasts. Superoxide dismutase (SOD) inhibited NBT reduction and bone resorption by isolated osteoclasts but to a lesser degree than DMnC. Inhibition of bone resorption in the isolated osteoclast system increased in parallel to the concentration of DMnC in cultures. Desferal without Mn had no effect on bone resorption by isolated osteoclasts. These results support the hypothesis that osteoclasts produce oxygen radicals as part of the process of bone resorption.

INTRODUCTION oxygen-derived free radicals. The reduction of nitroblue tetrazolium (NBT) to diformazan within the ruffled border"] suggests that osteoclasts produce superoxide at the site where bone resorption occurs. The localization of the formazan granules to this area suggests that oxygen radicals may be necessary for normal osteoclastic function. Furthermore, abnormal osteoclastic function has been associated with defective superoxide production in patients with malignant osteopelrosis.l','l Inhibition of bone resorption by compounds that reduce concentrations of superoxide would corroborate the functional importance of superoxide in bone resorption. Superoxide dismutase (SOD) and Desferal-manganese complex (DMnC), a superoxide scavenger,151reduce superoxide concentrations. SOD is unable to inhibit NBT reduction by osteoclasts in neonatal mouse calvarial organ cul-

0

STEOCLASTS PRODUCE

tures.f1)This failure to inhibit NBT reduction within calvarial osteoclasts may result from the large molecular weight of SOD (=4O,OOO) and the inability of SOD to diffuse across cell membranes. SOD inhibits NBT reduction in cultures of isolated osteoclasts, but this inhibition appears to result from a reduction in staining outside the osteoclastic cell membrane. DMnC (molecular weight =6CM) has been shown to cross cellular membranes. I 5 . w Thus, DMnC can traverse membranes, allowing it t o reach the osteoclast/bone interface or ruffled border space, where active bone resorption occurs. Several mechanisms for the role of superoxide in bone resorption have been proposed.lll One of the proposed mechanisms is that superoxide generated by osteoclasts next to the bone is a necessary biochemical event in bone resorption. To test this hypothesis, we employed systems using calvarial osteoclasts in situ and isolated osteoclasts resorbing prepared bone slabs.

Division of Endocrinology, Department of Pediatrics, Medical University of South Carolina, Charleston.

931

IUD ET AL.

932

MATERIALS AND METHODS Desferal-manganese complex preparation A 50 mM stock solution of a complex of Desferal (deferoxamine mesylate; Sigma Chemical Company, St. Louis, MO) and manganese, described as a low-molecular-weight ( ~ 6 0 0 )scavanger of superoxide within a cell,"' was prepared by dissolving 328 mg Desferal in 10 ml deionized water. MnO, (56 mg, Sigma) was added to this mixture and stirred for 4 h at room temperature. The resultant green complex was centrifuged at 1400 x g for 15 minutes and passed through a 0.22 pm filter (Millipore Corporation, Bedford, MA) to remove undissolved MnO,.

cells (American Type Culture Collection) were plated in PSO culture flasks (Corning, Corning, NY) and incubated for 48 h with Eagle's minimum essential medium (MEM, Sigma) containing 10% FBS and 100, ux), and 500 pM DMnC. The time to reach confluence was determined when the cells were plated at a density of SO00 cells per flask. The viability of osteoclasts isolated from the proximal periosteal surfaces of young adult Sprague-Dawley rat (150 g in weight) tibia"." was assessed after 48 h of incubation in MEM and 10% heat-inactivated FBS with or without 250 pM DMnC.

NBT reduction chemistry and DMnC exposure Determination of superoxide scavenger concentrations SOD (bovine erythrocyte, Sigma) at 10 ng/ml has been reported to inhibit bone resorption in neonatal mouse calvarial organ cultures. To determine the concentration of DMnC that would reduce superoxide by an amount equivalent to SOD, we performed the following analysis. Xanthine and xanthine oxidase (60 and 20 pM, respectively; Sigma), chemical reactants that together produce superoxide, were added to microtiter wells with 0.11 ml NBT solution [5 mg NBT (Sigma) dissolved in 60 pl dimethylsulfoxide (Sigma) and placed in 1.6 ml fetal bovine serum (FBS; Flow, McLean VA)] along with 0.9 ml of RPMI1640 (RPMI; Flow). To this mixture, either SOD at a final concentration of 10 ng/ml or DMnC at concentrations of 50, 250, and 500 pM was added and incubated for 15 minutes. Optical density was determined using a V, kinetics microplate reader (Molecular Devices, Menlo Park, CA) at 595 nm (the peak absorbance of diformazan stain). The means of five replicates at each concentration were taken as the optical density (OD) for a given determination (Table 1). DMnC at 500 pM showed greater inhibition of NBT reduction than 10 ng/ml of SOD, and this DMnC concentration was used in the organ culture experiments. Isolated osteoclast cultures were treated with a dose range of 50, 250, and 500 pM DMnC.

To determine an effect of 500 pM DMnC on non-superoxide-induced reduction of NBT. imprints of freshly sliced rat liver tissue were stained for succinic dehydrogenase activity using NBT as a histochemical coupling agent."," Imprints were made onto glass slides, air dried at 4"C, and fixed in cold 10% neutral buffered formalin for 5 minutes. After a brief wash in cold H,O, the imprints were incubated in medium containing 5 mg NBT dissolved in 0.2 M disodium succinate (Sigma) and 2.2 M Tris (Sigma), at pH 7.4 and 37°C with or without DMnC.

Analysis of calvarial cultures

The 4-day-old C57/b16 mice (Jackson Laboratories, Baultenbor, ME) received an intraperitoneal (IP) injection of 1.O pCi 4sCaCl, (45Ca; Amersham, Arlington Heights, IL). On day 4 after radiolabeling, calvarial bones (frontal and parietal bones) were harvested into RPMI and halved at the sagittal suture. Each half was placed on a permeable collagen-treated microporous membrane of a 16 mm well in a Transwell plate (Costar, Cambridge, MA). Calvariae were preincubated in 2 ml RPMI with 10% FBS for 24 h at 37°C in 5% CO, and air. After the preincubation period, one of a pair of calvarial bone halves was incubated in medium consisting of 2 ml RPMI (without phenol) and 10% FBS with no added parathyroid hormone. The other halfcalvaria was incubated in similar medium or in media containing the following test agents: (1) 3 pg/ml of parathyCell viability to DMnC exposure roid hormone [recombinant human PTH-( 1-34); Sigma], To assess the effects of DMnC on cellular viability using (2) PTH and 10 ng/ml of SOD, (3) PTH and 500 pM the trypan blue exclusion method, U20S osteosarcoma DMnC, and (4) 250 pM DMnC without PTH supplementation. The calvarial halves were incubated in 5% CO, and air at 37°C for 48 h. At the end of the incubation, calTABLE1. EFFECTOF SOD AND DMNCON NBT REDUCTIONvariae were transferred to a solution of freshly prepared NBT (as described earlier) and incubated at 37°C in 5% BY THE XANTHINE-XANTHINE OXIDASEREACTION C 0 2 and air. After a 1 h incubation. calvariae were fixed in FOR SUPEROXIDE 2.5% glutaraldehyde in 0.1 M cacodylate buffer for 30 NBT reduction minutes. Densitometric measurements were made of indiGroup Concentration (OD units) vidually stained cells in situ on the endocranial surface according to published methods,"' and the tissue was dehyControl 0.763 drated with ethanol and embedded in epoxy resin. Thick 0.524 SOD 10 ng/ml sections were prepared and stained with toluidine blue for DMnC 50 pM 0.600 light microscopic study. 0.492 250 pM To assess bone-resorptive activity, the amount of "Ca 0.254 5 0 0 pM released into the media from the labeled calvariae in 48 h

933

SUPEROXIDE DISMUTASE MIMIC INHIBITS OSTEOCLASTS was determined using the method described by Ries et al.i91A 1 ml aliquot of incubation medium was removed before NBT staining and counted in a Beckman LS lOOC scintillation counter. Each calvarial half and the remaining incubation medium were extracted in 10% trichloroacetic acid (Sigma) for 24 h at 4°C and the calvarial digests counted. The percentage of resorption was determined from the ratio of the media counts per minute (cpm, 2 times the cpm in the aliquot of the medium) to the total cpm in the calvaria before calcium release (cpm in the medium plus cpm from the digest). The results of the resorption assay are reported as the ratios of the experimental to the control percentage of resorption.

crographs onto a digitizing tablet (Kurta, Phoenix, AZ) with input to a microcomputer-based image analysis system (CUE 2, Opelco, Washington, DC) to calculate the plan area (jd) of each resorption bay. Another group of bone slabs was stained by the NBT reaction and fixed as described for the calvarial preparations. The NBT staining intensity of individual osteoclasts was measured (OD) by microdensitometry using an inverted microscope (Olympus) equipped with a CCD camera. The camera image was displayed on a high-resolution monitor (Sony), stored in a microcomputer (ALR, Inc., Irvine, CA), and analyzed using densitometry software (Cue 2d, Opelco).

Analysis of isolated osteoclasts

Statistical analysis

Using a modified version of the osteoclast isolation method described by Thomson et al.,'''] femora and tibiae were harvested aseptically from 36-h-old neonatal rats (Charles River, Wilrnington, MA) after pentobarbital euthanasia. Cartilage and soft tissues were removed from the bones. Bones (16 bones per 30 rnm well) were split longitudinally and minced with a knife in 2.5 ml HEPESbuffered medium 199 (M-199; Sigma). The bone fragments were pipetted vigorously and allowed to settle for 10 s before a 1 ml cellular suspension was placed into a 16 mm well containing two bovine cortical bone slabs (see bone slab preparation) in 1 ml M-199. Cells and bone slabs were incubated together for 1 h at 37°C in 5 % CO, and air to allow the adherence of osteoclasts to the bone surfaces. Bone slabs (5 x 5 x 0.2 mm in dimension) were cut from specimens of adult bovine Femora using an Isomet diamond saw (Buehler, Evanston, IL). The slabs were cleaned ultrasonically in distilled H,O for 20 minutes and treated with 3.0 pg/ml of penicillin, streptomycin (Sigma), and Fungizone (Flow) for a total of 1 h, with a change to fresh antibiotic solution every 20 minutes. The slabs were then transferred to wells for the osteoclastic adherence procedure. After the 1 h adherence phase, each bone slab was rinsed vigorously in MEM for 20 s. Slabs were incubated for 36 h in 2 ml MEM plus 10% heat-inactivated FBS and 1 % penicillin and streptomycin with or without one of the following test agents: 500 pM Desferal, 10 ng/ml of SOD, 0.5 pg/ml of thyrocalcitonin (hCT), and DMnC (SO, 250, and 500 pM). After culturing, one set of bone slabs was either fixed in 2.5% glutaraldehyde in 0.1 M cacodylate buffer for 30 minutes or brushed with a tine paintbrush and ultrasonicated in 0.25 M N H 4 0 H for 10 minutes and fixed. The bone slabs were briefly stored in cold 0.1 M cacodylate buffer, dehydrated through a graded series of ethanol, and treated with two changes of hexamethyldisilazane (Polysciences Inc., Warrington, PA). After the sputter-coating step, the bone slabs were examined in a Phillips 515 scanning electron microscope. Resorption bays were photomicrographed and analyzed. Resorption bays in either singular or complex configurations were identified as having both a fibrillar base without canalicular channel openings and a well-defined rim. The borders of the bony concavities and magnification bar lengths were traced from photomi-

Data are expressed as the mean standard error of the mean (SEM). Differences between experimental versus control and between experimental groups were determined using Student's 1-test. Analysis of variance (ANOVA) was used to analyze the DMnC dose-response curve.

*

RESULTS Calvarial osteoclastic NB T reduction NBT reduction by PTH-stimulated calvarial osteoclasts in culture was inhibited by 500 pM DMnC (Table 2). Figure l a and b shows dense NBT staining of PTH-stimulated calvarial osteoclasts in the absence of DMnC, but osteoclasts in the presence of DMnC (500 pM) are virtually unstained (Fig. l c and d). DMnC at 250 pM inhibited basal (non-PTH-stimulated) NBT reduction by osteoclasts to 40% of control calvarial osteoclasts (experimental, 0.395 f 0.039 versus control, 0.654 f 0,019; O D units, mean + SEM, N = 10, p < 0.WOl).

Calvarial bone resorption PTH-stimulated bone resorption was inhibited by 500 pM DMnC (Table 2). Addition of SOD to the stimulated calvarial cultures had no inhibitory effect on 45Ca release TABLE 2. EFFECTOF SOD AND DMNC ON PTH-STIMULATED BONERESORPTION AND NBT REDUCTION BY CALVARIAL OSTEOCLASTS ~

Group PTH only (3 pg/ml) PTH + SOD (10 ng/ml) PTH + DMnC (500 pM)

~~~

"Ca release

1.3

* O.la

1.3 f 0.2 0.6 f 0 . 1 ~

*

NBT reduction 1.4 f 0.2b 1.4 f 0.2 0.9 f 0.ld

'Values are the mean ratios SEM,N = 3 pairs of half-calvaria, treated versus control bone ratio of the percentage "Ca release. bValues are the mean ratios f SEM,N = 8-10 individual osteoclasts, treated versus control osteoclast ratio of the OD unit values. 5ignificantly different from the PTH-stimulated control group, p < 0.01. dSignificantly different from the PTH-stimulated control group, p < 0.05.

RIES ET AL.

934

8

FIG. 1. Serial sections (a and b) through a neonatal mouse calvaria stained with NBT. (a) The osteoclast (arrows) was counterstained with toluidine blue to show the presence of multiple nuclei. (b) The same osteoclast (arrows) with no counterstain. Staining in b is due to the formazan granules resulting from NBT reduction. Sections through a calvaria (c and d) cultured in media containing DMnC (500 pM) and stained with NBT. (c) The osteoclast (arrows) was counterstained. (d) The same osteoclast (arrows) was not counterstained. Note the absence of formazan granules in c and d, suggesting DMnC inhibited NBT reduction. (Original magnification: x 1ooO.) compared to positive control cultures stimulated with PTH Isolated osteoclastic bone resorption only (Table 2). In a separate experiment, basal bone reThe bone-resorptive activity of isolated osteoclasts was sorption by non-PTH-stimulated control calvarial cultures inhibited by 500 pM DMnC and 10 ng/ml of SOD as indiwas not inhibited by 250 pM DMnC [experimental percentage resorption (DMnC only) to control ratio, 0.97 f cated by a reduction in the resorption bay planar area 0.09, N = 8 versus control (no DMnC) to control ratio, (Table 3). However, DMnC produced a greater reduction (p < 0.03) in resorption bay area than SOD. Resorption 0.90 f 0.06, f S E M , N = 81. bay formation was completely inhibited by 0.5 pg/ml of hCT (Table 3). The inhibitory effect of DMnC on the bone-resorptive activity by isolated osteoclasts was dose dependent, with significant inhibition at 250 pM and no Isolated osteoclastic NBT reactivity significant inhibition at 50 pM DMnC (Table 4). In a sepaNBT reduction by osteoclasts isolated on bone slabs was rate experiment, 5 0 0 pM Desferal without Mn had no efinhibited by 10 ng/ml of SOD and 500 pM DMnC (Table fect on the resorptive activity of isolated osteoclasts com3). However, the reaction inhibition by DMnC was two- pared to control cultures without Desferal (data not fold greater than the inhibition by SOD. shown).

SUPEROXIDE DISMUTASE MIMIC INHIBITS OSTEOCLASTS TABLE 3. EFFECTOF SOD, DMNC, AND HCT ON NBT AND BONERESORPTION BY REDUCTION ISOLATEDOSTEOCLASTS

93s

Cellular viability and NBT reactivity

Using the trypan blue exclusion method, U20S osteosarcoma cells survived a 48 h exposure to 100, 200, and 500 Resorption NBT reduction pM DMnC in culture to a level of 99, 98, and 95% viabilGroup (OD units) bay area ity, respectively. Culture flasks of DMnC-treated (100 and 200 pM DMnC) and untreated osteosarcoma cells reached 5060 f 1309a 0.507 0.019b Control confluence in 6 and 5 days, respectively. 0.413 f 0.011d SOD (10 ng/ml) 1074 f 444c Isolated tibia1 periosteal osteoclasts incubated in MEM 0.282 f 0.028B-h with 250 pM DMnC or MEM only for 48 h were % and DMnC (500 pM) 57 f 14e.f hCT (0.5 pg/ml) 0' 89% viable, respectively, using trypan blue exclusion as a aValues are the mean plan areas, pm* f SEM,N = 10-18 indi- criterion. NBT reduction to formazan by hepatocyte succinic devidual resorption bays. bValues are the mean OD units f SEM, N = 12-15 individual hydrogenase activity was not inhibited by 500 pM DMnC osteoclasts. in the reaction medium (DMnC. 0.363 0.017 and withsignificantly different from the control group, p < 0.01. 0.399 & 0.030; O D units, mean SEM. N = out DMnC, dSignificantly different from the control group, p < 0.05. 8, p < 0.2). These data show no effect of DMnC on NBT CSignificantly different from the control group, p < 0.003. reduction independent of the ability of DMnC to reduce rSignificantly different from the SOD group, p < 0.03. rsignificantly different from the control group, p < O.OOO1. superoxide concentrations.

*

*

*

hSignificantly different from the SOD group, p < 0.001. Compleie inhibition of resorption bay formation.

DISCUSSION TABLE 4. EFFECTOF DIFFERENT CONCENTRATIONS OF DMNC ON THE RESORPTIVE ACTIVITY OF ISOLATED OSTEOCLASTS Group

Resorption bay area ~~

Control, media alone DMnC, pM 50 250 500

~

894.1

* 59.4a.b

436.8 f 20.2' 411.9 15.8d 227.5 17.6

* *

aValues are the mean plan areas, pm2 f SEM, N 10-1s individual resorption bays. hANOVA performed comparing the control group to the different DMnC concentrations, p < =

0.01, F = 5.58.

'Not significantly different from the control group. dSignificantly different from the control group, p < 0.04. CSignificantly different from the control group, p

< 0.01.

Both DMnC and SOD reduced levels of NBT reduction and bone resorption in isolated osteoclasts. DMnC inhibited NBT reduction and bone resorption by calvarial osteoclasts, but SOD did not. DMnC did not reduce bone cell viability. Therefore, the inhibition of isolated osteoclastic NBT reduction and bone resorption by DMnC suggests that a reduction in superoxide within the osteoclast resulted in reduced bone resorption. Osteoclastic superoxide concentrations were lowered by SOD and DMnC. SOD reduces the concentration of superoxide by catalyzing its conversion t o hydrogen peroxide; DMnC scavenges superoxide by reacting it with Mn'*.(61 The formation of DMnC from Desferal and MnO, involves reduction of the Mn4' t o Mn3* and the concomitant oxidation of a hydroxamate ligand group on the Desferal molecule.(5) DMnC contains one atom of Mn'* for each molecule of Desferal. (') DMnC removes superoxide from solution by reacting it with the Mn3+ t o form manganese dioxide. The complex is stable enough to survive the intracellular environment and retains SOD-like activity in the presence of serum albumin.(6,11) Less than 20% of the total manganese in DMnC is present as Mn2+."I Mn''Desfera1 Z= Mn3* + Desferal 2Mn'* + 2H,O * Mn" + MnO,

In addition to decreasing resorption bay area with DMnC treatment, the bases of experimental resorptive bays were frequently shallower than controls and showed more highly fibrillar collagen as opposed to a more compact fibrillar collagen at the base of control resorption bays (Fig. 2). The scanning electron microscope ultrastructural characteristics of osteoclasts on bone slabs after exposing the cells t o 500 pM DMnC, which include cytoplasmic pseudopodia and membrane surface microvilli, were similar in appearance to control osteoclasts (Fig. 3). The presence of these structures indicates that a relatively high concentration of DMnC in the culture media had no effect on osteoclastic morphologic integrity.

+

4H+

Mn" and Mn" alone have no effect on reducing superoxide formation. (') Furthermore, nonenzymatically linked manganese complexes have been shown not to catalyze the formation of highly reactive hydroxyl radicals, commonly associated with increased intracellular concentrations of copper and iron ions."2) Hence, the manganese cations are noncytotoxic. Differential adherence provides enrichment of osteoclasts. but the bone cells isolated in our experiments included a variety of other cells. Although not characterized in this study, we presume these cells t o include fibroblasts, marrow stromal cells, bone marrow macrophages, osteo-

FIG.2. (a) Complex resorption bay made by isolated osteoclast(s) in control medium (minus DMnC) over a 48 h incubation. (b) A small resorption bay made by an osteoclast over a period of 48 h in medium containing DMnC ( 5 0 0 pM). Note the resorption bay in a has a relatively compact fibrillar base compared to the more loosely constructed fibrillar base in b. A portion of an osteocyte lacuna with canalicular channel openings appears at the top of the resorption bay in b. (Original magnification: a, x 2165; b, x 4450.)

SUPEROXIDE DISMUTASE MIMIC INHIBITS OSTEOCLASTS

937

FIG. 3. (a) Control (minus DMnC) osteoclast in culture for 48 h partially covering a resorption bay. Another resorption bay appears at the top left in a. (b) Osteoclast cultured in 500 pM DMnC for 48 h and partially covering a shallow, loosely fibrillar resorption bay at the lower right. The osteoclasts have microvilli and pseudopodal cytoplasmic extensions indicating morphologic integrity during the incubation period. (Original magnification: a, x 1435; b, x 1980.)

RlES ET AL. clast precursors, and osteoblasts. Isolated osteoclastic tion. Independent of the the possible mechanistic implicabone resorption may be dependent on conditions produced tions, our data demonstrate that cellular permeability by these additional adherent cells. ( I o 1 Although our iso- should be considered when selecting an inhibitor of superlated osteoclast model proves that authentic osteoclasts are oxide availability in bone resorption. In summary, we have shown that lowering intracellular inhibited by the DMnC, the precise mechanism of this inhibition cannot be deduced from these experiments. An at- concentrations of superoxide within osteoclasts using tractive hypothesis to explain the role of oxygen radicals in DMnC inhibits bone resorption. Second, we find that bone resorption is that oxygen radicals help degrade bone DMnC is superior to SOD in inhibiting bone resorption. matrix. This mechanism would be analogous to bacterial This superiority appears to result from the greater cellular or tissue degradative effects resulting from leukocyte permeability of DMnC. Previously, inhibition of bone resuperoxide production. ("1 If direct protein destruction is sorption by reducing levels of superoxide has been demonthe major mechanism of the superoxide effect, then it strated in organ culture. (11 The present studies extend these would be likely that DMnC reduces the concentration of observations to isolated osteoclasts, demonstrating that the oxygen radicals adjacent to the bone surface, limiting bone resorption by osteoclasts is inhibited by reducing the degradation of matrix proteins. The decrease in the levels of superoxide. Furthermore, the differential effects level of intracellular NBT reduction products quantified in of SOD and DMnC on intracellular NBT concentrations, isolated osteoclasts treated with DMnC and SOD are con- taken together with the known mechanisms of leukocyte sistent with this hypothesis. In addition, significantly protein degradation, suggest that superoxide and the cassmaller but demonstrable resorptive bays are identified in cade of oxygen-derived free radicals generated from supertreated isolated osteoclastic cultures. These bays are shal- oxide are involved in the direct degradation of bone. low and have more highly fibrillar collagen bases, compared to deeper, smoother bays found in control cultures. The highly fibrillar bases suggest that mineral has been reREFERENCES moved from the surface of the bone without adequate degradation of matrix. Thus, our data are consistent with the 1 Key LL, Ries WL, Taylor RG, Hays BD, Pitzer BL 1990 Oxygen derived free radicals in osteoclasts: The specificity hypothesis proposed, but other mechanisms within the and location of the nitroblue tetrazolium reaction. Bone 11: osteoclast or outside the cell could explain the observed re115-119. duction in bone resorption. 2. Garrett IR, Boyce BF, Oreffo ROC, Bonewald L, Poser J , Garrett et al.I2] proposed a variety of potential mechaMundy GR 1990 Oxygen-derived free radicals stimulate nisms of superoxide action on bone resorption. These inosteoclastic bone resorption in rodent bone in vitro and in clude potential roles of superoxide within the ruffled borvivo. J Clin Invest 85:632-639. der space but not involving a direct biochemical attack on 3. Beard CJ, Key LL, Newburger PE, Ezekowitz AB, Arceci R, matrix proteins. Such mechanisms could include altering Miller B, Proto P, Ryan T, Anast C, Simons ER 1986 Neuthe activity of proteinases or their inhibitors. Proposed trophil defect associated with malignant infantile osteopetroextracellular mechanisms are stimulation of osteoclastic sis. J Lab Clin Med 108:498-505. numbers, alterations in activators or inhibitors of osteo- 4. Reeves JD. Augus CS, Humbert JR, Weston WL 1979 Host defense in infantile osteopetrosis. Pediatrics 64:202-206. clastic function, and effects on cells producing these osteo5. Beyer WF, Fridovich 1 1989 Characterization of a superoxide clast-active substances. There is evidence that both intradismutase mimic prepared from desferrioxamine and MnO,. cellular and extracellular mechanisms coexist. Arch Biochem Biophys 271:149-156. The dichotomy of effects between DMnC and SOD may 6. Rabinowitch HD, Privalle CT. Fridovich 1 1987 Effects of result from this combination of mechanisms. SOD inhibits paraquat on the green alga Dunaliella salina: Protection by bone resorption in the pit assay but not in stimulated calthe mimic of superoxide dismutase, Desferal-Mn(1V). Free varial osteoclasts. This appears to be because SOD does Radic Biol Med 3:125-131. not readily cross cellular membranes but DMnC 7. Ries WL. Gong JK 1982 A comparative study of osteoclasts: Differences in permeability would explain why DMnC deIn situ versus smear specimens. Anat Rec 203:221-232. creases the level of NBT reduction within the intracellular 8. Burstone MS I%2 Enzyme Histochemistry and Its Application in the Study of Neoplasms. Academic Press, New York. compartment of the osteoclast but SOD does not. Inhibip. 513. tion of bone resorption by isolated osteoclasts was ob9. Ries WL, Seeds MC, Key, LL 1989 Interleukin-2 stimulates served in the presence of both DMnC and SOD; however, osteoclastic activity: Increased acid production and radioacDMnC was markedly superior. The superiority of DMnC, tive calcium release. J Periodont Res 24:242-246. which readily crosses the cellular membrane, to SOD, 10. Thomson BM, Saklatvala J , Chambers TJ 1986 Osteoblasts which does not, suggests an intracellular mechanism. mediate interleukin 1 stimulation of bone resorption by rat Nonetheless, SOD reduced resorption bay size, although osteoclasts. J Exp Med 164:104-112. less effectively than DMnC. This SOD effect suggests that 11. Darr D. Zarilla KA, Fridovich I 1987 A mimic of superoxide extracellular superoxide also affects bone resorption. Aldismutase activity based upon desferrioxamine B and mangaternatively, the differential effect may result solely from a nese(1V). Arch Biochem Biophy 258:351-355. relative inability of SOD to reach the site of bone resorp- 12. Cheton PLB, Archibald FS 1988 Manganese complexes and

SUPEROXIDE DISMUTASE MIMIC INHIBITS OSTEOCLASTS the generation and scavenging of hydroxyl free radicals. Free Radic Biol Med 9325-333. 13. Babior BM 1984 The respiratory burst of phagocytes. J Clin Invest 75599-601.

Address reprint requests to:

W.L. Ries Room 453 Basic Sciences Building Medical University of South Carolina I71 Ashley A venue Charleston. SC 29425-3307 Received for publication August 27, 1991; in revised form March 2. 1992; accepted March 17, 1992.

939

Nitroblue tetrazolium reduction and bone resorption by osteoclasts in vitro inhibited by a manganese-based superoxide dismutase mimic.

Oxygen-derived free radicals are produced by osteoclasts. Oxygen radical formation occurs at the osteoclast/bone surface interface. This location next...
3MB Sizes 0 Downloads 0 Views