JOURNAL OF BONE AND MINERAL RESEARCH Volume 6, Number 6, 1991 Mary Ann Liebert, Inc., Publishers

Superoxide Dismutase and Catalase Activities in the Growth Cartilage: Relationship Between Oxidoreductase Activity and Chondrocyte Maturation H. MATSUMOTO,' S.F. SILVERTON,' K. DEBOLT,' and I.M. SHAPIRO'

ABSTRACT Superoxide dismutase (SOD) and catalase are enzymes that protect cells from radical attack. Catalase disproportionates hydrogen peroxide, and SOD is an oxidoreductase that serves to dismutate the superoxide anion. The objective of this communication was to measure the activity of these disproportionating enzymes in the chick tibia1 growth cartilage and to relate enzyme activity to chondrocyte maturation and tissue calcification. Analytic techniques were optimized for the measurement of both enzymes; p:articular care was taken to ensure that the values obtained were due to SOD and catalase, not to the presence of other oxidases or contaminants. Catalase and SOD had similar profiles of activity in cartilage. For both enzymes, the highest levels of activity were observed in premineralized cartilage; as chondrocytes matured there was a progressive decrease in the activity of SOD and catalase. Comparison of chondrocyte SOD activity with nonmineralizing tissues indicated that the activity of cultured cartilage cells was low. We also measured the SOD activity of avascular chondrodystrophic cartilage and found it to be less than that of proliferating cartilage. When cartilage was electrofocused, three SOD isozymes were detected. The PI of the major isozyme corresponded to the copper-zinc isoform. We suggest that the observed changes in enzymatic activity are dependent on a number of cartilage-specific factors that include the vascular supply, the local production of oxygen radicals by chondrocytes, and the oxidative state of the tissue.

INTRODUCTION

E

that generate and disproportionate reactive oxygen species play a significant role in protecting the cell against radical attack. The importance of these enzymes has been documented in a wide range of tissues"'; however, little is known of their role in connective tissues, such as bone or cartilage. Articular cartilage in particular is subjected to oxidative damage during inflammation, and there is some evidence to indicate that the severity of the attack can be mitigated by dismutating enzymes, antioxidants, and radical scavengers.(2)In biochemical terms, the presence of oxygen radicals can influence collagenase activity, hyaluronate synthesis, and ground substance polyNZYME SYSTEMS

merization. Moreover, collagen and collagen breakdown products can trigger superoxide release. 1 8 , 9 ) Disproportionating enzymes that may be present in cartilage are superoxide dismutase (SOD, EC 1.15.1.1) and catalase (EC 1.11.1.6). Catalase disproportionates hydrogen peroxide and protects membrane lipids and proteins from attack by peroxy radicals. SOD is an oxidoreductase that serves to dismutate superoxide anions that are generated during oxidative metabolisni and in response to noxious A single immunohistochemica1 study of SOD in the growth plate showed that the enzyme is present in the tissue and that the concentration of the immunoreactive protein was elevated in hypertrophic cartilage.(") The objective of this communication was to measure the ~~

-

'Department of Oral Surgery, School of Dental Medicine, Showa University, Tokyo, Japan. *Department of Oral Medicine and Skeletal Biology Research Group, School of Dental Medicine, University of Pennsylvania, Philadelphia.

569

MATSUMOTO ET AL.

570

location and activity of SOD and catalase in the growth cartilage. We chose to study this tissue because it shares a number of structural and biochemical characteristics with both articular cartilage and bone. Moreover, as the cartilage cells mature and the structure of the cartilage under~ . ~are ~) goes a series of well-characterized ~ h a n g e s , ' ~we able to examine the influence of vascularity and mineralization on the activities of these enzymes.

Catalase activity

Catalase activity was determined by a modification of the method of Johansson and Borg.('O) Sections of cartilage were minced and washed extensively to remove erythrocyte contaminants and then extracted for 16 h at 4°C with 0.1% Nonadet P-40 (NP-40, Sigma) in 25 mM KH,PO, buffer (pH 7.0). Aliquots of solution (50-100 pl) were used for the catalase assay and for protein assays. The sample was made up to 100 pl volume with 25 mM phosphate buffer. The assay mixture, consisting of sample, MATERIALS AND METHODS 50 pl methanol (chromatographic purity, Fisher), and 10 p1 Tissue and cell samples of 0.3% hydrogen peroxide (wt/vol), was incubated at Cartilage was isolated from the epiphysis of 8-week-old 15°C for 30 minutes. The reaction was halted by adding white rock chicks. The method that we employed has been concentrated base (7.8 M KOH). Following incubation for a described in detail elsewhere.'") Briefly, following the further 15 minutes at 20°C with 100 liters of 34.2 mM Purdeath of the chick by cervical dislocation, the proximal pald solution in HCI (adequate to neutralize the KOH previtibial growth cartilage was removed and washed in 50 mM ously added), the absorption at 550 nm was determined. phosphate buffer, pH 7.8, on ice. The cartilage was then Catalase standards (Sigma) from 0.06 to 0.55 IU were prefrozen in liquid nitrogen and transverse sections 50 pm pared from lyophilized enzyme for each assay. To assess the level of hemoglobin in the cartilage slices after washing, carthick collected using a Harris CTD cryostat at -40°C. These sections were utilized for histologic identification or tilage suspensions were treated with Drabkin's reagent and for biochemical analysis. We use the same procedure to absorption determined spectrophometrically at 559 nm. collect and analyze the tibial cartilage of chicks that were chondrodystrophic. The growth plates of these chicks have RESULTS been extensively characterized by ourselves and others.(1S) For preparation of chondrocytes, epiphyseal tissue was diTo measure the actual SOD activity in selected areas of gested with collagenase and the cells isolated using conventional tissue culture techniques. (I6) From each identified the growth plate, considerable time was spent developing region of the growth plate, tissue sections of 50 pM were the extraction and analytic protocols. The effects of matrix homogenized with a Polytron and then sonicated twice for components, as well as reagents used in the reaction mix20 s. The tissue suspension was centrifuged at 3000 x g for ture, were evaluated. The optimal concentrations of 20 minutes. Supernatant was used for SOD assay, and the EDTA and Triton X-100 were 500 pM and 0.1%, respecpellet was lyophilized for dry weight determinations. Pel- tively. We found that the presence of chondroitin sulfate, lets contained no detectable SOD activity. For measure- cartilage proteoglycan, and hyaluronate in the reaction ment of chondrocyte enzyme activity, cultured cells were mixture had no effect on SOD activity. Enzyme contamihomogenized with a Polytron, centrifuged at 3000 x g for nation due to the presence of blood cells was determined. 20 minutes, and the supernatant collected. The sediment In each section, very low levels of hemoglobin (< 1 pmol/ mg dry weight) were detected. There was no significant difwas used for protein determinations.(I4) ference in the hemoglobin content of sections isolated from the nonmineralized regions of the plate. Some hemoSuperoxide dismutase measurements globin was present in the calcified cartilage-bone-forming To distinguish between nonspecific oxidoreductase and region, and catalase activity of this region of the cartilage SOD activities, the ferricytochrome c was acetyIated.(I') This is not provided. To validate the analytic procedure and to ascertain procedure maintained dismutase activity but reduced by more than 95qo the activities of other cytochrome c oxidases and re- whether catalase and SOD activities were proportional to ductases. In this way, it was possible to determine the total cy- the actual amount of cartilage tissue that was used for each tochrome c oxidase activity as well as the SOD activity. Super- assay, dose-response curves were constructed (Fig. 1 ) . For oxide dismutase activity was measured by observing the de- these studies, the activity of aliquots of homogenized crease in absorbance of ferricytochrome c at 550 nm using a growth cartilage of different volumes was determined. Figxanthine-xanthine oxidase system to generate the superoxide ure 1 shows that the activities of the two enyzmes are proanion at a constant rate."') Pure SOD (Sigma) was used as portional to the volume of the cartilage sample. In addia standard for these assays. Cartilage superoxide dismutase tion, we examined procedures for extracting catalase from was also characterized by isoelectric focusing of the protein cartilage. It was noted that the use of 0.1% NP-40 yielded in 4% acrylamide and 2% ampholines for 4 h at 400 V; the more catalase activity than extraction with Triton-X ( 1 or gels were reduced with riboflavin in the presence of N,N,N,N 0.1%) or 1% NP-40; further, we noted that the long extetramethylethylenediamine (TEMED) and then allowed to traction period was preferable to the recommended short reoxidize in air. Superoxide formed in this way reduced ni- incubation time (0.5-4 h). trobluetetrazolium (NBT) to form blue insoluble formazan. We examined the inhibitory effect of cyanide and azide Assays were performed in the presence and absence of the on dismutase activity. Figure 2 shows that in both preminSOD inhibitors cyanide (2 mM) and azide (10 ITM).("-'~)eralized (RC-PC) and mineralizing (HC-CC) cartilage,

571

SOD AND CATALASE IN CARTILAGE

50

50

0

40

30

o x

g-

9> 20 "0,-.

20

5.

10

10

u"

0

FIG. 1. Dose-response curve for SOD and catalase activity in cartilage. A sample of cartilage was prepared for analysis as described in the text. The SOD and catalase activities of serially diluted aliquots (10-50 liters) of this extract were then measured.

100 x

w

.- 7 5 '5 4

a,

0 N

0 0

0

50

a,

-0 .N

U

a, -

U .N 0

4

a,

.-" N 7-7

0

0

4

X 25

0 RC-PC

HC-CC

FIG. 2. Inhibition of cartilage SOD activity by azide and cyanide. Extracts of cartilage from :he resting proliferative (RC-PC) and hypertrophic and calcifying cartilage (HC-CC) regions of the epiphysis were prepared. The extracts were treated with azide (10 mM) and cyanide (2 mM) and the percentage change in activity determined.

both of these compounds inhibit SOD activity. In both preparations, cyanide is a more effective inhibitor than azide. Together, these agents almost completely inhibit SOD activity. To assess the contribution of oxidase enzymes other than SOD, we measured the reduction of cytochrome c by cartilage extracts from sequential sections of the growth plate. Using the reduction of nonacetylated cytochrome c as a measure of the total oxidoreductase activity, the highest activity values are seen in the resting cartilage zone of the growth plate (Table 1). There is a three to fourfold de-

crease in total activity as the tissue matures and calcifies. It is likely that much of this activity is due to cytochrome c oxidase. When acetylated cytochrome c was used as a substrate for SOD, much lower levels of activity were seen. These values ranged from 0.76 to 0.18 unitslmg dry weight. As Table 1 indicates, SOD accounts for 7-14% of total oxidase activity. An SOD activity gradient exists in the growth plate. Figure 3 shows that as chondrocytes proliferate and hypertrophy there is a progressive decrease in SOD activity. The SOD activity of cartilage obtained from the epiphysis of

572

MATSUMOTO ET AL. TABLE1. SOD AND TOTALOXIDASE ACTIVITY IN SEQUENTIAL SECTIONS OF CHICKGROWTHCARTILAGE^ Zone

Total activity fCJ/mg dry weight)

SOD activity (U/mg dry weight)

To SOD activity fSOD/total activity)

5.16 3.22 2.69 2.06 1.75

0.76 0.45 0.32 0.21 0.18

14.72 13.97 11.89 10.19 10.29

1 2 3 4 5

aA chick growth plate was sectioned in a microtome. Successive transverse sections 1-5, commencing in resting proliferating cartilage, were homogenized, and the SOD and total oxidase activity were compared. Total activity was measured by observing the reduction of ferricytochrome c at 540 nm. SOD activity was determined using acetylated ferricytochrome c as a substrate. Values for both SOD and total activity were calculated using E ~ ~ ~ , =, , ,83, x 103.

2.0

cline in catalase values. Although the rate of decline is not as steep as SOD, cartilage in the resting-proliferating region has a higher activity than hypertrophic cartilage. Isoelectric focusing permitted the separation of SOD isoforms in growth plate chondrocytes (Fig. 5a) and comparison of these isoforms with those of chick liver (Fig. 51). Two major bands were present in cartilage (shown in the figure as a lightly stained areas). The major SOD bands focused at pH 6.9 and 6.7 (Fig. 5a). There was evidence of a minor component with a pl at 8.3; this isoform was absent from liver. In the presence of cyanide there was considerable loss of the 6.7 band and a decrease in the component that focused at pH 6.9 (Fig. 5b). Azide treatment (not shown) resulted in loss of the pH 6.9 band. Comparison of staining intensities indicated that the profile of isoforms in the precalcifying and calcifying regions were very similar.

r

1.5

0.0

,

I

I

I

I

I

FIG. 3. SOD activity of chick epiphyseal cartilage. SOD activity in sections of histologically identified cartilage were measured using acetylated cytochrome c: I and 11, resting proliferating cartilage; Ill and IV, hypertrophic cartilage; IV and V, calcified cartilage and bone. Error bars show standard error of the mean. *Significantly different from region I (p < 0.001).

tibial chondrodystrophic chicks was also analyzed. (Values for catalase activity in tibial chondrodystrophy are not given because insufficient tissue was available for analysis.) Compared with normal premineralized cartilage (a comparable region would be normal proliferating cartilage), on a dry weight basis the SOD activity of chondrodystrophic cartilage is low (0.58 0.23 U/mg dry weight, n = 4). We also measured the SOD activity in cultured chondrocytes. The value of 1.6 U/106 cells was considerably lower than that of isolated cells of other tissues (liver, 19.5 U/106 cells). The activity profile of catalase in cartilage is similar to SOD. Figure 4 shows that as the cells mature there is a de-

*

DISCUSSION Results of the study showed that avian cartilage contains the dismutating enzymes catalase and SOD. Three isoforms of SOD were present in cartilage. The PI of the major isozymes were pH 6.9 and 6.7 and corresponded to the copper-zinc isoforms.('O) The presence of a third band with a PI of 8.3 suggested that at least one of these isoforms was unique to cartilage. The function of SOD and catalase is to protect the cell from oxygen radical damage. Disproportionation of the superoxide anion radical by SOD would result in the formation of hydrogen peroxide; catalase would serve to degrade the peroxide to water and oxygen. In this way, these two enzymes provide protection from oxidative attack. The study clearly showed temporal and spatial differences in the distribution of enzyme activities in the growth plate. We found that the activities of both these enzymes declined as chondrocytes matured and the cartilage became calcified. The highest activities of SOD and catalase were evident in premineralized cartilage. In the terminally differentiated hypertrophic and calcified zones of the growth plate, however, activities were low.

573

SOD AND CATALASE IN CARTILAGE

0

I

II

Ill

IV

FIG. 4. Catalase activity in epiphyseal cartilage. Enzyme activity was determined using sections that were identified histologically: I and 11, resting proliferating cartilage; I11 and IV, hypertrophic cartilage. Error bars show standard error of the mean. *Significantly different from region I (p < 0.01).

C

C

8.36.96.7-

6.3-

Cells in this region would have lii tle endogenous protection against oxygen radical attack. It is likely that the demonstrated changes in enzyme activity are dependent on a number of factors: these include the vascular supply, the local production of oxygen radicals by cells, and the oxidative state of the tissue.(2L) Considering oxidative metabolism first, the activities of SOD correlated with the extent of oxidative energy conservation by chondrocytes in each of the regions of the epiphysis. In premineralized cartilage, cells have high energy charge and 221; chondrocyte dependence on NAD/NADH mitochondrial activity for energy metabolism would result in the formation of superoxide and hydroxyl radicals as by-products of electron transport. An elevated level of disproportionating activity would protect the cells from toxic oxygen radicals. As the chondrctcytes mature and become hypertrophic, there is an increase in nonmitochondrial energy metabolism. As a resiilt, mitochondrial activity is diminished and minimal levels of oxygen radicals would be expected to be generated through this pathway. The low SOD and catalase activities in hypertrophic and calcified cartilage cells parallel the decreased level of mitochondrial oxidative activity in these regions. In terms of the production of oxygen radicals by chondrocytes, none have been reported thus far for cells of the growth plate. Nevertheless, it is likely that chondrocytes are capable of generating reactivc: oxygen species by activation of the cyclooxygenase pathway and/or a NADPH oxidase. Indeed, we have previously shown that cells of the growth plate can generate NADPH through the pentose phosphate shunt and that the highest shunt activity is evident in the hypertrophic zone.(2-1The relative importance of each of these separate pathways for the generation of oxygen radicals (mitochondrial oxidative activity, pentose phosphate shunt, and NADPH oxidase and cyclooxygenase metabolism) remains to be established. Finally, with respect the chontlrodystrophic cartilage, it was found that the SOD activity was relatively low. The etiology of this condition is not fully understood, but the histologic structure indicates that the tissue contains very few blood vessels and many of the cells are necrotic. A possible cause of cellular necrosis, may be attack by oxygen radicals and minimal protective disproportionating enzyme activity. Measurement of lipid peroxides levels in chondrodystrophic chondrocytes would provide support for this mechanism.

ACKNOWLEDiGMENTS

a

b

This work was supported by NIH Grants AR 3441 1, DE 08239, and DE 06533. Shapiro i:j a recipient of a Clinical FIG. 5. Isoelectric focusing gel of cartilage SOD. Follow- Associate Physician Award (5-M01 RR-01224). ing focusing. the gel was stained with riboflavin and NBT. LLer homogenate: staining of SOD isoforms that focus at pH 6.9, 6.7, 6.6, and 6.3. Cartilage extract: lane a, staining REFERENCES of SOD isozymes that focus at pH 8.3, 6.9, and 6.7; lane b, staining of SOD isozymes after treatment with cyanide. 1. Klebanoff SJ 1975 Antimicrobial mechanisms in neutrophilic Note with the exception of the pH 6.9 band there is almost complete loss of staining of all the isozymes. polymorphonuclear leukocytes. Semin Hematol 12: 117-141. I ,

-

~-

MATSUMOTO ET AL.

514 2. Weiss SJ 1989 Tissue destruction by neutrophils. N Engl J Med 320:365-376. 3. Bates EJ, Johnson CC, Lowther DA 1985 Inhibition of pro-

4.

5.

6.

7.

teoglycan synthesis by hydrogen peroxide in cultured bovine articular cartilage. Biochim Biophys Acta 838:221-228. Bates EJ, Lowther DA, Handley CJ 1984 Oxygen free radicals mediate an inhibition of proteoglycan synthesis in cultured articular cartilage. Ann Rheum Dis 43:462-469. Schalkwijk J, van den Berg WB, van de Putte LBA, et al. 1985 Hydrogen peroxide suppresses the proteoglycan synthesis of intact articular cartilage. Arthritis Rheum 12:205-210. Greenwald RA, Moy WW, Lazarus D 1976 Degradation of cartilage proteoglycans and collagen by superoxide radical. Arthritis Rheum 19:799. Greenwald RA, Moy WW 1980 Effects of oxygen derived free radicals on hyaluronic acid. Arthritis Rheum 23455-

463. 8. Weiss SJ, Peppin 0, Ortiz X, Ragsdale C, Test ST 1985 Oxi-

9.

10. 11.

12.

dative autoactivation of latent collagenase by human neutrophils. Science 227:747-749. Monboisse J-C, Bellon G, Randoux A, Bore1 J-P 1987 Collagen activates superoxide anion production by human polymorphonuclear neutrophils. Biochem J 246:599-603. Weissinger RA, Fridovitch I 1973 Superoxide dismutase: Organelle specificity. J Biol Chem 248:3582-3592. Davis WL, Kipnis M, Shibata K, Farmer GR, Cortinas E, Matthews JL, Goodman DBP 1989 The immunohistochemical localization of superoxide dismutase activity in the avian epithelial growth plate. Histochem J 21:210-215. Boyde A, Shapiro IM 1987 Morphological observations concerning the pattern of mineralization of the normal and the rachitic chick growth cartilage. Anat Embryo1 (Berl) 175:

457-466. 13. Howlett CR 1980 The fine structure of the proximal growth plate of the avian tibia. J Anat 128:377-399. 14. Kakuta S, Golub EE, Haselgrove JC, Chance B, Frasca P, Shapiro IM 1986 Redox studies of the epiphyseal growth car-

tilage: Pyridine nucleotide metabolism and the development of mineralization. J Bone Miner Res 1:433-440. 15. Lilburn MS, Leach RM 1980 Metabolism of abnormal cartilage cells associated with tibia1 dyschondroplasia. Poultry Sci 591892-1896.

16. Golub EE, Schattschneider SC, Berthold P, Burke A, Shapiro 1M 1983 Induction of chondrocyte vesiculation in vitro. J Biol Chem 258:616-621. 17. Azzi A, Montecucco C, Richter C 1975 The use of acetylated

18. 19. 20.

21.

ferricytochrome c for the detection of superoxide radicals produced in biological membranes. Biochem Biophys Res Commun 65597-603. Misra HP, Fridovitch I 1978 Inhibition of superoxide dismutases by azide. Arch Biochem Biophys 189317-322. Flohe L, Otting F 1984 Superoxide dismutase assays. Methods Enzymol 10593-104. Johansson LH, Borg LAK 1985 A spectrophotometric method for the determination of catalase activity in small tissue samples. Anal Biochem 174:331-336. Peterson DA, Asinger RW, Elsperger KJ, Hosman DC, Eaton JW 1985 Reactive oxygen species may cause myocardial reperfusion injury. Biochem Biophys Res Commun 127:

87-93. 22. Shapiro IM, Golub EE, Kakuta S, Haselgrove JC, Havery J, Chance B, Frasca P 1982 Initiation of endochondral calcifi-

cation is related to changes in the redox state of hypertrophic chondrocytes. Science 217:950-952. 23. Silverton SF, Matsumoto H, DeBolt K, Reginato A, Shapiro IM 1989 Pentose phosphate shunt metabolism by cells of the chick growth cartilage. Bone 10:45-51. 24. Brighton CT, Heppenstall RB 1971 Oxygen tension in zones of the epiphyseal plate, the metaphysis and diaphysis. J Bone Joint Surg [Am] 53719-728.

Address reprint requests to: Dr. Irving M. Shapiro Department of Biochemistry School of Dental Medicine University of Pennsylvania 4001 Spruce St. Philadelphia, PA 19104-6002

Received in original form June 1, 1990; in revised form December 21, 1990; accepted December 26, 1990.