Endocytosis of Superoxide Dismutase Is Required in Order for the ...

5 downloads 0 Views 664KB Size Report
the medium by glucose oxidase. Five inhibitors of en- docytosis, methylamine, monensin, benzyl alcohol, cy- tochalasin B, and oligomycin, each abolished the ...
THEJOURNAL OF BIOLOGICAL CHEMISTRY

Vol. 263,No. 8, Issue of March 15, pp. 3784-3789, 1988 Printed in U.S.A.

0 1988 by The American Society for Biochemistry and Molecular Biology, Inc.

Endocytosis of Superoxide DismutaseIs Required in Order for the Enzyme to Protect Hepatocytes from theCytotoxicity of Hydrogen Peroxide* (Received for publication, September 24,

1987)

Marlene E. Kyle$, Dai Nakaes, Isao SakaidalI,Stefania MiccadeiII, andJohn L. Farber** From the Department of Pathobgy, Thomas Jefferson University, Philadelphia, Pennsylvania 19107

Inhibitors of endocytosis have been used to show that internalization of superoxide dismutase is required for the enzyme to protect hepatocytes from the cytotoxicity of hydrogen peroxide. As shown previously (Starke, P. E., and Farber, J. L. (1985) J. Biol. Chem. 260, 10099-10104), superoxide dismutase prevented the killing of cultured hepatocytes by HzOzgenerated in the medium by glucose oxidase. Five inhibitors of endocytosis, methylamine, monensin, benzyl alcohol, cytochalasin B, and oligomycin, each abolished the protective effect of superoxide dismutase. Cell-associated superoxide dismutase activity was increased 4-fold in hepatocytes after exposure to superoxide dismutase for 1 h. Each of the inhibitors abolished this increase in the cell-associated superoxide dismutase activity. The uptake of horseradish peroxidase, a measure of fluid phase endocytosis, differed from that of superoxide dismutase inits lower rate, reduced sensitivity to methylamine, and its insensitivity to cytochalasin B. The results of the present study demonstrate that endocytosis of superoxide dismutase is required to protect hepatocytes from the cytotoxicity of hydrogen peroxide. This conclusion may account for some of the conflicting results in the literature with respect to the protective action of superoxide dismutase.

superoxide anions are needed to reduce ferric to ferrous iron; the latter then reacts with hydrogen peroxide to form the ultimate mediator of cell injury, the hydroxyl radical. In recent years there have been many studiesof the protective effect of superoxide dismutase. The enzyme has been reported to prevent thecell injury mediatedby oxygen radicals under avariety of circumstancesinboth eukaryotic and prokaryotic systems (for recent reviews see Refs. 2 and 3). However, inall of thesestudies, as well as our own, the question as to whether exogenous superoxide dismutase acted inside or outside the cell when it protected against oxidative injury has been left unanswered. Cells are generally impermeable to such large exogenous biomolecules as superoxidedismutase.Nevertheless, these molecules can enter the cytoplasm of cells by at least two mechanisms, namely receptor-mediated endocytosis or fluidphase endocytosis (4). Both of these closely related pathways involve the invagination and subsequent pinchingoff of the plasma membrane to form vesicles containing extracellular material. While uptake via receptor-mediated endocytosis requires the bindingof ligand to specific cell surface receptors, fluid phase endocytosis is a nonsaturable receptor-independent process. A number of agents are known to inhibit the cellular uptake of exogenous compounds by interfering with one or more steps of endocytosis. In the present study, methylamine, monensin, benzyl alcohol, cytochalasin B, and oligomycin, known inhibitors of endocytosis (4-8), were used to The killing of cultured hepatocytes by hydrogen peroxide determine whether the uptake of superoxide dismutase is a depends on a cellular source of superoxide anions (1). The prerequisite for its protective actionagainst oxidativecell addition of superoxide dismutase, theenzyme which promotes injury. The results demonstrate that the endocytosis of suthe dismutation of superoxide anions, prevented the cytotox- peroxide dismutase is indeed necessary in cultured hepatoicity of hydrogen peroxide that was generated in themedium cytes for the prevention of the cell killing by hydrogen perby glucose oxidase. Glucose oxidase catalyzes a two-electron oxide. reduction of molecular oxygen to hydrogen peroxide without the intermediate formation of superoxide anions. Therefore, MATERIALS ANDMETHODS the superoxide anions required for the cell killing by hydrogen Male Sprague-Dawleyrats (150-200g) wereobtained from Charles peroxide are produced by the hepatocytes. A cellular source River Breeding Laboratories, Inc. All animals were fedad libitum and of ferric iron is alsorequired for thefull expression of hydro- fasted overnight prior to use. Hepatocytes were isolated by collagenase gen peroxide toxicity. It was concluded that the endogenous (Sigma) perfusion according to Seglen (9).Yields of 3-5 X 10’ cells/ * This work was supported by Grant DK38305 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $Recipient of an individual National Research Service Award (1F32ESGM 05418). 5 On leave fromthe Department of Oncological Pathology, Cancer Center, Nara Medical University, Kashihara, Nara 634,Japan. 11 On leave from the Department of Internal Medicine, Yamaguchi University School of Medicine, Ube 755,Japan. )I Onleave from the Dipartimento di Medicini Sperimentale, Universiti Degli Stud; di Roma “la Sapienza,” Rome, Italy. ** To whom all correspondence should be addressed.

liver with 85-90% viability by trypan blue exclusion were routinely obtained. The hepatocytes were plated in 25-cm2 flasks (Corning Glass Works, Corning, NY) at a density of 1.33 X IO6 cells/flask in 3 ml of Williams E medium (Gibco Laboratories) containing 10 IU/ml penicillin, 10 pg/ml streptomycin, 0.5 mg/ml gentamicin, 0.02 unit/ ml insulin, and 10% heat-inactivated (55“C for 15 min) fetal calf serum (Hazelton Research Products, Lenexa, KS) (completeWilliams E). After incubation for 2 h at 37 “C in an atmosphere of 5% COZ, 95% air, the cultures were rinsed twice with a prewarmed Hepes’ (Sigma) buffer, pH 7.4 (0.14 M NaCI, 6.7 mM KC], 1.2mM CaCL, and 2.4 mM Hepes) to remove unattached dead cells. Complete Williams E (5 ml) was replaced, and the cells were incubated for 24 h. The The abbreviation used is: Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid.

3784

Endocytosis Dismutase of Superoxide cultures were then washed and incubated in Williams E minus fetal calf serum with the additions indicated in the text. Oligomycin (Sigma) and cytochalasin B (Sigma) were dissolved in dimethyl sulfoxide and added to the cultures a t a final concentrationof 0.1 and 25 pg/ml, respectively (the final concentration of dimethyl sulfoxide was 0.1%). Monensin (Sigma) was dissolved in ethanol and added to the cultures in concentrations up to 10 p~ (the final concentration of ethanol was 0.5%). Methylamine (Sigma) was dissolved in 0.9% NaC1, and benzyl alcohol (Sigma) was added directly to thecultures. Glucose oxidase (Sigma) and bovine erythrocyte superoxide dismutase (Sigma) were dissolved in 0.9% saline and added to the cultures a t a final concentration of 5 and 500 units/ml, respectively. Glucose oxidase alone or glucose oxidase plus superoxide dismutase was added 15 min after addition of one of the above inhibitors of endocytosis. The viability of the hepatocytes was measured 1h following exposure to glucose oxidase by the release of lactate dehydrogenase into the medium (10). Superoxide dismutase activity was measured spectrophotometrically by monitoring the inhibition of the reduction of ferricytochrome c at 550 nm as described by Flohe and Otting (11).The superoxide dismutase activity of the hepatocytes was determined by washing the cultures several times with a Hepes buffer and then lysing the cells in 6 ml of 0.5% Triton X-100, 50 mM sodium-potassium phosphate, 0.1 mM EDTA, pH 7.8. An aliquot of the lysate was then assayed for superoxide dismutase as above. Horseradish peroxidase (Sigma) activity was measured according to Steinman and Cohn (12). Hepatocytes exposed to horseradish peroxidase (1 mg/ml) for 1 h were washed and lysed as described above except that 2 ml of Triton X-100-phosphate buffer was used. Aliquots of lysate were then assayed for horseradish peroxidase activity. Protein contentwas determined by the method of Lowry et al. (13) with bovine serum albumin serving as standard. RESULTS

Inhibitors of Endocytosis Preuent the Cytoprotection by Superoxide Dismutase-Hepatocytes in culture for 24 h were treated with glucose oxidase, and 1 h later the viability of the cells was measured by the release of lactate dehydrogenase into the medium. Whereas only 5% of the cells in control cultures died (Table I, line l),55% of the hepatocytes were killed by the hydrogen peroxide that was generated in the medium by glucose oxidase (line 2). As reported previously (l), theaddition of superoxide dismutase to the culture medium protected the hepatocytes from the toxicity of Hz02(line 3). However, the addition to the medium of the inhibitors of endocytosis, methylamine, monensin, benzyl alcohol, cyto-

3785

chalasin B, or oligomycin, prevented the ability of superoxide dismutase to protect the hepatocytes (lines 4-8). Importantly, the loss of the ability of superoxide dismutase to prevent the cell killing by hydrogen peroxide was not a consequence of the toxicity of the combination of the enzyme and any of the inhibitors. Superoxide dismutase in the presence of any one of the inhibitors had no effect on the viability of the cells in the absence of glucose oxidase (lines 9-13). Furthermore, the extent of the cell killing by glucose oxidase alone was not altered by the presence of any of the inhibitors (data not shown). We examined the action of monensin and methylamine in greater detail. Figs. 1 and 2 illustrate the dependence of the inhibition of the protective effect of superoxide dismutase on the log of the concentration of monensin (Fig. 1) and methylamine (Fig. 2). The concentration of monensin required for half-maximal inhibition of the protective effect of superoxide dismutase was about 0.5 PM,while the half-maximal concentration of methylamine was about 0.2 mM. These concentrations are similar to those reported in the literature to affect one or more steps of endocytosis (14-16). Inhibitors of Endocytosis Do Not Inhibit Superoxide Dismutase Actiuity-The prevention by methylamine, monensin, benzyl alcohol, cytochalasin B, and oligomycin of the protective effect of superoxide dismutase could be simply the consequence of an inhibition of the enzyme. However, none of the agents inhibited superoxide dismutase activity. Superoxide dismutase activity was assayed as the extent of the inhibition of the rate of the reduction of ferricytochrome c by the superoxide anions generated with xanthinelxanthine oxidase. Upon the oxidation of xanthine touric acid, superoxide anions are formed, which species, in turn, reduce ferricytochrome c. The reduction of ferricytochrome c is quantitated by the increase in absorbance at 550 nM. A linear relationship be-

1

Glucose Oxidase Alone

TABLEI Protection Ly superoxide dismutase depends on its endocytosis by cultured hepatocytes Hepatocytes in culture for 24 h were washed and placed in fresh medium or inmedium with methylamine, benzyl alcohol, cytochalasin B, or oligomycin. Following preincubation for 15 min, the cultures receivedglucose oxidase (GO) orglucose oxidase plus superoxide dismutase (SOD). In all cases, viability was measured 1h after glucose oxidase by the release of lactate dehydrogenase into the medium. Treatment

Dead cells"

-

%

1. Control 2. GO (5 units/ml for 1 h) GO 3. plus SOD (500 units/ml) 4. GO plus SOD plus methylamine 5. GO plus SOD plus benzyl alcohol 6. GO plus SOD plus cytochalasin B 7. GO plus SOD plus oligomycin 8. GO plus SOD plus monensin 9. SOD plus methylamine (40 mM) 10. SOD plus benzyl alcohol (30 mM) 11. SOD plus cytochalasin B (25 pg/ml) 12. SOD plus oligomycin (0.1pg/ml) 13. SOD plus monensin (50 p M )

4.6 f 0.6 54.6 f 4.4 12.8 f 4.3 52.9 f 3.2 54.2 f 4.3 52.9 f 3.7 53.6 f 5.7 55.0 f 3.4 4.4 f 0.6 5.9 f 1.0 4.3 f 0.5 3.9 f 0.4 4.7 f 0.7 Results represent the mean f S.D. of the determination on three separate cultures.

A

GIucose Oxidase SOD

I I 11

.os

1

.5 1

I

I

s

10

Monensin (pM)

FIG. 1. Dose dependence of the ability of monensin to inhibit the protective effect of superoxide dismutase. Hepatocytes in culture for 24 h were washed, fresh Williams E replaced, and then treated with increasing concentrations of monensin. Following a 15min preincubation period, the cultures received glucose oxidase or glucose oxidase plus superoxide dismutase (SOD). The viability of the hepatocytes was measured 1 h later by the release of lactate dehydrogenase into the medium. Values represent the mean f S.D. of the determinations of three separate cultures.

3786

Endocytosis of Superoxide Dismutase

uptake of superoxide dismutase by the cells. Hepatocytes were incubated for 1 h in the presence or absence of superoxide dismutase. The cells were then washed several times with Hepes buffer and lysed by Triton X-100. The lysate was then assayedforsuperoxide dismutase activity (Table 111). The specific activity of superoxide dismutase in those cultures not incubated with added enzyme was 11 units/mg protein (line l),a value similar to the activity of hepatocellular superoxide dismutasereported elsewhere (17). By contrast, when the liver cellswere incubated for 1h with 500 units/ml superoxide dismutase, the cell-associated superoxide dismutase activity, after washing and lysing the hepatocytes, increased to 42.9 units/mgprotein (line 2). Furthermore, methylamine, monensin, benzyl alcohol, cytochalasin B, and oligomycin prevented the increase in superoxide dismutase activity associated with the presence of exogenous enzyme in the culture medium (lines 3-6). Importantly, none of these compounds alone (in the absenceof added superoxide dismutase) had anyeffect on Glucose Oxidase + SO0 the superoxide dismutase activityof the hepatocytes (lines 710). 1 I , , , I Figs. 3 and 4 illustrate in further detail the effect of monen.01 .05.1 .5 1 102040 sin and methylamine, respectively, on the uptake of superMethylamine (mM) oxide dismutase by the cultured hepatocytes. In both cases FIG.2. Dose dependence of the ability of methylamine to the inhibition of the endocytosis of superoxide dismutase inhibit the protective effect of superoxide dismutase. Hepato- depends on the log of the concentration of monensin and cytes in culture for 24 h were washed, fresh WilliamsE replaced, and then treated with increasing concentrationsof methylamine. Follow- methylamine. The concentration of monensin requiredfor half-maximal inhibitionof the uptakeof superoxide dismutase ing a 15-minpreincubationperiod, the cultures receivedglucose oxidase or glucoseoxidaseplussuperoxide dismutase (SOD). The was about 0.8 p ~ a ,level nearly identical to that needed to viability of the hepatocytes was measured 1 h later by the release of produce half-maximal inhibition of the protective effect of lactate dehydrogenase into the medium. Values represent the mean superoxide dismutase illustrated in Fig. 1. The concentration f S.D. of the determinations on three separate-cultures. of methylamine required for half-maximal inhibition of the uptake of superoxide dismutase was 200 pM. This concentraTABLEI1 tion of methylamine isagain identical to thatassociated with The activity of superoxide dismutase is not affected by inhibitors the half-maximal inhibitionof the protective effect of superof endocytosis biphasic nature oxide dismutase (Fig. 2). Of further note is the Superoxide dismutase (SOD) activity was assayed by monitoring the rate of ferricytochrome c reduction by superoxide anions at 550 of the action of methylamine illustrated in bothFigs. 1B and nm. The assays contained per ml of sodium/potassium phosphate, 2B. Measurement of the Endocytosis of Horseradish PeroxiEDTA, pH 7.8: 20 nmol cytochrome c , 45 nmol of xanthine, and 0.2 unit of xanthine oxidase. Superoxide dismutase, methylamine, benzyldase-The uptake of superoxide dismutase was compared to alcohol, cytochalasin B, and oligomycin were present at the concen- that of horseradish peroxidase, an enzyme that is internalized trations indicated below. by fluid phase endocytosis. Following incubation with 1 mg/ Inhibition of the reduction” ml horseradish peroxidase for 1 h, cultured hepatocytes seTreatment of ferricytochrome c

iA

i-

I

,

~~

~

~~

~~

%

SOD (0.3unit/ml) 65 f 1 SOD plus monensin (10p ~ ) 66 f 5 65 f 2 SOD plus methylamine (40mM) 98 f 3 SOD plus benzyl alcohol (30mM) SOD plus cytochalasin B (25pg/ml) 56 f 2 SOD plus oligomycin (0.1pg/ml) 65 f 1 Results representthe mean f S.D. of the determinations on three separate assays.

TABLE111 Endocytosis of superoxide dismutase by cultured hepatocytes Hepatocytes in culture for 24 h were washed and placed in fresh medium or in medium containing either methylamine, benzyl alcohol, cytochalasin B, or oligomycin. After 15 min, superoxide dismutase (SOD) was added to the medium, and the cells were incubated for 1 h. The cultures were washed 4 times with buffer and then lysed with Triton X-100.Superoxide dismutase activity the in lysates was measured as described under “Materials and Methods.” Treatment

SOD activity

tween the concentration of superoxide dismutase and the unitlmg protein inhibition of ferricytochrome c reduction was observed from 11.1 f 1.6 1. Control 0.1 to 1unit/ml superoxide dismutase (data not shown). Using 42.9 f 3.5 2. SOD (500 units/ml) this assay, neither methylamine, cytochalasin B, oligomycin, 3. SOD plus methylamine (40mM) 13.9 k 3.3 13.0 f 2.0 4. SOD plus monensin (10pM) nor monensin, at the concentrations they were used in Table 13.9 f 1.7 5. SOD plus benzyl alcohol (30 mM) I, affected the ability of superoxide dismutase (0.3 unit/ml) 13.9 f 1.7 6. SOD plus cytochalasin B (25pg/ml) to inhibit the reduction of ferricytochrome c (Table 11). In 7. SOD plus oligomycin (0.1pg/ml) 13.0 k 2.8 addition, benzyl alcohol appeared to enhance the inhibition 14.9 f 1.7 8. Methylamine of ferricytochrome c reduction. As a control,none of the 9. Monensin 11.0f 2.2 inhibitors were shown t o affect the reduction of ferricyto10. alcohol Benzyl 11.1 f 1.5 11. Cytochalasin B 13.9 f 2.8 chrome c by xanthine/xanthine oxidase in the absence of 12.9 f 2.8 12. Oligomycin superoxide dismutase (data not shown). Measurement of the Endocytosis of Superoxide Dismutase“Results are the mean f S.D. of the determinations on three separate cultures. Theinhibitors of endocytosis were shown to prevent the

Endocytosis Dismutase of Superoxide

TABLE IV Endocytosis of horseradish peroxidase by cultured hepatocytes Hepatocytes in culture for 24 h were washed and placed in fresh medium or in medium containing either methylamine, benzyl alcohol, cytochalasin B, or oligomycin. After 15 min, horseradish peroxidase (HRP) was added to the medium and the cells were incubated for 1 h. The cultures were washed 5 times with buffer and then lysed with Triton X-100. Horseradish peroxidase activity inthe lysates was measured as described under “Materials and Methods.”

SOD Alone

C-

3787

a

HRP uptake”

Treatment

nglmg protein

1. HRP (1mg/ml) 2. HRP plus monensin (10 p M ) 3. HRP plus oligomycin (0.1 pg/ml) 4. HRP plus benzyl alcohol (30 mM) 5. HRP plus methylamine (40 mM) 6. HRP plus cytochalasin B (25 pg/ml)

35.0 13.7 12.3 16.5 18.3 30.0

“Results are the mean of the determinations on two separate cultures. Variance between values was less than 5%.

.05 .1

.5 1

40

5 10

1

Monensin (OM) FIG. 3. Dose dependence of the abilityof monensin to inhibit the uptake of superoxide dismutase. Hepatocytes in culture for 24 h were washed, fresh Williams E replaced, and then treated with increasing concentrations of monensin. After 15 min, the hepatocytes were treated with saline or500 units/ml superoxide dismutase (SOD 1. Following 1h of incubation, the cells were washed several times with buffer andthen lysed with 0.5% Triton X-100. The superoxide dismutase activity of the lysates was measured spectrophotometrically as described under “Materials and Methods.” Values represent the mean & S.D.of the determinations on three separate cultures.

.05 .1

.5 1

5 10

Monensin (pM)

FIG. 5. Dose dependence of the ability of monensin to inhibit the uptakeof horseradish peroxidase.Hepatocytes in culture for 24 h were washed, fresh Williams E replaced, and then treated with increasing concentrations of monensin. After 15 min, the hepatocytes were treated with saline or 1 mg/ml horseradish peroxidase ( H R P ) . Following 1h of incubation, the cells were washed several times with buffer and then lysed with 0.5% Triton X-100. The horseradish peroxidase activity of the lysates was measured spectrophotometrically as described under ”Materials and Methods.” Values represent the mean of the determinations on two separate cultures. Variance between the duplicate was not greater than 5%.

I

.01

I

,

.05 .I

I

/

.5 1

I

,

I

10 2 0 4 0

Methylamine (mM) FIG. 4. Dose dependence of the ability of methylamine to inhibittheuptake of superoxidedismutase. Hepatocytes in culture for 24 h were washed, fresh Williams E replaced, and then treated with increasing concentrations of methylamine. After 15 min, the hepatocytes were treated with saline or 500 units/ml superoxide dismutase (SOD). Following 1h of incubation, the cells were washed several times with buffer and then lysed with 0.5% Triton X-100. The superoxide dismutase activity of the lysates was measured spectrophotometrically as described under “Materialsand Methods.” Values represent the mean f S.D. of the determinations on three separate cultures.

questered 35 ng of enzyme/mg of protein (Table IV, line I), an amount similar to values described previously (12, 18). Like superoxide dismutase, the endocytosis of horseradish peroxidase was inhibited by monensin, methylamine, oligomycin, and benzyl alcohol (Table IV). For example, monensin and oligomycin decreased the endocytosis of horseradish peroxidase by 60-65% (lines 2 and 3), while methylamine and benzyl alcohol decreased its uptake by 50% (lines 4 and 5 ) . By contrast with superoxide dismutase, cytochalasin B only slightly reduced the intracellular content of horseradish peroxidase (line 6). The inhibitory action of monensin (Fig. 5) and methylamine (Fig. 6) on the endocytosis of horseradish peroxidase was dependent on the log of their concentration. From Figs. 3 and 5 it can be seen that theinhibitory effect of monensin on the hepatocellular uptake of horseradish peroxidase is identical to thatof superoxide dismutase. However, methylamine was less effective in reducing the uptake of horseradish peroxidase (Fig. 6) as compared to superoxide dismutase (Fig.

3788

Endocytosis of Superoxide Dismutase

radish peroxidase represents an exogenous substance that is sequestered by this fluid phase process (12, 18). However, enhancedrates of endocytosis above theintrinsicrate of membrane internalization have also been demonstrated, particularly for ligands that bind to the cell surface prior to entry (4, 19). In the present study, the cell-associated enzymatic activity observed after 1 h of incubation withhorseradish peroxidase was 35 ng/mg protein. However, the amount of superoxide dismutase internalized by hepatocytes in 1 h was 8 pg/mg protein, a level more than 200 times higher than that 2 p 20 described for horseradish peroxidase. Therefore, it would ap.pear that the uptakeof superoxide dismutase by the hepatoa I I cytes is not simply a consequence of the constitutiverecycling of plasma membrane. Two other facts support the conclusion that superoxide dismutaseisnotinternalized by a mechanismsimilar to horseradish peroxidase. The uptake of superoxide dismutase .01 .05.1 .5 1 la 2 0 4 0 by hepatocytes was more sensitive to the inhibitoryaction of Methylamine (mM) methylamine as compared to that of horseradish peroxidase. FIG. 6. Dose dependency of the ability of methylamine to Such a result is consistent with previous observations that inhibit the uptake of horseradish peroxidase. Hepatocytes in culture for 24 h were washed, fresh Williams E replaced, and then receptor-mediated processes are inhibited by low concentratreated with increasing concentrations of methylamine. After 15 min, tions of methylamine that have little effect on the rate of the hepatocytes were treated with saline or 1 mg/ml horseradish fluid phase endocytosis (14, 15, 20, 21). peroxidase (HRP).Following 1h of incubation, the cells werewashed Of particular importance is the inability of cytochalasin B several times with buffer and then lysed with 0.5% Triton X-100. to inhibit the uptake of horseradish peroxidase under condiThe horseradishperoxidaseactivity of the lysates was measured tions where it completely prevented theendocytosis of superspectrophotometricallyas described under “Materials and Methods.” of memValues represent the mean of the determinations on two separate oxide dismutase. It is well documented that the rate brane recycling is not influenced by cytochalasin B. Consecultures. Variance between the duplicate was not greater than 5%. quently, cytochalasinBdoes not significantly inhibitthe internalization of horseradishperoxidase (22-24), albumin 4). Theconcentration of methylaminethatinhibitedthe uptake of horseradish peroxidase by 50% was 40 mM (Fig. 6), (24), and ferritin (25), compounds believed to enter cells by numerous a concentration 200-fold higher than thatwhich produced an fluid phase endocytosis. By contrast,thereare equivalent inhibition of the uptake of superoxide dismutase reports of an inhibitoryeffect of cytochalasin B on the endocytosis of certain ligands internalized by receptor-dependent (0.2 mM). processes. For example,cytochalasinB has been demonstrated to reduce the cellular uptake of various peptides and DISCUSSION proteins (26-30), immunoglobulins (31), andlow density lipThe results of the present study show that five different oprotein (32). In light of the potent inhibitoryeffect of cytocompounds, which share an ability to inhibit endocytosis, chalasin B on the uptake of superoxide dismutase observed prevented superoxide dismutase from protecting cultured hepin this study,would it appear thatsuperoxide dismutase enters atocytes against the toxicity of hydrogen peroxide. Addition the hepatocyteby a mechanism involving its association with of superoxide dismutaseto the medium reducedthe sensitivity receptors foundin coated pits. of cultured hepatocytes to hydrogen peroxide. By contrast, no The abilityof methylamine to prevent the protective effect protection by superoxide dismutase was observed when the of superoxide dismutase should not be confused with a prehepatocytes were pretreated with either methylamine, monensin, benzyl alcohol, cytochalasin B, or oligomycin, inhibitors vious observation from this laboratory (33). In that report, hepatocytes from of endocytosis. The abilityof these agents to prevent protec-exposure to methylamine protected cultured the toxicity of hydrogen peroxide generated in the medium by tion by superoxide dismutase cannot be attributed to inhibiglucose oxidase. However, such a protective action required a tion of this enzyme (Table 11). That cultured hepatocytes internalize exogenous superox- 4-h pretreatment with methylamine prior to the addition of ide dismutase is indicated by the &fold increase in the cell- glucose oxidase. No protection was observed when methylaassociated superoxide dismutase activity following a 1-h ex- mine was added simultaneouslywith glucose oxidase. The posure to the enzyme in the culture medium. When methyl- protective effect of methylamine was interpreted as an attenamine, monensin, benzyl alcohol, cytochalasin B, or oligo- uation of the release of ferric iron from a lysosomal pool. In of methylaminetoprevent mycin were added 15 min prior to superoxide dismutase, no thepresentstudy,theability superoxide dismutase from protecting hepatocytes against increase in cell-associated superoxide dismutase activity was observed. Furthermore, the concentrations of methylamine hydrogen peroxide toxicity is correlated with an inhibition of and monensin required for 50% inhibition of the protective endocytosis. It is concluded that endocytosis of superoxide dismutase is action of superoxide dismutase were similar to those that produced a 50% inhibition of the internalization of the en- required to protect cultured hepatocytes from the toxicity of zyme. Taken together, these data document that the endocy- hydrogen peroxide. Importantly, this paper represents the tosis of superoxide dismutase added to the medium is essential first demonstration that the protection afforded by adding exogenous superoxide dismutase to hepatocytes, and perhaps for protection against thetoxicity of hydrogen peroxide. other cell types, is an intracellular event and notdue simply Eukaryotic cells continuously internalize their membrane surfaces. Thus, exogenous material dissolved in the extracel- to scavenging of superoxid. anions in the extracellular fluid. lular fluid will be internalized at a rate governed by the rate The present resultsmay also account for some of the discrepof membrane recycling (for review see Refs. 4 and 19). Horse- ancies in the literature regarding the protectiveeffects of I

I

I‘ t

I

,

,

3789

Endocytosis of Superoxide Dismutase superoxide dismutase. Addition of this enzyme would only protect those cells actively carrying out endocytosis. Thus, the inability of superoxide dismutase to protect against various types of cell injuries may not be due to the lack of involvement of superoxide anions but may simply represent a failure of the uptakeof superoxide dismutase. REFERENCES 1. Starke, P. E., and Farber, J. L. (1985) J. Biol. Chern. 260,10099-

10104 2. DiGuiseppi, J., and Fridovich, I. (1984) CRC Crit. Reu. Toxicol. 12,315-342 3. Fridovich, I. (1983) Annu. Rev. Pharmacal. Tozicol. 23, 239-257 4. Mellman, I., Fuchs, R., and Helenius, A. (1986) Annu. Reu. Biochern. 55, 663-700 5. Dean, R. T., Jessup, W., and Roberts, C.R. (1984) Biochern. J . 2 1 7 , 27-40 6. Poste, G., and Papahadjopoulos, D. (1976) Proc. Natl. Acad. Sci. U. S. A . 73, 1603-1607 7. Tolleshaug, H., and Berg, T. (1982) Biochern. Phurmucol. 31, 593-595 8. Pontecorvi, A., and Robbins, J. (1986) Endocrinology 119,27552761 9. Seglen, P. 0.(1976) Methods Cell Biol. 13,29-83 10. Casini, A., Giorli, M., Hyland, R. J., Serroni, A., Gilfor, D., and Farber, J. L. (1982) J . Biol. Chern. 257, 6721-6728 11. Flohe, L., and Otting, F. (1984) Methods E n z y m l . 106, 93-104 12. Steinman, R. M., and Cohn, Z. A. (1972) J. Cell Biol. 5 5 , 186204 13. Lowry, 0.H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chern. 193, 265-275 14. Maxfield, F. R. (1982) J . Cell Biol. 95,676-681

15. Schwartz, A. L., Bolognesi, A., and Fridovich, S. E. (1984) J . Cell Biol. 98, 732-738 16. Basu, S. K., Goldstein, J. L., Anderson, R. G. W., and Brown, M. S. (1981) Cell 24,493-502 17. Marvelli, I., Rigo, A., Frederico, R., Ciriolo, M. R., and Rotilio, G. (1982) Bwchem. J. 204,535-540 18. Steinman. R. M.. Silver. J. M.. and Cohn. Z. A. (1974) J . Cell Biol. 63,949-969 ' 19. Goldstein, J. L., Anderson, R. G . W., and Brown, M. S. (1979) Nature 279,679-685 20. Kaplan, J., and Keogh, E. A. (1981) Cell 2 4 , 925-932 21. Sibille, J., Octave, J., Schneider, Y., Trouet, A., and Crichton, R. R. (1982) FEBS Lett. 150,365-369 22. Piasek, A., and Thyberg, J. (1980) J. Cell Sci. 45, 59-71 23. Blok, J., Scheven, B. A. A., Mulder-Stapel, A.A., Ginsel, L. A,, and Daems, W. T. (1982) Cell Tissue Res. 222, 113-126 24. Klaus, G. G. (1973) Exp. Cell Res. 79, 73-78 25. Wills, E. J., Davies, P., Allison, A. C., and Haswell, A. D. (1972) Nat. New Bwl. 240,58-60 26. Beem, R. L., Niederlehner, A., and Targan, S. R. (1986) Cell. Immunol. 102, 187-197 27. Brouwer, A., and Knook, D. L. (1982) J. Reticuloendothel. SOC. 32, 259-268 28. Sumpio, B. E., and Maack, T. (1982)Am. J. Physiol. 243, F379F392 29. Smedsrod, B., Johansson, S., and Pertoft,H. (1985) Biochemistry 228,415-424 30. Coombs. K.. Mann. E.. Edwards. J.. and Brown, D. T. (1981) J. Virol.37,' 1060-1065 31. Mackenzie, N. M. (1985) Cell Biol. Znt. Rep. 9,923-937 32. Miller, N. E.. and Yin, J. A. (1978) Biochim. Biophys. Acta 5 3 0 , 145-150 33. Starke, P. E., Gilbertson, J. D., and Farber, J. L. (1985) Biochern. Biophys. Res. Commun. 1 3 3 , 371-379 I

,

~~