Temporal Relationship of Free Radical-induced Lipid Peroxidation and Loss of Latent Enzyme Activity in Highly Enriched Hepatic Lysosomes*

Loss of latency due to membrane lipid peroxidation induced in vitro was studied in highly purified rat liver lysosomes. Enriched fractions of lysosomes were isolated by free flow electrophoresis. Lipid peroxidation of lysosomes, assayed as malondialdehyde formation, was catalyzed by a radical generating system consist- ing of dihydroxyfumaric acid and Fe3+-ADP. The peroxidation reaction occurred readily at 37 “C and reached a plateau at 10 min; however, the loss of lysosomal latency, determined as increased percentage free 0-N-acetylglucosaminidase activity, occurred more gradually and reached a maximum after 30 min. Scavengers of superoxide, hydrogen peroxide, singlet oxygen, and hydroxyl radicals did not inhibit the peroxidation reaction nor prevent the loss of lysosomal latency. However, preincubation of the lysosomes with a-tocopherol effectively blocked the induction of per- oxidation and substantially reduced the loss of lysosomal latency. These results indicate that the lysosomal membrane is susceptible to free radical-induced lipid peroxidation; further, this process may be the imme- diate cause of the subsequent disintegration of the lysosome. The nature of the protective effect of a-tocoph- erol is unclear but may be due to its interaction with the unsaturated membrane lipids and the subsequent interruption of the chain-reaction initiated by free radicals. were and centrifuged, The accumulation of thiobarbi- turic acid-reactive products was measured at 535 and the lipid peroxide formed was estimated by using a standard curve of MDA in a similar reaction mixture. This procedure provides a similar extinc-tion coefficient for MDA (1.56 X lo5 M" cm") as described by Wills (18).

Lipid peroxidation has been suggested to be associated with a variety of pathological processes such as liver necrosis, hemolytic anemia, lung damage (1-4), and, more recently, ischemic heart disease (5, 6). The peroxidation of lipids may involve the direct reaction of oxygen radicals with polyunsaturated fatty acids resulting in deterioration of the lipid (4). Biomembranes are major sites of lipid peroxidation damage due to the presence of polyunsaturated fatty acids in their membrane phospholipids. The pathological consequence of lipid peroxidation may thus reflect the alterations of membrane integrity and/or membrane-associated function in subcellular organelles such as mitochondria, microsomes, and lysosomes.
Lysosomes contain many hydrolytic enzymes and their liberation may result in extensive intracellular digestion. Work of Allison and Young (7) provided evidence that hydrolases released in vivo in cells can cause severe damage to *This work was supported by National Heart, Lung, and Blood Institute Grants T32-HL-07244 and HL-28985. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. cellular structure. Thus, the susceptibility of the lysosomal membrane to lipid peroxidation is of apparent importance to cellular pathology. Studies of the effect of lipid peroxidation induced by irradiation (8) and by oxidation of NADPH by the microsomal monoxygenase system (9) on lysosomal membranes have been reported. However, in those studies of modestly enriched lysosomal preparations, there was heavy contamination by other subcellular organelles. In an attempt to circumvent such cross-contamination, we have used highly purified "native" hepatic lysosomes prepared by free flow electrophoresis to study their susceptibility to free radicalinduced lipid peroxidation. The free radicals were generated during the autoxidation of DHF' in the presence of Fe3'-ADP.
Preparation and Characterization of Lysosomes-Lysosomes were isolated by a free flow electrophoresis procedure similar to that described by Henning and Heidrich (12) and Beckman et al. (13). Sprague-Dawley rats (200-300 g) were decapitated and the livers were perfused with ice-cold 0.25 M sucrose, 0.003 M MgClz, 0.001 M EDTA, 0.01 M MOPS, pH 7.2. All subsequent steps were performed at 0-4 "C unless stated otherwise. The perfused livers were homogenized in the above buffer first with a VirTis tissue disrupter for 2 X 3 s (at setting 50) followed by two up and down strokes in a Thomas Teflonpestle homogenizer. After centrifugation for 15 min, the 5-20,000 X g pellet was resuspended and extracted in 0.12 M K-gluconate, 0.05 M sucrose, 0.001 M EDTA, 0.01 M MOPS (pH 7.4) and centrifuged again a t 20,000 X g for 15 min. The replacement of KC1 by Kthan those obtained by the method of Beckman et al. (13). The pellet gluconate in this step of extraction resulted in more latent lysosomes was then "washed twice with 0.25 M sucrose, 0.001 M EDTA, 0.01 M MOPS (pH 7.4). The final pellet was resuspended in the electrophoresis buffer (0.3 M sucrose, 0.01 M triethanolamine, 0.001 M EDTA, and 0.01 M acetate adjusted to pH 7.4 by NaOH) and aggregated particles were removed by centrifuging at 1,000 X g for 10 min. The supernatant was then applied to a free flow electrophoresis unit (Desaga 48, Heidelberg, West Germany). The electrode buffer consisted of 0.1 M triethanolamine and 0.1 M acetate, pH 7.4 (NaOH).
The unit was run at 800 V and 100 mA at 3 "C. The sample pump was set at 4 ml/h and the collecting pump was at a setting of 3.0. After electrophoresis, the fractions were concentrated by immediate centrifugation at 20,000 X g for 20 min and analyzed for free and total specific P-N-acetylglucosaminidase activities as described by Ruth and Weglicki (14). Percentage free activity was calculated as (activity in 0.25 M sucrose)/(activity in 0.1% Triton X-100) X 100, Fractions were isolated between tubes 8 to 40. The elution profile of the 6-N-acetylglucosaminidase activities was similar to that reported earlier (131, and fractions highly enriched in P-N-acetylglucosaminidase were nearer to the anode. Marker enzymes in the various fractions were assayed as an indicator of the degree of subcellular particulate separation. Glucose-6-phosphatase was assayed according to the method of Baginski et al. (15). Cytochrome oxidase was assayed according to the procedure of Wharton and Tzagoloff (16). Protein determinations were performed according to Lowry et al. (17).
Incubation Procedure and Measurement of Lipid Peroxidation- The initial 5 to 7 fractions containing highly enriched lysosomes (0-N-acetylglucosaminidase > 30-fold, cytochrome oxidase c 0.3-fold, glucose-6-phosphatase < 0.3-fold) were pooled and pelleted at 20,000 X g for 20 min. The purified lysosomes were resuspended in the reaction buffer (0.12 M KCl, 0.05 M sucrose, 0.01 M potassium phosphate, pH 7.2) and were used immediately for free radical-induced lipid peroxidation studies. The incubation mixture usually contained 150-200 pg of lysosomal protein, 0.1 mM FeCb, 1 mM ADP, and 3.33 mM DHF in 1 ml of the reaction buffer. Specific conditions or additional components in the incubation mixture are described in the text. Reactions were initiated by the final additions of Fe3+-ADP and DHF. Incubations were carried out at 37 "C in air in a shaking water bath. At the indicated times of incubation, aliquots of the reaction mixtures were assayed for free and total activities of 0-N-acetylglucosaminidase.
The rates of lipid peroxidation were measured by the formation of thiobarbituric acid-reactive material (18). After incubation, the reaction was stopped by the addition of 0.1 ml of 5% (w/v) trichloro-   L None acetic acid/ml of reaction mixture. One ml of 0.5% (w/v) 2-thiobarbituric acid was then added and the mixture was heated at 80 "C for 30 min. After cooling, 1 ml of 70% (w/v) trichloroacetic acid followed by 3 ml of chloroform were added to each sample tube. The samples were then vortexed and centrifuged, The accumulation of thiobarbituric acid-reactive products was measured at 535 nm, and the lipid peroxide formed was estimated by using a standard curve of MDA in a similar reaction mixture. This procedure provides a similar extinction coefficient for MDA (1.56 X lo5 M" cm") as described by Wills (18).

RESULTS AND DISCUSSION
Characterization of the Purified Lysosomes-Our methods of tissue extraction and differential centrifugation routinely produced enriched (9-to 10-fold) lysosomes which were highly latent (16.7 3.2% free). By using free flow electrophoresis (Table I), we were able to isolate more highly purified (up to 60-fold enriched) native rat liver lysosomes with very low percentage free activities (about 20% free) of P-N-acetylglucosaminidase. The purified lysosomes were observed to exhibit relatively low levels of cytochrome c oxidase and glucose-6-phosphatase activities (Table I). These fractions contained only trace levels of catalase activity (data not shown). Thus, in agreement with the reports of other groups (12,19), lysosomes isolated by free flow electrophoresis were reasonably free of other subcellular contamination.
Peroxidation of Lysosomes-Aerobic oxidation of DHF generates large steady state levels of superoxide anions (OF), which have been suggested to generate additional active oxygen radicals capable of inducing lipid peroxidation (20, 21). The effect of the induced peroxidation on lysosomal membrane integrity was examined (Fig. lb). Incubation of the lysosomes in buffer alone resulted in 29.3 f 2.8% free activity, indicating that the lysosomes still remained relatively intact. Neither Fe"-ADP alone nor DHF alone significantly in- creased the percentage free activities. However, addition of DHF and Fe3+-ADP together resulted in greatly elevated percentage free activity of P-N-acetylglucosaminidase (81.1 f 5.6% free) that was highly significant ( p < 0.001). The possibility of activation of 8-N-acetylglucosaminidase by the radical-generating system was ruled out as follows. At the end of the incubation, samples incubated with or without the radical-generating system exhibited similar total activities of P-N-acetylglucosaminidase which did not change appreciably over the time of incubation.
Time-course studies were conducted to compare the rate of lipid peroxidation with changes in free activity of P-N-acetylglucosaminidase (Fig. 2). With the additions of both Fe3+-ADP and DHF, lipid peroxidation occurred very rapidly and appeared to have reached a maximum after 10 min of incubation. However, the loss of lysosomal latency occurred a t a relatively slower rate and reached a maximum a t about 30 min. As a comparison, only an insignificant increase occurred in the level of MDA formed (from 1.13 -+ 0.16 to 1.66 f 0.64 nmol of MDA/mg of protein) and a modest, though significant Loss of lysosomal latency is known to be pH dependent (23,24). In a separate experiment, incubation of the purified lysosomes for 30 min at 37 "C in a buffer of 0.12 M KCl, 0.05 M sucrose, and 0.01 M potassium phosphate, pH 5, resulted in 60-70% free activity of P-N-acetylglucosaminidase. The loss of lysosomal integrity under the acidic condition is probably due to endogenous lipolytic degradation of the lysosomal membrane (13). In neutral pH, a similar incubation procedure resulted in only a modest increase in percentage free activity (Fig. 2). Presumably, lysosomal lipolytic enzymes are active in acidic pH range (pH 4 to 51, but not in neutral pH range (25). However, in neutral pH, when the lysosomes were incubated with DHF plus Fe3+-ADP, lipid peroxidation was induced and the loss of latency followed (Figs. 1 and 2). If any of the components were absent, lipid peroxidation did not occur, nor did the subsequent loss of the latency (Fig. 1). If all the components were present and a-tocopherol was introduced to neutralize the free radicals (Table 111), lipid peroxidation was inhibited and the loss of latency was also greatly reduced. Thus, these data strongly suggest that the loss of lysosomal latency is more of a consequence of the free radicalinduced lipid peroxidation and not due to the nonspecific combined effect of the components.
Lack of Protection by Various Active Oxygen Scavengers against Lipid Peroxidation-Once formed, 0; leads to the generation of other active oxygen species such as hydroxyl radical (-OH), hydrogen peroxide (H202) and singlet oxygen (IO2) through the following reaction scheme (9,26,27).
The Reaction b may be catalyzed by Fe3+-ADP as follows (9,27). In an attempt to identify the radical species directly responsible for the induced peroxidation, scavengers for different active oxygen radicals were added to the lysosomal samples before the addition of the radical-generating system. As indicated in Table 11, scavengers of O;, H202, 'OH, and '0, did not appear to prevent the lipid peroxidation or the loss of latent activity of P-N-acetylglucosaminidase. Longer preincubation (up to 5 min at 37 "C) of the scavengers with the lysosomes did not affect the subsequent results presented in Table 11. None of the scavengers tested altered the specific activity of P-N-acetylglucosaminidase.
In studying the effect of lipid peroxidation initiated by the oxidation of NADPH by liver microsomes on lysosomes, Fong et al. (9) presented evidence that 'OH was the radical species responsible for oxidative attack of the membrane lipids. To the contrary, our present data indicate that mannitol and ethanol, both 'OH scavengers, were without effect. One possible interpretation of our results is the inaccessibility of the site of 'OH generation to the scavengers. The O;, H2O2, and Fe"+/Fe2+ that are produced may enter the lysosomal membrane freely and generate .OH there, but the scavengers may not enter as efficiently and thus no protection would be provided. We include 0.05 M sucrose in our system, and since sucrose was reported to scavenge .OH at a tremendous rate ( K = 2.8 X lo9 M-' s-') at low pH (28), we investigated the effect of -OH scavengers in a system containing phosphate buffered saline, pH 7.4, without sucrose. In this system, 6 mM mannitol and 0.85 M ethanol were also without protective effect against lysosomal lipid peroxidation. Alternatively, the radical species promoting the lipid peroxidation might be a

+ Fe3'-ADP-induced lipid peroxidation and loss of lysosomal latency
Lysosomes (150-200 pg of protein/ml) were preincubated with each scavenger for 2 min at 37 "C before the additions of Fe3+-ADP and DHF. After 30 min of incubation, samples were assayed for MDA formation and percentage free acitivity of 0-N-acetylglucosaminidase as described under "Materials and Methods." Values are means of two to four separate preparations. Thus, the data suggest that the damage to lysosomal integrity correlates closely with the extent of lipid peroxidation.

TABLE 111
Protective effect of aTC on DHF + Fe3+-ADP-induced lipid peroxidation and loss of lysosomal latency Lysosomes (150-200 pg of protein/ml) were preincubated with the indicated levels of fvTC in 50 pl of ethanol for 3-5 min at 37 "C before the additions of Fe3+-ADP and DHF. Other conditions were the same as described in Table 11 ADP-iron-oxygen complex similar to that suggested by Svingen et al. (29). Therefore, none of those agents which are specific for active oxygen radicals would be expected to be effective. However, at the moment, there is no conclusive evidence of such a complex. Further experiments will be required to clarify the free radical species responsible for the peroxidation induced in our system.
Protective Effect of a-Tocopherol on the Peroxidation-a-Tocopherol is well known as an anti-oxidant (30). Preincubation of the lysosomes with a-tocopherol inhibited up to 83% of the induced lipid peroxidation (Table 111). Additionally, in our system, preincubation of the a-tocopherol with the lysosomes required at least 3 min to produce the subsequent effective and reproducible extent of protection against the lipid peroxidation. Such a requirement of the preincubation condition seems to suggest that initial interaction of atocopherol with the membrane lipids is important for the subsequent protective effect. Concomitantly, pretreatment with a-tocopherol substantially reduced the percentage free activity of P-N-acetylglucosaminidase of the lysosomes. However, the EtOH in which a-tocopherol was dissolved, partially labilized the lysosomes; this was probably due to the fluidizing effect of ethanol on the membrane (31), as lysosomes incubated with 50 p1 of ethanol alone for 30 min exhibited 51.1% free activity of P-N-acetylglucosaminidase, This is comparable to the level of percentage free activity (52.1%) exhibited by the samples incubated with 3.50 mol of a-tocopherol in 50 pl of ethanol ( Table 111).
The dose-dependent effects of a-tocopherol on the extent of lipid peroxidation and on the changes in percentage free P-N-acetylglucosaminidase activity of the lysosomes were further examined. As shown in Fig. 3a, levels of a-tocopherol that were lower than 1.75 pmol/ml, provided proportionately decreasing degrees of protection against lysosomal lipid peroxidation; levels of a-tocopherol that were greater than 3.50 pmol/mI did not inhibit further the induced peroxidation (data not shown). At different concentrations, a-tocopherol also reduced the percentage free activity of /3-N-acetylglucosaminidase to an extent which was closely associated with the various degrees of reduction in lipid peroxidation (Fig. 3b). Conditions of preincubation and incubation were as described in Table 111. Values are means -C S.D. of three separate preparations, out major effect in our system, the inhibitory effect of atocopherol might be due to its ability to interrupt the propagation of the free radical chain reaction already initiated in the membrane lipids (30). Our observations seem to agree with the proposal of Lucy (33) which suggests that tocopherol physiocochemically forms a complex with the fatty acyl chains of polyunsaturated phospholipids, particularly those derived from arachidonic acid in lipid membranes with the following functional consequences: i) inhibition of peroxidative destruction of polyunsaturated fatty acids in the membranes; ii) prevention of permeability (or leakage) of biological membranes containing relatively high levels of polyunsaturated fatty acids; and iii) possible prevention of degradation of the membrane phospholipids by membrane-boundphospholipase.
Our experiments (Fig. 2) indicate that the formation of lipid peroxides in the membrane preceded the disintegration of the lysosomes. The data, however, do not distinguish whether the accumulation of the lipid peroxides (or the loss of membrane polyunsaturated fatty acids) and/or a subsequent enzymatic process, such as lipolysis, would be directly responsible for the degradation of the lysosomal membrane.
In conclusion, our results clearly demonstrate the susceptibility of lysosomes to lipid peroxidation induced by free radicals. Since free radicals are believed to exist in uiuo (34) and are capable of promoting lipid peroxidation via a NADPH-dependent microsomal enzyme system in vitro (9, 35), we have carried out a preliminary study of the susceptibility of the purified lysosomes to free radicals generated by the microsomal NADPH oxidase system. The liver microsomes were prepared as previously described (9,35). In a typical experiment, purified lysosomes (0.2 mg of protein) in 1 ml of the same buffer described in Fig. 1 were incubated with 0.1 mg of the microsomes in the presence of 0.4 mM NADPH and Fe3'-ADP (0.1 mM FeCb, 1 mM ADP). Lipid peroxidation was greatly induced as indicated by the formation of 16-20 nmol of MDA in 30 min at 37 "C. Concomitantly, the lysosomal latency was also lost as indicated by the elevated percentage free (80-90%) activity of P-N-acetylglucosaminidase. The lysosomes incubated with the microsomes alone or in systems containing no NADPH or Fe3+-ADP exhibited no more than 5% of the MDA formed in the complete system and showed only a modest level of free activity of 0-Nacetylglucosaminidase (33-40% free). These data further support the conclusion that the observed susceptibility of the purified lysosomes to the free radicals generated by the DHF plus Fe3+-ADP system is not unique to that system; other systems, whether enzymatic or nonenzymatic, capable of generating free radicals may induce similar peroxidative damage on the lysosomal membrane. Our data demonstrate a similar facilitory effect of Fe3+-ADP on the peroxidative reactions as reported by other studies (9,35). The concentration of inorganic iron in the liver cytosol has been estimated in the range of 1.01 X to 1.28 X M, whereas the cytosolic concentration of ADP is believed to be in the millimolar range (36). Since the concentrations of both components are at levels compatible with promoting the process described, it is reasonable to consider the possibility that a similar destructive reaction on lysosomes may be catalyzed in uiuo by certain free radical-generating oxidoreductase systems, especially when free radical scavengers such as a-tocopherol are absent.
helpful discussions and review of this work.