Eur. J. Biochem. 50,595-602 (1975)

Lipid Peroxidation in Isolated Hepatocytes Johan HOGBERG, Sten ORRENIUS, and Robert E. LARSON Department of Forensic Medicine, Karolinska Institutet, Stockholm (Received July I 1 /September 30,1974)

Intracellular lipid peroxidation was initiated by the addition of ADP-complexed ferric iron to isolated rat hepatocytes and the reaction monitored by the thiobarbituric acid method or by measurement of the formation of conjugated dienes. Both the production of malondialdehyde (thiobarbituricacid-reacting substances) and of conjugated dienes was dependent, on the ADP . Fe3+ concentration in a dose-related fashion. Malondialdehyde formation stopped spontaneously within 20 min after the initiation of the reaction and the plateau reached was also related to the ADP . Fe3+ concentration. Control experiments revealed that more than 90 % of the malondialdehyde accumulating during the incubation period could be ascribed to intracellular production. The cellular NADPH/NADP+ ratio was always high and only slightly decreased upon ADP . Fe3+-induced lipid peroxidation which, however, was associated with a marked decrease in the cellular glutat hione concentration. The rate of accumulation of malondialdehyde as well as the final level reached during ADP . Fe3+-initiated lipid peroxidation was increased by the addition of chloral hydrate. This apparent stimulatory effect could, however, be ascribed to the inhibition of the mitochondria1 oxidation of the malondialdehyde formed during cellular lipid peroxidation, thus allowing more malondialdehyde to accumulate during the process. ADP . Fe3+-induced cellular lipid peroxidation was associated with a decrease in the concentration of glutathione. Also, lowering of the intracellular glutathione level by the addition of diethyl maleate or by simply preincubating the hepatocytes (up to 50 min) promoted the ADP .Fe3 malondialdehyde production and formation of conjugated dienes. Furthermore, when cellular glutathione concentration had been lowered by preincubation of the hepatocytes, significant malondialdehyde production could be observed even at ADP . Fe3+ concentrations which were too low to induce measurable lipid peroxidation in fresh hepatocytes. It is thus concluded that glutathione has an important role in the cell defence against lipid peroxidation and suggested that the isolated hepatocytes provide a suitable experimental model system for the characterization of this and other possible cellular defence mechanisms and how they are affected by the nutritional status of the donor animal. +

That peroxidation of the polyunsaturated lipids of intracellular membraneous structures can occur and is associated with the hepatotoxicity of several chemicals is well documented [I -41. Heretofore, investigations of lipid peroxidation have employed either isolated cellular organelles (mitochondria, microsomes or lysosomes) or isolated, perfused whole livers, or intact animals. Studies using isolated organelles, while providing an ability to measure the rapid, short-course events associated with lipid peroxidation, suffer from the elimination of potentially protective -~ Enzymes. Hyaluronidase or hyaluronate 4-glucanohydrolase (EC 3.2.1.35); collagenase (EC 3.4.24.3).

Eur. J. Biochem. 50 (1975)

intracellular systems. The significance of the observations in terms of real impact on the cell is thus open for interpretation. Studies which employ isolated whole organs or intact animals, on the other hand, because of sampling difficulties cannot follow closely the dynamic changes associated with lipid peroxidation. The development of procedures for the isolation of relatively intact hepatocytes by enzymatic perfusion of rat liver [5 - 71 has provided a potential for studying intracellukarlipidperoxi ation which might circumvent many of the short-col lings of earlier models. In the studies presented herein lipid peroxidation was initiated in suspensions of isolated rat hepato-

Lipid Peroxidation in Isolated Hepatocytes

596

cytes by the addition of ferric iron complexed to adenosine diphosphate (ADP . Fe3+). The primary aim was to determine whether peroxidation of intracellular lipids could be induced by this means in a manner analogous to that described by Hochstein and Ernster [l] for enzymatic peroxidation of rat liver microsomes. These initial studies focus upon the measurability of the products of lipid peroxidation in the cells, upon some of the temporal aspects of the peroxidation as initiated by ADP . Fe3+, and upon some of the intracellular characteristics that apparently influence the extent or measurability of the peroxidative attack.

MATERIALS AND METHODS Isolation of Hepatocytes Intact hepatocytes were isolated from enzymatically perfused rat livers according to a method described by Quistorff et al. [6]but with several modifications, essentially according to Seglen [7]. Unfasted male Sprague-Dawley rats weighing 200- 250 g were anesthetized with ether and the liver, portal vein and abdominal vena cava exposed via a midventral incision. Heparin sodium (200 units in 0.2 ml) was injected into the vena cam and the portal vein was rapidly cannulated. In situ perfusion of the liver was immediately begun with a modified Locke solution (120 ml) containing 2 % defatted bovine serum albumin (method of Chen [8]) and 0.5 mM ethyleneglycol bis(2-aminoethylether)N,N'-tetraacetic acid. The perfusion medium at 37 "C was passed from the reservoir via a roller pump into a thermostated oxygenation cylinder prior to its passage through the liver. The liver was then rapidly excised under continuous perfusion. Livers which did not clear completely upon initiation of the in situ perfusion were immediately discarded and a new donor rat was prepared for surgery. Perfusions were conducted at 40cm H 2 0 which is considerably higher than physiological pressures in order to ensure as complete perfusion as possible. After 2 min perfusion with the original Lockesalbumin solution the perfusion medium was changed to a Hanks buffer solution (100ml) containing 2% albumin, 4 mM CaCI2, 75 mg hyaluronidase (Boehringer Mannheim GmbH, Mannheim, Germany), and 120 mg collagenase (Boehringer Mannheim GmbH). During this perfusion the entire liver was placed in the perfusion reservoir and allowed to bath in the medium. That perfusion was continuing could be checked periodically by turning off the pump and observing the fall of the level of the fluid in the oxy-

genation chamber. Enzymatic perfusion was carried out for 8 min. At the end of the perfusion period the liver was markedly swollen and the parenchyma seen to be loosely suspended within the capsule. The liver was placed in a large petri dish while still being perfused and the capsule quickly incised several times using fine scissors. The perfusion was terminated and the liver was transferred to a large conical flask containing 50 ml of the enzymatic perfusion medium and shaken vigorously in a water bath at 37 "C under a carbogen atmosphere for 5 min. The cells were then harvested by gentle centrifugation (50 x g for 2 min) at room temperature. The supernatant was aspirated off and discarded and the softly pelleted cell suspension so obtained was gently resuspended in Krebs-Henseleit buffer containing 2% defatted albumin and 5 mM glucose and the cells again harvested. The cells were thus washed two times and then finally resuspended to a final volume of 12- 15 ml in the Krebs-Henseleit buffer in order to give a stock suspension containing 3 - 4 x 10: cells/ml. The washing procedure essentially follows the one described by Berg et al. [9], who harvested hepatocytes with only about 2 % contamination by Kupffer cells. The entire harvesting and washing procedure was conducted at room temperature and the stock suspension of cells was kept at room temperature under a carbogen atmosphere. In this way 4- 6 x lo8 cells which were 98 - 100% viable (as estimated by trypan blue exclusion) were obtained per liver. The basal oxygen consumption of the cells, as measured with a Clarke-type oxygen electrode, was not increased upon the addition of 10 mM succinate to incubating cell suspensions. The stock suspension remained 98 - 100 % viable for up to 3 h at room temperature under carbogen but in all cases the experiments were completed within 2 h of harvesting. Experimental Earlier work by Moldeus et ul. [lo] has shown that isolated cells (from the livers of fasted rats) were enriched in NADPH content when glucose was present in the incubation medium. Since we were studying a reaction possibly supported by NADPH-dependent cytochrome c reductase activity, we wished to assure the presence of sufficient amounts of the reduced nucleotide. Under these conditions incubations could be carried out for up to 120 min with no significant changes either in total cell count or viability as estimated by the commonly employed trypan blue exclusion criteria. Incubations were carried out at 37 "C in a shaking bath undcr a carbogen atmosphere using flasks or Eur. J. Biochem. 50 (1975)

J. Hoigberg, S. Orrcnius, and R. E. Larson

tubes. The cells were suspended in Krebs-Henseleit buffer (4x lo6 cells/ml) which contained glucose (5 mM). The lipid peroxidation was started by addition of iron complex after periods of preincubation of the cell suspensions ranging from 0 to 50 min. The iron complexes used were ADP . Fe3+ (100 mM ADP and 1.2mM Fe3+[l]) or pyrophosphate.Fe3+ (1.2mM sodium pyrophosphate and 1.2mM Fe3+[1,4]). Lipidperoxidation was followed by periodically removing aliquots (0.2 ml) from the incubation mixtures and measuring malondialdehyde (thiobarbituric-acid-reacting substances), content by the method of Bernheim et af. [ll]. Experiments were conducted in which these measurements were made on whole incubate aliquots or upon reharvested cells from the incubates and upon the supernatant above the reharvested cells. Additionally, the formation of diene conjugates in reharvested cells was followed by measuring the increase in absorbance of extracted lipids at 234nm. The cells in 1 ml incubate were pelleted by centrifugation at 100 x g for 2-3 min and lipids were directly extracted from the separated pellets by the addition of 20 volumes (10 ml) of chloroform - methanol (2: 1). After at least 1 h of extraction with intermittent shaking, the solution was filtered. The filtrate was washed by adding 2 ml distilled water and shaken vigorously. The phases were allowed to separate by standing overnight in the cold under a nitrogen atmosphere and the lower phase was used for diene measurement. The absorption at 535 nm (malondialdehyde) was measured using a Beckman model B spectrophotometer and for absorbance of the lipid extracts at 234 nm (dienes) an Aminco DW-2 UV-VIS spectrophotometer was used. The content of NADPH and NADP' was also estimated in aliquots from whole incubate, reharvested cells and the supernatant above the reharvested cells. The method employed was that described by Klingenberg [12]. Glutathione content of 1 ml aliquots of the incubates was estimated by the fluorometric method described by Cohn and Lyle [13] using an AmincoBowman model spectrophotofluorometer. For comparative purposes, microsomes were isolated from the whole liver of nonfasted rats essentially according to Dallner [14]. All chemicals and reagents employed were of commercially available reagents quality.

RESULTS AND DISCUSSION Since much of the work reported in the literature and many of our own studies of lipid peroxidation in isolated organelles have employed ferric iron complexed with ADP, we elected to initiate our studies Eur. J. Biochem. 50 (1975)

591

o.20

1

20 30 40 Time (min) Fig. 1. Ejficted of ADP . Fe3+ complex on mulondiuldehyde production in isokuted hepatocytes. Each point is the mean of results from three different incubations and cell preparations. About 40 x lo6 cells were incubated in a final volume of 12 ml. At the times indicated samples of 0.2 ml were taken for malondialdehyde measurements. Incubation was started on addition of ADP . Fe3+ complex. The final Fe3+ concentra), (w), 40pM tions were: 10pM (M20pM (0-0) and 80 pM (u) 0

10

on isolated hepatocytes with this complex. In Fig. 1 are presented the effects of increasing concentrations of Fe" + upon malondialdehyde (thiobarbituric-acidreacting substances) production in suspensions of isolated hepatocytes. Small increases in malondialdehyde were usually evident at a concentration of Fe3+ as low as 20 pM.While a marked increase in malondialdehyde production was evident at 80 pM Fe3+, we were concerned that this might be accompanied by outright destruction of the cell and therefore confined our further studies to the milder levels of peroxidation afforded by 20 - 60 pM concentrations of iron. Evidence of diene conjugation, as estimated by the increased absorbance of lipid extracts at 234 nm, was found to be associated with lipid peroxidation both in isolated microsomes and in hepatocytes (Fig. 3). Both malondialdehyde and conjugated dienes increased in a dose-related fashion with increasing concentrations of ADP Fe3+ (Table 1). Because such small amounts of lipids could be extracted from the amount of total liver tissue represented by the cells in our incubates, the absolute absorbance differences obtained at 234nm were quite low. They were, nevertheless, measurable and reproducible within a given experiment and the spectrum was quite clear when suitable instrumentation was employed. A broad band of intensely absorbant material with an absorbance

-

Lipid Peroxidation in Isolated Hepatocytes

598

Table 1. Effect of ADP-Fe3+ complex on malondialdehyde and conjugated diene production in isolated hepatocytes About 80 x 106 cells were incubated in a final volume of 12 ml for 40min. After 40min, 0.2ml was taken for malondialdehyde measurements and the remaining cells were spun down, and lipids extracted from the pellet. The AA234figures are calculated from difference spectra (see Fig.3) and the reference always was a lipid extract from cells, incubated in half that concentration of ADP . Fe3+ as in the sample

Final Fe3+ conc.

A A234

Expt-control

AA535 Exptcon trol

0 0 0.004 0.009

0 0 0.060 0.195

PM 10 20 40 80

01, maximum at about 260nm was also found in the lipid extracts. As lipid peroxidation increased this band tended to mask the 234 nm absorbance peak. This difficulty was overcome to a certain extent by using each successive sample as a reference material for the next higher level of peroxidation. The exact nature of this material and its relationship to the peroxidation are the subject of continuing investigations. Since there appeared to be good correlation between malondialdehyde and conjugated diene production the bulk of our observations reported herein are in terms of malondialdehyde. Experiments in which the cells were reharvested by gentle centrifugation and the pellct and supernatant separately analyzed for malondialdehyde revealed that only about 10% of this material was present in the supernatant. This was taken as evidence that the peroxidation was initiated within the cells and not upon extraneous debris or dead, lysed cells. Moreover, it was found that rehartesting the cells under the conditions used in their original isolation did not result in complete recovery of them after 40min of incubation. As many as 10% of the cells remained suspended in the supernatant at that time. Harder centrifugation resulted in greater recovery and revealed that much of the malondialdehyde measured in the supernatant was contained in cells suspended therein. Thus, it was concluded that most of the peroxidation was actually occurring within the cells. Time-course studies revealed that the total cell count in the incubation suspensions and cell viability, as measured by trypan blue exclusion, remained rather constant for at least 1.5 h of incubation. A decline of only 5-6% in the number of viable cells was observed and repeated experiments in which cell counts and viability estimates have been periodically made regularly bear this out. The same picture was

1

I

I

7

I

I

1

0 10 20 30 40 50 60 70 80 Time (min) Fig. 2. Effect of preincubation upon malondialdehyde production in isolated hepatocytes in the presence of A D P . Fe3+. About 40 x lo6 cells in a final volume of 10 ml was incubated in the presence of ADP . Fe3' (20 1M final Fe3+ conc.). The ADP .Fe3+ solution was added after different preincubation periods (0, 25 and 50 min). Samples for malondialdehyde measurements were taken at times indicated. The preincubation was performed under the same conditions except for the absence of ADP . Fe3+.Each point represents the mean of results from 3-5 different cell preparations. Arrows indicate time of ADP . Fe3+addition

obtained in either control incubates, in which no iron was present, or in the incubates in which peroxidation had occurred. Malondialdehyde production was essentially completed within about 10 min of the addition of ADP Fe3+ to the incubates (Fig. 1). However, in the timecourse studies an interesting relationship was observed. Under conditions stated above only relatively low levels of malondialdehyde and conjugated dienes were produced when the ADP . Fe3' was added to the fresh cells (Fig. 1, Table 1). The production of these materials was sequentially increased, however, with longer preincubation times prior to the addition of iron (Fig.2). That this increase was not due to endogenous production associated with the preincubation but was related to the addition of iron is demonstrated in Fig.2 by the fact that the initial malondialdehyde content of the cells remained essentially unchanged over the 50-min preincubation period. Furthermore, diene conjugate production followed an analogous pattern to that obtained for malondialdehyde (Fig. 3). These observations led to the speculation that as the incubation proceeded the cells became more Eur. J. Biochem. 50 (1975)

J. Hogberg, S. Orrenius, and R. E. Larson

599

f

A

B

AA.0 002

I

2x)

I

I

230 240 X (nrn)

I

220

250

J

/

230

I

I

240

250

X (nm)

Fig. 3. Dqference spectra of lipid extract rom (A) microsomes incubated in the presence of ADP . F and ascorbic acid, and ( B ) isolated hepatocytes undergoing lipid peroxidation in the presence of A D P . F$+. (A) About 8 mg microsomal protein was incubated for 20 min at 37°C in the presence of ascorbic acid (90 pM final conc.) Fe3' (120 pM final conc.) and pyrophosphate (120 pM final conc.) in a final volume of 1 ml. The same mixture was used as control but it was not incubated. (B) About 40 x lo6 cells were incubated in a final

volume of 12 ml. Different preincubation times preceded the addition of ADP . Fe3+ (40 pM final Fe3' conc.). The cells were reharvested by centrifugation after 10 min exposure to the iron and lipids extracted from the pelleted cells. The dashed line represents a difference spectra between samples preincubated 40 rnin (sample cuvette) and 20 rnin (reference cuvette). The unbroken line represents a difference spectra between samples preincubated 40 min (sample) and 0 rnin (reference)

permeable to the ADP . Fe3+ complex. This would be in the face of the fact that they were still apparently able to exclude trypan blue. Therefore, the more water-soluble pyrophosphate complex of iron that was known to initiate lipid peroxidation in microsomes [4] was compared with ADP complexed iron. In Table 2 are given the comparative activities of ADP and sodium pyrophosphate (PPi) complexed Fe3+.It was a constant finding that ADP . Fe3+ gave a higher malondialdehyde production, both in fresh cells and in preincubated cells, than did the PPi .Fe3+ complex. The response to P P i . Fe3+ was always very low. Of further interest was the finding that the effect of preincubation always seemed to be the same for the two complexers as is illustrated by the ratio of the malondialdehyde production in the 50-min preincubated cells to that in fresh cells (Table 2). Since PPi . Fe3+ was found to be on the order of four-fold more active in stimulating malondialdehyde production in isolated microsomes (Table 2), it seemed that this should be reflected in the cells if the plasma membrane had merely become unselectively permeable during preincubation. Furthermore, Saltman et al. [15] have reported that the uptake of organically complexed iron by cells in liver slices is not so dependent upon the permeability characteristics of the cell membrane as upon the complexer. An additional argument against the idea that increased permeability was responsible for the apparent enhancement of per-

Table 2. Comparison of ADP and pyrophosphate f P P i ) complexed iron on malondialdehyde production in isolated rat hepatocytes and liver microsomes About 32 x 106 cells were incubated in a final volume of 8 ml. Either initially (0 min) or after 50 rnin (50 min) of preincubation, lipid peroxidation was started by addition of the iron complexes to a final Fe3+ concentration of 60 pM. Aliquots were taken for malondialdehyde assay after 10 min incubation in the presence of the iron. The malondialdehyde production rate is expressed as AA535 x (106 cells)-' xmin-'. Microsomes (3 mg protein) were incubated for 3 min at 37 "C in the presence of NADPH (final conc. 80 pM) and iron complex (final Fe3' conc. 105 pM). The malondialdehyde production rate is expressed as AA53s x mg microsomal protein-' x min-'

Eur. J . Biochem. 50 (1975)

Complex

Hepatocytes Omin

50min

Microsomes 50/0min

AA,?, x(106 cells)-' x min-'

A D P . Fe3' P P i . Fe3+

0.019

0.137

0.007

0.048

7.2 6.9

Af4535 -

x

y

protein- x min-' 0.150 0.677

oxidation can be found in Fig. 1. For the concentrations of iron employed (30-60 pM), once malondialdehyde concentrations had stabilized after 10 rnin of incubation of fresh cells, no further increase was observed for more than 40 min of incubation. Given these observations, it appears that the preincubation

Lipid Peroxidation in Isolated Hepatocytes

600

effects seen in our cell suspensions are more probably related to intracellular factors which can either alter the apparent response to the iron or alter the redox state of the iron than to an increased permeation of the iron into the cell. This area, too, is the subject of continuing study. Recknagel and Ghoshal[3] pointed out that malonaldialdehyde could be rapidly metabolized by mitochondrial oxidation. This reaction is, however, effi-

0.4

L

1

t

I

0'1

I

I

I

I

-

I

1

1

0 10 X) 3 0 4 0 5 0 6 0 7 0 8 0 Time (rnin) Fig. 4. Effect of' ADP . Fe3+ and chloral hydrate upon rnalundialdehyde production in isolated hepatocytes. About 4.9 x lo6 cells were incubated in a final volume of 1.2 ml. After preincubation (0, 20 and 50 min) the peroxidation was started by addition of ADP . Fe3+ (final Fe3+ conc. or)addition of ADP . Fe3+ and chloral 60 pM) (M hydrate (final F$+ conc. 60 pM and chloral hydrate conc. 10 mM) (M Malondialdehyde ). was estimated from 0.2 ml samples. Arrows indicated time of ADP . Fe3+ or ADP . Fe3+ chloral hydrate addition

+

ciently inhibited by chloral hydrate (K. 0. Lindros, personal communication). In order to investigate the potential role of mitochondrial oxidation of malondialdehyde in incubated cells relative to the apparent resistance of fresh cells, experiments were designed in which chloral hydrate was added to the cell incubates to block aldehyde metabolism. Chloral hydrate did not influence malondialdehyde production in isolated microsomal systems in concentrations up to 20 mM. The addition of chloral hydrate (10 mM) to liver cell incubates which were concomitantly exposed to ADP . Fe3+ resulted in about a 50% increase in measurable malondialdehyde in these cells (Fig. 4). However, this increase was observed to be superimposed on the apparent malondialdehyde production at all preincubation intervals. Furthermore, the malondialdehyde production followed a similar time-course after the addition of iron either in the presence or absence of chloral hydrate (not documented) i.e. malondialdehyde production was apparently completed within about 10 min. Assuming that chloral hydrate did, in fact, only block aldehyde metabolism in the cells, it appeared that mitochondrial oxidation was a factor in the assayability of malondialdehyde in the peroxidating cells. Thus, it could be estimated that in the presence of mitochondrial oxidation the malondialdehyde assays underestimated the true level of lipid peroxidation by about 50%. Active mitochondrial oxidation of malondialdehyde did appear to be ongoing at all stages of the aging of the incubating cells, however. To further study the roles of various intracellular controlling systcms for lipid peroxidation the availability of NADPH for enzymatic peroxidation was measured. The results of one typical experiment are given in Table 3. Total NADP+ and NADPH values

Table 3. Eflect ojpreincubation and ADP.Fe3' treatments on NADPH and NADPt concentrations in isolated hepatocytes About 50 x lo6 cells were incubated in a final volume of 12 ml. ADP . FeJt was added either directly or after 25 and 50 min preincubation. Final Fe3' concentration was 30 pM.Cells were reharvested (from 1.0 ml samples) by centrifugation and the amount of NADPH in the pellel (cellular NADPH) and the supernatant (unsedimented NADPH) was estiimated. NADP+ was measured on the total incubation medium Treatment

None

Time of treatment

Cellular NADPH

min

nmol/ml

LJnsedimented NADPH

Total NADP+

Malondialdehyde production dAS,,/I06 cells

0

25

1.5

0.5

Preincubation

25 50

22 22

0.7 1.0

1.1 0.7

ADP . Fe3+

10

19

1 .o

1.1

0.03

21

0.8 1.3

1 .I 0.9

0.04

20

Preincubation

+ ADP . Fe"

25 50

+ 10 + 10

-

.

0.07

Eur. J . Biochem. 50 (1975)

J. Hogberg. S. Orrenius. and R. E. Larson

in the incubates ranged from 25-26 nmol/ml incubation medium or approximately 6- 6.5 nmol/106 cells. These values are somewhat higher than the 4- 5 nmol/106 cells reported by MoldCus et al. [lo] who used cells isolated from the livers of fasted rats. Moreover, the NADPH/NADP+ ratio was also somewhat higher (over 90% was in the reduced form) in our incubates. The fact that nonfasted rats were used as the hepatocyte donors and glucose was included in the washing medium as well as the incubation medium in the present studies may account for this difference. Moldeus et al. [lo] observed that when glucose was present an increased amount of NADPH was maintained. In any event, the availability of NADPH did not appear to be limiting to lipid peroxidation at any preincubation interval and, indeed the changes during peroxidation were restricted to some few per cent of the available amount during incubation. Neither did the cells appear to become strikingly more leaky to the nucleotides. Reharvesting of the cells revealed that only very small proportions of the nucleotides were apparently present in the medium (Table 3). The small progressive increase in unsedimented NADPH that was seen can also be accounted for in part by the earlier observations that as much as 10% of the cells may not be reharvested in the usual way after 40 or 50 rnin of incubation. Thus, a considerable portion of the NADPH which was found in the medium might, in fact, have been contained within cells that remained suspended therein. The rather sharp decrease in the NADPH/NADP+ ratio associated with the peroxidation of fresh cells was a constant finding, while changes in preincubated cells were of a more uncertain nature, and in any event less. This might possibly be a reflection of greater activity of NADPH-dependent protective devices (e.g. glutathione reductase) in the fresh cells. Fluorometric assay of the glutathione content of the incubates revealed a decline in glutathione during incubation of non-peroxidized cells to 75% of the initial value and a further decline to 58% upon ADP . Fe3+ addition (Table 4). When an alkylator of glutathione, diethyl maleate [17] was added to fresh cells a prompt diminution of glutathione was noted (Table 4). By 5 rnin exposure to the diethyl maleate the glutathione content of the fresh cell incubate was lowered to about 25 % of the initial value. Moreover, the addition of diethyl maleate to fresh cell incubate 2 min prior to the addition of ADP . Fe3+ greatly enhanced the rate of malondialdehyde production in these cells to the extent that the differences previously seen associated with varying preincubation periods were no longer evident (Fig.5). Diethyl maleate in these concentrations did not stimulate microsomal malondialdehyde production. Eur. J. Biochem. 50 (1975)

601

Table 4. Efcct of diethjdmaleate,preincubation and A D P . Fc3 + treatments in vitro on glutathione content in isolated hepatoCJlCS

About 4 . 0 lo6 ~ cells were incubated in a final volume of 1.2 ml. Diethylmaleate (2 pl) was added at the start of the incubation. ADP . Fe3+ (final Fe3+ conc. was 50 pM) was added either directly or after 20 rnin preincubation. 1 ml was taken for glutathione assay. The glutathione content of the untreated control was 5.1 pg/106 cells Treatment

Time of treatment min

None Diethylmaleate

5 10 15

30

10 30 10

ADP-Fe3

+

Preincubation

+ ADP-Fe3+

0'1

I

20

I

2l 100 25 23 21 25 98 75 83

0

Preincubation

Glutathione content

+ 10

58

I

20 30 40 Time (rnin) Fig. 5. Elimination of prcincubarion eJects on malondialdchyde production by diethyl maleate treatment in vitro in isolated hepatocytes. About 38 x lo6 cells were incubated in a final volume of 8 ml. Diethylmaleate (50 pl) was added 5min before the incubation was begun. After 0 and 20 rnin incubation peroxidation was started by the addition of ADP . Fe3+ (final Fe3+ conc. 60 pM). 0.2 ml was taken for nialondialdehyde measurement at times indicated. Arrows indicate time of ADP . Fe3+ addition. ADP . Fe3+ (W) and ADP . Fe3+ diethyl maleate (M)

0

10

+

It appears that the status of such factors as glutathione which can serve in a protective capacity against lipid peroxidation in the cell can be quite determinant of the cells vulnerability to peroxidative attack. This could already be anticipated from studies which have employed isolated organelles [18] and erythrocytes [19,20]. It will continue to be of importance to acknowledge this in the interpretation of studies which

602

J. Hogberg, S. Orrenius, and R. E. Larson: Lipid Peroxidation in Isolated Hepatocytes

use the isolated hepatocytes as a model system. Other potential redox systems in the cell, for example, ascorbate or tocopherol also must be considered in this light. From the findings presented herein it seems safe to conclude that intracellular lipid peroxidation can be initiated by the addition of iron complexes to suspensions of isolated hepatocytes. The dynamics of the process can be readily followed by monitoring malondialdehyde or conjugated diene production but these dynamics are modified dramatically according to the apparent integrity of intracellular systems. At this point the sites of peroxidative attack have not been localized nor has the toxic impact on the cell of the levels of peroxidation studied here been analyzed. Nevertheless the apparent oxidative consumption of NADPH during the process suggests the likelihood that the lipid peroxidation was at least in partenzymatically mediated as described by Hochstein and Ernster [l ] for microsomes. That mitochondrial oxidation of the malondialdehyde formed during peroxidation can occur even in incubated, aged cells seems to be demonstrated by the studies in which chloral hydrate was employed. Tappel et al. [20] have related malondialdehyde production to the toxic denaturation of proteins through the formation of Shiffs base products. Whether this toxicity can be demonstrated over short time-courses in isolated cell incubates remains to be seen. In any event, active mitochondrial oxidation of this toxic product (malondialdehyde) would be protective to the cell. Of obvious importance in the interpretation of future investigations on lipid peroxidation in isolated liver cells will be the nutritional status and biochemical integrity of the donor animal. Fasting and other factors may serve to lower the available NADPH for enzymatically mediated peroxidation. Conceivably the presence of substrates and inhibitors of the hepatic microsomdl monooxygenase pathways would also influence any enzymatically mediated component of intracellular lipid peroxidation [21]. Moreover, the level of glutathione (and probably other antioxidant systems) is of great importance in the manifestation of peroxidative attack. Recently Tateishi et al. [22] reported decreased levels of glutathione in the livers of starved rats. This finding may bear some relationship to the present observation that glutathione levels decreased in isolated hepatocytes as the preincubation progressed. The striking effect of diethyl maleate upon the intracellular glutathione content and the associated

enhancement of malondialdehyde production upon the addition of small amounts of complexed iron emphasizes this. It seems quite probable then, that the falling levels of glutathione were involved in the progressively increased susceptibility of the cells to peroxidative attack with ageing during incubation. This study was supported by a grant from the Swedish Medical Research Council (proj. no. 03X-2471). The authors wish to thank Mrs Annika Kristoferson for the technical assistance. We also want to thank Dr B. Arborgh for valuable discussions during development of the cell preparation technique.

REFERENCES 1. Hochstein, P. & Ernster, L. (1964) Ciha Foundation Symposium on Cellular Injury, pp. 123- 134. 2. Tappel, A. L. (1973) Fed. Proc. 32, 1870-1874. 3. Recknagel, R. 0. & Ghoshal, A. K. (1966) Lab. Invest. 15, 132- 146. 4. Hogberg, J., Bergstrand, A. & Jakobsson, S. V. (1973) Eur. J . Biochem. 37, 51 - 59. 5. Berry, M. N. & Friend, D. S. (1969) J. Cell Biol. 43, 506- 520. 6. Quistorff, B., Bondesen, S. & Grunnet, N. (1973) Biochim. Biophys. Acta, 320, 503- 516. 7. Seglen, P. 0. (1972) Exp. Cell Res. 74, 450-454. 8. Chen. R. F. (1967) J. Biol. Chem. 242. 173-181. 9. Berg,’T., Boman,’D. & Seglen, P. 0: (1972) Exp. Cell Res. 72. 571 - 574. 10. MoldCus,’P., Grundin, R., Vadi, H. & Orrenius, S. (1974) Eur. J. Biochem. 46, 351 - 360. 11. Bernheim, F., Bernheim, M. L. C. & Wilbur, K. M. (1948) .IBiol. . Chem. 174, 251- 264. 12. Klingenberg, M. (1970) Methoden der Enzymatischen Analyse (Bergmeyer, H. U., ed.) pp. 1975-1990, Verlag Chemie, Weinheim. 13. Cohn, V. H. & Lyle, J. (1966) Anal. Biochem. 14, 434440. 14. Dallner, G. (1963) Acta Parhol. Microbiol. Scand. Suppl. 166.

15. Charley, P., Rosenstein, M., Shore, E. & Saltman, P. (1960) Arch. Biochem. Biophys. 88, 222 - 226. 16. Reference deleted. 17. Boyland, E. & Chasseaud, L. F. (1970) Biochem. Pharmacol. 19, 1528-1533. 18. Christoffersen, B. 0. (1968) Biochern. J. 106, 515-522. 19. Cohen, G. & Hochstein, P. (1965) Biochemistry, 2, 1420 - 1424. 20. Goldstein, B. D. (1973) Arch. Environ. Health, 26, 279 - 280. 21. Bidlack, W. R. & Tappel, A. L. (1973) Lipid.7, 8, 177182. 22. Orrenius, S., Dallner, G. & Ernster. L. (1964) Biochem. Biophys. Res. Commun. 14, 329- 334. 23. Tateishi, N., Higashi, T., Shinyd, S., Naruse, A. & Sakamoto, Y. (1974) J . Biochem. 75.93- 103.

J. Hogberg and S . Orrenius, Department of Forensic Medicine, Karolinska Institutet, S-104 01 Stockholm 60, Sweden R. E. Larson’s present address : Department of Pharmacology and Toxicology, Oregon State University, Corvallis, Oregon, U.S.A. 97331

Eur. J. Biochem. 50 (1975)

Lipid peroxidation in isolated hepatocytes.

Eur. J. Biochem. 50,595-602 (1975) Lipid Peroxidation in Isolated Hepatocytes Johan HOGBERG, Sten ORRENIUS, and Robert E. LARSON Department of Forens...
756KB Sizes 0 Downloads 0 Views