Mechanisms of Cell Injury in the Killing of Cultured Hepatocytes by Bromobenzene*

Mechanisms producing lethal cell injury by bromo- benzene were explored in primary cultures of rat hepatocytes. Cells from phenobarbital-pretreated males were killed by bromobenzene from 0.1 to 0.7 m ~ . Cell death was preceded by a decline in glutathione and was prevented by SKF 525A. Cell killing was also reduced by promethazine, 1-a-tocopherol acetate, butyl- ated hydroxytoluene, N, W-diphenyl-p-phenylenedia-mine (DPPD), cysteine, N-acetylcysteine, a-mercapto- propionyl glycine, and 8-mercaptoethanol. After 18-20 h, cell death was proportional to the extent of the covalent binding of [*4C]bromobenzene to total cellular proteins in all cases. The time course, however, of the accumulation of covalently bound metabolites did not correlate with the time course of the cell death. Fur- thermore, the antioxidant DPPD prevented the cell death for 8-10 h without reducing the extent of covalent binding. Cell death during the first 6-10 h was accompanied by lipid peroxidation assessed by accumulation of malondialdehyde and by the appearance of conjugated dienes in cellular phospholipids. DPPD as well as the other protective agents reduced this lipid peroxi- dation in proportion to the to 8.35 of methanol, aliquots of this containing 1.25 to 5 nmol of malondialdehyde were added to 4.5 ml of complete Williams’ medium and treated as the regular cultures. In all cases, the excitation and emission spectra of the experimental samples were identical with that of the standards. Conjugated dienes were determined on phospholipids isolated from total lipid extracts. The incubation medium was aspirated, and 5 ml of methanol were added. The cells were scraped into this solution, and 2 volumes of chloroform were added to the methanol suspension of hepatocytes, and the lipids were extracted according to Folch et al (18). Phospholipids were precipitated from the total lipid extract according to Borgstrom (19). The precipitate was resuspended in 15 ml of chloroform-methanol (2:1, v/v). Three ml of water were added, and the phospholipids were recovered in the chloroform phase. Phospholipids were quantitated according to Harris and Popat (20). To determine the content of conjugated dienes, an aliquot of the chloroform phase was evaporated under oxygen-free N2, and the phos- pholipids were resuspended in cyclohexane to a concentration of 0.5 mg/ml. Spectrophotometric examination over the ultraviolet range of the phospholipids in cyclohexane was carried out in a Shimadzu Spectronic 210 UV spectrophotometer.

The mechanisms whereby chemicals lethally injure cells remain a central concern of biochemical toxicology. This problem has been pursued largely by following the intracellular fate of particular chemicals. Such an approach has been applied to a variety of toxins acting on a number of cell systems. The results have been generally quite consistent and have produced two major conclusions. First, most toxic chem-ic& are not biologically active but must be converted to reactive metabolites usually by the target cell itself. Target cell specificity and the actual extent of cell injury is, in turn, dependent upon this metabolism. Second, the critical metabolites exert their biological action by covalently binding to cellular macromolecules. Cellular injury mediated by the interaction of chemically reactive metabolites with macromolecules has been invoked to account for the liver cell death produced by a wide variety of xenobiotic chemicals and drugs (for a recent review see Ref. 1). The evidence to support such a role for covalent binding is largely circumstantial and based on the persistent correlation between the extent of binding and the severity of the accompanying liver cell necrosis. There is little direct evidence to substantiate the hypothesis that covalent binding to macromolecules can produce cell injury that results in necrosis. In particular, the molecular targets which interact with reactive metabolites and lead to cell death have not been identified. In addition, the functional consequences of such interactions between the chemical toxin and key cellular targets have rarely been considered.
We have begun to study the mechanism(s) mediating the bromobenzene killing of cultured hepatocytes in the attempt to define the cellular events linking the metabolism of chemical toxins with their biological consequences. Much of the evidence implicating the covalent binding of reactive metabolites to the subsequent cell death is based on bromobenzeneinduced liver necrosis in intact rats (2-7). Previous studies have shown that isolated hepatocytes metabolize bromobenzene and that this metabolism can result in lethal cellular injury (8-11).
In the present study monolayer cultures of hepatocytes were utilized rather than short term suspensions of these cells.
Using this system data are presented estabhhing that the killing of cultured hepatocytes reproduces the essential features of the toxicology of bromobenzene in the intact animal. A number of ways were then defined of protecting the hepatocytes from the cell killing by bromobenzene. The relationship between cell death and the covalent binding of bromobenzene was assessed with each protective agent. Conditions are described with which the killing of the hepatocytes could be dissociated from the covalent binding of metabolites of 6721 by guest on March 17, 2020 http://www.jbc.org/ Downloaded from 1'4CJbromobenzene to total cell proteins. An alternative mechanism underlying the lethal cell injury is proposed. Data documenting the correlation between lipid peroxidation and the extent of cell killing indicate that the former can account for the toxicity of bromobenzene quite apart from the interaction of metabolites with macromolecules. Finally, these data allow an initial assessment of the role that each alternative mechanism, covalent binding or lipid peroxidation, plays in the time course of the development of lethal cell injury with bromobenzene.
MATERIALS AND METHODS Sprague-Dawley rats (150-200 g) were obtained from Charles River Breeding Laboratories, Inc. Male rata pretreated with phenobarbital were given sodium phenobarbital by intraperitoneal injections (80 mg/kg body weight/day) for 3 days. All animals were fasted overnight prior to use. Isolated hepatocytes were prepared by collagenase perfusion according to Seglen (12). Yields of 2-4 X 10' cells/liver with 8540% viability (trypan blue) were routinely obtained. The hepatocytes were plated in either plastic 25-em2 flasks (Corning) at a density of 10,000 ceU/cm2 or in 75-cm2 flasks (Coming) at a density of 53,000 cells/cm2 in Williams' E medium (Flow Laboratories, Inc.) containing 10% heatinactivated (56 "C for 10 min) fetal calfserum (Grand Island Biological Co.), 10 IU/ml of penicillin, 10 pg/ml of streptomycin, and 0.02 unit/ ml of insulin. Williams' E plus serum, penicillin, streptomycin, and insulii is referred to below as complete Williams'. After incubation in an atmosphere of 5% C0,-95% air for 2 h to allow the attachment of viable cells, the cultures were rinsed once with prewanned Hanks' balanced salt solution (Flow Laboratories, Xnc.) to remove unattached dead cells. The cultures were then incubated in complete Williams' with the additions indicated in the text, Viability of the cultured ceUs was assayed either by trypan blue exclusion (13) or by the release of lactic dehydrogenase (14) as described previously.
Bromobenzene (Aldrich) was diluted in dimethyl sulfoxide and added to the cultures at the concentrations indicated in the text. SKF 525A was the generous gift of the Smith Kline and French Laboratories (Phdadelphia). It was dissolved in culture medium and added to the cells at a final concentration of 1 X M. 1-Cysteine (Sigma), I?-acetyl-1-cysteine (Calbiochem-Behring), and cu-mercaptopropionyl-I-glycine (Calbiochem-Behring) were dissolved in 0.9% NaCl and added to the cultures at a final concentration of 4,2.5, and 4 mM, respectively. /3-Mercaptoethanol (Sigma) was diluted in saline and added to the cultures at a final concentration of 2 mM. a-Tocopherol acetate (Sigma), butylated hydroxytoluene (Sigma), butylated hydroxyanisole (Sigma), and iV,N'-diphenyl-pph,henylenediamjne (Eastman) were dissolved in dimethyl sulfoxide and added to the cultures at a final concentration of 32,0.9, and 1 p~, respectively. Promethazine (Elkins-Sinn, Inc.) was obtained as a 25 m g / d solution, diluted with 0.9% saline, and added to the cultures at a final concentration of 10 GSH' was determined by the fluorometric method of Hissin and H i l f (15). The covalent binding of ["C]bromobenzene to protein was measured after incubating the cells for the times indicated in the text with a total of 1 X 10' cpm of br~mo-[U-'~CJbenzene (5.8 mCi/mmol, Amersham Corp.). Except for the concentration dependence study (Fig. 5), the final concentration of [14C]bromobenzene was always 0.5 n m . At the end of the incubation, the medium was aspirated and 5 ml of 10% trichloroacetic acid were added. The hepatocytes were scraped, the suspension was recovered by centrifugation, and the proteins were prepared according to h o and Recknagel (16). The dry protein residue was weighed in tared scintillation vials, dissolved in I ml of NCS tissue solubilizer (Amersham Corp.) and counted in 15 ml of a toluene-based solution.
Malondialdehyde was measured fluorimetrically by adaptation of the method of Yagi (17). At the times indicated, 0.5 rnl of the culture medium was removed for lactic dehydrogenase determination, and 5 0 % trichloroacetic acid was added to achieve a final concentration of  Conjugated dienes were determined on phospholipids isolated from total lipid extracts. The incubation medium was aspirated, and 5 ml of methanol were added. The cells were scraped into this solution, and 2 volumes of chloroform were added to the methanol suspension of hepatocytes, and the lipids were extracted according to Folch et al (18). Phospholipids were precipitated from the total lipid extract according to Borgstrom (19). The precipitate was resuspended in 15 ml of chloroform-methanol (2:1, v/v). Three ml of water were added, and the phospholipids were recovered in the chloroform phase. Phospholipids were quantitated according to Harris and Popat (20). To determine the content of conjugated dienes, an aliquot of the chloroform phase was evaporated under oxygen-free N2, and the phospholipids were resuspended in cyclohexane to a concentration of 0.5 mg/ml. Spectrophotometric examination over the ultraviolet range of the phospholipids in cyclohexane was carried out in a Shimadzu Spectronic 210 UV spectrophotometer.

Bromobenzene-inducedLiver Cell
Death-Freshly isolated hepatocytes were plated in plastic flasks and allowed to attach for 2 h prior to treatment with bromobenzene. Hepatocytes from male or female rats were insensitive to 0.5 rn bromobenzene over the course of a 20-h exposure. When male rats were pretreated for 3 days with phenobarbital (80 mg/kg body weight/day), their cultured hepatocytes were now sensitive to bromobenzene. following hydrolysis of fluorescein diacetate (13). Fig. 1 documents that release of lactic dehydrogenase similarly correlates with the ability to exclude trypan blue. The two assays gave virtually identical results. In contrast, the loss of intracellular K' or the content of ATP can be dissociated from the ability to exclude trypan blue or the release of lactic dehydrogenase? The 10s of K' or the ATP content, therefore, do not necessarily reflect viability of cultured hepatocytes and were not used in the present study.
As shown in Fig. 1, the viability of control cells (closed circles) was greater than 90% of the initial value after 20 h. Bromobenzene reduced the viability of the cells in a time-and dose-dependent manner. While there was some variability among cell preparations, most of the cell killing generally occurred by 6-8 h with only a slight increase in the number of dead cells between 8 and 20 h. Fig. 2 summarizes the relationship between the extent of cell killing and the concentration of bromobenzene.
The following data suggest that sensitivity to bromobenzene is a consequence of its enhanced metabolism. Hepatocytes isolated from phenobarbital-treated males have an increased content of cytochrome P-450. Within 30 min, bromobenzene reduced the GSH content in sensitive hepatocytes to only 25% of the initial level (Fig. 3). There were no differences in the initial content of GSH in hepatocytes from untreated male rats, and bromobenzene had no effect on their GSH concentration. The cytochrome P-450-dependent mixed function oxidase of the liver endoplasmic reticulum is inhibited by the phenothiazine SKF 525A (21). Liver cell death in the intact phenobarbital-pretreated rat is also prevented by SKF 525A (22). Similarly, addition of 10 PM SKF 525A to the culture medium prevented the bromobenzene-induced killing of the cultured hepatocytes (Fig. 4).
The above data are consistent with previous investigations employing suspensions of isolated hepatocytes (8-11). Cultured hepatocytes reproduce the major features of the toxic effect of bromobenzene in the intact animal. Sensitivity to bromobenzene in culture is dependent upon prior induction of cytochrome P-450 by phenobarbital administration to the intact animal. Bromobenzene depletes induced hepatocytes of GSH prior to the appearance of cell death, and the cell killing can be prevented by an inhibitor of mixed function oxidase activity, SKF 525A.
The experimental strategy employed in this study was to alter the reaction of cultured hepatocytes to bromobenzene in a number of ways and to then assess the effect of such manipulations on the cell killing and the covalent binding of metabolites of ['4C]bromobenzene. In addition to SKF 525A, two broad classes of chemicals, sdthydlyl compounds and antioxidants, were found to be effective in reducing the extent of the killing of cultured hepatocyks by bromobenzene. The specific compounds and their actions are detailed below. We note, however, that including SKF 525A a number of different chemicals were thus available with which to perturb the reaction of cultured hepatocytes to bromobenzene. The effect of these agents on the covalent binding of [14C]bromobemene to cellular macromolecules was examined next.
Cell to total cellular proteins was determined. In this manner, cell killing and the extent of covalent binding were quantitated in the same cultures. The dose dependence of the cell killing was essentially as illustrated in Fig. 2. The dose dependence of the covalent binding to protein illustrated in Fig. 5 closely paralleled that of the cell killing. The amount of covalent binding of ['?]bromobenzene rose sharply with increasing concentration to about 0.3 mu. The rate of increase then declined with a plateau reached in the extent of binding between 0.5 and 0.7 ~IU bromobenzene. Fig. 6 illustrates a representative time course of the covalent binding to protein with 0.5 mu ['?Yjbromobenzene.
The numbers in parentheses are the percentage of dead cells at the respective times. The greater proportion of the total covalent binding occurred within 2 h of exposing the cells to bromobenzene. At this time, however, only about 10% of the cells were dead. This represented only about 20% of the total number of cells that died by the end of the experiment. Between the 2nd and the 18th hours, the covalent binding increased only very slightly. It was during this time, however, that 80% of the cell death occurred.
The insensitivity of uninduced hepatocytes from male or female rats was reflected in very little covalent binding (Table  I). Each suIfhydry1 compound reduced the extent of covalent binding in proportion to the degree of protection against the killing (Table I). Reduction in covalent binding with each of the compounds listed in Table I was observed as early as 30 min after exposure to .['%]bromobenzene.
The percentage decrease in covalent binding was then constant between 30 min and 18 h. In Fig. 7 are plotted the data from Fig. 5 and Table I   h later), both measurements were reduced proportionately by vitamin E and promethazine.
There was no dissociation between covalent binding and cell death with these two antioxidants as with the sulfhydryl compounds. The antioxidanta butylated hydroxytoluene, butylated hydroxyanisole, and DPPD differed. With these agents there was dissociation to a variable extent between the covalent binding of ['4C]bromobenzene metabolites and the death of the hepatocytes. DPPD was the most striking in its ability to effect such a dissociation, and we concentrate in what follows on it. The data in Table II   effect on the extent of covalent binding with the low dose of bromobenzene and actually increased the amount of such binding with the larger dose of bromobenzene. After 18 h, DPPD no longer prevented the cell killing with either dose of bromobenzene. These data suggest that the cell killing by bromobenzene for up to 10 h is not necessarily a consequence of the extent of covalent binding of reactive metabolites as assessed by the binding of ['4C]bromobenzene to total cellular proteins.
Lipid Peroxidation in Bromobenzene-intoxicated Hepatocytes-Since DPPD is a potent antioxidant, its ability to prevent cell death between 0 and 10 h could result from an inhibition of lipid peroxidation. Such an action would then explain the lack of a correlation during these times between covalent binding and cell death in the DPPD-protected cells. The presence of lipid peroxidation in bromobenzene-intoxicated hepatocytes was demonstrated by the accumulation of malondialdehyde in the cultures and by the appearance of conjugated dienes in isolated cellular phospholipids, Fig. 8 compares the time course of the accumulation of malondialdehyde with that of the death of the cells during the first 6 h after treatment of the hepatocytes with 0.5 mM bromobenzene. The data document a close tempord association between accumulation and cell death. This association was further documented by examination of the effect of the protective agents on the formation of malondialdehyde. Table  111 compares the cell death and the formation of malondialdehyde in cells treated with bromobenzene and four different protective agents including DPPD. Prevention of cell death was accompanied by prevention of malondialdehyde formation in each case. The presence of any of the sulfhydryl protective compounds in the tissue culture medium interfered with the malondialdehyde assay. The effect of these compounds on lipid peroxidation was examined by their influence on the formation of conjugated dienes in cellular phospholipids.
When lipids containing fatty acid dienes or polyenes are peroxidized, there is a shift in double bond position leading to conjugation. This involves initial abstraction of H. from a doubly allylic position followed by a double bond migration. Conjugated dienes result which show an intense absorption at 233 nm. The presence of conjugated dienes has been widely used for detection and estimation of lipid peroxidation in liver cell injury by hepatotoxic agents. Fig. 9 illustrates the absorp- Conjugated dienes could be detected in the phospholipids extracted from the cultured hepatocytes as early as 30 min following treatment with bromobenzene. This is prior to the onset of cell death and before significant formation of malondialdehyde could be detected as illustrated in Fig. 8. The formation of conjugated dienes was then used to assess the effect of sulfhydryl agents on the course of lipid peroxidation in the bromobenzene-intoxicated cells. The difference between the of phospholipids from control and treated cells was used to quantitate the extent of formation of conjugated dienes. The data in Table IV  This close correlation between the cell killing and the peroxidation of cellular phospholipids that was evident with the protective agents held over the entire course of an 18-to 20-h exposure to bromobenzene. The exception was again DPPD. The data in Table I1 indicate that DPPD was effective in preventing the cell death produced by bromobenzene up to 10 h after exposure to the toxin. By 18 h DPPD had no effect  In each case the cultures were treated with bromobenzene alone or with bromobenzene plus the additions indicated for 6 h. Control cells received Me2SO.
'Results are the difference between the AZ33 of phospholipids isolated from the pooled cells in 2-3 separate 75-cm2 flasks of control cultures and cultures treated as indicated.
on the extent of the cell killing. DPPD did, however, prevent the formation of malondialdehyde and the appearance of conjugated dienes throughout this entire time. At times less than 10 h, therefore, DPPD could dissociate the cell killing from the extent of the covalent binding of ['4C]bromobenzene. After 10 h, it could similarly dissociate the cell killing from the manifestations of lipid peroxidation.
This action of DPPD suggested the final series of experiments. It was possible that the peroxidation of cellular phospholipids in the bromobenzene-intoxicated cells was an effect of their death rather than presumably a cause. Such an interpretation was also suggested by the presence of ferric iron and ascorbic acid in the tissue culture medium in all of the above studies. This possibility was considered by killing the hepatocytes by nphanisms that very likely did not involve Lipid peroxidatlon and then looking for evidence of peroxidation in the lipids of the dead cells. The results of such a study are illustrated in Fig. 10. The cultured hepatocytes were treated with the toxic calcium ionophore A23187. Within 15 min, over 40% of the cells were dead. By 30 min, almost 70% of the cells had died. Malondialdehyde could not be significantly detected in the cultures until at least 2 h after exposure to A23187. The amount of malondialdehyde was significantly less when virtually all the cells were killed by A23187 than the amount measured with bromobenzene when only 50% of the cells were dead (Table 11). A difference in the relationship between the lipid peroxidation and the cell death with A23187 and bromobenzene was also evident in the effect of the protective agents. None of the agents effective against bromobenzene-induced liver cell death had any activity against the toxicity of A23187. The antioxidants promethazine, vitamin E, and DPPD did, however, prevent the lipid peroxidation that followed the death of the cells induced by A23187. The sulfhdryl agents did not reduce the lipid peroxidation following upon the A23187-induced cell death.
Similar results were obtained with a second agent that killed the cells by a mechanism that is also probably unrelated to lipid peroxidation. Melittin is the main cytolytic component of bee venom (22). Fig. 10 indicates that 0.05 p g / d of melittin killed 50% of the hepatocytes within 1 h. At this time there was a little or no detectable malondialdehyde. Over the course, however, of the next 5 h, there was a significant production of malondialdehyde without any further increase in the number of dead cells. As with A23187, antioxidants added to the culture medium effectively prevented the formation of malmdialdehyde without reducing the number of dead cells. There is clearly, therefore, lipid peroxidation in cultures containing dead hepatocytes. This lipid peroxidation, however, can be readily dissociated from the death of the cells produced by A23187 and melittin. Such lipid peroxidation is not necessarily related to the presence of ferric iron or ascorbic acid in the culture medium. The same extent of malondialdehyde formation was observed in medium containing neither iron nor ascorbic acid or ascorbic acid alone.

DISCUSSION
The objective of the studies presented above was to define the mechanisms whereby the metabolism of the model hepatotoxin bromobenzene leads to lethal liver cell injury. Previous studies of bromobenzene hepatotoxicity suggested that the cell death is related to the covalent binding of reactive metabolites to cellular macromolecules (2)(3)(4)(5)(6)(7). The data in this report throw new light on the role of such covalent binding and suggest that the induction of lipid peroxidation may also participate in the pathogenesis of lethal cell injury. Use of a system as employed here is dependent upon its reproducing the essential features of the action of bromobenzene in the intact animal. Without such a relationship, analysis of the cell culture model would not be relevant to previous studies. The initial results reported here establish that the killing of cultured hepatocytes by bromobenzene exhibits the same conditions which determine its toxicity for liver cells in situ. With cultured hepatocytes, the toxicity of bromobenzene is dependent upon its metabolism. Pretreatment of rats with phenobarbital enhances the metabolism of bromobenzene in vivo and determines liver cell injury (2). In contrast, SKF 525A inhibits this metabolism and protects from liver injury (5). Pretreatment of rats with phenobarbital rendered the hepatocytes cultured from their liver sensitive to bromobenzene. The cell killing in culture was prevented by addition of SKF 525A to the medium.
The reaction of the cultured hepatocytes to bromobenzene was then manipulated by addition to the medium of a number of sulfhydryl compounds and antioxidants. The resulting data are best interpreted as consistent with, at least, 2 separate mechanisms of lethal cell injury. During the first 6 to 10 h, the predominant mechanism producing lethal cell injury seems related to the peroxidation of phospholipids. Lipid peroxidation was detected in the formation of conjugated dienes in total phospholipids as early as '/i h after treatment with bromobenzene. Malondialdehyde accumulation was first measurable at 1 h. The amount of malondkddehyde in the cells and medium increased in parallel with the increasing number of dead cells. During this time, each of the agents that modified the reaction of the liver cells to bromobenzene reduced the manifestations of the peroxidation of lipids in proportion to the reduction in the number of dead cells. These results, however, do not unequivocally establish a causal relationship between lipid peroxidation and the development of lethal cellinjury. The major alternative explanation of the same data is that lipid peroxidation is the effect rather than the cause of the cell death. The kiUing of the cultured hepatocytes by A23187 and melittin was used to evaluate this alternative interpretation. With both these toxins lipid peroxidation developed but was delayed following the death of the hepatocytes. In addition, the cell death could be clearly dissociated from resulting lipid peroxidation. Antioxidants prevented lipid peroxidation without having any effect on the cell killing by either A23187 or melittin. Lipid peroxidation occurring with bromobenzene, therefore, differs in two ways from that which can be established as an effect of cell death. Bromobenzene-induced lipid peroxidation is detected simultaneously with or even possibly before the death of the cells, and it cannot be dissociated from the cell killing during the first 10 h of bromobenzene intoxication.
With the antioxidants promethazine and vitamin E and all the sulfhydryl compounds, a reduction in lipid peroxidation is very likely the result of a reduction in bromobenzene metabolism. This is suggested by the parallel decrease in both lipid peroxidation and the covalent binding of ['4C]bromobenzene. Reduction in covalent binding by the antioxidants is difficult to explain if their action is related to events subsequent to the formation of bromobenzene epoxides. More likely they are reacting with activated oxygen species that are responsible for the oxidation of bromobenzene and, most likely, also for the induction of lipid peroxidation. The sulfhydryl compounds could be acting independently (as nucleophilic traps or by maintaining GSH levels) to prevent both covalent binding and lipid peroxidation. The close association, however, between these two effects suggests that the sulfhydryls are also interacting with the mixed function oxidase in a manner that again inhibits bromobenzene metabolism.
With DPPD, metabolism was dissociated from lipid peroxidation and cell death. In this case, covalent binding was unaffected or actually increased at a time when lipid peroxidation and cell death were reduced. Two consequences of the metabolism of bromobenzene are suspected as responsible for the appearance of lipid peroxidation. The first is an acute oxidative stress that results from the generation of activated oxygen species (0; or Hz02) accompanying the metabolism of many mixed function oxidase substrates. A branch point exists in the P-450 catalytic cycle at the introduction of the second electron (23). Active turnover by the P450 system involves production of hydrogen peroxide. As much as 55% of consumed oxygen appears as Hz02 in the presence of substrate trophilic products of bromobenzene metabolism. The depletion of GSH overwhelms the cellular defenses against oxidative stresses. The result is peroxidation of cellular phospholipids.
While lipid peroxidation may play a role in the pathogenesis of lethal cell injury during the fist 6-10 h after exposure to bromobenzene, there must be a second mechanism that is responsible for the death of the cells at later times. Again with DPPD the peroxidation of cellular lipids could be dissociated from the death of the cells at times between 8 and 20 h. The most obvious candidate for the responsible second mechanism is the covalent binding of reactive metabolites of bromobenzene to cellular macromolecules. The data presented above, however, place some restraints on the extent to which covalent binding can account for the cell death.
The present report confums the conclusion of previous studies in intact animals (2-7) that the extent of liver necrosis is closely correlated with the extent of the covalent binding of ['4C]bromobenzene to total cellular proteins. With cultured hepatocytes there was a parallelism between the extent of cell death and the extent of the covalent binding of ['*C]bromobenzene when both were assessed at times later than 10 h after treatment with the toxin. The correlation between the extent of covalent binding and cell killing does not, however, necessarily establish a cause and effect relationship. The data in this report indicate that the mere extent of the covalent binding to protein cannot be a sole determinant of the loss of viability. This conclusion is implied by the lack of a correlation between the time course of the covalent binding and that of the death of the cells and by the effect of DPPD. Covalent binding preceded signiikantly the death of the cells. DPPD protected the cells during the time that the covalent binding to total protein developed and reached its maximum extent.
The extent of covalent binding cannot, therefore, be a direct measure of the extent of a mechanism producing lethal cell injury with bromobenzene. If covalent binding is related to the pathogenesis of irreversible cell injury, there must be a portion of the measured binding that develops with a different time course than that shown by the total. Alternatively, there would have to be some additional and as yet unknown mechanism that is initiated by and a function of the covalent binding. This mechanism would, in turn, be responsible for lethal cell injury. Our data provide no clues as to the possible nature of such an additional mechanism nor do they necessarily imply its existence. The data simply imply that such a mechanism would have to exist if covalent binding has a uniform time course of development and is causally related to the lethal cell injury.
In conclusion, the data in the present report define conditions with which the peroxidation of cellular lipids and the 6728 Mechanisms of Cell Injury covalent binding of reactive metabolites may and may not be relevant to the genesis of lethal cell injury in bromobenzeneintoxicated hepatocytes. Continued analysis of this system in pursuit of the implications of the present study may help to define more precisely the role that each mechanism plays in the death of the cultured hepatocytes and, in turn, in the necrosis of liver cells in the intact animal.