Comparative cytopathology of primary rat hepatocyte cultures exposed to aflatoxin B1, acetaminophen, and other hepatotoxins

TOXICOLOGY

AND

APPLIED

PHARMACOLOGY

80,

345-356 (1985)

Comparative Cytopathology of Primary Rat Hepatocyte Cultures Exposed to Aflatoxin B, , Acetaminophen, and Other Hepatotoxins’ M. A. HAYES* AND D. B. PICKERING Department

of Pathology, University of Guelph, Guelph. Ontario Medical Sciences Building, University of Toronto,

Received

February

4, 1985;

NlG 2 WI, Canada and Department Toronto, Ontario M5S IA8, Canada

accepted April

of Pathology,

30. 1985

Comparative Cytopathology of Primary Rat Hepatocyte Cultures Exposed to Aflatoxin B,. Acetaminophen, and Other Hepatotoxins. HAYES, M. A., AND PICKERING, D. B. (1985). Toxicol. Appl. Pharmacol. 80, 345-356. The cytomorphological alterations in primary monolayer cultures of Fischer 344 rat hepatocytes exposed continuously to lethal concentrations of various hepatotoxins were compared. Aflatoxin B, (AFB,) at concentrations of 1 and IO pM killed 67 and 97% hepatocytes, respectively, at 48 hr, as determined by release of lactate dehydrogenase (LDH) into culture medium, or by trypan blue uptake. AFB, produced numerous radial fingerlike projections (blebs) at attached margins of cytoplasm in almost all hepatocytes between 6 and 24 hr. Formation of blebs was dose dependent and preceded release of LDH and trypan blue uptake. Similar marginal blebs were produced by cycloheximide. 2-acetylaminofluorene, N-hydroxy-2-acetylaminofluorene, and senecionine. By comparison, acetaminophen, at equivalently lethal concentrations (4 or 16 mM) did not cause marginal blebs, but resulted in a similar time course of LDH release and trypan blue uptake. Carbon tetrachloride and bromobenzene also failed to produce blebs at lethal concentrations. Phalloidin and cytochalasin B rapidly produced smaller spherical blebs over the entire surface of all hepatocytes, distinct from the finger-like blebs produced only at the free attached margins of cells exposed to AFB, . Blebs and lethal injury by AFB, were reduced by phenobarbitone or 3-methylcholanthrene pretreatments and by SKF-525-A in culture. The demonstration of morphological prelethal changes in hepatocytes injured by different classes of hepatotoxins in culture provides a means of differentiatidg the early biochemical mechanisms by which hepatotoxins and hepatocarcinogens lethally injure hepatocytes before membrane permeability to LDH and trypan blue is detectable. 0 1985 Academic

Press. Inc.

Primary monolayer cultures of isolated hepatocytes are widely used in quantitative assays for lethal hepatotoxicity of xenobiotics (Acosta et al., 1980; Green et al., 1982; Roberts et al., 1983; Metcalfe and Neal, 1983a,b; Hayes et al., 1984a,b) and to investigate the mechanisms by which hepatocellular necrosis occurs (Schanne et al., 1979; Casini and Farber, 198 1; Casini et al., 1982; ’ This study was supported by Grant A2761 from the Natural Sciences and Engineering Research Council of Canada. ’ To whom reprint requests should be addressed.

George et al., 1982; Russo et al., 1982). Lethal responses to many agents such as carbon tetrachloride (Casini et al., 1982) bromobenzene (Hayes et al., 1984a), acetaminophen (Acosta et al., 1980; Hayes et al., 1984a), aflatoxins (Green et al., 1982; Loury et al., 1984; Roberts et al., 1983; Metcalfe and Neal, 1983a,b), galactosamine (Farber et al., 1982) and pyrrolizidine alkaloids (Hayes et al., 1984b) generally reflect the hepatotoxicity of these agents in vivo. However, some agents, notably cycloheximide kill monolayer hepatocytes (Car-r and Laishes, 198 1) at concentrations that are not particularly necro345

0041-008X/85

$3.00

Copyright 0 1985 by Academic Press. Inc. All rights of reproduction in any form reserved.

346

HAYES AND PICKERING

genie in vivo (Verbin et al., 1969; Castro et al., 1977; Popp et al., 1978). Monolayer hepatocytes are frequently used in studies of hepatocellular metabolism of various xenobiotics (Green et al., 1982; Metcalfe and Neal, 1984b; Loury et al., 1984) and in investigations of non-lethal hepatotoxic effects, such as unscheduled DNA synthesis (Williams, 1980; Green et al., 1982) and inhibition of hepatocellular DNA synthesis (Metcalfe and Neal, 1983a; Novicki et al., 1985). Measurement of lethal injury of monolayer hepatocytes is generally achieved by determining the degree of permeability of the plasma membrane after chemical injury, for example, leakage of cytosolic enzymes including lactate dehydrogenase (LDH) and glutamate-pyruvate transaminase (GPT) (Acosta et al., 1980; Jauregui et al., 198 1; Casini et al., 1982; Hayes et al., 1984a; Loury et al., 1984) and trypan blue staining (Casini and Farber, 1981; Jauregui et al., 1981: Roberts et al., 1983; Metcalfe and Neal, 1983a). Similar methods to quantify lethal hepatocfle injury have been used in short-term suspension cultures of isolated hepatocytes exposed to various hepatotoxins (Smith et al., 1981; Jewel1 et al., 1982; Stacey and Klaassen, 1982; Story et al., 1983). Although indices of plasma membrane permeability are easily quantifiable endpoints that reflect lethal hepatocyte injury, they do not discriminate among different early mechanisms leading ultimately to membrane failure. There is presently no clear understanding of the necrogenic roles of covalent binding of metabolites of hepatotoxins to various macromolecules (Hinson et al., 198 I; Farber and Gerson, 1984), peroxidative membrane injury (Recknagel, 1983), calcium ion influx (Trump et al., 1980; Farber, 1982; Stacey and Klaassen, 1982; Jewel1 et al., 1982), and surface membrane blebs (Hoer1 and Scott, 1978; Russo et al., 1982; George et al., 1982; Jewel1 et al., 1982) during necrosis of isolated or monolayer hepatocytes. In previous studies on hepatotoxicity of

various hepatotoxins and hepatocarcinogens (Roberts et al., 1983; Hayes et al., 1984a-c), we recognized various early morphological changes in lethally injured hepatocytes isolated from Fischer 344 rats. Accordingly, we considered that different morphological changes preceding lethal injury may reflect separate biological and biochemical mechanisms for different classes of hepatotoxins. This report describes further studies on the relationship between different cytomorphological changes induced by necrogenic concentrations of acetaminophen and aflatoxin B1 (AFB1) and the subsequent release of LDH or uptake of trypan blue dye. It also compares responses to various other carcinogenic and noncarcinogenic hepatotoxins. These studies were performed in a monolayer culture assay in which the modulations in toxicity of acetaminophen or bromobenzene by different selective inducers or inhibitors of cytochrome P-450 were qualitatively and quantitatively similar to modulations of necrosis caused by bromobenzene and acetaminophen in vivo (Hayes et al., 1984a). METHODS Animals and chemicals. Rats were young (160 to 220 g) adult Fischer 344 males from Charles River Laboratories, Wilmington, Massachusetts. They were fed pelleted rat chow (5002 Ralston-Purina Co.. St. Louis, MO.) and maintained in suspension cages according to the guidelines of the Canadian Council on Animal Care. Aflatoxin B1 was purchased from Makor Chemicals. Jerusalem, Israel; acetaminophen from Sigma Chemical Company (St. Louis, MO.); phenobarbitone sodium (PB) from BDH Chemicals, Toronto, Canada; carbon tetrachloride and DMSO (dimethyl sulfoxide) from Fisher Chemicals, Toronto, Canada: 2-acetylaminofluorene from Aldrich Chemical Company (Milwaukee. Wise.). N-hydroxy-2acetylaminofluorene from Chemical Repository, IIT (Chicago, III.); cycloheximide, phalloidin, cytochalasin B, methotrexate, a-naphthoflavone (ANF), and 3-methylcholanthrene (3-MC) from Sigma. Senecionine was a gift of Dr. C. C. J. Culvenor, CSIRO, Parkville, Victoria, Australia. SKF-525-A was a gift of Smith Kline and French, Toronto, Canada. Hepatocyte isolation and culture. Hepatocytes were obtained by collagenase perfusion from rats anesthetized with sodium pentobarbital, as described in detail previously (Hayes et al.. 1984a). Most of the experiments

MODES OF NECROSIS IN MONOLAYER were performed on normal hepatocytes isolated from untreated control rats without prior feed withdrawal. However, the responses of hepatocytes from some rats given inducing xenobiotics in vivo were also studied. Rats induced with phenobarbitone were given three daily doses of 80 mg/kg in 0.9% saline ip and perfused on Day 4. Rats induced with 3-methylcholanthrene were given three daily doses of 40 mg/kg in corn oil by gavage. Dissociated normal or induced hepatocytes were purified by slow speed centrifugation and plated in 25cm2 flasks (Falcon) in William’s E Medium supplemented with insulin (20 U/liter), Hepes buffer (0.01 M), Lglutamine (266 mg/liter). penicillin (100 U/liter), streptomycin (100 mg/liter), and 10% fetal bovine serum (Flow Inc., Mississauga, Ontario, Canada). Plating densities were 1 X lo6 or 0.5 X lo6 viable ceils (trypan blue excluders) in 5 ml of medium per 25-cm2 flask. Cells were allowed to attach for 3 hr at 37°C in air/CO2 (95/ 5), after which they were washed in buffered Hanks’ solution (including 0.01 M Hepes) and resupplied with fresh medium containing the various hepatotoxins. Treatment with hepatotoxins. All test agents except senecionine were dissolved in DMSO at concentrations such that all flasks received a similar quantity of DMSO (20 ~1 per flask). At least two flasks per experiment were exposed to each concentration of hepatotoxin used. Control flasks (six per experiment) were given DMSO alone. Senecionine was administered to flasks in an aqueous vehicle as previously described (Hayes et al., 1984b). All treatments with hepatotoxins were applied 1 hr after hepatocytes were washed (4 hr after isolation). Cotreatments with the selective cytochrome P-450 inhibitors SKF-525-A or a-naphthoflavone were added (final concentration 10 ELM) 30 min before the hepatotoxins or DMSO. These concentrations of SKF-525-A and ANF. which were not cytotoxic in DMSO-treated control cultures, were evaluated as potential modulators of AFB, cytotoxicity because, in previous studies (Hayes et al., 1984a). SKF-525-A at IO pM strongly inhibited PBinduced cytotoxicity of acetaminophen and bromobenzene. whereas ANF at 10 WM strongly inhibited 3-MCinduced cytotoxicity of acetaminophen. Evaluation of cytotoxicity. The percentage of LDH released from cultured hepatocytes at various times in culture was determined as previously described (Hayes et al., 1984a). Trypan blue uptake was assessed by nuclear staining of washed monolayer hepatocytes after 5 min incubation with trypan blue (final concentration 0.06% in medium). The percentage of trypan bluenegative attached cells was determined as described previously (Roberts ef al., 1983) by counting six randomly chosen 0.28 mm* microscopic fields in each of three replicate flasks. Morphological appearance of monolayers was assessed by phase contrast light microscopy. Six premarked randomly selected fields were serially photographed after various periods of incubation to evaluate the sequential progression of the morphological alterations

347

HEPATOCYTES

of control or toxin-treated hepatocytes. Data are presented as X f SE for values compiled from several separate rat perfusion experiments (N). Statistical comparisons were made by Student’s t tests where indicated in the text.

RESULTS Monolayer hepatocytes in this culture system were lethally injured, as determined by LDH release, in a dose-dependent fashion by AFBr (Fig. 1) and acetaminophen (Hayes et al., 1984a). Hepatocytes induced in vivo by PB or 3-MC were less susceptible (p < 0.05) to lethal injury by AFBr but only at low concentrations in culture between 0.1 and 3.2 PM (Fig. 1). SKF-525-A (10 PM) consistently inhibited killing by AFB, but only at concentrations of 10 FM or above both in control hepatocytes (Fig. 1) and in hepatocytes induced in vivo with PB or 3-MC (data

+

011

0$2

Concentration

lb

312

of Aflatoxm

Ib

312

Id0

Bj @M)

FIG. 1. Dose-response curves for LDH release from monolayer hepatocytes (4 X 104/cm2) exposed to AFB, for 24 hr. Points are X f SE from numbers of separate experiments indicated in parentheses. Phenobarbitone (PB) or 3-methylcholanthrene (3-MC) pretreatments in vivo are protective (p < 0.025) in the low concentration ranges (0. I to 3.2 PM),whereas SKF-525-A in culture is protective (p < 0.05) against killing by higher concentrations (10 or 32 FM).

348

HAYES AND PICKERING

not included). ANF at the 10 PM concentration used had minimal effects on cytocidal response curves of control or induced hepatocytes exposed to AFBl (data not included). The dose-dependent toxicity of hepatocytes exposed continuously to equivalently toxic concentrations of AFB, or acetaminophen had qualitatively similar time-response curves during 48 hr of culture (Figs. 2 and 3). However, the concentrations required for equivalent lethal response curves differed by about 1500-fold, since 10 PM AFB, was equivalent to 16 mM of acetaminophen (Figs. 3 and 4). Lethal injury measured by LDH release for each agent was inversely proportional both quantitatively and chronologically to the numbers of trypan blue-negative attached hepatocytes (Figs. 2 and 3). Attached hepatocytes in control cultures spread during the first 18 hr of culture and then migrated together to form attached clusters of cells in monolayers between 18 and 24 hr (Fig. 4). Subsequently. elongated cytoplasmic projections formed at the free margins of individual or clustered hepatocytes. Many of these projections appeared to form cell-cell contacts at 48 hr. During this 48-hr culture period, the degree of cell mortality in

Hours

in Culture

FIG. 2. Time sequence of LDH release (solid lines) and trypan blue exclusion (broken lines) of monolayer hepatocytes (2 X 104/cmz) exposed to aflatoxin B, (1 or IO PM). Values are f & SE from four separate experiments.

‘\

4mM \, ‘,

.

16mM

60 50

50

2

40

40

30

30

m 5 F

20 10 0 Hours

in Culture

I

FIG. 3. Time sequence of LDH release (solid lines) and trypan blue exclusion (broken lines) of monolayer hepatocyte (2 X 104/cmZ) exposed to acetaminophen (4 or 16 PM). Values are X = SE from four separate experiments.

control cultures was low, as determined by LDH release or trypan blue staining (Figs. l-3). Sequential comparative morphological examinations of uninduced hepatocytes exposed to high or low lethal concentrations of AFB, or acetaminophen revealed a marked difference in their cytomorphological responses preceding LDH release or trypan blue staining. Hepatocytes .exposed to AFBl displayed a characteristic ruffling of the spreading margins of cytoplasm within 6 hr. The ruffled cytoplasm gradually segregated into characteristic finger-like blebs which mostly remained attached to the plastic dish at 18 hr (Fig. 5). These blebs projected radially from the free margins of cytoplasm, but were not seen at cell margins attached to adjacent hepatocytes. Many blebs subsequently detached from the hepatocytes and were observed as minute spherules less than 2 PM in diameter. These spherules were either attached to the plastic surrounding injured hepatocytes, or became detached and floated free in the culture medium. Many of the hepatocytes with blebs became detached from the plastic between 24 and 36 hr and were frequently seen as trypan blue-negative re-

MODES OF NECROSIS IN MONOLAYER

HEPATOCYTES

FIG. 4. Phase-control micrograph of normal control monolayer rat hepatocytes 18 hr after addition of DMSO (4 &ml).

FIG. 5. Phase-contrast micrograph of monolayer rat hepatocytes exposed to aflatoxin B, (10 pM) for 18 hr. Note the numerous finger-like blebs projecting from the peripheral free margins of attached cytoplasm (arrows).

349

350

HAYES AND PICKERING

lethal concentrations of either acetaminophen f&tile clusters of small spherules. Eventually, or AFB,. by 48 hr in cultures with virtually 100% The proportion of hepatocytes with the release of LDH (i.e., 10 PM AFB,), trypan characteristic peripheral blebbing induced by blue uptake was observed in nearly 100% of AFB, was dose dependent and peaked within hepatocytes, most of which formed clusters 6 to 12 hr (Fig. 7), well in advance of LDH of small stained blebs, without a clearly visible release or trypan blue staining (Refer to Fig. nucleus. By comparison, hepatocytes exposed to 2). Various other hepatotoxins were surveyed acetaminophen did not have typical large for their ability to induce the marginal blebs peripheral cytoplasmic blebs (Fig. 6) although seen in AFBr-exposed cultures. Cyclohexoccasional minute blebs were observed. imide, 2-acetylaminofluorene, N-hydroxy-2Rather, lethal injury was preceded by a nor- acetylaminofluorene, and senecionine at lemal degree of initial spreading by 12 to 18 thal concentrations all caused marginal blebs hr, followed by an increasing proportion of similar to those seen for AFBr (Table 1). trypan blue staining of attached spread cells However, this peripheral blebbing response in which the nucleus was clearly visible. to AFBr and these other agents was morphoBetween 24 and 48 hr hepatocytes killed by logically different from the blebbing response acetaminophen were partially rounded with elicited by phalloidin or cytochalasin B. These a refractile appearance under phase-contrast latter agents resulted in the rapid formation microscopy (Fig. 6). The normal cytoplasmic within 6 hr on all hepatocytes (Table 1) of projections associated with cell spreading ob- numerous minute spherical blebs that proserved between 24 and 48 hr in control jected mainly from the unattached top surface cultures did not occur in cultures exposed to of hepatocytes (Fig. 8) as distinct from the

FIG. 6. Phase-contrast micrograph of monolayer rat hepatocytes exposed to acetaminophen (16 for 18 hr. Cells are mostly round rather than spread and do not exhibit marginal blebs.

pM)

MODES

0

10

OF NECROSIS IN MONOLAYER

20 30 Hours in Culture

40

exposed to warm paraformaldehyde (3%) shed numerous surface vesicles within 30 min and were distinct from the protrusions caused by phalloidin, cytochalasin B, and AFB, . Lethal injury of hepatocytes by carbon tetrachloride in these studies or by bromobenzene in previous experiments (Hayes et al., 1984a) resembled that caused by acetaminophen in that cells rounded up and stained with trypan blue without previously forming marginal blebs. The peripheral blebbing caused by AFB, was clearly a prelude to cell detachment in serial photographs of the same fields of hepatocytes in culture flasks. For this reason, additional experiments were conducted using collagen-coated flasks and rat serum to improve cell attachment. However, AFBr resulted in identical marginal blebbing in these cultures, and in cultures that were serum free or supplemented with 5% rat serum. Control cultures exposed for 48 hr to CaC12 (up to 5 mM Ca*+) or EDTA (up to 5 mM) did not exhibit marginal blebs at any stage, nor did

I 50

FIG. 7. Time sequence of marginal bleb formation in monolayer rat hepatocytes (2 X 104/cm2) exposed to aflatoxin B, (AFB) or acetaminophen (AA). Values represent the mean percentage of affected hepatocytes (*SE) for three separate experiments. Acetaminophen at 4 or 16 mM consistently failed to produce marginal blebs.

much larger finger-like blebs attached to the peripheral cytoplasm of hepatocytes exposed to AFB, (see Fig. 5). Monolayer hepatocytes TABLE MARGINAL

351

HEPATOCYTES

I

CYTOPLASMIC BLEB FORMATION AND LDH RELEASE IN HEPAT~CYTE MONOLAYER EXPOSED TO LETHAL CONCENTRATIONS OF VARIOUS HEPATOTOXINS

Agent

Concentration”

Aflatoxin B,

1 PM 10

Acetaminophen Cycloheximide 2-acetylaminofluorene N-hydroxy-2-acetylaminofluorene Senecionine Phalloidin Cytochalasin B Carbon tetrachloride Methotrexate

,LLM

4mM

16 IIIM 500fiM 200 /AM 3.2 40

PM /.tM

1 a/ml 5rtiml

4 IIIM 800

ELM

Percentage of hepatocytes with blebs at 18 hr 34 f 53 f

Percentage of LDH release at 24 hr

9 (4)b 9 (6) 0 (4) 3 + 3 (6) 37 f 7(S) 17 f 17 (4)

40 + 6(7) 46 + 15 (2) 34 f 11 (6)

48 + 15 24 f 15 100' 100' 3 * 3

47 37 23 22 93

I1 *

(5) (6) (2) (2) (3)

5 (5)

18 * 21 f

9(7) 4(6)

10 f

3 (3)

+ 11 (6) + 6 (4) + 5 (3) * 15 (2) f 2 (2)

31 + 2 (2)

CULTURES

Percentage of LDH release at 48 hr 60 zk 16(4) 8lk S(5) 41 f 8 (3) 80 + 6 (5) 9-l (1)

-

91 f 4(2) 83 + 14(4) 70 f 10 (3) 53 f 24 (2) 104 (1) 62 + l(2)

a Concentrations used were those producing between 20 and 50% release of LDH at 24 hr except for additional lower concentrations of AFB, and acetaminophen. * Values are X k SD with numbers in parentheses representing the number of liver preparations. ’ Minute protrusion blebs on all hepatocytes but no marginal blebs.

352

HAYES AND PICKERING

FIG. 8. Phase-contrast micrograph of hepatocytes exposed to phalloidin (1 &ml) numerous round spherical blebs over all surfaces of hepatocytes (arrows).

these additives affect the appearance of blebs produced by Al%,. DISCUSSION These studies demonstrated three distinct modes of toxic cytologic injury in primary monolayer hepatocytes before lethal injury was detected by LDH release, trypan blue uptake, or cell detachment. One class of agents, of which AFB, was representative, caused characteristic large finger-like blebs at the attached cytoplasmic margins free of contact with adjacent cells. Phalloidin and cytochalasin B, each of which are well-recognized causes of hepatocyte surface blebs (Phillips et al., 1974; Prentki et al., 1979; Russo et al., 1982) produced minute spherical blebs that projected from all attached, free, and intercellular surfaces. By comparison, other hepatotoxins such as acetaminophen, carbon tetrachloride, bromobenzene, and

for 18 hr. Note the

methotrexate at equivalently lethal concentrations produced few or no blebs in attached monolayer hepatocytes. These different morphological modes of prelethal cell injury suggest different early mechanisms that cannot be differentiated by later endpoints of cell death such as LDH release, trypan blue uptake, or cell detachment. Since there is considerable evidence that hepatocellular necrosis in vivo can occur by several different morphological or biological mechanisms (Popp et al., 1978; Mitchell et al., 1982; Zimmerman, 1982; Recknagel, 1983; Columbano et al., 1984). the present observations in cell culture may be relevant to the various mechanisms by which chemicals kill hepatocytes in vivo.

Although there are many biological and biochemical endpoints that have been used to detect chemical injury of monolayer hepatocytes (Grisham, 1979; Grisham and Smith, 1984), most investigators use an index

MODES

OF NECROSIS IN MONOLAYER

of membrane permeability, because of the evidence that calcium ion influx after membrane damage is the best demonstrated irreversible step leading to necrosis (Farber, 1982; Recknagel, 1983). The exact mechanisms by which various xenobiotics ultimately lead to critically impaired membrane function have yet to be defined. There is evidence that peroxidative membrane injury (Recknagel, 1983; Recknagel and Glende, 1973; Casini et al., 1982) and covalent binding of metabolites to macromolecules (Mitchell et al., 1982; Casini et al., 1982) may act independently or concurrently before membranes become permeable. Clearly, if membrane leakage is a common critical consequence of separate prelethal mechanisms, the latter can perhaps be better analyzed in assays for the different early cytomorphological events rather than the common later events reflecting membrane leakage. Bleb formation in isolated hepatocytes or other cells is a well-recognized phenomenon (Hoer1 and Scott, 1978; Trump et al., 1980; Chenery et al., 1981; Russo et al., 1982; George et al., 1982; Jewel1 et al., 1982) but the characteristic marginal blebs induced by AFB, and the other hepatocarcinogens differ from those previously described in monolayer or suspension cultures. Phalloidin and cytochalasin B, respectively, stabilize or prevent polymerization of actin into filaments (MacLean-Fletcher and Pollard, 1980) required for various cytokinetic functions of hepatocytes (Prentki et al., 1979). These agents cause blebs rapidly on hepatocytes but this process is reversible and not inevitably a precedent to membrane permeability and calcium influx (Farber, 1982; Russo et al., 1982). Although the appearance, location, time course, and extent of the characteristic blebbing produced by AFB, and similar compounds were clearly different from those of phalloidin- or cytochalasin B-induced blebs, the fact that the marginal finger-like blebs occurred before membrane leakage or trypan blue uptake suggests that they too are the

HEPATOCYTES

353

consequence of a prelethal cytoskeletal perturbation. AFB, (McIntosh et al., 1976; Campbell and Hayes, 1976; Ch’ih et al., 1983; Metcalfe and Neal, 1983b) and cycloheximide (Verbin et al., 1969; Castro et al., 1977) each inhibit protein synthesis in hepatocytes and might thereby interfere with numerous functions including maintenance of cell shape and attachment to the plastic dish (Grinnell, 1978), or secretory functions (Evans, 1981). However, we were unable to prevent AFB1-induced blebbing either by coating the dishes with collagen or by supplementing the medium with rat serum to improve cell attachment to the plastic. The absence of membrane blebs in monolayer cultures of hepatocytes exposed to acetaminophen, carbon tetrachloride, methotrexate or, in previous studies, bromobenzene (Hayes et al., 1984a) indicate that characteristic marginal blebbing is not a general phenomenon associated with lethal cell injury in monolayer cultures. However, these agents can produce surface blebbing in suspension hepatocyte cultures (Jewel1 et al., 1982). Such blebs are minute and appear as cytoplasmic protrusions, whereas the marginal blebs caused by AFB, appeared to be due to segmental contraction of intervening portions of attached cytoplasm and as such are not protrusions. In our experience, freshly isolated untreated hepatocytes sometimes exhibit minute protrusion blebs over the entire surface before attachment but these disappear soon after hepatocytes form monolayers. Protrusion blebs of suspended hepatocytes have been attributed to depletion of the hepatocyte thiol status and a concurrent disturbance on extra mitochondrial calcium ion homeostasis (Jewel1 et al., 1982). The biological relationship between the protrusion blebs observed in suspended hepatocytes within 30 min of exposure to many agents (Jewel1 et al., 1982) and the characteristics larger marginal blebs produced after 6 to 12 hr in monolayers exposed to AFB, and other hepatocarcinogens have not been determined.

354

HAYES

AND

Isolated hepatocytes in suspension cultures in which minute protrusion blebs are perhaps a non-specific rapid phenomenon, are frequently used in quantitative assays or mechanistic studies on chemical hepatocidal injury (Grisham, 1979; George et al., 1982; Smith et al., 198 1; Stacey and Klaassen, 1982; Story et al., 1983). While membrane injury can be detected in these cultures by LDH release or trypan blue uptake within several hours, there is evidence that such rapid lethal injury in suspended hepatocytes is less dependent on calcium ion influx than is lethal injury in monolayer hepatocyte cultures or in viva (Schanne et al., 1979; Casini and Farber, 198 1; Smith et al., 198 1; Stacey and Klaassen, 1982; Farber, 1982). Also, control or chemically injured hepatocytes in suspension survive for much shorter periods than do hepatocytes in monolayers or in vivo. Hepatocytes killed by acetaminophen or carbon tetrachloride were generally stained with trypan blue while attached flat to the plastic dish, suggesting that lethal injury occurred either too rapidly for, or independently of the cytoskeletal alterations involved in marginal blebs induced by AFB, . The fact that the time course of LDH release or trypan blue uptake was similar for the concentrations of AFB, and acetaminophen used suggests that the absence of blebs during acetaminophen toxicity cannot be explained on the basis that lethal injury occurs too rapidly. However, hepatocytes are resistant to acetaminophen cytotoxicity until its metabolites conjugate to and deplete glutathione to critical concentrations (Hinson et al., 198 1; Mitchell et al., 1973, 1982) so there might yet be a brief interval between glutathione depletion and cell death which precludes the formation of marginal blebs in monolayers. Although various phase II systems, including glutathione-S-transferase concentrations are considered important protective mechanisms against AFB,-induced necrosis in cultures (Metcalfe and Neal, 1983a,b), depletion of constitutive amounts of glutathione is prob-

PICKERING

ably a less important determinant for killing by AFB, than by acetaminophen because of the much lower lethal concentrations of AFBl several orders of magnitude below constitutive cell glutathione concentrations. An interesting observation was the frequency with which hepatocarcinogens caused marginal blebs. Unlike acetaminophen, which apparently has a limited range of sublethal toxic effects on hepatocytes, AFB, may be lethal (Campbell and Hayes, 1976; Roberts et al., 1983; Metcalfe and Neal, 1983a,b), mitoinhibitory (Rogers and Newberne, 1967; Novicki et al., 1985), genotoxic (Williams et al., 1980) and carcinogenic (McLean and Marshall, 197 1; Campbell and Hayes, 1976). The relationship between the characteristic cytomorphological changes elicited by these hepatocarcinogens and these various other toxic effects remains to be evaluated. In the present experiments with hepatocytes induced by phenobarbitone or 3-methylcholanthrene, evidence of separately modulated phases of the dose-response curve for hepatocytes killed with AFB, was obtained. Concentrations of AFB, below 10 pM were less lethal in both types of induced hepatocytes but SKF-525A was not protective at all. However, at higher concentrations, SKF-525-A was consistently protective, but phenobarbitone or 3-methylcholanthrene were not. Such a twostage response pattern has not been observed with acetaminophen, bromobenzene, or pyrrolizidine alkaloids (Hayes et al., 1984a,b), but a similar response to AFB, has been reported by Metcalfe and Neal (1983a). The high concentration range for AFB, killing is in the range of the acute LD50 in which zonal liver necrosis is prominent (Campbell and Hayes, 1976). whereas the low concentrations are closer to the sublethal concentrations involved in AFB, carcinogenesis (McLean and Marshall, 197 1). These findings suggest that correlation of data concerning induced metabolic pathways for AFB, with its biological effects should be done with careful analysis of the chemical metabolites

MODES

OF

NECROSIS

IN

MONOLAYER

involved and the specific biological endpoints being considered. We anticipate that early detection of prelethal cytomorphological changes elicited by AFBl will facilitate such a mechanistic analysis. ACKNOWLEDGMENTS These experiments were performed in the laboratories of Dr. E. Farber and Dr. R. Cameron at the University of Toronto. We appreciate the assistance of Esther Roberts, Doris Oesterle, Luisa Di Lorenzo, and Fran Keen.

REFERENCES ACOSTA, D., ANUFURNO, D. C., AND SMITH, R. V. (1980). Cytotoxicity of acetaminophen and papaverine in primary cultures of rat hepatocytes. Toxicol. Appt. Pharmacol.

53, 306-3

14.

CAMPBELL, T. C., AND HAYES. J. R. (1976). The role of aflatoxin metabolism in its toxic lesion. Toxicol. Appl. Pharmacol.

35, 199-222.

CARR, B. I., AND LAISHES, B. A. (1981). Carcinogeninduced drug resistance in rat hepatocytes. Cancer Res. 41, 1715-1719. CASINI, A. F., AND FARBER, J. L. (1981). Dependence of the carbon-tetrachloride-induced death of cultured hepatocytes on the extracellular calcium concentration. Amer.

J. Pathol.

105, 138-148.

CASINI, A., GIORLI, M.. HYLAND, R. J., SERRONI, A., GILFOR, D., AND FARBER, J. L. (1982). Mechanisms of cell injury in the killing of cultured hepatocytes by bromobenzene. J. Biol. Chem. 257, 6721-6728. CASTRO, J. A., DE FERREYA, E. C., DE CASTRO, C. R., DE FENOS. 0. M., DIAZ GOMEZ, M. I., GRAM, T. E., REAGAN, R. L., ANDGUARINO, A. M. (1977). Studies on the role of protein synthesis in cell injury by toxic agents. I. Effect of cycloheximide administration on several factors modulating carbon tetrachloride induced liver necrosis. Toxicol. Appl. Pharmacol. 41, 305-320. CHENERY, R., GEORGE, M., AND KRISHNA, G. (1981). The effect of ionophore A23 187 and calcium on carbon tetrachloride-induced toxicity in cultured rat hepatocytes. Toxicol. Appl. Pharmacol. 60, 241-252. CH’IH, J. J., LIN, T., AND DEVLIN, T. M. (1983). Effect of inhibitors of microsomal enzymes on aflatoxin B,induced cytotoxicity and inhibition of RNA synthesis in isolated rat hepatocytes. Biochem. Biophys. Res. Cummun.

115, 15-21.

COLUMBANO, A., LEDDA-COLUMBANO, G. M., RAO, P. M., RAJALAKSHMI, S.. ANDSARMA, D. S. R. (1984).

355

HEPATOCYTES

Occurrence of cell death (apoptosis) in preneoplastic and neoplastic liver cells: A sequential study. Amer. J. Pathol.

116, 44 I-446.

EVANS, W. H. (1981). Membrane traffic at the hepatocyte’s sinusoidal and canalicular surface domains. Hepatology

1, 452-457.

FARBER, J. L. (1982). Membrane injury and calcium homeostasis in the pathogenesis of coagulative necrosis. Lab. Invest. 47, 114-123. FARBER, J. L., AND GERSON, R. J. (1984). Mechanisms of cell injury with hepatotoxic chemicals. Pharmacol. Rev. 36, 7 I S-74s.

GEORGE, M. G., CHENERY, R. J.. AND KRISHNA, G. (1982). The effect of ionophore A23187 and 2,4dinitrophenol on structure and function of cultured liver cells. Toxicol. Appi. Pharmacol. 66, 349-360. GREEN, C. E.. RICE, D. W.. HSIEH, D. P. H., AND BYARD. J. L. (1982). The comparative metabolism and toxic potency of aflatoxin B, and aflatoxin M, in primary cultures of adult-rat hepatocytes. Food Chem. Toxicol.

20, 53-60.

GRINNELL, F. (1978). Cellular adhesiveness and extracellular substrata. Int. Rev. C.vfo/. 53, 65-144. GRISHAM, J. W. (1979). Use of hepatic cell cultures to detect and evaluate mechanisms of action of toxic chemical. Int. Rev. Exp. Pachol. 20, 123-210. GRISHAM, J. W., AND SMITH, G. J. (1984). Predictive and mechanistic evaluation of toxic responses in mammalian cell culture systems. Pharmacol. Rev. 36, 15lS-171s. HAYES, M. A., ROBERTS, E.. ROOMI, M. W.. SAFE, S. H., FARBER, E., AND CAMERON, R. G. (1984a). Comparative influences of different PB-type and 3MC-type polychlorinated biphenyl-induced phenotypes on cytocidal hepatotoxicity of bromobenzene and acetaminophen. Toxicol. Appi. Pharmacol. 76, 118127. HAYES, M. A.. ROBERTS, E., JAGO. M. V., SAFE, S. H., FARBER, E.. AND CAMERON, R. G. (1984b). Influences of various xenobiotic inducers on cytocidal toxicity of lasiocarpine and senecionine in primary cultures of rat hepatocytes. J. Toxicol. Environ. Health 14, 683694. HAYES, M. A., CAMERON. R. G., ROBERTS. E., SAFE. S. H., AND FARBER,E. (1984~). Modulation of cytocidal and mitoinhibitory toxicity of hepatocarcinogens in rats pretreated with different PCBs and other inducers. Proc. Amer.

Assoc. Cancer

Res. 25, 505.

HINSON,J. A., POHL, L. R.. MONKS, T. J., ANDGILLETTE. J. R. (198 I). Acetaminophen-induced hepatotoxicity. Life Sri. 29, 107-l 16. HOERL. B. J., AND SCOTT, R. E. (1978). Plasma membrane vesiculation: A cellular response to injury. Virchows Arch. B Cell Pathol.

21, 335-345.

JAUREGUI. H. 0.. HAYNER, N. T., DRISCOLL, J. L.. WILLIAMS-HOLLAND, R.. LIPSKY, M. H., AND GAL-

356

HAYES AND PICKERING

LETTI, P. M. (1981). Trypan blue uptake and lactate dehydrogenase in adult rat hepatccytes-freshly isolated cells, cell suspensions, and primary monolayer cultures. In Vitro 17, 1100-I 1IO. JEWELL, S. A., BELLOMO, G., THOR, H., ORRENIUS, S., AND SMITH, M. T. (1982). Bleb formation in hepatocytes during drug metabolism is caused by disturbances in thiol and calcium ion homeostatis. Science (Washington. D.C.) 217, 1257-1259. LOURY, D. J.. HSIEH, D. P. H., AND BYARD, J. L. (1984). The effect of phenobarbitone pretreatment on the metabolism, covalent binding and cytotoxicity of aflatoxin Bi in primary cultures of rat hepatocytes. J. Toxicol. Environ. Health 13, 145-159. MACLEAN-FLETCHER, S., AND POLLARD, T. D. (1980). Mechanism of action of cytochalasin B on actin. Cell 20, 329-34 1. MCINTOSH, D. E.. EVANS, I. M., AND RABIN. B. R. (I 976). The effect of aflatoxins on the incorporation of RNA and protein precursors by isolated hepatocytes. Brit. J. Cancer 33, 440-449. MCLEAN, A. E. M., AND MARSHALL, A. ( 197 I). Reduced carcinogenic effects of aflatoxin in rats given phenobarbitone. Brit. J. Exp. Pathol. 52, 322-329. METCALF’E, S. A.. AND NEAL, G. E. (1983a). The metabolism of aflatoxin B, by hepatocytes isolated from rats following the in vivo administration by some xenobiotics. Carcinogenesis 4, lOO7- IO 12. METCALFE, S. A., AND NEAL, G. E. (1983b). Some studies on the relationship between cytotoxicity of aflatoxin Bi and metabolism of the toxin. Carcinogenesis 4, 1013-1019. MITCHELL, J. R., HUGHES, H., LALJTERBURG, B. H., AND SMITH, C. V. (1982). Chemical nature of reactive intermediates as determinant of toxicologic responses. Drug Metab. Rev 13, 539-553. MITCHELL, J. R., JOLLOW, D. J., POTTER, W. Z., GILLETTE, J. R., AND BRODIE, B. B. (1973). Acetaminophen-induced hepatic necrosis. IV. Protective role of glutathione. J. Pharmacol. Exp. Thu. 187,

PRENTKI, M., CHAPONNIER. C.. JEANRENALJD,B., AND GABBIANI, G. (1979). Actin microfilaments cell shape, and secretory processes in isolated rat hepatocytes. Effect of phalloidin and cytochalasin D. J. Cc/l. Biol. 81, 592-607.

RECKNAGEL, R. 0. (1983). A new direction in the study of carbon tetrachloride hepatotoxicity. L@ Sci. 33, 401-408.

RECKNAGEL, R. O., AND GLENDE, E. S. (1973). Carbon tetrachloride hepatotoxicity: An example of lethal cleavage. Crit. Rev. Toxicol. 2, 263-297. ROBERTS, E., AHLUWAHLIA, M. B., LEE, G., CHAN, C., SARMA, D. S. R., AND FARBER, E. (I 983). Resistance to hepatotoxins acquired by hepatocytes during liver regeneration. Cancer Res. 43, 28-34. ROGERS, A. E., AND NEWBERNE, P. M. (1967). The effects of aflatoxin B, and dimethylsulfoxide on thymidine-3H uptake and mitosis in rat liver. Cancer Res. 27, 855-864.

Russo, M. A., KANE, A. B.. AND FARBER, J. L. (1982). Ultrastructural pathology of phalloidin-intoxicated hepatocytes in the presence and absence of extracellular calcium. Amer. J. Path& 109, 133-144. SCHANNE, F. A. X., KANE. A. B.. YOUNG, E. E., AND FARBER, J. (1979). Calcium dependence of toxic cell death: A final common pathway. Science (Washington, D.C.) 206, 700-702. SMITH, M. T.. THOR, H., AND ORRENIUS, S. (198 I). Toxic injury to isolated hepatocytes is not dependent on extracellular calcium. Science (Washington, D.C.) 213, 1257-1259. STACEY, N. H., AND KLAASSEN, C. D. (1982). Lack of protection against chemically induced injury to isolated hepatocytes by omission of calcium from the incubation medium. J. Toxicol. Environ. Health 9, 267-276. STORY, D. L.. GEE, S. J.. TYSON, C. A., AND GOULD, D. H. (1983). Response of isolated hepatocytes to organic and inorganic cytotoxins. J. Toxicol. Environ. Health 11, 483-501. TRUMP, B. F., BEREZESKY,I. K., LAIHO, K. U., OSORNIO, 21 l-217. A. R., MERGNER, W. J., AND SMITH, M. W. (1980). The role of calcium in cell injury: A review. In NOVICKI, D. L., ROSENBERG,M. R., AND MICHALOPOULOS, G. (1985). Inhibition of DNA synthesis by chem“Scanning Electron Microscopy,” (A. M. F. O’Hare, Ed.), Vol. 2, pp. 437-458. SEM Inc., Chicago. ical carcinogens in cultures of initiated and normal proliferating rat hepatocytes. Cancer Res. 45, 337VERBIN, R. S., GOLDBLATT, P. J., AND FARBER, E. (1969). The biochemical pathology of inhibition of 344. protein synthesis in vivo. The effects of cycloheximide PHILLIPS,M. L., ODA, M., EDWARDS,V. O., GREENBERG, on hepatic parenchymal cell ultrastructure. Lab. Invest. G. R., AND JEWEEBHOY, N. (1974). Ultrastructural 20, 529-536. and functional studies of cultured hepatocytes. Lab. WILLIAMS, G. M. (I 980). Classification of genotoxic and Invest. 31, 533-542. epigenetic hepatocarcinogens using liver culture assays. POPP. J. A., SHINOZUKA, H., AND FARBER, E. (1978). The protective effects of diethyldithiocarbamate and Ann NY Acad. Sci. 349, 273-282. ZIMMERMAN, H. J. (1982). Chemical hepatic injury and cycloheximide on the multiple hepatic lesions induced its detection. In “Toxicology of the Liver” (G. L. Plaa by carbon tetrachloride in the rat. Toxicol. Appl. and N. R. Hewitt, Eds.), Raven Press, New York. Pharmacol. 45, 549-564.