- Email: [email protected]
Potentiation of vitamin A hepatotoxicity by butylated hydroxytoluene
TOXICOLOGY AND APPLIED PHARMACOLOGY 90,1-9
Potentiation DAVID Laboratory
(1987)
of Vitamin A Hepatotoxicity L. MCCORMICK,
ofPathophysiology.
Received
THERESA
Life Sciences
November
by Butylated
A. HULTIN,
Research,
AND CAROL J. DETRISAC
IIT Research
10, 1986; acceptedApril
Hydroxytoluene
Institute,
Chicago,
Illinois
60616
14, 1987
Potentiation of Vitamin A Hepatotoxicity by Butylated Hydroxytoluene. MCCORMICK, D. L., HULTIN, T. A., AND DETRISAC, C. J. (1987). Toxicol. Appl. Pharmacol. 90, l-9. The interaction between the natural vitamin A ester retinyl acetate (BA) and the phenolic antioxidant butylated hydroxytoluene (BHT) in the induction of biliary hyperplasia and hepatic fibrosis in female Sprague-Dawley rats was characterized. Using a 3 X 3 matrix design, rats were fed diets supplemented with (per kilogram diet) 0, 125, or 250 mg BA and/or 0, 2500, or 5000 mg BHT. The 125mg dose of RA induced no gross hepatotoxicity, while the 250-mg dose of RA induced a low incidence of hepatic fibrosis in rats examined after 120 and 180 days of exposure. Exposure to BHT alone induced hepatocellular hypertrophy and dose-related increases in liver weight, but no hepatocellular pathology. Simultaneous administration of RA plus BHT resulted in significant increases in the incidence of biliary hyperplasia and hepatic fibrosis compared to that induced by RA alone. BHT reduced total hepatic vitamin A content at all BA dose levels. Thus, mechanisms other than increases in liver vitamin A levels must underlie the potentiation by BHT of RA hepatotoxicity. o 1987 Academic PKSS, hc.
gland (Wattenberg, 1972; McCormick et al., 1984). As seen with RA, BHT inhibits mammary carcinogenesis when administered at dose levels which induce no gross or organspecific toxicity. Chronic exposure to BHT is associated with hepatic enzyme induction and dose-related increases in liver weight (Peraino et al., 1977; Hirose et al., 1981); however, the hepatic effects of BHT are reversible upon cessation of exposure (Botham et al., 1970) and thus are not considered to be of toxicologic significance (Nakagawa et al., 1984). Although both RA and BHT have significant anticarcinogenic activity in the rat mammary gland, their activity as single agents is imperfect; neither RA nor BHT can reduce cancer incidence to zero. One mechanism through which the magnitude of cancer inhibition may be increased is by the use of “combination chemoprevention” protocols, in which multiple inhibitors of carcinogenesis are administered simultaneously or se-
The natural vitamin A ester retinyl acetate (RA) has significant activity as an inhibitor of mammary carcinogenesis in the rat. Previous studies have demonstrated the efficacy of RA in suppressing rat mammary carcinogenesis induced by chemical carcinogens (Moon et al., 1976, 1977; McCormick et al., 1981) and by radiation (Stone et al., 1984). Furthermore, chronic exposure to RA can inhibit the occurrence of “spontaneous” mammary cancers in rats (Stone et al., 1984). In all of these studies, inhibition of mammary carcinogenesis was achieved using RA dose levels which were without significant systemic or organspecific toxicity; however, in several experiments, the liver and bone have been identified as the most sensitive target organs for the toxic effects of RA and other retinoids (Leelaprute et al., 1973; Sani and Meeks, 1983; McCormick et al., 1986). Butylated hydroxytoluene (BHT), a phenolic antioxidant, is also an effective inhibitor of cancer induction in the rat mammary 1
0041-008X/87
$3.00
Copyright 0 1987 by Academic Press, Inc. All rights of reproduction in any form reserved
2
MC CORMICK,
HULTIN,
quentially. We have recently reported that an enhanced inhibition of mammary carcinogenesis can be achieved by combined administration of RA and BHT (McCormick et al., 1986); however, the increased anticarcinogenie activity achieved by chronic exposure to RA and BHT was accompanied by the induction of biliary hyperplasia and hepatic fibrosis. The present study was performed to characterize the hepatotoxic interaction between RA and BHT and to study the temporal, pharmacologic, and histopathologic parameters associated with its induction.
AND DETRISAC
or choke&tic damage to the liver or as a result of hepatic dysfunction (Plaa and Hewitt, 1982). Serum levels of ornithine carbamyl transferase (GCT) were quantitated using the calorimetric method described by Plaa and Hewitt (1982). Levels of sorbitol dehydrogenase (SDH), glutamic-pyruvic transaminase (alanine transaminase; GPT/ALT), glutamic-oxaloacetic transaminase (aspartate transaminase; GGT/AST), lactate dehydrogenase (LDH), alkaline phosphatase (AP), 5’-nucleotidase, yglutamyl transpeptidase (GGT), and cholinesterase were analyzed using a Centrifichem 500 autoanalyzer system (Baker Instruments, Allentown, PA). Statistics. Intergroup comparisons of mean body weight, mean liver weight, hepatic vitamin A concentration, total liver vitamin A content, and serum enzyme levels were performed using analysis of variance. Incidence comparisons were made using x2 analysis.
METHODS RESULTS Experimental animals. Female Sprague-Dawley rats were obtained at 28 days of age from Harlan/SpragueDawley (Indianapolis, IN). Rats were housed in groups of three in polycarbonate cages on hardwood bedding in a windowless room maintained at 22 + 1°C on a 14-hr light/lo-hr dark cycle. Animals were quarantined for 1 week prior to the initiation of the experiment. All rats had free access to diet and drinking water throughout the study. Experimental diets. Basal diet used in the experiment was Wayne Lab Blox MRH 22/5 (Allied Mills, Chicago, IL); this diet contains 8 mg vitamin A/kg, as added retinyl palmitate, and contains no BHT, BHA, ethoxyquin, propyl gallate, or other synthetic antioxidants. RA (in gel beadlet form) was obtained from Hoffmann-LaRoche (Nutley, NJ) and was administered as a dietary supplement as required by the protocol. BHT was purchased from Sigma Chemical Co. (St. Louis, MO) and was also administered as a dietary supplement. Experimental protocol. Using a 3 X 3 matrix design, groups of 25 rats were fed diets supplemented with (per kilogram diet) 0, 125, or 250 mg RA and/or 0,2500, or 5000 mg BHT. At these dietary levels, daily intake of RA is approximately 0, 2, and 4 mg/day for the 0, 125, and 250 RA doses, respectively; BHT intake is approximately 0,40, and 80 mg/day for the 0,2500, and 5000 BHT dose groups. After 30,60,90, 120, and 180 days on diet, 5 rats per group were terminally bled via the inferior vena cava; serum was collected and kept on ice for enzyme assays, as described below. Livers were excised and weighed, and sections were taken and fixed in 10% buffered formalin for analysis via light microscopy. At the 180-day time point, liver samples from each animal were frozen in liquid nitrogen for analysis of vitamin A content using the trifluoroacetic acid method (Neeld and Pearson, 1963). Serum enzymes were selected for assay on the basis of reported alterations in their levels as a result of cytotoxic
Gross observations. The overall health status of rats fed supplemental RA and/or BHT was good throughout the 180-day study period. No mortality or indications of systemic vitamin A toxicity (skin or fur changes, brittle bones) were observed in any animals fed supplemental RA nor was gross toxicity associated with BHT exposure. Terminal body weights ranged from 89 + 3% of control in the group receiving 125 mg RA + 2500 mg BHT to 100 f 4% of control in the group fed 250 mg RA + 2500 mg BHT. Although RA induced no obvious gross toxicity, a low incidence of hepatic fibrosis was observed at necropsy in rats receiving chronic exposure to the retinoid. Dietary administration of BHT had no such effects on liver morphology. However, when administered in combination with RA, BHT resulted in a significant enhancement of RA hepatotoxicity: changes in hepatic morphology in animals exposed to RA + BHT occurred earlier, were seen in higher incidence, and were more severe than were alterations in groups exposed to RA alone. The effects of RA alone on gross hepatic morphology were dose-related. Dietary administration of RA at a level of 125 mg/kg diet resulted in no grossly detectable hepatic pathology during the 180-day study period.
BHT-VITAMIN TABLE
A HEPATOTOXIC
1
IINFLUENCE OF RETINYL ACETATE AND BHT ON INCIDENCEOF GROSS HEPATIC FIBROSIS Days on diet RA dose (mg/kg diet)
BHT dose (mg/kgdiet)
0 0 0
0 2500 5000
0000 0000 0000
125 125 125
0 2500 5000
0000 000 0 0
0
0 215
250 250 250
0 2500 5000
0 0 0
0 0 l/5
l/5 l/5 5156 l/5 5156 5156
30 60
90
120
180 0 0 0
0 0 0
0 315” 4/5b
‘p < 0.05 versus control (0 RA, 0 BHT), appropriate BHT control group (0 RA, same BHT dose), and appropriate RA control (same RA dose, 0 BHT). bp < 0.0 1 versus control (0 RA, 0 BHT), appropriate BHT control group (0 RA, same BHT dose), and appropriate RA control (same RA dose, 0 BHT).
By contrast, administration of RA at 250 mg/ kg diet was associated with alterations in hepatic morphology at later time points (Table 1); one of five rats examined at both the 120and the 180-day time points displayed focal hepatic fibrosis. At 120 days, the only affected animal demonstrated a single focus on one liver lobe. Hepatic involvement at 180 days was more extensive than at 120 days, as multiple foci of fibrosis were diffusely scattered over several liver lobes of the one affected rat. Dietary exposure to RA alone also resulted in a slight degree of hepatomegaly in treated groups. A 6% increase in mean liver weight was observed in groups exposed to the 125mg dose of RA, while the 250-mg dose of RA increased mean liver weight by 12% (data not shown). However, no consistent pattern of statistical significance or trend over time was observed when liver weights in these groups were compared to control. Administration of BHT at levels of 2500 or 5000 mg/kg diet for up to 180 days had no effect on gross hepatic morphology. Consistent with previous reports, BHT administra-
3
INTERACTION
tion was associated with the rapid induction of hepatomegaly; increases in liver weight in groups exposed to BHT were dose-related. After 30 days of exposure, mean liver weights were increased by approximately 30 and 40% from control levels by the 2500 and 5000 mg BHT doses, respectively (p < 0.01); after 30 days, liver weight remained elevated, but no further increases were observed (Fig. 1). Administration of RA had no effect on the induction of hepatomegaly by BHT. In contrast to the apparent lack of effect of RA on BHT induction of hepatomegaly, exposure to BHT did have an enhancing influence on the hepatic toxicity of RA. When administration of the 5000-mg dose of BHT was combined with the 250-mg dose of RA, the incidence of gross hepatic fibrosis was increased from 20 to 100% at 120 and 180 days of exposure. Liver lobes containing more than 10 discrete foci of fibrosis on the organ surface were a common finding at both the 120- and 1 go-day time points in the group exposed to 250 mg RA + 5000 mg BHT. Furthermore, simultaneous exposure to BHT appeared to accelerate the onset of liver damage induced by RA: while fibrosis
15
r
0
I 30
I 60
I 90 Days On Cue?
I 120
I 150
I 180
FIG. 1. InlIuence of BHT dose on liver weight in Sprague-Dawley rats. Each point represents the mean f SD for data from five animals. 0, 0 mg BHT/kg diet (control); A, 2500 mg BHT/kg diet; n , 5000 mg BHT/kg diet. p < 0.0 1 versus control for each data point except 2500 may BHT/ke diet at 120 davs. < I where D < 0.05.
4
MC CORMICK,
HLJLTIN. AND DETRISAC
FIG. 2. Liver section from female Sprague-Dawley rat exposed to 250 mg RA/kg diet for 180 days. Note the numerous enlarged Ito cells (arrows). Hematoxylin and eosin, X 120.
was first observed at 120 days in rats receiving the 250 mg RA supplement alone, such changes were seen at 90 days in the group treated with 250 mg RA + 5000 mg BHT. The effects of the lower dose of BHT were similar, although the increased incidence of hepatic fibrosis in the 250 mg RA + 2500 mg BHT group was statistically significant in comparison to the 250 mg RA group only at the 1SO-day time point. A parallel series of observations was made in groups exposed to the 125-mg dose of RA. Although no gross alterations in hepatic morphology were observed in animals exposed to this dose of RA only, when RA was administered in combination with BHT, morphologic changes were observed. At the 180-day time point, the incidence of hepatic fibrosis in groups exposed to 125 mg RA in combination with either dose of BHT was significantly increased in comparison to the effects of the retinoid alone (Table 1).
Light microscopic observations. Chronic administration of RA to Sprague-Dawley rats induced a characteristic pattern of hepatic alterations visible at the light microscope level. Prominent in the livers of RAtreated animals were large numbers of lipidcontaining perisinusoidal stellate cells (Ito cells; Fig. 2). Ito cells have been proposed to be the primary site of hepatic storage of vitamin A in both rats and humans (Bronfenmajer et al., 1966; Kobayashi et al., 1973; Wake, 1980) and have been implicated in the process of hepatic fibrogenesis (Wake, 1980). Although the number of Ito cells per tissue section was not clearly related to the dose of RA, proliferation of It0 cells in response to hepatotoxic agents has been reported (McGee and Patrick, 1972; Enzan, 1985). The gross lesions noted at 120 and 180 days in groups exposed to 250 mg RA were classified histopathologically as diffuse periportal fibrosis, which was frequently associ-
BHT-VITAMIN
A HEPATOTOXIC
INTERACTION
5
FIG. 3. Liver section from rat exposed to 5000 mg BHT/kg diet for 180 days. Hepatocellular hypertrophy is evident in cells surrounding the central veins (CV). Hematoxylin and eosin, X 120.
ated with bile duct hyperplasia. In all cases, areas of fibrosis observed in animals exposed to 250 mg RA were located at the periphery of the liver. Alterations in liver histology which were observed in rats exposed to dietary BHT were consistent with the gross findings of increased liver weight. BHT-induced hepatomegaly was characterized by hepatocellular hypertrophy which was distributed asymmetrically across the liver lobule. As shown in Fig. 3, hepatocyte size was greatest in regions adjacent to the central vein, while cells in the periportal regions exhibited little or no alteration in cellular dimensions as a result of BHT exposure. Histopathologic alterations observed in livers from rats exposed to RA + BHT consisted of increases in the incidence and severity of hepatic fibrosis compared to that induced by RA only. Although usually located peripherally, focal fibrosis could be demonstrated
throughout the liver parenchyma of rats fed RA + BHT. These lesions were most commonly located in periportal regions of the liver lobule and contained an abnormal number of bile ducts and tissue histiocytes (Fig. 4). Focal portal hepatic necrosis accompanied by an inflammatory infiltrate was also commonly observed. In addition to the increased number and size of fibrotic regions and associated biliary hyperplasia, livers from animals administered RA and BHT in combination also demonstrated enlarged Ito cells and asymmetric hepatocellular hypertrophy, indicative of exposure to RA or BHT alone. Clinical chemistries. Prior to the 180-day time point, neither RA nor BHT, administered alone or in combination, induced any consistent pattern of alterations in levels of serum enzymes. At the 180-day time point, significant increases in serum levels of LDH were observed in groups exposed to 125 mg RA + 5000 mg BHT, 250 mg RA + 2500 mg
6
MC CORMICK,
HULTIN,
AND DETRISAC
FIG. 4. Focus of hepatic fibrosis from rat exposed to 250 mg RA + 5000 mg BHT/kg diet for 180 days. Note abnormal number of bile ducts (arrows). Hematoxylin and eosin, X 120.
BHT, and 250 mg RA + 5000 mg BHT (Table 2). LDH levels in these groups ranged from 400 to 1100% of control values (p < 0.0 1) and were also greater than LDH levels in groups exposed to either RA only or BHT only. Statistically significant, but quantitatively smaller, increases from control levels of AP and GPT/ALT were observed in groups exposed to 250 mg RA in combination with either dose of BHT (Table 2). However, serum activity of these enzymes was not significantly increased from levels observed in the group receiving 250 mg RA alone. In contrast to the observed increases in levels of LDH, AP, and GPT/ALT, administration of RA in combination with BHT induced no statistically significant increases in serum levels of enzymes (OCT, SDH) which are considered to be highly specific indicators of hepatic damage. Conversely, although significant decreases in cholinesterase or increases in 5’-nucleotidase or GOT/AST activity were observed in several groups, these alterations
were not correlated with histologically detectable liver toxicity. Hepatic vitamin A levels. The influence of RA and/or BHT on hepatic levels of vitamin A compounds is presented in Table 3. As expected, administration of 125 or 250 mg RA/ kg diet resulted in significant increases in liver vitamin A stores: vitamin A concentration in these two groups was increased more than lofold in comparison to control. By contrast, administration of BHT reduced hepatic vitamin A concentration at all RA doses used. In animals receiving no RA supplementation, liver vitamin A concentration was reduced by approximately 50% in animals exposed to BHT. In groups receiving 125 or 250 mg RA/kg diet, the 2500-mg dose of BHT reduced hepatic vitamin A concentration by 17 to 19% from control levels, while the 5000-mg dose of BHT resulted in a 46 to 48% reduction. The decreases in hepatic vitamin A concentration associated with BHT exposure cannot be explained solely on the basis of the
BHT-VITAMIN
A HEPATOTOXIC TABLE
ALTERATIONS
RA dose (mg/kg diet)
IN SERUM ENZYME
BHT (m&kg
dose diet)
AP 19 9 27
137+ 114+ 1172
GPT/ALT
0 0 0
0 2500 5000
125 125 125
0 2500 5000
49% 17 93+ 9 272 + 127c,e
193+ 30 1482 25 197 xi 145
250 250 250
0 2500 5000
lOOk 35 598+217”’ 213+204b
217+ 235-t 214+
a All bp < ‘p < dp < ‘p i
2
LEVEU IN SPRAGUE-DAWLEY RATS RECEIVING DIETARY RETINYL ACETATE AND/OR BHT”
LDH 54* 31-c 70+
7
INTERACTION
28 29 27
44’ 63’ 32”
43-c 40+ 43+
GOT/AST
12 5 8
180 DAYS EXPOSURE
S-Nucleotidase
TO
Cholinesterase
74f 12 48 f 4’ 45 + 4’
14* 12* 12+
2 1 1
31102568 2580 + 388’ 2546 f 291b
67 & 15b 46+ 10 55 f 12
81 f 17 54 f 7b.d 66 2 sd
17+ 2 15* I 31*296
2582 + 282’ 2783 -c 199 2804 3~ 382
75 f 24’ 79 + 18’ 83+21’
87& 8 90 f 126 75*22
18? 2 35 + 16’ 25+ 5
2574 + 459’ 2588 + 156b 2279 zk 586’
values are expressed as milliunits per milliliter serum. 0.05 versus control (0 RA, 0 BHT). 0.0 1 versus control (0 RA, 0 BHT). 0.05 versus appropriate RA control (same dose RA, 0 BHT). 0.01 versus appropriate RA control (same dose RA, 0 BHT).
increases in liver weight induced by BHT; when total liver vitamin A content, rather than concentration of vitamin A per gram of tissue, is considered, the reduction in vitamin A stores as a function of BHT exposure persists in five of six BHT dose groups, although the magnitude of the reduction is decreased (Table 3).
(GPT and GOT) as a result of BHT administration. In the present experiment, neither hepatocellular toxicity nor elevation of serum transaminases was observed in groups exposed to BHT only. It should be noted, how-
TABLE
3
INFLUENCE OF RETINYL ACETATE AND BHT ON LIVER VITAMIN A CONTENT
DISCUSSION
Liver
The dose-response and time-response data obtained in the present experiment suggest that the increased incidence of hepatic damage observed in animals exposed to RA in combination with BHT reflects a potentiation by BHT of the hepatotoxic activity of RA. Although BHT administration did result in the induction of hepatomegaly, it appears that this compound has essentially no other effects on hepatic morphology at the dose levels used in the present study. The induction of hepatic necrosis in rats by BHT exposure has been reported by Nakagawa et al. (1984, 1985); these authors also reported significant increases in serum levels of transaminases
Ra dose (mgjkg
diet)
BHT dose (mg/kg
diet)
vitamin A concentration
Total liver vitamin A content
(pm00
(rmok)
0 0 0
0 2500 5ooo
1.9 kO.1 I.0 f 0.26 0.9 + 0.2b
125
0 2500 5ooo
21.7 + 1.9” 17.5 * 2.1” 11.8 +- 0.7”,b
214.8 + 17.9’ 185.8 f 22.3” 156.9 f 9.3-
0 2500
25.4 f 0.5” 21.1 f 1.2” 13.2 f 2.0”,b
238.8 272.2 171.6
125
125 250 250 250
5000
15.4* 10.8? 10.8 2
0.9 2.2c 4.8’
k 4.7” iz 15.5” +-26P’
’ p < 0.0 1 versus control (0 RA, 0 BHT). bp < 0.01 versus appropriate FU control group (same RA dose, 0 BHT). ‘p < 0.05 versus appropriate RA control group (same RA dose. 0 BHT).
8
MC CORMICK,
HULTIN,
ever, that the doses of BHT found to be necrogenic via gavage administration in the studies of Nakagawa and colleagues ( 1984, 1985) were more than threefold higher than the highest dietary exposure level used in the present study. The lack of hepatic toxicity of dietary BHT observed in the present experiment is in agreement with the data of Hirose et al. ( 198 1), who conducted a 2-year feeding study in rats using BHT doses of 2500 and 10,000 mg/kg diet. By contrast to the apparent lack of hepatotoxic effects associated with exposure to BHT as a single agent, a low incidence of hepatotoxicity was observed in rats exposed to RA alone; the histologic features of liver damage induced by RA in rats are similar to those reported in cases of hypervitaminosis A in humans (Russell et al., 1974; Jacques et al., 1979). It appears that the hepatotoxic activity of the doses of RA used in the present study is near threshold, since gross liver damage was observed in only one of five rats receiving either 120 or 180 days of exposure to the 250 mg/kg diet RA supplement. This incidence of diffuse hepatic fibrosis at 6 months is consistent with our previous report (McCormick et al., 1986). In quantitative terms based on data obtained for high dose groups at the 180-day time point, the interaction between BHT and RA in the induction of liver toxicity can be expressed as 0 + 1 = 5; while liver damage was seen in O/5 rats exposed to the 5000 mg/ kg diet supplement of BHT, and gross hepatic fibrosis was seen in only l/5 rats exposed to RA at a level of 250 mg/kg diet, administration of these doses of RA and BHT in combination induced grossly detectable liver damage in 5/5 animals. The observations that BHT has no apparent hepatic toxicity as a single agent, yet can increase the magnitude of damage induced by RA, conform exactly to the definition of toxicologic potentiation. Similar quantitative comparisons can be made among other dose groups (see Table 1). Because these data indicate that the observed liver toxicity is a potentiation by BHT
AND DETRISAC
of a toxic manifestation of a retinoid, one possible mechanism for this effect could lie in alterations by BHT of vitamin A accumulation in the liver. Should exposure to BHT result in increases in hepatic vitamin A concentration, the functional result of the BHT-RA interaction would be to increase the dose of a toxic agent at its site of action. In order to address this hypothesis, the influence of BHT on hepatic vitamin A concentration was determined. As indicated in Table 3, exposure to BHT decreased, rather than increased, hepatic vitamin A levels at all RA doses used. On this basis, the hypothesis that BHT potentiates retinoid hepatotoxicity through alterations in retinoid metabolism, resulting in increased liver levels of vitamin A, cannot be supported. Thus, the mechanism by which BHT acts to increase RA toxicity remains unknown. It appears, however, that BHT enhancement of RA liver toxicity occurs via potentiation of an as yet undefined biological activity of the retinoid, rather than through a pharmacologic mechanism resulting in increased target organ concentrations of the toxic agent. This increased biological activity is particularly interesting in that the potentiation of RA hepatotoxicity occurs in the presence of decreased, rather than increased, vitamin A levels in the liver. ACKNOWLEDGMENTS This work was supported by Contract NO l-CP-4 1063 from the National Cancer Institute, DHHS. We acknowledge the excellent technical assistance of Mary Ann Cahill, Mark S. Filla, Carrie Martinez, and Steven J. Moore. Patricia Moser provided expert secretarial assistance.
REFERENCES BOTHAM, C. M., CONNING, D. M., HAYES, J., LITCHFIELD, M. H., AND MCELLIG~TT, T. F. (1970). Effects
of butylated hydroxytoluene on the enzyme activity and ultrastructure of rat hepatocytes. Food Cosmet. Toxicol. 8, 1-8.
BHT-VITAMIN
A HEPATOTOXIC
BRONFENMAJER, S., SCHAFFNER, F., AND POPPER, H. (1966). Fat-storing cells (lipocytes) in human liver. Arch. Pathol. 82,447-453. ENZAN, H. (1985). Proliferation of Ito cells (fat-storing cells) in acute carbon tetrachloride liver injury. Acta Pathol. Japan. 35, 1301-1308. HIROSE,
M.,
SHIBATA,
M.,
HAGIWARA,
A., IMAIDA,
K.,
AND ITO, N. (198 1). Chronic toxicity of butylated hydroxytoluene in Wistar rats. Food Cosmet. Toxicol. 19,147-151. JACQUES, E. A., BUSCHMANN, R. J., AND LAYDEN, T. J. ( 1979). The histopathologic progression of vitamin Ainduced hepatic injury. Gastroenterology 76,599-602. KOBAYASHI, K., TAKAHASHI, Y., AND SHIBASAKI, S. (1973). Cytological studies of fat-storing cells in the liver ofrats given large doses of vitamin A. Nature New Biol. 243, 186-l 89. LEELAPRUTE, V., BOONPUCKNAVIG, V., BHAMARAPRAVATI, N., AND WEERAPRADIST, W. (1973). Hypervitaminosis A in rats. Arch. Pathol. 96,5-9. MCCORMICK, D. L., BURNS, F. J., AND ALBERT, R. E. (198 1). Inhibition of benzo[a]pyrene-induced mammary carcinogenesis by retinyl acetate. J. Natl. Cancer Inst. 66,559-564. MCCORMICK, D. L., MAJOR, N. M., AND MOON, R. C. (1984). Inhibition of 7,12-dimethylbenz[a]anthracene-induced rat mammary carcinogenesis by concomitant or post-carcinogen antioxidant exposure. Cancer Res. 44,2858-2863. MCCORMICK, D. L., MAY, C. M., THOMAS, C. F., AND DETRISAC, C. J. (1986). Anticarcinogenic and hepatotoxic interactions between retinyl acetate and butylated hydroxytoluene in rats. Cancer Res. 46, 52645269. MCGEE. J. O., AND PATRICK, R. S. (1972). The role of p&sinusoidal cells in hepatic fibrogenesis. Lab. Invest. 26,429-440. MOON, R. C., GRUBBS, C. J., AND SPORN, M. B. (1976). Inhibition of 7,12 - dimethylbenz[a]anthracene - induced mammary carcinogenesis by retinyl acetate. Cancer Res. 36,2626-2630. MOON, R. C.. GRUBBS, C. J., SFQRN, M. B., AND GOODMAN, D. G. (1977). Retinyl acetate inhibits mammary
INTERACTION
9
carcinogenesis induced by N-methyl-N-nitrosourea. Nature (London) 267,620-62 1. NAKAGAWA, Y., TAYAMA, K., NAKAO, T., AND HIRAGA, K. (1984). On the mechanism of butylated hydroxytoluene-induced hepatic toxicity in rats. Biothem. Pharmacol. 33,2669-2674. NAKAGAWA, Y., TAYAMA, K., NAKAO, T., AND HIRAGA, K. (1985). Effect of cobaltous chloride on butylated hydroxytoluene-induced hepatic necrosis in rats. Toxicol. Lett. 24,85-89. NEELD, J. B., AND PEARSON, W. N. (1963). Macro- and micromethods for determination of serum vitamin A using trifluoroacetic acid. J. Nutr. 79,454-462. PERAINO, C., FRYE, R. J. M., STAFFELDT, E., AND CHRISTOPHER, J. P. (1977). Enhancing effects of phenobarbitone and butylated hydroxytoluene on 2-acetylaminofluorene-induced hepatic tumorigenesis in the rat. Food Cosmet. Toxicol. 15,93-96. PLAA, G. L., AND HEWITT, W. R. (1982). Evaluation of liver injury. In Principles and Methods of Toxicology (A. W. Hayes, Ed.), pp. 407-445. Raven Press, New York. RUSSELL, R. M., BOYER, J. L., BAGHERI, S. A., AND HRUBAN, Z. (1974). Hepatic injury from chronic hypervitaminosis A resulting in portal hypertension and ascites. N. Engl. J. Med. 291,435-440. SANI, B. P., AND MEEKS, R. G. (1983). Subacute toxicity of all-trans- and 13-cis-isomers of N-ethyl retinamide, N-2-hydroxyethyl retinamide, and N-4-hydroxyphenyl retinamide. Toxicol. Appl. Pharmacol. 70, 228235. STONE, J. P., SHELLABARGER, C. J., AND HOLTZMAN, S. (1984). The long-term inhibition of induced and spontaneous rat breast carcinogenesis by retinyl acetate: Interim results. Proc. Amer. Assoc. Cancer Res. 25, 126. WAKE, K. (1980). Perisinusoidal stellate cells (fat-storing cells. interstitial cells, lipocytes), their related structure in and around the liver sinusoids, and vitamin A-storing cells in extrahepatic organs. Znt. Rev. Cytol. 66, 303-353. WATTENBERG, L. W. (1972). Inhibition of carcinogenic and toxic effects of polycyclic hydrocarbons by phenolit antioxidants and ethoxyquin. J. Natl. Cancer Inst. 48,1425-1430.
Potentiation DAVID Laboratory
(1987)
of Vitamin A Hepatotoxicity L. MCCORMICK,
ofPathophysiology.
Received
THERESA
Life Sciences
November
by Butylated
A. HULTIN,
Research,
AND CAROL J. DETRISAC
IIT Research
10, 1986; acceptedApril
Hydroxytoluene
Institute,
Chicago,
Illinois
60616
14, 1987
Potentiation of Vitamin A Hepatotoxicity by Butylated Hydroxytoluene. MCCORMICK, D. L., HULTIN, T. A., AND DETRISAC, C. J. (1987). Toxicol. Appl. Pharmacol. 90, l-9. The interaction between the natural vitamin A ester retinyl acetate (BA) and the phenolic antioxidant butylated hydroxytoluene (BHT) in the induction of biliary hyperplasia and hepatic fibrosis in female Sprague-Dawley rats was characterized. Using a 3 X 3 matrix design, rats were fed diets supplemented with (per kilogram diet) 0, 125, or 250 mg BA and/or 0, 2500, or 5000 mg BHT. The 125mg dose of RA induced no gross hepatotoxicity, while the 250-mg dose of RA induced a low incidence of hepatic fibrosis in rats examined after 120 and 180 days of exposure. Exposure to BHT alone induced hepatocellular hypertrophy and dose-related increases in liver weight, but no hepatocellular pathology. Simultaneous administration of RA plus BHT resulted in significant increases in the incidence of biliary hyperplasia and hepatic fibrosis compared to that induced by RA alone. BHT reduced total hepatic vitamin A content at all BA dose levels. Thus, mechanisms other than increases in liver vitamin A levels must underlie the potentiation by BHT of RA hepatotoxicity. o 1987 Academic PKSS, hc.
gland (Wattenberg, 1972; McCormick et al., 1984). As seen with RA, BHT inhibits mammary carcinogenesis when administered at dose levels which induce no gross or organspecific toxicity. Chronic exposure to BHT is associated with hepatic enzyme induction and dose-related increases in liver weight (Peraino et al., 1977; Hirose et al., 1981); however, the hepatic effects of BHT are reversible upon cessation of exposure (Botham et al., 1970) and thus are not considered to be of toxicologic significance (Nakagawa et al., 1984). Although both RA and BHT have significant anticarcinogenic activity in the rat mammary gland, their activity as single agents is imperfect; neither RA nor BHT can reduce cancer incidence to zero. One mechanism through which the magnitude of cancer inhibition may be increased is by the use of “combination chemoprevention” protocols, in which multiple inhibitors of carcinogenesis are administered simultaneously or se-
The natural vitamin A ester retinyl acetate (RA) has significant activity as an inhibitor of mammary carcinogenesis in the rat. Previous studies have demonstrated the efficacy of RA in suppressing rat mammary carcinogenesis induced by chemical carcinogens (Moon et al., 1976, 1977; McCormick et al., 1981) and by radiation (Stone et al., 1984). Furthermore, chronic exposure to RA can inhibit the occurrence of “spontaneous” mammary cancers in rats (Stone et al., 1984). In all of these studies, inhibition of mammary carcinogenesis was achieved using RA dose levels which were without significant systemic or organspecific toxicity; however, in several experiments, the liver and bone have been identified as the most sensitive target organs for the toxic effects of RA and other retinoids (Leelaprute et al., 1973; Sani and Meeks, 1983; McCormick et al., 1986). Butylated hydroxytoluene (BHT), a phenolic antioxidant, is also an effective inhibitor of cancer induction in the rat mammary 1
0041-008X/87
$3.00
Copyright 0 1987 by Academic Press, Inc. All rights of reproduction in any form reserved
2
MC CORMICK,
HULTIN,
quentially. We have recently reported that an enhanced inhibition of mammary carcinogenesis can be achieved by combined administration of RA and BHT (McCormick et al., 1986); however, the increased anticarcinogenie activity achieved by chronic exposure to RA and BHT was accompanied by the induction of biliary hyperplasia and hepatic fibrosis. The present study was performed to characterize the hepatotoxic interaction between RA and BHT and to study the temporal, pharmacologic, and histopathologic parameters associated with its induction.
AND DETRISAC
or choke&tic damage to the liver or as a result of hepatic dysfunction (Plaa and Hewitt, 1982). Serum levels of ornithine carbamyl transferase (GCT) were quantitated using the calorimetric method described by Plaa and Hewitt (1982). Levels of sorbitol dehydrogenase (SDH), glutamic-pyruvic transaminase (alanine transaminase; GPT/ALT), glutamic-oxaloacetic transaminase (aspartate transaminase; GGT/AST), lactate dehydrogenase (LDH), alkaline phosphatase (AP), 5’-nucleotidase, yglutamyl transpeptidase (GGT), and cholinesterase were analyzed using a Centrifichem 500 autoanalyzer system (Baker Instruments, Allentown, PA). Statistics. Intergroup comparisons of mean body weight, mean liver weight, hepatic vitamin A concentration, total liver vitamin A content, and serum enzyme levels were performed using analysis of variance. Incidence comparisons were made using x2 analysis.
METHODS RESULTS Experimental animals. Female Sprague-Dawley rats were obtained at 28 days of age from Harlan/SpragueDawley (Indianapolis, IN). Rats were housed in groups of three in polycarbonate cages on hardwood bedding in a windowless room maintained at 22 + 1°C on a 14-hr light/lo-hr dark cycle. Animals were quarantined for 1 week prior to the initiation of the experiment. All rats had free access to diet and drinking water throughout the study. Experimental diets. Basal diet used in the experiment was Wayne Lab Blox MRH 22/5 (Allied Mills, Chicago, IL); this diet contains 8 mg vitamin A/kg, as added retinyl palmitate, and contains no BHT, BHA, ethoxyquin, propyl gallate, or other synthetic antioxidants. RA (in gel beadlet form) was obtained from Hoffmann-LaRoche (Nutley, NJ) and was administered as a dietary supplement as required by the protocol. BHT was purchased from Sigma Chemical Co. (St. Louis, MO) and was also administered as a dietary supplement. Experimental protocol. Using a 3 X 3 matrix design, groups of 25 rats were fed diets supplemented with (per kilogram diet) 0, 125, or 250 mg RA and/or 0,2500, or 5000 mg BHT. At these dietary levels, daily intake of RA is approximately 0, 2, and 4 mg/day for the 0, 125, and 250 RA doses, respectively; BHT intake is approximately 0,40, and 80 mg/day for the 0,2500, and 5000 BHT dose groups. After 30,60,90, 120, and 180 days on diet, 5 rats per group were terminally bled via the inferior vena cava; serum was collected and kept on ice for enzyme assays, as described below. Livers were excised and weighed, and sections were taken and fixed in 10% buffered formalin for analysis via light microscopy. At the 180-day time point, liver samples from each animal were frozen in liquid nitrogen for analysis of vitamin A content using the trifluoroacetic acid method (Neeld and Pearson, 1963). Serum enzymes were selected for assay on the basis of reported alterations in their levels as a result of cytotoxic
Gross observations. The overall health status of rats fed supplemental RA and/or BHT was good throughout the 180-day study period. No mortality or indications of systemic vitamin A toxicity (skin or fur changes, brittle bones) were observed in any animals fed supplemental RA nor was gross toxicity associated with BHT exposure. Terminal body weights ranged from 89 + 3% of control in the group receiving 125 mg RA + 2500 mg BHT to 100 f 4% of control in the group fed 250 mg RA + 2500 mg BHT. Although RA induced no obvious gross toxicity, a low incidence of hepatic fibrosis was observed at necropsy in rats receiving chronic exposure to the retinoid. Dietary administration of BHT had no such effects on liver morphology. However, when administered in combination with RA, BHT resulted in a significant enhancement of RA hepatotoxicity: changes in hepatic morphology in animals exposed to RA + BHT occurred earlier, were seen in higher incidence, and were more severe than were alterations in groups exposed to RA alone. The effects of RA alone on gross hepatic morphology were dose-related. Dietary administration of RA at a level of 125 mg/kg diet resulted in no grossly detectable hepatic pathology during the 180-day study period.
BHT-VITAMIN TABLE
A HEPATOTOXIC
1
IINFLUENCE OF RETINYL ACETATE AND BHT ON INCIDENCEOF GROSS HEPATIC FIBROSIS Days on diet RA dose (mg/kg diet)
BHT dose (mg/kgdiet)
0 0 0
0 2500 5000
0000 0000 0000
125 125 125
0 2500 5000
0000 000 0 0
0
0 215
250 250 250
0 2500 5000
0 0 0
0 0 l/5
l/5 l/5 5156 l/5 5156 5156
30 60
90
120
180 0 0 0
0 0 0
0 315” 4/5b
‘p < 0.05 versus control (0 RA, 0 BHT), appropriate BHT control group (0 RA, same BHT dose), and appropriate RA control (same RA dose, 0 BHT). bp < 0.0 1 versus control (0 RA, 0 BHT), appropriate BHT control group (0 RA, same BHT dose), and appropriate RA control (same RA dose, 0 BHT).
By contrast, administration of RA at 250 mg/ kg diet was associated with alterations in hepatic morphology at later time points (Table 1); one of five rats examined at both the 120and the 180-day time points displayed focal hepatic fibrosis. At 120 days, the only affected animal demonstrated a single focus on one liver lobe. Hepatic involvement at 180 days was more extensive than at 120 days, as multiple foci of fibrosis were diffusely scattered over several liver lobes of the one affected rat. Dietary exposure to RA alone also resulted in a slight degree of hepatomegaly in treated groups. A 6% increase in mean liver weight was observed in groups exposed to the 125mg dose of RA, while the 250-mg dose of RA increased mean liver weight by 12% (data not shown). However, no consistent pattern of statistical significance or trend over time was observed when liver weights in these groups were compared to control. Administration of BHT at levels of 2500 or 5000 mg/kg diet for up to 180 days had no effect on gross hepatic morphology. Consistent with previous reports, BHT administra-
3
INTERACTION
tion was associated with the rapid induction of hepatomegaly; increases in liver weight in groups exposed to BHT were dose-related. After 30 days of exposure, mean liver weights were increased by approximately 30 and 40% from control levels by the 2500 and 5000 mg BHT doses, respectively (p < 0.01); after 30 days, liver weight remained elevated, but no further increases were observed (Fig. 1). Administration of RA had no effect on the induction of hepatomegaly by BHT. In contrast to the apparent lack of effect of RA on BHT induction of hepatomegaly, exposure to BHT did have an enhancing influence on the hepatic toxicity of RA. When administration of the 5000-mg dose of BHT was combined with the 250-mg dose of RA, the incidence of gross hepatic fibrosis was increased from 20 to 100% at 120 and 180 days of exposure. Liver lobes containing more than 10 discrete foci of fibrosis on the organ surface were a common finding at both the 120- and 1 go-day time points in the group exposed to 250 mg RA + 5000 mg BHT. Furthermore, simultaneous exposure to BHT appeared to accelerate the onset of liver damage induced by RA: while fibrosis
15
r
0
I 30
I 60
I 90 Days On Cue?
I 120
I 150
I 180
FIG. 1. InlIuence of BHT dose on liver weight in Sprague-Dawley rats. Each point represents the mean f SD for data from five animals. 0, 0 mg BHT/kg diet (control); A, 2500 mg BHT/kg diet; n , 5000 mg BHT/kg diet. p < 0.0 1 versus control for each data point except 2500 may BHT/ke diet at 120 davs. < I where D < 0.05.
4
MC CORMICK,
HLJLTIN. AND DETRISAC
FIG. 2. Liver section from female Sprague-Dawley rat exposed to 250 mg RA/kg diet for 180 days. Note the numerous enlarged Ito cells (arrows). Hematoxylin and eosin, X 120.
was first observed at 120 days in rats receiving the 250 mg RA supplement alone, such changes were seen at 90 days in the group treated with 250 mg RA + 5000 mg BHT. The effects of the lower dose of BHT were similar, although the increased incidence of hepatic fibrosis in the 250 mg RA + 2500 mg BHT group was statistically significant in comparison to the 250 mg RA group only at the 1SO-day time point. A parallel series of observations was made in groups exposed to the 125-mg dose of RA. Although no gross alterations in hepatic morphology were observed in animals exposed to this dose of RA only, when RA was administered in combination with BHT, morphologic changes were observed. At the 180-day time point, the incidence of hepatic fibrosis in groups exposed to 125 mg RA in combination with either dose of BHT was significantly increased in comparison to the effects of the retinoid alone (Table 1).
Light microscopic observations. Chronic administration of RA to Sprague-Dawley rats induced a characteristic pattern of hepatic alterations visible at the light microscope level. Prominent in the livers of RAtreated animals were large numbers of lipidcontaining perisinusoidal stellate cells (Ito cells; Fig. 2). Ito cells have been proposed to be the primary site of hepatic storage of vitamin A in both rats and humans (Bronfenmajer et al., 1966; Kobayashi et al., 1973; Wake, 1980) and have been implicated in the process of hepatic fibrogenesis (Wake, 1980). Although the number of Ito cells per tissue section was not clearly related to the dose of RA, proliferation of It0 cells in response to hepatotoxic agents has been reported (McGee and Patrick, 1972; Enzan, 1985). The gross lesions noted at 120 and 180 days in groups exposed to 250 mg RA were classified histopathologically as diffuse periportal fibrosis, which was frequently associ-
BHT-VITAMIN
A HEPATOTOXIC
INTERACTION
5
FIG. 3. Liver section from rat exposed to 5000 mg BHT/kg diet for 180 days. Hepatocellular hypertrophy is evident in cells surrounding the central veins (CV). Hematoxylin and eosin, X 120.
ated with bile duct hyperplasia. In all cases, areas of fibrosis observed in animals exposed to 250 mg RA were located at the periphery of the liver. Alterations in liver histology which were observed in rats exposed to dietary BHT were consistent with the gross findings of increased liver weight. BHT-induced hepatomegaly was characterized by hepatocellular hypertrophy which was distributed asymmetrically across the liver lobule. As shown in Fig. 3, hepatocyte size was greatest in regions adjacent to the central vein, while cells in the periportal regions exhibited little or no alteration in cellular dimensions as a result of BHT exposure. Histopathologic alterations observed in livers from rats exposed to RA + BHT consisted of increases in the incidence and severity of hepatic fibrosis compared to that induced by RA only. Although usually located peripherally, focal fibrosis could be demonstrated
throughout the liver parenchyma of rats fed RA + BHT. These lesions were most commonly located in periportal regions of the liver lobule and contained an abnormal number of bile ducts and tissue histiocytes (Fig. 4). Focal portal hepatic necrosis accompanied by an inflammatory infiltrate was also commonly observed. In addition to the increased number and size of fibrotic regions and associated biliary hyperplasia, livers from animals administered RA and BHT in combination also demonstrated enlarged Ito cells and asymmetric hepatocellular hypertrophy, indicative of exposure to RA or BHT alone. Clinical chemistries. Prior to the 180-day time point, neither RA nor BHT, administered alone or in combination, induced any consistent pattern of alterations in levels of serum enzymes. At the 180-day time point, significant increases in serum levels of LDH were observed in groups exposed to 125 mg RA + 5000 mg BHT, 250 mg RA + 2500 mg
6
MC CORMICK,
HULTIN,
AND DETRISAC
FIG. 4. Focus of hepatic fibrosis from rat exposed to 250 mg RA + 5000 mg BHT/kg diet for 180 days. Note abnormal number of bile ducts (arrows). Hematoxylin and eosin, X 120.
BHT, and 250 mg RA + 5000 mg BHT (Table 2). LDH levels in these groups ranged from 400 to 1100% of control values (p < 0.0 1) and were also greater than LDH levels in groups exposed to either RA only or BHT only. Statistically significant, but quantitatively smaller, increases from control levels of AP and GPT/ALT were observed in groups exposed to 250 mg RA in combination with either dose of BHT (Table 2). However, serum activity of these enzymes was not significantly increased from levels observed in the group receiving 250 mg RA alone. In contrast to the observed increases in levels of LDH, AP, and GPT/ALT, administration of RA in combination with BHT induced no statistically significant increases in serum levels of enzymes (OCT, SDH) which are considered to be highly specific indicators of hepatic damage. Conversely, although significant decreases in cholinesterase or increases in 5’-nucleotidase or GOT/AST activity were observed in several groups, these alterations
were not correlated with histologically detectable liver toxicity. Hepatic vitamin A levels. The influence of RA and/or BHT on hepatic levels of vitamin A compounds is presented in Table 3. As expected, administration of 125 or 250 mg RA/ kg diet resulted in significant increases in liver vitamin A stores: vitamin A concentration in these two groups was increased more than lofold in comparison to control. By contrast, administration of BHT reduced hepatic vitamin A concentration at all RA doses used. In animals receiving no RA supplementation, liver vitamin A concentration was reduced by approximately 50% in animals exposed to BHT. In groups receiving 125 or 250 mg RA/kg diet, the 2500-mg dose of BHT reduced hepatic vitamin A concentration by 17 to 19% from control levels, while the 5000-mg dose of BHT resulted in a 46 to 48% reduction. The decreases in hepatic vitamin A concentration associated with BHT exposure cannot be explained solely on the basis of the
BHT-VITAMIN
A HEPATOTOXIC TABLE
ALTERATIONS
RA dose (mg/kg diet)
IN SERUM ENZYME
BHT (m&kg
dose diet)
AP 19 9 27
137+ 114+ 1172
GPT/ALT
0 0 0
0 2500 5000
125 125 125
0 2500 5000
49% 17 93+ 9 272 + 127c,e
193+ 30 1482 25 197 xi 145
250 250 250
0 2500 5000
lOOk 35 598+217”’ 213+204b
217+ 235-t 214+
a All bp < ‘p < dp < ‘p i
2
LEVEU IN SPRAGUE-DAWLEY RATS RECEIVING DIETARY RETINYL ACETATE AND/OR BHT”
LDH 54* 31-c 70+
7
INTERACTION
28 29 27
44’ 63’ 32”
43-c 40+ 43+
GOT/AST
12 5 8
180 DAYS EXPOSURE
S-Nucleotidase
TO
Cholinesterase
74f 12 48 f 4’ 45 + 4’
14* 12* 12+
2 1 1
31102568 2580 + 388’ 2546 f 291b
67 & 15b 46+ 10 55 f 12
81 f 17 54 f 7b.d 66 2 sd
17+ 2 15* I 31*296
2582 + 282’ 2783 -c 199 2804 3~ 382
75 f 24’ 79 + 18’ 83+21’
87& 8 90 f 126 75*22
18? 2 35 + 16’ 25+ 5
2574 + 459’ 2588 + 156b 2279 zk 586’
values are expressed as milliunits per milliliter serum. 0.05 versus control (0 RA, 0 BHT). 0.0 1 versus control (0 RA, 0 BHT). 0.05 versus appropriate RA control (same dose RA, 0 BHT). 0.01 versus appropriate RA control (same dose RA, 0 BHT).
increases in liver weight induced by BHT; when total liver vitamin A content, rather than concentration of vitamin A per gram of tissue, is considered, the reduction in vitamin A stores as a function of BHT exposure persists in five of six BHT dose groups, although the magnitude of the reduction is decreased (Table 3).
(GPT and GOT) as a result of BHT administration. In the present experiment, neither hepatocellular toxicity nor elevation of serum transaminases was observed in groups exposed to BHT only. It should be noted, how-
TABLE
3
INFLUENCE OF RETINYL ACETATE AND BHT ON LIVER VITAMIN A CONTENT
DISCUSSION
Liver
The dose-response and time-response data obtained in the present experiment suggest that the increased incidence of hepatic damage observed in animals exposed to RA in combination with BHT reflects a potentiation by BHT of the hepatotoxic activity of RA. Although BHT administration did result in the induction of hepatomegaly, it appears that this compound has essentially no other effects on hepatic morphology at the dose levels used in the present study. The induction of hepatic necrosis in rats by BHT exposure has been reported by Nakagawa et al. (1984, 1985); these authors also reported significant increases in serum levels of transaminases
Ra dose (mgjkg
diet)
BHT dose (mg/kg
diet)
vitamin A concentration
Total liver vitamin A content
(pm00
(rmok)
0 0 0
0 2500 5ooo
1.9 kO.1 I.0 f 0.26 0.9 + 0.2b
125
0 2500 5ooo
21.7 + 1.9” 17.5 * 2.1” 11.8 +- 0.7”,b
214.8 + 17.9’ 185.8 f 22.3” 156.9 f 9.3-
0 2500
25.4 f 0.5” 21.1 f 1.2” 13.2 f 2.0”,b
238.8 272.2 171.6
125
125 250 250 250
5000
15.4* 10.8? 10.8 2
0.9 2.2c 4.8’
k 4.7” iz 15.5” +-26P’
’ p < 0.0 1 versus control (0 RA, 0 BHT). bp < 0.01 versus appropriate FU control group (same RA dose, 0 BHT). ‘p < 0.05 versus appropriate RA control group (same RA dose. 0 BHT).
8
MC CORMICK,
HULTIN,
ever, that the doses of BHT found to be necrogenic via gavage administration in the studies of Nakagawa and colleagues ( 1984, 1985) were more than threefold higher than the highest dietary exposure level used in the present study. The lack of hepatic toxicity of dietary BHT observed in the present experiment is in agreement with the data of Hirose et al. ( 198 1), who conducted a 2-year feeding study in rats using BHT doses of 2500 and 10,000 mg/kg diet. By contrast to the apparent lack of hepatotoxic effects associated with exposure to BHT as a single agent, a low incidence of hepatotoxicity was observed in rats exposed to RA alone; the histologic features of liver damage induced by RA in rats are similar to those reported in cases of hypervitaminosis A in humans (Russell et al., 1974; Jacques et al., 1979). It appears that the hepatotoxic activity of the doses of RA used in the present study is near threshold, since gross liver damage was observed in only one of five rats receiving either 120 or 180 days of exposure to the 250 mg/kg diet RA supplement. This incidence of diffuse hepatic fibrosis at 6 months is consistent with our previous report (McCormick et al., 1986). In quantitative terms based on data obtained for high dose groups at the 180-day time point, the interaction between BHT and RA in the induction of liver toxicity can be expressed as 0 + 1 = 5; while liver damage was seen in O/5 rats exposed to the 5000 mg/ kg diet supplement of BHT, and gross hepatic fibrosis was seen in only l/5 rats exposed to RA at a level of 250 mg/kg diet, administration of these doses of RA and BHT in combination induced grossly detectable liver damage in 5/5 animals. The observations that BHT has no apparent hepatic toxicity as a single agent, yet can increase the magnitude of damage induced by RA, conform exactly to the definition of toxicologic potentiation. Similar quantitative comparisons can be made among other dose groups (see Table 1). Because these data indicate that the observed liver toxicity is a potentiation by BHT
AND DETRISAC
of a toxic manifestation of a retinoid, one possible mechanism for this effect could lie in alterations by BHT of vitamin A accumulation in the liver. Should exposure to BHT result in increases in hepatic vitamin A concentration, the functional result of the BHT-RA interaction would be to increase the dose of a toxic agent at its site of action. In order to address this hypothesis, the influence of BHT on hepatic vitamin A concentration was determined. As indicated in Table 3, exposure to BHT decreased, rather than increased, hepatic vitamin A levels at all RA doses used. On this basis, the hypothesis that BHT potentiates retinoid hepatotoxicity through alterations in retinoid metabolism, resulting in increased liver levels of vitamin A, cannot be supported. Thus, the mechanism by which BHT acts to increase RA toxicity remains unknown. It appears, however, that BHT enhancement of RA liver toxicity occurs via potentiation of an as yet undefined biological activity of the retinoid, rather than through a pharmacologic mechanism resulting in increased target organ concentrations of the toxic agent. This increased biological activity is particularly interesting in that the potentiation of RA hepatotoxicity occurs in the presence of decreased, rather than increased, vitamin A levels in the liver. ACKNOWLEDGMENTS This work was supported by Contract NO l-CP-4 1063 from the National Cancer Institute, DHHS. We acknowledge the excellent technical assistance of Mary Ann Cahill, Mark S. Filla, Carrie Martinez, and Steven J. Moore. Patricia Moser provided expert secretarial assistance.
REFERENCES BOTHAM, C. M., CONNING, D. M., HAYES, J., LITCHFIELD, M. H., AND MCELLIG~TT, T. F. (1970). Effects
of butylated hydroxytoluene on the enzyme activity and ultrastructure of rat hepatocytes. Food Cosmet. Toxicol. 8, 1-8.
BHT-VITAMIN
A HEPATOTOXIC
BRONFENMAJER, S., SCHAFFNER, F., AND POPPER, H. (1966). Fat-storing cells (lipocytes) in human liver. Arch. Pathol. 82,447-453. ENZAN, H. (1985). Proliferation of Ito cells (fat-storing cells) in acute carbon tetrachloride liver injury. Acta Pathol. Japan. 35, 1301-1308. HIROSE,
M.,
SHIBATA,
M.,
HAGIWARA,
A., IMAIDA,
K.,
AND ITO, N. (198 1). Chronic toxicity of butylated hydroxytoluene in Wistar rats. Food Cosmet. Toxicol. 19,147-151. JACQUES, E. A., BUSCHMANN, R. J., AND LAYDEN, T. J. ( 1979). The histopathologic progression of vitamin Ainduced hepatic injury. Gastroenterology 76,599-602. KOBAYASHI, K., TAKAHASHI, Y., AND SHIBASAKI, S. (1973). Cytological studies of fat-storing cells in the liver ofrats given large doses of vitamin A. Nature New Biol. 243, 186-l 89. LEELAPRUTE, V., BOONPUCKNAVIG, V., BHAMARAPRAVATI, N., AND WEERAPRADIST, W. (1973). Hypervitaminosis A in rats. Arch. Pathol. 96,5-9. MCCORMICK, D. L., BURNS, F. J., AND ALBERT, R. E. (198 1). Inhibition of benzo[a]pyrene-induced mammary carcinogenesis by retinyl acetate. J. Natl. Cancer Inst. 66,559-564. MCCORMICK, D. L., MAJOR, N. M., AND MOON, R. C. (1984). Inhibition of 7,12-dimethylbenz[a]anthracene-induced rat mammary carcinogenesis by concomitant or post-carcinogen antioxidant exposure. Cancer Res. 44,2858-2863. MCCORMICK, D. L., MAY, C. M., THOMAS, C. F., AND DETRISAC, C. J. (1986). Anticarcinogenic and hepatotoxic interactions between retinyl acetate and butylated hydroxytoluene in rats. Cancer Res. 46, 52645269. MCGEE. J. O., AND PATRICK, R. S. (1972). The role of p&sinusoidal cells in hepatic fibrogenesis. Lab. Invest. 26,429-440. MOON, R. C., GRUBBS, C. J., AND SPORN, M. B. (1976). Inhibition of 7,12 - dimethylbenz[a]anthracene - induced mammary carcinogenesis by retinyl acetate. Cancer Res. 36,2626-2630. MOON, R. C.. GRUBBS, C. J., SFQRN, M. B., AND GOODMAN, D. G. (1977). Retinyl acetate inhibits mammary
INTERACTION
9
carcinogenesis induced by N-methyl-N-nitrosourea. Nature (London) 267,620-62 1. NAKAGAWA, Y., TAYAMA, K., NAKAO, T., AND HIRAGA, K. (1984). On the mechanism of butylated hydroxytoluene-induced hepatic toxicity in rats. Biothem. Pharmacol. 33,2669-2674. NAKAGAWA, Y., TAYAMA, K., NAKAO, T., AND HIRAGA, K. (1985). Effect of cobaltous chloride on butylated hydroxytoluene-induced hepatic necrosis in rats. Toxicol. Lett. 24,85-89. NEELD, J. B., AND PEARSON, W. N. (1963). Macro- and micromethods for determination of serum vitamin A using trifluoroacetic acid. J. Nutr. 79,454-462. PERAINO, C., FRYE, R. J. M., STAFFELDT, E., AND CHRISTOPHER, J. P. (1977). Enhancing effects of phenobarbitone and butylated hydroxytoluene on 2-acetylaminofluorene-induced hepatic tumorigenesis in the rat. Food Cosmet. Toxicol. 15,93-96. PLAA, G. L., AND HEWITT, W. R. (1982). Evaluation of liver injury. In Principles and Methods of Toxicology (A. W. Hayes, Ed.), pp. 407-445. Raven Press, New York. RUSSELL, R. M., BOYER, J. L., BAGHERI, S. A., AND HRUBAN, Z. (1974). Hepatic injury from chronic hypervitaminosis A resulting in portal hypertension and ascites. N. Engl. J. Med. 291,435-440. SANI, B. P., AND MEEKS, R. G. (1983). Subacute toxicity of all-trans- and 13-cis-isomers of N-ethyl retinamide, N-2-hydroxyethyl retinamide, and N-4-hydroxyphenyl retinamide. Toxicol. Appl. Pharmacol. 70, 228235. STONE, J. P., SHELLABARGER, C. J., AND HOLTZMAN, S. (1984). The long-term inhibition of induced and spontaneous rat breast carcinogenesis by retinyl acetate: Interim results. Proc. Amer. Assoc. Cancer Res. 25, 126. WAKE, K. (1980). Perisinusoidal stellate cells (fat-storing cells. interstitial cells, lipocytes), their related structure in and around the liver sinusoids, and vitamin A-storing cells in extrahepatic organs. Znt. Rev. Cytol. 66, 303-353. WATTENBERG, L. W. (1972). Inhibition of carcinogenic and toxic effects of polycyclic hydrocarbons by phenolit antioxidants and ethoxyquin. J. Natl. Cancer Inst. 48,1425-1430.