Early Liver Cell Lesions in Rats Induced by Thioacetamide An Ultrastructural, Cytophotometric and Autoradiographic Study

Path. Res. Pract. 184, 69-76 (1989)

Early Liver Cell Lesions in Rats Induced by Thioacetamide An Ultrastructural, Cytophotometric and Autoradiographic Study Praet Marleen and Roels Hendrik N. Goormaghtigh Institute of Pathology, University Hospital Ghent, Belgium

P. van Oostveldt Department of Biochemistry, State University of Ghent, Belgium

SUMMARY A morphological, cytophotometrical and autoradiographical study was carried out on the early liver cell lesions present after one month of thioacetamide (TAA) exposure. We wanted to determine the extent of cell damage in the hepatocytes in relation to the later occurring cholangiocarcinoma. Cytoplasmic and nuclear alterations were present in the hepatocytes. They were mainly due to the toxic influence of the metabolites of TAA. No Feulgen-DNA changes have been observed in the hepatocytes in any of the studied zones. The increased TdR H3 + uptake and mitotic activity in the periportal and midzonal areas represented regenerative activity. In addition to these hepatocytic changes, the centrolobular area was infiltrated by oval cells. These cells appeared at first singly, later on they formed clusters. The electron microscopy of these cells revealed phenotypic characteristics, which were different from the hepatocytes. When separately, these cells resembled undifferentiated cells with cytoplasmic extensions and absent basement membrane. When arranged in clusters, a definite canalicular arrangement was present with characteristics of bile canalicular cells with microvillous extensions at the apical border of the cytoplasm and the presence of a basement membrane. A transition from the first oval cell type to the second oval cell type was suggested. This transition might represent a differentiation process of cells, which are regarded as the target cell and the precursor cell in the development of the cholangiocarcinoma. This is the first study reporting oval cell proliferation in the centrolobular area in a multistep model of livercarcinogenesis in rats. Introduction

fied polar metabolites by oxydation38 • Several authors 26• have investigated the early changes after TAA administration. Cytoplasmi27, 53, nuclear and nucleolar changes28 , 52 have been reported. In an earlier study we have shown that the development of cholangiocarcinoma after TAA treatment also follows a well defined multistep sequence39 • The present study deals with the early changes present after one month of TAA administration. Light microscopic and ultratructural studies of the early lesions in hepatocytes and of the oval cells

27, 52, 53

Thioacetamide is a thionol-S-compound which induces liver tumors in rats and mice in conditions of prolonged administration 15 , 16. Most authors report the development of liver cell adenomas or hepatocarcinomas5, 15. Others describe also the occurrence of cholangiomas and cholangiocarcinomas 1, 39. Thioacetamide (TAA) is mainly metabolized in the liver43 with the concurrent formation of thioacetamide-S-oxide (TASO), acetamide and non-identi© 1989 by Gustav Fischer Verlag, Stuttgart

0344-0338/89/0184-0069$3.5010

70 . M. Praet, H. Roels and P. van Oostveldt

are completed by cytophotometrical and autoradiographical investigations. Our main purpose is the evaluation of the observed changes in relation to the development of the tumou~9.

Material and Methods Fifty, 3 month old male albino Lewis rats, weighing 250 grams, were on a semi-synsthetic, low protein diet consisting of casein 24%, dextrose 65%, salts 4%, corn oil 5% and halibut liver oil 0.03% to which vitamin B was added 55 • This diet is low in protein content but adequate for growth. Twenty-five experimental animals got a dietary concentration of 0.5 mg % TAA for one month, resulting in an approximate intake of 5.14 mg TAA per day. Twenty-five control animals got the diet and water ad libitum only. After one month the rats were killed simultaneously by cervical dislocation between 3.45 p.m. and 4.15 p.m. Liver fragments were studied by light microscopy, electron microscopy, cytophotometry and autoradiography in the different zones of the liver lobule according to the concept of Rappaport42 •

a) Light microscopy Liver fragments taken from the right and left lobe in 10 experimental and 10 control animals were fixed in AFA (ethylalcohol 75%, formalin 20%, acetic acid 5%) and processed. Five micron paraffin sections were cut and stained by H.E., Masson's Trichrome, PAS, Perl's reaction for iron and Giemsa.

An SP 560, LP 530 and a BG 36 filter were used in combination with a dichroic mirror with 50% reflection at 560 nm for excitation. An LP 580 and a RG 2 (4 mm) were used as barrier filters. During the measuring time of 1 second a sample of 200 readings was taken from the photomultiplier by the AD converter connected to the PDP 11/03 Digital microcomputer and the mean reading stored on a floppy disc. Standardization of the fluorimeter was carried out with uranylglass and the DNA content was calculated as a relative amount and expressed in arbitrary units. The measured data were analysed in a program, ranking the data in ascending order and after conversion, a histogram was plotted and the following statistical parameters were calculated: mean, standard deviation and standard error of the mean. During the statistical evaluation of the measured data the rejection level was fixed at p < 0.01.

d) Autoradiography Two groups, each consisting of 5 control animals and 5 experimental animals received an intraperitoneal injection of 100 microCie HHthymidine/lOO gr body weight (S.A. 20-30 Ci/ mmol, Amersham), respectively 6 hours and 30 minutes before they were killed. All animals were killed between 3.45 pm and 4.15 pm. Small liver fragments, taken from the right lobe were fixed in AFA. Five micron paraffin sections were made. They were covered with an Ilford L4 photographic emulsion by means

b) Electron microscopy A small liver fragment was immediately immersed in a buffered paraformaldehyde-glutaraldehyde solution for 4 hours at 4°C (2% glutaraldehyde + 1.2% formaldehyde 8% + 50 mg CaChin a cacodylate buffer) for routine electron microscopy. The small tissue blocs were rinsed overnight in a 0.1 M phosphate buffer with 6% sucrose. Postfixation was carried out with osmium tetroxide. The material was embedded in epon 812. Sections were made with an LKB ultratome I using glass knives, put on a bare 300 mesh grid and covered with a carbon film. They were stained by the Reynolds' method and alcoholic uranyl acetate. The sections were examined with a Siemens Elmiskop I electron microscope, at 80 KV.

c) Cytophotometry Small liver fragments were taken from the right lobe of the liver of 5 control and 5 experimental animals. After simultaneous fixation in AFA and processing, 12 micron paraffin sections were put on the same slide in order to standardize staining conditions. The slides were stained by the pararosaniline Feulgen method: 12 min hydrolysis in 1N HCl., pararosaniline 0.01 % pH 17• The nuclear DNA content of 100 hepatocytes was assessed in each zone of the liver lobule. The nuclear DNA content of 20 oval cells was assessed in zone III of the experimental livers as this is the only area where these cells are present after one month of TAA treatment. The fluorescence intensity of pararosaniline-Feulgen stained sections was measured with the MPV II cytofluorometer (Leitz). Photofading during measurements was reduced to a minimum by focusing the nucleus in the measuring diaphragm with phasecontrast using an orange light filter (OG2 4 mm). The excitation by light from a mercury high pressure lamp (HBO 100 W) was reduced to 1 second with electronically guided shutters.

Fig. 1. In the centrolobular area a marked depletion of glycogen is observed. The periportal area is not altered. (PAS, original magnification x 160)

Thioacetamide Induced Liver Carcinogenesis in Rats . 71 of the dipping method, exposed during 6 weeks at - 20°C, developed and stained with hematoxylin-eosin (31). In sections with background counts of less than 5 grains, cells with 10 grains and more were scored as labeled. In the zones I, II, III random fields were selected and for each zone 1000-3000 nuclei were counted. The numbers of the labeled and unlabeled interphase nuclei in hepatocytes and in the oval cells were counted separately and the results expressed as a percentage of the totals of labeled and unlabeled nuclei of each cell type. We also counted the number of labeled and unlabeled mitoses.

Results

a) Light microscopy The lobular arrangement of the liver was preserved. The main alterations were localised in zone III. The hepatocytes showed severe cytoplasmic and nuclear alterations with an infiltration by microvesicular lipid droplets and a severe depletion of the glycogen content (Figure 1). An increased number of binucleate hepatocytes was observed. Several necrotic hepatocytes were noticed (Figure 2). The nuclei were swollen and pale, the nucleoli were huge. They were surrounded by macrophages. Only a few mitoses were present: 0.35%0.

Fig. 2. The hepatocytes show huge nucleoli in the centrolobular area. Some necrotic liver cells (arrows) are present. The oval cells proliferate and form clusters (arrowheads) between the hepatocytes. (Masson's Trichrome, original magnification x 400)

The zones I and II did not show cytoplasmic nor nuclear alterations. A few portal tracts revealed a slight infiltration by lymphocytes and plasmocytes. The hepatocytes showed an increased number of mitoses: respectively 6.25%0 and 4.23%0. The oval cell proliferation was exclusively present in the zone III either as single cells or in small clusters. These cells were not found in the zone I and II. The oval cells resembled undifferentiated cells with a scant amount of cytoplasm and oval-shaped blunt ending basophilic nuclei. Characteristically, the oval cells had a high nucleus/cytoplasmic ratio. They were actively dividing as is shown by their mitotic index of 8.720/00. It was difficult to differentiate these cells from fibroblasts on light-microscopy. An ultrastructural study was carried out.

b) Electron microscopy The hepatocytes of the zone III showed severe cytoplasmatic changes (Figure 3) characterized by the accumulation of lipid droplets, the distension of the endoplasmic reticulum and the displacement of the lysosomes. The mitochondria displayed a great diversity of shape and size. The heterochromatin was condensed along the nuclear membrane, the nucleoli were enlarged. The hepatocytes of the zone I and II were not altered by the treatment.

Fig. 3. Severe cytoplasmic alterations are present in the hepatocytes around the central vein. Displacement of the lysosomes and a microvesicular lipid infiltration (arrow) are displayed. (E M, uranyl acetate x 4000)

72 . M. Praet, H. Roels and P. van Oostveldt

Fig. 4..One single oval cell (*) and 2 clusters of oval cells (U) in a canalicular arrangement infiltrate between the hepatocytes of the centrolobular area. Lipid droplets are present in the cytoplasm of the hepatocytes. (E M, uranyl acetate x 3000) The oval cells were exclusively located in the zone III. They appeared either as single cells or in clusters (Figure ~). When ~hese. cells appeared separately, they showed Irregular mIcroVIllous cytoplasmic extensions. A basement membrane was not present. These cells cannot be confused with fat-storing cells as they lack the characteristic fat droplets (Figure 5). When arranged in clusters, a distinctly ductular configuration cou.ld be observed, surrounded by a basement membrane (FIgure 6). Numerous microvilli were present along the apical side of the ductular cells. The ~ohesion of a~jacent ductular cells by lateral interdigitatmg ~ytoplasmlC processes and the tight junctions closing the mtercellular space towards the bile canalicular lumen w~re well d~tect~ble. The cytoplasm contained many large mItochondna, nbosomes and tonofilaments. The nuclei were very irregular with coarse chromatin, condensed along the nuclear membrane. The nucleolus was distinct but small.

c) Cytophotometry

Hepatoeytes (Table 1) The results were different according to the area: - zone III .The nU!llber of diploid cells seemed to increase only shg.htly.. Smc.e the treatment induced swelling of the nuclei maI~ly m thIS area and because we had carefully avoided sectioned nucleI, we might have underestimated the number of cells with larger volume such as the tetraploid nuclei.

Fig. 5.. One si~gle oval cell shows irregular microvillous cytoplasmIC extenSIOns. No basement membrane is present. (E M, Uranyl acetate x 8000) The Feulgen-DNA content in this area showed no significant alterations. - zone II A marked increase of the number of diploid cells was observed, together with an equal decrease of the number of tetraploid cells. No significant changes in the FeulgenDNA content between the control and the experimental animals were observed. - zone I A slight increase of the number of diploid cells was observed, together with a decrease in equal proportion of the number of tetraploid cells. No significant Feulgen-DNA changes were observed between the control and the experimental group.

Oval cells (Table 2) These cells could only be observed in the experimental animals and were located in zone III. The Feulgen-DNA content of the oval cells was usually diploid. Tetraploid values were rare.

d) Autoradiography (Table 3) Hepatocytes

- Thirty minutes after the administration of TdR-HH the control animals demonstrated an almost identicai

Thioacetamide Induced Liver Carcinogenesis in Rats . 73 4.23%0. In zone III a mitotic index of 0.35%0 was observed. By this time most of the mitotic figures were labeled. The labeling index strongly increased in zone I and zone II in contrast with the moderate uptake of TdR-W+ in zone III: in zone I a labeling index of 24.24%0 and in zone II a labeling index of 37.85%0 was counted with a slight increase to 10.9%0 in zone III.

Oval cells Thirty minutes after the administration of TdR-HH , a mitotic index of 0.97%0 and a labeling index of 7.33%0 were observed. After 6 hours the mitotic index inreased to 8.72%0 and the labeling index to 17.98%0. Discussion

Fig. 6. A cluster of 3 oval cells (*) is arranged in a canalicular configuration surrounded by a basement membrane. A few irregular mitochondria (arrow), ribosomes (small arrowheads) and tonofilaments (small arrow) can be distinguished. Between the oval cells tight junctions (double small arrows) are present. (E M, uranyl acetate x 8000) mitotic rate in the different zones of the liver. The labeling index in the zone III was very low, whereas a value of 1.060/00 and 2.50/00 was found respectively in the zone I and II. Since oval cells were absent in control conditions, a mitotic and labeling index could not be determined. The experimental animals showed an increased mitotic index of 2.92%0 in zone I and 7.920/00 in zone II in the liver cells and no changes in zone III. All mitoses were unlabeled. The labeling index of the interphase nuclei increased in all areas: in zone I a labeling index of 2.74%0, in zone II a labeling index of 10.16%0 and in zone III a labeling index of 2.560/00 was found. - Six hours after the administration of TdR-HH , the mitotic index in the experimental group was markedly increased in zone II and zone I to respectively 6.25%0 and

Our light microscoic observations fit in with the earlier findings described in studies dealing with the early effects of TAA administration 4, 20, 53. The predilection of the lesions for the centrolobular area can be explained by the heterogeneous structure of the liver lobule, with focal and zonal variations of enzyme activities as shown by Novikoff and Essner32 • Consequently, as demonstrated in numerous carcinogenic experiments, a different histologic pattern can be expected dependinr. on the target site of the carcinogen in the liver lobule' II, 14. The alterations that were described and which we also observed, are the fatty infiltration, the glycogen depletion and the liver cell necrosis. These changes are by no means specific for TAA and are also observed in conditions of low-protein diet, CCl4 and chronic passive congestion. The preferential localisation of the lesions in the centrolobular area is closely related to the oxidation of TAA and TASO, which needs the bioactivation by the mixedfunction monooxygenase system: the cytochrome P450 and the FAD containing monooxygenases 9, 24. Our findings support the opinion that cytoplasmic changes of the central toxic pattern are due to a "concentration action" of the carcinogen2, 3. A similar centrolobular lesion has been described in livers intoxicated with p-dimethylaminobenzene (butter yellow)36, 3-Me-4-diMe-aminobenzene8, DL-ethionine 13 , CcV9 , diethylnitrosaminel 9 • Our cytophotometrical results also confirm the heterogeneous reaction of the different areas of the liver lobule: there is a marked increase in diploid cells in the midzonal and periportal areas, while the centrolobular

Table 1. Cytophotometric Results in Hepatocytes in Control and Experimental Livers after One Month of TAA Application Area

Ploidy

Number of cells Control rats

Number of cells Exp. rats

F-DNA content Control rats

F-DNA content Exp. rats

P

Zone III

2n 4n 2n 4n 2n 4n

21 65 30 62 62 32

25 57 60 26 80 9

106.58 180.37 93.95 172.41 86.51 179

123.77 191.33 107.56 190.59 102.66 191.22

0.35 0.19 0.47 0.28 0.30 0.47

Zone II Zone I

74 . M. Praet, H. Roels and P. van Oostveldt Table 2. Cytophotometric Results in Oval Cell After One Month of TAA Application Ploidy

Number of cells Exp. rats

F-DNA content Exp. rats

2n 4n

19

106.34

area seems less affected. The increase of diploid cells in the periportal area is in accordance with the hypothesis of the growth zone in this area 12, 17. The increased cell degeneration in the centrolobular area induces a regenerative reaction which starts at the periportal area and results ultimately in cell replacement around the central vein. We did not observe significant changes of the Feulgen-DNA content of the nuclei in any of the zones of the lobule. This is in keeping with earlier observations40 • The changes described by L. Thienpont52 therefore seem to be transient and not indicative of a possible stable preneoplastic change. The observations in control animals after 30 minutes of exposure of TdR H3+ demonstrate the periportal growth zone, which is confirmed by the findings of 6 hours oftdR H3+ exposure revealing an increased uptake of TdR-H3+ in the periportal and midzonal area together with an increased mitotic index. These observations also fit in with the hypothesis of a periportal growth zone. From these observations we may conclude that the changes in the liver cells are epiphenomena of an aspecific drug-induced cytotoxic reaction and not preneoplastic. The necrosis, however, may be important and even crucial for the development of later lesions as is shown in other experiments5 • Since necrosis of the liver cells always

precedes- the proliferation of the oval cells, both phenomena seem to be closely related and even suggest that the necrosis triggers the proliferation of the oval cells. Indeed the administration of a tenfold lower concentration of thioacetamide induces fatty infiltration only; without cell necrosis, no oval cell proliferation nor tumor development is observed (unpublished results). These observations stress the importance of cell necrosis for the development of the tumour. The oval cells at first proliferate discretely and separately. After two weeks they form small clusters in the centrolobular region. This location is exceptional and unusual. Most authors, who studied the effect of other carcinogens6, 13,25,29,33,48, observed that the proliferation of oval cells occurred in the periportal area. The origin and the fate of the oval cells have been the subject of numerous speculations. Farber 14 suggested that these cells ori~inated from the Hering canal. Leduc30 and Popper et al. 7 suggested a hepatocytic origin. Other authors 10, 34, 46 put forward the theory of a stemcell located in the portal zone. Our ultrastructural findings are in agreement with the suggestion of a stemcell: indeed the undifferentiated nature of the single oval cells lacking hepatocytic and bile canalicular characteristics are in favour of this hypothesis. Later when clusters become apparent, another differentiation is present. The ductular arrangement of these cells, surrounded by a basement membrane, represent advanced differentiation characteristics. The presence of tonofilaments, tight junctions point in the same direction. A transition from the undifferentiated type of the ocal cell to the bile canalicular type is suggested. Our results demonstrate the heterogeneity within the oval cell population. Similar observations have been made by immunocytochemical studies23 , 45, 48 and by determination of isozyme profiles22, 56

Table 3. Autoradiographic Results in Control and Experimental Animals After One Month of TAA application Area

Celltype

Zone III

Hepatocytes

Administration time

Index %0

30 min

Control TAA Control TAA Control TAA

6H. Oval Cells

30 min 6H.

Zone II

Hepatocytes

30 min 6H.

Zone I

Hepatocytes

30 min 6H.

Control TAA Control TAA Control TAA Control TAA Control TAA

Mitoses n labeled

0.38 0.37 0.22 0.35 0 0.97 0 8.72 0.36 7.92 1.41 6.25 0.35 2.92 1.35 4.23

n not labeled

Interphase Index %0 2 2

1 1

1 1

5

6 31 2 5 18

4 2 46 4 2 16 1 5

0.19 2.56 3.14 10.9 0 7.33 0 17.98 2.5 10.16 9.56 37.85 1.06 2.74 7.20 24.24

Thioacetamide Induced Liver Carcinogenesis in Rats . 75

We presume that the oval cell tyre reEresents a stemcell, as is shown by other experiments2 ,34, 7. The explanation for a bile canalicular differentiation of the oval cells in this experiment is not clear. As is reported in the literature, the ultimate fate of the oval cells may vary. Some authors report on the differentiation towards hepatocytes 29, 33, 35, 54, 58, others believe these cells undergo atrophy and disappear l8 , 51. In our experiments an evolution to bile canalicular cells is observed41 • The cytophotometric measurements reveal a predominantly diploid population of cells with a high labeling and mitotic index. These findings aAree with those observations made by Stocker50 , Rubin and Tsao et a1. 54 • Since the oval cells actively proliferate in this experiment, they should be considered as the target cells which eventually become preneoplastic under continuous TAA administration. Their ultrastructural arrangement, the evolution to cholangiomas and cholangiocarcinomas in long term experiments39 , their high rroliferation rate, their different chromatin composition4 strongly supports this hypothesis. Further immunohistochemical work is, however, needed to trace these cells at the early moments of their existence in the liver lobule and during the further developments of preneoplasia and neoplasia.

References 1 Anghileri LJ, Heidbreder M, Weiler G, Dermietzel P (1977) Hepatocarcinogenesis by thioacetamide: correlation of histological and biochemical changes and possible role of cell injury. Exp Cell Bioi 45: 34-47 2 Bannasch P, Muller HA (1964) Lichtmikroskopische Untersuchungen uber die Wirkung von N-Nitrosomorpholin auf die Leber von Ratte und Maus. Arzneimittel-Forsch 14: 805-814 3 Bannasch P (1967) Nitrosamin-induzierte Glykogenose und Geschwulstbildung in der Rattenleber. Verh Dtsch Ges Path 51: 343-349 4 Bannasch P (1968) The cytoplasm of hepatocytes during carcinogenesis. Recent Results in Cancer Res 19: 66 5 Bannasch P, Hesse I, Angerer H (1974) Hepatozelluliire Glykogenose und die Genese sogenannter hyperplastischer Knoten in der Thiocetamid-vergifteten Rattenleber. Virchows Arch (Cell Pathol) 17: 29-50 6 Bannasch P, Massner B (1976) Histogenese und Cytogenese von Cholangiofibromen und Cholangiocarcinomen bei Nitrosomorphilin-vergifteten Ratten. Z Krebsforsch 87: 239-255 7 B6hm N, Sprenger E (1968) Fluorescence cytophotometry: a valuable method for the quantitative determination of nuclear Feulgen-DNA. Histochemie 16: 100-111 8 Chang JP, Spain JD, Griffin AC (1958) Histochemical manifestation of early changes in rat liver during carcinogenesis induced by 3-Me-4-dimethylaminoazobenzene. Cancer Res 18: 670-674 9 Chieli E, Malvaldi G (1984) Role of the microsomal FADcontaining monooxygenases in the liver toxicity of TASO. Toxicology 31: 41-52 10 Dempo K, Chisaka N, Yoshida Y, Kaneko A, Onoe T (1975) Immunofluorescent study on alpha-fetoprotein-producing cells in the early stage of 3-Me-4-diMEaminobenzene carcinogenesis. Cancer Res 35: 1282-1287 11 Emmelot P, Benedetti EL (1960) Changes in the fine struc-

ture of rat liver cells brought about by dimethylnitrosamine. J Biophys Biochem Cytol 7: 393-396 12 Fabrikant JI (1968) The kinetics of cellular proliferation in re~enerating liver. J Cell Bioi 36: 551-565 3 Farber E (1956) Similarities in the sequence of early histological changes induced in the liver of the rat by etionine, 2-acetylaminofluorene and 3'Me-4-dimethylaminoazobenzene. Cancer Res 16: 142-149 14 Farber E (1963) Ethionine Carcinogenesis. Advances in Cancer Res 7: 383-473 15 Fitzhugh OG, Nielson AA (1948) Liver tumors in rats fed thiourea or thioacetamide. Science 108: 626-628 16 Gothoskar SV, Talwalkar GV, Bhide SV (1970) The tumorigenic effect of thioacetamide in Swiss strain mice. Brit J Cancer 24: 498-503 17 Grisham JW (1962) A morphological study of DNA synthesis and cell proliferation in regenerating rat liver: Autoradiograph~ with thymidine H3+. Cancer Res 22: 842-849 8 Grisham JW, Porta EA (1964) Origin and fate of proliferated ductal cells in the rat: electronmicroscopic and autoradiogra~hic studies. Exp Mol Pathol 3: 242-261 9 Grundmann E, Sieburg H (1962) Die Histogenese und Cytogenese des Leberkarzinoms der Ratte durch Diiithyinitrosamin im lichtmikroskopischen Bild. Beitr Path Anat 126: 57-90 20 Gupta DN (1956) Acute changes in the liver after administration of thioacetamide. Nature 175: 257-259 21 Hadjiolov DC (1965) Histochemical studies of some oxidative enzymes in the early proliferative stage of liver cancerogenesis. Zeitschr rur Krebsforschung 66: 473-477 22 Hayner NT, Braun !-' Yaswen P, Brooks M, Fausto N (1984) Isozyme profiles of oval cells, parenchymal cells and biliary cells isolated by centrifugal elutriation from normal and preneoiastic livers. Cancer Res 44: 332-338 2 Hixson DC, Allison JP (1985) Monoclonal antibodies recognizing oval cells induced in the liver of rats by N-2-Fluorenylacetamide or Ethionine in a choline-deficient diet. Cancer Res 45: 3750-3760 24 Hunter AL, Holscher MA, Neal RA (1977) Thioacetamide induced hepatic necrosis. I. Involvement of the mixed-function oxidase enzyme system. J Pharmacol Exp Ther 200: 439-448 25 Inaoka Y (1967) Significance of so-called oval cell proliferation during azo-dye hepatocarcinogenesis. Gann 58: 355-366 26 Kleinfeld RG (1957) Early changes in rat liver and kidney cells induced by thioacetamide. Cancer Res 17: 954-962 27 Kleinfeld RG, Von Haam E (1959a) The effect of thioacetamide on rat liver regeneration. I. Cytologic studies. Cancer Res 19: 769-778 28 Kleinfeld RG, Von Haam E (1959b) The effect of thioacetamide on rat liver regeneration. II. Nuclear RNA in mitosis. J Biophys and Biochem Cytol 6: 393-397 29 Kuhlman WD (1978) Localisation of al-fetoprotein and DNA synthesis in liver cell population during experimental hePtatocarcinogenesis in rats. Int J Cancer 21: 368-380 o Leduc EH (1959) Cell modulations in liver pathology. J Histochem Cytochem 7: 253-257 31 Messier B, Leblond CP (1957) The preparation of coated radioautographs by dipping sections in fluid emulsion. Proc Soc Bioi Med 96: 7-11 Novikoff AB, Essner E (1960) The liver cell. Am J Med 19: 102-131 33 Ogawa K, Takashi M, Onoe T (1974) Demonstration of G-6-phosphatase activity in the oval cells of rat liver and the significance of the oval cells in azo-dye carcinogenesis. Cancer Res 34: 3379-3386 34 Onda H (1976) Immunohistochemical studies on arfetoprotein and ai-acid glycoproteins during azo-dye hepatocarcinogenesis in rats. Gann 67: 253-262

hE.

76 . M. Praet, H. Roels and P. van Oostveldt 35 Onoe T, Kaneko A, Dempo K, Ogawa K, Minase T (1975) at-fetoprotein and early histological changes of hepatic tissue in DAB hepatocarcinogenesis. Ann NY Acad Sci 259: 168-180 36 Opie EL (1946) Mobilisation of basophilic substance (Ribonucleic Acid) in the cytoplasm of liver cells with the production of tumors by butter yellow. J Exp Med 84: 91-106 37 Popper H, Kent G, Stein R (1957) Duetular cell reaction in the liver in hepatic injury. J Mt Sinai Hosp 24: 551-556 38 Porter WR, Neal RA (1978) The metabolism of thioacetamide and thioacetamide-S-oxide by rat liver microsomes. Drug Met Dispos 6: 379-388 39 Praet MM, Roels HJ (1984a) Histogenesis of cholangiomas and cholangiocarcinomas in thioacetamide fed rats. A Review. EX£ Path 26: 3-14 o Praet MM, Roels HJ (1984b) Cytochemical analysis of the chromatin of liver cell nuclei and nuclei of oval cells by means of the Feulgen-Hydrolysis curve in rats treated by Thioacetamide. Cytometry 5: 364-367 4t Praet MM, Roels HJ (1987) Thioacetamide-induced livercarcinogenesis. The oval cell as precursorcell in this multistep model. Arch Int de Physiol et de Bioch 95: 2 42 Rappaport AM (1976) The microcirculatory acinar concept of normal and pathological hepatic structure. Beitr Pathol 157: 215-243 43 Rees KR, Rowland GF, Varcoe JS (1966) The metabolism of tritiated thioacetamide in the rat. Int J Cancer 1: 197-206 44 Rubin E (1964) The origin and fate of proliferated bile ductular cells. Exp and Mol Pathol 3: 279-280 45 Sell S (1978) Distribution of alpha-fetoprotein and albumincontaining cells in the livers of Fischer rats fed four cycles of N-2fluorenylacetamide. Cancer Res 38: 3107-3113 46 Sell S, Osborn K, Leffert HL (1981) Autoradiography of "oval cells" appearing rapidly in the livers of rats fed N-2fluorenylacetamide in a choline-devoid diet. Carcinogenesis 2: 7-14 47 Sell S (1983) Comparison of oval cells induced in rat liver by feeding N-2-fluorenylacetamide in a choline-devoid diet and bile duct cells induced by feeding 4,4'-Diaminodiphenylmethane. Cancer Res 43: 1761-1767

48 Shinozuki H, Lombordi B, Sell S, lammarino RM (1978) Early histological and functional alterations of ethionine liver carcinogenesis in rats fed a choline-deficient diet. Cancer Res 38: 1092-1098 49 Spain J, Griffin AC (1957) A histochemical study of glycogen alterations in the liver of rats following azo dye administration. Cancer Res 17: 200-204 50 Stocker E (1963-1964) Autoradiographische Studien zum EinfluB von TAA auf den zelluliiren EiweiB- und Nukleinsiiurestoffwechsel in Leberzellen der Ratte. Beitr zur Path Anat und zur All~ Path 129: 247-295 5 Tatematsu M, Ho R, Kaku T, Ekem J, Farber E (1984) Studies on the proliferation and fate of oval cells in the liver of rats treated with 2-acetylaminofluorene and partial hepatectomy. Am J Path 114: 418-430 52 Thienpont L (1973) Etude cytophotometrique des alterations nucleaires du foie apres traitement au thioacetamide. C R Seances Soc Bioi 167: 1058-1061 53 Thoenes W, Bannasch P (1962) Elektronen- und lichtmikroskopische Untersuchungen am Cytoplasma der Leberzellen nach akuter und chronischer Thioacetamid-Vergiftung. Virch Arch 335: 556-583 54 Tsao M, Smith JD, Nelson KG, Grisham JW (1984) A diploid epithelial cell line from normal adult rat liver with phenotypic properties of oval cells. Exp Cell Res 154: 38-52 55 UAR: Usines d'Alimentation Rationelle. Villemoisson-sur-

°v~e

Yaswen P, Hayner NT, Fausto N (1984) Isolation of oval cells by centrifugal elutriation and comparison with other cell types purified from normal and preneoplastic livers. Cancer Res 44:324-331 57 Ying TS, Sarma DSR (1979) Role of liver cell necrosis in the induction of preneoplastic lesions. Proceedings of AACR 56 58 Yoshimura H, Harris R, Yokoyama S, Takahashi S, Sells M, Pan S, Lombardi D (1983) Anaplastic carcinomas in nude mice and in original donor strain rats inoculated with cultured oval cells. Am J Pathol11 0: 322-332

Received February 9, 1988 . Accepted in revised form April 28, 1988

Key words: Thioacetamide - Hepatocytes - Oval cells - Liver cell lessions - Cholangiocarcinoma M. M. Praet, N. Goormaghtigh Institute of Pathology, University Hospital, De Pintelaan 185, B-9000 GHENT, Belgium