Development of in vitro toxicity tests with cultures of freshly isolated rat hepatocytes

Can the In Vivo Maximum Tolerated Dose be Predicted Using In Vitro Techniques? A Working Hypothesis

All new chemical entities synthesised in our laboratories have routinely been subjected to in vitro toxicity tests. Out of curiosity, we established a working hypothesis in which the in vitro data could be empirically transformed to predict the in vivo four-week standard maximum tolerated dose (MTD) studies in rats and dogs.

As a first step to verifying this hypothesis, we report here the findings of an in vitro cytotoxicity study of 25 compounds randomly selected from our files, possessing a wide range of pharmacological activities and for which data from standard four-week MTD studies were available. Single blind in vitro toxicity studies in three carefully selected types of primary and cell line cultures were carried out. In vitro CT50 (concentration inducing 50% cell death) and CT100 (concentration inducing 100% cell death) values were obtained for each of the three cell types and, using empirical assumptions, these results were used to predict the MTD in vivo in the rat and dog. The actual in vivo threshold and toxic doses were obtained from the MTD study reports. The in vivo toxicity values predicted from the in vitro toxicity results with this series of 25 compounds showed a better than 80% correlation with the actual in vivo results obtained in the MTD studies.

Whether or not in vitro cytotoxicity predictions are ultimately found to be directly and consistently related to the MTD in vivo for all pharmacological classes of compounds will require many additional studies, but it is hoped that these results will stimulate the necessary research effort required to answer this question.

Hormonal regulation of microsomal flavin-containing monooxygenase activity by sex steroids and growth hormone in co-cultured adult male rat hepatocytes

To investigate the hormonal control of the expression of flavin-containing monooxygenase (FMO; EC 1.14.13.8) under defined in vitro conditions, adult male rat hepatocytes were isolated by collagenase perfusion and co-cultured with rat liver epithelial cells of primitive biliary origin. The direct effect of 17beta-estradiol, testosterone, 5alpha-dihydrotestosterone (5alpha-DHT) and human growth hormone (hGH) on FMO activity was studied using this in vitro model. Optimal, non-cytotoxic hormonal concentrations were determined by measuring the lactate dehydrogenase (LDH) index. In addition, the microsomal protein content of the cultured hepatocytes was determined as a function of culture time. The female sex hormone 17beta-estradiol caused a significant decrease in FMO as a function of culture time. After 14 days of exposure, FMO activity decreased by 56%. Neither of the male sex hormones or human growth hormone had an effect on FMO activity. These results in co-cultured male rat hepatocytes support in vivo observation that 17beta-estradiol is a potent hormone involved in the negative regulation of the expression of FMO in male rat liver.

A specific microassay for evaluating hepatic LDH activity in co-cultures of hepatocytes with other cells

This study describes the development of a simple, rapid and reproducible microassay for determining the intracellular LDH activity of rat hepatocytes present in a co-culture system with other cells. The procedure involves treatment of cellular homogenates with an anti-LDH antiserum that specifically inhibits the LDH activity of rat hepatocytes. The assay is performed in 96-well plates and LDH activity can be measured directly in the same wells using a colorimetric method. The difference in LDH activity values measured before and after antiserum incubation reflects the LDH content of the hepatocytes in the sample. The advantages of this method are the small number of cells required, a reduction in sample handling and the possibility of differentiating LDH activity in hepatic and non-hepatic cells. The possible applications of this technique as a parameter for biochemical data and as a test for cytotoxicity studies in co-cultures are also discussed.

Rat hepatocyte cultures and co-cultures in biotransformation studies of xenobiotics

Long-term cultures of hepatocytes could represent a suitable in vitro model for biotransformation studies of xenobiotics. At present however, no ideal culture system can be proposed since, in all existing models, phenotypic changes occur, affecting selectively some components of phase I and/or phase II xenobiotic metabolism. From the authors' own results and recent studies of several other investigators, carried out on rat hepatocytes, it becomes clear that four groups of factors may affect biotransformation capacity: soluble medium factors, extracellular matrix components, cell-cell interactions and factors affecting replication. For the maintenance of liver-specific functions, it seems of utmost importance that the tridimensional shape of the hepatocytes is kept. Usually, phase II enzymatic activity is better kept than that of phase I. The cytochrome P450 dependent monoxygenases, in particular, are easily lost. Interesting is the observation that co-cultures of rat hepatocytes with rat liver epithelial cells exhibit higher and much better preserved phase I and phase II biotransformation than monocultures. Clearly, further research is needed to improve this promising in vitro model.

Cell-cycle and ploidy analysis in bone marrow and liver cells of rats after long-term consumption of irradiated wheat

Rats were fed for 4 or 90 days either with 70% freshly irradiated wheat (0.25, 0.75 or 2.25 kGy) and 30% complementary feed or with a control diet. None of the parameters examined (food consumption, body weight, haematological analysis, histopathological inspection of thymus, lung, liver, spleen and kidney, DNA analysis of bone marrow cells and nuclei from liver cells by flow cytometry) showed any statistically significant association with the feeding regimen. Minor changes in ploidy of liver cells and cell cycling of bone marrow cells were detectable (wheat-irradiation dose-dependent increase in G2/M-phase bone marrow cells up to 0.6%, decrease of 8C nuclei up to 1.1% in liver cells). From the pattern of alterations observed in our study, radiolytic by-products of wheat irradiation with a spindle poison-like activity can be excluded. Polyploid cells do not accumulate within the 90-day feeding period. The minor effects on cell cycle and ploidy observed are qualitatively comparable with the effects seen after food restriction in animal studies. It is suggested that an altered composition of fatty acids (the components of wheat most sensitive to irradiation) is responsible for these marginal effects. Our findings may explain the earlier findings of Bhaskaram and Sadasivan (American Journal of Clinical Nutrition 1975, 28, 130-135) who reported an increase in the number of polyploid cells in the lymphocytes of malnourished children fed irradiated wheat. The most likely mechanisms for such an effect are adaptive, constitutively regulated processes, similar to those which respond to food restriction. It is concluded that the consumption of irradiated wheat does not, therefore, pose any health risk to humans.

Maintenance and induction in co-cultured rat hepatocytes of components of the cytochrome P450-mediated mono-oxygenase

Hepatocytes grown in culture rapidly lose many of the cytochromes P450 (CYP) responsible for metabolizing foreign compounds. Among the proteins most readily lost are members of the CYP2B subfamily. We have investigated, by RNase protection assays, the ability of rat hepatocytes, cultured conventionally or co-cultured with rat liver epithelial cells, to maintain the expression of genes encoding members of the CYP2B subfamily, and the inducibility of this expression by phenobarbital. After 4 days of conventional hepatocyte culture CYP2B mRNAs were undetectable, but remained inducible by phenobarbital. In co-cultured hepatocytes the abundance of the mRNAs remained relatively constant from 4-14 days. After 7 days of co-culture the concentration of the mRNAs was increased 12-15-fold by phenobarbital. RNase protection assays with probes capable of distinguishing between CYP2B1 and 2B2 mRNAs demonstrated that the ratios of the abundance and inducibility of the two mRNAs were the same in co-culture as in vivo. Co-cultured hepatocytes also maintained the expression of genes coding for two other components of the cytochrome P450-mediated mono-oxygenase, namely cytochrome P450 reductase and cytochrome b5.

Crude liver membrane fractions and extracellular matrix components as substrata regulate differentially the preservation and inducibility of cytochrome P-450 isoenzymes in cultured rat hepatocytes

The influence of cell-substrata interactions on the preservation of basal or in-vivo-induced microsomal cytochrome P-450 isoenzyme contents in cultured rat hepatocytes and on the adaptive responses after exposure to phenobarbital or 3-methylcholanthrene in vitro, was investigated. Hepatocytes from untreated or phenobarbital-treated rats were cultured in serum-free, aprotinin-supplemented culture medium in 96-well microtiter plates coated with collagen type I (COL), laminin, fibronectin or crude liver membrane fractions/collagen type I (CMF/COL). Basal cell functions were characterized by measuring the total protein content and lactate dehydrogenase release. The relative contributions of CYP1A1/2, CYP2B1/2, CYP2C6, CYP2C11, CYP3A and CYP4A isoenzymes were determined with ELISA using monoclonal antibodies raised against purified cytochromes P-450 from rat liver microsomes. The characterization of the CMF revealed that contaminations with mitochondria, nuclei and lysosomes are relatively low. Among these, membranes derived from the endoplasmic reticulum appeared to be the major organelle contaminant of the CMF. The matrix components laminin, fibronectin and collagen type IV were found in appreciable amounts. Hepatocytes from untreated rats, cultured for up to nine days on CMF/COL-coated plates, retained their relative cytochrome P-450 contents at 1.5-3-fold higher levels when compared to cells cultured on COL, fibronectin or laminin. Similarly, hepatocytes from phenobarbital-treated rats preserved the contents of barbiturate-inducible CYP2B1/2 and CYP3A proteins best when cultured on CMF/COL. After exposure of hepatocytes cultured on CMF/COL to phenobarbital from days 3-6, CYP3A proteins were enhanced more than twofold and CYP2B1/2, depending on the exposure level, increased 1.3-6-fold. After exposure to 3-methylcholanthrene, a threefold increase of CYP1A proteins was found in CMF/COL and laminin cultures. These results indicate that CMF/COL, as a substratum in rat hepatocyte cultures, regulates gene expression of cytochromes P-450 isoenzymes for up to 9 days and provides a matrix which enables the cells to respond qualitatively similar to the response observed in different zones of the liver. This activity cannot be replaced by single-matrix components.

Crude liver membrane fractions as substrate preserve liver-specific functions in long-term, serum-free rat hepatocyte cultures

Over time, rat hepatocytes cultured on collagen lose the capacity to express liver-specific functions. The influence on this degradation process of an alternative substratum--crude membrane fractions prepared from the liver of the same rat strain--was investigated. Freshly isolated rat hepatocytes were cultured in serum-free Williams E medium supplemented with aprotinin, selenium, dexamethasone, and insulin in flasks coated with a mixture of rat liver crude membrane fractions:collagen type I (100:1). The cells adhered firmly, exhibiting minimal spreading and remaining grouped in columns or in cell islands, and retained their liver-specific functions for more than 1 wk. Hepatocytes secreted substantially higher amounts of albumin than cells cultured on collagen-coated dishes, and on Days 1 and 9 in culture the total P-450 content was 72 and 40%, respectively, of that of freshly isolated cells. On Day 6, the 7-ethoxyresorufin-O-deethylase and the aldrin epoxidase activities were still more than 50% that of freshly isolated hepatocytes. Exposure to phenobarbital on Days 3 to 6 increased the total cytochrome P-450 content twofold; exposure to 3-methylcholanthrene increased the activity of the corresponding cytochrome P-450 isoforms to 20 times that observed in untreated cultures and 6 times that observed in freshly isolated cells. Thus, given the ease with which they are prepared, the use of crude membrane fractions combined with culture medium supplemented with aprotinin and selenium can facilitate the preparation of reproducible cultures suitable for long-term in vitro pharmacotoxicologic studies using rat hepatocytes.

Calcium antagonists can be classified using in vitro toxicity and potency indices

The toxicity of eleven calcium antagonists from different chemical families was determined in rat hepatocyte primary cultures. The calcium antagonist potency of the same compounds was also determined in isolated rabbit aortic rings contracted with high K+. The hepatocytotoxicity of the calcium antagonists was not directly linked to blockade of voltage-operated calcium channels, since there was no correlation between the rank order of hepatotoxicity and that for calcium antagonist potency. The toxicity and calcium antagonist potency of each calcium antagonist examined were used to calculate an in vitro therapeutic index value for each compound. It was observed that therapeutic indices fell into three distinct groups and we therefore propose that the in vitro therapeutic index can be used to subclassify the calcium antagonist group of drugs. The proposed classification corresponds very closely with one already suggested by Spedding on pharmacological grounds. In conclusion, the in vitro therapeutic index may provide a useful tool in the characterization and subclassification of novel calcium antagonist compounds.

Dependency of the in vitro stabilization of differentiated functions in liver parenchymal cells on the type of cell line used for co-culture

The differentiation status in cultures of primary rat liver parenchymal cells was determined by measuring the activities of various xenobiotic metabolizing enzymes. Most enzyme activities dropped rather rapidly in monocultures of parenchymal cells. The protein content and the activities of cytosolic epoxide hydrolase, glutathione S-transferase, and alpha-naphthol UDP-glucuronosyl transferase were, however, well stabilized in 7-day-old co-cultures of parenchymal cells with two different lines of rat liver nonparenchymal epithelial cells (NEC1 and NEC2). Phenol sulfotransferase and microsomal epoxide hydrolase activity were reduced in this coculture system after 7 days to about 30 and 20% of the initial activity. Generally, higher enzyme activities were measured in co-cultures with one specific epithelial cell line (NEC2) as compared to those with the other line (NEC1). C3H 10T1/2 mouse embryo fibroblasts supported the parenchymal cells even better than the two epithelial lines, because the activity of microsomal epoxide hydrolase was also stabilized. Glutathione transferase activity was increased over time in this co-culture system. Our results show that the differentiation status of liver parenchymal cells was much better stabilized in co-cultures than in monocultures but that, depending on the type of cells used for co-culture, great quantitative differences existed. The entire pattern of xenobiotic metabolizing enzyme activities could not be stabilized at the kind of levels found in freshly isolated parenchymal cells.

The stem cells of the liver ? a selective review

The current status of the much-debated question of the still-hypothetical stem cells of the liver is reviewed, with an emphasis on their role in hepatocarcinogenesis. The widely held view of the primacy of the hepatocyte, notably of the mononuclear diploid type, in this process--the "hepatocytic theory"--has been compared with variants of the "stem cell hypothesis" based on the "non-parenchymal epithelial cells" of the liver--the "oval" or biliary ductular cells, the "nondescript periductular" cells and the "primitive" bipotential epithelial cells. An attempt has been made to concentrate mainly on the more recent publications, in an effort to balance the conflicting opinions expressed by comparing results obtained by the newer procedures currently in use. Despite some interesting and relevant findings it appears that the evidence in favour of the stem-cell hypothesis is still circumstantial and that the hepatocytic theory has not been invalidated. Presumably the question of the hepatic stem cells will be answered when the riddle of hepatocarcinogenesis has been solved.

Differential stabilization of cytochrome P-450 isoenzymes in primary cultures of adult rat liver parenchymal cells

Cytochrome P-450 dependent hydroxylation of testosterone was measured in 7-day-old cultures of primary rat liver parenchymal cells. Determinations were carried out in monocultures of parenchymal cells and co-cultures of parenchymal cells with rat liver nonparenchymal epithelial cells, or mouse embryo fibroblasts. In the monoculture system, testosterone metabolism was drastically reduced and hardly measurable after 7 days in culture. In the co-culture systems, individual P-450 isoenzymes were stabilized on different levels. P-450s p and presumably c were well preserved, P-450 a was reduced but clearly measurable, P-450 h was totally lost whereas P-450s b and e were not measurable after 7 days (the activities of these isoenzymes however were already low in freshly isolated parenchymal cells). The results were independent of the cell line used for co-cultivation and of the method of parenchymal cell isolation, that is whether collagenase or EDTA was used as the agent for dissociating the cells from the liver. The results showed that the co-cultivation of liver parenchymal cells with other nonparenchymal cells significantly improved the differentiated status of the former. In this cell culture system however, not every parameter was equally well stabilized.

Single cell analysis in toxicity testing: the mitogenic activity of thioacetamide in cultured rat hepatocytes analyzed by DNA/protein flow cytometry

Rat hepatocytes were cultured at 4% O2 and 13% O2 and exposed to the nongenotoxic rodent carcinogen thioacetamide (TA) from 24 to 72 h after isolation at exposure levels between 0.01 and 0.33 mM. Hepatocytes and isolated nuclei were analyzed by DNA-protein flow cytometry. An aggregate correction procedure was applied and the proportion of S-phase, diploid, tetraploid or octoploid hepatocytes as well as binucleated cells, were measured or calculated. The proportion of S-phase cells within the diploid hepatocytes increased with increasing concentration of TA up to 3.9-fold, whereas the corresponding increase in S-phase mononucleated tetraploid cells was only 1.8-fold. S-phase binucleate tetraploid cells showed no increase. In the tetraploid hepatocytes, the mitogenic stimuli was detectable only in cultures maintained at 4% O2. The relative contribution of binuclear cells was increased 1.5-fold in the octoploid cells. It is concluded that the mitogenic activity of TA initiates DNA synthesis in diploid hepatocytes in the G1 and in the following G2 cell-cycle phase, omitting karyogenesis. The cellular protein content is not affected which indicates that the mitogenic activity of the chemical is not necessarily associated with an increase in cellular protein content. The results obtained correspond well with data of in vivo studies. The method applied therefore allows the mitogenic activity of nongenotoxic carcinogens to be detected in vitro within 48 h and their mode of action to be elucidated.

Rat adult hepatocytes in primary pure and mixed monolayer culture. Comparison of the maintenance of mixed function oxidase and conjugation pathways of drug metabolism

The stabilities of several drug oxidation and conjugation pathways in adult rat hepatocytes were investigated in two systems: a primary pure culture lasting 3 days and a primary mixed culture (hepatocytes co-cultured with epithelial cells) lasting 10 days. The cytochrome P450 content in hepatocytes drastically declined within 48 hr in both culture systems. Cytochrome P450-dependent mixed function oxidase was measured by the O-dealkylation of ethoxyresorufin (EROD) and of pentoxyresorufin (PROD). UPD-glucuronosyl transferase (UDP-GT) activity was measured using 1-naphthol and morphine as substrates. In both culture systems, the activities of enzymes belonging to the 3-methylcholanthrene-inducible family, namely EROD and 1-naphthol UDP-GT, were much better maintained than those of PROD and morphine UDP-GT, which belong to the phenobarbitone-inducible family: in pure cultures, EROD and 1-naphthol UDP-GT activities declined to 60% of initial values within 3 days; in mixed cultures, EROD activity was stable throughout the 10 day culture period, whereas that of 1-naphthol UDP-GT was stable until day 4 but had declined to 70% of the initial value by day 8. In contrast, PROD and morphine UDP-GT activities declined to approx. 30% of the initial values within 2 days in both culture systems, and had dropped to approx. 10% of the initial value within 8 days in mixed culture. Reduced glutathione (GSH) levels fluctuated, but remained high throughout culture. GSH conjugation declined to 40% of initial values within 3 days in pure culture, whereas it remained relatively constant in mixed culture. Comparison of these two culture systems therefore showed that although the inclusion of epithelial cells did prolong hepatocyte viability, there was a change in relative enzyme activities in both systems, suggesting a shift towards a more de-differentiated drug metabolism pattern.

Co-cultures of hepatocytes with epithelial-like cell lines: expression of drug-biotransformation activities by hepatocytes

To improve long-term expression of drug biotransformation activities in hepatocytes, we have examined the suitability of several epithelial-like cell lines (MDCK, MS and L-132) for supporting functional co-cultures with rat hepatocytes. Cells were selected on the basis of their compatibility with hepatocytes, formation of stable monolayers in the absence of serum and lack of drug biotransformation activities. The expression of individual elements of the biotransformation system was evaluated in these co-cultures. Co-cultured hepatocytes remained viable and showed a characteristic polygonal shape for more than a week. Depending on the cell line used, levels of aryl hydrocarbon hydroxylase and 7-ethoxycoumarin O-deethylase activities of co-cultured hepatocytes oscillated between 24-47% of their initial value after 4 days in culture. The highest levels of monooxygenase activity were found in hepatocytes co-cultured with MS cells (41-47%). In contrast, these activities decreased to 6% when hepatocytes were maintained in pure culture for the same period. The activities of the conjugating enzymes UDP-glucuronyltransferase and glutathione S-transferase were maintained at nearly the initial levels during the complete period of study, both in pure and mixed-cultures, regardless of the cell line used. MS cells adapted themselves much better to serum-free culture conditions, and the co-culture with rat hepatocyte was technically easier. After one week, total cytochrome P450 and reduced glutathione in rat hepatocytes/MS co-cultures were 31% and 127% respectively of the day O values, whereas they were undetectable in pure culture. A clear induction of monooxygenase activities by methylcholanthrene, phenobarbital and ethanol could be observed by the 5th day in MS cells/hepatocyte co-cultures. The fact that the results of our work show the suitability of MS cells, an epithelial-derived cell line, for improving the expression of biotransformation enzymes of cultured hepatocytes opens new possibilities of simplifying co-cultures for their use in drug-metabolism studies.

Drug metabolizing enzymes in rat hepatocytes co-cultured with cell lines

We have developed new co-cultures of continuous cell lines 3T3 (clone A31) and C3H/10T1/2 (clone 8) with hepatocytes as an alternative to co-cultures with noncontinuous epithelial cells. In this biological system we studied in detail the expression of the hepatic biotransformation system. After 7 d in culture, total cytochrome P-450 content and the monooxygenase activities aryl hydrocarbon hydroxylase and 7-ethoxycoumarin o-deethylase still maintained about 30% of their initial value, whereas in pure cultured hepatocytes these activities were undetectable. A significant response to induction by methylcholanthrene and phenobarbital of monooxygenase activities was observed in co-cultures for 7 d. NADPH-cytochrome c reductase activity remained unchanged for at least 7 d in co-cultured hepatocytes, whereas in pure cultures this activity was reduced to about 75% of the initial value after only 24 h. Finally, the activity of the conjugating enzymes UDP-Gt and GSH-t was maintained at nearly the initial levels during the complete period of study. The easy handling of continuous cell lines and the maintenance of the biotransformation system of hepatocytes in co-culture make this approach simpler and easier to standardize.

Phase I and phase II xenobiotic biotransformation in cultures and co-cultures of adult rat hepatocytes

The aim of this study was to measure the activity of phase I and II key enzymes in the biotransformation of xenobiotics and their inducibility by phenobarbital (2 mM) in two currently used in vitro models, namely adult rat hepatocytes, conventionally cultured or co-cultured with rat epithelial cells derived from primitive biliary duct cells. For phase I, the cytochrome P450 content and the enzymic activities of 7-ethoxycoumarin O-deethylase and aldrin epoxidase have been determined, for phase II glutathione S-transferase activity was measured. In conventional cultures, all phase I parameters investigated declined continuously as a function of culture time. Two mM phenobarbital had inducing effects on 7-ethoxycoumarin O-deethylase and glutathione S-transferases but not on aldrin epoxidase. In co-cultures, after an initial decrease, a steady state situation developed for all the parameters measured, lasting for at least 10 days. The cytochrome P450 content, the 7-ethoxycoumarin O-deethylase, aldrin epoxidase and glutathione S-transferase activities were maintained from 3 to 4 days on at 25, 100, 15 and 50%, respectively, of their corresponding value obtained for freshly isolated hepatocytes. After phenobarbital treatment, the parameters mentioned were significantly increased with the exception of the aldrin epoxidase activity of which the inducibility was nearly completely lost.

Influence of hormones and drugs on glutathione-S-transferase levels in primary culture of adult rat hepatocytes

GST activities against 1-Chloro-2,4-dinitrobenzene (CDNB) and 1,2-dichloro-4-nitrobenzene (DCNB) were measured in isolated and cultured adult rat hepatocytes. Within 24 h in culture, both GST activities decreased to about 70% and either stabilized at this level (CDNB) or recovered (DCNB) to the initial level. Use of hyaluronidase in addition to collagenase during the isolation of the cells strongly reduced both activities and its stimulation by various drugs for up to 168 h. The hormones insulin, glucagon, triiodothyronine, estradiol, testosterone, and progesterone did not affect GST activity, while dexamethasone showed some interference. In the presence of dexamethasone the activity against CDNB was mainly stimulated by the combination of methylcholanthrene (MC) and phenobarbital (PB) to about 260% within 168 h. The activity against DCNB was stimulated predominantly by MC alone reaching 170% after 168 h. Quantification of the GST subunits Ya, Yb1 and Yp by an ELISA technique revealed a strong decrease of Ya, a transient increase of Yb1 after 24 h followed by a moderate decrease, and a stable low level of the transformation marker Yp during cultivation. The level of Ya was markedly induced by PB, particularly in combination with MC. The level of Yb1 was equally induced by MC or PB with no synergistic effect. Yp was not affected by these drugs. None of the hormones affected the level of these GST subunits. These results indicate that the physiological type of regulation of the GSTs is maintained during primary culture and no signs of dedifferentiation or transformation are observed. Furthermore, they demonstrate that the interaction of drugs and hormones and their inducing potential can be efficiently studied in the cultured hepatocytes.