Statin-induced myopathy in the rat: relationship between systemic exposure, muscle exposure and myopathy.

Rare instances of myopathy are associated with all statins, but cerivastatin was withdrawn from clinical use due to a greater incidence of myopathy. The mechanism of statin-induced myopathy with respect to tissue disposition was investigated by measuring the systemic, hepatic, and skeletal muscle exposure of cerivastatin, rosuvastatin, and simvastatin in rats before and after muscle damage. The development of myopathy was not associated with the accumulation of statins in skeletal muscle. For each statin exposure was equivalent in muscles irrespective of their fibre-type sensitivity to myopathy. The low amount of each statin in skeletal muscle relative to the liver does not support a significant role for transporters in the disposition of statins in skeletal muscle. Finally, the concentration of cerivastatin necessary to cause necrosis in skeletal muscle was considerably lower than rosuvastatin or simvastatin, supporting the concept cerivastatin is intrinsically more myotoxic than other statins.

Novel imprinted DLK1/GTL2 domain on human chromosome 14 contains motifs that mimic those implicated in IGF2/H19 regulation.

The evolution of genomic imprinting in mammals occurred more than 100 million years ago, and resulted in the formation of genes that are functionally haploid because of parent-of-origin-dependent expression. Despite ample evidence from studies in a number of species suggesting the presence of imprinted genes on human chromosome 14, their identity has remained elusive. Here we report the identification of two reciprocally imprinted genes, GTL2 and DLK1, which together define a novel imprinting cluster on human chromosome 14q32. The maternally expressed GTL2 (gene trap locus 2) gene encodes for a nontranslated RNA. DLK1 (delta, Drosophila, homolog-like 1) is a paternally expressed gene that encodes for a transmembrane protein containing six epidermal growth factor (EGF) repeat motifs closely related to those present in the delta/notch/serrate family of signaling molecules. The paternal expression, chromosomal localization, and biological function of DLK1 also make it a likely candidate gene for the callipyge phenotype in sheep. Many of the predicted structural and regulatory features of the DLK1/GTL2 domain are highly analogous to those implicated in IGF2/H19 imprint regulation, including two hemimethylated consensus binding sites for the vertebrate enhancer blocking protein, CTCF. These results provide evidence that a common mechanism and domain organization may be used for juxtapositioned, reciprocally imprinted genes.

Potentiation of epidermal growth factor-induced DNA synthesis in rat hepatocytes by phenobarbitone: possible involvement of oxidative stress and kinase activation.

A transient induction of S phase DNA synthesis is a common feature of non-genotoxic rodent hepatocarcinogens when administered in vivo. In the present study the ability of phenobarbitone (PB) to induce S phase DNA synthesis in primary cultures of rat hepatocytes was investigated. In the absence of serum or growth factors PB was not a mitogen per se. However, stimulation of S phase DNA synthesis by epidermal growth factor (EGF) was enhanced by co-culture with PB. This effect was both time and concentration dependent. The lowest concentration of PB that significantly enhanced the effect of EGF was 10 microM and the effect was maximal at 1.0 mM. At a concentration of 2.0 mM PB no longer enhanced EGF-induced S phase DNA synthesis. Hepatocyte cultures pretreated with PB (0.1 mM) for 2 days were more responsive to the induction of S phase DNA synthesis by EGF for the subsequent 2 days. Despite the inhibition of PB enhancement of S phase DNA synthesis by the antioxidant dimethylthiourea, reduced glutathione was not depleted by PB treatment nor were oxidized glutathione or lipid peroxides elevated. Western blotting analysis showed that PB had no effect on epidermal growth factor receptor (EGFR) autophosphorylation per se after 1 and 48 h culture, enhanced sensitization of EGFR therefore does not appear to contribute to the enhancement of S phase DNA synthesis by PB. In contrast, treatment of hepatocytes with PB for 12 h resulted in a small but statistically significant activation of p42/44 MAP kinase activity and activation of protein kinase C, as measured by redistribution of enzyme activity from a soluble to a particulate compartment of hepatocytes. Therefore, PB-mediated changes in protein kinase activity may contribute to the potentiation this compound affords.

Identification of phenobarbitone-modulated genes in mouse liver by differential display.

The molecular basis of how rodent nongenotoxic hepatocarcinogens such as phenobarbitone cause liver-tumor formation is poorly understood. An early effect of phenobarbitone exposure is to induce hepatocyte proliferation transiently, and there is evidence that this may be important for subsequent tumor development. In this investigation, we have used the differential display reverse transcriptase polymerase chain reaction technique to analyze differential gene expression in male C57B1/10J mouse liver during the mitogenic phase of the phenobarbitone response. Seventy-seven putative differentially expressed cDNAs were isolated by differential display, and 13 of them were subsequently confirmed as being differentially expressed (both increased and decreased by phenobarbitone). Seven of the cDNAs were homologous to known mouse or human genes (carboxylesterase, coagulation factor X, amine N-sulphotransferase, human protein disulphide isomerase-related protein, cytochrome c oxidase subunit IV, golgin-245, thioredoxin reductase, betaine-homocysteine methyl transferase) and the remainder were novel. The expression pattern of the sulphotransferase was further characterized, and in mouse liver it was found to be significantly induced by phenobarbitone and not by five other rodent nongenotoxic hepatocarcinogens. In summary, the technique has enabled the identification of previously uncharacterized genes whose expression patterns are differentially altered by phenobarbitone in the mouse liver.

EGF-Induced receptor autophosphorylation in primary hepatocytes isolated from phenobarbitone-treated mice.

Phenobarbitone (PB) treatment of mice causes a decrease in the growth factor responsiveness of hepatocytes. Here, epidermal growth factor receptor (EGFR) expression and receptor autophosphorylation was determined in hepatocytes isolated from control and PB-treated mice. There was a decrease in the level of EGFR expression in hepatocytes isolated from mice following PB administration when compared to controls. EGF caused an approximate 20-fold increase of the 170 kD phosphotyrosine band in control hepatocytes, which was inhibited by the EGFR specific tyrosine kinase inhibitor 4, 5-dianilinopthalamide. Following PB treatment, the degree of basal receptor phosphorylation (in the absence of EGF) was significantly greater and therefore the fold rise in EGFR phosphorylation in isolated hepatocytes was lower than in controls. However, the overall extent of EGF-induced receptor phosphorylation was not diminished in hepatocytes isolated from PB-treated mice. Therefore the reduction in responsiveness to growth factors seen in hepatocytes ex vivo or the cessation of proliferation observed in vivo following PB administration is unlikely to be attributed to a decrease in ligand binding and subsequent receptor autophosphorylation.

The response of hepatocytes isolated from phenobarbitone treated mice to mitogenic growth factors.

The ability was investigated of epidermal growth factor (EGF) and hepatocyte growth factor (HGF) to stimulate DNA synthesis in hepatocytes isolated from C57B1/6J mice following 1, 3, 7, 30 and 90 days pre-treatment with the hepatomegalic drug, phenobarbitone (PB). A 3-fold increase in S-phase labelled hepatocytes was observed in the absence of growth factors after 3 days treatment with PB, which was not seen at other investigated time points. This suggests that the proliferative influence present in vivo at this time interval is maintained in the ex vivo model. Maximum labelling indices of > 5-fold the unstimulated control value were observed in hepatocytes isolated from control and 1 day PB pre-treated mice when cultured in the presence of 5 or 10 ng/ml EGF or HGF. Hepatocytes isolated from 3, 7, 30 or 90 day treated mice showed a considerably reduced responsiveness to growth factors; maximum labelling indices did not exceed by a factor of 2 the value obtained in the absence of growth factors. However, the apparent decrease in responsiveness to growth factors in hepatocytes isolated from 3 day pre-treated mice was due to an increased background level of proliferation and the attainment of a 'ceiling level' of DNA synthesis at approx. 35%. DNA synthesis was not further enhanced by addition of both EGF and HGF. This maximal level of stimulation may indicate that only a specific hepatocyte sub-population is capable of responding to growth factors under the conditions employed. The loss in sensitivity to mitogenic stimuli after 7 days PB pre-treatment correlates with a reported decrease in receptor protein and mRNA levels in rats and coincides with the in vivo shift from hyperplasia to hypertrophy.

Carcinogenicity of cyproterone acetate in the mouse.

The synthetic progestin cyproterone acetate (CPA) has been shown to be a hepatocarcinogen in the rat, but little is known of its effects in mice. A 52 week CPA study in the mouse strain C57Bl/10J has been reported not to produce liver tumours, although CPA induced significant liver enlargement and induction of the mixed function oxidase CYP3A. The present study is a further investigation of the effects of CPA in mice of the C57Bl/10J strain dosed for 104 weeks. A group of 40 mice/sex were fed 800 p.p.m. CPA in the diet for 104 weeks with a control group of eight/sex. Mortality was high in females after 40 weeks due to hormonal effects in the uterus; no female and only four CPA-dosed males survived to 104 weeks. Liver cell hypertrophy with increased fat and glycogen and single cell or small multifocal areas of hepatocellular necrosis were universal. Proliferating cell nuclear antigen demonstrated an increase in proliferating cells within tumours and within the non-tumour bearing liver of CPA-dosed mice compared with normal livers of control mice. Hepatocellular tumours developed in 44% of males and 22% of females dosed with CPA, compared with none in the controls (the strain has a low, <10%, incidence of spontaneous liver tumours compared with other mouse strains). In addition, over 85% of both sexes dosed with CPA developed adenomatous polyps of the pyloric antrum and pancreatic islet cell hyperplasia, shown by immunostaining to be chiefly of insulin-secreting cells. Adrenocortical atrophy was also observed with other widespread effects in the endocrine system. The results suggest that the liver tumours, as in the rat, are likely to be related to effects on liver growth and mitogenesis. It is suggested that the tumours of the stomach and the pancreatic islet cell hyperplasia are manifestations of the effects of CPA in the endocrine system.

Expression of growth factors and growth factor receptors in the liver of C57BL/10J mice following administration of phenobarbitone.

Liver enlargement is a common feature of non-genotoxic rodent hepatocarcinogens administered at high doses. In the present study, the expression of growth factors and growth factor receptors was investigated in the C57BL/1OJ mouse during liver enlargement induced by the non-genotoxic rodent hepatocarcinogen, sodium phenobarbitone (PB). Male mice were dosed 0-2500 p.p.m. PB in the diet for 1, 4 and 13 weeks. There was a dose and time dependent increase in liver weight. Hepatocyte replication, assessed by incorporation of bromodeoxyuridine, was increased in a dose-dependent manner at week 1 only (18-fold increase at 2000 p.p.m.) and was predominantly localized in the centrilobular region. At week 1, PB (2500 p.p.m.) caused transient increases in transforming growth factor alpha (TGFalpha) and epidermal growth factor receptor (EGFR) and decreases in transforming growth factor beta1 (TGF-beta1) and mannose-6-phosphate receptor (M6PR) in centrilobular hepatocytes which correlated with the replication in this region. At week 1, there was an increase in both hepatocyte growth factor (HGF) and hepatocyte growth factor receptor (HGFR) which colocalized in centrilobular hepatocytes; in some mice or periportal hepatocytes in other mice. After 13 weeks, HGF and HGFR were localized in the cytoplasm of centrilobular hepatocytes of all mice but exhibited a differential intracellular distribution across the lobule. At 2500 p.p.m. PB, EGFR and HGFR mRNA were essentially unchanged over the 13 week dosing period whilst M6PR mRNA was increased 2- to 4-fold. At 2500 p.p.m. PB, EGFR protein levels from immunoblots showed a consistent decrease over the 13 weeks whilst M6PR and HGFR protein levels were essentially unchanged. The protein level and mRNA data for EGFR suggest post-transcriptional modification. Thus, phenobarbitone caused transient replication of hepatocytes and modulation of growth stimulatory and inhibitory factors and their associated receptors in terms of overall levels and regional distribution in the liver.

M6P/IGF2R gene is mutated in human hepatocellular carcinomas with loss of heterozygosity.

The mannose 6-phosphate/insulin-like growth factor-II receptor (M6P/IGF2R) functions in the intracellular trafficking of lysosomal enzymes, the activation of the potent growth inhibitor, transforming growth factor beta 2, and the degradation of IGF2 (ref. 1), a mitogen often overproduced in tumours. We have recently shown that 70% of human hepatocellular tumours have loss of heterozygosity (LOH) at the M6P/IGF2R locus which maps to chromosome 6q26-q27 (ref. 8). Using a coarse screen, we have now identified point mutations in the remaining allele of 25% of human hepatocellular carcinomas (HCCs) with LOH. These mutations give rise to truncated receptor protein and significant amino acid substitutions, and provide evidence that the M6P/IGF2R gene functions as a tumour suppressor in human liver carcinogenesis.

Frequent loss of heterozygosity on 6q at the mannose 6-phosphate/insulin-like growth factor II receptor locus in human hepatocellular tumors.

The mannose 6-phosphate/insulin-like growth factor II receptor (M6P/IGFIIr) is required for the activation of transforming growth factor beta, and previously we have found its expression to be significantly reduced in both rat and human hepatocellular carcinomas (HCCs). Therefore, we have postulated that loss of the M6P/IGFIIr gene may be mechanistically involved in liver carcinogenesis. Using the polymerase chain reaction, we utilized two polymorphisms in the 3' untranslated region of the M6P/IGFIIr gene to screen non-cirrhotic, hepatitis virus negative patients with hepatocellular tumors for LOH. Twenty-two of 36 (61%) patients were informative (heterozygous), and 14/22 (64%) liver tumors had LOH; 11/16 (69%) carcinomas, 1/3 (33%) fibrolamellar tumors and 2/3 (67%) adenomas. This is the first report of LOH at the M6P/IGFIIr locus in human hepatocellular tumors, and the presence of LOH in adenomas suggests that allelic loss may be an early event in the etiology of HCCs. These results support the hypothesis that the M6P/IGFIIr gene may function as a tumor suppressor gene in the liver.

Ras mutations in methylclofenapate-induced B6C3F1 and C57BL/10J mouse liver tumours.

The majority of genotoxic carcinogen-induced liver tumours of the sensitive B6C3F1 mouse contain activated H-ras oncogenes. Such mutations also occur in hepatocarcinogenesis-resistant strains. In order to determine whether this is true of non-genotoxic carcinogen-induced tumours, liver tumours induced in B6C3F1 and C57BL/10J mice by methylclofenapate (MCP) were compared. Polymerase chain reaction (PCR) analysis revealed H-ras codon 61 mutations in 11/46 B6C3F1 and 4/31 C57BL/10J liver tumours. The nude mouse tumorigenicity (NMT) assay was used to analyse tumours without codon 61 mutations. Of the 12 B6C3F1 liver tumour DNAs subjected to this assay, one contained a H-ras codon 117 mutation. Further PCR analysis on frozen tumour samples (46 B6C3F1 and 15 C57BL/10J) revealed no codon 12 mutations; one additional codon 117 mutation was identified in a B6C3F1 tumour. Overall, then, H-ras codon 61 mutations were detected in MCP-induced B6C3F1 tumours less frequently than in genotoxin-induced tumours. Two B6C3F1 tumours contained codon 117 mutations similar to those previously found in tumours induced by ciprofibrate, furan and furfural, and in at least one spontaneous tumour. Ras mutations were also detected in some C57BL/10J tumours, providing further evidence that ras oncogenes can participate in hepatocarcinogenesis in resistant mice.

Comparison of the acute and chronic mitogenic effects of the peroxisome proliferators methylclofenapate and clofibric acid in rat liver.

Peroxisome proliferators are well known to cause liver enlargement in rodents. In this investigation, we have examined the effect of acute (1 week) and chronic (26 week) exposure to the peroxisome proliferators methylclofenapate (MCP) and clofibric acid (CA), at 0.05 and 0.5% in the diet respectively, on hepatocyte replication in the Sprague-Dawley rat. Both compounds induced an early increase in hepatocyte replication, with a concomitant increase in peroxisome proliferation as assessed by induction of palmitoyl CoA oxidation. However, after 26 weeks of treatment, there was no difference in the labelling index (LI) of control and CA-treated rat livers, while in MCP-treated rats the LI was 5- to 6-fold above control. Palmitoyl CoA oxidation remained elevated in both treated groups at 26 weeks. Analysis of the slides by a 'zonal' scoring procedure demonstrated that the induced replication was predominantly periportal after 1 week of treatment with either compound. The number of 5-bromo-2'-deoxyuridine (BrdU)-positive hepatocyte nuclei per field in the periportal region increased approximately 4-fold after CA treatment and 7-fold after MCP treatment. There was no significant difference in the number of BrdU-positive nuclei per field in the centrilobular areas of control and treated rats after 1 week. After 26 weeks of treatment, periportal replication was still elevated in the MCP-treated animals (approximately 10-fold above control), but there was no difference in periportal replication between control and CA-treated rats. CA induced a significant reduction in the replication of centrilobular areas at 26 weeks, while there was no effect of MCP. In summary, these results demonstrate that the acute mitogenic effects of MCP and CA are predominantly periportal, and, in the case of MCP, the mitogenicity is sustained up to 26 weeks of treatment.

Proto-oncogene activation in liver tumors of hepatocarcinogenesis-resistant strains of mice.

Activation of the ras family of oncogenes occurs frequently in liver tumors of the B6C3F1 mouse, a strain which is highly sensitive to hepatocarcinogenesis. Many other mouse strains are much more resistant to hepatocarcinogenesis; the aim of this study was to determine the frequency and pattern of oncogene activation in spontaneous and chemically induced liver tumors of three such strains, the C57BL/6J, the C57BL/6 x DBA/2 F1 hybrid (B6D2F1) and the C57BL/6 x Balb/c F1 hybrid (B6BCF1). The C57BL/6, DBA/2 and Balb/c strains are all relatively resistant to spontaneous hepatocarcinogenesis (1.5-3.6% of animals develop liver tumors in 2 years); with regard to chemically induced hepatocarcinogenesis the Balb/c is highly resistant, the C57BL/6 has low susceptibility and the DBA/2 has low to moderate susceptibility. The nude mouse tumorigenicity assay was used to search for activated oncogenes in 15 C57BL/6J liver tumors induced by a single neonatal dose of vinyl carbamate (VC, 0.15 mumol/g body weight). Three tumors contained H-ras genes activated by point mutations at codon 61 and one contained a non-ras oncogene. The polymerase chain reaction and allele-specific oligonucleotide hybridization were used to study H-ras mutations in spontaneous and VC-induced tumors from all three strains of mice. The frequency of H-ras codon 61 mutations in tumors induced by 0.15 mumol/g body weight VC in the C57BL/6J mouse (5/37) was similar to that in spontaneous tumors (2/9); surprisingly, tumors induced by a lower dose of VC (0.03 mumol/g body weight) had a higher frequency of H-ras mutations (12/28). The frequencies of H-ras activation detected in VC (0.03 mumol/g body weight)-induced tumors from the two F1 hybrids studied differed markedly. Only one VC-induced B6BCF1 tumor contained a mutated H-ras gene (1/10), whereas the majority of B6D2F1 tumors contained such mutations (23/33). Several spontaneous B6D2F1 liver tumors contained H-ras codon 61 mutations (6/15). Thus, H-ras activation frequency does not determine susceptibility to hepatocarcinogenesis in inbred mice and their F1 hybrids, since a relatively high frequency of H-ras mutations was observed in two resistant strains and a low frequency was found in the other strain.

Point mutation analysis of ras genes in spontaneous and chemically induced C57Bl/10J mouse liver tumours.

The high incidence and profile of ras gene mutations reported in spontaneous and chemically induced liver tumours of the B6C3F1 mouse provides a potential means of determining in vivo genotoxicity and its relevance to carcinogenicity. We analysed spontaneous and chemically induced [with 4-amino-biphenyl (ABP), 2-acetylaminofluorene (AAF) and diethylnitrosamine (DEN)] hepatocellular tumours of the C57Bl/10J mouse for H-ras, K-ras and N-ras gene mutations to see if mutational analysis of the ras genes could be useful for such a determination in this strain. Regions of DNA spanning codons 12, 13 and 61 of the ras genes were amplified from formalin fixed liver tumour sections using the polymerase chain reaction. Mutations were detected using allele specific oligonucleotide probing and confirmed by sequencing. We have found that there are few ras mutations in either spontaneous or chemically induced liver tumours in the C57Bl/10J mouse. Out of 25 spontaneous tumours two contained an A to T transversion and one contained an A to G transition in base 2 of H-ras codon 61 and two contained a G to A transition in base 2 of K-ras codon 13 (the K-ras mutations were only faintly detectable and may be present in a subpopulation of the tumour cells). In the case of the 18 ABP induced tumours one contained a C to A transversion in base 1 of H-ras codon 61, and one contained an A to T transversion in base 2 of H-ras codon 61 and one contained a G to C transversion in base 1 of K-ras codon 13. One C to A transversion in base 1 of H-ras codon 61 was detected out of eight AAF induced tumours. Of the 25 DEN induced tumours, one contained an A to G transition and one contained an A to C transversion in base 2 of H-ras codon 61. The data indicate that at least in hepatocellular tumours of the C57Bl/10J strain and using chronic dosing regimes the ras genes do not represent markers for in vivo genotoxic activity.

Hepatocarcinogenic effect of vinyl carbamate in the C57Bl/10J strain mouse.

The genotoxic carcinogen vinyl carbamate was dosed to C57Bl/10J strain mice for 35 weeks, and the study terminated after week 59. A main study group of 55 males and 50 females was dosed 6 mg/kg vinyl carbamate once weekly by intraperitoneal injection, whilst a reference group of 10 animals/sex were kept undosed. From week 39 onwards there was a high incidence of mortality, which was often associated with acute internal abdominal hemorrhage. Mice of both sexes killed from 34 weeks onwards frequently showed macroscopic evidence of blood-filled, cyst-like structures in the liver. Upon histopathological examination widespread peliosis hepatis was observed with frequent progression to hemangiomata and hemangiosarcomata. Trabecular hepatocellular carcinomata were also apparent, often within the same liver sections and invariably associated with peliosis hepatis. As a consequence of its tumor burden, the liver often showed hepatocyte atrophy, fibrosis, and coagulation necrosis. A small number of livers revealed hepatocellular adenomata and altered hepatocyte foci.

Effect of butylated hydroxyanisole on the level of DNA adduction by aristolochic acid in the rat forestomach and liver.

Administration of butylated hydroxyanisole (BHA) orally at either 0.5 g or 1 g/kg daily for 14 days to rats did not produce any DNA adducts in the forestomach as measured by the 32P-postlabeling method using (1) limiting concentrations of 32P-ATP; (2) nuclease P1 enhancement; or (3) butanol extraction. Experiments were conducted to establish the effects of BHA administration on aristolochic acid (AA) DNA adduct formation in the forestomach and liver, when BHA was administered prior to, together with or after AA administration. Adduct levels per 10(9) nucleotides in the liver after oral dosing daily for 5 days with 1 mg/kg AA and BHA (1 g/kg) or corn oil (5 ml/kg) for 7 days were as follows: (a) BHA and AA given simultaneously; 235 +/- 71, (b) AA + corn oil; 63 +/- 39, (c) AA followed by BHA; 57 +/- 13, (d) AA followed by corn oil; 91 +/- 38, (e) BHA followed by AA; 90 +/- 12, (f) corn oil followed by AA; 83 +/- 24. For the forestomach the values were: (a) 236 +/- 86, (b) 77 +/- 25, (c) 367 +/- 97, (d) 296 +/- 47, (e) 217 +/- 81, (f) 70 +/- 64. These data suggest that BHA could have an enhancing effect on AA-induced lesions in the forestomach if dosed together with, or prior to, AA as adduct levels are significantly higher than in controls.

Effect of tumour promoters and non-genotoxic carcinogens on terminal differentiation and proliferation in mouse teratoma XB2 cells cultured in low calcium medium.

XB2 cells, a teratocarcinoma derived cell line of keratinocyte lineage, have been shown to proliferate and differentiate in low calcium medium (0.03 mM Ca2+) at a density of 500 cells/9.6 cm2 without the need for fibroblast feeder layers or conditioned medium. The degree of differentiation can be assessed by measurements of keratin production and stratification of colonies of cells. Both of these parameters, as well as the proliferative capacity of the cells, can be altered by treatment of the cultures with various promoting agents and non-genotoxic carcinogens. Treatment with teleocidin and 12-O-tetradecanoyl phorbol-13-acetate induced large increases in proliferation, stratification and keratinization; mezerein-treated cells showed increased stratification at higher doses; butylated hydroxyanisole treatment resulted in hyperkeratinization and hyperstratification to lower levels than that seen with the phorbol-ester-like promoters; butylated hydroxytoluene had purely hyperproliferative effects. We suggest that this culture system may provide a useful model for studies of the mechanism of promotion and non-genotoxic carcinogenesis in epithelial tissues.

Teratogenicity of phenylhydantoins in an in vitro system: molecular orbital-generated quantitative structure-toxicity relationships.

1. The ability of 20 mono- and di-phenylhydantoin derivatives to inhibit differentiation of rat embryo mid-brain and limb bud cells in culture has been used as an index of the teratogenic hazard represented by these compounds. 2. Molecular orbital calculations on these compounds, using the MINDO-3 (modified intermediate neglect of differential overlap) and CNDO-2 (complete neglect of differential overlap) methods, were combined with indices of teratogenicity in the two cell types, to generate a coherent structure-toxicity relationship. 3. Teratogenicity correlated with frontier orbital electron density of the N1 hydantoin ring atom (HOMO-N1) in a sub-series of 12 monophenylhydantoins, whereas the corresponding toxicity for both mono- and di-phenylhydantoins related more to the molecular polarizability (alpha mol) of the molecule. 4. Furthermore the same structural parameter (alpha mol) exhibited a parallelism with log P values of these 20 compounds, indicating the importance of lipophilicity in the toxicity of these compounds. 5. Overall, the data emphasize the ability of electronic structural calculations to identify chemical descriptors of toxicity.