The atmospheric aerosol-forming potential of whole gasoline vapor.

A series of sunlight-irradiated, smog-chamber experiments confirmed that the atmospheric organic aerosol formation potential of whole gasoline vapor cna be accounted for solely in terms of the aromatic fraction of the fuel. The total amount of secondary organic aerosol produced from the atmospheric oxidation of whole gasoline vapor can be represented as the sum of the contributions of the individual aromatic molecular constituents of the fuel. The urban atmospheric, anthropogenic hydrocarbon profile is approximated well by evaporated whole gasoline, and thus these results suggest that it is possible to model atmospheric secondary organic aerosol formation.

Evidence of strong earthquake shaking in the lower wabash valley from prehistoric liquefaction features.

Earthquake-induced liquefaction features in Holocene sediments provide evidence of strong prehistoric shaking, magnitude m(b) 6.2 to 6.7, in the Wabash Valley bordering Indiana and Illinois. The source of the one or more earthquakes responsible was almost certainly in or near the Wabash Valley. The largest event is interpreted to have occurred between 7500 and 1500 years ago on the basis of archeological, pedological, and stratigraphic relations.

Hepatic microsomal glucuronidation of clofibric acid in the adult and neonate albino rat.

Hepatic microsomal UDP glucuronyltransferase activity towards the acid substrate clofibric acid has been described in the adult and neonate albino rat. The enzyme was maximally activated, approximately 2-fold, in the presence of 0.1-0.4% (w/v) digitonin. Induction of the digitonin activated clofibric acid glucuronyltransferase was observed following phenobarbitone treatment in vivo (2.2-fold), and to a lesser extent, following beta-naphthoflavone treatment (1.3-fold). Clofibrate treatment in vivo (of which clofibric acid is the ester hydrolysis product) had no effect on clofibric acid glucuronidation in vitro. The activity of clofibric acid glucuronyltransferase in the liver of rat before and at birth was low (approx. 0.08 nmoles glucuronide formed/min/mg microsomal protein). The activity increased 5-fold during the first three post-natal days. After this time, the activity increased linearly reaching adult levels by four weeks after birth. The data indicated that clofibric acid glucuronyltransferase belongs to the neonatal cluster of enzymes and clofibric acid is a group 2 substrate. Clofibric acid, a common therapeutic agent, is a useful, acid substrate for the estimation of mammalian hepatic microsomal glucuronyltransferase activity.

Obstacles to the prediction of estrogenicity from chemical structure: assay-mediated metabolic transformation and the apparent promiscuous nature of the estrogen receptor.

Information on structure-activity relationships (SAR) and pathways of metabolic activation would facilitate the preliminary screening of chemicals for estrogenic potential. Published crystallographic studies of the estrogen receptor (ER) imply an essential role of the two hydroxyl groups on estradiol (17beta-E(2)) for its binding to ER. The influence of these hydroxyl groups on ER binding and estrogenicity was evaluated by the study of 17beta-E(2) with one or both of these hydroxyl groups removed (17beta-desoxyestradiol and 3, 17beta-bisdesoxyestradiol, respectively). 6-Hydroxytetralin (17beta-E(2) with its C- and D-rings removed) and other synthetic estrogens were also studied. The estrogenicity assays comprised a yeast ER-mediated transcription assay, mammalian cell transcription assays incorporating either ER alpha or ER beta, and the immature rat uterotrophic assay. With the exception of 6-hydroxytetralin in the uterotrophic assay, all the chemicals were active in all the assays. Hydroxylation of the two desoxy compounds to estradiol was shown to occur in immature female rats, but metabolism was not implicated in the responses observed in the ER-binding and yeast systems. It is concluded that the 3-hydroxyl and 17beta-hydroxyl groups of 17beta-E(2) are not absolute requirements for estrogenicity. It would therefore be of value to the derivation of SAR for estrogenicity were the crystal structure of the bisdesoxy-E(2)/ER complex to be evaluated.

Neonatal exposure of male rats to nonylphenol has no effect on the reproductive tract.

P. C. Lee (1998, ENDOCRINE: 9, 105-111) has reported that neonatal exposure of SD rats to nonylphenol (NP; 8 mg/kg/day) by daily intraperitoneal (ip) injection in DMSO results in decreased ventral prostate and epididymides weights, and delayed testes descent, at post natal day (pnd) 31. These effects were surprising given that similar effects were not reported in an earlier multi-generation study of NP. We have repeated the central experiment described by Lee and were unable to confirm the effects reported. Alpk (Wistar derived) rats were exposed to NP (8mg/kg/day by ip injection in either arachis oil or DMSO) from pnd 1-10 and assessed on pnd 34-36. No significant effects on animal body weights were observed. The weights of the epididymides, seminal vesicles, testes, and ventral prostate were unaffected using either vehicle. Testes descent proceeded normally, with both test and control testes fully descended by pnd 29. Possible reasons for this divergence in findings for NP are discussed.

Gene expression changes in the immature rat uterus: effects of uterotrophic and sub-uterotrophic doses of bisphenol A.

J. C. Gould et al., 1998, Mol. Cell Endocrinol. 142, 203-214, have reported that administration of 5-150 mg/kg/day BPA to immature rats leads to increases in uterine peroxidase activity and progesterone receptor (PR) protein levels in the absence of a uterotrophic response. These observations are of interest given current concerns regarding the adequacy of the uterotrophic assay to act as a sentinel for the estrogenic activity of chemicals in vivo. Therefore, the uterotrophic activity of BPA to the immature rat has been re-evaluated over the dose range 2 microg/kg-800 mg/kg/day. Expression levels of three estrogen responsive uterine genes were determined using real-time RT-PCR--namely, complement component 3, lipocalin 2, and PR. 18S rRNA and RNA polymerase II large subunit acted as control genes. Observations of gene expression were made 4 h and 72 h after the first of three daily po administrations of BPA. Increases in gene expression were observed over the uterotrophic dose range (approximately 200-800 mg/kg BPA). Over the dose range 2 microg/kg-20 mg/kg BPA there was no uterotrophic response and no increase in gene expression. We conclude that BPA does not produce reproducible changes in gene expression in the uterus of immature rats at dose levels that are not also uterotrophic. Therefore, in the present study, the no effect level for uterotrophic activity for BPA coincided with the no transcriptional effect level for uterine genes.

Comparison of the developmental and reproductive toxicity of diethylstilbestrol administered to rats in utero, lactationally, preweaning, or postweaning.

The objective of the study was to determine which period of exposure produces the most marked effects on the reproductive capacity and sexual development of the rat, with particular emphasis on the relative sensitivity of in utero and postnatal exposures. The endocrine active chemical, diethylstilbestrol (DES) was used as an agent known to affect many of the endpoints examined. Hitherto, such comparisons have been made between studies, rather than within a study. Our data will be helpful in the interpretation of future multigenerational assay data. In preliminary studies, DES was shown to be active in the immature rat uterotrophic assay with a lowest detected dose of 0.05 mg DES/kg body weight by sc injection and 10 mg DES/l (1.6 mg DES/kg body weight) by administration in drinking water. A dose of 60 microg DES/l drinking water ( approximately 6.5mg DES/kg body weight/day) was selected for the main study since this represented the midpoint of the drinking water uterotrophic dose response and produced no overt maternal toxicity. The study used 10 groups of concomitantly pregnant animals, including 2 control groups. The first comparison was between the effects of exposure to DES in utero, and exposure from conception to weaning. Another group of animals was exposed to DES in utero and cross-fostered to untreated pregnant females to prevent lactational transfer of DES to pups. Two groups were exposed to DES neonatally, either from birth to postnatal day (PND) 10 (pups thus having only lactational exposure), or from birth until weaning (PND 21; pups thus having both lactational exposure and self-exposure via drinking water). In addition, a dose response study to DES was conducted on animals exposed from weaning to PND 100, when the first phase of the study was terminated. Pups exposed to DES in utero and pups exposed from weaning to PND 100 were bred to assess fertility of the F1 animals and the sexual development of F2 offspring. This last comparison was to determine the extent to which weanling rats could be used in endocrine toxicity studies to assess their potential to show activity in utero. The most sensitive period of exposure for inducing developmental effects in F1 animals was from weaning onwards. The neonatal to weaning period (PND 1-21) was the next most sensitive. Essentially no effects were induced in F1 animals exposed in utero. No effects of any kind were observed in animals only exposed over the early neonatal period of PND 1-10. The mean day of vaginal opening, testes descent, and prepuce separation was only altered in groups where postnatal exposure to DES continued beyond PND 10, or was started at weaning. No changes were observed in anogenital distance or caudal sperm counts. Some changes in organ weights were observed, but the interpretation of these was often confused by concomitant changes in body weight. In general, histopathological examination of tissues yielded no additional information. In breeding studies with animals exposed to DES in utero, or from weaning, reduced litter sizes and marginal advances in the day of vaginal opening were observed in the offspring, together with changes in organ weights. However, no unique sensitivity was noted for exposure in utero. Evaluation of the several exposure periods and the many markers monitored in this study may have individual strengths in individual cases, but when rigorously compared using the reference estrogen DES, many preconceptions regarding their absolute or relative value were not upheld. Further, each of these markers is subject to natural variability, as demonstrated by comparisons made among the 5 separate control groups available in parts of the present study. This variability increases the chance that small changes observed in endocrine toxicity studies employing small group sizes and a single control group, or no concomitant control group, may be artifactual. The most marked effects observed in this study were on the developmental landmarks in the F1 animals induced by exposures after PND 10. Some effects on developmental landmarks and organ weights were observed in F2 animals following exposure either in utero or postweaning. This study therefore does not establish a unique role for exposures in utero or during the early neonatal period.

Effect of rodent diets on the sexual development of the rat.

Five rodent diets have been evaluated for their possible effect on the sexual development of the rat. Groups of 12 pregnant Alpk rats were fed one of the following combinations of diets during pregnancy and postnatally: RM3/RM1, AIN-76A/AIN-76A, RM3/AIN-76A, Teklad Global 2016 (Global)/Global and Purina 5001/Purina 5001. AIN-76A is phytoestrogen-free while the other diets contained varying amounts of phytoestrogens. The phytoestrogens genistein and daidzein were determined in the diets studied, and the concentrations found agreed with earlier estimates. RM3/RM1 was selected as the control group, as this has been used routinely in this laboratory for the past decade. Determinations were made in offspring of the times of vaginal opening and first estrus among the females, and of prepuce separation and testes descent among the males. At postnatal day (PND) 26 the females from 6 of the 12 litters were terminated and tissue weights measured. Males from 6 of the 12 litters were similarly studied at PND 68. Animals from the remaining litters were transferred to RM1 diet at PND 70. Termination of the study was at PND 128 (males) and PND 140 (females) when body weights and tissue weights were determined. Marked differences in body weight, sexual development, and reproductive tissue weights were observed for rats maintained on AIN-76A or Purina 5001, with only minimal effects among rats maintained on the Global diet. These comparisons were against RM3/RM1 as the reference diet. However, using Purina 5001 as the reference diet reversed the direction of the differences seen when using RM3/RM1 as the reference diet. The differences observed when using RM3/RM1 as reference diet occurred mainly postnatally. In addition, the fact that similar differences were seen for the phytoestrogen-free diet, AIN-76A, and the phytoestrogen-rich diet, Purina 5001, indicate that these effects are more likely to be caused by nutritional differences between the diets that then have centrally mediated effects on rodent sexual development, rather than individual dietary components affecting peripheral estrogen receptors (ER). This proposal is supported by abolition of the uterotrophic activity of AIN-76A and Purina 5001 (relative to RM3/RM1) in the immature rat by coadministration of the gonadotrophin-releasing hormone (GnRH) antagonist ANTARELIX: The present data indicate that choice of diet may influence the timing of sexual development in the rat, and consequently, that when evaluating the potential endocrine toxicity of chemicals, the components of rodent diets used should be known, and as far as is possible, controlled.

Structure/activity studies of the nephrotoxic and mutagenic action of cysteine conjugates of chloro- and fluoroalkenes.

The cysteine conjugates of the nephrotoxins hexachlorobutadiene (HCBD), tetrafluoroethylene (TFE) and hexafluoropropene (HFP), together with those of trichloroethylene and perchloroethylene, have been chemically synthesized and a relationship determined between their structures and their nephrotoxicity and mutagenicity in vitro. All of the conjugates had a marked effect on the uptake of both the organic anion p-aminohippuric acid (PAH) and the cation tetraethylammonium bromide (TEA) into rat kidney slices, suggesting activation of the conjugates in the slices to a toxic species which interferes with ion transport. This observation is consistent with the known nephrotoxicity of HCBD, TFE and HFP in vivo. Each of the conjugates was found to be metabolised by rat kidney slices and by semi-purified rat kidney beta-lyase to pyruvate, ammonia and an unidentified reactive metabolite. When semi-purified beta-lyase was used stoichiometric amounts of pyruvate and ammonia were produced. Although all of the conjugates were activated by beta-lyase and had a similar effect on ion transport their mutagenicity differed markedly. The conjugates of HCBD, trichloroethylene and perchloroethylene were mutagenic in the Ames bacterial mutation assay when activated by rat kidney S9. Metabolic cofactors were not required suggesting that activation was due to the enzyme beta-lyase. In the same assay the conjugates of TFE and HFP were not mutagenic either in the presence or absence of rat kidney S9 and cofactors. With a limited number of cysteine conjugates a clear distinction has been identified between the conjugates of chloroalkenes which were were similarly nephrotoxic but were not mutagenic. The mutagenicity of the cysteine conjugate of HCBD is consistent with the known renal carcinogenicity of this chemical.

A mechanism for the development of Clara cell lesions in the mouse lung after exposure to trichloroethylene.

Female CD-1 mice exposed to trichloroethylene (6 h/day) at concentrations from 20-2000 ppm developed a highly specific lung lesion after a single exposure, characterised by vacuolation of the Clara cells, the number of cells affected increasing with increasing dose level. At the highest dose levels pyknosis of the Clara cells was apparent. After 5 days of repeated exposures the lesion had resolved but exposure of mice following a 2-day break resulted in recurrence of the lesion. The changes in mouse lung Clara cells were accompanied by a marked loss of cytochrome P-450 activities. No morphological changes were seen in the lungs of rats exposed to either 500 or 1000 ppm trichloroethylene. Isolated mouse lung Clara cells were shown to metabolize trichloroethylene to chloral, trichloroethanol and trichloroacetic acid. Chloral was the major metabolite. Trichloroethanol glucuronide was not detected. In comparative experiments using mouse hepatocytes the major metabolites were trichloroethanol and its glucuronide conjugate. The activity of UDP-glucuronosyltransferase was compared in mouse lung Clara cells and hepatocytes using two phenolic substrates and trichloroethanol. Hepatocytes readily formed glucuronides from all three substrates whereas Clara cells were only active with the two phenolic substrates. The three major metabolites of trichloroethylene, chloral, trichloroethanol and trichloroacetic acid were each dosed to mice and of these metabolites, only chloral had an effect on mouse lung causing a lesion (Clara cell) identical to that seen with trichloroethylene. It is proposed that the failure of Clara cells to conjugate trichloroethanol leads to an accumulation of chloral which results in cytotoxicity. The known genotoxicity of chloral suggests that this lesion may be related to the development of lung tumours in mice exposed to trichloroethylene by inhalation.

The role of glutathione conjugation in the development of kidney tumours in rats exposed to trichloroethylene.

Trichloroethylene is metabolised to a very minor extent (< 0.01% of the dose) by conjugation with glutathione, a metabolic pathway which leads to the formation of S-(1,2-dichlorovinyl)-L-cysteine (DCVC), a bacterial mutagen and nephrotoxin activated by the renal enzyme beta-lyase. The role of this metabolic pathway in the development of the nephrotoxicity and subsequent tumour formation seen in rats exposed to trichloroethylene has been evaluated. The pathway has been assessed quantitatively in vivo in rats, and in rats, mice and humans in vitro. Trichloroethylene was found to be a very weak nephrotoxin. There was no evidence of morphological change in the kidneys and only small increases in biochemical markers of kidney damage in rats dosed with 2000 mg/kg trichloroethylene by gavage for 42 days. N-acetyl-S-(1,2-dichlorovinyl)-L-cysteine was detected in the urine of rats dosed with 500 and 2000 mg/kg trichloroethylene for up to 10 days at levels equivalent to 0.001-0.008% of the dose. In vitro, the rate of conjugation of trichloroethylene with glutathione in the liver was higher in the mouse, 2.5 pmol/min per mg protein, than the rat, 1.6 pmol/min per mg protein, and in human liver the rates were extremely low, 0.02-0.37 pmol/min per mg protein. Comparisons of the metabolism of DCVC by renal beta-lyase and N-acetyl transferase showed that metabolism by N-acetyl transferase was two orders of magnitude greater than that by beta-lyase and that beta-lyase activity in rat kidney was 11-fold greater than that in human kidney. When the nephrotoxicity of DCVC was compared in rats and mice, the mouse was found to be 5-10 fold more sensitive than the rat. The no effect level in the rat was 10 mg/kg, a dose which is three orders of magnitude higher than the amount of DCVC formed from trichloroethylene in vivo. The lack of correlation between metabolism by this pathway and the rat specific tumours, together with questions concerning the potency of DCVC at the levels formed from trichloroethylene, suggests that DCVC may not be involved in the renal toxicity and subsequent tumour development seen in rats and that further evaluation of the mechanism(s) involved in the nephrotoxic response is warranted.

Detection of aromatase inhibitors in vitro using rat ovary microsomes.

Inhibition of aromatase activity in vitro is one of the Tier 1 screening assays proposed by the Endocrine Disrupter Screening and Testing Advisory Committee (EDSTAC) for the detection of potential endocrine disrupters. In this report a rat ovarian aromatase inhibition assay has been evaluated using the reference aromatase inhibitors anastrozole, fadrozole, letrozole and CGS 18320B. Rat ovary microsomes were used as the enzyme source, as endocrine disruption studies are most commonly carried out in this species. Aromatase activity was inhibited in vitro by anastrozole, fadrozole, letrozole and CGS 18320B with IC(50)s of 25, 7, 7 and 5 nM, respectively. This assay, therefore, appears to have good sensitivity to aromatase inhibitors and may be useful as a general screening assay and in mechanistic studies.

The oral administration of polysorbate 80 to the immature female rat does not increase uterine weight.

Some xenobiotics display estrogenic activity in in vitro and/or in vivo systems. Previous studies by Gajdova et al. have shown that polysorbate 80 (also known as Tween 80) administered by intraperitoneal injection to neonatal female rats on days 4-7 after birth produced estrogenic effects including earlier vaginal opening, prolongation of the estrus cycle and persistent vaginal estrus. Some of these effects were evident many weeks after cessation of administration of polysorbate 80. The present study has evaluated the estrogenic properties of polysorbate 80 following oral administration, a route of exposure which is more relevant for the consideration of human health hazard. The effects of polysorbate 80 at oral gavage doses of up to 5 g/kg/day for 3 consecutive days on uterine growth of immature female rats, a commonly used in vivo mammalian assay for estrogenic activity, have been determined. Estradiol benzoate administered subcutaneously was used as a positive control and significantly increased uterine weight in this age and strain of female rat (21-23 days, Alpk:APfSD Wistar derived) by up to 4.5-fold above vehicle control values. Polysorbate 80 administered orally to rats had no effect on uterine weight. Thus, intrinsic estrogenic effects of polysorbate 80 reported following its intraperitoneal injection to neonatal 4-day-old female rats are not manifest when it is administered by oral gavage to immature 20- to 22-day-old female rats. This latter route of exposure is of more relevance to human exposure scenarios and these data are, therefore, important in assessing hazard/risk of polysorbate 80 to man.

The uterotrophic activity of nonylphenol in the rat is not mediated by aromatase enzyme induction.

p-Nonylphenol (NP) is weakly estrogenic to rodents and to some species of fish. All evidence to date has indicated that the estrogenic effects of NP are due to the interaction of NP with the estrogen receptor. Recent findings of increased plasma estradiol in fish exposed to NP have, however, led to the proposal of an alternative mechanism for NP-induced estrogenicity in this species, possibly via induction of aromatase enzymes. In the present studies, this hypothesis was investigated in rats using the aromatase inhibitor anastrozole. The results indicated that the uterotrophic action of NP, as with estradiol used as a positive control, is mediated directly by its interaction with uterine ER, rather than an indirect effect via aromatase enzyme induction. Circulating levels of estradiol were unchanged after NP treatment and the aromatase inhibitor anastrozole failed to inhibit NP-induced uterine growth. These results are consistent with previous published data on NP in rodents.

Re-evaluation of the first synthetic estrogen, 1-keto-1,2,3, 4-tetrahydrophenanthrene, and bisphenol A, using both the ovariectomised rat model used in 1933 and additional assays.

1-Keto-1,2,3,4-tetrahydrophenanthrene (THP-1) was reported by Cook et al. in 1933 as the first synthetic estrogen. Estrogenic activity was assessed by the induction of vaginal cornification in ovariectomised rats. The corresponding 4-isomer (THP-4) was shown to be inactive. Both chemicals have been re-synthesised and assessed for hormonal activity. Each chemical bound weakly and to the same extent to isolated estrogen receptors, but only at high concentrations. However, they each lacked estrogenic or anti-estrogenic activity when evaluated in vitro using a yeast hER assay, and both failed to induce vaginal cornification or uterotrophic effects in ovariectomised rats. THP-1, and to a lesser extent THP-4, were shown to possess weak androgenic and anti-androgenic activity in vitro when evaluated using an hAR yeast assay. Estrogenic activity for bisphenol A (BPA) was subsequently demonstrated by [Dodds and Lawson, Synthetic, oestrogenic agents without the phenanthrene nucleus, Nature 137, (1936)] using the same ovariectomised rat protocol, and this activity has been confirmed and supplemented by positive uterotrophic effects for BPA in the same bioassays. The present results illustrate the complexity of deriving conclusions regarding the hormonal activities of chemicals. First, some activities observed in isolated hormonal receptor binding assays may not be expressed in functional hormonal assays. This indicates the need for functional hormonal assays in any screening programme. Second, that activities observed for a chemical in one hormonal assay may not be reflected in related hormonal assays. This indicates the need to define assay protocols with some precision when incorporating them into screening batteries. Finally, that some literature reports of hormonal activity for chemicals may not be capable of independent confirmation under apparently identical conditions of test. This illustrates the need to use lists of hormonally active chemicals with care.

Lack of activity of estradiol in rodent bone marrow micronucleus assays.

Estradiol has been evaluated in five independent rodent bone marrow micronucleus assays and has been found to be inactive. The dose-range evaluated extended from three daily doses of 20 micrograms/kg to the rat, a regimen that elicited a potent uterotrophic response from the animals, to single doses of between 10-150 mg/kg to the mouse. The mouse assays simulated and extended the conditions of test employed by earlier investigators who had found estradiol, and three structurally-related synthetic estrogens, to be active in mouse micronucleus assays over the dose range 1-10 mg/kg. It is concluded that estradiol is not genotoxic to the bone marrow of rodents. The top dose-level used in the present micronucleus assays (150 mg/kg) represented approximately 150,000 times the minimum estrogenic dose of this chemical to rodents, and that was considered to be above the dose at which useful genetic toxicity data would be generated for this potent estrogen. The maximum tolerated dose (MTD) of estradiol to rodents remains to be established.

Limitations of the mcf-7 cell proliferation assay for detecting xenobiotic oestrogens.

The MCF-7 cell assay has recently been proposed as a primary screen for the detection of xenobiotic oestrogens. The assay is based on the ability of human breast cancer cells (MCF-7) to proliferate in response to oestrogens. The use of this assay has a number of known drawbacks, such as the differing responses of cells from different sources. This paper describes the sources of variability which we have encountered with this assay. There was significant variability between experiments in the response of MCF-7 cells to oestradiol, from twofold to sevenfold increases, sometimes with no evidence of a dose response. The background rate of cell proliferation (in the absence of oestradiol) also varied between experiments, from a low basal rate in some experiments to rates approaching that of oestradiol-treated cells in others. The variability in our experiments may be due to selection of sub-clones of cells at passage. We also determined the effects of oestrogens on the non-oestrogen responsive human breast cancer cell line MDA-MB-231. These cells did not proliferate in response to oestradiol, but nonylphenol promoted a two- to threefold increase in cell number. Our data illustrate the potentially unpredictable nature of breast cancer cell lines when used routinely as a screening test for xenobiotic oestrogens.

Mechanistically-based human hazard assessment of peroxisome proliferator-induced hepatocarcinogenesis.

In this review we have evaluated the relationship between peroxisome proliferation and hepatocarcinogenesis. To do so, we identified all chemicals known to produce peroxisome proliferation and selected those for which there are data (on peroxisome proliferation and hepatocarcinogenesis) which meet certain criteria chosen to facilitate comparison of these phenomena. The summarised data and definition of the methodology used has been collected in appendices. These comparisons enabled us to evaluate the relationship between these phenomena using reliable data. As there is a good correlation between them, we further explored the mechanisms of action that have been proposed (direct genotoxic activity, production of hydrogen peroxide, cell proliferation and receptor activation). The relationship between these events in other species, including humans, was also reviewed and finally an overview of the assessment of human hazard is presented in section IX. Some of the first chemicals which were shown to produce peroxisome proliferation were also hepatocarcinogens whose carcinogenicity could not be readily explained by genotoxic activity. This raised the suggestion that the unusual phenomenon of peroxisome proliferation was intricately linked to the carcinogenic activity of these agents. Three questions have exercised the attention of regulatory, industrial and academic toxicology since then; are chemicals which elicit peroxisome proliferation in the liver actually a coherent class of chemical carcinogens?; does the early biological phenomenon of peroxisome proliferation have real predictive value for and mechanistic association with rodent carcinogenesis?; and what hazard/risk do these agents pose to humans that may be exposed to them? Whether peroxisome proliferators are indeed a discrete class of rodent carcinogens would appear to be the single, most important question. If so, then the assumptions and procedures relevant to human hazard and risk assessment should be applied to the class and should be essentially generic; if not, each chemical should be considered independently. Our critical analysis of the published data for over 70 agents which have been shown to possess intrinsic ability to induce peroxisome proliferation in the livers of rodents has led to the conclusion that there exists a strong correlation between peroxisome proliferation as n early effect in the liver and hepatocarcinogenicity in chronic exposure studies. An almost perfect correlation was observed between the induction of peroxisomes in the rodent liver and the eventual appearance of tumours following chronic exposure The few exceptions to this were largely explainable (section II).(ABSTRACT TRUNCATED AT 400 WORDS)