Effects of 0.5% and 2.0% Sodium Lauryl Sulfate in Male CD-1 Mice From a 3-Month Oral Gavage Toxicity Study.

The tolerability of single daily gavage doses of 0.5% or 2.0% (wt/vol) sodium lauryl sulfate (SLS) in 11- to 12-week-old male CD-1 mice was evaluated in a study of 3 months in duration. Live-phase, gross necropsy, and histopathologic parameters were evaluated. Mortality of 14% occurred in mice administered formulations containing SLS. Clinical observations in mice administered SLS included abnormal respiration (audible, irregular, and/or labored), swollen abdomen, rough haircoat, hunched appearance, and hypoactivity. Necropsy findings in mice administered SLS consisted of enlarged intestines containing abnormal contents with gas. There were no instances of mechanical gavage-related injury. Histologic evaluation of the respiratory tract revealed injury to the nasal passages and nasopharynx, including, but not limited to, inflammation, exudate, apoptosis/necrosis of epithelium, and atrophy of epithelium or olfactory nerves. Collectively, the data indicated that under the experimental conditions of our 3-month study in male CD-1 mice, once-daily gavage administration of vehicle formulations containing SLS at 0.5% or 2.0% resulted in nasal injury and 14% mortality supportive of gastroesophageal reflux. Sponsors utilizing formulations containing SLS in toxicity studies in CD-1 mice should exclude gastroesophageal reflux as a confounding factor in studies with morbidity or mortality associated with respiratory distress or evidence of aerophagia.

Reevaluation of Hepatocellular Neoplasms in CD-1 Mice from a 2-year Oral Carcinogenicity Study with Permethrin.

A 24-month oral carcinogenicity study of permethrin was conducted by feeding male and female CD-1 mice diets containing concentrations of 0, 20, 500, and 2,000 ppm of permethrin (males) or 0, 20, 2,500, and 5,000 ppm of permethrin (females). After approximately two years on study, surviving mice were sacrificed for the evaluation of chronic toxicity and/or carcinogenicity. An expert panel of pathologists was convened as a Pathology Working Group (PWG) to review coded liver histology sections from male and female mice and to classify all liver neoplasms according to current nomenclature and diagnostic criteria guidelines. The PWG results indicate that permethrin induced a significant dose-dependent increase in the incidence of hepatocellular neoplasms in treated female mice ( p < .01) as well as a nonstatistically significant increase in the incidence of hepatocellular tumors in treated male mice. Given the continuum of the diagnoses of adenoma and carcinoma, and the difficulty in distinguishing some of the lesions, it is appropriate to consider only the combined incidences of hepatocellular tumors (adenoma and/or carcinoma) for biological significance and risk assessment.

Regulatory Forum Opinion Piece*: The Value of Publishing Negative Scientific Study Data.

Historically it has been easier to publish positive scientific results than negative data not supporting the research hypothesis. This appears to be increasing, with fewer negative studies appearing in the literature across many disciplines. Failure to recognize the value of negative results has important implications for the toxicology community. Implications include perpetuating scientific fields based upon selective or occasionally erroneous, positive results. One example is decreased vaccination rates and increased measles infections that can lead to childhood mortality following one erroneous positive study linking vaccination to adverse effects despite multiple negative studies. Publication of negative data that challenges existing paradigms enhances progress by stopping further investment in scientifically barren topics, decreases the use of animals, and focuses research in more fruitful areas. The National Toxicology Program (NTP) publishes both positive and negative rodent data. Retrospective analysis of the NTP database has provided insights on the carcinogenic process and in the gradual acceptance of using fewer animals in safety studies. This article proposes that careful publication of both positive and negative data can enhance product safety assessment, add robustness to safety determinations in the regulatory decision-making process, and should be actively encouraged by those determining journal editorial policy.

Distinguishing cystic degeneration from other aging lesions in the adrenal cortex of Sprague-Dawley rats.

Cystic degeneration of the adrenal cortex is a common age-related finding in the Sprague-Dawley (SD) rat strain occurring more frequently in females. Compression of the adjacent cortex, a common hallmark of benign adrenal cortical tumors, often accompanies foci of cystic degeneration, creating a diagnostic challenge. Accurately differentiating these relatively common degenerative changes from proliferative lesions is critical in safety assessment studies. Cystic degeneration typically arises in the zona fasciculata of the adrenal cortex and often causes compression along the margin of the lesion. The degenerating cells are large, with abundant eosinophilic cytoplasm, or contain clear cytoplasmic vacuoles. Mitotic figures are generally uncommon. In many cases, cystic degeneration appears to arise in areas of hypertrophy in the zona fasciculata. In contrast, adrenal cortical hyperplasia and adrenal cortical adenoma are frequently comprised of smaller cells that cause compression of adjacent cortex, and in some cases mitotic figures are observed. Cytological detail and growth patterns should be considered more useful criteria than compression alone for separating degenerative cystic lesions from proliferative lesions in the adrenal cortex of SD rats.

Molecular mechanisms of fibrosis-associated promotion of liver carcinogenesis.

Hepatocellular carcinoma (HCC) mostly develops in patients with advanced fibrosis; however, the mechanisms of interaction between a genotoxic insult and fibrogenesis are not well understood. This study tested a hypothesis that fibrosis promotes HCC via a mechanism that involves activation of liver stem cells. First, B6C3F1 mice were administered diethylnitrosamine (DEN; single ip injection of 1mg/kg at 14 days of age). Second, carbon tetrachloride (CCl(4); 0.2ml/kg, 2/week ip starting at 8 weeks of age) was administered for 9 or 14 weeks to develop advanced liver fibrosis. In animals treated with DEN as neonates, presence of liver fibrosis led to more than doubling (to 100%) of the liver tumor incidence as early as 5 months of age. This effect was associated with activation of cells with progenitor features in noncancerous liver tissue, including markers of replicative senescence (p16), oncofetal transformation (Afp, H19, and Bex1), and increased "stemness" (Prom1 and Epcam). In contrast, the dose of DEN used did not modify the extent of liver inflammation, fibrogenesis, oxidative stress, proliferation, or apoptosis induced by subchronic CCl(4) administration. This study demonstrates the potential role of liver stem-like cells in the mechanisms of chemical-induced, fibrosis-promoted HCC. We posit that the combination of genotoxic and fibrogenic insults is a sensible approach to model liver carcinogenesis in experimental animals. These results may contribute to identification of cirrhotic patients predisposed to HCC by analyzing the expression of hepatic progenitor cell markers in the noncancerous liver tissue.

Acetaminophen-induced acute liver injury in HCV transgenic mice.

The exact etiology of clinical cases of acute liver failure is difficult to ascertain and it is likely that various co-morbidity factors play a role. For example, epidemiological evidence suggests that coexistent hepatitis C virus (HCV) infection increased the risk of acetaminophen-induced acute liver injury, and was associated with an increased risk of progression to acute liver failure. However, little is known about possible mechanisms of enhanced acetaminophen hepatotoxicity in HCV-infected subjects. In this study, we tested a hypothesis that HCV-Tg mice may be more susceptible to acetaminophen hepatotoxicity, and also evaluated the mechanisms of acetaminophen-induced liver damage in wild type and HCV-Tg mice expressing core, E1 and E2 proteins. Male mice were treated with a single dose of acetaminophen (300 or 500 mg/kg in fed animals; or 200 mg/kg in fasted animals; i.g.) and liver and serum endpoints were evaluated at 4 and 24h after dosing. Our results suggest that in fed mice, liver toxicity in HCV-Tg mice is not markedly exaggerated as compared to the wild-type mice. In fasted mice, greater liver injury was observed in HCV-Tg mice. In fed mice dosed with 300 mg/kg acetaminophen, we observed that liver mitochondria in HCV-Tg mice exhibited signs of dysfunction showing the potential mechanism for increased susceptibility.

Increased incidence of aflatoxin B1-induced liver tumors in hepatitis virus C transgenic mice.

Viral hepatitis and aflatoxin B1 (AFB1) exposure are common risk factors for hepatocellular carcinoma (HCC). The incidence of HCC in individuals coexposed to hepatitis C (HCV) or B virus and AFB1 is greater than could be explained by the additive effect; yet, the mechanisms are poorly understood because of the lack of an animal model. Our study investigated the outcomes and mechanisms of combined exposure to HCV and AFB1. We hypothesized that HCV transgenic (HCV-Tg; expressing core, E1, E2 and p7, nucleotides 342-2771) mice will be prone to hepatocarcinogenesis when exposed to AFB1. Neonatal (7 days old) HCV-Tg or C57BL/6J wild-type (WT) mice were exposed to AFB1 (6 µg/g bw) or tricaprylin vehicle (15 µl/g bw), and male offspring were followed for up to 12 months. No liver lesions were observed in vehicle-treated WT or HCV-Tg mice. Tumors (adenomas or carcinomas) and preneoplastic lesions (hyperplasia or foci) were observed in 22.5% (9 of 40) of AFB1-treated WT mice. In AFB1-treated HCV-Tg mice, the incidence of tumorous or pretumorous lesions was significantly elevated (50%, 18 of 36), with the difference largely due to a 2.5-fold increase in the incidence of adenomas (30.5 vs. 12.5%). Although oxidative stress and steatohepatitis were observed in both AFB1-treated groups, molecular changes indicative of the enhanced inflammatory response and altered lipid metabolism were more pronounced in HCV-Tg mice. In summary, HCV proteins core, E1, E2 and p7 are sufficient to reproduce the cocarcinogenic effect of HCV and AFB1, which is a known clinical phenomenon.

Predicting the hepatocarcinogenic potential of alkenylbenzene flavoring agents using toxicogenomics and machine learning.

Identification of carcinogenic activity is the primary goal of the 2-year bioassay. The expense of these studies limits the number of chemicals that can be studied and therefore chemicals need to be prioritized based on a variety of parameters. We have developed an ensemble of support vector machine classification models based on male F344 rat liver gene expression following 2, 14 or 90 days of exposure to a collection of hepatocarcinogens (aflatoxin B1, 1-amino-2,4-dibromoanthraquinone, N-nitrosodimethylamine, methyleugenol) and non-hepatocarcinogens (acetaminophen, ascorbic acid, tryptophan). Seven models were generated based on individual exposure durations (2, 14 or 90 days) or a combination of exposures (2+14, 2+90, 14+90 and 2+14+90 days). All sets of data, with the exception of one yielded models with 0% cross-validation error. Independent validation of the models was performed using expression data from the liver of rats exposed at 2 dose levels to a collection of alkenylbenzene flavoring agents. Depending on the model used and the exposure duration of the test data, independent validation error rates ranged from 47% to 10%. The variable with the most notable effect on independent validation accuracy was exposure duration of the alkenylbenzene test data. All models generally exhibited improved performance as the exposure duration of the alkenylbenzene data increased. The models differentiated between hepatocarcinogenic (estragole and safrole) and non-hepatocarcinogenic (anethole, eugenol and isoeugenol) alkenylbenzenes previously studied in a carcinogenicity bioassay. In the case of safrole the models correctly differentiated between carcinogenic and non-carcinogenic dose levels. The models predict that two alkenylbenzenes not previously assessed in a carcinogenicity bioassay, myristicin and isosafrole, would be weakly hepatocarcinogenic if studied at a dose level of 2 mmol/kg bw/day for 2 years in male F344 rats; therefore suggesting that these chemicals should be a higher priority relative to other untested alkenylbenzenes for evaluation in the carcinogenicity bioassay. The results of the study indicate that gene expression-based predictive models are an effective tool for identifying hepatocarcinogens. Furthermore, we find that exposure duration is a critical variable in the success or failure of such an approach, particularly when evaluating chemicals with unknown carcinogenic potency.

Mouse population-guided resequencing reveals that variants in CD44 contribute to acetaminophen-induced liver injury in humans.

Interindividual variability in response to chemicals and drugs is a common regulatory concern. It is assumed that xenobiotic-induced adverse reactions have a strong genetic basis, but many mechanism-based investigations have not been successful in identifying susceptible individuals. While recent advances in pharmacogenetics of adverse drug reactions show promise, the small size of the populations susceptible to important adverse events limits the utility of whole-genome association studies conducted entirely in humans. We present a strategy to identify genetic polymorphisms that may underlie susceptibility to adverse drug reactions. First, in a cohort of healthy adults who received the maximum recommended dose of acetaminophen (4 g/d x 7 d), we confirm that about one third of subjects develop elevations in serum alanine aminotransferase, indicative of liver injury. To identify the genetic basis for this susceptibility, a panel of 36 inbred mouse strains was used to model genetic diversity. Mice were treated with 300 mg/kg or a range of additional acetaminophen doses, and the extent of liver injury was quantified. We then employed whole-genome association analysis and targeted sequencing to determine that polymorphisms in Ly86, Cd44, Cd59a, and Capn8 correlate strongly with liver injury and demonstrated that dose-curves vary with background. Finally, we demonstrated that variation in the orthologous human gene, CD44, is associated with susceptibility to acetaminophen in two independent cohorts. Our results indicate a role for CD44 in modulation of susceptibility to acetaminophen hepatotoxicity. These studies demonstrate that a diverse mouse population can be used to understand and predict adverse toxicity in heterogeneous human populations through guided resequencing.

Evaluation of dichloroacetic acid for carcinogenicity in genetically modified Tg.AC hemizygous and p53 haploinsufficient mice.

There has been considerable interest in the use of genetically modified mice for detecting potential environmental carcinogens. For this reason, the National Toxicology Program has been evaluating Tg.AC hemizygous and p53 haploinsufficient mice as models to detect potential carcinogens. It was reasoned that these mouse models might also prove more effective than standard rodent models in evaluating the numerous disinfection byproducts that are found in low concentrations in drinking water. Dichloroacetic acid (DCA) is one of the most frequently found disinfection byproducts and DCA has been consistently shown to cause hepatocellular tumors in rats and mice in standard rodent studies. Tg.AC hemizygous and p53 haploinsufficient mice were exposed in the drinking water to DCA for up to 41 weeks. In a second study Tg.AC mice were subjected to dermal DCA exposure for up to 39 weeks. Increased incidences and severity of cytoplasmic vacuolization of hepatocytes were seen in the p53 mice, but there was no evidence of carcinogenic activity at exposures of up to 2000 mg/l in the drinking water. Increased incidences and severity of cytoplasmic vacuolization of hepatocytes were seen in the drinking water study with Tg.AC mice and a modest non-dose-related increase in pulmonary adenomas was observed in males exposed to 1000 mg/l in the drinking water. Dermal exposure up to 500 mg/kg for 39 weeks resulted in increased dermal papillomas at the site of application in Tg.AC mice. No significant increase in papillomas under the same study conditions was seen in the 26-week study. For DCA under these study conditions, the p53 and Tg.AC mice appear less sensitive to hepatocarcinogenesis than standard rodent models. These results suggest caution for the use of Tg.AC and p53 mice to screen unknown chemicals in drinking water for potential carcinogenicity.

Gene set enrichment analysis for non-monotone association and multiple experimental categories.

BACKGROUND: Recently, microarray data analyses using functional pathway information, e.g., gene set enrichment analysis (GSEA) and significance analysis of function and expression (SAFE), have gained recognition as a way to identify biological pathways/processes associated with a phenotypic endpoint. In these analyses, a local statistic is used to assess the association between the expression level of a gene and the value of a phenotypic endpoint. Then these gene-specific local statistics are combined to evaluate association for pre-selected sets of genes. Commonly used local statistics include t-statistics for binary phenotypes and correlation coefficients that assume a linear or monotone relationship between a continuous phenotype and gene expression level. Methods applicable to continuous non-monotone relationships are needed. Furthermore, for multiple experimental categories, methods that combine multiple GSEA/SAFE analyses are needed. RESULTS: For continuous or ordinal phenotypic outcome, we propose to use as the local statistic the coefficient of multiple determination (i.e., the square of multiple correlation coefficient) R2 from fitting natural cubic spline models to the phenotype-expression relationship. Next, we incorporate this association measure into the GSEA/SAFE framework to identify significant gene sets. Unsigned local statistics, signed global statistics and one-sided p-values are used to reflect our inferential interest. Furthermore, we describe a procedure for inference across multiple GSEA/SAFE analyses. We illustrate our approach using gene expression and liver injury data from liver and blood samples from rats treated with eight hepatotoxicants under multiple time and dose combinations. We set out to identify biological pathways/processes associated with liver injury as manifested by increased blood levels of alanine transaminase in common for most of the eight compounds. Potential statistical dependency resulting from the experimental design is addressed in permutation based hypothesis testing. CONCLUSION: The proposed framework captures both linear and non-linear association between gene expression level and a phenotypic endpoint and thus can be viewed as extending the current GSEA/SAFE methodology. The framework for combining results from multiple GSEA/SAFE analyses is flexible to address practical inference interests. Our methods can be applied to microarray data with continuous phenotypes with multi-level design or the meta-analysis of multiple microarray data sets.

Gene expression response in target organ and whole blood varies as a function of target organ injury phenotype.

This report details the standardized experimental design and the different data streams that were collected (histopathology, clinical chemistry, hematology and gene expression from the target tissue (liver) and a bio-available tissue (blood)) after treatment with eight known hepatotoxicants (at multiple time points and doses with multiple biological replicates). The results of the study demonstrate the classification of histopathological differences, likely reflecting differences in mechanisms of cell-specific toxicity, using either liver tissue or blood transcriptomic data.

Heat map visualization of high-density clinical chemistry data.

Clinical chemistry data are routinely generated as part of preclinical animal toxicity studies and human clinical studies. With large-scale studies involving hundreds or even thousands of samples in multiple treatment groups, it is currently difficult to interpret the resulting complex, high-density clinical chemistry data. Accordingly, we conducted this study to investigate methods for easy visualization of complex, high-density data. Clinical chemistry data were obtained from male rats each treated with one of eight different acute hepatotoxicants from a large-scale toxicogenomics study. The raw data underwent a Z-score transformation comparing each individual animal's clinical chemistry values to that of reference controls from all eight studies and then were visualized in a single graphic using a heat map. The utility of using a heat map to visualize high-density clinical chemistry data was explored by clustering changes in clinical chemistry values for >400 animals. A clear distinction was observed in animals displaying hepatotoxicity from those that did not. Additionally, while animals experiencing hepatotoxicity showed many similarities in the observed clinical chemistry alterations, distinct differences were noted in the heat map profile for the different compounds. Using a heat map to visualize complex, high-density clinical chemistry data in a single graphic facilitates the identification of previously unrecognized trends. This method is simple to implement and maintains the biological integrity of the data. The value of this clinical chemistry data transformation and visualization will manifest itself through integration with other high-density data, such as genomics data, to study physiology at the systems level.

Hepatic transcript levels for genes coding for enzymes associated with xenobiotic metabolism are altered with age.

Metabolism studies are crucial for data interpretation from rodent toxicity and carcinogenicity studies. Metabolism studies are usually conducted in 6 to 8 week old rodents. Long-term studies often continue beyond 100 weeks of age. The potential for age-related changes in transcript levels of genes encoding for enzymes associated with metabolism was evaluated in the liver of male F344/N rats at 32, 58, and 84 weeks of age. Differential expression was found between the young and old rats for genes whose products are involved in both phase I and phase II metabolic pathways. Thirteen cytochrome P450 genes from CYP families 1-3 showed alterations in expression in the older rats. A marked age-related decrease in expression was found for 4 members of the Cyp3a family that are critical for drug metabolism in the rat. Immunohistochemical results confirmed a significant decrease in Cyp3a2 and Cyp2c11 protein levels with age. This indicates that the metabolic capacity of male rats changes throughout a long-term study. Conducting multiple hepatic microarray analyses during the conduct of a long-term study can provide a global view of potential metabolic changes that might occur. Alterations that are considered crucial to the interpretation of long-term study results could then be confirmed by subsequent metabolic studies.

Gene expression analysis offers unique advantages to histopathology in liver biopsy evaluations.

Liver diseases that induce nonuniform lesions often give rise to greatly varying histopathology results in needle biopsy samples from the same patient. This study examines whether gene expression analysis of such biopsies could provide a more representative picture of the overall condition of the liver. We utilized acetaminophen (APAP) as a model hepatotoxicant that gives a multifocal pattern of necrosis following toxic doses. Rats were treated with a single toxic or subtoxic dose of APAP and sacrificed 6, 24, or 48 hours after exposure. Left liver lobes were harvested, and both gene expression and histopathological analysis were performed on biopsy-sized samples. While histopathological evaluation of such small samples revealed significant sample to sample differences after toxic doses of APAP, gene expression analysis provided a very homogeneous picture and allowed clear distinction between subtoxic and toxic doses. The main biological function differentiating animals that received sub-toxic from those that had received toxic doses was an acute stress response at 6 hours and signs of energy depletion at later time points. Our results suggest that the use of genomic analysis of biopsy samples together with histopathological analysis could provide a more precise representation of the overall condition of a patient's liver than histopathological evaluation alone.

Application of visualization tools to the analysis of histopathological data enhances biological insight and interpretation.

Gene expression profiling, metabolomic screens, and other high-dimensional methods have become an integral part of many biological investigations. To facilitate interpretation of these data, it is important to have detailed phenotypic data--including histopathology--to which these data can be associated, or anchored. However, as the amount of phenotypic data increases, associations within and across these data can be difficult to visualize and interpret. We have developed an approach for categorizing and clustering biologically related histopathological diagnoses to facilitate their visualization, thereby increasing the possibility of identifying associations and facilitating the comparison with other data streams. In this study, we utilize histopathological data generated as part of a standardized toxicogenomics compendium study to generate composite histopathological scores and to develop visualizations that facilitate biological insight. The validity of this approach is illustrated by the identification of transcripts that correlate with the pathology diagnoses that comprise the categories of "response to hepatocellular injury" and "repair." This approach is broadly applicable to studies in which histopathology is used to phenotypically anchor other data, and results in visualizations that facilitate biological interpretation and the identification of associations and relationships within the data.

Optimal sampling of rat liver tissue for toxicogenomic studies.

Different degrees of a toxic response between and within the various lobes of the liver have been observed in rodents following treatment with acetaminophen. This study was designed to compare 2 sampling methods of the rat liver for gene-expression analysis. Ten male Fischer 344/N rats, 12-14 weeks of age, were treated with vehicle (0.5% aqueous ethyl cellulose) or acetaminophen (APAP, 1500 mg/kg) and sacrificed 24 hours following dose administration. Two representative sections were collected from the left liver lobe, stained with hematoxylin and eosin (H&E), and evaluated independently by 2 pathologists. The central core of the left lobe was cubed and frozen. Five random cubes were conserved, while the remaining left lobe core was pulverized. From each of the 10 animals, 2 random cubes and 2 samples from the homogeneous, pulverized samples were prepared for microarray analysis. Histopathologic evaluation revealed a variable response of centrilobular necrosis within the left lobe. Multiple methods used to analyze the microarray data indicated that sampling technique was not a major contributor to the variability observed in the gene expression data; however, only the powdered samples clustered for all animals, even those with disparate histopathologic results. Additionally, a powdered sample provided the advantages of a homogenous sample pool and the ability to use sample aliquots for other analyses to include proteomics, metabonomics, and other molecular techniques.

Phenotypic anchoring of acetaminophen-induced oxidative stress with gene expression profiles in rat liver.

Toxicogenomics provides the ability to examine in greater detail the underlying molecular events that precede and accompany toxicity, thus allowing prediction of adverse events at much earlier times compared to classical toxicological end points. Acetaminophen (APAP) is a pharmaceutical that has similar metabolic and toxic responses in rodents and humans. Recent gene expression profiling studies with APAP found an oxidative stress signature at a subtoxic dose that we hypothesized can be phenotypically anchored to conventional biomarkers of oxidative stress. Liver tissue was obtained from experimental animals used to generate microarray data, where male rats were given APAP at subtoxic (150 mg/kg) or overtly toxic (1500 and 2000 mg/kg) doses and sacrificed at 6, 24, or 48 h. Oxidative stress in liver was evaluated by a diverse panel of markers that included assessing expression of base excision repair (BER) genes, quantifying oxidative lesions in genomic DNA, and evaluating protein and lipid oxidation. A subtoxic dose of APAP produced significant accumulation of nitrotyrosine protein adducts. Both subtoxic and toxic doses caused a significant increase in 8-hydroxy-deoxyguanosine (8-OH-dG) as well as a significant decrease in glutathione (GSH) content. Only toxic doses of APAP significantly induced expression levels of BER genes. None of the doses examined resulted in a significant increase in the number of abasic sites or in the amount of lipid peroxidation. The accumulation of nitrotyrosine and 8-OH-dG adducts along with reduced GSH content in the liver phenotypically anchors the oxidative stress gene expression signature observed with a subtoxic dose of APAP, lending support to the validity of gene expression studies as a sensitive and biologically meaningful end point in toxicology.

Gene interaction network analysis suggests differences between high and low doses of acetaminophen.

Bayesian networks for quantifying linkages between genes were applied to detect differences in gene expression interaction networks between multiple doses of acetaminophen at multiple time points. Seventeen (17) genes were selected from the gene expression profiles from livers of rats orally exposed to 50, 150 and 1500 mg/kg acetaminophen (APAP) at 6, 24 and 48 h after exposure using a variety of statistical and bioinformatics approaches. The selected genes are related to three biological categories: apoptosis, oxidative stress and other. Gene interaction networks between all 17 genes were identified for the nine dose-time observation points by the TAO-Gen algorithm. Using k-means clustering analysis, the estimated nine networks could be clustered into two consensus networks, the first consisting of the low and middle dose groups, and the second consisting of the high dose. The analysis suggests that the networks could be segregated by doses and were consistent in structure over time of observation within grouped doses. The consensus networks were quantified to calculate the probability distribution for the strength of the linkage between genes connected in the networks. The quantifying analysis showed that, at lower doses, the genes related to the oxidative stress signaling pathway did not interact with the apoptosis-related genes. In contrast, the high-dose network demonstrated significant interactions between the oxidative stress genes and the apoptosis genes and also demonstrated a different network between genes in the oxidative stress pathway. The approaches shown here could provide predictive information to understand high- versus low-dose mechanisms of toxicity.