* National Institute of Environmental Health Sciences, P.O. Box 12233, MD2-04, Research Triangle Park, North Carolina 27709; and
Boehringer-Ingelheim Pharmaceuticals, Inc., Ridgefield, Connecticut 06877
Received November 26, 2001; accepted January 8, 2002
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key Words: gene expression; toxicogenomics; DNA arrays; classification; rat liver; pattern recognition.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Structurally unrelated compounds may belong to the same class of chemicals because of similarity in the pharmacological or toxicological endpoints they elicit. For example, at doses of diethylhexylphthalate (DEHP) and Wyeth 14,643 that produce similar levels of peroxisome proliferation in rat liver, Wyeth 14,643 produces an earlier and much greater liver tumor response than does DEHP (Biegel et al., 1992; Melnick et al., 1987
).
In this study, we tested the hypothesis that structurally unrelated compounds from the same chemical class produce similar, yet distinguishable, gene expression profiles. We also hypothesized that intraclass profiles are more similar to each other than to profiles corresponding to agents from different chemical classes. In order to test whether specific patterns of gene expression can be defined for a class of compounds and whether distinguishable patterns can be discerned within that class, we studied the expression profiles of 3 well-studied agents belonging to the peroxisome proliferator class of compounds [clofibrate (ethyl-p-chlorophenoxyisobutyrate), Wyeth 14,643 (4-chloro-6[2,3-xylidino)-2-pyrimidinylthio]acetic acid) and gemfibrozil (52[2,5-dimethylphenoxy]2-2-dimethylpentanoic acid)]. We also studied the expression profile of a well-studied enzyme inducer, phenobarbital, in order to determine whether a distinction could be made between it and the peroxisome proliferators. Microarray analyses were performed using liver RNA derived from chemically exposed Sprague-Dawley rats at multiple time points of exposure.
This work highlights several important points for the utility of toxicogenomics studies. First, the data confirm that compound classification based on gene expression profiles is feasible. In addition, the data illustrate the differences in the gene expression elicited by chemicals that may be related in many aspects but differ with respect to toxicological effects. Further investigation of these differences might offer an explanation of the dissimilarity in adverse effects associated with various peroxisome proliferators. In addition, our report also addresses the influence of the time of exposure on gene expression by highlighting transient and delayed gene expression events in response to the 2 classes of compounds studied. As toxicogenomics databases evolve, these distinctions will be important for understanding mechanisms and developing signatures of toxicity or adaptation. Finally, the data in this paper provides important information on gene expression changes, including time-independent changes that may be used to develop signatures of the compound classes of peroxisome proliferators. Underlying the analyses of these signature genes is the potential to develop hypotheses about the potential mechanism of action of these agents.
![]() |
MATERIALS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Histopathological analysis.
The liver tissues collected in formalin at necropsy were processed, embedded in paraffin, sectioned at 5 microns, and stained with hematoxylin and eosin (H&E). Histopathologic examinations of the liver sections were conducted by a pathologist and peer-reviewed.
RNA isolation.
Total RNA was isolated using QIAGEN (Qiagen, Valencia, CA) RNeasy kits. Liver sections of 130250 mg were used for midipreps and liver sections of approximately 800 mg were used for maxipreps. Homogenization buffer was added to frozen liver sections, and the tissue was immediately homogenized on ice (tissue did not thaw prior to homogenization) using a Cyclone homogenizer equipped with a rotor/stator shaft (VirTis Company, Gardiner, NY). Homogenates were processed as per the standard QIAGEN 3/99 protocol. Final product yielded 260 nm/280 nm ratios of 1.62.0, purity was confirmed via gels, and concentration was determined based on 260 nm absorbances.
cDNA microarray hybridization and analysis.
A cDNA NIEHS Rat Chip, v1.0, developed in-house at NIEHS, was used for gene expression profiling experiments. A complete listing of the genes on this chip is available at the following Web site: http://dir.niehs.nih.gov/microarray/chips.htm. cDNA microarray chips were prepared according to DeRisi et al., 1996. The spotted cDNAs were derived from a collection of sequence-verified clones that covered the 3` end of the gene and ranged in size from 500 to 2000 base pairs (Research Genetics). M13 primers were used to amplify insert cDNAs from purified plasmid DNA in a 100 µl polymerase chain reaction (PCR) mixture. A sample of the PCR products (10 µl) was separated on 2% agarose gels to ensure quality of the amplifications. The remaining PCR products were purified by ethanol precipitation, resuspended in ArrayIt buffer (Telechem, San Jose CA) and spotted onto poly-L-lysine-coated glass slides using a modified, robotic DNA arrayer (Beecher Instruments, Bethesda MD).
For microarray hybridizations, each total RNA sample (3575 µg) was labeled with Cyanine 3 (Cy3)-or Cyanine 5 (Cy5)-conjugated dUTP (Amersham, Piscataway, NJ) by a reverse transcription reaction using reverse transcriptase, SuperScript (Invitrogen, Carlsbad, California), and the primer, Oligo dT (Amersham, Piscataway, NJ). Control samples were labeled with Cy3 while samples derived from chemically exposed animals were labeled with Cy5. The fluorescently labeled cDNAs were mixed and hybridized simultaneously to the cDNA microarray chip. Each RNA pair was labeled and hybridized independently in triplicate to a total of 3 arrays. The cDNA chips were scanned with an Axon Scanner (Axon Instruments, Foster City CA) using independent laser excitation of the 2 fluors at 532 and 635 nm wavelengths for the Cy3 and Cy5 labels, respectively.
The raw pixel intensity images were analyzed using the ArraySuite v1.3 extensions of the IPLab image processing software package (Scanalytics, Fairfax, VA). This program uses methods that were developed and previously described by Chen et al. (1997) to locate targets on the array, measure local background for each target, and subtract it from the target intensity value, and to identify differentially expressed genes using a probability-based method. After pixel intensity determination and background subtraction, the ratio of the intensity of the treated cells to the intensity of the control was calculated. We have previously determined that significant autofluorescence of the gene features on the array, attributed to spotting solution, occurs at high scanning power (Tucker et al., unpublished). We measured the pixel intensity level of "blank" spots comprised of spotting solution. The data was then filtered to provide a cut off at the intensity level just above the blank measurement values in order to remove from further analyses those genes having one or more intensity values in the background range. The ratio intensity data from all of the 1700 spots printed on the NIEHS Rat Chip v1.0 was used to fit a probability distribution to the ratio intensity values and estimate the normalization constants (m and c) that this distribution provides. The constant m, which provides a measure of the intensity gain between the two channels, was approximately equal to 1 for all arrays, indicating that the channels were approximately balanced. For each array, the ratio intensity values were normalized to account for the imbalance between the 2 fluorescent dyes by multiplying the ratio intensity value by m. A probability distribution was fit to the data and used to calculate a 95% confidence interval for the ratio intensity values. Genes having normalized ratio intensity values outside of this interval were considered significantly differentially expressed.
For each of the 3 replicate arrays for each sample, lists of differentially expressed genes at the 95% confidence level were created and deposited into the NIEHS MAPS database (Bushel et al., 2001). For each time point and each animal, a query of the database yielded a list of genes that were differentially expressed in at least 2 of the 3 replicate hybridizations. A calculation using the binomial probability distribution indicated that the probability of a single gene appearing on this list when there was no real differential expression is approximately 0.0006. Hierarchical cluster analysis was carried out with the Cluster/TreeView package (Eisen et al., 1998
). The entire data set is available at http://dir.niehs.nih.gov/microarray/datasets.
Real-time quantitative PCR.
RNA samples representing single animals treated with a peroxisome proliferator or phenobarbital for 24 h or 2 weeks (1852 [clofibrate, 24 hr], 1868 [Wyeth 14,643, 24 hr], 1878 [gemfibrozil, 24 hr], 1890 [phenobarbital, 24 h], 888 [clofibrate, 2 weeks], 898 [Wyeth 14,643, 2 weeks], 912 [gemfibrozil, 2 weeks], and 926 [phenobarbital, 2 weeks]) were used to validate the expression profile of 10 genes obtained using cDNA microarray data [AA818412 p450 2B2; AA996791 carnitine palmitoyl transferase 1; AI111901 tripeptidylpeptidase II; AA923966 Aflatoxin aldehyde reductase; AA957359 p55cdc; AA957519 stathmin cytosolic phosphoprotein p19; AA965078 enoyl CoA isomerase; AA818188 ketoacyl thiolase; AA963928 Ah receptor; AI070587 carboxylesterase precursor].
The primers for the aforementioned genes were designed using Primer Express® software (Applied Biosystems, Foster City, CA) and custom made (Research Genetics, Huntsville, AL). Primers that resulted in a single product which could be visualized on a 2% agarose gel were as follows: p450 2B2 [forward primer AGTGCATCACAGCCAACATCA, reverse primer GAGGGAAAAGGTCCGGTAGAA]; carboxylesterase precursor [forward primer AGTACTGGGCCAATTTTGCAA, reverse primer TGGGTGTCCAACTGCAGGTA]; Ah receptor [forward primer CATCCTGGAAATTCGAACCAA, reverse primer TGCAAGAAGCCGGAAAACT]; carnitine palmitoyl transferase 1 [forward primer CGGTTCAAGAATGGCATCATC, reverse primer ATCACACCCACCACCACGATA]; ketoacyl thiolase [forward primer ACGTGAGTGGAGGTGCCATAG, reverse primer CTCGACGCCTTAACTCGTGAAC]; stathmin p19 [forward primer CACAATCCACTGGCAAGGAA, reverse primer TGCCATGTTGGACAGAAGACA].
Real-time PCR targeting the message corresponding to these 10 genes was performed using the ABI prism 7700 Sequence Detection System (Applied Biosystems, Foster City, CA) according to the manufacturer's instructions. The SYBR® Green I labeling kit (Applied Biosystems, Foster City, CA), was used to detect double-stranded DNA generated during PCR amplification, used according to the manufacturer's instructions. Reverse transcription and PCR reactions were performed at the same time in a 50 µl reaction containing 4 mM MgCl2, 0.8 mM of each dNTP, 100 ng total RNA, 0.4 µM reverse primer and 0.4 µM forward primer, 0.4 units/µl RNasin, 0.025 units/µl AmpliTaq Gold DNA polymerase (Roche, Basel, Switzerland) and 0.25 units/µl MulV Reverse Transcriptase (Roche, Basel, Switzerland). Amplification reactions were carried out using the following temperature profile: 48°C, 30 min; 95°C, 10 min; 95°C, 15 s; 60°C, 1 min) for 40 cycles. Fluorescence emission was detected for each PCR cycle and the threshold cycle (CT) values were determined. The CT value was defined as the actual PCR cycle when the fluorescence signal increased above the background threshold. Induction or repression of a gene in a treated sample relative to control was calculated as follows: Fold increase/decrease = e (CT(exposed) CT(control)). Values were reported as an average of triplicate analyses.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Gene expression.
In order to determine gene expression changes associated with chemical exposure, liver mRNA was collected 24 h following a single exposure or after 2 weeks of daily exposures to the compound. Competitive hybridizations of fluorescently labeled cDNA (Duggan et al., 1999) derived from control vs. treated livers were used to measure relative abundance of each mRNA on the custom NIEHS cDNA microarray Rat Chip v1.0, which contained
1700 sequence-verified rat genes. We conducted statistical analyses of the microarray data and determined significantly changed genes using a ratio distribution model at the 95% confidence level, as mentioned in the Materials and Methods section. We were able to reduce the probability of false positives in our data set by performing triplicate hybridizations on each of at least 3 independent biological samples and by including only genes exhibiting a binomial distribution probability
0.0006 (Bushel et al., 2001
). These genes were utilized for further higher-level comparative analyses (e.g., clustering).
Exploratory interpretation.
The results obtained from the collective microarray analysis of the peroxisome proliferator-treated rat livers revealed significant gene expression changes in approximately 25% of the genes on the rat chip and elucidated interesting molecular changes and pathway relationships associated with peroxisome proliferator exposure (Table 1). These pathways include stimulation of triglyceride hydrolysis, fatty acid uptake and conversion to acyl CoA derivatives, and stimulation of the ß-oxidation pathway. Observation of alteration of these pathways corroborates past data (Amacher et al., 1997
; Schoonjans et al., 1996
) and serves as a validation of the use of microarrays to rapidly interrogate effector pathways of toxicants.
|
|
|
|
|
|
|
|
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Whether animals should be grouped together as a pool or examined individually represents one issue in the design of toxicogenomics studies. Some investigators advocate pooling for the costly microarray analysis and using individual animals for the follow-up verification steps. However, pooling may cause misinterpretation of data if one animal shows a remarkably distinct response, or lack of response. In this study we analyzed individual chemically exposed animals against a pool of control animals. The generation of gene expression profiles corresponding to 3 animals exposed to the same compound, as opposed to pooling, allowed for the detection of interanimal differences. This facilitated the testing of the robustness of DNA microarray technology and pattern recognition algorithms to generate distinguishable gene expression profiles, despite the existence of differences in response across similarly treated animals. Although we have observed similarities as high as about 90% between animals exposed to the same agents, this similarity showed chemical dependence, reaching about 80% with some of the compounds. The best interanimal correlation was observed among Wyeth 14,643-treated animals, where we saw the greatest number of statistically significant gene expression modulations, suggesting that compound potency may be positively related to decreased variation in gene expression response.
There is high concordance of the expression changes found in the microarray analyses in phenobarbital-exposed rats with the results obtained by scientists using other methodologies over many years of study. This concordance is illustrated in a display for gene expression changes in the form of an "effector pathway map" (EPM) for chemical action. The phenobarbital gene expression profile data set was mapped into previously defined cellular response pathways to demonstrate the informational power of this type of data presentation (Fig. 7). Organization of data in this integrated diagram facilitates clear visualization of regulatory and molecular events occurring as a response to chemical exposure and visualization of pathways that were affected by phenobarbital. At a glance, this diagram illustrates how the gene expression profile data has corroborated past findings (Fig. 7
, yellow targets) and may contribute new mechanistic insights (Fig. 7
, blue targets).
|
Analysis of transient, delayed, and time-independent alterations suggested a relationship between the pattern of expression and gene function. The majority of genes that were expressed in a time-independent fashion in response to phenobarbital or peroxisome proliferator treatment corresponded to enzymes that function in compound metabolism (Fig. 5A; p450 2B2, epoxide hydrolase 1, GST, aflatoxin B1 aldehyde reductase) or cell biochemical processes (Figs. 5A5C
;Acyl CoA oxidase, Carnitine palmitoyltransferase 1, histidine decarboxylase, stearyl-CoA desaturase). This makes sense when one considers that animals were treated with the studied chemicals on a daily basis, furnishing recurring surges in blood levels of the compounds in the exposed animals, and thus affecting compound metabolism genes or downstream effects.
Metabolism-related genes were notably absent from the transiently altered transcripts, the majority of which represented signaling related genes (Fig. 6A). This is consistent with the transient nature of the response and these genes probably constituted an initial response in the liver upon exposure to the specific toxicants for the first time. Alterations in gene expression that were present at 2 weeks of exposure but not at the 24-h time point (Fig. 6B
) constituted delayed responses to toxicant exposure. These responses could be associated with adaptation events or with the relation to the histopathological observation of hypertrophy noted in all animals treated by chemicals for 2 weeks. These changes could also be due to overt toxicity that may be manifested in gene expression responses but not necessarily detected by histopathological examination until a much longer period of exposure.
We have successfully generated gene expression profiles for 3 peroxisome proliferators and an enzyme inducer, and we were able to show, through the application of pattern recognition algorithms and computational analyses, that these patterns were distinct. We demonstrated that chemicals from the same class of compounds give rise to discernable gene expression profiles that bear more similarity to each other than to patterns corresponding to exposure to compounds from a different class. Systematic development of an expanded database for gene expression responses to drugs and environmental pollutants may yield compound-specific signature patterns that would also provide insights into affected regulatory and proliferative and repair/adaptive pathways. We demonstrated the validity of our expression data by corroborating published reports on the chemicals that we utilized. In addition, our data revealed gene expression that has not been previously associated with the compounds we used and suggest that such results will provide valuable targets for further investigations of the mechanism of action of chemical hazards.
![]() |
ACKNOWLEDGMENTS |
---|
![]() |
NOTES |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Argaud, D., Halimi, S., Catelloni, F., and Leverve, X. M. (1991). Inhibition of gluconeogenesis in isolated rat hepatocytes after chronic treatment with phenobarbital. Biochem. J. 280, 663669.[ISI][Medline]
Barbason, H., Rassenfosse, C., and Betz, E. H. (1983). Promotion mechanism of phenobarbital and partial hepatectomy in DENA hepatocarcinogenesis cell-kinetics effect. Br. J. Cancer 47, 517525.[ISI][Medline]
Biegel, L. B., Hurtt, M. E., Frame, S. R., Applegate, M., O`Connor, J. C., and Cook, J. C. (1992). Comparison of the effects of Wyeth 14,643 in Crl:CD BR and Fisher-344 rats. Fundam. Appl. Toxicol. 19, 590597.[ISI][Medline]
Burchiel, S. K., Knall, C. M. Davis, J. W., III, Paules, R. S. Boggs, S. E. Afshari, C. A. (2001). Analysis of genetic and epigenetic mechanisms of toxicity: Potential roles of toxicogenomics and proteomics in toxicology. Toxicol. Sci. 59, 193195.
Bushel, P. R., Hamadeh, H., Bennett, L., Sieber, S., Martin, K., Nuwaysir, E. F., Johnson, K., Reynolds, K., Paules, R. S., and Afshari, C. A. (2001). MAPS: A microarray project system for gene expression experiment information and data validation. Bioinformatics 17, 564565.
Busser, M. T., and Lutz, W. K. (1987). Stimulation of DNA synthesis in rat and mouse liver by various tumor promoters. Carcinogenesis 8, 14331437.[Abstract]
Butterworth, B. E., Conolly, R. B., and Morgan, K. T. (1995). A strategy for establishing mode of action of chemical carcinogens as a guide for approaches to risk assessments. Cancer Lett. 93, 129146.[ISI][Medline]
Chen, Y., Dougherty, E, R., and Bittner, M. L. (1997). Ratio-based decisions and the quantitative analysis of cDNA microarray images. J. Biomed. Optics 2, 364374.
DeRisi, J., Penland, L., Brown, P. O., Bittner, M. L., Meltzer, P. S., Ray, M., Chen, Y., Su, Y. A., and Trent, J. M. (1996). Use of a cDNA microarray to analyse gene expression patterns in human cancer. Nat. Genet. 14, 457460.[ISI][Medline]
Duggan, D. J., Bittner, M., Chen, Y., Meltzer, P., and Trent, J. M. (1999). Expression profiling using cDNA microarrays. Nat. Genet. 21(Suppl. 1), 1014.[ISI][Medline]
Eisen, M. B., Spellman, P. T., Brown, P. O., and Botstein, D. (1998). Cluster analysis and display of genome-wide expression patterns. Proc. Natl. Acad. Sci. U.S.A 95, 1486314868.
Feldman, D., Swarm, R. L., and Becker, J. (1981). Ultrastructural study of rat liver and liver neoplasms after long-term treatment with phenobarbital. Cancer Res. 41, 21512162.[Abstract]
Fielden, M. R., and Zacharewski, T. R. (2001). Challenges and limitations of gene expression profiling in mechanistic and predictive toxicology. Toxicol. Sci. 60, 610.
Furukawa, K., Numoto, S., Furuya, K., Furukawa, N. T., and Williams, G. M. (1985). Effects of the hepatocarcinogen nafenopin, a peroxisome proliferator, on the activities of rat liver glutathione-requiring enzymes and catalase in comparison to the action of phenobarbital. Cancer Res. 45, 50115019.[Abstract]
Griffin, M. J., Kirsten, E., Carubelli, R., Palakodety, R. B., McLick, J., and Kun, E. (1984). The in vivo effect of benzamide and phenobarbital on liver enzymes: Poly(ADP-ribose) polymerase, cytochrome P-450, styrene oxide hydrolase, cholesterol oxide hydrolase, glutathione S-transferase and UDP-glucuronyl transferase. Biochem. Biophys. Res. Commun. 122, 770775.[ISI][Medline]
Hamadeh, H. K., Bushel, P. B., Paules, R., and Afshari, C. A. (2001). Discovery in toxicology: Mediation by gene expression array technology. J. Biochem. Mol. Toxicol. 15, 231242.[ISI][Medline]
Hug, G., McGraw, C. A., Bates, S. R., and Landrigan, E. A. (1991). Reduction of serum carnitine concentrations during anticonvulsant therapy with phenobarbital, valproic acid, phenytoin, and carbamazepine in children. J. Pediatr. 119, 799802.[ISI][Medline]
IARC (1977). IARC monographs on the evaluation of the carcinogenic risk of chemicals to man: Some miscellaneous pharmaceutical substances. IARC Monogr. Eval. Carcinog. Risk Chem. Man. 13, 1255.[Medline]
IARC (1987).Overall evaluations of carcinogenicity: An updating of IARC Monographs volumes 1 to 42. IARC Monogr. Eval. Carcinog. Risks Hum. 7 (Suppl.), 1440.
Koppel, J., Loyer, P., Maucuer, A., Rehak, P., Manceau, V., Guguen-Guillouzo, C., and Sobel, A. (1993). Induction of stathmin expression during liver regeneration. FEBS Lett. 331, 6570.[ISI][Medline]
Melnick, R. L., Morrissey, R. E., and Tomaszewski, K. E. (1987). Studies by the National Toxicology Program on di(2-ethylhexyl)phthalate. Toxicol. Ind. Health 3, 99118.[ISI][Medline]
Nuwaysir, E. F., Bittner, M., Trent, J., Barrett, J. C., and Afshari, C. A. (1999). Microarrays and toxicology: The advent of toxicogenomics. Mol. Carcinog. 24, 153159.[ISI][Medline]
Ohashi, K., Nagata, K., Maekawa, M., Ishizaki, T., Narumiya, S., and Mizuno, K. (2000). Rho-associated kinase ROCK activates LIM-kinase 1 by phosphorylation at threonine 508 within the activation loop. J. Biol. Chem. 275, 35773582.
Schoonjans, K., Staels, B., and Auwerx, J. (1996). Role of the peroxisome proliferator-activated receptor (PPAR) in mediating the effects of fibrates and fatty acids on gene expression. J. Lipid Res. 37, 907925.[Abstract]
Tavoloni, N., Jones, M. J., and Berk, P. D. (1983). Dose-related effects of phenobarbital on hepatic microsomal enzymes. Proc. Soc. Exp. Biol. Med. 174, 2027.[Abstract]
Thomas, R. S., Rank, D. R., Penn, S. G., Zastrow, G. M., Hayes, K. R., Pande, K., Glover, E., Silander, T., Craven, M. W., Reddy, J. K., Jovanovich, S. B., and Bradfield, C. A. (2001). Identification of toxicologically predictive gene sets using cDNA microarrays. Mol. Pharmacol. 60, 11891194.
Thurman, R. G., and Marazzo, D. P. (1975). Mixed-function oxidation and intermediary metabolism: Metabolic interdependencies in the liver. Adv. Exp. Med. Biol. 58, 355367.[Medline]
Waring, J. F., Ciurlionis, R., Jolly, R.A., Heindel, M., and Ulrich, R. G. (2001a). Microarray analysis of hepatotoxins in vitro reveals a correlation between gene expression profiles and mechanisms of toxicity. Toxicol. Lett. 120, 359368.[ISI][Medline]
Waring, J. F., Jolly, R. A., Ciurlionis, R., Lum, P. Y., Praestgaard, J. T., Morfitt, D. C., Buratto, B., Roberts, C., Schadt, E., and Ulrich, R. G. (2001b). Clustering of hepatotoxins based on mechanism of toxicity using gene expression profiles. Toxicol. Appl. Pharmacol. 175, 2842.[ISI][Medline]
Watanabe, N., Kato, T., Fujita, A., Ishizaki, T., and Narumiya, S. (1999). Cooperation between mDia1 and ROCK in Rho-induced actin reorganization. Nat. Cell Biol. 1, 136143.[ISI][Medline]
Whysner, J., Ross, P. M., and Williams, G. M. (1996). Phenobarbital mechanistic data and risk assessment: Enzyme induction, enhanced cell proliferation, and tumor promotion. Pharmacol. Ther. 71, 153191.[ISI][Medline]