Grantee Research Project Results
Final Report: Fetal Metabolism of Aflatoxin B1 and Susceptibility to Childhood Cancer
EPA Grant Number: R827441Title: Fetal Metabolism of Aflatoxin B1 and Susceptibility to Childhood Cancer
Investigators: Gallagher, Evan
Institution: University of Florida
EPA Project Officer: Aja, Hayley
Project Period: July 1, 1999 through June 30, 2002 (Extended to March 30, 2004)
Project Amount: $523,123
RFA: Children's Vulnerability to Toxic Substances in the Environment (1999) RFA Text | Recipients Lists
Research Category: Children's Health , Human Health
Objective:
Cancer is the second leading cause of death for children under 14 years of age in the United States. The initial peak of cancer incidence occurs during the first 5 years of life, and available evidence indicates that a primary risk factor for childhood cancer involves transplacental exposure to either mutagenic or promutagenic agents. The rapid changes that occur during prenatal development may result in critical windows of susceptibility to toxic injury. A primary determinant of this susceptibility is the balance among in utero conversion of procarcinogens to DNA-reactive metabolites (e.g., by cytochrome P450s, lipoxygenases) and the detoxification of reactive intermediates (e.g., by glutathione S-transferases (GSTs) and other phase II enzymes). The overall objective of this research project was to characterize the metabolism of a model transplacental dietary carcinogen (aflatoxin B1 [AFB1]) in human prenatal liver to understand the genetic and developmental risk factors for dietary procarcinogens. In addition to AFB1, other drugs and environmental chemicals of relevance to the in utero environment that have been linked to maternal exposures during pregnancy were examined in studies using model cell systems relevant to human prenatal liver.
Summary/Accomplishments (Outputs/Outcomes):
Relationship Among Ontogenic Expression of Aflatoxin Metabolizing Enzymes and In Vitro DNA and Protein Binding
A substantial amount of work effort has been devoted to characterizing interindividual and ontogenic variation in the expression of AFB1 metabolizing enzymes, as such variation may contribute to individual and developmental differences in the susceptibility to AFB1-induced cancers. The approach involved using semiquantitative Western blotting for seven different enzymes involved in the biotransformation of AFB1. Substantial individual and ontogenic variation in the hepatic expression of certain AFB1 metabolizing enzymes were noted. In particular, we observed interindividual variation (> threefold) in the expression of enzymes that bioactivate AFB1 to the potent DNA-binding intermediate, aflatoxin B1-8,9-exo-epoxide (AFBO), including adult cytochrome P4501A2 (CYP1A2) and 3A4, and prenatal liver CYP3A7 (shown in Figure 1A). The expression of lipoxygenase, another AFB1 bioactivating enzyme, was ninefold higher in prenatal liver tissues than in adult liver tissues. The expression of hGSTM1-1, a GST enzyme purported to be protective against AFBO-DNA injury, was detected in 30 percent of the adult livers, as well as 40 percent of the prenatal livers examined (Figure 1B). The levels of hGSTM1-1 expression were higher in the adult hGSTM1-1-positive tissues compared to hGSTM1-1-positive prenatal tissues. The expression of other enzymes that may be important in protecting against AFB1, including aldehyde keto-reductase 7A and microsomal epoxide hydrolase, did not markedly vary (< threefold variation) among tissue donors. Thus, this study demonstrated important ontogenic and individual differences in the expression of AFB1 metabolizing enzymes.
Figure 1. Western Blot Analysis of Cytochrome P4503A (CYP3A-, Left) and hGSTM1-1 (Right) Protein Expression in Human Liver Microsomal Fractions. The CYP3A- and hGSTM1-1 expression levels were normalized against the housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH).
In vitro AFBO binding studies were conducted to examine the relationship among interindividual- and age-dependent variation in AFBO binding to DNA and proteins. In this regard, AFBO binding to DNA and proteins represent the initial steps in the pathogenesis of AFB1 liver cancer and liver injury, respectively. We had originally hypothesized that differences in AFB1 metabolizing enzyme expression, if present, would underlie a marked difference in susceptibility among adult and prenatal liver tissues with regards to in vitro AFB1-macromolecular binding. The prenatal liver tissues, however, did not exhibit a marked susceptibility to the formation of AFBO-macromolecular adducts in vitro (Figure 2). In addition, high interindividual variability was observed with respect to rates of in vitro formation of AFBO-DNA and AFBO-protein adducts among a panel of 10 adult donors, but little variation was observed among the 10 prenatal liver tissue donors (Figure 2). Correlation analysis revealed that the expression of some AFB1 oxidative enzymes (e.g., prenatal liver CYP3A7 and adult liver CYP1A2 and CYP3A- proteins) were correlated with AFBO-macromolecular adduct formation, however, the levels of putative AFBO detoxification enzymes hGSTM1-1, microsomal epoxide hydrolase, and aldehyde keto-reductase 7A were not. Collectively, these results indicate that despite ontogenic differences in the expression of AFB1 metabolizing enzymes, prenatal liver tissues do not exhibit a marked susceptibility to the in vitro formation of AFBO-DNA or AFBO-protein adducts relative to adults. Collectively, the results of these experiments indicated that the level of CYP3A7 expression in utero is an important determinant of AFBO-DNA binding and suggested that those individuals with higher CYP3A7 expression may be at risk to forming higher levels of AFBO-DNA adducts than those with low CYP3A7 expression. The results of these experiments have important ramifications with regard to the potential for the formation of AFB-DNA adducts in utero. In general, however, these data were not supportive of our original hypothesis that the prenatal livers would form substantially high levels of AFBO-DNA adducts in vitro.
Figure 2. Individual and Ontogenic Variation in Microsomal-Mediated In Vitro AFB1-DNA and AFB1-Protein Binding in the Presence of Cytosol. Left panel: individual data. Right panel: box and whisker plot of the data by age group (open squares prenatal liver, filled squares adult liver).
Based on our observations of substantial interindividual variability in human prenatal liver expression of AFB1 metabolizing enzymes, we conducted a comparative study of the expression of other biotransformation enzymes relevant to the detoxification of maternally transferred chemicals. These experiments targeted the protein products of genes involved in the detoxification of toxic intermediates involved in the etiology of in utero oxidative injury. These efforts centered on alpha class GST isozymes, which have been demonstrated to be important in protecting against the toxic effects of certain carcinogens and oxidative damage products, and included studies at the mRNA and catalytic activity levels. These studies revealed the presence of alpha class GST isozyme hGSTA4-4 and hGSTA1/2 mRNAs in prenatal liver cytosol and mitochondria, whereas multiple tissue array analysis demonstrated considerable tissue-specific and developmental variation in hGSTA4 and hGSTA1/2 mRNA expression. Although prenatal liver cytosolic GSTs were active toward a variety of GST substrates, glutathione-dependent peroxidase and GST-dependent peroxidase activities were ninefold and eighteenfold lower in prenatal liver relative to adult liver (Figure 3). Thus, although prenatal liver tissues were generally not highly susceptible to the procarcinogen AFB1, it is possible that the relatively inefficient prenatal reduction of hydroperoxides may underlie an increased susceptibility to maternally transferred pro-oxidants. The potential for susceptibility of prenatal liver to oxidative injury was followed up by the subsequent experiments using the prenatal hematopoietic stem cell model.
Figure 3. Comparative Initial Rate Cytosolic Glutathione Peroxidase (GPX) Activities in Prenatal and Adult Hepatic Cytosolic Fractions. (A) Total GPX specific activity (nmol CuOOH reduced/minute/mg protein) in 10 prenatal liver samples and 11 adult liver samples; (B) selenium-dependent GPX activity (nmol H2O2 reduced/minute/mg protein) in the same samples assayed in B; and (C) subtracted values of (B) from (A) yielding an estimate of GST-dependent peroxidase activities for these samples. Asterisks (*) denote activity values below the practical limit of detection (2 nmol/minute/mg).
Comparative Gene Array Study of Stress Responses Associated With AFB1 Exposure in Cultured Human Liver Slices
A key question posed in the studies was if there was a marked difference in stress responses associated with AFB1 exposure among cultured adult and prenatal precision liver slices. To answer this question, we prepared 250 µ m precision liver slices from second trimester prenatal liver tissue and also from adult liver and cultured the liver slices in the presence of 0.5 µ M AFB1 for 18 hours. Control and AFB1-exposed slices were harvested for preparation and labeling of poly(A)+ RNA, which was used to probe a commercial nylon stress gene microarray comprising 234 cDNA clones involved in cellular defense against chemical toxicity. A comparison of the control adult and prenatal liver slices revealed that a number of genes involved in chemical biotransformation, excision repair, and oxidative defense were expressed at markedly lower levels in prenatal liver compared to adult liver. Specifically, the prenatal liver slices had elevated expression of human glutathione S-transferases pi-1 (hGSTP1) and omega-1 (hGSTO1), as well as DNA excision repair ERCC1 and heat shock protein 27 (HSP27) (Figure 4). These data were supportive of the results of the experiments demonstrating differences in AFB1 activating and detoxification enzymes among prenatal and adult liver tissues.
Figure 4. Comparison of Relative Control Gene Expression in Second Trimester Prenatal Liver Slices as Compared to Adult Liver Slices
In addition, prenatal liver slices exposed to AFB1 were more sensitive to the acute toxic effects of AFB1 than were adult liver slices (Figure 5). This experiment indicated that a number of protective pathways, including GST, HSP, and DNA-excision repair, may be increased in prenatal liver on exposure to AFB1; however, in general, cultured prenatal liver slices appear to be more sensitive to the acute toxic effects of AFB1 when compared to adult liver. This latter observation was consistent with our earlier in vitro studies of AFBO-DNA binding. The results of this study also indicated that DNA microarray technology may be used in conjunction with cultured liver slices to examine ontogenic differences in chemical-mediated expression of relevant human genes. This approach has potential applications for evaluating human risk to transplacental exposure to dietary carcinogens.
Figure 5. Effect of AFB1 on Slice K+ Concentrations in Prenatal Liver (Left) and Adult Liver (Right)
Effects of a Model Teratogen on Oxidative Stress Gene Expression and Viability of Cultured Prenatal Liver Slices
Cultured precision human prenatal liver slices also were used to study the effects of a model human teratogen phenytoin (diphenylhydantoin; Dilantin) on cell toxicity, glutathione redox status, and steady state mRNA expression of a panel of oxidative stress-related biomarker genes. This compound was selected as a model because of the demonstrated involvement of oxidative stress as a component of embryotoxicity at high doses of the compound in animal studies. The biomarker genes analyzed were p53, bcl-2, alpha class glutathione S-transferases isozymes A1 and A4 (hGSTA1 and hGSTA4), and the catalytic subunit of γ-glutamylcysteine synthetase (γGCS-HS). In these experiments, liver slices were prepared from second trimester prenatal livers and cultured in the presence of 0, 250, and 1,000 µM phenytoin for 18 hours. Exposure to 1,000 µM phenytoin elicited 41 percent and 34 percent reductions in slice intracellular potassium and reduced glutathione concentrations, respectively. The reduction in slice glutathione concentrations at 1000 µM phenytoin was accompanied by a 2.2-fold increase in the percentage of total slice glutathione consisting of GSSG and a 3.9-fold increase in hGSTA1 steady state mRNA expression. Exposure to 250 µM or 1,000 µM phenytoin also elicited a relatively minor (less than twofold) but significant increase in p53 steady state mRNA expression. In contrast, the steady state levels of γGCS-HS, hGSTA4, and bcl-2 mRNAs were not affected by phenytoin exposure. These experiments demonstrated a role of glutathione and hGSTA1 against phenytoin toxicity and teratogenesis. These studies further demonstrated the utility of using cultured human prenatal liver slices as a relevant tool for developmental toxicology studies.
Induction of Phase II Enzymes in Cultured Human Liver Slices
Experiments associated with aim three were directed toward potential chemoprotection applications using the liver slice model, specifically, whether protective enzymes could be induced in prenatal liver by exposure to dietary anticarcinogens. If so, such observations could justify potential chemoprotective strategies relevant to protection of in utero injury during pregnancy. A series of studies initially was conducted using the adult liver slice model system. In these experiments, precision human liver slices prepared from an adult donor were incubated up to 24 hours in the presence of three different doses of t-butyl-hydroquinone (TBHQ), ethoxyquin, and butylated hydroxyanisole (BHA) compounds, which have been shown to induce protective phase II enzymes in rodent model systems. Cultured liver slices generally remained highly viable throughout the exposures, with the exception of an 11 percent loss in slice K + concentrations after 24 hours exposure to 200 mM TBHQ. Although exposure to all three compounds stimulated liver slice glutathione biosynthesis, the expression of the mRNA encoding the catalytic subunit of glutamate-cysteine ligase (GCLC, the rate limiting enzyme in glutathione biosynthesis) generally was unaffected by antioxidant exposures. Exposure to 10 mM BHA elicited a modest (< 20 %) increase in steady state hGSTA1 and hGSTA4 mRNA expression at 24 hour of exposure. In contrast, exposure to higher doses of BHA and also TBHQ elicited modest (< 20 %) decreases in steady state hGSTA1 mRNA expression. In summary, our results suggest that the expression of alpha GST isozymes and GCLC may not be modulated strongly by exposure to synthetic antioxidants in cultured human liver slices. These results may be a reflection of: (1) an overall poor induction response of human alpha GST by antioxidants relative to rodents; (2) a limited responsiveness of these genes to inducing agents in cultured human liver slices; or (3) nonresponsive genetics of some individual liver donors. Because of the relative nonresponsiveness of the adult tissues and the development of quality control issues associated with procurement of viable human prenatal liver studies, further chemoprotection studies were not pursued in cultured prenatal liver tissues. In lieu of these studies, we elected to pursue other relevant in utero model systems (see below).
The Effects of Transplacental Agents (AFB1, 4-Hydroxynonenal, and Pesticides) on Prenatal Liver Hematopoietic Stem Cell Injury
Studies under aim two were directed toward the effects of AFB1 and other pertinent transplacental carcinogens on critical cell populations of relevance in human prenatal liver. In this regard, hematopoietic cells and precursors (e.g., hematopoietic CD34+ stem cells) represent a significant proportion of the cell populations of human prenatal liver. These cells are of particular importance because they form the basis of hematopoiesis during development. Accordingly, injury to prenatal liver hematopoietic stem cells is highly relevant to potential development of hematopoietic injuries, as these cells are capable of initiating long-term hematopoiesis. Culture conditions were established for prenatal liver hematopoietic stem cells that allowed for adequate cell proliferation while maintaining the progenitor nature of these stem cells. Based on these results, we conducted toxicant challenge experiments after 10 days of culture. At day 10 of culture, cells still are proliferating and undergoing differentiation into myeloid and lymphoid lineages. On establishing culture conditions for the cells, several toxicant challenge experiments were conducted. Exposure to physiologically relevant levels of AFB1 did not result in loss of viability or in the formation of AFBO-DNA or AFBO-protein adduct formation in prenatal hematopoietic stem cells. Western blotting experiments revealed a potential mechanism for the lack of sensitivity to AFB1. It appears that the prenatal liver hematopoietic stem cells do not appreciably express the cytochrome P450 isoenzymes (CYP1A2 and CYP 3A4) that are primarily responsible for activation of AFB1 to the toxic DNA-binding derivative AFBO.
In contrast to our results with AFB1, subsequent dose-response studies revealed that prenatal liver hematopoietic stem cells lose viability when cultured in the presence of 4-hydroxynonenal, a model mutagenic and cytotoxic aldehyde produced upon oxidative injury (Figure 6A). The levels of 4-hydroxynonenal that caused injury were relevant to those encountered under cellular oxidative stress. Furthermore, exposure to 4-hydroxynonenal resulted in the formation of multiple 4-hydroxynonenal-adducted protein adducts (Figure 6B). These results indicated a high sensitivity of prenatal liver hematopoietic stem cells to 4-hydroxynonenal and indicated that 4-hydroxynonenal-protein binding may be a key mechanism of injury to the cells.
Figure 6. Sensitivity of Cultured Prenatal Liver Hematopoietic Stem Cells to 4-Hydroxynonenal Toxicity and 4-Hydroxynonenal-Protein Adduct Formation. (A) Viability falls sharply upon treatment with low concentrations (5 µM) of 4-hydroxynonenal. Error bars indicate standard error of three experiments. (B) SDS-PAGE followed by Western analysis using a 4-hydroxynonenal-protein adduct antibody reveals formation of high molecular weight 4-hydroxynonenal-protein adducts with discrete banding. Electrophoretic standards were used to discern the noted molecular weights (kDa); molecular weight of the primary band subsequently was interpolated.
Collectively, our results using the prenatal liver fractions, liver slices, and hematopoietic stem cells strongly suggest that human prenatal liver cells are not disproportionally sensitive to injury to AFB1 as originally hypothesized. Metabolic byproducts of transplacental compounds, however, that cause oxidative injury (e.g., 4-hydroxynonenal production from maternal alcohol exposure) may be more toxicologically relevant than AFB1 with regards to fetal liver-derived hematopoietic progenitor cell injury.
Conclusions:
The results of this research yielded several important findings. We have established that the level of cytochrome P4503A7 (CYP3A7), the major prenatal hepatic CYP isozyme, is a primary determinant of AFBO-DNA and AFBO-protein binding in vitro in human prenatal liver. In contrast, the level of protective phase II detoxification enzymes such as GSTs or epoxide hydrolases, which are involved in the detoxification of epoxide intermediates, do not markedly reduce AFBO macromolecular binding in prenatal liver. Furthermore, the level of CYP3A7 shows extensive individual variation among prenatal liver donors, with higher CYP3A-expressing donors forming more AFBO-macromolecular adducts in vitro. Comparison studies using subcellular fractions from different age groups, however, indicate that second trimester human prenatal liver tissues do not appear to be highly sensitive to the formation of AFBO-macromolecular adduct formation. These results were supported by experiments with cultured prenatal liver slices and gene expression array analysis. Collectively, these data did not support our original hypothesis that the fetus may be at disproportionately high risk to AFB1 injury relative to adults. In addition to the interindividual variation observed in the expression of enzymes that mediate AFB1 biotransformation, we also demonstrated individual variability in the expression of alpha class GST-associated catalytic activities towards oxidative damage products. Collectively, such interindividual variation in biotransformation enzyme pathways may be a determinant of sensitivity to in utero injury to maternally transferred toxicants. Subsequent studies on the project focused on liver-based hematopoietic cells that could be important targets of transplacental compounds. These studies demonstrated the utility of using human prenatal hematopoietic stem cells as a tool for screening the effects of transplacental chemicals that are of relevance in childhood hematological diseases. We have shown that hematopoietic stem cells do not appear to be sensitive targets of AFB1, and this insensitivity to the formation of AFBO-DNA adducts appears to be the result of low CYP3A- expression. Thus, when considered collectively, the studies using relevant human cell models, including subcellular fractions, liver slices, and hematopoietic stem cells, underscore the importance of in utero gene-environment interactions as a determinant of susceptibility to chemical injury.
Journal Articles on this Report : 5 Displayed | Download in RIS Format
Other project views: | All 18 publications | 5 publications in selected types | All 5 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Doi AM, Patterson PE, Gallagher EP. Variability in aflatoxin B1-macromolecular binding and relationship to biotransformation enzyme expression in human prenatal and adult liver. Toxicology and Applied Pharmacology 2002;181(1):48-59. |
R827441 (2002) R827441 (Final) |
Exit |
|
Gallagher EP, Sheehy KM. Effects of phenytoin on glutathione status and oxidative stress biomarker gene mRNA levels in cultured precision human liver slices. Toxicological Sciences 2001;59(1):118-126. |
R827441 (2000) R827441 (Final) |
Exit Exit |
|
Gallagher EP, Gardner JL. Comparative expression of two alpha class glutathione S-transferases in human adult and prenatal liver tissues. Biochemical Pharmacology 2002;63(11):2025-2036. |
R827441 (2001) R827441 (Final) |
Exit Exit |
|
Gardner JL, Gallagher EP. Development of a peptide antibody specific to human glutathione S-transferase alpha 4-4 (hGSTA4-4) reveals preferential localization in human liver mitochondria. Archives of Biochemistry and Biophysics 2001;390(1):19-27. |
R827441 (Final) |
Exit |
|
Gardner JL, Doi AM, Pham RT, Huisden CM, Gallagher EP. Ontogenic differences in human liver 4-hydroxynonenal detoxification are associated with in vitro injury to fetal hematopoietic stem cells. Toxicology and Applied Pharmacology 2003;191(2):95-106. |
R827441 (2002) R827441 (Final) |
Exit Exit |
Supplemental Keywords:
risk, risk assessment, health effects, human health, metabolism, vulnerability, sensitive populations, carcinogen, teratogen, mutagen, cellular, population, enzymes, infants, children, age, diet, genetic predisposition, susceptibility, chemicals, toxics, decisionmaking,, RFA, Scientific Discipline, Health, Genetics, Health Risk Assessment, Epidemiology, Risk Assessments, Susceptibility/Sensitive Population/Genetic Susceptibility, Biochemistry, Children's Health, genetic susceptability, Molecular Biology/Genetics, sensitive populations, aflatoxin B1, heterocyclic amines, adolescents, carcinogenesis, childhood cancer, fetal metabolism, cytochrome P450, detoxification enzymes, exposure, DNA reactive metabolites, dietary procarcinogens, children, cancer risks, human exposure, susceptibility, children's vulnerablity, assessment of exposure, genetic risk factors, biotransformation, epidemeology, environmentally caused disease, bioactivated environmental toxicants, transplacental exposure to mutagenic agents, aflotoxin, biomedical research, genetic susceptibility, toxicsRelevant Websites:
http://www.ufbi.ufl.edu/physdept/gallagher.htm Exit
http://www.floridatox.org/ Exit
http://depts.washington.edu/envhlth/about/faculty.html Exit
Progress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.