Michael C. Kohn

EFFECTS OF TCDD ON THYROID HORMONE HOMEOSTASIS IN THE RAT

A physiological dosimetric model was constructed to describe the effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on circulating thyroid hormones in the rat and to test the hypothesis that these hormonal changes cause chronically elevated serum thyrotropin (thyroid stimulating hormone, TSH), which mediates growth promotion and may lead to thyroid tumors in TCDD-treated rats. The model included diffusion restricted distribution of TCDD among compartments for liver, kidney, white fat, slowly and rapidly perfused tissues, and the thyroxine-sensitive tissues brown fat, pituitary, and thyroid. Blood was distributed among major vessels and the capillary beds of the tissues. Metabolism of TCDD was limited to the liver. Secretion of 3,5,3′-triiodothyronine (T3) and thyroxine (3,5,3′,5′-tetraiodothyronine, T4) from the thyroid was modeled as stimulated by circulating TSH, whose release from the pituitary was regulated by the hypothalamic peptides thyrotropin releasing hormone (activating) and somatostatin (inhibiting). Release of these peptides was represented as inhibited and activated, respectively, by circulating T4. Binding proteins for T3 and T4 and metabolism of the hormones by deiodination were included in thyroxine-sensitive tissues. Induction of hepatic UDP-glucuronosyltransferase-1*6 (UGT), the enzyme which glucuronidates T4, was modeled as induced by the complex formed between TCDD and the aryl hydrocarbon receptor. The computed extent of deiodination, primacy of the thyroid in generating T3 from T4, dependence of liver and kidney on locally produced T3, and export of T3 formed in the pituitary agreed with experimental observations. The model reproduced the observed decrease in circulating T4 and elevated serum TSH following chronic administration of TCDD. The altered levels were attributed to the increased clearance of T4 by the induced UGT and the consequent modification of feedback control of hormone releases. These results are consistent with the hypothesis of growth stimulation by elevated TSH, but measured values of this hormone in blood of rats vary over a large range, and the change induced by TCDD is often small. Measured UGT levels are less variable and the increase in this protein is much greater, suggesting that this response may be a more reliable biomarker for effects of TCDD on the thyroid.

Physiological modeling of a proposed mechanism of enzyme induction by TCDD

A physiological model was previously constructed to facilitate extrapolation of surrogates for the effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in rat liver to doses comparable to human environmental exposures. The model included induction of P450 isozymes and suggested the presence of multiple binding sites with different affinities for the TCDD-liganded Ah receptor at CYP1A1 dioxin responsive elements. The model also indicated that protein synthesis on the mRNA template exhibited saturation kinetics with respect to message levels. In the present work the earlier model was revised to include the increased proteolysis of the Ah receptor on binding TCDD, more realistic representations of gene transcription and mRNA translation, and different stability for each mRNA. The revised model includes multiple TCDD-liganded Ah receptor binding sites for CYP1A1 and CYP1B1 genes, a lag of 0.2 day for production of mRNA and induced proteins, and stabilization of mRNA by a poly(A) tail. The model reproduced the transient depletion of the Ah receptor subsequent to binding ligand and the dose-response of the receptor in rats treated with biweekly oral doses of TCDD in corn oil. The model reproduced tissue TCDD concentrations observed for several dosing scenarios. Such robustness indicates the utility of the model in estimating internal dose. The model also reproduced the observed dose-response patterns for mRNA and protein for CYP1A1, CYP1A2, and CYP1B1 after repeated dosing. Neither of the two dissociation constants for the Ah receptor bound to the CYP1B1 gene is negligible, supporting the assumption of multiple response elements for this gene. The poorer induction of CYP1B1 was predicted to be due to lower affinity of the dioxin responsive elements for binding the liganded Ah receptor, suggesting the involvement of other regulatory factors, and a shorter poly(A) tail on CYP1B1 mRNA, leading to a shorter lifetime. Saturation in the kinetics of protein synthesis was linked to the limited number of ribosomes that could bind to each message molecule, resulting in fewer ribosomes bound per message at higher doses. Predicted induction at low doses was found to vary widely with the assumptions used in the construction of a model. More detailed descriptions of biological processes might provide more reliable predictions of enzyme induction.

Achieving credibility in risk assessment models

Validation of a mathematical model requires demonstrating that a model is free of mathematical errors (internal consistency), is sensitive to large but not small errors or uncertainties in parameter values (verifiability and robustness), reproduces experimental observations on the system being modeled (external consistency), and leads to testable predictions of the systems biological properties. To be heuristically valid, a model also must be a realistic representation of the actual biological system. Only then would the models predictions be credible to the wider community of biological scientists who would use the model for risk assessment and dose or species extrapolation. Owing to incomplete data, most current dosimetric models are insufficiently realistic to pass this test of credibility. Enhancements to such models that would help achieve credibility are presented, and suggestions are offered for institutionalizing realistic modeling practices in risk assessment.

Absolute estimation of initial concentrations of amplicon in a real-time RT-PCR process

BackgroundSince real time PCR was first developed, several approaches to estimating the initial quantity of template in an RT-PCR reaction have been tried. While initially only the early thermal cycles corresponding to exponential duplication were used, lately there has been an effort to use all of the cycles in a PCR. The efforts have included both fitting empirical sigmoid curves and more elaborate mechanistic models that explore the chemical reactions taking place during each cycle. The more elaborate mechanistic models require many more parameters than can be fit from a single amplification, while the empirical models provide little insight and are difficult to tailor to specific reactants.ResultsWe directly estimate the initial amount of amplicon using a simplified mechanistic model based on chemical reactions in the annealing step of the PCR. The basic model includes the duplication of DNA with the digestion of Taqman probe and the re-annealing between previously synthesized DNA strands of opposite orientation. By modelling the amount of Taqman probe digested and matching that with the observed fluorescence, the conversion factor between the number of fluorescing dye molecules and observed fluorescent emission can be estimated, along with the absolute initial amount of amplicon and the rate parameter for re-annealing. The model is applied to several PCR reactions with known amounts of amplicon and is shown to work reasonably well. An expanded version of the model allows duplication of amplicon without release of fluorescent dye, by adding 1 more parameter to the model. The additional process is helpful in most cases where the initial primer concentration exceeds the initial probe concentration. Software for applying the algorithm to data may be downloaded at http://www.niehs.nih.gov/research/resources/software/pcranalyzer/ConclusionWe present proof of the principle that a mechanistically based model can be fit to observations from a single PCR amplification. Initial amounts of amplicon are well estimated without using a standard solution. Using the ratio of the predicted initial amounts of amplicon from 2 PCRs is shown to work well even when the absolute amounts of amplicon are underestimated in the individual PCRs.

Physiological modeling of butadiene disposition in mice and rats

The earliest physiological models of 1,3-butadiene disposition reproduced uptake of the gas from closed chambers but over-predicted steady-state circulating concentrations of the mutagenic intermediates 1,2-epoxybut-3-ene and 1,2:3,4-diepoxybutane. A preliminary model based on the observation of a transient complex between cytochrome P450 and microsomal epoxide hydrolase on the endoplasmic reticulum membrane reproduced the blood epoxide concentrations as well as the chamber uptake data. This model was enhanced by the addition of equations for the production and detoxication of 3,4-epoxybutane-1,2-diol in the liver, lungs, and kidneys. The model includes flow-restricted delivery of butadiene and its metabolites to compartments for lungs, liver, fat, kidneys, gastrointestinal tract, other rapidly perfused tissues, and other slowly perfused tissues. Blood was distributed among compartments for arterial, venous, and tissue capillary spaces. Channeling of the three bound epoxides to epoxide hydrolase and their release from the endoplasmic reticulum are competing processes in this model. Parameters were estimated to fit data for chamber uptake of butadiene and epoxybutene, steady-state blood concentrations of epoxybutene and diepoxybutane, and the fractions of the inhaled dose of butadiene that appears as various excreted metabolites. The optimal values of the apparent K(m)s of membrane-bound epoxides for epoxide hydrolase were only 5% of the values for the cytosolic substrate, consistent with the observation of a transient complex between epoxide hydrolase and the cytochrome P450 that produces the epoxide. This proximity effect corresponds to the notion that epoxides produced in situ have privileged access to epoxide hydrolase. The model also predicts considerable accumulation of epoxybutanediol, in agreement with the observation that most of the DNA adducts in animals exposed to butadiene arise from this metabolite.

Effects of the structure of a toxicokinetic model of butadiene inhalation exposure on computed production of carcinogenic intermediates

A flow-limited physiologically based toxicokinetic model was constructed for uptake, metabolism, and clearance of butadiene (BD) and its principal metabolite 1,2-epoxy-3-butene (EB), using physiological and biochemical parameters from the literature where available. The model includes compartments for blood, liver, lung, fat, GI tract, other rapidly perfused tissues, and slowly perfused tissues. The blood was distributed among compartments for arterial plus venous blood and subcompartments for vascular spaces associated with each of the tissue compartments. The lung contained a subcompartment for the alveolar space. Metabolic activation of BD by cytochrome P450-catalyzed epoxidation was modeled as occurring in liver, lung, and the rapidly perfused tissue compartments. The detoxication of EB catalyzed by epoxide hydrolase and glutathione S-transferase (GST) was modeled as occurring in liver, lung, and the rapidly perfused tissues compartments and by blood GST activity. The model also includes depletion of glutathione (GSH) by GST-catalyzed conjugation of EB and 3-butene-1,2-diol and resynthesis of GSH from cysteine. Values of biochemical parameters that were unavailable in the literature were estimated by iteratively reweighted least squares optimization to reproduce data for uptake of BD and EB by rats and mice in closed chambers. The resulting model also reproduced the depletion of GSH in liver and lung in flow-through systems. It reproduced the concentrations of expired EB produced from BD in closed chambers but overpredicted separately measured blood EB concentrations in flow-through systems, indicating an inconsistency between these two experiments that cannot be resolved by this model or an inadequacy in the model. Equilibration of chamber gases with the alveolar space and alveolar gas with lung capillary blood results in much less dilution of the inhaled gas in the blood compared with the predictions of models in which chamber gas equilibrates directly with the total circulation. The production of EB predicted by the present model was found to be sensitive to a number of physiological and biochemical parameters. A valid and useful toxicokinetic model must have reliable physiological and enzymological data for BD biotransformation before it can be credibly used for human risk assessment.

DOSE-RESPONSE ANALYSES OF EXPERIMENTAL CANCER DATA

Dose-response analysis provides a powerful tool to determine causality from experimental cancer data, estimate low-dose risk, and evaluate mechanistic hypotheses. However, the interpretation of cancer dose-response data can be influenced by how the dose and response terms are characterized. Using the poly-3 quantal response method to adjust for the extensive and early development of lethal lymphomas in butadiene-exposed mice provided a means of obtaining a better representation of dose-response relationships for late-developing tumors induced by this chemical. Fitting a Weibull model to survival-adjusted tumor data for chloroprene and butadiene indicated similar carcinogenic potencies for these chemicals in mice. In conjunction with the rodent toxicity and carcinogenicity studies conducted by the National Toxicology Program, toxicokinetic studies are performed to characterize relationships between exposure and tissue concentrations of parent compound and metabolites. A physiologically based pharmacokinetic model (PBPK) of butadiene dosimetry indicated that differences in carcinogenic response between rats and mice are not simply due to differences in tissue concentrations of epoxybutene, a mutagenic metabolic intermediate. Thus, factors beyond tissue dosimetry of this metabolite must be important in butadiene-induced carcinogenesis. A PBPK model for isoprene indicated that blood concentrations of isoprene epoxides are a better indicator of kidney cancer risk than care measurements of isoprene-exposure concentrations. An evaluation of dose-response relationships for cytotoxicity, regenerative hyperplasia, and tumor induction by trihalomethanes indicates that for this family of chemicals, cell proliferation is not a reliable predictor of tumor response.

Weight of evidence versus weight of speculation to evaluate the alpha2u-globulin hypothesis.

whereas the severity of spontaneous chronic nephropathy in treated male rats relative to controls increased from 52 and 104 weeks. As a result of the current discussion in the literature (3), the kidney specimens were reanalyzed for the hallmarks of x2u-globulin nephropathy sequelae, namely the presence of granular casts and linear mineralization (8). These changes were not observed in male rats receiving 250, 1,000, or 2,000 mg/kg of gabapentin for 2 years. In terms of proliferative effects, there were 3 foci of atypical hyperplasia of renal tubules: 2/50 were found in male rats at 1,000 mg/kg and 1/50 male rats at 2,000 mg/kg of gabapentin. These kidney examinations in the 2-year study included four sections (two from each kidney) from every animal in the group (50/sex/group), amounting to 1,600 kidney sections in the 2-year rat study. This standard procedure seems to adequately sample the kidney. Reference was made to renal accumulation of gabapentin and to its ability to bind a2u-globulin. Metabolic disposition data revealed no differences in the accumulation of gabapentin in the kidney of male and female rats. Gabapentin distributes rapidly to the kidney in both sexes, and approximately 10-12% of the dose is found at 2hr post-dose. The elimination decay is rapid, and 0.20-0.14 gabapentin radioequivalents in micrograms per gram are found at 12 hr post-dose. Full profile toxicokinetic studies have shown linear kinetics without accumulation, and elimination follows rapid clearance from the plasma and organ compartments. Recoveries in the mass balance studies yielded 99% or more of the administered dose. Why such a diverse series of chemicals causes a putatively similar microglobulin nephropathy with (i.e., unleaded gasoline) and without (i.e., gabapentin) nephrocarcinogenesis is intriguing. Evaluation of the gabapentin data does not prove or disprove that a2u-globulin accumulation precedes, promotes, or causes renal cancer in rats over a 2-year constant exposure at toxicity-limiting doses. We propose that although morphofunctional features may appear similar for all documented cases of this xenobioticinduced nephropathy, different mechanisms may be operative. Alternatively, a structural comparison of the x2u microglobulin species would assert this microglobulin not to be the same in all cases when induced by the diverse chemicals. The work by Hildebrand et al. (9) would indicate some support to this notion. In the case of gabapentin, in which we have substantial data on hand, it is difficult to advocate existing mechanisms because of the drugs lack of metabolism or biotransformation, lack of serum protein binding, or lack of significant organ accumulation. It remains unanswered whether data on cell proliferation or specific 4x2uglobulin binding to renal tubule proteins would substantiate current mechanisms. Clearly new data, rather than new assumptions, will give fresh insight to our understanding of the a2u nephropathy puzzle.

Biochemical mechanisms and cancer risk assessment models for dioxin

Biologically realistic mechanistic models of carcinogenesis by TCDD are composed of equations representing biochemical events leading to altered expression of proteins involved in the response or equations representing the kinetics of proliferation of clones of mutant cells. A biochemically augmented physiological dosimetry model reproduces the observed altered expression of liver proteins in female rats exposed to dioxin. The model suggests that oxidation of estradiol to DNA reactive quinones or semiquinones by CYP1A2 protein induced by TCDD may contribute to an increased mutational rate. It suggests that TCDD-stimulated production of a peptide ligand of the epidermal growth factor (EGF) receptor and subsequent activation of the receptors tyrosine kinase activity may increase the rate of proliferation of susceptible cells. These calculated quantities can serve as indices of toxicity and can be used to predict tumor incidence as a function of exposure.

Use of sensitivity analysis to assess reliability of metabolic and physiological models

Because ethical considerations often preclude directly determining the human health effects of treatments or interventions by experimentation, such effects are estimated by extrapolating reactions predicted from animal experiments. Under such conditions, it must be demonstrated that the reliability of the extrapolated predictions is not excessively affected by inherent data limitations and other components of model specification. This is especially true of high-level models composed of ad hoc algebraic equations whose parameters do not correspond to specific physical properties or processes. Models based on independent experimental data restricting the numerical space of parameters that do represent actual physical properties can be represented at a more detailed level. Sensitivities of the computed trajectories to parameter variations permit more detailed attribution of uncertainties in the predictions to these low-level properties. S-systems, in which parameters are estimated empirically, and physiological models, whose parameters can be estimated accurately from independent data, are used to illustrate the applicability of trajectory sensitivity analysis to lower-level models.