The use of mode of action information in risk assessment: Quantitative key events/dose-response framework for modeling the dose-response for key events

Figures & data

Figure 1. The HESI RISK21 Roadmap and Matrix.

Figure 2. Quantitative Key Events/Dose-Response Framework (Q-KEDRF) and Its Relationship to the Mode of Action/Human Relevance Framework.

Table 1. Modulating Factors (ModFs) potentially affecting KEs for dose-response in humans. ModFs fall into three general categories shown in the left column. The middle column shows subcategories and the right hand column shows some aspects to consider.

Category	Sub-category	Aspects
Host Factors	Genetic Variation	Polymorphisms
	Disease/Illness	Chronic
		Acute
	Defense mechanisms	Immune responsiveness
		DNA repair
		Cell proliferation
		Cell death
	Physiology	Sex
		Life stage
		ADME
		Hormonal status
Life Style	Diet	Calories
		Fat content
	Tobacco	Usage
	Alcohol	Usage
	Exercise	Frequency
		Intensity
	Pharmaceuticals	Usage
	Illegal drugs	Usage
	Dietary supplements	Vitamins
		Anti-oxidants
Environment	Co-Exposures	Duration
		Air
		Water
		Food
		Dust
		Occupational

Table 2. Dose-time concordance table for dimethylarsinic acid.

	Table —Dose-Time Concordance
Time		2 weeks	2–3 weeks	10 weeks	25 weeks	104 weeks
Dose (ppm in diet)		Increasing time
	2	Metabolism*	Metabolism*	Metabolism* Cytotoxicity	Metabolism* Cytotoxicity*	Metabolism* Cytotoxicity*
	10	Metabolism*	Metabolism* Cytotoxicity	Metabolism* Cytotoxicity	Metabolism* Cytotoxicity*	Metabolism* Cytotoxicity*
	40	Metabolism*	Metabolism* Cytotoxicity	Metabolism* Cytotoxicity Proliferation Hyperplasia	Metabolism* Cytotoxicity* Proliferation* Hyperplasia	Metabolism* Cytotoxicity* Proliferation* Hyperplasia Carcinomas
	100	Metabolism*	Metabolism Cytotoxicity Proliferation Hyperplasia	Metabolism Cytotoxicity Proliferation Hyperplasia	Metabolism Cytotoxicity* Proliferation Hyperplasia	Metabolism* Cytotoxicity* Proliferation* Hyperplasia Carcinomas

The asterisk means that the key event has not been observed at the specific dose/time point but is presumed to have occurred. Although not used here, shading of the table may be helpful with a shading scheme based on the number of KEs. Figure 5 in Meek et al. (2014b) provides another organizational scheme for the dose-time concordance table (Please see Figure 3 for the MOA and text for details).

Table 3. Dose-Response Species Concordance Table for Key Events (KEs) in the MOA of dimethylarsinic acid (DMA^V) (Adapted from USEPA, 2005c).

	Qualitative concordance			Quantitative concordance and quantitative Dose-response
Event or factor	Animals	Humans	Concord-ance	Str.*	Animals	Humans
Key events
Key Event #1 Metabolism to DMA^III	DMA^III detected in urine following 26 weeks treatment with 100 ppm DMA^V	Evidence following DMA^V exposure too limited to draw conclusions, but DMA^III shown to be present following human exposure to iAs	Plausible	+/−		NA
Key Event #2 Urothelial Cytotoxicity	Urothelial toxicity observed in vivo in rats at 2 ppm but not enough for successive key events	Potential to occur in humans but unknown if sufficient DMA^III formed	Plausible	+/−		NA
Key Event #3 Urothelial Proliferation	observed at 0.5 mg/kg/d DMA^V	Potential to occur in humans but unknown if sufficient DMA^III formed	Plausible	+/−		NA
Key Event #4 Hyperplasia	observed at 2 mg/kg/d or 0.3 to 2 μmol DMA^III in urine	Potential to occur in humans but unknown if sufficient DMA^III formed	Plausible	+/−		NA
Apical Event Tumors	observed at 5 mg/kg/d DMA^V or 0.8 to 5.05 μmol DMA^III in urine	No data in humans	Concordance cannot be made because there is no human data	−		NA

*Str. = strength.

Figure 3. Use of MOA in the HESI RISK21 Matrix. Left: MOA for Tumor Induction by Dimethylarsinic Acid (DMA^V; Cacodylic Acid) that includes cytotoxicity, regenerative proliferation, and hyperplasia. This MOA is used to illustrate the dose-time concordance table and dose-response species concordance table (Tables 2 and 3). Right: Matrix showing the exposure estimates and toxicity range (BMDL₁₀ to RfD) for chronic dietary exposure, data from EPA, 2006.

Figure 4. Mode of Action of Chlorpyrifos showing metabolic activation to CPF-oxon and inhibition of acetylcholinesterase as the critical effect. (Figure courtesy of Dr. Alan Boobis).

Figure 5. PON1-mediated Vmax values vs. age (upper plot). PON1 functional phenotypes are represented by open circles, open triangles, and open squares for QQ, QR, and RR, respectively (see text for definitions). CPF- oxon hydrolysis V_max values in plasma over paraoxon hydrolysis activity (lower plot) resolves QQ and QR, but not QR and RR. (From Smith et al. 2011; permission to reproduce figures granted by Dr. Jordan Smith, 22 March 2013.).

Figure 6. Modeled chlorpyrifos pharmacokinetics in adults and children and resulting AChE inhibition in erythrocytes. A. RBC AChE inhibition from the AUC (left) and maximum CPF concentrations (right) in blood (from Hinderliter et al. 2011). B. Modeled time courses of CPF and CPF oxon in blood from dietary exposures (upper panel) and corresponding RBC AChE inhibition (lower panel). (Reprinted from Regulatory Toxicology and Pharmacology (Hinderliter, P.M., Price P.S., Bartels M.J., Timchalk C., Poet T.S. 2011. Development of a source-to-outcome model for dietary exposures to insecticide residues: An example using chlorpyrifos, Regul. Toxicol. Pharmacol. 61, 82–92) with permission from Elsevier.).

Table 4. Dose-response concordance table for Modulating Factors (MFs) in the MOA of chlorpyrifos.

	Qualitative concordance			Quantitative concordance and quantitative Dose-response
Event or factor	Animals	Humans	Concordance	Strength	Animals	Humans
Modulating Factors
MF and affected KE	Animals	Humans	Concordance	Strength	Effects in Animals	Effects in Humans
MF #1 Genetic Polymorphism	NA	R vs. Q allele	NA			QQ genotype more sensitive, but at current exposure levels this difference is not a factor
MF #2 Use of Statin drugs	NA	Statins increase PON1 activity	NA			Statins modestly increase PON1 activity, but the effect is not consistently observed
MF #3 Alcohol Use	NA	Alcohol use increases PON1 activity	NA			This effect is likely not a factor at current exposure levels

Figure 7. Putative MOA for the uterotrophic response.

Table 5. Cellular effects of modulating factors.

Effect	Sub-effect
Gene Structure	Mutations
	Deletions
	Duplications
Gene Expression	Transcription factors
	Co-activators/accelerators
	Co-repressors/decelerators
	Co-modulators
Post-translational modifications	Acetylation
	Methylation
	Phosphorylation
	Others

Figure 8. Dose-response plots for putative key events in the MOA for the uterotrophic response.

Table 6. Quantitative aspects of the dose-response of key events in the uterotrophic response.

			Transition dose values
Key event	Study	Hill model parameters	Starting points and slope-based TDVs (Murrell et al., 1998)	BMD21 as a transitional dose (Sand et al., 2006)
Binding to ERa	Levin et al. (1993) (cytosol) (fractional binding response)	Kd = 31.2 nM Log Kd = 1.49 n = 0.76	(13.8, 102.8) (5.8, 62.4) 1.26 nM	5.41 nM
	Levin et al. (1993) (nucleus) (fractional binding response	Kd = 2.04 nM Log Kd = 0.310 n = 1.94	(1.49, 27.3) (2.85, 51.8) 1.05 nM	1.03 nM
	Notides et al. (1981)^a (fractional binding response)	Kd = 3.25 nM Log Kd = 0.512 n = 1.61	(1.853, 7.48) (4.99, 20.7) 1.11 nM	1.43 nM
Gene expression changes in relation to uterine weight gain (Naciff et al., 2003)	Naciff et al. (2003) (fold increase in Ca binding protein)	Bmax = 22.82 fold Kd = 0.807 μg/kg/d Log Kd = -0.0930 n = 0.755	(0.1, 3.5) (1.0, 12.5) 0.082 μg/kg/d	0.140 μg/kg/d
	Naciff et al. (2003) (fold increase in uterine weight)	Bmax = 5.48 fold Kd = 1.26 μg/kg/d Log Kd = 0.010 n = 0.914	(0.1, 1.12) (1.0, 4.63) 0.171 μg/kg/d	0.240 μg/kg/d
Gene expression changes in relation to uterine weight gain (Heneweer et al., 2007)	Heneweer et al. (2007) (fold increase in Ca binding protein)	Bmax = 12.66 fold Kd = 5.35 μg/kg/d Log Kd = 0.728 n = 1.54	(0.3, 2.62) (1.0, 9.1) 0.290 μg/kg/d	2.26 μg/kg/d
	Heneweer et al. (2007) (fold increase)	Bmax = 3.8 fold Kd = 15.65 μg/kg/d Log Kd = 1.195 n = 1.191	(1.0, 2.02) (10.0, 3.79) 0.220 μg/kg/d	5.15 μg/kg/d
Cell proliferation	Kaye et al. (1971) (increase in mitotic index	Bmax = 276 figures Kd = 0.809 μg/kg Log Kd = -0.092 n = 2.21	(1.5, 166) (15, 227) 0.0073 μg/animal/d	0.444 μg/animal/d
Increase in blood flow measure by uterine peroxidase	Lyttle and DeSombre (1977)	Bmax = 69 units/g Kd = 17.3 μg/animal Log Kd = 1.24 N = 0.561	(1.0, 13.8) (10.0, 37.6) 0.053 μg/animal/d	1.64 μg/animal/d
Uterine weight gain	Branham et al. (1985) (mg wet weight)	Bmax = 5.4 fold Kd = 1.85 μg/animal/d Log Kd = 0.268 n = 0.271	(0.1, 1.73) (10.0, 5.31) 0.078 μg/animal/d	0.014 μg/animal/d

Here, the results of two methods for determining transitional dose values (TDVs) are shown. Details of the calculation methods are provided in the text. The form of the Hill model used here is shown in Table 7.

^aNotides et al. (1981) observe the Hill coefficient for E2 binding to cell-free preparations ERα varies with the concentration of the receptor (from n = 1.1 at 0.3 nM ERα to 1.6 at 3.0 and 4.8 nM ERα, indicating that the Hill coefficient increases with increasing concentrations of ERα.

Table 7. Inverse equations and slope equations of dose-response models from EPA’s benchmark dose software (USEPA 2012) to enable estimation of baseline projection values.

Figure 9. Details of one of the heterologous expression systems that could be used to substitute for the uterotrophic assay. Left: Stably transfected Luc reporter plasmid BG1Luc4E2 cell line from ICCVAM. Right: Concentration-response of the BG1Luc4E2 cells to estradiol showing fits to both first- and second-order Hill functions and the results of the transitional dose value calculation using the baseline projection method (Eq. 3,4 and 5). Please see Supplementary Content for another example.

Meek ME, Boobis A, Cote I, Dellarco V, Fotakis G, Munn S, et al. (2014b). New developments in the evolution and application of the WHO/IPCS framework on mode of action/species concordance analysis. J Appl Toxicol, 34, 1–18.

 PubMed Web of Science ®Google Scholar

USEPA (United States Environmental Protection Agency). (2005c). Science Issue Paper: Mode of Carcinogenic Action for Cacodylic Acid V (Dimethylarsinic Acid, DMA) and Recommendations for Dose Response Extrapolation.

 Google Scholar

Smith JN, Timchalk C, Bartels MJ, Poet TS. (2011). In vitro age-dependent enzymatic metabolism of chlorpyrifos and chlorpyrifos-oxon in human hepatic microsomes and chlorpyrifos-oxon in plasma. Drug. Metab. Dispos, 39, 1353–62.

 PubMed Web of Science ®Google Scholar

Hinderliter PM, Price PS, Bartels MJ, Timchalk C, Poet TS. (2011). Development of a source-to-outcome model for dietary exposures to insecticide residues: an example using chlorpyrifos. Regul Toxicol Pharmacol, 61, 82–92.

 PubMed Web of Science ®Google Scholar

Murrell JA, Portier CJ, Morris RW. (1998). Characterizing dose-response: I: Critical assessment of the benchmark dose concept. Risk Anal, 18, 13–26.

 PubMed Web of Science ®Google Scholar

Sand S, Von Rosen D, Victorin K, Filipsson AF. (2006). Identification of a critical dose level for risk assessment: developments in benchmark dose analysis of continuous endpoints. Toxicol Sci, 90, 241–51.

 PubMed Web of Science ®Google Scholar

Levin E, Actis AM, Lopez S. (1993). Characterization of rat uterine estrogen receptors in vivo. J Steroid Biochem Mol Biol, 44, 277–85.

 PubMed Web of Science ®Google Scholar

Notides AC, Lerner N, Hamilton DE. (1981). Positive cooperativity of the estrogen receptor. Proc Natl Acad Sci U. S. A, 78, 4926–30.

 PubMed Web of Science ®Google Scholar

Naciff JM, Overmann GJ, Torontali SM, Carr GJ, Tiesman JP, Richardson BD, Daston GP. (2003). Gene expression profile induced by 17 alpha-ethynyl estradiol in the prepubertal female reproductive system of the rat. Toxicol Sci, 72, 314–30.

 PubMed Web of Science ®Google Scholar

Heneweer M, Houtman R, Poortman J, Groot M, Maliepaard C, Peijnenburg A. (2007). Estrogenic effects in the immature rat uterus after dietary exposure to ethinylestradiol and zearalenone using a systems biology approach. Toxicol Sci, 99, 303–14.

 PubMed Web of Science ®Google Scholar

Kaye AM, Icekson I, Lindner HR. (1971). Stimulation by estrogens of ornithine and S-adenosylmethionine decarboxylases in the immature rat uterus. Biochim Biophys Acta, 252, 150–9.

 PubMed Web of Science ®Google Scholar

Lyttle CR, Desombre ER. (1977). Uterine peroxidase as a marker for estrogen action. Proc Natl Acad Sci U. S. A, 74, 3162–6.

 PubMed Web of Science ®Google Scholar