Environmental and Molecular Mutagenesis 55:292^298 (2014)
Commentary Are We Ready to Consider Transgenerational Epigenetic Effects in Human Health Risk Assessment? Rebecca A. Alyea,1 B. Bhaskar Gollapudi2 and Reza J. Rasoulpour1* 1
Toxicology and Environmental Research and Consulting, The Dow Chemical Company, Midland, Michigan 2 Center for Toxicology and Mechanistic Biology, Exponent, Inc., Midland, Michigan
Recently, there has been a growing concern that chemically or nutritionally mediated epigenetic changes might lead to adverse health outcomes. The natural question is whether the existing chemical safety assessment paradigm is or is not protective of epigenetic-mediated effects, and if there is a need to incorporate new endpoints to specifically address epigenetics. Of particular interest are transgenerational epigenetic effects, which can be passed on through multiple generations. To investigate these questions, a comparison was performed between OECD guideline rat toxicology studies versus several rat transgenerational epigenetic studies. This analysis focused on vinclozolin owing to the availability of a comprehensive suite of dose-response data (NOAEL, reference dose, and human exposure estimates) for both conventional and epigenetic endpoints. This analysis revealed that vinclozolin transgenerational effects were demonstrated at a dose level
(100 mg/kg/day) that was: (1) 40-fold higher than the overall lowest-observed-adverse-effect level (LOAEL) from rat guideline studies, (2) 80-fold higher than the lowest NOAEL from rat guideline studies, (3) 80,000-fold higher than the reference dose for the molecule, and (4) 1.2-million fold above human exposure estimates. Through this analysis, we conclude that additional research across a spectrum of doses is necessary to elucidate the interplay between epigenetics and apical endpoints before considering epigenetics in human health risk assessment. Therefore, we recommend focusing future research toward (1) examining for potential causal relationships between epigenetic alterations and adverse apical endpoints, and (2) understanding the dose-response relationship of these causal epigenetic alterations when compared with those of the apical endpoints. Environ. Mol. Mutagen. 55:292–298, C 2013 Wiley Periodicals, Inc. 2014. V
Key words: epigenetics; product safety assessment; transgenerational effects
INTRODUCTION Epigenetics can be defined as heritable marks that are superimposed on the genome in the absence of direct changes in the DNA sequence. The mechanisms controlling these modifications are numerous and include, but are not limited to: DNA marks such as methyl-, hydroxymethy-, formyl-, and carboxyl-cytosine, histone modifications such as methylation, acetylation, phosphorylation, sumoylation, and ubiquitination, and variable miRNA expression, all of which act in concert to regulate gene expression [Cantone and Fisher, 2013; Harmston and Lenhard, 2013]. These epigenetic modifications are controlled by a higher level of regulatory machinery, which includes DNA (de-) methyltransferases, (de-) acetylases, ubiquitin lyases, chromatin remodeling machinery, and methyl-CpG-binding proteins [De Carvalho et al., 2010; Mohammad and Baylin, 2010]. These epigenetic modifications are key regulatory C 2013 Wiley Periodicals, Inc. V
steps in cellular differentiation and commitment to embryonic lineages as part of the normal biological processes. Within embryonic development this seemingly infinite combination of coordinated epigenetic events was described by Waddington in 1957 as a rolling landscape that represents the backdrop of molecular events guiding differentiation in the causal progression of zygote to adult
*Correspondence to: Reza J. Rasoulpour, PhD, The Dow Chemical Company, Toxicology & Environmental Research and Consulting, 1803 Building, Midland, MI 48674, USA. E-mail: [email protected] Received 31 July 2013; provisionally accepted 25 October 2013; and and in final form 28 November 2013 DOI 10.1002/em.21831 Published online 21 November 2013 in Wiley Online Library (wileyonlinelibrary.com).
Environmental and Molecular Mutagenesis. DOI 10.1002/em Epigenetics in Human Health Risk Assessment
[Waddington, 1957]. Epigenetic mechanisms play a critical role in determining the timing and progression of developmental processes within this rolling landscape. Epigenetic modulation of gene expression during development is evidenced by events such as X-chromosome inactivation, genomic imprinting, maternal and paternal allelic expression, tissue-specific gene expression, and cellular differentiation. During mammalian development, there are two periods of epigenetic reprogramming, one occurring during early embryonic development and another during primordial germ cell development. After fertilization both the paternal and maternal genomes undergo a period of DNA demethylation, followed by a remethylation phase coincident with the development of the somatic lineages [Hackett and Surani, 2013]. The second phase of reprogramming occurs in primordial germ cells during late embryogenesis. Presumably, the first phase of epigenetic reprogramming erases epigenetic modifications created during gametogenesis; however, there is evidence linking alterations in epigenetic mechanisms to various human disease states including immunodeficiency, Rett syndrome, Angelman, and Beckwith–Weideman Syndromes [Amir et al., 1999; Choufani et al., 2013; Lu, 2013]. Although the potential for perturbations within these processes leading to adverse health outcomes is a major area of research, many questions still remain for characterizing the role, temporal and spatial regulation, and modifications of epigenetic mechanisms during normal development. One area of particular interest within the field of epigenetics as it pertains to toxicology is the concept of a transgenerational epigenetic phenomenon. In these situations, it is hypothesized that an initial exposure during in utero development results in adverse apical outcomes not only for the parent, F1, and F2 generations, but also extends out to the F3 and later generations. This transgenerational phenomenon assumes that the epigenetic heritability can be sustained through generations in the absence of persistent exposure. For the purpose of this commentary, we will define the causative factor of the perturbation as a result of a chemically and/or nutritionally induced alteration. To observe a heritable transgenerational epigenetic effect following in utero exposure, the study should characterize stable epigenetic modifications that are causative of adverse health outcomes through three generations [Skinner, 2008]. Exposure of the dam leads to direct exposure of the developing fetus or the first generation as well as the germ cells within the fetus, which will potentially become the second generation. The third generation represents the first nonexposed generation, thus representing the true transgenerational population possible of displaying heritable modifications. Although our understanding of the mechanisms is continually evolving as we discover new states and interactions through basic research, questions surrounding the
293
inclusion of epigenetic modifications in human health risk assessment continue [Goodman et al., 2010; LeBaron et al., 2010; Rasoulpour et al., 2011; Alyea et al., 2012]. This commentary addresses if we are ready to consider epigenetic endpoints in human health risk assessment, with a particular focus on transgenerational effects. Before integrating a new testing strategy or evaluating additional endpoints within a product safety assessment paradigm, we must first evaluate what information is required and how such information is used in the risk assessment process.
PRODUCT SAFETY ASSESSMENT Fundamentally, risk is a function of the dose-response relationship of a hazard combined with potential human exposure. This concept is the tenant for safety assessment for industrial and agrochemicals, wherein the intrinsic hazards of a particular molecule are elucidated through toxicity studies, exposure is estimated based upon a combination of real-world data, computer modeling, and conservative assumptions, and these two sets of data are compared to determine the margin of exposure (i.e., distance between exposure and the lowest relevant noobserved-adverse-effect level). Multiple endpoints are evaluated and studies are performed to fully characterize the potential risk of a particular chemical, which include QSAR approaches, in vitro testing, and animal testing, which all help define the hazard fingerprint and the doseresponse relationship for each of the potential hazards identified. These data are used both in risk assessments as well as in establishing the appropriate hazard warning labels per programs such as the Globally Harmonized System (GHS) of Classification, Labeling, and Packaging (CLP). Additional factors are examined including potential environmental impact, toxicokinetic properties, and exposure assessments. The hazard identification component of the assessment relied on toxicity testing to characterize the potential hazards of a compound by evaluating adverse apical or phenotypic outcomes such as changes in body weight, body weight gain, feed consumption, organ weights, histopathological observations, and reproductive or developmental toxicity endpoints. This testing strategy integrates acute toxicity, immunotoxicity, systemic toxicity, specific target organ toxicity, reproductive toxicity, developmental toxicity, and carcinogenicity data to determine the lowest relevant no-observed-adverse-effect levels (NOAELs), which are then used in performing the risk assessment [OECD. 2013]. Toxicology hazard identification starts with evaluation of the physical chemical properties of a molecule, presence of reactive functional groups, and structure–activity relationships either through QSAR or expert judgment. These initial investigations can be facilitated through in silico technologies such as database mining or computational
Environmental and Molecular Mutagenesis. DOI 10.1002/em 294
Alyea et al.
modeling programs (e.g., DEREK or OASIS). Following these in silico approaches, acute toxicity parameters of oral, inhalation, and dermal lethality, eye and skin irritancy, and sensitization potential have been increasingly elucidated through in vitro tools. These core data can be quite useful in informing the material safety datasheet (MSDS) as well as understanding potential hazard for occupational exposure. Higher tier evaluations such as in vivo short-term, subchronic, chronic, and oncogenicity studies can also be performed in rodent species for agrochemicals and higher production volume industrial chemicals. Within these studies, clinical pathology (e.g., clinical chemistry, hematology, and urinalysis) combines with gross pathological and histopathological examination of all tissues and organs to understand the potential adaptive and adverse effects as well as carcinogenicity potential (see OECD test guidelines 407, 408, 451, 452, and 453 for more information). Often endpoints that drive the LOAEL are microscopic alterations within a particular tissue that can be identified by histopathological examination from long-term chronic or oncogenicity studies (i.e., 1- to 2-year studies). Developmental and reproductive toxicity studies are performed in rodents (and rabbits for developmental toxicity) to understand the potential of molecules to specifically cause adverse effects in neonatal growth, offspring survival, malformations (teratogenicity), functional deficits, puberty onset, and parental reproductive performance. These studies are typically performed under the OECD test guidelines 421, 422, 416, or 443. In addition to developmental and reproductive parameters, histopathology is performed on the major reproductive tissues, which is often the most sensitive endpoint for identification of endocrine active molecules. For a more thorough discussion and review of each of these testing strategies and guidelines that outline the evaluated endpoints refer to OECD guidance [OECD. 2013]. To perform the risk or product safety assessment, all of the relevant and available toxicology data are evaluated within a weight-of-evidence approach. During this evaluation, the lowest NOAEL across species and study types with findings of toxicological concern that have relevance to humans are used to derive such values as reference dose (RfD), allowable daily intake (ADI), and permitted daily exposure (PDE) for a molecule. In this process, the NOAEL is typically divided uncertainty factors to account for toxicokinetic and toxicodynamic difference between the species (interspecies uncertainty factor; rodent to human) and within the species (intraspecies uncertainty factor; human to human). Other potential uncertainty factors could be short-term to chronic extrapolations, database insufficiencies, LOAEL to NOAEL derivations, as well as the Food Quality Protection Act (FQPA) uncertainty factor. From a standpoint of hazard identification, the utility of characterizing adverse endpoints primarily relies on
anchoring adverse effects to the LOAEL and then determining the NOAEL for these apical endpoints. Characterization of these apical endpoints relies on expert judgment, biological plausibility, regulatory guidance documents, and precedent from the scientific literature or other laboratories. The core of a product safety assessment depends on characterizing the adverse apical endpoint and not necessarily identifying the mechanisms underlying these events. The reliance on the apical endpoint is critical to product safety assessment as it represents an effect at a dose that could not be regulated homeostatically or through adaptive mechanisms. If we conceptually imagine the genome as forming the base of a triangle (Fig. 1), exposure to exogenous agents at higher doses is likely to result in a large number of changes at the level of the transcriptome, which work together or in opposition to regulate fewer changes at the level of the proteome, which work again together or in opposition to regulate the physiome and finally culminating in an apical effect. At lower dose levels, there may be changes at the level of the transcriptome, which do not result in adverse apical effects. These latter changes would be classified as adaptive changes that are a normal part of biology and would be expected whenever we eat certain foods, exercise, or even breathe. It is the culmination of these changes at a particular dose level to an adverse apical effect that is generally used to characterize the dose-response relationship for the purposes of public health protection. However, what is currently unknown is the potential role or roles that the epigenome can play within the triangle of biological responses (Fig. 1). EPIGENETIC CASE STUDY As previously stated, epigenetic programming plays a critical role in normal developmental processes by altering gene expression in a temporally and spatially regulated fashion. The underlying factors that control these processes are by definition mechanisms and as such are
Fig. 1. Culmination of multiple molecular events at the level of the genome, transcriptome, and physiome to apical endpoints.
Environmental and Molecular Mutagenesis. DOI 10.1002/em Epigenetics in Human Health Risk Assessment
not incorporated under the current product safety assessment process. As stated earlier, it is the adverse apical outcome resulting from these alterations that forms the foundation of the safety assessment. Recent reports of adverse transgenerational effects following in utero exposure to environmental compounds such as vinclozolin, BPA, dioxin (TCDD), and jet fuel [Anway et al., 2005, 2006a; Manikkam et al., 2012, 2013; Tracey et al., 2013] suggest that an evaluation of epigenetic mechanisms as an indicator of adverse transgenerational effects within the product safety assessment may be warranted. In addition to transgenerational studies, as a fungicide, vinclozolin has a complete toxicology data package that includes developmental (OECD 414) and reproductive (OECD 416) toxicity studies as well as cancer bioassays (OECD 451 and 453). Finally, vinclozolin also has a published risk assessment that contains reference dose information as well as human exposure estimates from the US EPA [EPA, 2000]. Taken together, vinclozolin represents an ideal molecule to perform a case study comparing conventional toxicology data to epigenetic data and placing these findings into context with human exposure. For these benefits, this case study focuses only on vinclozolin as an example to provide recommendations more broadly for the field. Vinclozolin Toxicity Studies Vinclozolin is a fungicide that, along with its major metabolites, acts as an antiandrogen leading to demasculinized male rat offspring as demonstrated by reduced anogenital distance, retained nipples, a vaginal pouch, cleft phallus with hypospadias, suprainguinal ectopic scrota/ testes, and altered sex accessory glands in male offspring following in utero exposure [Gray et al., 1994, 1999; Matsuura et al., 2005]. In assessing the data presented within guideline regulatory studies to determine the lowest relevant NOAEL and LOAEL, it is clear that both the dose and developmental stage of exposure determine the extent of adverse effects [Gray et al., 1994, 1999; Wolf et al., 2000; Gray and Furr, 2008]. A multigenerational study in Wistar rats fed 0, 50, 300, 1,000, or 3,000 ppm vinclozolin reported adverse effects on fertility at 1,000 and 3,000 ppm [WHO-JMPR, 1995]. A marginal effect level of 50 ppm (4.5 mg/kg/day) was established based on the possibility of subfertility and signs of delayed development at 300 ppm. Additional multigenerational studies report a NOAEL of 40 ppm (4 mg/kg/day) following an assessment and establishment of no treatmentrelated effects in clinical signs, weight, feed consumption, reproduction, organ weight changes, or gross pathological findings [WHO-JMPR, 1995]. Exposure to 50 and 100 mg/kg/day via oral gavage from gestation day (GD) 14 to postnatal day (PND) 3 causes reduced fertility and 3.125 mg/kg/day causes a slight effect on anogenital distance,
295
establishing a LOAEL in this study of 3.125 mg/kg/day [Gray et al., 1999]. Based on a combined chronic toxicity and carcinogenicity study in rat, the lowest LOAEL and NOAEL is 2.3 and 1.2 mg/kg/day, respectively, for the development of Leydig cell tumors, which is most likely related to the antiandrogenic activity of vinclozolin [EPA, 2000]. The data presented in these toxicity evaluations are typical of the type of regulatory studies used to derive reference doses for use in product safety assessment and include establishment of a dose-response curve of adverse effects. Combining these studies together to examine the overall vinclozolin dose response reveals that: Dose levels 50 mg/kg/day causes reduced fertility in multigenerational reproductive toxicity studies as well as studies with GD 14–PND 3 exposure Dose levels between 3.125 and 50 mg/kg/day induce delayed preputial separation, slight effects on anogenital distance, and potential subfertility Dose level of 2.3 mg/kg/day in a cancer bioassay resulted in development of Leydig cell tumors likely owing to the antiandrogenic properties of the molecule Dose level of 1.2 mg/kg/day from the rat cancer bioassay was the lowest overall NOAEL for toxicity studies Therefore, the dose level of 1.2 mg/kg/day vinclozolin was used to derive the chronic reference dose, chronic population adjusted dose, and acceptable daily intake for use in risk assessment for protection of human health.
Vinclozolin Transgenerational Studies Interestingly, transgenerational research studies reported that following daily intraperitoneal (IP) injection of 100 mg/kg/day during GD 8–14 in SD rats causes adverse transgenerational effects presumably via an epigenetic alteration in the germ line [Anway et al., 2005]. The male offspring in the F1, F2, F3, and F4 generations displayed reduced sperm numbers and forward motility as well as an increase in the number of apoptotic germ cells per seminiferous tubule cross section. Follow-up studies further characterized a transgenerational reduction in spermatogenic capacity in both SD and Fisher F344 rats given 100 mg/kg/day [Anway et al., 2006b]. The epigenetic correlation was investigated in Fisher F344 rats given 100 mg/kg/day vinclozolin using methylation-specific restriction endonuclease, which identified 25 different PCR products that were differentially methylated [Anway et al., 2005]. Other investigators have reported a lack of transgenerational male reproductive effects following both IP injection and oral exposure to vinclozolin during the period of gonadal differentiation in male rats [(Gray and Furr, 2008; Schneider et al., 2008, 2013; Inawaka et al., 2009]. A recent study using IP injection of dams from GD 6–15
Environmental and Molecular Mutagenesis. DOI 10.1002/em 296
Alyea et al.
to 4 and 100 mg/kg/day found no effect on spermatogenesis or development of the male reproductive tract in offspring over four generations [Schneider et al., 2013]. A similar report of a lack of adverse transgenerational effects was also demonstrated over three generations via oral exposure of 4 and 100 mg/kg/day of vinclozolin, which is a more relevant route of exposure for human health assessment [Schneider et al., 2008]. The reason for the discordance between these studies may be due to the use of different strains, route of exposure, exposure window, undocumented protocol differences, diet, breeding, and/or colony variability. Comparison of Transgenerational versus Toxicity Studies Questions still remain in regard to defining an adverse transgenerational phenotype resulting from vinclozolin exposure. Furthermore, a causative relationship or correlated association to explain the mechanism (e.g., methylation changes with associated gene expression) by which these transgenerational effects could occur is not yet understood. In the absence of a causative association or correlation between the variations in methylation patterns and an adverse phenotype that results from this alteration, it is difficult to understand where this information would fit in the product safety assessment. In addition to questions surrounding variations in methylation patterns, if we review the current evidence of adverse apical effects and establishment of lowest relevant NOAEL and LOAELs, the overall NOAEL and LOAEL from the rat cancer bioassay is 1.2 and 2.3 mg/kg/day by the oral route of administration. Therefore, the transgenerational studies administering an IP injection of 100 mg/kg/day represent a dose level >40-fold higher than the LOAEL or >80fold higher than the NOAEL from apical endpoint studies. Therefore, it is unknown if potential transgenerational effects would occur at dose levels in the range of apical endpoint findings. In fact, the two studies that used dose levels within the range of the apical LOAELs (4 mg/kg/ day) by both oral and IP injection routes did not demonstrate a transgenerational effect [Schneider et al., 2008, 2013]. Comparing the animal data to human exposures within the risk assessment helps to further place these findings into context. The lowest NOAEL from the guideline animal studies was 1.2 mg/kg/day from the rat 2-year cancer bioassay. This overall NOAEL was used to derive a chronic reference dose (cRfD) by applying three sets of uncertainly factors (i.e., 103-intraspecies, 103-interspecies, and 103-FQPA) to derive 0.0012 mg/kg/day (Fig. 2) [EPA, 2000]. Moreover, human exposure for the US population, women, children, and infants is in the order of 34–78 ng/kg/day, which means that the reported transgenerational effects at 100 mg/kg/day are 1.2 millionfold above human exposures. Certainly, the data are
Fig. 2. Vinclozolin margin of exposure. Presented on a logarithmic axis of mg/kg/day dose are the rat IP injection transgenerational effects at 100 mg/kg/day as well as selected apical endpoint LOAELs and overall lowest NOAEL from the series of guideline toxicity studies. Rat study data present: 14.1 mg/kg/day is lowest LOAEL causing adverse reproductive effects, 22.3 mg/kg/day lowest LOAEL from chronic carcinogenicity assay, and 31.4 mg/kg/day is the lowest NOAEL [EPA, 2000]. Also presented are the chronic reference dose or population adjusted dose (cPAD) that applies 1,000-fold uncertainty factors to the 1.4 mg/kg/day lowest NOAEL, as well as human dietary exposure estimates in the total US population (40 ng/kg/day), women (34 ng/kg/day), infants (68 ng/kg/day), and children (78 ng/kg/day), which are data collected from the US EPA Reregistration Eligibility Decision (RED) [EPA, 2000].
scientifically interesting, but based upon a lack of doseresponse relationship data, the information within these studies does not provide evidence that would in fact alter the existing risk assessment for vinclozolin. In other words, there are no data to suggest that the risk assessment performed with apical endpoints on vinclozolin would not also be protective for any potential transgenerational epigenetic effects. RECOMMENDATIONS The field of epigenetics has grown exponentially in the past several years, as evidenced by a large increase in the scientific publications as well as coverage of this area within the popular press. The concept of transgenerational epigenetic effects in and of themselves even raises simplistic questions with regard to Darwinian versus Lamarkian evolution. Whether the genome or epigenome “wins” the battle of biological dominance remains to be seen, but this area of research has raised a large number of questions in multiple specialties within biological sciences. As it pertains to toxicology, there appear to be several data gaps with respect to causality as well as dose-response relationships that would help reveal the implications for epigenetic effects within product safety assessment. Our recommendations are for more research toward addressing two key areas, namely: (1) establishing causal relationships between epigenetic alterations and adverse apical endpoints, and (2) understanding the dose-response relationship of these causal epigenetic alterations with the apical endpoints. A key challenge to establish causal relationships between epigenetic changes and apical endpoints lies in the sheer number of potential epigenetic alterations that can occur. Moreover, a thorough understanding of the
Environmental and Molecular Mutagenesis. DOI 10.1002/em Epigenetics in Human Health Risk Assessment
normal biological variability within different tissues, organisms, and species, for methylation, histone modifications, and miRNA is still needed. By understanding the boundaries in the “noise” of biological variability for different epigenetic modifications at different areas of the genome would help provide the appropriate canvas to shape our understanding of adverse versus adaptive changes. Once normal variability is understood, identification of which epigenetic changes lead to which particular disease state can help establish the causal relationships needed to develop a concrete and scientific approach to understanding this area of research. Once epigenetic changes that are predictive of apical effects have been identified, examination of the doseresponse relationships of these epigenetic modifications in comparison to apical endpoints may provide toxicological insight into causality of these relationships. Therefore,
297
epigenetic studies with full dose-response relationships that anchor these changes to apical effect(s) would be extremely valuable in advancing our understanding of the contribution of epigenetics in toxicology. If the dose response for causal epigenetic changes lies on top (Fig. 3A) or to the right (Fig. 3B) of that for a given apical endpoint, then utilizing those apical endpoints for the purposes of product safety assessment would be appropriate. In other words, the product safety assessment based upon apical endpoints would also be protective of epigenetic changes similar to any molecular changes within our triangle from Figure 1. However, if persistent, causal epigenetic changes occur at dose levels below those of apical endpoints (Fig. 3C), then perhaps an alternative approach should be used to understand the implications to human health protection for those particular findings. Overall, future studies that examine this dose-response relationship using relevant routes of exposure at doses that are within both environmental exposure levels as well as doses used in hazard identification studies will aid in discerning what is meaningful for product safety assessment and human health protection. Although current evidence of epigenetic modulations causing adverse phenotypic outcomes is lacking in regard to establishing causality, this only implies that more research is needed to address these points. In going forward, we should learn from the existing studies and consider revising study designs to evaluate the potential correlative versus causal associations between adverse phenotypic effects and molecular endpoints by evaluating the relevance of the dose level and route of exposure, as well as considering the dose-response relationships for identification of NOAELs for causative molecular endpoints. ACKNOWLEDGMENTS The authors thank Drs. Matthew J. LeBaron, Michael R. Woolhiser, and Edward W. Carney for their review of the manuscript. REFERENCES
Fig. 3. Hypothetical dose-response curves for any given apical versus epigenetic effects. These dose-response curves can be similar to one another (A) as the early molecular changes are leading to apical endpoints, or there could be apical endpoints more sensitive than particular epigenetic changes (B). In contrast, causal epigenetic marks could be leftshifted of apical endpoints (C).
Alyea RA, Moore NP, LeBaron MJ, Gollapudi BB, Rasoulpour RJ. 2012. Is the current product safety assessment paradigm protective for epigenetic mechanisms? J Pharmacol Toxicol Methods 66:207–214. Amir RE, Van den Veyver IB, Wan M, Tran CQ, Francke U, Zoghbi HY. 1999. Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2. Nat Genet 23: 185–188. Anway MD, Cupp AS, Uzumcu M, Skinner MK. 2005. Epigenetic transgenerational actions of endocrine disruptors and male fertility. Science 308:1466–1469. Anway MD, Memon MA, Uzumcu M, Skinner MK. 2006a. Transgenerational effect of the endocrine disruptor vinclozolin on male spermatogenesis. J Androl 27:868–879. Anway MD, Leathers C, Skinner MK. 2006b. Endocrine disruptor vinclozolin induced epigenetic transgenerational adult-onset disease. Endocrinology 147:5515–5523.
Environmental and Molecular Mutagenesis. DOI 10.1002/em 298
Alyea et al.
Cantone I, Fisher AG. 2013. Epigenetic programming and reprogramming during development. Nat Struct Mol Biol 20:282–289. Choufani S, Shuman C, Weksberg R. 2013. Molecular findings in Beckwith–Wiedemann syndrome. Am J Med Genet C Semin Med Genet 163:131–140. De Carvalho DD, You JS, Jones PA. 2010. DNA methylation and cellular reprogramming. Trends Cell Biol 20:609–617. EPA US. 2000. Reregistration Eligibility Decision (RED) Vinclozolin. EPA 738-R-00–023. Available at: http://www.epa.gov/opp00001/ reregistration/REDs/2740red.pdf. Accessed on July 8, 2013. Goodman JI, Augustine KA, Cunnningham ML, Dixon D, Dragan YP, Falls JG, Rasoulpour RJ, Sills RC, Storer RD, Wolf DC, et al. 2010. What do we need to know prior to thinking about incorporating an epigenetic evaluation into safety assessments? Toxicol Sci 116:375–381. Gray L, Furr J. 2008. Vinclozolin treatment induces reproductive malformations and infertility in male rats when administered during sexual but not gonadal differentiation; However, the effects are not transmitted to the subsequent generations. Biol Reprod 78: 227–228. Gray LE Jr, Ostby JS, Kelce WR. 1994. Developmental effects of an environmental antiandrogen: The fungicide vinclozolin alters sex differentiation of the male rat. Toxicol Appl Pharmacol 129:46–52. Gray LE Jr, Ostby J, Monosson E, Kelce WR. 1999. Environmental antiandrogens: Low doses of the fungicide vinclozolin alter sexual differentiation of the male rat. Toxicol Ind Health 15:48–64. Hackett JA, Surani MA. 2013. DNA methylation dynamics during the mammalian life cycle. Philos Trans R Soc Lond B Biol Sci 368: 1–8. Harmston N, Lenhard B. 2013. Chromatin and epigenetic features of long-range gene regulation. Nucleic Acids Res 41:7185–7199. Inawaka K, Kawabe M, Takahashi S, Doi Y, Tomigahara Y, Tarui H, Abe J, Kawamura S, Shirai T. 2009. Maternal exposure to antiandrogenic compounds, vinclozolin, flutamide and procymidone, has no effects on spermatogenesis and DNA methylation in male rats of subsequent generations. Toxicol Appl Pharmacol 237: 178–187. LeBaron MJ, Rasoulpour RJ, Klapacz J, Ellis-Hutchings RG, Hollnagel HM, Gollapudi BB. 2010. Epigenetics and chemical safety assessment. Mutat Res 705:83–95.
Lu QJ. 2013. The critical importance of epigenetics in autoimmunity. J Autoimmun 41:1–5. Manikkam M, Tracey R, Guerrero-Bosagna C, Skinner MK. 2012. Dioxin (TCDD) induces epigenetic transgenerational inheritance of adult onset disease and sperm epimutations. PLoS One 7:e46249. Manikkam M, Tracey R, Guerrero-Bosagna C, Skinner M. 2013. Plastics derived endocrine disruptors (BPA, DEHP and DBP) induce epigenetic transgenerational inheritance of adult-onset disease and sperm epimutations. PLoS One 8:e55387. Matsuura I, Saitoh T, Ashina M, Wako Y, Iwata H, Toyota N, Ishizuka Y, Namiki M, Hoshino N, Tsuchitani M. 2005. Evaluation of a two-generation reproduction toxicity study adding endpoints to detect endocrine disrupting activity using vinclozolin. J Toxicol Sci 30 Spec No.:163–188. Mohammad HP, Baylin SB. 2010. Linking cell signaling and the epigenetic machinery. Nat Biotechnol 28:1033–1038. OECD. 2013. Chemicals testing guidelines. Available at: http:// www.oecd.org/env/ehs/testing/ [accessed 18 July 2013]. Rasoulpour RJ, LeBaron MJ, Ellis-Hutchings RG, Klapacz J, Gollapudi BB. 2011. Epigenetic screening in product safety assessment: Are we there yet? Toxicol Mech Methods 21:298–311. Schneider S, Kaufmann W, Buesen R, van Ravenzwaay B. 2008. Vinclozolin—The lack of a transgenerational effect after oral maternal exposure during organogenesis. Reprod Toxicol 25:352–360. Schneider S, Marxfeld H, Groters S, Buesen R, van Ravenzwaay B. 2013. Vinclozolin-No transgenerational inheritance of antiandrogenic effects after maternal exposure during organogenesis via the intraperitoneal route. Reprod Toxicol 37:6–14. Skinner MK. 2008. What is an epigenetic transgenerational phenotype? F3 or F2. Reprod Toxicol 25:2–6. Tracey R, Manikkam M, Guerrero-Bosagna C, Skinner M. 2013. Hydrocarbon (Jet Fuel JP-8) induces epigenetic transgenerational inheritance of adult-onset disease and sperm epimutations. Reprod Toxicol 36:104–116. Waddington CH. 1957. The Strategy of the Genes. London: George Allen & Unwin. pp 26–41. WHO-JMPR. 1995. Vinclozolin. pp 375–404. Wolf CJ, LeBlanc GA, Ostby JS, Gray LE Jr. 2000. Characterization of the period of sensitivity of fetal male sexual development to vinclozolin. Toxicol Sci 55:152–161.
Commentary Are We Ready to Consider Transgenerational Epigenetic Effects in Human Health Risk Assessment? Rebecca A. Alyea,1 B. Bhaskar Gollapudi2 and Reza J. Rasoulpour1* 1
Toxicology and Environmental Research and Consulting, The Dow Chemical Company, Midland, Michigan 2 Center for Toxicology and Mechanistic Biology, Exponent, Inc., Midland, Michigan
Recently, there has been a growing concern that chemically or nutritionally mediated epigenetic changes might lead to adverse health outcomes. The natural question is whether the existing chemical safety assessment paradigm is or is not protective of epigenetic-mediated effects, and if there is a need to incorporate new endpoints to specifically address epigenetics. Of particular interest are transgenerational epigenetic effects, which can be passed on through multiple generations. To investigate these questions, a comparison was performed between OECD guideline rat toxicology studies versus several rat transgenerational epigenetic studies. This analysis focused on vinclozolin owing to the availability of a comprehensive suite of dose-response data (NOAEL, reference dose, and human exposure estimates) for both conventional and epigenetic endpoints. This analysis revealed that vinclozolin transgenerational effects were demonstrated at a dose level
(100 mg/kg/day) that was: (1) 40-fold higher than the overall lowest-observed-adverse-effect level (LOAEL) from rat guideline studies, (2) 80-fold higher than the lowest NOAEL from rat guideline studies, (3) 80,000-fold higher than the reference dose for the molecule, and (4) 1.2-million fold above human exposure estimates. Through this analysis, we conclude that additional research across a spectrum of doses is necessary to elucidate the interplay between epigenetics and apical endpoints before considering epigenetics in human health risk assessment. Therefore, we recommend focusing future research toward (1) examining for potential causal relationships between epigenetic alterations and adverse apical endpoints, and (2) understanding the dose-response relationship of these causal epigenetic alterations when compared with those of the apical endpoints. Environ. Mol. Mutagen. 55:292–298, C 2013 Wiley Periodicals, Inc. 2014. V
Key words: epigenetics; product safety assessment; transgenerational effects
INTRODUCTION Epigenetics can be defined as heritable marks that are superimposed on the genome in the absence of direct changes in the DNA sequence. The mechanisms controlling these modifications are numerous and include, but are not limited to: DNA marks such as methyl-, hydroxymethy-, formyl-, and carboxyl-cytosine, histone modifications such as methylation, acetylation, phosphorylation, sumoylation, and ubiquitination, and variable miRNA expression, all of which act in concert to regulate gene expression [Cantone and Fisher, 2013; Harmston and Lenhard, 2013]. These epigenetic modifications are controlled by a higher level of regulatory machinery, which includes DNA (de-) methyltransferases, (de-) acetylases, ubiquitin lyases, chromatin remodeling machinery, and methyl-CpG-binding proteins [De Carvalho et al., 2010; Mohammad and Baylin, 2010]. These epigenetic modifications are key regulatory C 2013 Wiley Periodicals, Inc. V
steps in cellular differentiation and commitment to embryonic lineages as part of the normal biological processes. Within embryonic development this seemingly infinite combination of coordinated epigenetic events was described by Waddington in 1957 as a rolling landscape that represents the backdrop of molecular events guiding differentiation in the causal progression of zygote to adult
*Correspondence to: Reza J. Rasoulpour, PhD, The Dow Chemical Company, Toxicology & Environmental Research and Consulting, 1803 Building, Midland, MI 48674, USA. E-mail: [email protected] Received 31 July 2013; provisionally accepted 25 October 2013; and and in final form 28 November 2013 DOI 10.1002/em.21831 Published online 21 November 2013 in Wiley Online Library (wileyonlinelibrary.com).
Environmental and Molecular Mutagenesis. DOI 10.1002/em Epigenetics in Human Health Risk Assessment
[Waddington, 1957]. Epigenetic mechanisms play a critical role in determining the timing and progression of developmental processes within this rolling landscape. Epigenetic modulation of gene expression during development is evidenced by events such as X-chromosome inactivation, genomic imprinting, maternal and paternal allelic expression, tissue-specific gene expression, and cellular differentiation. During mammalian development, there are two periods of epigenetic reprogramming, one occurring during early embryonic development and another during primordial germ cell development. After fertilization both the paternal and maternal genomes undergo a period of DNA demethylation, followed by a remethylation phase coincident with the development of the somatic lineages [Hackett and Surani, 2013]. The second phase of reprogramming occurs in primordial germ cells during late embryogenesis. Presumably, the first phase of epigenetic reprogramming erases epigenetic modifications created during gametogenesis; however, there is evidence linking alterations in epigenetic mechanisms to various human disease states including immunodeficiency, Rett syndrome, Angelman, and Beckwith–Weideman Syndromes [Amir et al., 1999; Choufani et al., 2013; Lu, 2013]. Although the potential for perturbations within these processes leading to adverse health outcomes is a major area of research, many questions still remain for characterizing the role, temporal and spatial regulation, and modifications of epigenetic mechanisms during normal development. One area of particular interest within the field of epigenetics as it pertains to toxicology is the concept of a transgenerational epigenetic phenomenon. In these situations, it is hypothesized that an initial exposure during in utero development results in adverse apical outcomes not only for the parent, F1, and F2 generations, but also extends out to the F3 and later generations. This transgenerational phenomenon assumes that the epigenetic heritability can be sustained through generations in the absence of persistent exposure. For the purpose of this commentary, we will define the causative factor of the perturbation as a result of a chemically and/or nutritionally induced alteration. To observe a heritable transgenerational epigenetic effect following in utero exposure, the study should characterize stable epigenetic modifications that are causative of adverse health outcomes through three generations [Skinner, 2008]. Exposure of the dam leads to direct exposure of the developing fetus or the first generation as well as the germ cells within the fetus, which will potentially become the second generation. The third generation represents the first nonexposed generation, thus representing the true transgenerational population possible of displaying heritable modifications. Although our understanding of the mechanisms is continually evolving as we discover new states and interactions through basic research, questions surrounding the
293
inclusion of epigenetic modifications in human health risk assessment continue [Goodman et al., 2010; LeBaron et al., 2010; Rasoulpour et al., 2011; Alyea et al., 2012]. This commentary addresses if we are ready to consider epigenetic endpoints in human health risk assessment, with a particular focus on transgenerational effects. Before integrating a new testing strategy or evaluating additional endpoints within a product safety assessment paradigm, we must first evaluate what information is required and how such information is used in the risk assessment process.
PRODUCT SAFETY ASSESSMENT Fundamentally, risk is a function of the dose-response relationship of a hazard combined with potential human exposure. This concept is the tenant for safety assessment for industrial and agrochemicals, wherein the intrinsic hazards of a particular molecule are elucidated through toxicity studies, exposure is estimated based upon a combination of real-world data, computer modeling, and conservative assumptions, and these two sets of data are compared to determine the margin of exposure (i.e., distance between exposure and the lowest relevant noobserved-adverse-effect level). Multiple endpoints are evaluated and studies are performed to fully characterize the potential risk of a particular chemical, which include QSAR approaches, in vitro testing, and animal testing, which all help define the hazard fingerprint and the doseresponse relationship for each of the potential hazards identified. These data are used both in risk assessments as well as in establishing the appropriate hazard warning labels per programs such as the Globally Harmonized System (GHS) of Classification, Labeling, and Packaging (CLP). Additional factors are examined including potential environmental impact, toxicokinetic properties, and exposure assessments. The hazard identification component of the assessment relied on toxicity testing to characterize the potential hazards of a compound by evaluating adverse apical or phenotypic outcomes such as changes in body weight, body weight gain, feed consumption, organ weights, histopathological observations, and reproductive or developmental toxicity endpoints. This testing strategy integrates acute toxicity, immunotoxicity, systemic toxicity, specific target organ toxicity, reproductive toxicity, developmental toxicity, and carcinogenicity data to determine the lowest relevant no-observed-adverse-effect levels (NOAELs), which are then used in performing the risk assessment [OECD. 2013]. Toxicology hazard identification starts with evaluation of the physical chemical properties of a molecule, presence of reactive functional groups, and structure–activity relationships either through QSAR or expert judgment. These initial investigations can be facilitated through in silico technologies such as database mining or computational
Environmental and Molecular Mutagenesis. DOI 10.1002/em 294
Alyea et al.
modeling programs (e.g., DEREK or OASIS). Following these in silico approaches, acute toxicity parameters of oral, inhalation, and dermal lethality, eye and skin irritancy, and sensitization potential have been increasingly elucidated through in vitro tools. These core data can be quite useful in informing the material safety datasheet (MSDS) as well as understanding potential hazard for occupational exposure. Higher tier evaluations such as in vivo short-term, subchronic, chronic, and oncogenicity studies can also be performed in rodent species for agrochemicals and higher production volume industrial chemicals. Within these studies, clinical pathology (e.g., clinical chemistry, hematology, and urinalysis) combines with gross pathological and histopathological examination of all tissues and organs to understand the potential adaptive and adverse effects as well as carcinogenicity potential (see OECD test guidelines 407, 408, 451, 452, and 453 for more information). Often endpoints that drive the LOAEL are microscopic alterations within a particular tissue that can be identified by histopathological examination from long-term chronic or oncogenicity studies (i.e., 1- to 2-year studies). Developmental and reproductive toxicity studies are performed in rodents (and rabbits for developmental toxicity) to understand the potential of molecules to specifically cause adverse effects in neonatal growth, offspring survival, malformations (teratogenicity), functional deficits, puberty onset, and parental reproductive performance. These studies are typically performed under the OECD test guidelines 421, 422, 416, or 443. In addition to developmental and reproductive parameters, histopathology is performed on the major reproductive tissues, which is often the most sensitive endpoint for identification of endocrine active molecules. For a more thorough discussion and review of each of these testing strategies and guidelines that outline the evaluated endpoints refer to OECD guidance [OECD. 2013]. To perform the risk or product safety assessment, all of the relevant and available toxicology data are evaluated within a weight-of-evidence approach. During this evaluation, the lowest NOAEL across species and study types with findings of toxicological concern that have relevance to humans are used to derive such values as reference dose (RfD), allowable daily intake (ADI), and permitted daily exposure (PDE) for a molecule. In this process, the NOAEL is typically divided uncertainty factors to account for toxicokinetic and toxicodynamic difference between the species (interspecies uncertainty factor; rodent to human) and within the species (intraspecies uncertainty factor; human to human). Other potential uncertainty factors could be short-term to chronic extrapolations, database insufficiencies, LOAEL to NOAEL derivations, as well as the Food Quality Protection Act (FQPA) uncertainty factor. From a standpoint of hazard identification, the utility of characterizing adverse endpoints primarily relies on
anchoring adverse effects to the LOAEL and then determining the NOAEL for these apical endpoints. Characterization of these apical endpoints relies on expert judgment, biological plausibility, regulatory guidance documents, and precedent from the scientific literature or other laboratories. The core of a product safety assessment depends on characterizing the adverse apical endpoint and not necessarily identifying the mechanisms underlying these events. The reliance on the apical endpoint is critical to product safety assessment as it represents an effect at a dose that could not be regulated homeostatically or through adaptive mechanisms. If we conceptually imagine the genome as forming the base of a triangle (Fig. 1), exposure to exogenous agents at higher doses is likely to result in a large number of changes at the level of the transcriptome, which work together or in opposition to regulate fewer changes at the level of the proteome, which work again together or in opposition to regulate the physiome and finally culminating in an apical effect. At lower dose levels, there may be changes at the level of the transcriptome, which do not result in adverse apical effects. These latter changes would be classified as adaptive changes that are a normal part of biology and would be expected whenever we eat certain foods, exercise, or even breathe. It is the culmination of these changes at a particular dose level to an adverse apical effect that is generally used to characterize the dose-response relationship for the purposes of public health protection. However, what is currently unknown is the potential role or roles that the epigenome can play within the triangle of biological responses (Fig. 1). EPIGENETIC CASE STUDY As previously stated, epigenetic programming plays a critical role in normal developmental processes by altering gene expression in a temporally and spatially regulated fashion. The underlying factors that control these processes are by definition mechanisms and as such are
Fig. 1. Culmination of multiple molecular events at the level of the genome, transcriptome, and physiome to apical endpoints.
Environmental and Molecular Mutagenesis. DOI 10.1002/em Epigenetics in Human Health Risk Assessment
not incorporated under the current product safety assessment process. As stated earlier, it is the adverse apical outcome resulting from these alterations that forms the foundation of the safety assessment. Recent reports of adverse transgenerational effects following in utero exposure to environmental compounds such as vinclozolin, BPA, dioxin (TCDD), and jet fuel [Anway et al., 2005, 2006a; Manikkam et al., 2012, 2013; Tracey et al., 2013] suggest that an evaluation of epigenetic mechanisms as an indicator of adverse transgenerational effects within the product safety assessment may be warranted. In addition to transgenerational studies, as a fungicide, vinclozolin has a complete toxicology data package that includes developmental (OECD 414) and reproductive (OECD 416) toxicity studies as well as cancer bioassays (OECD 451 and 453). Finally, vinclozolin also has a published risk assessment that contains reference dose information as well as human exposure estimates from the US EPA [EPA, 2000]. Taken together, vinclozolin represents an ideal molecule to perform a case study comparing conventional toxicology data to epigenetic data and placing these findings into context with human exposure. For these benefits, this case study focuses only on vinclozolin as an example to provide recommendations more broadly for the field. Vinclozolin Toxicity Studies Vinclozolin is a fungicide that, along with its major metabolites, acts as an antiandrogen leading to demasculinized male rat offspring as demonstrated by reduced anogenital distance, retained nipples, a vaginal pouch, cleft phallus with hypospadias, suprainguinal ectopic scrota/ testes, and altered sex accessory glands in male offspring following in utero exposure [Gray et al., 1994, 1999; Matsuura et al., 2005]. In assessing the data presented within guideline regulatory studies to determine the lowest relevant NOAEL and LOAEL, it is clear that both the dose and developmental stage of exposure determine the extent of adverse effects [Gray et al., 1994, 1999; Wolf et al., 2000; Gray and Furr, 2008]. A multigenerational study in Wistar rats fed 0, 50, 300, 1,000, or 3,000 ppm vinclozolin reported adverse effects on fertility at 1,000 and 3,000 ppm [WHO-JMPR, 1995]. A marginal effect level of 50 ppm (4.5 mg/kg/day) was established based on the possibility of subfertility and signs of delayed development at 300 ppm. Additional multigenerational studies report a NOAEL of 40 ppm (4 mg/kg/day) following an assessment and establishment of no treatmentrelated effects in clinical signs, weight, feed consumption, reproduction, organ weight changes, or gross pathological findings [WHO-JMPR, 1995]. Exposure to 50 and 100 mg/kg/day via oral gavage from gestation day (GD) 14 to postnatal day (PND) 3 causes reduced fertility and 3.125 mg/kg/day causes a slight effect on anogenital distance,
295
establishing a LOAEL in this study of 3.125 mg/kg/day [Gray et al., 1999]. Based on a combined chronic toxicity and carcinogenicity study in rat, the lowest LOAEL and NOAEL is 2.3 and 1.2 mg/kg/day, respectively, for the development of Leydig cell tumors, which is most likely related to the antiandrogenic activity of vinclozolin [EPA, 2000]. The data presented in these toxicity evaluations are typical of the type of regulatory studies used to derive reference doses for use in product safety assessment and include establishment of a dose-response curve of adverse effects. Combining these studies together to examine the overall vinclozolin dose response reveals that: Dose levels 50 mg/kg/day causes reduced fertility in multigenerational reproductive toxicity studies as well as studies with GD 14–PND 3 exposure Dose levels between 3.125 and 50 mg/kg/day induce delayed preputial separation, slight effects on anogenital distance, and potential subfertility Dose level of 2.3 mg/kg/day in a cancer bioassay resulted in development of Leydig cell tumors likely owing to the antiandrogenic properties of the molecule Dose level of 1.2 mg/kg/day from the rat cancer bioassay was the lowest overall NOAEL for toxicity studies Therefore, the dose level of 1.2 mg/kg/day vinclozolin was used to derive the chronic reference dose, chronic population adjusted dose, and acceptable daily intake for use in risk assessment for protection of human health.
Vinclozolin Transgenerational Studies Interestingly, transgenerational research studies reported that following daily intraperitoneal (IP) injection of 100 mg/kg/day during GD 8–14 in SD rats causes adverse transgenerational effects presumably via an epigenetic alteration in the germ line [Anway et al., 2005]. The male offspring in the F1, F2, F3, and F4 generations displayed reduced sperm numbers and forward motility as well as an increase in the number of apoptotic germ cells per seminiferous tubule cross section. Follow-up studies further characterized a transgenerational reduction in spermatogenic capacity in both SD and Fisher F344 rats given 100 mg/kg/day [Anway et al., 2006b]. The epigenetic correlation was investigated in Fisher F344 rats given 100 mg/kg/day vinclozolin using methylation-specific restriction endonuclease, which identified 25 different PCR products that were differentially methylated [Anway et al., 2005]. Other investigators have reported a lack of transgenerational male reproductive effects following both IP injection and oral exposure to vinclozolin during the period of gonadal differentiation in male rats [(Gray and Furr, 2008; Schneider et al., 2008, 2013; Inawaka et al., 2009]. A recent study using IP injection of dams from GD 6–15
Environmental and Molecular Mutagenesis. DOI 10.1002/em 296
Alyea et al.
to 4 and 100 mg/kg/day found no effect on spermatogenesis or development of the male reproductive tract in offspring over four generations [Schneider et al., 2013]. A similar report of a lack of adverse transgenerational effects was also demonstrated over three generations via oral exposure of 4 and 100 mg/kg/day of vinclozolin, which is a more relevant route of exposure for human health assessment [Schneider et al., 2008]. The reason for the discordance between these studies may be due to the use of different strains, route of exposure, exposure window, undocumented protocol differences, diet, breeding, and/or colony variability. Comparison of Transgenerational versus Toxicity Studies Questions still remain in regard to defining an adverse transgenerational phenotype resulting from vinclozolin exposure. Furthermore, a causative relationship or correlated association to explain the mechanism (e.g., methylation changes with associated gene expression) by which these transgenerational effects could occur is not yet understood. In the absence of a causative association or correlation between the variations in methylation patterns and an adverse phenotype that results from this alteration, it is difficult to understand where this information would fit in the product safety assessment. In addition to questions surrounding variations in methylation patterns, if we review the current evidence of adverse apical effects and establishment of lowest relevant NOAEL and LOAELs, the overall NOAEL and LOAEL from the rat cancer bioassay is 1.2 and 2.3 mg/kg/day by the oral route of administration. Therefore, the transgenerational studies administering an IP injection of 100 mg/kg/day represent a dose level >40-fold higher than the LOAEL or >80fold higher than the NOAEL from apical endpoint studies. Therefore, it is unknown if potential transgenerational effects would occur at dose levels in the range of apical endpoint findings. In fact, the two studies that used dose levels within the range of the apical LOAELs (4 mg/kg/ day) by both oral and IP injection routes did not demonstrate a transgenerational effect [Schneider et al., 2008, 2013]. Comparing the animal data to human exposures within the risk assessment helps to further place these findings into context. The lowest NOAEL from the guideline animal studies was 1.2 mg/kg/day from the rat 2-year cancer bioassay. This overall NOAEL was used to derive a chronic reference dose (cRfD) by applying three sets of uncertainly factors (i.e., 103-intraspecies, 103-interspecies, and 103-FQPA) to derive 0.0012 mg/kg/day (Fig. 2) [EPA, 2000]. Moreover, human exposure for the US population, women, children, and infants is in the order of 34–78 ng/kg/day, which means that the reported transgenerational effects at 100 mg/kg/day are 1.2 millionfold above human exposures. Certainly, the data are
Fig. 2. Vinclozolin margin of exposure. Presented on a logarithmic axis of mg/kg/day dose are the rat IP injection transgenerational effects at 100 mg/kg/day as well as selected apical endpoint LOAELs and overall lowest NOAEL from the series of guideline toxicity studies. Rat study data present: 14.1 mg/kg/day is lowest LOAEL causing adverse reproductive effects, 22.3 mg/kg/day lowest LOAEL from chronic carcinogenicity assay, and 31.4 mg/kg/day is the lowest NOAEL [EPA, 2000]. Also presented are the chronic reference dose or population adjusted dose (cPAD) that applies 1,000-fold uncertainty factors to the 1.4 mg/kg/day lowest NOAEL, as well as human dietary exposure estimates in the total US population (40 ng/kg/day), women (34 ng/kg/day), infants (68 ng/kg/day), and children (78 ng/kg/day), which are data collected from the US EPA Reregistration Eligibility Decision (RED) [EPA, 2000].
scientifically interesting, but based upon a lack of doseresponse relationship data, the information within these studies does not provide evidence that would in fact alter the existing risk assessment for vinclozolin. In other words, there are no data to suggest that the risk assessment performed with apical endpoints on vinclozolin would not also be protective for any potential transgenerational epigenetic effects. RECOMMENDATIONS The field of epigenetics has grown exponentially in the past several years, as evidenced by a large increase in the scientific publications as well as coverage of this area within the popular press. The concept of transgenerational epigenetic effects in and of themselves even raises simplistic questions with regard to Darwinian versus Lamarkian evolution. Whether the genome or epigenome “wins” the battle of biological dominance remains to be seen, but this area of research has raised a large number of questions in multiple specialties within biological sciences. As it pertains to toxicology, there appear to be several data gaps with respect to causality as well as dose-response relationships that would help reveal the implications for epigenetic effects within product safety assessment. Our recommendations are for more research toward addressing two key areas, namely: (1) establishing causal relationships between epigenetic alterations and adverse apical endpoints, and (2) understanding the dose-response relationship of these causal epigenetic alterations with the apical endpoints. A key challenge to establish causal relationships between epigenetic changes and apical endpoints lies in the sheer number of potential epigenetic alterations that can occur. Moreover, a thorough understanding of the
Environmental and Molecular Mutagenesis. DOI 10.1002/em Epigenetics in Human Health Risk Assessment
normal biological variability within different tissues, organisms, and species, for methylation, histone modifications, and miRNA is still needed. By understanding the boundaries in the “noise” of biological variability for different epigenetic modifications at different areas of the genome would help provide the appropriate canvas to shape our understanding of adverse versus adaptive changes. Once normal variability is understood, identification of which epigenetic changes lead to which particular disease state can help establish the causal relationships needed to develop a concrete and scientific approach to understanding this area of research. Once epigenetic changes that are predictive of apical effects have been identified, examination of the doseresponse relationships of these epigenetic modifications in comparison to apical endpoints may provide toxicological insight into causality of these relationships. Therefore,
297
epigenetic studies with full dose-response relationships that anchor these changes to apical effect(s) would be extremely valuable in advancing our understanding of the contribution of epigenetics in toxicology. If the dose response for causal epigenetic changes lies on top (Fig. 3A) or to the right (Fig. 3B) of that for a given apical endpoint, then utilizing those apical endpoints for the purposes of product safety assessment would be appropriate. In other words, the product safety assessment based upon apical endpoints would also be protective of epigenetic changes similar to any molecular changes within our triangle from Figure 1. However, if persistent, causal epigenetic changes occur at dose levels below those of apical endpoints (Fig. 3C), then perhaps an alternative approach should be used to understand the implications to human health protection for those particular findings. Overall, future studies that examine this dose-response relationship using relevant routes of exposure at doses that are within both environmental exposure levels as well as doses used in hazard identification studies will aid in discerning what is meaningful for product safety assessment and human health protection. Although current evidence of epigenetic modulations causing adverse phenotypic outcomes is lacking in regard to establishing causality, this only implies that more research is needed to address these points. In going forward, we should learn from the existing studies and consider revising study designs to evaluate the potential correlative versus causal associations between adverse phenotypic effects and molecular endpoints by evaluating the relevance of the dose level and route of exposure, as well as considering the dose-response relationships for identification of NOAELs for causative molecular endpoints. ACKNOWLEDGMENTS The authors thank Drs. Matthew J. LeBaron, Michael R. Woolhiser, and Edward W. Carney for their review of the manuscript. REFERENCES
Fig. 3. Hypothetical dose-response curves for any given apical versus epigenetic effects. These dose-response curves can be similar to one another (A) as the early molecular changes are leading to apical endpoints, or there could be apical endpoints more sensitive than particular epigenetic changes (B). In contrast, causal epigenetic marks could be leftshifted of apical endpoints (C).
Alyea RA, Moore NP, LeBaron MJ, Gollapudi BB, Rasoulpour RJ. 2012. Is the current product safety assessment paradigm protective for epigenetic mechanisms? J Pharmacol Toxicol Methods 66:207–214. Amir RE, Van den Veyver IB, Wan M, Tran CQ, Francke U, Zoghbi HY. 1999. Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2. Nat Genet 23: 185–188. Anway MD, Cupp AS, Uzumcu M, Skinner MK. 2005. Epigenetic transgenerational actions of endocrine disruptors and male fertility. Science 308:1466–1469. Anway MD, Memon MA, Uzumcu M, Skinner MK. 2006a. Transgenerational effect of the endocrine disruptor vinclozolin on male spermatogenesis. J Androl 27:868–879. Anway MD, Leathers C, Skinner MK. 2006b. Endocrine disruptor vinclozolin induced epigenetic transgenerational adult-onset disease. Endocrinology 147:5515–5523.
Environmental and Molecular Mutagenesis. DOI 10.1002/em 298
Alyea et al.
Cantone I, Fisher AG. 2013. Epigenetic programming and reprogramming during development. Nat Struct Mol Biol 20:282–289. Choufani S, Shuman C, Weksberg R. 2013. Molecular findings in Beckwith–Wiedemann syndrome. Am J Med Genet C Semin Med Genet 163:131–140. De Carvalho DD, You JS, Jones PA. 2010. DNA methylation and cellular reprogramming. Trends Cell Biol 20:609–617. EPA US. 2000. Reregistration Eligibility Decision (RED) Vinclozolin. EPA 738-R-00–023. Available at: http://www.epa.gov/opp00001/ reregistration/REDs/2740red.pdf. Accessed on July 8, 2013. Goodman JI, Augustine KA, Cunnningham ML, Dixon D, Dragan YP, Falls JG, Rasoulpour RJ, Sills RC, Storer RD, Wolf DC, et al. 2010. What do we need to know prior to thinking about incorporating an epigenetic evaluation into safety assessments? Toxicol Sci 116:375–381. Gray L, Furr J. 2008. Vinclozolin treatment induces reproductive malformations and infertility in male rats when administered during sexual but not gonadal differentiation; However, the effects are not transmitted to the subsequent generations. Biol Reprod 78: 227–228. Gray LE Jr, Ostby JS, Kelce WR. 1994. Developmental effects of an environmental antiandrogen: The fungicide vinclozolin alters sex differentiation of the male rat. Toxicol Appl Pharmacol 129:46–52. Gray LE Jr, Ostby J, Monosson E, Kelce WR. 1999. Environmental antiandrogens: Low doses of the fungicide vinclozolin alter sexual differentiation of the male rat. Toxicol Ind Health 15:48–64. Hackett JA, Surani MA. 2013. DNA methylation dynamics during the mammalian life cycle. Philos Trans R Soc Lond B Biol Sci 368: 1–8. Harmston N, Lenhard B. 2013. Chromatin and epigenetic features of long-range gene regulation. Nucleic Acids Res 41:7185–7199. Inawaka K, Kawabe M, Takahashi S, Doi Y, Tomigahara Y, Tarui H, Abe J, Kawamura S, Shirai T. 2009. Maternal exposure to antiandrogenic compounds, vinclozolin, flutamide and procymidone, has no effects on spermatogenesis and DNA methylation in male rats of subsequent generations. Toxicol Appl Pharmacol 237: 178–187. LeBaron MJ, Rasoulpour RJ, Klapacz J, Ellis-Hutchings RG, Hollnagel HM, Gollapudi BB. 2010. Epigenetics and chemical safety assessment. Mutat Res 705:83–95.
Lu QJ. 2013. The critical importance of epigenetics in autoimmunity. J Autoimmun 41:1–5. Manikkam M, Tracey R, Guerrero-Bosagna C, Skinner MK. 2012. Dioxin (TCDD) induces epigenetic transgenerational inheritance of adult onset disease and sperm epimutations. PLoS One 7:e46249. Manikkam M, Tracey R, Guerrero-Bosagna C, Skinner M. 2013. Plastics derived endocrine disruptors (BPA, DEHP and DBP) induce epigenetic transgenerational inheritance of adult-onset disease and sperm epimutations. PLoS One 8:e55387. Matsuura I, Saitoh T, Ashina M, Wako Y, Iwata H, Toyota N, Ishizuka Y, Namiki M, Hoshino N, Tsuchitani M. 2005. Evaluation of a two-generation reproduction toxicity study adding endpoints to detect endocrine disrupting activity using vinclozolin. J Toxicol Sci 30 Spec No.:163–188. Mohammad HP, Baylin SB. 2010. Linking cell signaling and the epigenetic machinery. Nat Biotechnol 28:1033–1038. OECD. 2013. Chemicals testing guidelines. Available at: http:// www.oecd.org/env/ehs/testing/ [accessed 18 July 2013]. Rasoulpour RJ, LeBaron MJ, Ellis-Hutchings RG, Klapacz J, Gollapudi BB. 2011. Epigenetic screening in product safety assessment: Are we there yet? Toxicol Mech Methods 21:298–311. Schneider S, Kaufmann W, Buesen R, van Ravenzwaay B. 2008. Vinclozolin—The lack of a transgenerational effect after oral maternal exposure during organogenesis. Reprod Toxicol 25:352–360. Schneider S, Marxfeld H, Groters S, Buesen R, van Ravenzwaay B. 2013. Vinclozolin-No transgenerational inheritance of antiandrogenic effects after maternal exposure during organogenesis via the intraperitoneal route. Reprod Toxicol 37:6–14. Skinner MK. 2008. What is an epigenetic transgenerational phenotype? F3 or F2. Reprod Toxicol 25:2–6. Tracey R, Manikkam M, Guerrero-Bosagna C, Skinner M. 2013. Hydrocarbon (Jet Fuel JP-8) induces epigenetic transgenerational inheritance of adult-onset disease and sperm epimutations. Reprod Toxicol 36:104–116. Waddington CH. 1957. The Strategy of the Genes. London: George Allen & Unwin. pp 26–41. WHO-JMPR. 1995. Vinclozolin. pp 375–404. Wolf CJ, LeBlanc GA, Ostby JS, Gray LE Jr. 2000. Characterization of the period of sensitivity of fetal male sexual development to vinclozolin. Toxicol Sci 55:152–161.
