Thyroid hormone links environmental signals to DNA methylation

Abstract

Environmental conditions experienced within and across generations can impact individual phenotypes via so-called ‘epigenetic' processes. Here we suggest that endocrine signalling acts as a ‘sensor' linking environmental inputs to epigenetic modifications. We focus on thyroid hormone signalling and DNA methylation, but other mechanisms are likely to act in a similar manner. DNA methylation is one of the most important epigenetic mechanisms, which alters gene expression patterns by methylating cytosine bases via DNA methyltransferase enzymes. Thyroid hormone is mechanistically linked to DNA methylation, at least partly by regulating the activity of DNA methyltransferase 3a, which is the principal enzyme that mediates epigenetic responses to environmental change. Thyroid signalling is sensitive to natural and anthropogenic environmental impacts (e.g. light, temperature, endocrine-disrupting pollution), and here we propose that thyroid hormone acts as an environmental sensor to mediate epigenetic modifications. The nexus between thyroid hormone signalling and DNA methylation can integrate multiple environmental signals to modify phenotypes, and coordinate phenotypic plasticity at different time scales, such as within and across generations. These dynamics can have wide-ranging effects on health and fitness of animals, because they influence the time course of phenotypic adjustments and potentially the range of environmental stimuli that can elicit epigenetic responses.

This article is part of the theme issue ‘Endocrine responses to environmental variation: conceptual approaches and recent developments’.

1. Introduction

Most hormones are multi-functional and respond to a range of environmental signals to modify phenotypic traits [1–3]. In vertebrates, for example, thyroid hormones regulate development and life history, as well responses to environmental temperature variation [4,5]. Similarly, glucocorticoids mediate animal responses (e.g. metabolism and growth) to environmental signals [6], and hormones such as thyroid hormones and glucocorticoids may interact to mediate phenotypic responses [7]. In addition to natural variation, anthropogenic environmental signals modify endocrine-mediated functions [8,9]. The combined effects of these natural and anthropogenic environmental sources of variation (and disruption) are likely to determine the success of populations and species, so that there is a strong conservation dimension in unravelling environment–endocrine interactions [10].

Environmental variability can modify phenotypes during the lifetime of an organism and across generations via epigenetic modifications [11,12]. Importantly, endocrine signalling can be linked to epigenetic processes such as DNA methylation [13], which together can alter physiology, behaviour and ecology of animals [14–16]. Epigenetic processes modify gene expression programmes without altering respective DNA sequences but by the addition (or removal) of functional groups (e.g. methyl groups) to DNA or histone molecules [17,18]. One of the best-known epigenetic mechanisms involves the methylation (or de-methylation) of cytosine residues on DNA, and the addition or removal of a methyl group can repress or activate transcription, respectively [19,20]. Linking DNA methylation to environmental inputs provides a mechanism by which phenotypes could be remodelled to alter performance in variable environments [21,22]. Such phenotypic plasticity can be beneficial by compensating for negative environmental effects, or detrimental when it leads to phenotype–environment mismatches [23].

Repression of transcription by DNA methylation involves recruitment of methyl-CpG-binding proteins to the methylated site, which in turn recruit histone deacetylases and their associated repressive machinery [24,25]. DNA methylation is mediated by DNA methyltransferases (DNMTs), of which DNMT1 maintains methylation patterns across cell divisions, and DNMT3 alters methylation patterns in response to environmental cues [17,26,27]. The action and regulation of DNMT3 is therefore interesting in understanding induction of alternative phenotypes in response to environmental change (phenotypic plasticity). In reverse to the actions of DNMTs, methylated regions of DNA may be de-methylated by ten–eleven translocase (TET) enzymes, although many regions of DNA are protected from de-methylation, so that methylation patterns can cause long-term changes to animal phenotypes [28].

While endocrine signalling pathways and epigenetic regulators are often invoked as distinct processes, epigenetic effects often have endocrine origins. Epigenetic modifiers can be recruited and functionally regulated through direct or indirect interactions with hormone receptors [29–32]. Nuclear receptors (NRs), which are central to many endocrine pathways, in particular are principal regulators of transcription by orchestrating key components of the transcriptional initiation complex [33–35]. However, concomitant interactions of NRs with chromatin remodelling machinery, including histone (de)acetylases, histone methyltransferases and DNMTs, mean that many NRs also exert epigenetic control [31,36,37]. In addition to this direct mode of regulation, NRs can also regulate the expression of genes encoding chromatin modifiers by interacting with response elements in their promoters. Expression of DNMT1 in mice, for example, is regulated by cross-talk between the NR oestrogen-related receptor γ and its cofactor SHP [38]. Epigenetic remodelling therefore often represents an intrinsic component of endocrine signal transduction pathways, and not necessarily a parallel or independent processes. Within- and between-pathway feedback and cross-talk can also have an epigenetic dimension, because epigenetic modifiers can regulate the activities and sensitivities of upstream processes in their own and parallel pathways [39]. Given this complexity of evolved endocrine responses to ‘natural' variation, anthropogenic disruption of hormone signalling is likely to produce unexpected and unpredictable effects [1,40].

Most ecosystems in the world are experiencing unprecedented environmental change because of human activity. In addition to climate change, artificial light-at-night (ALAN) and endocrine-disrupting compounds (EDCs) are two of the most dominant anthropogenic impacts. All three are currently increasing at alarming rates [41,42], and undermine hormone-mediated signalling [43–45]. Importantly, these anthropogenic drivers can alter qualitatively how animals respond to natural variation, and existing knowledge of responses to natural environmental variability may now be inadequate [1,46,47]. Hence, in order to facilitate effective conservation there is considerable urgency to assess how animal responses are altered in the presence of endocrine disruption [48].

In this Opinion piece, we suggest that endocrine signalling acts as an environmental ‘sensor' to link epigenetic modifications of phenotypes to environmental variation (figure 1). We focus on the interactions between thyroid hormone and DNA methyltransferase 3 as two important signalling systems in animals. We note, however, that there are other potential endocrine (e.g. glucocorticoid hormones and receptors) and epigenetic (e.g. histone modifications) mechanisms that could act in a similar manner [18,32]. Thyroid hormone is closely associated with environmental sensing by its integration with the hypothalamus and hence temperature and light perception [2,49]. We briefly summarize the mode of action of thyroid hormone, and how it can interact with DNA methylation via DNMT3. We then discuss how these mechanistic relationships can alter epigenetic responses to environmental change.

Figure 1.

Conceptual outline. (a) We propose that endocrine mechanisms act as a sensor to link environmental signals to epigenetic processes and thereby phenotypic plasticity. The question mark indicates that this proposition is as yet unresolved experimentally. (b) In particular, we propose that thyroid hormone transmits light and temperature signals via the hypothalamus. Thyroid signalling is modified by differential expression of different deiodinase enzymes (dio), which control production of active hormone (T3). Signalling via T3 occurs via thyroid receptors (TR), which bind to thyroid response (TRE) elements in the promoters of genes. Genes for DNA methyltransferase 3 (DNMT3), responsible for adding methyl groups to DNA, and ten–eleven translocases (TET), responsible for removing methyl groups from DNA, contain TRE and are thereby regulated by thyroid hormone. Conversely, TR transcription is regulated by DNMT and TET, and together these mechanisms modify gene expression programmes in an environmentally sensitive manner. However, this pathway is vulnerable to endocrine disruption by anthropogenic inputs, such as endocrine-disrupting compounds (EDCs). (Online version in colour.)

2. Actions and interactions between thyroid hormone and DNMT3

(a) . Environmental signal transduction

Changed DNA methylation patterns under different environmental conditions imply that DNA methylation is linked to environmental sensing. However, to the best of our knowledge a causal, sensory link between DNMT3 activity and environmental change has not been established. Here, we suggest that environmental signals such as light and temperature act on the thyroid hormone signalling cascade, which then induces de novo methylation or de-methylation. The hypothalamus is crucial for processing temperature and light signals received via transient receptor potential ion channels (TRPs) and the retina (via the retino- and geniculo-hypothalamic tracts) and pineal gland, respectively [2,49]. The hypothalamic sensitivity to thyroid hormones or the hypothalamus–pituitary–thyroid (HPT) axis itself can thereby represent a mechanistic link connecting environmental signals to epigenetic changes in DNA methylation patterns (figure 1). The specificity of DNA methylation patterns in different organs could be explained by tissue-specific responses to thyroid signals [50].

(b) . Thyroid hormone action

Thyroid hormone production and release are controlled via the hypothalamic–pituitary–thyroid axis through the sequential actions of thyroid-releasing hormone and thyroid-stimulating hormone. There is some variation in this canonical pathway, and during metamorphosis in amphibians and fish, for example, the stress-induced corticotropin-releasing hormone controls thyroid-stimulating hormone, and metamorphosis is closely linked to environmental factors such as tadpole density [51]. During development in anemone fish (Amphiprion ocellaris), thyroid hormones orchestrate metabolism, morphological changes and behaviour to match juveniles to their ecological context [52]. Beyond metamorphosis, thyroid hormone more generally regulates development, at least in chordates [53]. Thyroid hormone itself can act both centrally through the central nervous system and peripherally on target tissues [54]. The genomic actions of thyroid hormone are mediated via thyroid receptors (TRs) which regulate gene expression by binding to thyroid response elements (TREs) on promoters of target genes [4]. Unliganded TRs may interact with co-repressors to repress transcription of genes that are regulated by triiodothyronine (T3). By contrast, T3-liganded TRs induce transcription by recruiting co-activators [55]. Repression of transcription by unliganded TRs involves indirect protein interactions with histone methylases and deacetylases, which alter chromatin structure and thereby modify access to DNA by transcriptional regulators; binding of T3 to TR reverses these processes [56,57].

Feedback controls help modulate plasma concentrations of thyroid hormones at a whole-animal level [58], and target tissues also regulate and remodel their sensitivities to circulating concentrations of thyroid hormone through local mechanisms that regulate the bioavailability and bioactivity of the hormone [50]. There are a number of cell-surface transporters (e.g. MCT8, OATPC1), for instance, that modulate intracellular concentrations of thyroid hormone [50]. The three different paralogues of deiodinase enzyme can also cleave different iodo-groups from thyroid hormone and its metabolites, thereby altering physiological activity [59]. Additionally, cells can fine-tune their sensitivities to thyroid hormone by regulating the expression and isoform profiles of TRs. While the main receptors for thyroid hormones are nuclear receptors (i.e. TRs), a cell-surface receptor (integrin αVβ3) also supports important roles in thyroid hormone signalling [60]. Lastly, thyroid receptors can bind to positive and negative TREs in the promoters of target genes. Positive TREs represent the traditional subtype (described above), whereas negative TREs reverse the transcriptional effects of TRs—they activate transcription in their unliganded state and repress transcription in the presence of thyroid hormone [34].

(c) . Thyroid hormone regulates DNA methyltransferase transcription

Thyroid hormone can regulate transcription of DNA methyltransferase genes via TRE at the dnmt locus. In frog (Xenopus laevis) tadpoles, T3 controls metamorphosis, during which global DNA methylation patterns change [61]. The DNMT3a gene of X. laevis contains a TRE, and DNMT3a (but not DNMT1) mRNA levels increased with the administration of T3 to tank water [62]. Similarly, the dnmt3a locus in mice has two functional TREs, and administration of T3 increased brain DNMT3a transcript abundance [63]. These data from frogs and mice indicate that TREs on dnmt3 are conserved at least among tetrapods [63]. The link between thyroid hormone and DNMT3 is also seen during mouse liver differentiation, where T3 spikes were associated with alterations in DNA methylation patterns [57]. Additionally, grooming of mouse pups by mothers increased DNMT3a mRNA levels in the hippocampus of pups. These changes in mRNA levels were accompanied by an association between circulating T3 levels and DNMT3a abundance, and increased TR binding by T3 led to increased DNA methylation at the oxytocin promoter [64]. Together, these data indicate that DNMT3 transcription and DNA methylation patterns can be regulated by T3 and TR binding to TRE in diverse groups of vertebrates and in different contexts. We note, however, that the number of species and contexts studied is sparse as yet.

Thyroid hormone can also induce DNMT3 transcription by mediating de-methylation of the dnmt3 locus. In the X. laevis tadpole brain, administration of exogenous T3 caused de-methylation of the dnmt3a gene, and recruitment of TET3 to the locus [61]. Additionally, the TET2 gene contains a TRE that activates transcription in the presence of bound TRs [61]. By contrast, induced hypothyroidism for 5 days after birth increased the total activity of DNMT enzymes in the hippocampus of rats, suggesting a T3-mediated pathway to repress DNMT [65]. However, total activity of DNMT may be determined primarily by the maintenance DNMT1 enzyme, so that the functional significance of total DNMT activity for induction of different phenotypes via DNMT3 is unresolved.

Interestingly, during ageing in humans there is increased methylation of the thyroid receptor β (TRβ) promoter regions, which is associated with lower TRβ mRNA levels [66]. Similarly, decreased mRNA levels of TRα in the hippocampus of adolescent rats following excess folate consumption may have been caused by increased methylation of the TRα promoter region [67]. The thyroid axis itself, therefore, can be regulated by DNA methylation, and there is a potential for a feedback loop where T3 regulates DNMT3, which in turn regulates TRs.

Together, these data show that epigenetic regulation can represent an intrinsic component of thyroid hormone signal transduction pathways, and that there is cross-talk between thyroid hormone signalling and epigenetic regulation of phenotypes. It is less clear to what extent environmental signals modulate this cross-talk. Thyroid hormone signalling is clearly linked to environmental inputs, and DNA methylation patterns change in response to environmental change. However, whether or not the environment acts independently on each process or whether the environmental signal is transduced by a linked response remains to be demonstrated experimentally. Uncovering these relationships is particularly important for wildlife populations that are exposed to increasingly unstable environments, and where the temporal resolution of responses can impact ecological relationships [68]. Currently, knowledge of interactions between hormones and epigenetic processes stems almost entirely from model organisms, and expanding this knowledge to wildlife is an important future direction.

3. Responses to environmental variation

(a) . ‘Natural' environmental variation

Environmental variation is a selection pressure that has shaped genotypes, and thereby phenotypes, over evolutionary time. Adaptation to different environments can underlie biogeography by determining optimal mean environments of different species [69]. However, environments are rarely stable, and species must respond to mean conditions as well as to variations around the mean. The relative timing of perturbations and physiological responses is therefore crucial in determining the efficacy of matching phenotypes to prevailing conditions [70]. The time course of responses depends very much on the underlying mechanisms. Endocrine responses can be rapid and elicit physiological responses within seconds to hours, such as in glucocorticoid-mediated responses to stress [71]. However, hormones can also have longer-lasting effects [71,72], and thyroid hormone for example regulates thermal acclimation over periods of several weeks [5]. Epigenetic responses such as DNA methylation concurrently alter phenotypes relatively permanently, for example by mediating developmental effects that can last throughout life and even across generations [73]. The temporal resolution of physiological responses to environmental change—with respect to both onset and duration of the response—would depend on the interaction and synchrony between these mechanisms. Endocrine and epigenetic processes can be intrinsically linked, as we summarized above, and mechanistically resolving the extent to which they co-regulate plastic responses by wildlife is critical to determine whether animal responses will be sufficient to buffer the increased environmental variation and novel interactions they face. These dynamics are particularly important in the context of anthropogenic impacts such as ALAN and EDCs, which can disrupt hormone signalling but may also have an epigenetic dimension because of endocrine–epigenetic interactions [15,74].

Regular fluctuations such as annual light and temperature cycles drove the evolution of endocrine and epigenetic regulatory systems to coordinate physiological processes (e.g. reproduction, development, acclimatization) to coincide with favourable environmental conditions [75–77]. In many seasonally breeding vertebrates, winter conditions inhibit neuroendocrine functions associated with reproduction, and seasonal changes in photoperiod or temperature can also elicit changes in DNA methylation patterns [78]. For example, in hamsters (Phodopus sungorus), responses to seasonally changing light cycles were mediated by thyroid hormone signalling through tanycytes of the hypothalamus. Winter photoperiods upregulated tanycyte deiodinase enzyme 3 (dio3) and downregulated deiodinase 2 (dio2), thereby promoting T4-conversion to its inactive form, reverse T3 (rT3). Consequently, thyroid hormone signalling and reproductive hormone (e.g. gonadotropins, Gn) production were reduced [79]. By contrast, spring and summer conditions elevated tanycyte dio2 (and depressed dio3) to convert T4 into active T3, thereby increasing thyroid hormone signalling and production of Gn through activation of kisspeptin neurons [79]. Interestingly, DNA methylation of the dio3 promoter is critical for photoperiodic time measurement, and methylation was reversible in the hypothalamus of hamsters: short photoperiod decreased DNMT3a and b mRNA levels, inhibited dio3 methylation, and increased its mRNA levels [76,79] leading to a reduction in T3 signalling. Importantly, these trends were reversed with increasing photophase in spring. Hence, the interplay between TH signalling and DNA methylation in the hypothalamus of hamsters optimized the timing of reproduction, which, however, also leaves it vulnerable to EDCs and ALAN [9].

Temperature on its own impacts both thyroid signalling and DNA methylation, although it is possible that these processes are linked ultimately. Thyroid hormone regulated thermal acclimation of cardiac and muscle function, and metabolism in zebrafish and lake whitefish (Coregonus clupeaformes) [77,80]. For example, zebrafish compensated perfectly for a 10°C temperature difference, with no significant difference in swimming performance between 18- and 28°C-acclimated fish when tested at their respective acclimation temperatures [81]. Pharmacologically induced hypothyroidism eliminated this compensatory response, which, however, could be re-instated with experimental supplementation of active T2 and T3 [81].

DNA methylation patterns also differ between different thermal regimes [82–84], and changes in global DNA methylation patterns can parallel the effects of thermal acclimation [85–87]. Temperature affected methylation of the myogenin promoter in Senegalese sole (Solea senegalensis), where increased DNMT3b mRNA levels at cooler temperatures were associated with increased promoter methylation, and decreased myogenin mRNA levels and muscle growth in embryos [88]. These correlations between DNA methylation patterns and thermal responses hint at a functional relationship that could be integrated with thyroid hormone signalling. DNMT3a also played a functional role in mediating developmental and transgenerational thermal plasticity in zebrafish. Unlike wild-type fish, offspring of dnmt3a knock-out zebrafish (raised and bred at 28°C) did not survive rearing at a relatively cool temperature (23°C), but survived equally as well as wild-type fish when reared at the parental temperature (28°C) [11]. Interestingly, acclimating parents to 23°C for two months prior to breeding restored survivorship of dnmt3a knock-out fish reared at 23°C to wild-type levels, indicating that different mechanisms can interact to match phenotypes to their prevailing conditions [11]. It is possible that our earlier studies showing a regulatory role of thyroid hormone [5] and of DNMT3a [11] in thermal plasticity in fact detailed different aspects of the same pathway that integrates endocrine signalling with epigenetic modifications. Understanding these mechanistic links is important because they determine the temporal resolution of animal responses to environmental variation—in this case coordinating transgenerational and within-generation responses.

(b) . Anthropogenic impacts: compounded perturbations

Environmental variability can drive adaptation or plasticity so that disturbance–recovery cycles have little long-term effect [69,89]. However, when several perturbations coincide, their compounded effect may exceed the capacity for compensatory responses. Compounded effects are particularly likely with human impacts that can occur relatively suddenly and concurrently with natural variation [89,90]. Three of the major anthropogenic impacts are climate change, pollution and ALAN. Climate change alters mean temperatures and temperature variability, both of which can impact thyroid hormone signalling and hence DNA methylation. For example, unusually warm winters cause a mismatch between seasonal light cycles and temperature cycles that impacted thyroid hormone signalling in mosquitofish (Gambusia holbrooki) [91] and caused physiological dysfunction in mice [92]. Similarly, heatwaves can have a broad range of epigenetic effects [93] and caused changes to methylation patterns in a polychaete (Spiophanes tcherniai) [85] that can potentially be passed across generations [94].

Thyroid hormone is also responsive to light and its signalling is important to mediate physiological responses to natural cycles in photoperiod and temperature (see above). Not surprisingly, therefore, ALAN disrupts thyroid hormone signalling and thereby its downstream targets [9,95]. Electric light from street lights, buildings and vehicles, for example, means that built-up areas are rarely completely dark [42,96]. In addition to direct electric light, skyglow caused by artificial light scattered by dust, water and gas molecules extends increased light levels to areas not directly illuminated [97]. ALAN has substantially increased globally [98], and the trend of increased ALAN is likely to increase along with increasing urbanization [99]. ALAN is an endocrine disruptor [2,45] that causes a broad range of physiological and behavioural consequences [100,101]. ALAN also caused changes to DNA methylation patterns, notably global hypomethylation in mouse oocytes [102], and in the pancreas in adult rats [14]. These effects are similar to overexposure to thyroid hormone, which decreased DNA methylation in male mouse germ cells [13]. Compared with the natural photoperiod, exposure to constant light also increased DNMT1 and 3 mRNA levels in fast muscle of juvenile Atlantic cod (Gadus moruha), which was accompanied by increases in body mass [103].

A vast range of human-manufactured chemicals are now present in the environment and many disrupt endocrine function [104–106]. The thyroid axis is particularly susceptible to chemical endocrine disruption by industrial chemicals that are commonly found in the environment globally, including bisphenols, halogenated organochlorides and polychlorinated biphenyls [8,107,108]. EDCs can disrupt nearly all steps in the thyroid signalling cascade from thyroid production and transport to disruption of deiodinase enzymes and thyroid receptors [8,109–111]. Bisphenols, in particular, can be TR antagonists by binding TRs to block T3 access, or can act as ligands to stimulate signalling [107,112]. The downstream consequences of thyroid hormone disruption by EDCs are not well characterized but may involve impaired locomotor function [113], and neurodevelopmental disruption within and across generations [114]. For example, the pesticide chlorpyrifos decreased T3 and T4 levels and disrupted sensory development in surgeonfish (Acanthurus triostegus). Exposed fish thereby experienced higher levels of predation, but these effects were reversed with supplementation of TH [44].

Synthetic chemicals can also have epigenetic effects [115], but it is not clear to what extent these are linked to endocrine signalling, if at all. Numerous chemicals (e.g. bisphenols) that interfere with thyroid hormone synthesis and transport, and TR signalling are also known to cause epigenetic effects, including changes in DNA methylation, histone marks, and noncoding RNAs [115]. There may be an endocrine dimension, and exposure of embryonic zebrafish to brominated flame retardant (TDCIPP), for example, caused abnormal TH levels in their offspring (increased T3 but decreased T4), which was associated with global hypermethylation, including of genes associated with thyroid hormone transport [114]. These data indicate a link between EDC, thyroid hormone levels and DNA methylation.

4. Conclusion

Here we propose that thyroid signalling and DNA methylation are part of the same pathway, with thyroid hormone acting as the environmental sensor that can regulate DNMT3 and TET enzyme transcription and recruitment. Although direct experimental evidence demonstrating an environment–endocrine–epigenetic axis is sparse, sufficient knowledge of these processes has now accumulated to provide circumstantial evidence that a sensory endocrine–epigenetic link is likely. Signal transduction between the environment and phenotype is crucial to optimize plastic responses to environmental change. It is important to understand the mechanisms that underpin these plastic responses because they determine the range of environmental signals that organisms can respond to, the timing of the response, and its persistence. We focused on the interaction between thyroid hormone and DNA methylation here, which could explain epigenetic responses to light and temperature. However, there are other pathways that can broaden the range of environmental signals which can induce epigenetic responses. For example, glucocorticoid receptors respond to a broad range of environmental signals via circulating glucocorticoids—including social stress and anxiety, and the reproductive and energetic environments [116–119]—and can change epigenetic states in response to the environment [32]. Endocrine responses to environmental inputs can occur at different time scales (e.g. from seconds to seasons). Different epigenetically mediated phenotypes that can persist across generations [73,120] may be induced therefore by environmental signals of varying length. However, an endocrine–epigenetic axis also means that epigenetic modifications of phenotypes are sensitive to endocrine disruption such as by EDCs and ALAN [15,45,74]. Hence, the underlying mechanisms very much determine the dynamics of phenotypic responses to environmental variation (phenotypic plasticity). Conservation and management rely on assessing animal responses to environmental change, which requires an understanding of these dynamics, particularly considering that change in multiple environmental drivers occurs concurrently and at different time scales [121,122]. It is important therefore to link the molecular mechanisms we describe here to broader-scale responses relevant for ecology and conservation, particularly when multiple stressors act through overlapping signalling pathways [123].

Even though there is a lot of excellent work on endocrine-mediated responses to the environment and on epigenetic mechanisms underlying phenotypic change, the mechanistic link between the two remains unresolved. Interesting research areas that remain to be addressed experimentally include (i) to determine whether or not the environment acts independently on each process or whether the environmental signal is transduced by a linked response; (ii) to establish the breadth of a potential endocrine–epigenetic axis with respect to different hormones and epigenetic processes. Addressing this last point would also contribute to assessing how environmental disruption such as by EDCs can affect phenotypes. Also, (iii) knowledge of interactions between hormones and epigenetic processes stems almost entirely from model organisms, and expanding this knowledge to wildlife is an important future direction.

Data accessibility

This article has no additional data.

Declaration of AI use

We have used AI-assisted technologies in creating figure 1.

Authors' contributions

F.S.: conceptualization, funding acquisition, investigation, project administration, writing—original draft, writing—review and editing; A.G.L.: conceptualization, funding acquisition, investigation, writing—review and editing.

Both authors gave final approval for publication and agreed to be held accountable for the work performed herein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

F.S. and A.G.L. were supported by Australian Research Council Discovery (grant no. DP220101342), and A.G.L. was supported by a Natural Sciences and Engineering Council of Canada (NSERC) Discovery (grant no. DGECR 2021-00123) and Alliance Collaboration (grant no. ALLRP 571888-21).