Decoding the language of epigenetics during neural development is key for understanding development as well as developmental neurotoxicity

Abstract

Neural development is a delicate process that can be disrupted by pollution that exerts detrimental impact on neural signaling. This commentary highlights recent discoveries in the arena of research at the interface of environmental toxicology and developmental neuroscience relating to toxicity mechanisms of bisphenol A (BPA), a ubiquitous chemical used in manufacturing of plastics and epoxy resins that is known to bind to and interfere with estrogen receptors, estrogen-receptor-related receptors and other receptors for gonadal steroids. It was recently observed that BPA disrupts the perinatal chloride shift, a key neurodevelopmental mechanism that brings down neuronal chloride from ~100 mM to ~20 mM within weeks. The chloride shift happens in all central nervous systems of vertebrates around parturition. High neuronal chloride supports neuron precursors’ migrations, low neuronal chloride is the prerequisite for inhibitory action of neurotransmitters GABA and glycine, and thus an absolute requisite for normal functioning of the mature CNS. One critical contributor to the neuronal chloride shift is the concomitant upregulation of expression of the chloride-extruding transporter molecule, KCC2. We highlight recent findings including our discovery that BPA disrupts the chloride shift in a sex-specific manner by recruiting epigenetics mechanisms. These could be relevant for childhood neuropsychiatric disorders as well as for liability to develop chronic neuropsychiatric diseases later in life.

Keywords: :

• neurodevelopmental toxicity
• bisphenol A
• chloride shift
• KCC2
• SLC12A5

This article refers to:

Development of the central nervous system (CNS), in particular the brain, is driven by highly coordinated cascades of sequential and concurrent events that begin very early in human gestation and continue through adolescence. During early development, brain cells proliferate, migrate to the appropriate location, differentiate into the correct cell type, and establish synapses with nearby and distant cells, ultimately forming complex neuronal circuits.^1-5 Disruption of this sequence of events at any point during brain development can have lasting consequences for brain function, which ultimately manifests as altered physiology, behavior and liability to respond to stress in a maladaptive manner and develop disease. Clear linkages between developmental disruption by environmental factors, including exposures to chemical pollution, and long-term neural and psychological outcomes, however, remain poorly characterized and thus a critical data gap toward improved understanding of neuropsychiatric disease.

Kcc2/KCC2 is a critical neuronal-specific gene of the CNS that codes for a chloride-extruding transporter molecule.^6-10 Developmental upregulation of Kcc2 expression is conserved in vertebrates, essential for the chloride shift in CNS neurons, proper migration of precursor neurons to final locations, and synchronization of maturation at the cellular and neural network level.^11-16 This is especially apparent for the delicate architecture of the mammalian cerebral cortex, where inhibitory precursor neurons originating from the median preganglionic eminence migrate considerable distances to their final destination in specific cortical layers.¹¹ This migration is critically dependent on the chloride shift and its appropriate developmental timing. Deceleration/attenuation of the ontogenetic upregulation of Kcc2 expression during the sensitive perinatal period can alter cortical organization and thus be regarded as an exemplar of developmental neurotoxicity.¹¹ In an earlier publication, we reported that a histone-deacetylase inhibitor, trichostatin-A (TSA) can accelerate the ontogenetic chloride shift by upregulating Kcc2 expression.¹⁵ This demonstrated for the first time the relevance of epigenetic regulation for Kcc2 gene expression.

Bisphenol A (BPA) is an endocrine disrupting compound that has been primarily characterized regarding its interactions with estrogen signaling on a diversity of cell types, but is now recognized to also have other possible mechanisms of action.¹⁷ BPA is a high production–volume plastic monomer produced at a rate of 3 million metric tons globally per year in the manufacturing of polycarbonate plastics, epoxy resins and plastic surface coatings, and is thus found in a wide variety of consumer items.¹⁸ It is estimated that more than 90% of the US population have detectable levels of BPA in their urine.^19-21 Concern has been expressed over the potential toxic and quasi-hormonal properties of BPA.

We recently demonstrated that BPA enhanced repression of Kcc2 transcription²² (Fig. 1). However, its effects did not result in an elimination of Kcc2 expression, which would be lethal, nor did it eliminate the upregulation of Kcc2. It did, however, decelerate and delay it, established in cultured neurons, organotypic brain slice culture and in vivo after gestational exposure. Therefore, BPA exposure could disrupt the layered architecture of the cerebral cortex because of over-migration of inhibitory precursor neurons. It is further anticipated that at this particularly sensitive time-point of neurodevelopment, BPA will disrupt dendritic spine formation and thereafter synaptic maturation, effectively compromising the functional set-up of inhibitory cortical neural circuits as well. This reasoning is consistent with studies showing disruption of cortical layer architecture in brains of mouse pups exposed to BPA through gestational feeding to their dams^23,24 These neuroanatomical abnormalities in turn dovetail extremely well with behavioral deficits that have been observed in mice and rats gestationally exposed to BPA.^17,25 Thus, we are describing a novel mechanism of neurodevelopmental toxicity by which altered neuronal chloride homeostasis and KCC2 expression subsequently compromise cortical neuronal and network structure. Because the chloride shift is important for development of the CNS elsewhere, by extension, other central neural structures may also be similarly impacted.

Figure 1. Exposure to BPA may disrupt development of the central nervous system by slowing down the removal of chloride from neurons. As an organism matures and the brain develops, chloride levels inside neurons decrease. However, when exposed to BPA, the chloride is removed more slowly from neurons. Female neurons appear to be more susceptible to the effects of BPA. In regards to the abscissa (left-hand), note a relevant ontological difference between rodents and humans: the chloride shift that coincides with the upregulation of KCC2/Kcc2 gene expression occurs between developmental stages E17 and P20 in rat neurons.^11,62 In humans, KCC2 levels start increasing by the second trimester and reaches a peak at birth.^15,22,63

Links between perinatal BPA exposure and behavioral pathology are increasingly evident in laboratory animals and humans, and BPA effects frequently differ in males and females.¹⁷ For instance, a number of studies reported anxiogenic effects of prenatal exposure to BPA in rodent species.^26-28 For many behavioral outcomes, a consistent finding is that perinatal exposure to BPA disrupts the development of normal sexually-dimorphic behaviors, including anxiety, exploration, social interaction, aggression, reward sensitivity and spatial memory.^28-32 In humans, sexually dimorphic disorders, such as attention-deficit/hyperactivity disorder and depression may be clinically correlated with such animal behaviors and may be related to disruption of the neuroendocrine system early in life.^33-35 Interestingly, a recent epidemiological study reported that gestational BPA exposure is associated with increased hyperactivity and aggression scores in young girls, but not in boys^36,37 (see also refs. 34 and 38 for recent data on human gestational exposure).

Understanding the mechanisms by which BPA alters CNS development via epigenetic effects on Kcc2 broadens our insight into how the environment can influence gene expression patterns and subsequent phenotype. In view of the possibly long-lasting duration of epigenetic changes it is reasonable to assume that such changes could be relevant for disease susceptibility later in life (Fig. 2). This reasoning is especially applicable to stresses and injuries that can potentially downregulate Kcc2 expression, such as traumatic brain injury, epilepsy or chronic pain states.^9,39-43 Subsequent upregulation would re-normalize chloride homeostasis in the CNS and lead to clinical improvement. However, re-normalization of KCC2/Kcc2 expression after stress- or injury-mediated downregulation of KCC2/Kcc2 could be delayed by perinatal epigenetic changes of the KCC2/Kcc2 promoter, thus contributing to disease chronicity for diseases in which KCC2/Kcc2 downregulation has been demonstrated to be a part of the molecular pathology in the CNS.^39,44

Figure 2. A model for neurodevelopmental disorders and susceptibility to suffer from maladaptive neural plasticity later in life: perinatal epigenetic changes of key genes for nervous system functioning, such as Kcc2, plasticity and repair mechanisms counteract. The degree and effectiveness of these regulatory (i.e., damage-limiting) mechanisms is further affected by how much they are affected by BPA exposure. In case these mechanisms are severely impaired themselves, then epigenetic dysregulation is more likely to lead to manifest neurodevelopmental disorders. In case counter-regulatory repair mechanisms are rather effective, it can lead to complete restitution, or it can lead to normal development, but a predisposition to show maladaptive changes in response to stress/ injury/ hormonal changes later in life. For example, development of a chronic pain syndrome, a post-traumatic brain injury syndrome, post-injury epilepsy and autism disorders.

Against this background, our discovery of the epigenetic effects of BPA on chloride regulation in neurons during the perinatal period²² raises several points of discussion that we are addressing here.

1) The methylating effects of BPA in in vivo and in vitro experiments appear to be different. Perinatal exposure to BPA may activate a specific series of biochemical pathways involving DNA synthesis and chromatin organization within maternal and fetal tissues that ultimately lead to activation of DNA-methyl-transferases and other components of the epigenetic machinery that target the Kcc2 gene promoter in cortical neurons and very likely other CNS neurons. In contrast, these biochemical pathways may not be similarly affected in a homogenous culture of cortical neurons. The critical difference could be rooted in non-neuronal cells in the developing brain, such as developing astrocytes. For example, developing glial cells, in response to BPA exposure, could facilitate neuronal epigenetic changes.⁴⁵ Additionally, differences could result from the function of the placenta as it is exposed to BPA. The developing brain could be affected by maternal humoral factors via the placenta so that, ultimately, epigenetic changes in developing neurons result.

2) Pertaining to relevance of our findings for human exposures, the mice in our experiments were fed chow containing 50 mg BPA/kg pellet; therefore, in our experiments, mouse exposure to BPA was effectively 10 mg/kg body weight/d through diet (≈250 µg). The amount fed to the mice, though on the high side, may still be physiologically relevant. A recent study measured epigenetic responses following gestational exposure to a range of BPA levels (ng-µg-mg). They showed overall global methylation across all BPA exposure groups in comparison to controls; however, there were no statistically significant differences in global methylation among the three BPA exposed groups.⁴⁶ Another group exposed rats to BPA in drinking water, leading to exposures of 35–55 µg per animal per day; by this method serum BPA levels were assessed to be equivalent to those reported in humans. They reported that BPA exposure decreased overall expression of genes known to affect socio-sexual behavior by about 1.5- to 2-fold. Interestingly, the sex differences of two sexually dimorphic genes (Kiss-1 and Bdnf) were eliminated with BPA exposure.²⁷ However, it remains to be seen if disappearance of sexual dimorphism of gene expression is mediated by epigenetic mechanisms. Yet, another group recently reported epigenetic effects in the form of DNA methylation and altered gene regulation in the brain in response to very low-dose gestational BPA exposures, applying 2, 2,0 and 200 µg BPA/kg/d.⁴⁷ Genes of the epigenetic machinery were affected as well as sex-hormone nuclear receptors, some of them in a sex-specific manner, also in the cerebral cortex and the hypothalamus. Offspring’s behavior was significantly altered at all three doses. This study documents epigenetic effects on gene regulation in the brain at the low end of the dose, thus providing counter-argument to previously published positions.⁴⁸

3) The sexual dimorphism effects of BPA that have been observed in our experiments point toward up-stream signaling mechanisms mediated by sex-hormone receptors. BPA is a documented estrogen disruptor.^49-52 Sex-specific estrogen receptor expression is present in perinatal brains of rodents and primates (including humans). For example, in the perinatal rat brain, estrogen is masculinizing. Perinatal exposure to BPA has also been shown to affect the organization of numerous estrogen-sensitive structures including the neocortex, amygdala, preoptic area, and hypothalamic regions important for sex-specific reproductive physiology and behavior.^{24,30,47,49,53-55} Viewing these results in concert, there is cause for concern that BPA exposure in humans may potentially result in adverse neuropsychiatric and other health outcomes.^56-58 For example, BPA exposure can also increase expression of leptin-receptors, thus predisposing affected subjects to future metabolic dysregulation, possibly obesity, if this dysregulation were affecting the brain.^59,60

Finally, our own results show that BPA led to increased CpG methylation and MECP2 occupancy in the upstream repressor-element 1 (RE1) site, a region ~1.8 kb upstream of the transcription start site (TSS) of the KCC2/Kcc2 gene, also known to harbor a CpG shore. Future experiments will be performed to ascertain if epigenetic alteration of the KCC2/Kcc2 shore region is causal for increased repression of transcription. For instance, using cultured cortical neurons, we can make use of the CRISPR-Cas system⁶¹ to engineer a deletion of the 500 bp promoter shore region containing the upstream RE1 site, then assess Kcc2 mRNA abundance, in the presence and absence of BPA.

In conclusion, our data revealing that BPA can delay the perinatal chloride shift in cortical neurons via epigenetic mechanisms unveils a novel mechanism by which environmental factors can shape neural development. Perturbation of the chloride shift likely affects neuronal migration and synaptic maturation in the developing brain’s cortex. Therefore, we have identified a novel example of neurodevelopmental toxicity, mediated by BPA. Our study evokes pertinent follow-up questions related to molecular and cellular mechanisms of DNA-methylation and other epigenetic mechanisms that affect Kcc2 expression. In addition, critical involvement of sex-hormone receptors upstream of epigenetic effects will be assessed.

Acknowledgments

Supported by funding from Duke University, the Klingenstein Fund, New York, NY, and the NIH (R21NS066307) to Liedtke W.

10.4161/epi.26406

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.