Mechanisms and modifiers of methylmercury-induced neurotoxicity

Abstract

The neurotoxic consequences of methylmercury (MeHg) exposure have long been known, however a complete understanding of the mechanisms underlying this toxicity is elusive. Recent epidemiological and experimental studies have provided mechanistic insights into the contribution of genetic and environmental factors that interact with MeHg to modify toxicity. This review will outline cellular processes directly and indirectly affected by MeHg, including oxidative stress, cellular signaling and gene expression, and discuss epigenetic modifications, genetic polymorphisms and gene–environment interactions capable of modifying MeHg neurotoxicity.

---

Introduction

The toxicity of mercury (Hg) is widely accepted, and while several Hg species induce neurotoxicity methylmercury (MeHg, CH₃Hg⁺) in particular is highly neurotoxic.¹ The majority of Hg is found in inorganic forms such as cinnabar ore. Inorganic mercuric salts enter the environment naturally through geocycling and anthropogenically through mining, industrial processes and waste incineration. Microorganisms generate organic mercury by methylation of elemental and inorganic mercury to form MeHg, which is then biomagnified in the food chain, particularly in aquatic fish and mammals. MeHg is the most common form of Hg that humans are exposed to, primarily through consumption of fish containing MeHg.

As early as the mid 1800s the neurotoxic effects of MeHg were evident from poisonings and deaths caused by industrial exposure. In Minamata, Japan, industrial wastewater containing MeHg was released into the environment over a number of years and bioaccumulated in fish consumed by the local population. In the 1950s many adults developed Minamata disease, characterized by paresthesia, sensory deficits, slurred speech, unsteady gait, muscle weakness, irritability, memory loss, depression, and sleeping disturbance.² Similar symptoms were also observed in Iraq in 1956 and 1960 after consumption of bread made from grain treated with ethylmercury and again in 1971 from grain treated with methylmercury.³

The first recorded case of prenatal MeHg neurotoxicity in humans occurred in the 1950s, a few years later children in Minamata, Japan, presented with severe neurodevelopmental deficits that were later related to maternal consumption of fish contaminated with MeHg. Studies from the Iraq exposures provided further evidence that the developing brain is particularly sensitive to MeHg.^3,4 Since these incidences, several large-scale epidemiological studies have examined prenatal MeHg toxicity in humans from Minamata and fish-eating regions in the Seychelles, Faroe Islands and New Zealand.

This review will outline cellular processes directly and indirectly affected by MeHg and discuss genetic, environmental and nutritional factors capable of modifying MeHg toxicity.

Mechanisms of toxicity

The toxicological properties of MeHg are largely related to its pro-oxidative effects. In the nervous system, MeHg disturbs oxidative balance, mitochondrial health, induces apoptosis, disrupts calcium homeostasis and alters glutamate and γ-aminobutyric acid (GABA) signaling. In addition to directly promoting neurotoxicity, these biochemical consequences of MeHg exposure cause indirect dysregulation of cellular signaling and gene expression, which also contributes to toxicity.

Oxidative stress

MeHg is a soft electrophile and interacts with thiol (–SH) and selenol (–SeH) groups, which are the only biological soft nucleophiles. Thiols and selenols play a fundamental role in MeHg-induced toxicity.⁵ Hg compounds react specifically with sulfhydryls and selenols, forming stable complexes with defined stoichiometry. Hg's affinity for these groups is extremely high⁶ and in biological media MeHg is always complexed to –SH-containing ligands.^7–10 Mechanisms of MeHg's membrane transport also invoke –SH-containing amino acids.^11–17 MeHg conjugates with cysteine (Cys) and is transported via the large amino acid transporter (LAT), which transports methionine.¹⁷

Several important regulators of the antioxidant response are rich in thiol and selenol groups. Glutathione (γ-glutamyl-cysteinyl-glycine; GSH) is a highly abundant tripeptide, which serves as a reducing agent for reactive oxygen species (ROS) and other unstable molecules, a reaction catalyzed by the selenoprotein glutathione peroxidase (GPx). GSH can also be conjugated to toxins to facilitate excretion during detoxification. The majority of GSH is present in the reduced form (GSH) while approximately 10% is present as the oxidized form glutathione disulfide (GSSG), the GSSG : GSH ratio reflects the oxidative state of the cell.¹⁸ MeHg interacts with GSH to form an excretable GS–HgCH₃ complex,¹⁴ increasing the GSSG : GSH ratio and reducing the antioxidant capacity both in astrocytes and microglia.^19,20 Furthermore, thioredoxins (Trxs), a family of proteins with abundant thiol groups, are essential in regulating cellular thiol/disulfide redox status. Trxs contain an invariant active center, -Cys-Gly-Pro-Cys- where the two Cys residues undergo reversible oxidation/reduction in the presence of the selenoprotein Trx reductase (TrxR) and NADPH.²¹ The cytosolic Trx (Trx 1) is also implicated in growth stimulation, transcriptional regulation and apoptosis.²² The mitochondrial Trx (Trx 2) is a critical component of the mitochondrial antioxidant system and protects against oxidative stress-induced cell death.²³In vitro, MeHg inhibits GPx and TrxR activity.²⁴ Prior to inducing cell death, MeHg inhibits GPx activity in cultured cerebellar granule cells, and overexpression of GPx prevents cell death.²⁵ These effects also occur in vivo in mice and fish.^26,27

In addition to impairing the antioxidant response of the cell, MeHg increases ROS production. Hydrogen peroxide (H₂O₂) levels are increased concomitant with altered motor reflexes in mouse pups exposed to MeHg through lactation.²⁸ H₂O₂ levels are also increased by MeHg in cultured astrocytes²⁹ and in mitochondria isolated from mouse and rat brain.^30,31 Decreased H₂O₂ detoxification resulting from reduced GSH and GPx activity contributes to these increases, although Mori and colleagues observed increased H₂O₂ production from MeHg-induced changes in the mitochondrial electron transport chain (ETC) complexes II and III in mitochondria isolated from cerebellum, but not liver or cerebrum.^31,32 Superoxide anion (O₂˙⁻) levels also increased due to changes in ETC complexes I and III respiration.³³ Nitric oxide (NO) production rises as well following MeHg exposure,^34–36 likely through stimulation of nitric oxide synthase (NOS) by increased Ca²⁺.

The resulting pro-oxidative shift has wide-reaching consequences including lipid peroxidation, protein oxidation, DNA oxidation, altered Ca²⁺ homeostasis, impaired mitochondria and apoptosis. Owing to the high content of polyunsaturated fatty acids (PUFAs) in the central nervous system, lipid peroxidation is a major consequence of free radical-mediated injury. H₂O₂ undergoes Fenton chemistry, producing the hydroxyl radical (˙OH), a primary cause of lipid oxidation.³³ The oxidative shift caused by MeHg also affects DNA. A study of Chinese mine workers and local residents chronically exposed to various forms of MeHg from Hg mining plants revealed increased urinary 8-oxo-2′-deoxyguanosine (8-OHdG) levels, a metabolite of oxidized DNA, as well as increased serum markers of oxidation in exposed individuals compared to age and gender matched control individuals.³⁷

Mitochondrial function is also affected by oxidation of thiol rich proteins, such as mitochondrial creatine kinase and respiratory chain complexes.^38,39 Mori et al. demonstrated that MeHg disrupts ETC function in cerebellar mitochondria from young adult rats exposed to MeHg for five days.³² Consistent with mitochondrial dysfunction, Hg²⁺ increased mitochondrial permeability and swelling, promoted de-coupling, and caused depolarization in isolated rat liver mitochondria.⁴⁰ Furthermore, MeHg inhibits uptake of Ca²⁺ and stimulates Ca²⁺ release in neuronal mitochondria.⁴¹

The pro-oxidative chemical properties and biological effects of MeHg have been shown to induce neuronal apoptosis in a variety of in vitro and in vivo models and neuronal cell types (for comprehensive review see ref. 42). MeHg-induced apoptosis is reduced in cultured primary cortical neurons and rat cerebral cortical neurons by the cysteine derivative N-acetylcysteine,⁴³ suggesting that the high affinity of MeHg for thiol groups (including cysteine) may contribute to neuronal apoptosis. Dysregulation of Ca²⁺ homeostasis is another important factor mediating MeHg-induced apoptosis. In addition to mitochondrial Ca²⁺, MeHg increases intracellular Ca²⁺ through N-methyl-D-aspartate (NMDA) receptors and inositol trisphosphate (IP₃) sensitive endoplasmic reticulum stores (discussed below). In an in vivo rat model, inhibition of NMDA receptors reduces apoptosis in the frontal cortex.⁴⁹ Altogether, MeHg reduces ETC capacity, increases ROS production and alters Ca²⁺ causing opening of the mitochondrial permeability transition pore (mPTP).^44,45 The mPTP allows release of cytochrome c oxidase from the inner mitochondrial membrane and triggers apoptotic cell death (for review see ref. 46).

Altered neurotransmitters

In addition to increasing oxidative stress, MeHg directly inhibits proteins necessary for calcium homeostasis, glutamate transport and GABA signaling, processes that are both exacerbated by and contribute to oxidative stress. Astrocytes have a central role in maintaining the synapse through regulation of neurotransmitter metabolism and reuptake and supporting neuronal metabolism through the glutamate–glutamine cycle.⁴⁷ MeHg preferentially accumulates in astrocytes and increases extracellular glutamate levels by reducing glutamate uptake and increasing glutamate release in astrocytes.⁴⁸ In cerebellar slices, Manfroi et al. demonstrated a nearly 50% reduction in glutamate uptake in slices from mice exposed to MeHg through milk compared to unexposed controls²⁸ This increase in glutamate results in excitotoxicity, which can be blocked in part by blocking NMDA receptors.⁴⁹ Mercuric chloride (HgCl₂), which accumulates in the brain following MeHg exposure, decreases glutamine synthetase activity and inhibits glutamine transport in cultured astrocytes.^50,51

Along with glutamate stimulated Ca²⁺ entry through NMDA receptors, MeHg causes release of IP₃ sensitive endoplasmic reticulum Ca²⁺ stores through direct inhibition of the thiol-rich endoplasmic Ca²⁺-ATPase.^52,53 Disrupted Ca²⁺ homeostasis contributes to toxicity by increasing neurotransmitter release, altering mitochondrial membrane potential, and promoting Ca²⁺-mediated intracellular signaling.^52,53

In contrast to increased excitatory glutamate signaling, MeHg decreases GABA signaling, the most abundant inhibitory neurotransmitter. In primary cerebellar granule cells, MeHg interacts with thiol resides on GABA_A receptors, which modifies receptor conformation.⁵⁴ Chronic low-level MeHg exposure in captive minks reduces the activity of the synthetic enzyme glutamic acid decarboxylase, and significantly decreases of GABA_A receptors and transport activity in the brain stem and basal ganglia.⁵⁵ Glutamate and GABA signaling are crucial for establishing equilibrium between excitation and inhibition. Shifts in this equilibrium are associated with regulation of normative critical periods in neurodevelopment⁵⁶ as well as with neurologic disorders such as epilepsy.⁵⁷ In combination, these direct effects of MeHg on the regulation of neuronal signaling likely contribute to MeHg-induced neurotoxicity.

Dysregulation of intracellular signaling

MeHg affects the regulation of several signaling pathways including phospholipase C (PLC), calcium signaling and phosphatidylinositol 3-kinases/protein kinase (PI3K/Akt). MeHg in MDCK cells activates PLC and PI3K/Akt, blocking PLC activation reduced toxicity.⁵⁸ PLC activation, in addition to the effects described above, increases intracellular Ca²⁺. Chang et al. demonstrated that a PLC-stimulated Ca²⁺ increase mediates interleukin-6 (IL-6) release in cerebellar granule cells exposed to MeHg, triggering an inflammatory response.⁵⁹ This is consistent with upregulation of inflammatory gene expression observed in a microarray from cerebellum of young adult mice injected with MeHg for seven days.⁶⁰

In pancreatic β-cells, MeHg increases PI3K/Akt activity,⁶¹ in part due to increased ROS production and in part through inactivation of PTEN, a thiol-rich PI3K inhibitor.⁶² Low concentrations of MeHg trigger ROS production and increase PI3K activity and its downstream effector phospho-Akt,⁶³ pharmacological inhibition of PI3K attenuates these changes in astrocytes.¹⁹ PI3K is activated by a G-protein-coupled receptor or receptor tyrosine kinase such as the insulin receptor.⁶⁴ Once activated, PI3K phosphorylates phosphatidylinositol 4,5-bisphosphate [PtdIns(4,5)P2] to form phosphatidylinositol (3,4,5)-triphosphate [PtdIns(3,4,5)P3]. Akt functions downstream of PI3K and is activated by phosphorylation.⁶⁴ Akt activation has many consequences including regulation of the protective transcription factors forkhead box protein O (FOXO) and nuclear factor (erythroid-derived 2)-like 2 (Nrf2) and additional signaling cascades regulating cellular growth and morphology. These observations are consistent with MeHg's widespread damage in fetal and neonatal brain, characterized by hypoplastic and symmetrical brain atrophy, and reflective of aberrant cell division, migration, differentiation and synaptogenesis.^2,65–67

Aberrant gene expression

Alterations in the redox status of a cell can activate several pathways. Nrf2 is a basic leucine zipper transcription factor that under unstressed conditions resides in the cytoplasm bound by the Keap1 protein, an adapter protein for Cullin 3-dependent ubiquitination and proteasomal degradation.⁶⁸ However, during oxidative stress, such as exposure to MeHg, Nrf2 is released from kelch-like ECH-associated protein1 (Keap1) by disruption of cysteine residues on Keap1 or through oxidant-dependent kinase signaling, and translocates into the nucleus, where it can bind to antioxidant response element (ARE) promoters as a heterodimer with Maf proteins.⁶⁸ In addition to redox regulation, a crosstalk exists between the protein kinase pathways and the Nrf2-dependent antioxidant system.⁶⁹ Nrf2 interacts with p38 mitogen-activated protein kinase,^70,71 ER-resident kinase PERK⁷² and a Src family tyrosine kinase, Fyn.⁷³ The PI3K/Akt pathway controls Nrf2's function upstream.

Nrf2 induces the expression of several cytoprotective proteins including NAD(P)H quinine oxidoreductase (NQO1), glutamate-cysteine ligase catalytic subunit and modifier subunit (GCLC and GCLM), and heme oxygenase-1 (HO-1).⁶⁸ MeHg has been shown to activate Nrf2 in both cell lines and primary cells, allowing for an up-regulation of NQO1 and HO-1.¹⁹ In addition to oxidative stress derived from MeHg exposure, Toyama et al. have shown MeHg's ability to bind to recombinant Keap1 and to activate Nrf2.⁷⁴ Nrf2 has been shown to be protective in a Caenorhabditis elegans model of dopaminergic neurodegeneration induced by MeHg, where reduction of SKN-1, the Nrf2 homologue, increases the vulnerability for neurodegeneration.⁷⁵ Activation of Nrf2 by MeHg is a protective response; however MeHg often alters gene expression resulting in neurotoxic effects.

Recently, several toxicogenomics studies have been performed to characterize changes in gene expression upon MeHg exposure.^60,76,77 In the cerebellum of eight-week-old mice exposed to 10 mg kg⁻¹ per day for 7 days, several genes involved in inflammation (such as the chemokines, CCL2 and CCL4), metabolism, anti-apoptosis (Bcl2a1b) and signal transduction were increased above control, while genes involved in cell proliferation, oxidative stress (selenoprotein H) and apoptosis (Ifi27) were decreased.⁶⁰ Increased oxidative stress can cause inflammation through the activation of the nuclear factor kappa-light-chain-enhancer of activated B cells (NF-kB) transcription factor, and inflammation contributes to several neurologic diseases, such as Alzheimer's disease and Parkinson's disease.⁷⁸ MeHg has been shown to increase IL-6 release from glial cells⁷⁹ and activation of NF-kB has been observed in rat cerebrum and cerebellum.⁸⁰

Changes in gene expression can be more severe if the exposure occurs during neurodevelopment. The adult brain has a complex structure formed by several cell types and numerous connections. Disruption of normal neurodevelopment or death of cells caused by exposure to a xenobiotic can have consequences on the functioning of the adult brain. There are several reports of genome wide analyses of genes that are up-regulated and down-regulated in adults or in various stages of development. In a comparative study, pooling transcriptomics data sets from studies involving mouse embryos exposed to MeHg in utero, young mice exposed both in utero and postnatally, adult rats chronically exposed to MeHg, whole embryo cultures exposed to MeHg, embryonic stem cells induced into cardiomyocytes or neurons and mouse embryonic fibroblasts exposed to MeHg, Robinson et al. found that although different models were used, as well as experimental variables of dose and time of exposure, there is a set of genes that are consistently altered by MeHg that effect development of brain as well as heart and kidney.⁷⁶ Using neural cells differentiated from murine embryonic stem cells exposed to 25 nM MeHg for 8 consecutive days, Theunissen et al. found MeHg to increase late differentiation gene and neuroectodermal gene sets, and to decrease early differentiation gene, mesodermal and endodermal derived tissue gene sets, resulting in cultures with retarded development that deviated from their normal differentiation track.⁷⁷ Large studies like these are important in identifying pathways affected by MeHg exposure although exact mechanisms of how an individual gene is up or down regulated may not be clear.

Epigenetic modifications

Changes in gene expression, whether up regulated or down regulated, can occur not just through inhibiting or activating transcription factors, but also through epigenetic modifications. Epigenetic regulation of gene expression does not occur through changes in DNA sequence but from post-translational modifications of histone proteins that package DNA or through methylation of the DNA. Orishchenko et al. investigated the changes in expression of brain-derived neurotrophic factor (BDNF), a gene known to be regulated by epigenetic modifications and associated with neuronal development, learning, and long-lasting depression-like behavior. Mice exposed to 0.5 mg kg⁻¹ per day MeHg via drinking water from gestational day seven to postnatal day seven show depressed behavior through measures of immobility in forced swim tests.⁸¹ MeHg decreased BDNF expression in the dentate gyrus of the hippocampus.⁸¹

The BDNF promoter in hippocampus of mice exposed to MeHg showed decreased histone H3 acetylation on K9 and K14, a marker of active chromatin state, and increased trimethylation on histone H3 K27, a marker of repressed gene expression.^81,82 Additionally, DNA methylation was enriched in four separate CpG sites in the BDNF promoter after exposure to MeHg.⁸¹ Hypomethylated DNA correlates with increased DNA binding to transcription factors whereas hypermethylated DNA is associated with reduced transcription factor binding.⁸³ All of these markers cumulatively suggest that the reduction in BDNF levels in the hippocampus after MeHg exposure is a result of epigenetic modifications, suggesting that MeHg may affect other genes through alteration of epigenetic markers. The extent to which epigenetic modifications contribute to MeHg-induced gene expression and the mechanisms through which MeHg alters epigenetic markers are not known.

Modifiers of toxicity

Genetic polymorphisms

The susceptibility of an individual to a toxic chemical is determined not solely by the toxicant itself but also by variations in alleles and inherited defects. Recent studies suggest that the body's response to Hg may be mediated by polymorphisms in several genes responsible for absorption, distribution, metabolism and excretion.

A study in Austria by Gundacker and colleagues, examined in medical students the potential associations between Hg exposure and the glutathione S-transferases (GSTs), GST theta 1, GSTT1; GST mu 1, GSTM1; GSTA1, GST pi 1, GSTP1; glutamate-cysteine ligase catalytic subunit, GCLC; and metallothionine (MT) polymorphisms. The levels of Hg exposure of students were low as determined by analysis of blood, urine, and hair. GSTP1-114 and MT4 allele carriers as well as GSTT1−/− and GSTM1−/− jointly deleted phenotypes were associated with higher hair mercury levels than homozygous wild types. GSTP1-114/GSTT1 and GSTP1-105/GCLC combinations showed synergistic effects on hair Hg levels compared to single-gene variants and the authors suggest that the GSTP1 variants are more important in mercury toxicokinetics than the other GST polymorphisms.⁸⁴

In a study of Michigan dental professionals, Goodrich and colleagues evaluated polymorphisms in key glutathione synthesizing enzymes, glutathione S-transferase, and selenoprotein genes underlying inter-individual differences in Hg body burden. They measured urine and hair Hg levels. Five polymorphisms were significantly associated with urine Hg levels (GSTT1 deletion), hair Hg levels (GSTP1-105, GSTP1-114, GSS 5′), or both (SEPP1 3′UTR). The authors suggest that polymorphisms in these genes may influence elimination of Hg in the urine and hair or retention following exposures to elemental Hg through dental amalgams and MeHg through fish consumption.⁸⁵ Similar data was presented by Schlawicke Engstrom and colleagues, demonstrating that carriers of GSTP1-105G and GSTP1-114T accumulated less Hg in erythrocytes after controlling for levels of polyunsaturated fatty acids.⁸⁶ Custodio and colleagues reported that the T allele of the GSTP1-114 polymorphism was associated with increased blood mercury levels.⁸⁷

Autism is a neurodevelopmental disorder with genetic and environmental components, and genetic susceptibility to high Hg has been suggested as a potential risk factor. Owens and colleagues tested the hypothesis that Hg could be implicated in the etiology of autism if genetic susceptibility altered Hg's metabolism or intracellular compartmentalization. Genetic sequences of four genes implicated in the transport [LAT1 and divalent metal transporter1 (DMT1)] and response to Hg [metal regulatory transcription factor1 (MTF1) and metallothionein1a (MT1a)] were screened for variation and association with autism. The group identified and characterized 74 variants in MT1a, DMT1, LAT1 and MTF1. Polymorphisms identified were evaluated for differences in allele frequencies. They reported no association of any variant evaluated with autism suggesting that variations in these genes may not play a significant role in the etiology of autism.⁸⁸ Another group using the C. elegans model system to study synapse formation and function as related to autism showed that neuroligin mutants (postsynaptic cell adhesion proteins that bind specifically to presynaptic membrane proteins-neurexins) are hypersensitive to Hg compounds and are defective in sensory behaviors and processing. They suggest that the behavioral deficits are similar to traits frequently associated with autism spectrum disorders and report a possible link between genetic defects in neuroligin and sensitivity to Hg in the development of autism.⁸⁹

Hg exposure is associated with cardiovascular problems. A common polymorphism of matrix metalloproteinase (MMP)-2 gene is the C(−1306)T which disrupts a promoter site and is associated with lower promoter activity when the T allele is present affecting the expression and activity of the enzyme. A study by Jacob-Ferreira and colleagues examined how this polymorphism affects the circulating MMP-2 levels and tissue inhibitor of metalloproteinase-2 (TIMP-2) following environmental exposure to Hg. They reported a positive association between plasma Hg concentrations and the ratio of MMP-2/TIMP-2. The C(−1306)T polymorphism modified MMP-2 concentrations and MMP-2/TIMP-2 ratio in subjects exposed to Hg, with higher MMP-2 levels found in subjects carrying the C allele. Their findings suggest a significant interaction between the C(−1306)T polymorphism and Hg exposure, with possible ramifications for those carrying the C allele.⁹⁰

Another study examined the role polymorphisms of the endothelial nitric oxide (eNOS) gene (T-786C and Glu298Asp) on nitrite concentrations following Hg exposure in humans. The hypothesis tested was that Hg exposure might decrease circulating nitrite concentrations and these polymorphisms might enhance the effects of Hg resulting in increased risk of cardiovascular disease. Decreased nitric oxide (NO) production was found to be predominantly related to Hg, age and gender rather than polymorphisms. These findings suggest that there is not an association between increased risk for cardiovascular disease in MeHg-exposed subjects and eNOS gene polymorphisms (T-786C and Glu298Asp).⁹¹ The same group investigated the contribution of the 27 nt tandem repeat of intron 4 of the eNOS gene to NO production, which could enhance susceptibility to cardiovascular disease in the MeHg-exposed study population. They found no significant differences in age, arterial blood pressure, body mass index or cardiac frequency between genotype groups, but observed different nitrite concentrations, with lower nitrite levels for the 4a4a genotype carriers. Age, gender and the presence of intron 4 polymorphism contributed to nitrite reduction as a result of blood Hg concentration. Their results suggest increased susceptibility to cardiovascular diseases after MeHg exposure in carriers of the 27 nt repeat polymorphism of intron 4 in the eNOS gene.⁹²

In a Swedish study, Engström and colleagues studied whether genetic polymorphisms in glutathione-related genes modify the association between the PUFAs eicosapentaenoic and docosahexaenoic acid (DHA) or MeHg and risk of first ever myocardial infarction. Polymorphisms in glutathione-synthesizing (glutamyl-cysteine ligase catalytic subunit, GCLC and glutamyl-cysteine ligase modifier subunit, GCLM) or glutathione-conjugating (glutathione S-transferase P, GSTP1) genes were assessed. The authors evaluated the impact of these polymorphisms on the association between erythrocyte-Hg and risk of myocardial infarction, as well as between plasma eicosapentaenoic and DHA and risk of myocardial infarction. They reported no statistically significant genetic modifying effects for the association between plasma eicosapentaenoic and DHA or erythrocyte-Hg and risk of myocardial infarction but do report a relatively rare GCLM-588 TT genotype that may have an impact.⁹³

An association between the Hg resistance merA gene and antibiotic resistance has been described in the literature and has been shown to be widely distributed among bacteria (both Escherichia coli and Staphylococcus aureus) isolated from healthy adults and children.^94–99 To study this further, Skurnik and colleagues analyzed Hg resistance in collections of strains from two populations with different levels of Hg exposure and various levels of antibiotic resistance. Hg-resistant E. coli was found significantly more frequently in the population that had the highest antibiotic-resistant E. coli and antibiotic resistance was higher in the population living in an environment with a high exposure to Hg.¹⁰⁰

Developmental vulnerability

In addition to genetic polymorphisms, susceptibility to MeHg is significantly altered by the age and developmental stage of the exposed individual. The increased vulnerability of the developing nervous system to MeHg is not disputed. Developmental exposure to MeHg, even at very low levels, results in more severe and widespread damage than the same exposure levels in adults. Neurodevelopment is characterized by rapid progression of developmental processes within defined critical periods including neural tube closure, cellular proliferation and differentiation and synaptic development. MeHg can alter many of these events, Robinson et al. observed delayed neural tube closure in embryos of two different mouse lines exposed to MeHg.¹⁰¹ In zebrafish embryos, MeHg delayed hatching and reduced cellular proliferation in the neural tube.¹⁰² In neuronally differentiating mouse embryonic stem cell culture, expression of neuronal differentiation markers, neurotransmitter receptors and transporters as well as growth factors was reduced by MeHg.¹⁰³

Optimal neurodevelopment requires large amounts of energy, and as a result mitochondrial activity and subsequent ROS production is higher in the developing nervous system than in the adult nervous system.¹⁰⁴ Accordingly, in mice the GSH system develops rapidly, with significant increases in GSH, glutathione reductase (GR) and GPx occurring in the first three postnatal weeks.¹⁰⁵ Prenatal MeHg exposure not only prevents these increases in antioxidant capacity, but also continues to suppress the system at postnatal day 21, when MeHg levels were no longer elevated, GSH, GR and GPx remained decreased relative to controls.¹⁰⁵ This suggests that the developing nervous system is vulnerable to the acute pro-oxidative effects of MeHg, which may affect the antioxidant response throughout life.

In addition to promoting plasticity and survival throughout life, growth factors such as BDNF and neuronal growth factor influence cell migration, differentiation and survival, axonal and dendrite structure and targeting, and synapse formation during development.¹⁰⁶ Most neurotrophic factors signal through intracellular cascades susceptible to the oxidative effects of MeHg as described above. Together these early changes not only cause immediate developmental abnormalities such as observed in congenital Minamata disease, but may also contribute to latent effects. For example, neurological deficits have emerged months and years following MeHg exposure in Minamata as well as in other instances of human exposure and experimental primate studies.^107,108

Nutrition and additional toxicants

In some instances, variation and manipulation of external factors can enhance or attenuate the toxicity of MeHg. For example the benefits of nutrients found in fish such as Se, and PUFAs including omega-3 fatty acids like DHA may counterbalance some of the toxic effects of MeHg. Inclusion of maternal PUFA levels during gestation and shortly after birth in the Seychelles studies revealed beneficial effects of PUFAs on developmental outcomes, particularly at 9 months of age.¹⁰⁹

Experimental studies have shown that Se supplementation attenuates growth and motor deficits in rats chronically exposed to MeHg following weaning, the severity of deficits was not directly related to MeHg levels in the brain but rather by the Hg : Se ratio.¹¹⁰ In 15-day-old mice born to dams fed MeHg with or without Se supplementation through gestation and lactation, combined Se and MeHg regulated expression of 63 genes in whole genome cerebral microarray compared to 8 and 5 genes affected by Se or MeHg alone.¹¹¹ This is consistent with functional Se deficiency due to MeHg inhibition of selenoproteins, effects that can be lessened by Se supplementation.¹¹² In a similar study, Jayashankar and colleagues exposed mice to MeHg and/or DHA through maternal diet. DHA decreased accumulation of MeHg in the brain and eliminated reflexive deficits observed in non-DHA supplemented MeHg pups, as well as altering gene expression.¹¹³

On the other hand, the contribution of concomitant exposure to additional environmental toxins such as arsenic, pesticides, and polychorinated biphenyls (PCB) to MeHg toxicity has also been investigated. There is limited support for interactions between MeHg and these additional toxins; however many studies have failed to demonstrate consistent interactive toxicity and more research is needed.¹¹⁴

Summary

MeHg-induced neurotoxicity is mediated at the cellular level by a complex interplay between direct and indirect consequences of the pro-oxidative nature of MeHg, including oxidative stress, disrupted neuronal signaling, intracellular signaling, and gene expression. However, these effectors alone do not provide a satisfactory understanding of MeHg toxicity. The contribution of gene-environment interactions, such as polymorphisms, developmental stage, and additional nutritional and chemical factors, are significant and may allow the development of more effective treatments to proactively attenuate the effects of MeHg.

Acknowledgements

We are grateful for support from NIEHS R01ES07331, the Center in Molecular Toxicology NIH grant P30ES00267 and the Training Program in Environmental Toxicology grant T32ES007028.

Notes and references

1. Toxicological Profile for Mercury, ed. U. S. D. o. H. a. H. Services, P. H. Service and A. f. T. S. a. D. Registry, 1999.
2. M. Harada, Crit. Rev. Toxicol., 1995, 25, 1–24.
3. F. Bakir, S. F. Damluji, L. Amin-Zaki, M. Murtadha, A. Khalidi, N. Y. al-Rawi, S. Tikriti, H. I. Dahahir, T. W. Clarkson, J. C. Smith and R. A. Doherty, Science, 1973, 181, 230–241.
4. P. Grandjean and K. T. Herz, Mt. Sinai J. Med., 2011, 78, 107–118.
5. W. L. Hughes, Ann. N. Y. Acad. Sci., 1957, 65, 454–460.
6. N. J. Taylor and A. J. Carty, J. Am. Chem. Soc., 1977, 99, 6143–6145.
7. A. Naganuma and N. Imura, Toxicol. Appl. Pharmacol., 1979, 47, 613–616.
8. S. Omata, K. Sakimura, T. Ishii and H. Sugano, Biochem. Pharmacol., 1978, 27, 1700–1702.
9. D. L. Rabenstein and M. T. Fairhurst, J. Am. Chem. Soc., 1975, 97, 2086–2092.
10. T. Refsvik and T. Norseth, Acta Pharmacol. Toxicol. (Copenh), 1975, 36, 67–78.
11. M. Aschner and J. L. Aschner, Neurosci. Biobehav. Rev., 1990, 14, 169–176.
12. M. Aschner and T. W. Clarkson, Brain Res., 1988, 462, 31–39.
13. M. Aschner and T. W. Clarkson, Pharmacol. Toxicol., 1989, 64, 293–297.
14. N. Ballatori and T. W. Clarkson, Fundam. Appl. Toxicol., 1985, 5, 816–831.
15. L. E. Kerper, N. Ballatori and T. W. Clarkson, Am. J. Physiol., 1992, 262, R761–765.
16. L. E. Kerper, E. M. Mokrzan, T. W. Clarkson and N. Ballatori, Toxicol. Appl. Pharmacol., 1996, 141, 526–531.
17. Z. Yin, H. Jiang, T. Syversen, J. B. Rocha, M. Farina and M. Aschner, J. Neurochem., 2008, 107, 1083–1090.
18. N. Ballatori, S. M. Krance, S. Notenboom, S. Shi, K. Tieu and C. L. Hammond, Biol. Chem., 2009, 390, 191–214.
19. L. Wang, H. Jiang, Z. Yin, M. Aschner and J. Cai, Toxicol. Sci., 2009, 107, 135–143.
20. M. Ni, X. Li, Z. Yin, H. Jiang, M. Sidoryk-Wegrzynowicz, D. Milatovic, J. Cai and M. Aschner, Toxicol. Sci., 2010, 116, 590–603.
21. Y. Meyer, B. B. Buchanan, F. Vignols and J. P. Reichheld, Annu. Rev. Genet., 2009, 43, 335–367.
22. G. Powis and W. R. Montfort, Annu. Rev. Biophys. Biomol. Struct., 2001, 30, 421–455.
23. Y. Chen, J. Cai, T. J. Murphy and D. P. Jones, J. Biol. Chem., 2002, 277, 33242–33248.
24. C. M. Carvalho, E. H. Chew, S. I. Hashemy, J. Lu and A. Holmgren, J. Biol. Chem., 2008, 283, 11913–11923.
25. M. Farina, F. Campos, I. Vendrell, J. Berenguer, M. Barzi, S. Pons and C. Sunol, Toxicol. Sci., 2009, 112, 416–426.
26. C. Wagner, J. H. Sudati, C. W. Nogueira and J. B. Rocha, BioMetals, 2010, 23, 1171–1177.
27. V. Branco, J. Canario, A. Holmgren and C. Carvalho, Toxicol. Appl. Pharmacol., 2011, 251, 95–103.
28. C. B. Manfroi, F. D. Schwalm, V. Cereser, F. Abreu, A. Oliveira, L. Bizarro, J. B. Rocha, M. E. Frizzo, D. O. Souza and M. Farina, Toxicol. Sci., 2004, 81, 172–178.
29. G. Shanker, J. L. Aschner, T. Syversen and M. Aschner, Mol. Brain Res., 2004, 128, 48–57.
30. J. L. Franco, H. C. Braga, J. Stringari, F. C. Missau, T. Posser, B. G. Mendes, R. B. Leal, A. R. Santos, A. L. Dafre, M. G. Pizzolatti and M. Farina, Chem. Res. Toxicol., 2007, 20, 1919–1926.
31. N. Mori, A. Yasutake and K. Hirayama, Arch. Toxicol., 2007, 81, 769–776.
32. N. Mori, A. Yasutake, M. Marumoto and K. Hirayama, J. Toxicol. Sci., 2011, 36, 253–259.
33. B. Chance, H. Sies and A. Boveris, Physiol. Rev., 1979, 59, 527–605.
34. A. M. Herculano, M. E. Crespo-Lopez, S. M. Lima, D. L. Picanco-Diniz and J. L. Do Nascimento, Braz. J. Med. Biol. Res., 2006, 39, 415–418.
35. T. Himi, M. Ikeda, I. Sato, T. Yuasa and S. Murota, Brain Res., 1996, 718, 189–192.
36. T. Yamashita, Y. Ando, N. Sakashita, K. Hirayama, Y. Tanaka, K. Tashima, M. Uchino and M. Ando, Biochim. Biophys. Acta, Gen. Subj., 1997, 1334, 303–311.
37. C. Chen, L. Qu, B. Li, L. Xing, G. Jia, T. Wang, Y. Gao, P. Zhang, M. Li, W. Chen and Z. Chai, Clin. Chem., 2005, 51, 759–767.
38. V. Glaser, G. Leipnitz, M. R. Straliotto, J. Oliveira, V. V. dos Santos, C. M. Wannmacher, A. F. de Bem, J. B. Rocha, M. Farina and A. Latini, NeuroToxicology, 2010, 31, 454–460.
39. V. Glaser, E. M. Nazari, Y. M. Muller, L. Feksa, C. M. Wannmacher, J. B. Rocha, A. F. de Bem, M. Farina and A. Latini, Int. J. Dev. Neurosci., 2010, 28, 631–637.
40. E. A. Belyaeva, V. V. Glazunov and S. M. Korotkov, Chem.-Biol. Interact., 2004, 150, 253–270.
41. P. C. Levesque and W. D. Atchison, J. Pharmacol. Exp. Ther., 1991, 256, 236–242.
42. S. Ceccatelli, E. Dare and M. Moors, Chem.-Biol. Interact., 2010, 188, 301–308.
43. A. Falluel-Morel, L. Lin, K. Sokolowski, E. McCandlish, B. Buckley and E. DiCicco-Bloom, J. Neurosci. Res., 2012, 90, 743–750.
44. T. L. Limke, S. R. Heidemann and W. D. Atchison, NeuroToxicology, 2004, 25, 741–760.
45. E. A. Belyaeva, S. M. Korotkov and N. E. Saris, J. Trace Elem. Med. Biol., 2011, 25, S63–S73.
46. A. Rasola and P. Bernardi, Cell Calcium, 2011, 50, 222–233.
47. L. Hertz and H. R. Zielke, Trends Neurosci., 2004, 27, 735–743.
48. M. Aschner, Neurotoxicology, 1996, 17, 663–669.
49. B. I. Juarez, H. Portillo-Salazar, R. Gonzalez-Amaro, P. Mandeville, J. R. Aguirre and M. E. Jimenez, Toxicology, 2005, 207, 223–229.
50. Z. Yin, D. Milatovic, J. L. Aschner, T. Syversen, J. B. Rocha, D. O. Souza, M. Sidoryk, J. Albrecht and M. Aschner, Brain Res., 2007, 1131, 1–10.
51. J. W. Allen, L. A. Mutkus and M. Aschner, Brain Res., 2001, 891, 148–157.
52. M. F. Hare and W. D. Atchison, J. Pharmacol. Exp. Ther., 1995, 272, 1016–1023.
53. M. F. Hare, K. M. McGinnis and W. D. Atchison, J. Pharmacol. Exp. Ther., 1993, 266, 1626–1635.
54. E. Fonfria, E. Rodriguez-Farre and C. Sunol, Neuropharmacology, 2001, 41, 819–833.
55. N. Basu, A. M. Scheuhammer, K. Rouvinen-Watt, R. D. Evans, V. L. Trudeau and L. H. Chan, Comp. Biochem. Physiol., Part C: Toxicol. Pharmacol., 2010, 151, 379–385.
56. T. K. Hensch, Annu. Rev. Neurosci., 2004, 27, 549–579.
57. Y. Yuan, NeuroToxicology, 2012, 33, 119–126.
58. M. S. Kang, J. Y. Jeong, J. H. Seo, H. J. Jeon, K. M. Jung, M. R. Chin, C. K. Moon, J. V. Bonventre, S. Y. Jung and D. K. Kim, Toxicol. Appl. Pharmacol., 2006, 216, 206–215.
59. J. Y. Chang, Neurosci. Lett., 2011, 496, 152–156.
60. G. W. Hwang, J. Y. Lee, K. Ryoke, F. Matsuyama, J. M. Kim, T. Takahashi and A. Naganuma, J. Toxicol. Sci., 2011, 36, 389–391.
61. Y. W. Chen, C. F. Huang, K. S. Tsai, R. S. Yang, C. C. Yen, C. Y. Yang, S. Y. Lin-Shiau and S. H. Liu, Diabetes, 2006, 55, 1614–1624.
62. N. R. Leslie, D. Bennett, Y. E. Lindsay, H. Stewart, A. Gray and C. P. Downes, EMBO J., 2003, 22, 5501–5510.
63. X. L. Chen, G. Dodd, S. Thomas, X. Zhang, M. A. Wasserman, B. H. Rovin and C. Kunsch, Am. J. Physiol.: Heart Circ. Physiol., 2006, 290, H1862–1870.
64. E. Fayard, L. A. Tintignac, A. Baudry and B. A. Hemmings, J. Cell Sci., 2005, 118, 5675–5678.
65. L. G. Costa, M. Aschner, A. Vitalone, T. Syversen and O. P. Soldin, Annu. Rev. Pharmacol., 2004, 44, 87–110.
66. D. Mergler, H. A. Anderson, L. H. Chan, K. R. Mahaffey, M. Murray, M. Sakamoto and A. H. Stern, Ambio, 2007, 36, 3–11.
67. M. A. Philbert, M. L. Billingsley and K. R. Reuhl, Toxicol. Pathol., 2000, 28, 43–53.
68. L. Baird and A. T. Dinkova-Kostova, Arch. Toxicol., 2011, 85, 241–272.
69. P. J. Sherratt, H. C. Huang, T. Nguyen and C. B. Pickett, Methods Enzymol., 2004, 378, 286–301.
70. J. Alam, C. Wicks, D. Stewart, P. Gong, C. Touchard, S. Otterbein, A. M. Choi, M. E. Burow and J. Tou, J. Biol. Chem., 2000, 275, 27694–27702.
71. R. Yu, S. Mandlekar, W. Lei, W. E. Fahl, T. H. Tan and A. N. Kong, J. Biol. Chem., 2000, 275, 2322–2327.
72. A. Kobayashi, M. I. Kang, Y. Watai, K. I. Tong, T. Shibata, K. Uchida and M. Yamamoto, Mol. Cell. Biol., 2006, 26, 221–229.
73. A. K. Jain and A. K. Jaiswal, J. Biol. Chem., 2006, 281, 12132–12142.
74. T. Toyama, D. Sumi, Y. Shinkai, A. Yasutake, K. Taguchi, K. I. Tong, M. Yamamoto and Y. Kumagai, Biochem. Biophys. Res. Commun., 2007, 363, 645–650.
75. N. Vanduyn, R. Settivari, G. Wong and R. Nass, Toxicol. Sci., 2010, 118, 613–624.
76. J. F. Robinson, P. T. Theunissen, D. A. van Dartel, J. L. Pennings, E. M. Faustman and A. H. Piersma, Reprod. Toxicol., 2011, 32, 180–188.
77. P. T. Theunissen, J. L. Pennings, J. F. Robinson, S. M. Claessen, J. C. Kleinjans and A. H. Piersma, Toxicol. Sci., 2011, 122, 437–447.
78. P. J. Khandelwal, A. M. Herman and C. E. Moussa, J. Neuroimmunol., 2011, 238, 1–11.
79. J. Y. Chang, Neurosci. Lett., 2007, 416, 217–220.
80. L. Dong, Z. Li, X. Bi and L. Ling, Wei Sheng Yan Jiu, 2001, 30, 7–9.
81. N. Onishchenko, N. Karpova, F. Sabri, E. Castren and S. Ceccatelli, J. Neurochem., 2008, 106, 1378–1387.
82. T. Kouzarides, Cell, 2007, 128, 693–705.
83. K. Martinowich, D. Hattori, H. Wu, S. Fouse, F. He, Y. Hu, G. Fan and Y. E. Sun, Science, 2003, 302, 890–893.
84. C. Gundacker, K. J. Wittmann, M. Kukuckova, G. Komarnicki, I. Hikkel and M. Gencik, Environ. Res., 2009, 109, 786–796.
85. J. M. Goodrich, Y. Wang, B. Gillespie, R. Werner, A. Franzblau and N. Basu, Toxicol. Appl. Pharmacol., 2011, 257, 301–308.
86. K. Schlawicke Engstrom, U. Stromberg, T. Lundh, I. Johansson, B. Vessby, G. Hallmans, S. Skerfving and K. Broberg, Environ. Health Perspect., 2008, 116, 734–739.
87. H. M. Custodio, K. Broberg, M. Wennberg, J. H. Jansson, B. Vessby, G. Hallmans, B. Stegmayr and S. Skerfving, Arch. Environ. Health, 2004, 59, 588–595.
88. S. E. Owens, M. L. Summar, K. K. Ryckman, J. L. Haines, S. Reiss, S. R. Summar and M. Aschner, NeuroToxicology, 2011, 32, 769–775.
89. J. W. Hunter, G. P. Mullen, J. R. McManus, J. M. Heatherly, A. Duke and J. B. Rand, Dis. Models Mech., 2010, 3, 366–376.
90. A. L. Jacob-Ferreira, R. Lacchini, R. F. Gerlach, C. J. Passos, F. Barbosa Jr. and J. E. Tanus-Santos, Sci. Total Environ., 2011, 409, 4242–4246.
91. K. C. de Marco, G. U. Braga and F. Barbosa Jr., J. Toxicol. Environ. Health, Part A, 2011, 74, 1323–1333.
92. K. C. de Marco, L. M. Antunes, J. E. Tanus-Santos and F. Barbosa Jr., Sci. Total Environ., 2011, 414, 708–712.
93. K. S. Engstrom, M. Wennberg, U. Stromberg, I. A. Bergdahl, G. Hallmans, J. H. Jansson, T. Lundh, M. Norberg, G. Rentschler, B. Vessby, S. Skerfving and K. Broberg, Environ. Health, 2011, 10, 33.
94. C. Edlund, L. Bjorkman, J. Ekstrand, G. Sandborgh-Englund and C. E. Nord, Clin. Infect. Dis., 1996, 22, 944–950.
95. C. A. Liebert, J. Wireman, T. Smith and A. O. Summers, Met. Ions Biol. Syst., 1997, 34, 441–460.
96. R. Pike, V. Lucas, P. Stapleton, M. S. Gilthorpe, G. Roberts, R. Rowbury, H. Richards, P. Mullany and M. Wilson, J. Antimicrob. Chemother., 2002, 49, 777–783.
97. D. Ready, J. Pratten, N. Mordan, E. Watts and M. Wilson, Int. J. Antimicrob. Agents, 2007, 30, 34–39.
98. D. Ready, F. Qureshi, R. Bedi, P. Mullany and M. Wilson, FEMS Microbiol. Lett., 2003, 223, 107–111.
99. J. Wireman, C. A. Liebert, T. Smith and A. O. Summers, Appl. Environ. Microbiol., 1997, 63, 4494–4503.
100. D. Skurnik, R. Ruimy, D. Ready, E. Ruppe, C. Bernede-Bauduin, F. Djossou, D. Guillemot, G. B. Pier and A. Andremont, J. Med. Microbiol., 2010, 59, 804–807.
101. J. F. Robinson, W. C. Griffith, X. Yu, S. Hong, E. Kim and E. M. Faustman, Reprod. Toxicol., 2010, 30, 284–291.
102. S. A. Hassan, E. A. Moussa and L. C. Abbott, J. Appl. Toxicol., 2011 DOI:10.1002/jat.1675.
103. B. Zimmer, S. Schildknecht, P. B. Kuegler, V. Tanavde, S. Kadereit and M. Leist, Toxicol. Sci., 2011, 121, 357–367.
104. M. Erecinska, S. Cherian and I. A. Silver, Prog. Neurobiol., 2004, 73, 397–445.
105. J. Stringari, A. K. Nunes, J. L. Franco, D. Bohrer, S. C. Garcia, A. L. Dafre, D. Milatovic, D. O. Souza, J. B. Rocha, M. Aschner and M. Farina, Toxicol. Appl. Pharmacol., 2008, 227, 147–154.
106. S. Cohen-Cory, A. H. Kidane, N. J. Shirkey and S. Marshak, Dev. Neurobiol., 2010, 70, 271–288.
107. T. Yorifuji, T. Tsuda, S. Inoue, S. Takao and M. Harada, Environ. Int., 2011, 37, 907–913.
108. B. Weiss, T. W. Clarkson and W. Simon, Environ. Health Perspect., 2002, 110(s5), 851–854.
109. J. J. Strain, P. W. Davidson, M. P. Bonham, E. M. Duffy, A. Stokes-Riner, S. W. Thurston, J. M. Wallace, P. J. Robson, C. F. Shamlaye, L. A. Georger, J. Sloane-Reeves, E. Cernichiari, R. L. Canfield, C. Cox, L. S. Huang, J. Janciuras, G. J. Myers and T. W. Clarkson, NeuroToxicology, 2008, 29, 776–782.
110. N. V. Ralston, C. R. Ralston, J. L. Blackwell 3rd and L. J. Raymond, NeuroToxicology, 2008, 29, 802–811.
111. S. Jayashankar, C. N. Glover, K. I. Folven, T. Brattelid, C. Hogstrand and A. K. Lundebye, Cell Biol. Toxicol., 2011, 27, 181–197.
112. F. Usuki, A. Yamashita and M. Fujimura, J. Biol. Chem., 2011, 286, 6641–6649.
113. S. Jayashankar, C. N. Glover, K. I. Folven, T. Brattelid, C. Hogstrand and A. K. Lundebye, Environ. Toxicol. Pharmacol., 2012, 33, 26–38.
114. D. C. Rice, NeuroToxicology, 2008, 29, 761–766.

---