S11 (II) Ecotoxicology of mercury

Methylmercury is a potent neurotoxin causing numerous neurological symptoms. Methylmercury can pass through the blood-brain barrier and the placenta and has higher toxicity during early development. Prenatal methylmercury exposure has been correlated with adverse effects in complex behaviors such as language skills and memory. A recent model system that is being used to study the neurological development of a complex behavior is song learning in the zebra finch (Taeniopygia guttata). In addition to humans, songbirds are one of only a few vertebrate groups that learn their vocalizations. The process of vocal learning is analogous in songbirds and humans, with similar timing, required inputs, and even similar genes involved. Because of these similarities, songbirds, and zebra finches in particular, have recently become a model system for studying speech learning in humans. In addition, mercury contamination in the field has been shown to correlate with aberrant song in three songbird species that learn their song but not in a species with innate song. We used a dosing experiment with environmentally relevant mercury exposures to investigate the effects of methylmercury exposure on birdsong in the lab. Pairs of zebra finches were exposed to methylmercury at a constant level in the diet and allowed to raise offspring normally. Male offspring exposed to methylmercury prenatally through the maternal diet and postnatally through juvenile diet produced significantly different songs than those raised on a control diet. Exposed males sing significantly less complex and lower pitched songs than control males. This study demonstrates a developmental effect of methylmercury on a complex behavior and opens the door to a study of the mechanisms through which mercury affects behavior in this model system.

Although it was recently documented that terrestrial songbirds are bioaccumulating mercury (Hg) in their blood and feathers at similar rates to those feeding directly on the aquatic ecosystem, there is little research about how this body burden of Hg could affect their reproductive success. From 2007 – 2010, we monitored reproductive success of Carolina wrens (Thryothorus ludovicianus) along two Hg-contaminated rivers in Virginia, USA. Using an information theoretic approach, we determined that Hg contamination explained a large amount of variation in nest success (the probability of at least one nestling surviving to fledge). Carolina wrens suffer a nearly 40% reduction in nest success on Hg-contaminated sites compared to reference uncontaminated sites. Although this approach is helpful in understanding how contamination affects the population, ecotoxicologists are often interested in how individual birds are affected by their particular Hg load. We, therefore, modeled nest success based on attending female blood Hg level and found that nest success decreases as blood Hg level increases; females suffer 20% reduction in nest survival at blood levels equal to 1.3 ppm (wet weight). This information is the first field-based study to link individual songbird mercury levels to reproductive success and is a critical step in determining the effects concentration gradient for Hg in insectivorous, terrestrial songbirds.

Methylmercury cycling in the Pacific Ocean has garnered significant attention in recent years, especially with regard to rising mercury emissions from Asia. Uncertainty exists over the extent to which mercury in biota may have resulted from increases in anthropogenic emissions over time. To address this, we assessed exposure to total mercury and, for a subset of samples, methylmercury (the bioaccumulated form of mercury) in an endangered seabird, the Black-footed Albatross (Phoebastria nigripes), using museum feathers spanning the past 130 years. We analyzed stable isotopes of nitrogen (d¹⁵N) and carbon (d¹³C) to control for temporal changes in trophic structure and diet. After both 1940 and 1990, significantly higher mean methylmercury concentrations and higher proportions of samples exhibiting above deleterious threshold levels (~40,000 ng/g) of methylmercury were observed relative to prior time points. We also found higher levels of (presumably curator-mediated) inorganic mercury in older specimens of albatross as well as two non-pelagic species lacking historical sources of bioavailable mercury exposure, which suggests that studies on bioaccumulation should measure methylmercury rather than total mercury when utilizing museum collections. After controlling for a significant trend in d¹³C over time, year remained a significant independent covariate with methylmercury exposure for the albatrosses. Changes in methylmercury levels were consistent with historical global and recent regional increases observed among published estimates and proxies of anthropogenic mercury emissions. Mercury toxicity may undermine current and future reproductive effort in the species.

Environmental contaminants are a potential threat to the health of Arctic ecosystems. However, little information is available on the toxic effects on wildlife health. We will provide an update on mercury (Hg) concentrations found in the brains of three marine mammal species in the Canadian Arctic and compare these data to threshold levels established from previous studies on mercury neurotoxicity. The concentration and proportion of total mercury (THg), soluble methylmercury (MeHg) and soluble inorganic mercury (iHg) vary greatly for beluga whales (Delphinapterus leucas), ringed seals (Phoca hispida), and polar bears (Ursus maritimus). The THg concentration was 15.65±9.96 mg/kg dry weight (dw) in the cerebellum of beluga whale (N=21), 0.7+/-0.47 mg/kg dw in ringed seal cerebellum (N=51) and 0.25±0.07 mg/kg dw in polar bear cerebellum (N=36). The proportion of soluble MeHg and iHg also varied between animal species; only 16% of THg was soluble MeHg in beluga whale cerebellum, yet 100% of THg was in the form of MeHg in polar bear cerebellum. The lowest observable effect level (LOEL) of Hg neurotoxicity has not been established for marine mammals. Subtle neurochemical changes were reported in captive mink exposed to low doses of MeHg at brain THg concentrations between 1.5±0.34 and 7.13±0.94 mg/kg dw. This is lower than THg concentrations linked to clinical effects in animal feeding trials; for example 8.15±5.73 mg/kg in wild mink brain, 12.3 and 7.3 mg/kg in male and female rat brain respectively and 16.3 mg/kg in cat brain, although to our knowledge, there is no established LOEL for THg concentration in brain tissue for neurotoxic effects. The conservative THg concentration associated with neurochemical changes is selected for assessing the risk associated with Hg exposure in marine mammals. The concentration of THg in polar bear was lower than this threshold; however, the concentration of THg in the cerebellum of beluga whales and in some ringed seals exceeded this threshold and these species therefore may be at risk of Hg neurotoxic effects such as subtle neurochemical changes. There are many uncertainties regarding the implications of Hg exposure to the health of marine mammals. We will address several data gaps concerning neurotoxic effects of mercury in both laboratory and wildlife animals and suggest directions for future research.

Arctic marine mammals accumulate some of the highest mercury (Hg) levels among all organisms. Despite the fact Hg is a proven neurotoxicant, little is known about its neurotoxic risk to Arctic marine mammals. Here we determined the distribution of Hg (total and organic) in ten brain regions of functional significance (frontal cortex, temporal cortex, occipital cortex, basal ganglia, brain stem, cerebellum, hippocampus, hypothalamus, thalamus, and pituitary) of harbor porpoise (Phocoena phocoena; n=12; Denmark), pilot whale (Globicephala melas; n=6; Faroe Islands), and polar bear (Ursus maritimus; n=37; East Greenland). To explore if brain Hg (total and/or organic) levels are of neurological concern to these marine mammals, we assessed neurochemical enzymes (glutamic acid decarboxylase (GAD), glutamine synthetase (GS)) and receptors (N-methyl-D-aspartic acid (NMDA), GABA(A) (benzodiazepine)) that play critical roles in learning and coordination among the different brain regions. Highest Hg levels were found in the pituitary (mean: 1.6ppm dry weight; range: 0.4–10ppm) of polar bears, pituitary (7.6ppm, 0.55–58 ppm) and frontal cortex (8.2ppm, 0.51–34ppm) of harbor porpoises, and three cortex regions (9.8ppm, 0.17–28ppm) of pilot whales. The brain levels of Hg found in harbor porpoise and pilot whale were comparable to levels that exceed subclinical thresholds in mammalian wildlife. In polar bear, despite relatively low levels, total Hg in only the pituitary was significantly correlated with age categories. Organic Hg comprised about 50% of the total Hg in the pituitary (compared to other brain regions where >80% of total was organic), and this may be of toxicological significance. When brain Hg was related to neurochemical biomarkers, some significant correlations were found. For example, in the harbor porpoise, pituitary Hg and GABA(A) receptor levels (n=8; r=0.923; p<0.01), and frontal cortex Hg and GAD activity (n=5; r=-0.934; p<0.01) significantly correlated. In the polar bear, organic Hg negatively associated with NMDA receptor levels in frontal cortex (n=17; r=-0.545; p<0.05). Overall, these results suggest that certain marine mammals (pilot whales, harbor porpoise) in the Arctic region are exposed to Hg at levels capable of causing subclinical neurological effect, but others (polar bears) may have developed unique schemes to limit brain exposures. Given that Hg is a potent neurotoxicant and that disruptions to neurochemistry precede structural and functional damage to the nervous system, it raises further questions about ecological and physiological impacts of Hg on Artic marine mammals health.

Elevated levels of mercury (Hg) have been detected in the brains of cetaceans. Hg levels in the brains of beluga whales in the western Canadian Arctic may exceed the lowest observable effect levels based on neurotoxicity observed in animal feeding trials involving cats, mink and rats. To elucidate the potential effects of Hg on the nervous system of belugas, brains were sampled from hunter-harvested whales in 2008 and 2010 on Hendrickson Island in the Beaufort Sea (N=35). A series of neurochemical and genetic biomarkers were selected to assess potential neurochemical disruption in the cerebellum. Six genes that encode N-methyl d-aspartate subunits 2b and 2c, gamma-aminobutyric acid (GABA) receptors 2 and 4, monoamineoxidase-A and the muscarinic acetylcholine receptor subunit 1, were sequenced and primers were designed within conserved nucleotide regions of the genes. Levels of cDNA for six target genes were measured by reverse transcription quantitative polymerase chain reaction in samples collected in 2008 and 2010 (N=33). Total monoamineoxidase activity was quantified by fluorescence assays in samples collected in 2008. In addition, hunters were asked to respond to specific questions regarding whale behaviour in 2010 (N=11) to assess whether belugas with elevated mercury were harpooned more quickly than animals with low exposure. Mean (± se) mercury concentration in cerebellar tissue was 17.6 ± 4.0 mg/kg dw (1.5 – 84.5 mg/kg dw, N=20) in 2008 and 9.1 ± 2.0 (1.7 – 29.6 mg/kg dw, N=15) in 2010. These concentrations exceeded the concentrations associated with neurochemical disruption in mink (between 1.5 ± 0.34 and 7.13 ± 0.94 mg/kg dw). GABA-2 expression decreased with increasing Hg exposure (χ2 value of 10.325 and p = 0.01600). MAO activity did not vary with Hg exposure (F_1,18= 0.05, p= 0.8213).There was no relationship between the ease of hunting success observed by the hunters and the brain Hg concentrations of the beluga. The results from this study suggest that gene expression for the GABA-2 receptor may vary with Hg exposure in harvested whales. Although THg reached 84.4 mg/kg in the cerebellum of belugas sampled in 2008 and 2010, this study did not reveal systematic effects of Hg on neurochemistry or behaviour.

Depending on their life history, organisms may be exposed both maternally and environmentally to contaminants, potentially placing them at greater risk of adverse effects than when exposed via either of the two pathways independently. Although adverse effects of maternally-derived contaminant exposure have been well documented, our study is one of the few to investigate the individual and combined effects of maternally- and environmentally- derived contaminants during critical periods of development. Amphibians, with their complex life cycles, provide an especially good model to study the context specificity of immediate and latent effects of one or more routes of contaminant exposure. We examined whether embryonic exposure to maternally-derived mercury (Hg) interacts additively or synergistically with dietary exposure to negatively influence larval development in American toads (Bufo americanus). We collected eggs from breeding pairs at reference and Hg-contaminated sites (20.6 ± 1.3 and 149.1 ± 18.0 ng/g total Hg, respectively) from our field site on the South River, VA (USA) and monitored performance, development, and survival of larvae fed three experimental Hg diets (total Hg: 0.01, 2.5, and 10 µg/g) with environmentally relevant methylmercury (MeHg) concentrations. The resulting average total Hg concentrations in metamorphs ranged from 50 ng/g (75% MeHg) to 1,800 ng/g (30% MeHg). We observed negative sublethal effects of both maternal and dietary Hg exposure routes on larval development, specifically growth, duration of metamorphic climax, and swimming performance. The effects of maternal and/or dietary Hg manifested differently, but maternal Hg exposure had a greater overall influence on offspring health than dietary exposure. However, the combination of sublethal effects of the two exposure routes acted synergistically, with lethal consequences; larvae exposed to maternal Hg and high dietary Hg experienced 50% greater mortality compared to larvae from reference mothers fed the control diet. Our study is one of the first to demonstrate that the latent effects of maternally transferred contaminants may be exacerbated by further exposure later in ontogeny, findings that may have important implications for both wildlife and human health, and further emphasizes the importance of investigating the effects of environmentally relevant routes of contaminant exposure over multiple early life stages.

Although it is well documented that exposure to methylmercury has adverse effects on a range of different organ systems and metabolic processes, the central nervous system is particularly sensitive to the detrimental effects of this toxin. Ongoing research has also demonstrated that developing organisms at embryonic and fetal stages are particularly sensitive to mercury exposure—with lower doses having a far more profound effect than similar doses at adult stages. However, the molecular and biochemical mechanisms that result in neural defects remain unclear. This study seeks to develop a Xenopus model system and utilize several approaches to understanding the pathways involved in methylmercury’s neurotoxicity. Interestingly, the effects of embryonic mercury exposure are not linearly dose-dependent and instead seem to be phenotypically and genetically expressed in an all-or-nothing manner around a very tight threshold. In addition, density of the embryos has a profound effect on the observed phenotype. Bioaccumulation of methylmercury in treatment groups is quantified using a Milestone Direct Mercury Analyzer and reveals that accumulations for individual embryos in less dense plates are significantly higher. To investigate the molecular mechanisms of toxicity, mRNA levels and expression patterns were assessed using in situ hybridization to visualize endogenous changes in the expression of calcium channel and neurotransmitter phenotype specification genes in mercury-exposed embryos. Again, here an all-or-nothing signal is seen, with embryos displaying either no expression when they are developmentally arrested, or normal gene expression when treated embryos seem morphologically normal. The exception is the rate-limiting enzyme for GABA synthesis, glutamic acid decarboxylase (GAD), which shows a difference between control and treatment embryos, with the signal being less prominent in the exposed embryos, and Delta, a protein involved in Notch signaling, which is detected in embryos with severe morphological abnormalities. To further the understanding of global gene expression changes, transcriptome analysis will reveal quantifiable information about differential mRNA levels. The ability of some embryos in the same treatment group to normally develop while others are completely retarded may be indicative of endogenous resistance mechanisms.