whales

Accepted Manuscript Title: Differences in the responses of three plasma selenium-containing proteins in relation to methylmercury-exposure through consumption of fish/whales Author: Ping Han Ser Sanae Omi Hana Shimizu-Furusawa Akira Yasutake Mineshi Sakamoto Noriyuki Hachiya Shoko Konishi Masaaki Nakamura Chiho Watanabe PII: DOI: Reference:

S0378-4274(16)33335-5 http://dx.doi.org/doi:10.1016/j.toxlet.2016.12.001 TOXLET 9654

To appear in:

Toxicology Letters

Received date: Revised date: Accepted date:

9-5-2016 18-11-2016 3-12-2016

Please cite this article as: Ser, Ping Han, Omi, Sanae, Shimizu-Furusawa, Hana, Yasutake, Akira, Sakamoto, Mineshi, Hachiya, Noriyuki, Konishi, Shoko, Nakamura, Masaaki, Watanabe, Chiho, Differences in the responses of three plasma seleniumcontaining proteins in relation to methylmercury-exposure through consumption of fish/whales.Toxicology Letters http://dx.doi.org/10.1016/j.toxlet.2016.12.001 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Differences in the responses of three plasma selenium-containing proteins in relation to methylmercury-exposure through consumption of fish/whales

Ping Han Sera, Sanae Omia, Hana Shimizu-Furusawaa, Akira Yasutakeb, Mineshi Sakamotob, Noriyuki Hachiyab, Shoko Konishia, Masaaki Nakamurab, Chiho Watanabea a. Department of Human Ecology, School of International Health, Graduate School of Medicine, The University of Tokyo,Hongo 7-3-1, Bunkyo-ku, Tokyo, Japan 113-0033

b. National Institute for Minamata Disease Hama, Minamata city, Kumamoto prefecture, Japan 867-0008 TOX LETTER

Highlights

 Se-containing proteins in plasma of methyl Hg-exposed population were analyzed.  Two plasma selenoproteins showed differences in the response to Hg exposure.  Blood Hg but not plasma Se positively correlated with consumption of fish/whales.  Increase in plasma Se might be associated with an increase in selenoprotein P.  Increased demand for Se in Hg-exposed population was suggested.

Abstract Putative protective effects of selenium (Se) against methylmercury (MeHg) toxicity have been examined but no conclusion has been reached. We recently reported the lack of serious neurological symptoms in a Japanese fish-eating population with high intakes of MeHg and suggested a potential protective role for Se. Here, relationships between levels of Hg and Se in the blood and plasma samples, with a quantitative evaluation of Se-containing proteins, obtained from this population were examined. While levels of the whole-blood Hg (WB-Hg) and plasma Se (P-Se) showed a positive correlation, stratified analysis revealed that they

2

correlated only in samples with higher (greater than the median) levels of MeHg. A food frequency questionnaire showed that consumption of fish/whales correlated with WB-Hg, but not with P-Se, suggesting that the positive correlation between WB-Hg and P-Se might not be the result of co-intake of these elements from seafood. Speciation of plasma Se revealed the differences in the responses of two plasma selenoproteins, glutathione peroxidase (GPx) and selenoprotein P (SePP), in relation to Hg exposure. In the high-Hg group, SePP showed a positive correlation with WB-Hg, but GPx did not. In the low-Hg group, neither SePP nor GPx showed any correlation with WB-Hg. These observations suggest that the increase in P-Se in the high-Hg group might be associated with an increase in SePP, which may, in turn, suggest an increased demand for one or more selenoproteins in various organs, for which SePP supplies the element.

Key words methylmercury, selenium, selenoprotein P, glutathione peroxidase, fish consumption

3

1. Introduction Methylmercury (MeHg) becomes biomagnified in the marine food chain, and the potential health risks caused by consumption of certain types of fish and marine mammals, including whales, have become a worldwide public health concern. Although most studies on MeHg and fish or marine mammal consumption have focused on prenatal exposure and health impacts after birth (Choi et al. 2007, Davidson et al. 2011), the adult population can also be affected, depending on exposure levels. We recently reported on the neurological outcomes of high MeHg exposure in a population from the coastal city of Taiji, Japan. This population is exposed to MeHg through consumption of fish and whales, the latter being hunted by traditional practices and containing high concentrations of MeHg (Endo and Haraguchi 2010). Despite the high level of MeHg exposure, neurological symptoms did not correlate with Hg exposure, suggesting the existence of one or more protective factors against MeHg toxicity (Nakamura et al. 2014). Among many factors that could attenuate MeHg neurotoxicity, protective effects of selenium (Se) against MeHg toxicity have been demonstrated in various experimental settings in vivo (Ralston et al. 2008, Sakamoto et al. 2013) and in vitro (Farina et al. 2011). The role of Se in MeHg toxicity in human populations has been suggested by many researchers, but no firm conclusion has yet been reached (Watanabe 2002). While many studies conducted on human populations with various level of Hg exposure have shown positive correlations between Hg and Se in blood/plasma (Grandjean et al. 1992, Svensson et al. 1992, Bensryd et al. 1994, Hagmar et al. 1998, Bates et al. 2006, Lemire et al. 2006, Valera et al. 2009, Lemire et al. 2010), in a cohort study in the Faroe islands, focusing on in utero exposure, Se showed no significant modifying effect on the neurobehavioral toxicity of MeHg (Choi et al. 2007). In the Taiji population (Nakamura et al. 2014), a positive correlation between blood

4

Hg and Se in whole blood was found, although the mechanism and any toxicological significance remain unclear. The positive correlation could be due to co-intake of Hg and Se through seafood, because seafood is generally relatively rich in both elements. Alternatively, it could be due to a kinetic or dynamic response of the physiologically available Se pool against ingested Hg, as suggested by experimental studies (Björkman et al. 1995, Ralston et al. 2008, Sakamoto et al. 2013). In this study, we focused on the chemical speciation of Se in plasma samples collected from the Taiji population. In human plasma, Se is known to associate with three proteins: selenoprotein P (SePP), (extracellular) glutathione peroxidase (GPx), and albumin (Alb), which can be detected by HPLC-ICP-MS (Koyama et al. 1999). Of these, SePP and GPx are so-called selenoproteins that incorporate Se as selenocysteine, a specifically encoded unusual amino acid, whereas Alb contains Se as selenomethionine, in positions of methionine residues, which is considered to be a “non-specific” form of Se, unrelated to specific selenium metabolic processes (Burk et al. 2001). It was expected that an examination of the quantitative changes of these three proteins would help in understanding the mechanism and relevance of the increase in plasma Se observed in MeHg-exposed populations. Previous studies that have examined the relationship between Hg and Se-containing proteins obtained mixed results. A study on a Latvian community with a moderate MeHg exposure (Hg in red blood cells (RBCs) in the range 0.1-43.7 µg/L) reported significant positive correlations between Hg in RBCs and plasma GPx, as well as SePP (Hagmar et al. 1998). Another study, conducted in an Amazonian population (blood Hg level was 47.8±36.3 µg/L), reported a significant negative association between blood Hg levels and GPx activity (Grotto et al. 2010). Elevated concentrations of GPx and SePP in a population of Hg miners have been reported (Chen et al. 2006), suggesting that a positive correlation may be observed even when the co-intake of Hg and Se from seafood consumption is unlikely

5

(although in this case the chemical form of Hg was metallic, not methyl, mercury). In this study, we examined the quantitative relationship between Hg exposure and plasma Se-containing proteins among a Japanese population with a high intake of MeHg from the consumption of whales and/or fish. We found differences in the behavior of the three Se-containing proteins, suggesting that the reported positive correlation between Hg and Se may be the result of a physiological response of Se to MeHg, rather than simple co-intake of the two elements.

2. Methods 2.1 Study site Our study was conducted in the town of Taiji, located in the Wakayama Prefecture of Japan (Nakamura et al. 2014). The town has a population of approximately 3,400 (Endo and Haraguchi 2010). Traditional whale hunting is practiced in the winter season, and the residents consume baleen and toothed whales, as well as fish species, such as tuna and swordfish. Toothed whale red meat and internal organs from Taiji have been reported to have total Hg levels exceeding Japanese safety standards (0.4 µg/wet g) for marine organisms (Endo and Haraguchi 2010). The median hair Hg contents of the study population were reported previously to be 18.7 and 15.1 µg/g for males (n = 114) and females (n =77), respectively (Nakamura et al. 2014), far exceeding the corresponding values in the general Japanese population: 2.46 and 1.63 µg/g for males and females, respectively (Yasutake et al. 2004).

2.2 Participants and biological samples We recruited volunteers from the town of Taiji. Recruitment was carried out in July 2010

6

during a health check-up, to which residents aged 30 and above were invited by local authorities. No specific criterion other than age was set for the recruitment, and 111 people aged from 33 to 82 years (48 males, 63 females) participated. Hair, blood, and plasma samples were collected to determine Hg and Se concentrations. Hair samples were collected from 99 participants, by cutting 10 strands of hair ~3 cm from the scalp, and storing them; 12 hair samples were not collected because the participants were absent during the collection procedure. Blood (2 mL) was collected via venipuncture: 1 mL was stored as whole blood and the remaining volume was centrifuged to obtain plasma. All samples were transported on dry ice and stored at -80°C until analysis. The protocols for this study were approved by the Research Ethics Committee of the Graduate School of Medicine, The University of Tokyo.

2.3 Analysis of Hg and Se Total Hg was measured using the oxygen combustion-gold amalgamation method (Ohkawa et al. 1977, Yasutake et al. 2004). Hair samples were washed with detergent in an ultrasonic washer, soaked in acetone twice to remove water, and left to dry on filter paper. An aliquot (2 mg) of the sample was then dissolved in 0.5 mL sodium hydroxide by heating at 60°C for 30 min. Next, 50 µL of this sample solution was subjected to total Hg analysis with the oxygen combustion-gold amalgamation methods using an atomic absorption mercury detector (MD-A, Nippon Instruments Co., Ltd., Osaka, Japan). Total Hg in 50 µL blood samples was also measured with the same instrument, and with no pretreatment. Se was measured with inductively coupled plasma mass spectrometry (ICP-MS) with a collision cell (Agilent 7500ce, Agilent Technologies, Santa Clara, CA, USA), as described previously (Parajuli et al. 2012). Samples (100 mg) were wet-digested with nitric acid. Filtration was performed on the samples using 0.45 µm pore size membranes to remove

7

contaminants. For quality control, the following reference materials were used: NIES CRM No. 13 (National Institute of Environment Science, Tsukuba, Japan) for Hg, and Trace Elements Whole Blood L-2 and Trace Elements Serum L-2 (SeronormTM) for Se in the whole blood and plasma. The results obtained fell within the certified ranges for both Hg and Se.

2.4 Speciation of plasma Se Chemical speciation of plasma Se was conducted with high-performance liquid chromatography coupled with ICP-MS (HPLC-ICP-MS), as described elsewhere (Koyama et al. 1999). Samples (50 µL) were passed through a heparin affinity column (AFpak AHR-894, Showa Denko, Tokyo, Japan) and a size exclusion column (Asahipak GS-520 HQ, Showa Denko, Tokyo, Japan) in tandem. The eluent was monitored for Se, and three major peaks were observed, as described previously, considered to be the Se peaks associated with Alb, GPx, and SePP. The area under each peak was calculated, and the proportion (%) of each peak against the sum of the three peaks was determined. Then, by multiplying the concentration of plasma Se (P-Se), the proportion of each peak was converted into the concentration of Se that was associated with each peak, referred to as Alb-Se, GPx-Se, and SePP-Se, respectively, in this paper. Speciation of P-Se was performed on the plasma samples with the highest and lowest Hg levels (n = 19 for each group).

2.5 Food consumption questionnaire Frequency of consumption in the last 3 months was assessed using a questionnaire for the following seafood items: tuna and/or swordfish (consumption levels: “low” = less than once per month, “high” = more than once per week, and “medium” = in between), skipjack tuna (using the same frequency categories as tuna and/or swordfish), baleen whale meat and/or

8

organs (yes = consumed at least once, no), and toothed whale meat and/or organs (yes / no). Although the original questionnaire had more categories for frequency of consumption, they were collapsed into two or three categories, depending on the distribution of the responses.

2.6 Statistical analyses Concentrations of Hg and Se were log-transformed for statistical tests. Relationships between seafood consumption and Hg and Se were tested by ANOVA, where frequency data were used as the ordinal scale. Statistical analyses were performed using the ‘R’ software (ver. 2.15.1; Team 2011) and the JMP software (SAS Institute, USA). P values < 0.05 were considered to indicate statistical significance.

3. Results 3.1 Hg and Se levels in blood samples Table 1 shows Hg and Se concentrations in hair, whole blood, and plasma samples, together with the age range of the participants. Hg levels in hair and whole blood (WB-Hg) samples indicated elevated Hg concentrations in the participants. Males had significantly higher Hg (in hair and blood), but not Se, levels than females. While the individual variation in Hg levels was as high as 70-fold, that of Se levels was only 2- and 3-fold in plasma and whole blood, respectively. As expected, highly positive correlations were found between hair Hg and WB-Hg (r = 0.96, p < 0.001) and between WB-Se and P-Se concentrations (r = 0.46, p < 0.001), consistent with our previous results (Nakamura et al. 2014). Also, WB-Hg showed significant correlations with WB-Se (r = 0.37, p < 0.001) and with P-Se (r = 0.41, p < 0.01). Because the scatter plots of WB-Hg versus WB-Se and P-Se suggested that the relationships might be

9

non-linear, the correlations were re-evaluated in the two groups, stratified by the level of WB-Hg: higher or lower than the median Hg of 38 µg/L. WB-Hg correlated significantly with both WB-Se (p = 0.31, p < 0.05) and P-Se (p = 0.40, p < 0.01) in the high-Hg group (n = 55), while it showed no significant correlation with these Se indices in the low-Hg group (n = 56; Fig. 1).

3.2 Relationships between levels of Hg, Se, and Se-containing proteins Among the 38 samples examined for Se-containing proteins (19 in the low- and 19 in the high-Hg groups), P-Se, WB-Se, and all the Se-containing proteins were significantly higher in the high- than the low-Hg group (Table 2). Considering the non-linear relationship between P-Se and blood Hg, the correlations between each Se-containing protein and blood Hg were examined in the high- and low-Hg groups separately. In the high-Hg group, SePP-Se showed a significant and positive regression with blood Hg (β = 0.197, p = 0.012). Alb-Se showed a positive but non-significant (β = 0.157, p = 0.108) regression, and GPx-Se showed no significant regression (Fig. 2a-c). Thus, the three proteins in the high-Hg group showed distinct differences regarding the regression with Hg exposure. In contrast, none of the three proteins correlated with blood Hg in the low-Hg group.

3.3 Relationship between consumption frequency of marine species and Hg and Se levels In total, 90 participants answered all questions and were used for analyses. Of the respondents, 64% and 56% had consumed tooth whale or baleen whale (meat or organ), respectively, at least once in the last 3 months. For all the seafood items examined, significantly higher WB-Hg values were associated with consumption of whales and a higher frequency of tuna and/or swordfish and skipjack consumption. In contrast, no difference in seafood consumption was reflected in P-Se. Although a

10

significant difference in P-Se was found among the three levels of tuna and/or swordfish consumption, it showed no dose-response relationship (Table 3).

4. Discussion Our results showed positive correlations between WB-Hg and P-Se and WB-Se, confirming the results of our previous study in this same population (Nakamura et al. 2014) with a larger sample size. Our observations also agreed with the previous study in that the molar ratio (WB-Hg/WB-Se) was far less than unity (< 1), and correlated positively with WB-Hg. The present results also add some new observations. Specifically, while WB-Hg and P-Se correlated significantly with each other, analyses stratified by WB-Hg level showed that these correlations were only found in the group with higher WB-Hg levels, suggesting the presence of a threshold of Hg exposure above which an increase in plasma selenium was observed. Speciation of P-Se revealed that the relationship between the level of Hg exposure and that of Se-containing protein differed across the three proteins. To our knowledge, such differences in responses, combined with the putative existence of a threshold for the responses, have not been reported before. The significance and implications of these observations with regard to potentially alleviating MeHg toxicity by Se, as suggested for this population (Nakamura et al. 2014), are discussed below. High hair Hg and WB-Hg concentrations suggested that the study population had high exposure to mercury, which was presumably due to seafood consumption. Hair Hg levels were five times higher than that of the general Japanese population (Yasutake et al. 2004), and the WB-Hg level was one of the highest reported among communities consuming marine mammals (Grandjean et al. 1997, Choi et al. 2008, Valera et al. 2009, Ayotte et al. 2011). The P-Se level was also higher than in some other Japanese populations living in

11

coastal areas (Imai et al. 1990, Deguchi and Ogata 1991). The results of the seafood consumption survey suggested that higher WB-Hg was associated with higher consumption of fish/whale, but neither WB-Se nor P-Se showed any association. Thus, any relationship between seafood consumption and blood/plasma Se levels is not straightforward in this population. Previous studies also reported mixed results regarding the relationship between Se levels and fish consumption (Grandjean et al. 1992, Svensson et al. 1992, Bensryd et al. 1994, Hagmar et al. 1998, Barany et al. 2003, Bates et al. 2006, Lemire et al. 2006). This might be taken for granted because fish/marine mammals are the major, and virtually the only, sources of MeHg, whereas dietary sources of Se are not limited to fish/marine mammals. Positive correlations between blood Hg and Se levels have been documented repeatedly in fish-eating populations. This has been considered to be the result of co-intake of the two elements from seafood or ascribed to the formation of a Hg-Se complex (Grotto et al. 2010). The stratified analysis in the present study, however, showed that a positive relationship was found only for the groups with higher WB-Hg (higher than the median of 1.16 µmole/L), supporting the notion that the correlation is not explained simply by co-intake and/or co-accumulation. Significant correlations between blood Hg and Se levels have also been reported in populations with limited fish consumption (Osman et al. 1998, Lindberg et al. 2004), suggesting that the correlation can occur without co-intake from seafood consumption. Speciation of the Se-containing peaks in the plasma revealed that SePP was the most prevalent, followed by albumin and GPx, confirming previous results (Koyama et al. 1999, Burk and Hill 2005). At high Hg exposure, concentrations of the two selenoproteins, GPx and SePP, showed a distinct pattern in terms of the relationship with Hg exposure. No such contrast was observed in the low Hg exposure group, supporting the existence of a threshold for Hg exposure, above which the metabolism (e.g., synthesis and/or release) of Se or

12

selenoproteins could change. Albumin showed no significant association with Hg exposure, although it showed a marginal and positive correlation at high Hg levels. Because albumin is not a selenoprotein, lacking selenocysteine residues, the Alb-Se monitored in this study would reflect changes, if any, in plasma albumin concentrations per se. Why the two selenoproteins showed different responses remains unknown. Direct competition between SePP and GPx for available Se seems unlikely because these proteins are secreted primarily by different organs: the liver and kidney, respectively (Reeves and Hoffman 2009). Moreover, the synthesis of GPx depends, at least partly, on the supply of Se to the kidney by the Se-transporting protein, SePP (Schomburg and Schweizer 2009). Additionally, this study population was relatively well nourished in terms of Se, as indicated by the high P-Se level, as well as the relative abundance of Se over Hg, as demonstrated by the blood Hg/Se ratio in this population (Nakamura et al. 2014). Previous studies reported differences in the relationship between different selenoproteins and MeHg exposure. In a Latvian population, fish consumption was associated with an increase in selenoprotein levels, where the increase in SePP was greater than that in GPx (Hagmar et al. 1998). In a Swedish population with various levels of fish intake, plasma Se levels increased with higher fish consumption, whereas the proportion of SePP remained constant, while that of GPx decreased (Huang et al. 1995). In an Amazonian population, an inverse relationship between GPx activity and blood Hg levels was reported (Grotto et al. 2010). Together, these reports, including the present study, are consistent in that plasma SePP levels may increase with higher Hg exposure (or higher fish consumption), but GPx may not. Because it is generally accepted that SePP serves as a Se-transporting protein (Burk and Hill 2005, Reeves and Hoffman 2009, Schomburg and Schweizer 2009), the specific increase of SePP at higher Hg exposure might reflect elevated physiological demand for Se transport by the organs of Hg-exposed individuals. The brain is one of the organs that

13

requires SePP for the transportation of Se (Schomburg and Schweizer 2009). Alternatively, the increase in SePP might be associated with its known anti-oxidant properties (Burk and Hill 2005, Reeves and Hoffman 2009). Koyama et al. (Koyama et al. 2009) found in a nested case-control study that patients experiencing stroke attack, an oxidative stress-related condition, had significantly lower (by ~14%) serum SePP levels than controls, while their GPx and albumin concentrations were similar to the controls. In the autopsy brains of two Se-depleted children on total parenteral nutrition, (cellular) GPx immunostaining showed a very weak signal, and the brains showed considerable pathological changes, illustrating the possibility of brain lesions in a severely Se-depleted state (Hirato et al. 2003). The authors noted that the brain lesions observed, at least partially, resembled the lesions in methylmercury poisoning (Hirato et al. 2003). Recently, we reported the absence of MeHg poisoning in this population, and suggested that the lack of neurological symptoms may be associated with adequate selenium intake (Nakamura et al. 2014). While we cannot show a direct connection between the absence of symptoms and the increased SePP level, the present results suggest that the increase in SePP might be an active physiological response in the metabolic handling of Se against Hg exposure rather than a purely passive change due to the consumption of seafood. Two major limitations of the study should be addressed. First, the speciation study involved a relatively small number of samples. The results obtained, however, showed distinct differences between GPx and SePP as well as a weak, but positive, correlation between SePP and Hg at higher Hg levels, despite the small sample size, warranting further examination of the responses of various selenoproteins in MeHg exposure. Second, in the speciation analysis, we monitored the amount of Se associated with each peak. Monitoring of the amount of selenoproteins would be also informative in exploring the biological response

14

to Hg exposure.

5. Conclusions In the plasma of a fish-eating population in Japan, in whom a relatively high intake of MeHg was observed, blood Hg and plasma Se showed a positive correlation, but only in the samples of the high (greater than the median) Hg group. Seafood consumption was also correlated with WB-Hg levels, but not with plasma Se levels, suggesting that the positive correlation is not the result of co-intake of these elements from seafood. Speciation of the plasma Se revealed differences in the responses of the two selenoproteins in relation to Hg exposure: higher Hg exposure caused an increase in SePP, whereas GPx was relatively insensitive to MeHg exposure. This may suggest that the increase in SePP is specific to this protein, indicating increased demand for one or more selenoproteins, for which SePP supplies the element, in MeHg exposed individuals.

6. Acknowledgements We thank the town of Taiji for facilitating this research and the participants who provided us with biological samples and information. This study was partly supported by a Grant-in-aid from the Ministry of Education, Culture, Science, and Technology, Japan (grant no. 23659325).

15

The English in this document has been checked by at least two professional editors, both native speakers of English. For a certificate, please see: http://www.textcheck.com/certificate/PiaYUA

16

References Ayotte, P., A. Carrier, N. Ouellet, V. Boiteau, B. Abdous, E. A. L. Sidi, M. L. Chateau-Degat and E. Dewailly (2011). "Relation between Methylmercury Exposure and Plasma Paraoxonase Activity in Inuit Adults from Nunavik. ." Environ Health Persp 119: 1077-1083. Barany, E., I. A. Bergdahl, L. E. Bratteby, T. Lundh, G. Samuelson, S. Skerfving and A. Oskarsson (2003. ). "Mercury and selenium in whole blood and serum in relation to fish consumption and amalgam fillings in adolescents." J Trace Elem Med Bio 17: 165-170. Bates, C. J., A. Prentice, M. C. Birch, H. T. Delves and K. A. Sinclair (2006). "Blood indices of selenium and mercury, and their correlations with fish intake, in young people living in britain. ." Brit J Nutr 96(523-531). Bensryd, I., L. Rylander, B. Hogstedt, P. Aprea, I. Bratt, C. Fahraeus, A. Holmen, A. Karlsson, A. Nilsson, B. L. Svensson, A. Schutz, Y. Thomassen and S. Skerfving (1994). "Effect of acid precipitation on retention and excretion of elements in man. ." Sci Total Environ. 145: 81-102. Björkman, L., K. Mottet, M. Nylander, M. Vahter, B. Lind and L. Friberg (1995). "Selenium concentrations in brain after exposure to methylmercury: relations between the inorganic mercury fraction and selenium." Arch Toxicol 69: 228-234. Burk, R. F. and K. E. Hill (2005). "Selenoprotein P: an extracellular protein with unique physical characteristics and a role in selenium homeostasis." Annu Rev Nutr 25: 215-235. Burk, R. F., K. E. Hill and A. K. Motley (2001). "Plasma selenium in specific and non-specific forms." Biofactors 14(1-4): 107-114. Chen, C., H. Yu, J. Zhao, B. Li, L. Qu, S. Liu, P. Zhang and Z. Chai (2006). "The roles of serum selenium and selenoproteins on mercury toxicity in environmental and occupational exposure." Environ Health Perspect 114(2): 297-301. Choi, A., E. Budtz-Jørgensen, P. Jørgensen, U. Steuerwald, F. Debes, P. Weihe and P. Grandjean (2007).

"Selenium

as

a potential

protective

factor

against mercury

developmental

neurotoxicity." Environmental Research 107: 45-52. Choi, A. L., E. Budtz-Jorgensen, P. J. Jorgensen, U. Steuerwald, F. Debes, P. Weihe and P. Grandjean (2008). "Selenium as a potential protective factor against mercury developmental neurotoxicity. ." Environ Res. 107: 45-52. Davidson, P., D. Cory-Slechta, S. Thurston, L. Huang, C. Shamlaye, D. Gunzler, G. Watson, E. van Wijngaarden, G. Zareba, J. LKlein, T. Clarkson, J. Strain and G. Myers (2011). "Fish consumption and prenatal methylmercury exposure: Cognitive and behavioral outcomes in the main cohort at 17 years from the Seychelles child development study." Neurotoxicology 32: 711-717. Deguchi, Y. and A. Ogata (1991). "Relationship between serum selenium concentration and atherogenic index in japanese adults. ." Tohoku J Exp Med. 165: 247-251. Endo, T. and K. Haraguchi (2010). "High mercury levels in hair samples from residents of taiji, a

17

japanese whaling town. ." Mar Pollut Bull. 60: 743-747. Farina, M., J. Rocha and M. Aschner (2011). "Mechanisms of methylmercury-induced neurotoxicity: evidence from experimental studies." Life Science 89: 555-563. Grandjean, P., P. Weihe, P. J. Jorgensen, T. Clarkson, E. Cernichiari and T. Videro (1992). " Impact of maternal seafood diet on fetal exposure to mercury, selenium, and lead. ." Arch Environ Health. 47: 185-195. Grandjean, P., P. Weihe, R. F. White, F. Debes, S. Araki, K. Yokoyama, K. Murata, N. Sorensen, R. Dahl and P. J. Jorgensen (1997). "Cognitive deficit in 7-year-old children with prenatal exposure to methylmercury. ." Neurotoxicol Teratol. 19 417-428. Grotto, D., J. Valentini, M. Fillion, C. J. Passos, S. C. Garcia, D. Mergler and F. Barbosa, Jr. (2010). "Mercury exposure and oxidative stress in communities of the Brazilian Amazon." Sci Total Environ 408(4): 806-811. Hagmar, L., M. Persson-Moschos, B. Akesson and A. Schutz (1998). "Plasma levels of selenium, selenoprotein P and glutathione peroxidase and their correlations to fish intake and serum levels of thyrotropin and thyroid hormones: a study on Latvian fish consumers." Eur J Clin Nutr 52(11): 796-800. Hirato, J., Y. Nakazato, H. Koyama, A. Yamada, N. Suzuki, M. Kuroiwa, A. Takahashi, S. Matsuyama and K. Asayama (2003). "Encephalopathy in megacystis-microcolon-intestinal hypoperistalsis syndrome patients on long-term total parenteral nutrition possibly due to selenium deficiency." Acta Neuropathol 106: 234-242. Huang, W., B. Akesson, B. G. Svensson, A. Schutz, R. F. Burk and S. Skerfving (1995). "Selenoprotein-p and glutathione-peroxidase (ec-1.11.1.9) in plasma as indexes of selenium status in relation to the intake of fish. ." Brit J Nutr. 73: 455-461. Imai, H., T. Suzuki, H. Kashiwazaki, T. Takemoto, T. Izumi and K. Moji (1990). "Dietary habit and selenium concentrations in erythrocyte and serum in a group of middle-aged and elderly japanese. ." Nutr Res. 10: 1205-1214. Koyama, H., R. Abdulah, T. Ohkubo, Y. Imai, H. Satoh and K. Nagai (2009). "Selenoprotein P: possible new predictor of increased risk for cerebrovascular events. ." Nutrition Research 29: 94-99. Koyama, H., K. Omura, A. Ejima, Y. Kasanuma, C. Watanabe and H. Satoh (1999). "Separation of selenium-containing proteins in human and mouse plasma using tandem high-performance liquid chromatography columns coupled with inductively coupled plasma-mass spectrometry." Anal Biochem. 267: 84-91. Lemire, M., M. Fillion, B. Frenette, A. Mayer, A. Philibert, C. J. S. Passos, J. R. D. Guimaraes, F. Barbosa and D. Mergler (2010). "Selenium and Mercury in the Brazilian Amazon: Opposing Influences on Age-Related Cataracts. ." Environ Health Persp 118: 1584-1589. Lemire, M., D. Mergler, M. Fillion, C. J. Sousa Passos, J. R. D. Guimaraes, R. Davidson and M. Lucotte (2006). "Elevated blood selenium levels in the Brazilian Amazon. ." Science of the Total

18

Environment 366: 101-111. Lindberg, A., K. A. Bjornberg, M. Vahter and M. Berglund (2004). "Exposure to methylmercury in non-fish-eating people in sweden. ." Environ Res. 96: 28-33. Nakamura, M., N. Hachiya, K. Murata, I. Nakanishi, T. Kondo, A. Yasutake, K. Miyamoto, P. Ser, S. Omi, H. Furusawa, C. Watanabe, F. Usuki and M. Sakamoto (2014). "Methylmercury exposure and neurological outcomes in Taiji residents accustomed to consuming whale meat." Environmental International 68: 25-32. Ohkawa, T., H. Uenoyama, K. Tanida and T. Ohmae (1977). "Ultra trace mercury analysis by dry thermal decomposition in alumina porcelain tube. ." The Journal of Hygienic Chemistry. 23: 13-22. Osman, K., A. Schutz, B. Akesson, A. Maciag and M. Vahter (1998). "Interactions between essential and toxic elements in lead exposed children in Katowice, Poland." Clin Biochem 31(8): 657-665. Parajuli, R. P., T. Fujiwara, M. Umezaki, H. Furusawa, P. H. Ser and C. Watanabe (2012). "Cord Blood Levels of Toxic and Essential Trace Elements and Their Determinants in the Terai Region of Nepal: A Birth Cohort Study." Biol Trace Elem Res 147: 75-83. Ralston, N., C. Ralston, J. r. Blackwell and L. Raymond (2008 ). "Dietary and tissue selenium in relation to methylmercury toxicity. ." Neurotoxicology 29: 802-811. Reeves, M. and P. Hoffman (2009). "The human selenoproteome: recent insights into functions and regulation." Cellular and Molecular Life Science 66: 245702478. Sakamoto, M., A. Yasutake, A. Kakita, M. Ryufuku, H. Chan, M. Yamamoto, S. Oumi, S. Kobayashi and C. Watanabe (2013). "Selenomethionine protects against neuronal degeneration by methylmercury in the developing rat cerebrum." environmental Science and Technology 47: 2862-2868. Schomburg, L. and U. Schweizer (2009). "Hirarchical regulation of selenoprotein expression and sex-specific effecs of selenium, ." Biochimica et Biophysica Acta 1790: 1453-1462. Svensson, B. G., A. Schutz, A. Nilsson, I. Akesson, B. Akesson and S. Skerfving (1992). "Fish as a Source of Exposure to Mercury and Selenium. ." Science of the Total Environment 126: 61-74. Team, R. D. C. (2011). "R: A language and environment for statistical computing, R Foundation for Statistical Computing, Vienna, Austria." Valera, B., E. Dewailly and P. Poirier (2009. ). "Environmental Mercury Exposure and Blood Pressure Among Nunavik Inuit Adults. ." Hypertension 54: 981-986. Watanabe, C. (2002). "Modification of mercury toxicity by selenium: proctical importance?" Tohoku J Exp Medicine 196: 71-77. Yasutake, A., M. Matsumoto, M. Yamaguchi and N. Hachiya (2004). "Current hair mercury levels in japanese for estimation of methylmercury exposure. ." J Health Sci. 50: 120-125.

19

Figure legends

Figure 1. Relationship between plasma Se concentration (ordinate) and blood Hg concentration (abscissa) in the low (open triangle, n = 56) and high (closed circle, n = 55) Hg exposure groups. Both values are expressed as the logarithm of [µmol/L]. The solid line is the regression for the high Hg exposure group, which was statistically significant (p = 0.031). Regression for the low Hg group was not significant and is not shown.

Figure 2. Relationship between the plasma concentrations of SePP (2a), GPx (2b), and albumin (2c) (ordinate), and blood Hg concentration (abscissa) in the lowest (open circle, n = 19) and highest (closed circle, n = 19) Hg exposure groups. Values are expressed as the logarithm of [µmol/L]. The solid line is the regression for the high Hg exposure group; only the regression line for SePP was statistically significant (p = 0.012). No significant regression

20

was observed in the low Hg group.

21

22

23

Table 1 Level of exposure indices for Hg and Se by gender

a

Total (n=111)

Males (n=45)

Females (n=66)

pa

Age [yr]b

63±11

65±11

61±11

Ns

Hair Hg c

0.045±2.31 d

0.068±2.21

0.036±2.19

**

Blood Hg

0.17± 2.44

0.27±2.35

0.13±2.21

**

Blood Se

3.09±1.27

3.24±1.28

3.00±1.25

ns

Blood Hg/Se

0.05±2.3

0.08±2.2

0.02±2.1

**

Plasma Se

1.80±1.14

1.85±1.15

1.77±1.14

ns

Statistical difference between gender. ** p<0.01. For age, arithmetic mean and standard deviations are shown. c For hair Hg, n=99 (n=37 for males; n=62 for females). d For variables other than age, geometric mean and standard deviations in [µmole/g], [µmole/kg] , or [µmole/L] for Hair Hg, Blood Hg and Se, or plasma Se, respectively, are shown. b

24

Table 2 Levels of Hg and Se indices in high vs low Hg exposure groups Hg exposure level a

Low (n=19)

Blood Hgb.

0.05±1.69

0.52±1.69

Blood Se

2.75±1.26

3.56±1.30*

**

Hg/Se ratio

0.03±1.73

0.11±1.65*

***

Plasma Se

1.71±1.10

1.95±1.15*

**

SePP-Se

0.83±1.20

0.94±1.20

*

GPx-Se

0.34±1.22

0.38±1.20

ns

Alb-Se

0.52±1.19

0.61±1.24

*

High (n=19)

p ***

Blood

a b

Participants with lowest and highest B-Hg values were selected (n-19 for each group)

Geometric means and their standard deviations are shown in [µmole/kg] or [µmole/L] (except for the Hg/Se ratio, which is dimensionless). * p<0.05; ** significant difference between high vs low Hg exposure. SePP, GPx, and Alb: selenoprotein P, (plasma) glutathione peroxidase, and albumin, respectively.

25

Table 3. Effect of seafood consumption on the levels of Hg/Se indices Blood Hg Frequency

a

n

µmole/L

b

Tuna and Sward fish

low medium high

15 57 22

0.097±2.39 c 0.150±2.20 c 0.275±2.47 d

SkipJack

low medium high

40 37 17

0.148±2.22 c 0.131±2.38 c 0.287±2.61 d

No

32

0.109±2.25 c

Yes

58

0.190±2.34 d

Toothed whales

Plasma Se µmole/L b

p

1.81±1.11 c,d *** 1.75±1.15 c 1.92±1.13 d

*

p

ns

*

1.79±1.13 1.76±1.15 1.82±1.17

**

1.73±1.12

ns

1.82±1.15

a; Details of the Frequency categories: b; geometric means and SDs are shown. c, d; those with different superscripts are significantly different within the same food category. P refers to significant difference among the different frequency groups (* p<0.05, ** p<0.01, and *** p<0.001, respectively).