Abstract
Background:
Fish contains methylmercury (MeHg) which can cause oxidative stress and neurodevelopmental toxicity at sufficiently high doses. Fish also contains polyunsaturated fatty acids (PUFA) which have both antioxidant (n-3) and oxidant (n-6) properties. Mitochondrial DNA (mtDNA) is sensitive to oxidative stress but has not been previously studied in relation to MeHg exposure or PUFA status.
Objective:
To investigate the associations between MeHg exposure and PUFA status during pregnancy with relative mitochondrial DNA copy number (RmtDNAcn) in mothers and their newborns.
Methods:
In total, 1488 mother-child pairs from the Seychelles Child Development Study Nutrition Cohort 2 were included in this study. Total Hg was measured in maternal blood collected at 28 weeks’ gestation, maternal hair at delivery, and in fetal cord blood. PUFA (n-3 and n-6) were measured only in maternal blood. RmtDNAcn was measured by qPCR in both maternal and cord blood.
Results:
Increasing maternal blood Hg (β=0.001, 95%CI 0.000, 0.002) and n-3 PUFA concentrations (β=0.183, 95%CI 0.048, 0.317) were associated with higher maternal RmtDNAcn. Increasing maternal n-6 PUFA (β=−0.103, 95%CI: −0.145, −0.062) and n-6/n-3 ratio (β=−0.011, 95%CI: −0.017, −0.004) were associated with lower maternal RmtDNAcn. Increasing fetal cord blood Hg was associated with lower fetal RmtDNAcn (β=−0.002, 95%CI: −0.004, −0.000). Neither maternal blood Hg nor PUFA status was associated with fetal RmtDNAcn.
Conclusions:
Our findings suggest that MeHg and PUFA may influence mitochondrial homeostasis although the magnitude of these associations are small. Future studies should confirm the findings and explore the underlying mechanisms.
Keywords: mitochondrial DNA, methylmercury, mercury, polyunsaturated fatty acids
1. Introduction
Methylmercury (MeHg) is present in all fish to a varying extent and is a well-known developmental neurotoxicant in sufficiently high doses (Hong et al. 2012). However, the associations between low to moderate levels of MeHg exposure from consuming fish with naturally-occurring MeHg levels and neurodevelopmental outcomes have been contradictory in different epidemiological studies (Barbone et al. 2018; Daniels et al. 2004; Davidson et al. 1998; Grandjean et al. 1997; Llop et al. 2012; Strain et al. 2015; Vejrup et al. 2016). A better understanding of the mechanisms behind MeHg toxicity, as well as the identification of factors that modify the potential toxicity are necessary for accurate risk assessment of MeHg exposure.
MeHg is a mitochondrial toxicant that can bind to proteins, and in experimental animals results in a loss of mitochondrial membrane potential (Zalups 2000). MeHg can also induce Ca2+ uptake and generation of reactive oxygen species (Fonnum and Lock 2004; Jia et al. 2015), which indirectly impair mitochondrial DNA (mtDNA). Mitochondria have multiple copies of small circular DNA (mtDNA), whose copy number (cn) is regulated through the processes of biogenesis, i.e. production of new mitochondria, and mitophagy (elimination of damaged mitochondria) to appropriately satisfy the metabolic and energy demands of tissues and cells (Byun and Baccarelli 2014; Carelli et al. 2015). Since the main function of mitochondria is to provide energy to cells, disruption of mitochondria may affect a variety of organs, particularly those with the highest energetic demands, such as muscle, heart, liver, and brain (Magner et al. 2015). Thus, we hypothesized that mitochondrial function may be involved in the mechanisms of MeHg neurotoxicity.
mtDNA is sensitive to reactive oxygen species insults due to the absence of protecting histones and DNA repair mechanisms, and resulting DNA alterations can affect mitochondrial biogenesis (Lee and Wei 2005). The content of mtDNA per cell is mainly measured in copy number and it has been suggested as a surrogate marker of mitochondrial biogenesis (Cho et al. 2007; Malik and Czajka 2013). MtDNA damage has been linked to environmental exposures (Roubicek and Souza-Pinto 2017), including metals (Zhang et al. 2011), and a different response of mtDNA in relation to different levels of exposure has been suggested where a low dose of exposure triggers mitochondrial biogenesis, while a high dose leads to deletions of the mtDNA (Hou et al. 2010). To date, there are no studies in human populations assessing mtDNAcn in relation to MeHg exposure and toxicity. An in vitro study of human neural progenitor cell cultures showed increased mtDNAcn when the cells where exposed to a low dose of MeHg (Wang et al. 2016). An in vivo study of bats exposed to MeHg showed reduced mtDNAcn (Karouna-Renier et al. 2014). The reason for these contrasting findings is not known.
Fish is rich in polyunsaturated fatty acids (PUFA), particularly the n-3 PUFA (Gribble et al. 2016). N-3 PUFA can act as antioxidants and may, therefore, counteract toxicity associated with MeHg-induced oxidative stress (Strain et al. 2015). In contrast, the n-6 PUFA, mainly found in vegetable oil but also present in fish, have pro-oxidative properties (Russo 2009). Studies addressing the influence of PUFA status on mtDNA in humans to date have been limited to cell culture studies. These have reported that n-3 PUFA can increase mtDNAcn in a dose and time dependent manner in human skeletal muscle cells (Vaughan et al. 2012), as well as in mouse muscle myoblasts (C2C12 cells) (Lee et al. 2016). This suggests a direct role for n-3 in mtDNA replication. In contrast, n-6 PUFA may have the opposite effect. Experimental evidence in male wistar rats showed increased oxidative damage to mtDNA following an n-6 PUFA rich diet (Ghosh et al. 2006).
To date, there have been no in vivo human studies of early-life exposure to MeHg and its effects on the mitochondrion. In this study, we used a large mother-child cohort from the Seychelles characterized by a fish-rich diet. We analyzed associations between maternal and fetal MeHg biomarkers, maternal n-3 and n-6 PUFA status, and mtDNAcn in maternal and fetal (cord) blood. We hypothesized that increasing exposure to MeHg and n-6 PUFA could impair mitochondrial function and biogenesis via increased pro-oxidative effects with reduced mtDNAcn as a consequence, whereas increasing exposure to n-3 PUFA would have a protective effect which may show as an increased mtDNAcn.
2. Materials and methods
2.1. Study population
The Nutrition Cohort 2 (NC2) of the Seychelles Child Development Study (SCDS) includes mother-child pairs of mixed African, European and East Asian origin. The aim of the SCDS is to evaluate whether MeHg exposure from maternal fish consumption during pregnancy is associated with child developmental outcomes and if the relationship is influenced by nutrition and genetics. Between 2008 and 2011 a total of 1535 mothers were recruited at their first antenatal visit at eight health centers across Mahé, the main island of the Seychelles (Strain et al. 2015). Inclusion criteria included being native Seychellois, being ≥16 y of age, having a singleton pregnancy, and having no obvious health concerns. Biological sampling included maternal blood samples collected at 28 weeks of gestation, fetal blood collected from cord at delivery and maternal hair collected at delivery. The following exclusions were made for this analysis: being a twin or lacking all of maternal hair, maternal blood, and cord blood for measurement of Hg and mtDNAcn, leaving 1488 eligible mother-child pairs. Maternal hair, maternal blood and cord blood were collected at different visits, leading to different numbers of responses for each measurement; there are 1447 maternal and 1055 cord blood mtDNAcn measurements (see counts in Table 1). The study was conducted according to guidelines laid down in the Declaration of Helsinki and all study procedures involving participants were reviewed and approved by the Seychelles Ethics Board, the Research Subjects Review Board at the University of Rochester, and the Regional Ethics Committee at Lund University, Sweden.
Table 1.
Summary statistics for relative mitochondrial DNA copy number (RmtDNAcn), MeHg biomarkers and PUFA status among pregnant women and their newborns (SD, standard deviation).
| Counts | Mean | SD | Min | Median | Max | |
|---|---|---|---|---|---|---|
| Maternal RmtDNAcn | 1447 | 0.36 | 0.21 | 0.034 | 0.32 | 2.79 |
| Fetal RmtDNAcn | 1055 | 0.91 | 0.55 | 0.21 | 0.78 | 5.71 |
| Maternal hair Hg (ng/mL) | 1339 | 3.95 | 3.49 | 0.010 | 2.93 | 31.66 |
| Maternal blood Hg (ng/mL) | 1385 | 18.37 | 10.82 | 1.87 | 15.97 | 84.15 |
| Cord blood Hg (ng/mL) | 1036 | 34.79 | 20.89 | 1.91 | 30.25 | 181.27 |
| Maternal n-3 (mg/mL) | 1450 | 0.27 | 0.09 | 0.12 | 0.27 | 0.64 |
| Maternal n-6 (mg/mL) | 1450 | 1.10 | 0.29 | 0.43 | 1.09 | 2.71 |
| Maternal age (year) | 1482 | 27.12 | 6.30 | 16.27 | 26.15 | 46.76 |
| Maternal BMI at 20 month | 1364 | 26.95 | 6.57 | 14.67 | 26.15 | 49.65 |
| Hollingshead SES at 20 month | 1427 | 31.99 | 10.37 | 11 | 31.5 | 63 |
| Gestational age at birth (week) | 1474 | 38.94 | 1.68 | 28 | 39 | 41 |
| Weight at birth (kg) | 1482 | 3.17 | 0.51 | 1.10 | 3.19 | 5.20 |
2.2. Measurements of biomarkers of methylmercury exposure
We measured total Hg (in maternal blood and hair) as a proxy of MeHg exposure in the mother as well as in cord blood as a proxy of fetal exposure to MeHg, since more than 80% of total Hg in hair and blood is MeHg (Cernichiari et al. 1995a; National Research Council 2000). Maternal hair was collected at delivery and the longest available segment reflecting the gestational period (assuming a growth rate of 1.1 cm/month) was identified for analysis. Total Hg content in the hair was measured as previously described (Cernichiari et al. 1995b) and reported in parts per million (ppm). Total Hg was measured in whole blood from mothers at 28 weeks’ gestation and in cord blood by atomic fluorescence spectrometry using a PSA Millennium Merlin System (PS Analytical, Kent, UK) with a limit of detection of approximately 0.01 ng/mL (McSorley et al. 2018). Estimated fish meals consumed by cohort mothers during pregnancy has been previously reported (8.52 ± 4.56 fish meals per week; Strain et al., 2015), therefore hair and blood mercury measures were presumed to recapitulate average steady-state exposure from habitual fish consumption during gestation.
2.3. Measurements of fatty acid
Maternal non-fasting blood samples (28 weeks’ gestation) were processed at the Public Health Laboratory of the Ministry of Health at Mahé, Seychelles and serum samples were shipped at −80 °C to the Ulster University for fatty acid measurements. Following total lipid extraction from serum, fatty acid methyl esters were prepared by boron trifluroide methanol according to an adaptation of the Folch method (Folch et al. 1957) and quantified by Gas Chromatography–Mass Spectrometry (Agilent 7890A-5975C, UK) as previously described (Strain et al. 2008).
2.4. Measurements of mitochondrial DNA copy number
DNA was isolated from maternal blood (28 weeks’ gestation) and cord blood samples by Qiagen DNA Blood Mini kit (Qiagen, Hilden, Germany) and diluted to 4 ng/μL. As previously described (Xu et al. 2017), mtDNAcn and single copy hemoglobin beta (HBB) gene were measured based on real-time quantitative polymerase chain reaction (PCR) and SYBR Green technology. In short, master mixes for mtDNA copy number and HBB were prepared with KAPA SYBR FAST qPCR Kit Master Mix (2X) ABI Prism (Kapa Biosystems, Woburn, MA, USA) and corresponding primers reported by Hou et al (Hou et al. 2010). A real-time PCR machine (7900HT, Applied Biosystems, Foster City, CA, USA) was used to perform PCR and each run included a standard curve, a control sample and a blank. For the standard curve, one reference DNA sample (from three persons) was diluted to 5 concentrations: 1, 2, 4, 8 and 16 ng/μL, The control sample was to monitor the variance between runs, and the coefficient of variation (CV) of the control sample was 5.3% based on 16 runs. All samples and standard curve points were run in triplicates. R2 for each standard curve was >0.99. The relative mtDNAcn (RmtDNAcn) was calculated by SDS 2.4.1 software (Life Technologies) and indicates average mtDNAcn in each cell.
2.5. Statistical analyses
Linear regression models were fit to investigate the pre-specified associations between RmtDNAcn and MeHg exposure, as well as between RmtDNAcn and maternal n-3 and n-6 PUFA status. The associations with PUFA were analyzed alone or together with MeHg, and all the association analyses were performed in mothers and newborns separately. In the maternal analyses, eight models examined associations between the maternal RmtDNAcn and concentrations. First, we evaluated the association of MeHg and PUFA status in separate models: 1) maternal hair Hg; 2) maternal blood Hg; 3) maternal blood n-3 and n-6 PUFA; 4) n-6/n-3 ratio. Subsequently, we ran models that included both MeHg and PUFA status: 5) hair Hg together with n-3 and n-6 PUFA; 6) hair Hg together with the n-6/n-3 ratio; 7) blood Hg together with n-3 and n-6 PUFA; and 8) blood Hg together with n-6/n-3 ratio. In the fetal analyses, 11 models were performed to analyze the associations with fetal RmtDNAcn. Among them, 8 models use the same predictors as in the maternal analyses as described above, and 3 additional models used the following predictors: 1) the concentration of cord blood Hg; 2) cord blood Hg together with maternal n-3 and n-6 PUFA status; and 3) cord blood Hg together with the maternal n-6/n-3 ratio. Covariates in the adjusted models were chosen a priori based on the literature and their known or suspected association with mtDNAcn (Brunst et al. 2017; Hosnijeh et al. 2014; Kaaman et al. 2007; Liu et al. 2003). In the maternal analyses they included age at enrollment, BMI at 20 months (as a surrogate for pre-pregnancy BMI), and Hollingshead Socioeconomic Status (SES). In the fetal analyses they included maternal age at delivery, BMI at 20 months, and Hollingshead SES and child’s sex, birth weight and gestational age. Smoking and alcohol use during pregnancy were not included in the models due to their very low prevalence in our study population.
Model assumptions were checked through the distribution of the residuals and the residuals from each linear regression showed symmetric distributions. All the statistical analyses were performed using the software R (version 3.3.2; The R Foundation for Statistical Computing). A P-value of 0.05 (2-tailed) was chosen as the criterion for statistical significance in all analysis.
3. Results
3.1. Characteristics of the mothers and newborns
Summary statistics for the biomarkers of MeHg exposure and PUFA status, as well as RmtDNAcn for pregnant women and their newborns are presented in Table 1. In this cohort of 1488 mother-child pairs, the mothers were on average 27.1 years old at childbirth. Their children were 51.7% male, had a mean gestational age of 38.9 weeks, and their mean weight was 3165 g at birth. Total Hg in cord blood (34.76 ng/mL) was 1.89 times higher than that in maternal blood (18.36 ng/mL). Fetal RmtDNAcn in cord blood (0.91 ng/mL) was 2.53 times higher than maternal blood RmtDNAcn (0.36 ng/mL).
Maternal blood Hg was significantly positively correlated with hair Hg and cord blood Hg (Table 2). N-3 PUFA was positively correlated with n-6 and weakly with Hg in maternal hair and blood; and Hg in cord blood. N-6 PUFA had a weak, negative correlation with cord blood Hg. There was no correlation between maternal and fetal RmtDNAcn. Maternal RmtDNAcn was positively correlated with maternal blood Hg and negatively with n-6 PUFA. Fetal RmtDNAcn was negatively correlated with cord blood Hg.
Table 2.
Correlations between relative mitochondrial DNA copy number (RmtDNAcn), MeHg biomarkers, and n-3, n-6 PUFAs among pregnant women and their newborns.
| Maternal RmtDNAcn |
Fetal RmtDNAcn |
Maternal hair Hg |
Maternal blood Hg |
Cord blood Hg |
Maternal n-3 |
Maternal n-6 |
|
|---|---|---|---|---|---|---|---|
| Maternal RmtDNAcn | 1.00 | ||||||
| Fetal RmtDNAcn | 0.049 | 1.00 | |||||
| Maternal hair Hg | 0.021 | 0.0037 | 1.00 | ||||
| Maternal blood Hg | 0.069* | −0.045 | 0.43** | 1.00 | |||
| Cord blood Hg | 0.073* | −0.084** | 0.36** | 0.66** | 1.00 | ||
| Maternal n-3 | 0.016 | −0.048 | 0.10** | 0.13** | 0.11** | 1.00 | |
| Maternal n-6 | −0.11** | 0.015 | 0.031 | −0.016 | −0.063* | 0.44** | 1.00 |
for p<0.05
for p<0.01
3.2. RmtDNAcn and biomarkers of MeHg exposure and PUFA status
In mothers, RmtDNAcn in blood was positively associated with blood Hg. The effect estimate for hair Hg was similar to that of blood Hg, but the association was not significant and not changed by adjustment for n-3 and n-6 PUFA (Table 3). Moreover, a positive association was found for maternal RmtDNAcn with n-3, while negative associations were found with n-6 PUFA and the n-6/n-3 ratio. The associations between RmtDNAcn and Hg exposure or PUFA status were largely independent of each other. For example, adjustment for maternal hair or blood Hg caused little change in effect estimates and p-values for the associations between RmtDNAcn and PUFA status in the mothers (Table 3).
Table 3.
Associations between maternal relative mitochondrial DNA copy number (RmtDNAcn) and MeHg biomarkers and PUFA status. Effect estimates from linear regression analyses are presented as β along with 95% confidence intervals (CI). Significant associations are marked in bold.
| Unadjusted model | Adjusted model b | ||||||
|---|---|---|---|---|---|---|---|
| Model no. a | Predictors | Counts c | β | 95% CI | Counts | β | 95% CI |
| 1) | Maternal hair Hg | 1308 | 0.001 | (−0.002, 0.005) | 1096 | 0.002 | (−0.001, 0.006) |
| 2) | Maternal blood Hg | 1377 | 0.001 | (0.000, 0.002) | 1136 | 0.001 | (0.000, 0.002) |
| 3) | Maternal n-3 | 1432 | 0.183 | (0.048, 0.317) | 1177 | 0.20 | (0.049, 0.354) |
| Maternal n-6 | −0.103 | (−0.145, −0.062) | −0.104 | (−0.150, −0.058) | |||
| 4) | Maternal n-6/n-3 | 1432 | −0.011 | (−0.017, −0.004) | 1177 | −0.012 | (−0.020, −0.005) |
| 5) | Maternal hair Hg | 1296 | 0.001 | (−0.002, 0.005) | 1085 | 0.002 | (−0.002, 0.006) |
| Maternal n-3 | 0.14 | (−0.004, 0.284) | 0.14 | (−0.020, 0.298) | |||
| Maternal n-6 | −0.095 | (−0.139, −0.051) | −0.094 | (−0.142, −0.046) | |||
| 6) | Maternal hair Hg | 1296 | 0.001 | (−0.002, 0.004) | 1085 | 0.002 | (−0.002, 0.006) |
| Maternal n-6/n-3 | −0.008 | (−0.015, −0.001) | −0.009 | (−0.017, −0.001) | |||
| 7) | Maternal blood Hg | 1366 | 0.001 | (0.000, 0.002) | 1127 | 0.001 | (−0.000, 0.002) |
| Maternal n-3 | 0.18 | (0.035, 0.315) | 0.198 | (0.040, 0.355) | |||
| Maternal n-6 | −0.101 | (−0.144, −0.058) | −0.103 | (−0.150, −0.056) | |||
| 8) | Maternal blood Hg | 1366 | 0.001 | (0.000, 0.002) | 1127 | 0.001 | (−0.000, 0.002) |
| Maternal n-6/n-3 | −0.011 | (−0.017, −0.004) | −0.012 | (−0.020, −0.005) | |||
Model 1): Maternal RmtDNAcn ~ Maternal hair Hg
Model 2): Maternal RmtDNAcn ~ Maternal 28 week blood Hg
Model 3): Maternal RmtDNAcn ~ Maternal 28 week blood n-3 + Maternal 28 week blood n-6
Model 4): Maternal RmtDNAcn ~ Maternal 28 week blood n-6/n-3 ratio
Model 5): Maternal RmtDNAcn ~ Maternal hair Hg + Maternal blood n-3 + Maternal blood n-6
Model 6): Maternal RmtDNAcn ~ Maternal hair Hg + Maternal blood n-6/n-3 ratio
Model 7): Maternal RmtDNAcn ~ Maternal blood Hg + Maternal blood n-3 + Maternal blood n-6
Model 8): Maternal RmtDNAcn ~ Maternal blood Hg + Maternal blood n-6/n-3 ratio
Adjusted for maternal age at enrollment, BMI at 20 months (as a surrogate for pre-pregnancy BMI and Hollingshead Socioeconomic Status (SES)
Counts of records in each regression models were different due to missing data on the measurements of total Hg, n-3 and n-6 PUFA and availability of the information related to potential confounders.
In the newborns, RmtDNAcn in cord blood was weakly negatively associated with cord blood Hg, even after including maternal PUFA and further adjustments. However, no associations were observed with maternal Hg concentrations (hair or blood) or PUFA status (Table 4).
Table 4.
Associations between fetal (cord blood) relative mitochondrial DNA copy number (RmtDNAcn) and MeHg biomarkers and PUFA status. Effect estimates from linear regression analyses are presented as β along with 95% confidence intervals (CI). Significant associations are marked in bold.
| Unadjusted model | Adjusted model b | ||||||
|---|---|---|---|---|---|---|---|
| Model no. a | Predictors | Counts c | β | 95% CI | Counts | β | 95% CI |
| 1 | Cord blood Hg | 956 | −0.002 | (−0.004, −0.001) | 889 | −0.002 | (−0.004, −0.001) |
| 2 | Maternal hair Hg | 961 | 0.001 | (−0.009, 0.010) | 902 | −0.002 | (−0.012, 0.009) |
| 3 | Maternal blood Hg | 974 | −0.002 | (−0.006, 0.001) | 907 | −0.003 | (−0.006, 0.000) |
| 4 | Maternal n-3 | 1025 | −0.41 | (−0.827, 0.008) | 951 | −0.37 | (−0.802, 0.067) |
| Maternal n-6 | 0.084 | (−0.046, 0.214) | 0.063 | (−0.071, 0.196) | |||
| 5 | Maternal n-6/n-3 | 1025 | 0.016 | (−0.005, 0.036) | 951 | 0.011 | (−0.010, 0.033) |
| 6 | Cord blood Hg | 930 | −0.002 | (−0.004, −0.000) | 864 | −0.002 | (−0.004, −0.001) |
| Maternal n-3 | −0.43 | (−0.877, 0.018) | −0.42 | (−0.885, 0.048) | |||
| Maternal n-6 | 0.037 | (−0.102, 0.175) | 0.035 | (−0.108, 0.177) | |||
| 7 | Cord blood Hg | 930 | −0.002 | (−0.004, −0.001) | 864 | −0.003 | (−0.004, −0.001) |
| Maternal n-6/n-3 | 0.016 | (−0.007, 0.038) | 0.013 | (−0.010, 0.036) | |||
| 8 | Maternal hair Hg | 939 | 0.0001 | (−0.010, 0.010) | 881 | −0.002 | (−0.012, 0.008) |
| Maternal n-3 | −0.38 | (−0.823, 0.055) | −0.41 | (−0.859, 0.050) | |||
| Maternal n-6 | 0.075 | (−0.061, 0.210) | 0.076 | (−0.062, 0.215) | |||
| 9 | Maternal hair Hg | 939 | −0.0002 | (−0.010, 0.010) | 881 | −0.002 | (−0.013, 0.008) |
| Maternal n-6/n-3 | 0.014 | (−0.008, 0.035) | 0.013 | (−0.009, 0.036) | |||
| 10 | Maternal blood Hg | 968 | −0.002 | (−0.005, 0.001) | 901 | −0.003 | (−0.006, 0.001) |
| Maternal n-3 | −0.39 | (−0.820, 0.032) | −0.34 | (−0.783, 0.100) | |||
| Maternal n-6 | 0.067 | (−0.065, 0.198) | 0.046 | (−0.089, 0.181) | |||
| 11 | Maternal blood Hg | 968 | −0.002 | (−0.006, 0.001) | 901 | −0.003 | (−0.006, 0.000) |
| Maternal n-6/n-3 | 0.014 | (−0.008, 0.035) | 0.009 | (−0.013, 0.031) | |||
Model 1): Fetal RmtDNAcn ~ Cord blood Hg
Model 2): Fetal RmtDNAcn ~ Maternal hair Hg
Model 3): Fetal RmtDNAcn ~ Maternal 28 week blood Hg
Model 4): Fetal RmtDNAcn ~ Maternal 28 week blood n-3 + Maternal 28 week blood n-6
Model 5): Fetal RmtDNAcn ~ Maternal 28 week blood n-6/n-3 ratio
Model 6): Fetal RmtDNAcn ~ Cord blood Hg + Maternal blood n-3 + Maternal blood n-6
Model 7): Fetal RmtDNAcn ~ Cord blood Hg + Maternal blood n-6/n-3 ratio
Model 8): Fetal RmtDNAcn ~ Maternal hair Hg + Maternal blood n-3 + Maternal blood n-6
Model 9): Fetal RmtDNAcn ~ Maternal hair Hg + Maternal blood n-6/n-3 ratio
Model 10): Fetal RmtDNAcn ~ Maternal blood Hg + Maternal blood n-3 + Maternal blood n-6
Model 11): Fetal RmtDNAcn ~ Maternal blood Hg + Maternal blood n-6/n-3 ratio
Adjusted for maternal at delivery, BMI at 20 months, Hollingshead SES and childs sex, birth weight and gestational age.
Counts of records in each regression models were different due to missing data on the measurements of total Hg, n-3 and n-6 PUFA and availability of the information related to potential confounders.
4. Discussion
In this study, we report for the first time associations between exposure to MeHg and mitochondrial DNA status in newborns and their mothers. Consistent with our hypothesis, we observed that in cord blood, increasing Hg concentration was associated with lower RmtDNAcn. Increasing Hg in maternal blood was associated with higher RmtDNAcn, counter to the hypothesis. However, the results should be interpreted very cautiously since the effect estimates for the associations between Hg and RmtDNAcn were overall quite small. We also evaluated the associations of RmtDNAcn with n-3 and n-6 PUFA which are present in fish. Congruent with our hypothesis we found that the anti-oxidative n-3 PUFA were positively associated with RmtDNAcn in the mothers suggesting a protective effect on mitochondria prior to other oxidative stress impacting mtDNA, whereas the pro-oxidative n-6 PUFA showed the opposite association. Together these findings suggest that dietary intake of fish may influence mitochondria homeostasis both via MeHg and PUFA. To note, the timing of assessment of exposure seems important as concurrent blood PUFA and Hg were associated with RmtDNAcn, but not the non-concurrent blood PUFA and hair Hg (e.g. maternal hair Hg or maternal blood Hg vs child RmtDNAcn).
In the newborns lower RmtDNAcn correlated with higher exposure to MeHg, whereas in the mothers the opposite was found. This result may stem from the timing of exposure as newborns and young children are often more susceptible to toxic-induced oxidative stress due to organs structural and functional immaturity, rapidly growing energy demand for aerobic tissues metabolism, and lack of antioxidant systems (Dobbing 1990). As an adult, the human body is better able to regulate the redox balance tightly, partly through mtDNA replication to compensate for mitochondrial dysfunction and generate sufficient ATP. However, exposure occurring in the early stages of life has been hypothesized to lead to a condition of overloading of mitochondria and subsequent adverse health outcomes, including neurodevelopmental disorders (Perrone et al. 2010). Moreover, maternal blood Hg could be associated with unmeasured fish consumption benefits that are beyond n-3 PUFA. Whereas newborns are sensitivity to toxic-insult it may overshadow nutritional benefits. Another, and more speculative explanation for the contrasting findings is that the difference in maternal and child RmtDNAcn effects stem from the different levels of MeHg exposure. For example, we observed that the cord blood contained nearly double the concentration of Hg compared to maternal blood, indicating that the fetus is at higher exposure than the mothers. A biphasic effect of oxidative stress on the mitochondrial proliferation has been suggested whereby low to moderate oxidative stress can stimulate mtDNA synthesis, while higher levels lead to a decompensation stage resulting in depletion of mtDNA and loss of mitochondrial function (Lee et al. 2005; Hou et al. 2010). Such a biphasic response may be present in the mothers and children, respectively. However, this hypothesis needs further experimental support.
Independently of MeHg exposure, n-3 and n-6 PUFA were also found to be associated with maternal RmtDNAcn, but in opposite directions. A mechanism for the influence of PUFA on mtDNA is not fully understood, but may involve n-3 or n-6 specific lipid metabolites, which have different activities towards inflammation and oxidation (Yeates et al. 2017). N-3 PUFA have been reported to regulate mitochondrial biogenesis partially through increased mtDNA replication in skeletal muscle cells and C2C12 myoblast cells (Lee et al. 2016; Vaughan et al. 2012). N-6 PUFA, on the other hand, have been reported to promote oxidative damage on mtDNA in the male wistar rats (Ghosh et al. 2006), which could result in decreasing in mtDNAcn. The negative association we found between n-6/n-3 ratio to the maternal RmtDNAcn suggests that n-6 PUFA may attenuate the positive effects on n-3 PUFA. This relationship is consistent with evidence that a high n-6/n-3 PUFA ratio in the diet can promote the pathogenesis of many chronic disease (Simopoulos 2006).
Strengths of this study include the large cohort of NC2 mother-child dyads in a high fish consuming population (meaning a high intake of both MeHg and n-3, n-6 PUFA), thus a better opportunity to detect subtle associations with mtDNA in both mothers and newborns. Additionally, we were able to account for important relevant covariates. Limitations of this study include the analyses of RmtDNAcn in blood cell populations may not accurately reflect mitochondrial homeostasis in the relevant MeHg target tissues, such as the brain, where energy demands are greater. Additionally, the study lacked information on PUFA levels in the newborns and additional measures of mitochondrial function or oxidative stress are not currently available.
5. Conclusion
In conclusion, our findings indicate that both MeHg and PUFA may alter mitochondria homeostasis as measured by RmtDNAcn. Future studies are needed to confirm these findings and if confirmed, explore the potential causal mechanisms.
Highlights:
We investigated, for the first time, the associations between MeHg exposure, PUFA status and mtDNA copy number in a large mother-child cohort
Maternal blood Hg and n-3 PUFA were positively associated with maternal mtDNA copy number
Maternal blood n-6 PUFA and n-6/n-3 ratio were negatively associated with maternal mtDNA copy number
Fetal cord blood Hg was negatively associated with fetal mtDNA copy number
MeHg and PUFA seem to influence mitochondrial homeostasis
Acknowledgements
We gratefully acknowledge the participation of all Seychellois citizens who took part in the study and the staff from the Child Development Centre, Seychelles for their assistance with data collection.
Funding
This study was supported by the US National Institute of Health (grants R01-ES010219 and P30-ES01247), Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (FORMAS), and in kind support from the Government of Seychelles. The study sponsors had no role in the design, collection, analysis, or interpretation of data, in the writing of this article, or in the decision to submit the article for publication.
Footnotes
Publisher's Disclaimer: 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 citable 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.
Competing interests
The authors declare that they have no competing financial interests.
References
- Barbone F; Rosolen V; Mariuz M; Parpinel M; Casetta A; Sammartano F; Ronfani L; Brumatti LV; Bin M; Castriotta L Prenatal mercury exposure and child neurodevelopment outcomes at 18 months: Results from the Mediterranean PHIME cohort. Int J Hyg Environ Health 2019; 222:9–21 [DOI] [PubMed] [Google Scholar]
- Brunst KJ; Sanchez Guerra M; Gennings C; Hacker M; Jara C; Bosquet Enlow M; Wright RO; Baccarelli A; Wright RJ Maternal lifetime stress and prenatal psychological functioning and decreased placental mitochondrial DNA copy number in the PRISM study. Am J Epidemiol 2017;186:1227–1236 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Byun HM; Baccarelli AA Environmental exposure and mitochondrial epigenetics: study design and analytical challenges. Hum Genet 2014;133:247–257 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carelli V; Maresca A; Caporali L; Trifunov S; Zanna C; Rugolo M Mitochondria: biogenesis and mitophagy balance in segregation and clonal expansion of mitochondrial DNA mutations. Int J Biochem Cell Biol 2015;63:21–24 [DOI] [PubMed] [Google Scholar]
- Cernichiari E; Brewer R; Myers GJ; Marsh DO; Lapham LW; Cox C; Shamlaye CF; Berlin M; Davidson P; Clarkson TW Monitoring methylmercury during pregnancy: maternal hair predicts fetal brain exposure. Neurotoxicology 1995a;16:705–710 [PubMed] [Google Scholar]
- Cernichiari E; Toribara TY; Liang L; Marsh DO; Berlin MW; Myers GJ; Cox C; Shamlaye CF; Choisy O; Davidson P; et al. The biological monitoring of mercury in the Seychelles study. Neurotoxicology 1995b;16:613–628 [PubMed] [Google Scholar]
- Cho YM; Park KS; Lee HK Genetic factors related to mitochondrial function and risk of diabetes mellitus. Diabetes Res Clin Pract 2007;77:S172–S177 [DOI] [PubMed] [Google Scholar]
- National Research Council. Toxicological effects of methylmercury. National Research Council, National Academy Press, Washington, DC: 2000 [Google Scholar]
- Daniels JL; Longnecker MP; Rowland AS; Golding J Fish intake during pregnancy and early cognitive development of offspring. Epidemiology 2004;15:394–402 [DOI] [PubMed] [Google Scholar]
- Davidson PW; Myers GJ; Cox C; Axtell C; Shamlaye C; Sloane-Reeves J; Cernichiari E; Needham L; Choi A; Wang Y; Berlin M; Clarkson TW Effects of prenatal and postnatal methylmercury exposure from fish consumption on neurodevelopment: outcomes at 66 months of age in the Seychelles Child Development Study. JAMA 1998;280:701–707 [DOI] [PubMed] [Google Scholar]
- Dobbing J Vulnerable Periods in Developing Brain In: Dobbing J (eds) Brain, Behaviour, and Iron in the Infant Diet. Springer, London: 1990. 10.1007/978-1-4471-1766-7_1 [DOI] [Google Scholar]
- Folch J; Lees M; Sloane Stanley GH A simple method for the isolation and purification of total lipides from animal tissues. J Biol Chem 1957;226:497–509 [PubMed] [Google Scholar]
- Fonnum F; Lock E The contributions of excitotoxicity, glutathione depletion and DNA repair in chemically induced injury to neurones: exemplified with toxic effects on cerebellar granule cells. J Neurochem 2004;88:513–531 [DOI] [PubMed] [Google Scholar]
- Ghosh S; Kewalramani G; Yuen G; Pulinilkunnil T; An D; Innis SM; Allard MF; Wambolt RB; Qi D; Abrahani A Induction of mitochondrial nitrative damage and cardiac dysfunction by chronic provision of dietary ω-6 polyunsaturated fatty acids. Free Radical Biol Med 2006;41:1413–1424 [DOI] [PubMed] [Google Scholar]
- Grandjean P; Weihe P; White RF; Debes F; Araki S; Yokoyama K; Murata K; Sorensen N; Dahl R; Jorgensen PJ Cognitive deficit in 7-year-old children with prenatal exposure to methylmercury. Neurotoxicol Teratol 1997;19:417–428 [DOI] [PubMed] [Google Scholar]
- Gribble MO; Karimi R; Feingold BJ; Nyland JF; O'Hara TM; Gladyshev MI; Chen CY Mercury, selenium and fish oils in marine food webs and implications for human health. J Mar Biol Assoc U.K 2016;96:43–59 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hong YS; Kim YM; Lee KE Methylmercury exposure and health effects. Journal of preventive medicine and public health. J Prev Med Public Health 2012;45:353–363 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hosnijeh FS; Lan Q; Rothman N; San Liu C; Cheng W-L; Nieters A; Guldberg P; Tjønneland A; Campa D; Martino A Mitochondrial DNA copy number and future risk of B-cell lymphoma in a nested case-control study in the prospective EPIC cohort. Blood 2014;124:530–535 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hou L; Zhu ZZ; Zhang X; Nordio F; Bonzini M; Schwartz J; Hoxha M; Dioni L; Marinelli B; Pegoraro V; Apostoli P; Bertazzi PA; Baccarelli A Airborne particulate matter and mitochondrial damage: a cross-sectional study. Environ Health 2010;9:48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jia G; Aroor AR; Martinez-Lemus LA; Sowers JR Mitochondrial functional impairment in response to environmental toxins in the cardiorenal metabolic syndrome. Arch Toxicol 2015;89:147–153 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kaaman M; Sparks L; Van Harmelen V; Smith S; Sjölin E; Dahlman I; Arner P Strong association between mitochondrial DNA copy number and lipogenesis in human white adipose tissue. Diabetologia 2007;50:2526–2533 [DOI] [PubMed] [Google Scholar]
- Karouna-Renier NK; White C; Perkins CR; Schmerfeld JJ; Yates D Assessment of mitochondrial DNA damage in little brown bats (Myotis lucifugus) collected near a mercury-contaminated river. Ecotoxicology 2014;23:1419–1429 [DOI] [PubMed] [Google Scholar]
- Lee HC; Wei YH Mitochondrial biogenesis and mitochondrial DNA maintenance of mammalian cells under oxidative stress. Int J Biochem Cell Biol 2005;37:822–834 [DOI] [PubMed] [Google Scholar]
- Lee MS; Shin Y; Moon S; Kim S; Kim Y Effects of Eicosapentaenoic Acid and Docosahexaenoic Acid on Mitochondrial DNA Replication and PGC-1α Gene Expression in C2C12 Muscle Cells. Prev Nutr Food Sci 2016;21:317. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Liu CS; Tsai C-S; Kuo C-L; Chen H-W; Lii C-K; Ma Y-S; Wei Y-H Oxidative stress-related alteration of the copy number of mitochondrial DNA in human leukocytes. Free Radic Res 2003;37:1307–1317 [DOI] [PubMed] [Google Scholar]
- Llop S; Guxens M; Murcia M; Lertxundi A; Ramon R; Riano I; Rebagliato M; Ibarluzea J; Tardon A; Sunyer J; Ballester F Prenatal exposure to mercury and infant neurodevelopment in a multicenter cohort in Spain: study of potential modifiers. Am J Epidemiol 2012;175:451–465 [DOI] [PubMed] [Google Scholar]
- Magner M; Kolárová H; Honzík T; Švandová I; Zeman J Clinical manifestation of mitochondrial diseases. Dev Period Med 2015;19:441–449 [PubMed] [Google Scholar]
- Malik AN; Czajka A Is mitochondrial DNA content a potential biomarker of mitochondrial dysfunction? Mitochondrion 2013;13:481–492 [DOI] [PubMed] [Google Scholar]
- McSorley EM; Yeates AJ; Mulhem MS; van Wijngaarden E; Grzesik K; Thurston SW; Spence T; Crowe W; Davidson PW; Zareba G Associations of maternal immune response with MeHg exposure at 28 weeks’ gestation in the Seychelles Child Development Study. Am J Reprod Immunol 2018:e13046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Perrone S; Negro S; Tataranno ML; Buonocore G Oxidative stress and antioxidant strategies in newborns. J Matem Fetal Neonatal Med 2010;23:63–65 [DOI] [PubMed] [Google Scholar]
- Roubicek DA; Souza-Pinto NC Mitochondria and mitochondrial DNA as relevant targets for environmental contaminants. Toxicology 2017;391:100–108 [DOI] [PubMed] [Google Scholar]
- Russo GL Dietary n− 6 and n− 3 polyunsaturated fatty acids: from biochemistry to clinical implications in cardiovascular prevention. Biochem Pharmacol 2009;77:937–946 [DOI] [PubMed] [Google Scholar]
- Simopoulos A Evolutionary aspects of diet, the omega-6/omega-3 ratio and genetic variation:nutritional implications for chronic diseases. Biomed Pharmacother 2006;60:502–507 [DOI] [PubMed] [Google Scholar]
- Strain JJ; Yeates AJ; van Wijngaarden E; Thurston SW; Mulhem MS; McSorley EM; Watson GE; Love TM; Smith TH; Yost K Prenatal exposure to methyl mercury from fish consumption and polyunsaturated fatty acids: associations with child development at 20 mo of age in an observational study in the Republic of Seychelles. Am J Clin Nutr 2015;101:530–537 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Strain JJ; Davidson PW; Bonham MP; Duffy EM; Stokes-Riner A; Thurston SW; Wallace JM; Robson PJ; Shamlaye CF; Georger LA; Sloane-Reeves J; Cernichiari E; Canfield RL; Cox C; Huang LS; Janciuras J; Myers GJ; Clarkson TW Associations of maternal long-chain polyunsaturated fatty acids, methyl mercury, and infant development in the Seychelles Child Development Nutrition Study. Neurotoxicology 2008;29:776–782 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang X; Yan M; Zhao L; Wu Q; Wu C; Chang X; Zhou Z Low-dose methylmercury-induced apoptosis and mitochondrial DNA mutation in human embryonic neural progenitor cells. Oxid Med Cell Longev 2016;2016: :5137042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Vaughan RA; Garcia-Smith R; Bisoffi M; Conn CA; Trujillo KA Conjugated linoleic acid or omega 3 fatty acids increase mitochondrial biosynthesis and metabolism in skeletal muscle cells. Lipids Health Dis 2012;11:142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Vejrup K; Schjolberg S; Knutsen HK; Kvalem HE; Brantsaeter AL; Meltzer HM; Alexander J; Magnus P; Haugen M Prenatal methylmercury exposure and language delay at three years of age in the Norwegian Mother and Child Cohort Study. Environ Int 2016;92-93:63–69 [DOI] [PubMed] [Google Scholar]
- Xu Y; Li H; Hedmer M; Hossain MB; Tinnerberg H; Broberg K; Albin M Occupational exposure to particles and mitochondrial DNA-relevance for blood pressure. Environ Health 2017;16:22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yeates AJ; Thurston SW; Li H; Mulhern MS; McSorley EM; Watson GE; Shamlaye CF; Strain JJ; Myers GJ; Davidson PW; van Wijngaarden E; Broberg K PUFA Status and Methylmercury Exposure Are Not Associated with Leukocyte Telomere Length in Mothers or Their Children in the Seychelles Child Development Study. J Nutr 2017;147:2018–2024 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zalups RK Molecular interactions with mercury in the kidney. Pharmacol Rev 2000;52:113–144 [PubMed] [Google Scholar]
- Zhang W; Liu Y; An Z; Huang D; Qi Y; Zhang Y Mediating effect of ROS on mtDNA damage and low ATP content induced by arsenic trioxide in mouse oocytes. Toxicol In Vitro 2011;25:979–98 [DOI] [PubMed] [Google Scholar]