Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Dec 29.
Published in final edited form as: Neurotoxicology. 2023 Dec 20;100:117–123. doi: 10.1016/j.neuro.2023.12.011

Synchrotron speciation of umbilical cord mercury and selenium after environmental exposure in Niigata

Monica Weng a, Natalia V Dolgova b, Linda I Vogt a, Muhammad Qureshi c, Dimosthenis Sokaras c, Thomas Kroll c, Hisashi Saitō d, John L O’Donoghue e, Gene E Watson e,f, Gary J Myers e,g, Tomoko Sekikawa j, Ingrid J Pickering a,h,i, Graham N George a,h,i,*
PMCID: PMC11682714  NIHMSID: NIHMS2040643  PMID: 38128735

Abstract

The insidious and deadly nature of mercury’s organometallic compounds is informed by two large scale poisonings due to industrial mercury pollution that occurred decades ago in Minamata and Niigata, Japan. The present study examined chemical speciation for both mercury and selenium in a historic umbilical cord sample from a child born to a mother who lived near the Agano River in Niigata. The mother had experienced mercury exposure leading to more than 50 ppm mercury measured in her hair and was symptomatic 9 years prior to the birth. We sought to determine the mercury and selenium speciation in the child’s cord using Hg Lα1 and Se Kα1 high-energy resolution fluorescence detected X-ray absorption spectroscopy, the chemical speciation of mercury was found to be predominantly organometallic and coordinated to a thiolate. The selenium was found to be primarily in an organic form and at levels higher than those of mercury, with no evidence of mercury-selenium chemical species. Our results are consistent with mercury exposure at Niigata being due to exposure to organometallic mercury species.

Keywords: Mercury, Selenium, High-energy resolution fluorescence detected X-ray absorption spectroscopy, Neurotoxicology, Synchrotron light, Umbilical cord, Niigata, Minamata, Stem cell

1. Introduction

Mercury’s compounds are well known for their toxic nature, with two broad types – inorganic and organometallic compounds – exhibiting distinct toxicology and pathology (O’Donoghue et al., 2020). Organometallic mercury compounds are widely recognized for their potent neurotoxicity, the realization of which has been defined by a number of poisonings of human populations (Clarkson, 1998). Perhaps the best known of these occurred at Minamata (Kurland, et al., 1960, Takeuchi et al., 1962, Eto et al., 2010). It was caused by pollution with mercury-containing industrial waste derived from the chemical manufacture of acetaldehyde, using a process that employed inorganic mercury to catalyze the hydration of acetylene (Yokoyama, 2018, James et al., 2020). The release of the mercury-containing industrial waste into the Minamata river occurred a short distance from the outlet into Minamata bay and led to contamination of fish and other seafood. Local villagers, consuming the contaminated seafood, were impacted with varying degrees of severity, with symptoms discussed below.

Less well-known than Minamata is the pollution and poisoning at Niigata (Saitō, 2009, Saitō et al., 2015) which also had its origins in industrial waste from the manufacture of acetaldehyde at a chemical plant. The Niigata poisoning began after that at Minamata, with the first cases being recognized in 1965. Early recognition of the etiology of the exposure led to women at Niigata being offered abortions if their blood mercury exceeded 50 ppm during gestation (Saitō et al., 2020). In Niigata, the chemical plant was located some 65 km from the outlet of the Agano River into the Sea of Japan. The majority of human cases of what subsequently became known as Niigata-Minamata Disease derived predominantly from consumption of contaminated freshwater fish and shellfish from the Agano River, rather than marine fish and shellfish, as was the case at Minamata. These two industrial events tied to pollution of fish have led to concern about consumption of fish in general. All fish naturally contain methylmercury at some level, and low-level exposure occurs in everyone who consumes fish. At Minamata fish averaged 10 ppm mercury, with some containing over 100 ppm mercury (Yokoyama, 2018), while fish with only normal environmental exposures usually average <1 ppm, up to 2–3 ppm. The level of exposure associated with harm is not presently known but is likely to be complex and involve chemical speciation both of mercury and of other elements such as selenium (Dolgova et al., 2019, James et al., 2022).

Much of what we know about the effects of exposure to organometallic mercury compounds comes from the Minamata and, to a lesser extent, the Niigata events. The disproportionate sensitivity of children, and especially fetuses exposed in utero, was first noted in Minamata. In some cases, children exposed in utero were born severely disabled and developmentally impaired, while their mothers were relatively unaffected. The signs and symptoms of those exposed to a toxic dosage are now known to be typical of organometallic mercury poisoning and include paresthesia, dysarthria, visual and auditory disturbances, ataxia, seizures and in some cases death.

Despite extensive study, the details of both the Minamata and Niigata episodes have been marked by controversy. Minamata disease is frequently presented in undergraduate toxicology classes as a prototypical example of heavy metal biotransformation in the environment. The scenario often offered is that the initial pollution from the chemical plant was inorganic mercury, which was converted to methylmercury forms by microbial biomethylation in anaerobic muds and sediments, followed by magnification through trophic levels (Klassen, 2013). While there is evidence suggesting that biomethylation of inorganic forms may not have played a major role in Minamata (Eto et al., 2001, James et al., 2020). Biomethylation of inorganic mercury is primarily carried out by anaerobic prokaryotes, although complex biota including benthic algae may also play a role (Desrosiers, et al., 2006). Suggestions that the industrial waste itself contained methylmercury (Eto, et al., 2001) have altered modern opinions. These suggestions derive mostly from contemporary records and modern analyses of the preserved brain tissues of an experimental animal, Cat 717, which had been fed with the plant effluent by the company physician in an effort to determine the source of the outbreak (Eto, et al., 2001). The written records at the time of the experiment show that the pathology exhibited by Cat 717 was consistent with what is now recognized as organometallic mercury poisoning and not with inorganic mercury poisoning. Likewise, the subsequent histology of the preserved tissues was highly characteristic of organometallic mercury exposure (Eto, et al., 2001). Because Cat 717 was fed the chemical plant waste directly, environmental biotransformation could not have played a role. Hence, the mercury in the plant effluent was at least partly organometallic, with the specific chemical form of the organometallic mercury widely assumed to be methylmercury. However, seemingly at odds with this idea, were subsequent conventional chemical analyses of both the preserved cat brain tissues and the industrial waste that were carried out in the same study, which indicated that inorganic mercury species, and not methylmercury, were the predominant forms in both (Eto, et al., 2001). This was attributed to demethylation of the stored samples over time, although no chemical basis for this was postulated.

In previous work (James et al., 2020), we have used newly-available advanced X-ray spectroscopy methods to provide in-situ speciation, without chemical pretreatment, combined with computational chemistry to examine some of the chemistry behind the Minamata episode. Our study examined historical brain tissue samples from the same Cat 717 that was exposed to the industrial waste from the Minamata chemical plant. At the outset we expected our results to mirror those of conventional chemical analysis (Eto, et al., 2001) and to show a predominance of inorganic forms of mercury. Unexpectedly, we found that the mercury in the brain of Cat 717 was predominantly organometallic, with smaller quantities of inorganic β-mercuric sulfide (James et al., 2020). Early studies trying to isolate mercury with organic solvents from tissue of animals thought to be exposed to mercury failed until it was discovered that the mercury was bound to tissue thiolates from which it had to be released in order to be detected. In addition, conventional chemical analyses often use mobility-based techniques, and because such methods might miss unexpected chemical forms of mercury, it seemed plausible that forms of organometallic mercury other than methylmercury might be involved (James et al., 2020). Synchrotron X-ray speciation methods depend upon basic atomic physics and hence are not affected by these complications, as mercury species can be directly studied in situ (James et al., 2023). Moreover, modern high-resolution methods (James et al., 2023) afford quantifiably greater precisions in speciation, by a factor of two or three for selenium (Nehzati, et al., 2021) and for mercury (Nehzati, et al., 2022), respectively.

In addition, computational chemistry suggested that a different organometallic compound might have been important, α-mercuri-acetaldehyde, or its oxidation product α-mercuri-acetic acid, and which, consistent with advanced X-ray spectroscopy results, could be additionally bound to a thiolate (James et al., 2020). This differs from the dominant form naturally found in marine fish lacking substantial exposure to pollution, which is methymercury-L-cysteinate (Harris, et al., 2003, George, et al., 2008). Compounds involving the α-mercuri-acetaldehyde group are expected to form by keto-enol tautomerization of the organometallic mercury catalytic intermediate in the industrial process (James et al., 2020). They are known in the chemical literature as high-melting stable solids and have even been structurally characterized (Halfpenny and Small, 1979). A role for forms other than methylmercury had been previously postulated by others as long ago as 1982 (Taylor, 1982) but despite this, our suggestion generated some controversy.

In Japan, the safekeeping of the umbilical cord of a newborn child is a well-kept and time-honored tradition (Selin and Stone, 2009). Japanese hospitals routinely place part of the umbilical cord in a special box designed to preserve it, which is carried home by the mother, with the dried stump often kept throughout adulthood. However, to our knowledge there is no standard way of collecting specimens. Since the mercury present may be in blood, and since the cord can contain a variable amount of this, it is not possible to determine whether remaining cord blood might account for some or all of the mercury in cord samples. In this study we have used advanced X-ray absorption spectroscopy to examine the chemical speciation of mercury and selenium in a small sample of an umbilical cord from a child born to a mother who had been exposed to mercury during the Niigata event, and who had recovered prior to giving birth.

2. Materials and Methods

2.1. Tissue Samples

A sample of dried umbilical cord of a child was donated by a woman who consumed Agano River fish daily. Like others living downstream of the chemical plant, the family earned their living through fishing the Agano River. The fish that were routinely eaten by the family had all been caught in the river downstream from the chemical plant. Although catching fish from the lower reaches of the Agano River was restricted in 1965, fishing in the upper reaches of the river was not restricted, and people continued to eat river fish even with restrictions. In 1965, her hair mercury was determined to be over 50 ppm and she had paresthesia of hands and feet, muscle cramps, dizziness, and fatigue. These are symptoms that could be attributable to organometallic mercury exposure. However, she was considered to have recovered from the mercury exposure prior to this pregnancy and did not consume significantly contaminated fish during gestation. The child, a female, was born in 1974 and appeared normal and in good health. The umbilical cord tissue was subjected to conventional chemical analysis for total mercury and was found to contain a mercury content of 0.2 ppm. This value is within the ranges reported by other authors (Grandjean et al., 2005, Sakamoto et al., 2007) and would not be considered elevated.

2.2. X-ray Spectroscopy

High energy resolution fluorescence detected X-ray absorption spectroscopy (HERFD-XAS) experiments were carried out at the Stanford Synchrotron Radiation Lightsource (SSRL) with the SPEAR3 storage ring containing 500 mA at 3.0 GeV and using beamlines 6–2 and 15–2, both of which employ a Si(311) double crystal monochromator with downstream specular optics to focus the beam. For all measurements, samples were maintained at a temperature of 10 K using a helium flow cryostat (Oxford instruments, Abingdon, UK) containing sample cuvettes inclined at an angle of 45° to the incident X-ray beam to facilitate measurement of X-ray fluorescence, giving an incident X-ray path length of 2.8 mm. An in-hutch photon shutter was employed to prevent exposure of the sample to the X-ray beam when data were not being actively recorded. The incident and transmitted X-ray intensities were measured using a helium-filled gas ionization chamber. Transmitted X-ray intensities were measured either using a nitrogen-filled gas ionization chamber on beamline 6–2, or a photodiode placed after the sample on 15–2. The photodiode employed was specially thinned so as to transmit a significant fraction of the incident X-ray beam in order to illuminate a calibration foil, with another photodiode placed after the foil to measure the calibration foil transmittance. The incident beam energy was calibrated relative to the lowest-energy inflections of the absorbance of the calibration foils, measured as the first peak in the first derivative. For the Hg LIII edge a Hg-Sn amalgam calibration foil was used and for the Se K-edge a grey hexagonal selenium foil was used, with first inflection values assumed as 12284.0 eV and 12658.0 eV, respectively.

High energy resolution X-ray fluorescence was measured by means of a 7-element array of spherically bent crystal analyzers in the Johann geometry, using Si(555) and Si(844) crystals (Sokaras et al, 2013) for the Hg Lα1 or Se Kα1 emission, respectively. Both cuts give the near-90° Bragg angle that is needed to obtain proper resolution enhancement with HERFD-XAS; Si(555) and Si(844) have respectively θB=81.8∘ and 85.2°, at the Hg Lα1 and Se Kα1 energies. The emission intensity was measured using a single-element silicon drift Vortex detector (Hitachi High-Technologies Science America Inc., Northridge, CA, USA). Standard compounds are often concentrated, giving high-count rates and in some cases self-absorption (Nehzati et al., 2021). For high count-rate samples, aluminum filters positioned upstream of the incident ion chamber were used to maintain the count-rates of the silicon-drift detector within the pseudo-linear regime. Self-absorption effects in the HERFD-XAS signal, H(E), were corrected using the measured X-ray transmittance as previously described (Nehzati et al., 2021). In brief, the sample absorbance A(E) is measured simultaneously with H(E) and the sample absorbance at the fluorescence energy AF is also measured. The self-absorption-free undistorted HERFD-XAS signal H′(E) can then be computed using eq. 1 (Nehzati et al., 2021).

H′(E)∝H(E)A(E)+AF1−exp−A(E)+AF (1)

Data were collected using the SPEC data acquisition software (Certified Scientific Software, Cambridge, Massachusetts, USA). Data reduction and analysis were carried out as previously described (James et al., 2020, Nehzati et al., 2021, Nehzati et al., 2022), using the EXAFSPAK suite of computer programs (https://www-ssrl.slac.stanford.edu/exafspak.html). Data normalization was carried out using the EXAFSPAK program BACKSUB (James et al., 2020), which employs tabulated X-ray cross sections (McMaster et al., 1969).

3. Results

Fig. 1 shows the Hg Lα1-HERFD-XAS data of the umbilical cord sample compared with the spectra of a series of biologically relevant standard compounds. The spectrum of the umbilical cord shows an intense pre-edge peak at 12287.2 eV which is characteristic of two-coordinate Hg(II) (Nehzati et al., 2022). As previously noted, organometallic mercury bound to nitrogen and selenium show shifts in the pre-edge peak energies to higher and lower energies, respectively, compared with that bound to sulfur (Fig. 1c). Fig. 1 also compares the Hg Lα1-HERFD-XAS of some four-coordinate species (Fig. 1e, f, and g), which lack the characteristic two-coordinate peak (Nehzati et al., 2022) and appear different in shape to the umbilical cord spectrum. Overall, the umbilical cord spectrum (Fig. 1a) bears a strong resemblance to that of methylmercury L-cysteinate (Fig. 1b), albeit with a slightly less intense pre-edge peak for the umbilical cord spectrum, suggesting the predominance of organometallic Hg bound to a thiolate donor in the umbilical cord. Since the speciation capabilities of Hg Lα1-HERFD-XAS are largely limited to the first coordination shell, the method cannot distinguish between different organic donors to Hg (e.g. ethylmercury and methylmercury compounds show very similar spectra) (Nehzati et al., 2022). Thus, while we have employed methylmercury standards in our analyses, our study only relates to the presence of organometallic mercury, and not specific organometallic mercury forms.

Fig. 1.

Fig. 1.

Open in a new tab

Comparison of Hg Lα1-HERFD-XAS data of the umbilical cord sample with spectra of biologically relevant standard compounds (from Nehzati et al., 2022): a, Umbilical cord sample; b, methylmercury L-cysteinate (RHgSR′); c, methylmercury L-selenocysteinate (RHgSeR′, brown line) and methylmercury 1-methylimidazole (RHgNR′2, blue line); d, mercury bis-L-cysteinate (Hg(SR)2, orange line) and mercury bis-(1-methylimidazole, blue line); e, β-HgS; f, nano-HgSe (solid line) and HgSe (broken line); and g, mercury tetrakis-L-cysteinate (Hg(SR)4). Samples b-d and g were measured as solutions. The vertical red broken line is included to guide the eye to small energy shifts in the spectral features.

Quantitative chemical speciation of spectra of complex mixtures can be obtained from HERFD-XAS using linear combination analysis (James et al., 2022, 2023). Here, spectra of unknowns are fitted to the sum of the spectra of standard compounds, such as those shown in Fig. 1. There are some obvious limitations of such analyses: the ability to quantify components in mixtures will depend on how distinctive the individual standard spectra are; small quantities of a minor component may be difficult or impossible to quantify; and there is an absolute requirement for the spectra of a range of appropriate standards. However, as we have previously discussed (Nehzati et al., 2022), the approximately 3-fold improvement in spectroscopic energy resolution of Hg Lα1-HERFD-XAS relative to conventional XAS translates to a 2.3-fold quantitative speciation advantage for complex mixtures (Nehzati et al., 2022). A linear combination analysis of the umbilical cord Hg Lα1-HERFD-XAS data is shown in Fig. 2. This analysis used the standard spectra of Fig. 1 and the method of James et al., 2022 which excludes components for which the fraction in the fit is greater than their estimated standard deviation, obtained from the diagonal of the variance-covariance matrix. The data fit best to only two components; a majority of RHgSR′, modelled as methylmercury L-cysteinate (80±3%), with a minor component for which β-HgS (20±3%) gave the best fit. Somewhat less adequate fits were obtained when β-HgS was excluded, and spectra of the other four-coordinate species were employed (i.e., data shown in traces f and g of Fig. 1). Hence, determination of β-HgS as a minor component is not unambiguous, but our confidence is increased by previous observations of β-HgS in samples related to the Minamata accident (James et al., 2020). The absence of selenium-mercury species determined from the selenium perspective using the Se Kα1-HERFD-XAS, discussed below, further supports this determination.

Fig. 2.

Fig. 2.

Open in a new tab

Linear combination analysis of Hg Lα1-HERFD-XAS data of the umbilical cord sample. The best fit (green line) to the experimental data (yellow points) is shown at top, with the fit residual (red) offset below. The component spectra (middle) are scaled by their contributions to the fit: RHgSR′, modelled as methylmercury L-cysteinate, 80±3% (blue); and β-HgS, 20±3% (orange). See text for details.

Because the toxicology of mercury and selenium are closely linked (Gailer, et al., 2000, George et al., 2008, Korbas et al., 2010, MacDonald et al., 2015, Spiller, 2018, Manceau et al., 2021a, b, Dolgova et al., 2019, James et al., 2022) we additionally examined the Se Kα1-HERFD-XAS data of the umbilical cord sample. Fig. 3 compares the Se Kα1-HERFD-XAS of the umbilical cord sample with spectra of selected biologically relevant standard compounds (Nehzati et al, 2021), with a linear-combination analysis (James et al., 2022) shown in Fig. 4.

Fig. 3.

Fig. 3.

Open in a new tab

Comparison of Se Kα1-HERFD-XAS data of the umbilical cord sample with the spectra of biologically relevant standard compounds (from Nehzati et al., 2021): a, umbilical cord sample; b, Selenocysteine-Cysteine selenylsulfide (RSeSR); c, L-selenomethionine (RSeR); d, seleno-bis-diglutathione (RSSeSR); e, α-Se; f, L-selenocysteine (RSeH) (solid line) and L-selenocysteinate (RSe−) (broken line); g, nano-HgSe (solid line) and HgSe (broken line); and h, trimethylselenonium [R3Se]−. The vertical red broken line is included to guide the eye to small energy shifts in the spectral features.

Fig. 4.

Fig. 4.

Open in a new tab

Linear combination analysis of Se Kα1-HERFD-XAS data of the umbilical cord sample. The upper part of the plot (a) shows the experimental data (points) and the linear combination fit. The three components used in the best fit, scaled by their contributions to the fit, are shown in b and were: RSeR modeled as selenomethionine (SeMet), blue line; RSeSR modeled as selenocysteine-cysteine selenylsulfide (Cys–Se–S–Cys), orange line; and RSSeSR modeled as seleno-bis-diglutathione (GS–Se–SG), dark green line. The lower red trace shows the fit residual.

The Se Kα1 fluorescence line shows distinctive chemical shifts with formal oxidation states (Nehzati et al., 2021), unlike for Hg Lα1 line which lacks chemical sensitivity due to electronic shielding effects (Nehzati et al., 2022). Hence, analysis of Se Kα1-HERFD-XAS presents an additional complexity compared to that of the Hg Lα1-HERFD-XAS. In Se Kα1-HERFD-XAS cases where there are mixtures with different formal oxidation states (e.g., Se(–II) and Se(VI)), multiple Se Kα1-HERFD-XAS must be obtained using different emission energies corresponding to the chemically shifted emission peaks (Nehzati et al., 2021). Fortunately, in the present case, the Se Kα1-HERFD-XAS of the umbilical cord sample shows no evidence of higher oxidation state selenium species, and a predominance of reduced Se(–II) forms, so these complications can be neglected. The linear combination analysis of the selenium spectra from the umbilical cord, conducted using the same methods as described for the Hg Lα1 analysis of Fig. 2, required three components, and suggested a predominance of organo-selenium species with little contribution from mercury-selenium complexes. The best fit algorithm suggested 41±1% RSeR modelled as L-selenomethionine, 40±2% RSeSR (a selenylsulfide) and 19±1% RSSeSR. Notably, the analysis detected no significant contribution from HgSe forms (James et al., 2022), which is in agreement with our interpretation of Hg Lα1 HERFD-XAS, discussed above. However, it is possible that HgSe or related species, in the form of nano-particulates, is present but below the detection limit for bulk measurements. HgSe and the mixed form with sulfur HgSxSe(1-x), have now been detected in a wide range of samples (e.g., Marjota and Berry, 1980, MacDonald et al., 2015) including vertebrate brain tissue (e.g., Devabathini et al., 2023). Thus, it seems plausible that small HgSe or HgSxSe(1-x) deposits may be present even when no bulk contribution of such species is detected.

From the magnitude of the Hg LIII and Se K-edge-jumps, we were able to estimate the relative molar levels of mercury and selenium from our data, with selenium being approximately 2-fold higher than mercury. This is consistent with the absence of significant Hg-Se species in the umbilical cord sample, as observed from both Hg and Se points of view.

4. Discussion

We have carried out a spectroscopic speciation analysis of both mercury and selenium of a historic umbilical cord sample from a child of a mother who had recovered from mercury exposure in the Niigata event. We found that the mercury species present in the umbilical cord was predominantly thiolate-bound organometallic mercury with a smaller fraction of black β-mercuric sulfide, which resembles the determination of the brain tissue from Minamata Cat 717 (James et al., 2020). Our findings on the umbilical cord sample are consistent with maternal exposure to organometallic mercury and are also consistent with the literature and with previous studies.

Three possible scenarios can be envisioned to explain the presence of organometallic mercury in the umbilical cord that we have analyzed. A first scenario would be exposure to natural levels of methylmercury from the environment and unrelated to any industrial release. We have previously noted that the mercury levels of the umbilical cord do not fall into an elevated range when compared with levels reported for communities consuming marine seafood. However, freshwater fish normally have lower mercury than marine fish (Zupo et al., 2019) and it seems likely that past maternal exposure might have played a role. Secondly, organometallic compounds might have been released from the chemical plant, and thirdly inorganic compounds might have been released and biomethylation could also have occurred.

At Minamata, as previously discussed (James et al., 2020, Eto et al., 2001), it seems that the chemical form of mercury in the industrial waste generated by the Chisso plant was organometallic. At least early on, the Chisso chemical plant at Minamata and the Shōwa Denkō Kanose plant at Niigata employed similar processes for the manufacture of acetaldehyde from hydration of acetylene (Othmer et al, 1956). Inorganic mercury Hg(II) was used as a catalyst by forming an organometallic mercury intermediate, α-mercuri-vinyl alcohol. Subsequent protonolysis of the Hg–C bond yielded regenerated catalyst and vinyl alcohol, with vinyl alcohol spontaneously converting to acetaldehyde. Side reactions were responsible for reduction of a fraction of the Hg(II) to elemental mercury; both plants would have employed manganese or iron compounds to re-oxidize elemental mercury, regenerating Hg(II) catalyst. Since the mercury catalyst was expensive yet essential for the chemistry used, the industrial process itself was intended not to release substantial mercury in any form, if operated as designed.

The Chisso plant in Minamata had been operating for two decades before poisoning of the local population was officially recognized in 1952. Up to 1952 the Minamata plant used manganese (IV) oxide for catalyst regeneration (Yokoyama, 2018); in 1952 the regeneration material was changed from MnO2 to nitric acid and ferric sulfate, Fe2(SO4)3 (Yokoyama, 2018). Subsequently the Fe2(SO4)3 was in turn replaced by pyrite cinders, waste derived from sulfuric acid manufacture, a poorly-defined material containing Fe2O3 but with variable composition. Then in 1955, the plant began adding seawater instead of freshwater for process hydration (Yokoyama, 2018). Overall, these changes may have altered the controlled chemistry originally envisioned (Othmer et al., 1956). It seems likely that one or more of these process changes may have led to catalyst recovery problems.

The history of the waste discharge at Niigata is less clear than at Minamata (Saitō et al., 2009). At Niigata, organometallic mercury was detected in water waste discharges from the plant, but whether this represented the majority mercury species present is unclear. If the bulk of the mercury discharged into the Agano river was inorganic then the conventional process of biotransformation to organometallic forms by anaerobic microbiota in muds and sediments, with accumulation through trophic levels, may have played a role (Klassen, 2013). Notably, the environmental circumstances at Niigata differed from those at Minamata. In Minamata, factory effluent was discharged directly into the Minamata river in close proximity to Minamata bay, while the factory at Niigata had a holding pond for waste. In January 1959, six years before the Niigata poisoning was recognized, there was a large-scale fish die-off in the Agano River. While the exact circumstances are unclear, there was suspicion that the holding pond breached, allowing the waste to flow into the river which was transformed from clear water to a milky consistency with evident fish die-off (Saitō et al., 2009, p. 163). Whether the plant discharged inorganic mercury with environmental biotransformation to methylmercury, or the plant discharged some other organometallic mercury species, cannot be determined from the present study.

The sample analyzed in the current study has been kept in a dried form in a family home for more than four decades at room temperature, which gives rise to questions related to the stability of toxicologically relevant mercury compounds. In general, two-coordinate organometallic mercury complexes with sulfur donors have high stability, especially if kept dry, and their persistence in a dried tissue sample is not surprising. We have observed organometallic mercury complexes in even older samples of wet tissues preserved in formaldehyde (James et al., 2020). The four-coordinate black mercury sulfide β-HgS (metacinnabar) is less stable than the two-coordinate red polymorph α-HgS but traces of other elements such as Fe stabilize the black polymorph (Dickson and Tunell, 1959), and black deposits on decades old dental amalgam have been observed to be β-HgS (George et al., 2009). The same is true of many selenium complexes (James et al., 2020). While we are reassured by these previous findings, we cannot exclude the possibility that some other mercury or selenium species may have been present in the original umbilical cord sample and have altered their chemical speciation over time.

During pregnancy the umbilical cord attaches to the placenta, which in turn attaches to the wall of the uterus. Both consist of fetal tissue, derived from the trophectoderm of the pre-implantation embryo at ~5 days post-fertilization (Turco and Moffett, 2019). The placenta functions to transport nutrients to the growing fetus and to remove fetal excretory products via the maternal blood. The umbilical cord consists of a tubular sheath of amnion enclosing arteries and a vein surrounded by mesenchymal stromal cells used as a source of stem cells. While our results do not provide any information about cellular location, this report raises the issue of whether organometallic mercury may localize in mesenchymal stem cells. The organometallic compounds of mercury, however, appear to be transported preferentially by the placenta to the developing fetus, probably via the LAT system of transporters (Simmons-Willis et al., 2002), providing an explanation for their presence in the sample investigated here.

The lack of any detectible Hg-Se species is of interest. Recently, there have been suggestions that consumption of high-mercury marine fish can be safe irrespective of the mercury level, providing that the fish selenium levels exceed that of the mercury (Ralston et al., 2016). It is known that selenium may effectively protect against the toxic effects of mercury, partly through formation of relatively inert HgSe species (Gailer et al., 2000). However, some caution may be appropriate because, as pointed out recently, both inorganic and organometallic mercury block selenium transport in a vertebrate model system (Dolgova, et al., 2019). Moreover, pre-administration of selenium species has been found to increase the toxic effects of organometallic mercury relative to organometallic mercury alone (Dolgova, et al., 2019). In addition, a recent analysis of human brain tissue samples suggests that the presence of substantial HgSe may be a signature of acute poisoning caused by short-term high-level organometallic mercury exposure (James et al., 2022). In that same study, we examined high-mercury brain tissues from individuals with a history of more than six decades of marine fish consumption and found no evidence of any substantial levels of HgSe species (James et al., 2022). Moreover, the greatest abundance of HgSe was observed in tissue from individuals who had survived for years following acute exposure to high levels of organometallic mercury. In contrast, the umbilical cord is a young organ, being only 9 months old, and hence the lack of HgSe or related species might not be surprising.

5. Acknowledgements

We acknowledge grant support from the Natural Sciences and Engineering Research Council of Canada (NSERC) (RGPIN-2022–04959 to I.J.P. and RGPIN-2019–05351 to G.N.G.) a Canada Foundation for Innovation (CFI) John Evans Leader’s Fund award (I.J.P., JELF-CFI-228315) and Canada Research Chairs (I.J.P., CRC-2019–00162 and G.N.G., CRC-2016–00092). M.W. and L.I.V. are Fellows in the NSERC CREATE to INSPIRE training grant (CREATE 555378–2021 to I.J.P. and others). Support for research at the University of Rochester was provided by the National Institute of Environmental Health Sciences (ES-015578 and ES-001247). Use of the Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, is supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences under Contract No. DE-AC02–76SF00515. The SSRL Structural Molecular Biology Program is supported by the DOE Office of Biological and Environmental Research, and by the National Institutes of Health, National Institute of General Medical Sciences (including P41GM103393 and P30GM133894). The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of NIGMS or NIH.

References

  1. Clarkson TW, 1998. Human toxicology of mercury. J. Tr. Elem. Exp. Med. 11(2–3), 303–317. [Google Scholar]
  2. Desrosiers M, Planas D, Mucci A, 2006. Mercury methylation in the epilithon of boreal shield aquatic ecosystems. Environ. Sci. Technol, 40(5), 1540–1546. [DOI] [PubMed] [Google Scholar]
  3. Devabathini P, Fischer DL, Chen S, Pattammattel A, Bury G, Antipova O, Huang X, Chu YS, Horai S, Pushkar Y, 2023. High-resolution imaging of Hg/Se aggregates in the brain of small Indian mongoose, a wild terrestrial species: insights into intracellular Hg detoxification. Environ. Chem. Lett, DOI: 10.1007/s10311-023-01666-3 [DOI] [Google Scholar]
  4. Dickson FW, Tunnell G, 1959. The stability relations of cinnabar and metacinnabar. Am Minerol. 44(5–6), 471–487. [Google Scholar]
  5. Dolgova NV, Nehzati S, MacDonald TC, Summers KL, Crawford AM, Krone PH, George GN, Pickering IJ, 2019. Disruption of selenium transport and function is a major contributor to mercury toxicity in zebrafish larvae. Metallomics, 11(3), 621–631. [DOI] [PubMed] [Google Scholar]
  6. Eto K, Marumoto M, Takeya M, 2010. The pathology of methylmercury poisoning (Minamata disease). Neuropathology 30(5), 471–479. [DOI] [PubMed] [Google Scholar]
  7. Eto K, Yasutake A, Nakano A, Akagi H, Tokunaga H, Kojima T, 2001. Reappraisal of the historic 1959 cat experiment in Minamata by the Chisso Factory. Tohoku. J. Exp. Med. 194(4), 197–203. [DOI] [PubMed] [Google Scholar]
  8. Gailer J, George GN, Pickering IJ, Madden S, Prince RC, Yu EY, Denton MB, Younis HS, Aposhian HV, 2000. Structural basis of the antagonism between inorganic mercury and selenium in mammals. Chem. Res. Toxicol. 13(11), 1135–1142. [DOI] [PubMed] [Google Scholar]
  9. George GN, Singh SP, Hoover J, Pickering IJ 2009. The chemical forms of mercury in aged and fresh dental amalgam surfaces. Chem. Res. Toxicol. 22(11), 1761–1764. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. George GN, Singh SP, Prince RC, Pickering IJ, 2008. The chemical forms of mercury and selenium in fish following digestion with simulated gastric fluid. Chem. Res. Toxicol. 21(11), 2106–2110. [DOI] [PubMed] [Google Scholar]
  11. Grandjean P, Budtz-Jørgensen E, Jørgensen PJ, Weihe P, 2005. Umbilical cord mercury concentration as biomarker of prenatal exposure to methylmercury. Environ. Health. Perspect. 113(7), 905–908. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Halfpenny J, Small RWH, 1979., Structure of (chloromercurio)acetaldehyde. Acta Cryst. B 35, 1239–1241. [Google Scholar]
  13. Harada M, 1978. Congenital Minamata disease: intrauterine methylmercury poisoning. Teratology, 18(2), 285–288. [DOI] [PubMed] [Google Scholar]
  14. Harris HH, Pickering IJ, George GN, 2003. The chemical form of mercury in fish. Science, 381(5637), 1203. [DOI] [PubMed] [Google Scholar]
  15. James AK, Dolgova NV, Nehzati S, Korbas M, Cotelesage JJH, Sokaras D, Kroll T, O’Donoghue JL, Watson GE, Myers GJ, Pickering IJ, George GN, 2022. Molecular fates of organometallic mercury in human brain. ACS Chem. Neurosci. 13(12), 1756–1768. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. James AK, Nehzati S, Dolgova NV, Sokaras D, Kroll T, Eto K, O’Donoghue JL, Watson GE, Myers GJ, Krone PH, Pickering IJ, George GN, 2020. Rethinking the Minamata tragedy: What mercury species was really responsible? Environ. Sci. Technol. 54(5), 2726–2733. [DOI] [PubMed] [Google Scholar]
  17. James AK, Popescu BF, Weng M, Myers GJ, O’Donoghue JL, Watson GE, Pickering IJ, George GN, 2023. Synchrotron X-ray methods in the study of mercury neurotoxicology. Neurotoxicology, 99, 129–138. [DOI] [PubMed] [Google Scholar]
  18. Klaassen CD, 2013. Toxic responses of the ocular and visual system. In Casarett & Doull’s Toxicology: The Basic Science of Poisons; Fox DA, Boyes WK, Eds.; McGraw-Hill: New York, 2013, Ed. 8. [Google Scholar]
  19. Korbas M, O’Donoghue JL, Watson GE, Pickering IJ, Singh SP, Myers GJ, Clarkson TW, George GN, 2010. The chemical nature of mercury in human brain following poisoning or environmental exposure. ACS Chem. Neurosci. 1(12), 810–818. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kurland LT, Faro SN, Siedler H, 1960. Minamata disease. The outbreak of a neurologic disorder in Minamata, Japan, and its relationship to the ingestion of seafood contaminated by mercuric compounds. World Neurol. 1, 370–395. [PubMed] [Google Scholar]
  21. MacDonald TC, Korbas M, James AK, Sylvain NJ, Hackett MJ, Nehzati S, Krone PH, George GN, Pickering IJ, 2015. Interaction of mercury and selenium in the larval stage zebrafish vertebrate model. Metallomics, 7(8), 1247–1255. [DOI] [PubMed] [Google Scholar]
  22. Manceau A, Bourdineaud J-P, Oliveira RB, Sarrazin SLF, Krabbenhoft DP, Eagles-Smith CA, Ackerman JT, Stewart AR, Ward-Deitrich C, del Castillo Busto ME, Goenaga-Infante H, Wack A, Retegan M, Detlefs B, Glatzel P, Bustamante P, Nagy KL, Poulin BA, 2021. Demethylation of methylmercury in bird, fish, and earthworm. Environ. Sci. Technol. 55(3), 1527–1534. [DOI] [PubMed] [Google Scholar]
  23. Manceau A, Gaillot A-C, Glatzel P, Cherel Y, Bustamante P, 2021. In vivo formation of HgSe nanoparticles and Hg–tetraselenolate complex from methylmercury in seabirds–Implications for the Hg–Se antagonism. Environ. Sci. Technol. 55(3), 1515–1526. [DOI] [PubMed] [Google Scholar]
  24. Martoja R, Berry JP, 1980. Identification of tiemannite as a probable product of demethylation of mercury by selenium in cetaceans. A complement to the scheme of the biological cycle of mercury, Vie Milieu 30, 7–10. [Google Scholar]
  25. McMaster WH, Del Grande NK, Mallett JH, Hubbell JH, 1969. Compilation of X-ray cross sections. Lawrence Livermore National Laboratory Report UCRL-50174, Sec II, Rev.1. [Google Scholar]
  26. Nehzati S, Dolgova NV, James AK, Cotelesage JJH, Sokaras D, Kroll T, George GN, Pickering IJ, 2021. High energy resolution fluorescence detected X-ray absorption spectroscopy: An analytical method for selenium speciation. Anal. Chem. 93(26), 9235–9243. [DOI] [PubMed] [Google Scholar]
  27. Nehzati S, Dolgova NV, Young CG, James AK, Cotelesage JJH, Sokaras D, Kroll T, Qureshi M, Pickering IJ, George GN, 2022. Mercury Lα1 High energy resolution fluorescence detected X-ray absorption spectroscopy: A versatile speciation probe for mercury. Inorg. Chem. 61(13), 5201–5214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. O’Donoghue JL, Watson GE, Brewer R, Zareba G, Eto K, Takahashi H, Marumoto M, Love T, Harrington D, Myers GJ, 2020. Neuropathology associated with exposure to different concentrations and species of mercury: A review of autopsy cases and the literature. Neurotoxicology 78, 88–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Othmer DF, Kon K, Igarashi T, 1956. Acetaldehyde by the Chisso process. Ind. Eng. Chem, 48(8), 1258–1262. [Google Scholar]
  30. Ralston NVC, Ralston CR, Raymond LJ, 2016. Selenium health benefit values: Updated criteria for mercury risk assessment. Biol. Trace Elem. Res, 171(2), 262–269. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Sakamoto M, Kaneoka T, Murata K, Nakai K, Satoh H, Akagi H, 2007. Correlations between mercury concentrations in umbilical cord tissue and other biomarkers of fetal exposure to methylmercury in the Japanese population. Environmental Research, 103(1), 106–111. [DOI] [PubMed] [Google Scholar]
  32. Saitō H, Maruyama K, Hagino N, 2015. Sensory disturbance as a characteristic symptom of methylmercury poisoning in Niigata, Japan J. Env. Sci. Eng. A 4, 532–540. [Google Scholar]
  33. Saitō H, 2009. Niigata Minamata Disease: Methyl Mercury Poisoning in Niigata, Japan. Niigata Nippō (English translation and editing Smith AM., George TS., Kanaya N., Rodgers K., Katagiri MA., Palmer M.). [Google Scholar]
  34. Saitō H, Sekikawa T, Taguchi J, Shozawa T, Kinoshita Y, Matsumara K, Yanagihara K, Nikaido K, Urasaki S, Imaizumi H, Hatano H 2020. Prenatal and postnatal methyl mercury exposure in Niigata, Japan: adult outcomes. Neurotoxicology 81, 364–372. [DOI] [PubMed] [Google Scholar]
  35. Selin H, Stone PK, 2009. Childbirth across cultures ideas and practices of pregnancy, childbirth and the postpartum. Dordrecht New York: Springer. ISBN 9789048125982. [Google Scholar]
  36. Simmons-Willis TA, Koh AS, Clarkson TW, Ballatori N, 2002. Transport of a neurotoxicant by molecular mimicry: the methylmercury-L-cysteine complex is a substrate for the human L-type large neutral amino acid transporter (LAT) 1 and LAT2. Biochem J, 367(1), 239–246. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Sokaras D, Weng T-C, Nordlund D, Alonso-Mori R, Velikov P, Wenger D, Garachtchenko A, George MJ, Borzenets V, Johnson B, Rabedeau T, Bergmann U, 2013. A seven-crystal Johann-type hard X-ray spectrometer at the Stanford Synchrotron Radiation Lightsource. Rev. Sci. Instrum. 84(5), 053102/1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Spiller HA, 2018. Rethinking mercury: the role of selenium in the pathophysiology of mercury toxicity. Clin. Toxicol. 56(5), 313–326. [DOI] [PubMed] [Google Scholar]
  39. Takeuchi T, Morikawa N, Matsumoto H, Shiraishi Y, 1962. A pathological study of Minamata disease in Japan. Acta Neuropath. 2, 40–57. [Google Scholar]
  40. Taylor D, 1982. Minamata disease. Environ Sci. Technol. 16(2), 81A. [DOI] [PubMed] [Google Scholar]
  41. Turco MY, Moffett A, 2019, Development of the human placenta. 2019. Development. 146(22), dev163428/1–14. [DOI] [PubMed] [Google Scholar]
  42. Yokoyama H, 2018. Mercury Pollution in Minamata (Springer Briefs in Environmental Science, Springer; ). [Google Scholar]
  43. Zupo V, Graber G, Kamel S, Plichta V, Granitzer S, Gundacker C, Wittmann KJ 2019. Mercury accumulation in freshwater and marine fish from the wild and from aquaculture ponds. Environ. Pollut. 255(1), 112975/1–13. [DOI] [PubMed] [Google Scholar]

RESOURCES