2026 Volume 51 Issue 4 Pages 259-266
Methylmercury (MeHg) is a well-established environmental neurotoxicant and the primary cause of Minamata disease. Recently, α-mercuri-acetaldehyde (HgCH2CHO) and α-mercuri-acetic acid (HgCH2COOH) have been proposed to be relevant to the Minamata tragedy. However, their in vivo toxicity has not been compared with that of MeHg under identical conditions. We conducted a comparative in vivo toxicity study of HgCH2CHO and HgCH2COOH using the dosing regimen previously established for MeHg. Male Wistar rats were orally administered MeHg, HgCH2CHO, or HgCH2COOH (26.6 µmol/kg/day) for 5 days, which was followed by a 2-day drug-free period and a single repeat of this cycle. Systemic toxicity was evaluated using the change in body mass, neurobehavioral effects were assessed using the hindlimb crossing test, and the total mercury accumulation in blood and organs was quantified. MeHg exposure resulted in marked weight loss and significant neurobehavioral impairment. In contrast, rats exposed to HgCH2CHO or HgCH2COOH exhibited only mild weight loss and substantially attenuated hindlimb crossing responses. The total mercury levels in the blood, liver, brain, muscle, and spleen were considerably lower in the HgCH2CHO- and HgCH2COOH-treated groups than in the MeHg group. The renal accumulation of mercury did not differ among the groups, despite the blood mercury levels in the HgCH2CHO and HgCH2COOH groups being extremely low, suggesting distinct toxicokinetic properties. Overall, HgCH2CHO and HgCH2COOH demonstrated much lower in vivo toxicity and systemic mercury burden than MeHg under equivalent dosing conditions. These findings do not challenge the established role of MeHg as the primary causative agent of Minamata disease.
Methylmercury (MeHg) is a well-recognized environmental neurotoxicant, causing severe neurological impairment of both the central and peripheral nervous systems, and the principal cause of Minamata disease (Eto and Takeuchi, 1977; Eto and Takeuchi, 1978; Takeuchi et al., 1978; Eto et al., 2010). Over recent decades, extensive research has been conducted to elucidate the mechanisms underlying MeHg-induced neurotoxicity (Farina et al., 2011; Antunes dos Santos et al., 2016; Novo et al., 2021; Shinoda et al., 2023; Leal-Nazaré et al., 2024). In 2020, James and colleagues re-examined the Minamata tragedy and suggested that α-mercuri-acetaldehyde (HgCH2CHO) and α-mercuri-acetic acid (HgCH2COOH) may be novel causes of Minamata disease (James et al., 2020). This study was subsequently featured in a news report in Science and generated broad interest in the field (Sokol, 2020). Recently, we reported that the neurotoxicity of HgCH2CHO and HgCH2COOH, evaluated in vitro using human neuroblastoma SH-SY5Y cells, is approximately 2–5 times lower than that of MeHg (Yamashiro et al., 2026). These compounds also caused less cellular accumulation, suggesting that they are less neurotoxic than MeHg. However, their in vivo toxicities have not been directly compared with that of MeHg.
Body mass change is an endpoint that is routinely included in animal toxicity assessments (OECD, 2002; Mostert et al., 2021), and weight loss has been reported in previous in vivo studies of exposure to MeHg (Shigematsu et al., 2000; Ferrer et al., 2021). The hindlimb crossing/clasping behavior of rodents is assessed as an indicator of neurobehavioral impairment and motor dysfunction, and this has been widely used in models of neurodegeneration and neurotoxicity (Guyenet et al., 2010; Lalonde and Strazielle, 2011; Yerger et al., 2022). Hindlimb crossing is known to be induced by various defects, including dysfunction of the basal ganglia/dopaminergic system, cerebellar ataxia, impairment of the corticospinal tract and spinal motor circuits, and peripheral neuropathy and muscle weakness (Lalonde and Strazielle, 2011). Although this phenotype cannot be attributed to a single underlying cause, it represents a simple and practical behavioral endpoint and has also been assessed in rodents following MeHg exposure (Chakrabarti and Bai, 2000; Fujimura et al., 2011). The tissue/organ distribution and accumulation of hazardous substances is considered to be a fundamental component of the assessment of their toxicity (OECD, 2010), including that of mercury compounds (Rodrigues et al., 2010; ATSDR, 2024). Indeed, the mercury levels of autopsy specimens from patients with Minamata disease have been measured, and characteristic patterns of mercury distribution have been reported (Marumoto et al., 2020; Sakamoto et al., 2025).
In the present study, we performed an in vivo comparative toxicity assessment of HgCH2CHO and HgCH2COOH using the same dosing regimen as for MeHg, in terms of dose, route, and duration. We compared the body mass changes, hindlimb crossing behavior, and tissue/organ mercury accumulation of rats administered each of these compounds.
Methylmercury chloride was purchased from Merck (Rahway, NJ, USA). α-Mercuri-acetaldehyde (HgCH2CHO), formally known as (2-oxoethyl)mercury(II) chloride, and α-mercuri-acetic acid (HgCH2COOH), formally known as (carboxymethyl)mercury(II) chloride, were synthesized as previously described (Yamashiro et al., 2026). MeHg (7.97 mM), HgCH2CHO, and HgCH2COOH (5 mM each) were dissolved in sterile water by sonication for 1 hr. For HgCH2COOH, 1 N NaOH was added to facilitate its complete dissolution, resulting in a final NaOH concentration of 0.1% (v/v) in the 5 mM solution.
Animals and the administration of mercury compoundsEight-week-old male Wistar rats were purchased from Japan SLC Inc. (Hamamatsu, Japan) and housed under standard laboratory conditions under a 12-hr light/dark cycle and with ad libitum access to food and water. Beginning at 9 weeks of age, MeHg, HgCH2CHO, or HgCH2COOH solutions were administered via gastric gavage at a dose of 26.6 µmol/kg/day for 5 consecutive days, which was followed by a 2-day drug-free period. This dosing schedule was then repeated once. Age-matched control rats were administered an equivalent volume of water per kg body mass.
A total of 31 rats were used in the study: 17 rats for the hindlimb crossing test (5 from the MeHg group and 4 each from the control, HgCH2CHO, and HgCH2COOH groups) and 14 rats for the total mercury content analysis (5 from the MeHg and HgCH2CHO groups and 4 from the HgCH2COOH group). All the experimental procedures were conducted in accordance with the Regulations for Animal Research of Tokyo University of Pharmacy and Life Sciences and were approved by the institutional ethics committee (Approval Nos.: P20-19 (April 13, 2020), P21-07 (April 19, 2021), P22-70 (May 6, 2022), P23-10 (May 8, 2023), P24-08 (May 16, 2024), P25-14 (May 19, 2025)). Every effort was made to minimize the number of animals used and to reduce any discomfort or distress.
Hindlimb crossing testThe hindlimb crossing test was performed as previously described (Shinoda et al., 2021). Briefly, the rat was gently grasped by the tail and suspended approximately 50 cm above its cage. Its hindlimb posture was evaluated as follows: complete crossing or bending of both hindlimbs was scored as 3; bending of one hindlimb was scored as 2; a hindlimb abduction angle <90° was scored as 1; and a hindlimb abduction angle >90° was scored as 0.
Total mercury accumulationTotal mercury accumulation in tissues was measured using the method previously reported (Nakano et al., 2024). On day 14 after the start of dosing, the rats were deeply anesthetized with carbon dioxide, and blood was collected by cardiac puncture. Their spleen, liver, kidneys, femoral muscles, and brain were rapidly excised, weighed immediately, and then stored at −80°C until analysis. The tissue mercury level in each (m/z = 202) was quantified using inductively coupled plasma mass spectrometry (ICP-MS; NexION 300S, PerkinElmer, Waltham, MA, USA).
Statistical analysisStatistical analyses were performed using Statcel5 software (OMS, Tokyo, Japan) and Microsoft Excel (Microsoft, Redmond, WA, USA), except for the results of the hindlimb crossing test. Data are expressed as the mean ± standard deviation (SD). Body mass data were analyzed using two-way repeated-measures analysis of variance (ANOVA), followed by the Tukey–Kramer post-hoc test. Between-group differences in body mass on the final day were assessed using one-way ANOVA, followed by the Tukey–Kramer post-hoc test. Hindlimb crossing test data (ordinal scores measured repeatedly over time) were analyzed using a cumulative link mixed model with a logit link function in R (version 4.5.2; package ordinal; R Foundation for Statistical Computing, Vienna, Austria). Treatment, time (day, centered), and their interaction were included as fixed effects, and the identity of the animal was included as a random intercept. Model selection was based on likelihood ratio tests. Mercury accumulation was analyzed using one-way ANOVA, followed by the Tukey–Kramer post-hoc test. p values less than 0.05 were considered statistically significant.
Use of artificial intelligence toolsR code was generated and revised using ChatGPT (version 5.2, OpenAI, San Francisco, CA, USA). The authors reviewed and are responsible for all the content.
Body mass change is routinely evaluated as a general indicator of systemic toxicity in repeated-dose studies (OECD, 2002; Mostert et al., 2021). In addition, hindlimb crossing is a characteristic phenotype in rodents exposed to MeHg (Chakrabarti and Bai, 2000; Takahashi et al., 2017). Therefore, we administered each of the two novel mercury compounds using the experimental protocol previously employed in our studies of the mechanisms of MeHg-induced toxicity, with respect to the concentration, treatment duration, and route of administration (Shinoda et al., 2019), and compared the changes in body mass and hindlimb crossing scores of the three groups. The MeHg-exposed group exhibited a pronounced decrease in body mass during the dosing period, as previously reported, and only a modest recovery after the end of the treatment (Fig. 1A). In contrast, the HgCH2CHO- and HgCH2COOH-exposed groups lost less weight, but still tended to have lower body masses than the control group (Fig. 1A). In the hindlimb crossing test, the MeHg-exposed group exhibited high scores, beginning immediately after the end of the exposure period (Fig. 1B). Under the same conditions, both the HgCH2CHO and HgCH2COOH groups showed significantly lower scores and much smaller increases in the scores over time than the MeHg group.
Body masses and hindlimb crossing test results for rats administered mercury compounds. (A) Time course of normalized body mass following treatment with the mercury compounds. Two-way repeated-measures ANOVA was used for the analysis. Significant main effects of treatment (F(3,12) = 37.75, p < 0.00001) and time (F(16,192) = 176.81, p < 0.0001), as well as a significant treatment-by-time interaction (F(48,192) = 27.27, p < 0.0001), were identified. Between-group differences in body mass on the final day were assessed using one-way ANOVA followed by the Tukey–Kramer post-hoc test. *p < 0.05, **p <0.01 vs. control. ++p < 0.01 vs. MeHg. Data are presented as mean ± SD. (B) Time course of the predicted severity score on the hindlimb crossing test following exposure to mercury compounds. The predicted probabilities of severe neurological impairment (score ≥2) are shown for rats treated with MeHg, HgCH2CHO, or HgCH2COOH over time. Probabilities were estimated using a cumulative link mixed model with a logit link function, including treatment, time (day, centered), and their interaction as fixed effects, and the identity of the rat as a random intercept. A significant treatment-by-time interaction was detected (likelihood ratio test: LR = 92.05, df = 2, p < 0.0001). The lines represent model-based population-level predictions.
Next, we evaluated the mercury accumulation in blood and organs following exposure to each mercury compound. On day 14 after the start of the exposure, the kidneys, blood, liver, brain, muscles, and spleen were collected from each rat, and the mercury level in each was quantified. In all the tissues/organs except the kidneys of rats in the HgCH2CHO- and HgCH2COOH-exposed groups, mercury accumulation was significantly lower than that in the tissues/organs of the MeHg-exposed group (Fig. 2). In contrast, there were no significant differences in the renal mercury accumulation in the exposure groups.
Mercury concentrations in the blood and organs of the rats. Concentrations in the (A) kidney, (B) blood, (C) liver, (D) brain, (E) muscle, and (F) spleen. The data were analyzed using one-way ANOVA, followed by the Tukey–Kramer post-hoc test. **p < 0.01 vs. MeHg. Data are presented as mean ± SD.
We conducted an in vivo comparative toxicity study of the mercury compounds HgCH2CHO, HgCH2COOH, and MeHg. Compared with MeHg, these compounds induced (1) a smaller reduction in body mass, (2) almost no detectable hindlimb crossing behavior, and (3) substantially lower mercury accumulation in multiple tissues/organs, but not in the kidney. Overall, these findings indicate that the in vivo toxicity and systemic mercury burden associated with these compounds are considerably lower than those for MeHg, consistent with our previous in vitro findings (Yamashiro et al., 2026).
The rats administered the novel mercury compounds showed much less weight loss than those exposed to MeHg under the same dosing conditions. This observation is consistent with our in vitro findings and supports the notion that these compounds are less toxic than MeHg (Yamashiro et al., 2026). However, analysis of the body masses of the rats on the final day (day 67) indicated that HgCH2CHO did induce a significant reduction in body mass. Therefore, although the established MeHg exposure protocol was used and the toxicity of these compounds appeared relatively mild under these conditions, the results suggest that HgCH2CHO does have measurable toxic effects in vivo.
According to the results of the hindlimb crossing test, neither of the two novel mercury compounds caused overt motor impairment. In contrast, MeHg induced significant hindlimb crossing, consistent with previous reports (Chakrabarti and Bai, 2000; Fujimura et al., 2011; Shinoda et al., 2021). It should be noted, however, that Fig. 1B presents the predicted probability of obtaining a score ≥2, which may give the impression that the novel compounds had little or no effect. In fact, in the groups treated with these two compounds, a few animals exhibited a very mild response, with a score of 1. As for the body mass results, this observation suggests that although the toxicity of these compounds is much lower than that of MeHg, they may be somewhat toxic in vivo.
It has been reported that patients with acute Minamata disease have very high mercury concentrations in their liver and kidneys, whereas those in the brain are lower (Marumoto et al., 2020; Sakamoto et al., 2025). A similar distribution pattern has also been identified in MeHg exposure studies performed in rodents (Rodrigues et al., 2010; Nakano et al., 2024). In the present study, MeHg administration was associated with a tissue mercury distribution profile that is consistent with those characterized in previous studies. In contrast, the total mercury accumulation in the HgCH2CHO- and HgCH2COOH-treated rats was extremely low in all the analyzed tissues/organs, in contrast to those treated with MeHg, except in the kidney. The total mercury accumulation levels were approximately 30–7,000 times lower in the HgCH2CHO group, and 20–1,500 times lower in the HgCH2COOH group, than in the MeHg group. It is noteworthy that the amounts of total mercury accumulated in the kidneys of rats treated with the two novel mercury compounds tended to be as high as that observed for MeHg administration, whereas their blood mercury concentrations were extremely low. This suggests that, although all of these substances are organomercury compounds, they exhibit markedly different toxicokinetic behaviors, such as rapid clearance from the systemic circulation and preferential renal accumulation.
One plausible explanation for this is that HgCH2CHO and HgCH2COOH, unlike MeHg, may not be retained in blood, owing to weaker binding to erythrocytes or plasma proteins, leading to more rapid renal elimination. In addition, their chemical structures may favor active uptake into renal tubular cells via organic anion transporters or related uptake systems, leading to disproportionate kidney accumulation, despite very low circulating mercury levels, a pattern that has been reported for other small, polar organometallic species. Similar dissociations of the blood and kidney concentrations have been described for inorganic mercury and certain low-molecular-weight mercury conjugates that undergo rapid glomerular filtration followed by tubular reabsorption and intracellular sequestration in the kidney (Zalups, 2000; Bridges and Zalups, 2010). However, the physicochemical properties of HgCH2CHO and HgCH2COOH have been poorly characterized, and therefore further studies are required to clarify their stability, protein-binding characteristics, and interactions with renal transport mechanisms. Based on our experience, the water solubilities of HgCH2CHO and HgCH2COOH are much lower than that of MeHg, and their lipophilicity is not particularly high. These differences in physicochemical properties are also likely to contribute to the observed differences in the distribution profiles of these compounds.
In the present study, we conducted a comparative in vivo toxicological evaluation of two mercury compounds, HgCH2CHO and HgCH2COOH, which were proposed to be involved in the etiology of Minamata disease by James et al. (2020). This work follows on from our previous in vitro study, in which we compared their cytotoxicity with that of MeHg (Yamashiro et al., 2026). As also noted in our previous in vitro study, the present study also does not provide evidence that these two compounds were present in effluents from the Chisso Minamata factory. Moreover, the in vivo assessment was designed to provide a direct comparison with MeHg exposure, using the same protocol that we previously established for MeHg. Therefore, we did not evaluate the effects of higher doses of or longer-term exposures to HgCH2CHO and HgCH2COOH. In addition, the mechanisms underlying their toxicity, as well as whether or not they induce neurotoxicity, remain unclear. However, as described above, mild neurotoxicity was suggested by the hindlimb crossing scores of the rats that were noted under the experimental conditions of the present study. Therefore, more comprehensive neuropathological examinations should be performed after higher levels of exposure of rats to these compounds in future studies.
Importantly, none of our findings are inconsistent with the established role of MeHg as the principal etiologic agent of Minamata disease. Although it was MeHg that was detected in factory effluent, it remains possible that HgCH2CHO and HgCH2COOH could have been intermediate byproducts of MeHg production. Therefore, it may be important to process these compounds using the same extraction and identification procedures that were historically used for MeHg (Uchida et al., 1961a, 1961b; Irukayama et al., 1962), to determine whether their presence could have been overlooked. If such analyses do not conclusively exclude the presence of these compounds, more detailed toxicological investigations would be warranted to clarify their potential contributions to the overall toxicological profile of relevant factory effluents.
We thank Mark Cleasby, PhD from Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.
FundingNo funding was provided for the work.
Conflict of interestThe authors declare no competing interests.
Data availabilityContact the corresponding authors directly to request the underlying data.
Author contributionsConceptualization: Yo Shinoda
Formal analysis: Yo Shinoda, and Eiko Yoshida
Investigation: Yo Shinoda, Eiko Yoshida, Sachie Arae, Ryo Irie, and Yuuki Fujimoto
Supervision: Yo Shinoda, and Toshiyuki Kaji
Visualization: Yo Shinoda, and Eiko Yoshida
Writing – original draft: Yo Shinoda
Writing – review & editing: Yo Shinoda, Eiko Yoshida, Kaito Yamashiro, Tsutomu Takahashi, Yasuyuki Fujiwara, and Toshiyuki Kaji
Ethical approval and consent to participateNot applicable.
Patient consent for publicationNot applicable.
