2019 年 44 巻 3 号 p. 191-199
Exposure to organic mercury, especially methylmercury (MeHg), causes Minamata disease, a severe chronic neurological disorder. Minamata disease predominantly affects the central nervous system, and therefore, studies on the mechanisms of MeHg neurotoxicity have focused primarily on the brain. Although the peripheral nervous system is also affected by the organometallic compound and shows signs of neural degeneration, the mechanisms of peripheral MeHg neurotoxicity remain unclear. In the present study, we performed quantitative immunohistochemical analyses of the dorsal root ganglion (DRG) and associated sensory and motor fibers to clarify the mechanisms of MeHg-induced peripheral neurotoxicity in Wistar rats. Methylmercury chloride (6.7 mg/kg/day) was orally administrated for 5 days, followed by 2 days without administration, and this cycle was repeated once again. Seven and 14 days after the beginning of MeHg exposure, rats were anesthetized, and their DRGs and sensory and motor nerve fibers were removed and processed for immunohistochemical analyses. The frozen sections were immunostained for neuronal, Schwann cell, microglial and macrophage markers. DRG sensory neuron somata and axons showed significant degeneration on day 14. At the same time, an accumulation of microglia and the infiltration of macrophages were observed in the DRGs and sensory nerve fibers. In addition, MeHg caused significant Schwann cell proliferation in the sensory nerve fibers. In comparison, there was no noticeable change in the motor fibers. Our findings suggest that in the peripheral nervous system, MeHg toxicity is associated with neurodegenerative changes to DRG sensory neurons and the induction of a neuroprotective and/or enhancement of neurodegenerative host response.
Methylmercury (MeHg) is an organic mercury compound that causes Minamata disease (Eto, 1997; McAlpine and Araki, 1958). Patients with Minamata disease typically exhibit ataxia, dysarthria, visual impairment, acoustic disturbance and hypesthesia (Harada, 1995). These symptoms are thought to be caused by damage to the central nervous system, and accordingly, the neurotoxic effects of MeHg have been studied primarily in the brain (Antunes Dos Santos et al., 2016; Costa et al., 2004; Unoki et al., 2018; Patel and Reynolds, 2013; Farina et al., 2011). Despite observations of peripheral nerve injury in humans, the main features and mechanisms of peripheral MeHg neurotoxicity remain unclear.
Pathological studies show that severe axonal degeneration, myelin degradation and disappearance of dorsal root ganglion (DRG) neurons are major features in Minamata disease patients (Eto et al., 2002; Takeuchi et al., 1978; Eto and Takeuchi, 1978). Similar pathological changes are found in MeHg-exposed rats, including extensive DRG neuronal degeneration (Delio et al., 1992; Sakamoto et al., 1998; Cao et al., 2013), and damage to sensory axons (Cavanagh and Chen, 1971; Miyakawa et al., 1974; Yip and Chang, 1981; Arimura et al., 1988; Cao et al., 2013) and myelin (Miyakawa et al., 1970; Munro et al., 1980; Yip and Chang, 1981; Cao et al., 2013). However, most of these studies lack appropriate controls and do not present quantitative data. Therefore, in the present study, we performed quantitative immunohistochemical analyses to clarify the primary features and mechanisms of MeHg-induced peripheral neurotoxicity using DRGs and their sensory and motor nerve fibers in Wistar rats.
All experimental protocols were evaluated and approved by the Regulations for Animal Research at Tokyo University of Pharmacy and Life Sciences. All efforts were made to minimize the number of animals used and their suffering. Eight-week-old male Wistar rats (Tokyo Laboratory Animals Science, Tokyo, Japan) were housed in cages under a 12/12-hr light-dark cycle, with ad libitum access to water and food. Methylmercuric chloride (MeHgCl; Sigma-Aldrich, Tokyo, Japan) was dissolved in MilliQ water to a concentration of 2 mg/mL. The MeHg solution was administrated orally using a gastric tube at a daily dosage of 6.7 mg/kg for 5 days, followed by no administration for 2 days (to obtain the day 7 tissue samples). This cycle was repeated once to obtain the day 14 samples (Fig. 1A). Age-matched rats given no treatment were used as controls.
MeHg-induced neural cell death in the rat DRG. (A) The MeHg administration protocol. (B) Representative images of DRGs immunostained for the neuronal marker NeuN. (C) The number of neurons revealed by NeuN immunostaining. Note that the number of neurons was significantly decreased on Day 14, but not Day 7 (n = 3). *p < 0.05 compared with control and Day 7.
Immunohistochemistry was performed as described previously with minor modification (Sadakata et al., 2014). Briefly, deeply anesthetized control, day 7 and day 14 rats were transcardially perfused, initially with PBS and then with 4% PFA in 0.1 M phosphate buffer. Then, the DRGs and sensory and motor nerves were removed from the lumbar region. Tissues were postfixed in 4% PFA in 0.1 M phosphate buffer overnight at 4°C, then cryoprotected by immersion in 20% sucrose in PBS overnight at 4°C. Tissues were embedded in Tissue-Tek OCT compound (Sakura Finetek, Tokyo, Japan) and frozen at −78°C. Frozen tissues were sectioned at a thickness of 14 µm using a cryostat (HM550; Thermo Fisher Scientific, Tokyo, Japan) at −18°C. Sections were air-dried overnight at 37°C, and then stored at −80°C until use. Sections were washed in PBS and blocked in antibody buffer (2 × PBS containing 2% donkey serum, 0.1% Triton X-100 and 0.05% NaN3) for 30 min. Sections were then incubated with primary antibodies at room temperature for 1 hr or overnight at 4°C, washed with PBS, and then incubated with secondary antibodies at room temperature for 1 hr. Stained sections were mounted in Hoechst 33342-containing Fluoromount/Plus (Thermo Fisher Scientific) and observed under a fluorescence microscope (TS-100; Nikon, Tokyo, Japan) equipped with a CMOS camera (Zyla5.5; ANDOR, Tokyo, Japan). Images were taken and processed with NIS-Elements software (Nikon, Tokyo, Japan) and analyzed with ImageJ software (NIH, MD, USA). The primary antibodies used were: CD31 (BD Biosciences, Tokyo, Japan; 1:1,000), CD68 (Bio-Rad, CA, USA; 1:1,000), GFAP (Thermo Fisher Scientific; 1:1,000), Iba1 (Abcam, Cambridge, UK; 1:1,000), Ki67 (Abcam; 1:1,000), NeuN (Merck Millipore, MA, USA; 1:1,000), neurofilament (Sigma-Aldrich; 1:1,000), MBP (IBL, Gunma, Japan; 1:1,000), SOX10 (R&D Systems, MN, USA; 1:1,000) and Vimentin (Abcam; 1:1,000). The secondary antibodies used were: Alexa Fluor 488, 555 or 647-conjugated goat anti-mouse, anti-rabbit or anti-chicken IgG (Thermo Fisher Scientific; 1:2,000).
StatisticsAll statistics were performed using Excel software (Microsoft, Redmond, WA, USA) with the add-in software Statcel (OMS, Tokyo, Japan). Data are expressed as the mean ± SEM. Differences between multiple data sets were assessed using one-way ANOVA with post-hoc Tukey–Kramer test. All data were collected and analyzed using a double-blind approach.
First, we examined the effects of MeHg exposure on the rat DRG. MeHg was orally administrated daily and rats were sacrificed at two time points (days 7 and 14; Fig. 1A). Neurons were quantified using the neuronal marker NeuN (Fig. 1B and C). MeHg did not induce any neural degeneration by 1 week (Day 7), but did result in significant neural degeneration by 2 weeks (Day 14; Fig. 1B and C). These results suggest that under our experimental conditions, MeHg induces neural degeneration in the rat DRG within 2 weeks. Next, we examined for the presence of axonal damage in sensory and motor neurons (Fig. 2A and B). Sensory nerve fibers were severely damaged on day 14. In contrast, motor nerve fibers were not affected by MeHg (Fig. 2). The number of sensory axons, but not motor axons, was significantly decreased on day 14 (Fig. 2C). In addition, Wallerian degeneration-like changes were observed in sensory fibers on day 14 (Fig. 2A, B and D). These results suggest that the MeHg administration protocol used here results in sensory neuron-specific injury.
MeHg-induced neural injury was observed in the sensory, but not motor, nerve. (A) Representative images of longitudinal sections of sensory (left) and motor (right) fibers immunostained for the axonal marker neurofilament (NF) and the myelin marker MBP. Note that sensory, but not motor, axons had severely degenerated by Day 14. (B) Representative images of transverse sections of sensory (left) and motor (right) fibers. (C) The number of axons counted in transverse sections was significantly decreased in sensory fibers on Day 14 (n = 3). **p < 0.01 compared with control and Day 7. (D) The average area of axons in transverse sections was increased in sensory fibers on Day 14 (n = 3). *p < 0.05 compared with control.
Generally, damaged neural tissues express and release cytokines that activate microglia. These cytokines also attract macrophages to the damaged tissue. Therefore, we investigated whether microglia and macrophages are activated or recruited to the DRGs following MeHg exposure. The number of Iba1-positive microglia/macrophages in the DRGs was unchanged on day 7 compared with control; however, a significant increase in these cells was observed on day 14 (Fig. 3A and B). Interestingly, Iba1-positive cells accumulated in certain areas, especially on day 14 (Fig. 3A and C). To distinguish between microglia and macrophages, we used a relatively macrophage-specific marker, CD68, together with Iba1. CD68 is known to express in both activated microglia and macrophage, but the expression level in macrophage is approximately 10-fold higher than microglia. In the present study, we chose the cells strongly stained with anti-CD68 antibody as macrophages. All of the CD68-positive cells expressed Iba1 (data not shown), and the number of CD68-positive macrophages was also increased by MeHg administration (Fig. 3A and D). Microglia were approximately 10-fold more numerous than macrophages (Fig. 3B and D). These results suggest that MeHg increases the number of microglia/macrophages in the DRG.
MeHg-induced accumulation of microglia/macrophages in the rat DRG. (A) Representative images of DRG sections immunostained for the microglia/macrophage marker Iba1 and the macrophage marker CD68. The accumulations of Iba1-positive cells are depicted by arrows. (B) The number of Iba1-positive cells (n = 3). **p < 0.01 compared with control and Day 7. (C) The number of accumulations of Iba1-positive cells (n = 3). **p < 0.01 compared with control and Day 7. (D) The number of CD68-positive cells. **p < 0.01 compared with control and Day 7.
As shown in Fig. 2, sensory fibers were damaged by MeHg exposure. Therefore, we investigated whether microglia/macrophages in sensory fibers were increased by MeHg. Transverse sections of sensory and motor fibers were immunostained for Iba1 and CD68. Iba1 and CD68-positive microglia and macrophages were significantly increased on day 14 in sensory fibers, similar to DRGs; however, there were no changes in motor fibers (Fig. 4). These findings suggest that MeHg exposure increases microglia/macrophages, especially in sensory neural tissue.
MeHg-induced accumulation of microglia/macrophages in sensory, but not motor, nerve fibers. (A) Representative images of sensory (left) and motor (right) nerve fibers immunostained for Iba1 and CD68. (B) The number of Iba1-positive cells in the sensory (left) and motor (right) nerve (n = 3). **p < 0.01 compared with control and Day 7. (C) The number of CD68-positive cells in the sensory (left) and motor (right) nerve (n = 3). *p < 0.05 compared with control and Day 7.
We next examined whether the increase in macrophages in the damaged tissue was caused by infiltration from the blood and/or by proliferation in situ. We co-stained sensory fiber sections for CD68 and the cell division marker Ki67 (Fig. 5A). The number of dividing cells was drastically increased in sensory, but not motor, fibers (Fig. 5B). However, this increase was not caused by macrophage proliferation in the sensory fibers (Fig. 5C). Together, these results suggest that the MeHg-induced increase in macrophages was caused by their infiltration from the blood.
The MeHg-induced accumulation of macrophages in sensory nerve fibers is caused by their infiltration. (A) Representative images of sensory (left) and motor (right) nerve fibers immunostained for the cell proliferation marker Ki67 and for CD68. Note that most Ki67-positive cells were not CD68-positive. (B) The number of Ki67-positive cells in the sensory (left) and motor (right) nerve (n = 3). **p < 0.01 compared with control and Day 7. (C) The number of Ki67/CD68-double positive cells in the sensory (left) and motor (right) nerve (n = 3).
MeHg exposure resulted in a significant increase in Ki67-positive (proliferating) cells in sensory fibers. To identify the cells undergoing proliferation, we co-immunostained for Ki67 and specific markers for various cell types, including astrocytes (GFAP), fibroblasts (Vimentin), blood vessel endothelial cells (CD31), microglia/macrophages (Iba1) and Schwann cells (SOX10) (Fig. 6A). Each of the cell types showed a degree of co-labeling with Ki67; however, most Ki67-positive cells were also positive for SOX10 (Fig. 6B). These results suggest that MeHg exposure induces the proliferation of Schwann cells in sensory fibers.
MeHg induces the proliferation of Schwann cells in Day 14 sensory nerve fibers. (A) Representative images of sensory nerve fibers co-immunostained for Ki67 and the astrocyte marker GFAP, the fibroblast marker vimentin, the endothelial cell marker CD31, Iba1, and the Schwann cell marker SOX10 (also stained with Hoechst dye). Triple-labeled cells are indicated with arrows. (B) The number of cells co-positive for each cell-type-specific marker and Ki67 (n = 3). **p < 0.01 compared with all others, except Ki67.
In the present study, we used quantitative immunohistochemical analyses to investigate the neurotoxic effects of MeHg on rat DRGs and peripheral sensory and motor nerve fibers. Our findings suggest that oral administration of MeHg causes the following: (1) significant neural degeneration in DRGs; (2) drastic axonal degeneration in sensory, but not motor, nerve fibers; (3) a significant increase in the number of microglia and macrophages in DRGs and sensory nerve fibers; and (4) the proliferation of Schwann cells in the sensory nerve. We therefore quantitatively confirmed the previously-reported qualitative histopathological changes in the peripheral nervous system caused by MeHg exposure. In addition, our study is the first to show that the accumulation of microglia and the proliferation of Schwann cells are two key features of MeHg-induced toxicity in the peripheral nervous system.
In this study, we used a dose of 6.7 mg/kg/day of methylmercury chloride (total dose of 33.5 mg for day 7 samples and 67 mg for day 14 samples) to evaluate the effect of MeHg on the peripheral nervous system. We chose this dose because rats given 10 mg/kg/day usually died within 10 days of the first administration, and a dose of 5 mg/kg/day produced no noticeable behavioral abnormalities, such as hind limb incoordination, for at least 3 weeks (data not shown).
While previous reports on the effects of MeHg on the peripheral nervous system differ in daily dose, administration regimen, the counter anion, solvent, animal strain, and sampling schedule, our findings are consistent with these prior studies. For example, in the present study, significant neuronal degeneration was observed on day 14 (total of 10 days of administration at 6.7 mg/kg/day). Similar degeneration in the DRG was reported for days 32 and 34 (total of 30 days of administration at 5 mg/kg/day) by Sakamoto and colleagues (Sakamoto et al., 1998). Axonal degeneration was observed on day 14 in the current study, similar to the Wallerian degeneration of sensory axons observed on day 10 in animals given 7.5 mg/kg/day MeHg-dicyandiamide (equivalent to 6.3 mg/kg/day MeHgCl) (Cavanagh and Chen, 1971). Motor nerve fibers were not affected by MeHg exposure, similar to that reported previously (Arimura et al., 1988). MeHg-induced myelin degeneration has also been reported previously (Miyakawa et al., 1970; Munro et al., 1980; Yip and Chang, 1981); however, myelin degeneration was not observed under our present experimental conditions. In the present study, we used a shorter period of administration (7–14 days) than in previous reports (20–56 days). Therefore, longer MeHg exposure may be required for myelin degeneration.
A MeHg-induced increase in microglia was reported in the mouse cortex (Fujimura et al., 2009), rat cerebellum (Sakamoto et al., 2008), and common marmoset cortex (Yamamoto et al., 2012). Here, we showed for the first time an increase in microglia in the rat DRGs and sensory nerve fibers caused by MeHg. Interestingly, the DRGs showed a significant accumulation of Iba1-positive microglia/macrophages. Microglia–astrocyte interactions and the release of cytokines are thought to represent a neuroprotective response to MeHg exposure (Eskes et al., 2002; Chang, 2007; Shinozaki et al., 2014), and therefore, the increase in microglia in DRGs and sensory fibers observed here might be neuroprotective. However, pathological activation of microglia is also thought to induce an inflammatory response, the disruption of the blood–brain barrier (Patel and Frey, 2015), and neuronal cell death (Ahmad et al., 2019). Amoeboid microglia, the activated form of microglia, were not distinguished from resting microglia in this study. However, the MeHg-induced increase in microglia might nonetheless produce inflammatory changes in the DRG and sensory fibers that result in neuronal cell death.
A MeHg-induced increase in macrophages was previously reported in the mouse cortex (Yamamoto et al., 2014), rat cerebellum (Nagashima, 1997), DRG (Cao et al., 2013; Schiønning and Danscher, 1999), and rat sensory nerve (Nagashima, 1997; Schiønning and Danscher, 1999). In the present study, we also found an increase in macrophages in the DRG and sensory fibers. Therefore, similar changes occur in both the central and peripheral nervous systems. Macrophages are derived from two different pathways. The first is the traditional pathway, in which these cells originate from circulating monocytes. The second is by self-proliferation within the local tissue (Gentek et al., 2014). The co-immunostaining results suggest that the macrophages were primarily derived from the monocyte pathway in the present study. Macrophages can be classified into two major phenotypes: proinflammatory macrophages (M1) and anti-inflammatory macrophages (M2) (Mantovani et al., 2013). M1 macrophages participate in host defense and secrete proinflammatory cytokines and molecules that cause tissue damage, whereas M2 macrophages express anti-inflammatory molecules that reduce inflammation and promote tissue recovery. We used an anti-CD163 antibody to specifically detect M2 macrophages (Supplemental Fig. 1). Approximately half of the CD68-positive macrophages were in the M2 state. This suggests that about half of the macrophages are associated with an anti-inflammatory response, while the other half are associated with a pro-inflammatory response in DRGs and sensory fibers following MeHg exposure. Therefore, the increase in macrophages produced by MeHg exposure might promote neuroprotective or neurodegenerative effect or both in the DRG.
A proliferation of Schwann cells was observed in sensory nerve fibers after MeHg exposure. Peripheral nerve injury sometimes induces the proliferation of Schwann cells (Pham and Gupta, 2009). This enhanced proliferation of Schwann cell is thought to be associated with neural recovery. Therefore, the proliferation of Schwann cells might be a neuroprotective response to MeHg toxicity.
Taken together, our findings suggest that MeHg-induced neural degeneration in rat DRG and sensory nerve fibers is associated with an increase in microglia/macrophages and the proliferation of Schwann cells. Several issues remain unresolved, such as the effects of different MeHg dosage and administration protocols, the temporal profile of the subcellular changes, the differences in sensitivity of the heterogeneous sensory neuron subpopulations in the DRG, and the role of microglia/macrophages (neuroprotective or neurotoxic). Further investigation of MeHg-induced changes in the peripheral nervous system is required to clarify the mechanisms of neurotoxicity.
This work was supported by the Study of the Health Effects of Heavy Metals organized by the Ministry of the Environment, Japan. We thank Barry Patel, PhD, from Edanz Group (www.edanzediting.com/ac) for editing a draft of this manuscript.
Conflict of interestThe authors declare that there is no conflict of interest.
