Chronic Exposure to Methylmercury Enhances the Anorexigenic Effects of Leptin in C57BL/6J Male Mice

Abstract

Several studies have demonstrated that heavy metals disrupt energy homeostasis. Leptin inhibits food intake and decreases body weight through activation of its receptor in the hypothalamus. The impact of heavy metals on leptin signaling in the hypothalamus is unclear. Here, we show that the environmental pollutant, methylmercury (MeHg), favors an anorexigenic profile in wild-type males. C57BL/6J mice were exposed to MeHg via drinking water (5 ppm) up to 30 days. Our data shows that MeHg exposure was associated with changes in leptin induced activation of Janus kinase 2 (JAK2)/signal transducer and activator of transcription 3 (STAT3) signaling pathway in the hypothalamus. In males, the activation of JAK2/STAT3 signaling pathway was sustained by an increase in SOCS3 protein levels. In females, MeHg-activated STAT3 was inhibited by a concomitant increase in PTP1B. Taken together, our data suggest that MeHg enhanced leptin effects in males, favoring an anorexigenic profile in males, which notably, have been shown to be more sensitive to the neurological effects of this organometal than females. A better understanding of MeHg-induced molecular mechanism alterations in the hypothalamus advances the understanding of its neurotoxicity and provides molecular sites for novel therapies.

1. Introduction

Mercury is a naturally occurring chemical element ubiquitously present in the environment. Its sources could be natural (forest fire and volcanoes) or anthropogenic (mining, coal factories and industry). In the environment, mercury may exist as elemental, inorganic and/or organic forms (Clarkson 2002). When inorganic mercury compounds reach aquatic systems, they undergo biomethylation by aquatic sulfate-reducing microorganisms forming methylmercury (MeHg) (Compeau and Bartha 1985). Thus, MeHg bioaccumulates and biomagnifies along the aquatic food chain reaching maximal levels in large predatory fishes such as swordfish, tuna or shark (Hintelmann 2010). The main route for human exposure to MeHg is via fish consumption, with populations that subside on seafood, especially large fish, being especially vulnerable to MeHg toxicity (Compeau and Bartha 1985; Antunes Dos Santos et al. 2016). Upon ingestion, MeHg is ubiquitously distributed in the organism, including the central nervous system (CNS). The CNS is the most susceptible organ to MeHg, especially when exposures occur during developmental stages (Marsh et al. 1995; Costa et al. 2004; Hassan et al. 2012). Although showing lesser severity, the adult brain is also vulnerable to MeHg toxicity.

In adults, after a variable latency period, symptoms begin to appear, including paresthesias, blurred vision, and ataxia, among many others. Notably, one of the initial symptoms is loss of body weight (Berthoud et al. 1976). In addition, weight loss has been described as a common event of MeHg exposure (Magos 1982; Shi et al. 2011; Li et al. 2014; Yamamoto et al. 2014; Bittencourt et al. 2017; Kong et al. 2019). However, the mechanisms of MeHg-induced disruption of body weight and food intake have yet to be elucidated. In our previous work, we reported that MeHg was able to induce an anorexigenic gene expression profile in a sex-dependent manner (Ferrer et al. 2018).

Leptin is a 167 aminoacid anorexigenic protein encoded by the obese gene (Zhang et al. 1994). It is mainly produced by adipocytes and exerts its functions in both the central and peripheral tissues. Circulating leptin levels correlate with body fat mass, serving as a sensor of the energy levels available in the organism (Frederich et al. 1995; Van Harmelen et al. 1998). Leptin levels fluctuate depending upon changes in food intake and circadian rhythms (Park and Ahima 2014). Leptin levels also display sexual dimorphisms. Females have higher circulating levels than males. These differences have been attributed to females’ tendency to accumulate more subcutaneous adipose tissue, which expresses higher levels of leptin than other fat depots, and the stimulatory effects of estrogen on leptin production versus the inhibitory actions of testosterone (Rosenbaum et al. 1996; Park and Ahima 2014; Zhang et al. 2014) .

Leptin has an important role in the regulation of energy homeostasis, metabolism, reproduction, angiogenesis and immune function (Park and Ahima 2014). Leptin readily crosses the brain-blood barrier (BBB) by a saturable transport system (Banks et al. 1996) and reaches the CNS. Leptin receptor is expressed in several brain regions, especially the arcuate nucleus of the hypothalamus, where leptin induces loss of appetite and increases energy expenditure, resulting in body weight loss (Schwartz et al. 2000; Morton et al. 2006). Upon binding to its receptor, leptin leads to activation of multiple intracellular signaling pathways, included janus kinase 2 (JAK2)/signal transducer and activator of transcription 3 (STAT3), which plays an important role controlling energy balance and mediating the anorexigenic/catabolic leptin actions (Vaisse et al. 1996; Bates et al. 2003; Gao et al. 2004; Jiang et al. 2008).

Understanding MeHg’s effects on molecular mechanisms that regulate body weight and energy homeostasis is important to better comprehend its toxic effects and to develop future treatments. The goal of this study was to test the hypothesis that MeHg interferes with the activity of leptin in the murine hypothalamus. To address this hypothesis, we used a mice model of MeHg exposure via drinking water. Our findings suggest that MeHg exposure modifies the effects of leptin on body weight and food intake in males, exacerbating the anorexigenic/catabolic functions of this hormone.

2. Material and methods

2.1. Mice

Seven-week-old C57BL/6J males (n=5 per group unless otherwise noted) and females (n=5 per group unless otherwise noted) were purchased from The Jackson Lab. All animals were kept under controlled conditions at constant temperature at 21±2°C, with a 12 h light/12 h dark cycle. Mice were provided ad libitum access to regular chow diet (Research Diets, Inc.) and water, unless otherwise stated.

All animal experiments were performed in compliance with the guidelines of Albert Einstein College of Medicine Institute of Animal Studies, and approved by the local Institutional animal care and use committee (IACUC).

2.2. Methylmercury exposure

Animals were exposed chronically to methylmercury via drinking water until a maximum of 30 days, which reproduces the most common exposure source in humans. After 1-week of acclimatization period, mice were assigned to the control or exposed groups. The control group received drinking water and the exposed group received drinking water with 5 ppm methylmercury (II) chloride (MeHg) (Sigma-Aldrich, 442534). Each week animals were provided with freshly prepared solutions. This MeHg dose was chosen based on previous experiments (Ferrer et al. 2018) and given its ability to reproduce sub-toxic threshold toxicity in mammalians, resulting in brain mercury concentrations around 25 μg Hg/g protein (Aschner 2012). In addition, adult mice exposed to 30 ppm MeHg in drinking water show accumulation of 23.2 μg Hg/g dry weight in the hippocampus and develop neurological changes, absent lethality (Fujimura et al. 2009). This protocol aims to mimic physiologically-relevantchronic exposures to subtoxic MeHg concentrations without the appearance of any observable toxic effect for the given dose and exposure times. Water consumption was measured periodically throughout the exposure time.

2.3. Tissue mercury content

Mice (female and male), exposed to 5 ppm MeHg or regular water, were euthanized on day 15 and on day 30 of exposure. Cortex and hypothalami were removed, frozen and stored at −80 °C to analyze MeHg accumulation.

Total Hg was measured in cortex and hypothalamus by atomic absorption spectroscopy, using a TricellDirect Mercury Analyzer (DMA-80, Milestone). Brain samples were weighed in a nickel boat and directly loaded into the DMA-80. They were then dried at 200°C for 1 minute and decomposed at 725°C for 3.5 minutes. Hg selectively binds to an amalgamator and is released upon rapid heating (12 s) to 850°C. The Hg vapor is then shuttled to a spectrophotometer, where absorbance is measured at 253.7 nM. Total Hg is quantified as a function of absorbance. Results were expressed as μg Hg/kg wet tissue.

2.4. Leptin treatment

2.4.1. For food intake and body weight changes evaluation:

On days 15 and 30 of MeHg exposure we performed a leptin administration experiment. Leptin was purchased from R&D systems (recombinant mouse leptin protein CF, 98-OB). Four days before the experiment, mice were individually housed and acclimated to their environment. Prior to the experiment, animals were fasted overnight (12–18 h). Then, they were administered intraperitoneally with 3 μg leptin/ g body weight or vehicle (saline, 0.9 % NaCl) twice a day (at 9 h and 17 h) for 2 days. Thus, in total there were 4 animal groups: control, MeHg, leptin, and MeHg plus leptin (n= 10 per group). Body weight and food intake were recorded every 8 h from the first administration through 72 h.

2.4.2. For tissue collection:

on day 30 of MeHg exposure, animals were fasted overnight (12–18 h). A single dose of 3 μg leptin/ g body weight or vehicle (saline, 0.9 % NaCl) was administered intraperitoneally. Mice were euthanized 30 min and 1 h from administration. Hypothalami were dissected, frozen in liquid nitrogen and stored at −80 °C. As in the previous experiment, we had 4 groups (control, MeHg, leptin and MeHg plus leptin).

Leptin times and doses chosen in these two protocols are based in reported studies and based in leptin’s ability to induce decreased body weight and food intake as well as induce leptin signaling pathways (Halaas et al. 1995; Ernst et al. 2009; Shen et al. 2009; Molinero et al. 2017; Tsunekawa et al. 2017).

2.5. Western blot

Hypothalami were homogenized in cold radioimmunoprecipitation assay (RIPA) buffer (Sigma Aldrich, R0278) supplemented with protease inhibitor cocktails 2 and 3 (Sigma Aldrich, P5726 and P0044, respectively) and Halt^™ phosphatase inhibitor cocktail (ThermoFisher Scientific, 78437) with a Kimble Pellet Pestle Motor (Fisherbrand, 12–141361). During tissue processing, samples were maintained on ice. The lysates were centrifuged at 10.000 g for 10 min at 4 °C. Whole-cell proteins were collected in the supernatant and quantified using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, 23225). Lysates were boiled at 100 °C for 5 minutes in laemmeli sample buffer (Bio-Rad, #1610737). Then, 20 μg of protein were resolved by 4– 20 % sodium dodecyl sulfate polyacrylamide gel (SDS-PAGE, Bio-Rad #4561096) and transferred to a nitrocellulose membrane. Non-specific sites were blocked with Tris buffer saline-Tween-20 (TBS-T) buffer containing 5 % bovine serum albumin (BSA, Sigma Aldrich, A3059) for 1 h at room temperature. The blots were incubated overnight at 4 °C with primary antibodies diluted in 5 % BSA in TBS-T buffer. Next, membranes were incubated 1 h at room temperature with the appropriate horseradish peroxidase-conjugated secondary antibody. The blots were visualized with a chemiluminescent method (Thermo Fisher Scientific 34579 or 37071). Band densitometries were quantified using ImageJ software (NIH) (Rueden et al. 2017), and normalized using β-actin as loading control.

The following antibodies were used: rabbit anti-phospho-STAT3 (Tyr705) (Cell Signaling Technology, #9145), rabbit anti-phospho-STAT3 (Ser727) (Cell Signaling Technology, #9134), mouse anti-STAT3 (Cell Signaling Technology, #9139), rabbit anti-phospho-JAK2 (Tyr1007/1008) (Thermo Fisher Scientific,44–426G), rabbit anti-JAK2 (Cell Signaling Technology, #3230), rabbit anti-SOCS3 (Cell Signaling Technology, #2932), goat anti-PTP1B (R&D Systems, AF3954), rabbit anti-phospho-AKT (Ser473) (Cell Signaling Technology, #9271), rabbit anti-AKT (Cell Signaling Technology,#9272), rabbit anti-phospho-p42/44 MAPK (ERK1/2) (Thr202/Tyr204) (Cell Signaling Technology, #9101), rabbit anti-p42/44 MAPK (ERK1/2) (Cell Signaling Technology, #9102), rabbit anti-phospho-mTOR (Ser2448) (Cell Signaling Technology, #2971), rabbit anti-mTOR (Cell Signaling Technology, #2983), rabbit anti-phospho-SAPK/JNK (Thr183/Tyr185) (Cell Signaling Technology, #9251), rabbit anti-SAPK/JNK (Cell Signaling Technology, #9252), rabbit anti-ROCK1 (Cell Signaling Technology, #4035), rabbit anti-ROCK1 cleavage (Thermo Fisher Scientific, MA1–41031), mouse anti-Bax (Santa Cruz Biotechnology, sc-7480), mouse anti-Bcl2 (Santa Cruz Biotechnology, sc-7382), mouse anti-cytochrome C (Santa Cruz Biotechnology, sc-13561), rabbit anti-SOD-2 (Cell Signaling Technology, #13141), rabbit anti-HO-1 (Cell Signaling Technology, #70081), rabbit anti-NeuN (Cell Signaling Technology, #12943), mouse anti-GFAP (Cell Signaling Technology, #3670), rabbit anti-Iba1/AIF-1 (Cell Signaling Technology, #17198), and mouse anti-β-actin (Sigma Aldrich, A1978).

2.6. Quantitative real time PCR

Total RNA isolation from hypothalami was performed with the TRIzol^™ reagent (Thermo Fisher Scientific, 15596026) according to the manufacturer’s protocol. Briefly, chloroform (Fisher Scientific, C298) was used to isolate RNA followed by precipitation with isopropyl alcohol (Sigma Aldrich, 67-63-0). Total RNA was quantified using NanoDrop2000 Spectrophotometer (Thermo Fisher Scientific, ND2000) and reverse transcribed with High Capacity cDNA Reverse Transcription kit (Applied Biosystems, 4368814) following manufacturer’s instructions. Quantitative Real Time PCR was carried out using an iCycler Thermal Cycler (BioRad). Predesigned TaqMan probes were selected from TaqMan Gene Expression Assays (Applied Biosystems). They are listed in table 1. The housekeeping gene Gapdh ID Mm99999915 was used to normalize the relative mRNA expression. The 2^−ΔΔCtmethod (Livak and Schmittgen 2001) was used to determine changes in the expression.

Table 1:

TaqMan probes for qPCR analysis

Gene symbol	Reference sequence	Official full name
Hmox1	Mn00516005	heme oxygenase 1
Lepr	Mm00440181	leptin receptor
Npy	Mm01410146	neuropeptide Y
Pome	Mm00435874	pro-opiomelanocortin-alpha
Ptpnl	Mn00448427	protein tyrosine phosphatase, non-receptor type 1
Rockl	Mn00485745	Rho-associated coiled-coil containing protein kinase 1
Socs3	MnO545913	suppressor of cytokine signaling 3
Sod2	Mn01313000	superoxide dismutase 2, mitochondrial
Stat3	MnO 1219775	signal transducer and activator of transcription 3
Gapdh	Mm99999915	glyceraldehydes-3 -phosphatase dehydrogenase

2.7. Statistical Analysis

Statistical calculations were performed using the Statistical Package for Social Sciences (SPSS) software. Mean daily water consumption was analyzed by two-way analysis of variance (ANOVA) with MeHg, and sex as main factors, followed by Tukey’s post-hoc test. Tissue accumulation was analyzed by three-way ANOVA with sex, MeHg and time as main factors, followed by Tukey’s post-hoc test. MeHg consumption was analyzed with a two-way repeated measure ANOVA with MeHg, sex and time as main factors. Western blot and qPCR data were analyzed by two-way ANOVA to determine the effect of MeHg and leptin within each sex separately, followed by Tukey’s post-hoc test. Body weight changes and cumulative food intake were analyzed with a three-way repeated measure ANOVA with MeHg, leptin and time (from leptin or vehicle administration to last measure point) as main factors. All data are presented as mean ± standard error of mean (SEM). Statistically significant differences were set at p < 0.05.

3. Results

3.1. MeHg exposure via drinking water increases mercury accumulation in brain in time- and sex-dependent manner

Water bottles were weighed periodically to ensure that the mice did not reject the MeHg treatments. No differences were observed in water consumption in both males and females (Fig 1a). However, females consumed more water than males (sex F_(1,76)= 88.740, p < 0.001). As expected, MeHg exposed mice (males and females) ingested significantly higher amounts of MeHg (F_(1,76) = 563,917, p < 0.001). As a result of the higher water intake, females consumed more MeHg than males compared to controls (F_(1,76) = 112,586, p < 0.001) (Fig 1b).

Fig. 1.

MeHg consumption and accumulation in exposed mice versus non-exposed. a) Mean water consumption per animal and per day (n = 26 animals per group). b) MeHg consumption per gram body weight (BW) in males and in females (n = 19 – 20 animals per group). c) Mercury accumulation in hypothalamus after 15 days (n = 5 animals per group). d) Mercury accumulation in cortex after 15 days (n = 5 animals per group). e) Mercury accumulation in hypothalamus after 30 days (n = 5 animals per group). f) Mercury accumulation in cortex after 30 days (n = 5 animals per group). ‡ p < 0.001

Due to the important role of the hypothalamus in the control of energy homeostasis and body weight, next, we analyzed whether MeHg accumulated in this brain region. After 15 and 30 days of MeHg exposure, female and male mice (n= 5 per group) were euthanized. Brains were dissected out and mercury accumulation was determined. As a positive control, we evaluated mercury levels in the cortex, a well known MeHg-targeted brain area in humans (Eto 1997), which is also known to accumulate mercury after 30 days of MeHg exposure in adult mice (Ferrer et al. 2018). Longer times of exposure led to greater mercury accumulation in the hypothalamus, and this increase was greater in females than in males (interaction MeHg, time and sex: F_(1,32) = 21.677, p < 0.001). The post hoc analysis of that interaction showed a significant difference between MeHg-exposed males and MeHg-exposed females (Fig 1c and 1e). As expected, mice exposed to MeHg displayed time- and sex-dependent mercury accumulation in the cortex (interaction of MeHg, time and sex: F_(1,32) = 117.820, p < 0.001) (Fig 1d and 1f) corroborating our previous observations (Ferrer et al. 2018). No significant difference was observed between brain areas.

3.2. MeHg exposure exacerbates leptin effects on physiological parameters in males

Next, we studied the effects of MeHg on the hypothalamus, with special emphasis on leptin activated pathways. Specifically, we examined the leptin response in animals exposed to MeHg via drinking water. We performed leptin treatment experiments on days 15 and 30 of MeHg exposure. The afternoon before, mice were fasted overnight (12–18 h) followed by leptin administration twice a day for two days. Upon the first administration, food intake and body weight were monitored. As expected, male mice treated with leptin lost greater weight than vehicle treated mice on day 15 (leptin: F_(1,34) = 6.212, p = 0.018) and on day 30 of MeHg exposure (leptin: F_(1,34) = 30.332, p < 0.001) (Fig 2a). Leptin also significantly reduced food consumption both on day 15 (leptin: F_(1,34) = 11.316, p = 0.002) and on day 30 of MeHg exposure (leptin: F_(1,34) = 8.393, p = 0.007) (Fig 2c). In this context, MeHg exacerbated the leptin effects in males, leading to a greater decrease in food consumption (on day15: MeHg: F_(1,34) = 4.535, p = 0.041 and day 30 MeHg: F_(1,34) = 11.919, p = 0.002) (Fig 2c). In addition, MeHg exacerbated leptin body weight loss in males exposed to MeHg for 30 days (MeHg: F_(1,34) = 6–855, p = 0.013) (Fig 2c). In this context, MeHg also decreased food consumption in males on day15 (MeHg: F_(1,34) = 4.535, p = 0.041) and day 30 MeHg (F_(1,34) = 11.919, p = 0.002), concomitant with a greater decrease in leptin injected males (Fig 2c).

Fig. 2.

MeHg disrupts exogenous leptin actions on body weight and food intake in males and did not have effects on females. On days 15 and 30 of MeHg exposure, animals were fasted overnight followed by 3 μg/g BW leptin intraperitoneal (ip) administration twice a day for 2 days. Body weight and food intake was measured. a) Body weight gain in males on 15 days (left) and on 30 days of MeHg exposure (right). b) Body weight gain in females on 15 days (left) and on 30 days of MeHg (right) exposure (right). c) Food intake in males on 15 days (left) and on 30 days of MeHg exposure (right). d) Food intake in females on 15 days (left) and on 30 days of MeHg exposure (right). n = 10 per group in all cases except in males on 30 days MeHg (n = 9–10). Data was analyzed by repeated measures ANOVA. Symbols in the graphs represent between subjects significant MeHg effects. * p < 0.05. † p < 0.01

In females, leptin significantly decreased body weight gain compared with vehicle-treated females both on day 15 (leptin: F_(1,36) = 13.285, p = 0.001) and on day 30 (leptin: F_(1,36) = 14.825, p < 0.001) of MeHg exposure (Fig 2b). In addition, leptin also decreased food intake in females, attaining statistical significance at day 30 of MeHg exposure (leptin: F_(1,33) = 13.381, p = 0.001) (Fig 2d). Notably, MeHg did not have a significant effect on females.

3.3. MeHg induces STAT3 phosphorylation in males

Next, we investigated if the changes observed in body weight and food intake were paralleled by altered molecular signaling. These studies were performed in mice exposed continuously to MeHg for 30 days. First, changes in transcription factor signal transducer and activator of transcription 3 (STAT3) phosphorylation were evaluated at 30 min and 1 h after leptin treatment. STAT3 has a key role mediating the anorexigenic leptin functions (Banks et al. 2000; Bates et al. 2003). In males, at 30 minutes post leptin administration, leptin significantly increased STAT3 phosphorylation on the tyrosine 705 residue (Tyr705) (leptin: F_(1,15) = 77.207, p < 0.001) and the ratio of phospho-STAT3 (Tyr705)/STAT3 (leptin: F_(1,15) = 59.115, p < 0.001)) (Fig 3a, right). Leptin’s effect remained significant 1 h after leptin injection in both phospho-STAT3 (Tyr705) (leptin: F_(1,14) = 5.125, p = 0.043) and the ratio of phospho-STAT3 (Tyr705)/STAT3 (leptin: F_(1,14) = 9.421, p = 0.008) (Fig 3a, left). Males exposed to MeHg via drinking water showed, 30 min after leptin injection, increased STAT3 phosphorylation on Tyr705 (MeHg: F_(1,15) = 7.342, p = 0.016) and increased ratio of phospho-STAT3 (Tyr705)/STAT3 (MeHg: F_(1,15) = 4.570, p = 0.05). One hour after leptin injection, MeHg-induced STAT3 activation remained significantly higher (phospho-STAT3 (Tyr705): MeHg: F_(1,14) = 8.018, p = 0.015 and the ratio of phospho-STAT3 (Tyr705)/STAT3:MeHg: F_(1,14) = 16.076, p = 0.001) (Fig 3a). In agreement with males, females also responded to leptin with a significant increase in STAT3 phosphorylation (leptin: F_(1,14)= 4.234, p = 0.040), and in the ratio of phospho-STAT3 (Tyr705)/STAT3 (leptin: F_(1,14) = 213.937, p < 0.001) at 30 min post leptin treatment. The effects of leptin remained significant 1 h post leptin administration for STAT3 phosphorylation (leptin: F_(1,13) = 18.251, p = 0.001) as well as in the ratio of phospho-STAT3 (Tyr705)/STAT3 (leptin: F_(1,13) = 15.639, p = 0.002) (Fig 4a). However, only at 30 min after leptin injection MeHg exposure led to a significant induction of STAT3 phosphorylation (MeHg: F_(1,14) = 7.420, p = 0.017) and also in the ratio of phospho-STAT3 (Tyr705)/STAT3 (MeHg: F_(1,14) = 8.045, p = 0.013) (Fig 4a).

Fig. 3.

MeHg- and leptin-induced effects on JAK2/STAT3 signaling pathway in males. Males were exposed to regular or 5 ppm MeHg in drinking water for 30 days. Then, they were fast overnight and injected ip with 3 μg/g BW leptin. Males were euthanized 30 min and 1 h after leptin administration. Protein levels and protein phosphorylation from hypothalamic lysates were measured by western blot. a) Phosphorylation of STAT3Tyr705 at 30 min (left) and at 1 h (right). b) Phosphorylation of JAK2 at 30 min (left) and at 1 h (right). c) SOCS3 levels at 30 min (left) and at 1 h (right). d) PTP1B at 30 min (left) and at 1 h (right). n = 4 – 5 per group. * p < 0.05. † p < 0.01. ‡ p < 0.001

Fig. 4.

MeHg- and leptin-induced effects on JAK2/STAT3 signaling pathway in females. Females were exposed to regular or 5 ppm MeHg in drinking water for 30 days. Then, they were fast overnight and injected ip with 3 μg/g BW leptin. Females were euthanized 30 min and 1 h after leptin administration. Protein levels and protein phosphorylation from hypothalamic lysates were measured by western blot. a) Phosphorylation of STAT3 Tyr705 at 30 min (left) and at 1 h (right). b) Phosphorylation of JAK2 at 30 min (left) and at 1 h (right). c) SOCS3 levels at 30 min (left) and at 1 h (right). d) PTP1B at 30 min (left) and at 1 h (right). n = 4 – 5 per group. * p < 0.05. † p < 0.01. ‡ p < 0.001

Phosphorylation of tyrosine janus kinase 2 (JAK2), an upstream protein that leads to STAT3 activation after the interaction of leptin with its receptor (Bjorbaek et al. 1997; Ghilardi and Skoda 1997), was also immunodetected by western blot analysis. In males, a significant interaction effect was observed between MeHg and leptin on phosphorylated JAK2 (interaction MeHg-leptin: F_(1,16) = 11.033, p = 0.004). Post-hoc analysis revealed significant differences between control and leptin-treated males, control and MeHg-exposed mice, as well as control and MeHg-exposed animals treated with leptin, with all showing higher values than control mice. (Fig 3b left). In addition, leptin and MeHg significantly increased phosphorylated JAK2/JAK2 ratio 30 min after leptin injection (leptin: F_(1,16) = 8.215, p = 0.011 and MeHg: F_(1,16) = 17.192, p = 0.001) (Fig 3b left). These differences were attenuated at 1 h post leptin administration (Fig 3b right). Surprisingly, in females, leptin treatment failed to reach any significant effect on JAK2, whereas at 30 minutes after leptin administration, MeHg decreased JAK2 activation, mitigating phospho-JAK2 (MeHg: F_(1,16) = 14.420, p = 0.002), as well as the ratio of phospho-JAK2/JAK2 (MeHg: F_(1,16) = 13.415, p = 0.002) (Fig 4b).

Suppressor of cytokine signaling 3 (SOCS3) is a protein capable of inhibiting leptin activities through a negative feedback mechanism induced by the same JAK2-STAT3 signaling pathway (Bjorbaek et al. 1998; Reed et al. 2010). In males, we failed to detect any leptin effect, whereas 30 min after leptin or vehicle administration MeHg significantly decreased SOCS3 (MeHg: F_(1,14) = 17.568, p = 0.001) (Fig 3c). This effect was not detected in males 1 h after leptin administration. On the other hand, in females at 1 h post leptin treatment, post-hoc analysis revealed a significant effect of leptin in animals not exposed to MeHg (interaction of leptin-MeHg: F_(1,14) = 12.105, p = 0.004) (Fig 4c).

Analogous to SOCS3, protein tyrosine phosphatase 1B (PTP1B) inhibits leptin signaling by dephosphorylation of JAK2 (Myers et al. 2001; Kaszubska et al. 2002; Zabolotny et al. 2002). In males, MeHg induced an increase of PTP1B in animals at 30 min of leptin treatment (MeHg: F_(1,16) = 27.113, p < 0.001) with no effect at 1 h of leptin administration (Fig 3d). In females, MeHg elevated PTP1B levels after 1 h of leptin injection (MeHg: F_(1,14) = 29.468, p < 0.001). Of note, at 30 min, leptin treatment increased PTP1B protein levels in females exposed to control drinking water. This increase was impaired in MeHg exposed females (interaction leptin-MeHg: F_(1,15) = 6.271, p = 0.024) (Fig 4d).

Upon STAT3 activation, the protein forms dimers and translocates into the nucleus, where it regulates the expression of several genes (Bates et al. 2003; Piper et al. 2008). To determine whether MeHg disrupts leptin-induced STAT3 transcriptional activity, we examined leptin effects on gene expression in mice exposed to MeHg for 30 days in drinking water. Animals were fasted overnight (12–18 h), treated with leptin or vehicle and euthanized at 30 minutes or 1 hour after administration. Gene expression was assessed by qPCR. We analyzed the gene expression of Socs3, which is a STAT3 target gene, and Ptp1b, both of them have a role suppressing JAK2/ STAT3 signaling (Bjorbak et al. 2000; Myers et al. 2001). In addition we also analyzed the gene expression of the orexigenic neuropeptide Npy and the anorexigenic neuropeptide Pomc that have a main role controlling food intake and energy homeostasis (Schwartz et al. 2000). In males, leptin significantly increased Socs3 gene expression at 30 min of treatment (leptin: F_(1,14) = 6.809, p = 0.021), whereas no changes in Ptp1b, and Npy genes were observed (Fig 5a I 5b). MeHg significantly decreased Socs3 (MeHg: F_(1,16) = 2.256, p = 0.032) and Npy (MeHg: F_(1,16) = 6.650, p = 0.020) gene expression in males at 1 h post-administration of leptin (Fig 5a). In addition, MeHg exposed males were more sensitive to the leptin effects, where leptin treatment increased Pomc gene expression (interaction leptin-MeHg: F_(1,14) = 16.300, p = 0.001). A similar trend was observed at 1 h after leptin injection, but it did not attain statistical significance. As for females, the leptin-induced increase in Ptp1b gene expression at 1 h of leptin administration was attenuated in MeHg exposed mice (interaction leptin-MeHg: F_(1,15) = 12.858, p = 0.003) (Fig 5b). Moreover, MeHg exposed females showed an increase in Stat3 mRNA at 30 min post-leptin or vehicle administration (MeHg: F_(1,15) = 4.938, p = 0.042) (Fig 6a). Interestingly, the leptin receptor (Lepr) mRNA was affected by MeHg in both sexes, decreasing in males (MeHg: F_(1,16) = 5.635, p = 0.030) and increasing in females (MeHg: F_(1,14) = 9.062, p = 0.009) (Fig 6b).

Fig. 5.

MeHg- and leptin-induced effects on gene expression. Males and female mice were exposed to regular water or 5 ppm MeHg in drinking water for 30 days. Then, they were fast overnight and injected ip with 3 μg/g BW leptin. Animals were euthanized 30 min and 1 h after leptin administration. MeHg’s and leptin’s effects on hypothalamic mRNA expression of Socs3, Ptp1b, Npy and Pomc at 30 min (A) or at 1 h (B) post-leptin administration. n = 4–5 animals per group. * p < 0.05. ‡ p < 0.001

Fig. 6.

MeHg- and leptin-induced effects on gene expression. Males and female mice were exposed to regular or 5 ppm MeHg in drinking water for 30 days. Then, they were fast overnight and injected ip with 3 μg/g BW leptin. Animals were euthanized at 30 min after leptin administration. MeHg’s and leptin’s effects on hypothalamic mRNA expression were analyzed by qPCR. a) Stat3 mRNA. b) Lepr mRNA. n = 4–5 animals per group. * p < 0.05. † p < 0.01

3.4. Leptin activity is mainly mediated by JAK2/STAT3 signaling pathways

Leptin mediates its actions via JAK2/STAT3 signaling pathway. Phosphorylation of STAT3 in the tyrosine 705 residue is needed to induce STAT3 dimerization and nuclear translocation. However, it is reported that STAT3 phosphorylation on residue serine 727 is implicated in DNA binding and leads a maximal activation (Wen et al. 1995). No significant effect of leptin or MeHg was detected on STAT3 phosphorylation at serine 727 residue, neither in males nor in females (Fig 7a).

Fig. 7.

MeHg- and leptin-induced effects on leptin-activated signaling pathways. Males and females were exposed to regular or 5 ppm MeHg in drinking water for 30 days. Then, they were fast overnight and injected ip with 3 μg/g BW leptin. Mice were euthanized 30 min after leptin administration. Protein levels and protein phosphorylation from hypothalamic lysates were measured by western blot. a) Phosphorylation of STAT3 Ser727 in males (left) and in females (right). b) Phosphorylation of JNK in males (left) and in females (right). c) Phosphorylation of mTOR in males (left) and in females (right). n = 4 – 5 per group. * p < 0.05. † p < 0.01. ‡ p < 0.001

In order to determine whether leptin and MeHg are modulating other pathways implicated in the regulation of energy homeostasis via leptin receptor (Ahima and Osei 2004; Park and Ahima 2014), we analyzed JNK, mTOR, ERK-1/2, and AKT phosphorylation by immunoblot (Fig 7 and 8). In males MeHg reduced phospho- JNK (MeHg: F_(1,16) = 7.973, p = 0.012) (Fig 7b left) at 30 min of leptin treatment, whereas leptin treatment in MeHg exposed mice led to a more pronounced decrease in phospho-JNK/JNK ratio compare with the other groups (interaction leptin-MeHg: F_(1,16) = 4.737, p = 0.046) (Fig 7b left). No differences were observed in mTOR in both males and females (Fig 7c). Contrary to males, at 30 min of leptin administration, MeHg decreased the phospho-JNK/JNK ratio and leptin reversed it (interaction leptin-MeHg: F_(1,16) = 5.965, p = 0.027) (Fig 7b right).

Fig. 8.

Effects of MeHg and leptin on leptin-activated signaling pathways. Males and females were exposed to regular or 5 ppm MeHg in drinking water for 30 days. Then, they were fast overnight and injected ip with 3 μg/g BW leptin. Mice were euthanized 30 min and 1 h after leptin administration. Protein and protein phosphorylation from hypothalamic lysates were measured by western blot. a) Phosphorylation of ERK1/2 at 30 min (left) and at 1 h (right) in males. b) Phosphorylation of AKT at 30 min (left) and at 1 h (right) in males. c) Phosphorylation of ERK1/2 at 30 min (left) and at 1 h (right) in females. n = 4 – 5 per group. d) Phosphorylation of AKT at 30 min (left) and at 1 h (right) in females. * p < 0.05. † p < 0.01. ‡ p < 0.001

In addition, in males, MeHg reduced ERK-1/2 phosphorylation at 30 min (MeHg: F_(1,15) = 6.003, p = 0.027) and at 1 h after leptin administration (MeHg: F_(1,14) = 8.483, p = 0.011). The same effect was detected when normalized by total ERK at 30 min (MeHg: F_(1,15) = 5.274, p = 0.038) and at 1 h of leptin injection (MeHg: F_(1,14) = 14.290, p = 0.002) (Fig 8a). Moreover, the ratio of phospho-ERK-1/2 vs ERK-1/2 at 1 h post leptin administration reveled a significant decrease in leptin treated males (leptin: F_(1,14) = 6.454, p = 0.024) (Fib 8a right). MeHg also reduced phospho-AKT (MeHg: F_(1,16) = 6.657, p = 0.020) (Fig 8b). However, the ratio of phospho-AKT/AKT was decreased in males exposed to MeHg and it was reverted at 30 min of leptin administration (interaction leptin-MeHg: F_(1,15) = 6.689, p = 0.021) (Fig 8b left).

Consistent with the results in males, in females MeHg decreased ERK-1/2 phosphorylation (MeHg: F_(1,15) = 17.894, p = 0.001) (Fig 8c left) and AKT phosphorylation. The interaction between leptin and MeHg revealed that all groups were different from control, showing reduction in AKT activation in both AKT phosphorylation (interaction leptin-MeHg: F_(1,16) = 8.584, p = 0.010) and the ratio of phospho-AKT/AKT (interaction leptin-MeHg: F_(1,15) = 6.003, p = 0.027) (Fig 8d left). In females, MeHg induced an increase in AKT phosphorylation at 1 h of leptin treatment (MeHg: F_(1,15) = 4.765, p = 0.048) (Fig 8d, right). Leptin-treated female mice also showed increased ratio of phospho-ERK-1/2/ERK-1-/2 in drinking water-exposed females, but not in MeHg-exposed females (interaction leptin-MeHg: F_(1,15) = 6.376, p = 0.023) (Fig 8c left).

3.5. MeHg induces cleavage/activation of ROCK1 in males

Since Rho-associated kinase 1 (ROCK1) might play a role in both mediating leptin-induced JAK2/STAT3 pathway (Huang et al. 2012) and MeHg-induced toxicity (Antunes Dos Santos et al. 2016), next, we assessed the effects of leptin in MeHg exposed animals on protein and gene expression at 1 h post leptin administration. Whereas MeHg did not affect ROCK1 protein levels, leptin decreased ROCK1 protein levels in males (leptin: F_(1,14) = 6.044, p = 0.028), and increased in females (leptin: F_(1,14) = 4.901, p = 0.024) (Fig 9a). There were no changes in Rock1 gene expression (Fig 9c). To assess ROCK1 activation, we performed immunodetection of ROCK1 cleavage. Whereas no changes were detected in females, in males both MeHg (MeHg: F_(1,14) = 13.267, p = 0.003) and leptin (leptin: F_(1,14) = 4.496, p = 0.05) increased ROCK1 cleavage (Fig 9b).

Fig. 9.

MeHg- and leptin-induced effects on ROCK1. Males and females were exposed to regular or 5 ppm MeHg in drinking water for 30 days. Then, they were fast overnight and injected ip with 3 μg/g BW leptin. Animals were euthanized 1 h after leptin administration. Protein levels from hypothalamic lysates were measured by western blot and gene expression was analyzed by qPCR. a) ROCK1 protein levels in males (left) and in females (right). b) ROCK1 cleavage in males (left) and in females (right). c) Rock1 mRNA expression. n = 4 – 5 per group. * p < 0.05. † p < 0.01

3.6. Leptin ameliorates MeHg alteration in apoptotic proteins and some anti-oxidant enzymes in a time and sex-dependent manner

Consistent with the toxic effect of MeHg, we noted in males analyzed 30 min after leptin treatment that MeHg significantly increased the ratio Bax/Bcl2 (MeHg: F_(1,16) = 19.059, p = 0.001), whereas leptin decreased it (leptin: F_(1,16) = 13.111, p = 0.003) (Fig 10a left). Analysis of Bax and Bcl2 proteins revealed that MeHg decreased protein levels of Bcl2. This decrease was reverted by leptin treatment (interaction leptin-MeHg: F_(1,16) = 13.671, p = 0.002). Unexpectedly, leptin increased the levels of pro-apoptotic Bax in MeHg exposed males (interaction of leptin-MeHg: F_(1,16) = 22.415, p < 0.001) (Fig 10a left). MeHg’s effects were also discerned 1 h after leptin treatment, with MeHg decreasing Bcl2 (MeHg: F_(1,12) = 4.984, p = 0.045) (Fig 10a right). However, no changes were detected in cytochrome C (Fig 10b). Contrary to males, in females leptin increased the ratio of Bax/Bcl2 at 30 min (leptin: F_(1,15) = 7.726, p = 0.014) and also at 1 h post leptin administration (interaction leptin-MeHg: F_(1,15) = 5.448, p = 0.036) (Fig 10c). No changes in cytochrome C were detected (Fig 10d).

Fig. 10.

MeHg- and leptin-induced effects on apoptotic proteins. Males and females were exposed to regular or 5 ppm MeHg in drinking water for 30 days. Then, they were fast overnight and injected ip with 3 μg/g BW leptin. Mice were euthanized 30 min or 1 h after leptin administration. Protein expression from hypothalamic lysates was measured by western blot. a) Bax and Bcl2 at 30 min (left) and at 1 h (right) in males. b) Cytochrome C at 30 min (left) and at 1 h (right) in males. c) Bax and Bcl2 at 30 min (left) and at 1 h (right) in females. d) Cytochrome C at 30 min (left) and at 1 h (right) in females. n = 4 – 5 per group. * p < 0.05. † p < 0.01. ‡ p < 0.001

Next, we assessed the expression of some anti-oxidant enzymes after 30 min of leptin treatment to test the hypothesis that leptin may protected against MeHg toxicity inducing anti-oxidant defense. Some authors have demonstrated that leptin have anti-oxidant properties (Zheng et al. 2010; Kaeidi et al. 2019). First, western blots showed that both leptin and MeHg increase superoxide dismutase 2 (SOD2), absent an effect of leptin in MeHg-exposed male mice (interaction leptin-MeHg: F_(1,16) = 9.603, p = 0.007) (Fig 11 a left). MeHg effects on Heme oxygenase 1 (HO1) were contrary, decreasing protein levels (by immunoblot detection), whereas leptin reverted MeHg’s effects (interaction leptin-MeHg: F_(1,16) = 13.438, p = 0.002) (Fig 11b left). No changes were detected in SOD2 and HO1 gene expression levels (Fig 11a and b right). In females, leptin had no effect on SOD2 and HO1 protein and gene expression. Moreover, MeHg decreased HO1 protein levels in females (MeHg: F_(1,16) = 6.825, p = 0.019) (Fig 11b right) without affecting its gene expression. In addition, MeHg did not have any effect on SOD2 gene and protein expression in females. (Fig 11c and d).

Fig. 11.

MeHg- and leptin-induced effects on anti-oxidant enzymes. Males and females were exposed to regular or 5 ppm MeHg in drinking water for 30 days. Then, they were fast overnight and injected ip with 3 μg/g BW leptin. Animals were euthanized 30 min after leptin administration. Proteins from hypothalamic lysates were measured by western blot gene expression was analyzed by qPCR. a) SOD2 protein levels in males (left) and in females (right). b) HO1 protein levels in males (left) and in females (right). c) Sod2 mRNA expression. d) Hmox1 mRNA expression n = 4 – 5 per group. * p < 0.05. † p < 0.01. ‡ p < 0.001

Finally, we performed western blot analyses for markers of different cellular populations (neurons (NeuN), astrocytes (GFAP) and microglia (Iba1)). In males, MeHg decreased NeuN, and this effect was reverted by leptin treatment (interaction leptin-MeHg: F_(1,16) = 6.066, p = 0.025) (Fig 12a left). No effects were detected on GFAP and Iba1 protein levels (Fig 12b and c left). In females, leptin decreased GFAP (leptin: F_(1,14) = 9.804, p = 0.007) (Fig 12 right) whereas MeHg significantly decreased Iba1 protein levels (MeHg: F_(1,16) = 5.229, p = 0.036) (Fig 12 right).

Fig. 12.

MeHg- and leptin-induced effects on neuronal, astrocytic and microglial markers. Males and females were exposed to regular or 5 ppm MeHg in drinking water for 30 days. Then, they were fast overnight and injected ip with 3 μg/g BW leptin. Animals were euthanized 30 min after leptin administration. Proteins from hypothalamic lysates were measured by western blot. a) NeuN in males (left) and in females (right). b) GFAP in males (left) and in females (right). c) Iba1 in males (left) and in females (right). n = 4 – 5 per group. * p < 0.05. † p < 0.01

4. Discussion

Body weight change is one of the earliest symptoms of MeHg poisoning (Schroeder and Mitchener 1975; Berthoud et al. 1976; Ralston and Raymond 2010), and it is frequently used as a toxicity marker (Magos 1982; Shi et al. 2011; Li et al. 2014; Yamamoto et al. 2014; Bittencourt et al. 2017). However, studies have yet to address whether MeHg is able to affect molecular pathways that regulate body weight and food intake. In this study we examined MeHg’s effects on leptin-induced signaling pathways in the hypothalamus in a mice model of MeHg toxicity. Our data, for the first time, provide evidence to support a novel anorexigenic role of MeHg in exacerbating leptin’s function.

The main area in the CNS responsible for energy homeostasis is the hypothalamus (Schwartz et al. 2000; Morton et al. 2006). It readily regulates energy necessities in the body. A wide-range of hormones and metabolites pass across the brain-blood barrier, providing signals to the hypothalamus on the body energy status (Schwartz et al. 2000; Haddad-Tovolli et al. 2017). In our model, animals exposed to 5 ppm MeHg in drinking water showed time-dependent accumulation of mercury in the hypothalamus. This observation corroborates earlier studies (Ernst et al. 1993; Oliveira et al. 2006). Concomitant with increased mercury levels in this structure, mercury levels also increased in the cerebral cortex. Thirty days after MeHg exposure females accumulated 5.3 μg/g and 5.4 μg/g in hypothalamus and cortex, respectively, and males accumulated 2 μg/g and 1.9 μg/g in hypothalamus and cortex, respectively. These levels are higher than those reported in low chronically-exposed humans, and higher than in non-exposed individuals (Weiner and Nylander 1993; O’Donoghue et al. 2020). The mice studied herein accumulated mercury at similar levels to Swiss albino mice administered for 60 days intraperitoneal 2 mg MeHg/kg body weight (Roos et al. 2010), as well as similar levels to females C57BL/6J exposed via diet to 0.01 mg Hg/kg body weight MeHg during gestation (Montgomery et al. 2008). Notably, females accumulated greater mercury concentrations than males in both hypothalamus and cortex. This observation is consistent with several studies that have demonstrated differences in mercury accumulation patterns between sexes (Hirayama and Yasutake 1986; Hirayama et al. 1987), and with our previous report, where we have shown higher accumulation in females’ than in males’ cortex upon exposure to different doses of MeHg in drinking water (Ferrer et al. 2018). Thus, after a single dose of MeHg, females showed a higher increase in mercury accumulation in the brain (Hirayama and Yasutake 1986; Hirayama et al. 1987). A similar sexual dimorphism was also observed in Wistar rats administrated MeHg via gastric gavage (Magos et al. 1981). Differences in sexual hormones have been suggested as possibly mediating these differences (Hirayama et al. 1987; Oliveira et al. 2006; Malagutti et al. 2009; Yamazaki et al. 2013). In our model, differences in water consumption may explain the higher accumulation in females compared with males.

The brain is the leptin principal site-of-action, particularly in the hypothalamus. The main function of leptin is to provide signaling to the CNS on the long-term energy status of the body (Schwartz et al. 2000). Once it reaches the hypothalamus, it activates molecular mechanisms to induce loss of appetite and body weight. Thus, leptin acts as an anorexigenic signal. In this context, we have shown that MeHg decreased body weight and food intake enhancing leptin’s effects. These effects were more pronounced after 30 days of MeHg exposure, when the changes were statistically significant for both reduced appetite and lower bodyweight, suggesting a time-dependent effect of MeHg. Of note, these effects were seen only in males, whereas females appeared resistant to MeHg’s effect.

Our results showed that MeHg impacted leptin’s actions, altering the JAK2/STAT3 signaling pathway. The main leptin effector is the transcription factor STAT3, which is activated via phosphorylation on its tyrosine 705 residue (Frederich et al. 1995; Vaisse et al. 1996; Banks et al. 2000; Bates et al. 2003). Body weight and food intake changes observed in males coincided with activation of STAT3. Of note, both leptin and MeHg increased STAT3 phosphorylation in males. Upstream STAT3, the main kinase activated by leptin is JAK2. Upon leptin binding to its receptor, JAK2 is phosphorylated and activated. Activated JAK2 phosphorylates the leptin receptor at the tyrosine 1138 residue, a docking site for STAT3. Upon binding to the receptor, STAT3 is phosphorylated by JAK2 and translocates to the nucleus. MeHg induced activation of JAK2/STAT3 signaling pathway, increasing both JAK2 and STAT3 phosphorylation. However, leptin signaling and MeHg alterations were different in females compared with males. Sexual dimorphism in response to leptin has been previously reported. Chronic leptin adenoviral overexpression in Sprague-Dawley male rats developed leptin resistance at an earlier stage than females (Cote et al. 2017). Likewise, Swiss male mice exposed to high fat diet developed leptin resistance before females, fed with an analogous diet (Harris et al. 2003). Contrary to peripheral administration, central acute leptin administration decreased food intake at 24 h in females, but not in males (Clegg et al. 2003). Combined, these results demonstrated that female brains are more sensitive to the acute and chronic leptin effects than males. It has been demonstrated that estradiol administration in males increased leptin sensitivity; whereas in ovariectomized females estradiol restored leptin response at the same levels as in tact females, suggesting a role of sexual hormones in leptin sensitivity (Clegg et al. 2006).

Though no effects of MeHg on food intake and body weight in females were observed, western blot analysis showed that MeHg activated STAT3 in the absence of changes in JAK2 phosphorylation. Although the main activator of STAT3 is JAK2, it has been shown that JAK2-null cells maintained the ability to phosphorylate STAT3 (Jiang et al. 2008). Similarly, other cellular models have also shown that STAT3 could be phosphorylated by JAK2- independent mechanisms (Singh 2011; Looyenga et al. 2012). Thus, it is feasible that in females MeHg activated STAT3 by mechanisms independent of JAK2.

MeHg effects on STAT3 have been reported in several cellular models. MeHg induced a concentration-dependent increase of phospho-STAT3 in N9 microglial cell line (Tan et al. 2019) mediated by JAK2 activation, whereas in SH-SH5Y neuroblastoma cell line and mouse cortical neuronal progenitor cells MeHg enhanced CNTF-induced STAT3 phosphorylation (Jebbett et al. 2013). How MeHg induces STAT3 activation or what are the mechanisms that are affected by MeHg-induced STAT3 has yet to be fully understood. Taken together, we demonstrate, for the first time, that in the hypothalamus MeHg induced JAK2/STAT3 signaling pathway in vivo in males.

Leptin signaling is tightly regulated at multiple points through different mechanisms. SOCS3 expression is induced by leptin-activated JAK2/STAT3 pathway, creating a negative feedback mechanism (Bjorbaek et al. 1999). SOCS3 binds JAK2 directly and inhibits its kinase activities (Sasaki et al. 1999). In addition, SOCS3 can also bind the leptin receptor at the tyrosine domains 985 and 1138, thereby mediating STAT3 inhibition (Bjorbak et al. 2000; Dunn et al. 2005). Another mechanism to attenuate leptin signaling is mediated by PTP1B, which suppresses JAK2 phosphorylation on tyrosine 1007 and 1008 (Zabolotny et al. 2002). In the present study, the increase in Socs3 gene expression in males suggests that MeHg did not impair transcriptional STAT3 functions. In this context, SOCS3 protein levels were decreased, suggesting that MeHg might mediate the induction of JAK2/STAT3 by mechanisms that abolish the inhibitory effects of SOCS3. Concomitant with the decreased levels of SOCS3, MeHg induced an increase in PTP1B. Despite the increased PTP1B protein levels, the JAK2/STAT3 pathway remained activated, suggesting that PTP1B failed may sustain signaling. Due to a MeHg induced decrease in PTP1B activity in human brain microvascular endothelial cells (Yoshida et al. 2017), we cannot discard the possibility that the activity of this protein might be affected by MeHg. To fully elucidate the effect of MeHg on JAK2/STAT3 inhibitors it will be necessary to analyze whether the activity of these proteins is affected by this organometal.

MeHg also increased PTP1B protein levels in females. In addition, we observed a biphasic regulation of the pathway in non-MeHg exposed females, where leptin-increased PTP1B protein levels were followed by an increase in SOCS3, concomitant with a decrease in PTP1B levels. All together, our results suggest that males and females differentially respond to leptin and MeHg treatments.

Huang and coworkers demonstrated that ROCK1 has a role in the hypothalamic leptin effects on energy homeostasis, by interacting with JAK2 and increasing JAK2 phosphorylation and nuclear translocation of STAT3 (Huang et al. 2012). Other studies in peripheral tissues support the role of ROCK proteins in leptin-activated JAK2. In cardiomyocytes and vascular cells leptin was involved in the modulation of cytoskeleton through the ROCK system (Zeidan et al. 2006; Zeidan et al. 2007). In chondrocytes leptin also induced cytoskeletal remodeling via activation of ROCK pathways (Liang et al. 2011). These studies support a role for ROCK1 in mediating leptin activation. MeHg is also able to induce ROCK1 both in vitro and in vivo. In mice microglial primary cultures exposed to MeHg induced ROCK1 expression (Shinozaki et al. 2019), whereas in mice primary culture of astrocytes MeHg exposure also induced ROCK1 cleavage (Dos Santos et al. 2018). Similarly, in vivo, in Sprague-Dawley rats MeHg exposure via drinking water increased ROCK1 and ROCK1 cleaved levels in spinal cord (Fujimura et al. 2019). Combined, these data suggest that ROCK1 mediates some of the toxic effects of MeHg. In our model, in males, leptin induced ROCK1 cleavage, suggesting that the protein was activated. In concordance with the observations on JAK2 phosphorylation, MeHg induced an increase of ROCK1 cleavage, supporting the toxicological effects of MeHg in the hypothalamus. MeHg induced ROCK1 activation might interact with JAK2/STAT3 pathway. Of note, MeHg did not have any effects on ROCK1 in females.

In our model, changes in body weight and food intake appear to be mediated mainly by JAK2/STAT3 signaling pathway. It is known that leptin induces other intracellular signaling pathways, such as PI3K/SHP2/MAPK, AKT-mTOR, and JNK. Whereas MeHg affected some of these pathways, leptin did not show any consistent effects on any of these measured mediators. These data suggest that in our model, MeHg and leptin’s effects on energy homeostasis are mediated predominantly by JAK2/STAT3 signaling pathway.

In addition to its effects in energy homeostasis regulation, leptin is involved in neuronal function, development and survival (de Candia and Matarese 2018; Lee et al. 2019). Leptin’s neuroprotective effects have been described in several studies. For example, leptin-induced anti-apoptotic mechanism in cortical neuronal primary culture after an oxygen-glucose deprivation insult (Zhang et al. 2020) and in glutamate-induced injury in cortical astrocytes culture (Park et al. 2017), increasing Bcl2 in vitro (Zhang et al. 2020). In vivo, intraperitoneal administration of leptin upregulated Bcl2 (Fernandez-Martos et al. 2012). Leptin’s protective effects have also been described in cerebral ischemia/reperfusion and cerebral arterial occlusion models (Amantea et al. 2011; Li et al. 2016). On the other hand, leptin’s effects on oxidative stress are contradictory. Whereas leptin decreased reactive oxygen species in Sprague-Dawley rats submitted to permanent cerebral arterial occlusion model (Hu et al. 2019), in SH-SY5Y neuroblastoma cells leptin did not have effects on MPP⁺-induced ROS production (Ho et al. 2010). Moreover, leptin enhanced SOD2 in serum-deprived cardiomyocytes (Zheng et al. 2010) and in glutamate-induced toxicity in hippocampal neurons (Guo et al. 2008). Thus, it is feasible that leptin might have a protective role against MeHg toxicity. In males, leptin reversed the MeHg-induced increase in the Bax/Bcl2 ratio and decrease in HO-1. In addition, leptin reversed the MeHg-induced decrease in the neuronal marker NeuN, suggesting that leptin is protecting against neuronal loss by anti-apoptotic and/or anti-oxidant mechanisms. Of note, these effects were only significant in males. Our results support a possible protective role of leptin against MeHg-induced toxicity. However, more studies are needed to support this hypothesis.

In addition, our data show that female mice were protected against MeHg insult to a greater extent than males, which has been reported to be linked, at least in part, to neuroprotective effects of sex steroids (Malagutti et al. 2009). The higher levels of neuronal marker NeuN combined with decreased expression of the microglial marker iba-1 suggest that this protection might be related to an anti-inflammatory mechanism. Indeed, evidence suggests that male brains are more vulnerable to toxicants than female brains (Kern et al. 2017) with several studies demonstrating gender-dependent vulnerability (Magos et al. 1981; Hirayama and Yasutake 1986; Malagutti et al. 2009; Biamonte et al. 2014). Little is known about this sexual dimorphism. Differences in anti-oxidants and defense mechanisms have been invoked to explain this different vulnerability. For example, it has been demonstrated that males have lower glutathione availability than females (Lavoie and Chessex 1997), rendering them more sensitive to changes in redox status. Furthermore, females seemed less susceptible to mitochondrial oxidative stress (Borras et al. 2003). These differences might be mediated by sexual hormones. De novo synthesized estradiol ameliorated MeHg-induced death in hippocampal organotypic cultures (Yamazaki et al. 2013). Similarly, in cerebral granular cells estradiol prevented MeHg-induced death (Dare et al. 2000). In vivo,in Swiss mice exposed to MeHg via drinking water the organometal impaired locomotor activities in both males and females, with greater severity in males, and co-treatment of males with estradiol and MeHg prevented the locomotor deficits (Malagutti et al. 2009). These data support the notion of a protective role of estradiol against MeHg-induced neurotoxicity and may explain the sexual dimorphic response to this organometal.

In summary, MeHg alters leptin-activated mechanisms in the murine hypothalamus in males and females (Fig 13). In males MeHg potentiated leptin-induced body weight decrease and appetite loss by enhancing leptin-induced JAK2/STAT3 signaling pathway. MeHg also promoted JAK2 and STAT3 phosphorylation and ROCK1 cleavage. In addition, MeHg-induced decrease in SOCS-3 sustained the MeHg-activation of JAK2/STAT3 pathway. In females, no effects of MeHg on body weight and food intake were observed. However, MeHg induced an increase in STAT3 phosphorylation, which was attenuated by a concomitant increase in PTP1B. In addition to leptin’s effects on energy homeostasis, leptin showed protection against MeHg in males. Leptin induced anti-apoptotic proteins and restored anti-oxidant HO1 protein levels. Figure 13 summarize leptin’s and MeHg’s effects.

Fig. 13.

Summary of MeHg effects on leptin signaling and mitochondrial proteins in males (A) and females (B) Black arrows indicate leptin’s effects. Red arrows indicate MeHg’s effects

In conclusion, this study demonstrates, for first time, that a subchronic MeHg exposure impairs leptin signaling favoring an anorexigenic profile in males. In the hypothalamus, MeHg induced JAK2/STAT3 signaling pathway enhancing leptin effects. These changes are concomitant with a higher mercury accumulation in hypothalamus. This novel hypothesis opens the possibility for additional studies and new approaches to mitigate MeHg-induced neuronal death.

Highlights.

• C57BL/6J female mice accumulated higher mercury concentrations in the hypothalamus than males.

• MeHg potentiated leptin-induced appetite loss and body weight decrease.

• In both males and females MeHg induced the hypothalamic JAK2/STAT3 signaling pathway.

• Leptin induced anti-apototic proteins and restored antioxidant HO1 protein levels in males.

Abbrebiations

GFAP

glial fibrillary acidic protein

HO1

Heme oxygenase 1

JAK2

janus kinase 2

MeHg

methylmercury

NPY

Neuropeptide Y

POMC

pro-opiomelanocortin

PTP1B

protein tyrosine phosphatase 1B

SOCS3

Suppressor of cytokine signaling 3

SOD2

Superoxide dismutase 2

STAT3

signal transducer and activator of transcription 3