Developmental mercury exposure elicits acute hippocampal cell death, reductions in neurogenesis, and severe learning deficits during puberty

Abstract

Normal brain development requires coordinated regulation of several processes including proliferation, differentiation, and cell death. Multiple factors from endogenous and exogenous sources interact to elicit positive as well as negative regulation of these processes. In particular, the perinatal rat brain is highly vulnerable to specific developmental insults that produce later cognitive abnormalities. We used this model to examine the developmental effects of an exogenous factor of great concern, methylmercury (MeHg). Seven-day-old rats received a single injection of MeHg (5 microg/gbw). MeHg inhibited DNA synthesis by 44% and reduced levels of cyclins D1, D3, and E at 24 h in the hippocampus, but not the cerebellum. Toxicity was associated acutely with caspase-dependent programmed cell death. MeHg exposure led to reductions in hippocampal size (21%) and cell numbers 2 weeks later, especially in the granule cell layer (16%) and hilus (50%) of the dentate gyrus defined stereologically, suggesting that neurons might be particularly vulnerable. Consistent with this, perinatal exposure led to profound deficits in juvenile hippocampal-dependent learning during training on a spatial navigation task. In aggregate, these studies indicate that exposure to one dose of MeHg during the perinatal period acutely induces apoptotic cell death, which results in later deficits in hippocampal structure and function.

Fig. 1

Methylmercury (MeHg) acutely inhibits DNA synthesis specifically in the hippocampus. (a) Dose-dependent effects of MeHg on [³H]-thymidine ([³H]-Thy) incorporation in the hippocampus. P7 rats were injected subcutaneously with vehicle or 2.5, 5.0, 7.5, or 10.0 μg/gbw MeHg at zero time, with [³H]-Thy at 22 h and killed at 24 h to assess incorporation. MeHg induced a dose-dependent decrease in [³H]-Thy incorporation. (b) Time course of MeHg effects on [³H]-Thy incorporation in the hippocampus. P7 rats were injected with 5 μg/gbw MeHg and killed 2, 6, 8, 10, and 24 h after injection. The effect of MeHg was detectable after 6 h, and maximal at 24 h. (c) Comparison of MeHg effects in hippocampus and cerebellum. P7 rats were injected with vehicle or 5.0 μg/gbw MeHg and killed 24 h after injections. The cerebellum was not affected by MeHg, suggesting that hippocampal DNA synthesis is specifically vulnerable to organomercurials. (d) Quantification of the number of cells engaged in mitotic S-phase in the hilus of the hippocampus. P7 rats were injected with 5 μg/gbw MeHg at zero time, with bromodeoxyuridine (BrdU) 22 h later and processed at 24 h for BrdU staining. MeHg exposure diminished the number of BrdU-positive cells in the hilus of the hippocampus, suggesting a possible block in the G1/S phase transition. Pictures show representative BrdU staining in control condition and after MeHg exposure. Arrows identify BrdU-positive cells. Values are expressed as the mean ± SEM of four independent experiments for all groups, with three animals per group in each experiment. *p < 0.05; **p < 0.01; and ***p < 0.001 versus control.

Fig. 2

Methylmercury (MeHg) exposure induces sustained modifications of hippocampal cell number. (a) Consequences of MeHg exposure on cells that were proliferating on P7. P7 rats were injected with vehicle or 5.0 μg/gbw MeHg at zero time and with bromodeoxyuridine (BrdU) 6 h later. Animals were killed 2 weeks after injections and processed for BrdU staining. BrdU-positive cells (arrows) were counted in granule cell layer (GL), hilus (H), and molecular layer (M). The mean number of immunopositive cells per section was significantly decreased following MeHg exposure in both granule cell layer and hilus, but not in molecular layer. (b) Effect of acute MeHg exposure on later DNA content in hippocampus and cerebellum. P7 rats were injected with vehicle or 5.0 μg/gbw MeHg and killed 2 weeks after injections. Mercury exposure reduced DNA content 2 weeks later specifically in the hippocampus. **p < 0.01 and ***p < 0.001 versus control.

Fig. 3

Methylmercury (MeHg) exposure affects hippocampal dentate gyrus structure and cellular composition. (a) Estimation of the size of different layers composing dentate gyrus and Ammon’s horn 2 weeks after MeHg exposure. P7 rats were injected with vehicle or 5.0 μg/gbw MeHg and areas were measured using the Bioquant morphometry analysis tool. Dentate gyrus subregion sizes were reduced in the MeHg exposed group, whereas Ammon’s horn was unchanged. Photomontage picture (left) shows the different hippocampal layers that were analyzed. fi, fimbria; GL, granule cell layer; H, hilus; M, molecular layer; P, pyramidal cell layer; SO, stratum oriens; and SR, stratum radiatum. Data are expressed as the mean area ± SEM per section of at least six sections per animal, three animals per group. (b) Estimation of dentate gyrus cell number 2 weeks after MeHg exposure. P7 rats were injected with vehicle or 5.0 μg/gbw MeHg and cell number was measured 2 weeks after injections in granule cell layer, hilus and molecular layer by unbiased stereology (See Materials and methods). Pictures show high magnification of cells from GL (top) and hilus (bottom). MeHg induced a decrease in cell number in granule cell layer and hilus, but not in the molecular layer, suggesting that MeHg toxicity is cell type-specific. Values are expressed as the mean cell number ± SEM per region of three independent experiments for all groups, with three animals per group in each experiment. *p < 0.05; **p < 0.01; and ***p < 0.001 versus control. Scale bars = 50 μm.

Fig. 4

Effect of methylmercury (MeHg) on hippocampal levels of cyclin D1 (a), D3 (b), and full length and cleaved forms of cylin E (c). P7 rats were injected with vehicle or 5.0 μg/gbw MeHg and processed after 24 h for protein extraction. MeHg induced moderate decreases in levels of cyclin D1, D3, and E. Notably, the decrease in cyclin E was associated with an increase in the level of cleavage fragment p18 (c). Densitometric quantification, expressed as arbitrary units, was performed on three independent experiments, with two animals per group in each experiment. *p < 0.05 and ***p < 0.001 versus control.

Fig. 5

Involvement of programmed cell death in methylmercury (MeHg) -induced toxicity. (a–d) Caspase 3 immunostaining in the dentate gyrus of the hippocampus. P7 rats were injected with vehicle or 5.0 μg/gbw MeHg, killed 24 h after injections and processed for immunostaining (see Materials and methods). (a and b) Low magnification of hippocampal dentate gyrus in control (a) and MeHg-treated (b) groups. In control animals, very few caspase 3-positive cells were detected in the hippocampus, while a vastly higher number was found in MeHg-treated animals as shown by arrows. (c and d) Boxed areas in (a and b) are presented at higher magnification in (c and d), respectively. Arrows identify caspase 3-positive cells. H, Hilus of the dentate gyrus; GL, granule layer; and M, molecular layer. Scale bars = 50 μm. (e) Effect of MeHg on hippocampal levels of active caspase 3. P7 rats were injected with vehicle or 5.0 μg/gbw MeHg and processed after 24 h for protein extractions. MeHg induced a marked increase in active caspase 3 levels. Densitometric quantification, expressed as arbitrary units, was performed on three independent experiments, with two animals per group in each experiment. (f) Levels of oxidative stress in rat hippocampus following MeHg exposure. P7 rats were injected with vehicle or 5.0 μg/gbw MeHg and brain reactive oxygen species levels were determined after 24 h by measuring dichlorofluorescein (DCF) diacetate oxidation (see Materials and methods). MeHg induced a significant increase in hippocampal reactive oxygen species levels. Values are expressed as the mean DCF fluorescence ± SEM of three independent experiments for all groups, with three animals per group in each experiment. ***p < 0.001 versus control.

Fig. 6

Involvement of caspases in the effect of methylmercury (MeHg) on cell cycle machinery. (a) Measurement of [³H]-thymidine ([³H]-Thy) incorporation in cultured neurons 20 h after treatment with MeHg (1.5 μmol/L), Z-VAD-FMK (60 μmol/L), or both. MeHg induced a decrease in thymidine incorporation which was almost completely reversed by addition of Z-VAD-FMK. Values are expressed as the mean ± SEM of five independent experiments for all groups, with four wells per group in each experiment. (b) Effect of MeHg and Z-VAD-FMK on cyclin E degradation. Western blot showing the effect of MeHg (1.5 μmol/L), Z-VAD-FMK (60 μmol/L), or both on cyclin E level. Densitometric quantification, expressed as arbitrary units, was performed on three independent experiments, with two dishes per group in each experiment. (c) Evidence of inhibitory activity of Z-VAD-FMK for MeHg-induced activation of caspase 3. Cultured neurons were treated 20 h with or without MeHg (1.5 μmol/L) and/or Z-VAD-FMK (60 μmol/L). MeHg treatment induced an increase in active caspase 3 level, which was prevented by Z-VAD-FMK. Blots were performed on two independent experiments, with two dishes per group in each experiment. **p < 0.01 and ***p < 0.001 versus control; ^#p< 0.05 versus MeHg.

Fig. 7

Detection of activated caspase 3 immunoreactivity in proliferative cells in the dentate gyrus following methylmercury (MeHg) exposure. P7 rats were injected with vehicle or 5.0 μg/gbw MeHg, killed at 8 and 24 h and processed for immunostaining. Double-labeling experiments demonstrate that 8 h after MeHg exposure a proportion of bromodeoxyuridine-labeled cells (green) is also caspase 3-positive (red). At 24 h, no double labeling was observed. Optical sectioning 3-dimensional analysis (Zeiss Apotome) of XZ and YZ orthogonal planes confirms colocalization of bromodeoxyuridine and caspase 3 signals. Arrowheads identify double-labeled cell. Scale bar = 25 μm.

Fig. 8

Effect of methylmercury (MeHg) on hippocampal-dependent learning processes. (a) Acquisition of place learning. P7 rats were injected with vehicle or 5.0 μg/gbw MeHg and trained at P35 four times per day to find a hidden platform in a Morris water maze (see Materials and methods). Control animals (green line) exhibited a progressive asymptotic decrease in escape latency suggesting a correct learning pattern. In contrast, the MeHg-treated animals (red line) had poor improvement of their performances. (b) Long-term recovery test. Long-term memory was measured 2 weeks later (P52) by a single trial under the same conditions as those used during training. The escape latency was twofold longer in MeHg-treated animals, suggesting that they failed to learn platform position. (c) Measurement of motor and visual abilities. Non-cognitive performances were compared by using a visible platform. No differences were found between groups suggesting that MeHg specifically alters memory processes, without damaging the animals vision and locomotion. Experiments were performed on two independent litters, with four to five animals per group in each litter. (d) Measurement of DNA content at 35 days of age. P7 rats were injected with vehicle or 5 μg/gbw MeHg and killed at P35. At the age animals were trained in water maze, hippocampal DNA was still reduced, suggesting the absence of recovery from the deficit observed at P21. *p < 0.05; **p < 0.01; and ***p < 0.001 versus control.

Fig. 9

Schematic representation of the signaling mechanisms likely involved in the toxic effects of MeHg during hippocampal neurodevelopment. MeHg exposure reduces cyclin D1 and D3 levels, and degrades cyclin E through a caspase-dependent mechanism. These effects result in a cell cycle arrest at the same time as an induction of cell death. As a consequence, hippocampal neurodevelopment is altered, resulting in later abnormalities in spatial learning abilities. MeHg, methylmercury; Z-VAD-FMK, general caspase inhibitor; ROS, reactive oxygen species; ↑, activation; ⊥, inhibition. Effects of MeHg are visualized in red.