Methylmercury Antagonizes the Survival-Promoting Activity of Insulin-like Growth Factor on Developing Cerebellar Granule Neurons

Methylmercury Antagonizes the Survival-Promoting Activity of Insulin-like Growth Factor on Developing Cerebellar Granule Neurons

TOXICOLOGY AND APPLIED PHARMACOLOGY ARTICLE NO. 153, 161–168 (1998) TO988561 Methylmercury Antagonizes the Survival-Promoting Activity of Insulinli...

552KB Sizes 5 Downloads 35 Views

TOXICOLOGY AND APPLIED PHARMACOLOGY ARTICLE NO.

153, 161–168 (1998)

TO988561

Methylmercury Antagonizes the Survival-Promoting Activity of Insulinlike Growth Factor on Developing Cerebellar Granule Neurons Robert F. Bulleit and Hong Cui Department of Pharmacology, University of Maryland School of Medicine, Baltimore, Maryland 21201 Received June 29, 1998; accepted September 2, 1998

Methylmercury Antagonizes the Survival-Promoting Activity of Insulin-like Growth Factor on Developing Cerebellar Granule Neurons. Bulleit, R. F., Cui, H. (1998). Toxicol. Appl. Pharmacol. 153, 161–168. Methylmercury (MeHg), a widely distributed environmental toxicant, has a profound effect on the developing central nervous system. Human exposure to MeHg in utero has led to severe neurological abnormalities in children, including cognitive and motor dysfunction. The abnormalities appear to result from death of neurons and altered cytoarchitecture in the developing CNS. Death of cerebellar granule neurons occurs following both adult and in utero exposure to MeHg, indicating the vulnerability of these cells to the toxic action of MeHg. The studies reported here use purified cultures of developing mouse cerebellar granule neurons to evaluate whether MeHg directly acts on these developing neurons to inhibit their survival. These experiments show that, in purified cultures of cerebellar granule neurons maintained in medium containing insulin-like growth factor I (IGF-I) as the only added trophic factor, low micromolar concentrations of MeHg inhibit granule neuron survival. The reduction in survival produced by MeHg can be partially reversed by increasing the concentration of IGF-I, suggesting an antagonism between MeHg and IGF-I. Inhibition of phosphoinositide 3-kinase (PI3-K), an intracellular mediator of IGF-I’s survival promoting action, can synergistically enhance MeHg’s effect on survival. Further studies indicate that MeHg’s inhibition of survival involves apoptotic death of granule neurons. This apoptosis appears to require activation of gene transcription and may involve an increase in expression of the immediate early transcription factor c-Jun. These studies suggest that MeHg can act on developing granule neurons to increase the expression of c-Jun and antagonize IGF-I’s survival promoting activity. © 1998 Academic Press Key Words: IGF-I; mouse; apoptosis; transcription factor; c-Jun.

Unlike the mature brain, those of the fetus and neonate undergo tremendous changes requiring the regulation of cellular growth, survival, migration, and differentiation. All of these processes may be disrupted by exposure to toxicants, suggesting that the developing brain may be more vulnerable to toxic insult. Methylmercury (MeHg) is an example of a widely distributed environmental toxicant that has a profound affect on the developing nervous system. Human exposure to MeHg

has led to catastrophic outcomes, particularly when the exposure occurs in utero (Takeuchi, 1977; Choi et al., 1978). Children exposed in utero were born with severe neurological abnormalities, including cognitive and motor dysfunction (Harada, 1978). In both humans and animals, MeHg causes similar neuropathologic symptoms, including CNS cell loss, altered cellular layering, and other cytoarchitectural abnormalities (Burbacher et al., 1990). A common observation following both adult and in utero exposure to MeHg is the loss of granule neurons in the cerebellum, suggesting that the cerebellum is particularly vulnerable to damage (Jacobs et al., 1986; Sager et al., 1982, 1984; Burbacher et al., 1990; Nagashima et al., 1996). Ultimately this reduction in neurons results from either a decrease in mitotic activity of granule neuron progenitors or an increase in death of developing granule neurons (Sager et al., 1982, 1984). Previous reports suggest that MeHg can produce granule cell death in both the developing and mature cerebellum (Sager et al., 1982; Syversen et al., 1981; Nagashima et al., 1996). These studies point to the potential importance of cell death in producing the abnormalities associated with MeHg intoxication. Death of cerebellar neurons would reduce the appropriate number of neurons and their connections, potentially interfering with cerebellar functions. In the developing cerebellum it could also reduce the appropriate number of cells needed for critical cell– cell interactions that regulate other developmental processes such as neuronal migration, proliferation, survival, and differentiation. An understanding of the mechanism by which MeHg produces granule cell death will thus provide insights into the molecular nature of MeHg’s developmental neurotoxicity. MeHg could either act directly on granule neurons to induce cell death or alternatively MeHg could inhibit the survival or function of other cell types, such as glia or cerebellar Purkinje neurons, reducing their ability to support granule cell survival. These possibilities can be distinguished by evaluating the effect of MeHg on a pure population of cerebellar granule cells in vitro. In order for neurons to be maintained in culture, trophic factors must be supplied in the culture medium. In vivo neuron survival depends on specific trophic factors secreted by the cells they innervate and/or by other neighboring cells such as glia (Purves, 1986; Korshing, 1993). Developing neurons that do not receive the appropriate amount of trophic support

161

0041-008X/98 $25.00 Copyright © 1998 by Academic Press All rights of reproduction in any form reserved.

162

BULLEIT AND CUI

die, usually by apoptotic mechanisms (Oppenheim, 1991; Johnson and Deckwerth, 1993). Insulin-like growth factor I (IGF-I) can promote in vitro survival of several CNS neuron types (Knusel et al., 1990; Ang et al., 1992; Beck et al., 1993; Hughes et al., 1993; Neff et al., 1993). In the developing cerebellum, IGF-I is predominantly produced by Purkinje neurons, the synaptic targets of granule neurons (Bondy, 1991; Lee et al., 1992). The peak production of IGF-I by Purkinje neurons occurs during the postnatal period of cerebellar granule neuron development. Thus, the temporal and spatial pattern of IGF-I expression suggests that it may function as a targetderived trophic factor for developing cerebellar granule neurons. In recent studies we showed that IGF-I added to pure cultures of cerebellar granule cells can, in the absence of other growth factors, maintain their survival and allow continued differentiation in culture (Lin and Bulleit, 1997). This culture system thus provides a way to minimize the complexity of granule neurons’ environment and influences of other cells in order to allow an evaluation of whether MeHg can directly act on developing granule neurons to inhibit their survival. In the present study, using purified cultures of developing granule neurons, we show that MeHg has a direct action on these developing neurons. MeHg at low micromolar concentrations can inhibit granule neuron survival. The degree of inhibition is reduced by increasing the concentration of IGF-I in the culture medium. We also observed that inhibition of phosphoinositide 3 kinase (PI3-K), an intracellular mediator of IGF-I’s survival-promoting action, acts synergistically with MeHg to produce cell death and that this cell death occurs by an apoptotic mechanism. These results suggest that MeHg induces apoptosis in these neurons, antagonizing IGF-I’s ability to support survival. These studies might also suggest that MeHg may initiate a cell death program. Previous studies suggest that an increase in expression of the immediate early transcription factor c-Jun may be part of a genetic program initiating apoptotic cell death in neurons (Ham et al., 1995; Miller and Johnson, 1996; Watson et al., 1998). Our studies show for the first time that MeHg can induce a prolonged increase in c-Jun expression in granule neurons, indicating a potential role for this transcription factor in MeHg-induced neuronal death.

estimates (Hatten, 1985). Granule cells were cultured at a density of 1 3 107 cells per 35-mm petri dish in MEM containing IGF-I. No serum was added to the culture medium. Methylmercury chloride (Alfa Products Inc., Danvers, MA) was dissolved in dimethyl sulfoxide at a concentration of 10 mM. This stock solution was diluted in culture medium and added at the start of culture. Determination of viable cell number. Viable cell number was determined as described previously (Volonte et al., 1994; Lin and Bulleit, 1997). Briefly, granule cells were lysed in a detergent-containing solution (0.5% ethylhexadecyldimethylammonium bromide, 0.28% acetic acid, 0.5% Triton X-100, 3 mM NaCl, 2 mM MgCl2, in PBS pH 7.4). Under these conditions viable nuclei remained intact and could be easily distinguished from the small, broken, or phase-bright apoptotic nuclei. Viable nuclei were counted in a hemocytometer. We determined viable nuclei counts from three to five separate cultures for each treatment. Statistical significance between different treatments was evaluated by a two-tailed Student’s t-test. DNA fragmentation assay. We isolated DNA from granule cell cultures using procedures described by Hockenbery et al. (1990). Briefly, granule cells were lysed in 0.5% Triton X-100, 5 mM Tris (pH 7.5), and 20 mM EDTA at 4°C for 20 min. The cell lysate was extracted with phenol and chloroform and incubated for 30 min at 37°C in the presence of RNase A. DNA was precipitated in ethanol and fractionated by electrophoresis using 1.8% agarose gel. The DNA fragmentation pattern was visualized by ethidium bromide staining. Thymidine incorporation. We cultured granule cells in MEM containing 50 ng/ml IGF-I and different concentrations of MeHg for 24 h. 3H-Thymidine (2.5 mCi/ml, NEN, Boston, MA) was added for the last 2 h of culture. This addition would label those granule neuron progenitors in S phase (the DNA synthesis phase) of the cell cycle after 24 h of culture. Following the addition of 3H-thymidine, granule cells were lysed by the addition of water, and DNA was precipitated with 10% trichloroacetic acid (TCA). TCA pellets were dissolved in 0.1 mM NaOH and incorporation of 3H-thymidine was determined by scintillation counting of an aliquot of the dissolved TCA pellets. Statistical significance of differences between treatment conditions was determined using a Student’s t-test. RNA blot analysis. RNA blot analysis was performed as described previously (Bulleit et al., 1994). Total RNA was isolated by acid guanidinium thiocyanate–phenol– chloroform extraction (Chomczynski and Sacchi, 1987). RNA was fractionated by formaldehyde agarose gel electrophoresis and subsequently transferred to nylon membranes. The RNA was crosslinked to the membrane by a UV-crosslinker (Strategene, La Jolla, CA). RNA blots were prehybridized for 2 to 4 h at 42°C in H buffer (5.63 SSPE [0.84 M NaCl, 64 mM Na2PO4, 6 mM EDTA], 50% formamide, 53 Denhardt’s solution [0.1% Ficoll, 0.1% polyvinylpyrrolidone, 0.1% bovine serum albumin], 1% SDS, and 200 mg/ml herring sperm DNA). The hybridization reaction used H buffer containing 1 to 3 3 106 cpm/ml of probe (specific activity of 1 to 2 3 109 cpm/mg) for 48 h at 42°C. The mouse c-Jun cDNA probe was a 2.6-kb cDNA, JAC.1, obtain from American type culture collection (Rockville, MD). Blots were washed under high stringency conditions (0.13 SSC, 1% SDS at 65°C) and exposed to X-ray film (DuPont).

METHODS

RESULTS

Granule cell culture. We established pure cerebellar granule cell cultures using a Percoll gradient procedure described previously (Hatten, 1985; Lin and Bulleit, 1997). Cerebellar cells were dissociated from Postnatal Day 7 CD-1 mice using trypsin and DNase. Dissociated cells were collected by centrifugation, and granule cells were separated from larger cell types by centrifugation through a 35%/60% Percoll gradient. The granule cell fraction was collected at the 35%/60% Percoll interface. To further purify the granule cell fraction, the cells were subjected to plating on poly-L-lysine-coated tissue culture dishes in minimal medium (MEM) supplemented with 5 mg/ml glucose (Sigma, St. Louis, MO), 2 mM glutamine, 50 U/ml penicillin, and 50 mg/ml streptomycin (Gibco/BRL, Grand Island, NY) containing 10% horse serum. Granule cells can be washed off the plates while the glia remain attached. This method yielded a granule cell purity of 95 to 98%, consistent with previous

We used purified cerebellar granule cells cultured in minimal medium containing IGF-I, 10 ng/ml, to evaluate the effects of MeHg on their survival. This concentration of IGF-I maintains survival but has only a minimal effect on stimulating granule cell proliferation (Lin and Bulleit, 1997). We evaluated survival over the first 2 days of culture. Over this period of time, cultured granule cells begin to differentiate and send out neurite processes. MeHg in a concentration-dependent fashion reduced the number of live cells (Fig. 1). After 1 day in culture, 2 mM MeHg reduced the number of live granule cells compared to cultures maintained in the absence of MeHg (Fig. 1A).

METHYLMERCURY ANTAGONIZES IGF-I-INDUCED NEURON SURVIVAL

FIG. 1. MeHg’s effect on granule cell survival promoted by IGF-I. Purified granule cell cultures were prepared as described in Methods and were maintained for (A) 24 h or (B) 48 h in MEM containing 10 ng/ml IGF-I and 0 to 2 mM MeHg. The number of live cells was determined by a nuclei counting assay (see Methods). The percent survival was determined by comparing the number of initially plated live cells with the number after 24 or 48 h of culture. Presented is the mean cell survival percentage 6 SEM for three separate experiments. *Statistically significant differences in mean percent survival from cultures maintained without MeHg, p , 0.05.

Following 2 days of culture, concentrations of MeHg from 0.25 to 2 mM significantly reduced granule cell survival, with a half maximal inhibition occurring between 1 and 2 mM MeHg (Fig. 1B). We further evaluated whether changing the concentration of IGF-I could alter MeHg’s effect on granule cell survival. In a concentration-dependent fashion, IGF-I reduced the survival inhibition produced by MeHg. Concentrations of IGF-I higher than 5 ng/ml increased the percent survival of granule cells cultured in 2 mM MeHg (Fig. 2). MeHg had its greatest effect on reducing survival when granule cells were maintained with-

FIG. 2. IGF-I attenuates the inhibition of granule cell survival induced by MeHg. Granule cells were maintained for 24 h in MEM containing 0 to 50 ng/ml IGF-I and 2 mM MeHg. The percent control survival is determined by comparing the number of live cells in control cultures maintained in the absence of MeHg with cultures maintained in the presence of 2 mM MeHg for each concentration of IGF-I. Presented is the mean percent control survival 6 SEM for three separate experiments. *Statistically significant differences in mean percent control survival from cultures maintain without IGF-I, p , 0.05.

163

FIG. 3. Effect of MeHg on 3H-thymidine incorporation in granule cell cultures maintained in 50 ng/ml IGF-I. Purified granule cell cultures were maintained for 24 h in MEM containing 50 ng/ml IGF-I and 0 to 2 mM MeHg. 3 H-Thymidine labeling and measurement of incorporation are described in Methods. Presented is the mean incorporation 6 SEM in three separate experiments. *Statistically significant differences from cultures maintained without MeHg, p , 0.05.

out IGF-I, suggesting MeHg may have a direct action on granule cell survival and not simply reduce their access to IGF-I. Previous studies indicate that increasing the concentrations of IGF-I can stimulate proliferation of granule neuron progenitors (Lin and Bulleit, 1997). We evaluated whether MeHg could also antagonize IGF-I’s ability to stimulate proliferation in cultured granule cells. Purified granule cells were incubated in MEM containing 50 ng/ml IGF-I and 0 to 2 mM MeHg. Cells were cultured under these conditions for 24 h. We added 3 H-thymidine during the last 2 hours of culture. This labeling method provides a measure of the number of cells remaining in the proliferative cycle after 24 h in culture. MeHg, at 1 and 2 mM, reduced thymidine incorporation in cultured granule cells (Fig. 3), indicating that it reduced the number of granule cells in the proliferative cycle. The degree of reduction in thymidine incorporation, 27%, induced by 2 mM MeHg is similar to the level of reduction in cell survival, 34% (Fig. 2). If MeHg directly induces granule cell death, then reducing IGF-I signaling may enhance the detrimental effect of MeHg on cell survival by removing the protective effect of IGF-I. We tested this hypothesis by using a selective inhibitor of PI3-K, LY294002 (Vlahos et al., 1994). IGF-I’s survival-promoting activity appears to be dependent on activation of PI3-K (Dudek et al., 1997). At a concentration of 10 mM, which is about seven times the IC50 value for inhibition of PI3-K, LY294002 could itself reduce IGF-I-dependent survival. This reduction was greatest after 3 days of exposure (Fig. 4A). After 24 h of exposure there was a 12% reduction in survival (Fig.4A). In contrast, exposure to MeHg for 24 h reduced survival by 35% (Fig. 2). We also observed that LY294002 in a concentrationdependent fashion enhanced MeHg’s ability to induce granule cell death (Fig. 4B). The half maximal enhancement for LY294002 occurred between 1 and 2 mM. This concentration is similar to the previously reported IC50 value for inhibition

164

BULLEIT AND CUI

FIG. 4. Inhibition of PI3-K enhances MeHg-induced granule cell death. Purified granule cell cultures were prepared as describe in Methods. (A) Granule cells were maintained for 24 h in MEM containing 50 ng/ml IGF-I and 10 mM LY294002. The percent control survival is determined by comparing the number of live cells in treated cultures with control cultures maintained 24 h in 50 ng/ml IGF-I. Presented are the means 6 SEM percent control survival in three separate experiments. *Statistically significant differences from control cultures, p , 0.05. (B) Granule cells were maintained for 24 h in MEM containing 50 ng/ml IGF-I, 2 mM MeHg, and 0 to 25 mM LY294002. The percent survival was determined as described in Fig. 1. Presented are the means 6 SEM percent survival in three separate experiments. *Statistically significant differences from cultures maintained in the absence of LY294002, p , 0.05.

of PI3-K (1.4 mM, Vlahos et al., 1994). A comparison of granule cells, maintained in 50 ng/ml IGF-1 and treated with 2 mM MeHg and LY294002, showed that together they increased cell death in a synergistic fashion (Fig. 5). MeHg’s toxic action on developing granule cells may involve apoptotic cell death. Consistent with this hypothesis is the observation that granule cell cultures exposed to MeHg contained a large number of small phase-bright apoptotic bodies (Fig. 6B). Moreover, an oligonucleosome-sized pattern of DNA fragmentation was observed in DNA isolated from 2-day-old cultures maintained in 2 mM MeHg (Fig. 6C). These characteristics are hallmarks of apoptotic death (Wyllie et al., 1980). An examination of granule cells treated with MeHg and LY294002 also showed a large number of apoptotic bodies compared to cultures maintained in IGF-I alone (Figs. 6D and 6E). Further, experiments examining DNA isolated from these cultures showed that MeHg and LY294002 produce fragmentation of DNA (Fig. 6F). These observations indicate that the combination of MeHg and LY294002 also produces apoptosis. Protein and RNA synthesis inhibitors can block apoptosis following survival factor withdrawal in cultured neurons, sug-

gesting apoptotic death may involve activation of gene expression (Martin et al., 1988; D’Mello et al., 1993; Milligan et al., 1994; Watson et al., 1998). MeHg’s action may also require an increase in the expression of certain genes. To test this possibility, we maintained granule cell cultures for 24 h in medium containing 10 ng/ml IGF-I, 2 mM MeHg, and 1 mg/ml actinomycin D, an RNA synthesis inhibitor. Previous studies showed that this concentration of actinomycin D could protect granule neurons from cell death induced by trophic factor withdrawal (D’Mello et al., 1993). MeHg reduced granule cell survival by 45% while cultures maintained in MeHg plus actinomycin D had only a 21% reduction in live cells (Fig. 7A). This difference was statistically significant. Increasing the concentration of actinomycin D to 10 or 100 mg/ml did not increase its effect on survival (data not shown). Although this experiment might suggest that MeHg’s induction of cell death may in part be dependent on increased gene transcription, it does not provide an indication of what specific gene might be induced following MeHg exposure. In several studies an increase in the expression of the transcription regulatory factor c-Jun has been associated with induction of apoptotic neuronal death (Ham et al., 1995; Miller and Johnson 1996; Bossy-Wetzel et al., 1997; Watson et al., 1998). We evaluated the level of c-Jun RNA following treatment of cultured granule cells with 10 ng/ml IGF-I and 0 or 2 mM MeHg. The level of c-Jun increased rapidly within 2 h of exposure following treatment of granule cells with 2 mM MeHg (Fig. 7B). The level of c-Jun continued to increase up to 8 h after exposure. In control cultures maintained in IGF-I without MeHg, the level of c-Jun remained stable over this same time of culture. The change in the level of c-Jun RNA was not due to a difference in RNA loaded on the gel since similar amounts of 28S RNA are present in all lanes of the gel. DISCUSSION

This report indicates that MeHg antagonizes the survival of developing cerebellar granule neurons maintained in medium

FIG. 5. MeHg and LY294002 synergistically enhance granule cell death. Granule cells were maintained for 24 h in MEM containing 50 ng/ml IGF-I, 0 or 2 mM MeHg, and 0, 10, or 25 mM LY294002. Percent survival was evaluated as described in Fig. 1. Presented are the means 6 SEM percent survival in three separate experiments. *Significant differences from cultures maintain in MeHg without LY294002, p , 0.05.

METHYLMERCURY ANTAGONIZES IGF-I-INDUCED NEURON SURVIVAL

165

FIG. 6. MeHg and MeHg plus LY294002 induce apoptotic death in developing cerebellar granule neurons. Phase-contrast photomicrographs at 2003 of granule cells maintained for 48 h in MEM containing (A) 10 ng/ml IGF-I or (B) 10 ng/ml IGF-I and 2 mM MeHg. (C) A comparison of DNA isolated from granule cells before culture (Control) and 48 h after the start of culture in cells treated with 10 ng/ml IGF-I (IGF-I) or 10 ng/ml IGF-I and 2 mM MeHg (IGF-I 1 MeHg). A 123-bp DNA ladder (123 bp DNA) was used as a molecular size standard. Phase-contrast photomicrographs at 2003 of granule cells maintained for 24 h in MEM containing (D) 50 ng/ml IGF-I or (E) 50 ng/ml IGF-I, 2 mM MeHg, and 10 mM LY294002. (F) A comparison of DNA isolated from granule cells maintained 24 h in 50 ng/ml IGF-I (IGF-I) or 50 ng/ml IGF-I, 2 mM MeHg, and 10 mM LY294002 (IGF-I 1 MeHg LY294002). A 123-bp DNA ladder (123 bp DNA) was used as a molecular size standard.

166

BULLEIT AND CUI

FIG. 7. MeHg increases new gene expression. Purified granule cell cultures were prepared as described in Methods. (A) Granule cells were maintained for 24 h in MEM containing 10 ng/ml IGF-I, 0 or 2 mM MeHg, and 0 or 1 mg/ml actinomycin D. Percent survival was evaluated as described in Fig. 1. Presented are the means 6 SEM for three separate experiments. *Statistically significant difference from cultures maintained in IGF-I and MeHg, p , 0.05. (B) Granule cells were maintained in minimal medium containing 10 ng/ml IGF-I and 0 or 2 mM MeHg. RNA was isolated from these cells after 1, 2, 4, or 8 h of culture and used for RNA blot analysis with a c-Jun cDNA probe. Presented is an autoradiograph showing the c-Jun hybridization signal. Also presented is a photograph of the ethidium bromide-stained RNA gel used for the blot. The staining of the 28S RNA indicates that similar amounts of RNA are in each lane of the gel.

containing IGF-I as the only added trophic factor. This inhibition of survival involves the induction of apoptotic cell death in these developing neurons. Previous studies reported that MeHg induces apoptotic death in mature granule neurons (Kunimoto 1994; Nagashima et al., 1996). However, these studies did not clearly define whether MeHg had a direct action on granule neurons. Our studies, using purified cultures of granule neurons in a minimal culture environment, suggest that MeHg’s effect on granule neuron survival is not dependent on reducing survival signals produced by other cerebellar cell types. Alternatively, and not unexpectedly, MeHg can act directly on granule neurons to inhibit their survival. Thus, the reduction in granule neuron number following MeHg exposure in vivo may be a consequence of MeHg directly producing cell death of developing granule neurons. It is also possible that MeHg can reduce granule cell proliferation. We observed that MeHg could reduce the incorporation of 3H-thymidine, a measure of the number of proliferative granule cells in culture, indicating a potential antimitotic effect. However, the percent reduction in DNA synthesis we observed (27%) was very similar to the percent reduction in total cell survival (34%). Thus, an alternative explanation for the decrease in thymidine incorporation

is that MeHg reduces survival of proliferative granule cells. However, we cannot completely rule out the possibility that MeHg also has a specific effect on inhibition of DNA synthesis. Previous analysis of murine erythroleukemic cells indicated that MeHg cytotoxicity may be related to a reduced rate of DNA synthesis in these cells (Zucker et al., 1990). Other studies also indicate that alteration in cell cycle progression may be associated with apoptosis in the developing CNS (Thomaidou et al., 1997). Thus, an alteration in cell cycle regulation by MeHg may be partially responsible for induction of apoptosis in proliferating granule cells. In either case, our experiments would indicate that cell death induced by MeHg in utero is likely to be an important action responsible for the loss of neurons in the developing cerebellum as well as other brain regions. The concentration of 2 mM MeHg used in these studies produces apoptosis in a majority of granule cells over 2 days of culture (Fig. 1). This concentration is equivalent to approximately 0.5 ppm in culture medium. This concentration is below the estimated brain concentration, 3 to 20 ppm, that produces cell loss and neuropathological effects in developing human and mammalian nervous systems (Burbacher et al., 1990). This observation suggests that this culture system, where neurons are removed from the environment of the brain and have minimal trophic support, may increase the susceptibility of granule neurons to methylmercury cytotoxicity. It is also possible that the local concentrations of MeHg around neurons in the brain are actually lower then would be predicted from concentrations measured in total brain tissue because of the presence of other cell types that might sequester or remove MeHg from the local environment. However, our studies suggest that concentrations of MeHg, close to in situ concentrations that produce neuropathologic effect in humans and other mammals, can directly produce apoptosis in cultured granule neurons, providing a model to further evaluate the intracellular mechanisms by which MeHg produces apoptotic neuronal death. In our granule cell cultures, IGF-I is the only added trophic factor. Previous in vitro studies indicate that IGF-I can function as a trophic factor for many cell types, including cultured cerebellar granule neurons (Gao et al., 1991; D’Mello et al., 1993; Lin and Bulleit, 1997). A number of in vivo studies also suggest a trophic role for IGF-I in maintaining granule cell survival (Kar et al., 1993; Bach et al., 1991; Lee et al., 1992; Beck et al., 1995; Ye et al., 1995). Although all these studies support a trophic role for IGF-I on granule neurons, it is likely that in vivo multiple trophic factors help to maintain granule cell survival. By using IGF-I as the only trophic factor, our culture system provides a means of testing between two possible mechanisms by which MeHg induces apoptotic cell death. MeHg could produce apoptosis either by inhibiting IGF-I’s trophic signaling pathway or alternatively by directly activating intracellular processes that induce apoptosis. Our studies suggest that it is likely that MeHg alters intracellular processes that directly activate mechanisms leading to apopto-

METHYLMERCURY ANTAGONIZES IGF-I-INDUCED NEURON SURVIVAL

sis. MeHg had its greatest effect on granule cell survival when cells were cultured in the absence of IGF-I (Fig. 2). This observation would suggest that it is unlikely that a major action of MeHg is a simple antagonism of IGF-I or IGF-I receptor activation. In the absence of added IGF-I it is unlikely that IGF-I is provided by other sources, since the culture medium contains no serum, granule cells were purified using several steps and washed several times before culture, and granule cells have not been shown to express IGF-I themselves (Lee et al., 1992). This hypothesis is also supported by experiments using the PI3-K inhibitor LY294002. Recent studies have identified a signaling mechanism by which IGF-I maintains survival of cerebellar granule neurons (Dudek et al., 1997; Datta et al., 1997). IGF-I receptor activation leads to stimulation of PI3-K activity, which ultimately leads to downstream events necessary for maintaining cell survival (Dudek et al., 1997; Datta et al., 1997). If MeHg blocks the IGF-I survival signaling pathway, then the combination of PI3-K inhibition and MeHg might have an additive effect on inhibition of survival. However, when we used the PI3-K inhibitor LY294002 in combination with MeHg, we observed a synergistic effect on granule cell survival (Fig. 5). This could be explained by MeHg and PI3-K inhibition (LY294002) affecting two separate pathways. We also observed that LY294002 at concentrations of 10 and 25 mM, seven and 18 times the reported IC50 value for inhibition of PI3-K (Vlahos et al., 1994), did not reduce granule cell survival to the same degree as cells exposed to MeHg (Figs. 2, 4, and 5). These experimental results are not consistent with MeHg inhibiting granule neuron survival simply by inhibiting IGF-I’s signaling pathway, but alternatively suggest that MeHg has a direct action that can lead to induction of apoptosis. MeHg exposure could directly lead to activation of a cell death program. Supporting this hypothesis are our data showing that MeHg induces a prolonged increase in the expression of the transcription factor c-Jun. Previous experiments have shown that an increase in c-Jun expression by itself, in NIH3T3 cells, can induce apoptosis, suggesting that this transcription factor could be part of a regulatory program that induces cell death (Bossy-Wetzel et al., 1997). MeHg could induce c-Jun expression as part of a direct mechanism that induces granule cell death. However c-Jun is also induced following growth factor withdrawal in neurons and has been shown to be part of the mechanism required for induction of neuronal death (Ham et al., 1995; Miller and Johnson 1996; Watson et al., 1998). Thus, an increase in c-jun expression could also be consistent with inhibition of the IGF-I survival signaling pathway reducing trophic support. This hypothesis is, however, not consistent with the time course of induction of c-Jun following MeHg exposure. Following growth factor withdrawal in cerebellar granule neurons, there is a transient increase in c-Jun RNA expression that peaks at 2 to 3 h after trophic factor removal and declines over the next 2 to 6 h (Miller and Johnson, 1996; Watson et al., 1998). Our experiments indicate that, following exposure of granule cells to MeHg, c-Jun expression continued

167

to increase up to 8 h of culture. This observation would suggest that the action of MeHg on c-Jun expression is more prolonged than that following growth factor withdrawal. Thus, MeHg may have a separate action that maintains the expression of c-Jun at a high level. MeHg’s effect on increasing c-Jun expression could be important in cell death induced in both CNS and peripheral neurons, since, like the death of cerebellar granule neurons, the death of sympathetic neurons also appears to involve an increase in c-Jun expression (Ham et al., 1995; Watson et al., 1998). The antagonism of IGF-I’s survivalpromoting activity is potentially due to MeHg activating a cell death program that involves c-Jun expression and compromises IGF-I’s ability to maintain survival. However, we cannot completely rule out the possibility that MeHg can also partially inhibit the IGF-I survival signaling pathway. Future studies measuring downstream events in the IGF-I survival pathway, such as the level of Bad phosphorylation, will help evaluate whether MeHg can also inhibit IGF-I signal transduction. ACKNOWLEDGMENTS This work was in part supported by NIH Grant ES/OD08087 to R.F. Bulleit.

REFERENCES Ang, L. C., Bhaumick, B., Munoz, D. G., Sass, J., and Jurlink, B. H. (1992). Effects of astrocytes, insulin and insulin-like growth factor I on the survival of motorneurons in vitro. J. Neurol. Sci. 109, 168 –172. Bach, M. A., Shen-Orr, Z., Lowe, W. L., Jr., Roberts, C. T., Jr., and LeRoith, D. (1991). Insulin-like growth factor I RNA levels are developmentally regulated in specific regions of the rat brain. Mol. Brain Res. 10, 43– 48. Beck, K. D., Knusel, B. and Hefti, F. (1993). The nature of the trophic action of BDNF, IGF-I, and bFGF on mesencephalic dopaminergic neurons developing in culture. Neuroscience 52, 855– 866. Beck, K. D., Powell-Braxton, L., Widmer, H., Valverde, J., and Hefti, F. (1995). IGF1 gene disruption results in reduced brain size, CNS hypomyelination, and loss of hippocampal granule and striatal parvalbumin-containing neurons. Neuron 14, 717–730. Bondy, C. A. (1991). Transient IGF-I gene expression during the maturation of functionally related central projection neurons. J. Neurosci. 11, 3442–3455. Bossy-Wetzel, E., Bakiri, L., and Yaniv, M. (1997). Induction of apoptosis by transcription factor c-Jun. EMBO J. 16, 1695–1709. Bulleit, R. F., Cui, H., Wang, J., and Lin, X. (1994). NMDA receptor activation in differentiating cerebellar cell cultures regulates the expression of a new POU gene, Cns-1. J. Neurosci. 14, 1584 –1595. Burbacher, T. M., Rodier, P. M., and Weiss, B. (1990). Methylmercury developmental neurotoxicity: A comparison of effects in humans and animals. Neurotoxicol. Teratol. 12, 191–202. Choi, B. H., Lapham, L. W., Amin-Zaki, L., and Saleem. T. (1978). Abnormal neuronal migration deranged cerebral cortical organization and diffuse white matter astrocytosis of humans fetal brain: A major effect of methylmercury poisoning in utero. J. Neuropathol. Exp. Neurol. 37, 719 –733. Chomczynski, P., and Sacchi, N. (1987). Single step method of RNA isolation by acid guanidinium thiocyanate–phenol– chloroform extraction. Anal. Biochem. 162, 156. Datta, S. R., Dudek, H., Tao, X., Masters, S., Fu, H., Gotoh, Y., and Greenberg, M. E. (1997). Akt phosphorylation of BAD couples survival signals to the cell-intrinsic death machinery. Cell 91, 231–241. D’Mello, S. R., Galli, C., Ciotti, T., and Calissano, P. (1993). Induction of

168

BULLEIT AND CUI

apoptosis in cerebellar granule neurons by low potassium: Inhibition of death by insulin-like growth factor I and cAMP. Proc. Natl. Acad. Sci. USA 90, 10989 –10993. Dudek, H., Datta, S. R., Franke, T. F., Brinbaum, M. J., Yao, R., Cooper, G. M., Segal, R. A., Kaplan, D. R., and Greenberg, M. E. (1997). Regulation of neuronal survival by the serine-threonine protein kinase Akt. Science (Washington, D.C.) 275, 661– 664. Gao, W. Q., Heintz, N., and Hatten, M. (1991). Cerebellar granule cell neurogenesis is regulated by cell– cell interactions in vitro. Neuron 6, 705–715. Ham, J., Babij, C., Whitfield, J., Pfarr, C. M., Lallemand, D., Yaniv, M., and Rubin, L. L. (1995). A c-Jun dominant negative mutant protects sympathetic neurons against programmed cell death. Neuron 14, 927–939. Harada, M. (1978). Congenital Minamata disease: Intrauterine methylmercury poisoning. Teratology 18, 285–288. Hatten, M. E. (1985). Neuronal regulation of astroglial morphology and proliferation in vitro. J. Cell Biol. 100, 384 –396. Hockenbery, D., Nunez, G., Milliman, C., Schreiber, R. D., and Korsmeyer, S. J. (1990). Bcl-2 is an inner mitochondrial membrane protein that blocks programmed cell death. Nature 348, 334 –336. Hughes, R. A., Sendtner, M., and Thoenen, H. (1993). Members of several gene families influence survival of rat motorneurons in vitro and in vivo. J. Neurosci. Res. 36, 663– 671. Jacobs, A. J., Maniscalco, W. M., and Finkelstein, J. N. (1986). Effects of methylmercury chloride, cycloheximide and colchicine on the reaggregation of dissociated mouse cerebellar cells. Toxicol. Appl. Pharmacol. 86, 362– 371. Johnson, E. M., and Deckwerth, T. L. (1993). Molecular mechanisms of developmental neuronal death. Annu. Rev. Neurosci. 16, 31– 46. Kar, S., Chabot, J. G., and Qutrion, R. (1993). Quantitative autoradiographic localization of [125I] insulin-like growth factor I, [125I] insulin-like growth factor II and [125I] insulin receptor binding sites in developing and adult rat brain. J. Comp. Neurol. 333, 375–397. Knusel, B., Michel, P. P., Schwaber, J., and Hefti, F. (1990). Selective and nonselective stimulation of central cholinergic and dopaminergic development in vitro by nerve growth factor, basic fibroblast growth factor, epidermal growth factor, and the insulin-like growth factors. J. Neurosci. 10, 558 –570. Korshing, S. (1993). The neurotrophic factor concept: A reexamination. J. Neurosci. 13, 2739 –2748. Kunimoto, M. (1994). Methylmercury induces apoptosis of rat cerebellar neurons in primary culture. Biochem. Biophys. Res. Commun. 204, 310 – 317. Lee, W., Javedan, S., and Bondy, C. A. (1992). Coordinate expression of insulin-like growth factor system components by neurons and neuroglia during retinal and cerebellar development. J. Neurosci. 12, 4737– 4744. Lin, X., and Bulleit, R. F. (1997). Insulin-like growth factor I (IGF-I) is a critical trophic factor for developing cerebellar granule cells. Dev. Brain Res. 99, 234 –242. Martin, D. P., Schmidt, R. E., Distefano, P. S., Lowry, O. H., Carter, J. G., and Johnson, E. M. (1988). Inhibitors of protein synthesis and RNA synthesis

prevent neuronal death caused by nerve growth factor deprivation. J. Cell Biol. 106, 829 – 844. Miller, T. M., and Johnson, E. M. (1996). Metabolic and genetic analysis of apoptosis is potassium/serum-deprived rat cerebellar granule cells. J. Neurosci. 16, 7487–7495. Milligan, C. E., Oppenheim, R. W., and Schwartz, L. M. (1994). Motorneurons deprived of trophic support in vitro require new gene expression to undergo programmed cell death. J. Neurobiol. 25, 1005–1016. Nagashima, K., Fujii, Y., Tsukamoto, T., Nukuzuma, S., Satoh, M., Fujita, M., Fujioka, Y., and Akagi, H. (1996). Apoptotic process of cerebellar degeneration in experimental methylmercury intoxication of rats. Acta Neuropathol. 91, 72–77. Neff, N. T., Prevette, D., Houenou, J. L., Lewis, M. E., Glicksman, M. A., Yin, Q-W., and Oppenheim, R. W. (1993). Insulin-like growth factors: Putative muscle-derived trophic agents that promote motorneuron survival. J. Neurobiol. 24, 1578 –1588. Oppenheim, R. W. (1991). Cell death during development of the nervous system. Annu. Rev. Neurosci. 14, 453–501. Purves, D. (1986). The trophic theory of neuronal connections. Trends Neurosci. 9, 486 – 489. Sager, P. R., Aschner, M., and Rodier, P. M. (1984). Persistent, differential alterations in developing cerebellar cortex of male and female mice after methylmercury exposure. Dev. Brain Res. 12, 1–11. Sager, P. R., Doherty, R. A., and Rodier, P. M. (1982). Effects of methylmercury on developing mouse cerebellar cortex. Exp. Neurol. 77, 179 –193. Syverson, T. L. M., Totland, G., and Flood, P. R. (1981). Early morphological changes in rat cerebellum caused by a single dose of methylmercury. Arch. Toxicol. 47, 101–111. Takeuchi, T. (1977). Pathology of fetal Minamata disease. Pediatrician 6, 69 – 87. Thomaidou, D., Mione, M. C., Cavanagh, J. F. R., and Parnavelas, J. G. (1997). Apoptosis and its relation to the cell cycle in the developing cerebral cortex. J. Neurosci. 17, 1075–1085. Vlahos, C. J., Matter, W. F., Hui, K. Y., and Brown, R. F. (1994). A specific inhibitor of phosphatidylinositol 3-kinase 2-(4-morpholinyl)-8-phenyl-4H1-benzopyran-4-one (LY294002). J. Biol. Chem. 269, 5241–5248. Volonte, C., Ciotti, M. T., and Battistini, L. (1994). Development of a method for measuring cell number: Application to CNS primary neuronal cultures. Cytometry 17, 274 –276. Watson, A., Eilers, A., Lallemand, D., Kyriakis, J., Rubin, L. L., and Ham, J. (1998). Phosphorylation of c-Jun is necessary for apoptosis induced by survival signal withdrawal in cerebellar granule neurons. J. Neurosci. 18, 751–762. Wyllie, A. H., Kerr, J. F. K., and Currie, A. R. (1980). Cell death: The significance of apoptosis. Int. Rev. Cytol. 68, 251–306. Ye, P., Carson, J., and D’Ercole, A. J. (1995). In vivo actions of insulin-like growth factor-1 (IGF-1) on brain myelination: Studies of IGF-1 and IGF-1 binding protein-1 (IGFBP-1) transgenic mice, J. Neurosci. 15, 7344 –7356. Zucker, R. M., Elstein, K. H., Easterling, R. E., and Massaro, E. J. (1990). Flow cytometric analysis of the mechanism of methylmercury cytoxicity. Am. J. Pathol. 137, 1187–1198.