- Email: [email protected]
The xenobiotic methionine sulfoximine modulates carbohydrate anabolism and related genes expression in rodent brain
Toxicology 153 (2000) 179 – 187 www.elsevier.com/locate/toxicol
The xenobiotic methionine sulfoximine modulates carbohydrate anabolism and related genes expression in rodent brain K. He´lary-Bernard *, M.-Y. Ardourel, J.F. Cloix, T. Hevor Laboratoire de Me´tabolisme Ce´re´bral et Neuropathologies-E.A. 2633, Uni6ersite´ d’Orle´ans, B.P. 6759, F-45067 Orle´ans Cedex 2, France
Abstract Methionine sulfoximine is a xenobiotic amino acid derived from methionine. One of its major properties is to display a glycogenic activity in the brain. After studying this property, we investigate here a possible action of this xenobiotic on the expression of genes related to carbohydrate anabolism in the brain. Glycogen was studied by the means of electron microscopy. Astrocytes were cultured and the influence of methionine sulfoximine on carbohydrate anabolism in these cells was investigated. In vivo, methionine sulfoximine induced a large increase in glycogen accumulation. It also enhanced the glycogen accumulation in cultured astrocytes principally, when the medium was enriched in glucose. The gluconeogenic enzyme fructose-1,6-bisphosphatase may account for glycogen accumulation. Plasmids were built using antisens cDNA to permanently block the expression of fructose-1,6-bisphosphatase. An eukaryotic vector was used and the expression of fructose-1,6-bisphosphatase gene was under the control of the promoter of the glial fibrillary acidic protein. In this case, the glycogen content in cultured astrocytes largely decreased. This work shows that methionine sulfoximine enhances energy carbohydrate synthesis in the brain. Since this xenobiotic also enhances the expression of some genes related to one of the key step of glucose synthesis, it is possible that genes may be one target of methionine sulfoximine. Next investigations will study the actual effect of methionine sulfoximine in the cells. © 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Glucose; Glycogen; Fructose-1,6-bisphosphatase; Gluconeogenesis; Energy metabolism; Astrocyte
1. The xenobiotic methionine sulfoximine Methionine sulfoximine is a xenobiotic amino acid, which was isolated from cereal proteins when cereal flours were processed using nitrogen * Corresponding author. Tel.: +33-2-38417079; fax: + 332-38417244. E-mail address: [email protected] (K. He´lary-Bernard).
trichloride. One of the first properties of methionine sulfoximine, which was discovered, is its effect on behaviour because it induced epileptiform activity in many animals when the latter were fed with nitrogen trichloride-bleached flour. A second major property of methionine sulfoximine is its high and irreversible inhibition of glutamine synthetase. This inhibition results from a direct link of the molecule of methionine sulfoximine to the
0300-483X/00/$ - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 0 0 - 4 8 3 X ( 0 0 ) 0 0 3 1 3 - 9
180
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187
molecule of glutamine synthetase so that methionine sulfoximine is largely utilized in either animal or plant studies, when glutamine synthetase has to be inhibited. A third property of methionine sulfoximine is a powerful glycogenic action in the brain. Indeed, it induces a very large accumulation of glycogen in the brain, a tissue that is normally pour in this polysaccharide. We have shown that methionine sulfoximine also activates the last key enzyme of the synthesis of glucose-6phosphate in the brain, fructose-1,6-bisphosphatase. Probably, the new synthesized glucose-6-phosphate may be utilized for glycogen accumulation. We hypothesized that these anabolic properties of the xenobiotic could result from a modulation of genes related to carbohydrate syntheses in the brain. In the present paper, we gain insight the problem of carbohydrate anabolism in the brain under the influence of the xenobiotic methionine sulfoximine.
2. Enhancement of glycogen accumulation in the brain Methionine sulfoximine is well known as a powerful glycogenic molecule in the brain in vivo (Folbergrova et al. 1969; Phelps, 1975). The cellular site of glycogen accumulation in the brain has been shown to be the astrocytes (Phelps, 1975). The glycogenic property of methionine sulfoximine has also been demonstrated in cultured astrocytes (Swanson et al., 1989). These show that methionine sulfoximine directly targets astrocytes and it does not need the cooperation of other cells such as neurons to induce glycogen accumulation in the brain (Hevor and Delorme, 1991). Because of the increasing utilization of cultured astrocytes to understand the cellular biology of the nervous system, we decided to carefully look inside the astrocytes and to compare the situation in vivo to that in vitro.
2.1. Preparation of animals for in 6i6o studies Male rats from Sprague – Dawley strain and male mice from Swiss strain were purchased from IFFA-Credo (Arbresle, France). They were
housed in the laboratory until the experiments. The access to food and to water were free ad libitum and the light cycle was 12 h of light (08:00–20:00 h) and 12 h of darkness (20:00– 08:00 am). In order to avoid changes in brain glycogen due to circadian rhythm, all the animals were sacrificed between 09:00 and 11:00 h am. Indeed, Hutchins and Rogers (1970) have shown that brain glycogen does not change between 09:00 and 11:00 despite its significant decrease by 08:00 pm. The animals were starved 24 h before their sacrifice in order to stabilize the glucose level in the blood. Then they were anaesthetized using halothane or diethyl ester. They were submitted to an intracardiac perfusion using glutaraldehyde buffered with phosphate buffer pH 0.15 M. The whole brain was dissected out of the skull and sectioned in small pieces. The latter were immersed in the same fixative for 2 h and then, were washed using the phosphate buffer. Then, they were post-fixed in 1% osmium tetroxide for 1 h, dehydrated in graded alcohols and in propylene oxide, embedded in Epon, and sectioned using an ultramicrotome. The sections were stained with uranyl acetate and lead citrate and examined in an electron microscope.
2.2. Astrocyte cultures Pregnant rats or pregnant mice were bred in the laboratory to get newborn animals. The latter were utilized between day 1 and 4 after their birth. The whole brain of the newborn rodent was dissected out of the skull in the aseptic atmosphere of a laminar flow hood for cell culture. The cerebral cortex was dissected out and was introduced in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% foetal calf serum, 2 mM glutamine, 100 units/ml penicillin, and 100 mg/ml dihydrostreptomycine. The meninges were carefully taken off and the tissue was triturated through a needle using a syringe by successive aspirations and rejections. The resulting cell suspension was diluted with the same medium and the number of cells/ml was determined using a haemocytometer. The dispersion of the tissue into single cells damaged some cells. The number of the latter can be known when the
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187
medium used for the cell counting contains trypan blue. An alive and viable cell does not incorporate trypan blue. This property is termed trypan blue exclusion. So, the blue cells that did not exclude the dye were not counted among the viable cells. Sixty millimeter diameter petri dishes containing 2.5 ml of the DMEM medium were incubated in a culture incubator where the carbon dioxide level was set to 5%, the temperature to 37°C, and the humidity to 95% before the beginning of the dissection in order to get a steady state before adding the cells. Two millileters and half of DMEM medium containing 2.5×105 cells were added to the Petri dishes and the latter were again incubated in the conditions described above. The medium was renewed 4 days after the seeding and then, twice a week. These primary cultures achieved maturity 3 – 4 weeks after the seeding and the cells were confluent (Fig. 1). The large majority, up to 90– 95% of the cells, were astrocytes as determined by Booher and Sensenbrenner (1972). Indeed, the neurons cannot adhere to the Petri dish because they need a coated surface (polylysine or collagen) to adhere. The medium was not suitable for the growth of endothelial cells and microglial cells. A few numbers of oligodendrocytes were present in the culture. The latter
Fig. 1. Astrocytes in primary cultured. Cells were dissociated from the cerebral cortex of the rat brain and plated in Dulbecco’s Modified Eagle’s Medium supplemented with glutamine and antibiotics. After 3 weeks, the astrocytes were photographed. Note that most cells have the common shape of protoplasmic astrocytes (arrows). G × 300
181
cells grew upon the layer constituted by astrocytes. Oligodendrocytes were easily recognized because of their shape and their darkness. They can be removed by shaking the dishes for some h in the incubator. All the cells were GFA positive. For electron microscopy studies, the astrocytes were directly fixed in the Petri dishes using glutaraldehyde buffered with 0.15 M phosphate buffer pH 7.4. Then, the procedure was the same as that for brain tissue. The solidified Epon was grossly sliced using a small saw and the pyramids for ultra slicing were made. The ultra slices were parallel to the astrocyte layer.
2.3. Glycogen in astrocytes For a long time, it was believed that glycogen was not present in the central nervous system. This was due to the way used for sacrificing the animals. When the latter were decapitated without special precautions before biochemical analysis, glycogen was undetectable. Indeed, the enzyme that hydrolyses glycogen in the brain, phophorylase, is very active even many min after the decapitation. This enzyme breaks down the brain glycogen when operating at room temperature. When the animals were killed using a high-energy microwave irradiation or when the heads were immediately dropped in liquid nitrogen after decapitation, glycogen content was measured and it was 2.729 0.08 mM glucosyl units/g tissue in the rat cerebral cortex and 3.219 0.16 mM glucosyl units/g in the cerebellum. Nevertheless, it was not possible to see glycogen particles in brain slice by the means of electron microscopy. In 1972, Phelps performed an intracardiac perfusion using glutaraldehyde fixative in rat. Then, the glycogen particles could be seen in the cerebral cortex slices of rats. In normal rat, the number of particles were principally smaller than that described in the liver by Drochmans (1962). Mice were submitted to the xenobiotic methionine sulfoximine by an intraperitoneal administration of 100 mg/kg body weight. The animals were sacrificed by an intracardiac perfusion of glutaraldehyde as described above 24 h after dosing. In the slices from control animals that received an intraperitoneal administration of 0.9% NaCl,
182
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187
thionine sulfoximine is able to increase the glycogen accumulation up to the rosette level. Glycogen accumulation was very intense in the perivascular end-feet of astrocytes and sometimes, all the areas of the end-feet were filled with glycogen particles. When rats were used, the results were the same as in mice. Glycogen particles were observed in astrocytes and exclusively in astrocytes. Other cells, neurons for example, never contained glycogen particles. For this reason, we cultured astrocytes and we submitted these cells to the xenobiotic action. The medium of cultured astrocytes was renewed with DMEM medium containing 1 mM of methionine sulfoximine. The astrocytes were fixed 8 or 24 h after the medium renewal, using glutaraldehyde fixative. As in the in vivo brain, methionine sulfoximine also increased the number of glycogen particles in cultured astrocyte cytoplasm. This increase was faint when the medium contained 1 g/l of glucose and it was obvious when the medium contained 4.5 g/l (Fig. 3). Fig. 2. Glycogen particles in the cerebral cortex of mice. The animals were submitted to NaCl 0.9% for control experiments (A) or to 100 mg/kg of methionine sulfoximine (B). Numerous glycogen particles accumulate in the cytoplasm of astrocytes (G). Note here the end foot against the capillary filled with the particles. C, blood capillary; M, mitochondrion; R, endoplasmic reticulum. G × 16 000.
there was a few number of glycogen particles, thus looking like the slices described by Phelps (1972). Contrarily, in the slices from animals submitted to methionine sulfoximine, there was a tremendous increase in the number and in the size of glycogen particles (Fig. 2). In the perikarya, these particles were often gathered in areas, which were clear without surrounding membrane. A major feature of the glycogen particles was the presence of rosettes. Rosette is the term ascribed by Drochmans (1962) to the morphology of glycogen in liver when the particles are disposed in a way looking like the petals of the flower rose. Drochmans (1962) mentioned that rosettes are the highest form of glycogen organization in liver, when the synthesis of the polysaccharide is intense. In normal brain, glycogen never achieves the step of rosette formation. The xenobiotic me-
3. Possible mechanism of glycogen accumulation under the effect of methionine sulfoximine The xenobiotic methionine sulfoximine is the most powerful glycogenic agent in the brain. The actual mechanism of such an anabolic effect is not yet well understood. One can suppose that the xenobiotic enhances the activity of the synthesizing enzyme glycogen synthase. Another possibility is the inhibition of the catabolizing enzyme phosphorylase. The work of Folbergrova (1973) has shown that methionine sulfoximine did not markedly influence the activities of the enzymes directly related to glycogen synthesis. We have made a series of investigations that showed a marked increase in the activities and quantities of the gluconeogenic enzyme fructose-1,6-bisphosphatase when animals were submitted to methionine sulfoximine. Since these increases were exclusively localized in astrocytes, we cultured the latter cells and submitted them to the xenobiotic. Again, in cultured astrocytes, methionine sulfoximine displayed its increasing effect on the activity of fructose-1,6-bisphosphatase. Taking into ac-
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187
count all the preceding data, we concluded that methionine sulfoximine enhances the gluconeogenic pathway in the rodent brain. Indeed, fructose-1,6-bisphosphatase is the last key enzyme before the synthesis of glucose-6-phosphate. As all key enzymes, the preceding one catalyses an irreversible step and glucose-6-phosphate cannot be transformed again in its generating compounds by the same enzyme. This observation means that glucose-6-phosphate has to follow another metabolic pathway. One possibility is that of glycolysis, the first key enzyme in this pathway being phosphofructokinase. We have already shown that methionine sulfoximine has almost no effect on the activity of the latter enzyme, despite the large increase in the activity of the opposite enzyme fructose-1,6-bisphosphatase which was dis-
Fig. 3. Glycogen particles in cultured astrocytes. The astrocytes were cultured in a medium containing 4.5 g/l of glucose (A) or 4.5 g/l of glucose and 1 mM methionine sulfoximine (B). Glycogen particles accumulate in the cytoplasm of astrocytes and this accumulation is enhanced by methionine sulfoximine. G, glycogen; M, mitochondrion; N, nucleus. G × 16,000.
183
played under the effect of the xenobiotic (Hevor and Gayet, 1978). So, the glycolysis pathway is improbable. A second possibility is that of free glucose synthesis, a pathway that was documented during this decade. Theoretically, glucose6-phosphate can be converted into free glucose. But the supposed absence of glucose-6-phosphate translocase (Fishman and Karnovsky, 1986) militated against such a conversion taking place in the brain. Burchell and Waddell (1991) have shown that antibodies raised against the liver glucose-6phosphatase did not allow the verification of the activity of this phosphatase in the brain. This may be due to the lack of cross-reaction in the immunolabelling of the two enzymes originating from two kinds of tissues. Recent investigations by Forsyth et al. (1993) produced some important data on the activity of glucose-6-phosphatase in the brain. These authors have shown that a glucose-6-phosphatase hydrolysing activity exists in homogenates of cultured astrocytes, 40% of this activity being specific to the catalytic subunit of glucose-6-phosphatase localized in microsomes. In many aspects, the enzyme looks like that of the liver, the kidney and the ß cells of the pancreas (molecular weight, sensitivity to pH, latency in activity) although the Km value was higher. This high Km may be attributed to the decrease in activity during cultivation. Moreover, Forsyth et al. (1993) have also shown that the microsomes are permeable to glucose-6-phosphate. These data strongly support the idea that gluconeogenesis in astrocytes can be complete, up to free glucose. So the idea of a free glucose synthesis through the catalysis of glucose-6-phosphatase in astrocytes appears likely, at least in cultured astrocytes. A third pathway that can be followed by glucose-6phosphate under the effect of methionine sulfoximine is that of glycogen synthesis. Our experiments showed that methionine sulfoximine induced a large increase in glycogen accumulation as well in vivo as in cultured astrocytes. A major property of the xenobiotic is a strong and irreversible inhibition of glutamine synthetase (Meister, 1978). This inhibition may normally results in the accumulation of the substrate glutamate. Paradoxically, the glutamate level does not increase, and contrarily, it decreased a little in some
184
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187
areas of the brain (Fonnum and Paulsen, 1990). Phelps (1975) suggested that glutamate might be massively incorporated in glycogen under the effect of methionine sulfoximine. Up to now, the actual pathway of such an incorporation was not demonstrated. We have hypothesized that the first steps of gluconeogenesis may be the major way followed by glutamate carbon toward glycogen synthesis. Indeed, if glutamate is converted to alpha-ketoglutarate, the latter can be incorporated in the cycle of Krebs and then, it can follow gluconeogenesis pathway. In the present work, taking advantage of the molecular biology means, we gain insight the problem of the possible involvement of gluconeogenic pathway in the glycogen accumulation in the brain. For this, we blocked the expression of fructose-1,6-bisphosphatase to see whether or not, changes will happen in the glycogen content of cultured astrocytes.
4. Fructose-1,6-bisphosphatase mRNA expression in eukaryotic vector The purpose of this investigation was to develop tools that allow the modulation of the specific expression of gluconeogenic enzymes fructose-1,6-bisphosphatase in a specific population of cells of the central nervous system, the astrocytes. The vector used was a construction built from different vectors in order to produce a plasmid that expresses the inserted cDNA under the control of a strong promoter. This promoter controls the expression of an intermediate filament protein, which is exclusively synthesized in astrocytes (Baba et al., 1997). This protein is the glial fibrillary acidic protein (GFAP). The GFAP promoter was cloned as a 2.7 kbp from mouse brain (Fig. 4). It contains a negative and a positive domain and two trans regulation domains. The trans regulation domain that is closed to the transcription initiation site ( + 1) contains two enhancer sequences (−455 and −306) and the sequences for various nuclear factors (AP1: −238 and − 169; AP2: − 339 and − 95; CRE: −154; NF1: − 118; heat choc: − 455). The CAT and TATA boxes were localized at −83 and −29, respectively
Fig. 4. Organisation of the promoter of the glial fibrillary acidic protein (GFAP). Designed from Miura et al. (1990) and from Baba et al. (1997).
(Miura et al., 1990). We produced the mouse brain GFAP promoter by PCR using specific primers from − 2215 to + 7, giving a 2222 bp cDNA that corresponds to the fully active GFAP promoter. The identity of the PCR product to the GFAP promoter sequence was assessed by restriction mapping and DNA sequencing. The primer sequences were modified in order to introduce restriction sites for cloning. The eukaryotic vector that we therefore produced is shown in Fig. 5. In addition to the GFAP promoter it contains the genes providing the resistance to ampicillin for selection in bacteria and to neomycin for selection in eukaryotic cells. The multiple cloning site (MCS) containing
Fig. 5. Plasmid pASTM-Neo. This plasmid contains the glial fibrillary acidic protein (GFAP) promoter upstream the multicloning sites. The ampicillin resistance gene (AmpR) allows selection of prokariotic cells, while neomycin resistance gene (NeoR) allows selection in eukaryotic cells.
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187
four unique restriction sites is located downstream the GFAP promoter. The sens and antisens cDNA of genes which expressions will be modulated in astrocytes, were cloned inside this MCS. This eukaryotic vector was called pASTM-Neo. The efficiency of the mouse GFAP promoter in the pASTM-Neo was assessed using two different ways. Firstly, the GFAP promoter was removed from the pASTM-Neo and cloned in the pEGFP1 plasmid that contains a reporter gene constituted of a green auto fluorescent protein called EGFP. Mouse astrocytes or COS cells were transfected with the latter plasmid. Only astrocytes showed fluorescence. Secondly, the gene coding EGFP protein was cloned in the multicloning site of pASTM-Neo and this plasmid was transfected in either mouse astrocytes or COS cells. Only the mouse astrocytes produced EGFP protein. These experiments demonstrated that the GFAP promoter inserted in the eukaryotic vector to build pASTM-Neo is active and could be considered as a strong promoter.
185
Fig. 6. Structure of fructose-1,6-bisphosphate gene (designed from El-Maghrabi et al., 1991 and El-Maghrabi et al., 1995).
The quality and integrity of the plasmids containing the fructose-1,6-bisphosphatase cDNAs were controlled by restriction mapping or DNA sequencing. The vectors containing antisens fructose-1,6-bisphosphatase cDNA was called pASTM-FBP.
6. Astrocytes culture and transfection 5. cDNA cloning Fructose-1,6-bisphosphatase was cloned as antisens cDNA inside the MCS of the pASTM-Neo. The fructose-1,6-bisphosphatase cDNA was obtained using RT-PCR as previously described (Cloix et al., 1997). The RT-PCR was realized on total RNA extracted from mouse brain using the primers, which were designed to span three introns (Fig. 6) accordingly to the published mouse fructose-1,6-bisphosphatase gene (Nomura et al., 1994; Cloix et al., 1997). This was done on the ground of a similarity between the mouse brain gene organization and that published for human (El-Maghrabi et al., 1995) and rat (El-Maghrabi et al., 1991) liver enzyme. The primers sequences were modified in order to introduce restriction site for cloning. The 5% and 3% primers contained additional bases to create restriction sites for AccI and BamHI, respectively. This allowed the insertion of the 547bp fructose-1,6-bisphosphatase cDNA in the antisens position in the pASTM-Neo MCS at the same sites.
Primary cultures of mouse astrocytes were performed as described above. The first attempts of transfection of these primary cultures with either pASTM-FBP or pASTM-Neo were unsuccessful. The reason for this failure was the low division rate of astrocytes in primary culture. Therefore, in order to improve the transfection yield of astrocytes, we used astrocytes after eight passages. The astrocytes were transfected with the plasmids using the co precipitation method of DNA with calcium phosphate, as follows. Twenty mg of ethanol-precipitated plasmid were dissolved in 0.5 ml of 2× HBS buffer (280 mM NaCl, 10 mM KCl, 1.5 mM Na2HPO4, 12 mM dextrose, 50 mM HEPES, pH 7.5) and co-precipitated with 0.5 ml of 0.25 mM CaCl2. The precipitate was mixed with the culture medium, and the astrocytes (105 cells) were incubated with this medium for 8 h at 37°C. Thereafter, the medium was renewed and 2 days later geneticin (400 mg/ml) was added to the medium for selection of geneticin-resistant astrocytes. When these cells were selected (3–4
186
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187
months), they were sub cultured (1 – 2 months) in order to obtain enough cells for experiments and for frozen stocks. Therefore, transfected astrocytes were obtained and used after a total of 12 – 15 passages.
7. Glycogen level in transfected astrocytes Mouse astrocytes were transfected with pASTM-FBP. The control of transfection was done using pASTM-Neo. When geneticin-resistant astrocytes were obtained as stable cell cultures, the activities of fructose-1,6-bisphosphatase (Verge´ et al., 1996) was measured as well as the cellular level of glycogen (Verge´ and He´vor, 1995). The results of these experiments are presented in Fig. 7. The activity of fructose-1,6-bisphosphatase in astrocytes transfected with fructose-1,6-bisphosphatase antisens cDNA was significantly reduced by 85%. A major result was a large decrease observed in glycogen content of the astrocytes transfected with fructose-1,6-bisphosphatase antisens cDNA, up to 75%. The preceding result demonstrated that it was possible to chronically decrease the cellular glycogen content of cultured astrocytes using antisens cDNA that inhibited the gluconeogenic enzyme fructose-1,6-bisphosphatase. The viability of these astrocytes was not apparently altered by this transfection. These permanently transfected astrocytes constitute precious tools for studying astro-
Fig. 7. Change in glycogen content of cultured astrocytes transfected with plasmids containing antisens cDNA of fructose-1,6-bisphosphatase.
cytic glucose and glycogen metabolisms, and to analyse the effect of methionine sulfoximine, which impairs glycogen metabolism in the brain. These genetically modified astrocytes could also be used to study the relationship between astrocyte glucose metabolism and neuron functions. These experiments will be done and the next step will be the effect of methionine sulfoximine on glycogen accumulation in astrocytes tranfected with fructose-1,6-bisphosphatase antisens cDNA.
Acknowledgements This work was supported by funds from La Ligue Nationale contre le Cancer, Comite´ du Loiret and from the Re´gion Centre. Katy He´laryBernard is a recipient of a grant from the Re´gion Centre. Many thanks to Philippe Moreau for his technical assistance.
References Baba, H., Nakahira, T.A., Morita, N., Tanaka, F., Akita, H., Ihenaka, K., 1997. GFAP gene expression during development of astrocytes. Dev. Neurosci. 19, 49 – 57. Booher, J., Sensenbrenner, M., 1972. Growth and cultivation of dissociated neurons and glial cells from embryonic chick, rat and human brain in flask cultures. Neurobiology 2, 97 – 105. Burchell, A., Waddell, I.D., 1991. The molecular basis of the hepatic microsomal glucose-6-phosphatase system. Biochim. Biophys. Acta 1092, 129 – 137. Cloix, J.F., Beaulieu, E., He´vor, T.K., 1997. Diversity of fructose-1,6-bisphosphatase transcripts in mouse brain, liver, kidney and heart. NeuroReport 8, 617 – 622. Drochmans, P., 1962. Morphologie du glycoge`ne: etude au microscope e´lectronique en coloration ne´gative du glycoge`ne particulaire. J. Ultrastruct. Res. 6, 141 – 163. El-Maghrabi, M.R., Lange, A.J., Jiang, W., Yamagata, K., Pilkis, S.J., 1995. Human fructose-1,6-bisphosphatase gene (FBP1): exon-intron organisation, localisation to chromosome bands 9q22.2 – 22.3, and mutation screening in subjects with fructose-1,6-bisphosphatase deficiency. Genomics 27, 520 – 525. El-Maghrabi, M.R., Lange, A.J., Ku¨mmel, L., Pilkis, S.J., 1991. The rat liver fructose-1,6-bisphosphatase. J. Biol. Chem. 266, 2115 – 2120. Fishman, R.S., Karnovsky, M.L., 1986. Apparent absence of a translocase in the cerebral glucose-6-phosphatase system. J. Neurochem. 46, 371 – 378.
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187 Folbergrova, J., Passonneau, J.V., Lowry, O.H., Schulz, D.W., 1969. Glycogen, amonia and related compounds in the brain during seizures evoked by methionine sulfoximine. J. Neurochem. 16, 191 –203. Folbergrova, J., 1973. Glycogen and glycogen phosphorylase in the cerebral cortex of mice under the influence of methionine sulfoximine. J. Neurochem. 20, 547–557. Fonnum, F., Paulsen, R.E., 1990. Comparison of transmitter amino acid levels in rat globus pallidus and neostriatum during hypoglycemia or after treatment with methionine sulfoximine or gamma-vinyl gamma-aminobutyric acid. J. Neurochem. 54, 1253–1257. Forsyth, R.J., Bartlett, K., Burchell, A., Scott, H.M., Eyre, J.A., 1993. Astrocytic glucose-6-phosphatase and the permeability of brain microsomes to glucose-6-phosphate. Biochem. J. 294, 145 –151. Hevor, T.K., Delorme, P., 1991. Biochemical and ultrastructural study of glycogen in cultured astrocytes submitted to the convulsant methionine sulfoximine. Glia 4, 64–69. Hevor, T.K., Gayet, J., 1978. Fructose-1,6-diphosphatase and phosphofructokinase activities in the brain of mice and rats submitted to the convulsant methionine sulfoximine. Brain Res. 150, 210 – 215. Hutchins, D.A., Rogers, K.J., 1970. Physiological and drug-induced changes in the glycogen content of mouse brain. Br. J. Pharmacol. 39, 1 – 9. Meister, A., 1978. Inhibition of glutamine synthetase and gamma-glutamylcysteine synthetase by methionine sulfox-
.
187
imine and related compounds. In: Seiller, N., Jung, M.J., Kosh-Weser, J. (Eds.), Enzyme Activated Irreversible Inhibitors. Elesevier, North Holland, pp. 187 – 210. Miura, M., Tamura, T.A., Mikoshiba, K., 1990. Cell-specific expression of the mouse glial fibrillary acidic protein genes: identification for astrocytes-specific expression. J. Neurochem. 55, 1180 – 1188. Nomura, M., Takihara, Y., Yasunaga, T., Shimada, K., 1994. One of the retinoic acid-inducible cDNA clones in mouse embryonal carcinoma F9 cells encodes a novel isoenzyme of fructose-1,6-bisphosphatase. FEBS Lett. 348, 201 – 205. Phelps, C.H., 1972. Barbiturate-induced glycogen accumulation in brain: an electron microscopic study. Brain Res. 39, 225 – 234. Phelps, C.H., 1975. An ultrastructural study of methionine sulfoximine-induced glycogen accumulation in astrocytes of the mouse brain cerebral cortex. J. Neurocytol. 4, 479 – 490. Swanson, R.A., Yu, A.C.H., Sharp, F.R., Chan, P.H., 1989. Regulation of glycogen content in primary astrocyte culture: effects of glucose analogues, phenobarbital, and methionine sulfoximine. J. Neurochem. 52, 1359 – 1365. Verge´, V., He´vor, T., 1995. Regulation of fructose-1,6-bisphosphatase activity in primary cultured astrocytes. Neurochem. Res. 20, 1049 – 1056. Verge´, V., Legrand, A., He´vor, T.K., 1996. Isolation of carbohydrate metabolic clones from cultured astrocytes. Glia 18, 244 – 254.
The xenobiotic methionine sulfoximine modulates carbohydrate anabolism and related genes expression in rodent brain K. He´lary-Bernard *, M.-Y. Ardourel, J.F. Cloix, T. Hevor Laboratoire de Me´tabolisme Ce´re´bral et Neuropathologies-E.A. 2633, Uni6ersite´ d’Orle´ans, B.P. 6759, F-45067 Orle´ans Cedex 2, France
Abstract Methionine sulfoximine is a xenobiotic amino acid derived from methionine. One of its major properties is to display a glycogenic activity in the brain. After studying this property, we investigate here a possible action of this xenobiotic on the expression of genes related to carbohydrate anabolism in the brain. Glycogen was studied by the means of electron microscopy. Astrocytes were cultured and the influence of methionine sulfoximine on carbohydrate anabolism in these cells was investigated. In vivo, methionine sulfoximine induced a large increase in glycogen accumulation. It also enhanced the glycogen accumulation in cultured astrocytes principally, when the medium was enriched in glucose. The gluconeogenic enzyme fructose-1,6-bisphosphatase may account for glycogen accumulation. Plasmids were built using antisens cDNA to permanently block the expression of fructose-1,6-bisphosphatase. An eukaryotic vector was used and the expression of fructose-1,6-bisphosphatase gene was under the control of the promoter of the glial fibrillary acidic protein. In this case, the glycogen content in cultured astrocytes largely decreased. This work shows that methionine sulfoximine enhances energy carbohydrate synthesis in the brain. Since this xenobiotic also enhances the expression of some genes related to one of the key step of glucose synthesis, it is possible that genes may be one target of methionine sulfoximine. Next investigations will study the actual effect of methionine sulfoximine in the cells. © 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Glucose; Glycogen; Fructose-1,6-bisphosphatase; Gluconeogenesis; Energy metabolism; Astrocyte
1. The xenobiotic methionine sulfoximine Methionine sulfoximine is a xenobiotic amino acid, which was isolated from cereal proteins when cereal flours were processed using nitrogen * Corresponding author. Tel.: +33-2-38417079; fax: + 332-38417244. E-mail address: [email protected] (K. He´lary-Bernard).
trichloride. One of the first properties of methionine sulfoximine, which was discovered, is its effect on behaviour because it induced epileptiform activity in many animals when the latter were fed with nitrogen trichloride-bleached flour. A second major property of methionine sulfoximine is its high and irreversible inhibition of glutamine synthetase. This inhibition results from a direct link of the molecule of methionine sulfoximine to the
0300-483X/00/$ - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 0 0 - 4 8 3 X ( 0 0 ) 0 0 3 1 3 - 9
180
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187
molecule of glutamine synthetase so that methionine sulfoximine is largely utilized in either animal or plant studies, when glutamine synthetase has to be inhibited. A third property of methionine sulfoximine is a powerful glycogenic action in the brain. Indeed, it induces a very large accumulation of glycogen in the brain, a tissue that is normally pour in this polysaccharide. We have shown that methionine sulfoximine also activates the last key enzyme of the synthesis of glucose-6phosphate in the brain, fructose-1,6-bisphosphatase. Probably, the new synthesized glucose-6-phosphate may be utilized for glycogen accumulation. We hypothesized that these anabolic properties of the xenobiotic could result from a modulation of genes related to carbohydrate syntheses in the brain. In the present paper, we gain insight the problem of carbohydrate anabolism in the brain under the influence of the xenobiotic methionine sulfoximine.
2. Enhancement of glycogen accumulation in the brain Methionine sulfoximine is well known as a powerful glycogenic molecule in the brain in vivo (Folbergrova et al. 1969; Phelps, 1975). The cellular site of glycogen accumulation in the brain has been shown to be the astrocytes (Phelps, 1975). The glycogenic property of methionine sulfoximine has also been demonstrated in cultured astrocytes (Swanson et al., 1989). These show that methionine sulfoximine directly targets astrocytes and it does not need the cooperation of other cells such as neurons to induce glycogen accumulation in the brain (Hevor and Delorme, 1991). Because of the increasing utilization of cultured astrocytes to understand the cellular biology of the nervous system, we decided to carefully look inside the astrocytes and to compare the situation in vivo to that in vitro.
2.1. Preparation of animals for in 6i6o studies Male rats from Sprague – Dawley strain and male mice from Swiss strain were purchased from IFFA-Credo (Arbresle, France). They were
housed in the laboratory until the experiments. The access to food and to water were free ad libitum and the light cycle was 12 h of light (08:00–20:00 h) and 12 h of darkness (20:00– 08:00 am). In order to avoid changes in brain glycogen due to circadian rhythm, all the animals were sacrificed between 09:00 and 11:00 h am. Indeed, Hutchins and Rogers (1970) have shown that brain glycogen does not change between 09:00 and 11:00 despite its significant decrease by 08:00 pm. The animals were starved 24 h before their sacrifice in order to stabilize the glucose level in the blood. Then they were anaesthetized using halothane or diethyl ester. They were submitted to an intracardiac perfusion using glutaraldehyde buffered with phosphate buffer pH 0.15 M. The whole brain was dissected out of the skull and sectioned in small pieces. The latter were immersed in the same fixative for 2 h and then, were washed using the phosphate buffer. Then, they were post-fixed in 1% osmium tetroxide for 1 h, dehydrated in graded alcohols and in propylene oxide, embedded in Epon, and sectioned using an ultramicrotome. The sections were stained with uranyl acetate and lead citrate and examined in an electron microscope.
2.2. Astrocyte cultures Pregnant rats or pregnant mice were bred in the laboratory to get newborn animals. The latter were utilized between day 1 and 4 after their birth. The whole brain of the newborn rodent was dissected out of the skull in the aseptic atmosphere of a laminar flow hood for cell culture. The cerebral cortex was dissected out and was introduced in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% foetal calf serum, 2 mM glutamine, 100 units/ml penicillin, and 100 mg/ml dihydrostreptomycine. The meninges were carefully taken off and the tissue was triturated through a needle using a syringe by successive aspirations and rejections. The resulting cell suspension was diluted with the same medium and the number of cells/ml was determined using a haemocytometer. The dispersion of the tissue into single cells damaged some cells. The number of the latter can be known when the
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187
medium used for the cell counting contains trypan blue. An alive and viable cell does not incorporate trypan blue. This property is termed trypan blue exclusion. So, the blue cells that did not exclude the dye were not counted among the viable cells. Sixty millimeter diameter petri dishes containing 2.5 ml of the DMEM medium were incubated in a culture incubator where the carbon dioxide level was set to 5%, the temperature to 37°C, and the humidity to 95% before the beginning of the dissection in order to get a steady state before adding the cells. Two millileters and half of DMEM medium containing 2.5×105 cells were added to the Petri dishes and the latter were again incubated in the conditions described above. The medium was renewed 4 days after the seeding and then, twice a week. These primary cultures achieved maturity 3 – 4 weeks after the seeding and the cells were confluent (Fig. 1). The large majority, up to 90– 95% of the cells, were astrocytes as determined by Booher and Sensenbrenner (1972). Indeed, the neurons cannot adhere to the Petri dish because they need a coated surface (polylysine or collagen) to adhere. The medium was not suitable for the growth of endothelial cells and microglial cells. A few numbers of oligodendrocytes were present in the culture. The latter
Fig. 1. Astrocytes in primary cultured. Cells were dissociated from the cerebral cortex of the rat brain and plated in Dulbecco’s Modified Eagle’s Medium supplemented with glutamine and antibiotics. After 3 weeks, the astrocytes were photographed. Note that most cells have the common shape of protoplasmic astrocytes (arrows). G × 300
181
cells grew upon the layer constituted by astrocytes. Oligodendrocytes were easily recognized because of their shape and their darkness. They can be removed by shaking the dishes for some h in the incubator. All the cells were GFA positive. For electron microscopy studies, the astrocytes were directly fixed in the Petri dishes using glutaraldehyde buffered with 0.15 M phosphate buffer pH 7.4. Then, the procedure was the same as that for brain tissue. The solidified Epon was grossly sliced using a small saw and the pyramids for ultra slicing were made. The ultra slices were parallel to the astrocyte layer.
2.3. Glycogen in astrocytes For a long time, it was believed that glycogen was not present in the central nervous system. This was due to the way used for sacrificing the animals. When the latter were decapitated without special precautions before biochemical analysis, glycogen was undetectable. Indeed, the enzyme that hydrolyses glycogen in the brain, phophorylase, is very active even many min after the decapitation. This enzyme breaks down the brain glycogen when operating at room temperature. When the animals were killed using a high-energy microwave irradiation or when the heads were immediately dropped in liquid nitrogen after decapitation, glycogen content was measured and it was 2.729 0.08 mM glucosyl units/g tissue in the rat cerebral cortex and 3.219 0.16 mM glucosyl units/g in the cerebellum. Nevertheless, it was not possible to see glycogen particles in brain slice by the means of electron microscopy. In 1972, Phelps performed an intracardiac perfusion using glutaraldehyde fixative in rat. Then, the glycogen particles could be seen in the cerebral cortex slices of rats. In normal rat, the number of particles were principally smaller than that described in the liver by Drochmans (1962). Mice were submitted to the xenobiotic methionine sulfoximine by an intraperitoneal administration of 100 mg/kg body weight. The animals were sacrificed by an intracardiac perfusion of glutaraldehyde as described above 24 h after dosing. In the slices from control animals that received an intraperitoneal administration of 0.9% NaCl,
182
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187
thionine sulfoximine is able to increase the glycogen accumulation up to the rosette level. Glycogen accumulation was very intense in the perivascular end-feet of astrocytes and sometimes, all the areas of the end-feet were filled with glycogen particles. When rats were used, the results were the same as in mice. Glycogen particles were observed in astrocytes and exclusively in astrocytes. Other cells, neurons for example, never contained glycogen particles. For this reason, we cultured astrocytes and we submitted these cells to the xenobiotic action. The medium of cultured astrocytes was renewed with DMEM medium containing 1 mM of methionine sulfoximine. The astrocytes were fixed 8 or 24 h after the medium renewal, using glutaraldehyde fixative. As in the in vivo brain, methionine sulfoximine also increased the number of glycogen particles in cultured astrocyte cytoplasm. This increase was faint when the medium contained 1 g/l of glucose and it was obvious when the medium contained 4.5 g/l (Fig. 3). Fig. 2. Glycogen particles in the cerebral cortex of mice. The animals were submitted to NaCl 0.9% for control experiments (A) or to 100 mg/kg of methionine sulfoximine (B). Numerous glycogen particles accumulate in the cytoplasm of astrocytes (G). Note here the end foot against the capillary filled with the particles. C, blood capillary; M, mitochondrion; R, endoplasmic reticulum. G × 16 000.
there was a few number of glycogen particles, thus looking like the slices described by Phelps (1972). Contrarily, in the slices from animals submitted to methionine sulfoximine, there was a tremendous increase in the number and in the size of glycogen particles (Fig. 2). In the perikarya, these particles were often gathered in areas, which were clear without surrounding membrane. A major feature of the glycogen particles was the presence of rosettes. Rosette is the term ascribed by Drochmans (1962) to the morphology of glycogen in liver when the particles are disposed in a way looking like the petals of the flower rose. Drochmans (1962) mentioned that rosettes are the highest form of glycogen organization in liver, when the synthesis of the polysaccharide is intense. In normal brain, glycogen never achieves the step of rosette formation. The xenobiotic me-
3. Possible mechanism of glycogen accumulation under the effect of methionine sulfoximine The xenobiotic methionine sulfoximine is the most powerful glycogenic agent in the brain. The actual mechanism of such an anabolic effect is not yet well understood. One can suppose that the xenobiotic enhances the activity of the synthesizing enzyme glycogen synthase. Another possibility is the inhibition of the catabolizing enzyme phosphorylase. The work of Folbergrova (1973) has shown that methionine sulfoximine did not markedly influence the activities of the enzymes directly related to glycogen synthesis. We have made a series of investigations that showed a marked increase in the activities and quantities of the gluconeogenic enzyme fructose-1,6-bisphosphatase when animals were submitted to methionine sulfoximine. Since these increases were exclusively localized in astrocytes, we cultured the latter cells and submitted them to the xenobiotic. Again, in cultured astrocytes, methionine sulfoximine displayed its increasing effect on the activity of fructose-1,6-bisphosphatase. Taking into ac-
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187
count all the preceding data, we concluded that methionine sulfoximine enhances the gluconeogenic pathway in the rodent brain. Indeed, fructose-1,6-bisphosphatase is the last key enzyme before the synthesis of glucose-6-phosphate. As all key enzymes, the preceding one catalyses an irreversible step and glucose-6-phosphate cannot be transformed again in its generating compounds by the same enzyme. This observation means that glucose-6-phosphate has to follow another metabolic pathway. One possibility is that of glycolysis, the first key enzyme in this pathway being phosphofructokinase. We have already shown that methionine sulfoximine has almost no effect on the activity of the latter enzyme, despite the large increase in the activity of the opposite enzyme fructose-1,6-bisphosphatase which was dis-
Fig. 3. Glycogen particles in cultured astrocytes. The astrocytes were cultured in a medium containing 4.5 g/l of glucose (A) or 4.5 g/l of glucose and 1 mM methionine sulfoximine (B). Glycogen particles accumulate in the cytoplasm of astrocytes and this accumulation is enhanced by methionine sulfoximine. G, glycogen; M, mitochondrion; N, nucleus. G × 16,000.
183
played under the effect of the xenobiotic (Hevor and Gayet, 1978). So, the glycolysis pathway is improbable. A second possibility is that of free glucose synthesis, a pathway that was documented during this decade. Theoretically, glucose6-phosphate can be converted into free glucose. But the supposed absence of glucose-6-phosphate translocase (Fishman and Karnovsky, 1986) militated against such a conversion taking place in the brain. Burchell and Waddell (1991) have shown that antibodies raised against the liver glucose-6phosphatase did not allow the verification of the activity of this phosphatase in the brain. This may be due to the lack of cross-reaction in the immunolabelling of the two enzymes originating from two kinds of tissues. Recent investigations by Forsyth et al. (1993) produced some important data on the activity of glucose-6-phosphatase in the brain. These authors have shown that a glucose-6-phosphatase hydrolysing activity exists in homogenates of cultured astrocytes, 40% of this activity being specific to the catalytic subunit of glucose-6-phosphatase localized in microsomes. In many aspects, the enzyme looks like that of the liver, the kidney and the ß cells of the pancreas (molecular weight, sensitivity to pH, latency in activity) although the Km value was higher. This high Km may be attributed to the decrease in activity during cultivation. Moreover, Forsyth et al. (1993) have also shown that the microsomes are permeable to glucose-6-phosphate. These data strongly support the idea that gluconeogenesis in astrocytes can be complete, up to free glucose. So the idea of a free glucose synthesis through the catalysis of glucose-6-phosphatase in astrocytes appears likely, at least in cultured astrocytes. A third pathway that can be followed by glucose-6phosphate under the effect of methionine sulfoximine is that of glycogen synthesis. Our experiments showed that methionine sulfoximine induced a large increase in glycogen accumulation as well in vivo as in cultured astrocytes. A major property of the xenobiotic is a strong and irreversible inhibition of glutamine synthetase (Meister, 1978). This inhibition may normally results in the accumulation of the substrate glutamate. Paradoxically, the glutamate level does not increase, and contrarily, it decreased a little in some
184
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187
areas of the brain (Fonnum and Paulsen, 1990). Phelps (1975) suggested that glutamate might be massively incorporated in glycogen under the effect of methionine sulfoximine. Up to now, the actual pathway of such an incorporation was not demonstrated. We have hypothesized that the first steps of gluconeogenesis may be the major way followed by glutamate carbon toward glycogen synthesis. Indeed, if glutamate is converted to alpha-ketoglutarate, the latter can be incorporated in the cycle of Krebs and then, it can follow gluconeogenesis pathway. In the present work, taking advantage of the molecular biology means, we gain insight the problem of the possible involvement of gluconeogenic pathway in the glycogen accumulation in the brain. For this, we blocked the expression of fructose-1,6-bisphosphatase to see whether or not, changes will happen in the glycogen content of cultured astrocytes.
4. Fructose-1,6-bisphosphatase mRNA expression in eukaryotic vector The purpose of this investigation was to develop tools that allow the modulation of the specific expression of gluconeogenic enzymes fructose-1,6-bisphosphatase in a specific population of cells of the central nervous system, the astrocytes. The vector used was a construction built from different vectors in order to produce a plasmid that expresses the inserted cDNA under the control of a strong promoter. This promoter controls the expression of an intermediate filament protein, which is exclusively synthesized in astrocytes (Baba et al., 1997). This protein is the glial fibrillary acidic protein (GFAP). The GFAP promoter was cloned as a 2.7 kbp from mouse brain (Fig. 4). It contains a negative and a positive domain and two trans regulation domains. The trans regulation domain that is closed to the transcription initiation site ( + 1) contains two enhancer sequences (−455 and −306) and the sequences for various nuclear factors (AP1: −238 and − 169; AP2: − 339 and − 95; CRE: −154; NF1: − 118; heat choc: − 455). The CAT and TATA boxes were localized at −83 and −29, respectively
Fig. 4. Organisation of the promoter of the glial fibrillary acidic protein (GFAP). Designed from Miura et al. (1990) and from Baba et al. (1997).
(Miura et al., 1990). We produced the mouse brain GFAP promoter by PCR using specific primers from − 2215 to + 7, giving a 2222 bp cDNA that corresponds to the fully active GFAP promoter. The identity of the PCR product to the GFAP promoter sequence was assessed by restriction mapping and DNA sequencing. The primer sequences were modified in order to introduce restriction sites for cloning. The eukaryotic vector that we therefore produced is shown in Fig. 5. In addition to the GFAP promoter it contains the genes providing the resistance to ampicillin for selection in bacteria and to neomycin for selection in eukaryotic cells. The multiple cloning site (MCS) containing
Fig. 5. Plasmid pASTM-Neo. This plasmid contains the glial fibrillary acidic protein (GFAP) promoter upstream the multicloning sites. The ampicillin resistance gene (AmpR) allows selection of prokariotic cells, while neomycin resistance gene (NeoR) allows selection in eukaryotic cells.
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187
four unique restriction sites is located downstream the GFAP promoter. The sens and antisens cDNA of genes which expressions will be modulated in astrocytes, were cloned inside this MCS. This eukaryotic vector was called pASTM-Neo. The efficiency of the mouse GFAP promoter in the pASTM-Neo was assessed using two different ways. Firstly, the GFAP promoter was removed from the pASTM-Neo and cloned in the pEGFP1 plasmid that contains a reporter gene constituted of a green auto fluorescent protein called EGFP. Mouse astrocytes or COS cells were transfected with the latter plasmid. Only astrocytes showed fluorescence. Secondly, the gene coding EGFP protein was cloned in the multicloning site of pASTM-Neo and this plasmid was transfected in either mouse astrocytes or COS cells. Only the mouse astrocytes produced EGFP protein. These experiments demonstrated that the GFAP promoter inserted in the eukaryotic vector to build pASTM-Neo is active and could be considered as a strong promoter.
185
Fig. 6. Structure of fructose-1,6-bisphosphate gene (designed from El-Maghrabi et al., 1991 and El-Maghrabi et al., 1995).
The quality and integrity of the plasmids containing the fructose-1,6-bisphosphatase cDNAs were controlled by restriction mapping or DNA sequencing. The vectors containing antisens fructose-1,6-bisphosphatase cDNA was called pASTM-FBP.
6. Astrocytes culture and transfection 5. cDNA cloning Fructose-1,6-bisphosphatase was cloned as antisens cDNA inside the MCS of the pASTM-Neo. The fructose-1,6-bisphosphatase cDNA was obtained using RT-PCR as previously described (Cloix et al., 1997). The RT-PCR was realized on total RNA extracted from mouse brain using the primers, which were designed to span three introns (Fig. 6) accordingly to the published mouse fructose-1,6-bisphosphatase gene (Nomura et al., 1994; Cloix et al., 1997). This was done on the ground of a similarity between the mouse brain gene organization and that published for human (El-Maghrabi et al., 1995) and rat (El-Maghrabi et al., 1991) liver enzyme. The primers sequences were modified in order to introduce restriction site for cloning. The 5% and 3% primers contained additional bases to create restriction sites for AccI and BamHI, respectively. This allowed the insertion of the 547bp fructose-1,6-bisphosphatase cDNA in the antisens position in the pASTM-Neo MCS at the same sites.
Primary cultures of mouse astrocytes were performed as described above. The first attempts of transfection of these primary cultures with either pASTM-FBP or pASTM-Neo were unsuccessful. The reason for this failure was the low division rate of astrocytes in primary culture. Therefore, in order to improve the transfection yield of astrocytes, we used astrocytes after eight passages. The astrocytes were transfected with the plasmids using the co precipitation method of DNA with calcium phosphate, as follows. Twenty mg of ethanol-precipitated plasmid were dissolved in 0.5 ml of 2× HBS buffer (280 mM NaCl, 10 mM KCl, 1.5 mM Na2HPO4, 12 mM dextrose, 50 mM HEPES, pH 7.5) and co-precipitated with 0.5 ml of 0.25 mM CaCl2. The precipitate was mixed with the culture medium, and the astrocytes (105 cells) were incubated with this medium for 8 h at 37°C. Thereafter, the medium was renewed and 2 days later geneticin (400 mg/ml) was added to the medium for selection of geneticin-resistant astrocytes. When these cells were selected (3–4
186
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187
months), they were sub cultured (1 – 2 months) in order to obtain enough cells for experiments and for frozen stocks. Therefore, transfected astrocytes were obtained and used after a total of 12 – 15 passages.
7. Glycogen level in transfected astrocytes Mouse astrocytes were transfected with pASTM-FBP. The control of transfection was done using pASTM-Neo. When geneticin-resistant astrocytes were obtained as stable cell cultures, the activities of fructose-1,6-bisphosphatase (Verge´ et al., 1996) was measured as well as the cellular level of glycogen (Verge´ and He´vor, 1995). The results of these experiments are presented in Fig. 7. The activity of fructose-1,6-bisphosphatase in astrocytes transfected with fructose-1,6-bisphosphatase antisens cDNA was significantly reduced by 85%. A major result was a large decrease observed in glycogen content of the astrocytes transfected with fructose-1,6-bisphosphatase antisens cDNA, up to 75%. The preceding result demonstrated that it was possible to chronically decrease the cellular glycogen content of cultured astrocytes using antisens cDNA that inhibited the gluconeogenic enzyme fructose-1,6-bisphosphatase. The viability of these astrocytes was not apparently altered by this transfection. These permanently transfected astrocytes constitute precious tools for studying astro-
Fig. 7. Change in glycogen content of cultured astrocytes transfected with plasmids containing antisens cDNA of fructose-1,6-bisphosphatase.
cytic glucose and glycogen metabolisms, and to analyse the effect of methionine sulfoximine, which impairs glycogen metabolism in the brain. These genetically modified astrocytes could also be used to study the relationship between astrocyte glucose metabolism and neuron functions. These experiments will be done and the next step will be the effect of methionine sulfoximine on glycogen accumulation in astrocytes tranfected with fructose-1,6-bisphosphatase antisens cDNA.
Acknowledgements This work was supported by funds from La Ligue Nationale contre le Cancer, Comite´ du Loiret and from the Re´gion Centre. Katy He´laryBernard is a recipient of a grant from the Re´gion Centre. Many thanks to Philippe Moreau for his technical assistance.
References Baba, H., Nakahira, T.A., Morita, N., Tanaka, F., Akita, H., Ihenaka, K., 1997. GFAP gene expression during development of astrocytes. Dev. Neurosci. 19, 49 – 57. Booher, J., Sensenbrenner, M., 1972. Growth and cultivation of dissociated neurons and glial cells from embryonic chick, rat and human brain in flask cultures. Neurobiology 2, 97 – 105. Burchell, A., Waddell, I.D., 1991. The molecular basis of the hepatic microsomal glucose-6-phosphatase system. Biochim. Biophys. Acta 1092, 129 – 137. Cloix, J.F., Beaulieu, E., He´vor, T.K., 1997. Diversity of fructose-1,6-bisphosphatase transcripts in mouse brain, liver, kidney and heart. NeuroReport 8, 617 – 622. Drochmans, P., 1962. Morphologie du glycoge`ne: etude au microscope e´lectronique en coloration ne´gative du glycoge`ne particulaire. J. Ultrastruct. Res. 6, 141 – 163. El-Maghrabi, M.R., Lange, A.J., Jiang, W., Yamagata, K., Pilkis, S.J., 1995. Human fructose-1,6-bisphosphatase gene (FBP1): exon-intron organisation, localisation to chromosome bands 9q22.2 – 22.3, and mutation screening in subjects with fructose-1,6-bisphosphatase deficiency. Genomics 27, 520 – 525. El-Maghrabi, M.R., Lange, A.J., Ku¨mmel, L., Pilkis, S.J., 1991. The rat liver fructose-1,6-bisphosphatase. J. Biol. Chem. 266, 2115 – 2120. Fishman, R.S., Karnovsky, M.L., 1986. Apparent absence of a translocase in the cerebral glucose-6-phosphatase system. J. Neurochem. 46, 371 – 378.
K. He´lary-Bernard et al. / Toxicology 153 (2000) 179–187 Folbergrova, J., Passonneau, J.V., Lowry, O.H., Schulz, D.W., 1969. Glycogen, amonia and related compounds in the brain during seizures evoked by methionine sulfoximine. J. Neurochem. 16, 191 –203. Folbergrova, J., 1973. Glycogen and glycogen phosphorylase in the cerebral cortex of mice under the influence of methionine sulfoximine. J. Neurochem. 20, 547–557. Fonnum, F., Paulsen, R.E., 1990. Comparison of transmitter amino acid levels in rat globus pallidus and neostriatum during hypoglycemia or after treatment with methionine sulfoximine or gamma-vinyl gamma-aminobutyric acid. J. Neurochem. 54, 1253–1257. Forsyth, R.J., Bartlett, K., Burchell, A., Scott, H.M., Eyre, J.A., 1993. Astrocytic glucose-6-phosphatase and the permeability of brain microsomes to glucose-6-phosphate. Biochem. J. 294, 145 –151. Hevor, T.K., Delorme, P., 1991. Biochemical and ultrastructural study of glycogen in cultured astrocytes submitted to the convulsant methionine sulfoximine. Glia 4, 64–69. Hevor, T.K., Gayet, J., 1978. Fructose-1,6-diphosphatase and phosphofructokinase activities in the brain of mice and rats submitted to the convulsant methionine sulfoximine. Brain Res. 150, 210 – 215. Hutchins, D.A., Rogers, K.J., 1970. Physiological and drug-induced changes in the glycogen content of mouse brain. Br. J. Pharmacol. 39, 1 – 9. Meister, A., 1978. Inhibition of glutamine synthetase and gamma-glutamylcysteine synthetase by methionine sulfox-
.
187
imine and related compounds. In: Seiller, N., Jung, M.J., Kosh-Weser, J. (Eds.), Enzyme Activated Irreversible Inhibitors. Elesevier, North Holland, pp. 187 – 210. Miura, M., Tamura, T.A., Mikoshiba, K., 1990. Cell-specific expression of the mouse glial fibrillary acidic protein genes: identification for astrocytes-specific expression. J. Neurochem. 55, 1180 – 1188. Nomura, M., Takihara, Y., Yasunaga, T., Shimada, K., 1994. One of the retinoic acid-inducible cDNA clones in mouse embryonal carcinoma F9 cells encodes a novel isoenzyme of fructose-1,6-bisphosphatase. FEBS Lett. 348, 201 – 205. Phelps, C.H., 1972. Barbiturate-induced glycogen accumulation in brain: an electron microscopic study. Brain Res. 39, 225 – 234. Phelps, C.H., 1975. An ultrastructural study of methionine sulfoximine-induced glycogen accumulation in astrocytes of the mouse brain cerebral cortex. J. Neurocytol. 4, 479 – 490. Swanson, R.A., Yu, A.C.H., Sharp, F.R., Chan, P.H., 1989. Regulation of glycogen content in primary astrocyte culture: effects of glucose analogues, phenobarbital, and methionine sulfoximine. J. Neurochem. 52, 1359 – 1365. Verge´, V., He´vor, T., 1995. Regulation of fructose-1,6-bisphosphatase activity in primary cultured astrocytes. Neurochem. Res. 20, 1049 – 1056. Verge´, V., Legrand, A., He´vor, T.K., 1996. Isolation of carbohydrate metabolic clones from cultured astrocytes. Glia 18, 244 – 254.