[1] 10w ATP induces leukemia inhibitory factor mRNA in cultured rat astrocytes

[1] 105w Leukemia inhibitory factor (LIF) is a cytokine involved in the survival and differentiation of the neural cells in the central and peripheral nervous systems. In the present study, we examined the effects of various neurotransmitter receptor agonists on LIF mRNA expression in cultured rat astrocytes, microglia and neurons to elucidate the cell types producing LIF and to clarify the neurotransmitter(s) regulating the mRNA expression. The results demonstrated that the expression of LIF mRNA was intensely induced by ATP in the cultured astrocytes. Experiments using ATP, UTP and related compounds showed the involvement of P2Y 2 and P2Y 4 purinoceptors in the expression induced by ATP.

[1] 117w Leukemia inhibitory factor (LIF) is a glycoprotein, which was identified as a differentiation factor for the murine myeloid leukemia cell line M1 (Gearing et al., 1987) and has been shown to be a pleiotropic cytokine that affects the proliferation, differentiation and survival of various types of cells. It stimulates the proliferation of murine myeloid IL-3dependent DA-1a leukemia cells (Moreau et al., 1988) and inhibits the differentiation of totipotent mouse embryonal stem cells (ES cells) (Williams et al., 1988;Smith et al., 1988). It also stimulates acute phase protein synthesis in hepatocytes (Baumann and Wong, 1989). Furthermore, a recent study with LIF knockout mice revealed that this cytokine plays an anti-inflammatory role in cutaneous inflammation (Zhu et al., 2001).

[2] 57w In the peripheral nervous system, LIF promotes the survival of sensory and motor neurons in vitro (Murphy et al., 1991;Martinou et al., 1992). Furthermore, LIF acts as a differentiation factor to change the phenotype of cultured sympathetic neurons from adrenergic to cholinergic (Yamamori et al., 1989), so it is also referred to as a ''cholinergic differentiation factor''.

[3] 149w In the central nervous system (CNS), LIF affects the survival and differentiation of astrocytes (Gadient et al., 1998;Bugga et al., 1998). In the adult rat brain, low levels of LIF mRNA are constitutively expressed (Yamamori, 1991) and the expression is up-regulated by brain injuries such as cortical lesions (Banner et al., 1997). In addition, LIF receptor (LIFR) mRNA is widely expressed in the adult rat brain, with most of the signals localized on neuronal cells (Yamakuni et al., 1996). These findings suggest that LIF may act on the neurons of the adult CNS to help their survival or regulate the synthesis of neurotransmitters and/or neuropeptides as is observed in the peripheral nervous system. Indeed, it has been reported that LIF prevents the axotomy-induced cell death of septal cholinergic neurons (Panni et al., 1999) and increases the mRNA expression of a neuropeptide, nociceptin (Buzas et al., 1999;Minami et al., 2001).

[4] 172w We previously showed that intraperitoneal injection of kainic acid at a convulsion-inducing dose induced the expression of LIF mRNA in the rat brain (Minami et al., 1991). This finding suggests that the intensive excitation of neurons, which is supposed to cause the release of various neurotransmitters, leads to the expression of LIF mRNA. However, little is known about the roles of neurotransmitters in the regulation of LIF mRNA expression, while some immunological stimuli such as lipopolysaccharide and inflammatory cytokines have been shown to induce the expression (Aloisi et al., 1994;Murphy et al., 1995). Furthermore, some controversy exists over which types of cells express LIF mRNA in the brain. In vivo and in vitro studies showed that LIF mRNA was expressed in cortical astrocytes following injury to the brain (Banner et al., 1997) and in cultured astrocytes, but not microglia (Murphy et al., 1995). On the other hand, Lemke et al. (1996) reported that LIF mRNA was predominantly expressed in neurons, not in glial fibrillary acidic protein (GFAP)-positive cells, in normal rat brain.

[5] 51w In the present study, we examined the effects of various neurotransmitter receptor agonists on the mRNA expression of LIF in cultured astrocytes, microglia and neurons. The results revealed that ATP strongly induced the expression in cultured astrocytes. Thus, we examined the regulation of LIF mRNA expression through P2 purinoceptors, ATP receptors.

[1] 26w In the nontreated control cultures, astrocytes expressed LIF mRNA weakly (Fig. 1A, upper panel), but no mRNA expression was detected in microglia and neurons (Fig. 1B

[2] 72w Table 1 Conditions of PCR for P2Y purinoceptors PCR primers (upper = forward, lower = reverse) Denature Anneal Extension Cycle no. P2Y 2 5VCTGGACTCCTGGAATAGTACC3V 94 jC, 30 s 60 jC, 60 s 72 jC, 90 s 35 5VAGCACCCACACAACCGCAG3V P2Y 4 5VCACCGATACCTGGGTATCTGCCAC3V 94 jC, 60 s 65 jC, 30 s 72 jC, 60 s 30 5VCAGACAGCAAAGACAGTCAGCACC3V P2Y 6 5VGCTTCCTCTTCTATGCCAAC3V 94 jC, 30 s 64 jC, 60 s 72 jC, 90 s 40 5VTAGGCTGTCTTGGTGATGTG3V

[3] 99w and C, upper panels). The cells were treated with 1 mM acetylcholine, dopamine, isoproterenol, serotonin, glutamic acid, NMDA, kainic acid, ATP or GABA for 1 h. In cultured astrocytes, LIF mRNA expression was increased markedly by ATP and moderately by dopamine and isoproterenol but not by the other agonists examined (Fig. 1A, upper panel). In microglia, LIF mRNA expression was increased slightly by ATP but not at all by the other agonists (Fig. 1B, upper panel). On the other hand, treatment with glutamate, but no other agonists, induced some LIF mRNA expression in neuronal cells (Fig. 1C, upper panel).

[4] 127w Effects of P2 purinoceptor antagonists, suramin and PPADS, on the LIF mRNA induction by ATP and UTP (100 AM each, 1 h) were examined. Fig. 7 shows that suramin concentration-dependently suppressed the LIF mRNA expression induced by ATP (Fig. 7A, upper panel) and UTP (Fig. 7B, upper panel). The antagonistic effect of suramin on the UTP-induced LIF mRNA expression was more potent than that on the ATP-induced one. The antagonistic effect of PPADS on the ATP-induced expression was much less potent than that of suramin (Fig. 7A) and PPADS did not suppress the UTP-induced expression (Fig. 7B). Fig. 5. Effects of ATP, ADP, AMP, adenosine and inosine (100 AM each, 1 h) on the expression of LIF (upper panel) and GAPDH (lower panel) mRNAs in cultured astrocytes.

[1] 267w LIF mRNA has been shown to be expressed in the brain in response to excitotoxin (Minami et al., 1991) and injury (Banner et al., 1997), suggesting that the release of neurotransmitter(s) evoked by neuronal excitation or the leakage of some substance(s) from injured neurons induces the expression. However, little is known about the regulation of LIF mRNA expression through neurotransmitter(s). In this study, we demonstrated that ATP markedly induced the expression of LIF mRNA in cultured astrocytes. This induction of mRNA expression led to the elevation of LIF protein levels. Western blot analysis revealed two major bands of 53 and 63 kDa. The mature forms of rat, mouse and human LIF are reported to be highly glycosylated molecules, varying in molecular weight from 38 to 67 kDa (Kurzrock et al., 1991). Dopamine and isoproterenol also induced the expression of LIF mRNA in astrocytes, but the induction was considerably weaker than that by ATP. Although ATP is well known as a high-energy phosphatecontaining intracellular molecule, a body of evidence has revealed that ATP is released from excited neurons and acts as a neurotransmitter or neuromodulator in the central and peripheral nervous systems (Evans et al., 1992;Edwards et al., 1992;Zimmermann, 1996). In addition, it is possible that a considerable amount of ATP is leaked from injured neurons because the intracellular concentration of ATP is much higher than the extracellular one (Neary et al., 1996;Zimmermann, 1996). The present results, combined with circumstantial evidence, suggest that ATP is a potential candidate for the neurotransmitter or substance, which mediates the induction of LIF mRNA expression by excitotoxin or injury in the brain.

[2] 97w In cultured microglia, no agonist examined induced the expression of LIF mRNA, except for a weak induction by ATP. This result is consistent with the report that LIF mRNA was induced in astrocytes, but not microglia, by pro-inflammatory agents such as LPS (Murphy et al., 1995). The microglial cultures used in this study contained 1-2% astrocytes. However, it is unlikely that the LIF mRNA expression observed in the ATP-treated microglial culture was due to the contaminating astrocytes, because the concentration dependency and time course of LIF mRNA induction by ATP were different between cultured astrocytes and microglia.

[3] 212w In cultured cortical neurons, LIF mRNA was not detected in the nontreated control culture. Although glutamate, but no other agonists, induced the expression, the effect was very weak. This result is consistent with the report by Banner et al. (1997) that LIF mRNA is expressed in astrocytes as well as in a small number of microglial cells, but not in neurons, following cortical brain injury. On the other hand, the same group reported that LIF mRNA was predominantly expressed in neurons, not in GFAP-positive cells, in the normal rat brain using a less stringent in situ hybridization protocol with the same antisense RNA probe (Lemke et al., 1996). It is possible that the level of LIF mRNA expression in neurons is too low to detect by highstringent in situ hybridization or Northern blot analysis. In this context, a RT-PCR method, which is considered to have greater sensitivity than Northern blotting, detected LIF mRNA in normal rat brain at different levels among several regions (Minami et al., 1991), suggesting the expression occurs in neurons rather than in glial cells. Collectively, these observations suggest that LIF is constitutively produced in neurons at low levels and induced in glial cells, especially astrocytes, at high levels in response to various insults such as excitotoxin and injury.

[4] 110w Because extracellular ATP is readily hydrolyzed to ADP, AMP, adenosine and then inosine by ectonucleotidases and adenosine deaminase (Zimmermann, 1996), it is possible that hydrolyzed products from ATP, not ATP itself, induced the expression of LIF mRNA when ATP was added to the cultured astrocytes. To test this possibility, the effects of ADP, AMP, adenosine and inosine on the expression of LIF mRNA were investigated. The results showed that the induction by ATP was more intense than that by any hydrolyzed product. Furthermore, ATPgS, a hydrolysisresistant ATP analog, induced the mRNA expression at a similar intensity to ATP. These results indicate that the induction was primarily due to ATP itself.

[5] 279w Cellular responses to extracellular ATP are mediated by specific receptors called P2 purinoceptors, which are classified into two families. The P2X and P2Y families contain seven ligand-gated, ionotropic receptors (P2X 1 -7 ) and six G protein-coupled, metabotropic receptors (P2Y 1, 2, 4, 6, 11, 12 ), respectively (Ralevic and Burnstock, 1998;von Kugelgen Fig. 8. Expression of P2Y 2 , P2Y 4 and P2Y 6 purinoceptor mRNAs in the cultured astrocytes detected by a RT-PCR method. and Wetter, 2000;Williams and Jarvis, 2000;Nicholas, 2001). In the present study, the effects of ATP, UTP and related compounds on the expression of LIF mRNA were examined at concentrations of 1 -100 AM. Reportedly, these compounds evoke no or very little activation of the P2X 7 receptor at these concentrations (Surprenant et al., 1996;Bianchi et al., 1999). a,h-MeATP, which is selective for P2X 1 and P2X 3 (Bianchi et al., 1999), did not induce the expression of LIF mRNA, indicating little involvement of these two subtypes in the induction. The induction by 2-MeSATP, which activates P2X 1 -6 , P2Y 1 and P2Y 11 (Bianchi et al., 1999;Williams and Jarvis, 2000;Nicholas, 2001), was much weaker than that by ATP. Furthermore, ADP, which is a stronger agonist of the P2Y 12 subtype than ATP and UTP (Nicholas, 2001), had a weaker effect at 100 AM (Fig. 3). These results suggested that the P2X 1 -7 , P2Y 1 , P2Y 11 and P2Y 12 subtypes contribute little to the induction, at least, at these ligand concentrations (1-100 AM). Consequently, P2Y 2 , P2Y 4 and P2Y 6 subtypes appear to be the receptors that play a crucial role in the expression of LIF mRNA.

[6] 200w RT-PCR analyses revealed the expression of P2Y 2 , P2Y 4 and P2Y 6 receptors in cultured astrocytes. Reportedly, the order of potency of ATP, UTP and related compounds in rat P2Y 2 , P2Y 4 and P2Y 6 receptors is ATP = UTP>AD-P>UDP>2-MeSATP in P2Y 2 (Chen et al., 1996), ATP = UT-P>UDP>ADP>2-MeSATP in P2Y 4 (Bogdanov et al., 1998;Webb et al., 1998) and UDP>UTP>ADP = 2-MeSATP> ADPhS>ATP = ATPgS in P2Y 6 (Chang et al., 1995;von Kugelgen and Wetter, 2000). In the present study, ATP and UTP at a concentration of 10 AM induced the expression of LIF mRNA to a similar degree, suggesting the involvement of P2Y 2 and P2Y 4 receptors. At a lower concentration (1 AM), UTP and UDP induced the expression more intensely than ATP. This result indicates that the P2Y 6 , rather than P2Y 2 or P2Y 4 , receptor is essential for the induction at the lower concentration. On the other hand, at a higher concentration (100 AM), the induction by ATP was more intense than that by UTP. The induction is probably mediated in part by P2X subtypes, which are activated by ATP, but not UTP, at the higher concentration.

[7] 294w These ideas were also supported by the results from the experiments using P2 purinoceptor antagonists, suramin and PPADS. According to recent articles (Ralevic and Burnstock, 1998;von Kugelgen and Wetter, 2000), suramin is active against P2X 1 -3, 5, 7 and P2Y 1, 2, 11, 12 , less active against P2Y 4, 6 and inactive against P2X 4, 6 . On the other hand, PPADS is active against P2X 1 -3, 5, 7 and P2Y 1 , less active against P2Y 6, 12 and inactive against P2X 4, 6 and P2Y 2, 4, 11 . In this study, LIF mRNA expression induced by UTP was suppressed by suramin, but not PPADS, suggesting the involvement of P2Y 2, 4, 11 . Because UTP is shown to be inactive on the P2Y 11 receptors (Ralevic and Burnstock, 1998), the induction of LIF mRNA expression is considered to be mediated principally by P2Y 2 and P2Y 4 subtypes. The antagonistic effect of suramin on the ATP-induced LIF mRNA expression was less potent than that on the UTPinduced one and PPADS somewhat suppressed the ATPinduced LIF mRNA expression. These results can be explained by the antagonistic activity of suramin and PPADS against P2X subtypes, which are activated by ATP, but not UTP. As described above, the induction of LIF mRNA expression by a higher concentration (100 AM) of ATP is probably mediated in part by P2X 1 -6 subtypes. The less antagonistic potency of suramin against the ATPinduced LIF mRNA expression is likely due to the induction through the suramin-insensitive P2X subtypes, P2X 4, 6 . On the other hand, the weak antagonistic activity of PPADS against the ATP-induced expression is explained as an antagonistic effect on the induction through the PPADSsensitive P2X subtypes, P2X 1 -3, 5 .

[8] 167w In this study, we demonstrated that (1) among various neurotransmitter receptor agonists, ATP markedly induced the expression of LIF mRNA in cultured astrocytes, (2) the induction by ATP was mediated by P2 purinoceptors, principally P2Y 2 and P2Y 4 subtypes, and (3) none of the neurotransmitter receptor agonists examined markedly induced the expression in cultured neurons or microglia. These results, combined with previous findings on the actions of LIF in the peripheral (Murphy et al., 1991;Martinou et al., 1992;Yamamori et al., 1989) and central (Panni et al., 1999;Buzas et al., 1999;Minami et al., 2001) nervous systems, suggest interaction between neurons and astrocytes via ATP and LIF. Namely, it is possible that ATP released from excited neurons or leaked from injured neurons acts on astrocytes to produce LIF and then LIF acts on neurons to prevent cell death or to affect neurotransmitter/neuropeptide synthesis. Further study is necessary to elucidate the involvement of ATP in the expression of LIF mRNA in vivo induced by excitotoxin injection or brain injury.

[1] 60w Acetylcholine was purchased from Daiichi Pharmaceutical (Tokyo, Japan). ADP was obtained from Oriental yeast (Tokyo, Japan). 2-Methylthio ATP (2-MeSATP) was from Research Biochemicals International (Natick, MA). GABA, serotonin, ATP, AMP, adenosine, inosine, adenosine 5V-O-(3-thiotriphosphate) (ATPgS), adenosine 5V-O-(2-thiodiphosphate) (ADPhS), a,h-methylene ATP (a,h-meATP), UTP and UDP were purchased from Sigma (St. Louis, MO). All other chemicals were from Nacalai tesque (Kyoto, Japan).

[2] 105w The cDNAs for rat LIF and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) were obtained by RT-PCR with poly (A) + RNA extracted from whole brains of Sprague -Dawley rats treated with lipopolysaccharide (1 mg/kg ip). Rat LIF and GAPDH cDNAs were subcloned into the plasmid vector pCRII (Invitrogen, San Diego, CA) and pBluescript II SK( À ) (Stratagene, La Jolla, CA), respectively, and their sequences were confirmed by fluorescent dye terminator sequencing. 32 P-labeled antisense RNA probes for LIF and GAPDH mRNAs were synthesized in the presence of [a 32 P]uridine triphosphate (15 TBq/mmol, Amersham Pharmacia Biotech UK, Buckinghamshire, UK) using SP6 and T7 RNA polymerases, respectively.

[3] 162w Total RNA was extracted from cultured cells by a singlestep acid phenol/chloroform method with ISOGEN (Nippon Gene, Tokyo, Japan). Total RNA samples (10 Ag each) were fractionated by electrophoresis in 1.2% agarose gels containing 6% formaldehyde, transferred onto nylon membrane (Biodyne, Pall, Glen Cove, NY) and baked at 80 jC for 2 h. The blots were prehybridized and then hybridized to 32 Plabeled antisense RNA probe for rat LIF mRNA at 65 jC. The membrane was washed twice in 2 Â SSC/0.1% SDS for 5 min each at room temperature and then twice in 0.1 Â SSC/0.1% SDS for 30 min each at 65 jC. The membrane was exposed to X-ray film at À 80 jC using an intensifying screen. In some experiments, autoradiograms were quantified using a bioimaging analyzer BAS2000 (Fuji, Tokyo, Japan). The membrane was rehybridized to 32 P-labeled antisense RNA probe for rat GAPDH mRNA and the expression levels of LIF mRNA were normalized to those of GAPDH mRNA.

[4] 305w Total RNA was extracted from cultured cells as described above. Each total RNA sample (2 Ag) was incubated with 300 ng of random primer hexamer at 70 jC for 10 min in a volume of 22 Al and then on ice for 5 min. Reverse transcription was started by adding 10 Al of 5 Â RT buffer, 5 Al of 100 mM dithiothreitol, 10 Al of dNTP mixture containing 2.5 mM each of dATP, dCTP, dGTP and dTTP and 3 Al of 200 units/Al SuperScript II/RNaseH À reverse transcriptase (Gibco BRL, Rockville, MD). The reaction mixture was incubated at 25 jC for 10 min and then at 37 jC for 2 h. The RT reaction was terminated by heating at 70 jC for 15 min and then the template RNA was degraded by incubation with 2 units of RNaseH (Gibco BRL) at 37 jC for 30 min. An aliquot of the RT product (1 Al) was mixed with 2.5 units of AmpliTaq DNA polymerase (Perkin Elmer, Branchburg, NJ), 20 pmol each of forward and reverse primers in 50 Al of 1 Â PCR buffer containing 10 mM Tris -HCl (pH 8.3), 50 mM KCl, 2 mM MgCl 2 and 200 AM each of dATP, dCTP, dGTP and dTTP. cDNA was amplified using a thermal cycler (GeneAmp PCR system 2400, Perkin Elmer). The sequences of forward and reverse primers and the conditions of PCR for rat P2Y 2 , P2Y 4 and P2Y 6 purinoceptors are shown in Table 1. An aliquot of PCR product was electrophoresed in 0.7% agarose/1.2% Synergel (Diversified Biotech, Boston, MA) in TAE buffer. The gel was stained with ethidium bromide and photographed. To confirm that the obtained PCR products were the targeted cDNAs, the cDNA fragments were subcloned into the plasmid vector pCR II and analyzed by fluorescent dye terminator sequencing.

[5] 207w Cultured astrocytes were collected by scraping in PBS containing 1 mM ethylenediaminetetraacetic acid (EDTA), precipitated and stored at À 80 jC until use. The astrocytes were resuspended and homogenized in 1 ml of PBS containing 2.5 mM EDTA and 0.05% Tween-20. The homogenates were centrifuged at 10,000 Â g for 15 min and the supernatants were used for Western blotting. The protein concentration in each sample was determined with a protein assay kit (Bio-Rad Laboratories, Hercules, CA). Samples containing 5 Ag of protein each were denatured at 100 jC for 5 min in a buffer containing 50 mM Tris (pH 6.8), 2% sodium dodecyl sulfate (SDS), 6% h-mercaptoethanol, 10% glycerol and 0.01% bromophenol blue, electrophoresed in a 10% polyacrylamide gel containing 0.1% SDS and then transferred to a PVDF membrane (Immobilon-P, Millipore, Bedford, MA) using a semidry electrotransfer apparatus (Atto, Tokyo, Japan) with a Tris -borate buffer (pH 9.5) containing 20% methanol. The membrane was incubated with 2 Ag/ml of polyclonal goat anti-LIF antibody (Santa Cruz Biotechnology, Santa Cruz, CA, cat. no. sc-1336) at 4 jC for 15 h. The primary antibody was visualized with a peroxidase-conjugated donkey anti-goat IgG (Jackson Immunoresearch Laboratories, West Grove, PA) and ECL plus Western blotting detection system (Amersham Pharmacia Biotech UK).

[1] 28w Brains from neonatal or fetal Sprague -Dawley rats were used to prepare glial and neuronal cultures. Cultures were maintained at 37 jC in 5% CO 2 /95% air.

[2] 148w Astrocyte cultures were prepared according to the method of Takuma et al. (1994). Primary cultures were prepared from the cerebral cortices of neonatal rats (1 day old) by digestion with dispase. Dissociated cells were plated onto 75 ml tissue culture flasks with 10 ml of Eagle's minimal essential salt medium (Eagle's MEM) supplemented with 10% heat-inactivated fetal bovine serum. After cultivation for 14 days, the flasks were shaken on an orbital shaker at 250 rpm for 17 h at 37 jC to remove microglia, oligodendrocytes and oligodendrocyte/type II astrocyte progenitor cells. After 13 -15 days of cultivation in 100 mm culture dishes, astrocytes at confluence were used for experiments. The purity of astrocytes was > 90%, as determined by immunostaining with anti-GFAP antibody. Astrocytes were maintained in Eagle's MEM supplemented with 10% heat-inactivated fetal bovine serum and treated with receptor agonists or related compounds in serum-free Eagle's MEM.

[3] 197w Cultured microglia were prepared according to the method of Suzumura et al. (1991). Primary cultures were prepared from the whole brains of neonatal rats (1 or 2 days old). Whole brains were freed of meninges and ependyma, minced and then sieved through 50 Am nylon mesh. Dispersed cells were plated to 75 ml tissue culture flasks with 10 ml of Eagle's MEM supplemented with 10% heatinactivated fetal bovine serum and 5 Ag/ml of insulin. After cultivation for 14 days, the flasks were shaken on an orbital shaker at 180 rpm for 3 h at 37 jC. From the medium containing the detached cells, microglial cells were recovered as the cells adhering to uncoated plastic dishes. The recovered microglial cells were resuspended and plated at a density of 1 Â 10 6 cells/dish on culture dishes. After 3 days of cultivation in 100 mm culture dishes, microglial cells were used for experiments. The purity of the microglia was >98%, as determined by immunostaining with OX-42 antibody. Microglial cultures were maintained and treated with receptor agonists in Eagle's MEM supplemented with 10% heat-inactivated fetal bovine serum, 5 Ag/ml of insulin and 50 ng/ml of recombinant human macrophage colony-stimulating factor.

[4] 117w Neuronal cultures were prepared from the cerebral cortex of fetal rats (17 -19 days gestation) as described by Kume et al. (1997). The cerebral cortices of fetal rats were minced and sieved through stainless steel mesh. The cells were plated onto 60 mm culture dishes. The culture was maintained in Eagle's MEM supplemented with additional glucose (final concentration 11 mM), HEPES (10 mM) and 10% heat-inactivated fetal bovine serum. After 8 days of plating, non-neuronal cells were removed by addition of 10 AM cytosine arabinoside and then the culture was maintained in Eagle's MEM supplemented with 10% heat-inactivated horse serum instead of fetal bovine serum. After 12 days of cultivation, the culture was treated with receptor agonists.

[5] 64w The concentration dependency of the induction of LIF mRNA expression was examined in cultured astrocytes and microglia. Fig. 2A (upper panel) shows the concentrationdependent induction of LIF mRNA by ATP at a range of 0.1 -1000 AM in the cultured astrocytes. On the other hand, only 1000 AM, but no lower concentrations, of ATP induced LIF mRNA expression in the cultured microglia (Fig. 2B).

[6] 92w The time course of the LIF mRNA expression induced by 100 AM ATP was examined in cultured astrocytes. The expression peaked at 1 h after the addition of ATP and disappeared at 4 h (Fig. 3A, upper panel). Thus, in subsequent experiments, the induction was examined at 1 h after the addition of ATP, UTP or related compounds. On the other hand, the LIF mRNA expression in the cultured microglia peaked at 2 h after the addition of ATP and had decreased to nearly the basal level by 8 h (Fig. 3B).

[7] 75w The time course of the LIF protein production induced by ATP was examined in cultured astrocytes (Fig. 4). Western blot analysis revealed two major bands of 53 and 63 kDa. A detectable level of LIF protein was produced in the nontreated culture and the protein level was elevated by the treatment with ATP. The elevation peaked at 1 h after the addition of ATP and had decreased to nearly the basal level by 4 h.

[8] 83w Effects of the hydrolyzed products from ATP, that is, ADP, AMP, adenosine and inosine, on LIF mRNA expression were examined. The cultured astrocytes were treated with 100 AM of each compound for 1 h. As shown in Fig. 5 (upper panel), ATP markedly induced the expression of LIF mRNA. The induction by ADP was intense but weaker than that by ATP. Moderate to weak induction was observed in the cultures treated with AMP or adenosine. Inosine showed no effect on LIF mRNA expression.

[9] 64w Effects of ATP, UTP and related compounds at concentrations of 1, 10 and 100 AM on LIF mRNA expression were examined (Fig. 6). In terms of potency to induce the expression, the compounds ranked UTP = UDP>ATP = ATPgS = ADPhS = 2-MeSATP>a,h-meATP at 1 AM, ATP = ATPgS = UTP = UDP>ADPhS>2-MeSATP>a,h-meATP at 10 AM and ATPgS = ATP>UTP>UDP>ADPh-S> 2-MeSATP>a,h-meATP at 100 AM.

[10] 22w RT-PCR analysis revealed the expression of P2Y 2 , P2Y 4 and P2Y 6 purinoceptor mRNAs in the cultured astrocytes (Fig. 8).