Expression and functional characterization of P2Y 1 and P2Y 12 nucleotide receptors in long-term serum-deprived glioma C6 cells Patryk Krzemin ´ ski, Dorota Supłat, Rafał Czajkowski, Paweł Pomorski and Jolanta Baran ´ ska Department of Molecular and Cellular Neurobiology, Nencki Institute of Experimental Biology, Polish Academy of Sciences, Warsaw, Poland The focus of this study is glioma C6 cells, which are chemically transformed glial cells derived from adult rat brain. They have oligodendrocytic and astrocytic progenitor properties and are often used as a bio- chemical model system for studies related to astrocytes [1]. They are also used as a model for studying glioma biology [2], e.g. glioblastoma multiforme, which is pathologically characterized by a marked increase in cell growth and proliferation. C6 cells can be easily implanted into rat brains in vivo to test invasiveness. To date, there is no better model for examining this process. The increased invasiveness of these cells was found to be associated with the constitutive up-regula- tion of phosphatidylinositol 3-kinase (PI3-K) ⁄ protein Keywords Akt; glioma C6 cells; mitogen-activated protein kinase; nucleotide receptors; serum deprivation Correspondence J. Baran ´ ska, Department of Molecular and Cellular Neurobiology, Nencki Institute of Experimental Biology, Polish Academy of Sciences, 3 Pasteur Street, Warsaw PL 02-093, Poland Fax: +48 22 8225342 Tel: +48 22 5892300 E-mail: j.baranska@nencki.gov.pl (Received 6 June 2006, revised 7 February 2007, accepted 14 February 2007) doi:10.1111/j.1742-4658.2007.05741.x We characterized the expression and functional properties of the ADP- sensitive P2Y 1 and P2Y 12 nucleotide receptors in glioma C6 cells cultured in medium devoid of serum for up to 96 h. During this long-term serum starvation, cell morphology changed from fibroblast-like flat to round, the adhesion pattern changed, cell-cycle arrest was induced, extracellular sig- nal-regulated kinase (ERK1 ⁄ 2) phosphorylation was reduced, Akt phos- phorylation was enhanced, and expression of the P2Y 12 receptor relative to P2Y 1 was increased. These processes did not reflect differentiation into astro- cytes or oligodendrocytes, as expression of glial fibrillary acidic protein and NG2 proteoglycan (standard markers of glial cell differentiation) was not increased during the serum deprivation. Transfer of the cells into fresh medium containing 10% fetal bovine serum reversed the changes. This demonstrates that serum starvation caused only temporary growth arrest of the glioma C6 cells, which were ready for rapid division as soon as the environment became more favorable. In cells starved for 72 and 96 h, expression of the P2Y 1 receptor was low, and the P2Y 12 receptor was the major player, responsible for ADP-evoked signal transduction. The P2Y 12 receptor activated ERK1 ⁄ 2 kinase phosphorylation (a known cell prolifer- ation regulator) and stimulated Akt activity. These effects were reduced by AR-C69931MX, a specific antagonist of the P2Y 12 receptor. On the other hand, Akt phosphorylation increased in parallel with the low expression of the P2Y 1 receptor, indicating the inhibitory role of P2Y 1 in Akt pathway signaling. The shift in nucleotide receptor expression from P2Y 1 to P2Y 12 would appear to be a new and important self-regulating mechanism that promotes cell growth rather than differentiation and is a defense mechan- ism against effects of serum deprivation. Abbreviations Akt, protein kinase B ⁄ Akt kinase; DIC, differential interference contrast; ERK, Ras extracellular signal-regulated kinase; GFAP, glial fibrillary acidic protein; IRM, interference reflection microscopy; MEM, minimum essential medium; 2MeSADP, 2-methylthio-ADP; NaCl ⁄ P i , phosphate-buffered saline; PI3-K, phosphatidylinositol 3-kinase; PTEN, phosphatase and tensin homolog. 1970 FEBS Journal 274 (2007) 1970–1982 ª 2007 The Authors Journal compilation ª 2007 FEBS kinase B ⁄ Akt kinase (Akt) signaling [3] and with glioma-induced angiogenesis [4]. Among other functions in the brain, nucleotide re- ceptors act as sensors of necrotic cell death. The abun- dance of ATP released from dying cells evokes a particular response in surviving neighbor cells, which may result in their transformation from normal to malignant cells. This is why studies of nucleotide receptors in transformed cells are so important. The large family of nucleotide receptors is divided into two major classes [5]: P1, responding to adenosine; P2, responding to ATP, ADP, UTP and UDP. P2 recep- tors are further divided into two groups [6,7]: P2X, the intrinsic ligand-gated ion channels; P2Y, members of the G-protein coupled receptor superfamily. To date, eight genuine mammalian P2Y receptor subtypes have been cloned and pharmacologically characterized. Of these, P2Y 2 responds to ATP and UTP, whereas P2Y 1 and P2Y 12 respond to ADP. P2Y 1 and P2Y 2 receptors are both coupled to phospholipase C [8]. Whereas the presence of the P2Y 2 receptor in glioma C6 cells is generally accepted, the presence of the P2Y 1 receptor, which stimulates phospholipase C, has been a subject of serious debate. Studies on glioma C6-2B and C6 cells have shown the presence of an additional P2Y 1 -like receptor, termed P2Y-C6 or P2Y AC [9–14]. This P2Y receptor, which is negatively coupled to adenylate cyclase, has been cloned and designated P2Y 12 [15–17]. Results from our laboratory revealed that both clas- sic P2Y 1 and P2Y 12 receptors coexist in glioma C6 cells: P2Y 1 , linked to phospholipase C, and P2Y 12 , inhibiting adenylate cyclase [18,19]. Observations on P2Y 1 receptor made by us and other authors [11,13,14,20] have suggested that its functional inactiv- ity can be caused by differences in culture conditions. We have shown that mRNA expression level, as well as the extent of the response to ADP and ADP ana- logs such as 2-methylthio-ADP (2MeSADP) depends on the presence of serum in the cell culture medium [20]. 2MeSADP has been chosen as a ligand because it is the agonist of both receptors studied, and ecto- enzymes present on the glioma C6 cells are unable to hydrolyze this compound [21]. In cells cultivated in medium devoid of serum, as well as in a serum-free chemically defined medium, expression of the P2Y 1 receptor is low [20,22]. We have also found that agon- ists that stimulate P2Y 1 and P2Y 12 receptors differen- tially affect the Ras extracellular signal-regulated kinase (ERK)1 ⁄ 2 and PI3-K signaling pathways and that it is the P2Y 12 receptor that predominates in ADP-evoked ERK1 ⁄ 2 activation. These results also suggest that P2Y 1 and P2Y 12 receptors may regulate each other’s activities by cross-talk mechanisms [20,23,24]. In carcinogenesis, several alterations in cell physio- logy have been proposed, which must cooperate for malignant growth to occur. In the first place, there must be self-sufficiency of the growth-promoting signal [25]. Therefore, in the present study, we used glioma C6 cells cultured in medium devoid of serum for up to 96 h and examined the effect of such long-term serum starvation on cellular morphology, adhesion, prolifera- tive and differentiation state, basal level of ERK1 ⁄ 2 and Akt kinase phosphorylation, and P2Y 1 and P2Y 12 receptor protein expression and functional activity. Examination of the effects of specific pharmacological agents (a single agonist and two specific antagonists of P2Y 1 and P2Y 12 receptors) on Akt and ERK1 ⁄ 2 phos- phorylation should help to clarify the effect of cross- talk between the two receptors on cell physiology. Our data show that the P2Y 12 receptor primarily contributes to glioma C6 cell growth and proliferation, as long-term serum deprivation strongly decreases expression of the P2Y 1 receptor, and under these con- ditions the concentration of the P2Y 12 receptor protein predominates. We also observed a simultaneous increase in Akt kinase phosphorylation. We suggest that this high level of Akt phosphorylation is an important, self-sufficient mechanism for survival under inhospitable conditions. These results may be useful in the development of therapeutic strategies targeting brain tumor angiogenesis. Results Glioma C6 cells are characterized by a typical fibro- blast-like flat morphology clearly visible by differential interference contrast (DIC) microscopy (also known as Nomarski contrast microscopy). As shown in Fig. 1A (control), the cells are flattened and their adhesion spots, visualized by interference reflection microscopy (IRM), are distributed under the periphery of the cell body (Fig. 2A, arrows) in a manner characteristic of motile cells. Vinculin staining reveals adhesion spots distributed evenly on the whole basolateral surface of the cell (Fig. 2E). However, during prolonged (48, 72 and 96 h) serum deprivation, changes in cell morphol- ogy from an epithelial, bipolar, to a rounded shape can be observed (Fig. 1B–D). The cells change from a fibroblast-like flat morphology to a rounded appear- ance characterized by numerous long extrusions. These particular changes were clearly visible when single enlarged cells are compared (Fig. 1H versus Fig. 1I). Figure 1G shows that, whereas the percentage of altered, rounded cells in the control amounted only P. Krzemin ´ ski et al. P2Y 1 and P2Y 12 in serum-deprived glioma C6 FEBS Journal 274 (2007) 1970–1982 ª 2007 The Authors Journal compilation ª 2007 FEBS 1971 to 8%, the percentage in cells cultivated for 96 h in serum-free medium rose to 75%. Also, cell adhesion was generally increased and adhesion regions were pre- sent under the center of cell body (Fig. 2B,F, arrows). Replacement of the medium with a fresh, but still serum-free, medium only partially reversed the mor- phological alterations (Fig. 1E) and adhesion pattern (Fig. 2C,G). However, replacement with complete medium containing 10% (v ⁄ v) fetal bovine serum led to fibroblast-like morphology and typical C6 cell body shape (Fig. 1F) and distribution of adhesion spots (Fig. 2D,H). Under these conditions, the percentage of altered cells decreased to  25%. These results suggest that serum deprivation, forcing glioma C6 cells to change their morphology, did not cause permanent alterations, and, even after 96 h of starvation, the cells were still ready to be activated. Serum deprivation did not cause significant loss of cell viability and the majority of cells remained adherent (not shown). Serum withdrawal can induce cell cycle arrest in many cell lines, including astrocytes [26,27]. In the pre- sent study, we used propidium iodide staining and flow ABC DEF GHI Fig. 1. Induction and reversal of a morphological change in serum-starved glioma cells. Cells were seeded at low density on Petri dishes (8 cm diameter) and allowed to recover for 24 h in MEM with 10% fetal bovine serum. (A) Nonstarved cells; (B) cells cultivated for 48 h in serum-free MEM; (C) cells cultivated for 72 h in serum-free MEM; (D) cells cultivated for 96 h in serum-free MEM; (E) cells cultivated for 96 h in serum-free MEM followed by 24 h recovery in fresh MEM without serum; (F) cells cultivated for 96 h in serum-free MEM followed by 24 h recovery in fresh MEM with serum. (G) Percentage changes in morphology during serum deprivation. Control, see A; 48 h, see B; 96 h, see D; 96 h +24 h M, see E; 96 h +24 h M+, see F. Cells were counted in five independent images for every experimental setup. In each image, 100–150 total cells were counted. (H) Single glioma C6 cell cultivated in the presence of serum. (I) Single glioma C6 cell cultiva- ted for 96 h in the absence of serum. Note appearance of cell extrusions indicated by arrows. The micrographs were taken with a Nikon Diaphot inverted microscope, equipped with 40· DIC Nomarski objective lens. A–D, bars ¼ 50 lm; H-I, bars ¼ 15 lm. P2Y 1 and P2Y 12 in serum-deprived glioma C6 P. Krzemin ´ ski et al. 1972 FEBS Journal 274 (2007) 1970–1982 ª 2007 The Authors Journal compilation ª 2007 FEBS cytometry analysis to determine how long serum dep- rivation was needed to cause full cell cycle arrest. Glioma C6 cells growing in medium supplemented with 10% (v ⁄ v) fetal bovine serum were used as a con- trol. Figure 3 shows that  60% could be synchron- ized in G 0 ⁄ G 1 , 25% in G 2 ⁄ M, and 15% in S phase. After 24 and 48 h of serum starvation,  80% and 90% of cells were growth arrested, respectively. However, there were still a number of unsynchronized cells, making it impossible to consider them as one population. However, incubation in serum-free condi- tions for 96 h resulted in 95% of cells arrested in G 0 ⁄ G 1 phases, with less than 5% of the cells in other cell cycle phases. Figure 3 shows that the addition of fresh serum-free medium to cells starved for 96 h and incubation for the next 24 h (Fig. 3, M-) increased the Fig. 3. Cell cycle analysis performed on propidium iodide-stained glioma C6 cells. Each phase (G o ⁄ G i , S and G 2 ⁄ M) is indicated. Control: cells were grown in MEM containing 10% fetal bovine serum. 24, 48, 72, 96 h: for serum deprivation, medium was changed to serum-free MEM and cells were cultivated without medium exchange for up to 96 h. M-: cells were cultivated for 96 h in serum-free MEM followed by 24 h recovery in fresh MEM without serum. M+: cells were cultivated for 96 h in serum-free MEM followed by 24 h recovery in fresh medium with serum. The data represent one typical experiment, similar to three other repetitions. ABCD EFGH Fig. 2. Changes in adhesion pattern during starvation and recovery of glioma cells. (A, E) Nonstarved cells, adhesion regions on the cell edge indicated by arrows (A). Vinculin staining on whole cell surface (green, E). (B, F) Cells cultivated for 96 h in serum-free MEM; arrows point to strongly adherent central region of the cell (B). Vinculin staining concentrated in the central part of the cell (green, F). (C, G) Cells cultiva- ted for 96 h in serum-free MEM followed by 24 h recovery in fresh MEM without serum; arrow points to fragmented adhesion region in the cell center. Vinculin staining on the whole cell surface in most of the cells (green, G); some cells still have vinculin staining similar to starved cells (arrow). (D, H) Cells cultivated for 96 h in serum-free MEM followed by 24 h recovery in fresh MEM with serum; arrows point to recov- ering adhesion regions on the cell edges. Vinculin staining similar to control cells (E). (A–D) The IRM micrographs were taken with a Nikon Diaphot inverted microscope, equipped with a 100· DIC Nomarski objective lens and IRM epifluorescence illumination. Bar ¼ 15 lm. (E–H) Fluorescence microphotographs were taken with a Nikon Diaphot inverted microscope, equipped with a 63· fluo lens. Specimens stained with antibodies against vinculin visualized by secondary Alexa Fluor 488-conjugated secondary antibodies (green) and phalloidin conjugated with TRITC (red) were closed with 4¢,6¢-diamino-2-phenylindole (DAPI)-containing mounting medium (blue). Bar ¼ 30 lm. P. Krzemin ´ ski et al. P2Y 1 and P2Y 12 in serum-deprived glioma C6 FEBS Journal 274 (2007) 1970–1982 ª 2007 The Authors Journal compilation ª 2007 FEBS 1973 number of cells in the S phase. On the other hand, addition of serum-containing medium (Fig. 3, M+) almost completely restored the control pattern of the cell cycle phases, with  70% of cells in G 0 ⁄ G 1 phases and  20% in S phase. In addition, no significant amounts of apoptotic sub-G 1 cells appeared in any of the experiments (not shown). Similar morphological changes concomitant with cell growth arrest and induction of synthesis of glial fibril- lary acidic protein (GFAP) were observed when glioma C6 cells were treated with agents that increase the intracellular concentration of cAMP [28,29]. GFAP is widely expressed in astrocytes and is used as a marker for the induction of differentiation into an astrocyte type II [30,31]. Therefore, in further experiments we measured the expression of this protein in serum- deprived cells. However, long-term serum starvation did not increase the low GFAP expression (Fig. 4) characteristic of transformed glioma C6 cells [31,32]. In the control glioma C6 cells cultivated in medium supplemented with serum, the GFAP concentration was about 1000-fold lower than in astrocytes (Fig. 4A). Similarly, the changes in glioma C6 cell morphology were not caused by differentiation of the cells into the oligodendrocytic lineage. In this study, the low expression of NG2 proteoglycan (a marker of oligodendrocyte progenitors [33]) was examined by im- munocytochemistry. No statistically significant change in staining intensity was found during serum starvation (not shown). Moreover, Messens & Slegers [31] des- cribed changes in glioma C6 cell morphology from bipolar to more round when cells grown in 10% fetal calf serum were cultivated in chemically defined serum- free medium. As, under these conditions, the concen- tration of P2Y 1 receptor mRNA greatly decreased [22], these data and those presented in this study show that the differentiation-like morphological changes are not connected with differentiation of the cells into astro- cytes or oligodendrocytes, but are connected with the lack of serum in the culture medium. Furthermore, these data indicate that growth arrest and GFAP expression proceed via different pathways. We have previously confirmed the presence of P2Y 1 , P2Y 2 and P2Y 12 receptors in glioma C6 cells using PCR on reverse-transcribed total mRNA [20,22]. We have found that, in cells serum-deprived for 48 h, the concentration of P2Y 1 mRNA decreased, whereas P2Y 12 mRNA expression slightly increased. In con- trast, serum starvation was without effect on P2Y 2 mRNA receptor expression [20]. To explore the effects of prolonged serum starvation on ADP-sensitive recep- tor expression, we measured changes in protein con- centration using Western blot analysis. The cells were cultivated for 24, 48, 72 and 96 h in serum-free med- ium in exactly the same conditions as used for the experiments on cell cycle, adhesion and morphology. Figure 5 shows that the concentration of P2Y 1 recep- tor protein gradually decreased. Changes in the con- centration of the P2Y 12 receptor protein were not so evident, although densitometry analysis of the blots from several experiments revealed a slight increase, particularly after 72 and 96 h of serum deprivation. However, given the fact that, in cells growing in serum-free medium for 72 and 96 h, the concentration of P2Y 1 receptor protein was low, under such condi- tions, P2Y 12 receptor expression must strongly pre- dominate (Fig. 5). These results suggest that, during prolonged serum starvation, the P2Y 12 receptor is pri- marily responsible for ADP-evoked signal transduc- tion. The addition of fresh medium for the next 24 h restored low expression of the P2Y 1 receptor (Fig. 5, M-). The increase in P2Y 1 receptor expression was especially noticeable when, after 96 h of serum starva- tion, the medium was replaced with complete medium containing 10% (v ⁄ v) fetal bovine serum (Fig. 5, M+). The serum-free medium also increased the B A Astrocytes Fig. 4. GFAP expression in serum-deprived glioma C6 cells. Total protein isolated from glioma cells was subjected to SDS ⁄ PAGE on 10% polyacrylamide gel and transferred to nitrocellulose mem- branes (A). Filters were probed with the indicated GFAP antiserum. Protein bands were detected with secondary antisera coupled to horseradish peroxidase by enhanced chemiluminescence, and then densitometry analysis was performed (B). Bars represent mean integrated band intensity taken from three independent experi- ments. Control cells cultivated in medium with 10% fetal bovine serum displayed as 100%; 48 h-, cells cultivated in medium with- out fetal bovine serum for 48 h; 96 h-, cells cultivated in medium without fetal bovine serum for 96 h. The results are mean ± SD and are representative of three separate experiments. P2Y 1 and P2Y 12 in serum-deprived glioma C6 P. Krzemin ´ ski et al. 1974 FEBS Journal 274 (2007) 1970–1982 ª 2007 The Authors Journal compilation ª 2007 FEBS concentration of P2Y 1 but to a much lesser extent. No such difference was observed in the case of the P2Y 12 receptor. Multiple bands observed for P2Y receptors are probably the effect of protein glycosylation. We have investigated this phenomenon in glioma C6 cells and have recently shown that preincubation of cell lysates with N-glycosidase F resulted in the reduction of bands detected by P2Y antibodies [34]. We have previously shown that glioma C6 cells growing in medium supplemented with serum exhibit high, constitutive concentrations of active, phosphoryl- ated ERK1 ⁄ 2 and PI3-K [22]. In the present study, we performed experiments with cells incubated with medium devoid of serum for up to 96 h. Figure 6A shows that, in the absence of serum stimulation, the level of p42 and p44 ERK1 ⁄ 2 phosphorylation was greatly reduced, as monitored by Western blot analysis with antibodies to p42 ⁄ p44 ERK. In contrast, pro- longed serum deprivation did not decrease Akt kinase phosphorylation. In fact, 24–72 h of serum starvation resulted in its increase, showing that exhibition of a high basal level of Akt activity is part of an important self-regulatory mechanism. This process was only stopped, and Akt phosphorylation decreased, in cells after very prolonged (96 h) serum starvation (Fig. 6A), suggesting that the stimulatory role of the P2Y 12 Fig. 5. Expression level of P2Y 1 and P2Y 12 receptor protein in nonstarved and serum- starved cells. Total proteins isolated from whole glioma cell lysates were subjected to SDS ⁄ PAGE on 10% polyacrylamide gel and transferred to nitrocellulose membranes. Fil- ters were probed with the indicated P2Y antisera. Protein bands were detected with secondary antisera coupled to horseradish peroxidase by enhanced chemilumines- cence. Equal loading of proteins was ensured by b-actin concentration. Control: cells were growing in MEM containing 10% fetal bovine serum. 24, 48, 72, 96 h: cells were cultivated in MEM without serum for 24, 48, 72 and 96 h, respectively. M-: cells were cultivated for 96 h in serum-free MEM followed by 24 h recovery in fresh MEM without serum. M+: cells were cultivated for 96 h in serum-free MEM followed by 24 h recovery in fresh medium with serum. The blot represents one typical experiment. Plots are integrated densitometry analysis results and are mean ± SD from three repetitions. P. Krzemin ´ ski et al. P2Y 1 and P2Y 12 in serum-deprived glioma C6 FEBS Journal 274 (2007) 1970–1982 ª 2007 The Authors Journal compilation ª 2007 FEBS 1975 receptor is only one of the factors that affect the mul- tiple and complex process of Akt activation. However, replacement of the medium with fresh but still serum- free medium (Fig. 6B, M-), and especially with fresh medium containing 10% (v ⁄ v) fetal bovine serum (Fig. 6B, M+), and subsequent incubation for the next 24 h restored ERK1 ⁄ 2 phosphorylation. In con- trast, fresh serum-free medium restored Akt phos- phorylation (Fig. 6B, M-), whereas fresh medium supplemented with serum had the opposite effect (Fig. 6B, M+). Thus, phosphorylation of Akt (Fig. 6A, p-Akt) increased up to 72 h, in parallel with the decrease in P2Y 1 receptor protein concentration (Fig. 5), and decreased (Fig. 6B, M+) when P2Y 1 receptor protein concentration increased after replace- ment with medium containing serum (Fig. 5). To check whether, in the cells adapted to serum-free medium for 96 h, P2Y 12 receptor is still functionally active and able to respond to agonist affecting ERK1 ⁄ 2 as well as Akt activity, in the next experiments the cells were stimulated with 2MeSADP. Within 15 min, 2MeSADP had induced a rapid,  35-fold increase in the concentration of p44 ⁄ p42 ERK1 ⁄ 2 (Fig. 6C) and 25-fold increase in Akt phosphorylation (Fig. 6D). Within 30 min of the addition of the agonist, phos- phorylation had declined; in the case of p-ERK1 ⁄ 2, concentrations fell to basal, whereas in the case of p- Akt, concentrations declined to half the observed increase and reached a plateau. Figure 6D shows that AR-C69931MX (a P2Y 12 -specific antagonist) almost completely inhibited Akt phosphorylation, whereas ERK1 ⁄ 2 phosphorylation was inhibited by  70%, returning to the level observed after the addition of fresh medium in the absence of agonist (Fig. 6C, inset). Incubation in media containing specific receptor antag- onists and in the absence of agonists shows that AR-C69931MX, the P2Y 12 receptor antagonist, greatly reduces the level of active ERK1 ⁄ 2 phosphorylation, whereas MRS2179, the P2Y 1 receptor antagonist, had a much weaker effect (Fig. 6E). These results again high- light the stimulatory role of P2Y 12 in mitogen-activated protein kinase and Akt kinase signaling pathways. On the other hand, the functional activity of the P2Y 1 receptor in long-term serum-starved cells was recently confirmed by us by their ability to generate calcium signals. We showed that, in glioma C6 cells serum- starved for 96 h, there was a tight correlation between low expression of this receptor and low C AB D E Fig. 6. Changes in constitutive ERK1 ⁄ 2 and Akt phosphorylation during long-term serum starvation (A, B) and the effect of P2Y 12 receptor stimulation on ERK1 ⁄ 2 (C, E) and Akt (D) phosphorylation in cells adapted for 96 h to serum-free medium. (A) p44 ⁄ p42 ERK1 ⁄ 2 and Akt-Ser473 phosphorylation in glioma C6 cells growing in MEM with 10% fetal bovine serum (control) or after incubation in serum-free MEM for 24, 48, 72 and 96 h. (B) Cells cultivated in serum-free MEM for 96 h followed by long-term recovery (24 h) in fresh MEM with 10% fetal bovine serum (M+) or in MEM without serum (M-). (C) p44 ⁄ p42 ERK1 ⁄ 2 phosphorylation in cells cultivated in serum-free MEM for 96 h and then stimulated by addition of 2MeSADP (10 l M) up to 180 min. Inset shows inhibition of 2MeSADP-dependent phosphorylation of ERK1 ⁄ 2by10l M AR- C69931MX, a P2Y 12 receptor antagonist. The cells were preincubat- ed for 2 min with the above receptor antagonists and, still in their presence, stimulated for 15 min with 2MeSADP. (D) Akt-Ser473 phosphorylation in glioma C6 cells cultivated in serum-free MEM for 96 h and then stimulated by addition of 2MeSADP (10 l M)or cultivated with 2MeSADP and 10 l M AR-C69931MX for up to 180 min (E), cells cultivated in serum-free MEM for 96 h followed by long-term recovery (24 h) in fresh serum-free medium with addi- tion of 10 l M AR-C69931MX [the P2Y 12 receptor antagonist (ARC)], 30 l M MRS2179 [the P2Y 1 receptor antagonist (MRS)] or both (ARC + MRS).The cell growing conditions and Western blot analy- sis were as described in Experimental procedures. b-actin was used as a control of equal protein loading. The results are represen- tative of three separate experiments and are mean ± SD. P2Y 1 and P2Y 12 in serum-deprived glioma C6 P. Krzemin ´ ski et al. 1976 FEBS Journal 274 (2007) 1970–1982 ª 2007 The Authors Journal compilation ª 2007 FEBS calcium response [24]. However, to eliminate the possi- bility that ATP secreted by glioma C6 cells, as well as products of its hydrolysis, might specifically desensitize P2Y 1 which could eventually modulate its expression at the transcriptional level, we measured the cons- titutive release of ATP into the extracellular medium. Figure 7 shows that, after careful exchange of the med- ium for one without serum, the ATP concentration increased slightly, reaching the low nanomolar range. It then increased to more than 100 nm at 24 h after the medium exchange. It then slowly decreased and stayed at  10 nm from 48 h until 96 h. As this value is low and similar to that observed in the presence of serum, possible receptor desensitization occurring through its continuous stimulation seems to be doubt- ful. What’s more, the ATP concentration in the med- ium fell over time, reaching control levels after 48 h. A similar low concentration of ATP released into the medium by glioma C6 cells and stable for up 12 h was described by Lazarowski et al. [35]. Discussion This study shows that long-term serum-starved glioma C6 cells is characterized by increased expression of the P2Y 12 receptor and low expression of the P2Y 1 recep- tor. In cells grown in medium containing 10% (v ⁄ v) fetal bovine serum, the expression ratio of the two receptors is reversed. The different expression of P2Y 1 and P2Y 12 recep- tors in serum-starved and nonstarved cells seems to have an important biological significance. Given the shift in receptor expression, there is simultaneous cell growth arrest and changes in the adhesion pattern and the cell morphology. The addition of fetal bovine serum to cells serum-deprived for 96 h that have become morphologically altered causes reversion to the normal fibroblast-like, flat morphology, restores the expression of the P2Y 1 receptor, and halts growth arrest. These data indicate that, during very prolonged serum starvation, glioma C6 cells enter a transient physiological state in which they stay until there is an improvement in the inhospitable environment. The P2Y 1 and P2Y 12 receptors appear to play an important role in this process. We believe that the shift in expres- sion of these receptors underlies a new defense mech- anism of transformed cells, which allows maintenance of self-sufficiency in growth signal and proliferation even in unfavorable conditions. We have previously reported that the P2Y 12 receptor in glioma C6 cells is negatively coupled to adenylate cyclase [18,19] and is predominantly involved in ERK1 ⁄ 2 activation. Furthermore, antagonizing the P2Y 12 receptor with AR-C69931MX counteracted its activation by ADP or 2MeSADP [18,22]. Moreover, it has a stimulatory effect on PI3-K activation, whereas the effect of P2Y 1 on this kinase has been charac- terized as inhibitory [22]. Antagonists of the P2Y 1 receptor (pyridoxal phosphate-6-azophenyl-2,4-disulfo- nic acid and MRS2179) inhibit calcium mobilization [18,19], but stimulate PI3-K activity [22]. Similar opposite effects of P2Y 1 and P2Y 12 receptors on PI3-K activity have been described in human platelets [36]. It is noteworthy that glioma C6 cells are charac- terized by a lack of measurable lipid phosphatase [phosphatase and tensin homolog (PTEN)] [3]. PTEN dephosphorylates the 3¢ position of phosphoinositides and acts as an antagonist of PI3-K, resulting in a reduction of the concentration of this kinase product. As a high concentration of phosphatidylinositol 3,4, 5-trisphosphate is necessary to increase tumor prolifer- ation and motility [3], PTEN phosphatase can be treated as a tumor suppressor [37] and PI3-K as a tumor enhancer. In primary astrocytes and glioma C6 cells, an increase in intracellular cAMP concentration inhibits ERK1 ⁄ 2 activity, blocks cell growth, and promotes differentiation [38–42]. Differentiation into an astro- cyte type II may be induced upon stimulation of b-adrenergic receptors, positively coupled to adenylate cyclase. In this case, glioma C6 cells undergo morpho- logical changes from a bipolar to a round shape, Time (hours) 0 24487296 nM)( noitartnecnoc PTA 1 10 100 1000 Fig. 7. ATP concentration in medium above serum-starved glioma C6 cells. ATP concentration was measured by the luciferin–lucif- erase assay (see Experimental procedures for details). Each time point is the mean ± SD consisting of six measurements repeats from three experiments. Zero time represents the ATP concentra- tion in the culture medium containing serum, before the experi- ment. Broken line, observed increase in ATP concentration after exchange of the medium for a serum-free one. Data are shown in logarithmic scale. P. Krzemin ´ ski et al. P2Y 1 and P2Y 12 in serum-deprived glioma C6 FEBS Journal 274 (2007) 1970–1982 ª 2007 The Authors Journal compilation ª 2007 FEBS 1977 concomitant with induction of GFAP synthesis, the astrocytic marker [28]. Furthermore, Claes et al. [28] indicated that, in glioma C6, the P2Y 12 receptor negat- ively coupled to adenylate cyclase and-stimulated by 2MeSADP abolished b-adrenergic receptor induced differentiation. Antagonists that inhibit PI3-K ⁄ Akt activity also induced C6 glioma differentiation [43]. Thus, serum starvation and the increase in intracellular cAMP concentration produce similar morphological changes, whereas serum withdrawal does not increase GFAP concentration, as shown above. Coyle [44] reported that glioma C6 cells growing in the presence of serum may be characterized by two popu- lations of cells (young oligodendrocyte phenotype and mixed astrocyte–oligodendrocyte phenotype), whereas cells adapted to serum-free conditions exhibit more mature morphology with only one population of cells, the mixed astrocyte–oligodendrocyte phenotype. Thus, serum withdrawal changed the phenotype of the cells but, as we also observed in this study, did not affect their differentiation, as investigated by oligodendrocyte progenitor or astrocyte markers. Concern arises if a low concentration of P2Y 1 can- not be interpreted as the result of receptor desensitiza- tion by ATP released by serum-starved cells themselves. We doubt whether it can, as, first of all, a raised extracellular concentration of ATP should first influence the P2Y 2 receptor, and furthermore its meta- bolite should influence the P2Y 12 receptor as strongly as P2Y 1 . We do know that the concentration of the P2Y 2 receptor does not change during long-term serum starvation [24], and the concentration of the P2Y 12 receptor even increases. What’s more, the observed changes in protein concentration follow parallel chan- ges in mRNA concentration [20,22], so these transcrip- tional changes are probably the result of growth factor deprivation, not receptor desensitization. Nevertheless, we cannot exclude the possibility that the ATP released as the result of the mechanical stress accom- panying medium exchange [35] is responsible for the effects observed after replacement of serum-free med- ium with fresh medium. Taken together, the data presented here indicate that the P2Y 12 receptor acts in favor of cell growth and proliferation because of (a) the reduction in cAMP concentration, (b) the stimulation of ERK1 ⁄ 2, and (c) the stimulation of PI3-K ⁄ Akt signaling cascades. As the P2Y 1 receptor inhibits PI3-K ⁄ Akt signaling, a decrease in its expression has a similar effect. We pos- tulate therefore that the shift in P2Y nucleotide recep- tor expression from P2Y 1 to P2Y 12 observed during prolonged serum starvation, which leads to a high, sta- ble basal level of Akt phosphorylation, is part of the self-sufficiency mechanism of glioma C6 cells, enabling survival in unfavorable metabolic condition and sta- ving off cell death. In conclusion, these data reveal the role of the P2Y 12 receptor in cell survival. Sak & Illes [8] have recently reported that a general property of neuronal malignant cells appears to be the absence of P2Y 1 receptor, which is present in healthy cells. Our results are consistent with this. Our findings suggest that, in malignant cells, the lack of the P2Y 1 receptor inhibit- ing PI3-K ⁄ Akt signaling may play an important role in maintaining intense pathological proliferation and invasiveness. Our results also demonstrate that the change in glioma C6 cell morphology is not part of the differentiation process to astrocytes or oligodendro- cytes. One can speculate that the observed changes may be malignant-like behavior. As, even after 96 h of serum starvation we can still observe functional cross- talk between P2Y 1 and P2Y 12 calcium signaling [24], it would be extremely interesting if similar cross-talk between P2Y 12 and P2Y 2 receptors exists in malignant cells devoid of P2Y 1 , or if P2Y 12 receptor function is limited in those cells to cAMP inhibition. Our data show that even long-term serum starvation can be readily reversed. Prasad et al. [45] suggested that most glioma cells do exhibit mutation, only down- regulation, of the differentiation gene and therefore they are capable of differentiating in the presence of experimental chemical stimuli, e.g. cAMP. However, in contrast with neuroblastoma, glioma cell s, owing to a de- fect in the regulatory genes, are unable to maintain a differentiated phenotype upon removal of the stimulus. Therefore, their differentiation into astrocyte type II induced by cAMP is reversible. According to the authors, such defects may have occurred at a later stage of tumor development [45]. As P2Y 1 inhibits PI3-K, a tumor enhancer, we assume that our findings will provide the rationale for the development of therapeutic strategies specifically targeting this nucleotide receptor. These data may also provide a new approach to the current use of glioma C6 cells as a model for glioma antiangiogenic therapies [46,47]. Experimental procedures Materials MEM, fetal bovine serum, antibiotics, MEM vitamins, MEM nonessential amino acids, and phosphate buffered saline (NaCl ⁄ P i ) were from Sigma Chemical Co. (St Louis, MO, USA). Complete EDTA-free protease inhibitors were from Roche Molecular Biochemicals (Mannheim, P2Y 1 and P2Y 12 in serum-deprived glioma C6 P. Krzemin ´ ski et al. 1978 FEBS Journal 274 (2007) 1970–1982 ª 2007 The Authors Journal compilation ª 2007 FEBS Germany). Propidium iodide was from Molecular Probes (Leiden, the Netherlands). Antibodies recognizing the C-terminus of P2Y 1 and P2Y 12 receptors, Tween 20 and cell-dissociation solution were obtained from Alomone Laboratories (Jerusalem, Israel) and Sigma Chemical Co. Antibodies against Akt and phosphorylated Akt (Ser473 ⁄ Thr308), phosphorylated p44 ⁄ 42 mitogen-activated protein kinase (Erk1 ⁄ 2; Thr202 ⁄ Tyr204) and horseradish peroxidase-conjugated anti-rabbit IgG came from Cell Sig- nalling (New England Biolabs, Hitchin, UK). Monoclonal antibodies against GFAP were from Calbiochem (Meudon, France). Polyclonal antibodies against NG2 proteoglycan, horseradish peroxidase-conjugated anti-mouse IgG, secon- dary antibodies against rat IgG and antibodies against b-actin were from Oncogene Research Products (Darms- tadt, Germany). Nitrocellulose membrane and enhanced chemiluminescence detection system came from Amersham Pharmacia Biotech (Little Chalfont, UK). AR-C69931MX was a gift from AstraZenca (Wilmington, DE, USA). MRS2179 tetra-ammonium salt was from Tocris (Ellisville, MO, USA). 2MeS-ADP and all other reagents were pur- chased from Sigma Chemical Co. Cell culture and morphology Rat glioma C6 cells (passages 40–55) were obtained from American Type Culture Collection and cultured in MEM supplemented with 10% (v ⁄ v) fetal bovine serum and penicillin (50 IUÆmL )1 ), streptomycin (10 lgÆmL )1 ), ampho- tericin B (25 lgÆmL )1 ) and 2 mml-glutamine in a humidified atmosphere of 5% CO 2 at 37 °C. For experi- ments, cells were cultivated in MEM supplemented with 10% (v ⁄ v) fetal bovine serum to reach 80% confluence. In the case of serum-starved cells, the medium was chan- ged to MEM without fetal bovine serum for up to 96 h before the experiment. For morphology changes, the cells were cultured in MEM with 10% (v ⁄ v) fetal bovine serum, or for 48, 72 and 96 h in MEM without fetal bovine serum, and finally allowed to recover for 24 h in MEM with or without fetal bovine serum. For statistical analysis of morphology changes during serum deprivation, altered and control cells were counted from five independ- ent images taken at each time point. The observations were performed with a Nikon Diaphot inverted micro- scope (DIC Nomarski optics equipped with 40 ⁄ 0.55 LWD objective lens). All experiments were performed at least twice. Flow cytometric cell cycle analysis After treatment, floating cells detached by treatment with cell-dissociation solution were pooled, centrifuged for 10 min at 200 g, washed with NaCl ⁄ P i , and fixed in 70% (v ⁄ v) ethanol at )20 °C overnight. Before analysis, cells were centrifuged, washed twice in NaCl ⁄ P i , subsequently suspended in a DNA staining solution (75 lgÆmL )1 propi- dium iodide and 50 lgÆmL )1 RNase A in NaCl ⁄ P i ), and incubated for 30 min at room temperature in the dark. Cells were then analyzed for DNA content with a FAC- Scalibur flow cytometer using cellquest software (Becton Dickinson, Warsaw, Poland). Western blot analysis For the experiment, cells were grown up to 80% confluency on 10 mm dishes in MEM containing 10% (v ⁄ v) fetal bovine serum. At the end of the experiment, cells were washed in NaCl ⁄ P i , detached with cell-dissociation solution (Sigma), centrifuged at 1500 g for 5 min, and then resuspended in lysis buffer [50 mm Tris ⁄ HCl, pH 7.5, 120 mm NaCl, 1% Nonidet P40 (Boehringer, Mannheim, Germany)] containing freshly added proteinase inhibitors (Roche). Proteins were separated by SDS ⁄ PAGE [10% (v ⁄ v) polyacrylamide gels]. They were then transferred to nitrocellulose membranes (Hybond C; Amersham Pharma- cia Biotech, Milan, Italy), blocked for 1 h at room tempera- ture with 5% milk in NaCl ⁄ P i ⁄ Tris-buffered solution (pH 7.6) ⁄ 0.005% Tween 20 (NaCl ⁄ P i -T) and incubated overnight at 4 °C with the following anti-rabbit sera diluted in 1% non-fat milk in NaCl ⁄ P i -T: P2Y 1 , P2Y 12 (both dilu- ted 1 : 2500) (Alomone and Sigma Chemical Co); Akt, phospho-Ser473 Akt, phospho-ERK1 ⁄ 2 (all diluted 1 : 2500) (Cell Signalling); GFAP (Calbiochem). The pri- mary antibody reaction was followed by 2 h incubation with the relevant secondary antibodies against rabbit IgG (1 : 2500) conjugated with horseradish peroxidase, or, in the case of GFAP, secondary antibody against rat IgG (1 : 2500) conjugated with horseradish peroxidase (Onco- gene Research Products). Immunoreactions were analyzed using enhanced chemiluminescence (Amersham Pharmacia Biotech) and membrane exposure to X-ray film (Foton, Warsaw, Poland). Protein molecular mass was determined with a prestained protein ladder (Fermentas Life Sciences, Vilnius, Lithuania). Band intensities were determined by densitometric analysis using ingenius genesnap genetools software (Syngene Bioimaging, Cambridge, UK). Only sta- tistically significant immunoblot band intensity data are reported. Adhesion measurements Distribution of the cell adhesion regions was assessed by the IRM technique at 37 °C. A Nikon Diaphot inverted microscope equipped with planapochromat 60·⁄1.4 NA strainless DIC and fluor 40·⁄1.3 NA oil immersion lenses was used. Interference images were formed by two perpen- dicular polarizers and 50 ⁄ 50 beamsplitter (Chroma, Rock- ingham, VT, USA). Images were acquired with a Retiga 1300 cooled digital camera and aqm advance 6 software (Kinetic Imaging, Rockingham, VT, USA). P. Krzemin ´ ski et al. P2Y 1 and P2Y 12 in serum-deprived glioma C6 FEBS Journal 274 (2007) 1970–1982 ª 2007 The Authors Journal compilation ª 2007 FEBS 1979 [...]... Krzeminski et al P2Y1 and P2Y12 in serum-deprived glioma C6 References Immunocytochemistry analysis of f-actin and vinculin distribution The distribution of vinculin was assessed with a primary antibody against vinculin (Sigma) detected by secondary antibody conjugated with Alexa Fluor 488 (Invitrogen, Leiden, the Netherlands) At the end of the experiment, cells were washed in NaCl ⁄ Pi and fixed in freshly... ecto -nucleotide pyrophosphatase is one of the main enzymes involved in the extracellular metabolism of ATP in rat C6 glioma J Neurochem 72, 826–834 Czajkowski R, Banachewicz W, Ilnytska O, Drobot LB & Baranska J (2004) Differential effects of P2Y1 and P2Y12 nucleotide receptors on ERK1 ⁄ ERK2 and phosphatidylinositol 3-kinase signalling and cell proliferation in serum-deprived and nonstarved glioma C6 cells. .. Cross-talk between the ATP and ADP nucleotide receptor signalling pathways in glioma C6 cells Acta Biochim Pol 49, 877– 889 ´ ´ Supłat D, Krzeminski P, Pomorski P & Baranska J (2007) P2Y1 and P2Y12 receptors cross-talk in calcium signalling: evidence from non-starved and long-term serum-deprived glioma C6 cells Purinergic Signalling in press Hanahan D & Weinberg RA (2000) The hallmarks of cancer Cell 100,... Reactive blue 2 inhibition of cyclic AMP-dependent differentiation of rat C6 glioma cells by purinergic receptor-independent inactivation of phosphatidylinositol 3-kinase Biochem Pharmacol 67, 1489–1498 29 Oey J (1975) Noradrealine induces morphological alterations in nucleated and enucleated rat C6 glioma cells Nature 257, 317–319 30 Dahl D (1981) The vimentin-GFA protein transition in rat neuroglia... Kaczmarek L & Baranska J (2001) Two subtypes of G protein-coupled nucleotide receptors, P2Y (1) and P2Y (2) are involved in calcium signalling in glioma C6 cells Br J Pharmacol 132, 393–402 Baranska J, Czajkowski R & Sabala P (2004) Crosstalks between nucleotide receptor-induced signaling pathways in serum-deprived and non-starved glioma C6 cells Adv Enzyme Regul 44, 219–232 Grobben B, Anciaux K, Roymans... AMP inhibits Akt activity by blocking the membrane localization of PDK1 J Biol Chem 276, 12864–12870 FEBS Journal 274 (2007) 1970–1982 ª 2007 The Authors Journal compilation ª 2007 FEBS 1981 ´ P Krzeminski et al P2Y1 and P2Y12 in serum-deprived glioma C6 40 Kurino M, Fukunaga K, Ushio Y & Miyamoto E (1996) Cyclic AMP inhibits activation of mitogenactivated protein kinase and cell proliferation in response... Slegers H (2004) P2Y12 receptor stimulation inhibits beta-adrenergic receptor-induced differentiation by reversing the cyclic AMP-dependent inhibition of protein kinase B J Neurochem 89, 442–453 44 Coyle DE (1995) Adaptation of C6 glioma cells to serum-free conditions leads to the expression of a mixed 1982 astrocyte-oligodendrocyte phenotype and increased production of neurite-promoting activity J Neurosci... paraformaldehyde in NaCl ⁄ Pi for 30 min at room temperature After the NaCl ⁄ Pi wash, cells were permeabilized in 0.01% (volume) Triton X-100 for 30 s and washed again in NaCl ⁄ Pi Before overnight incubation with primary anti-rabbit IgG against vinculin at 4 °C, cells were blocked in 2.5% fetal bovine serum for 1 h at room temperature After overnight incubation with antibody against vinculin (1 : 1000... cells of early and late passages and advanced passages of glial cells derived from aged mouse cerebral hemispheres Glia 6, 245–257 33 Deng W & Poretz RD (2003) Oligodendroglia in developmental neurotoxicity Neurotoxicology 24, 161–178 34 Krzeminski P, Misiewicz I, Pomorski P, KasprzyckaGuttman T & Baranska J (2007) Mitochondrial localization of P2Y1, P2Y2 and P2Y12 receptors in rat astrocytes and glioma. .. cancer Cell 100, 57–70 Chou RC & Langan TJ (2003) In vitro synchronization of mammalian astrocytic cultures by serum deprivation Brain Res Brain Res Protoc 11, 162–167 P2Y1 and P2Y12 in serum-deprived glioma C6 27 Tu MT, Luo SF, Wang CC, Chien CS, Chiu CT, Lin CC & Yang CM (2000) P2Y (2) receptor-mediated proliferation of C (6) glioma cells via activation of Ras ⁄ Raf ⁄ MEK ⁄ MAPK pathway Br J Pharmacol . Differential effects of P2Y1 and P2Y12 nucleotide receptors on ERK1 ⁄ ERK2 and phos- phatidylinositol 3-kinase signalling and cell proliferation in serum-deprived and nonstarved glioma C6 cells. Br J Pharmacol. characterized the expression and functional properties of the ADP- sensitive P2Y 1 and P2Y 12 nucleotide receptors in glioma C6 cells cultured in medium devoid of serum for up to 96 h. During this long-term. by increased expression of the P2Y 12 receptor and low expression of the P2Y 1 recep- tor. In cells grown in medium containing 10% (v ⁄ v) fetal bovine serum, the expression ratio of the two receptors