Gq Protein-induced Apoptosis Is Mediated by AKT Kinase Inhibition That Leads to Protein Kinase C-induced c-Jun N-terminal Kinase Activation*

Gq protein-coupled receptors (GqPCRs) regulate various cellular processes, including mainly proliferation and differentiation. In a previous study we found that in prostate cancer cells, the GqPCR of gonadotropin-releasing hormone (GnRH) induces apoptosis by reducing the PKC-dependent AKT activity and elevating JNK phosphorylation. Because it was thought that GqPCRs mainly induce activation of AKT, we first undertook to examine how general this phenomenon is. In a screen of 21 cell lines we found that PKC activation results in the reduction of AKT activity, which correlates nicely with JNK activation and in some cases with apoptosis. To understand further the signaling pathways involved in this stimulation, we studied in detail SVOG-4O and αT3-1 cells. We found that prostaglandin F2α and GnRH agonist (GnRH-a) indeed induce significant Gαq- and PKC-dependent apoptosis in these cells. This is mediated by two signaling branches downstream of PKC, which converge at the level of MLK3 upstream of JNK. One branch consists of c-Src activation of the JNK cascade, and the second involves reduction of AKT activity that alleviates its inhibitory effect on MLK3 to allow the flow of the c-Src signal to JNK. At the MAPKK level, we found that the signal is transmitted by MKK7 and not MKK4. Our results present a general mechanism that mediates a GqPCR-induced, death receptor-independent, apoptosis in physiological, as well as cancer-related systems.

G protein-coupled receptors (GPCRs), 4 also termed serpentine receptors, are the largest group of membranal proteins that mediate cellular responses to a wide variety of extracellular agents (1,2). The intracellular transmission of GPCR signals is mediated mainly by heterotrimeric G proteins, which are divided into four groups: G s , G i , G q , and G 12 , according to the downstream effectors of their ␣ subunit. The ␤␥ subunits of the G proteins (3), as well as G protein-independent signaling (4,5), also contribute to the wide functional array of the various receptors, which include proliferation, differentiation, vision, olfaction stress response, and more. Among all G proteins, the four members of the G q family (G q , G 11 , G 14 , and G 15/16 ) function primarily via activation of phospholipase C-␤ (6). This effector then produces the second messengers inositol trisphospate and diacylglycerol, which further elevate intracellular calcium levels as well as protein kinase C (PKC) activity to induce many intracellular signaling and the plethora of G q PCR-induced effects.
As the other members of the GPCR family, the G q PCRs induce a wide array of cellular processes. Studies using G q knock-out in mice demonstrated a role of this protein in platelet activation as well as in the development and functioning of the central nervous system and the heart (7,8). Other G q PCRregulated effects have been identified by other methods, including regulation of proliferation (9), the reproductive system (10), brain functioning (6), and more (11). Aside from these G q PCRinduced effects, these central receptors were also implicated in growth arrest or apoptosis in few pathological systems (11)(12)(13). For instance, cardiac hypertrophy seems to be mediated by auto/paracrine mediators acting through G q PCRs (14,15), and gonadotropin-releasing hormone (GnRH) was shown to induce apoptosis of prostate cancer cells through its G q PCR (16). Surprisingly, overexpression of constitutively active form of G q was found to induce apoptosis of COS7, CHO (17), and HeLa (18) cells as well.
Because G q PCRs are distinct from the classical death receptors and cannot recruit caspases or known apoptotic adaptors, the molecular mechanism by which these receptors induce apoptosis attracted considerable attention. Interestingly, the mechanisms identified seem to vary among the different systems. Thus, whereas the apoptotic effects in COS7 and CHO cells are mediated by PKC (17), AT2-induced myocyte apoptosis was found to be dependent on the elevation of intracellular Ca 2ϩ (14). In addition, it was proposed that G q PCR-mediated apoptosis in myocyte might be dependent on permeability transition, pore formation, and activation of the mitochondrial death pathway due to dysregulation of intracellular Ca 2ϩ levels (19). These effects may also be mediated by additional intracellular signaling mechanisms including JNK, p38, RhoK (12), and novel autophosphorylation of ERK that modulates its activity to induce hypertrophic gene expression (20). However, the full signaling in all of these systems need further clarification.
Another signaling pathway that seems to play a role in G q PCRs action is the one involving the lipid kinase phosphatidylinositol 3-kinase (PI3K) and the protein Ser/Thr kinase AKT (11). The mechanism of activation of this pathway is normally initiated by stimulus-dependent recruitment of PI3K to the plasma membrane, which is mediated either by binding to Try(P) residues in membranal proteins or to a GTP-bound Ras (21). This recruitment allows PI3K to phosphorylate phosphoinositol and further recruit to this location PH domain-containing proteins. One such protein is AKT, which upon shifting, is phosphorylated on Thr 308 by PDK1 and on Ser 473 by mTORC2, to induce its full activation. Activated AKT then phosphorylates a large number of substrates to execute many cellular processes, which include mainly cellular metabolism and survival (22). Interestingly, although this pathway seems to be activated by most G q PCRs (11), some reports showed that activation of certain G q proteins actually reduces AKT activation. These include a GnRH receptor that counteracts IGF-1 in ␣T3-1 cells (23), a m1 muscarinic receptor that counteracts insulin in HeLa cells (24), and constitutively active G q that counteracts EGF in HEK293 cells (25). However, the mechanism of G q PCRs effect on growth factor-induced AKT activation is not fully elucidated.
In a previous study we found that GnRH not only inhibits growth factor-stimulated AKT activation, but can also reduce the high basal phosphorylation of this kinase in prostate cancer cells (16). This reduced phosphorylation and activity alleviate the inhibitory effects of AKT on the proapoptotic JNK cascade and therefore induced cell death. Because it was previously thought that most stimuli elevate AKT activity and therefore lead to survival, it was not clear how general is the inhibitory effect observed in the prostate cancer cells. To investigate these points, we first conducted a screen on the effect of 12-O-tetradecanoylphorbol-13-acetate (TPA), which mimics the G q PCR activation, on 21 distinct cell lines. To our surprise, AKT phosphorylation was reduced in 11 cell lines, and this correlated with elevated JNK phosphorylation and with TPA-induced apoptosis. To elucidate the full signaling machinery that links these effects, we used the two cell lines in which TPA induced the strongest apoptosis, SVOG-4O and ␣T3-1. We found that stimulation of these cells with prostaglandin F2␣ (PGF2␣) or GnRH, respectively, results in a G q -dependent apoptosis. The signaling mechanisms involved in this process can be divided into two pathways downstream of PKC. One acts via c-Src to activate MLK3 and the rest of the JNK cascade, whereas the other acts via inactivation of AKT to alleviate its inhibitory MLK3 phosphorylation. Thus, G q PCRs can serve as important apoptotic receptors, acting via a common physiological and pathological PKC-dependent activation of c-Src and reduction of PI3K/AKT activity to induce JNK-dependent apoptosis. 6 GnRH (GnRH-a), protein A-Sepharose and TPA were obtained from Sigma. GF109203X, U0126, SB203580, PP2, SP600125, okadaic acid, and wortmannin were from Calbiochem. 8-iso-PGF2␣ was from Cayman Chemical. Anti-pERK and pJNK antibodies (Abs), and anti-ERK, JNK, and AKT Abs were from Sigma. Anti-Ser(P) 473 -AKT, rabbit polyclonal anti-Thr(P) 308 -AKT, cleaved caspase3, c-Src, Tyr(P) 416 -Src, Ser(P) 271 /Thr 275 -MKK7, Ser(P) 80 -MKK4, Thr(P) 261 -MKK4, Ser(P) 83 -ASK1, Thr(P) 277 /Ser 281 -MLK3, Ser(P)/Thr AKT substrate, MLK3, and ASK1 Abs were from Cell Signaling Technology. Anti-tubulin, MKK4, and MKK7 Abs were from Santa Cruz Biotechnology. HA-G␣ q and HA-Q209L-G␣ q in pCDNA3 were a gift from Dr. Ulrike Mende from Brown University (Providence, RI).
TUNEL Assay-To analyze apoptosis, subconfluent SVOG-4O, ␣T3-1, or HeLa cells were plated on glass coverslips in 12-well plates under the standard culture conditions as described above. Twenty-four hours after the initial seeding, cells were serum-starved and then treated. At different times after treatment the cells were fixed with paraformaldehyde solution (4% in PBS (pH 7.4) for 1 h at 15-25°C), washed with PBS and then incubated with 0.1% Triton X-100 in 0.1% sodium citrate (2 min, 4°C), washed again with PBS, and incubated with TUNEL mixture containing fluorescein-dUTP and terminal deoxynucleotidyl transferase (Roche Applied Science) for 30 min at 37°C. Preparations were analyzed by fluorescence microscopy. The incidence of apoptosis was determined by counting 200 cells.
Cell Harvesting, SDS-PAGE, and Western Blotting-Cells were grown to subconfluence, serum-starved, and stimulated for various times. After treatment the cells were rinsed twice with ice-cold PBS and either rinsed with ice-cold buffer A and scraped into buffer H and disrupted by sonication or scraped into radioimmuneprecipitation assay buffer. The extracts were centrifuged (15,000 ϫ g for 15 min at 4°C), and the supernatants were boiled (5 min) in sample buffer. The samples were then subjected to 10% SDS-PAGE and Western blotting with the appropriate Abs that were detected using alkaline phosphatase or ECL according to the manufacturer's instructions.
Immunoprecipitation (IP) and Co-immunoprecipitation-Cells were treated and extracted as above. The extracts were incubated (2 h at 4°C) with A/G-conjugated Ab, and for IP the beads were washed twice with radioimmuneprecipitation assay buffer, once with 0.5 M LiCl in 0.1 M Tris, pH 8.0, and once with PBS. For co-immunoprecipitation, the beads were washed three times with buffer A. Both IP and co-immunoprecipitation reactions were boiled in sample buffer, and proteins were then subjected to 10% SDS-PAGE and Western blotting.
SiRNA and Cell Transfection-SiRNAs were from Dharmacon (Lafayette, CO). Transfection of the siRNA was made by using DharmaFECT according to the manufacturer's instructions. Transfection of plasmids into cells was done using PEI (26).

Different Effects of TPA on AKT, JNK, and Apoptosis in Various Cell
Lines-In a previous study we found that GnRH reduces the high basal phosphorylation of AKT in starved prostate cancer cells, and this led to JNK activation and apoptosis (16). Because it is usually assumed that stimuli should activate AKT rather than inactivating it, we undertook first to study how general this phenomenon is. For this purpose we studied the effect of TPA on AKT, JNK, and caspase3 in 21 distinct cell lines. To our surprise, we found that AKT phosphorylation was significantly reduced in 11 of the cell lines ( Fig. 1), whereas in other cells the phosphorylation of AKT was not changed (five cell lines) or elevated (four cell lines). This reduced AKT activity correlated with elevated JNK phosphorylation that was detected in 10 of the 11 cells in which AKT phosphorylation was reduced, but not in any of the other cell lines. Importantly, 5 of the cell lines exhibited TPA-induced caspase3 cleavage that is a marker of apoptosis. Four of them were cells in which AKT phosphorylation was reduced and JNK phosphorylation was elevated. These results, together with our previous finding with DU145 cells (16), suggest a common linkage of PKC, AKT, and JNK to apoptosis in many cells. Thus, unlike previous thoughts, the PKC-induced AKT inactivation that leads to JNK phosphorylation and apoptosis seems to be a widespread signaling process that is worth further study.
PGF2␣ and GnRH-a Induce Apoptosis in SVOG-4O and ␣T3-1 Cells-To characterize further the signaling pathway and elucidate its importance, we chose to study the two cell lines in which TPA induced the strongest apoptosis, namely the SVOG-4O and ␣T3-1 cells. Interestingly, these lines represent cells that may undergo physiological apoptosis, upon G q PCR/ PKC stimulation. Thus, SVOG-4O are leutinized granulosa cells that were reported to undergo luteolytic apoptosis following PGF2␣ exposure (27,28). The other was based on the pituitary-derived ␣T3-1 cells that undergo apoptosis in response to GnRH. Although the exact function of apoptosis in the pituitary is not clear, it is likely to play a role in gonadotrope development, similar to the effect of GnRH on the development of hypothalamic GnRH neurons (29). Therefore, the study of these systems may contribute not only to understanding of PKC-related apoptosis, but also to the physiological importance of G q PCR/PKC-induced apoptotic signaling, which was previously thought to occur only under pathological settings (11)(12)(13).
For this purpose, we stimulated the two cell lines with their physiological stimuli, which, due to the interaction of their receptors with G q , are likely to induce PKC activation in a more physiological way than the TPA used in the screen. Thus, we first treated the SVOG-4O cells with PGF2␣ and found that this treatment resulted in a significant time-and dose-dependent apoptosis, as judged from TUNEL (up to 63 Ϯ 5% after 72 h with 1 M PGF2␣) and caspase3 cleavage (Fig. 2). Similar to the PGF2␣ effect, GnRH-a that was added to the ␣T3-1 cells induced significant apoptosis as well (Fig. 2). Because this apoptosis was also induced by TPA (Fig. 1) and was affected by PKC inhibitors (data not shown), it is likely that the effect was indeed G q PCR/PKC-dependent.
G␣ q Protein Induces Apoptosis in SVOG-4O and ␣T3-1 but Not HeLa Cells-We then undertook to verify further that PGF2␣ and GnRH-a induce their function via G␣ q , which may then induce further signaling through phospholipase C ␤ and PKC (6). For this purpose, we overexpressed WT-G␣ q as well as its constitutively active form Q209L-G␣ q in SVOG-4O, ␣T3-1, and HeLa cells. TUNEL, as well as caspase3 cleavage assays, revealed that both G␣ q constructs induced a significant level of apoptosis (Fig. 3). As with PGF2␣, this proapoptotic stimulation was accompanied by a reduced AKT and an elevated JNK phosphorylation. Similar results were obtained with ␣T3-1 cells, although the effects of the constitutively active isoform were stronger than that of WT-G␣ q . Interestingly, the overexpression of these constructs in HeLa cells had no significant effect on the parameters examined. This result opposes previous findings (18), which might be due to the use of serum starvation in our experiments. Thus, our data suggest that the ability of G␣ q to induce apoptosis is dependent on cellular contexts. Moreover, our results indicate that PGF2␣-and GnRH-a-induced apoptosis is indeed mediated primarily by G␣ q .  Q209L-G␣ q , or empty plasmids. Twenty-four hours after transfection, the cells were serum-starved for 16 h, and apoptosis was detected by TUNEL and DAPI (right panels). Alternatively, the cells were lysed, and the cell extracts were subject to Western blotting with anti-p473AKT, gAKT, pJNK, gJNK, cleaved caspase3, and G q Abs as indicated (left panels).

Signaling Pathways Affected by the Proapoptotic PGF2␣ and
GnRH-a-We then examined the effect of the two stimuli on AKT, JNK as well as p38 and ERK that have been demonstrated to be activated by GPCRs in many cell types (30). As expected, the phosphorylation of three MAPKs examined was increased upon both stimulations. However, only the kinetics of JNK phosphorylation was similar in both cell lines (peaking at 30 min; 10 Ϯ 2.3-fold induction; Fig. 4A), whereas the pattern of p38a and more so of ERK1/2 phosphorylation was stronger in GnRH-a-stimulated ␣T3-1 cells (Fig. 4, B and C). AKT phosphorylation was detected by Abs directed to both phosphorylation sites Thr 308 and Thr 473 . In all cases, AKT demonstrated high basal phosphorylation that was significantly reduced within 5-30 min of stimulation and remained low for at least 90 min of stimulation (Fig. 4D). This result is likely to be exerted by removal of phosphate from nuclear AKT (supplemental Fig.  S1A), and as expected, the reduction in AKT phosphorylation correlated with reduction in its catalytic activity (supplemental Fig. S1B). Inhibition or activation of PKC by its specific inhibitors GF109203X or TPA revealed that the stimulus-dependent dephosphorylation of both Ser 308 and Ser 473 is dependent on PKC (supplemental Fig. S2). On the other hand, c-Src that has previously been shown to act downstream of PKC in ␣T3-1 cells (30, 31) was indeed activated in both cell types (supplemental Fig. S3), but its inhibition affected only the basal and not stimulated reduction in AKT phosphorylation. This indicates that G q transmits signals to AKT via PKC, but not c-Src, although the latter is a mediator of the high basal AKT activity in these cells.
We then used selective inhibitors to identify the signaling components participating in the apoptotic effect in both systems. For this purpose, we pretreated both cell lines with specific inhibitors to JNK (SP600125), MEK (U0126), PI3K (wortmannin), and p38 (SB203580), followed by stimulation. TUNEL assay on the treated cells (Fig. 4E) revealed that in both cell lines the stimulated apoptosis was inhibited by the JNK inhibitor, activated by the PI3K inhibitor, and not affected by the two other inhibitors. Thus, the apoptotic effect observed in granulosa and pituitary cells is mediated by reduced PI3K/AKT activity downstream of PKC and by elevated JNK activity.
JNK Activation by GnRH-a and PGF2␣ Is Mediated by MLK3 Downstream of PKC, c-Src, and AKT-As shown above, in addition to PI3K/AKT inhibition, JNK activation plays a role in PGF2␣/GnRH-a-induced apoptosis. We therefore studied the mechanism of activation of JNK in both ␣T3-1 and SVOG-4O cells. As expected, JNK activation was dependent on PKC, c-Src, and inhibition of AKT as judged by pretreatment of the cells with the appropriate inhibitors (Fig. 5A) and with TPA (Fig. 5B). This indicates that the PKC signal to JNK branches into Src-dependent and AKT inactivation-dependent pathways. We then undertook to find out at what level these two pathways may converge, and we focused on a MAP3K component of the JNK cascade, MLK3, which had previously been reported to undergo inhibitory phosphorylation by AKT (13,32). Therefore, we assessed the effect of PGF2␣ and GnRH-a on the activity of MLK3 by anti-phosphoantibodies, which revealed an increase in MLK3 phosphorylation in both cell systems (Fig. 6A). The activation of MLK3 was clearly downstream of c-Src, PKC, and PI3K/AKT, as c-Src and PKC inhibitors reduced (supplemental Fig. S4), whereas TPA elevated the phosphorylation (Fig. 6B). The PI3K inhibitor wortmannin enhanced and prolonged the PGF2␣-induced activatory phosphorylation ( Fig. 6C and supplemental Fig. S4). A more compelling verification for the central role of MLK3 in the G q -JNK pathway was achieved with specific siRNA of MLK3, which markedly reduced the expression levels of MLK3. This knockdown abolished the stimulated JNK phosphorylation (Fig. 6, D  and E), indicating that MLK3 is the main, or even the only MAP3K, in the delineated pathway. Using Western blotting with anti pAKT-substrate Ab on immunoprecipitated MLK3 we found that the activation of MLK3 is correlated downstream with a reduced AKT phosphorylation (Fig. 6, F and G) down-stream of PKC but not c-Src (supplemental Fig. S4). Therefore, it is likely that the diminished AKT activity upon G q and PKC stimulation leads to a reduced inhibition of MLK3 and thereby allows an activatory phosphorylation to prevail and transmit the signal further to the JNK cascade.
Roles of MKK4 and MKK7 in JNK Activation-We then examined the involvement of the MAPKK level kinases of the JNK cascade, namely MKK4 and MKK7 in JNK activation. Because it was shown previously that AKT inhibits MKK4 activity by phosphorylating it on Ser 80 (33) we studied whether AKT might be involved in JNK activation via MKK4 as well. However, Western blotting with the appropriate Abs showed that the inhibitory MKK4 phosphorylation (Ser(P) 80 ) was in fact increased, whereas the activatory phosphorylation (Ser(P) 261 ) decreased upon stimulations (Fig. 7A), suggesting that MKK4 is not involved in the G q -AKT-JNK pathway. On the other hand, PGF2␣ and GnRH-a stimulation resulted in increased phosphorylation of Ser 271 /Thr 275 of MKK7 (Fig. 7B), indicating that this MAPKK may participate in the examined pathway. To confirm further that JNK activation is mediated by MKK7 and not by MKK4, we knocked down either MKK4 or MKK7 in ␣T3-1 cells using appropriate siRNAs. Indeed, when these cells were stimulated with GnRH-a, there was a clear reduction in JNK activation in the MKK7-but not in the MKK4-depleted cells (Fig. 7C). Taken together, these results indicate that JNK activation is dependent only on MKK7, whereas MKK4 is in fact inhibited upon the stimulations. These results do not support the notion that both MKK4 and MKK7 are simultaneously required for JNK activation (34) and suggest that each of them is sufficient for this purpose. These differences may be explained either by different cell contexts or varying scaffold proteins that allow full activation by MKK7 alone.

DISCUSSION
Although G q proteins mediate mostly proliferation, differentiation, and survival, some of them may participate in the induction of apoptosis as well (11). This was shown in pathological settings such as myocardium hypertrophy (12) or our finding that GnRH receptor can induce apoptosis in cancer cells (13). Some sporadic indications for apoptotic effects were suggested for more physiological systems as well (11), but their significance and mechanism of action have not been studied. Here, we showed that PKC, which is the main effector of G q , has the ability to reduce the activity of the survival protein AKT and elevate the activity of the proapoptotic kinase JNK. We then analyzed two systems in which G q PCRs were implicated in growth arrest and apoptosis. In both cell lines, stimulation of the examined G q PCR induced apoptosis by activating the same signaling pathway that seems similar to the one identified in prostate cancer cells (16). Thus, our results indicate that upon stimulation, G q transmit its apoptotic signal first by inducing activation of PKC that transmits it further by two separate signaling branches. One is the activation of c-Src that can activate the JNK cascade, and the other branch consists of dephosphorylation AKT. Our results indicate that in resting cells, AKT is highly active due to basal c-Src activity. Upon stimulation, c-Src is activated and induces phosphorylation of MLK3 upstream of the apoptotic JNK cascade. However, the apoptotic signal cannot proceed unless the inhibition at the level of MAP3K of the JNK cascade is alleviated. This is induced by reduction of the activity of AKT by a PKC-dependent mechanism, which, together with a phosphatase activity, strongly reduces the inhibitory phosphorylation of the MAP3K component MLK3. This further allows signal flow via MKK7 and JNK to induce apoptosis (see a model in Fig. 8).
One of the main regulatory steps of the apoptotic pathway observed here is the decreased AKT activity. Because AKT is a central mediator of cell survival (35), it is not surprising that its inhibition leads to induction of apoptosis. However, unlike other systems in which the activity of AKT in quiescent cells is low and is increased upon stimulation to induce survival, the common phenotype in the cells examined here is the relatively high basal AKT activity. This high activity seems to be induced mainly by a c-Src family member, as the specific c-Src inhibitor for the indicated times. In parallel, the cells were treated with VOOH for 20 min, which served as a positive control. Activatory phosphorylation of MLK3 was determined by Western blotting using the indicated Abs. D and E, ␣T3-1 cells were transfected with siRNA of MLK3, serum-starved, and stimulated with GnRH-a (10 Ϫ7 M, 30 min). The activatory phosphorylation of JNK (D) and the expression of MLK3 (E) were determined by Western blotting using the indicated Abs. F and G, serum-starved SVOG-4O (F) and ␣T3-1 (G) cells were treated with PGF2␣ (10 M) and GnRH-a (10 Ϫ7 M) for the indicated times or left untreated (0). Cells were then subjected to immunoprecipitation (IP) with anti-MLK3 Ab, and its inhibitory phosphorylation by AKT was detected by Western blot (IB) analysis using anti-Ser(P)/Thr-AKT substrate and MLK3 Abs.
PP2 dramatically inhibits it. Our study further demonstrates that c-Src effects on AKT are limited to quiescent cells, but are not involved in the induced inhibition. This is different from the central role of c-SRC in the activation of JNK (Fig. 5) and also ERK (36,37) downstream of G q PCRs. These differences in c-Src effects before and after stimulation could be a result of an alternative complex formation, as recently reported upon GnRH-a nonapoptotic stimulation of L␤T2 cells (38). Alternatively, it can also be due to interference by interacting downstream processes, such as phosphatases or additional inhibitory phosphorylations. Our findings strengthen our previous unique finding (16) that G q PCR may inhibit the basal activity of AKT. This effect might be due to the relatively high basal activity of AKT in these nonstimulated cells. Therefore, it is possible that the inhibition of AKT activity by G q -dependent stimulations in mitogenes (insulin, insulin-like growth factor-1, or EGF)-stimulated cells (23,24) occurs via a mechanism similar to the one identified here.
One of the mechanisms by which AKT exerts its antiapoptotic activity is by inhibiting the apoptotic JNK cascade (39), which was shown here to be a key player in the G q -induced apoptosis. The mechanism by which AKT inhibits JNK activity has been elucidated in few systems and found to involve mainly  inactivating phosphorylation of MAP3K or MAPKK components of the JNK cascade, including ASK1 (40), MLK3 (32), and MKK4 (33). We tested all three kinases and found that only MLK3 is involved in JNK activation in the cells examined here. In addition, we found that the activity of this kinase is suppressed in resting cells by AKT phosphorylation, and this inhibitory phosphorylation is alleviated by reduced AKT activity upon stimulation. Other mechanisms by which AKT regulates the JNK cascade is by regulating JNK scaffold proteins such as JIP1 and POSH (41,42). The involvement of these processes still needs to be clarified in our systems, but it is likely that the upstream effect on MLK3 is sufficient to explain the effects observed here. Independent of the full mechanistic details it is clear from our results that the high basal AKT activity maintains the low activity of JNK in resting cells, and this inhibition is released upon stimulation.
Due to the possible effect of AKT on MKK4 (33), we analyzed the effect of the latter in our systems. To our surprise we found not only that MKK4 was not activated following G q -stimulated AKT inactivation, but rather it was inhibited upon stimulation. Consistently, MKK4 was not involved in JNK activation, which seems to be fully mediated by G q stimulation of MKK7 in our systems. This is unlike the suggested synergistic and simultaneous activation of JNKs by MKK4 and MKK7 (43), which could be cell type-dependent (44). Given that MKK4 was active in quiescent cells and that G q stimulation induced its inactivation suggest that MKK4 affects a subset of JNK molecules, whereas the subset responsible for apoptosis is regulated by MKK7. In addition, the elevated Ser 80 phosphorylation, which occurred despite the lack of AKT activity, indicates that other kinases that participate in this phosphorylation should be identified.
In the present study, we evaluated the structural luteolysis effect of PGF2␣ in the human luteinized granulosa cell line SVOG-4O and found that PGF2␣ induced apoptosis of these cells. This result is consistent with previous studies, which also demonstrated that treatment of cultured human luteinized granulosa cells with PGF2␣ as well as other cytokines induced an apoptotic effect in such cells (27). Similar effects were detected in granulosa cells of corpus luteum in vivo (45). Revealing the new signaling pathways may provide new insight into the above physiological processes. Thus, in the human corpus luteum this pathway may help both in the diagnosis and treatment of various reproductive abnormalities, using specific signal transduction modulators. We also found here that GnRH-a induces apoptosis in the pituitary gonadotrope ␣T3-1 cells. The pathway is likely to be important for pituitary development and as a consequence to the treatment of prostate cancer by GnRH-a (13). These results elucidated a physiological mechanism of apoptosis, which involves G q -mediated and PKC-dependent AKT inactivation, leading to reversal of MLK3 inhibition, and thereby promotes JNK activation and apoptosis.
In summary, we show here that G q PCRs can act as apoptosismediating receptors that act independently from the well known death receptors in both physiological and pathological systems. The mechanism that allows G q PCR-induced apoptosis is general and includes a two-branched pathway downstream of PKC. One branch consists of c-Src activation that transmits the signal to the JNK cascade. However, the signal via this branch cannot proceed without the alleviation of an inhibitory phosphorylation on the MLK3, which is mediated by the second branch that consists on PKC-dependent inactivation of AKT and a phosphatase activity. From the MLK3 the signal is transmitted by MKK7, not MKK4, to JNK, which then induces apoptosis.