Acute activation of Erk1/Erk2 and protein kinase B/akt proceed by independent pathways in multiple cell types Doris Chiu, Kewei Ma, Alexander Scott and Vincent Duronio Department of Medicine, University of British Columbia and Vancouver Coastal Health Research Institute, Jack Bell Research Centre, Vancouver, Canada A complex network of signaling pathways is activated in cells in response to various cytokines and growth factors. Understanding the interactions that occur in these networks is complicated by the fact that, in var- ious cell types, the same type of stimulation may trig- ger different sets of signaling cascades. Thus, analysis of the same signaling pathways operating in different cell types can lead to a variety of conclusions. In our laboratory, we are interested in the relationship between activation of two major signaling pathways: one initiated by p21ras proteins leading to activation of the p44 Erk1 and p42 Erk2 members of the MAPK family [1,2], and the other initiated by the phosphati- dylinositol 3-kinase (PtdIns3K) group of enzymes that Keywords cross-talk; cytokine; inhibitors; kinase; phosphorylation Correspondence V. Duronio, Jack Bell Research Centre, Rm. 255, 2660 Oak St, Vancouver, British Columbia, V6H 3Z6, Canada Fax: +1 604 875 4497 Tel: +1 604 875 4707 E-mail: vduronio@interchange.ubc.ca (Received 17 June 2005, accepted 7 July 2005) doi:10.1111/j.1742-4658.2005.04850.x We used two inhibitors of the signaling enzyme phosphatidylinositol 3-kin- ase (PtdIns3K), wortmannin and LY294002, to evaluate the potential involvement of PtdIns3K in the activation of the MAP kinases (MAPK), Erk1 and Erk2. In dose–response studies carried out on six different cell lines and a primary cell culture, we analyzed the ability of the inhibitors to block phosphorylation of protein kinase B⁄ akt (PKB ⁄ akt) at Ser473 as a measure of PtdIns3K activity, or the phosphorylation of Erk1 ⁄ 2 at activa- ting Thr ⁄ Tyr sites as a measure of the extent of activation of MAPK ⁄ Erk kinase (MEK ⁄ Erk). In three different hemopoietic cell lines stimulated with cytokines, and in HEK293 cells, stimulated with serum, either wortmannin or LY294002, but never both, could partially block phosphorylation of Erks. The same observations were made in a B-cell line and in primary fibroblasts. In only one cell type, the A20 B cells, was there a closer corre- lation between the PtdIns3K inhibition by both inhibitors, and their corres- ponding effects on Erk phosphorylation. However, this stands out as an exception that gives clues to the mechanism by which cross-talk might occur. In all other cells, acute activation of the pathway leading to Erk phosphorylation could proceed independently of PtdIns3K activation. In a biological assay comparing these two pathways, the ability of LY294002 and the MEK inhibitor, U0126, to induce apoptosis were tested. Whereas LY294002 caused death of cytokine-dependent hemopoietic cells, U0126 had little effect, but both inhibitors together had a synergistic effect. The data show that these two pathways are regulating very different down- stream events involved in cell survival. Abbreviations BCR, B-cell antigen receptor; DMEM, Dulbecco’s modified Eagle’s medium; FBS, fetal bovine serum; GM-CSF, granulocyte–macrophage colony stimulating factor; IGF, insulin-like growth factor; IL, interleukin; MAPK, mitogen-activated protein kinase; MEK, MAPK ⁄ Erk kinase; PLC, phospholipase C; PMA, phorbol 12-myristate 13-acetate; PtdIns, phosphatidylinositol; PtdIns3K, phosphatidylinositol 3-kinase; PKB, protein kinase B ⁄ akt; PKC, protein kinase C. 4372 FEBS Journal 272 (2005) 4372–4384 ª 2005 FEBS lead to activation of downstream kinases including protein kinase B (PKB), also referred to as akt [3–5]. The possible relationship between PtdIns3K and Erk activation may differ among cell types, as alluded to above, based on numerous studies that seemed to have reached opposing conclusions. Some studies have sug- gested that PtdIns3K activation is required for activa- tion of Erk [6–11], whereas other studies have shown that PtdIns3K activation is independent of Erk activa- tion [12–14]. However, because a variety of techniques has been used to determine the connection between these pathways, the conclusions may or may not be correct in all cases, as discussed below. Understanding whether two pathways function independently may be important for potential attempts to target signaling events for therapeutic reasons. Clearly, if there is con- siderable cross-talk between the two pathways, drugs acting on one of the pathways may also alter the other. In other words, the effects of a drug acting at an early point in the PtdIns3K ⁄ PKB pathway may be very different depending upon whether the Erk signa- ling pathway is also activated by PtdIns3K-dependent events. Our study was designed to further address this issue of PtdIns3K activation and its relationship to Erk activation in a broad range of cell types. Activation of PKB ⁄ akt has been shown to depend upon the production of 3-phosphorylated inositol phospholipids formed as a result of PtdIns3K activity [15,16]. The phosphorylation of PKB at Thr308 by an intermediate PtdIns-dependent kinase, termed PDK-1, is a well-characterized event. PKB phosphorylation at Ser473 is also important for full activation, with the mechanism of this activation being clarified with the publication of the structure of PKB [17], although more than one kinase has been proposed to phos- phorylate this site [18,19]. There is no good evidence that components of the ras ⁄ Erk pathway have effects on the pathway leading to activation of PKB, except for the demonstration that p21ras, via its activation loop, can bind to and activate the p110 catalytic sub- unit of PtdIns3K [20]. The total contribution of p21ras to PtdIns3K activation is difficult to determine, but may primarily be seen with oncogenic forms of ras. However, it is clear that activation of PtdIns3K has been shown to occur in the absence of p21ras or Erk activation [21]. It has been suggested that the PtdIns3K pathway can lead to Erk activation, although in earlier studies, much of the evidence for the connection was based on the effects of wortmannin as a PtdIns3K inhibitor [6,22], which is now known to be insufficient proof of the role of PtdIns3K due to the effects of this inhibitor on multiple targets. Some studies have drawn conclusions based on use of dominant negative versions of p85 [10,23], but one must also determine whether nonspecific targets are being blocked due to binding of the overexpressed p85 SH2 domains to additional targets. Another potential cross-talk between the PtdIns3K and ras-Erk pathways has been shown to result from the direct phosphoryla- tion of raf by PKB, which can contribute to an inhibi- tion of raf and thus inhibition of Erk activation [24,25]. Intriguingly, these studies have shown that this inhibition is observed in differentiated myotubes, but not in the undifferentiated myoblast cells, implicating other regulatory events controlling the phosphoryla- tion. Perhaps the most convincing study to address the issue of cross-talk between PtdIns3K and Erk path- ways is that of Jacob et al. [26], which showed in sev- eral ways that in both primary B cells and a B-cell line, full activation of Erks was dependent upon the activation of PtdIns3K. Several years ago, our group showed, by use of two different PtdIns3K inhibitors, LY294002 and wortmannin, that LY294002 had no effect on Erk activation in response to cytokines when PtdIns3K was completely blocked [13]. The conclusions of that study relied on the fact that in different situations, one of the two inhibitors could inhibit Erks, whereas the other could not, at inhibitor concentrations that completely blocked PtdIns3K activity. Although those results seemed quite conclusive, we felt that the question had to be addressed again, because a subsequent study, using a different hemopoietic cell line, suggested that activation of PtdIns3K may impinge on different isoforms of raf and thus affect Erk activation [10]. These results were based on the use of wortmannin (but not LY294002) as well as dominant negative forms of PtdIns3K. The latter study used a different cytokine-dependent hemopoietic cell line and suggested that the differences with the study of Scheid & Duronio [13] were due to differ- ences in cell type. Therefore, in this study we addressed the question of how the PtdIns3K inhibi- tors affect activation of Erks in a number of differ- ent cell types, hemopoietic and nonhemopoietic, and including the FDC-P1 cells used by Sutor et al. [10]. An additional part of this study served to reinforce the distinction between PtdIns3K- and Erk-regulated events to verify that the PtdIns3K pathway was indeed the more dominant in determining cytokine- dependent survival of cells. Thus, we compared the apoptosis of the hemopoietic cells in the presence of PtdIns3K and MEK inhibitors to show that inhibi- tion of MEK could only have a significant contribu- tion to the onset of apoptosis when PtdIns3K activity was also blocked. D. Chiu et al. PtdIns3K activity is not required for Erk activation FEBS Journal 272 (2005) 4372–4384 ª 2005 FEBS 4373 Results and Discussion Differential effect of wortmannin and LY294002 on Erk activation in FDC-P1 cells In previous studies, we showed that cytokine-activated Erks were differentially affected by the two inhibitors of PtdIns3K, LY294002 and wortmannin, in MC ⁄ 9 mast cells [13], which led to our conclusion that the activation of PtdIns3K could be dissociated from acti- vation of the pathway leading to Erk activation. We decided to re-address this issue in multiple cell types, beginning with FDC-P1 cells, to determine whether our initial results were limited to a particular cell line. Cytokine-starved FDC-P1 cells were treated with PtdIns3K inhibitors, LY294002 and wortmannin, or the MEK inhibitor, U0126 at varying concentrations for 10 min. The cells were then stimulated with either granulocyte–macrophage colony stimulating factor (GM-CSF; 10 ngÆmL )1 ) or synthetic interleukin (IL)-3 (10 lgÆmL )1 ) for 5 min. The cells were then lysed and proteins were isolated for immunoblotting. From the immunoblots (Figs 1 and 2) we observed that in both GM-CSF and IL-3 treated cells the phosphorylation of Erks was not reduced by LY294002 even at concentra- tions higher than those that completely blocked phos- phorylation of PKB; however, wortmannin did reduce the level of Erk phosphorylation. Thus, these results are entirely consistent with those that we had obtained previously in MC ⁄ 9 cells [13]. The two bands indicated as phosphorylated Erk1 and Erk2 in FDC-P1 cell extracts (Figs 1 and 2) cor- responded to bands detected with anti-Erk serum (data not shown). The nature of the lower two bands, which were not detected with the anti-Erk serum, is not known, but these have only been detected with anti- (phospho-Erk) serum in FDC-P1 cells. Although we cannot be certain at this point, they may be degrada- tion products of phosphorylated Erk1 and Erk2 that are not detected with the anti-Erk serum we used, because their relative intensity correlated with that of the Erk bands in the same samples. In addition, detec- tion of both pairs of bands with the anti-(phospho- Erk) serum was eliminated when cells were treated with the MEK inhibitor (Fig. 2). The effect of wortmannin in reducing Erk phos- phorylation in FDC-P1 cells was also seen in a previ- ous study [10], but the effects of LY294002 were never compared. Both LY294002 and wortmannin blocked phosphorylation of PKB completely at Ser473 as expected, because it may be closely correlated to PtdIns3K inhibition. In FDC-P1 cells, complete inhibi- tion by the PtdIns3K inhibitors was seen at 10 lm LY294002, but the effective concentration of these Fig. 1. FDC-P1 cells pretreated with LY294002 and wortmannin fol- lowed by stimulation with GM-CSF. FDC-P1 cells were starved of cytokine overnight. Cells were pretreated with PtdIns3K inhibitors, LY294002 and wortmannin at the indicated concentrations for 10 min. Ccells were then stimulated with GM-CSF (10 ngÆmL )1 ) for 5 min. Cell lysates were fractionated using 9% SDS ⁄ PAGE and immunoblotted for anti-(phospho-PKB) (P-S473-PKB) or anti-(phos- pho-ERK1 ⁄ 2) (P-ERK1, P-ERK2). The levels of p85 were used for normalization. Fig. 2. FDC-P1 cells pretreated with LY294002 and wortmannin fol- lowed by stimulation with IL-3. FDC-P1 cells were starved of cyto- kine overnight. Cells were pretreated with PtdIns3K inhibitors, LY294002 and wortmannin. Also MEK inhibitor, U0126 was used at the indicated concentrations for 10 min. Cells were then stimulated with IL-3 (10 lgÆmL )1 ) for 5 min. Cell lysates were fractionated using 9% SDS ⁄ PAGE and immunoblotted as for Fig. 1. PtdIns3K activity is not required for Erk activation D. Chiu et al. 4374 FEBS Journal 272 (2005) 4372–4384 ª 2005 FEBS inhibitors may vary among different cell types, as shown below. Because we have found that the dose– response to either LY294002 or wortmannin can vary among different cell types, it is possible that nonselec- tive effects of the inhibitors can be avoided by a more careful use of dose–response studies to verify the con- centration at which PtdIns3K is inhibited. This is also relevant with respect to time of preincubation, because longer preincubation times would likely require lower concentration of inhibitors to achieve equivalent intra- cellular concentrations. The MEK inhibitor, U0126 completely inhibited the IL-3-induced activation of Erks without affecting PKB phosphorylation, as expec- ted (Fig. 2). In virtually all aspects, the data obtained with FDC-P1 confirmed our observations made in the murine mast cell line, MC ⁄ 9 [13]. Both wortmannin and LY294002 do not affect the activation of Erks in TF-1 cells Next we investigated the relationship of PtdIns3K acti- vation and Erk activation in a human hemopoietic cell line, TF-1 which was derived from an erythroleukemic patient [27]. The GM-CSF-starved TF-1 cells were treated with PtdIns3K inhibitors, LY294002 and wort- mannin, and the MEK inhibitor, U0126, at various concentrations for 10 min. The cells were then stimula- ted with GM-CSF for 5 min. The immunoblots of cell lysates (Fig. 3A) showed that neither LY294002 nor wortmannin reduced the phosphorylation of Erks at concentrations at which PKB phosphorylation was completely inhibited. U0126 completely blocked the activation of Erks, without affecting PKB phosphory- lation. Surprisingly, these results are somewhat differ- ent from those seen previously in the FDC-P1 and MC ⁄ 9 cells. In the TF-1 cells, neither PtdIns3K inhib- itor had effects on the Erks. It should be noted that in Fig. 3A the source of GM-CSF was the conditioned medium that was used to grow these cells. However, a more robust phosphorylation response was observed when recombinant GM-CSF was used to stimulate the cells (Fig. 3B). When a dose–response was performed with LY294002, it was apparent that 10 lm LY294002 was not sufficient to block PKB phosphorylation in the latter case. However, again, no effect was seen on Erk phosphorylation even when 50 lm LY294002 was used. In other experiments, use of a range of GM-CSF doses also showed that the addition of LY294002 had no effect on Erk phosphorylation (data not shown). Thus, these results suggest that the lack of effect of LY294002 on Erk is independent of the concentration of cytokine used for stimulation. This result is not entirely consistent with a study suggesting that differ- ential effects may be observed with the PtdIns3K inhibitor wortmannin, depending on the concentration of activator used [28]. Furthermore, it has been shown [29] that PtdIns3K lipid products may play a permis- sive role in activation of Erks in response to low con- centrations of EGF, but this does not appear to be the case with cytokine activation. Effect of PtdIns3K inhibitors in BAF-3 cells Another commonly used hemopoietic cell line, BAF-3, was also tested in similar assays. This is a pro-B-cell A B Fig. 3. TF-1 cells pretreated with various inhibitors followed by sti- mulation with GM-CSF. TF-1 cells were starved of cytokine over- night. (A) Cells were pretreated with LY294002 or MEK inhibitor, U0126, at the indicated concentrations for 10 min. Cells were then stimulated with GM-CSF (10% CGM1) for 5 min. Cell lysates were fractionated using 9% SDS ⁄ PAGE and immunoblotted for anti- (phospho-Ser473-PKB) or anti-(phospho-ERK1 ⁄ 2). The levels of p85 were used for normalization. (B) Cells were treated with a range of concentrations of LY294002, followed by stimulation with 10 ngÆmL )1 recombinant human GM-CSF. A control in which cells were incubated with solvent (dimethylsulfoxide) was included. D. Chiu et al. PtdIns3K activity is not required for Erk activation FEBS Journal 272 (2005) 4372–4384 ª 2005 FEBS 4375 line [30], which is dependent upon IL-3 for growth and survival, and is distinct from the usual myeloid cells used to study IL-3 responses. This cell line is also un- usual in that it is easily transfectable, unlike most other hemopoietic cell types, and thus it may be quite different from other cytokine-dependent hemopoietic cells. In our studies, we used these cells to compare the effects of LY294002 and wortmannin on PKB and Erk phosphorylation, and the results obtained were similar to those seen in TF-1 cells, because neither inhibitor, at the low concentrations used to completely block PKB phosphorylation, had any effect on Erk phos- phorylation (Fig. 4). Again, these results allow us to conclude that there is not a role for the PtdIns3K pathway in Erk activation. It is interesting to note that other studies have shown that a dominant negative form of the p85 subunit of PtdIns3K was able to block Erk activation in BAF-3 cells [23], which we argue may be due to a nonselective effect of expressing a p85 mutant whose SH2 domains could conceivably bind to other targets upstream of the MEK ⁄ erk pathway. In contrast to the results shown in Fig. 4, the study by Craddock et al. [20] paradoxically showed inhibitory effects of LY294002 on Erk phosphorylation, a dis- crepancy that may be attributed to differences in dose and time of treatment. Those authors utilized 10 lm LY294002 as the lowest concentration, with a preincu- bation time of 30 min. We show here that concentra- tions of LY294002 up to 10 lm, preincubated with cells for 10 min, had no effect on Erk phosphorylation, but almost completely blocked PKB phosphorylation. When we used higher concentrations (25 or 50 lm)of LY294002 under these conditions, inhibition of Erk phosphorylation was observed (data not shown). Most importantly, at concentrations of wortmannin where PtdIns3K was completely inhibited, we show that there was no change in phosphorylation of the Erks. Thus, it is important to re-emphasize that it is valuable to use both of the PtdIns3K inhibitors and compare their effects when trying to determine downstream events that may be regulated by PtdIns3K. Potential cross-talk in B cells The activation of B cells via the B-cell antigen receptor has been a widely used model system for signal trans- duction studies [31,32]. In a recent study, [26] the role of the PtdIns3K pathway in regulation of Erk activity was investigated using A20 as well as primary B cells. Thus, we tested murine A20 cells as well as another B-cell line, the human BJAB cells, which were both derived from B lymphomas. Consistent with previous results [26] we observed that inhibition of PtdIns3K activity with LY294002 in A20 cells caused a parallel inhibition of Erk phosphorylation (Fig. 5). Most nota- bly, at 5 lm LY294002, there was a decrease in PKB phosphorylation as well as a decrease in Erk phos- phorylation. Higher concentrations of LY294002 could completely inhibit PKB phosphorylation, and caused greater inhibition of ERK phosphorylation. However, in the human BJAB cell line, there was not as great an effect of LY294002 on Erk phosphorylation, because 0.5–2.0 lm LY294002 caused noticeable inhibition of PKB phosphorylation, whereas ERK phosphorylation was unaffected. There was some inhibition of ERK phosphorylation in BJAB cells in response to higher concentrations of LY294002. Parallel experiments car- ried out using wortmannin showed that this inhibitor affected both PKB and ERK phosphorylation in a similar manner to LY294002 in both of these cell lines (data not shown). Thus, in the two B-cell lines, there was clearly some involvement of PtdIns3K activation in activation of the Erks, even if it is much more obvi- ous in the A20 cells than BJAB cells. One can assume that the reason for the cross-talk between the two pathways in B cells, which is different from growth factor-induced activation, may be attri- buted to the involvement of Btk downstream of B-cell antigen receptor (BCR) activation [33]. This tyrosine kinase is known to be dependent upon PtdIns3K-gen- erated PtdIns lipids, because it has a PH domain that is necessary for its activation [34]. Furthermore, Btk is Fig. 4. BAF-3 cells pretreated with LY294002 and wortmannin fol- lowed by stimulation with IL-3. BAF-3 cells were starved of cyto- kine overnight. Cells were pretreated with PtdIns3K inhibitors, LY294002 and wortmannin or MEK inhibitor, U0126 at the indicated concentrations for 10 min. Cells were then stimulated with IL-3 (10 lgÆmL )1 ) for 5 min. Cell lysates were fractionated using 9% SDS ⁄ PAGE and immunoblotted for anti-(phospho-Ser473-PKB), anti- (phospho-ERK1 ⁄ 2). Ponceau S staining was used to verify equal loading of total protein between lanes (not shown). PtdIns3K activity is not required for Erk activation D. Chiu et al. 4376 FEBS Journal 272 (2005) 4372–4384 ª 2005 FEBS responsible for phosphorylation of phospholipase Cc (PLCc) and the subsequent generation of diacylglycerol and Ca 2+ release [35], which result in protein kinase C (PKC) activation. While other tyrosine kinases may also be involved in ras activation downstream of the BCR, it is clear that activation of PLCc and PKC are also intimately involved in the pathway leading to Erk activation in B cells [36]. Effects of PtdIns3K inhibitors in HEK293 cells We also wanted to examine nonhemopoietic cells to observe the relationship between PtdIns3K and Erks. We decided to use the commonly used human epi- thelial kidney cell line, HEK293. HEK293 cells were serum-starved, and treated with inhibitors at various concentrations for 10 min. The cells were then stimula- ted by the re-addition of 10% fetal bovine serum (FBS) for 10 min. The cells were lysed and proteins were isolated for immunoblotting. HEK293 cells dif- fered from the previous cell lines used as seen by the presence of low levels of endogenous phosphorylated Erks in the unstimulated lane (Fig. 6). However, no phosphorylated PKB was observed in the serum- starved cells. Concentrations of LY294002 from 10 to 50 lm, which blocked the phosphorylation of PKB, partially blocked the phosphorylation of Erks. When the effects of wortmannin were tested, concentrations of wort- mannin much higher than that needed to block PtdIns3K, did not seem to affect Erk phosphorylation. U0126, as seen in the previous three cell lines, com- pletely blocked the activation of Erks, without any effect on PKB. These results suggest either that LY294002 may have some nonspecific effects in this particular cell line, or it may be that LY294002 selec- tively affects a component of the pathway leading to Erk activation in response to serum stimulation. Inter- estingly, we showed previously that phorbol ester sti- mulated Erk activation in MC ⁄ 9 cells showed a similar response, because it could be partially inhibited by LY294002, whereas wortmannin had no effect [13]. In any event, the fact that one of the two PtdIns3K inhi- bitors blocked PKB phosphorylation without any A B Fig. 5. B-cell lines stimulated pretreated with LY294002 and stimulated via the B-cell antigen receptor. The two B-cell lines, A20 and BJAB, were stimulated via the B-cell receptor as described in Experimental proce- dures, following preincubation with various concentrations of LY294002. Samples were immunoblotted for anti-(phospho-Erks) or anti-(phospho-Ser473-PKB). D. Chiu et al. PtdIns3K activity is not required for Erk activation FEBS Journal 272 (2005) 4372–4384 ª 2005 FEBS 4377 effect on Erk phosphorylation supports the conclusion that the two pathways are independently controlled. Effects of PtdIns3K inhibitors in tendon fibroblast cells We also examined the effects of LY294002 and UO126 on Erk and PKB phosphorylation in primary tendon fibroblasts cells. The cells were extracted from porcine Achilles tendon (dissected free of peritendinous tissues) by collagenase digestion and cultured up to the fifth passage in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% FBS. Prior to the experi- ments, the cells were trypsinized and replated at 500 000 cells ⁄ 60 mm plate. After adhering overnight, the cells were stimulated with various doses of insulin- like growth factor (IGF)-I with or without 10 min pre- treatment with graded doses of LY294002 or UO126. After stimulation with IGF-I for 5 min, cells were scraped and pelleted, then lysed and immunoblotted as described above. There was considerable phosphory- lated PKB and Erks in unstimulated cells, mainly because no serum starvation was done with these cells (Fig. 7). Most importantly, at 25 lm LY294002, PKB phosphorylation was completely blocked, but no effect was observed on Erk phosphorylation. If anything, there was increased phosphorylation of Erks following treatment with LY294002, suggesting that the PtdIns3K pathway may have an inhibitory effect on Erk activation. We have not pursued this question suf- ficiently to determine its significance. However, cross- talk between PKB and raf has been suggested to be inhibitory towards Erk activation in specific cases [37,38]. We found that 10 lm UO126 was sufficient to block Erk phosphorylation, but the highest dose of UO126 tested (50 lm) caused no reduction in PKB phosphorylation. Conversely, 10 lm LY294992 only partially reduced PKB phosphorylation, whereas the higher doses eliminated it, without causing any reduc- tion in the levels of Erk phosphorylation. In support of our findings, another study showed that IGF-I, insulin and platelet-derived growth factor stimulation of Erks was not affected by a PtdIns3K inhibitor [39]. Analysis of inhibitor effects on induction of apoptosis In numerous studies [40–44], including our own 10 years ago using MC ⁄ 9 cells [45], the importance of the PtdIns3K pathway in blocking apoptosis has been highlighted. Furthermore, we showed that blocking the Erk pathway with a MEK inhibitor, PD98059, did not cause apoptosis over the same time frame as that seen with PtdIns3K inhibitors (VD, unpublished observa- tions). We wished to address this issue more carefully in cytokine-dependent cells using the MEK inhibitor, U0126, which is a more potent inhibitor of Erk activa- tion [46]. As can be seen from Fig. 8, which summar- izes flow cytometry analysis of propidium iodide staining for subdiploid DNA, the use of the PtdIns3K inhibitor LY294002, caused apoptosis of all three cell Fig. 7. Primary porcine tendon fibroblasts pretreated with LY294002 followed by stimulation with IGF-I. Porcine tendon fibro- blasts were preincubated with or without LY294002 at the indica- ted concenrations for 10 min followed by 5 min stimulation with IGF-I. Samples were fractionated on SDS ⁄ PAGE and immunoblot- ted for phospho-Erk or phospho-Ser473 PKB. Fig. 6. HEK293 cells pretreated with LY294002 and wortmannin followed by stimulation with FBS. HEK293 cells were starved of FBS overnight. Cells were pretreated with PtdIns3K inhibitors, LY294002 and wortmannin or MEK inhibitor, U0126 at the indicated concentrations for 10 min. Cells were then stimulated with 10% FBS for 10 min. Cell lysates were fractionated using 9% SDS ⁄ PAGE and immunoblotted for anti-(phospho-PKB), anti-(phos- pho-ERK1 ⁄ 2) and the levels of p85 were used for normalization. PtdIns3K activity is not required for Erk activation D. Chiu et al. 4378 FEBS Journal 272 (2005) 4372–4384 ª 2005 FEBS types, MC ⁄ 9, FDC-P1 and BAF-3, whereas U0126 had almost no effect over a 16 h period. Interestingly, the addition of both inhibitors together had a greater effect on apoptosis, which suggests that in the absence of PtdIns3K activity, Erk activation may provide some survival effect. This may not be surprising because the ras ⁄ Erk pathway has been shown to have antiapop- tosis effects [47]. Conclusions The two inhibitors of PtdIns3K, LY294002 and wort- mannin, are useful when used under the same condi- tions, in parallel experiments, to study the importance of PtdIns3K in various signaling pathways, because they have different mechanisms of action and different pharmacological profiles. However, as with any phar- macological agents, there are other targets affected by each of the inhibitors [48–50] and thus one has to be cautious with interpretation of data. We have shown that the effects of wortmannin and LY294002 on inhi- bition of Erk activation can vary, depending upon the cell type and agonist being used. When the two inhibi- tors both inhibit PtdIns3K in a dose-dependent fash- ion, and yet have differential effects on downstream events, this must dissociate PtdIns3K activation from being directly involved in these other events. If the activation of PtdIns3K was required for a certain downstream event, then complete inhibition of the PtdIns3K pathway with either inhibitor should have equivalent effects. We addressed the issue of cross-talk between the PtdIns3K and ras ⁄ Erk pathways in two previous stud- ies [13,51]. One of the reasons we decided to re-visit this problem was the publication of a study which sug- gested that in FDC-P1 cells, PtdIns3K impinged on the ras to Erk pathway via differential effects on raf isoforms [10]. In that study, it was suggested that the difference from our work using MC ⁄ 9 cells may be due to the difference in cell types, yet those authors failed to test the effects of LY294002, relying largely on the effect of dominant negative PtdIns3K con- structs and the inhibitor wortmannin. Thus, we initi- ated the study of a number of cell types, including the FDC-P1 cells used in a previous stusy [10] to compare the effects of wortmannin and LY294002 on activation of Erk1 ⁄ 2. The results support our previous conclu- sion, using MC ⁄ 9 cells, that in most cases, inhibition of PtdIns3K does not affect the activation of Erk1 ⁄ 2. In fact, the only case in which we have seen evidence for the involvement of PtdIns3K in Erk activation was the A20 B-cell line, which is explained by the role of PtdIns3K in activating Btk and PLCc, and subse- quently activation of Erks [33]. One of the more compelling results is the difference in inhibition seen with the inhibitors in the various cell types. Results in FDC-P1 were consistent with those seen previously in MC ⁄ 9 cells, because inhibition with wortmannin could partially reduce Erk activation, whereas LY294002 had no effect. However, neither inhibitor had any effect on Erk activation in TF-1 cells. In the BAF-3 cells, there was some effect of higher concentrations of LY294002 in blocking Erk activation, but no effect was observed with wort- mannin when PKB phosphorylation was blocked. In HEK293 cells, the serum-starved cells were stimulated by re-addition of serum rather than a specific cytokine. In this case, in which a number of factors may be con- tributing to activation of PtdIns3K and ras-Erk path- ways, LY294002 was able to partially block Erk1 ⁄ 2 activation, but wortmannin did not. The latter result is reminiscent of the effect of the inhibitors on blocking phorbol 12-myristate 13-acetate (PMA)-induced Erk1 ⁄ 2 activation. When MC ⁄ 9 cells were treated with PMA, a treatment in which a ras-independent activa- tion of the Erk pathway is implicated, wortmannin did not inhibit the activation of Erk1 ⁄ 2, whereas LY294002 did inhibit [13]. Therefore, the results CON LY LY + U0126 U0126 DMSO 80 60 40 20 0 % APOPTOSIS FDC-P1 TF-1 BAF-3 Fig. 8. Summary of apoptosis studies in the three hemopoietic cell types. FDC-P1, TF-1 or BAF-3 cells were incubated in complete medium containing appropriate cytokines, then treated with either 50 l M LY294002 (LY) or 10 lM U0126, both inhibitors together, or vehicle control (dimethylsulfoxide) for 16 h. Cells were analyzed for the extent of apoptosis by measuring the percentage of cells con- taining subdiploid DNA, based on propidium iodide staining. D. Chiu et al. PtdIns3K activity is not required for Erk activation FEBS Journal 272 (2005) 4372–4384 ª 2005 FEBS 4379 showing that in some cases, LY294002 can block Erk activation, but wortmannin does not, whereas in others the reverse is true, leads us to conclude that both wort- mannin and LY294002 can have some effect on var- ious upstream components leading to activation of Erk1 ⁄ 2. Because these effects on Erks do not correlate with the inhibition of PtdIns3K by the inhibitors, then PtdIns3K activation cannot be placed upstream of Erk activation. We have not provided new insight into the mechanism of the inhibitors’ effects on Erk activation, but we can make the important point that the effect of only one of the inhibitors should not be taken as proof for PtdIns3K dependence, and there will always be a need for a dose–response analysis to conclusively dem- onstrate a correlation of the inhibitors’ effects on PtdIns3K and other downstream targets. A recent study put forward a strong argument that activation of Erk was dependent upon PtdIns3K in response to c-kit receptor activation in stem cells [33]. The results in this study were compelling, because he- matopoietic stem cell lines were compared with more differentiated mast cells, in which blocking PtdIns3K had no effect on Erk phosphorylation. However, it should be noted that only single concentration of LY29402 was used, and based on our observations, a more careful analysis with a range of inhibitor doses might be warranted, as this might account for the dif- ferent effects in the different cell types. The analysis of the effects of the PtdIns3K inhibitors and the MEK inhibitor on induction of apoptosis pro- vides further evidence that the PtdIns3K pathway regulates survival events that are independent of the Erk pathway in cytokine-dependent hemopoietic cells. When PtdIns3K is blocked with LY294002, cells undergo apoptosis, as reported previously [45]. It should be noted that when cells are incubated with 50 lm LY294002 for 16 h, there may be some inhibi- tion of Erk activation, as we have shown, but obvi- ously this is not contributing substantially to the effect on apoptosis, because blocking with U0126 alone has little or no effect. We felt it was important to do these studies using U0126, which is a more effective inhib- itor of Erk activation than PD98059 [46], to verify that Erk activation was not necessary for cell survival over the 16 h time course. However, when both inhibitors were used together, there was a synergistic effect resulting in a greater extent of apoptosis, which sup- ports the fact that the Erk pathway can provide some signals important for cell survival. These results pro- vide a functional assay of downstream events regulated by the signaling pathways and thus support our other studies, because blocking the two pathways individu- ally have very different outcomes on these cells. As we have stated previously, and can now empha- size based on results using a variety of cell lines, it is clear is that when the effects of PtdIns3K inhibition are being studied, the use of only one of the two PtdIns3K inhibitors cannot be accepted as proof that any effects being blocked are regulated by activation of PtdIns3K. Furthermore, it is important to note that the type of experiment used will only assess the effects of immediate activation of PtdIns3K on the concurrent activation of Erk1 ⁄ 2. We have not addressed the pos- sible effects of longer term PtdIns3K activation on components of the Erk signaling pathway. For exam- ple, one could imagine that there are secondary events regulated by PtdIns3K (e.g. as seen in the B cells) that may contribute to prolonged activation of Erks. In conclusion, our results using pharmacological inhibitors suggest that activation of the PtdIns3K path- way is not required for the pathway leading to acute activation of Erk1 and Erk2. We have shown this in sev- eral cytokine-dependent cell types as well as an epithe- lial cell line and primary fibroblasts. Only in one specific B-cell line could we conclude that blocking PtdIns3K activation had an inhibitory effect on the activation of Erk, which is explainable based on what is known about BCR signaling. Our study has concentrated on acute activation of these signaling enzymes and thus has spe- cifically addressed possible cross-talk between the path- ways that may occur proximal to receptor activation. Further studies will be required to address possible inter-relationships that occur due to chronic activation or inhibition of one of the pathways. Experimental procedures Cells and reagents FDC-P1, BAF-3, TF-1, BJAB, A20 and HEK293 cells and primary porcine Achilles tendon fibroblasts (tenocytes) were grown at 37 °C and 5% (v ⁄ v) CO 2 with humidity. FDC-P1 and BAF-3 cells were grown in RPMI-1640 medium sup- plemented with 10% (v ⁄ v) FBS, 20 mm 2-mercaptoethanol. 5% WEHI-3-conditioned medium was added as a source of IL-3. The TF-1 cells were grown in RPMI-1640 medium supplemented with 10% FBS, 20 mm 2-mercaptoethanol and 1% conditioned medium containing recombinant human GM-CSF (from CGM1 cells; kind gift from C. Brown, University of Calgary, Canada). HEK293 and primary tendon cells were grown in DMEM medium sup- plemented with 10% (v ⁄ v) FBS, l-glutamine, sodium pyru- vate, 50 UÆmL )1 penicillin and 50 mgÆmL )1 streptomycin. The tendon fibroblasts were extracted from porcine Achilles tendon (dissected free of peritendinous tissues) by collage- nase digestion and cultured up to the fifth passage in PtdIns3K activity is not required for Erk activation D. Chiu et al. 4380 FEBS Journal 272 (2005) 4372–4384 ª 2005 FEBS DMEM containing 10% (v ⁄ v) FBS. Prior to the experi- ments, the cells were trypsinized and replated at 500 000 cells per 60 mm plate. After adhering overnight, the cells were stimulated with various doses of IGF-I. BJAB cells and A20 cells were cultured in RPMI-1640 medium supple- mented with 10% (v ⁄ v) FBS, 50 lm 2-mercaptoethanol (A20 only) and antibiotics. Antibodies to phosphorylated S473 of PKB ⁄ akt and dually phosphorylated Erk1 and Erk2 were from Cell Sign- aling Technologies (Beverly, MA, USA); antibody to the p85 subunit of PtdIns3K was from Upstate Biotechnology Inc. (Lake Placid, NY, USA) LY294002, wortmannin, and U0126 were from Calbiochem (San Diego, CA). Goat anti- [human IgM F(ab¢) 2 ] and intact goat anti-(human IgM), rabbit anti-[mouse IgG F(ab¢) 2 ] and intact rabbit anti- (mouse IgG) were purchased from Jackson Immuno- Research (West Grove, PA, USA). IGF-I was purchased from Sigma (St. Louis, MO, USA). Cell treatments and lysis conditions Cells were starved of cytokines or growth factors by over- night incubation in medium with 1% (v ⁄ v) WEHI-3-condi- tioned medium (for FDC-P1 and BAF-3 cells), or no GM-CSF (for TF-1 cells) or no FBS (for HEK293 cells). Cells were incubated at 37 °Cin20mm Hepes (pH 7.4)- buffered RPMI-1640 prior to assay. Cells were pretreated with PtdIns3K inhibitors, LY294002 and wortmannin or the MEK inhibitor, U0126, at varying concentrations for 10 min. The cells were then stimulated with synthetic IL-3 (10 lgÆmL )1 for FDC-P1 and BAF-3 cells) or recombinant GM-CSF (10 ngÆmL )1 ; murine GM-CSF for FDC-P1 cells and human GM-CSF for TF-1 cells) or recombinant IGF-I (tendon cells) for 5 or 10 min. BJAB cells were stimulated with goat anti-[human IgM F(ab¢) 2 ] at the specified time points. A20 cells were similarly stimulated with goat anti- [mouse IgG F(ab¢) 2 ]. The HEK293 or porcine tendon fibro- blast cells were first detached from the plates in NaCl ⁄ P i containing 1 mm EDTA, then stimulated with 10% FBS in microfuge tubes. In each case, to stop the reactions, cells were pelleted and then solubilized in ice-cold lysis buffer containing 50 mm Tris ⁄ HCl, pH 7.4, 1% (v ⁄ v) Triton X-100, 10% (v ⁄ v) glycerol, 100 mm NaCl, 2.5 mm EDTA, 10 mm NaF, 1 mm Na 3 VO 4 ,1mm Na 3 MoO 4 , 10 mm b-glycerophosphate, 1 lgÆmL )1 microcystin-LR, 1 lgÆmL )1 aprotinin, 40 lgÆmL )1 phenylmethylsulfonyl fluoride, 1 lm pepstatin, 0.5 lgÆmL )1 leupeptin and 10 lgÆmL )1 soybean trypsin inhibitor. Nuclei were pelleted by centrifuging at 16 000 g for 1 min at 4 °C and the super- natants were transferred to new tubes, mixed with SDS sample buffer, boiled and separated on polyacrylamide gels prior to immunoblotting analyses. In all figures, immuno- blots shown are representative of several experiments show- ing similar results. Each of the analyses (except for those done with primary fibroblasts which were done only twice) was done a minimum of three times (and in most cases at least five times). Immunoblotting Phosphorylated PKB (S473), phosphorylated Erks (pT-E-PY motif) and p85 were visualized in the samples by running equivalent amounts of total protein lysate (50–100 lg) on 9% SDS ⁄ PAGE gels, transferred to nitrocellulose blots. Detection of the p85 subunit of PtdIns3K was used to show equivalent loading. In all cases, staining of the immunoblots with Ponceau S was also used to verify the presence of equal amounts of protein per lane. The blots were blocked in 5% skim milk and incubated in 1 : 1000 or 1 : 2000 dilution of the primary antibody. The primary antibody was detected by incubating in the appropriate secondary antibody coupled to horseradish peroxidase and the bands were visualized using ECL (Amersham Pharmacia Biotech, Piscataway, NJ, USA) according to the manufacturer’s instructions. Apoptosis analysis by staining for subdiploid DNA content The FDC-P1 cells, BAF-3 cells and TF-1 cells were washed three times in sterile, 1· NaCl ⁄ P i prior to treatments. Cells were resuspended in growth medium to a concentration of 1 · 10 6 cellsÆmL )1 , without any cytokine and treated as fol- lows. The cells were either treated with or without cyto- kines (5% WEHI-3 and 1% GM-CSF) and with or without inhibitors (LY294002 and U0126). The cells were incubated at 37 °C, 5% CO 2 in humidified air for 16 or 24 h. Fixation of the cells was modified from Current Protocols of Immu- nology. One milliliter aliquots of the cells, following various treatments, were collected and pelleted. The supernatant was poured off and 1 mL of ice-cold, 70% ethanol was added to the cells drop-wise, while vortexing. Cells were then incubated at 4 °C for at least 18 h. The cells were pel- leted by spinning at 3000 g for 5 min. The ethanol was poured off and the cells resuspended in staining buffer containing 1· NaCl ⁄ P i , 0.1% (w ⁄ v) dextrose, 100 lgÆmL )1 RNaseA, 50 lgÆmL )1 propidium iodide. The samples were kept at room temperature, in the dark, for 45 min prior to flow cytometry analysis. The propidium iodide stained cells were analyzed using a Beckman Coulter EPICS XL-MCL, flow cytometer. expo 32 data acquisition software was used to collect the data and expo 32 analysis software was used to analyze the data. The background noise and autofluorescence of the cells were gated out using the negative control sample in the FL3 versus FS histogram. The same gate was applied to all subsequent treatment samples. The amount of sub- diploid DNA content was quantified using the number of events versus FL3 histogram by marking the area starting from the y-axis to the first largest peak, which corresponds to the G1 cycle. D. Chiu et al. PtdIns3K activity is not required for Erk activation FEBS Journal 272 (2005) 4372–4384 ª 2005 FEBS 4381 [...]... (1998) Mechanisms and consequences of activation of protein kinase B ⁄ Akt Curr Opin Cell Biol 10, 262–267 Scheid MP, Lauener RW & Duronio V (1995) Role of phosphatidylinositol 3-OH -kinase activity in the inhibition of apoptosis in haemopoietic cells: phosphatidylinositol 3-OH -kinase inhibitors reveal a difference in signalling between interleukin-3 and granulocyte-macrophage colony stimulating factor Biochem... regulation of protein kinase B ⁄ Akt by hydrophobic motif phosphorylation Mol Cell 9, 1227–1240 Delcommenne M, Tan C, Gray V, Rue L, Woodgett J & Dedhar S (1998) Phosphoinositide-3-OH kinasedependent regulation of glycogen synthase kinase 3 and protein kinase B ⁄ AKT by the integrin-linked kinase Proc Natl Acad Sci USA 95, 11211–11216 Feng J, Park J, Cron P, Hess D & Hemmings BA (2004) Identification of a... mitogen-activated protein kinase by insulin in rat adipocytes: comparison of insulin and protein kinase C modulators Biochem Biophys Res Commun 209, 1082–1088 7 Von Willebrand M, Jascur T, Bonnefoy-Berard N, Yano H, Altman A, Matsuda Y & Mustelin T (1996) Inhibition of phosphatidylinositol 3 -kinase blocks T cell antigen receptor ⁄ CD3-induced activation of the mitogen-activated kinase Erk2 Eur J Biochem... Davies SP, Reddy H, Caivano M & Cohen P (2000) Specificity and mechanism of action of some commonly used protein kinase inhibitors Biochem J 351, 95–105 51 Schubert KM & Duronio V (2001) Distinct roles for extracellular-signal-regulated protein kinase (ERK) mitogen-activated protein kinases and phosphatidylinositol 3 -kinase in the regulation of Mcl-1 synthesis Biochem J 356, 473–480 FEBS Journal 272... MEK -independent pathways regulating the prolonged activation of the ERK-MAP kinases Oncogene 14, 1635–1642 9 Bondeva T, Pirola L, Bulgarelli-Leva G, Rubio I, Wetzker R & Wymann MP (1998) Bifurcation of lipid and protein kinase signals of PI3Kgamma to the protein kinases PKB and MAPK Science 282, 293–296 10 Sutor SL, Vroman BT, Armstrong EA, Abraham RT & Karnitz LM (1999) A phosphatidylinositol 3-kinasedependent... Identification of a novel inhibitor of mitogen-activated protein kinase kinase J Biol Chem 273, 18623–18632 Ballif BA & Blenis J (2001) Molecular mechanisms mediating mammalian mitogen-activated protein kinase (MAPK) kinase (MEK)-MAPK cell survival signals Cell Growth Differ 12, 397–408 Cross MJ, Stewart A, Hodgkin MN, Kerr DJ & Wakelam MJ (1995) Wortmannin and its structural analogue demethoxyviridin inhibit... in Swiss 3T3 cells Wortmannin is not a 4383 PtdIns3K activity is not required for Erk activation specific inhibitor of phosphatidylinositol 3 -kinase J Biol Chem 270, 25352–25355 49 Brunn GJ, Williams J, Sabers C, Wiederrecht G, Lawrence JC Jr & Abraham RT (1996) Direct inhibition of the signaling functions of the mammalian target of rapamycin by the phosphoinositide 3 -kinase inhibitors, wortmannin and. .. and membrane trafficking Annu Rev Cell Dev Biol 14, 231–264 5 Duronio V, Scheid MP & Ettinger S (1998) Downstream signalling events regulated by phosphatidylinositol 3 -kinase activity Cell Signal 10, 233–239 6 Standaert ML, Bandyopadhyay G & Farese RV (1995) Studies with wortmannin suggest a role for phosphatidylinositol 3 -kinase in the activation of glycogen synthase and mitogen-activated protein kinase. .. for insulin stimulation of pp70, S6 kinase, DNA synthesis, and glucose transporter translocation Mol Cell Biol 14, 4902–4911 Scheid MP & Duronio V (1996) Phosphatidylinositol 3-OH kinase activity is not required for activation of mitogen-activated protein kinase by cytokines J Biol Chem 271, 18134–18139 Frevert EU & Kahn BB (1997) Differential effects of constitutively active phosphatidylinositol 3 -kinase. .. Greenberg S, Clark EA & Kurosaki T (1998) Involvement of guanosine triphosphatases and phospholipase C-gamma2 in extracellular signal-regulated kinase, c-Jun NH2-terminal kinase, and p38 mitogen-activated protein FEBS Journal 272 (2005) 4372–4384 ª 2005 FEBS PtdIns3K activity is not required for Erk activation 37 38 39 40 41 42 43 44 45 46 47 48 kinase activation by the B cell antigen receptor J Exp Med 188, . Acute activation of Erk1/Erk2 and protein kinase B/akt proceed by independent pathways in multiple cell types Doris Chiu, Kewei Ma, Alexander Scott and Vincent Duronio Department of Medicine,. PtdIns3K, phosphatidylinositol 3 -kinase; PKB, protein kinase B ⁄ akt; PKC, protein kinase C. 4372 FEBS Journal 272 (2005) 4372–4384 ª 2005 FEBS lead to activation of downstream kinases including protein. (1998) Involve- ment of guanosine triphosphatases and phospholipase C-gamma2 in extracellular signal-regulated kinase, c-Jun NH 2 -terminal kinase, and p38 mitogen-activated protein kinase activation