Eukaryotic elongation factor 2 kinase (eEF2K) is activated under diverse stress conditions, where it aids cell survival but also undergoes degradation. We show that degradation of eEF2K requires it to be in an active state but does not require its activity.
Eukaryotic elongation factor 2 kinase (eEF2K) plays roles in protecting cells against stressful conditions, including nutrient starvation [1–3] and also anti-cancer agents [4,5], as well as in neurological processes  such as dendritic stability , long-term potentiation  and learning/memory . eEF2K is a member of the small family of so-called α-kinases, rather than of the main protein kinase superfamily [10,11]. It is a Ca2+/calmodulin-dependent enzyme. Its activity is also regulated by multiple other signalling inputs including mammalian target of rapamycin complex 1 (mTORC1) [12,13], the classical mitogen-activated protein kinase (MAPK) [extracellular-signal-regulated kinase (ERK)] pathway , AMP-activated protein kinase (AMPK) [14,15] and stress-activated MAPK [16–18].
eEF2K is expressed at high levels in some types of cancer, such as gliomas, and can act to protect tumour cells against nutrient starvation , e.g. by lowering the rate of translation elongation, the stage in this energy-intensive process where most (>99%) of the energy is used. This suggests that inhibiting eEF2K activity may be of utility in solid tumours [2,5,19,20]. Cells of solid tumours may be particularly vulnerable to inhibition of eEF2K in combination with angiogenesis inhibitors, to restrict nutrient availability or in conjunction with metabolic inhibitors such as 2-deoxyglucose (2-DG). 2-DG is under evaluation as a cancer therapeutic . This is an analogue of glucose which enters cells and undergoes phosphorylation by hexokinase to the 6-phosphate in the first step of glycolysis. However, this product cannot be metabolized further by this pathway. The consequences of this can include depletion of ATP (used up in the hexokinase reaction). An earlier study showed that the effects of 2-DG on human glioma cells were enhanced by depleting cells of eEF2K by siRNA .
Since energy deprivation activates eEF2K via AMPK [14,15] and 2-DG treatment turns on AMPK , it is important to understand the relationship between treatment of cells with 2-DG and eEF2K expression levels, particularly since other studies have shown that activation of eEF2K (in that case in response to genotoxic stress) leads to its degradation via an autophosphorylation event which creates a phosphodegron . This is important in respect to the application of eEF2K inhibitors as therapeutic agents, since by inhibiting eEF2K, they might lead to its stabilization and increased expression, partly countering the effects of the inhibition of its activity and mitigating the effectiveness of eEF2K inhibitors.
Furthermore, the tumour microenvironment is frequently acidic, conditions which lead to activation of eEF2K , again prompting questions about the effect of low pH on the expression of eEF2K.
Given the importance of eEF2K in this range of settings and as a potential therapeutic target in oncology, it is clearly necessary to understand how its activity and expression levels are regulated. Our recent work showed that eEF2K activity is controlled by several oncogenic signalling pathways . The present study focuses on further delineating the mechanisms through which its expression of eEF2K is controlled at the level of the eEF2K protein itself.
Earlier work showed that the stability of the eEF2K protein can be regulated by the ubiquitin–proteasome system [23,26–28]. In the present study, we have explored the regulation of its stability to gain insight into the features of eEF2K which modulate its degradation. In particular, we have studied the regulation of eEF2K expression by 2-DG and low pH and the importance for this of both the activity and a key autophosphorylation site in eEF2K. Intriguingly, our data indicate that eEF2K needs to be in an active form to undergo efficient degradation but that activity itself is not needed for this. Small-molecule inhibitors of eEF2K, which may have value in cancer therapy, are not therefore expected to stabilize eEF2K and promote its expression.
MATERIALS AND METHODS
All chemicals were purchased from Sigma unless otherwise stated. Bradford assay reagent was from Bio-Rad Laboratories. Compounds JAN-849 and -978 were from Janssen Pharmaceutica. Patent protection for these compounds has been applied for and is in process.
Primary antibodies: the p-eEF2K Thr348 and p-eEF2 Thr56 antisera were generated by Eurogentec. The antibody for p-eEF2K Ser445 was kindly provided by Dr D. Guardavaccaro (Hubrecht Institute-KNAW, Utrecht, The Netherlands). Other antibodies were obtained as follows: anti-eEF2 (eukaryotic elongation factor 2), Cell Signaling Technology; anti-tubulin, anti-actin, anti-GST and anti-FLAG from Sigma; anti-HA from Roche. Anti-eEF2K was kindly provided by the Division of Signal Transduction Therapy, College of Life Sciences, University of Dundee, Dundee, U.K.
Generation of an eEF2K conditional knockout allele
Mice (C57BL/6J) in which the Eef2k gene is disrupted were created as described in . Homozygous Eef2k−/− mice were viable and showed no developmental abnormalities or problems with fertility.
Cell culture, transfection and treatment
Human embryonic kidney (HEK) 293 cells were grown in 4.5 g/l D-glucose-containing Dulbecco's modified Eagle's medium (DMEM) containing 10% (v/v) FBS and 1% penicillin/streptomycin. Cells were pre-cultured in 1 g/l D-glucose-containing DMEM for 1 h before 2-DG treatment (always using 10 mM 2-DG).
For pH experiments, HEK293 cells were cultured in DMEM containing 44 mM NaHCO3 (culture medium for maintenance) or the pH was adjusted by adding increasing concentrations of NaHCO3 to DMEM. In all cases, the medium contained 10% (v/v) FBS and 1% penicillin/streptomycin. Cells were pre-cultured in 1 g/l D-glucose-containing DMEM for 1 h before 2-DG treatment.
HEK293 cells were transfected using calcium phosphate. Cells were pre-treated with the proteasome inhibitor MG132 (10 μM) in DMEM containing 1 g/l D-glucose for 1 h before 2-DG treatment.
After treatment, cells were lysed in ice-cold Triton lysis buffer containing 1% (v/v) Triton X-100, 50 mM Tris/HCl, pH 7.4, 50 mM NaCl, 0.2 mM EDTA, 0.2 mM EGTA, 50 mM β-glycerophosphate, 1 mM Na3VO4, 15 mM β-mercaptoethanol and 1× protease inhibitor cocktail. Lysates were spun at 16 000 g for 10 min, the supernatants were kept and total protein concentration was quantified by the Bradford assay (Bio-Rad Laboratories) following the manufacturer's instructions. Normalized lysates were either kept at −80°C or subjected to further analysis.
SDS/PAGE and Western blot analysis
Laemmli sample buffer was added to the cell lysates and the mixture was boiled at 100°C for 4 min before being loaded on the SDS/PAGE gel. SDS/PAGE was performed using the Bio-Rad Laboratories Protean 3 mini-slab gel system followed by electrotransfer of proteins resolved by SDS/PAGE on to nitrocellulose membranes. Membranes were then blocked in PBS containing 0.1% Tween 20 (PBST) and 5% (w/v) skimmed milk for 1 h, washed thrice in PBST and then probed with indicated primary antibody overnight at 4°C. The next morning, membranes were incubated with fluorescently tagged secondary antibody for 1 h, washed again thrice in PBST and fluorescent signals were visualized using Li-Cor Odyssey® imaging system. Vertical lines in some panels indicate that the lanes shown are non-adjacent ones from the same gel and exposure of the immunoblot. Data from multiple experiments are given as means±S.E.M.
Binding of eEF2K to β-transducin repeat-containing protein (β-TrCP)
HEK293 cells were transfected with pcDNA3-FLAG-β-TrCP and haemagglutinin (HA)–eEF2K. Cells were pre-treated with the proteasome inhibitor MG132 (10 μM) in 1 g/l D-glucose-containing DMEM for 1 h before 2-DG treatment. Cells were harvested in buffer comprising 50 mM Tris/HCl, pH 7.4, 10 mM β-glycerophosphate, 150 mM NaCl, 0.5% (w/v) Nonidet P40 (NP40) 1 mM Na3VO4 plus 1× protease inhibitors (cocktail from Roche) and 15 mM β-mercaptoethanol.
A 3 μg amount of anti-FLAG antibody was pre-bound to 10 μl of Protein G–agarose beads at 4°C overnight. Beads were then washed thrice with PBS and then 1 mg (protein) of cell lysate was added and tumbled at 4°C for 90 min. Beads were finally washed thrice with NP40 lysis buffer and bound proteins were subjected to SDS/PAGE.
FACS analysis and cell survival
Following treatment, the cells were trypsinized using trypsin/EDTA (0.5%) and resuspended in 100 μl of ice-cold 1× PBS, fixed upon addition of 900 μl of methanol and kept at −20°C for 16–24 h before processing. Prior to flow cytometry analysis, cells were pelleted by centrifugation at 200 g for 5 min at room temperature and resuspended in 1 ml of PBS containing 20 μg/ml propidium iodide and 100 μg/ml RNase. Quantification of propidium iodide staining was performed using a FACSCalibur flow cytometer and Cellquest software (BD Biosciences). Experiments were recorded in logarithmic scale to separate sub-G1 population and live cells. For each sample, 20000 gated events were recorded.
Point mutations were created by PCR mutagenesis using the QuikChange® system (Stratagene). Vectors were always resequenced prior to use.
RESULTS AND DISCUSSION
2-DG induces the degradation of eEF2K
As mentioned above, 2-DG is a metabolic poison which inhibits glycolysis and depletes cells of ATP, leading to activation of AMPK , which can in turn cause the stimulation of eEF2K [14,15]. Consistent with this, treatment of HEK293 cells with 2-DG induced a sustained increase in the phosphorylation of eEF2 (Figure 1A). To study the effect on eEF2K levels and the mechanisms underlying their control, cells were transfected with a vector encoding FLAG- or HA-tagged eEF2K, as described below; this allowed us to compare the role of specific features of eEF2K, such as phosphorylation sites, in regulating its stability and expression.
2-DG induces phosphorylation of eEF2 and loss of eEF2K protein
Treatment with 2-DG caused a rapid and marked decrease in the levels of the FLAG-tagged eEF2K protein (Figure 1B), which was largely prevented by the proteasome inhibitor MG132 (Figure 1C). The effect of MG132 indicates (i) that the degradation of eEF2K involves the proteasome (consistent with earlier data for its loss under other conditions [23,27,28]), and (ii) that this is primarily an effect on the stability of the eEF2K protein rather than on its synthesis.
The known phosphodegron is not involved in 2-DG-induced decay of eEF2K
It has previously been reported that, in response to genotoxic stress, degradation of eEF2K occurs via a mechanism involving a phosphodegron that includes Ser441 and Ser445, a site of autophosphorylation . However, in contrast with this, eEF2K[S441A/S445A] underwent degradation in response to 2-DG treatment similarly to wild-type (WT) eEF2K (Figures 1D and 2A). Mutation to alanine of the second major autophosphorylation site in eEF2K, Thr348 [30,31], markedly diminished its 2-DG-induced degradation, although it did not completely prevent it (Figure 2A). Phosphorylation of Thr348 is required for the full activity of eEF2K against eEF2 , probably because it allows the enzyme to adopt an active conformation (Figure 2B), in a similar way to that described for another α-kinase, myosin heavy chain kinase A (MHCK A) . The observation that this mutant is autophosphorylated on Ser445 (Figure 2A) probably reflects the fact that it does possess some residual activity. The fact that the T348A mutant is stabilized provides further evidence that 2-DG-induced loss of eEF2k is not mediated through the phosphodegron that contains Ser445.
Identification of residues in eEF2K that affect its susceptibility to degradation
A recent report has shown that another site, Ser500, a target for cAMP-dependent protein kinase, can also play a role in the degradation of eEF2K under certain conditions . However, 2-DG still induced degradation of eEF2K[S500A] to a similar degree as for WT eEF2K (Figure 2A, right-hand section). Thus, the 2-DG-induced breakdown of eEF2K does not require any of the phosphorylation sites previously linked to its proteasome-mediated decay. Lastly, since 2-DG causes activation of AMPK, as indicated by enhanced phosphorylation of its substrate, acetyl-CoA carboxylase (ACC; Figure 1C), we tested whether mutating the reported AMPK-site in eEF2K, Ser398 , affected its stability in 2-DG-treated cells. It did not (results not shown).
Catalytic activity is required for 2-DG-induced loss of eEF2K
Since the T348A mutant of eEF2K, which shows only 10–15% of the activity than WT eEF2K , was less susceptible to 2-DG-induced degradation, we tested a kinase-dead mutant of eEF2K (eEF2K[K170M] ). The levels of this protein were not affected by 2-DG-treatment of HEK293 cells (Figure 2C).
We recently identified other mutants of eEF2K which show no or little catalytic activity (eEF2K[N242A] and eEF2K[N299A] ). Importantly, these mutations are in a quite different region of eEF2K from K170M, the so-called N/D-loop which may serve a similar function to the activation loop in classical protein kinases and play a role in substrate recognition . Mutations in this region reduce or eliminate activity. The two inactive N/D-loop mutants were also refractory to 2-DG-induced degradation (Figure 2C). Taken together, the data indicate that catalytic activity is required for the degradation of eEF2K. This is clearly not because autophosphorylation at Ser445 within the previously identified phosphodegron  is involved. However, mutation of another major autophosphorylation site at Thr348 does render eEF2K more stable in 2-DG-treated cells. This may reflect the fact that the T348A mutant cannot fully adopt the active conformation (as for MHCKA ); the stability of the catalytically inactive K170M, N242A and N299A mutants is also consistent with this explanation.
Degradation of eEF2K linked to the phosphorylation of Ser441/Ser445 involves the ubiquitin E3 ligase β-TrCP , which exists as two isoforms, β-TrCP1 and β-TrCP2 . Interestingly, 2-DG treatment increased the association of WT eEF2K with β-TrCP1 and β-TrCP2 (Figure 3A), as shown by their co-immunoprecipitation. This interaction and its enhancement by 2-DG were also found by mass spectrometric analyses of eEF2K immunoprecipitates. However, β-TrCP1 did not bind to the eEF2K[S441A/S445A] mutant (Figure 3B) and neither did β-TrCP2 (results not shown). This is in agreement with earlier data showing that β-TrCP requires this phosphodegron, but since the eEF2K[S441A/S445A] mutant still undergoes 2-DG-induced degradation, indicates that β-TrCP is not playing a major role in destabilizing eEF2K in this setting. Consistent with this, eEF2K[K170M] still bound β-TrCP2 similarly to WT eEF2K in 2-DG-treated cells, although this mutant is stable (results not shown).
Interaction of eEF2K and mutants with β-transducin repeat-containing protein (β-TrCP)
Given that the phosphodegron previously linked to degradation of eEF2K is not required for its degradation in response to 2-DG, we asked whether an alternative phosphodegron might be involved. There are several pairs of adjacent serines in eEF2K which might create a phosphodegron, e.g. Ser27/Ser31, Ser70/Ser74, Ser392/Ser396 and Ser470/Ser474. We created double alanine mutants at each pair and tested their stability. The S392A/S396A mutant was stable to 2-DG treatment, whereas all the others were degraded similarly to WT eEF2K (Figure 4A). This mutant protein underwent autophosphorylation at Thr348 at similar levels to WT eEF2K illustrating it is catalytically active (Figure 4B).
Effects of selected mutations in eEF2K on its stability and binding to β-TrCP
Although the eEF2K[S392A/S396A] mutant was not degraded in response to 2-DG (Figure 4B), this mutant protein still bound to β-TrCP1 and β-TrCP2 (Figure 4C), albeit slightly less than WT eEF2K. This interaction was enhanced by 2-DG. These data again strongly suggest that β-TrCP1/2 are not important in mediating the 2-DG-induced decay of eEF2K.
Proteasome-mediated loss of eEF2K also occurs at low pH and requires eEF2K activity
It was of interest to examine whether degradation of eEF2K also occurs in other settings where eEF2K is activated. eEF2K is activated at pH values slightly below the normal physiological value of 7.4 , such as may occur during conditions of acidosis. Shifting cells from pH 7.4 to 6.6 caused an increase in eEF2 phosphorylation consistent with activation of eEF2K. In contrast, we observed a drop in eEF2K levels (Figure 5A); this suggests (i) that eEF2K is strongly activated at this lower pH, as lower kinase levels support higher substrate phosphorylation, and (ii) its activation is associated with its degradation. The finding that MG132 prevented the loss of eEF2K (Figure 5B) is consistent with its loss being due to its proteasome-mediated degradation.
Extracellular acidosis leads to eEF2K degradation
Similarly to the situation for the 2-DG-induced degradation of eEF2K, the eEF2K[S441A/S445A] mutant underwent degradation to a similar extent to WT eEF2K (Figure 5C), whereas the levels of the kinase-inactive K170M mutant were not affected by shifting from pH 7.4 to 6.6 (Figure 5D). Thus, in this setting too, the previously identified phosphodegron involving these two serine residues is not required for control of eEF2K stability.
Finally, and again analogously to the situation for 2-DG, the eEF2K[S392A/S396A] mutant was stable at the lower pH, similarly to the kinase-inactive K170M mutant (Figure 5D).
The data above indicate that inactive mutants of eEF2K, such as K170M, T348A, N242A and N299A are stable to 2-DG and/or low pH. This could reflect a requirement of activity itself for degradation or an influence of conformation on stability (none of these mutants is phosphorylated at Thr348, a residue thought to be required for attaining the correct conformation in α-kinases ). The latter explanation would be consistent with the data for the eEF2K[S392A/S396A] mutant, which appear to have an altered conformation  and which is also stable. To test whether activity per se is needed for the degradation of eEF2K, we studied whether known inhibitors of eEF2K activity result in stabilization. An earlier study reported that A484954 inhibits eEF2K , whereas a previously described compound, NH-125, has off-target effects and actually enhances eEF2 phosphorylation . We tested the effect of A484954 on the stability of eEF2K. This commercially available compound is not a potent inhibitor of eEF2K, but did reduce p-eEF2 levels when used at a high concentration (100 μM; Figure 6A) confirming that it can impair eEF2K function within cells. However, it did not prevent the loss of eEF2K protein induced by 2-DG (Figure 6A), suggesting that ongoing eEF2K activity is not required for this. Notably, A484954 had no effect on the phosphorylation of eEF2K at Thr348 (Figure 6A), even though this is a site of autophosphorylation. Phospho-Thr348 may be protected from dephosphorylation because it is thought to reside within a binding pocket in analogous way to the corresponding autophosphorylation site in MHCK A  where it may be inaccessible to phosphatases. This point is discussed further below.
Pharmacological inhibition of eEF2K does not stabilize it
In addition, we therefore used a recently developed and potent eEF2K inhibitor, JAN-978 (generated by Janssen Pharmaceutica; see the Materials and methods section). This compound does effectively inhibit eEF2K within cells (as judged from the signal for p-eEF2, Figure 6B), but it did not prevent the loss of eEF2K protein induced by incubating cells at pH 6.6. Indeed, if anything, greater loss of eEF2K was seen. These data indicate that it is not the ongoing activity of eEF2K which is required for its degradation. One possibility, mentioned above, is that autophosphorylation of eEF2K and attainment of the active conformation are required.
Depriving HCT116 colorectal carcinoma cells of glucose also causes activation of eEF2K, as shown by increased phosphorylation of eEF2 (Figure 6C), and, by 24 h, causes a marked decrease in eEF2K protein levels (Figure 6C). This is consistent with the observation that nutrient deprivation activates eEF2K in other cell types . We made use of another eEF2K inhibitor JAN-849, related to JAN-978, to explore whether eEF2K activity was required for its decay in this situation. As shown in Figure 6(C), JAN-849 did not affect the loss of eEF2K protein that occurs during glucose starvation, again indicating that ongoing eEF2K activity is not needed for its decay.
Do specific ubiquitination sites mediate 2-DG-induced loss of eEF2K?
The observation that the proteasome mediates the degradation of eEF2K indicated that it was subject to ubiquitination. We therefore attempted to identify the ubiquitination events involved. To do so, we first inspected the information available at http://www.phosphosite.org/. This indicated that several lysines have been reported as probable sites of ubiquitination and we focused on the most frequently reported ones, Lys341, Lys347, Lys485, Lys603 and Lys684. Each was mutated to arginine, another positively charged residue but one which is not subject to ubiquitination. Expression of each of the five single lysine-to-arginine mutants was still decreased in response to 2-DG (Figure 7A), indicating that none was responsible for the 2-DG-induced destabilization of eEF2K. We next addressed this point by immunoprecipitating eEF2K from 2-DG-treated cells, subjecting it to tryptic digestion followed by mass spectrometric analysis. Ubiquitination sites are evident (i) by the failure of trypsin to cleave next to the lysine residue, and (ii) by a ‘tag’ comprising G-G, which arises from tryptic digestion of ubiquitin. By these criteria, only two sites showed increased modification following 2-DG treatment, Lys525 and Lys603 (already tested). Neither the K525R nor the K525R/K603R double mutant was stabilized. We also made and tested K341R/K347R and K603R/K684R mutants, but these were not stabilized either. Thus, despite these extensive studies, we were unable to identify the lysine residues involved in the 2-DG-induced degradation of eEF2K, perhaps because ubiquitination events can ‘jump’ from one location to another when the first is not available. In view of this feature, we have not studied the effects of mutating other combinations of lysine residues.
Testing the role of selected ubiquitination sites in controlling the stability of eEF2K
High levels of eEF2K increase 2-DG-induced cell death
How could cells benefit from down-regulating eEF2k under conditions where it is activated? This seems counterintuitive, unless it is either to help cells recover again  or because high levels of p-eEF2 are detrimental to cells by inhibiting protein synthesis too strongly and, e.g., compromising cell survival .
To address this, we studied the effect of overexpressing eEF2K on cell survival in control or 2-DG-treated conditions. As shown in Figure 8(A), excess WT eEF2K had little or no effect on cell survival under control conditions, but as judged by FACS analysis, caused a marked increase in cell death when cells were treated with 2-DG. Protein synthesis is a major driver of cell growth, as most of the dry mass of cells is protein. We therefore also used the FACS analysis to assess cell size. eEF2K caused a small (non-significant) decrease in cell size under control conditions, but a much larger (and significant) decrease in 2-DG-treated cells (Figure 8B). This is consistent with greater activity of eEF2K, a negative regulator of protein synthesis, in 2-DG-treated cells.
High levels of eEF2K enhance 2-DG-induced cell death
This prompted us to ask whether eEF2K plays any role in promoting cell survival in response to 2-DG; to do this, we used mouse embryonic fibroblasts (MEFs) from WT or eEF2K-KO mice. Such cells were treated with 2-DG for up to 56 h, and cells were analysed by FACS, to examine cell death, or by Western blot. The FACS analysis revealed accumulation of sub-G1 cells after 48 h of 2-DG treatment and that such accumulation was significantly greater in eEF2K-KO cells, indicating that endogenous eEF2K does protect these cells to some extent from 2-DG (Figure 8C). The levels of p-eEF2 rose by 48 h, and this was accompanied by a fall in eEF2K levels, again indicating that activation of eEF2K causes its loss (Figure 8D).
In the present study, we demonstrate that under a range of conditions where eEF2K is activated, i.e. treatment of cells with 2-DG, exposure to low pH or glucose starvation, eEF2K protein levels decrease and, based on the fact that the proteasome inhibitor MG132 prevents this loss, is mediated by the proteasome. Interestingly, all of these conditions are relevant to the consideration of eEF2K as an anti-cancer target: 2-DG is a potential anti-cancer agent which is under clinical evaluation ; glucose starvation resembles the nutrient deprivation which cancer cells may undergo; and the tumour microenvironment is often acidic.
A key observation of the present study is that the loss of eEF2K that we observe occurs via a pathway that is distinct from that engaged following genotoxic stress; the latter involves Ser441/Ser445 and β-TrCP , whereas the decay described in the present study is independent of that phosphodegron and occurs for mutants which cannot bind β-TrCP. However, 2-DG treatment does promote increased association of β-TrCP1/2 with WT eEF2K. This suggests that there are alternative pathways for the decay of eEF2K, one of which involves β-TrCP1/2 and the previously identified phosphodegron and at least one additional pathway which is engaged under the conditions studied in the present study. Thus, abrogating one pathway, e.g. by mutating the known phosphodegron which allows interaction with β-TrCP, does not prevent the loss of eEF2K.
Our data also show that inactive mutants of eEF2K, such as K170M , or mutants with low activity e.g., T348A , are stabilized, indicating activity is required for decay/degradation. However, none of the three pharmacological inhibitors of eEF2K we have used prevent its degradation. One explanation of this apparent paradox is that inactive mutants cannot undergo the autophosphorylation event (at Thr348) that is required for eEF2K and other α-kinases to attain the active conformation (Figure 2B) [33,34] and that this conformation renders eEF2K sensitive to proteasome-mediated degradation. The data for the S392A/S396A mutant, which our previous data imply has an altered conformation , support this idea, inasmuch as this variant displays activity but is stable.
Our data imply that pharmacological inhibitors of eEF2K, which are under investigation as potential anti-cancer therapies [1,2,19], are not expected to stabilize eEF2K, which could otherwise have mitigated against their suitability because, although on one hand they would impair eEF2K activity, on the other hand they would also increase its levels of expression.
Our data also imply that the role of eEF2K in cell survival is complex, with a certain level of eEF2K being beneficial whereas high levels can actually promote cell death. This could reflect a requirement to restrain eEF2K activity (to conserve resources) under stressful conditions, while not inhibiting protein synthesis so strongly that pro-survival proteins cannot be made.
AMP-activated protein kinase
Dulbecco's modified Eagle's medium
eukaryotic elongation factor 2
eukaryotic elongation factor 2 kinase
human embryonic kidney
mitogen-activated protein kinase
mouse embryonic fibroblast
A, myosin heavy chain kinase A
PBS containing 0.1% Tween 20
β-transducin repeat-containing protein
Xuemin Wang and Christopher Proud conceived the study. All authors contributed to the design of the experiments. Xuemin Wang, Jianling Xie, Sergio da Mota and Claire Moore conducted the experiments and Christopher Proud wrote the paper with assistance from the other authors.
We thank Dr Andrew Bottrill (University of Leicester, Leicester, U.K.) for mass spectrometric analyses and Dr D. Guardavaccaro (Hubrecht Institute-KNAW, Utrecht, The Netherlands) for the β-TrCP vectors and anti-eEF2K (p)-Ser445 antibodies.
This work was supported by the Wellcome Trust [grant number 086688 to C.G.P.].
Present address: South Australian Health & Medical Research Institute, P.O. Box 11060, SA5001, Adelaide, Australia.
Present address: Cancer Research UK Centre, Somers Building, Southampton General Hospital, Tremona Road, Southampton SO16 6YD, U.K.