Translocation activity of eEF2 is essential for cell viability due to its unique role in enabling polypeptide chain elongation. For example, irreversible inhibition of eEF2 by the Diphtheria Toxin infection, which inactivates eEF2 through permanent ADP-ribosylation of Histidine 715, results in cell death (Drazin, Kandel, and Collier 1971). Therefore, reversible inhibition eEF2 is an important part of the crisis management mechanism of the cell for its adaption to the acute cues of cellular stress.
Curiously, there have been no functionally-activating post-translational modifications described for maximal eEF2 activity. Rather, ever since its discovery in 1988, the phosphorylation of the major Thr56 and two other minor Threonine residues in the GTP-binding domain is known as the switchOFF (stalled protein synthesis) and, conversely, dephosphorylation of the same residue is known as the switchON (resumed protein synthesis) for the eEF2-dependent translocase activity. In our previous study, we described a novel mechanism for the regulation of eEF2, whereby phosphorylation of a distinct Ser595 residue in the C-terminus domain by Cyclin A/CDK2 is required for the augmentation of Thr56 phosphorylation, promoting a more robust and efficient inhibition of the translocase activity (Hizli et al. 2013).
In order to acquire further insight into the physiological role of Ser595 phosphorylation, we generated a point mutant variant of the HCT116 cells that express a Ser595Ala point mutant of eEF2. Several properties of the parental HCT 116 cells were compared to those of HCT116-KI cells, including proliferation, cell cycle progression, colony formation, migratory properties, and response to Cisplatin treatment. Cell cycle analysis indicated that there is no significant difference in the abundance of G0/G1 population between asynchronously growing HCT116-WT and HCT116-KI cells. However, the mutant cells displayed a significantly higher accumulation of cells in S phase with a concomitant decrease during G2/M population compared to the parental cells. A 3-fold-reduction in the number of HCT116-KI cells at the G2/M border reconciles with both our previous finding that maximal levels of both Ser595 and Thr56 phosphorylation are seen at the entry of mitosis. Moreover, the decrease in the G2/M population in HCT116-KI cells suggests that Cyclin A/CDK2-dependent augmentation of the T56 phosphorylation is required for the proper cell cycle progression and that AK2-dependent accumulation of S595-eEF2 could be underlying the suppression of protein synthesis during mitosis reported by other groups (Fan and Penman 1970) (Celis, Madsen, and Ryazanov 1990) (Sivan and Elroy-Stein 2008) (Tanenbaum et al. 2015). On the other hand, the basis for the accumulation of HCT116-KI cells in S phase is not well understood in the sense that hyperphosphorylation of Ser595 and Thr56 was rather observed in prometaphase cells than S phase cells.
With respect to the expression levels of cell cycle regulators an obvious reduction in CDK1 levels in the HCT116-KI cells supported the decrease in G2/M population. Molecular evidence for the increased S phase population of the knock-in cells clearly did not involve changes in the levels of key regulators of G1/S transition, such as CDK4 and CDK6, nor in E2F4, PCNA and the S phase regulator, CDK2. Nevertheless, a moderate increase in the Ki-67 levels in the knock-in cells was consistent with the higher proliferation propensity of the mutant cells.
In agreement with the finding of an increased S phase population in the HCT116-KI cells, there was a 3-fold-increase in their incorporation of EdU stain, which is indicative of higher levels of DNA synthesis, compared to HCT116-WT cells. The comparison of baseline apoptotic index with respect to their Caspase 3/7 activity between the parental and mutant cell lines revealed that the HCT-116-KI cells have reduced, but progressively increasing levels of cell death, whereas HCT116-WT cells have a higher and constant apoptotic index throughout the measurement. Lack of significant difference in the levels of major apoptosis regulators such as BRCA1, BRCA2, Bax, and BIRC5 did not explain reduced levels of apoptosis in the HCT116-KI cells. However, a remarkable increase in the expression of p53 by the mutant cells was noteworthy.
To address the potential involvement of the HIPPO Pathway in mediating the increased autophagic flux in the S595A-eEF2-expressing mutant cells, mRNA levels of the several factors of the HIPPO pathway such as TAZ, YAP1, MST-1, Merlin, SAV-1, CY61, CTGF and GADD45a were analyzed. There was a significant decrease in YAP1 mRNA levels and those of its target genes, such as CY61 and CTGF. Increase seen in the autophagy levels could be a direct consequence of heightened protein synthesis in the mutant eEF2-expressing cells. On the other hand, the decrease in YAP expression in the S595A-eEF2 mutant cells, which is concomitant with their improved malignant properties, could be due to the dual role of YAP both as an oncogene and tumor suppressor reported in colorectal cancer (Ou et al. 2017). In other words, YAP could act as a tumor suppressor in the context of the HCT116 cells, where its activity as a transcriptional co-activator is in charge of the upregulation of pro-apoptotic gene expression (Lamar et al. 2012). In this respect, HCT116-KI cells could reflect the pro-apoptotic capacity of YAP, the decrease of which in the knock-in cells could account for the decreased caspase activity measured in these cells. While there was no significant change in the mRNA levels of TAZ and Merlin, the drop in MST-1 transcript levels was at a lower magnitude compared to the drop in the message levels of SAV-1. Intriguingly, these changes in the mRNA levels were co-existent with a striking increase in MST-1-dependent fluorescent signal (modest decrease in transcript level) and a modest increase in SAV1 foci (potent decrease in transcript level) as altered features of the HIPPO pathway in the S595A-eEF2 HCT116 cells compared parental cells.
In terms of the response of the two cell lines to chemotherapeutics, there is no difference in the IC50 value of Cisplatin between the parental and HCT116-KI cells. However, striking increases in the levels of S (CDK2 2 fold) and G2/M cdk’s (CDK1 7 fold) as well as Ki67 (3 fold) accompanied with decreases in G1 cdk’s (CDK4 and CDK6) in the HCT-116-KI cells compared to the HCT116-WT cells are noteworthy in the sense that while unchallenged mutant cells appear to be underpopulated in the G2/M phase, they appear to reciprocate expression pattern of the cell cycle regulators in response to genotoxic stress. Nonetheless, further investigation of cell cycle analysis and cell death upon Cisplatin treatment are needed to conclude on the differences of the knock HCT116 cells when they respond to the drug.
When increased proliferation and decreased baseline apoptosis of HCT116-KI cells are put in perspective together with their enhanced migratory and colony formation capacity, expression of the S595A-eEF2 mutant by the knock-in cells could be associated with the progression of malignant properties of HCT116 parental cells. Since S595A-eEF2 mutant expressing knock-in cells have reduced T56 eEF2 phosphorylation and display increased malignancy, these observations are in agreement with the tumorigenic effects seen upon eEF2K knock down by Xie and colleagues in HCT116 cells (Xie et al. 2014). In other words, S595A-eEF2 expressing HCT116 cells phenocopy the mild increase in the malignant properties seen when the endogenous eEF2K levels are knocked down using siRNA-based approach in HCT116 cells as an in vitro model of colorectal cancer in that study.
More interestingly, our observations on the resultant impact of an eEF2-mutant with reduced Thr56 phosphorylation on autophagy also overlap with the elevated levels of autophagy these authors report upon silencing of eEF2K (Xie et al. 2014). Mechanistically, this study reports that the reduced inhibition of protein elongation through eEF2K knock down promotes autophagic response by upregulating both Beclin 1 and Atg7 protein levels independent of changes in mRNA levels of these genes, while there is no change in the levels of Atg5 (Xie et al. 2014). On the other hand, reduced inhibition of protein elongation through expressing an eEF2 mutant with dampened Thr56 phosphorylation levels promoted an induction of Atg5 and no significant change in LC3B foci formation, while significantly increasing Beclin 1 expression both at mRNA (5 fold increase) (Fig. 6b) and protein levels (data not shown). These findings confirm eEF2K as a negative regulator of autophagy in colon carcinoma cells and highlight that inactivation of its function directly liberates eEF2 translocase activity promoting heightened protein synthesis that could trigger activation of autophagy due to elevated demand on amino acids. Finally, increased colony formation upon eEF2K knock down in HCT116 cells reported in this study is also in agreement with our finding from the same assay as confirmed by the increased colony formation ability of the HCT11-KI cells. Clearly, expression of an eEF2 mutant that is incapacitated in its inhibition somehow contributed to the progression of malignant properties of colorectal cancer cells.
Increased eEF2K expression in the context of colon cancer appears to contrast to the cases of breast, ovarian, colon, glioma, medulloblastoma, hepatocellular carcinoma, and prostate cancer where increased eEF2K activity exerts a cytoprotective effect on cancer cells through acquisition of adaptation ability to the hypoxic, nutrient-deprived, and acidic tumor microenvironmet (Leprivier et al. 2013) (Leprivier et al. 2015). In the in vitro and in vivo models of colorectal cancer, manipulations that increase eEF2K activity promotes antitumorigenic effects. For example, Faller and his colleagues demonstarted that anti-tumorigenic activity of rapamycin is exerted through re-activation of eEF2K and decreased protein synthesis in Apcmin model of colorectal cancer (Faller et al. 2015). Similarly, De Gassart and colleagues provide evidence that anti-tumorigenic activity of a drug used in HIV treatment Nelfinavir (inhibitor of HIV aspartyl protease) is strictly dependent on presence of potent eEF2K activation in the sub-clones of HeLa cells and wt and knock out eEF2K MEFs where the drug becomes ineffective in reducing tumors in the absence of active eEF2K levels (De Gassart et al. 2016). Taken together with the evidence from colorectal cancer tumors from patients, where low expression of eEF2K is associated with poor survival outcome compared to patients whose tumors express higher levels of eEF2K, epxression of this gene elicits differential consequences on malignant properties of the cancers originating from different epithelia (Ng et al. 2019). It will be interesting to fully understand the molecular basis for eEF2K’s differential role in terms of contributing to malignancy in different cell systems. A major limitation of this study to demonstrate the role of S595-phosphorylation of eEF2 in normal cell cycle progression. Therefore, the role of S595-phosphorylation of eEF2 in proper G2/M transition as well as in other types of cellular stress such as DNA damage, nutrient deprivaton, and hypoxia should be investigated in a non-transformed cell line model where the S595A mutation in eEF2 gene is knocked-in.