Phosphorylation of Eukaryotic Initiation Factor-2α during Stress and Encystation in Entamoeba Species

Abstract

Entamoeba histolytica is an enteric pathogen responsible for amoebic dysentery and liver abscess. It alternates between the host-restricted trophozoite form and the infective environmentally-stable cyst stage. Throughout its lifecycle E. histolytica experiences stress, in part, from host immune pressure. Conversion to cysts is presumed to be a stress-response. In other systems, stress induces phosphorylation of a serine residue on eukaryotic translation initiation factor-2α (eIF2α). This inhibits eIF2α activity resulting in a general decline in protein synthesis. Genomic data reveal that E. histolytica possesses eIF2α (EheIF2α) with a conserved phosphorylatable serine at position 59 (Ser59). Thus, this pathogen may have the machinery for stress-induced translational control. To test this, we exposed cells to different stress conditions and measured the level of total and phospho-EheIF2α. Long-term serum starvation, long-term heat shock, and oxidative stress induced an increase in the level of phospho-EheIF2α, while short-term serum starvation, short-term heat shock, or glucose deprivation did not. Long-term serum starvation also caused a decrease in polyribosome abundance, which is in accordance with the observation that this condition induces phosphorylation of EheIF2α. We generated transgenic cells that overexpress wildtype EheIF2α, a non-phosphorylatable variant of eIF2α in which Ser59 was mutated to alanine (EheIF2α-S59A), or a phosphomimetic variant of eIF2α in which Ser59 was mutated to aspartic acid (EheIF2α-S59D). Consistent with the known functions of eIF2α, cells expressing wildtype or EheIF2α-S59D exhibited increased or decreased translation, respectively. Surprisingly, cells expressing EheIF2α-S59A also exhibited reduced translation. Cells expressing EheIF2α-S59D were more resistant to long-term serum starvation underscoring the significance of EheIF2α phosphorylation in managing stress. Finally, phospho-eIF2α accumulated during encystation in E. invadens, a model encystation system. Together, these data demonstrate that the eIF2α-dependent stress response system is operational in Entamoeba species.

Fig 1. Analysis of eukaryotic translation initiation factor 2α amino acid sequences.

Protein identity (A) and similarity (B) matrices were generated using BLOSUM 62 algorithm and Protein Blast. (C) The amino acids around the key serine residue, occurring at position 59 in E. histolytica were aligned using a Standard Protein BLAST. The key serine residue that becomes phosphorylated is indicated by shading. Fully conserved resides are noted by an asterisk (*) below the residues. Residues showing strongly similar properties are indicated by a colon (:). Amino acid sequences were identified using UniProtKB; UniProtKB accession number identified. Eh, Entamoeba histolytica (accession no. C4M0A4); Ei, E. invadens (accession no. S0AZW3); Tg, Toxoplasma gondii (accession no. S8GC56); Pf, Plasmodium falciparum (accession no. Q8IBH7); Sc, Saccharomyces cerevisiae (accession no. P20459); Dm, Drosophila melanogaster (accession no. P41374); Hs, Homo sapiens (accession no. P05198)

Fig 2. Viability of E. histolytica trophozoites during stress.

Log-phase trophozoites were exposed to a variety of stress conditions as described in the text. Cells were collected by centrifugation and live/dead cells were enumerated via microscopy and Trypan blue exclusion. Percent viable cells is given for each condition. The data represent the mean (± standard error) of at least three separate trials. Significant cell death occurs after long-term serum starvation or oxidative stress (***P<0.001)

Fig 3. Levels of phospho-eIF2α and total eIF2α levels in control and stressed cells.

Control or stressed cells were subjected to Western blot analysis using antibodies specific total eIF2α or phospho-eIF2α. The ratio of phospho-eIF2α to total eIF2α was determined by scanning densitometry and image analysis (Image J, NCBI). The data represent the mean (± standard error) of at least 3 separate trials. All densitometry values were corrected for load using actin and the ratio of phospho-eIF2α to total eIF2α in control unstressed cells was arbitrarily set to 100%. Representative Western blots shown below each stress condition. Only long-term serum starvation, long-term heat shock, and oxidative stress induced a statistically significant increase in the level of phospho-eIF2α (*P<0.05, ***P<0.001). Due to the lability of phospho-eIF2α, it was necessary to perform Western blots for the various conditions at different times. As such, differences in the intensities of bands representing total eIF2α are due to variations in exposure to film from trial to trial.

Fig 4. Polyribosome abundance in control and stressed cells.

Total RNA from control (A), serum-starved (B), or glucose-starved (C) cells were resolved by sucrose gradient (15–45%) ultracentrifugation, which separates free ribosomes and monosomes (light fractions) from polysomes (dense fractions). The gradients were fractionated and the fractions were analyzed by UV spectrometry (254 nm). Representative profiles of at least 3 separate trials are shown. Long-term serum starvation led to a decrease in large polysome abundance.

Fig 5. Western blot analysis confirming exogenous protein expression in transgenic cell lines.

Trophozoites were transfected with plasmids encoding luciferase (209-Luc), wildtype eIF2α (EheIF2α-S59), the non-phosphorylatable variant (EheIF2α-S59A), or the phosphomimetic variant (EheIF2α-S59D). All exogenous versions of eIF2α were FLAG-tagged at the N-terminus. Protein expression was induced using 5 μg mL^-1 tetracycline for 24 hours. Western blots of cell lysates were performed using antibodies specific for total EheIF2α, phospho-EheIF2α, the FLAG tag, or luciferase. Tetracycline induces expression of exogenous proteins.

Fig 6. Polyribosome abundance in transgenic cell lines after 72 h of induction of protein synthesis.

RNA was isolated from the four transgenic cell lines after incubation in 5 μg mL^-1 tetracycline for 72 h. The cell lines were the control cell line expressing luciferase, 209-Luc (A), the cell line overexpressing EheIF2α (B), the cell expressing the non-phosphorylable form of EheIF2α (C), and the cell line expressing the phosphomimetic form of EheIF2α (D). The RNA was resolved by sucrose gradient (15–45%) ultracentrifugation, which separates free ribosomes and monosomes (light fractions) from polysomes (dense fractions). The gradients were fractionated and the fractions were analyzed by UV spectrometry (254 nm). Representative profiles of at least three separate trials are shown. The percent of total absorbance in the dense polyribosome fractions (mean ± standard error, n≥ 3) is show in each panel. There is a statistically significant reduction in polyribosome abundance in cells expressing EheIF2α-S59D (*P<0.05) when compared to control cells (209-Luc).

Fig 7. SUnSET analysis of active protein biosynthesis in control and transgenic cell lines.

(A) Wildtype cells were incubated in normal growth medium with 100 μg mL^-1 cycloheximide (Cyclo), 10 μg mL^-1 puromycin (Puro), or both (Cyclo + Puro). Cell lysates were subjected to SDS-PAGE and Western blotting using antibody specific for puromycin. Trophozoites readily incorporated puromycin into proteins (Puro). The incorporation was authentic since it was blocked by cycloheximide treatment (Cyclo + Puro). This confirms the utility of the SUnSET procedure for assessing protein synthesis in E. histolytica. (B) Protein expression was induced in the four transgenic cell lines by tetracycline for 72 h and SUnSET was performed. Cell overexpressing the wildtype EheIF2α showed the highest level of puromycin incorporation, and thus, the highest level of protein synthesis. As expected, cells expressing the phosphomimetic protein (EheIF2α-S59D) exhibited the lowest level of puromycin incorporation in accordance with the known function of phosphorylated versions of eIF2α. Equal protein loads are demonstrated by Coomassie staining of gels (lower panels). Representative data for at least 3 separate trials are shown. Molecular weight (kDa) is shown.

Fig 8. Viability of control and transgenic cell line during stress.

(A) Expression of exogenous protein was induced with 5 μg ml^-1 tetracycline for 24 h after which the cells were exposed to stress as described (see Materials and Methods). Cells were collected by centrifugation and live/dead cells were enumerated via microscopy and Trypan blue exclusion. The cell line expressing EheIF2αS59D exhibited slightly higher viability in most stresses; however, the increases in viability were not statistically significant. (B) Protein expression was induced in the transgenic cells with 5 μg ml^-1 tetracycline for 24 h after which the cells were exposed to serum starvation for an additional 48 hours for a total of 72 hours. Although expression of any exogenous protein seemed to afford some protection to long-term serum starvation, the cell line expressing EheIF2αS59D exhibited a statistically significant increase in viability (*P<0.05). Data represents the mean (± standard error) for at least three separate trials.

Fig 9. Phospho-eIF2α accumulates during encystation in E. invadens.

(A) Western blot analysis of phospho-eIF2α and total eIF2α in E. invadens trophozoites (T) or trophozoites induced to encyst for 24 h. (B) Western blot analysis of phospho-eIF2α and total eIF2α in E. invadens trophozoites (T) induced to encyst for 72 h. The 72 h population was probed before (72) and after (72C) treatment with sarkosyl detergent, which eliminates un-encysted trophozoites from the population. (C) The ratio of phospho-eIF2α to total eIF2α was determined by scanning densitometry and image analysis (Image J, NCBI). The data represent the mean (± standard error) of at least three separate trials. All densitometry values were corrected for load using a single Coomassie stained band and the ratio of phospho-eIF2α to total eIF2α in control unencysted cells was arbitrarily set to 1.0. Representative Western blots are shown. There was significant accumulation of phospho-EieIF2α in E. invadens after 24 hours (24, *P<0.05) and 72 hours (72; **P<0.01) into encystation as well as in detergent-purified cysts (72C; ***P<0.001) (B). Coomassie blue staining of SDS-PAGE gels revealed equal protein loading. The lanes with the molecular weight markers (M) are indicated.