Regulation of starvation- and virus-induced autophagy by the eIF2alpha kinase signaling pathway

Abstract

The eIF2alpha kinases are a family of evolutionarily conserved serine/threonine kinases that regulate stress-induced translational arrest. Here, we demonstrate that the yeast eIF2alpha kinase, GCN2, the target phosphorylation site of Gcn2p, Ser-51 of eIF2alpha, and the eIF2alpha-regulated transcriptional transactivator, GCN4, are essential for another fundamental stress response, starvation-induced autophagy. The mammalian IFN-inducible eIF2alpha kinase, PKR, rescues starvation-induced autophagy in GCN2-disrupted yeast, and pkr null and Ser-51 nonphosphorylatable mutant eIF2alpha murine embryonic fibroblasts are defective in autophagy triggered by herpes simplex virus infection. Furthermore, PKR and eIF2alpha Ser-51-dependent autophagy is antagonized by the herpes simplex virus neurovirulence protein, ICP34.5. Thus, autophagy is a novel evolutionarily conserved function of the eIF2alpha kinase pathway that is targeted by viral virulence gene products.

Figure 1

Yeast with mutations in the GCN2-signaling pathway are defective in autophagy. (A) Western blot analysis of lysates from yeast grown in normal media (YEPD) or N deprivation media lacking amino acids and ammonium sulfate (SD-N), with an eIF2α Ser-51-phosphospecific Ab (1:100 dilution; Research Genetics). (B) Quantitation by DIC microscopy of autophagic body formation in yeast grown in normal growth conditions (open bars) or subjected to 4 h of nitrogen deprivation (black bars), treatment with 0.2 μg⋅ml⁻¹ rapamycin (light gray bars), or treatment with 10 μg⋅ml⁻¹ cycloheximide (dark gray bars). Cells with one or more autophagic bodies within the vacuole were scored as positive. A minimum of 100 cells was counted for each sample for each yeast strain. The results shown represent the mean ± SEM percentage of cells with autophagic bodies within the vacuole for triplicate samples. Similar results were obtained in five independent experiments. (C) Representative DIC micrographs (left columns of each panel) and electron micrographs (right columns of each panel) of yeast deprived of nitrogen for 4 h. Arrows in light micrographs denote representative cells which would be scored as positive in experiment shown in B. Arrows in electron micrograph denote representative autophagic bodies within the vacuole. (Scale bars, 0.5 μm.)

Figure 2

Mammalian pkr promotes autophagy in GCN2-disrupted yeast. (A) Western blot analysis of lysates from N-starved Δgcn2 yeast by using an eIF2α Ser-51-phosphospecific Ab. (B) Quantitative effects of GCN2 and pkr transformation on autophagic body formation in Δgcn2 yeast grown in normal growth conditions (open bars), or subjected to 4 h of nitrogen deprivation (black bars) or treatment with 0.2 μg⋅ml⁻¹ rapamycin (light gray bars). Cells with one or more autophagic bodies within the vacuole were scored as positive. A minimum of 100 cells was counted for each sample for each yeast strain. The results shown represent the mean ± SEM percentage of cells with autophagic bodies within the vacuole for triplicate samples. Similar results were obtained in five independent experiments. Similar results were also observed for five independent clones of pkr-transformed Δgcn2 yeast that displayed normal growth phenotypes in nutrient rich media. (C) Representative light micrographs (Left column) and electron micrographs (Right column) of GCN2- and pkr-transformed Δgcn2 yeast deprived of nitrogen for 4 h. Arrows in light micrographs denote representative cells that would be scored as positive in experiment shown in B. Arrows in electron micrographs denote representative autophagic bodies within the vacuole. (Scale bars, 0.5 μm.)

Figure 3

Pkr^−/− MEFs are deficient in autophagic protein degradation and autophagic vacuole accumulation induced by α-IFN and HSV-1Δ34.5 infection. (A and B) Cumulative percentage degradation of long-lived cellular proteins in pkr^+/+ and pkr^−/− MEFs, respectively. Results are mean (± SEM) of triplicate wells. Similar results were obtained in five independent experiments. (C and D) Electron micrographs showing examples of early (Avi) and late (Avd) autophagic vacuoles, respectively, in pkr^+/+ MEFs infected with HSV-1Δ34.5. (Scale bars, 0.5 μm.) (E) Volume density of early (Avi) and late (Avd) autophagic vacuoles in pkr^−/− and pkr^+/+ MEFs infected with wt HSV-1 and HSV-1Δ34.5. Error bars represent SEM of the volume densities of three to five grid squares. Similar results were obtained in two independent experiments.

Figure 4

Mutant eIF2α S51A MEFs are deficient in HSV-1Δ34.5 infection-induced and amino acid starvation-induced autophagic protein degradation and autophagic vacuole accumulation. (A) Western blot analysis of MEF lysates by using an eIF2α Ser-51-phosphospecific Ab. The phosphorylation of wt eIF2α at 0 h after starvation is presumed to result from eIF2α kinase activation that occurs during washing and cell lysis (29). (B, C, E, and F) Cumulative percentage degradation of long-lived cellular proteins in wt eIF2α MEFs (B and E) at serial time points after virus infection (B) or amino acid starvation (E), and in homozygous mutant eIF2α S51A MEFs (C and F) at serial time points after virus infection (C) or amino acid starvation (F). Results are mean (±SEM) of triplicate wells. Similar results were obtained in three independent experiments. (D and G) Volume density of autophagic vacuoles in wt and mutant eIF2α S51A MEFs subjected to 5 h of virus infection (D) or subjected to 2 h of amino acid starvation (G). Each bar represents volume density of both early (Avi) and late (Avd) autophagic vacuoles. The percentage of early (Avi) vacuoles is indicated above each bar. Error bars represent SEM of the volume densities of three to four grid squares.