2D-gel electrophoresis followed by Western blotting for eIF2A from BY4684 ΔeIF2A yeast transformed with YCplac111-YP with or without the indicated stress.

A, HA-eIF2A from lysate of BY4684 ΔeIF2A yeast transformed with either pTB328 vector alone or YCplac111-YP was immunoprecipitated with α-HA antibody. The immunoprecipitated protein complexes were analyzed by SDS-PAGE followed by mass spectrometry. Bands analyzed by mass spectrometry are highlighted by arrows or brackets. B, Some proteins identified in the IP-mass spectrometry experiment were analyzed for RNA dependence in either GST pulldown experiments (for Ded1p, Ssb2p and eEF1A/Tef1p) or immunoprecipitations followed by Western blotting (for Xrn1p/Kem1p, which cannot be expressed in bacteria) with lysates from eIF2A-HA yeast expressing a stable C-terminal HA-tagged eIF2A protein under the control of its endogenous promoter. Protein complexes were incubated in the absence (−) or presence (+) of RNase A followed by Western blotting for eIF2A-HA. In these experiments GST alone and GST-PIN1, a mammalian cis-trans prolyl isomerase, were analyzed as negative controls. Furthermore, controls for levels of GST-tagged proteins present in the pulldown reactions are shown by coomassie-staining of the membranes after Western blotting.

A, The domain architecture of eEF1A is presented with amino acid demarcations at the boundaries of individual domains, and color coding of domains that is consistent throughout the figure. Bold arrows represent the location of mutations employed in C. Regions of eEF1A previously reported to be important for interaction with other proteins are highlighted with brackets. B, Domain deletions of eEF1A fused to GST were used to precipitate eIF2A-HA from eIF2A-HA yeast lysate. Western blotting for HA-tagged eIF2A was conducted on the precipitated material (upper panel). Domains of eEF1A present in the GST-fusion protein are indicated above the Western blot. GST fusion protein added to each GST-pulldown reaction was analyzed after Western blotting by coomassie staining of the membrane (lower panel). C, Domain 3 of eEF1A was found to be critical for the interaction between eEF1A and eIF2A. GST-fusion protein point mutants of eEF1A were used to confirm the necessity for domain 3 in the interaction between eIF2A and eEF1A using eIF2A-HA yeast. Each mutant is indicated in the figure. Western blotting followed by Coomassie staining of the membrane were conducted as described above (B).

BY4684 ΔeIF2A yeast transformed with YCplac111-YP (Figure 1A) were grown to log-phase growth and shifted to medium lacking (−) or containing 6% ethanol (+) for one hour prior to preparation of lysates. GST-pulldown reactions were then conducted with full-length eEF1A fusion protein and equivalent eIF2A-HA yeast protein lysate. Western blots for HA-tagged eIF2A (upper panel) and control coomassie staining (lower panel) of the resulting membrane are shown with 8% inputs for each lysate. Bands for input and HA-eIF2A protein precipitated were quantified using Image J software. Input band intensities were set to 1 and HA-eIF2A pulled down for each condition were normalized relative to respective input controls. The ratio of Input∶HA-eIF2A protein is indicated below each lane.

A. Schematic diagram of the YCplac111-YP plasmid containing N-terminally HA-tagged eIF2A under the control of its native promoter. This plasmid contains a LEU2 selection cassette and is used throughout the experiments presented in this manuscript, except where otherwise denoted. B, mRNA analysis (upper panels) and Western blot analysis of eIF2A and PGK1 transcripts and protein levels (lower panels) indicating eIF2A mRNA is turned over during both 6% ethanol and 1M sorbitol stress in BY4684 ΔeIF2A yeast transformed with YCplac111-YP illustrated in A.

A, BY4684 ΔeIF2A yeast transformed with YCplac111-YP with or without 6% ethanol treatment were analyzed by polysome profiling. B, Subsequently, the top and 80S complex regions of the gradients (shown in A) were analyzed by 2D-gel electrophoresis and Western blotting with α-HA antibody to detect HA-eIF2A.

A, Endogenous HA-tagged protein was immunoprecipitated from lysate of eIF2A-HA yeast. In these experiments, α-Pgk1p antibody was used as a negative control for the interaction. B, eIF2A-HA yeast lysates were used for endogenous immunoprecipitation of eEF1A protein in reciprocal reactions to those presented in A. Immunoprecipitations were analyzed for HA-tagged eIF2A and eEF1A protein by Western blotting analysis. Inputs refer to 1% of total protein added to each reaction in these experiments. Of note is that eEF1A is highly abundant within the cell, so even less apparent interactions are likely significant.

A, A domain map of eIF2A, which indicates the presence of a WD-repeat that is predicted to fold into a seven-bladed β-propeller domain, indicates the amino acid size of yeast eIF2A and approximate location of the boundaries for the WD-repeat. HA-tagged C-terminal deletions of eIF2A were prepared, as described in Results, by homologous recombination. The location of each truncation is indicated below the schematic with the amino acid position and color coding that is used throughout the figure. B, Full-length GST-eEF1A was used in precipitation experiments with lysates from HA-tagged C-terminal eIF2A deletion yeast strains. Subsequent Western blotting for HA-tagged eIF2A truncations (upper panel; indicated with amino acid numbers) and coomassie staining (lower panel) of the reactions are indicated with 8% input controls for each reaction. C,URE2 IRES function was analyzed in the presence of eIF2A truncations using the previously reported p281–4 plasmid shown in the inset [10], [18]. The p281–4 plasmid contains a galactose-inducible promoter, a stable hairpin that was shown to block cap-dependent scanning in yeast [18], and the URE2 minimal IRES element upstream of the LacZ reporter sequence. Error bars indicate standard deviation from the mean. These experiments were repeated at least three times.

URE2 IRES activity of the minimal IRES element (nucleotides 305–309) was evaluated as above (Figure 7C). This construct was transformed into ΔeIF2A yeast expressing HA-eIF2A from the YCplac111-YP plasmid (BY4684; ΔeIF2A+HA-eIF2A), and cultures were grown in the absence (minimal medium) or presence (6% ethanol) of 6% ethanol for one hour. Cells were harvested and β-galactosidase activity was measured. We have previously shown that there is no change in mRNA levels of this construct in an eIF2A-dependent manner [6], [10]. Error bars represent 15% of the mean, which is empirically acceptable for this type of experiment. These experiments were repeated several times.