The interaction of the cap-binding complex (CBC) with eIF4G is dispensable for translation in yeast

Abstract

In eukaryotes, the m(7)GpppN cap structure is added to all nascent RNA polymerase II transcripts, and serves important functions at multiple steps of RNA metabolism. The predominantly nuclear cap-binding complex (CBC) binds to the cap during RNA synthesis. The predominantly cytoplasmic eukaryotic initiation factor 4F (eIF4F) is thought to replace CBC after export of mature mRNA to the cytoplasm, and mediates the bulk of cellular translation. Yeast as well as mammalian CBC interacts in vitro with eIF4G, a subunit of eIF4F. In this work, we investigate a potential role of this interaction during translation in yeast. We identify a mutation (DR548/9AA) in Tif4631p, one of two isoforms of yeast eIF4G, that abolishes its binding to CBC. Cells expressing this mutant protein as the sole source of eIF4G grow at wild-type rates, and bulk cellular translation, as assessed by metabolic labeling and polysome profile analysis, is unchanged. Importantly, we find that the DR548/9AA mutation neither diminishes nor delays the translation of newly induced reporter mRNA. Finally, microarray analysis reveals marked transcriptome alterations in CBC subunit deletion strains, whereas eIF4G point mutants have essentially a wild-type transcriptome composition. Collectively, these data suggest that in yeast, the phenotypic consequences of CBC deletions are separable from its interaction with eIF4G, and that the CBC-eIF4G interaction is dispensable for a potential "pioneering round" of translation in yeast.

FIGURE 1.

Mutation of amino acids 548 and 549 of eIF4G1 to alanine impairs binding to CBC in vitro. [³⁵S]-labeled wild-type fragment of eIF4G1 protein (amino acids 486–656) or various point mutants as indicated above the panel were translated in rabbit reticulocyte lysate. All reactions yielded similar amounts of labeled protein as confirmed by SDS-PAGE (not shown). Equal amounts of each eIF4G1 protein were then incubated with a column with antibody-immobilized CBC (+) or a control column, prepared with preimmune serum (−), as previously described (Fortes et al. 2000). After incubation and washing, the bound fractions were analyzed by 8% SDS-PAGE (top) and binding was quantified by fluorimetry. The relative binding efficiencies of the mutants in comparison to the wild-type eIF4G1 are shown (bottom). The integrity and amount of unbound material was also analyzed on a separate gel confirming that there was no selective loss of [³⁵S]-labeled eIF4G1 in any of the binding reactions (not shown).

FIGURE 2.

Phenotypic characterization of the eIF4G1–548 mutation. (A) Schematic representation of the eIF4G1 protein. Boxed regions represent binding sites for the proteins indicated above. Vertical bars indicate the nature and position of the mutations described in this study. (B) Immunoprecipitation using rabbit polyclonal α-CBP80 antibodies and extracts from indicated yeast strains. A Western blot analysis is shown revealing the amount of eIF4G (upper panel) and CBP80 (lower panel) proteins present in the input (left panels, 1% of the input is loaded), and bound fractions (right panels, 100% of the eluate) of the immunoprecipitation. A residual level of eIF4G binding to the antibody resin is seen even in the absence of CPB80 protein (lane 4). Variation of several parameters of the immunoprecipitation protocol failed to reduce this background binding (data not shown); however, robust binding of eIF4G1–459 protein and a lack of detectable binding of eIF4G1–459,548 protein to the CBC resin was consistently seen in three independent repeat experiments. (C) Serial dilutions (10-fold) of the indicated yeast strains growing exponentially were plated on YPD and incubated for 48 h at 30°C. Doubling times listed on the right were measured in liquid YPD medium over 8 h of incubation at exponential growth at 30°C.

FIGURE 3.

Effects of eIF4G mutants on general translation activity. (A) The amount of newly synthesized proteins in different strains was measured by pulse labeling with [³⁵S]-methionine. The data are expressed as a percentage of methionine incorporation measured with each mutant strain relative to the 4G1 wild-type control. Shown are averaged results from 10 to 16 measurements corresponding to 3 to 5 independent yeast cultures with standard deviations. (B) Polysome profile analysis. Yeast extracts from the indicated strains were analyzed by sucrose density gradient centrifugation. Absorbance profiles at 254 nm were recorded after centrifugation. The profiles are representative of several biological repeat experiments.

FIGURE 4.

Translation of newly made mRNA is independent of the eIF4G-CBC interaction. The yeast strains 4G1 and 4G1–548 were transformed with a plasmid expressing the luciferase reporter mRNA under the control of a galactose-inducible promoter. The resulting strains YBJ 220 and YBJ 221, respectively, were used to monitor the accumulation of luciferase activity over time after galactose induction as described in Materials and Methods. The luciferase activity was normalized against the number of cells per sample. The graph represents the average of three independent experiments.

FIGURE 5.

Transcriptome composition analysis. Scatter plot representation of the transcriptome microarray analysis. Each dot represents the averaged measurements for a single gene. On the vertical and horizontal axes, the intensities measured for the indicated probes corresponding to two fluorescent channels are plotted. Each panel represents the average of two or three array experiments (including dye-swap). (A) A self-to-self experiment using identical RNA samples to prepare the Cy3- and C5-labeled probes. Comparisons of probes derived from strains 4G1–459 (B) and 4G1–548 (C) to control strain 4G1, and from the deletion strains Δcbp20 (D) and Δcbp80 (E) to corresponding isogenic wild-type cells are also shown. The coefficient of pairwise correlation is a measure of the distance between a control-to-control and mutant-to-control comparison: a coefficient of 0 corresponds to the minimal distance between two identical arrays and a coefficient of 1 represents the maximal distance for two divergent experiments.