Depletion of cap-binding protein eIF4E dysregulates amino acid metabolic gene expression

Abstract

Protein synthesis is metabolically costly and must be tightly coordinated with changing cellular needs and nutrient availability. The cap-binding protein eIF4E makes the earliest contact between mRNAs and the translation machinery, offering a key regulatory nexus. We acutely depleted this essential protein and found surprisingly modest effects on cell growth and recovery of protein synthesis. Paradoxically, impaired protein biosynthesis upregulated genes involved in the catabolism of aromatic amino acids simultaneously with the induction of the amino acid biosynthetic regulon driven by the integrated stress response factor GCN4. We further identified the translational control of Pho85 cyclin 5 (PCL5), a negative regulator of Gcn4, that provides a consistent protein-to-mRNA ratio under varied translation environments. This regulation depended in part on a uniquely long poly(A) tract in the PCL5 5' UTR and poly(A) binding protein. Collectively, these results highlight how eIF4E connects protein synthesis to metabolic gene regulation, uncovering mechanisms controlling translation during environmental challenges.

Keywords: amino acid metabolism; cap-binding protein; translation.

Figure 1:. Transcript-level sensitivities to eIF4E depletion.

(A) Schematic of conditional depletion of eIF4E by auxin-inducible degron (AID). (B) Western blot for eIF4E-mAID-Flag expression levels over the course of Indole-3-acetic acid (IAA) depletion. (C) Bulk translation measured by nascent peptide metabolic labeling in eIF4E depleted cells. IAA or cycloheximide (CHX) was added for indicated durations and maintained during a 2-hour labeling period with L-Homopropargylglycine (HPG). Median intensities of the Alexa Fluor^™ 488 (HPG) signal normalized to that of DMSO treated cells. (**) p < 0.05 by Student’s t test. (D) Growth rates of eIF4E depleted cells. (E) Growth curves of eIF4E-AID CRISPRi cells with null gRNA or CDC33 gRNA, maintain in SCD-URA and IAA or DMSO. See also Figure S1.

Figure 2:. Transcript-level sensitivities to eIF4E depletion.

(A) Differential expression after 1 hour of eIF4E depletion measured by RNA-seq and ribosome profiling. IAA-treated cells are compared with DMSO-treated controls. Color represents significant (adjusted p-value < 0.05) and substantial (absolute fold-change (log₂) > 0.58) changes. Correlation coefficient (Pearson’s) calculated between log₂ fold changes of RPF and RNA abundance. (B) As in (A), for 8-hour IAA treatment. (C, D) Histograms of gene expression changes after (C) 1 hour or (D) 8 hours of eIF4E depletion. (E) RNA abundance fold-change (log₂) for transcripts grouped according to steady-state mRNA half-life. (**) p < 0.05; one-way ANOVA test followed by Tukey’s HSD test. (F) Translation efficiency fold change (log₂) for transcripts grouped based on length. (**) indicates p < 0.05; one-way ANOVA test followed by Tukey’s HSD test. See also Figure S2 and Table S1.

Figure 3:. CiBER-seq profile ARO10 expression regulation.

(A) RT-qPCR of ARO10 expression over the course of eIF4E-AID depletion. (**) p < 0.05 by Student’s t test. (B) RT-qPCR of P(Z) reporter expression over the course of eIF4E-AID depletion. (**) p < 0.05 by Student’s t test. (C) Schematic of CiBER-seq screen. (D) Genome-wide CiBER-seq results showing fold-change (log₂) in P(Z) reporter abundance, relative to P(TFC1) reporter levels, for each gRNA. Line indicates significance cutoff (adjusted p-value < 0.05). (E) GO analysis for genes targeted by gRNAs that up-regulated P(Z) expression (log₂ fold-change > 1 and adjusted p-value < 0.05). The most statistically significant entries were chosen and narrowed based on percentage of overlapping genes. (F) RT-qPCR of ARO10 expression following 1-hour of eIF4E-AID depletion in ARO80 and aro80Δ cells. (**) represents p < 0.05 calculated by Student’s t test. G) RT-qPCR of ARO10 expression following 1-hour of eIF4E-AID depletion. (**) p < 0.05 by Student’s t test. (H) Model for ARO10 upregulation and subsequent feedback in response to translational stress and aromatic amino acid availability. See also Figure S3 and Table S2.

Figure 4:. GCN4 activation in response to eIF4E depletion.

(A) GO analysis for genes which were significantly (adj. p-value < 0.05) up-regulated following eIF4E-depletion (8-hour treatment) in RNA-seq analysis. The most statistically significant entries were chosen and narrowed based on percentage of overlapping genes. (B) Ribosome footprints (adjusted to A-site) over the GCN4 locus for 8-hour IAA and DMSO control cells. Read counts were scaled based on library size. (C) Western blot for Gcn4 levels over the course of IAA depletion. 3-AT (3-amino-1,2,4-triazole) was added for 1 hour. (D) Luciferase assay for Gcn4-Firefly over the course of IAA depletion. Luminescence values were normalized to OD₆₀₀. (**) represents p < 0.05 calculated by Student’s t test. (E) Empirical cumulative distribution function showing relationship between change in mRNA expression following eIF4E depletion for genes categorized as Gcn4 transcriptional targets. P-value was calculated using the Mann-Whitney U test. (F) Western blot for eIF2α phosphorylation levels over the course of IAA depletion. (G) RT-qPCR of PCL5 expression following 1-hour of IAA treatment. (**) represents p < 0.05 calculated by Student’s t test. (H) Western blot for Gcn4 levels over the course of IAA depletion. 3-AT (3-amino-1,2,4-triazole) was added for 1 hour. See also Figure S4.

Figure 5:. Translation regulation of PCL5.

(A) Ribosome footprints (adjusted to A-site) over the PCL5 locus for 8-hour IAA and DMSO control cells. Footprint counts were scaled based on library size and normalized to PCL5 mRNA abundance from paired RNA-seq. (B) Schematic of PCL5-ZEM reporter. (C) RT-qPCR of ZEM transcript and mScarlet reporter expression of PCL5 5ÚTR mutants normalized to 5ÚTR^WT reporter. (**) p < 0.05 by Student’s t test. (D) As in (C), comparing reporter changes following eIF4E-AID depletion, with individual reporters normalized to un-depleted control. (**) p < 0.05 by Student’s t test. (E) As in (D), for eIF4G-AID depletion. See also Figure S5.

Figure 6:. CiBER-seq genetic screen for regulators of PCL5 expression.

(A) Schematic of CiBER-seq screen design. (B) Genome-wide CiBER-seq results showing fold-change (log₂) in P(Z) reporter abundance, relative to P(TFC1) reporter levels for each gRNA in GCN4 and gcn4Δ backgrounds. (C) Genome-wide CiBER-seq screen results showing fold-change (log₂) in P(Z) reporter abundance, relative to P(TFC1) reporter levels for each gRNA in GCN4 background. Line indicates significance cutoff (adjusted p-value < 0.05). (D) As in (C), for gcn4Δ. (E) RT-qPCR of ZEM transcript and mScarlet reporter expression of PCL5 5ÚTR mutants following HTS1 gRNA induction. Individual reporters were normalized to uninduced control. (**) represents p < 0.05 by Student’s t test. (F) As in (E), following PAB1 gRNA induction. See also Figures S6 and S7 and Table S3.

Figure 7:. Model of the role of eIF4E in maintenance of amino acid homeostasis.

(A) Model emphasizing the effects of eIF4E depletion on dysregulation of amino acid metabolism gene expression and mechanisms of PCL5 translational regulation.