eIF2D promotes 40S ribosomal subunit recycling during intrinsic ribosome destabilization

Figures

• 

  Figure 1.. Decreased TE of transcripts containing the D/E motif in eIF2D KO cells. (A) Immunoblotting analysis of HEK293T cells deficient in eIF2D (eIF2D KO), MCTS1 (MCTS1 KO), or DENR [DENR knockdown (KD)] and wild-type (WT) cells using the indicated antibodies. Heat shock protein (HSP)-90 was used as a loading control. (B) Number of transcripts with significantly altered TE (adjusted P-value <.05) in eIF2D KO, eIF2A KO, MCTS1 KO, and DENR KD cells compared to that in the control cells (n = 2 biologically independent samples). Transcripts not commonly downregulated or upregulated in two independent clones or with two different siRNAs were excluded. (C) Amino acid motif enriched in transcripts with significantly downregulated TE in eIF2D KO cells. (D) Schematic representation of the IRD process. (E) Histograms of TE fold changes in eIF2D KO cells relative to the control cells. Results are shown for randomly selected, short nascent chain, and long nascent chain transcripts. Effect sizes [Cliff’s delta (d)] and P-values (two-tailed Mann–Whitney U test) of these transcripts are indicated. (F) Effect sizes [Cliff’s delta (d)] in the indicated cell lines compared to those in the control cells.

• 

  Figure 2.. Loss of eIF2D leads to the accumulation of 40S ribosomes with subsequent stalling of 80S ribosomes during IRD. (A) Schematic representation of 40S and 80S ribosome footprinting. Normalized metagene plots of the 40S (B, D) and 80S (C, E) ribosome footprints on mRNAs with D/E-rich sequences in HEK293T cells deficient in eIF2D (eIF2D KO) (B, C) or MCTS1 (MCTS1 KO) (D, E) and WT cells (n = 2 biologically independent samples). For each transcript, ribosome footprints were aggregated within a window extending from –150 to +250 nt relative to the start of the D/E-rich motif. Read counts within this region were normalized by the total number of reads across the entire window for each sample to account for differences in the sequencing depth. Analyses were performed separately for transcripts encoding short (n = 455) and long (n = 4974) nascent chains. The average of two replicates (dots) is shown. Yellow areas show the positions of D/E-rich sequences. Normalized metagene plots of the 5′-ends of the 40S (F) and 80S (H) ribosome footprints on mRNAs with D/E-rich sequences in HEK293T cells deficient in eIF2D (eIF2D KO) and WT cells, along with the corresponding summed read counts. The 40S…

• 

  Figure 3.. eIF2D preferentially binds to 40S ribosomes during IRD. (A) Schematic representation of the Sel-TCP-seq protocol for analyzing eIF2D-, MCTS1-, and DENR-V5-bound 40S ribosomes. IP, immunoprecipitation. (B) Normalized metagene plots of 40S ribosome footprints in eIF2D KO #1 cells stably expressing eIF2D-V5, untagged eIF2D, or an empty vector control. Footprints were aggregated within a –150 to +250 nt window relative to the start of the D/E-rich motif and normalized to the total read count within the window for each sample. Only transcripts encoding short nascent chains (n = 455) were analyzed. The average of two replicates is shown. Yellow areas show the positions of D/E-rich sequences. (C) Normalized metagene plots of the footprints of eIF2D-, MCTS1-, or DENR-V5-bound 40S ribosomes on mRNAs with D/E-rich sequences in HEK293T cells (n = 2 biologically independent samples). For each transcript, footprints were aggregated within a window extending from –150 to +250 nt relative to the start of the D/E-rich motif. Read counts within this region were normalized by the total number of reads across the entire window for each sample to account for differences in the sequencing depth.…

• 

  Figure 4.. eIF2D-dependent regulation by D/E-rich motifs is independent of the uORFs. (A) NH2-terminal amino acid sequences of BRISC and BRCA1 A complex member 1 (BABAM1), X-ray repair cross-complementing 6 (XRCC6), and proline-rich mitotic checkpoint control factor (PRCC). Acidic amino acids are indicated in red. The 10-amino acid window with seven or more D/E residues is indicated in yellow. Read aggregation plots of 40S and 80S ribosome footprints in HEK293T WT and eIF2D KO cells and eIF2D-, MCTS1-, or DENR-V5-bound 40S ribosome footprints for BABAM1 (B), XRCC6 (C), and PRCC (D). The average of two replicates is shown. Black dotted lines indicate the canonical translation initiation sites. uORFs predicted by RibORF are also shown. Pink shading marks the positions of D/E codons. RPM, reads per million. (E) Schematic of dual-luciferase reporter constructs containing WT or mutant BABAM1 NH2-terminal sequences. Acidic residues are indicated in red; yellow shading marks the 10-residue windows with ≥7 D/E residues. (F) Luciferase assay comparing the TEs of the BABAM1 (WT) and BABAM1 (Mut) reporter constructs in WT and eIF2D KO HEK293T cells (n = 3 biologically independent samples). ***P < .001…

• 

  Figure 5.. Role of the WH domain in 40S recycling. (A) Domain architecture of human eIF2D, MCTS1, and DENR. (B) Structural comparison of 40S ribosomes bound by the MCTS1–DENR complex (PDB: 5VYC) or eIF2D (PDB: 5OA3). WH domain of eIF2D sterically clashes with the FeS cluster of ABCE1 in the post-splitting complex (PDB: 6ZVJ). MCTS1–DENR complex does not exhibit this clash. (C) Close-up view of the eIF2D WH domain in contact with h44 of the 18S rRNA. (D) Immunoblotting analysis of the 40S ribosome fractions and input lysates of eIF2D KO #1 cells stably expressing eIF2D-V5 or eIF2D(ΔWH)-V5 using the anti-V5 antibody. (E) Quantification of V5 band intensity via densitometry (n = 3 biologically independent samples). *P < .05 and **P < .01 (two-tailed Welch’s t-test). (F) Normalized metagene plots of 40S ribosome footprints in eIF2D KO #1 cells expressing eIF2D-V5, eIF2D(ΔWH)-V5, or an empty vector. For each transcript, footprints were aggregated across a –150 to +250 nt window relative to the start of the D/E-rich motif and normalized to the total read counts within that window. Plots represent short nascent chain transcripts (n = 455). The average of two replicates is shown. Yellow shading…

• 

  Figure 6.. Decreased IRD-inducing protein levels and altered splicing patterns in eIF2D-deficient cells. (A) Cumulative fraction of the fold changes in protein levels in eIF2D KO cells relative to the control cells (n = 4 biologically independent samples). Protein levels were analyzed via DIA-MS, and the results are shown for all detected, short nascent chain, and long nascent chain proteins. P-values were calculated using the two-tailed Mann–Whitney U test. (B) Differences in the median log2 fold changes in protein levels between the indicated cell lines and control cells. **P < .01 and ***P < .001; NS, not significant (two-tailed Mann–Whitney U test). (C) Gene Ontology (GO) term analysis of IRD target genes with reduced protein levels in eIF2D KO cells. FDR, false discovery rate. (D) Volcano plots showing the differences in the PSI scores of skipped exons between the eIF2D KO and control cells (n = 2 biologically independent samples). Gray dashed lines indicate a ∆PSI of ±0.05 and an FDR of 0.05. (E) Numbers of skipped and retained introns in the eIF2D KO, eIF2A KO, and MCTS1 KO cells compared to those isplicing, and transcripn the control cells. Alternative splicing events not…