The C-terminal region of eukaryotic translation initiation factor 3a (eIF3a) promotes mRNA recruitment, scanning, and, together with eIF3j and the eIF3b RNA recognition motif, selection of AUG start codons

Abstract

The C-terminal domain (CTD) of the a/Tif32 subunit of budding yeast eukaryotic translation initiation factor 3 (eIF3) interacts with eIF3 subunits j/Hcr1 and b/Prt1 and can bind helices 16 to 18 of 18S rRNA, suggesting proximity to the mRNA entry channel of the 40S subunit. We have identified substitutions in the conserved Lys-Glu-Arg-Arg (KERR) motif and in residues of the nearby box6 element of the a/Tif32 CTD that impair mRNA recruitment by 43S preinitiation complexes (PICs) and confer phenotypes indicating defects in scanning and start codon recognition. The normally dispensable CTD of j/Hcr1 is required for its binding to a/Tif32 and to mitigate the growth defects of these a/Tif32 mutants, indicating physical and functional interactions between these two domains. The a/Tif32 CTD and the j/Hcr1 N-terminal domain (NTD) also interact with the RNA recognition motif (RRM) in b/Prt1, and mutations in both subunits that disrupt their interactions with the RRM increase leaky scanning of an AUG codon. These results, and our demonstration that the extreme CTD of a/Tif32 binds to Rps2 and Rps3, lead us to propose that the a/Tif32 CTD directly stabilizes 43S subunit-mRNA interaction and that the b/Prt1-RRM-j/Hcr1-a/Tif32-CTD module binds near the mRNA entry channel and regulates the transition between scanning-conducive and initiation-competent conformations of the PIC.

FIG. 1.

a/HLD substitutions impair cell growth. (A) Schematic representation of a/Tif32, indicating N- and C-terminal halves of the j/Hcr1-like domain (HLD). The CTD of a/Tif32 subjected to mutagenesis is indicated as a bar and string of ovals (boxes 1 to 15) representing 10 consecutive residues replaced by alanines. Shades of gray indicate the degree of Slg⁻ phenotype, and black indicates lethality. A sequence alignment of the C-terminal portion of the a/HLD with j/Hcr1 (residues 168 to 222) is shown, indicating the nature of tif32 mutations box6, box9, H725P, and R731I above and the KERR motif below. S.c., S. cerevisiae. Asterisks, amino acids conserved in all sequences; colons, homologous substitutions; periods, nonhomologous substitutions; boldface, box6 region and KERR motif conserved residues. (B) ClustalW2 was used to align regions of j/Hcr1 and the a/HLD using sequences from the indicated species (K.lactis, Kluyveromyces lactis; H.sapiens, Homo sapiens; M.musculus, Mus musculus; B.taurus, Bos taurus; S.pombe, Schizosaccharomyces pombe; A.thaliana, Arabidopsis thaliana) with the following accession numbers (in the order shown, from the top): NP_013293, Q6CMJ8, O75822, Q66JS6, Q0VCU8, NP_009635, CAA21076, Q9LD55, and Q14152. (C) Phenotypes of tif32 a/HLD mutants. Serial dilutions of GCN2 his4-301 tif32Δ strains harboring lc plasmids with TIF32⁺-His (WLCY01), tif32-H725P-His (WLCY02), tif32-R731I-His (WLCY03), or tif32-box6-His (H3715) were spotted on synthetic dextrose minimal medium supplemented with histidine, tryptophan, and uracil (SD+HWU) and incubated at the indicated temperatures for 3 days.

FIG. 2.

a/HLD substitutions impair translation initiation. (A) Polysome profiles of the strains shown in Fig. 1C cultured in YPD medium at 30°C and shifted to 36°C for 6 h, with cycloheximide added just prior to harvesting. WCEs were separated by velocity sedimentation through 4.5 to 45% sucrose gradient centrifugation, and fractions were collected while scanning at 254 nm to visualize ribosomal species and determine P/M ratios. (B) Western analysis of a/Tif32 proteins. (Left) WCEs were prepared from the strains in panel A after being cultured in YPD at 30°C or after being shifted to 36°C for 6 h. Aliquots (2.5, 5, and 10 μl in successive lanes) were separated by SDS-PAGE and subjected to Western analysis with monoclonal antibodies against His₈ epitope or polyclonal antibodies against Gcd6. The amounts of His₈-a/Tif32 were normalized to the Gcd6 amounts measured in the same lanes, and the resulting ratios were normalized to those measured in WT cells (set to 1.0). The mean and standard errors calculated from replicate determinations are plotted in the histogram below. (Right) Western analysis of WCEs from transformants of TIF32⁺ strain H2994 harboring lc WT TIF32-His or lc or hc tif32-box9-His and cultured in synthetic complete medium lacking uracil at 30°C. (Because of its lethality, expression of His₈-a/Tif32-box9 was examined in cells containing TIF32⁺).

FIG. 3.

a/Tif32 KERR substitutions diminish native 48S PICs containing RPL41A mRNA. (A to C) Strains shown in Fig. 1C were grown in YPD at 30°C and shifted to 36°C for 6 h, and the cells were cross-linked with HCHO for 1 h prior to being harvested. WCEs were sedimented through 7.5 to 30% sucrose gradients, and fractions were subjected to Western analysis with antibodies against the indicated proteins. The amounts of each factor in the 40S fractions (boxed) were normalized for the Rps22 level, and the ratios of the eIF/40S levels in the mutant to those in the WT were plotted in the adjoining histograms (means ± standard errors [SE]; n = 3). (D) (Left) rpl11bΔ strains isogenic to those in panels A to C were cultured and cross-linked as described above. Total RNA was extracted from each fraction, and the amounts of 18S rRNA and RPL41A mRNA were measured by real-time quantitative PCR (qPCR). The amounts of mRNA were calculated as 2^−CT × 10⁻⁷ for RPL41A mRNA and 2^−CT × 10⁻⁴ for 18S rRNA. (Right) The ratio of RPL41A mRNA in fractions 10 to 12 to 18S rRNA in fractions 9 to 11 was calculated for each mutant and normalized to the corresponding value for WT (means ± SE; n = 4). Student's t test indicated that the value for each mutant differed significantly from that for the WT (P < 0.01).

FIG. 4.

a/HLD mutations impair derepression of GCN4 translation and increase leaky scanning of uAUG-1. (A) a/HLD mutants have Gcn⁻ phenotypes. Serial dilutions of the strains in Fig. 1C were spotted on SD+HWU and SD+HWU containing 0.5 μg/ml SM and incubated at 30°C, 33°C, or 36°C for 3 days. (B) a/HLD mutations alter expression of GCN4-lacZ reporters. Shown are the strains in Fig. 1C harboring GCN4-lacZ reporter plasmid p180 (i), p227 (ii), pM226 (iii), or p209 (iv), containing the 5′ UTR configurations shown schematically, with Xs indicating AUG mutations in uORFs. Transformants were grown in SD+HWU (−) or SD+HWU containing 0.5 μg/ml SM (+), as indicated in the left-hand column, at 33°C for 6 h, and β-galactosidase activity (in nmol of o-nitrophenyl-β-d-galactopyranoside cleaved per min per mg) was assayed in WCEs. The means and SE from four independent transformants are reported, along with the means expressed as a percentage of the corresponding WT value. The values in the column “box6 norm.” are the results from column “box6” normalized to correct for the different expression of construct ii (without SM) in box6 versus WT cells.

FIG. 5.

Evidence that a/HLD substitutions confer scanning defects in vivo. (A) Effects of a/HLD mutations on expression of LUC reporters with 5′ UTRs of different lengths. The yeast strains in Fig. 1C and H3774 (ded1-57), harboring the plasmid-borne L0LUC or L2LUC reporters under the control of the GAL1 promoter with the indicated 5′ UTR lengths, were grown in SD supplemented with adenine, histidine, and tryptophan at 30°C to an A₆₀₀ of ∼0.6, shifted to the same medium containing galactose instead of glucose, and incubated at 36°C for 6 h. Luciferase activities were assayed in WCEs, and the means and SE from 6 independent transformants for L2LUC were normalized to the corresponding values for L0LUC. (B) The yeast strains from panel A were transformed with GCN4-lacZ reporter plasmid p227 (i), pWCB07 (ii), or pWCB06 (iii), indicated schematically as in Fig. 4B, except that constructs ii and iii contained the indicated sequences inserted 21 nt 5′ of the GCN4 AUG codon, with complementary bases underlined. The cells were cultured in SD+HW at 33°C for 6 h, and β-galactosidase activities were assayed in WCEs of four independent transformants. Means ± SE (n = 4) and activities as percentages of WT values are indicated. The values in the columns “box6 norm.” and “ded1 norm.” are the results from columns “box6” and “ded1-57” normalized to correct for the different expression levels of construct i in box6 or ded1-57 versus WT cells.

FIG. 6.

a/HLD mutations suppress UUG initiation in Sui⁻ mutants. (A) Suppression of His⁺/Sui⁻ phenotypes of SUI5. Serial dilutions of his4-301 strains in Fig. 1C harboring empty vector or a SUI5 plasmid were grown at 30°C on SD+H for 3 days or SD for 7 days. (B and C) Suppression of increased UUG/AUG initiation ratio. Strains from Fig. 1C carrying empty vector or the SUI5, SUI3-2, or tif11Δ125-153 plasmid and harboring HIS4-lacZ reporter plasmids with an AUG or UUG start codon were grown in SD+H medium and assayed for β-galactosidase activity in WCEs. The mean ratios and SE of UUG versus AUG reporter expression from six independent transformants are shown, with the fold increases relative to the WT for the vector transformants of each strain in the column “vector” and the fold increases relative to vector for each plasmid-borne Sui⁻ allele in each strain in the remaining three columns.

FIG. 7.

a/HLD substitutions weaken interactions of eIF3 with j/Hcr1 and eIF2. (A) Strains described in Fig. 1C were cultured in YPD medium at 30°C and shifted to 36°C for 6 h. WCEs were incubated with Ni-NTA-silica resin, and bound proteins were eluted and subjected to Western blot analysis with antibodies against the His₈ epitope (for a/Tif32) or against the factors listed on the right. Three percent of input WCE (In), 15% and 30% of the eluate (1× and 2×), and 3% of flowthrough (FT) were analyzed in successive lanes. (B) Western signals for each factor in the eluates from panel A were normalized to that of His₈-a/Tif32, and the ratio of each tif32 mutant to WT was plotted (means ± SE; n = 3).

FIG. 8.

Overexpression of j/Hcr1 partially suppresses the Ts⁻ phenotypes of tif32 KERR mutants and restores j/Hcr1 association with eIF3. (A) Serial dilutions of the strains from Fig. 1C harboring an hc empty vector or hc HCR1 plasmid were spotted on SD+HWU and incubated at 30°C or 36°C for 3 days. (B to E) Nickel chelation chromatography of WCEs from strains in panel A was conducted as for Fig. 7A. Western signals were quantified, and the ratios of normalized values for hc HCR1 versus empty vector are plotted in the histograms (means ± SE; n = 3).

FIG. 9.

The NTD, box6, and KERR motif in the CTD of j/Hcr1 are required for its association with eIF3. (A) WCEs were prepared from a transformant of YAH05 (prt1Δ hcr1Δ pRS316-PRT1 [PRT1 URA3]) containing plasmid YEp-j/HCR1-DS (lanes 1 to 4) or from YAH05 derivatives lacking pRS316-PRT1 and containing pRS-b/PRT1-His and YEp-j/HCR1-DS-U (lanes 5 to 8), pRS-b/PRT1-His and YEp-j/hcr1-NTD-U (lanes 9 to 12), pRS-b/PRT1-His and YEp-j/hcr1-CTD-U (lanes 13 to 16), or pRS-b/PRT1-His and YEp-j/hcr1-box6-U (lanes 17 to 20), cultured in SD at 30°C. Nickel chelation chromatography and Western blot analysis were conducted as for Fig. 7A, except that 5% of input WCEs (In), 30% (E1) or 60% (E2) of the eluates, and 5% of flowthrough (FT) was loaded. Mean Western signals were normalized to those of b/Prt1-His₈ and plotted as percentages of the corresponding values calculated for the HCR1⁺ strain. (B) Same as panel A, except that WCEs were prepared from a YAH05 transformant harboring YCp-j/HCR1-DS-L (lanes 1 to 4) or YAH05 derivatives lacking pRS316-PRT1 and containing pRS-b/PRT1-His and YCp-j/HCR1-DS-U (lanes 5 to 8) or pRS-b/PRT1-His and YCp-j/hcr1-R215I-U (lanes 9 to 12).

FIG. 10.

Effects of box6 and KERR substitutions in j/CTD and a/HLD on binary interactions in the b/RRM-j/Hcr1-a/Tif32-CTD module and leaky scanning of GCN4 uAUG-1. (A) Schematic of j/Hcr1 showing the positions of the N-terminal acidic (nta) and KERR (kerr) motifs, with arrows delimiting minimal binding domains for the indicated proteins. Shown below are sequence alignments of the box6 and KERR segments of j/Hcr1 and a/Tif32, indicating identical (*) or conserved (:) positions, the KERR residues (underlined), and substitutions present in the indicated hcr1 mutants. Lowercase letters indicate j/Hcr1 residues that were subjected to site-directed mutagenesis. (B) Summary of molecular interactions of components of the b/RRM-j/Hcr1-a/Tif32-CTD module. The shaded rectangles represent the b/RRM, a/HLD, and full-length j/Hcr1. The solid arrow indicates interaction between the NTA motif of j/Hcr1 and helix α1 and loop L5 of the b/RRM, whose structural determinants are known. The dashed arrows depict other interactions that were mapped previously or determined in this study. The results presented here establish the roles of box6 and KERR residues in the a/HLD in binding to the b/RRM and of the equivalent residues in j/CTD in binding a/Tif32 regions flanking the a/HLD. (C) Both the NTD and CTD of j/Hcr1 are required for binding to a/Tif32 in vitro. GST fusions to full-length j/Hcr1 (lane 3), the j/NTD (lane 4) or j/CTD (lane 5), or GST alone (lane 2), were tested for binding to full-length ³⁵S-a/TIF32 in pulldown assays. The GST fusions visualized by Coomassie blue staining (top) and the ³⁵S-a/TIF32 visualized by autoradiography (bottom) in the bound fractions are shown in lanes 2 to 5. Lane 1 contains 10% of the input ³⁵S-a/TIF32. (D) The KERR motif and box6 of j/Hcr1 are critical for its binding to a/Tif32. Same as panel C, except that GST fusions to full-length j/Hcr1 (lane 3) or its box6 (lane 4), box9 (lane 5), or R215I (lane 6) mutant derivatives were examined. (E) Substitutions in the KERR motif and box6 in the a/HLD strongly reduce its binding to the b/RRM. A GST fusion to the b/RRM (aa 1 to 136) or GST alone was tested for binding to the indicated WT and mutant derivatives of ³⁵S-a/HLD (aa 490 to 790). (F) Leaky-scanning phenotypes of substitutions in box6 or KERR residues of j/Hcr1 or a/Tif32. (Top) Transformants of hcr1Δ strain H428 bearing plasmid YEp-j/HCR1, YEp-j/hcr1-NTD, YEp-j/hcr1-R215I, YEp-j/hcr1-Box6, or YEp-j/hcr1-Box6-R215I and containing GCN4-lacZ reporter plasmid pM226 were analyzed for β-galactosidase activities as for Fig. 4B (iii). Mean values and standard errors from 6 or more measurements of three transformants containing pM226 are shown, along with activities in the hcr1 strains normalized to that in the HCR1 strain. All values were normalized to correct for any differences among the strains in expression of the reporter lacking uORF on p227. (Bottom) Same as above, but using AY51 (TIF32-His) or AY52 (tif32-R731I-His) transformants bearing YEp-j/HCR1-W or YEp-j/hcr1-R215I-DS-W, respectively.

FIG. 11.

a/Tif32 CTD interacts with Rps2 and Rps3 in vitro: a hypothetical model for binding of a/Tif32 CTD near the mRNA entry channel pore of the 40S subunit. (A and B) The extreme CTD of a/Tif32 interacts with Rps2 and Rps3 in vitro. (A) A GST fusion to the a/Tif32-CTD (aa 791 to 964) or GST alone was tested for binding to ³⁵S-labeled full-length Rps2, Rps3, Rps0, or Rps23, as shown in Fig. 10C. (B) Same as panel A, except that Rps2 fused to GST was examined for binding to ³⁵S-a/Tif32. (C) Hypothetical location of eIF3 on the solvent side of the S. cerevisiae 40S subunit based on cryo-EM reconstruction (adapted from reference 47). The 40S subunit is shown from the solvent side, with RNA segments in yellow and proteins in green. The positions of Rps2, Rps3, helices 16 and 18 of 18S rRNA, the a/HLD and CTD of a/Tif32, the b/RRM, and j/Hcr1 are highlighted in color and/or boldface. The mRNA exit channel is indicated by an arrow. The blue lines represent mRNA. The positions of Rps2 and Rps3 were modified according to reference .