Translation initiation factors eIF3 and HCR1 control translation termination and stop codon read-through in yeast cells

Abstract

Translation is divided into initiation, elongation, termination and ribosome recycling. Earlier work implicated several eukaryotic initiation factors (eIFs) in ribosomal recycling in vitro. Here, we uncover roles for HCR1 and eIF3 in translation termination in vivo. A substantial proportion of eIF3, HCR1 and eukaryotic release factor 3 (eRF3) but not eIF5 (a well-defined "initiation-specific" binding partner of eIF3) specifically co-sediments with 80S couples isolated from RNase-treated heavy polysomes in an eRF1-dependent manner, indicating the presence of eIF3 and HCR1 on terminating ribosomes. eIF3 and HCR1 also occur in ribosome- and RNA-free complexes with both eRFs and the recycling factor ABCE1/RLI1. Several eIF3 mutations reduce rates of stop codon read-through and genetically interact with mutant eRFs. In contrast, a slow growing deletion of hcr1 increases read-through and accumulates eRF3 in heavy polysomes in a manner suppressible by overexpressed ABCE1/RLI1. Based on these and other findings we propose that upon stop codon recognition, HCR1 promotes eRF3·GDP ejection from the post-termination complexes to allow binding of its interacting partner ABCE1/RLI1. Furthermore, the fact that high dosage of ABCE1/RLI1 fully suppresses the slow growth phenotype of hcr1Δ as well as its termination but not initiation defects implies that the termination function of HCR1 is more critical for optimal proliferation than its function in translation initiation. Based on these and other observations we suggest that the assignment of HCR1 as a bona fide eIF3 subunit should be reconsidered. Together our work characterizes novel roles of eIF3 and HCR1 in stop codon recognition, defining a communication bridge between the initiation and termination/recycling phases of translation.

Figure 1. Mutations reducing the activity of translation initiation factor eIF3 and HCR1 affect stop codon read-through.

Stop codon read-through was measured using dual luciferase reporter constructs as described in the main text. Plasmid-born mutant alleles of genes encoding eIF3 subunits and HCR1 were introduced into their respective shuffling strains, which are derived from a common strain background (for details please see Text S1 and Table S1). The wt strain background has unusually high levels of UGA read-through, due to the presence of an opal (UGA) suppressor tRNA in the genome. For each independently derived shuffling strain, read-through is shown for pairs of strains shuffled with wt or the indicated mutant alleles of the gene in question. For the non-essential HCR1 subunit, the H3675 hcr1Δ strain is shown. All investigated mutants showed significant (p<0.05) reductions in the level of stop codon read-through, with the exception of tif35^TKMQ, which showed no significant difference, and Δhcr1 which showed strong and significant increase in stop codon read-through.

Figure 2. Increased gene dosage of ABCE/RLI1 suppresses the slow growth and read-through defects of hcr1Δ.

(A) The hcr1Δ strain was transformed with either empty vector (EV), hc HCR1 or hc RLI1. The resulting transformants were subjected to a growth spot assay at 30°C for 2 days. (B) The hcr1Δ strain was transformed with hc vectors carrying either wt or mutant HCR1 and RLI1 alleles, and SUI1 (eIF1) and TIF11 (eIF1A). The resulting transformants were grown in SD and analyzed for stop codon read-through as described in Figure 1. Thus obtained values were normalized to the value obtained with the hcr1Δ strain transformed with wt HCR1, which was set to 100%.

Figure 3. Complexes containing eIF3, HCR1, ABCE1/RLI1 and both eRFs, free of ribosomes and RNA, occur in vivo; and the NTD of g/TIF35 and i/TIF34 directly interact with the N and M domains of eRF1.

(A) WCEs were prepared from YDH353 bearing chromosomal Myc-tagged RLI1 and immunoprecipitated with or without anti-Myc antibodies. The immune complexes were subjected to Western analysis. In, 5% of input; E, 100% of the elution fraction; W, 5% of the supernatant fraction. Also note that anti-RLI1 and -eRF1 antibodies were raised for the purpose of this study. (B) WCEs were prepared from HCHO-treated (1%) cells bearing wt (H2879) or TAP-tagged (H553) chromosomal alleles of HCR1 and incubated with IgG Sepharose 6 Fast Flow beads. The immune complexes were eluted by boiling in the SDS buffer and subjected to Western analysis. In, 1.5% of input; E, 50% of the elution fraction; W, 1.5% of the supernatant fraction. eRF1 is indicated by an asterisk below the immunoglobulins. (C) WCEs from HCHO-treated cells (1%) cells bearing wt (H2879) or TAP-tagged (H555) chromosomal alleles of TIF32 were processed as in panel B except that the immune complexes were eluted by TEV protease cleavage. In, 1.5% of input; E, 100% of the elution fraction; W, 1.5% of the supernatant fraction. (D) WCEs from HCHO-treated cells (1%) cells bearing wt (74D-694) or TAP-tagged (H517) chromosomal alleles of SUP35 were processed as in panel C. (E) Full-length i/TIF34 (lane 3), g/TIF35 (lane 4), and HCR1 (lane 5) fused to GST, and GST alone (lane 2), were tested for binding to ³⁵S-labeled individual domains of eRF1; 10% of input amounts added to each reaction is shown in lane 1 (In). (F) The RRM (lane 3) and N-terminal (lane 4) domains of g/TIF35 fused to GST, and GST alone (lane 2), were tested for binding to ³⁵S-labeled NM domains of eRF1; 10% of input amounts added to each reaction is shown in lane 1.

Figure 4. eIF3 associates with 80S couples isolated from heavy polysomes in an eRF1-dependent manner.

(A) The wt strain (H2819) was grown in SD medium at 30°C to an OD₆₀₀ of ∼1 and cross-linked with 0.5% HCHO prior to harvesting. WCEs were prepared, separated on a 5%–45% sucrose gradient by centrifugation at 39,000 rpm for 2.5 h and two collected fractions containing either disomes and trisomes or pentasomes and heavier polysomes were treated with RNase A to separate the initiating PICs from 80S couples on mRNAs and subjected to the sucrose gradient resedimentation protocol . Two fractions containing 43-48S PICs and 80S ribosomes from each polysomal pool were collected and subjected to Western blot analysis; the ratio of the 80S/40S ratios for heavy over light polysomes was calculated and plotted for each factor. This experiment was repeated four times. (B) eRF1 depletion reduced association of eRF3 and eIF3 with 80S ribosomes isolated from heavy polysomes. The Tet::SUP45 cells were grown in SD medium at 30°C in the presence or absence of 1 µg/ml doxycycline for six hours before harvesting, and treated as described in Figure 4A with the exception that only heavy polysomes were collected after the first round of centrifugation, and only the 80S couples were collected after the second round of centrifugation. Thus obtained samples were subsequently subjected to Western blot analysis; this experiment was conducted three times.

Figure 5. Deletion of hcr1 results in accumulation of eRF3 in heavy polysomes.

(A–C) The hcr1Δ strain (H3675) was transformed with either hc HCR1 (A), empty vector (B), or hc RLI1 (C), and the resulting transformants were grown in SD medium at 30°C to an OD₆₀₀ of ∼1 and cross-linked with 0.5% HCHO prior to harvesting. WCEs were prepared, separated on a 5%–45% sucrose gradient by centrifugation at 39,000 rpm for 2.5 h and subjected to Western blot analysis. Several fractions corresponding to the Top, 40S, 60S, and 80S plus polysomal species were pooled, as indicated. Asterisk indicates a non-specific band. (D) Statistical significance of the eRF3 accumulation in heavy polysomes in the hcr1 strain and its partial recovery by hc RLI1. Amounts of each individual factor in all fractions were quantified by fluorescence imaging. Thus obtained values for the fractions containing heavy polysomes (14–18) as well as all remaining fractions (1–13) were added up for each of these two groups. Values (mean±SE; n = 4) given in the table then represent relative amounts of factors in heavy polysomes divided by the compound value of the rest of the gradient. Changes in the redistribution of factors between the heavy polysomes and lighter fractions in all three strain were analyzed by the student's t-test and shown to be statistically significant only for eRF3 as shown in the table. (E) Statistical significance of the eIF3 shift from 40S-containing fractions to the top, which is independent of the effect of hc RLI1 on eRF3. Essential the same as in panel D, except that the values for the Top fractions (1–4) as well as the 40S fractions (5–6) were added up for each of these two groups. Values (mean±SE; n = 4) given in the table then represent relative amounts of factors in the Top divided by the 40S group. Changes in the redistribution of factors between the 40S and Top fractions in hcr1Δ+EV or +hc RLI1 strains vs. wt were analyzed by the student's t-test and shown to be statistically significant only for eIF3 as shown in the table.

Figure 6. Model of eIF3 and HCR1 involvement in yeast translation termination.

Upon stop codon entry into the ribosomal A-site the pre-TC forms, composed of the canonical release factors eRF1 and eRF3·GTP, and eIF3 and HCR1. eRFs and eIF3 may associate with the pre-TC as a pre-formed unit or alone. In the pre-TC, eIF3 interacts with the N domain of eRF1, via its two small g/TIF35 and i/TI34 subunits, and modulates, perhaps inhibits its stop codon recognition activity during the proofreading step. Upon stop codon recognition the GTP molecule on eRF3 is hydrolyzed. Subsequently, HCR1 promotes eRF3·GDP ejection to allow the ABCE1/RLI1·ATP recruitment to begin the accommodation phase of termination – the eRF1 GGQ motif is pushed to the peptidyl-transferase center (PTC) – during which HCR1 interacts with ABCE1/RLI1. Subsequently, both factors together with eIF3 participate in ribosomal recycling to enable and promote initiation of the next translational cycle (the elongation step is shown only for illustration purposes).

Figure 7. The sup45^Y410S mutation prevents stable association of eRF3 and HCR1 with polyribosomes.

(A–B) The sup45^Y410S mutant and its corresponding wt strain were subjected to HCHO cross-linking (0.5%) and polysomal gradient analysis as described in Figure 5. (C) Statistical significance of the reduction of polysomes-associated amounts of eRF3 and HCR1 in sup45^Y410S. Amounts of each individual factor in all fractions were quantified by fluorescence imaging. Thus obtained values for the Top fractions as well as fractions containing 80S couples and polysomes were added up for each of these two groups. Values (mean±SE; n = 4) given in the table then represent relative amounts of factors in the Top divided by the 80S+polysomes group. Changes in the redistribution of factors between the Top and 80S+polysomes fractions in the sup45^Y410S mutant vs. wt were analyzed by the student's t-test and shown to be statistically significant for HCR1 (P<0.05) and SUP35 (P<0.1).

Figure 8. hcr1Δ and eIF3 mutants genetically interact with release factor mutants.

(A–B) The sup45^Y410S mutation eliminates the negative impact of hcr1Δ on (A) read-through and (B) growth rates. The hcr1Δ strain was crossed with the sup45^Y410S mutant strain and the resulting double mutant was transformed with sc SUP45, hc HCR1, or empty vector (EV), respectively, and (A) processed for stop codon read-through as described in Figure 1 (hcr1Δ read-through values were set to 100%) or (B) subjected to a growth spot assay at indicated temperatures for 2 or 3 days. (C–D) Combining the selected TIF32 mutants with sup35^N536T and sup45^Y410S (C) reduces their read-through defects and (D) produces synthetic growth phenotypes. The wt and mutant alleles of TIF32 were introduced into tif32Δ, sup35^N536T tif32Δ, and sup45^Y410S tif32Δ strains, respectively, by plasmid shuffling. (C) The resulting double mutant strains were grown in SD and processed for the stop codon read-through as described in Figure 1 (the read-through values of both single eRF mutants were set to 100%), or (D) spotted in four serial 10-fold dilutions on SD medium and incubated at indicated temperatures for 4 days. ND; not determined due to severe growth deficiency.