Conserved heterodimeric GTPase Rbg1/Tma46 promotes efficient translation in eukaryotic cells

Abstract

Conserved developmentally regulated guanosine triphosphate (GTP)-binding proteins (Drgs) and their binding partner Drg family regulatory proteins (Dfrps) are important for embryonic development, cellular growth control, differentiation, and proliferation. Here, we report that the yeast Drg1/Dfrp1 ortholog Rbg1/Tma46 facilitates translational initiation, elongation, and termination by suppressing prolonged ribosome pausing. Consistent with the genome-wide observations, deletion of Rbg1 exacerbates the growth defect resulting from translation stalling, and Rbg1 stabilizes mRNAs against no-go decay. Furthermore, we provide a cryoelectron microscopy (cryo-EM) structure of the 80S ribosome bound with Rbg1/Tma46 that reveals the molecular interactions responsible for Rbg1/Tma46 function. The Rbg1 subunit binds to the GTPase association center of the ribosome and the A-tRNA, and the N-terminal zinc finger domain of the Tma46 subunit binds to the 40S, establishing an interaction critical for the ribosomal association. Our results answer the fundamental question of how a paused ribosome resumes translation and show that Drg1/Dfrp1 play a critical role in ensuring orderly translation.

Keywords: Dfrp; Drg; Drg-family regulatory proteins; developmentally-regulated GTP-binding proteins; protein homeostasis; ribosome; ribosome stalling; translation.

Figure 1.. Rbg proteins play important roles in global translation

(A) Model of the auxin-inducible-degron system.To remove the Rbg1 protein from the cell, a GAL10 promoter was inserted into the 5′ UTR of the rbg1 gene so that Rbg1 transcription levels can be controlled by the presence of glucose in the media. A mini auxin-inducible degron (mAID) tag was fused to the N terminus of Rbg1. Cells were first grown in YPGR medium and then changed to YEPD medium containing 1 mM auxin to induce the degradation of the Rbg1 protein in the cell. Aliquots at different time points were acquired, and the Rbg1 protein levels were monitored by western blot with an anti-FLAG antibody. With the mAID degron tag, the Rbg1 protein degrades after 20 min, but its expression is restored after 45 min. The western blot represents one of three biological replicates. (B) Elimination of Rbg1 under Δrbg2Δslh1 background shows a temperature-independent growth defect. Serial dilutions of liquid cultures growing in exponential phase were spotted on YEPD plates containing 1 mM auxin and were incubated for 2 days at 30°C or 37°C or for 4 days at 19°C. (C) Polysome profiles from cells incubated for20 min in YEPD indicate that Rbg1 plays an important role in global translation. Cells were first grown in YPGR medium to an optical density 600 (OD₆₀₀) of 0.6 and then were changed to YEPD medium containing 1 mM auxin for 20 min. Cells were then quickly harvested and polysome profiling was performed. The profiles of wild-type (WT) and Δrbg2 strains are not shown here because WT, Δrbg2, and Δrbg1 have nearly identical profiles. These data represent one of three biological replicates.

Figure 2.. Deletion of Rbg1 results in slow translation initiation, elongation, and termination

(A) Translation initiation arrests in mutant cells. Normal translation pauses at start codons in WT cells; and increased ribosome pauses at start codons in Δrbg1, Δrbg2Δslh1, and Δrbg2Δslh1-Rbg1d cells. Metagene analysis displays the abundance of 5′P reads relative to start codons for WT, Δrbg1, Δrbg2Δslh1, and Δrbg2Δslh1-Rbg1d strains or after random fragmentation (5P-Seq control, dotted black line). Reads are represented by rpm, with the blue bar indicating the +1 frame and light blue bars indicating the 0 and +2 frames. The red peak at −14 nt corresponds to the protected region from a putative initiation-paused ribosome, and the blue peak at 4 nt is caused by peptide-induced ribosomal arrest. Biological replicates are averaged. (B) Effects on translation termination in mutant cells. Normal translation pauses at stop codons in WT cells and altered pauses in the three mutants Δrbg1, Δrbg2Δslh1, and Δrbg2Δslh1-Rbg1d. Red peaks at the −17-nt position denote ribosomes with a stop codon in the A-site; and the pause at the ‒50-nt position is indicative of a disome position, with the leading ribosome reaching translation termination. Other experimental details are the same as described in (A). (C) Slower translational elongation after depletion of Rbg1, Rbg2, and Slh1. i. Representative genome tracks of the 5′ ends of 5P-Seq reads in WT (black), Δrbg1 (red), Δrbg2Δslh1 (green), and Δrbg2Δslh1-Rbg1d strains (blue). Coverage is expressed in rpm. An obvious 3-nt periodicity pattern was observed for YER177W mRNA in the Δrbg2Δslh1-Rbg1d strain. ii–iv. The proportion of 5P-Seq reads in the ribosome-protected frame (frame 1 in Figure S2) shows that the 5′ end of coding regions was protected by ribosomes in the Δrbg2Δslh1-Rbg1d strain. Proportion scores were calculated for each codon by normalizing the read counts in frame 1 to the total reads in all three frames within the same codon. These are shown as dots with smoothed lines (polynomial fitting) for genes longer than 1,000 base pairs (bp) and with reads per million per kilobase (RPKM) values of >20. Only the region containing ‒20 nt to 501 nt with respect to the first base of start codon was used to calculate the proportion score. 5P-Seq samples from Drbg1 (ii, red), Δrbg2Δslh1 (iii, green), and Δrbg2Δslh1-Rbg1d (iv, blue) strains were compared to WT strains (black). Random fragmentation samples (control) are also shown. Biological replicates are averaged.

Figure 3.. Deletion of Rbg1 results in ribosome pauses at certain amino acids and R/K-rich regions in the mRNA

(A) Increased translation pauses at arginine, lysine, glutamic acid, and aspartic acids positions in the absence of Rbg1. The metagene (−40 to +10 window) shows the number of 5P intermediates from translation of arginine, lysine, glutamic acid, and aspartic acids amino acids and their ribosomal positions in Δrbg1 (red lines) and WT (black lines) cells. The total number of reads was normalized to facilitate data analysis and comparison, and data were analyzed as described (Pelechano and Alepuz, 2017). The peaks at ‒17 nt, ‒14 nt, and ‒11 nt represent indicated codons at the ribosomal A-, P-, and E-sites, respectively. The normalized rpms for each amino acid between the WT and Δrbg1 were compared, and the adjusted p values at the −17 position were shown. The adjusted p value was calculated using the Benjamin-Hochberg method in R package called DESeq2 (Love et al., 2014). The adjusted p < 0.01 was used as the criterion for a significantly regulated difference in the data. (B) Rbg1 alleviates ribosome pausing at arginine/lysine-rich regions. i. Representative genome tracks of 5′ ends of 5P-Seq reads. Genome tracks of the 5′ ends of 5P-Seq reads in WT (black), Δrbg1 (red), Δrbg2Δslh1 (green), and Δrbg2Δslh1-Rbg1d (blue) strains were shown around the arginine-/lysine-rich region. Coverage is expressed in average rpm of the biological duplicates. ii–iv. Translation pauses at the R-/K-rich region in Δrbg1, Δrbg2Δslh1, and Δrbg2Δslh1-Rbg1d strains. To identify specific peptide sequences that induce ribosome pausing while being translated in the knockout strains, pause scores were first calculated by dividing the rpm value at each position of the transcript by the mean rpm value for the 10 codons upstream and downstream of the same position. The 10 amino acids upstream with a pause score of >10 were compared to those with pause scores of <10 by using MEME (Bailey et al., 2009). E-values were defined by MEME. Amino acid sequences extracted from the WT strain show no consensus sequence.

Figure 4.. Rbg1 suppresses ribosome pauses and promotes efficient translation

(A) Deletion of Rbg1 results in the slower growth of yeast cells in the presence of anisomycin. Serial dilutions of liquid cultures growing in exponential phase were spotted on YEPD, in the absence or presence of different concentrations of anisomycin. Cells were incubated at 30°C, and pictures were taken after 2 days (without anisomycin) or 3 days (with anisomycin). (B) Rbg1 enhances translation of mRNAs harboring an intrinsic stalling sequence. Yeast cells expressing the indicated mRNA with the R12 sequence, fluorescent mCherry, and GFP reporters (top panel) were grown to exponential phase (OD₆₀₀, 0.6) and were analyzed by flow cytometry. mCherry and GFP fluorescence intensities in the cell were monitored simultaneously by using 561-nm and 488-nm excitation lasers, respectively. Scatterplots of 10,000 individual cells of the WT, Δrbg1, and Δrbg2Δslh1 background are displayed separately. The scatterplots are shown on a bi-exponential scale with pseudo-color in order to better visualize data across the wide range of expression levels seen in these experiments. According to the control cells that do not express the indicated mRNA construct (data not shown), only the cells with more than 10³ relative fluorescence unit (RFU) of mCherry fluorescence are considered to contain enough mRNAs and are subjected to further analysis. WT, Δrbg1, and Δrbg2Δslh1 cells with >10³ RFU mCherry intensities are divided into two groups and shown as percentages in gray dashed lines. Cells in the dashed box have high mCherry expression levels but low GFP expression levels, the amount of which in WT, Δbg1, and Δrbg2Δslh1 is shown accordingly as percentages (red) (ii, iii, and iv). The relative cell quantities at each position are indicated using a red-to-blue spectrum, for which the red color represents the largest cell population and blue represents the smallest one. v. Normalized reporter GFP levels. For WT, Δrbg1, and Δrbg2Δslh1 cells with >10³ RFU mCherry intensities, the GFP intensity is normalized to the corresponding mCherry intensity. The ratio of GFP/mCherry intensities obtained from flow cytometry data in WT, Δrbg1, and Δrbg2Δslh1 are compared using violin plots (center dots, median; boxes, 25th to 75th percentiles; whiskers, 1.5× interquartile range [IQR]). Median values for the GFP/mCherry ratios in WT, Δrbg1, and Δrbg2Δslh1 are 0.200, 0.158, and 0.482, respectively.

Figure 5.. Structural basis for enhanced translation when Rbg1/Tam46 binds to a paused ribosome

(A) An overall structure of Rbg1/Tma46-bound ribosomes. The ribosome is colored in gray. P-tRNAs, A-tRNAs-nascent peptides, eIF5A, Rbg1, and Tma46 are colored in green, pink, magenta, red, and blue, respectively. (B–E) Detailed essential interactions of the Rbg1 G domain with the sarcin-ricin loop (B), S5D2L domain with A-tRNA (C), TGS domain with h5 in the 40S subunit (D), and the second zinc finger in Tma46 with the head and shoulder regions of the 40S subunit (E) are shown. The insert in (B) shows conformational changes in the G domain of Rbg1 in this study (red surface) and eEF1A (gray surface, with GDP and Didemnin B bound, PDB: 5LZS; Shao et al., 2016) on the ribosome when the 25S rRNAs are aligned. Helices α3–α5 are shown in Rbg1 and eEF1A to illustrate conformational changes of the G domain in the ribosome. (F) The Rbg1/Tma46 complex binds to the ribosome by the zinc-finger domain of Tma46. Proteins co-expressed in the Escherichia coli BL21 strain were purified and incubated with pure 80S ribosomes, and then ribosomal complexes were separated by sedimentation through a 10%–50% sucrose gradient. Binding of Rbg1/Tma46 to the ribosome was determined by western blot. FLAG-tagged RPL25 was immunoblotted on the same membrane. Both the GTP- and GDP-bound heterodimeric GTPases bound to the 80S ribosome, but only the GTP-bound state is shown. The western blot represents one of three biological replicates. (G) Conformation of eIF5A in the Rbg1/Tma46-bound ribosomes. eIF5A binds to the E-site in the Rbg1/Tma46-bound ribosomal complex. eIF5A (purple), P-tRNA (green), and A-tRNA (pink) are represented as cartoons. The site of hypusination in eIF5A can be determined according to the density in the complex. Ribosome proteins are removed for clarity, and 25S rRNA (cyan) and 18S rRNA (orange) are shown.

Figure 6.. Functional interplay of Drg/Dfrp and ASCC3 in the translation pause-and-resume and quality-control pathways

When translation pauses, Drg/Dfrp senses the paused ribosome, stabilizes ribosomes in the productive conformation, and promotes efficient translation, thereby allowing ribosomes to continue translating the mRNA. In the absence of Drg/Dfrp, another protein called ASCC3 binds the stalled ribosome and triggers RQC involving ribosome subunit disassociation, mRNA degradation, and no-go decay.