Ageing exacerbates ribosome pausing to disrupt cotranslational proteostasis

Strains and growth conditions.

The Bristol N2 strain of C. elegans was grown at 20°C on nematode growth medium agar plates seeded with Escherichia coli strain OP50, following standard methods⁶⁷. All yeast experiments were performed using derivatives of BY4741 (MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0). Ribo-Seq experiments used diploid BY4743 (WT) and homozygous sch9Δ/sch9Δ (YSC6275–201917395), which were obtained from GE Healthcare Dharmacon. For immunoblotting and microscopy experiments, haploid strains of WT, ltn1Δ, rqc2Δ, hel2Δ, and sch9Δ were from the Yeast Knockout Collection⁶⁸. Strains from the N-terminally GFP tagged seamless collection⁶⁹ as well as from the C-terminally GFP tagged clone collection⁷⁰ were used for the YTM1 and HAT2 experiments. Yeast cells were grown at 30°C in variations of synthetic complete (SC) media⁶² with fivefold excess of nutrients to compensate for strain auxotrophies.

Plasmid and yeast strain construction.

Ribosome pausing reporter plasmids were derived from gifts from Onn Brandman⁵² and Toshifumi Inada²². The PGK1 promoter was amplified using oligonucleotides 5’GCGGAGCTCTGTTTGCAAAAAGAACAAAAC and 5’GCGTCTAGATGTTTTATATTTGTTGTAAAAAG, digested with SacI/XbaI, and ligated to create p416PGK-GFP-R12-FLAG-HIS3 and p416PGK-GFP-K12(AAA)-FLAG-HIS3. Correct clones were confirmed by colony PCR and sequencing. Experiments using GFP-R12-RFP and GFP-K12-RFP pausing reporters were performed after integrating these reporters into yeast. First, RFP was amplified from pTDH3-GFP-R12-RFP⁵² using oligonucleotides 5’GCGACTAGTATGGTGAGCGAGCTGATTAAG and 5’GCGGAATTCTTATCTGTGCCCCAGTTTG, digested with SpeI/EcoRI-HF, and ligated to replace the FLAG-HIS3 in the plasmids above to create p416PGK-GFP-R12-RFP and p416PGK-GFP-K12(AAA)-RFP. A XbaI/ClaI digest was used to clone the reporter into pAG306GPD-ccdB-chrI, a gift from Dan Gottschling (Addgene plasmid #41894)⁷¹. These constructs were then digested with NotI and transformed into yeast cells using standard methods and selected on SC-ura. Correct clones were confirmed by colony PCR, sequencing, and microscopy.

To construct ltn1Δ strains with integrated YTM1 GFP fusions, LTN1 was deleted from the knock-in strains via standard PCR-based homologous recombination. The YTM1-GFP ltn1Δ strain was constructed by replacing LTN1 with LEU2 using a cassette amplified from plasmid GTL-g (Addgene plasmid #81099) with oligonucleotides 5’GATTATGCCCCAACATGGAAAACTGAAAAATATTGATGAAGCGAGTCTGTAGGCGAACCTAACCGG and 5’TTTCCAGAATATCCGGGTGATGGGCTGGATTGGCAAGGTATTATATGAAGATTGTTCTACCATTCACAACTATAT. Similarly, the GFP-YTM1 ltn1Δ strain was constructed by replacing LTN1 with URA3 using a cassette amplified from plasmid pSH100 (Addgene plasmid #45930) with oligonucleotides 5’GATTATGCCCCAACATGGAAAACTGAAAAATATTGATGAAGCGAGCCACAGCTTTTCAATTCAATTCATCA and 5’TTTCCAGAATATCCGGGTGATGGGCTGGATTGGCAAGGTATCCTGATGCGGTATTTTCTCCTTA. Both cassettes were amplified, separated by electrophoresis, gel purified and transformed into yeast cells using the lithium acetate method. Correct transformants were verified by standard PCR using the oligonucleotides LTN1-VF: 5’ TCCGTTTTGGATTCGTTGGAGT, LTN1-VR: 5’ ACCGCCAAGCAGAAAATCC, LEU2-VF: 5’ AGCACGAGCCTCCTTTACCT, and URA3-VF: 5’ CGAATGCACACGGTGTGGT.

Ribo-Seq.

For C. elegans samples, after passaging for at least three generations, age-synchronized populations of L1 larvae were obtained and grown to the L4 larval stage (Day 0). The plates of L4 animals were then washed and the collected animals were transferred to plates containing 50 µg/mL 5-fluoro-2’-deoxyuridine (FUdR; Millipore Sigma) at ~1,000 worms per plate. Animals were carefully monitored to ensure adequate bacteria was in constant supply, adding concentrated stocks of bacteria as necessary. At each of the indicated ages, ~40,000 adult worms were collected by quickly washing subsets of plates into microcentrifuge tubes using modified lysis buffer (20 mM Tris-HCl pH 7.5, 140 mM KCl, 1.5 mM MgCl₂), followed by a short centrifugation (0.3 g at room temperature for 10 sec) that minimized the number of eggs and bacteria that pelleted with the animals. Supernatant from each tube was then aspirated and pelleted animals were dripped into liquid nitrogen. Lysis was performed by combining 1 mL of lysis buffer containing twice the concentration of DTT, cycloheximide (CHX), and Triton X-100 (20 mM Tris-HCl pH 7.5, 140 mM KCl, 1.5 mM MgCl₂, 1 mM DTT, 200 µg mL⁻¹ CHX, 2% Triton X-100) to compensate for the approximate volume of lysis buffer that remained when harvesting the animals. Worms were then pulverized using a MM-301 mixer mill at 20 Hz for 1 min, with lysate being thawed in a room temperature water bath at room temperature and centrifuged at 15,000 g at 4°C for 10 min. After quantifying RNA concentration, 200 µg RNA was incubated for 45 min with 0.75 µl 100 U µl⁻¹ RNase I (Ambion), with another 200 µg RNA being kept on ice undigested to use for polysome profiles. Sucrose gradient centrifugation and fractionation were then performed as described previously¹⁷. Total RNA was extracted from the digested 80S fraction using the hot SDS-phenol-chloroform method, 20–35 nucleotide RNA footprints were isolated, ribosomal RNA was removed using the Ribo-Zero kit (Illumina), with remaining library preparation steps described elsewhere⁷². Barcoded samples were pooled and sequenced with a HiSeq 4000 (Illumina). For yeast samples, overnight cultures were diluted to an OD₆₀₀ of 0.05 in 500 mL SC media and grown at 30°C. Part of each culture was harvested as Day 0 cells at an OD₆₀₀ of ~0.7, with the remaining culture placed back at 30°C for 48 and 96 hours for harvesting Day 2 and Day 4 cells, respectively. All cells were harvested by vacuum filtration and freezing in liquid nitrogen. Lysis was performed by combining 2 mL of lysis buffer (20 mM Tris-HCl pH 7.5, 140 mM KCl, 1.5 mM MgCl₂, 0.5 mM DTT, 100 µg mL⁻¹, 1% Triton X-100) frozen in liquid nitrogen with cell pellets, with the remaining steps of lysis and library preparation following that of worms described above.

Data processing and pause score calculation.

Demultiplexed sequencing reads were trimmed of adaptor sequences using Cutadapt v1.4.2, followed by removal of the 5’ nucleotide using FASTX-Trimmer. Reads that mapped to ribosomal RNAs using Bowtie v1.0.0 (http://bowtie-bio.sourceforge.net/index.shtml)⁷³ were removed. Remaining reads were aligned to reference libraries that consisted of coding sequences containing 21 nucleotides flanking upstream of the start codon and downstream of the stop codon, or the entire transcript sequence including untranslated regions. The C. elegans library consisted of the longest transcript of 20,222 genes (ce11 / WBcel235), and the yeast library consisted of 5,793 ORFs (sacCer3 / R64–1-1) that excluded ORFs characterized as dubious or that overlapped with other genes. Bowtie v1.0.0 alignment of sequencing reads to these libraries used the following parameters to allow for two mismatches and keep only uniquely mapped reads for further analysis: -y -a -m 1 -v 2 --norc --best --strata. For each footprint length, customized python scripts were used to sum reads at each nucleotide. Metagene analysis was performed separately on each fragment length, and lengths that did not exhibit the characteristic 3-nucleotide periodicity were removed. Remaining reads had a nucleotide offset empirically determined from the 5’ end of each fragment length, using the characteristic large ribosome density at the start codon, so that each read was assigned to the first A-site nucleotide. Nucleotide reads at each codon were then summed and used for all additional analysis.

To analyze gene expression, reads were summed for each gene after excluding the first 20 and last 20 sense codons, followed by calculating tpm. Statistical significance was calculated using DESeq2⁷⁴ for genes with greater than 64 reads in each biological replicate. Pause scores at each codon position in a transcript were calculated by dividing the number of reads at that position by the expected number of reads, which was defined as the average number of reads across the internal part of the transcript, i.e. excluding the first 20 and last 20 sense codons. Mathematically:

where RR is the number of ribosomal reads at position i of gene j that has length k codons.

Age-dependent pause site identification.

To identify positions where ribosome pausing was increased during ageing, we included genes that had an average sequencing coverage of ≥ 0.5 reads per codon (calculated as above), and ≥ 64 total reads in each replicate, which included 7,200 C. elegans genes and 4,082 yeast genes. Next, we adapted a strategy we used previously¹⁹ that used two-tailed Fisher’s exact tests to identify positions where statistically significant changes in ribosome pausing between the young (Day 1 worms and Day 0 yeast) and old (Day 12 worms and Day 4 yeast) samples. Briefly, reads at each position and for each transcript were averaged between replicates and rounded to the nearest integer. At each position of a transcript, 2 x 2 contingency tables were created to perform a two-tailed Fisher’s exact test to compare the ratio of reads in the young and aged fractions at a given position to the ratio at all other positions in that transcript (i.e. the summed reads in each fraction for the entire transcript minus the position of interest). In other words, this compares the observed ratio of ribosome reads from our young and aged samples at a given position to the expected ratio based on the total number of reads that map to the transcript. At each position, this allows us to calculate the odds ratio as a measure of enrichment, along with an adjusted p-value to test significance, using the Benjamini-Hochberg correction for multiple hypothesis testing for each gene. The odds ratio is:

where A and Y are the ribosomal reads of the aged and young translatome fractions, respectively, at position i of gene j that has length k codons. Age-dependent pause sites were identified as those that had: (1) Benjamini-Hochberg adjusted p-value < 0.05, (2) 1 < odds ratio < ∞, (3) pause score in the oldest sample greater than the pause score in the intermediate aged sample (Day 6 worms and Day 2 yeast), (4) reads in the oldest sample greater than the average number of reads across the transcript, to control for background, and (5) a position in the internal part of the transcript (i.e. not in the first 20 or last 20 sense codons).

Ribosome occupancy analysis.

Ribosome occupancy for individual genes was plotted using a 5-residue moving average of the pause scores. To analyze ribosome occupancy around polybasic regions, we identified polybasic regions in the genomes of worms and yeast as defined by a stretch of 3, 4, 5, or 6 consecutive residues that were either lysine or arginine (Supplementary Table 3). These regions were categorized into only one group, such that regions of 6 consecutive basic residues was not included in the analysis of the shorter regions. For metagene analysis of ribosome pausing, reads were aligned at the ribosome A-site around the region of interest (e.g. age-dependent pause sites, or the start of the polybasic region). We then calculated the mean and bootstrapped 95% confidence intervals at each position. To control for differences in gene expression and coverage, ribosome reads were normalized by dividing the reads at each codon by the mean number of reads per codon across the analysis window. We excluded low coverage genes that had an average reads per codon across the analysis window less than 0.5.

Pause site sequence analysis.

To examine the enrichment of amino acids and codons associated with greater ribosome pausing during ageing, we first calculated the average frequency of each residue/codon across coding sequences. Using this as background, we generated logo plots to analyze amino acid enrichment in two ways using the R package Logolas⁷⁵. First, we used the tripeptide motifs of age-dependent pause sites in the upper rank of ribosome pausing (pause score > 10 for Day 12 worms, or pause score > 6 for Day 4 yeast). Second, we calculated the average pause score for each of the 8,000 possible tri-peptide motifs across coding sequences and used the motifs that had a higher average pause score, as indicated, and filtered by count (≥ 100) and pause score (average pause score in older sample > 2 in worms and > 1.5 in yeast). We also calculated the residue and codon frequency in the ribosomal active site and compared it to the background frequency within coding sequences.

Other computational analysis.

To examine the association of yeast chronological lifespan with the ability to clear paused, truncated nascent polypeptides, we used previously published datasets^6,62. We binned yeast strains based on the distribution of the GFP-R12 reporter abundance, as indicated (Fig. 3c). To examine the association of age-dependent ribosome pausing and protein aggregation (Fig. 4e and Extended Data Fig. 10a), we compared our dataset of ribosome pausing to two previous datasets that identified age-dependent protein aggregates in C. elegans^37,65. For the higher coverage dataset³⁷, we used proteins that were identified in at least three of the four replicates of Day 1 and Day 12 worms, and had an average aggregate abundance at Day 12 higher than that at Day 1. To examine the association of age-dependent ribosome pausing and ubiquitination (Extended Data Fig. 3e), we compared our dataset of ribosome pausing to a previous dataset examining co-translational ubiquitination in yeast²⁷.

Immunoblotting.

Yeast strains were grown at 30°C in SC media containing 2% glucose (standard conditions), 3% glycerol, or buffered to pH 6.0 with citrate phosphate buffer (64.2 mM Na₂HPO₄, 17.9 mM citric acid, pH 6.0)⁷⁶, as indicated. Day 0 cells were harvested at an OD₆₀₀ of ~0.7, and Day 4 cells were harvested 96 hours later from the same culture. For media swap experiments, overnight cultures were diluted into either fresh SC media or the nutrient-depleted media of the supernatant collected after pelleting Day 4 cells, and Day 0 cells were harvested as above. Frozen cell pellets were resuspended in 200 µl of lysis buffer (50 mM Tris pH 7.5, 150mM NaCl, 1mM EDTA, 5% glycerol, 0.1% Triton X-100, 0.1% SDS, 1mM PMSF, 0.5mM DTT, Roche cOmplete EDTA-free Protease Inhibitor Cocktail), and lysed by vortexing with glass beads at 3000 rpm at 4°C for 3 min, incubating on ice for 3 min, and vortexing again at 3000 rpm at 4°C for 3 min. After clearing by centrifugation of cell lysates at 4,000 rpm at 4°C for 30 sec, protein concentration was normalized by Bradford Assay (Bio-Rad) and 4X NuPage™ LDS Sample Buffer (ThermoFisher) was added. Samples were then boiled for 5 min, run on a 12% SDS-PAGE gel, and transferred to nitrocellulose membrane. Membranes were blocked in 5% milk reconstituted in TBS (20 mM Tris pH 7.5, 150 mL NaCl, 0.1% NaN₃). Primary antibodies were diluted in Antibody Buffer (20 mM Tris pH 7.5, 150 mL NaCl, 0.1% NaN₃, 5% BSA) and secondary antibodies were prepared in 5% milk in TBS. As indicated, blots were subjected to immunblotting using Mouse anti-GFP antibody (Millipore Sigma 11814460001, 1:1000 dilution), Rabbit anti-FLAG antibody (Millipore Sigma F7425, 1:1000 dilution), and Rabbit anti-Histone H3 antibody (EpiCypher 13–0001, 1:2500 dilution), and visualized using the LI-COR system IRDye® 800CW Donkey anti-Mouse IgG (LI-COR 926–32212, 1:10,000 dilution) and IRDye® 680RD Donkey anti-Rabbit IgG (LI-COR 926–68073, 1:10,000 dilution). See Supplementary Data for uncropped immunoblots.

Microscopy.

Yeast cells were grown as above for immunoblotting. Upon harvesting, cells were fixed in 4% paraformaldehyde by incubating at room temperature for 15 min, and washed again in 1X PBS. Cells were then immobilized onto polylysine coated coverslips (Neuvrito) and mounted with ProLong™ Diamond antifade mountant with DAPI (Thermo Fisher). Fluorescence microscopy for quantification was acquired using a Zeiss Axio observer.Z1 inverted microscope equipped with a Plan-Apochromat 100x/1.4 oil DIC M27 objective (Zeiss), X-cite 120 LED system (Lumen Dynamics), filter set (HE) RFP/GFP/DAPI (Zeiss) and a digital Axiocam MRm camera (Zeiss) controlled with the Zen blue software. Raw data was collected as z-stacks and projected using ImageJ (NIH), and quantification was performed manually. Confocal microscopy of representative cells was collected on a Leica TCS SP8 inverted sSTED microscope equipped with a 100x/1.40 APO objective and using the following detection mirror settings: RFP 560–630nm, GFP 470–536. DAPI staining was detected using a blue diode 405nm laser (10%) and a photon multiplying tube. One representative middle slide was collected, and images were subsequently deconvolved and background subtracted using Huygens Professional (Scientific Volume Imaging).

Statistical analysis.

All analysis was performed in R (https://www.r-project.org). Gene ontology annotations and significance were obtained using the Database for Annotation, Visualization, and Integrated Discovery (DAVID v6.8)⁷⁷ using as background, as indicated, the transcripts included in the Ribo-Seq reference library or present in both this library and the mass spectrometry datasets^37,65. Statistical significance of categorical variable distributions shown in box plots used two-sided Wilcoxon rank-sum tests (e.g. Fig. 3c). For box plots, the center line represents the median, box limits indicate the upper and lower quartiles, whiskers indicate the 1.5x interquartile range, and points are outliers. Significance of age-dependent puncta formation (Fig. 3b) was assessed using two-sided Welch’s t-tests. Two-sided Fisher’s exact tests were used to calculate significance of residue and codon frequency in age-dependent pause sites (Fig. 2, Fig. 4, Extended Data Fig. 3, and Extended Data Fig. 9), association of ribosome pausing and ubiquitination (Extended Data Fig. 3e), association of ribosome pausing and aggregation (Fig. 4e), and sequence enrichment of protein aggregates (Extended Data Fig. 10b). Additional statistical details are mentioned in the figures or figure legends, including the values of n and P. None of the experiments involved blinding or randomization, and sample size was not predetermined.