Reducing the metabolic burden of rRNA synthesis promotes healthy longevity in Caenorhabditis elegans

Abstract

Ribosome biogenesis is initiated by RNA polymerase I (Pol I)-mediated synthesis of pre-ribosomal RNA (pre-rRNA). Pol I activity was previously linked to longevity, but the underlying mechanisms were not studied beyond effects on nucleolar structure and protein translation. Here we use multi-omics and functional tests to show that curtailment of Pol I activity remodels the lipidome and preserves mitochondrial function to promote longevity in Caenorhabditis elegans. Reduced pre-rRNA synthesis improves energy homeostasis and metabolic plasticity also in human primary cells. Conversely, the enhancement of pre-rRNA synthesis boosts growth and neuromuscular performance of young nematodes at the cost of accelerated metabolic decline, mitochondrial stress and premature aging. Moreover, restriction of Pol I activity extends lifespan more potently than direct repression of protein synthesis, and confers geroprotection even when initiated late in life, showcasing this intervention as an effective longevity and metabolic health treatment not limited by aging.

Fig. 1. Attenuation of Pol I activity increases lifespan.

a, b Comparison of tif-1A mRNA (a) and pre-rRNA (b) levels in C. elegans wild-type (N2 Bristol) strain and COP2239 strain harboring an ectopic tif-1A gene (ect. tif-1A). RT-qPCR data are shown relative to N2 and normalized to tgb-1 (tubulin gamma chain) mRNA. Each data point represents an independent biological sample of n = 60 worms. c Lifespan analysis of N2 and COP2239 (ect. tif-1A) strains, n = 140 worms. d, e RT-qPCR analysis of tif-1A mRNA (d) and pre-rRNA (e) levels in nematodes treated either with control (Ctrl) RNAi or tif-1A RNAi. Data are relative to Ctrl RNAi and normalized to tgb-1 mRNA. RNA from 3 independent experiments with n = 60 animals per sample was analyzed. f Lifespan analysis of animals treated with Ctrl RNAi or tif-1A RNAi, n = 140 worms. g, h Relative levels of rpoa-2 mRNA (g) and pre-rRNA (h) in nematodes exposed to control (Ctrl) or rpoa-2 RNAi. Samples from 4 (g) or 3 (h) biological replicates (n = 60 animals per sample) were used. i Lifespan analysis of worms treated with Ctrl RNAi or rpoa-2 RNAi, n = 140 worms. For a, b, d, e, g and h data represent mean values ± S.D. P values were determined using a Student’s unpaired t-test. For lifespan analyses in c, f and i, results representative of 3 independent experiments are shown. Survival was scored daily. P values were calculated using the Mantel-Cox test. ****P < 0.0001, **P < 0.01 and *P < 0.05. All statistical tests used were two-sided, the exact P values and statistical analyses are reported in the Source Data file. C. elegans culturing temperatures and RNAi treatment durations for this and other figures are provided in Supplementary Data 23.

Fig. 2. Pre-rRNA expression is inversely correlated with lifespan and healthspan.

a, b RT-qPCR analysis of ncl-1 mRNA (a) and pre-rRNA (b) levels of worms treated either with control (Ctrl) RNAi or ncl-1 RNAi. Values are relative to Ctrl RNAi and normalized to tgb-1 mRNA. Mean ± S.D. values of 3 independent experiments (n = 60 worms per sample) are shown. c Lifespan analysis of the nematodes upon treatment with Ctrl RNAi or ncl-1 RNAi, n = 140 worms. Survival was scored daily and P values were calculated using the Mantel-Cox test. Result is representative of 3 independent trials. d Body size (area) measurement of animals treated with Ctrl, tif-1A or ncl-1 RNAi. n = 30 worms. Data combine three independent trials and mean ± S.D. is shown. e Schematic of the locomotor performance test (burrowing assay). f, g Burrowing assays of animals treated with Ctrl, tif-1A or ncl-1 RNAi. The % of animals reaching the surface was determined at the indicated time points. Data show mean values ± S.D. of 3 independent assays performed on the second day of adulthood (AD2) with n = 40–45 worms (f) or on AD12 with n = 40–60 worms (g). P values were calculated with the Mantel-Cox test. h, i Representative images (n = 10 worms per image) and quantification of intestinal integrity monitored by the Brilliant Blue dye retention assay. Worms with staining outside the gut (‘Smurf’ phenotype) were scored following treatment with Ctrl, tif-1A or ncl-1 RNAi (n = 40–60 worms per treatment) at AD2 (h) or AD12 (i), scale bar 200 μm. Mean ± S.D. values of 3 independent experiments are shown. P values were determined using a Student’s unpaired t-test. ns = not significant, ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05. All statistical tests used were two-sided, the exact P values and statistical analyses are reported in the Source Data file.

Fig. 3. Proteomics links Pol I activity, ribosomal content and metabolism.

a–f Proteomes of nematodes treated with control (Ctrl) RNAi, tif-1A RNAi or ncl-1 RNAi were analyzed at AD2, AD6 and AD12. Each condition has five biological replicates of n = 700 worms. a, b Box plots showing log2 fold changes of selected immune response (CLEC), heat stress response (HSP), oxidative stress response (GST), and ribosomal (Ribosome) proteins as well as vitellogenins (VIT), comparing tif-1A KD (a) and ncl-1 KD (b) nematodes to age-matched controls. Each dot represents one protein, the median fold change within each group is shown as a bold line, the lower and upper limits of the boxplot indicate the first and third quartile, respectively, and the whiskers show the most up- and down-regulated proteins. The expression levels and Q values of individual proteins are reported in Supplementary Data 1–3. c, d Proteins differentially expressed (log2 fold change > 0.58, Q value < 0.05) in tif-1A KD (c) and ncl-1 KD (d) versus Ctrl RNAi animals on AD12 were classified using WormCat, and the two top enriched categories (Ribosome and Metabolism) with their subgroups are presented (analysis details are shown in Supplementary Data 8 and 9). e Box plots showing log2 fold changes of selected mitochondrial proteins between AD2 and AD12 Ctrl RNAi animals are presented. Either all measured mitochondrial proteins (‘mitochondrion’) or distinct categories (‘mito-protein complex’, ‘mito-respirasome’ and (‘mito-translation’) are shown. f Box plots showing log2 fold changes of selected mito-ribosome components in tif-1A RNAi and ncl-1 RNAi worms compared to age-matched controls. The expression levels and Q values of individual proteins are reported in Supplementary Data 10 (for e) and 11 (for f). In a, b, e and f the Wilcoxon rank-sum test was used for statistical analysis; Bonferroni correction was performed to adjust for multiple comparisons. ns = not significant, ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05. All statistical tests used were two-sided, the exact P values and statistical analyses are reported in the Source Data file.

Fig. 4. Pol I inhibition decelerates metabolic aging.

a Survival analysis of C. elegans wild-type (N2 Bristol) and ife-2 mutant strains treated with control (Ctrl) RNAi or tif-1A RNAi, n = 140 worms. b Comparison of ATP pools at AD2 between wild-type (N2) and ife-2 mutant worms treated with Ctrl RNAi or tif-1A RNAi. ATP levels were measured in a bioluminescence assay and normalized to total protein content. Values are shown relative to N2 Ctrl RNAi. Data are from 6 independent experiments with n = 50 worms in each sample. Mean ± S.D. values are shown. c Comparison of ATP pools at AD2 and AD12 between worms treated with Ctrl RNAi or tif-1A RNAi. ATP measurement was conducted as in b, and values are relative to Ctrl RNAi on AD12. Mean ± S.D. values of 6 independent experiments with n = 50 worms are shown. d Survival analysis of worms exposed to Ctrl RNAi or tif-1A RNAi and, from AD12 onwards, to 50 mM metformin (Met), n = 140 animals. A representative result of four independent experiments is shown. For a and d, survival was scored daily and P values were determined using the Mantel-Cox test. For b and c, a Student’s unpaired t-test was used for statistics. e, f BJ primary human skin fibroblasts were treated with indicated doses of metformin in the absence (DMSO) or presence of the indicated concentrations of ATP (positive survival control) or the Pol I inhibitors BMH-21 (e) and CX-5461 (f). Mitochondrial activity was assessed by MTT assay, and an unpaired Student’s t-test was used to compute P values, comparing each co-treatment to the respective DMSO + Metformin control condition, data represent mean ± S.D of 3 independent experiments. ns = not significant, ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05. All statistical tests used were two-sided, the exact P values and statistical analyses are reported in the Source Data file.

Fig. 5. Pol I activity affects the lipidome and mitochondria.

a Lipid abundance in control (Ctrl) RNAi, tif-1A RNAi and ncl-1 RNAi animals at different ages, normalized to the internal standard and animal number (n = 700). Values were computed from absolute intensities; mean of 5 biological replicates expressed as % of Ctrl RNAi treatment at AD2 is shown in each case (see Supplementary Data 12 for raw data). PC phosphatidylcholine, PE phosphatidylethanolamine, PS phosphatidylserine, PI phosphatidylinositol, PG phosphatidylglycerol, TAG triacylglycerol, FFA free fatty acid. b Volcano plots of lipid species separated by age (AD 2, 6, 12) and fatty acid desaturation. SFA saturated fatty acid, MUFA monounsaturated fatty acid, PUFA polyunsaturated fatty acid. Horizontal axis subtracts Ctrl RNAi mean log₁₀ values from respective tif-1A RNAi (left panel) or ncl-1 RNAi (right panel) values for each lipid, with statistics performed by multiple unpaired t-test on log-transformed mean values using a two-stage linear step-up procedure by Benjamini, Krieger and Yekutieli, FDR = 5%, data are from 5 biological replicates with n = 700 worms each. Calculations (from relative intensities) are shown in Supplementary Data 13–15. c Ratio of MUFAs to PUFAs in PCs at indicated age upon treatment with Ctrl RNAi, tif-1A RNAi or ncl-1 RNAi. Mean ± S.D. values from 5 replicates each with n = 700 worms are shown. Calculations (from relative intensities) are shown in Supplementary Data 19. d Fluorescence measurements in nematodes expressing hsp-6p::GFP following exposure to Ctrl, mip-1, ncl-1 or tif-1A RNAi. mip-1 RNAi served as a positive mito-stress control. n = 20-25 worms per condition. Representative result of 3 independent trials. For c and d, Student’s unpaired t-test compared ncl-1 and tif-1A RNAi effects to the age-matched Ctrl RNAi baseline. e–g Survival analysis of wild-type (N2 Bristol strain) and atfs-1 (gk3094) mutant nematodes (e), or N2 worms treated from AD1 onwards with the vehicle DMSO or 25 µM FCCP (f), or N2 and skn-1 (zj15) mutant nematodes (g), exposed to Ctrl or rpoa-2 RNAi, survival was scored daily. Representative results of three independent experiments, n = 140 worms, statistics assessed by Mantel-Cox test. ns = not significant, ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05. All statistical tests used were two-sided, the exact P values and statistical analyses are reported in the Source Data file.

Fig. 6. TAG lipolysis mediates longevity effects of the differential Pol I activity.

a, b Survival analysis of wild-type (N2 Bristol strain) nematodes exposed to control (Ctrl), rpoa-2 and atgl-1 RNAi, or simultaneously to rpoa-2 and atgl-1 RNAi (a) or Ctrl, ncl-1 and atgl-1 RNAi, or simultaneously to ncl-1 and atgl-1 RNAi (b). c Survival analysis of N2 and prx-5 (ku517) mutant nematodes exposed to Ctrl or rpoa-2 RNAi. d Survival analysis of wild-type worms exposed to Ctrl, rpoa-2, acs-2 RNAi or double RNAi targeting rpoa-2 and acs-2. Survival was scored daily, and n = 140 worms in all tests. Statistics were calculated using Mantel-Cox test. e Mitochondrial oxygen consumption rate (OCR) measured on AD12 in wild-type nematodes treated with Ctrl, tif-1A or ncl-1 RNAi. OCR was normalized to the number of worms and mean ± S.E.M. of n = 85 animals is shown. Statistical analysis was performed by using unpaired t-test on area under the curve values. ns = not significant, ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05. Representative results of 3 independent experiments are shown in all cases. All statistical tests used were two-sided, the exact P values and statistical analyses are reported in the Source Data file. f Schematic representation of the molecular mechanism connecting curbed Pol I activity with the attenuation of metabolic aging. The image was created with BioRender.com.

Fig. 7. Curbing of Pol I activity promotes healthy longevity independently of age.

a Schematic of the strategy for rpoa-2 inactivation at different ages and subsequent analysis of the animals. The image was created with BioRender.com. b, c RT-qPCR analysis of rpoa-2 mRNA (b) and pre-rRNA (c) levels at AD10 in nematodes treated with either control (Ctrl) RNAi or rpoa-2 RNAi from AD0, AD6 or AD8 onwards. Data are relative to Ctrl RNAi values and normalized to tgb-1 mRNA. Samples from 3 independent experiments with n = 60 animals per sample were analyzed, mean ± S.D. is shown. d Lifespan analysis of animals treated with Ctrl RNAi or with rpoa-2 RNAi from AD0, AD6 or AD8 onwards. n = 140 worms, survival was scored daily. Result is representative of 3 independent tests. e Locomotor performance tests (burrowing assays) at AD10 after Ctrl RNAi or rpoa-2 RNAi were administered as shown in (a). Percentage of animals that have reached the surface within indicated time is shown. Data are from 3 independent trials with n = 40 worms per group, mean ± S.D. values are shown. f Quantification of ‘Smurf’ animals showing intestinal leakage after RNAi treatments as indicated in (a). Percentage of worms with whole body staining from n = 40–50 worms per treatment is presented. Mean ± S.D. values from 3 independent trials are shown. P values were determined with an unpaired Student’s t-test (b, c, f) or Mantel-Cox test (d, e). ns = not significant, ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05. All statistical tests used were two-sided, the exact P values and statistical analyses are reported in the Source Data file.