The 18S rRNA Methyltransferase DIMT-1 Regulates Lifespan in the Germline Later in Life

Abstract

Ribosome heterogeneity has emerged as an important regulatory control feature for determining which proteins are synthesized, however, the influence of age on ribosome heterogeneity is not fully understood. Whether mRNA transcripts are selectively translated in young versus old cells and whether dysregulation of this process drives organismal aging is unknown. Here we examined the role of ribosomal RNA (rRNA) methylation in maintaining appropriate translation as organisms age. In a directed RNAi screen, we identified the 18S rRNA N6'-dimethyl adenosine (m^6,2A) methyltransferase, dimt-1, as a regulator of C. elegans lifespan and stress resistance. Lifespan extension induced by dimt-1 deficiency required a functional germline and was dependent on the known regulator of protein translation, the Rag GTPase, raga-1, which links amino acid sensing to the mechanistic target of rapamycin complex (mTORC)1. Using an auxin-inducible degron tagged version of dimt-1, we demonstrate that DIMT-1 functions in the germline after mid-life to regulate lifespan. We further found that knock-down of dimt-1 leads to selective translation of transcripts important for stress resistance and lifespan regulation in the C. elegans germline in mid-life including the cytochrome P450 daf-9, which synthesizes a steroid that signals from the germline to the soma to regulate lifespan. We found that dimt-1 induced lifespan extension was dependent on the daf-9 signaling pathway. This finding reveals a new layer of proteome dysfunction, beyond protein synthesis and degradation, as an important regulator of aging. Our findings highlight a new role for ribosome heterogeneity, and specific rRNA modifications, in maintaining appropriate translation later in life to promote healthy aging.

Fig. 1 |. Ribosomal RNA methyltransferases regulate lifespan, heat and UV stress response.

a-c, Directed RNAi screen of putative rRNA methyltransferases reveals changes in a) lifespan, b) UV stress survival, and c) heat stress survival relative to empty vector control worms. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. Some RNAi clones were not replicated (without error bars) due to no effect being observed.

Fig. 2 |. Ribosomal RNA modifications in 26S and 18S are dynamically regulated throughout life and change in response to UV and heat stress.

a, 26S and 18S rRNA extracted from C. elegans at different ages from 4 days to 19 days reveal changes in rRNA methylation as assessed by UHPLC-ms/ms. This heat map represents the relative changes of methylation in 4 biological replicates. Changes in individual modifications can be seen in fig. S2C. b-c, 26S and 18S rRNA extracted from C. elegans after UV stress (b) or 37°C heat stress (c) relative to worms not exposed to UV (ctl) or grown at 20°C reveal some changes in rRNA methylation as assessed by UHPLC-ms/ms. These heat maps represent the relative changes of methylation in 3 biological replicates. Changes in individual modifications can be seen in Extended Data Fig. 2e–f.

Fig. 3 |. Longevity induced by dimt-1 deficiency requires the FoxO and TOR signaling pathways and requires an intact germline.

a, Mutation of glutamic acid 79 to an alanine (E79A) in dimt-1 caused a complete elimination of 18S rRNA m^6,2A as assessed by UHPLC-ms/ms. Statistics represent an unpaired t-test with Welch’s correction. b, Mutation of E79A in dimt-1 caused lifespan extension relative to WT worms. c, dimt-1 knock-down extends the lifespan of both WT and eat-2(ad1116) mutant worm lifespan to a similar extent (p=0.7021 by 2-way ANOVA). d, dimt-1 knock-down extends the lifespan of both WT and daf-2(e1370) mutant worm lifespan to a similar extent (p=0.0806 by 2-way ANOVA). e, dimt-1 knock-down extends the lifespan of both WT and hsf-1(sy441) mutant worm lifespan to a similar extent (p=0.4245 by 2-way ANOVA). f, dimt-1 knock-down extends the lifespan of WT but not daf-16(mu86) mutant worm lifespan (p<0.0001 by 2-way ANOVA). g, dimt-1 knock-down extends the lifespan of WT but does not further extend the long lifespan of raga-1(ok386) mutant worms (p=0.0327 by 2-way ANOVA). h, dimt-1 knock-down extends the lifespan of WT but does not further extend the long lifespan of germline deficient glp-1(e2141ts) mutant worms that were shifted to the restrictive temperature at the L1 stage (p<0.0001 by 2-way ANOVA). i, dimt-1 knock-down extends the lifespan of WT but not sterile pgl-1(bn101ts) mutant worms whose mothers were shifted to the restrictive temperature (25.5°C) at the L4 stage (p<0.0001 by 2-way ANOVA). Statistics and replicate experiments are presented in Supplementary Table 1. ns; not-significant, *; p < 0.05, **; p < 0.01, ***; p < 0.001, ****; p < 0.0001 as calculated by log-rank (mantel-cox) statistical test.

Fig. 4 |. DIMT-1 functions in the germline to regulate lifespan and affects translation of specific mRNA transcripts.

a, Auxin-inducible degron (AID) ubiquitous degradation of DIMT-1 protein leads to significant decrease in m^6,2A levels in the 18S rRNA subunit as assessed by UHPLC-ms/ms. b, Ubiquitous and germline-specific AID-induced DIMT-1 protein degradation causes a lifespan extension while DIMT-1 depletion in the muscle, intestine, or neurons has no effect on lifespan extension. c-d, Tissue-specific knock-down of dimt-1 in the germline increases lifespan to a similar extent as in ubiquitous knock-down while knock-down of dimt-1 in the muscle, intestine, or neurons has no significant effect on longevity. e, dimt-1 knock-down does not extend the lifespan of worms treated with 5-fluorodeoxyuridine (FUdR), a drug that inhibits proliferation of germline stem cells and the production of intact eggs. Statistics and replicate longevity experiments are presented in Supplementary Table 2. ns; not-significant, *; p < 0.05, ***; p < 0.001, ****; p < 0.0001 as calculated by log-rank (mantel-cox) statistical test.

Fig. 5 |. DIMT-1 affects translation of specific mRNA transcripts.

a, Heat maps of the 2082 differentially ribosome bound transcripts in the ribosome after dimt-1 knock-down from day 7 worms. Ribosome binding was normalized to total RNA expression to give translation efficiency. Dimt-1 was knocked down from the L4 stage until day 7. Each column represents an independent biological replicate from ribosome sequencing after TRAP. b, Pathway analysis of differentially bound transcripts after dimt-1 knock-down revealed altered ribosome binding to transcripts involved in longevity regulation, degradation, fatty acid metabolism, the ribosome, and oxidative phosphorylation. RNAseq and translation efficiency significantly regulated genes and gene ontology categories are presented in Supplementary Tables 3 and 4. c, Sequence motifs enriched in the 5’ UTR of more bound mRNA transcripts after dimt-1 knock-down. d, dimt-1 knock-down extends the lifespan of WT but not daf-9 mutant worms (p=0.004 by 2-way ANOVA). e, dimt-1 knock-down extends the lifespan of WT but not daf-12 mutant worms (p=0.0005 by 2-way ANOVA).

Fig. 6 |. DIMT-1 regulates lifespan after mid-life.

a, Experimental design for the AID-tagged DIMT-1 temporal knock out experiments (y.a. = young adults). Red lines indicate when strains were placed on 150 uM auxin. b, AID-induced depletion of DIMT-1 in the germline extends lifespan when depleted in the previous generation and for the entirety of the assayed generation, starting at birth, or from young adulthood for the remainder of the lifespan but does not extend lifespan when depleted only from birth until young adulthood. c, AID-induced depletion of DIMT-1 ubiquitously extends lifespan when depleted in the previous generation and for the entirety of the assayed generation, starting at birth, or from young adulthood for the remainder of the lifespan and causes a less dramatic extension in lifespan when depleted only from birth until young adulthood. d, AID-induced depletion of DIMT-1 in the germline extends lifespan to a similar extent when depleted from young adulthood, after reproduction, or starting in mid-life. Statistics and replicate longevity experiments are presented in Supplementary Table 5. ns; not-significant, *; p < 0.05, ***; p < 0.001, ****; p < 0.0001 as calculated by log-rank (mantel-cox) statistical test.