Reduced insulin/IGF-1 signaling restores germ cell immortality to Caenorhabditis elegans Piwi mutants

Abstract

Defects in the Piwi/piRNA pathway lead to transposon desilencing and immediate sterility in many organisms. We found that the C. elegans Piwi mutant prg-1 became sterile after growth for many generations. This phenotype did not occur for RNAi mutants with strong transposon-silencing defects and was separable from the role of PRG-1 in transgene silencing. Brief periods of starvation extended the transgenerational lifespan of prg-1 mutants by stimulating the DAF-16/FOXO longevity transcription factor. Constitutive activation of DAF-16 via reduced daf-2 insulin/IGF-1 signaling immortalized prg-1 strains via RNAi proteins and histone H3 lysine 4 demethylases. In late-generation prg-1 mutants, desilencing of repetitive segments of the genome occurred, and silencing of repetitive loci was restored in prg-1; daf-2 mutants. This study reveals an unexpected interface between aging and transgenerational maintenance of germ cells, where somatic longevity is coupled to a genome-silencing pathway that promotes germ cell immortality in parallel to the Piwi/piRNA system.

Figure 1. Progressive sterility of prg-1 mutants

(A) Levels of fertility for different prg-1 mutants at 20°C and 25°C (mean±s.d., n=10 lines per strain). Individuals from independently derived lines were singled and their progeny counted. (B) prg-1 exhibits progressive sterility at both 20°C and 25°C. Animals at 25°C appear more robust than siblings at 20°C (Mantel-Cox log-rank test, P=0.001 for tm872, P=0.030 for pk2298, P=0.144 for n4357) (n=12). (C) Progressive sterility of prg-1(tm872); prg-2 double mutants (n=5 strains per genotype). (D) Starvation extends transgenerational lifespan of prg-1 strains propagated at 25°C. Extension by starvation is dependent on daf-16.

Figure 2. daf-2 signaling can suppress fertility defects of prg-1 and modulates germline remodeling at sterility

(A) Reduced daf-2 signaling suppresses prg-1 mediated progressive sterility. (B) lin-15B does not suppress transgenerational lifespan of prg-1 (n=30 strains per genotype). daf-16 and daf-18 are required for suppression of prg-1 by daf-2. Four prg-1 daf-16 double mutant strains were studied where prg-1 alleles tm872 and n4357 were combined with daf-16 alleles mgDf50 or mu86 (n=10 strains per genotype). Four prg-1; daf-18 double mutant strains were studied where prg-1 alleles tm872 and n4357 were combined with daf-18 alleles e1375 and ok480 (n=10 strains per genotype). For prg-1 daf-16; daf-2 triple mutants, prg-1(tm872) daf-16(mgDf50) and prg-1(n4357) daf-16(mu86) were each combined with three daf-2 alleles m41, e1368 and e1370 (n=10 strains scored per genotype) and examined for progressive sterility. Four prg-1; daf-2(e1368); daf-18 lines were constructed from the four allelic combinations of prg-1; daf-18 and examined for progressive sterility (n=10 strains scored per genotype). Data for all independent alleles was combined to show transgenerational lifespan for strains of the same genotype. (C) Lack of immediate sterility upon removal of daf-2 suggests suppression of the heritable epigenetic defect that causes sterility in prg-1 mutants. prg-1; daf-2(+/+) strains represent cross-progeny of late-generation prg-1; daf-2 double mutants where the daf-2 mutation has been removed. (D–E) Levels of spontaneous mutation assessed by reversion frequencies of unc-58 and unc-54. (F) Treatment of sterile late-generation prg-1 adults with daf-2 or age-1 RNAi restores fertility, and fertility can be maintained on these RNAi strains for at least 10 generations while RNAi is maintained.

Figure 3. piRNA loss affects expression of few genes targeted by piRNAs

(A) Boxplots are based on read frequencies for 4,839 piRNAs sequenced in one or more libraries. Boxes indicate interquartile ranges, horizontal bars medians, whiskers extend to the most extreme data points with distance from the box no more than 1.5 times the interquartile range, crosses indicate outliers. (B) Boxplots showing reduced median levels of 22G-RNAs for piRNA targets in later generation prg-1 animals. Levels of microRNAs do not show the same progressive reduction. P values are for Wilcoxon signed rank tests. (C) Analysis of genes whose expression changed more than 2-fold in late- versus early-generation prg-1 strains reveals few common genes are altered for prg-1 alleles. (D) Few altered genes that are cured by daf-2 are predicted piRNA targets.

Figure 4. Derepression of repetitive elements in prg-1 mutants is cured by daf-2

(A) Late-generation prg-1 mutants show a mean increase in transposon expression, compared to the expression of transposons in prg-1; daf-2 double mutants. (B) Density plot of the transposon expression changes shown in Figure 4A reveal increased expression of a subset of transposon sequences in late-generation prg-1 animals that is not seen in prg-1; daf-2 double mutants. (C) Simple repeats are upregulated in late-generation prg-1 mutants. (D) Simple repeats upregulated in prg-1 mutants are repressed in late generation prg-1; daf-2 mutants. (E–H) cDNA prepared from RNA was hybridized to microarrays revealed upregulation of tandem repeat tracts expression in late-generation prg-1 mutants but not in late generation prg-1; daf-2 mutants (E and G), as confirmed by RT-PCR analysis for late-generation wild-type as well as early-(E), and late-(L) generation mutant strains (F and H). F and H show expression of tandem repeats corresponding to E and G, respectively. (I) Genome-wide plots of 101 longest tandem repeat tracts defined by visually scanning the C. elegans genome in 70 kb sliding windows. Typically, tandem repeats display increased expression in late-generation prg-1(tm872) and prg-1(n4357) single mutants but silencing in prg-1 double mutants with daf-2 alleles e1368, e1370 or m41.

Figure 5. Altered 22G-RNA frequencies and gene expression changes in prg-1 mutants

(A) 22G-RNAs targeting transposons in prg-1 versus wild-type strains. (B) Increased 22G reads in late-generation prg-1 strains mapping to transposons are reduced in prg-1 mutants deficient for the Mutator pathway genes rde-2 or mut-7. (C) 22G-RNAs targeting tandem repeats in prg-1 versus wild-type strains. Note that it is easier to clearly identify transposon 22G-RNAs, because many permutations of each tandem repeat are found, so tandem repeat 22G-RNA data are almost certainly underestimated. (D) RT-PCR reveal increased expression of 174 mer tandem repeat for strains carrying the ypEx3 extrachromosomal array in comparison to sibling control strains lacking this array or prg-1 single mutant controls. (E) Repetitive extrachromosomal arrays containing CeRep59 and a 174 mer tandem repeats (ypEx3 array) or a histone locus cluster containing his-13, his-14, his-15, and his-16 genes (ypEx4 array) or a Helitron transposon (ypEx5 array) reveal that ypEx3 array accelerates the progressive sterility phenotype of prg-1. (F) rde-2 and ppw-1 are required for suppression of prg-1 by daf-2. Two ppw-1 prg-1 double mutant strains were studied where prg-1 alleles tm872 and n4357 were combined with ppw-1(pk1425) (n=10 strains per genotype). Four ppw-1 prg-1; daf-2 triple mutants were studied where both ppw-1 prg-1 mutants were combined with daf-2 alleles e1370 or e1368 (n=5 strains per genotype). Two rde-2 prg-1 double mutant strains were studied where prg-1 alleles tm872 and n4357 were combined with rde-2(ne221) (n=10 strains per genotype). Six rde-2 prg-1; daf-2 triple mutants were studied where both rde-2 prg-1 mutants were combined with daf-2 alleles e1370, e1368 and m41 (n=5 strains per genotype). (G) rde-2 is required for restoration of tandem silencing in prg-1 mutants by daf-2. (H) rbr-2 and spr-5 are required for suppression of prg-1 by daf-2. Four prg-1; rbr-2 double mutant strains were studied where prg-1 alleles tm872 and n4357 were combined with rbr-2 alleles ok2544 or tm1231 (n=10 strains per genotype). Eight prg-1; daf-2; rbr-2 triple mutants were studied where each prg-1; rbr-2 mutant was combined with daf-2 alleles e1370 or e1368 (n=5 strains per genotype). Two prg-1 spr-5 double mutant strains were studied where prg-1 alleles tm872 and n4357 were combined with spr-5(by134) (n=10 strains per genotype). Three prg-1; daf-2; rbr-2 triple mutants were studied where each prg-1; rbr-2 mutant was combined with daf-2(e1370) and prg-1(n4357) spr-5 was combined with daf-2(e1368) (n=10 strains per genotype).

Figure 6. Model of parallel small RNA silencing pathways that can repress transgenerational fertility defects

(A) Wild-type PRG-1 maintains transgenerational fertility by silencing repetitive RNA expression, and functions separately to initiate silencing of foreign transgenes and most transposons. (B) Deficiency for prg-1 results in transgenerational desilencing of repetitive loci and sterility. Transposon and transgene silencing is maintained by Mutator proteins. (C) Increased DAF-16 signaling via daf-2 mutation suppresses progressive sterility and desilencing of repetitive loci when prg-1 is mutant. Mutations are indicated by a red X. Light color tints indicate pathway dysfunction for prg-1 mutants or low levels of DAF-16 activity in response to wild-type DAF-2 signaling.