Environmental Temperature Differentially Modulates C. elegans Longevity through a Thermosensitive TRP Channel

Abstract

Temperature profoundly affects aging in both poikilotherms and homeotherms. A general belief is that lower temperatures extend lifespan, whereas higher temperatures shorten it. Although this "temperature law" is widely accepted, it has not been extensively tested. Here, we systematically evaluated the role of temperature in lifespan regulation in C. elegans. We found that, although exposure to low temperatures at the adult stage prolongs lifespan, low-temperature treatment at the larval stage surprisingly reduces lifespan. Interestingly, this differential effect of temperature on longevity in larvae and adults is mediated by the same thermosensitive TRP channel TRPA-1 that signals to the transcription factor DAF-16/FOXO. DAF-16/FOXO and TRPA-1 act in larva to shorten lifespan but extend lifespan in adulthood. DAF-16/FOXO differentially regulates gene expression in larva and adult in a temperature-dependent manner. Our results uncover complexity underlying temperature modulation of longevity, demonstrating that temperature differentially regulates lifespan at different stages of life.

Figure 1. Temperature treatment at different stages of worm life differentially affects lifespan but not fertility

(A) Exposure to low and high temperatures at the adult stage extends and shortens lifespan, respectively (p<0.001, log-rank test). The top panel describes the protocol of temperature treatment. (B) Exposure to low and high temperatures at the larval stage shortens and extends lifespan, respectively (p<0.001, log-rank test). The top panel describes the protocol of temperature treatment. (C–D) Exposure to different temperatures at the larval stage does not have a notable effect on fertility in adulthood. (C) Overall fertility. (D) Egg-laying pattern. The percentage of eggs laid every 12 hours (half a day) was plotted. Larvae were reared at 15, 20, and 25 °C, and their adults were moved to 20 °C to lay eggs. n=30. Error bars: SEM. Also see Figure S1. Representative data are shown here, and replicates and detailed statistics are shown in Table S1.

Figure 2. The temporal window critical for temperature modulation of lifespan

(A) The L1–L2 stage is a critical window for temperature modulation of lifespan at the larval stage. The protocol was described in the upper panel. Hermaphrodite mothers were allowed to lay eggs at 20 °C. The specific stage (egg, L1, L2, and L3) at which animals were up- or down-shifted from 20 °C to 25 °C or 15 °C was labeled in each panel. At the L4 stage, animals were all shifted back to 20 °C to score adult lifespan. (egg: p<0.001, log-rank test; L1: p<0.001, log-rank test; L2: p=0.066, log-rank test; L3: p=0.198, log-rank test) (B) A second protocol also shows that the L1–L2 stage is a window critical for temperature modulation of lifespan at the larval stage. The protocol was described in the upper panel. Eggs were allowed to hatch at 15 and 25 °C and developed to different stages of larvae or adults, which were shifted to 20 °C to score lifespan. The specific stage [L1, L2, L3, L4, D1 (Day 1 adult), D2 (Day 2 adult), and D3 (Day 3 adult)] at which animals were shifted to 20 °C were labeled in each panel (L1: p=0.399, log-rank test; L2: p=0.013, log-rank test; L3: p=0.001, log-rank test; L4: p<0.001, log-rank test; D1: p<0.001, log-rank test; D2: p=0.007, log-rank test; D3: p=0.788, log-rank test). Representative data are shown here; replicates and detailed statistics are described in Table S1.

Figure 3. TRPA-1 is required for temperature modulation of longevity at the larval stage

(A) Loss of trpa-1 nearly eliminated the effect of larval temperature treatment on lifespan (p=0.104, log-rank test). (B–C) trpa-1 mutant worms show a normal lifespan following treatment at 15 °C at the larval stage (B) (p=0.945, log-rank test), but are short-lived if their larvae were treated at 25 °C (C) (p<0.001, log-rank test). Temperature treatment was performed as shown in Figure 1B, and the adult lifespan was scored at 20 °C. (D–E) trpa-1 mutant worms are short-lived when adult animals are reared at 15 °C (D) (p<0.001, log-rank test), but show a normal lifespan when reared at 25 °C (E) (p=0.934, log-rank test). Temperature treatment was performed as shown in Figure 1A, and the adult lifespan was scored at 15 °C and 25 °C in (D) and (E), respectively. (F–G) trpa-1 mutant worms are long-lived following treatment at 15 °C at the larval stage (F) (p<0.001, log-rank test), but show a normal lifespan if their larvae were treated at 25 °C (G) (p=0.878, log-rank test). Temperature treatment was performed as shown in Figure 1B, but the adult lifespan was scored at 25 °C instead of 20 °C. Also see Figure S2. Representative data are shown here; replicates and detailed statistics are described in Table S1.

Figure 4. Overexpression of TRPA-1 enhances the effect of larval temperature treatment on lifespan

(A) The effect of larval temperature treatment on lifespan in wild-type worms (p<0.001, log-rank test). Temperature treatment was performed as described in Figure 1B. (B–F) Overexpression of TRPA-1 as a transgene in neurons (B) (p<0.001, log-rank test), but not in the intestine (C) (p<0.001, log-rank test), muscle (D) (p<0.001, log-rank test), or cuticle (E) (p=0.001, log-rank test) augments the effect of larval temperature treatment on lifespan. Data in (B–E) were summarized in (F). p values are listed (paired t test). The ges-1, rgef-1, myo-3, and dpy-7 promoter drives expression in the intestine, neurons, muscle, and cuticle, respectively (Aamodt et al., 1991; Altun-Gultekin et al., 2001; Fire and Waterston, 1989; Gilleard et al., 1997). (G–H) Overexpression of TRPA-1 as a transgene shortens lifespan following temperature treatment at 15 °C at the larval stage (G) (p<0.001, log-rank test), but the same transgene has no effect on lifespan if larvae are treated at 25 °C (H) (p=0.637, log-rank test). Temperature treatment was performed as shown in Figure 1B, but the adult lifespan was scored at 25 °C instead of 20 °C. Also see Figure S3. Representative data are shown here; replicates and detailed statistics are described in Table S1.

Figure 5. DAF-16 and HSF-1 mediate the temperature effect at the larval stage

(A) Loss of daf-16 nearly eliminated the effect of larval temperature treatment on lifespan (p=0.021, Cox proportional hazard regression analysis). The temperature treatment protocol was performed as described in Figure 1B. (B) RNAi of hsf-1 blocked the effect of larval temperature treatment on lifespan (p=0.002, Cox proportional hazard regression analysis). The temperature treatment protocol was performed as described in Figure 1B. (C–F) SKN-1 (p=0.788, Cox proportional hazard regression analysis), PHA-4 (p=0.390, Cox proportional hazard regression), DAF-9 (p=0.179, Cox proportional hazard regression analysis), and DAF-12 (p=0.190, Cox proportional hazard regression analysis) are not required for larval temperature treatment to affect lifespan. Representative data are shown here; replicates and detailed statistics are described in Table S1.

Figure 6. DAF-16 mediates the effect of TRPA-1 on lifespan at the larval stage

(A) Loss of daf-16 blocks the effect of trpa-1 transgene on temperature modulation of lifespan at the larval stage (p<0.001, Cox proportional hazard regression analysis). Temperature treatment protocol was performed as described in Figure 1B. (B) hsf-1 RNAi cannot block the effect of trpa-1 transgene on temperature modulation of lifespan at the larval stage (p=0.824, Cox proportional hazard regression analysis). (C–D) RNAi of daf-16 in adulthood shortens lifespan, independently of the cultivation temperature at the larval stage. Larvae were cultured at 15 °C (C) (p<0.001, log-rank test)] and 25 °C (D) (p<0.001, log-rank test) until L4, and then shifted to 20 °C to score lifespan. daf-16 RNAi was delivered at the adult stage only. At the larval stage, the same worms were fed vector RNAi. Control: worms were fed vector RNAi throughout the larval and adult stages. (E–F) RNAi of daf-16 in larvae extends lifespan in a temperature-dependent manner. Larvae were cultured at 15 °C (E) (p<0.001, log-rank test) and 25 °C (F) (p=0.877, log-rank test) until L4, and then shifted to 20 °C to score lifespan. daf-16 RNAi was delivered throughout the larval and adult stages (larva+adult) or adult stage only (adult). For those worms treated with daf-16 RNAi at the adult stage only (adult), their larvae were fed vector RNAi. Also see Figure S4. Representative data are shown here; replicates and detailed statistics are described in Table S1.

Figure 7. DAF-16-dependent genes are differentially regulated by temperature in larvae and adults

(A) Genes are differentially regulated by temperature in larvae and adults of wild-type (WT). Genes that were up-regulated (15 °C vs. 25 °C) in wild-type larvae are shown in left panel, while those genes that were down-regulated in wild-type larvae are shown in the right panel. As a comparison, the relative expression levels of the same set of genes in wild-type adults are also shown. Only differentially-regulated genes are included in the panels. FDR<0.05; log₂ FC>0.7. (B) DAF-16-dependent genes are differentially regulated by temperature in larvae and adults (15 °C vs. 25 °C). The left panel and right panel lists up- and down-regulated genes in larvae, respectively. Specifically, the first lane in the left panel lists DAF-16-dependent genes that were up-regulated by low temperature (15 °C vs. 25 °C) in wild-type larvae. The relative expression levels of the same set of genes in daf-16(mgDf47) mutant background are listed in the second lane. Each gene showed significant difference in expression level (p<0.05) between wild-type and daf-16 mutant, indicative of DAF-16 dependence. The third and fourth lane lists the relative expression levels of the same set of genes in wild-type adults and daf-16 adults, respectively. Some of these genes become DAF-16 independent in adults. The lanes in the right panel show down-regulated genes and are organized in a similar manner. FDR<0.05; log₂ FC>0.7. (C) qPCR quantification of DAF-16-dependent larval genes up-regulated by low temperature (15 °C vs 25 °C) in wild-type and daf-16 mutant worms. The expression level of each gene (8/10) was significantly different in daf-16(mgDf47) mutant larvae compared to wild-type, indicative of DAF-16 dependence. Their expression (10/10) is also significantly different in wild-type adults. Some of these genes (5/10), however, are no longer DAF-16 dependent in adults, indicating a possible switch in regulatory mechanisms. These genes were picked from Supple. Table 2A based on their high FC values. Experiments were repeated three times. Error bars: SEM. *p < 0.05; **p < 0.005 (ANOVA). Also see Figure S5 and Table S2.