Reducing translation through eIF4G/IFG-1 improves survival under ER stress that depends on heat shock factor HSF-1 in Caenorhabditis elegans

Abstract

Although certain methods of lowering and/or altering mRNA translation are associated with increased lifespan, the mechanisms underlying this effect remain largely unknown. We previously showed that the increased lifespan conferred by reducing expression of eukaryotic translation initiation factor 4G (eIF4G/IFG-1) enhances survival under starvation conditions while shifting protein expression toward factors involved with maintaining ER-dependent protein and lipid balance. In this study, we investigated changes in ER homeostasis and found that lower eIF4G/IFG-1 increased survival under conditions of ER stress. Enhanced survival required the ER stress sensor gene ire-1 and the ER calcium ATPase gene sca-1 and corresponded with increased translation of chaperones that mediate the ER unfolded protein response (UPR^ER ). Surprisingly, the heat-shock transcription factor gene hsf-1 was also required for enhanced survival, despite having little or no influence on the ability of wild-type animals to survive ER stress. The requirement for hsf-1 led us to re-evaluate the role of eIF4G/IFG-1 on thermotolerance. Results show that lowering expression of this translation factor enhanced thermotolerance, but only after prolonged attenuation, the timing of which corresponded to increased transcription of heat-shock factor transcriptional targets. Results indicate that restricting overall translation through eIF4G/IFG-1 enhances ER and cytoplasmic proteostasis through a mechanism that relies heavily on hsf-1.

Keywords: Caenorhabditis elegans; ifg-1; eIF4G; healthspan; lifespan; proteostasis.

Figure 1

Reduced translation through ifg‐1 promoted survival under ER stress. (A) Cellular component GO analysis of differentially translated genes associated with suppressing ifg‐1 (q = false discovery rate). Additional information is available in the Experimental procedures section. (B) Survival of N2 wild‐type animals fed bacteria expressing control (L4440) or ifg‐1 dsRNA starting on the first day of adulthood, followed by exposure to 25 μg mL⁻¹ tunicamycin 2 days later. ifg‐1(cxTi9279) mutants exposed to tunicamycin were included in the survival assay for comparison. Experiments were performed four or more times with similar results (see Table S2, Supporting information for additional data). Kaplan–Meier survival curves were compared using Mantel–Cox log‐rank test (P < 0.0001 for N2 survival on control vs. ifg‐1 RNAi; P < 0.0001 for N2 and ifg‐1(cxTi9279) on control RNAi; P = 0.94 for N2 on ifg‐1 RNAi vs. ifg‐1(cxTI9279) animals on control RNAi). (C) N2 wild‐type animals were fed control or ifg‐1 dsRNA for 4 days to allow for full effect of RNAi prior to exposure to 25 μg mL⁻¹ tunicamycin. Relative mRNA levels of atf‐6, pek‐1, and ire‐1 are shown after six or 24 h of exposure (6 h Tun and 24 h Tun, respectively). Samples were normalized to similarly prepared animals exposed to DMSO for 6 h (No Tun). Results are from three separate experiments (**P < 0.001; two‐tailed unpaired t‐test; error bars represent SEM). (D) N2 adults treated with either control or ifg‐1 RNAi starting at adulthood were exposed to 25 μg mL⁻¹ tunicamycin from day 4 to day 10 of adulthood (10 days Tun). Relative mRNA levels of atf‐6, pek‐1, and ire‐1 were measured at day 10. Samples were normalized to animals exposed to DMSO over the same period. Results were from three separate experiments (*P < 0.05, **P < 0.001, ***P < 0.0001, two‐tailed unpaired t‐test; error bars represent SEM). In all experiments, results are considered significant for P < 0.05.

Figure 2

Reducing expression of ifg‐1 promoted transcriptional and translational upregulation of UPR^ER‐responsive genes under ER stress. (A) Polysome profiles for wild‐type N2 adults fed bacteria expressing L4440 control (black dashes) or ifg‐1 dsRNA (red dashes) for 4 days prior to exposure to 25 μg mL⁻¹ tunicamycin for 6 and 24 h (6 h Tun and 24 h Tun, respectively). Nematode lysates were separated over a sucrose gradient and absorbance through the gradient was measured at 254 nm (UV254). Portions of the profile corresponding to polysomes are highlighted in blue and were collected from each sample for analysis of translated mRNA. Profiles are representative of four experiments performed. (B) Total and translated (polysome‐associated) mRNA levels of xbp‐1s and xbp‐1u were measured under conditions in (A) (*P < 0.05, **P < 0.001; two‐tailed t‐test; error bars indicate SEM). (C) Total and translated mRNA levels for hsp‐4 were measured under conditions in (A) (***P < 0.0001; two‐tailed t‐test; error bars indicate SEM). (D) Similar to (C), but for dnj‐7 (*P < 0.05, **P < 0.001; two‐tailed t‐test; error bars indicate SEM). Results in A–D are from the same four biological replicates and were considered significant for P < 0.05.

Figure 3

Enhanced survival in the ifg‐1 mutant required sca‐1, the ire‐1 branch of the UPR^ER, and the cytoplasmic UPR gene hsf‐1. Day 1 adults were subjected to RNAi as indicated for 2 days prior to exposure to 25 μg mL⁻¹ tunicamycin and survival was tracked daily. Kaplan–Meier survival curves were compared using Mantel–Cox log‐rank test, and average median lifespan for replicates was calculated using two‐tailed t‐test with Welch's correction. (A) The survival curve of ifg‐1(cxTi9279) fell below that of N2 when each were subjected to sca‐1 RNAi (P < 0.0001). In (B–D), survival was enhanced in ifg‐1(cxTi9279) compared to N2 at P < 0.0001 for RNAi test conditions shown. In (E–G), average median survival was not enhanced in ifg‐1(cxTi9279) compared to N2 when each were subjected to ire‐1 RNAi (P = 0.37), xbp‐1 RNAi (P > 0.99), and hsf‐1 RNAi (P = 0.18). (H) Survival was enhanced in ifg‐1(cxTi9279) compared to N2 when both were subjected to daf‐16 RNA (P < 0.0001). (I) Survival was enhanced in ifg‐1(cxTi9279) compared to N2 when day 1 adults were pretreated with cycloheximide for 2 days prior to exposure to 25 μg mL⁻¹ tunicamycin (P = 0.001). All experiments were performed three times and were considered significant for P < 0.05. See Table S3 (Supporting information) for additional data.

Figure 4

Enhanced thermotolerance was delayed compared with translation attenuation when ifg‐1 was attenuated with RNAi and corresponded with upregulation of heat‐shock protein gene expression. (A) ifg‐1 was inhibited for 2 or 7 days via RNAi beginning at adulthood in wild‐type N2 animals prior to incubation at 35 °C (i.e. heat stress conditions). Experiments were performed three times with similar results (see Table S5, Supporting information for additional data). Kaplan–Meier survival curves were compared using Mantel–Cox log‐rank test. Survival curve comparison was not significantly different after 2 days on RNAi for ifg‐1 RNAi compared to the control (P = 0.06) but was highly significant for enhanced thermotolerance after 7 days (P < 0.0001). (B) Protein expression of IFG‐1 was measured after 2 days of ifg‐1 RNAi treatment. Western blot quantification was from three independent experiments in which the average difference in IFG‐1 protein was standardized by the level of tubulin for each experiment. (*P < 0.05; Wilcoxon test; error bars indicate SEM). (C) Similar to (A), but for ifg‐1(cxTi9279) compared to N2. Experiments were performed three times with similar results (see Table S5, Supporting information for additional data). Kaplan–Meier survival curves were compared using Mantel–Cox log‐rank test. Results showed increased thermotolerance in the mutant compared to N2 for both time points (P < 0.0001 for each comparison). (D) Average median survival for three independent experiments corresponding to strains and conditions from (A) and (C) (**P < 0.001, ***P < 0.0001; ANOVA with post hoc Tukey test; error bars indicate SEM). (E) Expression of HSF‐1 target genes for the conditions and time points shown (*P < 0.05; ANOVA with post hoc Tukey test; error bars indicate SEM). All experiments were performed three times and were considered significant for P < 0.05.

Figure 5

Enhanced thermotolerance by reducing ifg‐1 did not require ER regulators and was not entirely dependent on hsf‐1. Kaplan–Meier survival curves in A–F were compared using Mantel–Cox log‐rank test. Survival at 35 °C was scored after 7 days under RNAi conditions indicated. Enhanced survival in ifg‐1(cxTi9279) compared to N2 was observed under the same RNAi treatment at P < 0.0001 for experiments in A–D. (E) Adults were pretreated with 0.5 mm cycloheximide for 7 days prior to scoring survival at 35 °C. Survival was enhanced in for cycloheximide‐treated N2 (P < 0.0001) but not for cycloheximide‐treated ifg‐1(cxTi9279) (P = 0.30). Despite lack of additional protection in ifg‐1(cxTi9279) treated with cycloheximide, survival was still enhanced over N2 under this condition (P < 0.0001). (F) Longevity of hsf‐1(sy441) mutants on ifg‐1 RNAi was enhanced compared to mutants under control RNAi (P < 0.0001). All experiments were performed three or more times and were considered significant for P < 0.05. See Tables S6 and S7 (Supporting information) for additional data.