Nuclear hormone receptor DHR96 mediates the resistance to xenobiotics but not the increased lifespan of insulin-mutant Drosophila

Abstract

Lifespan of laboratory animals can be increased by genetic, pharmacological, and dietary interventions. Increased expression of genes involved in xenobiotic metabolism, together with resistance to xenobiotics, are frequent correlates of lifespan extension in the nematode worm Caenorhabditis elegans, the fruit fly Drosophila, and mice. The Green Theory of Aging suggests that this association is causal, with the ability of cells to rid themselves of lipophilic toxins limiting normal lifespan. To test this idea, we experimentally increased resistance of Drosophila to the xenobiotic dichlordiphenyltrichlorethan (DDT), by artificial selection or by transgenic expression of a gene encoding a cytochrome P450. Although both interventions increased DDT resistance, neither increased lifespan. Furthermore, dietary restriction increased lifespan without increasing xenobiotic resistance, confirming that the two traits can be uncoupled. Reduced activity of the insulin/Igf signaling (IIS) pathway increases resistance to xenobiotics and extends lifespan in Drosophila, and can also increase longevity in C. elegans, mice, and possibly humans. We identified a nuclear hormone receptor, DHR96, as an essential mediator of the increased xenobiotic resistance of IIS mutant flies. However, the IIS mutants remained long-lived in the absence of DHR96 and the xenobiotic resistance that it conferred. Thus, in Drosophila IIS mutants, increased xenobiotic resistance and enhanced longevity are not causally connected. The frequent co-occurrence of the two traits may instead have evolved because, in nature, lowered IIS can signal the presence of pathogens. It will be important to determine whether enhanced xenobiotic metabolism is also a correlated, rather than a causal, trait in long-lived mice.

Keywords: DHR96; IIS; lifespan; nuclear hormone receptor; xenobiotic resistance.

Fig. 1.

Enhancing DDT resistance by artificial selection or overexpression of Cyp6g1 in Malpighian tubules did not extend fly lifespan. (A) Both selection lines (selection A and selection B) showed significant DDT resistance compared with three control populations (control lines X, Y, and Z) that had been maintained in parallel under nonselection conditions. (B) Lifespans of the same lines as in A, in the absence of DDT. The DDT-selected lines were shorter-lived than controls (P < 0.005 in all comparisons of selection vs. control populations, log-rank test). (C and D) Uo-GAL4 drove expression of Cyp6g1 in Malpighian tubules. This intervention increased resistance to DDT (P = 0.040 for comparison with Uo-GAL4/+ and P = 0.001 for comparison with UAS-Cyp6g1-8a/+, log-rank test) (C) but did not affect longevity (D) (P > 0.3 for all experimental lines vs. controls, log-rank test).

Fig. S1.

DR flies were not DDT-resistant, whereas IIS mutant flies showed increased DDT resistance. (A) Long-lived, dietarily restricted flies were not resistant to DDT. Age-synchronized female flies were maintained under dietary restriction (DR) or fully fed (FF) conditions as described in Grandison et al. (45). On day 7 of adult life, flies were transferred to the same food containing DDT. FF flies were significantly longer lived than DR flies under DDT stress (P < 0.001 in both trials, log-rank test). (B) Long-lived chico¹-heterozygote and insulin-producing mNSC-ablated flies were resistant to DDT (for any comparison of mutant versus control in either trial of resistance to either compound, P < 0.03, log-rank test).

Fig. 2.

Functionally related changes in gene expression in IIS mutants. Microarray data from chico¹ and mNSC-ablated females were analyzed by using CATMAP, which retrieves significant changes in functionally related groups of genes (44). The P values for genes with increased expression in common between the two mutants are plotted (P < 0.1, chico¹ compared with wild-type Dahomey control, mNSC-ablated flies compared with UAS-rpr control), where one data point represents a single functionally related gene, and the genes are labeled with the higher-level categories shown in the legend. P values from the chico¹ comparison are plotted on the x axis, those from the mNSC-ablation comparison are on the y axis. The equivalent data for genes with lower expression in common in the two mutants are shown in Fig. S2.

Fig. S2.

CATMAP categories from microarray data for chico and mNSC-ablated flies. Similar functional groups of genes identified by CAPMAP (44) were down-regulated in both long-lived IIS mutants. The P values for functional group changes that were found in common between the two mutants are plotted (P < 0.1), for chico¹ on the x axis and the mNSC ablation on the y axis.

Fig. 3.

DHR96 is a direct target of dFOXO and required for normal xenobiotic response and lifespan. (A) Relative enrichment of chromatin immunoprecipitated with a dFOXO-specific antibody. Higher levels in the precipitate of DNA neighboring DHR96 versus U6, a nonpolII-transcribed gene, indicate direct binding of dFOXO to DNA adjacent to the gene (P < 0.001, Welch t test). Relative enrichment was calculated as proportion of chromatin recovered in the IP for each region divided by the average of the two regions (HR96 and U6) for each chromatin (arbitrary scale). (B) Genetic deletion of DHR96 modestly decreased lifespan of female flies (P < 0.0001, log-rank test). (C and D) Tissue-specific overexpression of DHR96 in the Malpighian tubules (Uo-GAL4 driver) increased DDT resistance (C; P < 0.005, log-rank test), but did not affect lifespan (D).

Fig. S3.

Cyp6g1-8a and DHR96 are important mediators of the response to DDT. (A) Flies overexpressing Cyp6g1-8a in the Malpighian tubules were resistant to DDT compared with driver control (P < 0.05, log-rank test; repeat of the experiment shown in Fig. 1C). (B) Flies with genetic deletion of the DHR96 gene were sensitive to DDT compared with control wild-type flies (wDah) (P < 0.05, log-rank test).

Fig. S4.

Constitutive overexpression of DHR96 in the whole body caused developmental lethality. Overexpression of DHR96 using the daughterless-GAL4 driver resulted in lethality in different stages of Drosophila development and few survivors. Flies reared at 18 °C showed increased survival.

Fig. 4.

Analysis of the effects of DHR96 on the xenobiotic resistance of two IIS mutants. (A) Muscle-specific overexpression of dFOXO significantly enhanced resistance to DDT, phenobarbital, and malathion compared with control lines (Upper, log-rank test, P values for all comparisons with the matching driver and UAS lines <0.001, except for comparison of DDT resistance of dFOXO overexpressors with the MHC-GAL4 line, P = 0.61). Enhanced resistance was lost, when dFOXO was overexpressed in a DHR96 null background (Lower; P values for all comparisons with the matching driver and UAS lines >0.05). Cox proportional hazards (CPH) was used to test for a statistical interaction between the effects of dFOXO overexpression and genomic deletion of DHR96, and revealed that each significantly affected stress resistance, with a significant interaction between them (P < 0.01; Table S2). (B) Deletion of the mNSC cells significantly enhanced resistance to the three xenobiotics (Upper, log-rank test, P values for all comparisons with the matching driver and UAS lines <0.001), and this resistance was lost in a DHR96 null background (P values for all comparisons with the matching driver and UAS lines >0.05). CPH analysis revealed a significant interaction between the effect of mNSC ablation and genomic DHR96 deletion, indicating that xenobiotic resistance was significantly blocked by the genomic deletion of DHR96 (CPH, P < 0.001; Table S2).

Fig. S5.

Repeat xenobiotic stress assays with dFOXO overexpressing flies in wild-type and DHR96 null background. (A) dFOXO overexpressing flies were resistant to DDT (Left; 275 mg/L DDT, P values for all comparisons with the matching driver and UAS lines <0.001, log-rank test), whereas dFOXO over-expression in a DHR96 null background did not increase DDT resistance (Right; 175 mg/L DDT, P values for all comparisons with the matching driver and UAS lines >0.05). (B) dFOXO overexpressing flies in a wild-type or DHR96 null background were exposed to PB. dFOXO overexpression increased PB resistance, which entirely depended on the presence of DHR96. Two-way ANOVA revealed a significant interaction term (P = 0.016 for two-way ANOVA against the driver control and P = 0.0005 against the UAS control). Individual pairwise comparisons used Tukey’s multiple comparisons test (*P < 0.05; **P < 0.01; ***P < 0.001).

Fig. S6.

Repeat xenobiotic stress assay with mNSC-ablated flies in wild-type and DHR96 null background. Ablation of mNSCs enhanced DDT resistance but this resistance was lost when mNSCs were ablated in a DHR96 null background because two-way ANOVA revealed a significant interaction term (P = 0.0004 for two-way ANOVA against the driver control and P < 0.0001 against the UAS control). Individual pairwise comparisons used Tukey’s multiple comparisons test (*P < 0.05; **P < 0.01; ***P < 0.001).

Fig. 5.

DHR96 mediates the increased expression of detoxification genes in IIS mutants. mRNA expression of GstE1 (A) and Cyp6g1 (B) in the gut of mNSC-ablated flies was assessed by qRT-PCR to determine whether it was regulated by IIS or DHR96. Results represent fold changes in mRNA levels relative to the InsP3-GAL4 control (mean ± SEM). GstE1 and Cyp6g1 were significantly up-regulated in mNSC-ablated flies in a wild-type but not a DHR96 null background. Two-way ANOVA revealed a significant interaction term (P = 0.027 for GstE1 and P = 0.011 for Cyp6g1) with the response of both genes in the mNSC-ablated flies being entirely dependent on DHR96 (n ≥ 4). Individual pair-wise comparisons used Tukey’s multiple comparisons test (*P < 0.05, **P < 0.01, ***P < 0.001).

Fig. S7.

Regulation of detoxification genes by IIS is both common and model-specific. (A) Fold changes in mRNA expression of GstE1 and Cyp6g1 in guts and Malpighian tubules was assessed by qRT-PCR in dFOXO overexpessing flies and driver controls. dFOXO overexpression did not affect mRNA expression of either gene (P > 0.05 for both GstE1 and Cyp6g1, Student’s t test). (B) Correlation of fold changes in expression of genes within the GO term Defense in chico/+ and mNSC-ablated flies. Fifty-five genes were differentially regulated in chico/+ (green) and 17 in mNSC-ablated flies (red) with 8 being regulated in both datasets (yellow) with a significant overlap between them (P = 0.0085, Fisher’s exact test). (C) Differentially expressed (DE) genes within the GO term Defense common to both chico/+ and mNSC-ablated flies were generally up-regulated in both mutants (P < 0.0001 for both mutants, one-sample t test). Expression changes were significantly different for specific genes in the two mutants as revealed by Sidak’s multiple comparison test (P < 0.01). Two-way ANOVA revealed a significant interaction term between differentially regulated genes and the mutant genotype (P < 0.0001), showing that the two mutants produced different changes in expression of genes within the GO term Defense. *Although only differently regulated in mNSC-ablated flies, data for Cyp6g1 are included in this figure because this gene enhanced xenobiotic resistance when overexpressed in the Malpighian tubules (Fig. 1C and Fig. S3).

Fig. 6.

Lifespan extension by lowered IIS is independent of DHR96. Lifespan of females was significantly increased by muscle-specific overexpression of dFOXO or by targeted ablation of mNSC cells in both a wild-type (A and C, respectively) and a DHR96 null background (B and D, respectively) (P values for all comparisons with the matching driver and UAS lines <0.001, log-rank test). CPH analysis revealed that genomic DHR96 and overexpression of dFOXO or ablation of mNSC each significantly affected lifespan, but these effects did not show a significant interaction (Table S2).

Fig. S8.

Repeat of lifespan experiment with mNSC-ablated flies in wild-type and DHR96 null background. Ablation of mNSCs significantly increased lifespan in a wild-type background (A; P values for all comparisons with the matching driver and UAS lines <0.0001, log-rank test) and this lifespan extension was not affected by DHR96 null mutation (B; P = 0.017 compared with driver control InsP3-GAL3/DHR96Δ and P < 0.001 compared with UAS control UAS-rpr/DHR96Δ, log-rank test).