Abstract
Recent studies have revealed that parental diet can affect offspring metabolism and longevity in Drosophila. However, the underlying mechanisms are still unknown. Here we demonstrate that Sir2 encoding an NAD+-dependent histone deacetylase is required for the intergenerational effects of low nutrition diet (1:5 dilution of standard diet). We observed an increased amount of triacylglyceride (TAG) in the offspring when fathers were maintained on a low nutrition diet for 2 days. The offspring had increased levels of metabolites of glycolysis and TCA cycle, the primary energy producing pathways. We found that Sir2 mutant fathers showed no intergenerational effects. RNAi-mediated knockdown of Sir2 in the fat body was sufficient to mimic the Sir2 mutant phenotype, and the phenotype was rescued by transgenic expression of wild-type Sir2 in the fat body. Interestingly, even fathers had no experience of low nutrition diet, overexpression of Sir2 in their fat bodies induced a high level of TAG in the offspring. These findings indicated that Sir2 is essential in the fat body of fathers to induce intergenerational effects of low nutrition diet.
Introduction
Nutritional conditions of parents have significant impacts on offspring growth and development. Recent studies have reported maternal and paternal inter/transgenerational phenotypes, including stress-induced epigenetic changes (Seong et al., 2011), endocrine disruptors (Anway et al., 2016), behavior (Dias & Ressler, 2014), and energy metabolism (Ng et al., 2010). Inter/transgenerational inheritance of environment-induced physiology involves generation and transmission of that information to the next generation (Sharma, 2015). It has been shown that altered DNA methylation (Radford et al., 2014), histone modifications (Inoue et al., 2017; Zenk et al., 2017), and noncoding RNA transcripts (Chen et al., 2016) can be transmitted from parents to offspring. However, it is largely unknown how environmental stimuli generate signals leading to inheritable epigenetic modifications.
Transgenerational effects of nutritional conditions have been studied in flies and explored the underlying epigenetic mechanisms. Early-life low-protein diet induced a higher level of H3 Lys27 trimethylation (H3K27me3) by upregulation of the protein level of E(z), H3K27 specific methyltransferase. This altered H3K27me3 was associated with a short lifespan in the offspring (Xia et al., 2016). Two days of high/low-sugar diet elicited obesity in offspring, which involves H3K9/K27me3-dependent reprogramming of metabolic genes (Öst et al., 2014). However, how the nutritional stress induces epigenetic modification to be transmitted to the next generation remains unknown.
Sir2/sirtuin1, which encodes a NAD-dependent class III histone deacetylase is a nutrition sensor and an effective regulator of metabolism and stress responses (Imai et al., 2002). Sir2 activity depends on the cellular level of NAD+, which is increased by starvation. Activated Sir2 has many functions to maintain energy homeostasis, regulating insulin signaling, fat mobilization, and energy consumption (Banerjee et al., 2013; Banerjee et al., 2017; Palu & Thummel, 2016). Moreover, Sir2 reduction is related to dysfunction of lipid metabolism in offspring by the maternal dietary intervention (Nguyen et al., 2017). On the other hand, Sir2 overexpression in offspring attenuates intergenerational effects (Nguyen et al., 2018). These studies have suggested that Sir2 may respond to nutritional stress and lead to the generation of epigenetic modification in the parental body.
In this study, we investigated the role of Sir2 in the intergenerational effects caused by paternal dietary conditions. We found that Sir2 mutant males showed no intergenerational phenotypes such as an increase of TAG levels in their offspring. The RNAi-mediated knockdown of Sir2 in the fat body of father impaired the ability to induce intergenerational phenotype. This inability was rescued by overexpression of Sir2 in the fat body. Thus, our results indicated that that Sir2 is essential in the fat body for fathers to induce intergenerational effects of low nutrient diet on offspring physiology.
Results
Paternal low nutrition diet-induced intergenerational phenotypes
To induce intergenerational effects with paternal diet, 4-to 5-day-old male flies were kept on a low nutrient diet for 2 days, and allowed to mate with virgin females which have been maintained on a control diet. After 1 day, the parents were removed, and the offspring were left to develop on a control diet. Adult male offspring were transferred to a new control diet, and 4-to 5-day-old males were weighed and used to measure the amounts of TAG levels in the whole body. The amount of TAG in the offspring was increased when fathers were maintained on a low nutrition diet for 2 days (Fig. 1A). The dietary intervention in fathers had no significant effect on body weight of the offspring (Fig. 1B). Although the increase of TAG levels was reproducibly observed, we found that the weight-normalized TAG level was more consistent compared to TAG alone (Fig. 1C). We use TAG/BW values as intergenerational phenotypes induced by paternal low nutrition diet.
Since the paternal nutrient condition could affect the physiology of offspring in various aspects, we examined whether a paternal low nutrient diet affects starvation tolerance and lifespan of the offspring flies. The paternal intervention led to increased starvation tolerance (Fig. 2A). On the other hand, the lifespan was decreased by a low diet of the fathers (Fig. 2B). These results suggested that paternal low nutrient diet has intergenerational effects on a wide variety of physiology of offspring.
Paternal low nutrition diet altered offspring metabolic phenotype
It has been shown that obesity and lifespan are often associated with impaired energy metabolism. Therefore, we examined whether the intergenerational effects of the paternal dietary intervention were also associated with altered energy metabolism. Using a liquid chromatography-mass spectrometry (LC-MS/MS), we measured the levels of metabolites, especially those in the primary energy-producing pathways, such as glycolysis and TCA cycle. Metabolite profiles were obtained from the offspring of fathers which were maintained on a low nutrition diet or standard diet (control) for 2 days. To compare metabolite profiles of two groups, we performed an unsupervised method of Principal Component Analysis (PCA) after the data had been preprocessed by autoscaling. The scores plot of the PCA indicated that the PC1 scores of the control and low paternal diet were negative and positive, respectively (Supplementary Fig. 1A). The PC1 score appears to be related to the diet.
Next, we investigated critical metabolic pathways related to PC1 score. We performed statistical hypothesis testing for factor loading in PC1 (Yamamoto et al., 2014), and 24 metabolites were statistically significant(Correlation coefficient: R ≧ 0.7, p: Holm’s method). Pathway analysis was performed for these 22 metabolites. Tricarboxylic acid (TCA) cycle, glyoxylate, and dicarboxylate metabolism and glycolysis were significantly affected in the offspring of fathers fed with low nutrition diet (Supplementary Fig. 1B). Our analysis revealed that low paternal diet altered offspring metabolic states.
Paternal Sir2 is essential in the fat body for generating intergenerational effects
To identify a gene required for generating the intergenerational phenotype in fathers, we focused on Sir2, which encodes a NAD-dependent class III histone deacetylase. Sir2 is an evolutionarily conserved nutrient sensor, which is activated by starvation or low nutrient conditions (Brachmann et al., 1995; Frye, 2000). Moreover, Sir2 can regulate levels of H3K9/K27me3 (Furuyama et al., 2004; Vaquero et al., 2007). Therefore, it is possible that Sir2 is involved in the intergenerational effects by transmitting information from fathers to offspring via germline cells. The existing evidence suggests that Sir2 may have functions to mediate low-nutrient diet-induced intergenerational effects.
To examine the role of Sir2 in fathers, we determined a weight-normalized TAG level and starvation tolerance in the offspring. The offspring of Sir2 mutant fathers showed no TAG accumulation (Fig. 3A) and decreased starvation tolerance (Fig. 3B). Next, we performed the RNAi-mediated knockdown experiments. The knockdown of Sir2 in the whole body showed the same phenotype as Sir2 null mutant (Fig. 4A). Interestingly, the knockdown of Sir2 in the fat body showed no response in the offspring (Fig. 4A). These results suggested that parental Sir2 is required in the fat body to form intergenerational effects. To confirm that, we conducted the fat body-specific Sir2 overexpression experiment in Sir2 null mutant background. The TAG accumulation in the offspring was rescued by fat-body-specific Sir2 overexpression in the Sir2 null mutant fathers (Fig. 4B). The findings suggested that parental Sir2 is essential in the fat body for generating intergenerational effects.
Discussion
In the present study, we demonstrated that low nutrition diet for fathers induced intergenerational phenotypes, which include an increased amount of TAG in the offspring. The offspring also showed reduced life span, starvation tolerance, and altered energy metabolic pathways including glycolysis, TCA cycle, glyoxylate, and dicarboxylate metabolism. Furthermore, we found that loss of paternal Sir2 showed no TAG accumulation in the offspring. The same phenotype was observed when paternal Sir2 was knocked down in the fat body, and transgenic overexpression of Sir2 in the fat body of father was sufficient to rescue mutant phenotype. These results support our hypothesis that paternal Sir2 receives nutritional stress and plays an essential role in generating information leading to transmission of intergenerational effects in the paternal body.
Previous studies have suggested that the nutrient-induced intergenerational effects are transmitted to offspring by H3K9/K27me3 in germline cells (Guida et al., 2019; Öst et al., 2014; Xia et al., 2016). Recently, it has been reported that maternal H3K9me3 is directly inherited to next generation and required for normal embryogenesis in mice and Drosophila (Inoue et al., 2017a; Inoue et al., 2017; Zenk et al., 2017). However, it remains unclear how environmental signals control histone modification enzymes which produce H3K9/K27me3. It has not been clear whether environmental signals directly regulate histone modification enzymes in germline cells to induce environment-dependent intergenerational epigenetic changes. In our experiments, knockdown of Sir2 in the germline cells (nos > Sir2-IR) did not affect the intergenerational phenotype. Although germline cells are likely to have epigenetic modifications in response to low nutrition diet, they don’t seem to respond to the signal directly. Sir2 is responsible for receiving environmental stress and may generate signals that could cause epigenetic modification in the germline cells to alter the metabolic phenotype. It is crucial to identify the downstream components of Sir2, which are involved in the transmission of environmental stress information to germline cells.
Transgenerational inheritance has been observed in various organisms. In human, many studies have reported that parental and early-life nutritional conditions have effects throughout one’s life (Langley-Evans, 2015; Tarry-adkins & Ozanne, 2011). It should be determined how intergenerational effects are generated from parental nutritional conditions in parents. Although the process could be complicated, the discovery of Si2 as a critical regulator of intergenerational effects of low nutrition diet would simplify the approach to understanding the mechanism of epigenetic inheritance.
Experimental procedures
Fly strains and media
Flies were maintained on a standard glucose-yeast-agar medium containing 10% glucose, 4% dry yeast, 9% cornmeal, 0.8% agar, 0.3% propionic acid and 1% para-hydroxybenzoate in temperature-controlled environmental chambers at 25°C throughout development. Unless otherwise stated, the standard medium was used as a control medium in all experiments. Low nutrition diet was prepared by diluting the standard medium (1/5), and mixed with agar at the final concentration of 0.8%. Sir22A-7-11 (Xie & Golic, 2004), nos-Gal4 and lsp-Gal4 were obtained from the Bloomington Stock Center. Ptc-Gal4 and Tub-Gal4 were obtained from the Kyoto Stock Center. The Sir2RNAi line was obtained from the National Institute of Genetics. Since Sir2 mutant strain has w1118 background, we used w1118 flies as a genotype control.
Generation of transgenic fly
To generate transgenic flies expressing Sir2 in germline and somatic cells, we constructed pUASp-Sir2-attB using pUASp-attB vector (Kaneuchi et al., 2015) and introduced them into the fly genome using a phiC31-based integration system (Bischof et al., 2007).
Triacylglyceride (TAG) measurement
TAG content was determined by an enzymatic assay with lipase (Triglyceride Reagent, Wako), which breaks down the triacylglycerol to free glycerol and three fatty acids. Ten flies (3-to 5-day-old) were weighed and homogenized with 550 µl of 1% Triton X-100 (Wako). Homogenates were incubated for 10 min at 70°C and frozen at −80°C. The frozen samples were thawed on ice and centrifuged for 10 min at 15,000 g at room temperature. The supernatants were collected and centrifuged again under the same conditions, and the resulting supernatants were used to measure the amount of TAG. A 10 µl sample was mixed with 100 µl assay reagent (free glycerol reagent, Wako) and 20 µl lipase reagent (triglyceride reagent, Wako), and incubated for 30 min at 37°C. The absorbance at 540 nm was measured using a plate reader (Enspire, PerkinElmer).
Starvation tolerance
To determine the survival rate of flies under starvation, 3-to 5-day old male flies were kept in vials containing 2% agar, and the number of dead flies was counted every 4 h. At least 80 flies per paternal diet and genotype were used.
Lifespan
Three-to five-day-old males flies were maintained in vials containing a standard glucose-yeast-agar medium at 25°C (30 males per vial). Flies were transferred to fresh food every 3 days, and the number of dead flies was counted at the time of transfer.
Measurement of metabolites
Ten adult flies (3-to 5-day old) were homogenized in 200 ml of 75% acetonitrile on ice. Supernatants were collected after centrifugation for 10 min at 12,000 g and were dried in a miVac Sample Concentrator (Genevac, Stone Ridge, NY). The sample was resuspended in 10 mM dibutyl ammonium acetate (DBAA, pH 4.95). The metabolites were analyzed by using a liquid chromatography quadrupole time-of-flight mass spectrometry (LC-QTofMS) system, ACQUITY UPLC, and Xevo QTofMS (Waters, Milford, MA).
Statistical analysis
Analyses of metabolites data were carried out with MetaboAnalyst 4.0 (http://www.metaboanalyst.ca), a web-based platform for metabolomics data analysis (Chong et al., 2018). The data were initially analyzed using the unsupervised method of principal component analysis (PCA), and then further analyzed using statistical hypothesis testing for factor loading in PC1 (statistically significant: Correlation coefficient: R ≧ 0.7, p: Holm’s method). Pathway analysis for factor loading was performed independently for metabolites with a significant difference between the two groups.
Statistical significances are evaluated using Student’s t-test for comparison of the levels of TAG and metabolites. The log-rank test was used for comparison of survival curves. A Box and Whisker Plot was used to visualize the data of TAG levels, body weights, and TAG levels normalized by body weights. The middle lines in the boxes represent the medians. Box edges and whiskers indicate the 25th/75th and 2.5th/97.5th percentiles, respectively.
Acknowledgments
We are grateful to the members of the Cellular Genetics Lab for helpful discussions and critical reading of the manuscript. We also thank the Kyoto Stock Center, the Bloomington Stock Center, and the National Institute of Genetics for providing the stocks.