Effects of hunger on neuronal histone modifications slow aging in Drosophila

Abstract

Hunger is an ancient drive, yet the molecular nature of pressures of this sort and how they modulate physiology are unknown. We find that hunger modulates aging in Drosophila. Limitation of branched-chain amino acids (BCAAs) or activation of hunger-promoting neurons induced a hunger state that extended life span despite increased feeding. Alteration of the neuronal histone acetylome was associated with BCAA limitation, and preventing these alterations abrogated the effect of BCAA limitation to increase feeding and extend life span. Hunger acutely increased feeding through usage of the histone variant H3.3, whereas prolonged hunger seemed to decrease a hunger set point, resulting in beneficial consequences for aging. Demonstration of the sufficiency of hunger to extend life span reveals that motivational states alone can be deterministic drivers of aging.

Figure 1.. Hunger extends lifespan independent of appetite in dietary and genetic models of hungry flies.

(A) Schematic (left) and relative composition (right) of experimental diets. Details in Supplementary Tables S1–2 and Methods. Yaa=baseline diet as in (20), RD=reference diet, containing 51. 36g/L carbohydrate, 34. 9g/L non-essential amino acids, 22. 56g/L non-BCAA essential amino acids and, 12. 39g/L BCAAs. (B) Lifespan of Canton-S flies on diets of indicated BCAA concentration (log-rank test, p-values derived by comparison to RD, N=159–172). (C) Lifespan of Canton-S flies on low- or high-BCAA diets (log-rank test, N=173–175). (D-F) 24-hr Con-Ex measurement of w¹¹¹⁸ or PoxN flies on indicated diets. Volume (D), kcal (E), and total or BCAA-only amino acid intake (F) (two-tailed Student’s t-tests, N=10 flies per treatment). (G) Schematic of FLIC experimental designs used in H-I (top panel) and K (bottom panel). (H) FLIC interactions in common garden (one-way ANOVA with Tukey’s post-hoc, 2 experimental replicates are pooled, N=52–55 flies per treatment). (I) Con-Ex measurement of volumetric intake during re-feeding period (two-tailed Student’s t-test). (J) Con-Ex measurement on diets with individual BCAA reductions (one-way ANOVA with Tukey’s post-hoc, N=8–18 flies per treatment). (K) FLIC interactions in food choice environment (2% sucrose or [2% yeast + 1%sucrose]) after re-feeding SY food +/− 1% isoleucine (two-way ANOVA with Tukey’s post-hoc, 3 experimental replicates are pooled, N=59–60 flies per treatment). (L) Lifespan of Canton-S flies on RD, low-BCAA, or low-isoleucine diets (log-rank test, p-values derived by comparison to low-BCAA diet, N=171–177). (M-N) Con-Ex and lifespan measurements of flies carryingR50H05-GAL>UAS-CsChrimson or R50H05-GAL4/w-;CS controls exposed to red light for 12hours per day or kept in constant darkness. (M) 24-hr Con-Ex measurement (two-way ANOVA with Tukey’s post-poc, N=5–10 flies per treatment). (N) Lifespan measurement (log-rank test, N=101–123). All FLIC data are expressed as box-cox transformation to the 0. 25 power to achieve normality. *p<0. 05, **p<0. 01, ***p<0. 001. Error bars represent the mean +/− SEM.

Figure 2.. Dietary BCAAs reshape the neuronal epigenome by altering histone acetylation and histone H3 abundance.

(A) Representative western blots for H3K9ac or total histone H3 in Canton-S fly heads after exposure to indicated diets for 5–7 days, quantified in right panel (one-way ANOVAs. 10 heads were used per biological replicate, and 5 experimental replicates were performed, N=11 biological replicates per treatment). Values are normalized to low-BCAA treatment. (B) RT-qPCR measurement of relative mRNA abundance in Canton-S heads after 5–7 days on low- or high-BCAA, values are normalized to low-BCAA treatment (one-way ANOVA. 10 heads were used per biological replicate, and 2 experimental replicates were performed, N=10–11 biological replicates per treatment). (C) Western blots for histone PTMs in Canton-S heads after exposure to indicated diets +/− 100 mM sodium butyrate for 5–7 days, bands are quantified in bottom panel and normalized to low-BCAA treatment (10 heads were used per biological replicate, N=2 biological replicates). (D) Western blots for histone H3 and H4 in Canton-S heads on low-BCAA, high-BCAA, or [low-BCAA + high isoleucine] diets, quantified in bottom panel and normalized to low-BCAA treatment (one-way ANOVAs with Tukey’s post-hoc. 10 heads were used per biological replicate, N=2–3 biological replicates). (E) Western blots for histone H3 in heads of flies carrying R50H05-GAL>UAS-CsChrimson or UAS-CsChrimson/w-;CS controls exposed to red light for 12 hours, quantified in right panel and normalized to UAS-CsChrimson/w-;CS control (one-way ANOVA. 10 heads were used per biological replicate, N=4–5 biological replicates). Error bars represent the mean +/− SEM.

Figure 3.. Utilization of the histone variant H3.3A in the brain is modulated by BCAAs to stimulate feeding but is not required to extend lifespan.

(A) RT-qPCR of relative mRNA abundance of H3.3A in Canton-S heads after 5–7 days on BCAA diets, values are normalized to low-BCAA treatment (one-way ANOVA. 10 heads were used per biological replicate, 3 experimental replicates are pooled, N=12–13 biological replicates per treatment). (B-D) Pulse-chase of fluorescently labelled H3.3A in brains (B) or heads (C-D) of flies carrying Nsyb-GeneSwitch-GAL4>H3. 3A-GFP on BCAA diets. (B) Representative confocal images of brains immediately following the 2-day pulse of H3.3A-GFP induced by BCAA food + RU486 (top panel) and 1 week after the pulse (bottom panel). Scale bar 100μm. (C) Western blots of H3.3A-GFP (predicted molecular weight = 42 kDa) and histone H3 following the 2-day pulse of H3.3A-GFP induced by BCAA food + RU486, quantified in bottom panel and normalized to low-BCAA treatment (one-way ANOVAs, 10 heads were used per biological replicate, N=4 biological replicates per treatment). (D) Western blots of H3.3A-GFP (one-way ANOVA), histone H3(one-tailed Student’s t-test), and ubiquitin (one-way ANOVA, p=0. 571) 1 week after the H3.3A-GFP pulse, quantified in bottom panel and normalized to low-BCAA treatment (10 heads were-used per biological replicate, N=3 biological replicates per treatment). (E) Con-Ex measurement of 24-hr food intake after 5 days on BCAA diets +/− 100 mM sodium butyrate (two-way ANOVA with Tukey’s post-hoc). (F) Con-Ex measurement of 24-hr food intake after 5–7 days on BCAA diets from flies carrying Nsyb-GAL4>UAS-HIRA RNAi or controls (Nsyb-GAL4/w-;CS and UAS-HIRA RNAi/w-;CS) (two-way ANOVA with Tukey’s post-hoc, 3 experimental replicates are pooled, N=29–30 flies per treatment). (G) Lifespans of flies carrying Nsyb-GAL4>UAS-HIRA RNAi or controls (Nsyb-GAL4/w-;CS and UAS-HIRA RNAi/w-;CS) (log-rank test, N=184–202). (H) Lifespan of flies on BCAA diets +/− 100 mm sodium butyrate (log-rank test, N=186–196). (I) Lifespan of flies heterozygous for a null mutation in Rpd3 or w-;CS controls on BCAA diets (log-rank test, N=85–105). ***p<0. 001. Error bars represent the mean+/− SEM.

Figure 4.. Hunger neurons use BCAA metabolism to regulate histone abundance and modulate lifespan independent of feeding.

(A) RT-qPCR quantification of relative mRNA abundance of BCAT (CG1673) in Canton-S heads after 5–7 days on BCAA diets, values are normalized to low-BCAA treatment (one-way ANOVA, 10 heads were used per biological replicate, N=5–6 biological replicates). (B) Western blot of histone H3 in heads of flies carrying Nsyb-GAL>UAS-BCAT RNAi or control (Nsyb-GAL4/w-;CS), quantified in right panel and normalized to low-BCAA treatment within genotypes (two-way ANOVA with Tukey’s post-hoc, N=2 biological replicates per treatment). (C) Western blots of histone H3 (3 experimental replicates were pooled, N=5 biological replicates per treatment) and H3K9ac (N=2 biological replicates per treatment) in heads of flies carrying R50H05-GAL4>UAS-BCAT RNAi or control(R50H05-GAL4/w-;CS), quantified in right panel and normalized to low-BCAA treatment within genotypes (two-way ANOVAs with Tukey’s post-hoc, 10 heads were used per biological replicate). (D) Lifespans of flies carrying R50H05-GAL4>UAS-BCAT RNAi or controls (R50H05-GAL4/w-;CS or UAS-BCAT RNAi/w-;CS) on low- or high-BCAA diets (log-rank test, N=91–100, ***p<0. 001,). (E) Con-Ex measurement of 24-hr food intake after 5 days on BCAA diets in flies carrying R50H05-GAL4>UAS-BCAT RNAi or controls (R50H05-GAL4/w-;CS or UAS-BCAT RNAi/w-;CS) (two-way ANOVA, N=5–7 flies per treatment). Error bars represent the mean +/− SEM.

Figure 5.. Spatially and temporally distinct processes link hunger to feeding and aging.

(A) RT-qPCR quantification of relative mRNA abundance of Trh and Tph in Canton-S heads after 5–7 days on BCAA diets, values are normalized to low-BCAA treatment (one-way ANOVA. 10 heads were used per biological replicate, N=6 flies per treatment). (B-D) Immunostaining for serotonin in PLP neuron cell-bodies in flies carrying R50H05-GAL4>UAS-GTACR. eYFP after 5–7 days on BCAA diets. (B) Maximum intensity projection of ten 1. 5 μm stacks highlighting the PLP neurons (white box) in the posterior fly brain. Scale bar 100μm. (C) Representative confocal images of immunostaining for serotonin in PLP neurons of flies carrying R50H05-GAL4>UAS-GTACR. eYFP (green=nc82, magenta=serotonin, blue=DAPI, scale bar=10μm). Images are maximum intensity projections of ten 1. 5 μm stacks through the PLP neurons (D) Quantification of (C) as described in Methods (one-way ANOVA, 3 experimental replicates are pooled, N=5–7 brains per treatment). (E) Con-Ex measurement of 24-hr food intake after 5 days on BCAA diets in flies carrying Trh-GAL4>UAS-BCAT RNAi or control (Trh-GAL4/w-;CS) (two-way ANOVA, N=5 flies per treatment). (F-G) Con-Ex measurement of 24-hr food intake after 5 days or 2 weeks on BCAA diets in Canton-S flies; volume (F) and total amino acid intake (G) (two-tailed Student’s t-tests, N=8 flies per treatments). (H) Western blots of histone H3 and H3K9ac in heads of Canton-S flies after 5 days vs. 2 weeks on BCAA diets, quantified in right panel (one-tailed Student’s t-tests, 10 heads were used per biological replicate, 2 experimental replicates are pooled, N=3–5 biological replicates per treatment). Error bars represent the mean+/− SEM.