Cell non-autonomous control of autophagy and metabolism by glial cells

Abstract

Glia are the protectors of the nervous system, providing neurons with support and protection from cytotoxic insults. We previously discovered that four astrocyte-like glia can regulate organismal proteostasis and longevity in C. elegans. Expression of the UPR^ER transcription factor, XBP-1s, in these glia increases stress resistance, and longevity, and activates the UPR^ER in intestinal cells via neuropeptides. Autophagy, a key regulator of metabolism and aging, has been described as a cell autonomous process. Surprisingly, we find that glial XBP-1s enhances proteostasis and longevity by cell non-autonomously reprogramming organismal lipid metabolism and activating autophagy. Glial XBP-1s regulates the activation of another transcription factor, HLH-30/TFEB, in the intestine. HLH-30 activates intestinal autophagy, increases intestinal lipid catabolism, and upregulates a robust transcriptional program. Our study reveals a novel role for glia in regulating peripheral lipid metabolism, autophagy, and organellar health through peripheral activation of HLH-30 and autophagy.

Keywords: Biological sciences; Cell biology; Functional aspects of cell biology; Neuroscience.

Graphical abstract

Figure 1

Expression of XBP-1s in CEPsh glia modulates peripheral lipid metabolism and ER remodeling in C elegans (A) Fluorescent light micrographs of wild-type (N2) and glial XBP-1s animals stained with BODIPY 493/503 dye and imaged at day 2 of adulthood. Scale bar, 250 μm. (B) Whole animal fluorescence intensity quantification of BODIPY 493/503 dye in day 2 adults in wild-type (gray) and glial XBP-1s animals (green) using COPAS BioSorter. p < 0.0001 (∗∗∗∗), using the non-parametric Mann-Whitney test. Plots are representative of three biological replicates, n = 362 (wild type) and n = 302 (glial XBP-1s). Results were normalized to the mean fluorescent intensity of wild-type animals. Each dot represents one animal, and the boxplot shows the median (horizontal line), the first and third quartiles (box), and the smallest and largest data points (whiskers). (C) Fluorescent light micrographs of wild-type and glial XBP-1s animals of intestinal lipid droplets (dhs-3p::dhs-3::GFP) imaged at day 2 of adulthood. Scale bar, 250 μm. (D) Whole animal fluorescence intensity quantification of intestinal dhs-3p::dhs-3::GFP lipid droplet marker in wild-type (gray) and glial XBP-1s animals (green) using COPAS BioSorter. p < 0.0001 (∗∗∗∗), using the non-parametric Mann-Whitney test. Plots are representative of three biological replicates, n = 254 (wild type) and n = 532 (glial XBP-1s). Results were normalized to the mean fluorescent intensity of wild-type animals transgenic for dhs-3p::dhs-3::GFP. (E) Representative Airyscan micrographs of dhs-3p::dhs-3::GFP labeled lipid droplets from anterior intestines of day 2 wild-type and glial XBP-1s adult animals. Arrowheads point at lipid droplets: scale bar, 10 μm. (F) Quantification of intestinal dhs-3p::dhs-3::GFP labeled lipid droplets in wild-type (gray) and glial XBP-1s (green) animals. Lipid droplets were counted from three independent replicates using ImageJ and expressed as density. Density was determined by dividing the number of lipid droplets by the total area expressing dhs-3p::dhs-3::GFP in each respective image. N = 15 (wild type) and N = 22 (glial XBP-1s). p < 0.0001 (∗∗∗∗), using the non-parametric Mann-Whitney test. (G) Representative Airyscan micrographs of lmp-1::GFP labeled lysosomes in the anterior intestine of day 2 adults in both wild-type and glial XBP-1s animals. Arrowheads point at lysosomes: scale bar, 10 μm. (H) Quantification of intestinal lmp-1::GFP labeled lysosomes in wild-type (gray) and glial XBP-1s (green) animals. Lysosomes were counted from three independent replicates using ImageJ and expressed as density. Density was determined by dividing the number of lysosomes by the total area that expressed lmp-1::GFP in each respective image. N = 15 (wild type) and N = 16 (glial XBP-1s). p < 0.0001 (∗∗∗∗), using the non-parametric Mann-Whitney test. (I) Representative confocal micrographs of intestinal ER morphology (vha-6p::ERss:mRuby:HDEL, ERss = hsp- 4 ER signal sequence), wild-type and glial XBP-1s animals were imaged at day 2 of adulthood. Arrowheads mark ER puncta. Scale bar, 10 μm. Scale bar of inset, 5 μm. (J) Electron micrographs of intestine from wild-type and glial XBP-1s animals at day 2 of adulthood. Imaging was replicated in triplicate, for a total of 15–20 animals being imaged per condition. Arrowheads mark rough endoplasmic reticulum. Scale bar, 1μm. Scale bar of inset, 0.2 μm.

Figure 2

Lifespan screen of key metabolic regulators reveals HLH-30 requirement for longevity of glial XBP-1s animals (A–E) Survival of wild-type (N2) and glial XBP-1s animals on control (HT115 E. coli expressing empty vector), daf-16 (A), aak-1 (B), aak-2 (C), pha-4 (D), and hlh-30 (E) RNAi from L4 at 20°C. Lifespan is representative of three replicates. Graphs were plotted as Kaplan-Meier survival curves and p values were calculated by Mantel-Cox log rank test. See Table S1 for lifespan statistics. (F) Survival of wild-type and glial XBP-1s animals on control RNAi at 20°C, with and without hlh-30(tm1978) mutation. The graph was plotted as Kaplan-Meier survival curves and p values were calculated by the Mantel-Cox log rank test. See Table S1 for lifespan statistics.

Figure 3

Glial XBP-1s cell non-autonomously activates intestinal HLH-30 which transcriptionally regulates genes involved in lipid catabolism and lysosomal biogenesis (A) Fluorescent light micrographs of wild-type and glial XBP-1s animals transgenic for fluorescently tagged HLH-30 (HLH-30:GFP) grown on OP50 and imaged at day 2 of adulthood. HLH-30 translocates to the nucleus in glial XBP-1s animals. The inset shows posterior intestine. Scale bar, 250μm. Scale bar of inset, 100μm. (B) Schematic showing how HLH-30:GFP nuclear localization was scored. Categories include no nuclear enrichment (gray), weak nuclear enrichment (green), medium nuclear enrichment (blue), and strong nuclear enrichment (purple). Yellow circle indicates nucleolus, which transcription factors are excluded from. (C) Representative Airyscan micrographs of posterior intestinal cells of wild-type and glial XBP-1s animals transgenic for HLH-30:GFP grown on OP50 and imaged at day 2 of adulthood. Colored inset squares are representative of the scoring categories from (B). Scale bar, 10μm. Intestinal cells are outlined with a dashed gray line, intestinal nuclei are outlined with a dashed white line and intestinal nucleoli are outlined with a dashed yellow line. (D) Nuclear translocation of HLH-30:GFP in anterior intestinal cells of day 2 adult wild-type and glial XBP-1s animals with and without the loss-of-function unc-31 mutant, unc-31(e928), N > 40 animals per condition from three independent replicates. Statistics done by Chi-squared test for independence with adjusted residual and Bonferroni correction, statistics shown in Table S4. (E) Representative fluorescent micrograph of wild-type animals transgenic for HLH-30:GFP treated with either DMSO or tunicamycin. Inset shows posterior intestine. Scale bar, 250 μM. Scale bar of inset, 100 μM. Image is representative of two independent replicates, in n > 20 no animals showed nuclear localization. (F) Volcano plot demonstrating magnitude (log2[fold change]) and significance (-log10[adjusted p value]) of changes in gene expression from whole-animal RNA sequencing of glial XBP-1s versus wild-type animals at day 2 of adulthood. Differentially expressed genes (DEGs) shown in red (upregulated, adjusted p value <0.05 and log2[fold change] > 0.5, n = 86) and blue (downregulated, adjusted p value<0.05 and log2[fold change] < 0.5, n = 22). Labeled genes and DEGs with dark red (upregulated DEGs) and dark blue (down-regulated DEGs) dots correspond to HLH-30 targets which are labeled with their corresponding gene. (G) Comparison of log2(fold change) of the 86 upregulated DEGs in glial XBP-1s animals compared to wild-type animals (top) compared to the log2(fold change) of these genes in glial XBP-1s animals with a loss-of-function hlh-30 mutation, hlh-30(tm1978), compared to glial XBP-1s animals alone (bottom). log2(fold change) is color-coded via a heatmap from warm (upregulated) to cool (downregulated). Green dots above the heatmap represent HLH-30 target genes. Statistics shown in Table S5.

Figure 4

Glial XBP-1s cell non-autonomously activates peripheral autophagy, which is required for lifespan extension and reduced lipid levels (A) Representative Airyscan micrographs from intestines of day 2 and day 5 wild-type and glial XBP-1s transgenic animals expressing the tandem autophagy reporter mCherry::GFP::lgg-1 (green, GFP, magenta, mCherry). White arrowheads indicate autophagosomes, and yellow arrowheads indicate autolysosomes. Scale bar, 10 μm. (B and C) Quantification of mCherry puncta co-localized with GFP (autophagosomes [AP]) (B) or containing mCherry alone (autolysosome [AL]) (C) in the intestine of wild-type and glial XBP-1s transgenic animals at day 2 and day 5 of adulthood. Data are from three independent experiments, each with $\ge$ 5 animals. Statistics by Kruskal-Wallis with Dunn’s multiple comparison test, p < 0.0001 (∗∗∗∗). (D) Representative Airyscan micrographs of dhs-3p::dhs-3::GFP labeled lipid droplets from the intestine of day 2 wild-type and glial XBP-1s adult animals grown on HT115 E. coli expressing control or hlh-30 RNAi. Scale bar, 10 μm. Scale bar of inset, 5 μm. (E) Changes in lipid droplet density of wild-type and glial XBP-1s animals grown on control or hlh-30 RNAi from (D). Boxplot shows median, whiskers are minimum to maximum values, each dot is representative of one animal. Statistics by Kruskal-Wallis with Dunn’s multiple comparison test, p < 0.001 (∗∗∗), p > 0.05 (ns). N > 15, and from three independent replicates. (F) Representative fluorescent micrographs of age-dependent accumulation of transgenic polyQ₄₄-YFP aggregates in wild-type or glial XBP-1s animals grown on control or hlh-30 RNAi. Scale bar, 1 mm. Scale bar of inset, 200 μm. (G) Quantification of age-dependent polyQ₄₄ aggregates from (F), grouped into animals with 0 (gray), 1–3 (green), 4–9 (blue), or >10 (purple) aggregates. N > 100 per condition. Data are representative from three independent replicates. Statistics done by Chi-squared test for independence with adjusted residual and Bonferroni correction, statistics shown in Table S4. (H and I) Survival of wild-type and glial XBP-1s animals on control, bec-1 (D) or atg-18 RNAi (E) at 20°C. Lifespan is representative of three independent replicates. See Table S1for lifespan statistics. (J) Representative Airyscan micrographs of transgenic dhs-3p::dhs-3::GFP labeled lipid droplets from the intestine of day 2 wild-type and glial XBP-1s adults grown on control or bec-1 RNAi. Scale bar, 10 μm. Scale bar of inset, 5 μm. (K) Changes in lipid droplet density of wild-type and glial XBP-1s animals grown on control or bec-1 RNAi from (J). Boxplot shows median, whiskers are minimum to maximum values. Statistics by Kruskal-Wallis with Dunn’s multiple comparison test, p < 0.001 (∗∗∗). N > 15 per condition from three independent replicates.

Figure 5

Macroautophagy is required for changes in intestinal ER morphology and lipid depletion in glial XBP-1s animals (A) Representative Airyscan micrographs from intestine of day 2 wild-type and glial XBP-1s animals transgenic for the ER marker vha-6p::ERss:mRuby:HDEL, animals were grown on control or the corresponding RNAi. ER puncta is denoted by arrowheads. (B) Quantification of ER puncta from (A). Density was determined by counting the number of ER puncta per area of intestine expressing vha-6p::ERss:mRuby:HDEL. N > 15 per condition from three independent replicates. Statistics by Kruskal-Wallis with Dunn’s multiple comparison test, p < 0.001 (∗∗∗), p < 0.0001 (∗∗∗∗). (C) Representative Airyscan micrographs from intestines of day 2 wild-type and glial XBP-1s animals transgenic for the ER marker vha-6p::ERss:mRuby:HDEL (pseudo-colored in magenta) and stained with Lysotracker Blue DND-22 (pseudo-colored in green). Arrowheads mark colocalization. Scale bar, 10 μm. Scale bar of inset, 5 μm. (D) Representative Airyscan micrographs from intestine of day 2 wild-type and glial XBP-1s animals transgenic for the ER marker vha-6p::ERss:mRuby:HDEL (magenta) and autophagosome marker LGG-1:GFP (green). Arrowheads mark colocalization. Scale bar, 10 μm. Scale bar of inset, 5 μm. (E) Quantification of density of ER puncta and lysotracker colocalization. Statistics by Mann-Whitney test, p < 0.0001 (∗∗∗∗). N > 15 per condition from three independent replicates. (F) Quantification of density of ER puncta and autophagosome colocalization. Statistics by Mann-Whitney test, p < 0.0001 (∗∗∗∗). N > 15 per condition from three independent replicates. (G) Electron micrographs of intestine from wild-type and glial XBP-1s animals at day 2 of adulthood grown on either empty vector control or bec-1 RNAi. Imaging was replicated in triplicate for a total of 15–20 animals being imaged per condition. Arrowheads mark rough endoplasmic reticulum. Scale bar, 1μm. Scale bar of inset, 0.2 μm.

Figure 6

Schematic of glial XBP-1s effects on peripheral metabolism and autophagy Expression of xbp-1s in the four CEPsh glia induces activation of the UPR^ER and the transcription factor HLH-30/TFEB in the periphery, reliant on exocytosis of dense core vesicles (unc-31). HLH-30 regulates the transcription of lipid catabolism and lysosomal biogenesis genes in glial XBP-1s animals. HLH-30 is required for increased lipid catabolism, activated autophagy, decreased intestinal protein aggregation, lifespan extension, and ER remodeling in glial XBP-1s animals. Autophagy induction is also required for longevity, lipid catabolism, and ER remodeling phenotypes in the glial XBP-1s paradigm. These data suggest that glia provide cell non-autonomous control over peripheral autophagy and metabolic adaptation during stress.