Distinct responses to non-autonomous UPRER mediated by glutamatergic and octopaminergic neurons

Abstract

The capacity to deal with stress declines during the aging process, and preservation of cellular stress responses is critical to healthy aging. The unfolded protein response of the endoplasmic reticulum (UPR^ER) is one such conserved mechanism, which is critical for the maintenance of several major functions of the ER during stress, including protein folding and lipid metabolism. Hyperactivation of the UPR^ER by overexpression of the major transcription factor, xbp-1s, solely in neurons drives lifespan extension as neurons send a neurotransmitter-based signal to other tissues to activate UPR^ER in a non-autonomous fashion. Previous work identified serotonergic, dopaminergic, and tyraminergic neurons in this signaling paradigm. To further expand our understanding of the neural circuitry that underlies the non-autonomous signaling of ER stress, we activated UPR^ER solely in glutamatergic, octopaminergic, and GABAergic neurons in C. elegans and paired whole-body transcriptomic analysis with functional assays. We found that UPR^ER-induced signals from glutamatergic neurons increased expression of canonical protein homeostasis pathways and octopaminergic neurons promoted pathogen response pathways, while more modest changes were detected in GABAergic UPR^ER activation. These findings provide further evidence for the distinct role neuronal subtypes play in driving the diverse response to ER stress.

Fig. 1. Dopaminergic and serotonergic xbp-1s together do not recapitulate pan-neuronal xbp-1s overexpression.

A comparison of differentially expressed genes (p value ≤ 0.01) between worms expressing xbp-1s pan-neuronally (rab-3p), and in either A dopaminergic (dat-1p) neurons, B serotonergic (tph-1p) neurons, or C concurrently in both dopaminergic and serotonergic (dat-1/tph-1p) neurons. For a complete list of differentially expressed genes, see Supplementary Data 1. For a complete list of genes represented in Venn Diagrams, see Supplementary Data 2. Overrepresentation analysis for all Venn Diagrams was performed using GeneOverlap in R. D Heat map of differentially expressed genes in worms expressing xbp-1s pan-neuronally, with corresponding expression levels in serotonin, dopamine, and both serotonin and dopaminergic xbp-1s expressing animals. Warmer colors indicate increased expression, and cooler colors indicate decreased expression. See Supplementary Data 3 for a list of gene names and expression values. E Schematic of each neuronal type explored in this study: glutamatergic, eat-4p (n = 79, location: head, pharynx, ventral nerve cord, and body, tail), octopaminergic, tbh-1p (n = 2, location: head), GABAergic, unc-25p (n = 32, location: head, ventral nerve cord, and body, tail). F Heat map of differentially expressed genes in worms expressing xbp-1s pan-neuronally (rab-3p and rgef-1p), dopaminergic (dat-1p), serotonergic (tph-1p), glutamatergic (eat-4p), octopaminergic (tbh-1p), and GABAergic (unc-25p) neurons. Warmer colors indicate increased expression, and cooler colors indicate decreased expression. **p < 0.01, ***p < 0.001. See Supplementary Data 3 for a list of gene names and expression values.

Fig. 2. Glutamatergic, octopaminergic, and GABAergic xbp-1s modulate distinct transcriptional pathways.

Volcano plots of whole-body genome-wide changes in gene expression upon xbp-1s overexpression in A glutamatergic, B octopaminergic, and C GABAergic neurons. Red dots indicate significantly differentially expressed genes with p value ≤ 0.01. See Supplementary Data 1. D Comparison of differentially expressed genes (p value ≤ 0.01) between worms expressing xbp-1s in glutamatergic, octopaminergic, and GABAergic neurons. For a complete list of differentially expressed genes, see Supplementary Data 2. E Heat map of XBP-1s target gene expression under neuronal, glutamatergic, octopaminergic, and GABAergic xbp-1s. Warmer colors indicate increased expression, and cooler colors indicate decreased expression. See Supplementary Data 3. Top ten most enriched gene ontology terms of all differentially expressed genes (both up- and downregulated) upon xbp-1s overexpression in F glutamatergic, G octopaminergic, and H GABAergic neurons. See Supplementary Data 4.

Fig. 3. Octopaminergic xbp-1s, but not glutamatergic or GABAergic xbp-1s, is sufficient to extend lifespan.

A Lifespan measurements of control (blue) and a mixed population of both “normal” and “stunted” growth octopaminergic xbp-1s (yellow, tbh-1p, mixed) animals. B Lifespan measurements of control (blue) and octopaminergic xbp-1s animals. Octopaminergic xbp-1s animals were separated into normal size (yellow, tbh-1p) and stunted growth (purple, tbh-1p, small). C Lifespan measurements of control (blue, light blue) and octopaminergic xbp-1s animals (tbh-1, yellow, orange) grown on either EV or xbp-1 RNAi. D Lifespan measurements of control (blue), glutamatergic xbp-1s animals (eat-4p, green), and GABAergic xbp-1s (unc-25p, pink) animals. E Lifespan measurements of control (blue), octopaminergic xbp-1s (tbh-1, yellow), xbp-1(zc12) (light blue), and octopaminergic xbp-1s; xbp-1(zc12) (orange) animals grown on EV. ***p < 0.001 with tbh-1 compared to control and tbh-1, xbp-1(zc12) compared to xbp-1(zc12). Note: This lifespan was performed in tandem with Fig. S4A and thus controls are identical in each figure. F Lifespan measurements of control (blue), octopaminergic xbp-1s (tbh-1, yellow), tbh-1(n3247) (light blue), and octopaminergic xbp-1s; tbh-1(n3247) (orange) animals grown on EV. Pan-neuronal xbp-1s (rab-3, dark red) is shown as a positive control to compare octopaminergic xbp-1s to pan-neuronal xbp-1s. Lifespans were scored every 2 days, and data are representative of three biological replicates (N). Sample size (n) is written next to each condition, followed by significance measured using Log-Rank testing: ns not significant, ***p < 0.001. rab-3 and tbh-1 statistics are compared against control, tbh-1, tbh-1(n3247) statistics are compared against tbh-1(n3247). All statistical analyses are available in Supplementary Data 5. G Representative fluorescent images of day 2 hsp-4p::GFP animals. Wild-type (control) and octopaminergic xbp-1s (tbh-1p) animals were grown on EV plates until day 1 of adulthood and moved onto plates containing 1% DMSO or 25 ng/µL tunicamycin (TM) for 24 h, then imaged on day 2 of adulthood. Data is representative of 4 independent trials. Scale bar is 500 µm. H Quantification of fluorescence integrated density normalized to area was performed across two technical replicates, each of four biological replicates, for a total of eight replicates. Lines represent the mean and standard deviation. *p ≤ 0.05, **p ≤  0.01 using a non-parametric, multi-comparison Kruskal–Wallis testing. Note: TM treatment results in a massive induction of the hsp-4p::GFP reporter and thus is on an entirely different scale (10x brighter, note y-axis label) than untreated animals. DMSO- and TM-treated animals are contrasted and plotted separately to allow visibility of fluorescence in both unstressed and stressed conditions. However, all conditions were imaged and quantified together.

Fig. 4. Glutamatergic and octopaminergic xbp-1s enhance pathogen resistance and increase pathogen apathy.

A Heat map of immune response (GO:0006955) gene expression under pan-neuronal (rgef-1p), glutamatergic (eat-4p), octopaminergic (tbh-1p), and GABAergic (unc-25p) xbp-1s overexpression. Warmer colors indicate increased expression, and cooler colors indicate decreased expression. See Supplementary Data 3. B Survival analysis of control (N2, blue), glutamatergic xbp-1s (green, eat-4p), octopaminergic xbp-1s (yellow, tbh-1p), or GABAergic xbp-1s (pink, unc-25p) on PA14 fast kill assay plates for 2, 4, 6, and 8 h. Each fast kill assay is comprised of three technical replicates per biological replicate and at least three biological replicates per condition. Results were analyzed via two-way ANOVA test; **(p < 0.01) ***(p < 0.001) ****(p < 0.0001). C Pathogen avoidance behavior of control (N2, blue), glutamatergic xbp-1s (green, eat-4p), octopaminergic xbp-1s (yellow, tbh-1p), or GABAergic xbp-1s (pink, unc-25p) during “forced” food choice assays measured at 1, 2, 3, 6, and 8 h time points. Each forced food choice assay is comprised of three technical replicates per biological replicate and at least three biological replicates per condition. Results were analyzed via two-way ANOVA test; **(p < 0.01) ***(p < 0.001) ****(p < 0.0001). D Representative brightfield and fluorescent images of adult worms grown on bacteria expressing mCherry. Animals are moved to OP50 plates for 2 h to remove mCherry-expressing bacteria from the intestine before imaging. Any remaining mCherry signal after OP50 clarification is a sign of bacterial colonization. Note: Glutamatergic xbp-1s and octopaminergic xbp-1s animals have a myo-2p::mCherry co-injection marker, so the red pharynx signal should be ignored. Scale bar is 500 µm. E Quantification of the percent of animals displaying intestinal bacterial colonization was performed across two technical replicates for each of three biological replicates for a total of six replicates. Lines represent the mean and standard deviation. *p ≤ 0.05, **p ≤  0.01, ns = p >  0.05 using a Mann–Whitney test.

Fig. 5. Glutamatergic and octopaminergic xbp-1s results in depletion of lipids.

A Heat map of lipid homeostasis (GO:0055088) gene expression under pan-neuronal (rgef-1p), glutamatergic (eat-4p), octopaminergic (tbh-1p), and GABAergic (unc-25p) xbp-1s overexpression. Warmer colors indicate increased expression, and cooler colors indicate decreased expression. See Supplementary Data 3. B Representative fluorescent micrographs of day 3 adult animals of control, glutamatergic xbp-1s (eat-4p), octopaminergic xbp-1s (tbh-1p), and GABAergic xbp-1s (unc-25p) taken on a stereomicroscope (top) and on a confocal microscope (bottom). Confocal imaging is performed in the same region of all worms, midway between the pharynx and the vulva. All images are contrast-matched. Scale bar represents 10 μm. Note: GABAergic xbp-1s animals have a myo-2p::GFP co-injection marker, so the green pharynx signal should be ignored. Scale bar is 500 µm. C Quantification of (B) of fluorescence integrated density normalized to area was performed across four biological replicates for a total of six replicates. Lines represent the mean and standard deviation. All conditions showed no significant difference (p > 0.05) compared to wild-type control using a Mann-Whitney test. D Representative images of day 3 adult animals of control, glutamatergic xbp-1s (eat-4p), octopaminergic xbp-1s (tbh-1p), and GABAergic xbp-1s (unc-25p) of ORO-stained lipids. Quantification of lipid staining as non-lipid depletion (black) and lipid depletion (gray). E Representative fluorescent micrographs of day 3 adult mRuby::HDEL of control, glutamatergic xbp-1s (eat-4p), and GABAergic xbp-1s (unc-25p), animals (left), day 3 mRuby::HDEL of control or xbp-1 RNAi used as a positive control to show morphological changes to ER (middle, white arrowheads), or day 3 adult mCherry::HDEL of control and octopaminergic xbp-1s (tbh-1p), animals (right). Images are representative of three independent biological replicates and are independently contrast-enhanced for each individual image. Confocal imaging is performed in the same region of all worms, midway between the pharynx and the vulva. Scale bar represents 10 µm.

Fig. 6. Glutamatergic and octopaminergic xbp-1s enhance proteostasis.

Note: GABAergic xbp-1s animals have a myo-2p::GFP co-injection marker, so the green pharynx signal should be ignored. A Representative images of protein accumulation in animals expressing intestinal polyglutamine repeats (vha-6p::polyQ40::YFP) in glutamatergic (eat-4p), octopaminergic (tbh-1p), or GABAergic (unc-25p) xbp-1s animals. All animals were imaged on days 1, 5, and 9 of adulthood. Images were captured using a Leica M205 stereo microscope. Scale bar is 500 µm. B Quantification of fluorescence integrated density normalized to area was performed across two technical replicates, each of three biological replicates, for a total of six replicates. Lines represent the mean and standard deviation. *p ≤ 0.05, **p ≤  0.01, ns = p >  0.05 using a Mann–Whitney test. C Representative images of a second distinct integration line of glutamatergic, octopaminergic, or GABAergic xbp-1s animals expressing intestinal polyglutamine repeats (vha-6p::polyQ40::YFP) imaged as per (A). Scale bar is 500 µm. D Quantification of fluorescence integrated density normalized to area was performed across three biological replicates. A Shapiro–Wilk test was used to determine normality, and a Student’s t-test was used to assess significance.

Fig. 7. Glutamatergic, octopaminergic, and GABAergic xbp-1s reduce protein accumulation into foci.

A Representative images of protein foci in animals expressing intestinal polyglutamine repeats (vha-6p::polyQ44::YFP) in glutamatergic (eat-4p), octopaminergic (tbh-1p), or GABAergic (unc-25p) xbp-1s animals and grown on either EV or xbp-1 RNAi. All animals were imaged on days 1, 5, and 9 of adulthood. Images were captured using a Leica M205 stereo microscope. Data are representative of three biological replicates Note: GABAergic xbp-1s animals have a myo-2p::GFP co-injection marker, so the green pharynx signal should be ignored. Scale bar is 500 µm. B Quantification of the number of foci in the intestine of each animal in N2 control, glutamatergic (eat-4p), or GABAergic (unc-25p) xbp-1s animals grown on either EV or xbp-1 RNAi. Data are averaged across all three biological replicates. Lines represent the mean and standard deviation. *p ≤ 0.05, **p ≤  0.01, ****p ≤ 0.0001, ns = p >  0.05 using a Mann–Whitney test. C Quantification of the number of foci in the intestine of each animal in N2 control or octopaminergic (tbh-1p) xbp-1s animals grown on either EV or xbp-1 RNAi. Data is averaged across all three biological replicates. Lines represent the mean and standard deviation. *p ≤ 0.05, **p ≤  0.01, ****p ≤ 0.0001, ns = p >  0.05 using a Mann–Whitney test. Note: One biological replicate was performed with N2, glutamatergic, octopaminergic, and GABAergic xbp-1s animals altogether, which is represented in the images in (A), while two replicates were performed with octopaminergic xbp-1s performed separately from glutamatergic or GABAergic xbp-1s animals. Therefore, quantification is shown separately in (B, C).