Restoration of cellular homeostasis after stress is essential for maintaining cellular function and overall organismal health. Stress affecting specific organelles can trigger adaptive responses that restore proper homeostatic states. One such pathway is the unfolded protein response of the endoplasmic reticulum (UPR ^ER ) which is regulated by three distinct branches, of which the most well-studied is the IRE-1 / XBP-1 pathway. Inositol-requiring enzyme 1 ( IRE-1 ) is an ER-membrane protein that responds to ER stress by dimerizing and undergoing autophosphorylation, thereby activating its RNAse domain. This results in splicing of the mRNA encoding the transcription factor X-box binding protein ( xbp-1 ) into its active xbp-1 s (spliced) form. XBP-1 s regulates expression of genes involved in ER quality control including protein chaperones, lipid metabolism, autophagy, ER-associated degradation (ERAD), and the ubiquitin proteosome system (UPS) (Dutta et al., 2022).

When xbp-1 s is ectopically expressed in neurons of the nematode, Caenorhabditis elegans , the UPR ^ER is activated in distal tissues via a non-autonomous stress signal (Taylor & Dillin, 2013). This neuron-to-body signal is mediated by a diverse array of neuronal subtypes, including serotonergic and dopaminergic neurons (Higuchi-Sanabria et al., 2020), tyraminergic neurons (Özbey et al., 2020), and glutamatergic, octopaminergic, and GABAergic neurons (Coakley, Hruby et al., 2025). Recently published work found that xbp-1 s overexpression in glutamatergic neurons (driven by the eat-4 promoter), and to a lesser degree in octopaminergic neurons (driven by the tbh-1 promoter) and GABAergic neurons (driven by the unc-25 promoter) improves the proteostatic capacity of the intestine (Coakley, Hruby et al., 2025). However, the mechanisms by which these neurons signal to the periphery and what proteostatic pathways are induced by this signal has yet to be explored.

To investigate how these neuronal subtypes signal to distal tissue to improve proteostasis, we previously utilized transgenic strains with xbp-1 s overexpression in glutamatergic, octopaminergic, and GABAergic neurons (referred to as glutamatergic, octopaminergic, or GABAergic xbp-1 s for simplicity) crossed with strains expressing fluorescently-labeled polyglutamine 44 repeats in the intestine (referred to as polyQ44). PolyQ44 forms insoluble protein foci (Morley et al., 2002) whose abundance serves as a proxy for assessing proteostatic capacity. Glutamatergic, octopaminergic, and GABAergic xbp-1 s overexpression was found to reduce the number of polyQ44 foci, suggesting improved proteostasis (Coakley, Hruby et al., 2025).

Since glutamatergic xbp-1 s animals displayed the largest changes (Coakley, Hruby et al., 2025), we sought to further explore the mechanism driving the improved proteostasis in these animals. First, we aimed to determine whether xbp-1 was required in peripheral tissues, as previous studies have shown that neuronal xbp-1 s activates peripheral xbp-1 to promote its beneficial effects. Therefore, we crossed glutamatergic xbp-1 s polyQ44 expressing animals into the xbp-1 ( zc12 ) mutant, which contains a nonsense mutation in xbp-1 (Calfon et al., 2002). Surprisingly, improved polyQ44 clearance was not dependent on peripheral xbp-1 , as the xbp-1 ( zc12 ) mutant failed to suppress the reduction in polyQ44 foci in glutamatergic xbp-1 s animals ( Fig. 1A ). Non-autonomous UPR ^ER signaling from both neurons (Imanikia et al., 2019) and glia (Metcalf et al., 2024) have been shown to activate HLH-30 , the C. elegans ortholog of Transcription Factor EB (TFEB), in peripheral tissues to enhance lysosomal function and improved autophagy. Therefore, to determine whether glutamatergic xbp-1 s similarly activates HLH-30 to promote peripheral proteostasis, we crossed glutamatergic xbp-1 s polyQ44 expressing animals into the loss-of-function hlh-30 ( tm1978 ) mutant strain (Settembre et al., 2013). We found that HLH-30 is indeed required for the improvement in peripheral proteostasis driven by glutamatergic xbp-1 ( Fig. 1B ).

To continue to dissect the neuronal signaling required for the improved proteostasis of glutamatergic xbp-1 s animals, we crossed this strain into an eat-4 ( ky5 ) mutant strain which prevents glutamatergic signaling (Lee et al., 1999). Surprisingly, the eat-4 ( ky5 ) mutation did not suppress the reduction of polyQ44 foci in glutamatergic xbp-1 s animals ( Fig. 1C ), suggesting the improvement in proteostasis does not require glutamatergic signaling. The eat-4 promoter drives xbp-1 s expression in 79 neurons, some of which are not exclusively glutamatergic (Loer & Rand, 2022). Therefore, improvement in proteostasis may be driven by a subset of these neurons, potentially through xbp-1 s overexpression in serotonergic neurons which was previously demonstrated to drive chaperone induction in a non-autonomous fashion (Higuchi-Sanabria et al., 2020). Neurons reported to be both glutamatergic and serotonergic in hermaphrodites include: AIM, ASG, and I5 (Loer & Rand, 2022). To investigate the contribution serotonergic signaling plays in glutamatergic xbp-1 -mediated proteostasis, we crossed glutamatergic xbp-1 s polyQ44 expressing animals into a tph-1 ( mg280 ) mutant strain which blocks serotonergic signaling (Sze et al., 2000). Only a very minor increase in polyQ44 foci was observed at day 9 of adulthood in the tph-1 ( mg280 ) mutant animals expressing glutamatergic xbp-1 s , suggesting that serotonergic signaling does not play a major role in glutamatergic xbp-1 -mediated improvements in peripheral proteostasis ( Fig. 1D ). Thus, the signal mediating this effect remains to be determined. Future studies specifically overexpressing xbp-1 s in subpopulations of glutamatergic neurons (e.g., AIM, ASG, and I5) could prove fruitful for further dissection of non-autonomous xbp-1 s -mediated proteostasis.

The IRE-1 / XBP-1 pathway is known to activate several different processes which have the potential to resolve proteotoxic stress, including autophagy, ERAD, and the UPS. To determine which of these processes are required for the improvement in proteostasis, we first analyzed the transcriptome of glutamatergic, octopaminergic, and GABAergic xbp-1 s animals using previously published bulk RNA sequencing data (Coakley, Hruby et al., 2025). Gene ontology analysis of the differentially expressed genes revealed terms associated with proteostatic pathways ( Fig. 1E ). In particular, glutamatergic xbp-1 s was associated with terms related to ERAD, while octopaminergic and GABAergic xbp-1 s was associated with terms related to autophagy ( Fig. 1E ). To test the dependency of improved proteostasis on these processes, we performed an RNA interference (RNAi) screen. Genes involved in the following processes were knocked down: IRE-1 / XBP-1 ( xbp-1 ), autophagy ( bec-1 , lgg-1 , atg-18 ; Chen et al., 2017), ERAD ( sel-11 , hrdl-1 ; Sasagawa et al., 2007), the UPS ( rpn-6.1 ; Vilchez et al., 2012), and lipophagy ( ehbp-1 ; Shi et al., 2010). Representative images for glutamatergic xbp-1 s animals are shown in Fig. 1F . Quantification of polyQ44 foci revealed several hits ( Fig. 1G- H ). Knockdown of xbp-1 s , atg-18 , sel-11 , and hrdl-1 increased the number of polyQ44 foci in the glutamatergic xbp-1 s animals, although not to the same levels as the control ( Fig. 1G ). This suggests the potential for both autophagy and ERAD to be involved in glutamatergic xbp-1 s -mediated clearance of polyQ44 foci, mirroring the RNA sequencing data ( Fig. 1E ). The requirement for HLH-30 , a known regulator of autophagy in C. elegans (Lapierre et al., 2013), in glutamatergic xbp-1 s -mediated improved proteostasis ( Fig. 1B ) provides further evidence that autophagy is involved in this process. In contrast, knockdown of autophagy, ERAD, the UPS, or lipophagy components did not affect the number of polyQ44 foci in octopaminergic or GABAergic xbp-1 s animals. In addition, lgg-1 knockdown resulted in a mild, but statistically significant, reduction in polyQ44 foci in octopaminergic xbp-1 s animals. Only knockdown of xbp-1 and atg-18 in glutamatergic xbp-1 s overexpressing animals were found to significantly increase polyQ44 foci in the normalized data ( Fig. 1H ). Trends for increased polyQ44 foci were present under sel-11 and hrdl-1 knockdown, although these were not statistically significant. No significant differences in normalized polyQ44 foci counts were found in octopaminergic and GABAergic xbp-1 s overexpressing animals. These data suggest that glutamatergic xbp-1 s promotes proteostasis potentially via more canonical pathways including autophagy and potentially ERAD, while octopaminergic and GABAergic neurons employ other pathways.

Overall, this study further elucidates the pathways through which glutamatergic, octopaminergic, and GABAergic xbp-1 s improve proteostasis. We find that while glutamatergic xbp-1 s animals utilize canonical proteostasis mechanisms to promote protein processing, octopaminergic and GABAergic neurons do not. Surprisingly, the improvement in proteostasis found in glutamatergic xbp-1 s animals did not require a canonical XBP-1 dependent mechanism in the periphery nor glutamatergic or serotonergic signaling. However, the suppression of HLH-30 did block improved proteostasis, suggesting peripheral activation of HLH-30 is a likely mediator. This study has several limitations: first, we utilize the polyQ44 animal as a proxy for proteostasis; however, this is an artificial system that may not always recapitulate endogenous proteostasis pathways. Additionally, we focused on glutamatergic signaling of our eat-4 p:: xbp-1 s animals, although the eat-4 promoter is expressed in several neuronal subtypes. Despite these caveats, our data align well with our transcriptomic analysis and revealed the potential for non-autonomous glutamatergic xbp-1 s signaling to be activating ERAD as well as autophagy through HLH-30 to clear polyQ44 foci. This work furthers our understanding of the potential mechanisms whereby non-autonomous xbp-1 s signaling drives proteostasis.

C. elegans maintenance

C. elegans were maintained on standard nematode growth medium (NGM) plates fed with OP50 E. coli B strain at 15°C. All strains used in this study are viable across the 15–20°C range; however, stocks were maintained at 15°C to slow population growth and reduce the frequency of passaging, thereby minimizing genetic drift. Experimental assays were conducted at 20°C. Standard NGM plates contained the following: Bacto-Agar (Difco) 2% w/v, Bacto Peptone 0.25% w/v, NaCl ₂ 0.3% w/v, 1 mM CaCl ₂ , 5 µg/ml cholesterol, 0.625 mM KPO ₄ pH 6.0, 1 mM MgSO ₄ . Animals were bleached and L1 arrested to synchronize age. Briefly, worms were collected into a 15 mL conical tube using M9 solution (22 mM KH ₂ PO ₄ monobasic, 42.3 mM NaHPO ₄ , 85.6 mM NaCl, 1 mM MgSO ₄ ) and exposed to a bleaching solution (1.8% sodium hypochlorite, 0.375 M NaOH in M9) until complete disintegration of carcasses. Intact eggs were then washed four times with M9 solution by centrifugation at 1,100 x g for 30 seconds. After the final wash, animals were L1 arrested by incubating overnight in M9 at 20°C on a rotator for a maximum of 24 hours. After arresting, worms were transferred to growth conditions at 20°C utilizing HT115 E. coli K strain carrying an empty pL4440 vector, referred to as empty vector (EV), for all experiments. NGM plates for experimental conditions contained the following: Bacto-Agar (Difco) 2% w/v, Bacto Peptone 0.25% w/v, NaCl ₂ 0.3% w/v, 1 mM CaCl ₂ , 5 µg/ml cholesterol, 0.625 mM KPO ₄ pH 6.0, 1 mM MgSO ₄ , 100 µg/mL carbenicillin, 1 mM IPTG. All experiments were performed using worms grown at 20°C. For all aging experiments, 100 μL of 10 mg/mL (+)-5-Fluorodeoxyuridine (FUDR) was placed directly on the bacterial lawn and worms were moved onto FUDR-containing plates on day 1 of adulthood. For all experiments including hlh-30 ( tm1978 ) mutants, animals were not L1 arrested as HLH-30 is necessary for survival during starvation (Murphy et al., 2019); after bleaching, eggs were placed directly onto experimental plates.

Stereoscope imaging

For whole-worm imaging, synchronized animals were grown on experimental NGM plates seeded with EV bacteria or RNAi bacteria. Animals were imaged on day 1 of adulthood and for aging experiments also imaged on day 5 and day 9 of adulthood. To induce paralysis, 13 worms were placed in a drop of 100 mM sodium azide in M9 on standard NGM plates without bacteria. Paralyzed animals were then lined up alongside each other and imaged on a Leica M205FCA automated fluorescent stereomicroscope running LAS X software and equipped with a standard GFP filter, Leica LED3 light source, and Leica K5 camera. For all imaging experiments, at least 3 biological replicates were performed with 2 technical replicates each, aside from Fig. 1G in which only 1 technical replicate per biological replicate was performed. vha-6p::Q44::YFP quantification was performed by counting the number of foci in individual worms using ImageJ Fiji (Schindelin et al., 2012), and statistical analysis was performed with GraphPad Prism 10 software using a Mann-Whitney test.

Statistics and reproducibility

All statistical analyses were performed using GraphPad Prism 10 software. No assumptions were made about data distribution, and p-values less than 0.05 were considered significant. For all experiments, a Mann-Whitney test was used to determine significance, aside from Fig. 1H in which a Shapiro-Wilk test was used to determine normality and an unpaired t-test used to determine significance. At least 3 biological replicates were performed for each experiment. Representative images from Fig. 1A- D are contained in Extended Data 1 and all raw data is contained in Extended Data 2 . RNA-seq analysis was performed using a previously published dataset (Coakley, Hruby et al., 2025) which is available through Annotare 2.0 ArrayExpress Accession E-MTAB-14132. Gene ontology analysis was performed using WormEnrichr (Chen et al., 2013; Kuleshove et al., 2016).

Extended Data 1. Representative images for Figure 1A- D: Representative images for all experiments in Figure 1 are provided as extended data for data transparency. (A) Representative images of protein foci within animals expressing intestinal polyglutamine 44 repeats ( vha-6p::polyQ44::YFP ) in control or glutamatergic ( eat-4 p ) xbp-1 s animals with and without a nonsense mutation in xbp-1 , xbp-1 ( zc12 ). (B) Representative images of protein foci within animals expressing intestinal polyglutamine 44 repeats ( vha-6p::polyQ44::YFP ) in control or glutamatergic ( eat-4 p ) xbp-1 s animals with and without loss-of-function hlh-30 ( tm1978 ) mutation. (C) Representative images of protein foci within animals expressing intestinal polyglutamine 44 repeats ( vha-6p::polyQ44::YFP ) in control or glutamatergic ( eat-4 p ) xbp-1 s animals with and without mutation blocking glutamatergic signaling, eat-4 ( ky5 ). Animals were imaged on days 1, 5, and 9 of adulthood. (D) Representative images of protein foci within animals expressing intestinal polyglutamine 44 repeats ( vha-6p::polyQ44::YFP ) in control or glutamatergic ( eat-4 p ) xbp-1 s animals with and without mutation blocking serotonergic signaling, tph-1 ( mg280 ). For all panels animals were imaged on days 1, 5, and 9 of adulthood, and images were captured using a Leica M205 stereo microscope. Scale bar represents 500 µm.