Impaired Endosomal Recycling of Signaling Receptors Activates an Extracellular UPR

7 Mitochondrial dysfunction and extracellular protein aggregation occur in neurodegenerative diseases 8 such as Alzheimer’s disease (AD). However, it remains unclear if these processes are functionally 9 linked. Here, we identify a signaling pathway that is activated upon accumulation of aggregation-prone 10 proteins in the extracellular space. We find that the transcription factor ATFS-1, which regulates the 11 mitochondrial unfolded protein response, also regulates transcripts required for endosomal recycling, 12 multiple plasma membrane-localized signaling receptors, and secreted proteins that bind aggregation-13 prone proteins in the extracellular space, including transthyretin and A β, and promote their 14 degradation. Interestingly, A β(1-42) aggregation induces atfs-1-dependent transcription by promoting 15 degradation of the bZIP protein ZIP-3, which antagonizes ATFS-1. ZIP-3 accumulates in the cytosol 16 when it is phosphorylated by kinases that function downstream of plasma membrane-localized 17 signaling receptors, including the WNT and glutamate receptors. Upon ligand binding, the signaling 18 receptors stimulate the cognate kinase, many of which we found phosphorylate ZIP-3, impeding ZIP-3 19 degradation, allowing it to antagonize atfs-1-dependent transcription. However, accumulation of 20 aggregation-prone proteins such as A β (1–42) causes endosomal swellin g, which impairs endosomal 21 recycling, instead diverting signaling receptors to lysosomes for degradation. In turn, the depletion of 22 signaling receptors reduces the level of ZIP-3 phosphorylation, resulting in ZIP-3 degradation and 23 activation of atfs-1 -dependent transcription, which promotes extracellular proteostasis. Our findings 24 uncover an unexpected coupling between endocytic quality control and mitochondrial signaling, 25 revealing a circuit that preserves extracellular proteostasis and promotes organismal resilience. 26 27

C. elegans, extracellular proteostasis, protein aggregation, Alzheimer’s Disease, Amyloid 28 beta, ATFS-1, ZIP-3, extracellular unfolded protein response, endosomal recycling 29 Corresponding author: [email protected] 30 31 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 2

32 Canonical proteostasis pathways, including the endoplasmic reticulum unfolded protein response 33 (UPRER)1, heat shock response (HSR)2, and mitochondrial unfolded protein response (UPRmt)3, rely on 34 organelle-specific sensors and transcription factors to restore local protein homeostasis; however, 35 whether and how these systems detect and respond to proteostasis failure in the extracellular space 36 remains unclear. Disruption of extracellular proteostasis is a defining feature of neurodegenerative 37 diseases such as Alzheimer’s disease (AD), which is characterized by the accumulation of 38 extracellular amyloid-β (Aβ ) aggregates and widespread mitochondrial dysfunction 4–6. Analyses of AD 39 patient samples and disease models have revealed extensive metabolic reprogramming 7,8, including 40 increased glycolysis and activation of the UPR mt. Despite these observations, the mechanistic links 41 between extracellular proteostasis perturbations and mitochondrial stress signaling remain poorly 42 understood. 43 The UPR mt is a mitochondria-to-nucleus signaling pathway that promotes mitochondrial 44 recovery by activating transcription of numerous genes that promote mitochondrial biogenesis in 45 response to mitochondrial stress 3. In Caenorhabditis elegans , the UPR mt is mediated by the bZIP 46 transcription factor ATFS-1, which under basal conditions is mostly imported into mitochondria and 47 degraded9,10. However, during mitochondrial stress, ATFS-1 fails to import into mitochondria and 48 translocate to the nucleus, where it induces transcription of genes that promote mitochondrial 49 proteostasis and biogenesis 9–11. Importantly, ATFS-1 activity is also negatively regulated by the bZIP 50 protein ZIP-3, which forms a heterodimer with cytosolic ATFS-1 12,13. Inhibition of ZIP-3 expression is 51 sufficient to activate ATFS-1-dependent transcription, independent of mitochondrial dysfunction. 52 Endosomal dysfunction is an early and prominent pathological feature of AD 14,15. In patient 53 samples and disease models, endosomal swelling disr upts the recycling of endocytosed cargo, such 54 as membrane-localized signaling receptors, from endosomes back to the plasma membrane. 55 Endosomal recycling is required to maintain plasma membrane composition, sustain receptor-56 mediated signaling, and preserve extracellular homeostasis 14,16–18. In AD, accumulation of A β (1-42) 57 within endosomes promotes endosomal swelling, impairing car go recycling and diverting signaling 58 receptors to lysosomes for degradation, thereby attenuating cell-surface signaling14,15,19,20. 59 The retromer complex, consisting of VPS26, VPS29, and VPS35, mediates endosomal 60 recycling16, and perturbations in this pathway are causally implicated in AD 15,21. Mutations in VPS35 61 are associated with familial forms of AD 15,21 and Parkinson’s disease 22 and have also been shown to 62 cause mitochondrial dysfunction 23, yet how defects in endosomal trafficking intersect with 63 mitochondrial stress signaling remains unclear. 64 Here, we identify ATFS-1 as a regulator of extracellular proteostasis through transcriptional 65 control of endosomal recycling and sorting components, as well as secreted extracellular chaperones 66 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 3 and proteases that limit extracellular protein aggregation. We define a mechanism by which 67 extracellular protein aggregation promotes ZIP-3 degradation at recycling endosomes, leading to 68 ATFS-1 activation and induction of a transcriptional program that restricts extracellular aggregate 69 accumulation and preserves plasma membrane receptor expression. ZIP-3 degradation is inhibited by 70 phosphorylation mediated by kinases acting downstream of cell-surface receptors, including glutamate 71 receptors. In contrast, disruption of endosomal recycling or receptor-mediated signaling promotes ZIP-72 3 degradation, thereby activating ATFS-1. Together, these findings define a stress-responsive 73 signaling network in which endosomal trafficking, receptor signaling, and ATFS-1-dependent 74 transcription maintain or restore extracellular proteostasis. 75 76

77 ATFS-1-dependent transcription impairs the accumulation of extracellular protein aggregates 78 We previously demonstrated that ATFS-1 promotes transcription of genes required for 79 mitochondrial biogenesis, structure, and function. Interestingly, many genes encoding proteins 80 secreted to the extracellular space are reduced in atfs-1(null) worms, suggesting that ATFS-1 also 81 regulates transcription of genes that affect cellular activities beyond mitochondria (Supplementary 82 Figure 1A-B, Supplementary Table 1). A previous study identified 57 genes that impair extracellular 83 protein aggregation in C. elegans 24. Interestingly, atfs-1 is required to express many of these 57 84 extracellular protein aggregation regulators that promote extracellular proteostasis 24 (Supplementary 85 Figure 1A and 1D). 86 To gain insight into the role of ATFS-1 in regulating these genes that promote extracellular 87 proteostasis, we generated GFP-tagged reporter strains expressing W01A8.6 (an extracellular 88 metalloproteinase), F54E2.1 (a membrane glycoprotein), and dnj-20 (homolog of mammalian 89 extracellular chaperone DNAJB11) 25 (Figure 1A). RNAi inhibition of these genes resulted in elevated 90 extracellular protein aggregates (Figure 1A-B). As ATFS-1 is required for the expression of W01A8.6, 91 F54E2.1, and dnj-20, we hypothesized that the expression of these genes would respond to OXPHOS 92 perturbation, which increases ATFS-1-dependent transcription 3,11,26. As expected, W01A8.6pr::GFP, 93 F54E2.1pr::GFP, and dnj-20 pr::GFP transgenic strains exhibited increased GFP fluorescence when 94 grown on RNAi targeting cco-1 (complex IV gene) (Figure 1A-C). 95 To determine if atfs-1 promotes extracellular proteostasis, we generated atfs-1(null) worms 96 expressing either LBP-2::RFP or A β (1-42)-GFP. We find that atfs-1(null) worms have increased 97 aggregation, indicating that ATFS-1 is required to impair extracellular protein aggregation (Figure 1E-98 H). In contrast, cco-1 inhibition, which activates ATFS-1-dependent transcription, reduced the 99 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 4 accumulation of LBP-2::RFP and A β (1-42)-GFP aggregates (Figure 2D and supplementary Figure 100 2E). Together, these findings indicate that atfs-1 is required to impair extracellular protein aggregation. 101 Inhibition of ZIP-3 induces ECR transcription via ATFS-1 activation 102 We previously identified the bZIP protein ZIP-3 as a negative regulator of ATFS-1-dependent 103 transcription12. Transcriptomic analyses confirmed that ZIP-3 represses transcription of ATFS-1 target 104 genes while also regulating a broader transcriptional network independent of ATFS-1 and the 105 mitochondrial unfolded protein response (UPR mt)11,12. Consistent with a functional interaction, prior 106 studies demonstrated physical binding between ZIP-3 and ATFS-1 13. Notably, ZIP-3–regulated genes 107 were enriched for pathways involved in endosomal transport, endosome-to-lysosome trafficking, 108 endocytic recycling, regulation of endocytosis, and lysosomal transport (Figure 2A). 109 To determine the impact of ZIP-3 on extracellular proteostasis, we examined aggregation of 110 LBP-2::RFP and A β (1–42)::GFP in zip-3(null) worms. Interestingly, ZIP-3 deletion worms have 111 reduced aggregates than wild-type animals (Figure 2E-F), even under elevated temperature stress at 112 25°C (Figure 2B). In contrast, inhibition of atfs-1 increased aggregation in zip-3 –deletion animals, 113 suggesting that atfs-1 is required to promote extracellular proteostasis in zip-3(null) worms (Figure 2C-114 D). Consistent with improved extracellular protein aggregation, zip-3(RNAi) also extended lifespan at 115 25°C (Figure 2G). 116 To further assess ZIP-3 function in extracellular proteostasis, we generated a transgenic strain 117 expressing LBP-2::GFP levels under the muscle myosin ( myo-3) promoter. We confirmed that LBP-2 118 is secreted into the pseudocoelomic space, internalized by coelomocytes, and forms aggregates by 119 day 1 of adulthood, coinciding with reduced lifespan (Supplementary Figure 2A-B). The level of 120 aggregation is markedly enhanced compared to the lbp-2 pr::lbp-2-RFP (LBP-2-RFP) strain used 121 before. Strikingly, zip-3(RNAi) reduced LBP-2::GFP aggregation and extended lifespan (Figure 2H and 122 Supplementary Figure 2A). Supporting these findings, RNA-seq analysis revealed that zip-3(null) 123 mutants increased the expression of genes involved in endocytosis and endosomal recycling, 124 including 52 genes encoding secreted extracellular proteins such as collagens, oxidoreductases, 125 proteases, and protease inhibitors (Supplementary Table 2). Among these are seven genes previously 126 shown to limit extracellular aggregation 24,27, which are reduced in atfs-1(null) mutants but elevated in 127 zip-3 deletion mutants (Supplementary Figure 1E-F). Together, these results demonstrate that ZIP-3 128 deletion enhances ATFS-1–dependent extracellular proteostasis, linking transcriptional activation of 129 genes required to limit extracellular protein aggregation, promote endosomal trafficking, and improve 130 organismal survival. 131 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 5 ATFS-1 activation prevents the accumulation of extracellular protein aggregates and limits 132 neurodegeneration 133 To assess the physiological relevance of ZIP-3 in regulating extracellular proteostasis, we used 134 transgenic C. elegans models that secrete either human transthyretin (TTR) or A β (1–42) into the 135 extracellular space. To model TTR-associated proteotoxicity, we used the aggregation-prone 136 TTR(V30M) variant under the unc-54 promoter, while A β (1–42) is expressed in either neuronal or 137 muscle tissues. In both models, the secreted proteins accumulated in the extracellular space, 138 providing robust systems to examine mechanisms governing extracellular protein quality control28,29. 139 We evaluated functional outcomes, including lifespan, body bending, and thrashing behavior. 140 RNAi-mediated inhibition of zip-3 significantly rescued motor dysfunction and extended lifespan in 141 both A β (1–42)- and TTR(V30M)-secreting animals (Figure 3A-D and supplementary Figure 2C-D). 142 Given that zip-3 inhibition and ATFS-1 activation are each associated with improved healthspan, we 143 next examined whether extracellular aggregation could activate ATFS-1–dependent transcription. We 144 found that the neuronal and muscle A β (1–42)-expressing strains and TTR(V30M) worms exhibited 145 increased hsp-6p::GFP expression, indicative of ATFS-1–dependent transcriptional activation (Figure 146 3E). Consistent with this, neuronal and muscle-secreted A β worms have reduced ZIP-3 levels (Figure 147 3G-H). Moreover, multiple genes that limit extracellular protein aggregation and have a reduced 148 expression in atfs-1(null) animals are found to be induced in both A β (1–42) and TTR(V30M) 149 transgenic strains (Supplementary Figure 1J-K). 150 To directly assess neurodegenerative phenotypes in the presence of extracellular aggregation, 151 we utilised the FLP neuronal marker des-2p::Myristoyl::GFP in the TTR(V30M) strain, shown to have 152 impaired nociception and defective dendritic morphology 28. We find that inhibiting zip-3 improved the 153 dendritic morphology in these worms (Figure 3F). We also examined GABAergic and dopaminergic 154 neurons using unc-47 pr::GFP and dat-1 pr::GFP reporters 30,31. Both zip-3 overexpression and atfs-1 155 deletion enhanced neuronal puncta formation by day 1 of adulthood, consistent with accelerated 156 neurodegeneration (Supplementary Figure 2F-G). Consistent with this protective role of ATFS-1 and 157 the expression of ZIP-3 and ATFS-1 in the head (Supplementary Figures 1C and 1L), inhibition of zip-158 3 specifically in the neurons extended the lifespan of worms (Figure 3I). Together, these results show 159 that extracellular proteotoxic stress triggers an ATFS-1–dependent transcriptional program that 160 reduces extracellular aggregation, a process otherwise inhibited by ZIP-3. 161 Endocytic recycling is required to limit extracellular aggregates and maintain ZIP-3 levels. 162 Endosomal recycling is essential for maintaining plasma membrane integrity and extracellular 163 protein quality control. Also, impaired endosomal recycling is a common pathology in AD progression. 164 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 6 To determine whether A β (1–42) expressing worms impact the endos omal recycling, we utilised the 165 late endosomal marker RAB-7::GFP. Expression of A β (1–42) in C. elegans resulted in a significant 166 enlargement of RAB-7::GFP-labeled late endosomes (Figure 4A). This phenotype aligns with the 167 endocytic trafficking impairments typically induced by extracellular proteotoxic stress 15. Knockdown of 168 vps-35, a core component of the retromer complex, further exacerbated endosomal swelling, 169 indicating that impaired retromer-dependent recycling intensifies endosomal dysfunction (Figure 4A). 170 Interestingly, atfs-1 mutants exhibited disrupted endosomal morphology both basally and following 171 vps-35 depletion (Figure 4B), suggesting that ATFS-1 is required to preserve endosomal architecture. 172 Notably, RNAi-mediated inhibition of zip-3 or cco-1 reduced the endosomal enlargement in A β (1–42)–173 expressing worms (Figure 4C), consistent with a protective role of ATFS-1 activation in maintaining 174 endosomal homeostasis under extracellular proteotoxic stress. 175 The retromer complex, composed of VPS- 26, VPS-29, and VPS-35, coordinates endosomal 176 recycling to the plasma membrane 16,17 and retrograde trafficking to the trans-Golgi network via RME-8 177 and HSP-117. To test whether disruption of these pathways activates ATFS-1, we individually inhibited 178 the expression of retromer components ( vps-26, vps-29, and vps-35 ), retrograde regulators ( rme-8 179 and hsp-1), and additional endosomal recycling-associated factors ( rme-1, sdpn-1, alx-1, rab-10, rab-180 11.1, and rab-35)32. Each perturbation robustly induced the ATFS-1 reporter hsp-6pr::GFP (Figure 4D). 181 Consistent with ATFS-1–dependent extracellular proteostasis, inhibition of vps-26, vps-35, or rme-8 182 also induced expression of extracellular genes, including the protease W01A8.6pr::GFP and the 183 chaperone dnj-20pr::GFP (Figure 4E and supplementary Figure 1G-I). Importantly, mitochondrial 184 membrane potential remained intact following rme-8 knockdown, indicating that ATFS-1 activation in 185 this context occurs independently of overt mitochondrial dysfunction (Supplementary Figure 3B). 186 To understand whether ATFS-1-dependent transcription is specific to perturbation in the 187 retromer complex or in general related to the health of endosomal recycling, we perturbed the 188 phosphatidylinositol-3-phosphate [PI(3)P] synthesis pathway. PI(3)P is a key lipid signal that defines 189 early endosomes and drives endosomal recycling and is generated primarily by class III PI3-kinase 190 VPS-34, in complex with VPS-15 and BEC-1 3233. Consistent with a role for PI(3)P in endosomal 191 recycling34, RNAi-mediated knockdown of vps-34 , as well as piki-1 (phosphoinositide-3-kinase), 192 robustly induced hsp-6p::GFP expression (Figure 4D). In contrast, inhibition of the PI(3)P phosphatase 193 mtm-6, which increases PI(3)P levels, did not induce hsp-6pr::GFP expression (Figure 4D). We next 194 examined whether disruption of endosomal trafficki ng activates ATFS-1 through ZIP-3 destabilization. 195 ZIP-3::GFP abundance was significantly reduced following knockdown of rme-8, vps-26, vps-29, or 196 vps-35, a result confirmed in animals expressing endogenously tagged ZIP-3–FLAG (Figure 4F and 197 supplementary Figure 3A). Similarly, depletion of vps-34 caused enlargement of early and late 198 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 7 endosomes and decreased ZIP-3 levels, indicating that PI(3)P-dependent trafficking and retromer 199 function are required to maintain ZIP-3 stability under homeostatic conditions (Figure 4G). Together, 200 these results demonstrate that disruption of endosomal recycling, through impaired retromer activity or 201 PI(3)P synthesis, destabilizes ZIP-3, relieves repression of ATFS-1, and activates the unfolded protein 202 response in the extracellular compartment to restore extracellular proteostasis. 203 Endosomal recycling of plasma membrane-localized receptors promotes ZIP-3 stabilization via 204 phosphorylation 205 Inhibition of endosomal recycling reduced ZIP-3 levels and activated ATFS-1–dependent 206 transcription, suggesting that diminished cell surface receptor signaling initiates this response. To 207 identify receptors regulating ZIP-3, we performed an RNAi screen targeting plasma membrane 208 receptors and monitored ATFS-1 activity using the hsp-6pr::GFP reporter. Knockdown of multiple G 209 protein–coupled receptors (GPCRs), including fshr-1, mgl-2, npr-28, and paqr-2, as well as non-GPCR 210 receptors such as cam-1 (Ror), lin-12 (Notch), lin-17 (Wnt Frizzled), vab-1 (Ephrin), and ver-1 (EGFR), 211 induced hsp-6 expression (Figure 4H). WWP-1 recognizes a specific consensus binding motif, PPxY , 212 on its substrates 35, facilitating efficient internalization and endocytic recycling of the receptors 36 213 (Supplementary Figure 3G). Notably, all non-GPCR receptors that induced hsp-6pr::GFP when 214 inhibited using RNAi contained PPxY motifs. To determine whether the absence of interaction between 215 the receptors and WWP-1 activates ATFS-1-dependent transcription, we mutated the specific PPxY 216 sites to PPxA on the cam-1 and ver-1 receptors. Impressively, mutating the PPxY sites on cam-1 and 217 ver-1 induces hsp-6::GFP (Supplementary Figure 3F), suggesting that the ZIP-3 turnover and 218 activation of ATFS-1-dependent transcription initiates with the lack of WWP-1 interaction with the cell 219 surface receptors. This activation of ATFS-1-dependent transcription is suppressed in ZIP-3(YA) 220 mutants that lack the PPxY motif owing to the lack of ZIP-3 degradation by WWP-1 (Supplementary 221 Figure 3H). Moreover, ZIP-3::GFP colocalized with the recycling endosome marker RAB-11.1::mRuby 222 upon WWP-1 inhibition (Figure 5B), and WWP-1 depletion induced tubular ZIP-3::GFP structures 223 (Figure 5A). This data suggests that the WWP-1-mediated ZIP-3 turnover takes place on the 224 endosomal membrane. Similar to increased ZIP-3 accumulation, stalled tubular endosomal 225 morphology is observed in atfs-1(null) worms following vps-35 knockdown (Figure 4B), suggesting that 226 ATFS-1 is required to sustain endosomal recycling. 227 Because GPCRs lack PPxY motifs, we examined arrestin adaptor proteins that harbor PPxY 228 motifs and mediate WWP-1-dependent endocytosis of GPCR 36. RNAi depletion of arrd-8, arrd-10, 229 arrd-11, and arrd-22 , as well as disruption of G protein signaling via the G α subunit gsa-1, induced 230 hsp-6 (Supplementary Figure 3D), implicating GPCR-mediated signaling in the regulation of ZIP-3. 231 Consistent with impaired endosomal recycling leading to reduced receptor expression, inhibition of 232 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 8 vps-35 or rme-8 reduced the abundance of fluorescently tagged FSHR-1, LIN-17, and DAF-2 receptor 233 expression and concomitantly decreased ZIP-3 levels (Figure 5C and supplementary Figure 3A and 234 3E). We also observed a reduction in the receptor expression in the A β (1–42) expressing strain that 235 showed endosomal swelling similar to vps-35 inhibition (Figure 5D). During normal endosomal 236 recycling, receptors are trafficked back to the cell surface membrane via retromer-dependent 237 endosomal recycling or direct endosomal recycling. However, if the recycling pathway is perturbed, 238 endosomal cargoes are directed to lysosomes for degradation. To test whether inhibiting lysosomal 239 uptake can block the reduction in receptor expression, we inhibited cup-5 in A β (1–42) expressing 240 worms. CUP-5 is a mammalian orthologue of MCOLN1 expressed on the endo-lysosomal and 241 lysosomal membranes and is required for lysosomal uptake and function 37. We found that cup-5 242 inhibition in the A β (1–42) worms reverted and increased FSHR-1 and LIN-17 expression (Figure 5D), 243 suggesting that during perturbed endosomal swelling, receptors are degraded in the lysosomes, which 244 activates ATFS-1-dependent transcription. Supporting our model of UPREC, we found that inhibiting the 245 receptor expression (fshr-1, lin-17, and cam-1) reduced protein aggregation in the A β (1–42) worms by 246 activating ATFS-1 dependent transcription (Supplementary Figure 4A). 247 Because ZIP-3 degradation follo wed receptor loss, we determined whether ZIP-3 stability 248 depends on phosphorylation by kinases downstream of cell surface receptor signaling. Our 249 bioinformatics analysis shows that ZIP-3 contains multiple predicted phosphorylation sites (Figure 5E 250 and supplementary Figure 4B). We used CRISPR-Cas9 to generate site-directed mutagenesis of a 251 conserved phosphorylation cluster on ZIP-3 (RKRSSS → RKRAAA)38. This decreased ZIP-3 levels and 252 induced hsp-6pr::GFP expression similar to the zip-3 deletion mutant (Figure 5F-G). ZIP-3 abundance 253 in this mutant is restored by inhibiting wwp-1, indicating that WWP-1 selectively targets the 254 dephosphorylated ZIP-3 protein (Figure 5H). Co-immunoprecipitation of ZIP-3–FLAG followed by 255 phospho-specific immunoblotting revealed phosphorylation on both serine and tyrosine residues on 256 ZIP-3 (Figure 5I). This is also supported by an RNAi screen of families of kinases that are predicted to 257 have conserved phosphorylation sites on ZIP-3, which identified kin-1 (PKA), pkc-2 (PKC), akt-2 258 (AKT), and kgb-1 (JNK) as kinases required to main tain ZIP-3 phosphorylation and stability 259 (Supplementary Figure 4B-D). 260 WWP-1 activity is regulated by a conserved C2 domain that mediates cell membrane 261 association and WWP-1 self-autoinhibition through interactions with the HECT domain 39,40. To test 262 whether disruption of this domain alters WWP-1 activity, we targeted the predicted C2 region of wwp-1 263 using CRISPR–Cas9 (Supplementary Figure 5B). While complete deletion of this region was not 264 viable, we isolated a mutant containing an insertion of two amino acids (N, E) within the C2 domain 265 (wwp-1(NE)) (Supplementary Figure 5A-B). This mutant exhibited elevated hsp-6 expression 266 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 9 compared to wwp-1(null) animals, consistent with a partial gain-of-function phenotype, and showed 267 extended lifespan, in contrast to the reduced viability of wwp-1(null) mutants (Supplementary Figure 268 5A-D). Together, these results demonstrate that surface receptor signaling and endosomal recycling 269 maintain ZIP-3 stability through phosphorylation of ZIP-3 and C2 domain–regulated WWP-1 activity. In 270 contrast, impaired endosomal trafficking or loss of cell signaling receptors promotes WWP-1-mediated 271 ZIP-3 degradation, relieving inhibition of ATFS-1 by ZIP-3 and activating a transcriptional program that 272 limits extracellular protein aggregation. 273 Mitochondrial perturbations that extend lifespan also promote EC proteostasis 274 Previous studies have established that mild inhibition of mitochondrial electron transport chain 275 (ETC) components, such as nuo-6, isp-1, and cco-1, extends lifespan via an ATFS-1-dependent 276 transcriptional program 26,41–43. However, the precise downstream mechanisms translating this stress 277 response into longevity remain elusive. We hypothesized that ATFS-1 may promote organismal 278 survival by promoting extracellular proteostasis. Through a targeted screen for ATFS-1-dependent 279 chaperones, we identified dnj-20, W01A8.6, and F54E2.1 as a critical effector. Knockdown of dnj-20, 280 W01A8.6, and F54E2.1 resulted in a marked accumulation of A β (1–42)-positive aggregates (Figure 281 1D). Furthermore, chromatin immunoprecipitation confirmed that ATFS-1 directly binds the dnj-20 282 promoter, establishing dnj-20 as a primary transcriptional target likely responsible for stabilizing the 283 extracellular proteome under mitochondrial stress. 284 To determine the functional necessity of this extracellular proteostasis axis for the longevity of 285 isp-1(qm150) and clk-1(qm30) mutants, we inhibited dnj-20 and W01A8.6 . Notably, the knockdown of 286 dnj-20, or the deletion of W01A8.6, significantly suppressed the extended lifespan of these OXPHOS 287 mutants (Supplementary Figure 6A-C). Together, these findings establish the ATFS-1–mediated 288 extracellular unfolded protein response (UPR EC) as an essential regulator of proteostasis and 289 longevity. Our work uncovers a cell-non-autonomous signaling mechanism that aligns stress 290 responses with extracellular protein quality control to sustain organismal health and lifespan. 291

292 Although proteostasis has largely been examined through intracellular stress pathways, the 293 maintenance of the extracellular (EC) proteome is equally important for tissue homeostasis and 294 intercellular communication 44,45. Here, we define an extracellular unfolded protein response (UPR EC) 295 centered on ATFS-1 that is activated by inhibition or degradation of the bZIP transcription factor ZIP-3. 296 Inhibition of ZIP-3 activates an ATFS-1–dependent transcriptional program that coordinates 297 extracellular proteostasis through promoting expression of genes encoding extracellular chaperones, 298 extracellular proteases, cell surface receptors, and endosomal recycling components. These findings 299 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 10 position ATFS-1 as a central transcription factor linking extracellular protein quality control to 300 endosomal trafficking and nuclear gene regulation. 301 A key advance of this study is the demonstration that perturbations in endosomal recycling, an 302 early and conserved feature of Alzheimer’s disease (AD) pathology, directly activate UPR EC through 303 destabilization of ZIP-3. Disruption of retromer components such as VPS-35, or early endosomal 304 regulators including RME-8 and VPS-34, results in ZIP-3 degradation and robust induction of ATFS-1–305 dependent target genes, including those with extracellular functions. Importantly, this activation occurs 306 independently of mitochondrial damage, revealing a signaling pathway in which endosomal 307 dysfunction and stalled cargo recycling of cell surface receptors serve as primary triggers. We propose 308 that endosomal swelling and impaired membrane flux are sensed upstream of ZIP-3, initiating a 309 compensatory transcriptional response that preserves extracellular proteostasis and organismal 310 health. 311 Mechanistically, we show that ZIP-3 stability is controlled by post-translational mechanisms 312 involving ubiquitin-mediated degradation and kinase-dependent phosphorylation. Loss of surface 313 receptors or downstream kinase activity markedly reduces ZIP-3 abundance, indicating that 314 extracellular signaling actively maintains ZIP-3 as a repressor of UPR EC under basal conditions. Under 315 conditions of extracellular protein aggregation or trafficking failure, this repression is relieved through 316 ZIP-3 turnover mediated by the E3 ubiquitin ligase WWP-1. Notably, phosphorylation of ZIP-3 confers 317 protection from degradation, suggesting a regulatory switch by which extracellular cues calibrate 318 UPREC activation. 319 The functional consequences of UPR EC activation extend to organismal proteostasis and 320 longevity. ATFS-1 activation, or genetic inhibition of ZIP-3, enhances extracellular proteome integrity 321 and significantly improves healthspan in models expressing aggregation-prone proteins such as A β or 322 TTR. Rescue of neuronal and muscular dysfunction in these models highlights UPR EC as a protective 323 pathway that buffers extracellular proteotoxic stress and mitigates tissue decline. These findings 324 suggest that controlled activation of ATFS-1–dependent UPR EC may represent a conserved adaptive 325 strategy to counteract extracellular aggregation during aging. 326 Together, our findings determine an extracellular unfolded protein response (UPR EC) that links 327 endosomal trafficking defects to transcriptional programs that preserve extracellular proteostasis. By 328 identifying ZIP-3 degradation as a molecular switch that activates ATFS-1–dependent gene 329 expression, this work reveals how cells sense disruptions in membrane trafficking and extracellular 330 protein homeostasis to initiate adaptive protective responses. Because extracellular protein 331 aggregation and endosomal dysfunction are conserved features of aging and neurodegenerative 332 disorders such as Alzheimer's Disease, the mechanisms uncovered here suggest that related 333 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 11 surveillance pathways may operate in higher organisms to coordinate extracellular protein quality 334 control. Elucidating whether analogous signaling networks involving mammalian ATF family 335 transcription factors46,47 regulate extracellular proteostasis may therefore provide new insights into the 336 pathogenesis of protein aggregation diseases and identify strategies to enhance extracellular 337 proteome resilience during aging. 338 339 340 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 12

341 Worms, plasmids, and bacteria. N2 (wildtype), VC4372 W01A8.6 (gk5453), DCD23 LBP-2::tagRFP, 342 OH906 otIs39 [unc-47(delta)::GFP + lin-15(+)], and BZ555 dat-1pr::GFP strains were obtained from the 343 Caenorhabditis Genetics Center (Minneapolis, MN). LSD2104 xchIs015 [pLSD134-Phsp-344 16.2::ssSel1:FLAG::superfolderGFP::spacer::humanAmyloidBeta1-42::let-858–3’UTR; pRF4 rol-345 6(su1006)] was obtained from the Ewald lab 27. JKM2 Is [rgef-1p::Signalpeptide-Abeta(1-42)::hsp-346 3(IRES)::wrmScarlet-Abeta(1-42)::unc-54(3’UTR) + rps-0 p::HygroR] and JKM7 Is [myo-347 3p::Signalpeptide-Abeta(1-42)::hsp-3(IRES)::wrmScarlet-Abeta(1-42)::unc-54(3’UTR) + rps-348 0p::HygroR] strains were obtained from the Kirstein lab 29. SEE037 scrIs008[unc-54p::hTTR(WT) + rol-349 6], SEE034 uthIs378[unc-54p::hTTR(V30M) + rol-6] , SEE106 geIs101[rol-6(su1006)]; scrIs010[des-350 2p::myr::GFP + unc-122p::DsRed] and SEE145 uthIs378[unc-54p::hTTR(V30M) + rol-6]; scrIs010[des-351 2p::myr::gfp + unc-122p::DsRed strains were obtained from the Encalada lab 28. fshr-1p::fshr-352 1::SL2::mKate strain was obtained from the Beets lab 48. mRuby2::RAB-11.1 strain was obtained from 353 Douglas' lab49. 354 ZIP-3-3x-FLAG and ZIP-3(RAAA) strains were generated via CRISPR-Cas9 in wildtype worms 355 as described 12. The crRNAs (IDT) were co-injected with purified Cas9 protein, tracrRNA (Dharmacon), 356 repair templates (IDT), and the pRF4::rol-6(su1006) plasmid as described 50. The pRF4::rol-6 (su1006) 357 plasmid was a gift from Craig Mello 50. The vha-6pr::daf-2::GFP plasmid was gifted by Tian Xia and 358 Anbing Shi labs51. 359 The W01A8.6pr::GFP plasmid was generated using conventional cloning of the W01A8.6 360 promoter sequence and a portion of the first exon (2021 bp) into the pPD95.75 vector. The 361 F54E2.1pr::GFP plasmid was generated using conventional cloning of the F54E2.1 promoter sequence 362 and the full first exon (4025 bp) into the pPD95.75 vector. The myo-3 pr::LBP-2-GFP plasmid was 363 generated using conventional cloning of the LBP-2 coding sequence (584 bp) into the pPD136.64 364 vector in frame with the myo-3 promoter. Worms were raised with the HT115 strain of Escherichia coli, 365 and RNAi was performed as described. 366 Analysis of worm development. Worms were synchronized via bleaching and allowed to develop on 367 HT115 bacteria plates at 20°C. The developmental stage is quantified as a percentage of the total 368 number of animals that turned adult after an incubation of 58 hrs at 20°C as previously described 52. 369 Each experiment was performed three times. 370 Lifespan assay. Lifespan analysis was carried out following an established protocol 53,54. Each strain 371 was repeated at least twice. At least 50 animals were used per condition, and worms were scored for 372 viability every second day, from day 1 of adulthood (treating the pre-fertile day preceding adulthood as 373 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 13 t = 0). Young adult worms were transferred to fresh plates every other day, and the number of dead 374 worms was recorded as events scored. Animals that were lost or burrowed in the medium, exhibiting 375 protruding vulva (intestine protrudes from the vulva), or undergoing bagging (larvae hatching inside 376 the worm body) were censored. Prism 8 (GraphPad) software was used for statistical analysis, and p-377 values were calculated using the log-rank (Kaplan-Meier) method. 378 Protein analysis and antibodies. Synchronized worms were raised on plates with control(RNAi) or 379 cco-1(RNAi) to the L4 stage before harvesting. The whole worm lysate preparation was previously 380 described 55. Antibodies against β -actin (cell signaling), anti-FLAG M2 antibody (Sigma, F1804), RFP 381 antibody (Rockland, 600-401-379), GFP antibody (Abcam, AB 183734), anti-phosphoserine/threonine 382 antibody (BD Biosciences, 612548), and anti-phosphotyrosine antibody (Cell signaling, 8954S). All 383 antibodies were diluted 1:2000. Immunoblots were imaged using the ChemiDoc XRS+ system (Bio-384 Rad). All western blot experiments were performed at least three times. 385 Immunoprecipitation. Synchronized worms were raised on plates with wildtype (N2) or ZIP-3-386 3xFLAG strains to the L4 stage before harvesting. The whole worm lysate preparation was previously 387 described 55 with a protease inhibitor and phosphatase inhibitor cocktail. 388 Anti-FLAG magnetic beads were gently resuspended by pipetting, and 10 µL of bead 389 suspension was transferred to a microcentrifuge tube. The beads were washed twice with 0.5 mL TBS 390 buffer (50 mM Tris-HCl, 150 mM NaCl, pH 7.4) by gentle pipetting and magnetic separation. Washed 391 beads were incubated with 500 µL of cell lysate for 2 h at room temperature or overnight at 4 °C with 392 gentle rotation to allow binding of ZIP-3-3xFLAG. Anti-FLAG M2 Magnetic beads (Sigma, M8823-1mL) 393 are used for Immunoprecipitation using the ZIP-3-3xFLAG strain. Following incubation, beads were 394 separated magnetically, and the supernatant was collected to assess unbound protein. The beads 395 were then washed three times with PBST (136.89 mM NaCl, 2.67 mM KCl, 8.1 mM Na ₂ HPO₄ , 1.76 396 mM KH₂ PO₄ , 0.5% Tween-20), with 5 min rotation for each wash. For denaturing elution, beads were 397 resuspended in 50 µL of 1× protein sample loading buffer and boiled for 5 min. After magnetic 398 separation, the eluates were analyzed by SDS-PAGE with anti-phosphoserine/threonine and anti-399 phosphotyrosine antibodies. 400 TMRE staining. TMRE staining was performed by synchronizing worms and raising them on plates 401 pre-soaked with S-Basal buffer containing TMRE at a final concentration of 100 μ M (Sigma, 87917) 56. 402 Before imaging, TMRE-stained worms were transferred to plates seeded with control (RNAi) bacteria 403 and incubated for 3 h to eliminate TMRE-containing bacteria from the digestive tract. Images were 404 acquired using a Zeiss LSM800 confocal microscope equipped with Airyscan, with identical exposure 405 settings applied across all conditions. TMRE fluorescence analysis was performed as previously 406 described. Briefly, average pixel intensity values were obtained by sampling images from individual 407 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 14 worms, and the mean fluorescence intensity for each animal was quantified using ImageJ 408 (http://rsb.info.nih.gov/ij/). Fluorescence intensity was measured from threshold-adjusted images for 409 each condition using biological triplicates. Statistical comparisons were performed using Student’s t-410 test or one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test, as appropriate. 411 RNA isolation and qRT-PCR. RNA isolation and quantitative reverse transcriptase PCR (qRT-PCR) 412 analysis were previously described 57. Worms were synchronized by bleaching, raised on HT115 E. 413 coli, and harvested at the L4 stage. Total RNA was extracted from frozen worm pellets using Trizol 414 reagent, and 500 ng RNA was used for cDNA synthesis with qScript™ cDNA SuperMix (QuantaBio). 415 qPCR was performed using iQ™ SYBR® Green Supermix (Bio-Rad Laboratories). All qPCR results 416 were repeated at least 3 times and performed in triplicate. A two-tailed Student’s t-test was employed 417 to determine the level of statistical significance. 418 Microscopy. C. elegans were imaged using either a Zeiss AxioCam 506 mono camera mounted on a 419 Zeiss Axio Imager Z2 microscope or a Zeiss AxioCam MRc camera mounted on a Zeiss SteREO 420 Discovery.V12 stereoscope. Images with high magnification (×63) were obtained using the Zeiss 421 Apotome 2. Exposure times were the same in each experiment. Cell cultures were imaged with the 422 Zeiss LSM800 microscope. With Zen 2.3 Blue software. All images are representative of more than 423 three images. Quantification of fluorescent intensity as well as creating binary skeleton-like structures, 424 was done with ImageJ. 425 Gene set enrichment analysis. The gene set was downloaded from the WormBase Ontology 426 Browser. mRNA abundance was measured and ranked by reads per kilobase per million reads from 427 RNA-seq data. Pre-ranked gene set enrichment analysis was performed with GSEA3.0 software with 428 ‘classical’ scoring. 429 Statistics. All experiments were performed at least three times, yielding similar results, and comprised 430 biological replicates. The sample size and statistical tests were chosen based on previous studies with 431 similar methodologies, and the data met the assumptions for each test. No statistical method was 432 used in deciding sample sizes. No blinded experiments were performed, and randomization was not 433 used. For all figures, the mean ± SD is represented unless otherwise noted. Prism 8 (GraphPad) is 434 used for statistical analysis and graph creation. 435 Acknowledgements. 436 We thank Dr. Collin Ewald, Dr. Janine Kirtstein, Dr. Sandra Encalada, Dr. Peter Douglas, and Dr. 437 Isabel Beets for sharing some worm strains with us. We thank Dr. Barth Grant for his advice on and for 438 sharing the endosomal recycling reporter worm strains. We also thank Dr. Hira Goel for his advice and 439 guidance throughout the project. We thank the Caenorhabditis Genetics Center for providing C. 440 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 15 elegans strains, funded by NIH Office of Research 362 Infrastructure Programs (P40OD010440). This 441 work was supported by the National Institutes of Health grants (R01AG040061, R01AG047182, and 442 R37-AG047182-07) to C.M.H. and the Natural Sciences and Engineering Research Council (NSERC) 443 postdoctoral fellowship to A.M. The authors are solely responsible for the content. 444 Author's contributions: 445 A.M. and C.M.H. planned the experiments. A.M. and Y .D. generated worm strains. A.M. performed 446 experiments. A.M. and C.M.H. wrote the manuscript. 447 Conflict of Interest. 448 The authors declare no conflicts of interest. 449

450 1. Walter, P. & Ron, D. The unfolded protein response: From stress pathway to homeostatic 451 regulation. Science (80-. ). 334, 1081–1086 (2011). 452 2. Higuchi-Sanabria, R., Frankino, P. A., Paul, J. W., Tronnes, S. U. & Dillin, A. A Futile Battle? 453 Protein Quality Control and the Stress of Aging. Dev. Cell 44, 139–163 (2018). 454 3. Shpilka, T. & Haynes, C. M. The mitochondrial UPR: Mechanisms, physiological functions and 455 implications in ageing. Nat. Rev. Mol. Cell Biol. 19, 109–120 (2018). 456 4. Vaquer-Alicea, J. & Diamond, M. I. Propagation of protein aggregation in neurodegenerative 457 diseases. Annu. Rev. Biochem. 88, 785–810 (2019). 458 5. Gao, H. et al. Mitochondrial complex I deficiency induces Alzheimer’s disease–like signatures 459 that are reversible by targeted therapy. Alzheimer’s Dement. 21, 1–19 (2025). 460 6. Stojakovic, A. et al. Partial inhibition of mitochondrial complex I ameliorates Alzheimer’s disease 461 pathology and cognition in APP/PS1 female mice. Commun. Biol. 4, (2021). 462 7. Mathys, H. et al. Single-cell multiregion dissection of Alzheimer’s disease. Nature (2024). 463 doi:10.1038/s41586-024-07606-7 464 8. Sorrentino, V. et al. Enhancing mitochondrial proteostasis reduces amyloid-β proteotoxicity. 465 Nature 552, 187–193 (2017). 466 9. Nargund, A., Pellegrino, M. W., Fiorese, C. J., Baker, B. M. & Haynes, C. M. Mitochondrial 467 Import Efficiency of ATFS-1 Regulates Mitochondrial UPR Activation. Science (80-. ). 337, 587–468 590 (2012). 469 10. Nargund, A. M., Fiorese, C. J., Pellegrino, M. W., Deng, P. & Haynes, C. M. Mitochondrial and 470 nuclear accumulation of the transcription factor ATFS-1 promotes OXPHOS recovery during the 471 UPRmt. Mol. Cell 58, 123–133 (2015). 472 11. Shpilka, T. et al. UPRmt scales mitochondrial network expansion with protein synthesis via 473 mitochondrial import in Caenorhabditis elegans. Nat. Commun. 12, (2021). 474 12. Deng, P. et al. Mitochondrial UPR repression during Pseudomonas aeruginosa infection 475 requires the bZIP protein ZIP-3. Proc. Natl. Acad. Sci. U. S. A. 116, 6146–6151 (2019). 476 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 16 13. Reinke, A. W., Baek, J., Ashenberg, O. & Keating, A. E. Networks of bZIP Protein-Protein 477 Interactions Diversified Over a Billion Years of Evolution. Science (80-. ). 340, 730–735 (2013). 478 14. Szabo, M. P., Mishra, S., Knupp, A. & Young, J. E. The role of Alzheimer’s disease risk genes 479 in endolysosomal pathways. Neurobiol. Dis. 162, 105576 (2022). 480 15. Young, J. E., Holstege, H., Andersen, O. M., Petsko, G. A. & Small, S. A. On the causal role of 481 retromer-dependent endosomal recycling in Alzheimer’s disease. Nat. Cell Biol. 25, 1394–1397 482 (2023). 483 16. Cullen, P. J. & Steinberg, F. To degrade or not to degrade: mechanisms and significance of 484 endocytic recycling. Nat. Rev. Mol. Cell Biol. 19, 679–696 (2018). 485 17. Shi, A. et al. Regulation of endosomal clathrin and retromer-mediated endosome to Golgi 486 retrograde transport by the J-domain protein RME-8. EMBO J. 28, 3290–3302 (2009). 487 18. Sorkin, A. & Von Zastrow, M. Signal transduction and endocytosis: Close encounters of many 488 kinds. Nat. Rev. Mol. Cell Biol. 3, 600–614 (2002). 489 19. Fletcher, K. A., Alkurashi, M. H. & Lindsay, A. J. Endosomal recycling inhibitors downregulate 490 the androgen receptor and synergise with enzalutamide. Invest. New Drugs 42, 14–23 (2024). 491 20. Mishra, S. et al. The Alzheimer’s gene SORL1 is a regulator of endosomal traffic and recycling 492 in human neurons. Cell. Mol. Life Sci. 79, 1–22 (2022). 493 21. Small, S. A. et al. Model-Guided Microarray Implicates the Retromer Complex in Alzheimer ’ s 494 Disease. 909–919 (2005). doi:10.1002/ana.20667 495 22. Kachergus, J. M. et al. VPS35 Mutations in Parkinson Disease. 162–167 (2011). 496 doi:10.1016/j.ajhg.2011.06.001 497 23. Tang, F., Liu, W., Hu, J., Mei, L. & Xiong, W. VPS35 Deficiency or Mutation Causes 498 Dopaminergic Neuronal Loss by Impairing Mitochondrial Fusion and Article VPS35 Deficiency 499 or Mutation Causes Dopaminergic Neuronal Loss by Impairing Mitochondrial Fusion and 500 Function. CellReports 12, 1631–1643 (2015). 501 24. Gallotta, I. et al. Extracellular proteostasis prevents aggregation during pathogenic attack. 502 Nature 584, 410–414 (2020). 503 25. Braun, J. E. A. Extracellular chaperone networks and the export of J-domain proteins. J. Biol. 504 Chem. 299, 102840 (2023). 505 26. Durieux, J., Wolff, S. & Dillin, A. The cell-non-autonomous nature of electron transport chain-506 mediated longevity. Cell 144, 79–91 (2011). 507 27. Jongsma, E., Goyala, A., Mateos, J. M. & Ewald, C. Y. Removal of extracellular human amyloid 508 beta aggregates by extracellular proteases in C. elegans. Elife 12, 1–21 (2023). 509 28. Madhivanan, K. et al. Cellular clearance of circulating transthyretin decreases cell-510 nonautonomous proteotoxicity in Caenorhabditis elegans. Proc. Natl. Acad. Sci. U. S. A. 115, 511 E7710–E7719 (2018). 512 29. Gallrein, C. et al. Novel amyloid-beta pathology C. elegans model reveals distinct neurons as 513 seeds of pathogenicity. Prog. Neurobiol. 198, 101907 (2021). 514 30. Wang, C. et al. A neurotransmitter atlas of C . elegans males and hermaphrodites. 1–46 (2024). 515 31. Robinson, S. B. et al. Molecular and Cellular Neuroscience Sequence determinants of the 516 Caenhorhabditis elegans dopamine transporter dictating in vivo axonal export and synaptic 517 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 17 localization. Mol. Cell. Neurosci. 78, 41–51 (2017). 518 32. Sato, K., Norris, A., Sato, M. & Grant, B. D. C . elegans as a model for membrane traffic *. 519 (2014). doi:10.1895/wormbook.1.77.2 520 33. Rostislavleva, K. et al. Structure and flexibility of the endosomal Vps34 complex reveals the 521 basis of its function on membranes. 7365, (2015). 522 34. Wallroth, A. & Haucke, V. Phosphoinositide conversion in endocytosis and the endolysosomal 523 system. 293, 1526–1535 (2018). 524 35. Zhi, X. & Chen, C. WWP1: A versatile ubiquitin E3 ligase in signaling and diseases. Cell. Mol. 525 Life Sci. 69, 1425–1434 (2012). 526 36. Sardana, R. & Emr, S. D. Membrane Protein Quality Control Mechanisms in the Endo-527 Lysosome System. Trends Cell Biol. 31, 269–283 (2021). 528 37. Sun, T., Wang, X., Lu, Q., Ren, H. & Zhang, H. CUP-5, the C. elegans ortholog of the 529 mammalian lysosomal channel protein MLN1/TRPML1, is required for proteolytic degradation in 530 autolysosomes. Autophagy 7, 1308–1315 (2011). 531 38. Blom, N., Sicheritz-pontén, T., Gupta, R., Gammeltoft, S. & Brunak, S. Prediction of post-532 translational glycosylation and phosphorylation of proteins from the amino acid sequence. 533 Proteomics 4, 1633–1649 (2004). 534 39. Wiesner, S. et al. Autoinhibition of the HECT-Type Ubiquitin Ligase Smurf2 through Its C2 535 Domain. Cell 130, 651–662 (2007). 536 40. Courivaud, T. et al. Functional Characterization of a WWP1 / Tiul1 Tumor-derived Mutant 537 Reveals a Paradigm of Its Constitutive Activation in Human Cancer *. 290, 21007–21018 538 (2015). 539 41. Feng, J., Bussière, F. & Hekimi, S. Mitochondrial Electron Transport Is a Key Determinant of 540 Life Span in Caenorhabditis elegans. Dev. Cell 1, 633–644 (2001). 541 42. Felkai, S. et al. CLK-1 controls respiration, behavior and aging in the nematode Caenorhabditis 542 elegans. EMBO J. 18, 1783–1792 (1999). 543 43. Dillin, A. et al. Rates of behavior and aging specified by mitochondrial function during 544 development. Science (80-. ). 298, 2398–2401 (2002). 545 44. Sutherland, T. E., Dyer, D. P. & Allen, J. E. The extracellular matrix and the immune system: A 546 mutually dependent relationship. Science (80-. ). 379, (2023). 547 45. Bandzerewicz, A. & Gadomska-Gajadhur, A. Into the Tissues: Extracellular Matrix and Its 548 Artificial Substitutes: Cell Signalling Mechanisms. Cells 11, (2022). 549 46. Fiorese, C. J. et al. The Transcription Factor ATF5 Mediates a Mammalian Mitochondrial UPR. 550 Curr. Biol. 26, 2037–2043 (2016). 551 47. Anderson, N. S. & Haynes, C. M. Folding the Mitochondrial UPR into the Integrated Stress 552 Response. Trends Cell Biol. 30, 428–439 (2020). 553 48. Kenis, S. et al. Ancestral glycoprotein hormone-receptor pathway controls growth in C . 554 elegans. 1–15 (2023). doi:10.3389/fendo.2023.1200407 555 49. Watterson, A. et al. Intracellular lipid surveillance by small G protein geranylgeranylation. 556 Nature 605, 736–740 (2022). 557 50. Dokshin, G. A., Ghanta, K. S., Piscopo, K. M. & Mello, C. C. Robust genome editing with short 558 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 18 single-stranded and long, partially single-stranded DNA donors in caenorhabditis elegans. 559 Genetics 210, 781–787 (2018). 560 51. Chen, D. et al. Revealing Functional Crosstalk between Distinct Bioprocesses through 561 Reciprocal Functional Tests of Genetically Interacting Genes Article Revealing Functional 562 Crosstalk between Distinct Bioprocesses through Reciprocal Functional Tests of Genetically 563 Interacting Genes. 2646–2658 (2019). doi:10.1016/j.celrep.2019.10.076 564 52. Mallick, A. et al. AMP accumulation during mitochondrial stress induces transcription of 565 cytosolic and mitochondrial protein synthesis components via NHR-180. 1, (2025). 566 53. Amrit, F. R. G., Ratnappan, R., Keith, S. A. & Ghazi, A. The C. elegans lifespan assay toolkit. 567

68, 465–475 (2014). 568 54. Mallick, A., Ranawade, A., van den Berg, W. & Gupta, B. P. Axin-Mediated Regulation of 569 Lifespan and Muscle Health in C. elegans Requires AMPK-FOXO Signaling. iScience 23, 570 101843 (2020). 571 55. Yoneda, T. et al. Compartment-specific perturbation of protein handling activates genes 572 encoding mitochondrial chaperones. J. Cell Sci. 117, 4055–4066 (2004). 573 56. Mallick, A. & Haynes, C. M. Methods to analyze the mitochondrial unfolded protein response 574 (UPRmt). Methods Enzymol. 707, 543–564 (2024). 575 57. Lin, Y. F. et al. Maintenance and propagation of a deleterious mitochondrial genome by the 576 mitochondrial unfolded protein response. Nature 533, 416–419 (2016). 577 578 Figure legend. 579 Figure 1. Activation of ATFS-1 promotes the expression of extracellular genes that limit 580 aggregation. A-C. Representative images of worms expressing W01A8.6pr::W01A8.6-GFP, 581 F54e2.1pr::F54E2.1-GFP, and dnj-20pr::GFP following control or cco-1(RNAi). Scale bar, 200μ m. N =3, 582 biologically independent replicates. D. Representative images of worms expressing hsp-583 16.2pr::ssAβ (1-42)-GFP labelled protein aggregation puncta grown at 37°C for 1 hr during Day 1 adult 584 following control, W01A8.6(RNAi), F54E2.1(RNAi), and dnj-20(RNAi). Scale bar, 50 μ m. N =3, 585 biologically independent replicates. E. Representative images of worms expressing hsp-586 16.2pr::ssAβ (1-42)-GFP labelled protein aggregation puncta grown at 37°C for 1 hr during Day 1 adult 587 of wildtype and atfs-1(null) worms. Scale bar, 50 μ m. N =3, biologically independent replicates. F. 588 Representative image of a Western blot analysis of total and insoluble fraction of hsp-16.2pr::ssAβ (1-589 42)-GFP in wildtype and atfs-1(null) worms. G. Representative images of LBP-2-RFP labelled protein 590 aggregation puncta of wildtype and atfs-1(null) Day 1 adults. Scale bar, 50 μ m. N =3, biologically 591 independent replicates. H. Dot plots showing the number of LBP-2-RFP aggregation puncta, in 592 wildtype and atfs-1(null) worms in G . N = 3, biologically independent samples. *** p < 0.001 (two-tailed 593 Student’s t-test). 594 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 19 Figure 2. Inhibition of ZIP-3 reduces extracellular protein aggregation. A. Bubble plot showing the 595 normalized enrichment scores of gene ontology classes for differentially expressed genes in zip-3(null) 596 worms compared to wildtype. B. Representative images of LBP-2-RFP labelled protein aggregation 597 puncta in wildtype and zip-3(null) Day 1 adults grown at 25°C. Scale bar, 50 μ m. N =3, biologically 598 independent replicates. C. Representative images of LBP-2-RFP labelled protein aggregation puncta 599 in wildtype, zip-3(null), atfs-1(RNAi), and zip-3(null), atfs-1(RNAi) Day 1 adults grown at 25°C. Scale 600 bar, 50 μ m. N =3, biologically independent replicates. D. Representative image of a Western blot 601 analysis of the insoluble fraction of LBP-2-RFP in wildtype, zip-3(null), atfs-1(null), and cco-1 (RNAi) 602 worms. N =3, biologically independent replicates. E. Representative images of worms expressing hsp-603 16.2pr::ssAβ (1-42)-GFP labelled protein aggregation puncta grown at 37°C for 1 hr during Day 1 adult 604 of wildtype and zip-3(null) worms. Scale bar, 50 μ m. N =3, biologically independent replicates. F. 605 Representative image of a Western blot analysis of the insoluble fraction of hsp-16.2pr::ssAβ (1-42)-606 GFP in wildtype and zip-3(null) Day-2 and Day-4 adult worms. N =3, biologically independent 607 replicates. G-H. Lifespan graphs for wildtype and zip-3(RNAi) worms grown at 25°C ( G) and for myo-608 3pr::LBP-2-GFP worms grown at 20°C following control and zip-3 (RNAi) ( H). N =3, biologically 609 independent replicates. *p < 0.01 (G, H) (log-rank test). 610 Figure 3. Inhibiting zip-3 and the ETC complex component cco-1 ameliorates phenotypes in 611 extracellular protein aggregation models. A-B. Lifespan graphs for wildtype, neuronal Aβ (1-42) (A), 612 and TTR(V30M) ( B) worms grown following control and zip-3 (RNAi) ( H). N =3, biologically 613 independent replicates. * p < 0.01 ( A, B) (log-rank test). C-D Dot plots showing the rate of thrashing 614 and body bending of neuronal- and muscle secreted-A β (1-42) worms grown on control, zip-3(RNAi), 615 and cco-1(RNAi) plates. Data represents the mean of three independent biological replicates with 616 error bars showing the standard error of the mean. N = 3, biologically independent samples. ***p < 617 0.001 (one-way ANOVA). E. Representative image of hsp-6pr::GFP expression in wildtype, muscle A β 618 (not secreted), muscle secreted A β (1-42), neuron secreted A β (1-42), and muscle secreted 619 TTR(V30M). Scale bar, 200μ m. N =3, biologically independent replicates. F. Representative images of 620 the dendritic morphology of FLP neurons marked with des-2pr::Myristoyl::GFP of wildtype and 621 TTR(V30M) worms following control and zip-3(RNAi). Scale bar, 100 μ m. N =3, biologically 622 independent replicates. G. Representative image of zip-3pr::ZIP-3::GFP expression in wildtype, zip-623 3(RNAi) and muscle secreted A β (1-42) worms. Scale bar, 100 μ m. N =3, biologically independent 624 replicates. H. Representative image of a western blot analysis of ZIP-3-3x-FLAG in wildtype and 625 neuron-secreted A β (1-42) worms. N =3, biologically independent replicates. I. Lifespan graph of 626 neuron-specific RNAi strain (MAH677) grown on control and zip-3(RNAi). N =3, biologically 627 independent replicates. *p < 0.01 (log-rank test). 628 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 20 Figure 4. Inhibition of endosomal recycling leads to reduced ZIP-3 level and expression of 629 extracellular genes dependent on ATFS-1 activation. A. Representative images of late endosomal 630 marker RAB-7::GFP in wildtype, muscle secreted A β (1-42), vps-35(RNAi), and rab-11.1(RNAi). B. 631 Representative images of late endosomal marker RAB-7::GFP in atfs-1(null) worms grown on control 632 and vps-35(RNAi) HT115 bacteria. C. Representative images of late endosomal marker RAB-7::GFP 633 in muscle secreted A β (1-42) worms grown on zip-3(RNAi) and cco-1(RNAi) HT115 bacteria. Enlarged 634 endosomes seen in muscle secreted A β (1-42) worms ( A) are rescued following zip-3(RNAi) and cco-635 1(RNAi) (C). Scale bar, 50 μ m. N =3, biologically independent replicates. D. Representative image of 636 hsp-6pr::GFP expression following control RNAi, vps-35 (RNAi), vps-29(RNAi), piki-1(RNAi), vps-637 34(RNAi) and mtm-6(RNAi). Scale bar, 200 μ m. N =3, biologically independent replicates. E. 638 Representative image of extracellular protease w01a8.6pr::GFP and extracellular chaperone dnj-639 20pr::GFP expression following control RNAi, vps-35(RNAi), and rme-8 (RNAi). Scale bar, 200 μ m. N 640 =3, biologically independent replicates. F-G. Representative image of a western blot analysis of ZIP-3-641 3x-FLAG following inhibition of endosomal retromer complex components ( rme-8, vps-35, and vps-29) 642 and class III PI3K components ( vps-34, bec-1, and vps-15). N =3, biologically independent replicates. 643 H. Representative image of hsp-6pr::GFP expression in wildtype, fshr-1(ok778), mgl-2(tm355), paqr-644 2(tm3410), cam-1(RNAi) and ver-1 (RNAi). Scale bar, 200 μ m. N =3, biologically independent 645 replicates. 646 Figure 5. Kinases downstream of cell signaling r eceptors that undergo endosomal recycling 647 regulate ZIP-3 via phosphorylation. A. Representative image of zip-3pr::ZIP-3::GFP expression 648 following control and wwp-1(RNAi). Scale bar, 50 μ m. N =3, biologically independent replicates. B. 649 Representative image showing overlay of zip-3pr::ZIP-3::GFP and endosomal recycling marker RAB-650 11.1::mRuby expression following wwp-1(RNAi). Scale bar, 50 μ m. N =3, biologically independent 651 replicates. C. Representative images of cell signaling receptors LIN-17/Frizzled, DAF-2/Insulin-like 652 receptor, and FSHR-1/GPCR expression following inhibition of rme-8 and vps-35 . D. Representative 653 images of FSHR-1 and LIN-17 expression in wildtype and muscle secreted A β (1-42) worms following 654 control and cup-5(RNAi). C-D. Scale bar, 50 μ m. N =3, biologically independent replicates. E. 655 Schematic of ZIP-3 protein s howing the cons erved phosphorylat ion site RKRSSS (26-28aa), PPxY 656 motif (167aa), and bZIP domain (245-304aa). F. Representative image of a western blot analysis of 657 ZIP-3-3x-FLAG in wildtype worms lacking the FLAG tag and with the FLAG tag, and in ZIP-658 3(RKRAAA) mutant with the RKRSSS site muta ted. N =3, biologica lly independent replicates. G. 659 Representative image of a western blot analysis of ZIP-3-3x-FLAG in wildtype and ZIP-3(RKRAAA) 660 mutant following control RNAi and wwp-1 inhibition. N =3, biologically independent replicates. I. 661 Representative image of a western blot analysis of ZIP-3-3x-FLAG after coimmunoprecipitation (Co-662 IP) in wildtype worms lacking the FLAG tag and with the FLAG tag following control RNAi and wwp-1 663 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 21 inhibition. Phosphotyrosine and phosphoserine/threonine antibodies are used to blot for 664 phosphorylation of tyrosine, serine, and threonine residues on ZIP-3. N =3, biologically independent 665 replicates. 666 Supplementary Figures. 667 Supplementary Figure 1. ATFS-1 promotes ex pression of endosomal recycling components 668 and extracellular genes that limit protein aggregation. A. List of extracellular proteases and 669 chaperones that regulate extracellular protein aggregation. B. Representative fluorescent image of 670 zip-3pr::zip-3-gfp animals at the young adult stage. zip-3 is expressed ubiquitously throughout the 671 body, with notable high expression in the head, intestine, and hypodermis of animals. In the head 672 (pharynx) zip-3 is expressed in numerous neuronal cell bodies. Scale bar 50µm. C. Gene enrichment 673 analysis of genes that are reduced in atfs-1(null) and increased in zip-3(null); spg-7(RNAi) worms. 674 Dark orange color shows the categories vital to regulate extracellular proteostasis. D. Quantification of 675 mRNA transcripts of extracellular genes ( cyp-33c5, epic-1, F54E2.1, Y38A10A.2, ttc-17, bcf-1, 676 W01A8.6, and T21D12.12 ) by qPCR in wildtype and atfs-1(null) at the L3 larval stage. N = 3, 677 biologically independent samples. *** p < 0.001 (one-way ANOVA). E-F. Quantification of mRNA 678 transcripts of extracellular genes ( asp-4, sod-4, ttc-17, bcf-1, T21D12.12, clec-41, lys-2, lys-3, txt-8, 679 F54E2.1, dyf-4, W01A8.6 and asp-8 ) by qPCR in wildtype and zip-3(null) at L3 larval stage. N = 3, 680 biologically independent samples. *** p < 0.001 (one-way ANOVA). G. Bar graphs showing the 681 expression of genes by qPCR in the rme-1(b1023ts) mutants compared to wildtype (RFU=1) at the L3 682 stage, grown at 15°C. H. Bar graphs showing the expression of genes by qPCR following control and 683 vps-26(RNAi) (RFU=1) at the L3 stage. I. Bar graphs showing the expression of genes by qPCR in the 684 vps-35(ok1880) mutants compared to wildtype (RFU=1) at the L3 stage. J-K. Bar graphs showing the 685 expression of genes by qPCR in muscle secreted A β (1-42) and TTR(V30M) compared to wildtype 686 (RFU=1) at the L3 stage. G-K. Graphs show the mean of two independent replicates with error bars 687 showing the standard errors of the mean. *** p < 0.001 (one-way ANOVA). L. Representative 688 fluorescent image of atfs-1pr::atfs-1-gfp animals at the young adult stage (scale bar 200µm). atfs-1 is 689 expressed ubiquitously throughout the body, with notable high expression in the head of animals 690 (scale bar 25µm). 691 Supplementary Figure 2. Inhibition of atfs-1 and zip-3 overexpression leads to 692 neurodegeneration. A. Representative image of myo-3pr::LBP-2-GFP expression following control 693 and zip-3(RNAi). Scale bar, 50μ m. N =3, biologically independent replicates. B. Representative image 694 of myo-3pr::LBP-2-GFP worms with GFP detected in the coelomocytes showing that the coelomocyte 695 cells have endocytosed the extracellular LBP-2::GFP secreted from the muscle cells. C-D. Dot plots 696 showing the rate of thrashing and body bending of wildtype and TTR(V30M) worms grown on control, 697 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 22 zip-3(RNAi), and cco-1 (RNAi) plates. Data represents the mean of three independent biological 698 replicates with error bars showing the standard error of the mean. N = 3, biologically independent 699 samples. ***p < 0.001 (one-way ANOVA). E. Representative images of worms expressing hsp-700 16.2pr::ssAβ (1-42)-GFP labelled protein aggregation puncta grown at 37°C for 1 hr during Day 1 adult 701 following control, and cco-1(RNAi). Scale bar, 50 μ m. N =3, biologically independent replicates. F. 702 Representative image of unc-47pr::GFP expression in the GABAergic neurons of wildtype, atfs-1(null), 703 and zip-3pr::zip-3-GFP worms. Scale bar, 50 μ m. N =3, biologically independent replicates. G. 704 Representative image of dat-1pr::GFP expression in the dopaminergic neurons of wildtype, atfs-705 1(RNAi), and zip-3pr::zip-3-GFP worms. Scale bar, 50μ m. N =3, biologically independent replicates. 706 Supplementary Figure 3. Inhibiting cell surface r eceptors leads to ZIP-3 degradation and ATFS-707 1 activation. A. Representative image of zip-3pr::ZIP-3::GFP expression following control, rme-708 8(RNAi), vps-35(RNAi), and vps-29 (RNAi). Scale bar, 100 μ m. N =3, biologically independent 709 replicates. B. Representative images of TMRE staining on wildtype worms at the L4 stage grown on 710 control and rme-8(RNAi) HT115 bacteria. Scale bar, 50 μ m. N =3, biologically independent replicates. 711 C. Representative image of UPR mt reporter hsp-6 pr::GFP expression in wildtype, fshr-1(ok778), mgl-712 2(tm355), and paqr-2(tm3410) worms grown on control and wwp-1(RNAi) HT115 bacteria. Scale bar, 713 200μ m. N =3, biologically independent replicates. D. Representative image of UPR mt reporter hsp-714 6pr::GFP expression in wildtype, arrd-8(cmh21), arrd-11(cmh22), gsa-1 (RNAi), and gplb-1(RNAi) 715 worms grown on control and wwp-1 (RNAi) HT115 bacteria. Scale bar, 200 μ m. N =3, biologically 716 independent replicates. E. Representative image of zip-3pr::ZIP-3::GFP expression following control, 717 fshr-1(RNAi), paqr-2(RNAi), mgl-2(RNAi) and arrd-8 (RNAi). Scale bar, 100 μ m. N =3, biologically 718 independent replicates. F. Representative image of UPR mt reporter hsp-6 pr::GFP expression in 719 wildtype, cam-1(YA) and ver-1(YA) worms. PPxY motifs on the CAM-1 and VER-1 genes have been 720 mutated. Scale bar, 200 μ m. N =3, biologically independent replicates. G. A schematic showing WWP-721 1 interaction with its substrates, cell surface receptors, and ZIP-3. Lack of one substrate makes the 722 interaction with the other substrate more prominent. H. Representative image of UPR mt reporter hsp-723 6pr::GFP expression in ZIP-3(YA) mutants following control, cam-1(RNAi), ver-1(RNAi) and lin-724 17(RNAi). The PPxY motif of the ZIP-3 gene has been mutated. Scale bar, 200 μ m. N =3, biologically 725 independent replicates. 726 Supplementary Figure 4. Inhibiting cell surface receptors limits A β (1-42) aggregation. A. 727 Representative images of worms expressing hsp-16.2pr::ssAβ (1-42)-GFP labelled protein aggregation 728 puncta grown at 37°C for 1 hr during Day 1 adult following control, fshr-1(RNAi), cam-1(RNAi), and lin-729 17(RNAi). Scale bar, 50 μ m. N =3, biologically independent replicates. B. A schematic of the ZIP-3 730 protein showing the predicted phosphorylation sites by different families of kinases (JNK, PKA, and 731 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 23 PKC). Ligand binding at the cell surface receptors leads to activation of downstream kinases and ZIP-732 3 phosphorylation, which in turn leads to ZIP-3 stability. C. Representative image of UPR mt reporter 733 hsp-6pr::GFP expression following control, kin-1(RNAi), pkc-2(RNAi), and kgb-1(RNAi). Scale bar, 734 200μ m. N =3, biologically independent replicates. D. Representative image of zip-3pr::ZIP-3::GFP 735 expression following control, kin-1(RNAi), pkc-2(RNAi), and kgb-1(RNAi). Scale bar, 100 μ m. N =3, 736 biologically independent replicates. E. Developmental rate plotted as the time taken to become an 737 adult from egg. The graph represents the mean of two independent biological replicates, and the error 738 bars shows standard error of means. *** p < 0.001 (one-way ANOVA), all RNAi conditions vs zip-739 3pr::ZIP-3. 740 Supplementary Figure 5. Mutating the C2 domain on WWP-1 activates ATFS-1-dependent 741 transcription. A. Representative image of UPR mt reporter hsp-6 pr::GFP expression in wildtype and 742 wwp-1(NE) mutants grown on control and atfs-1(RNAi). Scale bar, 200 μ m. N =3, biologically 743 independent replicates. B. A schematic of the mutation on the C2 domain of WWP-1. The C2 domain 744 is required to facilitate cell membrane tethering of WWP-1 and interaction with its HECT domain to 745 promote autoinhibition. C. Survival graph of wildtype (N2), wwp-1(null) and wwp-1(NE) worms on 746 Pseudomonus aeruginosa (PA14) at 25°C. D. Lifespan graphs for wildtype, wwp-1(null), and wwp-747 1(NE). N =3, biologically independent replicates. *p < 0.01 (C, D) (log-rank test). 748 Supplementary Figure 6. Inhibiting AT FS-1-dependent extracellular protease W01A8.6 and 749 extracellular chaperone DNJ-20 limits the long lifespan of clk-1 and isp-1 mutants. A. Lifespan 750 graphs for wildtype, W01A8.6(null), and dnj-20(RNAi). N =2, biologically independent replicates. * p < 751 0.01 (wildtype vs W01A8.6 and wildtype vs dnj-20) (log-rank test). B. Lifespan graphs for isp-752 1(qm150) mutants compared to isp-1(qm150), W01A8.6(null), and isp-1(qm150), dnj-20(RNAi). N =2, 753 biologically independent replicates. ** p < 0.001 ( isp-1 vs isp-1, W01A8.6 and isp-1 vs isp-1, dnj-20) 754 (log-rank test). C. Lifespan graphs for clk-1(qm30) mutants compared to clk-1(qm30) , W01A8.6(null), 755 and clk-1(qm30), dnj-20(RNAi). N =2, biologically independent replicates. ** p < 0.001 ( clk-1 vs clk-1, 756 W01A8.6 and clk-1 vs clk-1, dnj-20) (log-rank test). 757 758 759 760 761 762 763 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 24 764 765 766 767 768 769 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 25 Figure 1. Activation of ATFS-1 promotes the expression of extracellular genes that limit 770 aggregation. 771 772 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 26 Figure 2. Inhibition of ZIP-3 reduces extracellular protein aggregation. 773 774 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 27 Figure 3. Inhibiting zip-3 and the ETC complex component cco-1 ameliorates phenotypes in 775 extracellular protein aggregation models. 776 777 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 28 Figure 4. Inhibition of endosomal recycling leads to reduced ZIP-3 level and expression of 778 extracellular genes dependent on ATFS-1 activation. 779 780 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 29 Figure 5. Kinases downstream of cell signaling r eceptors that undergo endosomal recycling 781 regulate ZIP-3 via phosphorylation. 782 783 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 30 Supplementary Figure. 784 Supplementary Figure 1. ATFS-1 promotes ex pression of endosomal recycling components 785 and extracellular genes that limit protein aggregation. 786 787 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 31 Supplementary Figure 2. Inhibition of atfs-1 and zip-3 overexpression leads to 788 neurodegeneration. 789 790 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 32 Supplementary Figure 3. Inhibiting cell surface r eceptors leads to ZIP-3 degradation and ATFS-791 1 activation. 792 793 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 33 Supplementary Figure 4. Inhibiting cell surface receptors limits Aβ (1-42) aggregation. 794 795 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 34 Supplementary Figure 5. Mutating the C2 domain on WWP-1 activates ATFS-1-dependent 796 transcription. 797 798 799 800 801 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint 35 Supplementary Figure 6. Inhibiting AT FS-1-dependent extracellular protease W01A8.6 and 802 extracellular chaperone DNJ-20 limits the long lifespan of clk-1 and isp-1 mutants. 803 804 805 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 13, 2026. ; https://doi.org/10.64898/2026.03.12.711310doi: bioRxiv preprint