Systemic stress signalling: understanding the cell non-autonomous control of proteostasis

Abstract

Proteome maintenance is crucial to cellular health and viability, and is typically thought to be controlled in a cell-autonomous manner. However, recent evidence indicates that protein-folding defects can systemically activate proteostasis mechanisms through signalling pathways that coordinate stress responses among tissues. Coordination of ageing rates between tissues may also be mediated by systemic modulation of proteostasis. These findings suggest that proteome maintenance is a systemically regulated process, a discovery that may have important therapeutic implications.

Figure 1. A model for the cell non-autonomous regulation of the cytosolic HSR in Caenorhabditis elegans

Release of the neurotransmitters GABA (γ-aminobutyric acid) or acetylcholine (ACh) from inhibitory and stimulatory motor neurons can activate or suppress, respectively, heat shock factor 1 (HSF-1)-mediated protection against proteotoxicity in muscle cells. Activation of neurons of the thermoregulatory circuit by increased temperature triggers the activation of the heat shock response (HSR) in intestinal cells, which is protective under conditions of acute stress. Neuronal genes involved in the activation of this pathway include guanylyl cyclase 8 (gcy-8), the LIM homeodomain-containing abnormal thermotaxis 3 (ttx-3), the G protein-coupled receptor (GPCR) GPCR thermal receptor 1 (gtr-1), and the CAPS homologue uncoordinated 31 (unc-31). However, the specific roles of these molecules are so far unclear. These neurons also inhibit the protective effects of HSF-1 against chronic proteotoxic stress. Activation of thermoregulatory neurons leads to increased longevity through the abnormal dauer formation 9 (DAF-9) and DAF-12 steroidal signalling pathway. Finally, expression levels of DAF-21 (the homologue of heat shock protein 90 (HSP90) in C. elegans), which mediates the folding of client proteins but also inhibits HSF-1 and its downstream protective effects, can be communicated between tissues dependent on the defective pharynx development 4 (PHA-4) transcription factor.

Figure 2. Cell non-autonomous regulation of the UPR^mt in Caenorhabditis elegans

Imbalance between mitochondrial- and nuclear-encoded proteins in neurons gives rise to mitochondrial unfolded protein response (UPR^mt) activation in intestinal cells, and increased longevity through the UPR^mt regulator ubiquitin-like 5 (UBL-5). The proposed secreted mediator of this communication between neurons and intestine has been termed a ‘mitokine’.

Figure 3. Cell non-autonomous regulation of the UPR^ER in Caenorhabditis elegans and mammalian cells

A | In Caenorhabditis elegans, two models for neuronal control of the endoplasmic reticulum (ER)-mediated unfolded protein response (UPR^ER) have been proposed. Aa | Expression of the spliced and activated form of the UPR^ER transcription factor X box-binding protein 1 (XBP-1), xbp-1s, in neurons leads to activation of inositol-requiring protein 1 (IRE-1) and XBP-1 and to transcription of UPR^ER targets in distal intestinal cells, which is dependent on the release of a secreted signal (the secreted ER stress signal (SERSS)) that is mediated by uncoordinated 13 (UNC-13), a regulator of syntaxin. Ab | The second model suggests that expression of octopamine receptor 1 (OCTR-1), a G protein-coupled receptor, and arrestin 1 (ARR-1; a homologue of β-arrestin) in the chemosensory ASH, ASI, AIY or ADE neurons functions to suppress XBP-1 activation and expression of the non-canonical activated in blocked unfolded protein response (ABU) UPR^ER genes in distal cells under basal unstressed conditions. B | In human cell culture and mouse models, induction of ER stress and splicing of XBP1 in tumour cells has been shown to induce UPR^ER activation and XBP1 splicing in macrophages and dendritic cells, which results in activation and a suppressive phenotype that favours tumour growth.

Figure 4. Cell non-autonomous regulation of proteostasis by insulin-like signalling and gonadal signalling

A | Insulin-like signalling. Aa | In Caenorhabditis elegans, reduced activity of the abnormal dauer formation 2 (DAF-2) insulin-receptor homologue in neurons results in activation of the DAF-16 insulin-like signalling transcription factor in intestinal and muscle cells. Activation of DAF-16 may occur through two mechanisms: changes in insulin-like peptide release and/or a novel forkhead box O (FOXO)–FOXO signalling pathway, which both lead to upregulation of different sets of DAF-16 target genes. DAF-16 activation in intestinal cells results in extended longevity, whereas DAF-16 activation in muscle cells leads to improvements in proteostasis. Bidirectional communication of DAF-16 activity between these tissues depends on the transcriptional co-regulator mediator 15 (MDT-15). Ab | In Drosophila melanogaster, activity of Foxo causes an increase in activity of the Foxo target 4E-BP, a regulator of translation. 4E-BP activity improves muscle cell proteostasis and delays muscle decline, which is mediated in part by autophagy. Moreover, 4E-BP activity results in changes in feeding behaviour, which alters insulin-like peptide release and therefore the regulation of Foxo activity and proteostasis in distal tissues, and ultimately influences longevity. Activation of Foxo in muscle cells also reduces activin signalling, which leads to an increase in autophagy, improved proteostasis and a reduction in insulin-like peptide release from the brain. Finally, mitochondrial stress activates the mitochondrial unfolded protein response (UPR^mt) in muscles, increasing proteostasis and the release of an insulin-binding protein, Ecdysone-inducible gene L2 (ImpL2), which reduces systemic insulin circulation and increases longevity. B | Gonadal signalling in C. elegans. Loss of germline stem cells activates signalling pathways involving DAF-12, DAF-9 and krev interaction trapped/cerebral cavernous malformation 1 (KRI-1) that upregulate or downregulate multiple pathways in distal cells to increase longevity (for example, target of rapamycin (TOR)). Proteostasis and stress resistance are also increased, and this is at least in part mediated by an increase in proteasome activity through DAF-16-mediated upregulation of the proteasome subunit regulatory particle non-ATPase 6 (RPN6). Solid arrows indicate direct regulatory interactions, dashed arrows represent regulation that may be indirect. Black arrows indicate pathways that regulate longevity and proteostasis; grey arrows indicate pathways known to influence lifespan, but that are not currently known to modify proteostasis. FAT, fatty acid desaturase; NHR, nuclear hormone receptor; PHA, defective pharynx development.