The discovery and consequences of the central role of the nervous system in the control of protein homeostasis

Abstract

Organisms function despite wide fluctuations in their environment through the maintenance of homeostasis. At the cellular level, the maintenance of proteins as functional entities at target expression levels is called protein homeostasis (or proteostasis). Cells implement proteostasis through universal and conserved quality control mechanisms that surveil and monitor protein conformation. Recent studies that exploit the powerful ability to genetically manipulate specific neurons in C. elegans have shown that cells within this metazoan lose their autonomy over this fundamental survival mechanism. These studies have uncovered novel roles for the nervous system in controlling how and when cells activate their protein quality control mechanisms. Here we discuss the conceptual underpinnings, experimental evidence and the possible consequences of such a control mechanism.

PRELUDE:

Whether the detailed examination of parts of the nervous system and their selective perturbation is sufficient to reconstruct how the brain generates behavior, mental disease, music and religion remains an open question. Yet, Sydney Brenner’s development of C. elegans as an experimental organism and his faith in the bold reductionist approach that ‘the understanding of wild-type behavior comes best after the discovery and analysis of mutations that alter it’, has led to discoveries of unexpected roles for neurons in the biology of organisms.

Introduction

Most protein-based biological mechanisms proceed optimally only within a narrow range of environmental conditions (Fields, Dong, Meng, & Somero, 2015; Hofmann & Somero, 1995; Somero, 1995). Despite this, organisms thrive in a variety of ecological niches. A mechanism by which organisms function despite wide fluctuations in their environment is through the maintenance of homeostasis (from Greek: ὅμoioς homoeos, ‘similar’ and στάσις stasis, ‘standing still’), whereby the internal milieu is maintained close to constant in the face of external perturbations. This concept, first described by Claude Bernard in 1865, has provided the foundation for our understanding of macroscopic physiological processes such as the maintenance of core body temperature, pH of blood (Bernard, 1965), regulation of food intake, body mass, and energy metabolism, all of which are under homeostatic control. More recently, the concept of homeostasis has been extended to mechanisms that occur at the sub-cellular level, such as the maintenance of protein function in cells, and has become popularly known as proteostasis (Balch, Morimoto, Dillin, & Kelly, 2008; Gidalevitz, Ben-Zvi, Ho, Brignull, & Morimoto, 2006; Powers, Morimoto, Dillin, Kelly, & Balch, 2009). In general, there appear to be two distinct kinds of strategies to achieve homeostasis. The first are servomechanisms (Leow, 2007), whereby error-sensing negative feedback loops correct the performance of a system to maintain some specific feature, called a regulated variable, constant or close to some set-point (Figure 1(A)). Such a mechanism requires an error detector that senses deviations from the set-point, and a mechanism to perform the necessary error corrections (often called the servomotor, as this terminology grew from mechanical devices such as steam engines). While servomechanisms can precisely maintain the regulated variable at the set-point, they are triggered only after the set-point, and homeostasis, have been perturbed. An alternate mechanism by which biological systems achieve homeostasis is through the activation of anticipatory mechanisms. These mechanisms, termed Cephalic mechanisms, are predictive, being implemented prior to the actual perturbation of the system. For instance, temperature homeostasis in an organism is maintained through mechanisms that prevent the core body temperature from deviating above or below a certain range through launching preemptive/anticipatory adjustments such as sweating, before core body temperatures are perturbed. The essence of such mechanisms is that, by definition, the command for the correction of the deviation comes from centers that can predict the oncoming perturbation using associated information, prior to themselves becoming perturbed. Thus, cephalic mechanisms can prevent that the system be perturbed in the first place (Lechan & Fekete, 2006; Mattes, 1997); however, they are not fail-safe as they can be ‘fooled’ and triggered even in the absence of any perturbations to homeostasis.

Figure 1.

(A) A brief definition of control theory terminology in relation to its use to describe proteostasis. (B) A simplified model of proteostasis. Protein conformation is the regulated variable that is maintained at a set-point through the regulated activities of protein biogenesis and protein degradation (controlled variables). CONTROLLERS (here mainly, molecular chaperones) that surveil and monitor protein conformation are present at specific amounts during the normal activities of the cells. They direct the activity of the controlled variables through PLANTS (e.g. ribosomes, proteasome, autophagy etc.) to generate homeostatic levels of the regulated variable (namely low levels of misfolded proteins). Extra CONTROLLERS can be generated in response to signals that change the levels of the regulated variable. However, as explained in the text, increasing the levels of the CONTROLLERS is under neuronal control in C. elegans.

Until about a decade ago, the mechanisms by which cells and organisms maintained protein homeostasis were best studied in isolated cells in culture or unicellular organisms such as yeast and bacteria (Akerfelt, Morimoto, & Sistonen, 2010; Balch et al., 2008; Goff & Goldberg, 1987; Goldberg, 1971, 1972; Hightower, 1980; Lindquist, 1986; Morano, Liu, & Thiele, 1998; Neef, Turski, & Thiele, 2010; Ritossa, 1996). In these systems, protein homeostasis is implemented mainly through feed-back loops that act as servomechanisms (Morimoto, 1998; Voellmy, 1994, 1996). In the last decade, however, experiments that may not have been so readily feasible were it not for the ability to manipulate neurons in C. elegans (Avery, Bargmann, & Horvitz, 1993; Chalfie et al., 1985; Walrond, Kass, Stretton, & Donmoyer, 1985; White, Southgate, Thomson, & Brenner, 1983, 1986), demonstrated that in this metazoan the nervous system exerts neuroendocrine control over the conserved cellular machinery that maintains protein quality control. Such control enables the animal to trigger these deeply conserved survival and defense mechanisms in anticipation of a threat. Here we briefly highlight how the genetic manipulation of C. elegans neurons enabled these insights and discuss some of the implications of the existence of cephalic control over cellular protein homeostasis. We focus mainly on the conceptual framework, and would like to direct interested readers to more thorough reviews for details (Miles, Scherz-Shouval, & van Oosten-Hawle, 2019; Wolff, Weissman, & Dillin, 2014).

Proteostasis

A mammalian cell synthesizes on average ~10,000 proteins at any given time, most of which can assume multiple conformations, many of which are unstructured, occur in multiple copies, have widely varying half-lives, and are in the continual process of being modified, transported and degraded (Jayaraj, Hipp, & Hartl, 2020; Mogk, Bukau, & Kampinga, 2018; Tyedmers, Mogk, & Bukau, 2010; Wolff et al., 2014). Despite this complexity, cells largely maintain a target expression level of any particular protein, and protein concentrations do not vary erratically. Moreover, cells can detect the presence of protein aggregates and non-native protein conformations (Geiler-Samerotte et al., 2011; Vendruscolo, Knowles, & Dobson, 2011), and respond in specific and conserved ways to restore conformation or trigger degradation of the abnormal species (Bednarska, Schymkowitz, Rousseau, & Van Eldere, 2013; Hipp, Park, & Hartl, 2014; Labbadia & Morimoto, 2015; Mattoo & Goloubinoff, 2014; Miller et al., 2015; Silverman et al., 2015; Vendruscolo et al., 2011; Walther et al., 2015). Since the conformation of proteins drives their stability and function, this ability of cells to maintain proteins as functional entities at specific concentrations occurs through mechanisms that surveil protein conformation (Ananthan, Goldberg, & Voellmy, 1986; Goff & Goldberg, 1985, 1987; Goldberg, 1972; Kandror, Busconi, Sherman, & Goldberg, 1994; Knowles, Gunn, Hanson, & Ballard, 1975; Knowles & Ballard, 1976, 1978; Prouty, Karnovsky, & Goldberg, 1975). These mechanisms are collectively called proteostasis (Balch et al., 2008; Gidalevitz et al., 2006). Exciting studies have demonstrated the existence of interconnected networks of surveillance and quality control mechanisms that monitor protein quality during normal protein synthesis. For instance, ribosome stalling that can occur due to stochastic perturbations of translation, mutations, or a number of biotic or abiotic fluctuations in the environment will lead to the aborted synthesis of a nascent polypeptide chains. These aborted polypeptides are recognized by the cellular surveillance machinery and modified by the addition of carboxy-terminal alanine and threonine residues (CAT tails) (Brandman et al., 2012; Kostova et al., 2017; Sitron, Park, & Brandman, 2017; Sitron, Park, Giafaglione, & Brandman, 2020; Sitron & Brandman, 2019). This, in turn, leads to their aggregation and targeted degradation through the sequestration of a group of specialized proteins called molecular chaperones that recognize the aberrant conformation of the CAT-tailed polypeptides (Sitron et al., 2017, 2020; Sitron & Brandman, 2019). The sequestration of molecular chaperones is also thought to activate a conserved transcription factor to increase the amounts of molecular chaperone proteins available to surveil and prevent the build-up of aberrant polypeptides. Translating these concepts into homeostatic control theory (Figure 1(A,B)) protein conformation is the regulated variable, being maintained at some set-point (optimal flux, optimal concentration etc.) by the activity of controlled variables: protein biogenesis and protein degradation (green boxes). The regulated variable is maintained by the activity of PLANTS: ribosomes for protein biogenesis and proteases, the proteasome, and autophagy pathways for degradation. The set-point for misfolded proteins is monitored and maintained through the activity of CONTROLLERS, the best known of which are molecular chaperones that surveil the folding state of proteins. If the CONTROLLER detects deviations from the set-point, signals to the PLANTS change the flux of proteins through the system restoring the regulated variable to its acceptable set-point (Goldstein & Kopin, 2017; Iberall & Cardon, 1964).

As with most biological processes, the levels of proteins in a cell, and therefore also misfolded proteins, varies with the types of protein synthesized, physiological needs and environmental changes. Accordingly, controlled variables i.e. the flux of proteins through the PLANTS, and even the concentration of PLANTS and CONTROLLERS required for maintaining protein homeostasis vary. The best understood regulatory mechanisms by which a sudden increase in the regulated variable, i.e. misfolded protein, impacts the PLANTS and CONTROLLERS, are the so-called heat shock response (HSR) in the cytoplasm and the unfolded protein response in the endoplasmic reticulum (Gomez-Pastor, Burchfiel, & Thiele, 2018; Vihervaara, Duarte, & Lis, 2018). First discovered as a heat-inducible transcriptional response when Drosophila busckii cells were accidently exposed to high temperatures (De Maio, Santoro, Tanguay, & Hightower, 2012; Ritossa, 1996), the HSR is a highly conserved, universal transcription program executed in response to protein damage (Akerfelt et al., 2010; Ananthan et al., 1986; Morimoto, 1998; Voellmy, 1994). The HSR is mediated by the heat shock transcription factor 1 (HSF1) whose activity results in the rapid expression of more molecular chaperones (CONTROLLERS) to help identify, and refold or degrade the damaged proteins. Molecular chaperones accomplish this by triggering changes in the controlled variables, typically promoting a decrease in protein biogenesis and an increase in degradation of the damaged proteins (Goldberg, 1971; Hightower, 1980; Lindquist, 1986; Morimoto, 1998). Similarly, the Unfolded Protein Response of the Endoplasmic Reticulum (UPR^ER) is a conserved cellular mechanism to detect and regulate proteostasis within the lumen of the endoplasmic reticulum (ER) if protein processing stalls due to the increase in aberrant ER proteins (Gething & Sambrook, 1992; Kozutsumi, Segal, Normington, Gething, & Sambrook, 1988; Munro & Pelham, 1986, 1987; Pelham & Munro, 1993). UPR^ER acts through the activation of all or a subset of three stress signal transducers found within the ER membrane, Inositol-requiring enzyme 1 (IRE1), Activating transcription factor 6 (ATF6), and protein kinase R-like endoplasmic reticulum kinase (PERK), which also increase the molecular chaperones (CONTROLLERS) that act in the ER and modulate the controlled variables (Mori, 2009; Mori, Kawahara, Yoshida, Yanagi, & Yura, 1996; Walter, 2010; Yoshida, Matsui, Yamamoto, Okada, & Mori, 2001). Parallel mechanisms exist for ensuring protein homeostasis in mitochondria (Haynes & Ron, 2010). Until recently, activation of the transcription factors that increase the amount of CONTROLLERS (HSF1 or IRE1, ATF6 and XBP1) was thought to be solely triggered by the individual cells in which protein homeostasis was perturbed. Thus, both HSF1 and the transcriptional activators of the UPR^ER are typically inhibited unless the flux of protein misfolding exceeds some set-point or threshold because they themselves are under the negative inhibition of the CONTROLLERS. Upon sensing an increase in protein misfolding, the CONTROLLERS are kinetically titrated away, dis-inhibiting (or permitting the activation of) the transcription factors, increasing the amounts of CONTROLLERS, and restoring homeostasis (Bernales, Papa, & Walter, 2006). Indeed, this is what is thought to occur during normal protein synthesis upon ribosome stalling. Under these conditions, the aborted synthesis of the nascent aggregation prone polypeptide titrates away the negative inhibition on HSF1 activity to promote an increase in chaperone gene transcription. This elegant model thus lay the foundation for the proteostasis network in isolated cells.

Proteostasis in C. elegans

C. elegans possesses dedicated neurons to sense threats in its environment and can detect environmental changes well before they are severe enough to disrupt protein homeostasis (Bargmann, 1998; Gally & Bessereau, 2003; Goodman & Sengupta, 2019; Hobert, 2005). One such sensory system that senses temperature change is the thermosensory circuitry, which involves at least three pairs of sensory neurons (AFD, AWC and ASI), that are extraordinarily sensitive to temperature changes, and a layer of interneurons (AIZ and AIY) that process and integrate the temperature information (Beverly, Anbil, & Sengupta, 2011; Clark, Biron, Sengupta, & Samuel, 2006; Goodman & Sengupta, 2018, 2019). It had been shown, once all the neurons of C. elegans had been catalogued and mapped, that AFD neurons can detect changes as small as 0.05 °C above ambient temperature; hence, they could arguably be excited by temperature increments well below those that cause cellular damage (Beverly et al., 2011; Clark et al., 2006; Colosimo et al., 2004). With such specialized machinery to sense the environment, had metazoans with a nervous system evolved mechanisms to activate their protein homeostasis machinery prior to experiencing temperatures where their proteins would misfold? This was the question that motivated the first experiments that demonstrated that, in addition to the elegant servomechanism of proteostasis that existed in individual cells, in C. elegans there also exists cephalic control over protein homeostasis (Prahlad, Cornelius, & Morimoto, 2008). In the first experiments we used mutations that impaired the thermosensory capacity of the AFD neurons, and asked whether chaperones would be induced upon a transient increase in temperatures, if the animal did not sense the temperature upshift. Indeed, mutations affecting the ability of thermosensory AFD neurons to sense heat also delayed HSF1 dependent molecular chaperone induction throughout C. elegans upon heat shock (Prahlad et al., 2008) (Figure 2(A)). However, mutations that impair neurons throughout the lifetime of animals could arguably have reset physiology and metabolism such that animals had altered the flux of proteins through the PLANTS or re-set their thresholds for tolerating misfolded proteins. Therefore, the more convincing demonstration that proteostasis was under cephalic control came from experiments using optogenetics to excite the AFD neurons in the absence of heat. Such a perturbation mimicked the sensation of heat and was sufficient to simulate a HSR, activate HSF1 and upregulate HSPs in distal tissues (Figure 2(A)) (Tatum et al., 2015). We found that the signal for AFD-dependent upregulation of HSPs was the ancient bioamine, serotonin that was released upon the excitation of the AFD neurons. Serotonin is best studied for its roles in mood, anxiety and cognition, and in almost all organisms real or perceived threats cause the release of neural serotonin (Cruz-Corchado, Ooi, Das, & Prahlad, 2020; Tatum et al., 2015). In C. elegans, the use of powerful genetic tools including neuronal mutants that lack all serotonin, had shown that serotonin played a pivotal role in learning and memory (Chase & Koelle, 2007; Curran & Chalasani, 2012). These studies have thus immediately linked anxiety, experience, learning—aspects that are fundamental to cephalic processes—to cellular changes in transcription and protein quality control. Indeed, in following experiments we have shown that C. elegans can ‘learn’, through prior experience of a threat, to enhance chaperone expression were it to encounter the specific threat in its environment (Ooi & Prahlad, 2017) (Figure 2(B)). Moreover, the release of serotonin that occurs upon the sensing of temperature change has a rapid effect on chromatin accessibility in the germline of animals conferring stress protection on future offspring (Das et al., 2020) (Figure 2(C)).

Figure 2.

Consequences of neuronal and cephalic control over homeostasis in C. elegans. (A) Neuronal circuits through serotonergic, dopaminergic, WNT, FLP-2 and octopaminergic signaling can activate transcription factors that are the master regulators of chaperones, and degradation machinery such as the ubiquitin-proteasome system in distal tissue of C. elegans. (B) Prior experience of an environmental threat can enhance the expression of chaperones upon actual encounter with the threat. (C) Maternal release of serotonin upon stress encounter increases the stress tolerance of progeny. (D) Neuronal control enables rapid response to stimuli at the cost of the normal proteostasis machinery.

Analogous neurohormonal signaling mechanisms have since been discovered for the CONTROLLERS of proteostasis in other cellular compartments in C. elegans (Berendzen et al., 2016; Durieux, Wolff, & Dillin, 2011; Frakes et al., 2020; Taylor, Berendzen, & Dillin, 2014; Taylor & Dillin, 2013; Zhang et al., 2018) (Figure 2(A)). Thus, the constitutive ectopic expression of XBP1 in the nervous system—in neurons or supporting glial cells—or the activation of the mitochondrial UPR in neurons induces a systemic UPR^ER or UPR^MT respectively, in other tissues (Durieux et al., 2011; Frakes et al., 2020; Taylor et al., 2014). In addition, it appears that neurons may not be the only sentinels of harm: activation of the heat shock response in one tissue results in the systemic induction of this stress response (Miles et al., 2019; O’Brien et al., 2018; van Oosten-Hawle, Porter, & Morimoto, 2013). Thus, it appears that in a multicellular organism, changes in proteostasis in one tissue can feedback onto protein metabolism in another tissue to systemically increase the quantity of the molecular chaperones or the CONTROLLERS that monitor set points for misfolded proteins.

Many studies have also shown, directly or indirectly, that the autonomy of cells over the controlled variables, protein biogenesis and protein degradation, is also outsourced to neurons (Ben-Gedalya & Cohen, 2012). The nervous system controls growth, metabolism and aging in response to nutrient and energy status of the animal (Broughton & Partridge, 2009; Burkewitz et al., 2015; Kawli, Wu, & Tan, 2010; Minnerly, Zhang, Parker, Kaul, & Jia, 2017; Weir et al., 2017; Wolkow, Kimura, Lee, & Ruvkun, 2000; Zhang et al., 2019). As a consequence, dietary restriction activates the conserved cap’n’collar transcription factor, SKN-1 in two neurons (called the ASI neurons) (Bishop & Guarente, 2007; Schmeisser et al., 2013) to increase metabolic activity (Bishop & Guarente, 2007; Schmeisser et al., 2013). SKN-1 is also a key regulator of the abundance of the ubiquitin-proteasome machinery within the animal (Lehrbach & Ruvkun, 2016; Li et al., 2011; Steinbaugh et al., 2015). Dietary restriction is thus known to increase lifespan through (amongst other mechanisms) the increase in proteasome activity and the E3 ubiquitin ligase WWP-1 (Carrano, Liu, Dillin, & Hunter, 2009). Remarkably, the smell of food is sufficient to reverse the benefits of dietary restriction and a recent study demonstrated rapid changes in protein degradation rates in response to food odors sensed by olfactory neurons (Finger et al., 2019). Similarly changes in the levels of the neuromodulator dopamine control the flux of proteins by modulating protein degradation through the ubiquitin-proteasome system (Joshi, Matlack, & Rongo, 2016). Moreover, since growth and differentiation of cells within metazoa such as C. elegans occur in a coordinated manner, it is perhaps not surprising that neurons are not the only cell-type that exerts hierarchical control over the proteostasis machinery of another cell: cells also manage the proteostasis control systems of each other. For instance, germline stem cells limit the capacity of muscle cells, and perhaps other cells, to respond to heat stress in multiple tissues by inhibiting the amounts of PLANTS, in this case the regulatory particle non-ATPase 6 (RPN6) proteasome subunit that these distal cells are allowed to express (Sinha & Rae, 2014; Vilchez et al., 2012). Similarly, specialized cells called coelomocytes play a key role in the clearance of aggregation-prone transthyretin expressed in C. elegans muscle (Madhivanan et al., 2018), and protein degradation in the muscle, albeit driven by neuronal activity, is the integrated outcome of signals from neurons and other tissues such as hypodermis and intestine. In turn, distal cells modulate controlled variables to feedback on neurons and the rest of the animal (Chikka, Anbalagan, Dvorak, Dombeck, & Prahlad, 2016; Dalton & Curran, 2018; Minnerly et al., 2017).

While neuronal networks confer the ability to respond rapidly to environmental threats, we and others have found that this ability comes at the expense of the ‘normal’ proteostasis machinery (Berendzen et al., 2016; Cao & Aballay, 2016; David, 2013; El-Ami et al., 2014; Imanikia, Ozbey, Krueger, Casanueva, & Taylor, 2019; Joshi et al., 2016; Maman et al., 2013; Moll, Ben-Gedalya, Reuveni, & Cohen, 2016; O’Brien et al., 2018; Prahlad & Morimoto, 2011; Ray, Zhang, Rentas, Caldwell, & Caldwell, 2014; Roitenberg et al., 2018; Taylor & Dillin, 2013) (Figure 2(D)). For example, the AFD-dependent circuity accelerates the activation of HSF1 upon sensing environmental stress but prevents HSF1 from being activated when cells chronically express aggregation prone proteins (Prahlad & Morimoto, 2011). Consequently, the activity of the AFD-thermosensory neurons needs to be inhibited in order to allow cells in the metazoan to activate HSF1 and increase the expression of CONTROLLERS needed to respond adequately to protein aggregation. Similarly, in wild-type animals, the presence of a G protein coupled receptor expressed on chemosensory neurons (gtr-1) is required for animals to increase chaperone expression upon exposure to heat, but its presence inhibits the ‘normal’ proteostasis machinery from rescuing the proteotoxicity of Alzheimer’s-disease-linked aggregative peptide Aβ_3–42 (Maman et al., 2013). A similar phenomenon is seen in studies on aging, a process that results in the systemic loss of protein quality control and the accumulation of misfolded proteins in various tissues: sensory neuronal function while allowing animals to respond to their environment, invariably inhibits the proteostasis machinery from acting on misfolded proteins (Boccitto, Lamitina, & Kalb, 2012; Broughton & Partridge, 2009; Burkewitz et al., 2015; Byrne, Wilhelm, & Richly, 2017; Jiang et al., 2015; Kumsta et al., 2014; Stein & Murphy, 2012; Taylor & Dillin, 2013; Yin, Liu, Yuan, Jiang, & Cai, 2014; Zhang et al., 2019). In these cases, the loss of normal neuronal control is required to allow protein misfolding to activate the normal proteostasis machinery, increase CONTROLLERS, and recalibrate the flow of proteins through the PLANTS to restore health and function (Figures 1 and 2). Why might such a seemingly counterintuitive mechanism have evolved? We believe this is because neuronal signals have higher physiological priority because they orchestrate the protective response to life threatening environmental changes. Therefore, in a metazoan, the ability of neurons to antagonize the mechanisms that could normally be used by the proteostasis machinery of individual cells allows neurons to disengage this very same machinery for use upon the sensing of a threat. The control over the activation of transcription factors that are responsible for the expression of CONTROLLERS in all compartments of cells allows neurons to play this dominant role. In addition, such a control mechanism also forces the normal proteostasis machinery—the CONTROLLERS and the PLANTS involved in protein biogenesis—to function within a range of stringent set points under normal physiological conditions even through the regulated variable, protein misfolding, itself may not be well defined. Indeed, experiments in C. elegans have shown that the proteostasis machinery operates within a narrow range which is not easily perturbed (Gidalevitz et al., 2006).

It is interesting to speculate on what could happen were cell intrinsic and cell extrinsic control mechanisms in disagreement. In human diseases such as Alzheimer’s disease and Parkinson’s disease, and in animal models of these diseases, cells accumulate misfolded and aggregated proteins but fail to naturally activate their cell-autonomous proteostasis response. This suggests that in the absence of external stimuli, the cell-extrinsic control dominates, preventing the increase in CONTROLLERS and PLANTS even at the cost of increasing the cellular burden of misfolded proteins. Indeed, experimentally increasing CONTROLLERS such as molecular chaperones in the affected cells, by transfection or other means, ameliorates misfolding and disease pathology. However, it is also telling that the cell intrinsic increase in the abundance of CONTROLLERS and PLANTS occurring in the apparent absence of cell extrinsic control is a prerequisite for cancer maintenance and malignancy in mammals (Dai, Whitesell, Rogers, & Lindquist, 2007; Gaglia et al., 2020; Mendillo et al., 2012; Scherz-Shouval et al., 2014). Since the most obvious function of cephalic control over a community of cells, where each cell has autonomous control over its own proteostasis machinery, is the maintenance of harmony and co-ordination required for multicellular function, one might image that an anarchy of cell intrinsic control over that exerted by neurons may result in defects in development and differentiation, and other such organismal functions.

Is what we have learnt from C. elegans more generally true?

The ability to manipulate C. elegans neurons has uncovered a novel role for the nervous system in the control of protein quality control mechanisms. Dysfunction of proteostasis is responsible for the devastating age-associated neurodegenerative diseases and other disorders prevalent in our societies today. Therefore, that need to understand the extent to which neurohormonal signaling pathways influence cellular proteostasis in mammals is urgent. Interventions to activate protective protein quality control mechanisms in the treatment of protein conformational diseases have, to date, been based largely on the assumption that protein misfolding, and its subsequent correction, are cell autonomous processes: namely the cells that need to be targeted are the cells within which protein misfolding occurs (Balch et al., 2008; Powers et al., 2009). However, if protein folding homeostasis in all multicellular organisms is under the cell non-autonomous regulation of sensory neuronal modalities, another method for treatment of neurodegenerative diseases could involve modulation of neurosensory systems: a method that can be used alone, or in combination with small molecules to effectively target protein misfolding. Evidence for such control in mammals does exist, and is growing. Prior to the studies in C. elegans, restraint stress had been shown to induce expression of HSP70 mRNA in the adrenal cortex of the rat through the activity of the by the hypothalamic-pituitary-adrenal (HPA) axis (Blake, Buckley, & Buckley, 1993; Blake, Udelsman, Feulner, Norton, & Holbrook, 1991; Fawcett, Sylvester, Sarge, Morimoto, & Holbrook, 1994). Notably, the control of the proteostasis machinery of cells by the nervous system shares features with the systemic inhibition of acute inflammation seen in mammals (Tracey, 2002) and invertebrates including C. elegans (Anderson, Laurenson-Schafer, Partridge, Hodgkin, & McMullan, 2013; Singh & Aballay, 2012; Styer et al., 2008; Sun, Liu, & Aballay, 2012; Sun, Singh, Kajino-Sakamoto, & Aballay, 2011; Yu, Zhi, Wu, Jing, & Wang, 2018), and treatments to modulate inflammatory processes are leveraging the existence of neuronal control. In mammals cholinergic activation within the efferent vagus nerve suppresses chronic inflammation, restricting the innate immune response to remain an acute response to local insults, and not flare up into chronic inflammation (Tracey, 2002). Moreover, as with the inflammatory response, recent experiments show that increased activity of the CONTROLLERS and PLANTS is protective only under a controlled regimen, and can be detrimental if unchecked (Styer et al., 2008; Sun et al., 2011). In mouse models of Alzheimer’s Disease and Prion Disease it was the sustained activation of proteostasis machinery, rather than its dysfunction that accelerated neurodegeneration (Ma et al., 2013; Moreno et al., 2012). Protein misfolding within the cell, if detected and responded to by the CONTROLLERS elicits not repair, but also a change in the activity of the PLANTS such as a transient attenuation of protein translation through the phosphorylation of the a-subunit of eukaryotic initiation factor-2 (eIF2α) (Harding et al., 2003), presumably to lower the burden of nascent polypeptides to an already stressed proteome (Ma et al., 2013; Moreno et al., 2012). This attenuation of translation has been shown to be toxic to neurons.

Thus, in retrospect, the finding that proteostasis is systemically controlled in a metazoan could perhaps have been anticipated. The maintenance of protein homeostasis ensures the presence of adequate amounts of functional proteins in a cell and is therefore directly responsible for cell growth, differentiation and function. It is possible that the nervous system, because of its role in regulating multicellular outcomes such as aging, behavior, metabolism, longevity and reproduction, also determines the set-points of misfolded proteins in the different cells of an organism maintaining them within some ‘optimal’ range. A better understanding of the mechanisms responsible for determining these set-points could go a long-way in treating protein conformation diseases.