When Proteins Go Berserk: The Unfolded Protein Response and ER Stress

Abstract

Background

The cellular proteostasis machinery is essential for maintaining protein homeostasis by employing quality control systems that identify, sequester, and eliminate damaged or misfolded proteins. However, the accumulation of misfolded proteins can overwhelm these protective mechanisms, disrupting proteostasis. This phenomenon is a hallmark of numerous pathologies, including a variety of genetic disorders. In the secretory pathway, the buildup of misfolded proteins triggers endoplasmic reticulum (ER) stress, which activates the unfolded protein response (UPR). The UPR serves as an adaptive mechanism, aiming to alleviate stress and restore cellular homeostasis. However, if ER stress is prolonged or severe, the UPR may fail to restore balance and apoptosis is induced.

Summary

This review introduces the intricate signaling pathways activated by the three UPR transmembrane sensors: protein-kinase R-like endoplasmic reticulum kinase (PERK), inositol requiring enzyme 1 (IRE1), and activating transcription factor 6 (ATF6). We briefly present the roles of the distinct transcriptional programs activated by each sensor in modulating the cellular response to protein stress and in determining cell fate. We discuss how genetic variants and environmental factors contribute to the heterogeneity observed in protein misfolding diseases. Finally, we critically evaluate select therapeutic strategies, specifically protein stabilization, trafficking modulation, and UPR sensor targeting approaches.

Key Messages

This review introduces the potential consequences of protein misfolding, which may not only impair protein function but can also lead to toxic protein accumulation and stress induction. Using Fabry disease as a compelling example, we suggest that future therapeutic intervention may require nuanced, combination approaches that address both loss and gain of protein function.

Proteostasis and ER Stress

Cellular proteostasis relies on quality control (QC) checkpoints aimed at maintaining and supporting the integrity of the proteome. These checkpoints identify damaged proteins, sequester them, and target them for disposal. In the secretory pathway, proteostasis machineries facilitate the biosynthesis of nearly one-third of the eukaryotic proteome [1]. Most secretory proteins are co-translationally translocated into the endoplasmic reticulum (ER) [2] where they are folded, matured, and then transported to the Golgi apparatus. In the Golgi, proteins undergo posttranslational modifications followed by sorting and transport to different intracellular compartments, including endosomes, lysosomes, and the plasma membrane, or secretion to the extracellular milieu [3–5].

A major challenge threatening the cellular proteostasis machinery and its overall proteome is exposure to misfolded proteins [6–9]. Protein misfolding can occur due to genetic mutations, translation errors, and environmental stresses [6, 10, 11]. It can broadly result in two non-mutually exclusive phenomena, namely, loss of protein function and gain of toxic function. Loss-of-function diseases arise from the failure of the faulty protein to perform its intended function, whereas gain of function refers to a protein acquiring novel and maladaptive functions (e.g., formation of aggregates or association with new binding partners) [12, 13].

In the secretory pathway, the toxic accumulation of misfolded proteins triggers ER stress [14]. ER stress plays a pivotal role in the development of many different pathologies, including neurodegenerative disorders, cancer, diabetes, and a variety of genetic diseases [15–18]. Notably, emerging evidence highlights a significant role of ER stress in the pathology of Fabry disease [19–21].

Unfolded Protein Response

To maintain cellular proteostasis and overcome the induced ER stress, a cellular response called the unfolded protein response (UPR) is activated. The UPR is an adaptive response mechanism designed to restore proteostasis and ER function by reducing protein synthesis and enhancing the folding and degradation capacity of the ER. However, when ER stress is prolonged or severe, the UPR may fail to restore balance, leading instead to a “terminal UPR” that drives apoptosis [22–24]. Understanding the role of the UPR in disease mechanism and progression may provide significant mechanistic insights into the pathology and identify potential therapeutic targets [25].

ER stress is sensed by three main transmembrane sensors that constitute the UPR: protein-kinase R-like endoplasmic reticulum kinase (PERK), inositol requiring enzyme 1 (IRE1), and activating transcription factor 6 (ATF6) (shown in Fig. 1). Each sensor has a distinct role and activates a specialized transcriptional program initially aimed to resolve the cellular stress and restore proteostasis [26].

Fig. 1.

Overview of the UPR signaling pathway. Misfolded protein accumulation induces ER stress and activates the three main UPR sensors: protein-kinase R-like endoplasmic reticulum kinase (PERK), inositol requiring enzyme 1 (IRE1), and activating transcription factor 6 (ATF6). Each sensor activates a distinct signaling pathway aimed at resolving the cellular stress and restoring proteostasis. If ER stress is prolonged or severe, the UPR may fail to restore balance, leading instead to a terminal UPR that drives apoptosis.

PERK Pathway

PERK (encoded by the gene EIF2AK3), is an ER-resident kinase. Under basal conditions, PERK is bound to the chaperone-binding immunoglobulin protein (BiP), which maintains it in an inactive state. However, upon misfolded protein accumulation and ER stress induction, BiP dissociates from PERK, allowing it to oligomerize, undergo autophosphorylation and subsequent activation [26] (shown in Fig. 1).

Activated PERK phosphorylates the α subunit of eukaryotic initiation factor 2 (eIF2α) at Ser51, leading to a global reduction in mRNA translation. This translation repression decreases the influx of nascent proteins into the ER, thereby reducing the load on the ER’s folding machinery and mitigating stress. At the same time, eIF2α phosphorylation selectively enhances the translation of specific mRNAs, most notably ATF4 (activating transcription factor 4). ATF4 translocates to the nucleus and activates the transcription of genes involved in protein folding, protein transport, and anti-oxidant response, thereby supporting cellular adaptation to stress [27, 28]. ATF4 also upregulates GADD34 (growth arrest and DNA damage-inducible protein 34), which in turn facilitates the dephosphorylation of eIF2α, thereby restoring general protein synthesis and preparing the cell to resume normal function if stress resolves [29, 30].

During prolonged, severe, or unresolved ER stress, PERK activation induces pro-apoptotic signaling. Under these settings, ATF4 upregulates CHOP (C/EBP homologous protein), a transcription factor that promotes apoptosis by upregulating genes involved in cell death pathways [29, 31, 32]. Moreover, sustained ATF4 induction of GADD34 and subsequent protein synthesis increases protein overload on stressed cells, leading to apoptosis [29]. Notably, while CHOP upregulation has a prominent role in terminal UPR and apoptosis induction, recent studies show that in certain instances CHOP can promote cell survival [33]. Collectively, PERK signaling has a dual role: initially protecting cells by reducing ER load and activating adaptive responses but ultimately contributing to apoptosis if ER stress is severe or prolonged. This highlights the complex balance between survival and death mediated by the UPR pathway [18, 23, 24, 28, 34].

IRE1 Pathway

IRE1α, encoded by the ERN1 gene, is the most conserved sensor [35], hinting at its crucial role in the UPR. IRE1α is the widely expressed paralog of IRE1, whereas IRE1β (ERN2 gene) is restricted to mucosal epithelial cells [34, 36]. IRE1α is a type-I transmembrane protein that contains both a serine/threonine kinase domain and an endoribonuclease (RNase) domain on its cytosolic side. The activation of IRE1α begins with the dissociation of BiP from IRE1α’s luminal domain (shown in Fig. 1). BiP dissociation enables misfolded proteins that accumulate in the ER to interact with the luminal domain of IRE1α, resulting in oligomerization and autophosphorylation of its kinase domain. This conformational change activates the RNase domain of IRE1α that selectively cleaves a 26-nucleotide segment from the mRNA of X-box-binding protein 1 (XBP1) [37]. The splicing of XBP1 induces a translational frameshift that produces a transcriptional activator (b-ZIP) domain in the C terminus of the spliced XBP1 (XBP1s), enabling it to regulate target gene transcription [38]. The genes regulated by the IRE1α-XBP1 pathway both enhance the processes of protein folding and transport, and also stimulate protein degradation pathways, thus promoting the clearance of misfolded proteins [39]. In a process known as regulated IRE1-dependent decay (RIDD), oligomerization of IRE1α mediates degradation of ER-associated mRNAs, aiming to further halt translation during a state of ER stress [18, 40].

Initial IRE1 activation causes splicing of XBP1 and RIDD-associated degradation of mRNA, which contribute to the adaptive UPR. However, when ER stress is sustained or severe, IRE1 triggers cell death by XBP1-mediated disruption of Ca²⁺ levels [38], RIDD degradation of anti-apoptotic mRNAs and miRNAs [24, 40], and activation of the pro-apoptotic c-Jun N-terminal kinase [41].

ATF6 Pathway

ATF6 (encoded by the ATF6 gene) is a type II transmembrane protein that straddles the ER membrane. Its C-terminus is located within the ER lumen, while the N-terminus faces the cytoplasm and contains a basic region/leucine zipper (bZIP) motif [42]. Mammals express two homologous ATF6 proteins, ATF6α and ATF6β, where ATF6α acts as the main driver of the ATF6 branch of the UPR (shown in Fig. 1). ATF6β appears to have some overlapping functions with ATF6α and is involved in the regulation of ATF6α, although its specific functions have yet to be fully elucidated [43, 44]. At its N-terminus, ATF6α contains an 8-amino acid region, termed VN8, which is important for optimal transcriptional activity and for its rapid degradation [43]. Upon ER stress, ATF6α dissociates from BiP and translocates to the Golgi. In the Golgi, it is cleaved by proteases S1P and S2P [43, 45]. This cleavage releases the cytosolic ATF6α fragment (ATF6f). ATF6f enters the nucleus, binds DNA through its bZIP region, and activates transcriptional programs involved in ER quality control. These programs include the induction of ER chaperones that assist in protein folding (e.g., BiP and other chaperones such as Grp94 and ERp72), and the promotion of XBP1 expression, facilitating feedback between the IRE1 and ATF6 branches of the UPR [46–48]. Additionally, XBP1 and ATF6 work together to induce transcription of ER-associated degradation components, which remove misfolded proteins from the ER and send them to the proteasome for degradation [46, 49, 50].

ATF6 signaling is generally regarded as a cytoprotective mechanism, yet upon chronic activation it can lead to cellular damage [51, 52]. In fact, chronic ATF6 activation has been shown to promote inflammation and trigger an autoimmune response [53].

Genetic Variants and Environmental/Nongenetic Factors Differentially Impact ER Stress Induction and UPR Activation

ER stress and UPR activation are involved in numerous pathologies. Many genetic diseases arise from the accumulation of misfolded proteins and their mislocalization [54]. Notably, different genetic mutations affecting a single gene may have distinct effects on the proteostasis machinery. Recently, a systematic study screened thousands of missense mutations and evaluated their effect on protein mislocalization. The study suggests that even in monogenic proteinopathies, the extent of variant mislocalization can impact disease severity and lead to disease pleiotropy [54]. Similarly, evidence indicates that in certain instances, the UPR branches may be differentially triggered depending on the specific genetic variant expressed [55]. Lastly, the distinct genetic background of individual patients is also a source of variability that may modulate the response to ER stress.

In addition to genetic variability, the environment plays a major role in ER stress induction and UPR activation. Throughout a lifetime, individuals are exposed to countless types of stress and injuries, such as inflammation, oxidative stress, hypoxia, and ultraviolet radiation. These environmental factors may each affect the cellular proteostasis and trigger ER stress and UPR signaling in their own right [56–59]. A common feature of many proteinopathies is that their pathologies are age dependent and are often exacerbated with aging [60, 61]. Therefore, environmental changes, in addition to age-related alterations, are thought to have a prominent role in disease onset and progression. Moreover, environmental conditions and nongenetic factors can potentially account for the often-observed phenotypic variability among patients who harbor the same disease-causing mutation [62–66]. Namely, a mutation with subclinical phenotypes at baseline conditions may have a severe impact under specific environmental settings [62, 67, 68].

Variability and Heterogeneity in Fabry Disease

Fabry disease, which is the main focus of this issue, has typically been understood as a disease caused by loss of function of the alpha galactosidase A (AGAL) protein. The lysosomal enzyme AGAL assists in the breakdown of glycoconjugates by cleaving off terminal galactose residues [69]. When AGAL is unable to carry out this function, lipids, most notably the glycosphingolipid globotriaosylceramide (Gb3), accumulate within the lysosome [70]. This lipid storage causes multisystemic and progressive dysfunction [71].

Despite being a monogenic disease, Fabry disease displays in highly heterogeneous ways, both genotypically and phenotypically. Nearly 1,000 unique AGAL mutations have been identified and symptom profiles vary greatly among different patients, even within a single family [72]. Not only does the severity of symptoms vary, but also the range of symptoms varies widely. While many patients present comorbidity with several organs affected, certain later-onset forms of Fabry disease are restricted to a single organ, most commonly the heart [73] or kidneys [74].

In recent years, compelling evidence has emerged that challenge the dogma that Fabry disease is merely a loss-of-function illness. Intriguingly, various AGAL mutations were shown to be associated with AGAL mistrafficking and ER stress induction [19–21]. Certain AGAL variants are incapable of navigating to the lysosome, accumulating instead in earlier compartments of the secretory pathway [19–21]. Further investigation revealed that AGAL retention in these compartments drives ER stress and UPR activation [19–21] and that this AGAL gain of function should be considered as a driver of disease in these particular AGAL mutants [19]. Notably, while UPR was demonstrated in a range of AGAL variants, its potential pathological implications are yet to be investigated.

Importantly, the degree of mislocalization and UPR activation varied among genetic variants and presented as a spectrum of proteostasis malfunction. The differential mislocalization and UPR activation associated with specific genetic variants, together with the unique exposure to distinct environmental factors throughout the patient’s life, may help explain the vast heterogeneity of Fabry disease severity and symptoms. This is even further complicated by the differential degree of loss of function of each variant, which may further accentuate the spectrum of symptoms and clinical presentations. Collectively, patients harboring different mutations, or even the same genetic defect, may experience a separate set of symptoms, or alternatively the lack thereof [65]. Interestingly, these phenomena are not restricted to Fabry disease and were previously demonstrated in other lysosomal storage diseases, most prominently in Gaucher disease [75–78].

While ER stress induction provides a compelling explanation for at least part of the variability observed in Fabry disease, it also highlights the need for the scientific community to develop tools that can accurately assess the degree of ER stress involvement in driving the pathology on a patient-specific basis. A promising step in this direction has been provided by Živná et al. [19], who demonstrated that certain variants, causing measurable ER stress, specifically activate the ATF6 branch of the UPR. Notably, the upregulation of ATF6 markers was observed not only in cell cultures but also in patient-derived kidney biopsies [19]. These findings suggest that ATF6 markers hold potential as biomarkers for evaluating the patient-specific contribution of ER stress to Fabry disease pathology. Lastly and most importantly, the emerging evidence of ER stress involvement in Fabry disease underlines the need to consider new therapeutic avenues that target ER stress induction and UPR activation.

Therapeutic Strategies to Combat ER Stress

As many devastating diseases are driven by misfolded protein accumulation and excessive ER stress, there is great importance in developing therapeutics that alleviate the stress and combat this pathological mechanism [79]. If left untreated, accumulation of mutant proteins in the early secretory pathway can overwhelm the cellular proteostasis machinery, causing apoptosis and leading to cell damage. While many different proteostasis regulating entities exist, we will focus here on three distinct therapeutic strategies, each targeting different aspects of the proteostasis machinery (shown in Fig. 2).

Fig. 2.

Selected therapeutic entities that target UPR activation and relieve misfolded protein accumulation. Three distinct therapeutic strategies are illustrated, each targeting different aspects of the proteostasis machinery: (i) UPR sensor, (ii) protein stabilization, and (iii) protein trafficking.

Strategy 1: Targeting UPR Sensors

Modulation of the response to ER stress induction can be achieved by targeting the function of the upstream UPR sensors, thereby enhancing the cell’s ability to handle misfolded protein accumulation. The drug salubrinal enhances the function of the PERK arm of the UPR by inhibiting the GADD34-protein phosphatase 1 (PP1) complex that is responsible for the dephosphorylation of eIF2α (shown in Fig. 2). GADD34-PP1 inhibition increases eIF2α phosphorylation levels, thereby maintaining the eIF2α inhibitory form and attenuating protein translation [80]. This temporary reduction in protein synthesis enables the ER to restore its protein folding capacity and has been shown to reduce markers for apoptosis [81]. By blocking ER stress-induced apoptosis, salubrinal shows potential for treating chronic ER stress conditions, including certain neurodegenerative diseases [80].

In contrast to salubrinal, the compound GSK2606414 works by inhibiting the PERK branch of the UPR (shown in Fig. 2). In prion disease, misfolded prion protein causes overactivation of the UPR, repressing translation for long stretches of time. This prevents the formation of essential proteins needed for neuronal cell health and survival. GSK2606414 was shown to prevent this neurotoxicity in prion disease mice by reducing PERK-mediated inhibition of translation [82]. Improvements to GSK2606414 led to the development of the compound GSK2656157 that also targets the UPR sensor PERK [83] (shown in Fig. 2). Notably, studies show that in addition to PERK inhibition, both compounds may more potently inhibit receptor-interacting serine/threonine-protein kinase 1 (RIPK1). Therefore, the pro-survival benefits of these compounds may partially be due to RIPK1 inhibition [84].

Strategy 2: Stabilizing Proteins

To prevent the formation of protein aggregates, several protein-folding chaperones have been developed. The chemical chaperone TUDCA (tauroursodeoxycholic acid) is a clinically approved ER stress inhibitor that stabilizes protein folding, prevents aggregation, and reduces ER stress (shown in Fig. 2). TUDCA was shown to bind heat shock protein 90 (HSP90), a molecular chaperone that assists in proper protein folding. The binding of TUDCA to HSP90 increases its protein folding capacity and helps reestablish proteostasis [85]. It also has antioxidant and anti-inflammatory properties, which help protect mitochondria and reduce inflammation. TUDCA’s broad cellular protection makes it a promising agent in treating conditions characterized by protein misfolding and ER stress, including diabetes [79].

The small molecule migalastat (deoxygalactonojirimycin) is an approved therapeutic for the treatment of Fabry disease, improving renal and cardiovascular symptoms and disease biomarkers [86]. Migalastat functions by mimicking galactose, specifically binding to unfolded/misfolded AGAL in the ER (shown in Fig. 2). This stabilizes the protein, helping it fold into a more energetically favorable and stable conformation. The properly folded enzyme is then able to traffic to its target organelle, the lysosome. Once in the lysosome, the low pH causes migalastat to dissociate from AGAL, thereby freeing the enzyme to function on its intended substrate [87].

Strategy 3: Targeting Trafficking

As described here, protein mislocalization causes profound defects in cellular health. Therefore, the modulation of protein trafficking may be a promising therapeutic strategy. Trafficking-based therapeutics focus on either reestablishing the native trafficking route of the targeted protein or, alternatively, directing the toxic protein for removal.

BRD4780 is a promising small molecule that was recently discovered to enhance misfolded protein removal and relieve ER stress [88, 89]. BRD4780 targets and engages the cargo receptor TMED9 [88, 90], which belongs to the p24 (transmembrane emp24 domain, i.e., TMED) family of cargo receptors [91]. TMED9 is a transmembrane protein that associates with COPI and COPII vesicles to shuttle protein cargos in the ER and Golgi interphase [25, 90, 92]. Emerging studies point to the involvement of TMED9 in various pathologies ranging from proteinopathies to various cancers [88, 92–96]. In different proteinopathies, TMED9 was shown to capture the misfolded cargo within the ER-Golgi interphase and promote its entrapment and retention in the early secretory pathway [88, 96]. Interestingly, the entrapment of the misfolded cargoes by TMED9 was associated with the induction of the ATF6 branch of the UPR [88]. Application of the small molecule BRD4780 targets the cargo receptor TMED9 and triggers its forward trafficking toward the lysosome, where it is degraded and cleared from the cell (shown in Fig. 2). Essentially, BRD4780 removes TMED9, which serves as the anchor for the misfolded protein entrapment. In various genetic proteinopathies, the removal of TMED9 is coupled with the transport of the mutant cargo to the lysosome. The clearance of the toxic cargo relieves the ER stress and UPR activation that was associated with the toxic accumulation [88]. BRD4780 shows therapeutic promise in kidney diseases like autosomal dominant tubulointerstitial kidney disease caused by mutations in MUC1 (ADTKD-MUC1) and ADTKD caused by mutations in UMOD (ADTKD-UMOD), as well as in retinitis pigmentosa, a blindness pathology [88]. In a recent study, Živná et al. [19] demonstrated the beneficial application of BRD4780 in certain GLA variants associated with ER retention and induction of ER stress in cell culture. Notably, the small molecule promoted the trafficking of AGAL into the lysosome, reducing ER stress by relieving the ER overload from faulty proteins. Specifically, BRD4780 mitigated excessive activation of the UPR, as seen by the reduction of UPR chaperones belonging to the ATF6 branch, such as Grp94. BRD4780 is therefore a promising candidate for targeting ER entrapment of misfolded proteins involved in Fabry disease and other proteinopathies.

The drug 4-phenylbutyrate (4-PBA) was shown to promote protein folding and to target protein trafficking components responsible for aberrant retention of misfolded proteins in the ER (shown in Fig. 2). The exact mechanism of action is not fully understood, but 4-PBA has shown promise in many diseases associated with protein misfolding, including cystic fibrosis, ADTKD-UMOD, and neurodegenerative diseases [97]. In the context of 4-PBA’s effect on trafficking machineries, 4-PBA was shown to compete with p24 cargo receptors for binding to COPII vesicles. COPII selectively sorts and transports properly folded protein cargos out of the ER. The binding of 4-PBA to COPII allows for less stringent cargo uptake, including proteins that are improperly folded, allowing their incorporation into COPII vesicles for export from the ER [98]. This 4-PBA-induced forward trafficking likely relieves ER stress by directing misfolded proteins out of the ER and into other cellular compartments, or by directing them for degradation.

The selected therapeutic agents described here modulate ER stress response by tackling distinct aspects of the UPR machinery. The selected agents represent diverse approaches to proteostasis regulation and associated pathologies (shown in Fig. 2).

Conclusion

ER stress induction and UPR activation is a prominent aspect of many proteinopathies. While adaptive UPR aims to mitigate stress outcomes, chronic and severe stress drives a terminal UPR that may lead to cellular apoptosis and tissue damage. The molecular complexity of this pathway and the range of potential outcomes demand a careful investigation of the specific involvement of this pathway and the most appropriate therapeutic strategy that should be applied in each pathology (shown in Fig. 2). Moreover, environmental/nongenetic contributions along with distinct genetic background and diverse mislocalization patterns differentially activate ER stress and induce the UPR. Considering the intricacy of the UPR process and the complex interplay between genetics and environment, there is a need to develop appropriate biomarkers aimed at evaluating the contribution of the UPR in pathological developments. Understanding the association between these factors is crucial for the development of effective therapeutic approaches.

Notably, in pathologies like Fabry disease, the currently approved treatments mainly address protein function deficiency. Yet emerging evidence suggests a toxic induction of ER stress and activation of a terminal UPR. These gain-of-function aspects should be carefully considered to create optimal treatment strategies. Such pathologies may require a combination of therapeutics, in which both protein loss of function and gain of function are addressed. Future studies are needed in order to elucidate the interaction frameworks between these two disease mechanisms.

Acknowledgments

We apologize to any authors whose important contributions to the subject were not cited in this piece due to space limitations.

Conflict of Interest Statement

The authors have no conflicts of interest to disclose.

Funding Sources

This study is supported by the Zuckerman STEM Leadership Program, by the Israel Science Foundation (1301/22), and by the David Fela Shapell Family Center for Genetic Disorders Research.

Author Contributions

Writing – reviewing and editing: D.M., N.A., and M.D.-L.

Funding Statement

This study is supported by the Zuckerman STEM Leadership Program, by the Israel Science Foundation (1301/22), and by the David Fela Shapell Family Center for Genetic Disorders Research.