Protein-rich foods, sea foods, and gut microbiota amplify immune responses in chronic diseases and cancers – Targeting PERK as a novel therapeutic strategy for chronic inflammatory diseases, neurodegenerative disorders, and cancer

Abstract

The endoplasmic reticulum (ER) is a cellular organelle that is physiologically responsible for protein folding, calcium homeostasis, and lipid biosynthesis. Pathological stimuli such as oxidative stress, ischemia, disruptions in calcium homeostasis, and increased production of normal and/or folding-defective proteins all contribute to the accumulation of misfolded proteins in the ER, causing ER stress. The adaptive response to ER stress is the activation of unfolded protein response (UPR), which affect a wide variety of cellular functions to maintain ER homeostasis or lead to apoptosis. Three different ER transmembrane sensors, including PKR-like ER kinase (PERK), activating transcription factor 6 (ATF6), and inositol-requiring enzyme-1 (IRE1), are responsible for initiating UPR. The UPR involves a variety of signal transduction pathways that reduce unfolded protein accumulation by boosting ER-resident chaperones, limiting protein translation, and accelerating unfolded protein degradation. ER is now acknowledged as a critical organelle in sensing dangers and determining cell life and death. On the other hand, UPR plays a critical role in the development and progression of several diseases such as cardiovascular diseases (CVD), metabolic disorders, chronic kidney diseases, neurological disorders, and cancer. Here, we critically analyze the most current knowledge of the master regulatory roles of ER stress particularly the PERK pathway as a conditional danger receptor, an organelle crosstalk regulator, and a regulator of protein translation. We highlighted that PERK is not only ER stress regulator by sensing UPR and ER stress but also a frontier sensor and direct senses for gut microbiota-generated metabolites. Our work also further highlighted the function of PERK as a central hub that leads to metabolic reprogramming and epigenetic modification which further enhanced inflammatory response and promoted trained immunity. Moreover, we highlighted the contribution of ER stress and PERK in the pathogenesis of several diseases such as cancer, CVD, kidney diseases, and neurodegenerative disorders. Finally, we discuss the therapeutic target of ER stress and PERK for cancer treatment and the potential novel therapeutic targets for CVD, metabolic disorders, and neurodegenerative disorders. Inhibition of ER stress, by the development of small molecules that target the PERK and UPR, represents a promising therapeutic strategy.

Introduction

The endoplasmic reticulum (ER) is a large vital intracellular organelle [1, 2] that contains the nuclear envelope, ER cisternae, and an interconnected tubular network [3]. ER has a folding enzymes, multifunctional integral membrane and luminal chaperones, as well as sensor molecules such as glucose-related protein 78 kDa (GRP78), also known as binding immunoglobulin protein (BiP), glucose-related protein 94 kDa (GRP94), inositol-1,4,5-trisphosphate (InsP3) receptor/calcium release channel, protein disulfide isomerases (PDI), lectins such as calreticulin and calnexin, sarco-ER calcium-ATPase (SERCA) pump, ryanodine receptors (RYRs), and ER-associated calcium sensors stromal interacting molecule (STIM) proteins [4]. ER provides an optimal environment for fundamental biological processes such as protein synthesis, protein folding, protein modification, ER retention, trafficking, and canonical secretome (secretory proteins with signal peptide) [5–7], as well as lipid metabolism and steroid synthesis [8], and calcium homeostasis and storage, making it an essential organelle for cell survival [9]. It has been reported that 11% of the 25,000 predicted human full-length open reading frames (ORFs) are secreted proteins, and 20% of single or multi-pass transmembrane proteins are trafficked into the ER lumen by their N-terminal signal sequences [10]. When a newly synthesized protein enters the ER, it undergoes a series of modifications and encounters a variety of molecular chaperones and folding enzymes, all of which aid in its proper folding and eventual release from the ER. ER dysfunction and the generation of large amounts of substances such as reactive oxygen species (ROS) [11] and calcium ions under stress conditions and disease risk factors/danger-associated molecular patterns (DAMPs) stimulation lead to reduced protein folding and the accumulation of unfolded or misfolded proteins in the ER lumen, which causes ER stress [12]. If the ER stress is mild and the inducers are temporary, the excess ER membranes and proteins are removed by autophagy in a process of ER remodeling and selective recycling called Reticulophagy to reduce ER stress [13]. However, the accumulation of unfolded proteins represents a key barrier in the secretory apparatus and is detrimental to cellular and organismal homeostasis. Thus, an evolutionarily conserved cellular checkpoint mechanism called the unfolding protein response serves to sense and promote adaptation or cellular execution in response to unfolding proteins [14].

The unfolded protein response (UPR) is a cellular stress response triggered by the accumulation of unfolded proteins in the ER lumen in eukaryotic cells. The UPR is an adaptive signaling pathway which is highly controlled and integrates information about the type, intensity, and duration of the stress stimuli, therefore determines the fate of cells. Three transmembrane protein sensors regulate the UPR, including protein kinase R-like endoplasmic reticulum kinase (PERK) [15], inositol-requiring kinase 1 (IRE1) [16], and activating transcription factor 6 (ATF6) [17], which, in response to the accumulation of unfolded and misfolded proteins, determine cellular fate. In unstressed cells, these sensors are associated with GRP78/BiP, which consistently and dynamically binds and arrests those proteins in an inactive status in the ER membrane [18]. However, in response to ER stress, those ER resident sensors are dissociated from BiP, undergo conformational changes, and activate the UPR and downstream signaling (Figure 1).

Figure 1. ER stress regulators IRE1 and ATF6 signaling are mediated by mRNA processing and proteolytic processing, respectively, whereas PERK signaling is mediated by translational and posttranslational mechanisms.

Under normal conditions, the three proteins (sensors), including inositol-requiring enzyme 1 (IRE1), activating transcription factor 6 (ATF6), and protein kinase RNA (PKR) like ER kinase (PERK), bind to the molecular chaperone protein glucose-related protein 78 kDa (GRP78)/binding immunoglobulin protein (BiP) and inhibit their activation. Under stress conditions, GRP78/BiP dissociates from the three ER stress sensors and binds to unfolded proteins, leading to the activation of the three sensors. Each activation pathway has a different signal transduction mechanism. (A) IRE1 pathway: IRE1α splices X-box binding protein 1 (XBP1) mRNA to encode for the transcription factor XBP1s, which promotes the expression of genes involved in protein folding and induces chaperones, endoplasmic reticulum-associated protein degradation (ERAD), and lipid synthesis. (B) ATF6 pathway: under ER stress, ATF6 translocates from the ER to the Golgi apparatus. ATF6 is then cleaved to produce the amino terminus of ATF6 (ATF6-N), which migrates to the nucleus to increase the transcription activity of XBP1 and regulate ERAD and lipid metabolism. (C) PERK pathway: PERK (eukaryotic translation initiation factor 2 alpha kinase 3, EIF2AK3) undergoes oligomerization and auto-phosphorylation, which then promotes the phosphorylation of the eukaryotic initiation factor 2 alpha subunit (eIF2α), resulting in a general inhibition of protein translation. However, phosphorylated eIF2α selectively increases the translation of activating transcription factor 4 (ATF4), which upregulates CHOP and GADD34 mRNA. Created with BioRender.com

The main objective of the UPR is to maintain cellular homeostasis and restore ER function [19]. However, if ER stress is persistent or cellular homeostasis cannot be reestablished, the UPR promotes cell death [20]. The UPR can be induced by pathophysiological conditions such as viral infection, ischemia, hyperhomocysteinemia, hypoxia, hypoglycemia, proteotoxicity, excessive oxidative stress, and changes in membrane lipid composition [21–25] or experimentally by agents that decrease the folding capacity of the ER (e.g., tunicamycin, which inhibits glycosylation of asparagine residues, and thapsigargin, which inhibits SERCA (sarco/ER-Ca²⁺-ATPase), thereby reducing ER calcium)

(i). IRE1 pathway.

IRE1 is a type I transmembrane protein, identical to PERK. It has serine/threonine kinase and endoribonuclease (RNase) activities in the cytosol. Under non-stressful circumstances, the IRE1α cytosolic domain is bound to heat shock proteins 90 (HSP90) and 72 (HSP72). After ER stress, unfolded proteins bind to BiP and release IRE1α, which is autophosphorylated and rapidly activated. The activated IRE1α activates and removes the non-coding sequence of the X box-binding protein 1 (XBP1) gene [26], which regulates protein folding and trafficking, ER-associated degradation (ERAD), autophagy, and lipid biosynthesis. ERAD is a quality control system that regulates the degradation of unfolded and misfolded proteins and targets them for clearance [27]. It is also responsible for the regulation of the endogenous levels of physiologically important proteins such as 3-hydroxy-3-methylglutaryl coenzyme-A (HMG-CoA) reductase, a rate-limiting enzyme for cholesterol biosynthesis [28]. In addition, when the IRE1 signal is activated, the tumor necrosis factor receptor-associated factor 2 (TRAF2) and IRE1 combine to form a complex and activate the apoptosis signal-regulatory kinase 1 (ASK1), the c-Jun N-terminal kinase (JNK) [29], and the nuclear factor kappa B (NF-κB) [30, 31], which trigger cell death [9].

(ii). ATF6 pathway.

The leucine zipper protein transcription factor, ATF6, is a type-II transmembrane receptor with its C-terminal domain located in the ER lumen and its N-terminal DNA-binding domain located in the cytoplasm [32]. under non-stressful conditions, ATF6 is constitutively localized in the ER and binds to BiP. Under conditions of ER stress, ATF6 separates from BiP and moves into the Golgi apparatus. There, it’s cleaved by site-1 protease (S1P) and site-2 protease (S2P), releasing functional ATF6 fragments into the cytoplasm and transferring them into the nucleus to start transcription [33]. In the nucleus, the cleaved ATF6 induces the expression of genes involved in protein folding, components of ERAD, protein secretion, and lipid biogenesis. Furthermore, ATF6 induces the expression of BiP, XBP1, and CCAAT-enhancer binding protein (C/EBP) homologous protein (CHOP) [34].

(iii). PERK pathway.

PERK (1116 amino acids, https://www.ncbi.nlm.nih.gov/protein/NP_004827.4), also referred to as eukaryotic translation initiation factor-2α kinase 3 (eIF2AK3) [35], is a major sensor ER-resident type-I transmembrane protein that is a member of serine/threonine kinase family. Its N-terminal luminal domain (LD) is responsive to the upstream ER stress signal, whereas its C-terminal cytoplasmic domain directly phosphorylates eukaryotic translation initiation factor 2 alpha (eIF2α) and linked to cytosolic kinases (kinomes) [15]. Six PERK structural domains include the N-terminal signal peptide (amino acids (aa) 1–29), N-terminal ER luminal eIF2AK3 domain (aa 104–420), transmembrane domain (aa 515–535), cytosolic PKC-like domain (aa 587–663), cytosolic serine-threonine kinase domain (STKc-EIF2AK3-PERK, aa873–1075), and four cytosolic phosphorylation sites (aa 619, 715, 982, and 1094) (https://www.ncbi.nlm.nih.gov/protein/NP_004827.4). Under normal conditions, PERK binds to BiP to exist as an inactive monomer. PERK is activated as a homodimer under pathological conditions such as hypoxia and oxidative stress by separating from BiP, which enables PERK oligomerization and activation. Phosphorylated and activated PERK phosphorylates the downstream mediators (substrates), including eIF2α and nuclear factor erythroid-derived 2-like 2 (Nrf2), impairing global protein translation [35, 36]. Meanwhile, phosphorylated eIF2α induces the selective translation of activating transcription factor 4 (ATF4) and further activates the transcription of downstream UPR target genes [37, 38]. ATF4 has dual functions: under mild to moderate ER stress, it is responsible for the restoration of cellular homeostasis, but under severe and prolonged ER stress conditions, it induces the activation of CHOP-mediated apoptosis [39–41].

Recent studies have identified new PERK substrates, including lipids that signal the second messenger diacylglycerol (DAG) [42] and protein substrates such as the forkhead box O protein (FOXO) [43]. We recently reported that PERK activates 12 more kinases in the cytosol, including cAMP response element-binding protein kinase (CREB), p38 mitogen-activated protein kinase (MAPK) α (P38αMAPK), mammalian target of rapamycin (mTOR), extracellular signal-regulated kinase (ERK1/2MAPK), c-Jun N-terminal kinase (JNK1/2/3MAPK), protein kinase B (AKT1/2/3), protein kinase AMP-activated catalytic subunit alpha 1 (AMPKα1, PRKAA1), AMPKα2 (PRKAA2), glycogen synthase-3 (GSK-3α/β), mitogen- and stress-activated kinases (MSK1/2), platelet-derived growth factor receptor β (PDGFRβ), and epidermal growth factor receptor (EGFR) [15]. The control of the UPR and its activity under ER stress is greatly dependent on the PERK protein pathway, and loss of PERK activity severely affects the ability of cells to withstand ER stress. Because of its many branching pathways, PERK has been implicated in the development of several diseases [39]. PERK has 18 disease-related pathways, including many discussed in this review (https://www.ncbi.nlm.nih.gov/gene/9451). The significance of the PERK pathway has also been highlighted by findings from human genetic diseases. For example, Wolcott-Rallison syndrome has been identified as a PERK mutation (https://databases.lovd.nl/shared/genes/EIF2AK3), which is an uncommon autosomal recessive disease that causes permanent neonatal or early infancy insulin-dependent diabetes (https://omim.org/entry/226980). Additional human phenotype ontology includes abnormalities of body height, kidney, urinary system, and renal insufficiency (https://www.genecards.org/cgi-bin/carddisp.pl?gene=EIF2AK3&keywords=EIF2AK3). Moreover, the results from gene-deficient mice have also significantly enhanced our understanding. PERK knockout cells are hypersensitive to the lethal effects of ER stress-inducing toxins such as thapsigargin (a sarcoplasmic and ER Ca²⁺-ATPase (SERCA) inhibitor, leading to ER calcium depletion and increased cytosolic calcium concentrations) and tunicamycin (a protein glycosylation inhibitor), which cause ER stress by reducing the protein folding capacity of the ER [44]. In addition to serving as a secondary (indirect) DAMP sensor for sensing ER stress induced by endogenous DAMPs such as ROS [11] and calcium ions under stress conditions, PERK can also become the primary (or frontier) DAMP sensor to directly sense exogenous pathogen-associated molecular patterns (PAMPs). Recently, Dr. Biddinger’s team, our team, and others reported that the gut microbiota-generated uremic toxin trimethylamine-N-oxide (TMAO) binds to the N-terminal luminal domain (LD) of PERK and activates PERK [15, 45–48].

There are several studies that indicate that the ER stress-related signaling pathways could serve as a promising target for developing innovative therapeutic approaches [49]. These approaches may help to decrease the adaptation of cancer cells to hypoxia, inflammation, and angiogenesis, thereby overcoming adaptation-related drug resistance [50]. The PERK-mediated UPR signaling pathways have a crucial role in promoting adaptation and survival during ER stress conditions resulting from hypoxia. Therefore, it should be the future target of anticancer therapeutic intervention [51]. In summary, the role of ER stress, particularly the PERK pathway, in several diseases, including CVDs, neurodegenerative diseases, and cancers, has been reviewed previously; however, the role of a protein-rich diet, a seafood diet, and gut microbiota-generated uremic toxins TMAO in PERK activation was not highlighted in detail in the previous reviews. In addition, how PERK activation affects the innate immune response and further exacerbates the inflammatory response was not discussed in previous publications. Therefore, in this review, we summarize the functions of the ER stress sensors, particularly PERK [52] as a conditional DAMP [53, 54] receptor for gut microbiota-generated uremic toxin TMAO [47], and focus on its role in the regulation of protein translation and organelle crosstalk regulation [1]. We highlight that PERK is not only an ER stress regulator by sensing UPR and ER stress but also a frontier sensor that directly senses gut microbiota-generated metabolites. We further highlight the function of PERK as a central hub that leads to metabolic reprogramming and epigenetic modification [55], which further enhance the inflammatory response [56, 57] and promote trained immunity (innate immune memory) [58–63], innate immunity [64, 65], cell death [66–68], reticulophagy (ER autophagy) [69, 70], and fibrogenesis [71, 72]. Moreover, we highlight the contribution of ER stress and PERK in the pathogenesis of several diseases, including cardiovascular diseases [72–74], inflammations [52], chronic kidney disease [75], Alzheimer’s disease and neurodegenerative diseases [76–78], type II diabetes [79], and cancers [8, 80–85]. Finally, we discuss the therapeutic targets of ER stress and PERK for cancer treatment and potential novel therapeutic targets for cardiovascular disease, metabolic disorders, chronic kidney diseases, and neurodegenerative disorders. Inhibition of ER stress/UPR activation and the PERK pathway by the development of small molecules that target the PERK and UPR represents a promising therapeutic strategy.

ER stress sensor PERK is a conditional danger receptor for the microbiota-generated uremic toxin TMAO.

Pathogen-associated molecular patterns (PAMPs) and DAMPs are molecules associated with groups of pathogens [86] and elevated endogenous metabolites [48, 53] and have an exceptionally high potential for triggering inflammatory responses [86–88]. Several types of molecules can serve as PAMPs, including lipopolysaccharides (LPS), endotoxins found on the cell membranes of gram-negative bacteria [89], which are considered to be the prototypical class of PAMPs and specifically recognized by toll-like receptor-4 (TLR4), a receptor of the innate immune system [60, 61, 86, 90]. DAMPs, also known as alarmins, are endogenous danger signals released by stressed cells that induce and amplify the immunological and inflammatory response [91]. These DAMPs include heat-shock proteins, high-mobility group box 1 (HMGB1) [5], and reactive oxygen intermediates [92]. DAMPs can be recognized by classical DAMP receptors, including Toll-like receptors (TLRs) and other pattern-recognition receptors. Nucleotide-binding oligomerization domain (NOD)-like receptors (NLRs) play essential roles in innate immunity by detecting intracellular PAMPs/DAMPs [87, 93–95]. Other DAMP receptors have been identified, including the receptor for advanced glycation end products (RAGE), a receptor for HMGB1, a retinoid acid-inducible gene I (RIG-I), transmembrane C-type lectin receptors, and the absent in melanoma 2 (AIM2) [96, 97]. Our team previously reported that the uremic toxins serve as conditional DAMPs and homeostasis-associated molecular patterns (HAMPs) that modulate inflammation [48, 54, 98]. In addition, we identified lysophospholipid receptors, G-protein-coupled receptors (GPCRs), as novel conditional DAMP receptors for the endogenous metabolite lysophospholipid [53, 99]. We also proposed that metabolism-associated danger signal (MADS) recognition is a novel mechanism for metabolic risk factor-induced inflammatory responses [100] and that the homocysteine-methionine (HM) cycle is a key metabolic sensor system controlling methylation-regulated pathological signaling [101]. Furthermore, we proposed the new concept that ROS systems create an integrated metabolic sensor network for sensing metabolic stresses and metabolic homeostasis [11, 102].

ER stress responses result in upregulation and ligand-independent activation of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) receptors 1 and 2 (also known as DR4 and DR5), leading to a FAS-associated death domain protein (FADD)/caspase-8/receptor-interacting serine/threonine protein kinase 1 (RIPK1) (FADDosome)-dependent pathway to NF-κB activation and inflammatory cytokine production, implying that TRAIL receptors behave as “stress-associated molecular patterns (SAMPs)” that link ER stress to NF-κB-dependent inflammation [103]. Interestingly, we and others recently reported that gut microbiota-generated uremic toxin TMAO is a conditional DAMP that binds to and activates its intrinsic receptor PERK rather than commonly shared DAMP receptors to initiate pathological signaling [15, 47], which is the first DAMP receptor localized in the ER [97]. TMAO is significantly increased in pathological conditions such as end-stage renal disease (ESRD), and induced endothelial cell dysfunction, and inflammation [104, 105]. Protein-rich, sea food, and gut microbiota-generated metabolites [46] has been linked to ER stress response [106]. In conclusion, conditional DAMPs, including lysophospholipids and uremic toxins, have several important features. They are (a) endogenous metabolites; (b) pathologically elevated; (c) have a physiological signaling function; and (d) bind to a unique receptor and amplify the signals rather than shared DAMP receptors such as TLRs and NLRs. As far as we are aware, our study was the first to demonstrate that PERK, as the sensor for UPR/ER response and mitochondrial UPR [107] and also as a sensor/receptor for gut microbiota-generated uremic toxin TMAO, serves as a significant driver for metabolic reprogramming and trained immunity (also termed innate immune memory) [15, 47, 60, 108].

ER stress sensors are organelle crosstalk regulators, enriched in MAMs, and interact with MAM-tethering proteins to regulate metabolite transfer and affect mitochondrial unfolded protein responses through inter-organelle signaling.

The ER has multiple contact sites with every membrane-bound organelle, including mitochondria, plasma membrane, Golgi, endosomes, and peroxisomes [1, 3]. Connections between the ER and other organelles are required for the integration of cellular signals originating from other cellular compartments. The most extensively studied structural connectivity is the junction between ER-mitochondria and the ER-plasma membrane. Mitochondria are highly sensitive to a variety of cellular insults, including unfolded proteins in the ER, calcium, lipids, and ROS [109–115], and the exchange of these metabolites from the ER to the mitochondria causes mitochondrial dysfunction, accumulation of mitochondrial unfolded protein response (UPR^mt) [116] and determines ultimate cell fate [107]. The sites of ER that are in contact with mitochondria are known as the mitochondria-associated ER membranes (MAM). Recently, MAMs have received much attention due to their functional significance in organelle communication at membrane contact sites. It is tightly tethered by different multifunctional proteins that integrally regulate lipid metabolism, calcium signaling, and ROS homeostasis [117, 118].

Several studies have shown that different mitochondrial proteins involved in tethering can regulate the ER stress and UPR pathways (Figure 2), including mitochondrial fusion protein mitofusin 2 (MFN2), inositol triphosphate receptor (IP3R), voltage-dependent anion-selective channel protein 1 (VDAC1), GRP75, vesicle-associated membrane protein-associated protein B (VAPB), protein tyrosine phosphatase interacting protein 51 (PTPIP51), mitochondrial fission 1 (FIS1), and B-cell receptor-associated protein 31 (BAP31) [119]. At the contact sites, calcium is released through the ER channels to the mitochondrial membrane, and this requires VDAC1 interaction with IP3R [3]. Other mediators that facilitate the interactions at the contact sites include the ER-mitochondria encounter structures (ERMES) complex identified in yeast and MFN2 in mammalian cells [120], which regulate the shape of ER-controlling ER-mitochondrial contacts. In addition to the accumulation of unfolded proteins, metabolite imbalance can activate PERK in the lumen of the ER, and sustained activation of PERK leads to mitochondrial dysfunction [35, 121]. Furthermore, PERK kinases communicate with mitochondria to regulate oxidative stress at the MAMs [1, 122]. As the second tier of ER-mitochondrial tethering, PERK is MAM-rich components that interact with multifunctional MAM tethering proteins (the first tier) and integrally regulate the exchange of lipids, ROS, and metabolites such as calcium at the contact sites [123]. To maintain ER-mitochondrial connections, PERK interacts with MFN2 [122]. This structural effect on MAMs appears to require the cytosolic kinase domain of PERK, as it does not occur because of PERK downstream signaling of eIF2 [122]. A new report showed that PERK recruits E-Syt1 at ER-mitochondrial contacts for mitochondrial lipid transport, respiration [124], and mitochondrial membrane biogenesis [125]. In addition to being a tethering protein, MFN2 regulates PERK functions involved in mitochondrial fusion, inflammation, and the effective uptake of calcium at the ER-mitochondrial contact sites [126–128]. Thus, PERK functions as a crucial UPR sensor to mediate the crosstalk between the ER and mitochondria [129]. MAM dysfunctions may be involved in the pathophysiology of a variety of diseases, including metabolic and neurodegenerative disorders [130, 131].

Figure 2. ER-mitochondria and ER-plasma membrane contact sites derive their close apposition from tethering proteins and mediate inter-organelle signaling pathways.

Several pairs of integral membrane proteins located on the mitochondria, ER, and plasma membrane that are important for the physical tethering of the organelles were identified, reflecting and supporting the diverse functions of the MAMs and PAMs. (1) IP3R-GRP75-VDAC1 ternary binding complex: IP3R at the ER interacts with the mitochondrial channel VDAC1 owing to the chaperone GRP75, which interacts directly with the two proteins. At the ER, SIGMAR1 is chaperoning IP3R. This complex is essential for the efficient calcium transfer from the ER to the mitochondrial matrix through the mitochondrial Ca²⁺ uniporter (MCU) in the inner mitochondrial membrane. (2) VAPB-PTPIP51 tether: VAPB located at the ER membrane interacts with PTPIP51 located at the mitochondrial outer membrane. (3) BAP31-FIS1 tether: ER-resident protein BAP31 interacts with the mitochondrial fission protein FIS1. (4) MFN1/2 tether: in addition to its predominant location at the outer mitochondrial membrane, MFN2 can be found in the ER and interact with mitochondrial MFN-1/2. (5) PERK can also interact with mitochondrial MFN2. ER-PM contact sites have emerged as key regulators of intracellular calcium dynamics. Stromal-interacting molecule 1 (STIM1) molecule mainly exists as a dimer on the ER membrane that regulates the formation of ER-PM contact sites in response to calcium depletion in the ER lumen. Calcium depletion from the ER store induces dimerization or oligomerization of STIM1. The STIM1 dimer in the active state may directly couple to and activate the Orai1 channel on the PM, followed by replenishing calcium levels in the ER lumen. Created with BioRender.com

In addition, the contact sites between the ER and plasma membrane play a significant role in calcium exchange. The stromal-interacting molecule (STIM) proteins can sense a reduction in calcium within the ER. This leads to a conformational change in the STIM proteins, which then relocate tubular structures within the ER towards the ER-plasma membrane junctions. Ultimately, this results in the direct activation of the calcium release-activated calcium modulator (Orai), the pore-forming component of the calcium release-activated calcium (CRAC) channel. This activation initiates the channel opening and the subsequent calcium influx [132]. Furthermore, ER-plasma membrane contact sites play a crucial role in the metabolism of phosphatidylinositol, particularly in the regulation of phosphatidylinositol 4-phosphate (PI4P). This process is regulated by the oxysterol-binding homology (Osh) protein family and the ER membrane integral proteins, such as vesicle-associated membrane protein-associated protein (VAP), following Sac1-like phosphatidylinositide phosphatase (SACM1L) activation [3]. Interestingly, PERK has been shown to interact with filamin-A and F-actin remodeling to coordinate the establishment of ER-plasma membrane contact sites [133].

Furthermore, recent studies have identified the role of IRE1 in ER-mitochondrial communication by controlling ER-mitochondrial dynamics, mitochondrial calcium uptake, and cell survivability. IRE1 acts as a scaffold to stabilize IP3R at MAMS [134]. All these indicate a complex interrelation of IRE1 and PERK, two membrane-bound ER stress sensors, in controlling ER-mitochondrial communication, which suggests that PERK functions via ER-mitochondrial tethering are a new branch of PERK pathways in addition to cytosol kinase activation and transcription activation via ATF4. In addition, ER is not only a network hub that physically connects membrane organelles but also plays a fundamental role in the regulation of membraneless organelles [135].

ER stress pathways may facilitate the dynamics and formation of cytoplasmic and nuclear membraneless organelles.

Membraneless organelles are relatively new cellular structures that are not surrounded by a sealed phospholipid membrane and are formed through phase separation in either a liquid-liquid or liquid-solid manner [136]. Liquid-liquid phase separation can result in the production of stable liquid droplets within another liquid; however, liquid-solid phase separation can result in gel-like stress granules and solid and crystalline structures [137]. Membraneless organelles can be found in both the nucleus, such as the nucleolus, Cajal bodies (sub-nuclear structures often associated with the nucleolus), nuclear stress bodies, nuclear speckles, interchromatin granule clusters, paraspeckles, and transcription histone locus bodies, and in the cytoplasm, such as the centrosome, P-granules, pyrenoid, and processing-bodies (P-bodies) [138, 139] (Figure 3). Most constitutive membraneless organelles are regulated according to the cell cycle phase [140]. Membraneless organelles have crucial roles in cell physiology, but when they develop as a result of the production of mutant proteins, they might develop pathogenic characteristics such as amyloids [141]. Cellular stress leads to the formation of reversible membraneless organelles that are known as stress assemblies, among which the best studied are the P-bodies and stress granules [137, 142]. Environmental stressors trigger cellular signaling, resulting in the formation of stress granules. Stress granules can be formed in the nucleus or in the cytoplasm [143]. ER stress inhibits mRNA translation initiation and polysome disassembly, leading to the accumulation of untranslated 80S ribosome-free mRNA in the cytoplasm, which binds to RNA-binding proteins [144] and combines into membraneless foci called reversible stress granules [145]. It contains polyadenylated mRNAs, eukaryotic translation initiation factors such as eIF2A and eIF3, 40S ribosomes, and RNA-binding proteins such as Tia1 cytotoxic granule-associated RNA-binding protein and G3BP1 stress granule assembly factor 1, which are the main drivers for stress granule formation.

Figure 3. ER stress pathways may facilitate the dynamics and formation of cytoplasmic and nuclear membraneless organelles.

The membrane-bound organelles in the cells, including mitochondria, Golgi, nucleus, lysosomes, and endosomes, are labeled in green. A growing number of membraneless organelles have been identified in the cytoplasm and nucleus. The cytosolic membraneless organelles such as P-bodies, U bodies, P-granules, centrosomes, and pyrenoid and the nuclear membraneless organelles such as the nucleolus, nuclear speckles, paraspeckles, cajal bodies, and promyelocytic leukemia (PML) bodies are labeled in red. Under cellular stress, the formation of stress granules and enlarged P-bodies is induced. Created with BioRender.com

The heterogeneity in the development and composition of the stress granule pathway is depending on the cell type, specific stressors, and the signaling pathways that are activated. Stress granule formation can triggered by ER stress, oxidative stress, heat shock, hypoxia, malnutrition, the presence of translation-blocking medications, viral infection, knockdown of particular translation initiation factors, and overexpression of particular RNA-binding proteins [146]. There are two types of stress granules: canonical and non-canonical stress granules (Figure 4) [142]. The canonical stress granules can be triggered by a variety of cell-damaging stimuli, including ER stress (heat chock and thapsigargin) [147], oxidative stress (sodium arsenite) [148], nutrient deprivation, and proteasome inhibition (MG132) [149]. Particular stress conditions are detected by specific kinases, including heme-regulated initiation factor 2α kinase (HRI, or eIF2α kinase 1 (eIF2αK1)), protein kinase RNA-activated (PKR, eIF2αK2), PERK (EIF2AK3), and general control non-derepressible 2 (GCN2, or eIF2α kinase 4 (eIF2αK4)) [150], which then become activated and phosphorylate eIF2α. The phosphorylation of eIF2α resulted in the accumulation of several types of RNA-binding proteins, such as poly(A) binding protein cytoplasmic 1 (PAB1), ras GTPase-activating protein-binding protein 1 (G3BP1), Caprin family member 2, tristetraprolin (TTP), T-cell intracellular antigen-1 (TIA-1), fragile X messenger ribonucleoprotein 1 (FMR1), AU-rich element binding protein (HuR), TAR DNA binding protein (TDP-43), as well as polyadenylated mRNA, and specific RNAs, including long RNA, arginines and glycines (RG) rich motifs (induced by heat shock) [151], and adenylate/uridylate (AU)–rich motifs (induced by ER stress) [152]. A non-canonical stress granule pathway is independent of eIF2α phosphorylation [153]. It is dependent on how pateamine A (PatA) and hippuristanol (Hipp) inhibits eIF4A activity (an RNA helicase needed for the ribosome recruitment phase of translation initiation). This inhibits ribosome activity and stops translation initiation by preventing the formation of the 48S initiation complex [153]. It also formed upon exposure to ultraviolet (UV) and osmotic chock. This type of stress granule contains RNA-binding proteins and non-polyadenylated mRNAs. In addition to their active role in the stress response, stress granules participate in mRNA triage and expression regulation, cell signaling and apoptosis, and inhibition of viral replication. Taken together, this indicates that the induction of ER stress and PERK kinase pathway activation plays a significant role in inducing canonical stress granule formation.

Figure 4. Membrane organelles and membraneless organelles are both integrated to mediate cellular responses to stress. Stress granules are multimolecular cytoplasmic foci that assemble as part of the cellular response to stress in two different pathways.

Canonical stress granules are induced by activation and phosphorylation of eukaryotic initiation factor 2 alpha (eIF2α) by four kinases, including general control non-derepressible-2 (GCN2, or eIF2α kinase 4 (EIF2AK4)); pancreatic eIF2α kinase (PKR-like ER kinase (PERK) or eIF2α kinase 3 (EIF2AK3)); protein kinase R (PKR), and haem-regulated inhibitor (HRI); or eIF2α kinase 1 (EIF2AK1). Each kinase contains specific regulatory regions that detect and recognize different stress stimuli, including viral infection, oxidative stress, ER stress, starvation, and arsenite. It is mainly comprised of eukaryotic translation initiation factor 2 subunit alpha (eIF2α), polyadenylated (poly-A) mRNA molecules, long RNAs, AU and RG motifs, and a vast array of proteins, including RNA-binding proteins (RBPs) such as ras GTPase-activating protein-binding protein 1 (G3BP1), T-cell intracellular antigen-1 (TIA-1) and TIA1-related protein (TIAR), polyadenylate-binding protein (PABP), AU-rich element binding protein (HuR), ataxin-2, fragile X mental retardation protein (FMRP), and tristetraprolin (TTP). The non-canonical stress granules are dependent on the inhibition of the eukaryotic initiation factor 4 alpha (eIF4α) by pateamine A (PatA) and hippuristanol (Hipp), which prevents the formation of the 48S initiation complex, inhibits ribosome activity, and arrests translation initiation. These events lead to SG formation independently of eIF2α phosphorylation. Created with BioRender.com.

Regulation of protein translation in the ER stress response.

Protein synthesis is an essential process in cells and responds to dynamic cellular and environmental changes. Translational changes occur more quickly and are better able to adapt to changing environments than transcriptional changes [154]. The ER is the main site for folding and posttranslational modification of secreted and integral membrane proteins. Regulation of protein translation occurs mostly at the level of initiation, which is the rate-limiting step in protein synthesis. Protein translation depends on the formation of the 43S preinitiation complex, a 40S ribosomal subunit, and the translation initiation factors, including eukaryotic translation initiation factor 1 (eIF1), eIF1A, and eIF3. Two distinct pathways are defined in protein translational processes, including (i) the regulation of ternary complex formation by eIF2 phosphorylation at its conserved residue, serine 51 of its α subunit (eIF2α) and suppression of the polypeptide biosynthesis, and (ii) the regulation of the eIF4F complex via the mammalian target of rapamycin complex 1 (mTORC1) [155, 156]. One of the most important regulatory functions of PERK is its role as a regulator of protein translation. Once PERK is activated by ER stress, it phosphorylates eIF2α. eIF2α regulates the binding of the methionyl tRNA to the small ribosomal subunit. Phosphorylation of eIF2α prevents the recycling of eIF2α to its active GTP-bound form and inhibits the delivery of the initiator Met-tRNAi to the translation initiation complex, thereby reducing the rate of general protein translation (global translation inhibition) and affecting the translation of both cytoplasmic and ER client proteins [157]. The formation of stress granules is dependent on eIF2α phosphorylation under certain conditions [158]. Meanwhile, PERK activation and eIF2α phosphorylation result in the translational upregulation of selected proteins, including ATF4 (PERK downstream transcription factor), because of the upstream short open reading frame (uORF) [35, 159–161]. PERK-dependent reduction of protein translation restricts the trafficking of nascent proteins into the ER lumen, thereby reducing potential chaperone loading and allowing chaperones to clear misfolded aggregates. In response to ER stress, PERK knockout cells are hypersensitive to the lethal doses of toxins and lose their ability to regulate protein translation associated with hyperactivation of the IRE1 pathway, implying that these cells experience more ER stress [44] and that PERK acts upstream of the IRE1 pathway. PERK knockout animals develop progressive destruction of multiple cell types normally responsible for high protein secretion. The most notable destructions are the endocrine and exocrine cells of the pancreas and the collagen I-secreting osteoblasts, which lead to diabetes mellitus and severe bone defects [162, 163].

ER stress sensor PERK activation promotes trained immunity, an implication for disease interaction.

Trained immunity is quit novel concept. DAMP receptors signaling can modulate immunity. Innate immune cells develop an enhanced long-term immune response and inflammatory phenotype after brief exposure to endogenous or exogenous DAMPs. As a result, an inflammatory response is induced and greatly enhanced after exposure to the second challenge, and this phenomenon is known as “innate immune memory or trained immunity” [15, 58, 60–63, 90, 108, 164]. ER stress profoundly affects the innate and adaptive immune responses and is associated with autoimmune and inflammatory diseases, including diabetes mellitus and atherosclerosis [60, 61, 64, 165]. Upon activation, the ER stress response activates immune signaling pathways, including signal transducer and activator of transcription (STAT), c-Jun N-terminal kinases (JNKs), and NF-κB, to increase inflammatory cytokine release and affect the activation and function/dysfunction of the innate immune cells [64]. It has been demonstrated that trained immunity is triggered by the activation of classical DAMP receptors like TLR by TLR agonists. This process involves metabolic reprogramming and epigenetic alterations, which lead to a significant enhancement of antimicrobial functions [15, 58–60, 62, 90, 108]. During respiratory syncytial virus (RSV) infection, ER stress was induced and PERK was activated in dendritic cells, resulting in an altered innate immune response phenotype and promoting a pathogenic response via transcriptional regulation of key innate immune cytokines, enhancement of eIF2α-phosphorylation signaling, and induction of CD4⁺ T cell recruitment associated with interleukin-13 (IL-13) and IL-17 production [166, 167]. Furthermore, we recently reported that priming human aortic endothelial cells with β-glucan followed by TMAO binds to its conditional DAMP receptor, PERK, enhances endothelial cell activation, and increases tumor necrosis factor-α (TNF-α), a classical readout for trained immunity [15]. However, the inhibition of PERK activation resulted in reduced TMAO-enhanced trained immunity. Chronic kidney disease (CKD) is characterized by marked increases in proinflammatory cytokine levels, particularly TNF-α and IL-6 [168], which are inversely correlated with decreased glomerular filtration rate (GFR) [169, 170]. Monocytes and neutrophils from CKD patients showed an exaggerated response to LPS stimulation [171]. This may be due to the uremic environment that induces increased expression of TLR2 and 4 [168, 172, 173]. However, persistent activation of monocytes during CKD may also be a direct consequence of uremic toxins [6, 48]. Our latest findings indicate that the PERK pathway and ER stress are responsible for mediating the angiotensin II-accelerated innate immune response, as well as the differences in disease susceptibilities between the thoracic and abdominal aortas [174]. In monocytes, indoxyl sulfate binds to and activates its conditional DAMP receptor, aryl hydrocarbon receptors [175], thereby promoting the production of proinflammatory cytokines [176], which provoke endothelial damage and increase the risk of cardiovascular complications [177, 178].

ER stress sensors are critical regulators for metabolic reprogramming.

The energetic metabolism reprogramming by ER stress usually interferes with mitochondrial respiration to limit the production of adenosine triphosphate (ATP), increase the production of ROS, and switch mitochondrial energetics to glycolysis, otherwise called the “Warburg effect” if it happens under normoxia, which favors cell transformation and cancer development [179]. The protein disulfide isomerase family A member 2 (PDIA2), an ER stress protein, translocates into mitochondria and engages in interaction with components of the electron transport chain, inhibiting mitochondrial respiration and switching energy metabolism to hyperglycolysis in the tumor microenvironment [180]. Furthermore, new research has shown a direct link between PERK-mediated control on insulin resistance [47]. This work is an example of innovative regulation of metabolic circuits through the PERK-ATF4-phosphoserine aminotransferase 1 (PSAT1) axis. At physiologically relevant concentrations, the gut microbiota-generated TMAO binds to PERK and specifically stimulates the PERK-mediated activation of the transcription factor FoxO1, which is a major driver of metabolic diseases. TMAO is increased by insulin resistance and is linked to a number of human metabolic disorders. However, at pathological concentrations, TMAO binds to and activates PERK, which in turn promotes hyperglycemia. In addition, inhibition of flavin-containing monooxygenase 3 (FMO3), a TMAO synthesizing enzyme, reduces PERK activation and insulin resistance in vivo, suggesting a potential target for a new therapeutic intervention for metabolic syndrome. Furthermore, PERK signaling regulates mitochondrial autophagy in immune cells via mitochondrial reprogramming and epigenetic changes [181]. In addition to the promotion of trained immunity and endothelial cell activation, as we reported [15], the PERK arm of UPR signaling enhances macrophage metabolic functions and promotes an immunosuppressive M2 macrophage phenotype [182, 183]. However, PERK deficiency suppresses mitochondrial respiration (oxygen consumption rate), decreased ATP production [113], reduced extracellular acidification rate (ECAR) and the rate of glycolysis [166]. In addition, PERK deficiency significantly downregulates the expression of key genes involved in lipid metabolism, glutamine metabolism, and amino acid metabolism, including peroxisome proliferator-activated receptor γ (PPAR-γ) [184], lysosomal acid lipase (LIPA), and PPAR-γ coactivator-1β (PGC-1β), resulting in a significant decrease in lipid intake, lipolysis, and energy intake in M2 macrophages [185]. Furthermore, stimulation of bone marrow-derived macrophages (BMDMs) isolated from mice with IL-4 (M2 macrophages) increases PERK activation [186]. However, future work needs to determine how PERK-promoted M2 macrophage phenotype is correlated with its trained immunity promotion.

ER stress sensor signaling pathway promotes epigenetic modifications.

Epigenetic changes [55] have been associated with the progression of kidney diseases. During the progression of CKD, several uremic toxins accumulated can stimulate the production of ROS, resulting in epigenetic alterations and further promoting the progression of CKD [187]. We recently reported that ER stress and mitochondrial genes are co-upregulated in CKD environments [15] and that the uremic toxin TMAO binds to and activates the ER stress protein PERK [15, 47]. PERK signaling can regulate mitochondrial autophagy by inducing mitochondrial reprogramming and distinct epigenetic modifications in T cells [181]. To meet cellular energy needs, PERK signaling promotes mitochondrial respirations and signaling through ATF4, which controls PSAT1 activity, mediates the pathway for serine biosynthesis, and results in enhanced mitochondrial function and α-ketoglutarate production required for JMJD3-dependent histone demethylation in M2 macrophages. This finding indicates PERK-directed rewiring of metabolic reprogramming and epigenetic alteration through PERK-ATF4-PSAT1. Furthermore, PERK has been shown to modulate histone 3 lysine 27 (H3K27) methylation in macrophages [185, 186]. Histone H4 acetylation is increased by the induction of the ER chaperone GRP78, which binds to unfolded proteins and promotes the activity of ER stress transducers [188]. In aortic stenosis, the levels of histone deacetylase-6 (HDAC6) are significantly reduced, which is associated with ER stress-mediated osteogenesis and calcification in aortic valve tissue [189]. However, the pharmacological inhibition of HDAC6 significantly reduced apoptosis by inhibiting ER stress in tubular epithelial cells in an acute kidney injury mouse model [190].

ER stress sensor signaling pathways increase inflammation and promote tumor progression.

A crucial characteristic of tumors is the uncontrolled growth of the transformed cells. As a result of poor vascularization, tumors are constantly challenged with a restricted supply of nutrients and oxygen [191]. Tumor growth is linked to increased ER stress inducers such as glucose and oxygen restriction, as well as increased ROS production [192, 193], hence targeting ER stress and UPR provides a new therapeutic target for tumors [194]. UPR activation has been reported in a number of human cancers as well as in various cellular and animal cancer models [195]. In tumors, the upregulation of UPR markers is commonly observed, indicating the development of ER stress, and all UPR signaling branches contribute to tumor growth, angiogenesis, and immune evasion. The high baseline levels of UPR activation in tumor cells give a survival advantage, but it also maintains tumor cells on a tight threshold of survival-death transition. Tumor cells likely require optimum UPR machinery for survival, and thus suppressing oncogenesis by blocking the UPR response or boosting ER stress levels may be an effective way to suppress oncogenesis. Substantial evidence suggests that ER stress pathways and UPR signaling are involved in the activation of different classical inflammatory processes such as NF-κB activation and cytokine and chemokine production within and surrounding the tumor microenvironment [196, 197]. The ER stress response in myeloid cells can be transmitted cell-to-cell, promoting macrophage activation and triggering proinflammatory responses in the tumor microenvironment [198]. Activation of the UPR in tumor dendritic cells results in impaired immune function associated with impaired lipid metabolism and a reduction of T cell anti-tumor immunity. Recent studies have shown that tumor-infiltrating dendritic cells exhibit a high level of spliced XBP1, which promotes the progression of primary and metastatic ovarian cancer [199]. The IRE1α/XBP1 signaling pathway is implicated in the development and progression of various tumor types, including triple-negative breast cancer (TNBC) [200], prostate cancer, and pancreatic cancer [201]. These tumor-promoting effects appear to extend beyond increased protein folding capacity and involve immunomodulation [202]. Treatment of breast cancer cells with the IRE1 RNase inhibitor MKC8866 decreased the expression of numerous immunomodulators, including IL-6, IL-8, CXCL1, TGFβ2, and GM-CSF.

The role of the PERK pathway in promoting cell survival has been established since its discovery [44]. During tumor initiation and expansion, PERK-dependent signaling is utilized to maintain redox homeostasis, thereby facilitating tumor growth [203]. Previous research has demonstrated that the PERK pathway is substantially activated in multiple myeloma, probably contributing to the development of treatment resistance in these patients. The PERK inhibitor GSK2606414 is very effective against myeloma cells. These results imply that the PERK pathway may be a possible therapeutic target for the treatment of multiple myeloma patients since they were more prominent when the inhibitor was combined with an anti-myeloma medication, such as the proteasome inhibitor bortezomib [204, 205]. The PERK-eIF2α-ATF4 pathway has been demonstrated to be a crucial factor in the ability of cancer cells to adjust to hypoxic stress both in vivo and in vitro [206]. Inhibition of PERK kinase function in human cancer cell lines has demonstrated an increase in apoptosis under hypoxia in vitro and impaired tumor growth in vivo [207]. Additionally, tumors derived from transformed PERK-deficient fibroblasts tend to exhibit significantly reduced tumor growth, and it was determined that the impaired angiogenesis and the PERK-deficient cells’ sensitivity to the resulting hypoxic environment were the causes of the decreased tumor growth [207, 208]. There has been further proof shown that PERK plays a role in promoting tumor growth in genetically engineered mouse mammary tumor virus (MMTV)-c-Neu oncogene transgenic mice crossed with PERK knockout mice, where no delay is found in tumor development but rather a significant defect in tumor progression and a significantly reduced rate of metastatic spread are identified as a result of extensive DNA damage caused by increased ROS accumulation [203]. Transgenic mice that expressed the SV40 T-antigen in the pancreatic insulin-secreting beta cells showed that insulinoma growth and angiogenesis were significantly reduced in PERK knockout mice compared with wild-type controls [209]. It has been noted that imatinib mesylate (STI571)-resistant BCR/ABL (a fused oncogene formed from the ABL gene from chromosome 9 fused to the BCR gene on chromosome 22) leukemia cell lines exhibit increased PERK activation and upregulation of the eIF2 pathway; therefore, it has been hypothesized that PERK inhibition may make chronic myeloid leukemia cells more responsive to therapy [210–213]. There is evidence that PERK signaling has effects on the tumor microenvironment in relation to tumor progression. For instance, proinflammatory cytokine production is associated with PERK activation [214, 215]. Thus, the PERK pathway seems to drive carcinogenesis by triggering prosurvival pathways that assist cancer cells in adjusting to the hostile tumor microenvironment. It is believed that the typical role of PERK is to protect secretory cells against ER stress. Pharmacological or hereditary suppression of PERK decelerates the growth of tumor xenografts in mice. Conventional PERK knockout mice exhibit diabetes as a result of pancreatic islet cell death as well as significant developmental abnormalities and growth retardation. The characteristics are similar to those observed in patients with Wolcott-Rallison syndrome who possess inactivating germline mutations in the PERK gene [162, 163]. Also, embryonic PERK deletion results in pancreatic failure, and glucose homeostasis ultimately results in prenatal death [162, 216]. In addition, conditional PERK deletion with a tamoxifen-inducible CRE enzyme showed that PERK deletion leads to the destruction of pancreatic tissue [217]. Although certain types of cancer and metabolic disorders exhibit excessive PERK activation, genetic ablation of PERK results in pancreatic inflammation and irregular glucose levels. This implies that a certain level of PERK function is necessary for maintaining pancreatic health. It could be feasible that a reduction in PERK activity, instead of complete inhibition, may offer therapeutic advantages while maintaining the regular functioning of the pancreas.

In addition, increased levels of ER chaperones, such as GRP78 and GRP94, which are classical UPR markers, are frequently seen in solid tumors. GRP78 has been linked to carcinogenesis [218]. GRP78 is highly expressed in a variety of tumors, presumably due to the ER stress triggered by the hypoxic and nutrient-deficient tumor microenvironment, and its expression correlates to tumor growth, invasion, and tumor metastases [218]. Previous studies have shown that the downregulation of GRP78 in fibrosarcoma cells impairs their ability to form tumors when transplanted into mice [219].

Targeting and inhibiting the ER stress pathways via chemical compounds has been demonstrated to be effective against multiple tumors in preclinical and clinical studies [220, 221] (Table 1). Therefore, the discovery of small-molecule inhibitors that target ER stress sensors, specifically PERK, will provide an attractive therapeutic approach for anticancer and antiangiogenic activity. Inhibiting the integrated stress response (ISR) by using ISRIB reduced tumor size in transgenic mice and patient-derived xenograft models of metastatic disease [222]. Recently, small-molecule inhibitors have entered clinical trials for various tumors, including clear cell renal cell carcinoma and multiple myeloma (https://clinicaltrials.gov/ct2/show/NCT04834778; https://clinicaltrials.gov/ct2/show/NCT05027594). The first identified small-molecule inhibitor of PERK is GSK2606414 [223]. GSK2606414 is an inhibitor that competes with ATP and is highly selective for PERK. Another compound, GSK2656157, was developed for preclinical studies. In various human tumor xenograft models, GSK2656157 was found to efficiently reduce cancer growth [224, 225]. In conclusion, depending on the significant role of PERK in promoting inflammation, tumor progression, angiogenesis, and tumor metastases, targeting the PERK signaling pathway presents a novel therapeutic strategy for cancers, particularly those of secretory function such as neuroendocrine tumors, adenocarcinoma, and multiple myeloma, as well as those with more significant hypoxia such as pancreatic adenocarcinoma and glioblastoma.

Table 1. New therapeutics targeting endoplasmic reticulum stress signaling pathways are in preclinical and clinical research to treat cancer and potentially other diseases in the future.

Emerging evidence suggests that agents affecting ER stress and UPR could be exploited as promising anticancer drug candidates. Several compounds that target ER stress and UPR have been confirmed in preclinical studies and clinical trials for the treatment of various cancer types, and some drugs have been approved by the Food and Drug Administration (FDA). Agents that either disrupt the cytoprotective function of the altered UPR of cancer cells or cause severe ER stress and cell death might be utilized alone or in conjunction with traditional anticancer therapies.

Experiments	Molecules	Targets	Mechanism	Tumor type	Phase	Pharmacological effects	PMID
Preclinical	MAL3–101	HSP70	Inhibition of molecular chaperones	Multiple myeloma	Preclinical	Triggers tumor-specific apoptosis and cell death	21977030
	MG-132	26S proteasome complex	Inhibition of proteasomes activity	Glioblastoma	Preclinical	Inhibits tumor cells growth Induces tumor cell apoptosis	22959274 22897979
	Toyocamycin	IRE1	Inhibition of XBP1 splicing	Leukemia	Preclinical	Inhibits tumor cells	18653996 22538852
	4μ8C	IRE1	Inhibition of XBP1 mRNA splicing	Multiple myeloma	Preclinical	Inhibits tumor growth Enhances the efficacy of chemotherapy to tumor cells	22315414 24821775
	MKC-3946	IRE1	Inhibition of XBP1 mRNA splicing	Multiple myeloma	Preclinical	Inhibits tumor growth Induces tumor cell apoptosis Enhances the efficacy of bortezomib or 17-AAG to tumor cells	14559994 25069067 22538852
	STF-083010	IRE1	Inhibition of XBP1 mRNA splicing	Multiple myeloma	Preclinical	Inhibits tumor growth Induces tumor cell apoptosis	21081713 19609461
	GSK2656157	PERK	Inhibition of PERK activity	Pancreatic cancer Multiple myeloma	Preclinical	Inhibits tumor growth and angiogenesis	23333938
	GSK2606414	PERK	Inhibition of PERK activity	Multiple myeloma	Preclinical	Inhibits tumor growth	33028016
	ISRIB	eIF2α	Inhibition of eIF2α phosphorylation	Chronic myeloid leukemia	Preclinical	Sensitizes CML cells to imatinib Eliminates Leukemia cells	36460969
	Ceapin-A7	ATF6	Inhibition of ATF6 activity		Preclinical	Sensitizes tumor cells to ER stressor	27435960
	CB-5083	P97 (a key regulator of ERAD)	Activation of UPR	Solid tumors	Preclinical	Induces tumor cell apoptosis	26555175
Clinical trials	Sunitinib	Multitargeted tyrosine kinase	Inhibition of IRE1 activity	Renal cell carcinoma	Phase III; FDA approved for renal cell carcinoma	Increases patient survival rate after nephrectomy	27718781
	Sorafenib	Multitargeted tyrosine kinase	Activation of UPR	Hepatocellular carcinoma Desmoid tumors	Phase II and Phase III; FDA approved for hepatocellular carcinoma and renal carcinoma	Inhibits tumor-cell proliferation and tumor angiogenesis Increases the rate of apoptosis	18650514 30575484
	Delanzomib CEP-18770	26S proteasome	Inhibition of proteasome activity	multiple myeloma	Phase I/II clinical trials	55% of patients had stable disease and 9% had a partial response	28140719
	Lobaplatin	PERK	Inhibition of PERK-eIF2α-ATF4-CHOP pathway	breast cancer Hepatic cancer	Phase II clinical trial NCT03210389	Reduce tumor recurrence and metastasis	29483583
	Bortezomib PS-341	26S proteasome	Inhibition of proteasome activity	Multiple myeloma Cell lymphoma	Phase II, III VISTA trial; FDA approved for multiple myeloma	Increases survival rate Increases response to Dexamethasone	20368561 14657528 15953001 15953004 12826635

ER stress sensors mediate inflammation in cardiovascular diseases.

Cardiovascular diseases (CVD), including atherosclerosis, ischemic heart disease, and heart failure, are the main causes of morbidity and death in CKD patients. Proper cardiovascular function is associated with ER homeostasis, and ER stress is a factor in the development of many different types of CVDs. There is mounting evidence that ER stress contributes significantly to the development and progression of atherosclerotic cardiovascular diseases (CVDs) by being linked to inflammatory signaling pathways via a variety of mechanisms [226, 227]. Specifically, prolonged ER stress activates the three ER transmembrane stress sensors, PERK, ATF6, and IRE1, to initiate UPR signaling that induces oxidative stress and specific inflammatory responses associated with CVD, particularly in ECs and macrophages [228, 229]. Human aortic endothelial cells that have experienced ER stress and cellular UPR activation produce more inflammatory molecules such as IL-8, IL-6, monocyte chemoattractant protein 1 (MCP1), TNF-α, and the chemokine CXC motif ligand 3 (CXCL3), which are blocked by gene silencing of ATF4 and/or XBP1 [230], resulting in vulnerable atherosclerosis plaques [231]. A recent report showed that ER stress plays a role in cardiac remodeling in hypertensive animals [232]. Additionally, it was demonstrated that in rats receiving aldosterone salt treatment, inhibiting ER stress reduces cardiovascular remodeling [233]. Increased PERK and IRE1 levels, as well as accumulated ROS, activate and augment the inflammatory response in atherosclerosis [74]. PERK-mediated translation inhibition results in phosphorylation of the inhibitor of nuclear factor-κB (IκB). Then NF-κB is released and translocated to the nucleus, where it promotes the activation and expression of genes involved in downstream inflammatory pathways, such as those that encode cytokines like TNF-α and IL-1 [227, 234]. Oxidized low-density lipoprotein (ox-LDL) is a key mediator of endothelial dysfunction and the initiation and progression of atherosclerosis [235]. Studies have shown that ox-LDL induces expression of PERK and phosphorylation of eIF2α in human umbilical vein endothelial cells (HUVECs) [236]. We recently showed that PERK activation increases endothelial cell activation, trained immunity, and vascular inflammation in human aortic endothelial cells [15]. In addition, PERK deletion in vascular smooth muscle cells (VSMCs) has been shown to prevent up to 80% of plaque formation in hyperlipidemic mice while also inhibiting VSMC phenotypic modulation [237] and migration [72, 238].

Given the importance of ER stress in the pathophysiology of CVD, strategies targeting ER stress, particularly the dysfunctional UPR, are emerging as potential therapeutic pathways for disease intervention. To prevent and treat CVD, selectively and effectively targeting ER stress in an organelle- or organ-specific manner, such as the mitochondria, holds promise.

ER stress sensor signaling pathways and neurodegenerative diseases.

In contrast to the majority of mammalian cell types, neurons have a limited ability to regenerate. Consequently, the loss of neurons due to cell death leads to neurodegeneration, a loss of neuronal function in the tissues of the central nervous system. The UPR increases the production and release of proinflammatory mediators such as NF-κB, which regulates the expression of the nuclear proto-oncogene SET. SET is directly implicated in the pathogenesis of neurodegenerative diseases. Among the most well-known neurodegenerative diseases are Alzheimer’s disease, Parkinson’s disease, and Huntington’s disease [239]. Alzheimer’s disease is characterized by hyperphosphorylation of the Tau protein (pTau), which destabilizes the microtubules in neurons, leading to the formation of intracellular neurofibrillary tangles (NFTs) and extracellular plaques due to the accumulation of Amyloid-β (Aβ) peptides as a result of mutations in the amyloid precursor protein. Consequently, the accumulation of pTau, NFTs, and Aβ-oligomers increases protein load in the ER, which triggers ER stress and initiates the UPR [240–242]. In models of Alzheimer’s disease, there has been an observed increase in the expression of heat shock protein family A (HSP70) member 5 (BiP), p-PERK, p-IRE1α, P-eIF2α, and ATF4. The activation of the apoptosis signal-regulating kinase 1 (ASK1) branch of the IRE1α pathway may also result from ER stress and oxidative stress caused by Aβ/pTau [243], suggesting that ASK1 might be a new therapeutic target to prevent or treat Alzheimer’s disease. Furthermore, prolonged PERK activation has been demonstrated to impact memory and promote neurodegeneration by impairing protein synthesis [76]. In Parkinson’s disease, the accumulation of mutant α-synuclein (lewy body dementia, abnormal deposits of a-synuclein in the brain) in the dopaminergic neurons results in the activation of the UPR and PERK pathways through increased levels of p-PERK and p-eIF2α [244, 245]. In Huntington’s disease, the accumulation of mutant huntingtin (mHtt) causes ER stress and upregulation of the IRE1, ATF6, and PERK pathways [246]. Although ER stress is a common characteristic of neurodegenerative disorders, it triggers all three pathways of UPR. Yet, it is evident that the PERK pathway plays the main role in the development and progression of neurodegenerative disorders. Therefore, the use of small-molecule drugs to modulate the PERK pathway shows great potential as a therapeutic strategy for various neurodegenerative disorders.

ER stress and UPR in kidney diseases

Healthcare professionals face several difficulties while treating patients with acute kidney injury (AKI) and chronic kidney disease (CKD). AKI is characterized by the rapid loss of renal function and represents an independent risk factor for CKD and end-stage renal disease (ESRD) [247]. CKD is a cardiometabolic disorder characterized by vascular calcification and the accumulation of toxic metabolites, ultimately leading to decreased renal function. Understanding the underlying causes of CKD can help reduce renal dysfunction and associated complications, such as cardiovascular abnormalities. The discovery of novel pharmacological targets for kidney disease is necessary for the development of more effective therapies. ER stress contributes to the development and progression of a wide variety of kidney diseases, including AKI, diabetic nephropathy, nephrotoxicity, and renal fibrosis [75]. The sustained UPR activation leads to the upregulation of inflammatory cytokines and fibrotic genes, which leads to progressive tissue injury and fibrosis. Increased expression of phosphorylated PERK, phosphorylated IRE1α, ATF6, XBP1, and CHOP was observed in the kidneys of diabetic db/db (leptin receptor mutation) mice and renal ischemia/reperfusion (I/R) mouse models. PERK-ATF4-CHOP and IRE1-XBP1 signaling pathways are activated in ischemic acute kidney injury and mediate immunological responses, autophagy, inflammation, apoptosis, and renal fibrosis [248–251] and inhibition of this pathway by chemical chaperones can mitigate renal dysfunction and fibrosis [252]. ATF6 knockout animals showed less lipid deposition and tubulointerstitial kidney fibrosis through decreased expression of proliferator-activated receptor alpha (PPARα), which causes mitochondrial malfunction, oxidative stress, inflammation, apoptotic cell death, and an increase in pro-fibrotic genes [250]. Modulation of CHOP expression may be the most effective target to slow the progression of kidney disease and fibrosis. CHOP deficiency attenuates renal fibrosis and kidney renal diseases [253]. Significant advancements have been achieved in treatments that target ER stress to attenuate the progression of renal diseases. Chemical chaperones approved by the Food and Drug Administration (FDA), such as ursodeoxycholic acid (TUDCA) and 4-phenylbutyric acid (4-PBA), are classic ER stress suppressors that improve insulin sensitivity in metabolic disorders associated with diabetes, restore glucose tolerance, and enhance ER protein folding capacity [254, 255].

Conclusion

The ER is a crucial organelle for maintaining protein homeostasis. Misfolded protein accumulations occur in the ER as a result of physiological or pathological stresses disrupting ER equilibrium. ER stress triggers the UPR, which activates the compensatory protection mechanism as the body’s adaptive and defensive reaction to damaging stimuli. ER stress can be triggered by various stressors that affect homeostatic cellular functions, and prolonged ER stress has been identified as a pathogenic mechanism in the development and progression of a large number of diseases. Emerging evidence indicates that ER stress is implicated in inflammatory responses and plays rather fundamental roles in the onset and progression of cardiovascular diseases [227], cancer [192], kidney diseases, and neurodegenerative diseases [242]. In response to ER stress, UPR is activated under the mediation of three ER membrane-associated proteins: PERK, IRE1, and ATF6 [227]. Pharmacologically, there are two ways to target ER stress: either directly by modifying the misfolded protein accumulation within the ER, or indirectly by controlling UPR signaling via ER stress sensors or the enzymes that mediate their downstream effects. Targeted inhibition of PERK/eIF2, ATF6, and IRE1, the three primary UPR branches, can reduce ER stress and, as a result, exert a protective effect.

In this review, we summarized the ER stress new sensor functions, particularly PERK as a conditional DAMP receptor, and its roles in organelle crosstalk regulation, regulation of protein translation, regulation of membraneless stress assemblies’ formation, as well as modulation of the innate immune response, regulation of metabolic reprogramming, epigenetic modification, and trained immunity. We also summarized the roles of ER stress in several diseases, including cardiovascular diseases, cancers, neurodegenerative diseases, and kidney diseases. We proposed that, in addition to the pattern recognition receptors (PRRs), such as Toll-like receptors (TLRs) and inflammasomes, the disease risk factors, DAMPs, and PAMPs can be sensed by protein folding as well as the three ER stress molecules, including IRE1, PERK, and ATF6. It has been well characterized that PRR signaling pathways are the initiators of cascades that eventually lead to the amplification of the inflammatory response.

Our new working model (Figure 4) proposes that the ER has two new categorized sensor types: protein folding and ER stress regulators as DAMP receptors (PERK). Under disease conditions, disease risk factors, DAMPs, and PAMPs can be recognized by pattern recognition receptors (PRRs), such as Toll-like receptors (TLRs), and inflammasome pathways to enhance inflammatory responses. DAMPs and PAMPs can also be sensed by protein folding and the three ER stress molecules, including IRE1, PERK, and ATF6. A fundamental factor in the development of many pathologies is the loss of protein folding homeostasis, and ER stress plays a significant role in this process. The activation of the ER stress pathways results in impaired calcium homeostasis, increased mitochondrial ROS production, and oxidative stress. The toxic accumulation of ROS within the mitochondria and ER and sustained ER stress trigger inflammatory responses. Dysfunctional UPR pathways have been linked to a variety of chronic diseases, such as inflammatory disease, cardiovascular disease, metabolic disorders, neurodegenerative disorders, cancer, and others. An improved understanding of the molecular mechanisms underlying the ER stress pathways may lead to the discovery of highly promising therapeutic targets for chronic inflammatory diseases, cardiovascular diseases, immune disorders, kidney diseases, neurodegenerative diseases, diabetes, and cancers. Given the detrimental effects of ER stress, the development of novel drugs that can specifically target ER stress in a cell- and disease-specific manner is critically needed. In summary, the advancement and knowledge gained would also support the treatment of diseases linked to ER stress.

Figure 5. Our new working model.

ER has two new categorized sensor types: protein folding and ER stress regulators as DAMP receptors (PERK). Under disease conditions, disease risk factors, DAMPs, and PAMPs from protein-rich food and sea foods can be recognized by pattern recognition receptors (PRRs), such as Toll-like receptors (TLRs), and inflammasome pathways to enhance inflammatory responses. DAMPs and PAMPs can also be sensed by protein folding and the three ER stress molecules, including IRE1, PERK, and ATF6. Loss of protein folding homeostasis is central to many diseases, and ER stress is a major contributor to the development of these pathologies. The activation of the ER stress pathways results in impaired calcium homeostasis, increased mitochondrial ROS production, and oxidative stress. The toxic accumulation of ROS within the mitochondria and ER and sustained ER stress trigger inflammatory responses. Dysfunctional UPR pathways have been linked to a variety of chronic diseases, such as neurodegenerative diseases, metabolic disorders, cancer, inflammatory disease, diabetes mellitus, cardiovascular disease, and others. Created with BioRender.com

Abbreviations:

AIM2

absent in melanoma 2

AKI

acute kidney injury

ATF4

activating transcription factor 4

ATF6

activating transcription factor 6

AKT1/2/3

protein kinase B

ATP

adenosine triphosphate

BAP31

B-cell receptor-associated protein 31

BiP

binding immunoglobulin protein

CHOP

C/EBP-homologous protein

C/EBP

CCAAT-enhancer binding protein

CKD

chronic kidney disease

CRAC

calcium release-activated calcium

CREB

cAMP response element binding protein kinase

CVD

cardiovascular diseases

CXCL3

chemokine CXC motif ligand 3

DAG

diacylglycerol

DAMPs

danger-associated molecular patterns

ECAR

extracellular acidification rate

eIF2AK3

eukaryotic translation initiation factor-2α kinase 3

EGFR

epidermal growth factor receptor

ER

endoplasmic reticulum

ERAD

ER-associated degradation

ERK1/2MAPK

extracellular signal-regulated kinase

ERMES

ER-mitochondria encounter structures

ESRD

end-stage renal disease

FADD

FAS-associated death domain protein

FIS1

mitochondrial fission 1 protein

FMO3

flavin-containing monooxygenase 3

FMR1

fragile X messenger ribonucleoprotein 1

FOXO

forkhead box O protein

G3BP1

GTPase-activating protein-binding protein 1

GCN2

general control non-derepressible 2

GFR

glomerular filtration rate

GPCRs

G-protein-coupled receptors

GRP75

glucose regulatory protein 75

GRP78

glucose-related protein 78

GSK-3

glycogen synthase-3

HAMPs

homeostasis-associated molecular patterns

HDAC6

histone deacetylase-6

H3K27

histone 3 lysine 27

HMGB1

high-mobility group box 1

HMG-CoA

3-hydroxy-3-methylglutaryl coenzyme-A

HRI

heme-regulated initiation factor 2α kinase

HSP

heat shock protein

IL-13

interleukin-13

InsP3

inositol-1,4,5-trisphosphate

IP3R

inositol triphosphate receptor

IRE1

inositol-requiring enzyme-1

ISR

integrated stress response

JNK1/2/3MAPK

c-Jun N-terminal kinase

LPS

lipopolysaccharides

MAM

mitochondria-associated ER membranes

MCP1

monocyte chemoattractant protein 1

MFN1

mitochondrial fusion protein 1

MFN2

mitofusin 2

MSK1/2

mitogen- and stress-activated kinases

mTOR

mammalian target of rapamycin

NFTs

neurofibrillary tangles

NLRs

Nucleotide-binding oligomerization domain (NOD)-like receptors

Nrf2

nuclear factor erythroid-derived 2-like 2

ORFs

open reading frames

Osh

oxysterol-binding homology

ox-LDL

Oxidized low-density lipoprotein

PAB1

poly(A) binding protein cytoplasmic 1

PAMs

plasma membrane-associated ER membranes

PAMPs

pathogen-associated molecular patterns

PDI

protein disulfide isomerases

PDGFRβ

platelet-derived growth factor receptor β

PERK

protein kinase R-like endoplasmic reticulum kinase

P38αMAPK

p38 mitogen-activated protein kinase (MAPK)

PI4P

phosphatidylinositol 4-phosphate

PKR

protein kinase RNA-activated

PPAR-γ

peroxisome proliferator-activated receptor γ

PM

plasma membrane

PRKAA1

protein kinase AMP-activated catalytic subunit alpha 1

PRRs

pattern recognition receptors

PTPIP51

protein tyrosine phosphatase interacting protein 51

RAGE

advanced glycation end products

RIG-I

retinoid acid-inducible gene I

RIPK1

receptor-interacting serine/threonine protein kinase 1

ROS

reactive oxygen species

RYRs

ryanodine receptors

SACM1L

Sac1-like phosphatidylinositide phosphatase

SAMPs

stress-associated molecular patterns

SERCA

sarco-ER calcium-ATPase

S1P

site-1 protease

S2P

site-2 protease

STAT

signal transducer and activator of transcription

STIM

stromal interacting molecule

TIA-1

T-cell intracellular antigen-1

TLR4

toll-like receptor 4

TMAO

trimethylamine-N-oxide

TNF-α

tumor necrosis factor-α

TRAF2

the tumor necrosis factor receptor-associated factor 2

TRAIL

tumor necrosis factor-related apoptosis-inducing ligand

TTP

tristetraprolin

UPR

unfolded protein response

VAPB

vesicle-associated membrane protein-associated protein B

VDAC1

voltage-dependent anion-selective channel protein 1

VSMCs

vascular smooth muscle cells

XBP1

X box-binding protein 1