The role and therapeutic implication of endoplasmic reticulum stress in inflammatory cancer transformation

Abstract

Endoplasmic reticulum (ER) stress occurs when proteins are affected by various factors, fail to fold properly into higher structures and accumulate in the lumen of the ER, which activates the unfolded protein response (UPR) to restore normal cellular function or induce apoptosis as a self-protective mechanism. However, a growing number of studies have shown that the three branches of ER stress and the UPR can mediate inflammation and cancer development by interacting with inflammatory transformation-related signaling pathways. Targeting the UPR, especially the use of small molecules that target the active sites of the enzymes IRE1α and PERK and BIP/GRP78 inhibitors are potential strategies for treating tumors and have shown promising results in some tumor models. Therefore, in this review, we summarize the progress of ER stress/UPR research and the signaling pathways associated with inflammatory cancer transformation, provide an in-depth description of the mechanisms of these pathways, and outline strategies in the field of UPR biology in tumor therapy to provide new ideas for the mechanisms of inflammatory cancer transformation and tumor-related treatment.

Introduction

The endoplasmic reticulum (ER) has essential cellular functions, including the synthesis and folding of secreted and transmembrane proteins, calcium storage, and lipid synthesis for membrane biogenesis or energy storage. Disruption of any of these functions leads to ER stress and subsequent activation of an elaborate network of adaptive responses that are collectively known as the unfolded protein response (UPR). The UPR re-establishes homeostasis in vivo through transcriptional and translational control, and the UPR signals through three mechanistically distinct branches that are initiated by the ER-resident protein folding sensor inositol-requiring enzyme 1 (IRE1), protein kinase RNA-like endoplasmic reticulum kinase (PERK), and activating transcription factor 6 (ATF6) [1]. However, protein folding is an error-prone process and is therefore tightly regulated by many molecular chaperones and enzymes [2,3]. Normally, three trans-ER membrane proteins bind specifically to the molecular chaperone heavy chain binding protein/glucose regulatory protein 78 κDa (BIP/GRP78) in the ER lumen [4], and when ER stress occurs, BIP/GRP78 dissociates, binds to unfolded or misfolded proteins in the lumen to aid in correct folding and triggers downstream pathways and effector mechanisms that remodel the ER to restore homeostasis in vivo [5,6]. However, the UPR is not always effective in regulating ER homeostasis [7], and if stress persists, it leads to ER dysfunction, which activates relevant apoptotic pathways and mediates cell death [8].

Multiple pro-inflammatory signaling pathways are involved in the transformation of inflammation to cancer: aberrant and sustained activation of nuclear factor-κB (NF-κB) signaling promotes the malignant transformation of inflammation to cancer, and NF-κB induces the production of cytokines that regulate the immune response (e.g., TNF-α, IL-1, IL-6 and IL-8), as well as adhesion molecules that recruit leukocytes to sites of inflammation [9]. In addition, NF-κB may also control the transformation and metastasis of epithelial cells to mesenchymal cells [10]. The upregulation of VEGF and its receptor controls tumor angiogenesis and promotes tumor progression [11]. Specific NF-κB inhibitors targeting the IKK complex are promising anticancer agents [12]. Growth factors (VEGF and TGF-β) and various cytokines, including IL-6, IL-17, IL-10, granulocyte-macrophage colony-stimulating factor (GM-CSF) and leukemia inhibitory factor (LIF), promote signal transducer and activator of transcription 3 (STAT3) activation, and STAT3 transduces signals from all members of the IL-6 and IL-10 families to promote tumor growth and disease progression [13].

Numerous studies in recent years have shown that ER stress is closely related to the inflammatory response, IRE1 activation in macrophages mediates a decisive pro-inflammatory response [14,15], and ATF6 was shown to be activated in inflammatory bowel disease [16]. In the sophisticated and complex process of inflammatory cancer transformation, ER stress is widely involved in different signaling pathways, such as NF-κB [17], STAT [18], PI3K/AKT [19,20], and Notch [21], to regulate inflammation to cancer transformation. Moreover, the molecular mechanism of inflammation-induced tumorigenesis is an important branch of tumor therapy, and so in this review, we will focus on ER stress, the three ER membrane receptors, and the signaling pathway associated with inflammatory cancer transformation, summarize the mechanism by which ER stress/the UPR affect inflammatory cancer transformation, outline the strategy of studying the UPR in tumor therapy, and provide new ideas for the diagnosis and treatment of tumors.

IRE1 and inflammatory cancer transformation

IRE1, which is a dual enzyme with serine/threonine kinase and ribonucleic acid endonuclease activity, is the most highly conserved gene from yeast to humans, and IRE1α is one of the isoforms that is normally expressed in most cells and tissues [22,23]. Under ER stress conditions, the structural domain of the kinase IRE1 phosphorylates itself and activates endoribonuclease (RNase) activity [24,25], followed by RNase domain binding to X-box binding protein 1 (XBP1) mRNA to translate the active transcription factor XBP1s [26,27], which induces the transcriptional chaperones and lipid synthesis involved in the expression of target genes [26,28,29] and increases the protein folding capacity and concomitant expansion of the ER membrane area. The RNase activity of IRE1 is associated with various pathologies, such as cancer and inflammatory, metabolic and degenerative diseases, for instance, inhibiting IRE1 activity is effective in preventing MYC-driven breast cancer development [30], and the regulation of IRE1 activity is emerging as an attractive therapeutic strategy against these diseases [31].

Prolonged ER stress activates IRE1 [32], IRE1α/XBP1 axis further promoted the phosphorylation of STAT3 Tyr705 in primary effusion lymphoma (PEL) by inducing the release of IL-6, IL-10 and VEGF to enable the survival of PEL cells [33]. In human melanoma cells, spliced/activated XBP1 induced IL-6 expression by directly binding to the IL-6 promoter, which further promoted STAT3 phosphorylation and enhanced the oncogenic effects of STAT3 [34]. However, IL-6 and its biological effects activated by STAT3 not only play an important part in tumor progression, but also are closely related to tumor-associated macrophages (TAMs) [35-37]. Some investigators have specifically observed XBP1s expression in TAMs, which promotes the growth and metastasis of colorectal cancer (CRC) by enhancing the expression of the protumor cytokines IL-6, VEGFα, and IL-4, triggering STAT3 signaling by activating downstream SIRPα [38,39]. 7S,15R-dihydroxy-16S,17S-epoxy-docosapentaenoic acid (diHEP-DPA) use reduced SIRPα expression in TAMs and effectively reversed TAM conditioned medium (TCCM)-induced epithelial-mesenchymal transition (EMT) [39,40]. In M1-like TAM, the mesenchymal/stem-like phenotype of oral squamous cell carcinoma (OSCC) was shown to be linked to the IL-6/STAT3/THBS1 feedback loop [41]. In addition, the RNase activity of IRE1α also regulates phosphatidylinositol (PI3K) levels during premiR2137 maturation and mediates ER stress-related degradation (ERAD) via the PI3K/AKT/mTOR pathway in breast cancer cells [42,43].

Notch signaling is an intercellular transmission mechanism in which the interactions between Notch receptors and membrane-bound Notch ligands expressed on juxtaposed cells regulate the fate of one cell via neighboring cells [44-46] and is a major factor in stem cell maintenance in many tumors [47]. Dll1 in mammary stem cells activates Notch signaling in macrophages and enhances mammary stem cell self-renewal through macrophage-derived WNT ligands [48]. Deletion of Notch3 under ER stress conditions allows BIP/GRP78 to interact with IRE1α, leading to its ubiquitination and inactivation [49]. Targeting Notch3 in with the presence of ER stress inducers could be a therapeutic approach for T-cell acute lymphoblastic leukemia (T-ALL): the Notch3-dependent IRE1α/XBP1 axis disrupts the conversion of prosurvival UPR signaling to the proapoptotic UPR signal, promoting CHOP-mediated apoptosis [50]. In addition, the NOTCH signaling pathway directly regulates the expression of ERAD genes (e.g., Sel1l, Hrd1, Os9, Edem1) [51]. Sel1l deletion significantly activates three branches of the UPR, leading to the accumulation of mis/unfolded proteins and apoptosis. The IRE1α/XBP1 signaling pathway can act as a compensatory mechanism to suppress Sell deletion-induced ER stress, and PERK deletion significantly restores immune cell numbers and inhibits thymocyte apoptosis in Sel1lCKO mice [52]. Despite the opposite effect of ERAD on Notch signaling [53,54], ER stress inhibits GLP-1 (a Notch receptor body)/Notch-mediated aberrant proliferation, promoting tumor formation, and the IRE1/XBP1 branch of the UPR is an important weapons. The UPR may not distinguish between gain-of-function or loss-of-function when correcting aberrantly folded mutant Notch receptors but simply identifies these proteins as “corrected misfold” [55,56]. More studies are needed to provide insight into whether the compensatory mechanism by which IRE1α/XBP1 promotes ER homeostasis after ERAD/Notch deletion is based on two conserved protein regulatory systems [57] or on a regulatory mechanism associated with inflammatory signaling pathways under different conditions.

PERK and inflammatory cancer transformation

PERK is a MAM-rich component that interacts with multifunctional mitochondria-associated ER membrane (MAM) tethering proteins and regulates the exchange of metabolites, such as lipids, reactive oxygen species (ROS) and Ca²⁺, at the contact site [58]. In contrast to IRE1, PERK has a different function in determining cell fate [59,60]. When stimulated by unfolded proteins, PERK oligomerizes, phosphorylates itself and subsequently phosphorylates eukaryotic translation initiation factor 2α (eIF2α) to promote the transcription of ATF4, which accumulates at the promoters of target genes, including CHOP, GADD34 and ATF3 [61-63]. The PERK/ATF4/CHOP signaling pathway is thought to play a key role in inducting apoptosis of liver cells, podocyte, PC12 cells in vitro and in vivo [64-66], and joint involvement of CHOP and ATF4 is required for the induction of cell death [61]. In other words, sustained PERK signaling can promote apoptosis [59]. To further support this hypothesis, one investigator used the ERα bioregulator BHPI to induce breast cancer cell apoptosis and found that this treatment activated PERK but suppressed IRE1α/XBP1 expression due to the functional inhibition of ERα, suggesting that PERK activation was sufficient to induce apoptosis [67].

On the other hand, sustained activation of PERK causes phosphorylation of the downstream signal eIF2α at serine 51 [68]. eIF2α activates NF-κB by reducing the level of IκBα in mouse embryonic fibroblasts [69]. NF-κB, a key molecule between chronic inflammation and cancer [70], acts as a central regulator of immune response and inflammation by upregulating many chemokines CXCL1, CXCL2, CXCL3 and cytokines TNF-α, IL-1β, IL-6, IL-8. Activated NF-κB also affects cell proliferation and apoptosis by targeting Bcl2, IAPs and cell cycle proteins [71], it also participates in adaptive immune responses through B cells and T cells, which further exacerbate inflammation [72]. This is associated with reduced genetic stability and gene mutations in tumorigenesis and progression [73]. But this effect is not absolute. NF-κB in LTED breast cancer cells, however, is activated independent of PERK and eIF2α phosphorylation, whereas STAT3, which is a stress response mediator, is phosphorylated by PERK to increase NF-κB DNA binding activity [69]. In addition, specific STAT3 nuclear translocation inhibitors significantly reduce NF-κB DNA binding activity in breast cancer cells, suggesting that DNA-level interactions could mediate full activation of NF-κB by STAT3. Thus, PERK transmits stress signals from the ER to the nucleus by activating STAT3 and promoting NF-κB activation [69,74,75]. However, not only PERK kinase, but also other inflammatory factors such as high levels of IL-6 stimulation provide a link between STAT3 and NF-κB in tumors [76,77], while synergistically participating with TNF-α in the activation of NF-κB and STAT3 signaling pathways. Together, they play an important role in inflammation and tumorigenesis [78], an event which has been shown to promote CRC cell growth [79].

As previously mentioned, downstream of PERK, CHOP and ATF4 work together to cause apoptosis, which is clear, but the downstream signal eIF2α also plays an important role, which shows that PERK can activate NF-κB in breast cancer cells through both eIF2α-dependent/independent phosphorylation, while STAT3 acts as a stress mediator that further promotes NF-κB activation and subsequently transmits stress signals to the nucleus in cancer cells to produce a strong regulatory effect. This finding suggests that PERK induces apoptosis when cells are subjected to various adverse stimuli, but in breast cancer cells, ER signaling is delivered to the nucleus through multiple inflammatory transformation signaling pathways, which ultimately results in reduced cancer cell proliferation and apoptosis, which may be related to tumor cell immune escape but is always the result of many factors.

ATF6 and inflammatory cancer transformation

ATF6 is one of the three sensors of the UPR and is anchored to the ER membrane with its C-terminus located in the ER lumen and its N-terminal end toward cytoplasmic side. In the absence of ER stress, inactive ATF6 is a 90-κDa ER transmembrane protein [80,81]. Under ER stress conditions, ATF6 interacts with BIP/GRP78 by cleaving the disulfide bond and translocating from the ER lumen to the Golgi apparatus, where it is sequentially cleaved by the Golgi apparatus by site 1 and site 2 proteases (S1P and S2P) [81,82]. These modifications release a 50-κDa N-terminal cytoplasmic fragment containing the transcriptionally active bZIP structural domain. Activated p50-ATF6 translocates into the nucleus and regulates the expression of downstream target genes such as GRP78 and XBP1 [81,83]. Sterol regulatory element binding protein (SREBP) regulates genes involved in cholesterol and fatty acid homeostasis and is also an ER membrane-bound transcription factor. SREBP is activated through the same protein hydrolysis mechanism as ATF6, and it is also cleaved by the proteases S1P and S2P in the ER [82,83]. ATF6 forms a complex, stimulated cholesterol gene by interacting with SREBP2 and synthesizing cholesterol de novo [84,85].

ATF6 regulates the inflammatory cancer transformation signaling pathway to exercise dual roles in tumor cells. ATF6 can mediate the PI3K/AKT/mTOR pathway, enabling ERAD-mediated increases in ER folding capacity in breast cancer cells [43]. However, inhibitor of DNA binding 1 (ID1) activates NF-κB signaling by promoting nuclear translocation of NF-κB p65, which enhances IL-6 expression and secretion in ovarian cancer cells, and subsequently activates STAT3 and promotes ATF6 transcription through protein phosphorylation at Y705, culminating in the induction of ER stress to promote cellular autophagy, which is one of the important reasons for making ovarian cancer harder to cure [86]. ATF6 is also expressed in ulcerative colitis-associated CRC as a marker of precancerous atypical changes [87]. In intestinal epithelial cells, ATF6 is regulated by CSNK2B and ACSL1, which exacerbates inflammation by reducing the transcription of its target genes HSP90B1 and HSPA5 and activating NF-κB [16]. CSNK2B has recently been reported to be a novel oncogene that promotes CRC cell proliferation by activating the mTOR signaling pathway [88] and regulates negative elongation factor complex member E (NELFE) overexpression to promote gastric cancer progression [89]. In hepatocellular carcinoma, tumor necrosis factor α-inducible protein 1 (TNFAIP1) blocks NF-κB activation through selective downregulation of CSNK2B [90]; therefore, we predict that CSNK2B may be a potential target gene for controlling CRC, gastric cancer and liver cancer in the future. In addition, after selective cleavage of BIP/GRP78 by chymotrypsin cytotoxin SubAB, investigators found that ATF6 preferentially dephosphorylates CCAAT/enhancer binding protein (C/EBP) and mTOR-dependent dephosphorylation of akt and subsequently blunts NF-κB activation by TNF-α, thereby protecting mice from endotoxic lethality and collagen arthritis [91].

Cancer-related regulation of inflammatory processes are not monolithic; Notch interacts with transcription factors such as STAT3, NF-κB, the proinflammatory factor IL-6 and its signaling pathway, and the JNK signaling pathway [92,93]; Notch1 and Notch target genes enhance NF-κB activity through multiple mechanisms [94-96], and the Notch and NF-κB signaling pathways exert bidirectional regulation through different mechanisms and exert different effects under physiological and pathological conditions. The different regulatory mechanisms of the branches of the UPR with the same targets need to be explored in depth, and as mentioned earlier, the deletion of PERK but not IRE1 rescues thymus development in mice [52]. The mechanisms by which different branches of the UPR jointly regulate inflammatory cancer transformation deserve to be explored at multiple levels.

ER stress inhibitors in inflammatory cancer transformation

Recent drug discoveries have opened the door to approaches that require selective regulation of UPR signaling, and this research has identified a number of novel small molecules that target the enzymatic activity of specific UPR regulators [97-101]. Considering the presence of aberrant UPR signaling in cancer, controlling its pharmacological output to limit tumor growth is considered a very reasonable strategy. IRE1α, PERK, and ATF6 are promising targets to control disease progression as the first responders to the UPR. In particular, there have been many efforts to develop small molecule modulators of the active sites of IRE1α and PERK, while only one ATF6 inhibitor has been reported. Ceapins, which are a new class of pyrazolamides, have been demonstrated to specifically inhibit the ATF6α branch of the UPR by blocking nuclear translocation in ATF6α-treated and ER-stressed U2-OS cell line [102].

First-generation inhibitors of IRE1 activity target the structural domain of ATP kinase, an ATP-competitive IRE1α kinase that inhibits RNase attenuators [103], such as sunitinib; sunitinib, in turn, inhibits VEGFR and PDGFR and effectively inhibits IRE1 phosphorylation by inhibiting autophosphorylation and the subsequent RNase activation [104,105]. In a study of renal cell carcinoma, sunitinib attenuated EGFR loss-induced MAPK (pERK1/2) and pAKT expression and further inhibited EGFR-/- cell proliferation [106]. In contrast to first-generation drugs, second-generation IRE1 inhibitors are compounds that directly target the RNase structural domain and dose-dependently inhibit endogenous IRE1α oligomerization, in vivo XBP1 mRNA cleavage and ER-localized mRNA decay [107]. The known inhibitors that directly target the RNase structural domain are characterized by the fact that they share a common hydroxyaryl aldehyde (HAA) fraction [108] and include B-I09 [109], STF-083010, 4μ8C, toyocamycin, and a series of MKC compounds [97]; among them, B-I09 has been shown to control the aggressiveness of chronic lymphocytic leukemia cells in vivo [109], and the small molecules STF-083010 and 4μ8c react specifically by forming a specific lysine residue (Lys907) with the Schiff base in the RNase structural domain [108]. In addition to forming reversible Schiff bases with Lys907, both agents establish hydrophobic contacts with His910 and Phe889 and form hydrogen bonds with Tyr892 in the IRE1 RNase structural domain [110], thereby blocking their function. This strategy has shown favorable therapeutic effects on diseases; for example, the use of STF-083010, MKC-3946 and toyocamycin inhibited the growth of multiple myeloma [99,109,111-113], and 4μ8c was shown to reduce inflammation in a mouse model of arthritis [114]. In addition, the salicylaldehyde MKC-8866 effectively inhibited IRE1 RNase activity and exerted tumor-suppressive effects to slow tumor recurrence in a mouse xenograft model (PDX) of triple-negative breast cancer (TNBC) [115] and in a glioblastoma multiforme (GBM) model [116]. Recent studies have shown that MKC-8866 also induces the regression of breast tumors caused by MYC overexpression [30]. Whether this covalent inhibitor can also produce desirable effects in other types of cancer is still being tested.

In addition to these pharmacological inhibitors, investigators found that kinase inhibitory RNA enzyme attenuators (KIRAs) inhibit IRE1 RNase activity as ATP-competing ligands, suggesting that inhibition of the kinase site may have an inhibitory effect on RNase activity of pancreatic β-cell tumors [103,117]; interestingly, the peptide fragments of the IRE1 cytoplasmic structural domain itself affect its oligomerization and subsequent RNase activity [118,119]. Therefore, a team has demonstrated that methotrexate, cefoperazone, folinic acid and fludarabine phosphate inhibit IRE1 RNase activity in vitro and in human GBM cell models based on IRE1 peptide fragment interactions; the researchers identified peptides and FDA-approved compounds and sensitized them to the established GBM chemotherapy treatment temozolomide to address the clinical relevance of IRE1 inhibition in oncology. Because these agents are structurally homologous to the region of IRE1 and are therefore complementary, they provide greater selectivity for IRE1 binding and reduce off-target effects than other treatments [120]. In conclusion, blocking IRE1 has been shown to be beneficial in inhibiting tumor progression in mouse models, emphasizing that specific targeting of the UPR can have beneficial effects on disease.

GSK2606414, a potent first-generation PERK inhibitor, has a half inhibitory concentration (IC50) that is less than 1 nM. However, the dose required to completely block PERK autophosphorylation is 30 nM under extreme ER stress conditions [121]. It was shown that the PERK-ATF4-CHOP pathway is involved in stress-induced neuronal damage in the hypothalamus. The PERK phosphorylation inhibitor GSK2606414 significantly decreased ATF4, CHOP and CHOP mRNA expression, blocking activation of the downstream ATF4-CHOP signaling pathway [122]. GSK2606414 treatment also inhibited RANKL-induced MAPK and NF-κB pathway activation and maintained osteoblast homeostasis [123]. The second-generation inhibitor GSK2656157 is an ATP-competitive inhibitor of PERK with an IC50 of 0.9 nmol/L. GSK2656157 is highly selective for PERK with an IC50 value of >100 nmol/L against 300 kinases [98], and there are efforts to improve the physical properties and pharmacokinetics by decreasing inhibitor lipophilicity [124]. This compound showed antitumor effects in immunocompromised mouse xenograft models of multiple myeloma and pancreatic cancer [98], and the inhibitory effect was independent of the inhibition of eIF2α phosphorylation [125]. In contrast, investigators also identified that the potent salubrinal derivative Sal003 could attenuate eIF2α phosphorylation without inhibiting PERK [126], and this agent specifically prevented the dephosphorylation of eIF2α by blocking eIF2α phosphatase activity of glioblastoma multiforme [127,128]. In addition, an integrative stress response inhibitor (ISRIB) was reported known as symmetrical bis (ethylene glycol) amide, which binds and activates elongation initiation factor 2β and subsequently releases eIF2α phosphorylation-mediated inhibition of protein translation [129]; this agent has been shown to cause tumor regression and prolong patient survival in patient-derived models of advanced prostate cancer [130]. In addition to these two generations of inhibitors, AMG’44 [131] has not been tested on various cancer models, although it is said to be the highly selective for PERK compared to the more than 300 kinases tested. Kinase inhibitor research for preclinical development is fraught with challenges. The reaction mechanisms usually occur in an unpredictable manner via multiple enzymatic reactions. However, although inhibitors that simultaneously target different kinases sound promising for cancer therapy, there are many difficulties in clarifying the mechanisms.

BIP/GRP78, which is an important molecular chaperone of the UPR, is essential for maintaining ER homeostasis. BIP/GRP78 is also one of the typical markers of tumor cells and is closely associated with the aggressive growth, invasiveness and metastasis of tumors [132], thus, the development of related inhibitors is a potential strategy for the treatment of cancer. Some established GRP78 inhibitors, such as (-)-epigallocatechin gallate (EGCG), bind to the ATP-binding structural domain of GRP78 and block its function [133,134] and have been shown to protect against hormone-related cancers (breast, prostate) [135]. Potassium-3-beta-hydroxy-20-oxo-substituted-5-en-17-α-yl sulfate (PHOS) inhibits the proliferation of CRC cells and induces apoptosis in cancer cells by inhibiting GRP78 activity [136]. In addition, KP1339/IT-139 can deplete key molecular chaperones such as GRP78 and further induce HCT116 cells apoptosis through the specific disruption of ER homeostasis in combination with enhanced KP1339-mediated protein damage [137]. However, in recent years, the novel inhibitor HA15, which targets GRP78, has been reported and is the the main compound of thiazole benzenesulfonamide, which acts by inhibiting ATPase activity [138], shows antitumor effects and can overcome drug resistance in melanoma and breast, pancreatic and adrenocortical cancers [138,139]. In addition, targeting GRP78 with HA15 also resulted in apoptosis in lung cancer cells, which was accompanied by ER stress and autophagy [140]. The inhibitor sFLT01 was shown to inhibit prostate cancer DU145 cell invasion and metastasis by regulating the VEGF/GRP78/MMP2&9 axis [141]. This protein consists of a second immunoglobulin (IgG)-like structural domain of Flt-1, which is fused to the CH3 structural domain of IgG1 Fc by fusion to human IgG1 Fc or IgG1 Fc using the polyglycine linker 9Gly [142] and has been shown to exhibit potent and significant antitumor activity in vivo in numerous preclinical subcutaneous tumor models, including H460 non-small-cell lung cancer, HT29 colon cancer, Karpas 299 lymphoma, MOLM-13 AML (acute myeloid leukemia), 786-O and RENCA renal cell carcinoma (RCC) [143].

Conclusion and future

A growing number of studies have revealed the dual roles of ER stress, especially the UPR, in inflammatory and cancer responses but have not yet elucidated the molecular mechanisms underlying the switch between survival and death, and the branches of the UPR pay multiple roles in regulating the transformation of inflammation to cancer. Targeting the UPR may be a plausible strategy for developing cancer drugs: a high level of basal ER stress and constitutive UPR activation exist in tumor cells [144,145]. However, drugs that exacerbate ER stress/the UPR may also be cytotoxic to cancer cells [138]. Understanding how the UPR regulates the balance of inflammatory and cancer responses is necessary.

Therefore, this review summarizes the different mechanisms and signaling pathways of the transformation of inflammation to cancer and UPR-targeting drugs that inhibit cancer development via the three branches of the UPR (Figure 1). IRE1 mediates tumorigenesis through the STAT, PI3K/AKT/mTOR, and NOTCH signaling pathways with the participation of multiple inflammatory factors, especially IL-6, IL-10, and VEGF. Its first-generation inhibitors are ATP-competitive IRE1α kinase inhibitors and RNase attenuators, and second-generation inhibitors include a number of compounds that directly target the structural domain of RNase. Another important receptor, PERK, not only induces apoptosis with downstream CHOP and ATF4 but also allows eIF2α to interact with NF-κB and STAT3 to transmit stress signals, and inhibitors of PERK inhibit PERK and eIF2α phosphorylation, ATF4 translation and CHOP mRNA expression. In addition, inhibitors of BIP/GRP78, EGCG, PHOS, KP1339/IT-139, and HA15 have shown antitumor effects in a variety of cancer models through targeted depletion (Table 1). Overall, IRE1 and PERK are potent targets for the treatment and prevention of tumors, and targeted studies of both branches may yield promising therapeutic strategies.

Figure 1.

The role and therapeutic implication of endoplasmic reticulum stress in inflammatory cancer transformation. UPR exerts pro-survival and pro-death effects by the dual role of in inflammatory and cancer responses. The mechanisms and signaling pathways of the transformation of inflammation to cancer and UPR-targeting drugs that inhibit cancer development via the three branches of the UPR. IRE1 mediates tumorigenesis through the STAT, PI3K/AKT/mTOR, and NOTCH signaling pathways with the participation of multiple inflammatory factors, especially IL-6, IL-10, and VEGF. IRE1 and PERK are potent targets for the treatment and prevention of tumors. Targeted therapies for UPR response inhibits cancer progression. IRE1α: inositol-requiring enzyme 1; PERK: protein kinase RNA-like endoplasmic reticulum kinase; ATF6: activating transcription factor 6; XBP1: X-box binding protein 1; GRP78: glucose regulatory protein 78 κDa; SREBP: Sterol regulatory element binding protein; IL-6: Interleukin-6; PI3K: Phosphatidylinositol 3-kinase; ATF4: activating transcription factor 4.

Table 1.

Candidates to target UPR for cancer therapy

Candidates	Molecular formula	Therapeutic effects associated with ER stress	Indications	References
Sunitinib	C22H27FN4O2	Inhibits VEGFR and PDGFR, inhibits autophosphorylation and the subsequent IRE1α RNase activation	Renal cell carcinoma	[104-106]
B-I09	C16H17NO5	Inhibits endogenous IRE1α oligomerization, in vivo XBP1 mRNA cleavage and ER-localized mRNA decay	B cell chronic lymphocytic leukemia	[109]
STF-083010	C15H11NO3S2	Forms Schiff bases with Lys907and hydrogen bonds with Tyr892 in the IRE1 RNase structural domain	Multiple myeloma	[97,99]
4μ8C	C11H8O4	Forms Schiff bases with Lys907and hydrogen bonds with Tyr892 in the IRE1 RNase structural domain	Rheumatoid arthritis	[97,108,110,114]
Toyocamycin	C12H13N5O4	Inhibits endogenous IRE1α oligomerization, in vivo XBP1 mRNA cleavage and ER-localized mRNA decay	Multiple myeloma	[97]
MKC-3946,	C21H20N2O3S	Inhibits endogenous IRE1α oligomerization, in vivo XBP1 mRNA cleavage and ER-localized mRNA decay	Multiple myeloma, triple-negative breast cancer, glioblastoma multiforme	[97,109,115]
MKC-8866	C18H19NO7
Methotrexate	C20H22N8O5	Inhibits IRE1 RNase activity	Glioblastoma multiforme	[120]
Cefoperazone	C25H27N9O8S2	Inhibits IRE1 RNase activity	Glioblastoma multiforme	[120]
Folinic acid	C20H23N7O7	Inhibits IRE1 RNase activity	Glioblastoma multiforme	[120]
Fludarabine phosphate	C10H13FN5O7P	Inhibits IRE1 RNase activity	Glioblastoma multiforme	[120]
GSK2606414,	C24H20F3N5O	Inhibits PERK, inhibits eIF2α phosphorylation (or not), ATF4 translation and CHOP mRNA expression	Hypothalamic neuronal injury, bone loss, multiple myeloma, pancreatic adenocarcinoma	[98,122,123,125]
GSK2656157	C23H21FN6O
Sal003	C18H15Cl4N3OS	Blocks eIF2α phosphatase activity	Glioblastoma multiforme	[126,127]
ISRIB	C22H24Cl2N2O4	Releases eIF2α phosphorylation-mediated inhibition of protein translation	Prostate cancer	[129,130]
AMG’44	C34H29ClN4O2	Inhibits PERK	/	[131]
EGCG	C22H18O11	Binds to the ATP-binding structural domain of GRP78 and block its function	Breast cancer, prostate cancer	[133-135]
PHOS	C21H31KO6S	Inhibits GRP78 activity	Colorectal cancer	[136]
KP1339/IT-139	C14H12Cl4N4NaRu	Depletes GRP78 and disrupts ER homeostasis specifically	Colorectal cancer	[137]
HA15	C23H22N4O3S2	Inhibits ATPase activity	Melanoma, breast cancer, pancreatic cancer, adrenocortical carcer, lung cancer	[138-140]

Disclosure of conflict of interest

None.