Protease-Activated Receptor-2 Promotes Metastasis: An Emerging Therapeutic Target.

Introduction
Metastasis is the spread of primary tumor cells to other sites in the body where they establish secondary growths and is the leading cause of cancer-related deaths(1,2). While primary tumors typically develop from the acquisition of oncogenic driver mutations that sustain localized, uncontrolled cell proliferation, and often respond well to surgical removal followed by radiation or chemotherapies(2,3), metastasized tumor cells have acquired new functional abilities. These can include enhanced migratory and invasive capacities, altered cell signaling pathways that promote survival and immune evasion, and the capacity create a tumor-supportive environment in distant tissues that nurtures successful colonization(1,2).
Metastatic tumors have long been associated with abnormally active proteolytic enzymes, which play a crucial role in cleaving specific protein substrates that are implicated in multiple functions required for successful metastasis(1,4,5). Proteases in the pericellular microenvironment communicate directly with cells through a unique class of G protein-coupled receptors (GPCRs) known as protease-activated receptors (PARs), of which there are 4 family members, PAR-1, PAR-2, PAR-3, and PAR-4(6). These transmembrane GPCRs are activated by proteolytic cleavage of their extracellular N-terminus, serving as sensors of the tumor microenvironment (TME) and initiating cellular responses through multiple context-dependent intracellular signaling cascades that regulate physiological and pathological processes(7,8). While PAR-2 activation can be triggered by a broad range of serine proteases, the specific sequence determinants responsible for distinct signaling outcomes remain poorly understood.
PAR-2 is widely expressed by many cell types including epithelial cells (e.g. skin, gastrointestinal tract, lung), endothelial and smooth muscle cells of the vasculature, hematopoietic cells of both the innate and adaptive immune system, and sensory neurons. Under physiological conditions, PAR-2 participates in regulating immune responses, maintaining epithelial barrier functions, transmission of pain sensation, and support of vascular functions(7,9–11). However, when over-expressed and hyperactivated, PAR-2 is associated with chronic inflammation, thus pharmacological inhibitors or modulators that target PAR-2 are being actively developed as potential therapeutics. Here we highlight recent evidence in support of targeting PAR-2-associated chronic inflammatory signaling for drug development in the treatment of metastatic cancers.
This review discusses PAR-2 proteolytic cleavage and activation, the physiological and pathological roles of PAR-2 in inflammation and chronic pain, PAR-2’s mechanistic involvement in cancer metastasis and preclinical evidence supporting a causal role for PAR-2 in metastatic progression. Lastly, we provide a forward-looking section on how next-generation PAR-2 modulators can be harnessed to target metastatic vulnerabilities.

Proteolytic activation of PAR-2
Of the over 800 human GPCRs, PAR-2 is one of only four that are activated by proteolytic cleavage that generates a self-activating ligand, rather than by exogenous soluble agonists present in the extracellular milieu. Proteolytic cleavage of the PAR-2 N-terminus unmasks a new N-terminus that serves as an endogenous tethered ligand (TL) that binds intramolecularly, inducing a PAR-2 conformation that triggers intracellular signal transduction(12,13). The classical PAR-2 activating protease, trypsin, cleaves at the canonical cleavage site of human PAR-2 at Arg36↓Ser37(12–14). Using cryo-electron microscopy a recent structure of PAR-2 cleaved after Arg36 revealed that the exposed TL binds to a shallow orthosteric pocket comprised of extracellular residues of transmembrane (TM) helices TM5-7, and extracellular loop 2 (ECL2), and induces conformational changes particularly in TM7 in the activated receptor(15). Within the past three decades, a large repertoire of proteases have been reported to cleave PAR-2 at the canonical activation site, including soluble plasmin, mast cell tryptase, coagulation factors thrombin, FVIIa and FXa, several kallikreins(7), and the membrane-anchored serine proteases (MASPs) matriptase and testisin(16,17) (Fig. 1). PAR-2 may also be cleaved at alternate N-terminal sites by proteases that stabilize different receptor conformations, known as biased activators (Fig. 1). Desensitization or disarming of PAR-2 can also occur due to non-canonical proteolytic cleavage C-terminal to the TL, thereby removing or shedding the TL(7,18) (Fig. 1). Synthetic peptides, known as PAR-2 activating peptides (PAR-2AP), can directly bind to the orthosteric site and cause PAR-2 signaling in the absence of proteolytic cleavage, mimicking the effects of proteases on various signaling pathways(7,18).

PAR-2 signaling
The molecular pathways and intracellular signaling responses induced by PAR-2 activation predominantly involve coupling to G-proteins and recruitment of β-arrestin, which have been recently reviewed in detail(7,13,19). Canonical PAR-2 cleavage can stimulate the recruitment and activation of various G-proteins containing different Gα subtypes (Gαq, Gαq/11, Gα12/13, Gαi/o and Gαs), leading to a vast array of possible intracellular signaling responses(7,13). For example, PAR-2 coupling to Gαq drives phospholipase C-β activation, IP3-mediated calcium mobilization, protein kinase C (PKC) activation, and downstream transcriptional programs involved in cytokine production, epithelial barrier regulation, and smooth muscle contraction. While coupling to Gαi/o activates the PI-3 kinase/Akt (PI3K/Akt) axis and MAPK/ERK signaling, contributing to inflammatory gene expression, chemotaxis, proliferation, and survival responses. Whereas coupling to Gα12/13 stimulates RhoA and cytoskeletal reorganization, facilitating cell shape changes, motility, and barrier remodeling(7,13,19).
Cleavage of PAR-2 by biased proteases at different sites can produce distinct TLs and induce diverse receptor conformations that activate selective signaling profiles, which has been comprehensively reviewed in(20,21). For example, unlike activation by trypsin at the canonical cleavage site of PAR-2, elastase cleavage at S67↓V68 (Fig. 1) does not initiate PAR-2 mediated calcium signaling or trigger β-arrestin recruitment and receptor internalization, but triggers PAR-2-dependent ERK phosphorylation by a pathway involving Gα12/13-mediated activation of Rho kinase(22,23). Similarly, cleavage by the cysteine protease cathepsin S at E56↓T57 (Fig. 1), activates Gαs and cAMP generation, but does not stimulate calcium mobilization, activate ERK1/2 or β-arrestin mediated PAR-2 endocytosis(20,24). Biased signaling can also be influenced by the presence of co-receptors or co-associated proteins, as shown for the Tissue Factor (TF)-FVIIa-FXa robust induction of pro-inflammatory cytokines in response to endotoxin, which requires an association with endothelial protein C receptor (EPCR)(25). It remains to be determined whether the localization of PAR-2 to specific membrane microdomains such as lipid rafts may influence signaling outcomes (26,27). The cytotoxic lymphocyte protease granzyme K was also recently reported to activate PAR-2 at the canonical cleavage site, but activate biased β-arrestin dependent signaling directed by mechanisms that are as yet unclear (28).
Termination of PAR-2 G-protein signaling after activation by trypsin at the surface of cells occurs by phosphorylation of Ser/Thr residues predominantly located in the C-terminal tail, which stimulates the recruitment of the multifunctional adaptor protein β-arrestin to uncouple G-protein signaling. This triggers receptor internalization via dynamin/clathrin-mediated endocytosis(29–31) and leads to the activation of c-Cbl, an E3 ligase that directly ubiquitinates PAR-2 which is targeted to lysosomes for degradation(29,32,33). Signal termination was also reported to be mediated by the E3 ubiquitin ligase ring finger 43 (RNF43) in colon cancer cells, which leads to PAR-2 ubiquitination after its activation, inducing loss of cell surface PAR-2, turnover and termination of signaling(34). Once internalized, PAR-2 can continue to signal via β-arrestin dependent mechanisms. PAR-2 is capable of forming an endosomal signaling scaffold containing β-arrestin-1/2 that induces sustained ERK activation(35–37) and may also involve activation of Gα subunits(38). β-arrestin-2-dependent ERK activation is believed to mediate the proinflammatory effects of PAR-2 by activating transcription factors such as NF-κB and the production of IL-4, IL-6, IL-13 and TNF-α(39,40) and was shown to be essential for serine protease-induced airway inflammation in a murine allergen-induced asthma model(41). Sustained endosomal ERK signaling after PAR-2 activation contributes to the persistent hyperexcitability of pain receptors associated with irritable bowel syndrome(38).

PAR-2 - a mediator of inflammation, pain and tissue repair
Inflammation is a fundamental defensive innate response to infection and tissue injury. Through numerous in vitro and in vivo studies, it is now well established that PAR-2 activation promotes inflammatory responses (reviewed in(42)). Early studies showed that that the onset of inflammation was delayed in PAR-2 deficient mice(43). PAR-2 activation on immune cells leads to induced expression and release of proinflammatory cytokines, as well as upregulation of leukocyte adhesion molecule intracellular adhesion molecule-1 (ICAM-1)(44,45). Many of these proinflammatory effects are believed to be mediated by the Ca2+/IP3/PKC signaling axis, leading to the activation and nuclear translocation of NFκB(40,46). In vivo studies have also directly linked PAR-2 activation with the activation or maturation of dendritic(47) and T cells(48). PAR-2 also plays a role in the innate response to bacterial pathogens by transactivation of toll-like receptor-4 (TLR4) that promotes proinflammatory NFκB signaling(49), as demonstrated using endotoxemia models in PAR-2 deficient rodents(50). While this evidence links PAR-2 to acute inflammatory responses, chronic inflammation occurs when an immune response goes unchecked and fails to resolve. In this context, over-expression and activation of PAR-2 is linked to the pathogenesis of chronic inflammation in the pathogenesis of several diseases including asthma(51), inflammatory bowel diseases(52), autoimmune diseases (diabetes, encephalomyelitis)(53,54) and inflammatory skin conditions(55).
Significant accumulating evidence implicates PAR-2 in the transmission of neuropathic pain (reviewed in(10)). PAR-2 is expressed on peripheral nerve endings, by nociceptive neurons, a subset of neurons present in the dorsal root ganglia (DRG) that transmit pain sensation(56), and studies in mouse models have shown that PAR-2 activation induces neuropathic pain and hyperalgesia(57,58). PAR-2 activation sensitizes transient receptor potential (TRP) ion channels including TRPV1, TRPV4 and TRPA1 to induce pain transmission(10,59). Animal studies have also demonstrated the contribution of PAR-2 to neuropathic pain in a variety of pathologies including chronic cancer pain(60), migraine(61), and chemotherapy peripheral neuropathy (62), where PAR-2 inhibition is being investigated as a therapeutic target for chronic pain.
While excessive PAR-2 activation in the immune system can induce inflammatory and autoimmune-based tissue injury, several studies have shown that PAR-2 expression outside of the immune compartment is important for tissue regeneration and repair in multiple pathological systems (reviewed in(11)). For example, in the gastrointestinal (GI) tract, PAR-2 deficiency exacerbates colitis and impairs colonic regeneration in several experimental murine models(63). Likewise, PAR-2 expression in murine pancreatic β-cells can be protective against the onset of autoimmune diabetes(64), and its expression in hepatocytes is essential for liver recovery after CCl4 induced hepatitis(64).
PAR-2 has emerged as a multifaceted modulator of immune and tissue responses in chronic inflammation. It appears that the effect of PAR-2 activation lies in its cell or tissue specific location: PAR-2 activation in the immune system intensifies inflammatory tissue injury, whereas its activation in damaged tissue promotes repair and tissue regeneration(11). Unresolving chronic inflammation is strongly linked to the development and progression of cancer, exemplified by clinical studies showing that long term use of anti-inflammatory drugs reduces cancer incidence and mortalities (reviewed in(65)). Chronic inflammation can lead to epigenetic alterations that are required for cancer initiation and progression, induce the accumulation of growth factors that can drive cancer progression(66,67) and modulate multiple oncogenic signaling pathways including NFκB, K-RAS, and p53(65,66).

PAR-2 in cancer and metastasis
PAR-2 expression is increased in many primary and advanced cancers and is often correlated with poor prognosis (68,69). Analysis of RNA-seq data for PAR-2 (F2RL1) gene expression in normal and tumor samples across 22 different human cancer types (Fig. 2) shows that in the majority of tumor types, PAR-2 expression is significantly elevated compared to normal tissue counterparts. Kaplan-Meyer survival analysis (KMPlotter) of the correlation between the expression of PAR-2 mRNA and patient survival for a range of tumor types (Fig. 3) indicate that high PAR-2 expression is correlated with poor survival. In women’s gynecological cancers and in cervical carcinomas there is a positive correlation between PAR-2 expression and metastatic characteristics in clinical samples, where upregulated PAR-2 is detected in lymphatic metastases compared to primary tumor tissue, localizing to cells in the invasive front(70,71). Similarly, in hepatocellular carcinoma tissues, high PAR-2 expression is associated with advanced stage and high microvascular invasion rate, which correlates with poor patient survival(72). In contrast, in colorectal cancer (CRC) it has been reported that high PAR-2 is predictive of a favorable prognosis(73), consistent with our analysis (Fig. 3). However, others have reported that high PAR-2 expression is detected in primary colorectal tumors of patients with associated liver metastases which correlated with poor overall survival(74). While global RNAseq data analysis can be used to assess broad changes in gene expression, its usefulness becomes limited without identifying cell-type specific PAR-2 expression. PAR-2 is not simply oncogenic or tumor suppressive, but rather acts as a powerful biosensor of active proteases in the microenvironment and its effects are dependent on the nature of its proteolytic activation in the TME. Thus, the efficacy of targeting PAR-2 for inhibition of metastasis will likely be tumor, cell type and tissue selective. Single-cell RNA sequencing and spatial transcriptomics may be more accurate for outcome predictions than bulk tumor tissue analysis.
There is limited data regarding mechanisms that regulate PAR-2 expression and its dysregulation in cancer. Genetic variations, specifically single nucleotide polymorphisms (SNPs) within the PAR-2 gene (F2RL1) that affect PAR-2 expression levels have been identified in patients with asthma(75,76). Increased expression of PAR-2 in cancer cells has been associated with demethylation of the PAR-2 promoter. PAR-2 promoter hypomethylation causes increased gene transcription in lung adenocarcinomas, where promoter hypermethylation is correlated with a better prognosis(77). PAR-2 promoter methylation is also reduced in gastric cancer where PAR-2 is over-expressed(78). Since PAR-2 activation can stimulate its own transcription(79), the elevated PAR-2 activating protease content in the tumor microenvironment may also promote PAR-2 upregulation in a positive feed-back mechanism. Interestingly, PAR-2 and the PAR-2 activating protease mesotrypsin (PRSS3/Trypsin IV) were found to be significantly upregulated at the mRNA and protein level in lung tumor cells exposed to sheer stress, as is encountered by circulating tumor cells in the vasculature(80), identifying a novel but uncharacterized mechanism of PAR2 upregulation.
In multiple murine experimental metastasis models, expression of PAR-2 in tumor cells promotes metastasis and depletion/inhibition reduces metastasis. Tumor cell PAR-2 activation in a murine model of spontaneous metastatic B16 melanoma, enhances metastasis to the lungs(81). PAR-2 deficiency in mice also delays tumor progression to invasive carcinoma in the polyoma middle T (PyMT) model of spontaneous breast cancer, leading to reduced lung metastasis(82). Recently, CRISPR/Cas9 mediated deletion of PAR-2 in a human breast carcinoma cell line was shown to inhibit lung metastases in vivo(83). Moreover, knock-down of tumor cell PAR-2 reduced the incidence of spleen-liver metastases in orthotopic models of murine CRC(34,74) and in a metastasis model of hepatocellular carcinoma(72). PAR-2 deficiency in mice also inhibited the development of colon adenocarcinomas in the chemically induced AOM-DSS model of colon cancer(34). In contrast, it was recently reported that PAR-2 knock-down in a syngeneic murine carcinoma cell line promoted tumor growth by impairing anti-tumor immunity(73). In agreement with these findings, PAR-2 and its activating MASP, matriptase, which are upregulated in ovarian cancers, together promote ovarian cancer metastasis in a murine metastasis model(16,84).

PAR-2 promotes metastatic processes in tumor cells
Recent data suggest that components of PAR-2-associated chronic inflammatory signaling create a tumor and metastasis friendly environment and could be targeted to suppress cancer growth, reduce metastasis, and enhance the effectiveness of existing anti-cancer therapies. Below we discuss the ways in which PAR-2 activation stimulates several essential steps required for successful tumor cell metastasis (summarized in Fig. 4).

Increased tumor cell motility and invasion.
Acquiring the ability to migrate and invade into the extracellular matrix and surrounding tissues is considered one of the first steps in the metastatic cascade(1). Numerous studies have shown that PAR-2 activation by proteases present in the TME stimulates signaling that promotes tumor cell migration and invasion in multiple cancer types. Consistent with the now well recognized role of β-arrestins in the promotion of tumor cell migration and invasion(85), a PAR-2-activated endosomal signaling complex was identified in breast cancer cell lines composed of β-arrestin, PAR-2, Raf and ERK1/2, that localizes to pseudopodia and prolongs ERK1/2 signaling at the leading edge of migrating cells(35,37). Additional studies confirmed the importance of both β-arrestin-1 and 2 for PAR-2-mediated ERK activation in pseudopodia and migratory properties of breast cancer cells(35,86), involving β-arrestin-dependent activation of the actin filament-severing protein cofilin(87). PAR-2-β-arrestin signaling was also central to the in vitro migration and invasion of the ovarian cell line OV90, comprising PAR-2 transactivation of the epidermal growth factor receptor (EGFR) through a combination of Gαq/11, Gα12/13, and β-arrestin1/2(88), while in lung adenocarcinoma, PAR-2 was found to induce migration through a β-arrestin-1-mediated activation of p38 MAPK(89). Breast cancer cell migration and invasion induced through PAR-2 activation by the coagulation proteases FVIIa and FXa is associated with phosphatidylinositol hydrolysis and ERK1/2 activation(90,91). Further it was demonstrated that FVIIa-PAR-2 induced breast cancer cell migration is dependent on PI3K-Akt activation, and that upregulated β-catenin is also essential for this process, likely via enhancing transcription of metastasis-related genes including cMyc, MMP7 and MMP14(91). Stimulation of PAR-2-mediated migration and invasion in other cancer types has also been shown dependent on activation of MEK/ERK and PI3K/Akt signaling pathways including esophageal(92), oral squamous cell carcinoma(93), renal cell carcinoma(94) and ovarian(84) cancers.
In multiple cell types, PAR-2 activation initiates the synthesis and secretion of the gelatinases MMP2 and MMP9(95), which mediate extracellular matrix degradation to facilitate tumor cell invasion of surrounding tissues(96). Production of MMP2 in response to PAR-2 activation by TF-FVIIa or trypsin in breast cancer cells was shown to be induced via PI3K-Akt-NF-ĸB pathway, which acts in an autocrine manor to promote cell invasion(97). Similarly, the membrane-anchored serine protease matriptase, a potent PAR-2 activator, was shown to drive dissemination of ovarian cancer spheroids by a PAR-2/PI3K/Akt/MMP9 signaling axis(84), and matriptase knock-down also inhibited PAR-2/PLCγ2/PKC-mediated invasion of MCF-7 breast cancer cells(98).

Promotion of an ‘epithelial-to-mesenchymal’ transition (EMT) facilitating tumor cell dissemination.
A developmental regulatory program known as EMT is co-opted by metastatic cells, and enables new capabilities that are essential for facilitating dissemination to distal sites(99) - one of the initial steps of the metastatic cascade. During EMT, epithelial characteristics are repressed, mesenchymal characteristics are acquired, and cells change shape, losing apical-basal polarity and acquiring an invasive phenotype. EMT is orchestrated by EMT-inducing transcription factors (e.g. Snail, Slug, Twist) and by activation of multiple signaling pathways, e.g. PI3K/PDK1/Akt and Ras-GTP/Raf/MEK/ERK/MAPK due to environmental cues such as activation of receptor tyrosine kinases(100). As previously noted, PAR-2 can activate ERK and Akt signaling pathways, both of which have been shown to induce EMT(101,102). Loss of E-cadherin expression and function is a characteristic of EMT, and it has been shown that activated PAR-2 in lung and ovarian cancers suppresses E-cadherin levels, which in turn decreases cell-cell adhesion and enhances cell migratory behavior(84,89). Similarly, in a study in which PAR-2 expression in hepatocellular carcinoma (HCC) cells was found to promote experimental metastasis in vivo, PAR-2 knockdown enhanced E-cadherin levels and reduced expression of the mesenchymal markers N-cadherin, EGR-1 and Snail(72). In recent lung cancer studies, PAR-2 was shown to promote Slug-mediated EMT in lung adenocarcinoma cells, through a pathway that depended on ERK2 stabilization of Slug via suppression of GSK3β activity(89). Bioinformatics data analysis of human lung cancers revealed that PAR-2 gene expression correlates with expression of EMT markers, the mesenchymal cytoskeletal protein vimentin, and Twist1 and Slug transcription factors(103). This study also showed that PAR-2 knockdown in lung cancer cell lines suppressed the expression of Snail, Slug and Twist and resulted in loss of vimentin, at the same time increasing expression of E-cadherin(103), further implicating PAR-2 in the promotion of EMT in cancers.

Stimulation of angiogenesis to support tumor growth.
An essential modification of the TME required for successful tumor metastasis is the production of factors that stimulate angiogenesis, to support the developing metastatic lesion with oxygen and nutrients(99). In vitro studies utilizing cancer cell lines derived from various tumor types have shown that PAR-2 activation plays a critical role in the induction of the potent angiogenic growth factor vascular endothelial growth factor (VEGF). In breast cancer cells, PAR-2 activation by FVIIa induces VEGF expression via a p38 MAPK-ERK1/2 dependent pathway(104). Similarly, in human glioblastoma and gastric cell lines, PAR-2 dependent VEGF production is ERK1/2 dependent(105,106). PAR-2 activation in pancreatic cancer cell lines stimulates synthesis and secretion of TGF-α and VEGF-A, which enhances endothelial cell tube formation in an in vitro angiogenesis model(107). Similarly, in lung adenocarcinoma cells PAR-2 over-expression stimulates VEGF production and promotes in vitro angiogenesis which was shown to be via an EGFR-dependent pathway(108). Other in vitro studies have shown that activation of PAR-2 also induces other pro-angiogenic factors such as C-X-C motif chemokine ligand 1 (CXCL1)(109) and IL-8(110). In vivo studies demonstrating a causal link between PAR-2 expression and enhanced tumor-induced angiogenesis are limited. However, using an orthotopic model with PyMT PAR2−/− mammary carcinoma cells, PAR-2 reconstitution was shown to increase angiogenic vessel density and enhance tumor growth when implanted in the mammary fat pad(111).

Modulation of the tumor microenvironment.
PAR-2 may also shape the immune cell composition of the TME. In vitro and in vivo studies have shown that TF-VIIa-PAR-2 signaling in breast cancer cells modulates the TME through the production of immunomodulatory cytokines such as GM-CSF and M-CSF(109), important regulators of myeloid cell recruitment and lymphangiogenesis, respectively(112). In vitro PAR-2 activation on breast cancer cell lines using PAR2AP stimulates production of G-CSF(113), a mobilizer of bone-marrow derived cells predominantly of the neutrophilic lineage, which have been shown to promote tumor metastasis in murine models(114). Neutrophil extracellular traps (NETs) are reported to induce TF expression in breast cancer cells resulting in activation of the TF/PAR-2 pathway to promote the expression of metastasis-related genes(115). Conversely, NETs containing elastase have been reported to disarm PAR-2 on macrophages to downregulate CD24 and enhance tumor phagocytosis to inhibit CRC liver metastasis(116). PAR-2 signaling is also involved in shaping the immune environment by repressing type I IFN secretion during CRC liver metastasis(117). In the tumor microenvironment, PAR-2 is also significantly upregulated in cancer-associated fibroblasts (CAF) in close contact with tumor cells(118) as assessed by immunohistochemical and in situ hybridization techniques. PAR-2 activation on fibroblasts promotes proliferation and collagen production (reviewed in(69)) which may contribute to the pro-tumorigenic activities of CAFs in the TME, including the development of fibrosis which is implicated in resistance to anti-tumor therapies(119). Apart from tumor cells and fibroblasts, studies identifying cell type specific expression of PAR-2 within the TME has been limited due to unreliable PAR-2 specific antibodies(120). Recently, using single cell RNAseq analysis of human colorectal tumors PAR-2 was found to be expressed in myeloid populations, where tumor-associated dendritic cell PAR-2 expression was shown to be important for priming the anti-tumor immunity of CD8+ T cells(73).

Promotion of immune evasion.
Studies show that metastatic tumor cells are much more resistant to being eliminated by the immune system than recently transformed cells and show enhanced resistance to immune-based therapies(121,122). PAR-2 induction of inflammatory cytokines, chemokines and immunomodulatory regulators likely creates a chronic inflammatory environment that is favorable to tumors by stimulating recruitment of immunosuppressive immune cells and the upregulation of immune checkpoint inhibitors to promote tumor evasion from the immune system(123). PAR-2 activation by TF/FVIIa was recently shown to promote immune-evasion in a murine model of breast cancer, by increasing the expression and stability of the T-cell inhibitory receptor programmed death-ligand 1 (PD-L1)(124). FVIIa-activation of PAR-2 trascriptionally upregulates PD-L1 through activation of Hippo signaling, which reduces lymphocyte proliferation and stimulates lymphocyte apoptosis, and in vivo, tumor cell deficiency of TF/PAR-2 expression enhanced the therapeutic efficacy of anti–PD-1 antibodies(124). In another recent study on hepatocellular carcinoma (HCC), PAR-2 activation by coagulation factor FXa was also shown to increase PD-L1 expression(125). Further, in vivo treatment of mice with the FXa inhibitor rivaroxaban was found to reduce HCC metastasis and enhance anti-PD-1 immunotherapy, and in a clinical trial, rivaroxaban improved the response rate of patients with HCC to immune checkpoint inhibitors(125) and enhanced overall patient survival. In other studies, activation of PAR-2 on tumor-associated macrophages by FXa was shown to induce an immune-evasive macrophage phenotype that promoted breast cancer progression in murine models, whereas treatment with rivaroxaban enhanced anti-tumor immunotherapy(126).

Metastatic colonization.
Colonization refers to the growth of metastatic tumor cells in a secondary tissue(1). Cell proliferation in the metastatic lesion generally follows similar mechanisms as the primary tumor. PAR-2 is a mitogenic factor; cancer cell lines of various tissue origins proliferate upon exposure to trypsin which is dependent upon PAR-2 expression(127),(128). By overexpression and stable knockdown studies, PAR-2 was also shown to promote tumor cell proliferation and growth in vivo in HCC tumors via ERK activation(72). Likewise, silencing of PAR-2 via microRNA suppressed the growth and proliferation of gastric carcinoma cells through inhibition of ERK activation(129), and inhibition of PAR-2 using a novel PAR-2 antagonist (9a) was also found to inhibit the proliferation of breast cancer cell lines(130). Tumor cells often contain activating mutations in EGFR(131), and transactivation of EGFR by PAR-2 activation has been shown to promote cell proliferation in various tumor cell types(88,128,132). In ovarian cancer, MEK-ERK1/2 signaling activated by PAR-2/EGFR transactivation activates several transcription factors including the key stimulators of cell proliferation; FOS, MYC, and STAT3(88). In colon cancer cells, PAR-2 was shown to promote cell proliferation by indirectly enhancing the expression of the G1/S checkpoint regulator cyclin D1(133). In human gastric carcinoma cells, PAR-2 activation by trypsin stimulates adhesion to fibronectin via integrin α5β1, which stimulates proliferation via a Src kinase dependent mechanism(134).

Promotion of chemoresistance.
PAR-2 expression and activation promotes resistance to chemotherapeutics commonly used to treat various tumor cell types. In in vitro studies on the cervical cancer cell line CASKI, PAR-2 activation by PAR-2AP or FVIIa was found to increase resistance to the DNA-intercalating drug cisplatin via a mechanism that was dependent on EGFR transactivation and likely involved the induction of cyclooxygenase-2 (COX-2)(135). In the colon cancer cell line HT29, PAR-2 activation by trypsin or PAR-2AP also attenuated doxorubicin-induced cell death which is associated with MEK1/2-ERK1/2 upregulation of the anti-apoptotic survival proteins MCL-1 and Bcl-xL, and PAR-2 deletion or inhibition sensitized cells to doxorubicin-induced cell death(136). Several recent studies have also demonstrated a role for PAR-2 in the resistance of lung cancer cells to EGFR targeting receptor tyrosine kinase inhibitors (TKIs). PAR-2 knockdown or inhibition in lung cancer cell lines increases sensitivity to the clinically used EGFR tyrosine kinase inhibitor (TKI) gefitinib(103), and in murine models, PAR-2 inhibition using the pepducin P2pal-18S (see below) was shown to sensitize non-small cell lung carcinoma (NSCLC) cells previously resistant to the EGFR TKIs gefitinib and osimertinib, reducing metastatic burden in vivo(137,138).

Targeting PAR-2 signaling for inhibition of metastasis
Pursuing PAR-2 as a drug target for metastatic disease offers several advantages over targeting proteases directly due to its greater specificity, role in signal transduction, versatility, and potential for immune system modulation. The development of metastatic disease requires the acquisition of new functional abilities, including those directed by protease activities(4). PAR-2 activating proteases such as trypsin, matriptase, urokinase-plasminogen activator (uPA), FXa, FVIIa and type II transmembrane serine protease (TMPRSS2) are frequently upregulated in many human cancers(60,84,139,140). However, the direct targeting of such tumor associated proteases as drug targets has been challenging for several reasons. The reaction mechanisms are highly conserved among each protease class and there exist many closely related family members rendering difficulty in developing drugs that target specific proteases. Many potent clinical lead compounds have been unsuccessful either because of lack of specificity or because of an underdeveloped understanding of their promiscuity and their pathophysiological roles(141,142).
The upregulation of PAR-2 in many cancers, its correlation with poor prognoses, and its ability to trigger signaling outcomes that promote metastasis and influence immune responses, suggest that PAR-2 represents an attractive anti-metastatic therapeutic target. In the following sections and summarized in Table 1, we highlight the major classes of PAR-2 antagonists that target different aspects of PAR-2 signaling and that have shown efficacy in preclinical models for the treatment of inflammatory conditions and pain. While studies utilizing PAR-2 inhibitors in preclinical experimental metastasis models remain scarce, these studies demonstrate effective alleviation of PAR-2-mediated pathologies in vivo without adverse side-effects, warranting future studies on their repurposing for treatment of metastatic cancers.

Small molecule antagonists-
The first PAR-2 targeting molecules developed were PAR-2 peptide antagonists that bound to the orthosteric tethered ligand binding site of PAR-2 to block the critical tethered ligand interaction with ECL2(12,143), however they showed poor in vivo stability and bioavailability. The first non-peptide small molecule PAR-2 antagonist ENMD-1068 was shown to inhibit PAR-2 responses in vitro and was successful in reducing joint inflammation in a murine model of arthritis(144), and reducing CCl4-induced liver fibrosis in mice by inhibiting PAR-2-TGFβ signaling(145), however, these effects required administration of very high doses. Two newer non-peptide small molecule antagonists AZ8838 and AZ3451 show good potency and in vivo efficacy(146,147). Both compounds have been shown to inhibit Ca2+ mobilization after activation of PAR-2, as well as attenuate PAR2-induced phosphorylation of ERK1/2 and β-arrestin-2 recruitment(148). In murine models, AZ8838 and AZ3451 significantly reduced PAR-2AP-induced paw swelling, and inhibited mast cell and neutrophil activation and degranulation(148). AZ3451 also showed efficacy in preventing inflammation and cartilage degradation osteoarthritis models in rats(149), and blocked PAR-2 induced resistance of colon cancer cells to doxorubicin-induced cell death in vitro(136). In addition, nanoparticle delivery of AZ3451 in three preclinical models of oral cancer in mice, was shown to significantly reduce oral cancer nociception via inhibition of PAR-2(150). Another small molecule inhibitor of PAR-2, C-391 which inhibits PAR-2 induced calcium mobilization, ERK activation and β-arrestin recruitment, was shown to be effective in preventing pain in murine models of thermal hyperalgesia(151), and lung inflammation in acute allergen induced asthma models in mice(152), but its efficacy preventing tumor progression has not been tested.

Antibody antagonists-
Antibodies have also been generated that target the TL sequence of PAR-2(153). The widely utilized PAR-2 antibody SAM11, blocks trypsin and PAR-2AP mediated signaling and reduces joint inflammation in murine models of arthritis(144,153,154). SAM11 was also effective in suppressing airway inflammation and hyperresponsiveness in allergen-induced asthma models in mice(155,156). Additionally, intranasal administration of SAM11 was effective in reducing lung inflammation and viral load in mice infected with influenza A virus(157). In recent years, AstraZeneca has developed a next generation fully humanized PAR-2 monoclonal antibody, MEDI0618, which shows enhanced pharmacological and pharmacokinetic properties. In murine models, subcutaneous injection of MEDI0618 was shown to be effective in preventing migraine-like pain in multiple preclinical migraine models(146,158), allowing progression into human clinical trials for migraine prevention (NTC05714254, NCT04198558). These Phase I clinical trials concluded that MEDI0618 is safe for use in healthy individuals(158), indicating that targeting PAR-2 may hold promise for the treatment of metastatic cancers.

PAR-2 endosomal inhibitors-
Endosomal targeting of GPCRs is an emerging area in drug development, since sustained signaling from endosomes after GPCR internalization is believed to contribute to chronic pathologies(159). In a recent study, Jimenez-Vargas et al. developed the first endosome-targeted PAR-2 antagonist to determine the contribution of sustained PAR-2 endosomal signaling to the hyperexcitability of pain receptors in irritable bowel syndrome(38). This tripartite PAR2 antagonist (MIPS15479) contains a novel small molecule PAR-2 inhibitor I-343 conjugated to polyethylene glycol (PEG) and the transmembrane lipid cholestanol(38), modifications that have been shown to link other GPCR inhibitors to the plasma membrane, and which are then retained inside early endosomes after GPCR internalization(159). MIS15479 was shown to inhibit the sustained (but not initial) nociceptor excitability induced by trypsin or IBS-derived proteases in mouse neurons in vitro, demonstrating that endosomal PAR-2 signaling mediates persistent pain signaling neurons(38). In addition, unconjugated I-343 was able to block the chronic excitability of colonic neurons that occurs after resolution of TNBS-induced colitis. These findings indicate further development of endosome-targeted PAR-2 inhibitors for in vivo applications is warranted.

Biased PAR-2 antagonists-
Recent data demonstrate that it is possible to modulate PAR-2 signaling in ways that could target the vulnerabilities of metastatic tumors while avoiding inhibition of physiological PAR-2 functions. The complete ablation of PAR-2 signaling has the potential to inhibit necessary functions of PAR-2 in acute immune responses, pain sensation and tissue healing. The development of biased antagonists that target specific signaling pathways could inhibit detrimental effects while sparing those that mediate beneficial outcomes(160). The PAR-2 small molecule biased antagonist, GB88, inhibits Gαq/11 activation, calcium release and proinflammatory signaling(161–163), while at the same time stimulating ERK activation and β-arrestin recruitment(21,146). GB88 was shown to be effective alleviating PAR-2-mediated inflammation in several rodent models including acute paw edema(161,162), experimental colitis(164) and collagen induced arthritis(165). In in vitro studies, the small molecule PAR-2 antagonist, I-287, was demonstrated to selectively target Gαq and Gα12/13 activation without affecting Gαi/o signaling or β-arrestin recruitment and PAR-2 internalization(166). In a murine model of complete Freund’s adjuvant (CFA)-induced paw edema model in mice, oral administration of I-287 was shown to be equally as effective in reducing inflammatory effects as NSAIDs(166). Another recently developed biased antagonist, C781, selectively inhibits β-arrestin recruitment and MAPK signaling without affecting Gαq activation and calcium responses(62,167). C781 attenuates nociception in mouse models of protease-induced pain, and also airway hyperresponsiveness and inflammation in an acute allergen-challenged mouse model(62,167). These studies reveal that PAR-2 biased antagonists can be useful tools for identifying the PAR-2 activated pathways underlying disease pathologies in vivo. Further development of selective PAR-2 antagonists and understanding the pathways that they inhibit will be important in the design of future PAR-2 targeted drugs for treatment of metastatic disease.

Pepducins-
Pepducins are cell penetrating lipidated synthetic peptides that mimic the amino acid sequence of the intracellular loops or C-terminus of GPCRs to block interactions with G-proteins, where the attached lipid moiety facilitates peptide movement through the lipid bilayer and anchors the peptide to the cytoplasmic side of the cell membrane(168). The PAR-2 targeting pepducin P2pal-18S (also known as PZ-235) is an N-palmitoylated peptide derived from PAR-2 ICL3 that blocks calcium signaling and ERK activation in response to PAR-2 activating proteases and PAR-2AP, but does not appear to inhibit β-arrestin recruitment(21,169,170). In several murine models, P2pal-18S was shown to successfully attenuate the severity of multiple PAR-2-dependent inflammatory challenges including lindlimb paw edema(169), atopic dermatitis and itch(171), experimental auto-immune encephalitis(54), as well as diet-induced non-alcholic fatty liver disease and CCl4-induced liver fibrosis(170). Recently P2pal-18S has been utilized to successfully target PAR-2 and overcome resistance of non-small-cell lung cancer (NSCLC) cells to EGFR TKIs(137). These authors demonstrated that PAR-2 actively contributes to NSCLC resistance to gefitinib, and that treatment with P2pal-18S reversed this resistance, increasing cell death in response to gefitinib in vitro(137). Importantly, using an in vivo model of gefitinib resistant NSCLC, the combination therapy of gefitinib with P2pal-18S significantly inhibited ERK activation and tumor growth(137). Furthermore, P2pal-18S was also found to sensitize resistant NSCLC cells to osimertinib by attenuating ERK-mediated EMT and PD-L1 expression, and this combination therapy dramatically inhibited growth of resistant lung cancer cells in vivo(138), with no overt signs of systematic toxicity. These exciting findings present the first preclinical evidence that targeting PAR-2 using this approach may be a promising adjuvant treatment for primary and metastatic cancers.

Conclusions
It is now understood that active proteases in the pericellular environment play a critical role in the landscape of cancer. PAR-2 is a crucial mediator of healthy acute inflammation, tissue repair and pain responses. Its cell signaling outcomes are mediated by both G-proteins and β-arrestins at the plasma membrane and internally in endosomes, and PAR-2 can mediate transactivation of other cell surface receptors. It is becoming clear that physiological signaling events activated by PAR-2 can be hijacked in chronic inflammatory diseases and for tumor metastasis. PAR-2 is over-expressed in many cancers where it correlates with poor prognoses. It serves as a molecular sentinel that detects proteolytic stimuli in the microenvironment and initiates cellular signaling responses. Preclinical studies have clearly shown that tumor cell PAR-2 expression promotes metastatic progression. The exciting recent studies highlighted above using the PAR-2 pepducin as an adjuvant therapy in mouse tumor models to overcome EGFR therapeutic resistance, reverse tumor EMT and immune cell inhibition, highlight the importance of pursuing PAR-2 antagonism for cancer treatment. The ability of PAR-2 activation to trigger signaling outcomes that promote metastasis and further, the ability to fine-tune signaling outcomes using selective, biased PAR-2 modulators together suggest that PAR-2 may be a very attractive therapeutic target for limiting the lethality of cancer metastasis.