Wnt/beta-catenin pathway: modulating anticancer immune response

Nusse and Varmus discovered the components of the Wnt/β-catenin pathway in 1982 while studying oncogenic mechanisms of mouse mammary tumor virus (MMTV) [1]. Proviral insertion at the “integration site” was thought to be the mechanism of carcinogenesis, giving the name to the first gene discovered in this pathway as INT1. Simultaneous work in developmental biology and work in drosophila established INT1 gene to be the homologue of the Drosophila segment polarity gene, Wingless [2]. Subsequently, human INT1 was shown to be very similar to mouse INT1, thereby revealing the highly conserved nature of this pathway across various species [3]. Additional screens for MMTV proviral insertion sites in tumors yielded several other activated genes that are related to other developmental gene pathways, such as INT2, INT3, and INT4 [4‐6]. For example, INT2 is a member of the fibroblast growth factor (FGF; INT2 is the same FGF-3 protein) family, and INT3 is related to the NOTCH gene family (INT3 protein is the same neurogenic locus notch homologue 4/NOTCH4) [7, 8]. With “INT” nomenclature turning out to be inadequate and confusing, consensus was reached to create the hybrid name “WNT” (for Wingless-related integration site) to denote genes belonging to the INT1/Wingless family. INT1—now called WNT1—became the founding member [9].

WNTs (translated products of WNT gene) are cysteine-rich glycoproteins, secreted by cells into the extracellular matrix, that activate receptor-mediated signaling with cells in immediate proximity [10]. The WNT protein family consists of at least 19 secreted glycoproteins (350–400 amino acids in length) highly conserved across species from invertebrates to mammals [11]. WNT binds to the N-terminal extra-cellular cysteine-rich domain of a Frizzled family receptor, a member of the superfamily of G-protein-coupled receptors. This disrupts the destruction complex of β-catenin (a tertiary complex formed by axin, adenomatous polyposis coli (APC), CK1α, and GSK3β) and triggers the cytoplasmic accumulation of β-catenin (Fig. 1).

T cell factor/lymphoid enhancer factor-1 (TCF/Lef1) is the transcription complex that mediates canonical WNT-triggered gene transcription [12, 13]. β-catenin translocates into the nucleus where it interacts with TCF/Lef1 and activates TCF/Lef1 transcription complex [14‐16]. β-catenin also localizes to multiple subcellular locations including the cytoplasm where its levels are tightly controlled. β-catenin also promotes cell-to-cell adhesion by accumulating in cell–cell contact sites, namely the adherens junctions [17, 18]. Figure 1 illustrates the canonical pathway of Wnt/β-catenin signaling. In addition to classical canonical WNT-induced activation of β-catenin–TCF/Lef1 transcriptional complexes, WNT can elicit alternative responses through β-catenin independent mechanisms which are collectively known as noncanonical pathways [19]. In an alternative concept known as integrated Wnt pathway, the canonical and noncanonical pathways are combined and multiple inputs at the level of both Wnt-receptor binding and the downstream, intracellular response have been integrated [20].

Wnt/β-catenin pathway is intricately involved in pathogenesis of several cancers. Recent findings of its role in regulating immunomodulation have renewed enthusiasm in the field.

Several therapeutic strategies have been developed with the aim to inhibit WNT-pathway, and many agents are undergoing early phase clinical trials (Table 1 and Fig. 2). Membrane bound O-acyl transferase porcupine (PORCN) has emerged as a molecular target of interest to treat WNT-driven cancers [56]. Addition of palmitoyl groups to WNT proteins is catalyzed by PORCN, a biochemical process known as palmitoylation or S-acylation in the endoplasmic reticulum, which enhances WNT secretion into the cytoplasm [57]. WNT974, a PORCN inhibitor, produced cytostatic effects in ovarian cancer cells in vitro [58] and decreased tumor growth and metastatic spread in head and neck squamous cell carcinoma models in vivo [59]. A phase I/II trial evaluating WNT974 in combination with LGX818, a specific BRAF inhibitor, and cetuximab in patients with metastatic colorectal cancer bearing WNT and BRAF mutations is ongoing (NCT02278133). Another trial is evaluating WNT974 in patients with metastatic head and neck squamous cell carcinoma (NCT02649530).

WNT agonists, specifically WNT-5a activation, have shown to inhibit metastases. Increased WNT-5a signaling suppressed endothelial tumor cell migration and invasion and inhibited metastasis in model of breast cancer in vivo. Foxy-5 is a WNT-5a mimic hexapeptide that binds and activates the WNT-5a receptors, Frizzled-2 and Frizzled-5 [60]. Two phase I trials are evaluating Foxy-5 in advanced solid tumors (breast, colorectal, and prostate) with loss or reduced expression of WNT-5a protein (NCT02020291 and NCT02655952).

CWP232291 is a novel small molecule that binds Src associated with mitosis 68K protein (Sam68). Preclinical data showed selective inhibitory activity on a WNT gene reporter and decreased expression of the β-catenin target genes, cyclin D1 and survivin. CWP232291 is being tested in phase I clinical trials in patients with acute myeloid leukemia (NCT01398462) and relapsed or refractory myeloma (NCT02426723) [61].

Genistein, a dietary compound present in soy-based foods, is thought to mediate anticancer activity through pleiotropic mechanisms that may include the inhibition of WNT pathway [62]. Phase I and II studies investigating the clinical activity of genistein and other soy isoflavones compounds in cancer treatment and chemoprevention have been published with disappointing results [63‐65].

OMP-54F28 (Ipafricept) is a fusion protein that combines the immunoglobulin Fc domain with the cysteine-rich domain of frizzled family receptor 8 (Fzd8) competing with the native Fzd8 receptor for its ligands and antagonizes WNT signaling [66]. A dose-escalation phase I showed that OMP-54F28 was well tolerated and dose escalation cohorts with standard therapy in advanced hepatocellular cancers (NCT02069145), pancreatic cancers (NCT02050178), and recurrent platinum-sensitive ovarian cancer (NCT02092363) are under way [67]. Another monoclonal antibody against frizzled receptors OMP-18R5 is in phase 1 clinical trials for patients with metastatic breast cancer in combination with paclitaxel. DKN-01 is a humanized monoclonal antibody (Mab) with neutralizing activity against Dickkopf-1 (Dkk-1). Strong preclinical evidence of synergistic activity with chemotherapy agents is the basis for clinical testing of these agents in combination.

Preclinical studies have addressed several other strategies to counteract Wnt/β-catenin-mediated carcinogenesis, although they have not been yet tested in clinical trials. Inducing or stabilizing the “destruction complex” of β-catenin, therefore reducing its intracellular levels and precluding its transcriptional activity, is a promising strategy. Axin is a concentration-limiting component of the β-catenin destruction complex, and its stability is regulated by tankyrase, the key regulator enzyme that is responsible for poly(ADP-ribosyl)ation (PARsylation) of axin and induces its proteasomal degradation [68, 69]. Waaler et al. demonstrated the antitumoral effect of inhibiting tankyrase in a model of colorectal adenoma and adenocarcinoma [70]. Potential adverse events of this therapy include diarrhea and intestinal toxicity [71]. Tankyrase inhibition has been shown to revert resistance to PI3K and AKT inhibitors in colorectal cancer patient-derived sphere cultures and mouse tumor xenografts [72]. High nuclear β-catenin expression predicted resistance to PI3K and AKT inhibitors. Combined treatment with a WNT/tankyrase inhibitor reduced nuclear β-catenin, reverted resistance to PI3K and AKT inhibitors, and repressed tumor growth.

Another possible strategy to reduce signaling through the canonical pathway of Wnt/β-catenin consists of antagonizing β-catenin/TCF-mediated transcription. Because transcriptional regulation of the β-catenin/TCF complex needs some coactivators, such as the cAMP response element-binding protein (CREB)-binding protein (CBP), this molecule was considered a potential target for inhibition [73]. INT-001, a specific antagonist of CBP, showed antitumoral activity in preclinical models of pancreatic, colon, and tamoxifen-resistant breast cancers [74‐76].

Multiple observational studies and randomized controlled trials have demonstrated a chemo-preventative role for aspirin, particularly in the development of colorectal neoplasia. Given the critical importance of WNT dysregulation in colorectal carcinogenesis, the interplay between aspirin and canonical WNT signaling has become a focus of investigation [77]. Through inhibiting cyclooxygenase-2 (COX-2), aspirin decreases the availability of prostaglandin E2 (PGE2), therefore reducing WNT levels [78]. Interestingly, cancer cells when exposed to aspirin in vitro led to the inhibition of WNT pathway through increasing β-catenin ubiquitination and consequent degradation, independent on COX inhibition as well [79]. Figure 2 summarizes the developmental strategies under investigation in Wnt/β-catenin pathway.

Overexpression of immune checkpoint molecules in the tumor microenvironment has a critical role in antitumor immunity evasion and cancer progression [80]. Currently, four immune checkpoint inhibitors (ICI) are approved for treatment of cancer, including ipilimumab (anti-cytotoxic T-lymphocyte-associated protein 4 (anti-CTLA4)), nivolumab, pembrolizumab (anti-programmed death-1 (anti-PD-1)), and atezolizumab (anti-programmed death ligand-1 (anti-PD-L1)). Anti-PD-1/PD-L1 antibodies have demonstrated clinical activity in more than 15 cancer types, but the majority of patients with advanced cancer still do not derive clinical benefit from these drugs suggesting that immunosuppressive mechanisms within the tumor microenvironment may play a role in de novo resistance to these therapies.

Clinical responses to multiple T cell-based therapies, including immune checkpoint inhibitors, have correlated with tumors with a T cell-inflamed microenvironment [81, 82]. These tumors are characterized by infiltration of CD8⁺ T cells, chemokines, and an interferon signature, which ultimately correlate with improved survival and responses to immunotherapies [83, 84]. On the other hand, it has been recognized that tumors without T cell infiltration have worse prognosis and do not benefit from immunotherapy [84]. Thus, understanding mechanisms driving T cell exclusion is critical to increase the number of patients that will benefit from immunotherapy.

Wnt/β-catenin pathway has been identified as one of the important oncogenic pathway signaling related to immune evasion (Fig. 3) [85, 86]. In a teratoma model, enhanced expression of Wnt correlated with impaired immune cell recruitment. Both T- and B cell infiltration was reduced which was independent of teratoma size and differentiation suggesting impaired immune surveillance [87]. Luke et al. evaluated gene expression data from The Cancer Genome Atlas (TCGA) to segregate tumors based on a T cell-inflamed gene expression signature and classified 8890 tumor samples into T cell-inflamed, non-T cell inflamed, and intermediate subtypes [88]. Approximately one third of the interrogated tumors were characterized as non-T cell inflamed. They further interrogated exomic sequencing to address the presence of mutations in Wnt/β-catenin pathway in this subtype. Activating mutations in CTNNB1 and inactivating mutations in negative regulators such as Axin1, Axin2, APC1, and APC2 were related to non-T cell inflamed gene signature, accounting for about 13% of the sequenced tumors. Most of the mutations were noted in Exon 3 of the CTNNB1 gene, with mis-sense mutation being the most common. Evidence of β-catenin pathway activation, in the presence of mutation or otherwise, was noted in 24 of 30 solid tumors of TCGA. Evidence for pathway activation without mutation was evidenced by increased expression of β-catenin pathway elements such as WNT ligands, Fzd receptors, or β-catenin itself. Immunohistochemistry demonstrated inverse correlations between β-catenin and CD8⁺ T cell infiltration.

These results suggest that increased expression of β-catenin in the tumor should be investigated and validated as a predictive tool to improve the selection of candidate patients for checkpoint inhibition therapy. In addition, one might hypothesize that inhibition of Wnt/β-catenin signaling could improve CD8⁺ T cell infiltration and priming, therefore possibly producing a more favorable scenario to immune checkpoint inhibition [89]. Thus, the investigation of the possible role of Wnt/β-catenin inhibitors as possible adjuvants to anti-PD1, anti-CTLA4, or anti-PD-L1 is an interesting strategy and should be evaluated in preclinical and clinical studies.

While tumor infiltration with CD8⁺ effector T cell is favorable, presence of regulatory T cells (Treg) has been associated with poor antitumor immunity. Tregs are known to suppress adaptive responses mainly by reducing CD8⁺ T cell proliferation, activation, and effector function [90, 91]. Some studies have shown that expression of β-catenin has been associated with Treg infiltration, survival, and activity [92, 93]. In an adoptive transfer mouse colorectal cancer model, β-catenin expression was enforced in intratumoral CD4⁺ T cells using lentiviral transduction. This led to increased IL-17a expression, enhanced proliferation, and inhibited apoptosis of colorectal cancer cells [87]. Active WNT signaling has been shown to disrupt Foxp3 transcriptional activity, a crucial step for both the development and function of regulatory T cells [94]. Hence, it has to be recognized that blocking WNT for immunomodulation is still controversial at this time. On the other hand, induction of Wnt signaling has shown to be important in maintenance of stemness of memory CD8⁺ T cells by blocking T cell differentiation [95]. This strategy may be used in generating memory T cell populations, thereby improving effectiveness of cancer vaccines, and may find applications in programming antitumor T cells for adoptive immunotherapy [96].

Wnt/β-catenin pathway has also been associated with modulation of innate immunity, such as dendritic cells [97]. Spranger et al. detected a correlation between activation of the Wnt/β-catenin signaling pathway and absence of a T cell gene expression signature in human metastatic melanoma [98]. Active β-catenin signaling was observed in 48% of non-T cell-infiltrated melanomas, which was thought to be the important pathway mediating immune exclusion. In mouse models, β-catenin induced expression of the transcription repressor ATF3, which suppressed the transcription of CCL4. The defective production of CCL4 led to impaired infiltration and activation of Batf3-lineage CD103⁺ dendritic cells, reduced CD8⁺ T cell priming and infiltration, and consequent non-response to immune check point blockade. In the absence of active β-catenin signaling, the normal production of CCL4 was restored, leading to CD103⁺ dendritic cell activation and infiltration and proficient priming of CD8⁺ T cell. Wnt/β-catenin signaling in intestinal dendritic cells has shown to regulate the balance between inflammatory versus regulatory responses in the gut [99]. β-catenin when expressed in intestinal dendritic cells was associated with expression of anti-inflammatory mediators such as retinoic acid-metabolizing enzymes, interleukin-10, and transforming growth factor-β, and the stimulation of Treg induction [100].

Wnt/Beta-catenin signaling has been shown to be involved in cross talk between cancer cells and tumor-associated macrophages. Kaler et al. demonstrated a possible role of interleukin-1β secreted by TAMs in the increased availability of β-catenin through phosphorylation of GSK3β in colon cancer cells, disrupting the function of β-catenin destruction complex [101]. Colorectal cancer cells stimulate macrophage production of IL-1β through Snail, a soluble factor product of a Wnt-regulated gene. These data demonstrated a novel Wnt-dependent cross talk mechanism between tumor cells and macrophages [102].

The Wnt/β-catenin pathway is increasingly recognized as a potentially important target for anticancer therapies, with several relevant inhibitors at various stages of clinical development. The expanding role of immunotherapies in cancer and recent insight into the role of Wnt-pathway in cancer-related immune-regulation may give a new dimension to this field of developmental therapeutics. Only tumors with upregulated Wnt/β-catenin signaling such as colorectal cancer have been explored as targets for Wnt inhibition. However, with its role in immunomodulation, Wnt inhibitors may have a broader role in cancers such as melanoma, lung, and renal cancers where immunotherapy has come to the forefront. There are some nuances to be addressed in the exact role of Wnt signaling in immunomodulation before this could be adopted in clinical practice.