Emergence of MUC1 in Mammals for Adaptation of Barrier Epithelia

Abstract

The mucin 1 (MUC1) gene was discovered based on its overexpression in human breast cancers. Subsequent work demonstrated that MUC1 is aberrantly expressed in cancers originating from other diverse organs, including skin and immune cells. These findings supported a role for MUC1 in the adaptation of barrier tissues to infection and environmental stress. Of fundamental importance for this evolutionary adaptation was inclusion of a SEA domain, which catalyzes autoproteolysis of the MUC1 protein and formation of a non-covalent heterodimeric complex. The resulting MUC1 heterodimer is poised at the apical cell membrane to respond to loss of homeostasis. Disruption of the complex releases the MUC1 N-terminal (MUC1-N) subunit into a protective mucous gel. Conversely, the transmembrane C-terminal (MUC1-C) subunit activates a program of lineage plasticity, epigenetic reprogramming and repair. This MUC1-C-activated program apparently evolved for barrier tissues to mount self-regulating proliferative, inflammatory and remodeling responses associated with wound healing. Emerging evidence indicates that MUC1-C underpins inflammatory adaptation of tissue stem cells and immune cells in the barrier niche. This review focuses on how prolonged activation of MUC1-C by chronic inflammation in these niches promotes the cancer stem cell (CSC) state by establishing auto-inductive nodes that drive self-renewal and tumorigenicity.

Keywords: MUC1; MUC1-C; barrier epithelia; cancer; inflammatory memory; loss of homeostasis.

Figure 5

The MUC1-C cytoplasmic domain (CD) functions as a node (NODE 1) in auto-induction of the WNT/β-catenin signaling pathway. The 72 aa MUC1-C/CD includes a serine-rich motif (aa 50–59, SAGNGGSSLS; SRM) that interacts directly with the β-catenin Armadillo repeats [60]. This interaction contributes to stabilization of β-catenin and is regulated by upstream phosphorylation of (i) T-41 by PKCδ [59], (ii) S-44 by GSK3β [60], and (iii) Y-46 by EGFR and SRC [57,58]. The MUC1-C/CD CQC motif is necessary for interactions with certain proteins that are mediated by disulfide bonds [48]. In this way, the MUC1-C/CD CQC motif binds to the TCF4 E-tail [61], facilitating the formation of β-catenin/TCF4 complexes that activate CCND1 and MYC. MUC1-C forms auto-inductive circuits with β-catenin/TCF4 and MYC in sustaining activation of this node. Notably, the CQC motif Cys residues are subject to palmitoylation [50], which could preclude their capacity to form disulfide bonds and function in this and other auto-inductive nodes.

Figure 8

MUC1-C promotes chromatin remodeling in coupling memory with progression of the CSC state. (A). The inflammatory MUC1-C→STAT3 auto-inductive node induces the EMT program, stem cell factors and CSC state as evidenced by self-renewal capacity and tumorigenicity [43]. MUC1-C/STAT3 complexes recruit the BAF chromatin remodeling complex to specific target genes to regulate chromatin accessibility (BioRender). (B). Proposed model in which MUC1-C and BAF recruit JUN/AP-1 to establish memory and activation of stemness-associated genes in sustaining the CSC state [68]. Modified from Bhattacharya using BioRender [68].

Figure 1

Structure of the MUC1 heterodimeric complex. MUC1 is translated as a single polypeptide that includes a signal sequence for membrane localization and an extended region of O-glycosylated 20 aa tandem repeats (TRs). MUC1 also includes a SEA domain and a transmembrane (TM) domain. Autoproteolysis within the SEA domain results in the generation of MUC1 N-terminal (MUC1-N) and C-terminal (MUC1-C) subunits that form a non-covalent complex. Figure modified from Kufe [9].

Figure 2

MUC1 evolved from MUC5B and HSPG2. The secreted MUC2, MUC6, MUC5AC and MUC5B mucins, which localize to chromosome 11p15.5 in humans, emerged from common ancestors that appeared in early metazoan evolution. MUC1 emerged in mammals in part from MUC5B and is located at chromosome 1q22. The MUC1 SEA domain evolved from HSPG2. MUC5AC and MUC5B form protective oligomeric structures in the response to inflammation [29]. MUC1 is devoid of the von Willebrand factor type C (VWC) and D (VWD), trypsin inhibitor-like cysteine-rich (TIL) and C-terminal cysteine knot (CT) domains. HSPG2/perlecan is activated by inflammation and has been linked to regulation of the tumor microenvironment and cancer progression [30]. Figure modified from Kufe [9].

Figure 3

MUC1 contributes to a protective physical barrier and activation of repair in epithelia. The MUC1 heterodimer is poised in a basal state at the apical cell membrane as a sensor of homeostasis. Entropic forces in the glycocalyx induced in association with loss of homeostasis, as well as proteolytic cleavage, disrupt the MUC1-N/MUC1-C complex with release of MUC1-N into a protective physical barrier. Activation of MUC1-C induces a program of repair associated with the wound healing response. Figure modified from Kufe [9].

Figure 4

Structure of the MUC1-C extracellular domain (ED). The 58 aa MUC1-C/ED includes the SVVVQLTLAFREGTIN sequence that forms a non-covalent interaction with MUC1-N [24]. Downstream to that region is the VHDVETQFNQ sequence which forms the alpha-3 helix. A QYK motif separates the alpha-3 and alpha-4 helices. Adjacent to the alpha-4 helix is a consensus NLT motif that is modified by N-glycosylation and functions as a galectin-3 ligand binding site. Galectin-3 acts as a bridge for the association of MUC1-C with EGFR and other RTKs at the cell membrane, as well as additional effectors in the cytoplasm and nucleus.

Figure 6

MUC1-C/CD functions as a node (NODE 2) for auto-induction of the TAK1/IKK/NF-κB pathway. The MUC1-C/CD CQC motif binds to TAK1 [69]. MUC1-C/CD(4–45) interacts with IKKβ and MUC1-C(46–72) forms a complex with IKKγ [81]. The MUC1-C/CD SRM GGSSLS sequence binds directly to the NF-κB p65 RHD/DBD, integrating activation of the TAK1/IKK/NF-κB pathway [77]. NF-κB induces MUC1 expression in an auto-inductive circuit.

Figure 7

MUC1-C activates JAK1→STAT3 signaling in an auto-inductive loop (NODE 3). As found for TAK1 in the NF-κB pathway, the MUC1-C/CD CQC motif binds to JAK1. In addition, like NF-κB p65, the MUC1-C/CD SRM functions as a site for direct binding to the STAT3 DBD and for JAK1-mediated pSTAT3 activation [82]. In turn, pSTAT3 induces MUC1 expression in another auto-inductive inflammatory circuit that parallels the node driving MUC1-C→NF-κB signaling.

Figure 9

Targeting the MUC1-C extracellular and cytoplasmic domains for disruption of auto-inductive nodes and elimination of CSCs. (A). Antibody 3D1 generated against the MUC1-C/ED alpha-3 helix has been developed for (i) allogeneic CAR T cells that are under clinical evaluation, and (ii) ADCs that are being advanced with IND-enabling studies by the NCI NExT Program. MUC1-C forms complexes with EGFR at the cell membrane that are mediated by galectin-3 [44]. In this way, MUC1-C contributes to EGFR activation and resistance to EGFR inhibitors [101]. Antibodies generated against the alpha-4 helix are being developed to block the MUC1-C/ED interaction with galectin-3 and thereby inhibit constitutive MUC1-C-driven RTK activation. (B). The MUC1-C/CD CQC motif is necessary for MUC1-C homodimerization and function as an oncoprotein. Targeting the MUC1-C CQCRRKN region with the GO-203 inhibitor blocks interactions with TCF4 [61], TAK1 [69] and JAK1 [82]. GO-203 treatment also inhibits the interactions of MUC1-C with STAT3 [82] and NF-κB [77]. As a result, targeting the MUC1-C CQC motif disrupts auto-induction of MUC1-C NODES 1–3. Ongoing work is addressing another MUC1-C node that may be of importance for RTK→RAS signaling in cancer. In this regard, MUC1-C forms complexes with effectors of the RTK→RAS pathway that include PI3K [102], SHC, PLCγ and GRB2/SOS [103].