|Home | About | Journals | Submit | Contact Us | Français|
The membrane mucin MUC4 (human) is abundantly expressed in many epithelia, where it is proposed to play a protective role, and is overexpressed in some epithelial tumors. Studies on the rat homologue, Muc4, indicate that it acts through anti-adhesive or signaling mechanisms. In particular, Muc4/MUC4 can serve as a ligand/modulator of the receptor tyrosine kinase ErbB2, regulating its phosphorylation and the phosphorylation of its partner ErbB3, with or without the involvement of the ErbB3 ligand neuregulin. Muc4/MUC4 can also modulate cell apoptosis via multiple mechanisms, both ErbB2 dependent and independent. Muc4/MUC4 expression is regulated by multiple mechanisms, ranging from transcriptional to post-translational. The roles of MUC4 in tumors suggest that it may be valuable as a tumor marker or target for therapy.
Muc4 is a membrane mucin originally discovered and characterized in a highly metastatic rat mammary adenocarcinoma [1,2]. It was initially called sialomucin complex (SMC)  as the mucin nomenclature had not been developed. Human MUC4 was first recognized by cloning a fragment from a tracheal library . Complete sequencing of the rat SMC (Muc4) [4,5] and human MUC4 demonstrated their homology . Rat Muc4/SMC is a heterodimeric glycoprotein composed of a mucin subunit and a transmembrane subunit (Figure 1A) . The mucin subunit, called ASGP-1 in the rat, contains 11 repeats of approximately 125 amino acids  and is highly O-glycosylated . The transmembrane subunit (ASGP-2 in the rat; β in the human) contains two EGF domains, a transmembrane sequence and a short cytoplasmic tail and multiple N-glycosylation sites [4,8]. The human MUC4 is also heterodimeric, but the mucin subunit (α) contains only three of the 125-amino acid repeats. However, it has extensive (up to 400) polymorphic repeats of 16 amino acids . Thus, it is much larger than the rat Muc4; it is one of the largest of the known membrane mucins. The Muc4 and MUC4 transmembrane subunits are more similar, with 60–70% amino acid identity and conservation of the domain structure .
Muc4 is transcribed from a single gene into a 9-kb message and translated into a 300-kDa precursor polypeptide, which is N-glycosylated . Cleavage of this precursor into the two subunits occurs before O-glycosylation of the mucin subunit, presumably in the endoplasmic reticulum. A second cleavage occurring at approximately the same time can release a soluble form of the mucin containing the ASGP-1 and most of the ASGP-2 . Because of its similar subunit and domain/sequence organization structure, similar processing mechanisms are presumed to occur with human MUC4. Moreover, numerous splice variants of MUC4, both soluble and membrane, have been proposed based on the structure of the gene , some of which have been found to be expressed. Thus, the expression of MUC4 isoforms in human tumors can potentially be very complex, but has not been analyzed in detail.
MUC4 has been implicated in multiple human cancers, including pancreatic, lung, breast, gall bladder, salivary gland, prostate, biliary tract and ovarian (Table 1), and may be a useful clinical marker for some of these . Understanding the expression and function of MUC4 in tumors is complicated by the variety of analytical methods used, ranging from transcript analysis to immunohistochemistry, and the difficulty of their application. Thus results obtained from different methods and from different laboratories and on different types of samples may vary widely and even indicate different outcomes of the expression of MUC4. These vagaries make the investigation of the function of MUC4 in tumors very difficult. However, some insights are beginning to emerge. Although the exact role of MUC4 in any cancer remains to be determined, silencing of MUC4 leads to decreased pancreatic tumor growth and metastasis . Thus, a contribution of MUC4 to the aggressiveness of pancreatic cancer appears to be important. Interestingly, a recent study showed the upregulation of MUC4 in breast tumor metastases, but not in primary tumors [Carraway KL 3rd. Pers. Commun.]. These results are consistent with a previous study showing the ability of Muc4 to promote tumor metastasis . In contrast, MUC4 appears to be associated with more differentiated and less aggressive head and neck cancers . Thus, the role of MUC4 appears to depend on the particular cancer and cell context.
Because of its large size and extended conformation, Muc4 can block access to the cell surface . In normal epithelial cells this mechanism can protect the epithelium from invasion by microbes . In human pancreatic tumor cells, MUC4 represses cell aggregation . Similarly, tumor cells are protected from killing mechanisms by immune cells , and even from antibodies that might bind to specific molecules on the cell surface , thus blocking important immune surveillance mechanisms and potential immune-based therapies. With regard to therapy, MUC4 was found to be overexpressed in a human breast cancer resistant to antibody therapy with trastuzumab (Herceptin®) . Silencing of the MUC4 expression restored trastuzumab binding. Moreover, since the action of antibody-based therapies is often based on cell killing by immune cells , the ability of Muc4/MUC4 to block cell killing  may be even more important than the effects on antibody binding.
An important part of this protection mechanism is the localization of the mucin. In polarized epithelia, Muc4 is located on the apical surfaces . In nonpolarized tumor cells the Muc4 is located circumferentially. Interestingly, Muc4 is able to disrupt cell–cell interactions in tumor cells that can form adhesions . Thus, the Muc4/MUC4 can contribute to dissociation of tumor cells from the primary tumor, facilitating invasion and metastasis. The disruption of adherens junctions and cadherin interactions in tumor cells abrogates contact inhibition in these cells, leading to increased signaling through the Erk pathway that promotes cell proliferation . Therefore, Muc4 can promote tumor cell growth by indirect, anti- adhesive mechanisms, as well as by the more direct signaling mechanisms described below.
Pro-adhesive effects for Muc4 can be envisioned because of its extensive glycosylation of both N- and O-linked oligosaccharides [7,8], which could interact with cellular lectins. Although such mechanisms have been described for MUC1 , no comparable interactions are known for MUC4.
Surprisingly, Muc4 can contribute to both differentiation and proliferation, depending on the cell context. The key to these disparate activities is the ability of Muc4 to act as an intramembrane ligand and modulator for the receptor tyrosine kinase ErbB2 [17,25]. Numerous localization and immunoprecipitation studies in Muc4-transfected cells and in cells naturally expressing Muc4/MUC4 have demonstrated a complex between the mucin and ErbB2 . Since Muc4 interacts with ErbB2 via an EGF domain on the Muc4  and promotes phosphorylation of the ErbB2 [26–28], the mucin has been proposed to be a unique type of ligand, which acts via an intramembrane mechanism (Figure 1) . Interestingly, the Muc4–ErbB2 complex forms shortly after synthesis of the two proteins , which has important implications for the function of the complex.
The EGF family of receptor tyrosine kinases plays important roles in development, oncogenesis and apoptosis, including epithelial growth control and wound healing . The family consists of four members, named ErbB1–4. ErbB1 and ErbB4 can be activated by ligands by a classical homodimerization mechanism, in which ligand binding converts an inactive tethered conformation to an active extended conformation. This conformational change frees a loop through which the ErbBs can dimerize, with consequent re-orientation of the kinase domains of the receptor and phosphorylation of the receptor cytoplasmic tails . ErbB2 is proposed to be unable to undergo such changes. It is already in the extended conformation with the loop accessible for dimer formation , primed for activation. Notably, ErbB3 has amino acid changes in its kinase domain that repress its kinase activity . ErbB3 thus serves as a docking protein, similar to the insulin receptor substrate, particularly for activation of the phosphatidylinositol 3 (PI3)-kinase/Akt pathway. The ErbB2–ErbB3 pair forms a potent heterodimer that undergoes phosphorylation in the presence of ErbB3 ligand neuregulin , and is a major contributor to cell proliferation as well as an inhibitor of apoptosis. However, whether they can form heterodimers depends on the presence and localization of the receptors .
In polarized epithelial cells that have no Muc4, ErbB2 is localized to the lateral surfaces of the cells along with ErbB3 (Figure 2A) . However, when Muc4 is present in polarized cells, the Muc4 and ErbB2 form a complex shortly after synthesis. This complex transits to the apical cell surface . These results are consistent with observations of co-localization of ErbB2 and Muc4 at apical surfaces in several epithelial tissues, such as the uterus, mammary gland and airway . This mechanism effectively segregates ErbB2 from ErbB3 (Figure 2B), repressing their ability to form the heterodimers that can drive cell proliferation and motility and inhibit apoptosis. ErbB2 of the Muc4–ErbB2 complex at the apical surface is phosphorylated on specific tyrosine residues, including 1139 and 1248, and coupled to the activation of the MAP kinase p38 , which has been implicated in differentiation . Thus, Muc4 can contribute to maintaining differentiation in polarized epithelia by segregating ErbB2 and ErbB3, and by activating p38 through an ErbB2-dependent mechanism .
This segregation mechanism breaks down when barriers establishing cell polarity are lost, as occurs when epithelia are damaged  or undergo neoplastic transformation  (Figure 2C). ErbB2 and ErbB3 can form complexes even in the absence of ligands . The presence of Muc4 stabilizes the ErbB2–ErbB3 complex and promotes phosphorylation of the ErbB2, but not ErbB3 (Figure 1). The presence of the ErbB3 ligand neuregulin likewise stabilizes the complex and, in addition, promotes phosphorylation of both ErbB2 and ErbB3  (Figure 1). Muc4 can modulate the effect of neuregulin on the receptor phosphorylation by regulating the localization and turnover of the complex .
A key issue is which downstream pathways are activated by the action of Muc4 on the ErbB2–ErbB3 complex in tumors. One important element is the potentiation of signaling through the PI3-kinase–Akt pathway [26,37], which is activated by phosphorylated ErbB3 and plays a role in both proliferation and repression of apoptosis. MUC4 silencing in pancreatic tumor cells has been shown to inhibit focal adhesion kinase (FAK) phosphorylation and activation of the Erk pathway , the former being involved in cell migration and the latter important to proliferative responses. The involvement of the Muc4 interaction with ErbB2 in cell migration is supported by microarray studies on changes in protein phosphorylation of cellular proteins in tumor cells transfected with a Muc4 derivative that does not show antiadhesive effects. These cells showed both increased FAK phosphorylation and β-catenin levels, suggesting that Muc4 can influence cell migration without directly altering cell adhesion through its action on the ErbB2–ErbB3 complex. MUC4 in ovarian cancer cells can promote FAK phosphorylation, cytoskeletal rearrangements and cell migration . The mechanism (signaling vs anti-adhesion) is unclear, but the implications for ovarian cancer invasiveness and metastasis are likely to be important.
Muc4/MUC4 can promote tumor progression by repressing apoptosis by multiple mechanisms, both ErbB2 dependent and independent . One of these mechanisms is likely to involve activation of the PI3-kinase–Akt pathway . These effects extend to apoptosis induced by differing insults, including loss of adhesion (anoikis), chemotherapeutic drugs  and serum starvation . The mechanisms involved in the apoptosis repression vary in different cell lines, but a key element appears to be phosphorylation and inactivation of the pro-apoptotic protein Bad and an increase in the expression of Bcl-xl , which promotes survival. The mucin subunit and cytoplasmic domain of Muc4 are not required for these antiapoptotic effects or for ErbB2 signaling.
Muc4/MUC4 is regulated by a variety of mechanisms, ranging from transcriptional to post-translational (Table 2). Transcriptional regulation of rat Muc4 has been studied in mammary epithelial cells and 13762 mammary adenocarcinoma cells, from which the membrane mucin was originally isolated . Upregulation of transcription in mammary epithelial cells was observed by treatment with IGF, but not EGF . Inhibition of the Erk pathway blocked the IGF effect. Regulation of Muc4 in the 13762 tumor cells similarly involved the Erk pathway, acting as a potentiator of the transcription factor PEA3 . Muc4 can also be transcriptionally regulated through the JNK pathway. PEA3 binding to the Muc4 promoter  was demonstrated to be specific, acting through one or more of the PEA3 binding sites on the Muc4 promoter [43,44]. Interestingly, ErbB2 is also regulated by PEA3 . Thus, PEA3 can be involved in the coordination of the regulation of the receptor and its modulator .
Transcription of the human MUC4 gene is complex and regulated by many signaling pathways [46,47]. Extensive studies have been performed on the promoter of human MUC4 and its transcriptional regulators [48,49], showing four transcriptional start sites, one in a proximal promoter region and three in a distal region. The distal promoter contains a TATA box, but the proximal does not. Multiple binding sites for factors initiating transcription were found in the proximal region, including the glucocorticoid receptor, Sp1, AP-1, PEA3, CACCC box and Med-1. The distal promoter contains many of the same sites and also has a cyclic AMP responsive element. Multiple sites for other signaling pathways, such as protein kinase A, protein kinase C, cytokines and TGFβ, have been found. Numerous transcription factors operate on the MUC4 promoter, including AP2, PEA3, STAT1, SMAD and forkhead box A.
Transcriptional regulation of MUC4 has been particularly studied in pancreatic cancer cells [12,49]. MUC4 is poorly expressed, if at all, in the normal pancreas, but is highly expressed in pancreatic cancer . Thus, it has been proposed as a potential target for pancreatic cancer therapy . Targeting mucins is difficult, however, as they have no activities, such as those of enzymes or receptors, that can be easily blocked by small-molecule inhibitors. Therefore, repression of expression of the mucins represents one way of developing targeting strategies. Retinoic acid, interferon-γ and TGFβ have all been identified as regulatory factors for MUC4. The TGFβ has both SMAD-dependent and independent pathways and is negatively regulated by SMAD7 . Regulation by interferon-γ involves STAT-1 . Interferon-γ and retinoic acid synergistically upregulate MUC4 in pancreatic cancer cells in a process that involves reprogramming of signaling pathways. Other cytokines, including TNF-α, have been implicated in MUC4 regulation . A synergy between the TNF and interferon-γ activates STATs and NFκB transcription factors that interact with the MUC4 promoter to enhance MUC4 expression .
Interleukins, such as IL4 and IL9, have been shown to contribute to MUC4 regulation in normal tissues, such as the airway, through the JAK–STAT pathway . A similar mechanism has been observed in a lung cancer cell line (Table 2). The STAT pathway has also been observed to be involved in upregulation of MUC4 in gastric cancer through IL6 (Table 2). In esophageal cancer, bile acids have been implicated in MUC4 expression through the PI3-kinase signaling pathway and hepatocyte nuclear factor 1α (Table 2).
Epigenetic regulation of MUC4 appears to be important in some tumors, and is probably dependent on the presence of two CpG islands in the 5´-flanking region . Five methylation sites were found in pancreatic tumor cells to contribute to this regulation. Moreover, inhibition of histone deacetylation in these cells demonstrated a role for this epigenetic process in the regulation of MUC4 in these tumors. Breast, lung and colon cancer cell lines have also been shown to regulate MUC4 by methylation . These studies may provide new approaches for targeting the involvement of MUC4 in tumor progression .
MUC4 in human airway epithelial cells is upregulated by transcript stabilization by elastase produced by neutrophils as part of the inflammatory process . In this case similar effects are observed for MUC4 and MUC5AC , a gel-forming mucin. Thus, the regulation of airway mucins appears to be complex. It is likely that the regulation of MUC4 in lung tumors is equally complex.
TGFβ is a potent repressor of Muc4 expression in rat mammary epithelial cells, but not in rat mammary tumor cells . Transcript and protein analyses indicate that the effect occurs at a post-translational level, involving the protein precursor of the heterodimeric Muc4 . Investigation of the signaling pathways involved demonstrated a role for SMAD2 of the canonical TGFβ pathway . Inteferon-γ is able to block the TGFβ effect by upregulating the inhibitory SMAD7. These results suggested a role for proteosomal degradation in the regulation of Muc4. Blocking proteosome activity increases levels of Muc4 . This block also increases the level of ubiquitinated Muc4 and of Muc4 associated with calnexin and calreticulum. Undegraded Muc4 in cells treated with proteosome inhibitors is shunted to aggresomes . Tumor cells do not show a similar downregulation of Muc4 because the tumors have lost their responsiveness to TGFβ . This loss may occur through mutations in the TGFβ receptors, the SMAD pathway components or other mechanisms. In the normal mammary gland the loss of TGFβ during pregnancy permits the upregulation and secretion of the Muc4 as a protective agent for the epithelia . In contrast, mammary tumors have lost the responsiveness to TGFβ, thus they continue to express Muc4, which can promote their survival and proliferation.
Rat Muc4 has been found in squamous epithelia in both the eye  and female reproductive tract , localized to specific layers of the stratified epithelia. Interestingly, in the eye the localization to the most superficial layers depends on the regulation of the Muc4 by TGFβ through the proteosomal degradation mechanism. How Muc4 is regulated in the female reproductive tract squamous epithelia is unknown, but should be investigated as it may provide insight into contributions of this mucin to cancers in these sites.
TGFβ has also been shown to downregulate Muc4 in rat uterine epithelial cells, thus contributing to preparing the uterus for embryo implantation . Although the molecular mechanism (transcriptional vs post-translational) is unclear, the effect has been shown to involve a hormone-driven paracrine cellular mechanism . Estrogen and progesterone act on uterine fibroblasts, which produce TGFβ. The TGFβ then acts on the uterine epithelial cells to block Muc4 expression by inhibiting precursor processing and shunting the Muc4 to the proteosome for degradation. It would be interesting to know whether a similar mechanism for Muc4 regulation is present in hormone-sensitive tumors, such as breast cancer. The multiple effects of TGFβ on rat Muc4 indicate the complexity of its functions, operating at different levels (transcriptional vs post-translational) in different tissues. Similarly, TGFβ has been shown to upregulate MUC4 in human pancreatic tumors, but downregulate it in rat mammary gland. Thus, the effects of TGFβ are very much context dependent.
In spite of great advances in understanding mechanisms of oncogenesis and tumor progression, many aspects are still poorly understood. Recent studies on the mechanisms by which MUC4 may contribute to multiple tumor types indicate that it should be seriously considered not only as a tumor marker, but as a potential target for therapy. In the former case the ability of MUC4 to contribute to resistance mechanisms by blocking immune mechanisms or repressing chemotherapeutic approaches that induce apoptosis suggests that MUC4 expression should be considered in stratifying patients for particular types of therapy. Therapeutically, repression of MUC4 expression might eliminate some of those resistance mechanisms and improve the success rates of some current treatments , such as trastuzumab therapy.
The contributions of MUC4 to many types of tumors is still unclear because of inadequate reagents for analysis. New antibodies are needed that eliminate problems with carbohydrate dependence. The mechanisms by which MUC4 represses apoptosis have not been well characterized and need to be investigated at the molecular level. Mechanisms for targeting MUC4 have not received much attention, but the involvement of epigenetic regulation may provide new avenues for blocking MUC4 expression in tumors.
We thank Isabelle van Seuningen for her comments on the manuscript.
Financial & competing interests disclosure
Original research was supported by NIH grant CA52498. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was utilized in the production of this manuscript.
Kermit L Carraway, Department of Cell Biology and Anatomy, University of Miami School of Medicine, Miami, FL 33136, USA, Tel.: +1 305 243 6512, Fax: +1 305 243 4431, Email: imaim.dem@awarraCK.
George Theodoropoulos, Department of Cell Biology and Anatomy, University of Miami School of Medicine, Miami, USA.
Goldi A Kozloski, Department of Cell Biology and Anatomy, University of Miami School of Medicine, Miami, USA.
Coralie A Carothers Carraway, Department of Biochemistry and Molecular Biology, University of Miami School of Medicine, Miami, USA.
Papers of special note have been highlighted as:
of considerable interest