Mucins — barrier glycoproteins that form mucus and other gel-like secretions — have long been a target of interest in oncology, but their thick coating of sugars and diverse functions have defied traditional drug discovery. Although multiple drug candidates that target mucins have been tested in clinical trials, none has shown compelling clinical benefit. Writing in Nature Biotechnology, Pedram et al.1 now propose a new drug modality to target cancer-associated mucins: an engineered mucin-degrading enzyme linked to an antibody against a cell-surface marker on cancer cells. Using a variety of in vitro and in vivo models, the authors show promising results in breast cancer that merit further preclinical investigation.

Mucins are defined as glycoproteins that have more than 50% of their mass as carbohydrate. They are expressed on most epithelial cells (cells that line the ducts in the body) and are predominantly, but not exclusively, overexpressed in the carcinomas that arise from this cell type2. Many mucin genes are considered oncogenes, and they exert their pro-tumorigenic effects in several ways. First, mucins can present a barrier around the cancer cell that prevents immune eradication and facilitates electrostatically mediated spread3. Second, through their intracellular signaling motif, membrane mucins can signal after interacting with cognate ligands, lectins, bacteria and viruses or after cleavage by several common proteases, which leads to activation of pro-proliferative and pro-survival pathways in epithelial cells4. Third, through their sugars, mucins can interact directly with lectins carried on immune cells in ways that support tumor survival, growth and spread5. Finally, they can promote metastasis, either through direct binding of their carbohydrates to receptors elsewhere in the body (for example, sLex-binding E-selectin, or core 1-binding galectin 3) or indirectly by enhancing the function of adhesion molecules6,7.

It is important to note that the increased expression of mucins commonly seen in cancer arises not from genetic mutations or gene amplification, but instead from local inflammatory factors released in the tumor microenvironment. Indeed, the growing consensus is that mucins — beyond their protective and biome-orientated functions — support healthy epithelial and epithelial–immune homeostasis by regulating re-epithelialization and local immunity; dysregulation of these processes is seen in diseases of chronic epithelial inflammation, such as carcinomas and interstitial lung disease. Targeting mucins for cancer therapy is of interest not only because they are biomarkers of disease but also because they have a pathological role.

A particularly popular target is the mucin MUC1 (ref. 8). Most therapeutic approaches have focused on the MUC1 extracellular domain —for example, by targeting it with antibodies to deliver toxins and radioisotopes or with chimeric antigen receptor T cells. Approaches directed to the cytoplasmic tail, which are aimed at inhibiting signaling, are also being investigated. The study by Pedram et al. describes a strategy that is applicable to most members of the mucin and mucin-like family. Here, the aim is to deliver a bacterial mucin-degrading enzyme (a mucinase) that is capable of degrading any protein carrying a ‘mucin motif’ at the surface of cancer cells. In its reliance on protein degradation, the approach resembles proteolysis-targeting chimeras (PROTACs)9. However, PROTACs induce intracellular proteolysis in the proteasome, whereas Pedram et al. use a protease against extracellular proteins.

To increase specificity toward cancer cells, the authors fuse the bacterial mucinase StcE to an antibody against HER2, a receptor that is commonly overexpressed by some cancers and is the target of the breast-cancer drug trastuzumab (Herceptin). There are two important aspects of this construct. First, engineering of the mucinase is used to lower its activity, which ensures that it is active only when concentrated at the cell surface. Second, the targeting antibody is a nanobody (an antibody derived from the camel family that contains only a heavy chain). Importantly, this means the antibody cannot bind immune cells, which prevents ‘mopping up’ of the construct and unwanted bystander effects.

Tests in mixed cell cultures show that the construct effectively degrades the mucins on a HER2-expressing cell line but not on a non-expressing line. These experiments demonstrate the absolute requirement for an engineered low-activity mucinase, as the wild-type enzyme construct degrades mucins on both HER2-expressing and non-expressing lines. Further in vitro experiments show that removal of mucin by eStcE–HER2 enhances natural killer cell killing and macrophage phagocytosis of HER2+ cells. In a mouse model, the lungs of BALB/cJ mice were seeded by intravenous injection with a mouse mammary carcinoma cell line (4T07 cells) transfected with human MUC1, without the cytoplasmic tail, and human HER2. In this in vivo model, eStcE–HER2 limits tumor growth and metastasis while also altering the local immune environment to a more ‘anti-tumor’ state, without any obvious toxicity to the animals (Fig. 1).

Fig. 1: Action of the HER2-targeted mucinase anti-HER2–eStcE and the resultant change in immunological landscape.
figure 1

The mucins and mucin-like proteins expressed by HER2+ cancer cells are degraded by anti-HER2–eStcE. The mucinase has been engineered to have low activity and so is active only when concentrated at the tumor cell via targeting of HER2, which ensures specificity. Removal of the mucins and mucin-like proteins results in a change of immune cells surrounding the tumor in an in vivo breast cancer model, which leads to slower primary tumor growth and reduced lung metastases.

This study is an extremely encouraging start to the use of targeted mucinases for cancer therapy. However, as is often the case, more preclinical work is needed before the approach can be taken forward into the clinic. As mentioned, mucins such as MUC1 can drive tumor growth and survival through signaling after ligation or cleavage via their intracellular domain. Quite correctly, Pedram et al. use models in which MUC1 lacks this intracellular domain so as to remove a variable and aid in interpretation. However, this means that we do not yet know the effect of the treatment on cancer growth or survival through this mechanism. The use of human transgenic models for mouse studies will also be an important next step. Although mouse and human HER2 have a high level of sequence identity, many antibodies, such as trastuzumab, do not cross-react with mouse HER2 (ref. 10). This is important as HER2 is known to be expressed at lower levels by some normal tissues and cells, including cardiomyocytes, which leads to toxicity when targeted.

A conceptual hurdle to be overcome is the fact that mucin production by cancers is a continuous process driven by local inflammation, which means that continuous treatment would be needed until a satisfactory clinical end point is reached. Other targeted therapies also require continuous treatment, but in this case the use of a bacterial, non-self protein would probably induce increasing anti-drug immunity over time, preventing long-term administration. However, the data from Pedram et al. demonstrate that a short treatment regime may indeed be sufficient to tip the immunological balance toward tumor eradication.

Finally, the authors suggest that their approach may be applicable to other diseases that overexpress mucins, such as cystic fibrosis and gut dysbiosis. However, these two conditions occur in organs where the mucin layer is crucial for protection or microbiome function, so caution will be required in developing such applications. Overall, the work of Pedram et al. is an exciting approach toward an intractable target, with great potential in cancer therapy.