Targeting CDCP1 dimerization in triple-negative breast cancer

This article refers to:

Breast cancer is the leading contributor to new cancer cases and the second leading cause of cancer deaths in women in the United States. Approximately 15% of breast cancer patients present with triple-negative breast cancer (TNBC) disease. TNBC lacks expression of the Estrogen (ER), Progesterone (PR), and Human Epidermal growth factor 2 (HER2) receptors, rendering it resistant to current FDA-approved breast cancer targeted therapies like tamoxifen, aromatase inhibitors, and trastuzumab (Herceptin), respectively. Clinical trials targeting other Receptor Tyrosine Kinase (RTK) pathways like Epidermal Growth Factor Receptor (EGFR, also known as HER1) with cetuximab and Vascular Endothelial Growth Factor A (VEGF-A) with bevacizumab in combination with standard-of-care chemotherapy gave unimpressive results.¹ Current clinical trials are now investigating the efficacy of immune-modulating therapeutics like pembrolizumab, nivolumab, and ipilimumab, as well as therapeutics targeting androgen receptor and PI3K pathway. Thus, because TNBC is typically more aggressive with higher rates of recurrence and metastasis than non-TNBC, it is important to map new therapeutic targets to treat TNBC progression and metastasis.

TNBC overexpresses a trans-membrane protein, CUB-domain containing protein 1 (CDCP1),^2,3 reported to drive metastasis in multiple in vivo models.⁴ Not surprisingly, a hypoxic microenvironment, known to trigger metastasis, upregulates CDCP1 at the mRNA and protein level in kidney cancer,⁵ and stimulates CDCP1 phosphorylation in TNBC, increasing its activity.² CDCP1 contains a large extracellular domain with 3 CUB domains, important for protein-protein interactions.⁶ CDCP1s extracellular domain can be cleaved by serine proteases resulting in a 65 kDa fragment containing CUB1 that is released into the extracellular space and a 75 kDa membrane-bound isoform containing CUB2 and CUB3, referred as cleaved CDCP1 (cCDCP1).⁴ Importantly, cCDCP1 represents an active form of CDCP1, which efficiently stimulates signaling through multiple kinases, including PKCδ, MAPK, ERK, and Akt,^4,7 in comparison to the full length CDCP1. In our recent publication in Oncogene Journal² we reported that TNBC expresses high levels of phosphorylated cCDCP1 that forms a homo-dimer. The importance of homo-dimerization for CDCP1s activity was underscored by the ability of a soluble extracellular fragment of cCDCP1 to inhibit dimerization of cCDCP1 and its downstream signaling. This cCDCP1 fragment decreased PKCδ phosphorylation, inhibited TNBC migration in 2D, inhibited invasion and proliferation and increased apoptosis in 3D hydrogels.

Thus, to date, the CDCP1 activation cascade includes 3 levels: phosphorylation, cleavage, and dimerization. Currently there are antibody therapeutics developed that inhibit CDCP1 cleavage.⁷ There are also therapeutics capable of indirect inhibition of CDCP1 cleavage, such as dexamethasone,³ a steroid therapeutic, and aprotinin (Trasylol) (NCT00354900, trial in advanced breast cancer terminated). However, the last 2 therapeutics have substantial off-target effects. Steroids in particular not only decrease patient quality of life, but may also cause disease conditions like hypertension and hyperglycemia, which increase the risk of co-morbidities. Our discovery of the involvement of cCDCP1 extracellular domain in dimerization opens up a new avenue for therapeutic development. This approach is justified by multiple successful FDA-approved therapies that inhibit homo- and hetero-dimerization of cell surface receptors driving tumor progression. These include trastuzumab targeting extracellular domain IV of HER2, inhibiting ligand-independent homo-dimerization (HER2-HER2), as well as ligand-independent hetero-dimerization (HER2-HER3 and also HER2-HER1 and HER2-HER4); pertuzumab targeting extracellular domain II of HER2, inhibiting ligand-dependent hetero-dimerization with the other HER family members, especially HER3; cetuximab targeting domain III of EGFR/HER1 inhibiting its interaction with the ligands, which is needed for a conformational change stimulated HER1 homo-dimerization or hetero-dimerization with the other HER family members. Thus, dimerization of cell surface receptors represents a rational therapeutic target and further research is needed to test the existing antibodies recognizing cCDCP1 for their ability to block its dimerization, as well as development of new antibodies specifically inhibiting dimerization. In addition the soluble extracellular fragment of cCDCP1, identified in our study as an inhibitor of cCDCP1 dimerization and its downstream signaling, represents starting material for development of a potent protein-based agent. Since CDCP1 cleavage stimulates its dimerization, therapeutics blocking CDCP1 cleavage should indirectly block its dimerization.

Importantly, CDCP1 knock-out mice are viable, have no developmental and postnatal pathology, and no apparent decrease in life span as shown by Spassov et al.⁴ and Wright et al. (unpublished observations), indicating that CDCP1 blocking agents should be well tolerated. In support of this, administration of antibodies that inhibit CDCP1 cleavage to mice bearing prostate tumors did not cause any notable side effects, but reduced metastatic burden.⁷ Thus, CDCP1 might represent an example of a protein exploited by cancer cells to promote their metastatic ability, the function of which is dispensable in normal cells.

In conclusion, CDCP1 represents a rational therapeutic target for the treatment of TNBC, where current targeted therapies are failing. Since CDCP1 is expressed and cleaved in multiple carcinomas in addition to TNBC, including prostate, colorectal, lung and pancreatic carcinomas, further research is needed to evaluate its dimerization status in those cancers, and therapeutic potential of dimerization-blocking agents.