Neuropilin (NRP)-1 is a co-receptor for vascular endothelial growth factor (VEGF). Preclinical data suggest that blockade of NRP1 suppresses tumour growth by inhibiting angiogenesis, in addition to directly inhibiting tumour cell proliferation in certain models. A humanized monoclonal antibody to NRP1 is currently being evaluated as a potential anti-cancer therapy in clinical trials. However, the expression of NRP1 in cancer and physiological angiogenesis has yet to be systematically described. Here we characterize the in situ expression of NRP1 in human cancer and during mammalian development. A monoclonal antibody to human NRP1 was generated and validated for immunohistochemistry by western blotting, use of formalin-fixed cell pellets transfected with NRP1, immunofluorescence, and comparison with in situ hybridization. NRP1 expression was assessed in whole sections of 65 primary breast carcinomas, 95 primary colorectal adenocarcinomas, and 90 primary lung carcinomas. An additional 59 human metastases, 16 xenografts, and three genetically engineered mouse tumour models were also evaluated. Immunoreactivity for NRP1 was seen in vessels from normal tissues adjacent to cancer and in 98-100% of carcinomas. Tumour cell expression of NRP1 was also observed in 36% of primary lung carcinomas and 6% of primary breast carcinomas, but no colorectal adenocarcinomas. NRP1 was evaluated in mouse embryos, where expression was limited to the nervous system, endocardium, vascular smooth muscle, and, focally, endothelium on subsets of vessels. Moreover, in a model of VEGF-dependent angiogenesis in the postnatal mouse trachea, blockade of NRP1 signalling resulted in defective angiogenesis and recapitulated the effects of anti-VEGF treatment. These observations confirm NRP1 as a valid anti-angiogenic target in malignancy, and as a potential direct anti-tumour target in a subset of cancers. The data also confirm a role for NRP1 in physiological, VEGF-mediated angiogenesis.

Vascular endothelial growth factor-A (VEGF) is an important regulator of vascular permeability. In preclinical studies, VEGF induces endothelial fenestrations in pre-existing and neo-vasculature, while inhibition of VEGF leads to a reduction in endothelial fenestrations. Recently, vascular regression in response to VEGF inhibition has been shown to correlate with the presence of endothelial fenestrations. Plasmalemmal vesicle-associated protein (PLVAP) is believed to be a component of diaphragmed endothelial fenestrations, but a direct relationship with VEGF signalling has not been established. The aim of this study was to characterize the expression pattern of PLVAP and investigate whether PLVAP is a transcriptional target of VEGF signal transduction. The expression pattern of PLVAP was characterized in normal and neoplastic human tissues by in situ hybridization and/or immunohistochemistry. The role of VEGF signal transduction in the regulation of PLVAP expression was investigated in vitro using receptor-selective engineered forms of VEGF, a neutralizing monoclonal antibody against VEGF, and inhibitors of downstream signalling pathways. PLVAP mRNA and protein were widely expressed in the endothelium of normal and neoplastic tissues. In cultured endothelial cells, VEGF signalling through receptor 2 stimulated expression of PLVAP total RNA and protein. This induction could be blocked with an anti-VEGF monoclonal antibody and by inhibitors of phosphatidylinositol 3-kinase (LY294002) or p38 mitogen-activated protein kinase (SB203580), but not by PD98059, a mitogen-activated protein/extracellular signal-regulated kinase 1 inhibitor. These data show that PLVAP is more widely expressed in the vasculature of normal tissues than previously thought and that it is expressed in the vasculature of most human tumours. We suggest that PLVAP is a downstream target of VEGF signalling. This work solidifies the association between VEGF and the appearance and maintenance of fenestrations by providing a potential mechanistic link.

Despite the availability of new targeted therapies, ductal pancreatic adenocarcinoma continues to carry a poor prognosis. Carcinoembryonic antigen-related cell adhesion molecule (CEACAM)6 has been reported as a potential biomarker and therapy target for this malignancy. We have evaluated CEACAM6 as a potential therapy target, using an antibody-drug conjugate (ADC). Expression of CEACAM6 in pancreatic adenocarcinomas was determined using immunohistochemistry on tissue microarrays. The expression pattern in granulocytes and granulocytic precursors was measured by flow cytometry. Murine xenograft and non-human primate models served to evaluate efficacy and safety, respectively. Robust expression of CEACAM6 was found in > 90% of invasive pancreatic adenocarcinomas as well as in intraepithelial neoplastic lesions. In the granulocytic lineage, CEACAM6 was expressed at all stages of granulocytic maturation except for the early lineage-committed precursor cell. The anti-CEACAM6 ADC showed efficacy against established CEACAM6-expressing tumours. In non-human primates, antigen-dependent toxicity of the ADC consisted of dose-dependent and reversible depletion of granulocytes and their precursors. This was associated with preferential and rapid localization of the antibody in bone marrow, as determined by sequential in vivo PET imaging of the radiolabelled anti-CEACAM6. Localization of the radiolabelled tracer could be attenuated by predosing with unlabelled antibody confirming specific accumulation in this compartment. Based on the expression pattern in normal and malignant pancreatic tissues, efficacy against established tumours and limited and reversible bone marrow toxicity, we propose that CEACAM6 should be considered for an ADC-based therapy approach against pancreatic adenocarcinomas and possibly other CEACAM6-positive neoplasms.

IntroductionBreast cancer, the most common cause of cancer-related deaths worldwide among women, is a molecularly and clinically heterogeneous disease. Extensive genetic and epigenetic profiling of breast tumors has recently revealed novel putative driver genes, including p21-activated kinase (PAK)1. PAK1 is a serine/threonine kinase downstream of small GTP-binding proteins, Rac1 and Cdc42, and is an integral component of growth factor signaling networks and cellular functions fundamental to tumorigenesis.MethodsPAK1 dysregulation (copy number gain, mRNA and protein expression) was evaluated in two cohorts of breast cancer tissues (n = 980 and 1,108). A novel small molecule inhibitor, FRAX1036, and RNA interference were used to examine PAK1 loss of function and combination with docetaxel in vitro. Mechanism of action for the therapeutic combination, both cellular and molecular, was assessed via time-lapse microscopy and immunoblotting.ResultsWe demonstrate that focal genomic amplification and overexpression of PAK1 are associated with poor clinical outcome in the luminal subtype of breast cancer (P = 1.29 × 10−4 and P = 0.015, respectively). Given the role for PAK1 in regulating cytoskeletal organization, we hypothesized that combination of PAK1 inhibition with taxane treatment could be combined to further interfere with microtubule dynamics and cell survival. Consistent with this, administration of docetaxel with either a novel small molecule inhibitor of group I PAKs, FRAX1036, or PAK1 small interfering RNA oligonucleotides dramatically altered signaling to cytoskeletal-associated proteins, such as stathmin, and induced microtubule disorganization and cellular apoptosis. Live-cell imaging revealed that the duration of mitotic arrest mediated by docetaxel was significantly reduced in the presence of FRAX1036, and this was associated with increased kinetics of apoptosis.ConclusionsTaken together, these findings further support PAK1 as a potential target in breast cancer and suggest combination with taxanes as a viable strategy to increase anti-tumor efficacy.Electronic supplementary materialThe online version of this article (doi:10.1186/s13058-015-0564-5) contains supplementary material, which is available to authorized users.

In the majority of lung, breast and colorectal cancers, the effects of anti-neuropilin-2 are likely to be restricted to the vasculature. These results will assist in pharmacokinetic evaluations, tolerability assessments and the choice of setting to evaluate the activity of anti-neuropilin-2 therapies.