Oxygen sensor molecules assist breast cancer cells to disseminate

In August Cell Reports published an article called “The Cancer Cell Oxygen Sensor PHD2 Promotes Metastasisvia Activation of Cancer-Associated Fibroblasts”. We asked the authors, Prof Mieke Dewerchin, Prof Peter Carmeliet and Dr Anna Kuchnio from the Laboratory of Angiogenesis and Neurovascular Link, Vesalius Research Center, to comment on their research.

The study

Cells possess systems to “sense” whether oxygen is present. An important oxygen sensing system are the enzymes named prolyl hydroxylases (also called prolyl hydroxylase domain containing enzymes or PHDs). These PHD enzymes act by modifying particular proline residues in their target proteins by hydroxylation, which can only happen in the presence of oxygen. In this study, we found that one of such PHDs, PHD2, is present in cancer cells and mediates the production by the cancer cells of molecules (the growth factor TGF-b1) that activate neighbouring non-cancer cells (stroma cells), more particularly the tumor fibroblast cells (cancer-associated fibroblasts or CAFs). Activated CAFs then deposit bundles of collagen fibres in the extracellular matrix, which represent “highways” via which the cancer cells invade the surrounding healthy tissue and reach blood vessels via which they can escape and disseminate to distant organs. This PHD2 mechanism was discovered using a mouse model of breast cancer combined with genetic techniques to reduce the expression of PHD2 in the cancer cells. The study showed that PHD2-deficiency in the cancer cells did not affect primary tumor growth but substantially reduced cancer invasion and metastasis (Figure 1). General PHD2-deficiency in all cell types had the same metastasis-inhibitory effect. This was in part due to reduced CAF activation and reduced formation of collagen fibre “highways” as described above, and in part to “normalization” of the otherwise malformed and leaky tumor blood vessels. This normalization effect results in more normally formed, stronger vessels with better oxygenation and tightened, less leaky, endothelium, contributing to reduced cancer cells escape and metastasis (Figure 1).

Figure 1: Loss of one PHD2 allele (PHD2+/-) reduces the activation of cancer-associated fibroblasts (CAF) and extracellular matrix (ECM) changes in tumors, and strengthens (“normalizes”) the tumor blood vessels. This dual effect results in reduced cancer cell escape and metastasis. Reprinted from Cell Reports 2(6), Kuchnio et al, “The cancer cell oxygen sensor PHD2 promotes metastasis”, pp 992-2005, Copyright 2015, with permission from Elsevier.

Background

This study was part of our priority research line into the molecular mechanisms underlying the formation of blood vessels in health and disease. We previously had shown a role of PHD2 in normalization of malformed tumor blood vessels and in tumor metastasis by using mouse models of transplanted cancer cells. Other groups also studied the role of PHD2 in cancer cells using tumor xenograft models, reporting opposing effects on tumor growth, likely due to differences among the tumor models and/or the approaches used to block PHD2 (general inhibition, cell type selective inhibition, etc). Of note, none of these previous studies used models of spontaneous tumor formation as occurs in patients. In this study, we wanted to address the role of PHD2 in a mouse model of spontaneous breast cancer, the most frequent cancer and second leading cause of cancer death in women, primarily due to metastasis.

Future directions

The discovery of the prominent dual role of the oxygen sensor PHD2 in CAF activation and tumor vessel normalization facilitating breast cancer metastasis provided important novel insights with potential clinical implications. To explore the clinical relevance, we also blocked PHD2 at a later time point, after tumor onset, to mimic better the patient situation. Also in these conditions, metastasis was reduced. Furthermore, analysis of patient databases revealed that lower levels of PHD2 in human breast cancer samples correlated with reduced expression of markers of CAF activation. Together, our findings warrant further investigation of the potential of inhibiting the CAF-dependent pro-metastatic activity of PHD2 for therapy of tumor metastasis.