Inflammation and angiogenesis are intimately related processes and contribute to essential physiological responses, such as host defence and wound healing, but also promote tumour growth and metastasis. In a recent edition of Nature, West et al. describe a new mechanistic link between these two processes; they show that inflammation leads to the release of oxidized phospholipids that drive angiogenesis by activating Toll-like receptor 2 (TLR2) on endothelial cells.

Leukocytes can promote oxidative stress at inflamed sites by contributing to the production of carboxyalkylpyrrole protein adducts (CAPs), such as ω-(2-carboxyethyl) pyrrole (CEP). The authors found that, during wound healing, the level of CAPs in the injured tissue increased transiently, with the peak production of CAPs coinciding with a period of intense neovascularization. In particular, the tissue levels of CEP were markedly increased during wound healing, and much of this CEP was associated with tissue macrophages. Futhermore, CEP levels were found to be persistently elevated in highly vascularized melanomas.

As these data suggested that CEP may be associated with inflammation-induced vascularization, the authors investigated whether CEP had direct pro-angiogenic effects. In in vitro assays using human or mouse endothelial cells, CEP was found to promote angiogenic responses as effectively as the classical pro-angiogenic mediator vascular endothelial growth factor (VEGF). However, CEP-induced responses occurred independently of VEGF receptor signalling, suggesting that CEP promotes angiogenesis through a different pathway. The fact that other adducts from the same family of CAPs could also promote angiogenesis suggested that the endothelial cells were responding to a particular molecular pattern, rather than a specific chemical moiety, and led the authors to speculate that TLRs could be involved in CEP recognition.

To explore this possibility, they examined the ability of CEP to promote angiogenesis in the absence of TLR signalling. In in vitro assays, antibody-mediated blockade of TLR2, but not TLR4, inhibited CEP-induced branching and growth of endothelial cells, suggesting that CEP might be recognized by TLR2 on endothelial cells. In further support of this, CEP could not promote angiogenic responses in TLR2-deficient endothelial cells, and a synthetic TLR2 ligand promoted angiogenic responses to a similar extent as CEP and VEGF in wild-type endothelial cells.

The authors extended these finding in vivo, showing that in mouse models of ischaemia and wounding, exogenous CEP promoted vascularization and accelerated tissue repair in wild-type mice but not in TLR2-deficient animals. In keeping with the in vitro data, experiments with bone marrow chimaeras showed that TLR2 expression by non-haematopoietic cells was necessary for the pro-angiogenic effects of CEPs. Administration of CEP-specific blocking antibodies to wild-type mice markedly impaired tissue repair and vascularization during wound recovery, indicating that endogenously generated CEP contributes to these responses. Similarly, treatment with CEP-specific antibodies decreased vascularization and delayed tumour progression in a mouse model of melanoma, suggesting that endogenous CEP can also promote cancer-associated angiogenesis. The authors suggest that blocking CEP could therefore be an effective treatment for patients with cancers that are resistant to current VEGF-targeting therapies.