Background Endothelial injury is a critical feature in the early stages of vascular disease. Inflammation and hypoxia are often associated with endothelial injury, stimulating the expression of several cytokines that include erythropoietin (EPO). Endothelial cell-derived EPO appears to be important for protecting the endothelium against ischaemic injury. A non-erythropoietic analogue of EPO; pyroglutamate helix B surface peptide (pHBSP) retains these protective properties of EPO without possessing its erythropoietic effects. The aim of our study was to assess the effects of these molecules in a model of endothelial injury under normoxic and hypoxic conditions.

Method The reparative effects of EPO and pHBSP were assessed under hypoxia (1% O₂) and normoxia (21% O₂) in an in vitro model of endothelial injury (scratch assay). A monolayer of bovine aortic endothelial cells (BAECs), grown to confluence in a multi-well plate, was scratched and the closure of the injured endothelial monolayer was assessed over 24 h. The effects of EPO and pHBSP on BAEC proliferation, chemotaxis and apoptosis were assessed under similar hypoxic conditions in separate experiments. The potential molecular mechanisms of these effects were also explored.

Results Both EPO and pHBSP enhanced scratch closure under hypoxic conditions by 13 ± 2.6%, and 10 ± 1.69% respectively (p < 0.01) compared to normoxic conditions (3.2 ± 0.9% and 2.9 ± 0.3% for EPO and pHBSP respectively p > 0.05). These effects appeared to be by promoting cell proliferation and migration of BAECs (p < 0.05). EPO also protected BAECs from staurosporine-induced apoptosis under hypoxic conditions. The priming effect of hypoxia was associated with stabilisation of HIF-1Î ±, EPO receptor (EPOR) up-regulation and decreased phosphorylation of endothelial nitric oxide synthase (eNOS) at the Ser-1177 residue. The effect of hypoxia on the latter was rescued by EPO. Hypoxia was associated with a reduction in nitric oxide (NO) production as assessed by its oxidation products nitrite and nitrate, and this was consistent with the oxygen requirement for the endogenous production of NO by NO synthase (NOS). Whilst EPO did not affect NO formation in normoxia, it markedly increased NO production under hypoxic conditions, in a NOS-dependent manner; its effects were inhibited by NOS inhibitors (e.g. L-NAME).

Conclusion and implication The tissue-protective properties of EPO-related cytokines are likely to be mediated by NO in pathophysiological settings associated with poor oxygenation. Further work should be directed towards an understanding of the cellular redox status and some of the signalling events down-stream of the emerging EPO/EPOR/NO axis that underpin its beneficial biological effects. These findings may be particularly relevant to atherogenesis and post-angioplasty restenosis.