Interactions between lipopolysaccharide and hypoxia in upregulation of equine endothelial cell activation

Andrew Brooks1, Nicola Menzies-Gow2, Simon Bailey3, Fiona Cunningham1, Jonathan Elliott1

1Veterinary Basic Sciences, Royal Veterinary College, London, UK, 2Veterinary Clinical Sciences, Royal Veterinary College, London, USA, 3Faculty of Veterinary Science, University of Melbourne, Melbourne, Australia

Lipopolysaccharide (LPS; maximal effect at 10 μg/ml) and hypoxia stabilise HIF-1α in equine digital vein endothelial cells (EDVEC) and pre-incubation with a sub-maximal concentration of LPS (10 ng/ml) prior to induction of hypoxia with 5% O2 synergistically enhanced the response (Brooks et al. 2008). Here we report the effects of exposure to LPS before induction of hypoxia on equine neutrophil (PMN) adhesion to EDVEC and on endothelial monolayer permeability. The roles of p38 MAPK and HIF-1α in mediating the responses to these stimuli were also investigated.

Cultured EDVEC obtained from mixed breed horses killed at an abattoir were exposed to LPS (10 pg/ml - 10 μg/ml; E. Coli 055:B5), 5% O2 or LPS at a sub-maximal concentration (10 pg/ml (permeability) or 10 ng/ml (adhesion)) and then to 5% O2 1h later. SB203580 (p38 MAPK inhibitor; 10μM) or the HIF-1α inhibitors, geldanamycin and YC-1 (10μM), were added 1h prior to LPS or 5% O2. PMN adhesion was quantitated by measuring the myeloperoxidase activity of adherent cells at 30min and EDVEC permeability by the passage of FITC-conjugated dextran. Results are expressed as mean % adherent PMN or % total FITC ± SEM (corrected for basal values) and analysed using 1- or 2-way repeated measures ANOVA and Bonferonni’s test. * = p<0.05 versus the appropriate control. N = 6 (adherence assays) or 4/5 (permeability assays without and with inhibitors, respectively).

LPS (10μg/ml) and 5% O2 each increased PMN adhesion, by a maximum of *16.7±0.5 and *4.7±0.5%, respectively (after 2h; basal adherence: 11.8±0.2%). Addition of LPS (10ng/ml) to EDVEC prior to 5% O2 induced a greater than additive increase at 2h (*20.9±1.7% versus 4.1±1.4% (LPS) or 4.7±0.5% (5% O2)). SB203580 decreased the response to LPS alone (*2.9±0.7% versus 11.7±0.2% adherent cells and LPS + 5% O2 (*4.8±0.6% versus 11.8±0.6% adherent cells) but not 5% O2. Inhibition of HIF-1α was without effect. LPS and 5% O2 also each increased endothelial cell permeability (maximum at 4h of *39.8±6.6% and *22.8±3.6% total FITC for LPS (10μg/ml) and 5% O2, respectively; basal permeability: 17.8±1.3%). Exposure of EDVEC to LPS (10pg/ml) prior to 5% O2 led to a more than additive response at 4h (*39.9±8.3% total FITC versus 6.4±1.3% (LPS) or 22.8±3.6% (5% O2)). Inhibition of either p38 MAPK or HIF-1α reduced the response to LPS (*-5.4±1.4% (SB203580), *7.1±3.9% (YC-1), *-11±0.7% (geldanamycin) versus 26.9±4.6% total FITC) and to 5% O2 (*-11.5± 0.5% (SB203580), *-3.8±2% (YC-1), *-7±2.5% (geldanamycin) versus 17±5% total FITC).

These data suggest that LPS in the circulation of endotoxaemic horses has the potential to cause local vascular leakage and increase PMN accumulation and enhance these responses in hypoxic tissue. Inhibition of p38 MAPK may limit these adverse actions on the endothelium but the effects of hypoxia are only likely to be partially reduced; targeting HIF-α would appear to be of more limited value.

Brooks et al. (2008) Fundam Clin Pharmacol 22 (suppl. 2):82