Effect of genetic deletion of cyclo-oxygeanse isoforms on endothelial cell morphology in the mouse aortic arch

Malak Alyamani1,2, Martina Helena Slingsby5, Nicholas S Kirkby1, Blerina Ahmetaj-Shala1, Timothy D Warner3, Jane A Mitchell1. 1National Heart and Lung Institute,Imperial college London, London, UK, 2King Fahad Cadiac Center, Riyadh, Saudi Arabia, 3William Harvey Research Institute,Barts and The London School of Medicine and Dentistry, Charterhouse Square,London, UK, 4Institute for Idraet og Ernaering, Bygningen, Universitetspaken 13,2001 Kobenhavn, Denmark, 5Department of Nutrition, Exercise and Sports Integrative Physiology University of Copenhagen Universitetsparken, Copenhagen, Denmark

Endothelial cells (EC) are subjected to shear stress (the frictional force of blood flow) that affects morphology and susceptibility to inflammation. ECs exposed to non-directional shear stress (such as in the lesser curvature (LC) of the aortic arch) have a cobblestone morphology and are prone to inflammation. By contrast, ECs exposed to unidirectional shear stress (such as in the greater curvature (GC) of the aortic arch) are elongated, aligned in the direction of shear and are protected from inflammation (1). Prostacyclin is an important cardio-protective hormone produced by ECs, whose synthesis is dependent on the concerted actions of cyclo-oxygenase (COX) and prostacyclin synthase enzymes. COX is present in two isoforms, COX-1 that is constitutively expressed and COX-2 that is induced at the site of inflammation. We have previously shown that COX-1 is abundantly expressed in ECs of all regions of the aortic arch whilst COX-2 expression is restricted to a discreet area within the LC (2). As a result the release of prostacyclin by whole aortic arches is very largely dependent upon COX-1 and not COX-2 (2). The effect of COX activity on EC morphology is poorly understood, while selective roles for COX-1 and COX-2 on EC morphology within complex vascular structures such as the aortic arch have never been addressed. Here we have used confocal imaging of mouse aortic arches to determine the influences of COX-1 and COX-2 on EC morphology. Wild type (WT, C57BL/6), COX-1-/- and COX-2-/- mice (n=6, 10-12 weeks old) were killed by CO2 and EC nuclei stained with 4’,6’-diamidino-2-phenylindole and luminal surfaces imaged as described previously2. EC morphology was studied by measuring nuclear alignment, density and size using standard applications in ImageJ (3). In wild type and COX-1-/- mice ECs of the GC were more aligned, had a lower nuclear density and smaller nuclear size than those lining the LC region of the arch. In COX-2-/- mice the ECs in the GC and LC were similarly aligned. Moreover, nuclear average size for ECs lining the LC was significantly reduced in COX-2-/- mice (Figure 1).

Figure 1: Effect of loss of COX-1 or COX-2 (-/-) on endothelial cell morphology in the greater (GC) or lesser (LC) curvature of the mouse aortic arch. Data is the mean±SEM for n=6 mice. #p<0.05, t-test; *p<0.05 one-way ANOVA followed by Dunnett\\'s post-tests.

This data suggests that COX-2 specifically affects EC morphology in regions associated with inflammation where genetic deletion results in morphological changes towards a ‘protected’ EC phenotype. The functional consequences of these observations remain the subject of investigation

1. Iiyama, K. et al (1999). Circ Res 85, 199-207.

2. Kirkby, N. S. et al. (2012) Proc Natl Acad Sci U S A 109, 17597-17602.

3. http://imagej.nih.gov/ij/.