Studying COX-2 in the cardiovascular system using a transcriptomic approach

Abstract

At the time of starting my thesis there were no agreed mechanisms that could explain how cyclooxygenase (COX)-2 protects the cardiovascular system and no ‘hypothesis generating’ approaches done to try to find new answers. I have used transcriptomics, to come up with targets for this problem. I used microarray and RNA-seq on the following experimental models; (i) tissue from COX-1/COX-2 knockout mice, (ii) wildtype mice treated with the COX-2 inhibitor paracoxib and (iii) whole blood from healthy male volunteers after taking the COX-2 inhibitor celecoxib. Transcriptomic data was also analysed from online databases including (i) Open TG-GATEs and (ii) Human Protein Atlas. Altered transcriprt lists generated using GeneSpring and StrandNGS where analysed using manual curation, g:Profiler and GeneOntology.org databases and using Gene Set Enrichment Analysis. No transcripts were altered in aorta, heart, kidney, blood or brain from COX-1 knock out mice but >260 transcripts were altered (>1.5 fold) in the renal medulla of COX-2 knockout mice or wildtype mice where COX-2 was blocked with paracoxib with expression of four cardiovascular genes, namely, Agxt2, Edn1, Cd34 and Ptgds being similarly altered in both approaches. Transcripts were also altered in blood of humans taking celecoxib or COX-2 knockout mice, although in blood it was transcripts associated with inflammation and interferon pathways that came up. As an unexpected observation, it was found that expression of the oncogene Rgl1 was down regulated in all tissues from the COX-2 knockout mice. Since drugs that block COX-2 also prevent cancer the Rgl1 hit was explored using human cancer cells. This work helps us understand new ways that COX-2 protects the cardiovascular system and causes cancer. Finally, after more tests, the pathways I found could be used to help identify and treat people at risk of having heart attacks when they take pain killers that work by blocking COX-2.