S94 Identification of Natural Targets of Nonsense-Mediated Decay Relevant to Pulmonary Vascular Diseases

Introduction and Objectives Nonsense-mediated decay (NMD) is a quality- control mechanism that degrades RNA transcripts harbouring premature stop codons and consequently reduces production of truncated proteins. Inhibition of NMD is being evaluated as a therapeutic approach for pulmonary vascular diseases caused by pathogenic nonsense substitutions causing hereditary haemorrhagic telangiectasia (HHT). As this non-specific approach might also affect the expression of transcripts that are naturally regulated by NMD, our aim was to identify exons that are controlled by NMD and the biological processes they are involved in. Methods Primary human microvascular endothelial cells (HMEC) were cultured to confluence in antibiotic-free medium before treatment for 1 hour with 100µg/ml cycloheximide to inhibit NMD, or fresh media. Ribosomal (r)-RNA-depleted total RNA was used to prepare strand-specific whole transcriptome libraries which were sequenced on an Illumina Genome Analyser II, aligned to hg18, counted using custom scripts, and normalized to total valid reads and exon size. Further scripts were written to identify exons present in HMEC treated with cycloheximide but not media-treated HMEC. Separately, blood outgrowth endothelial cells (BOECs) were established from 23 HHT patients with pathogenic nonsense substitutions in ENG, ACVRL1 and SMAD4. Results In the cycloheximide and media-treated normal HMEC, there were alignments to 15,756 RefSeq genes, and 113 micro (mi)RNAs. The 419 most differentially expressed RefSeq genes (p<0.15), clustered to Gene Ontology (GO) biological process compatible with the observed induction of membrane proteolysis in cycloheximide-treated cells, validating the methodological approach. There were overlaps between miRNAs that were differentially expressed, and their mRNA targets predicted by Targetscan. The approach also identified candidate alternate exons observed only in the cycloheximide-treated HMEC, including 333 alternate first exons, 662 mid exons, 275 terminal exons and 59 exon extensions. Candidate exons that introduced a premature stop codon into transcripts of genes involved in GO biological processes other than protein translation were validated by reverse transcriptase PCR, prior to selection as a panel to quantitatively evaluate NMD inhibition in BOECs from HHT patients. Conclusion Natural targets of nonsense-mediated decay in HMEC were identified. Further investigation should provide new insights into the role of NMD in cellular physiology.