Nanopore sequencing unveils the complexity of the cold-activated murine brown adipose tissue transcriptome

Abstract

Alternative transcription increases transcriptome complexity by expression of multiple transcripts per gene and thus fine tunes cellular identity and function. Annotation and quantification of transcripts at complex loci using short-read sequencing is non-trivial. Recent long-read sequencing methods such as those from Oxford Nanopore Technologies (ONT) and Pacific Biosciences aim at overcoming these problems by sequencing full length transcripts. Activation of BAT thermogenesis involves major transcriptomic remodelling and positively affects metabolism via increased energy expenditure and endocrine factors. Here we comprehensively benchmark features of ONT long-read sequencing protocols compared to Illumina shortread sequencing assessing alignment characteristics, gene and transcript detection and quantification, differential gene and transcript expression, transcriptome reannotation and differential transcript usage (DTU). We find that ONT sequencing is superior to Illumina for transcriptome reassembly and reduces the risk of false-positive events due to the ability to unambiguously map reads to transcripts, at the expense of statistical power for calling differentially expressed features. We identified novel isoforms of genes undergoing DTU in cold-activated BAT including Cars2, Adtrp, Acsl5, Scp2, Aldoa and Pde4d, validated by RT-qPCR. Finally, we provide a reannotation of the murine iBAT transcriptome as a valuable resource for researchers interested in the molecular biology underlying the regulation of BAT.