Non-canonical splicing and novel gene products revealed by direct RNA sequencing of adenovirus transcripts

Adenovirus is a common human pathogen of the gut, eye, and lung. During infection of the host cell, the cell nucleus is reorganized as the virus redirects cellular processes towards production of viral RNAs and viral DNA genome replication. Adenovirus is reliant on cellular processes for production and processing of viral RNA. Studying how this virus uses cellular machinery led to seminal discoveries of fundamental cellular processes such as RNA splicing. Although adenoviral promoters, splice junctions, and cleavage and polyadenylation sites have been previously mapped, previous studies relied on low-throughput biochemical techniques or short read cDNA-based sequencing technologies that do not fully capture the complexity of the adenoviral transcriptome. We used the Oxford Nanopore Technologies minION platform to perform long-read direct RNA sequencing during adenovirus infection. Analysis of these data allowed us to map both transcription start sites, and cleavage and polyadenylation sites genome-wide. This analysis confirmed the canonical viral early and late RNA cassettes, but direct analysis of splice junctions within long RNA reads also uncovered 20 unique viral RNAs expressed during infection. These RNAs include 7 new splice junctions which lead to expression of canonical open reading frames, as well as 13 RNAs encoding for messages that potentially alter protein functions through truncations in open reading frames or fusion of genes. We detected RNAs that bypass canonical RNA cleavage sites and generate potential chimeric proteins by linking separate gene transcription units. Among these novel chimeric proteins, we discovered the E4orf6/DBP gene product generated from both the E4 and E2 transcriptional units. Our work highlights how long-read sequencing technologies can reveal further complexity within viral transcriptomes.