Minnis, C. J.; Zhang, H.-Y.; Gustavsson, E.; Lee, J.; Anderson, C.; Gardner, E.; Schulz, A.; Nickel, M.; Arno, G.; Jurkute, N.; Webster, A. R.; Gammaldi, N.; Santorelli, F. M.; Gissen, P.; Mills, P.; Ryten, M.; Mole, S. E.

Most genes are not yet fully annotated, and the extent of their transcript diversity and the roles and significance of specific isoforms is not understood. This information is therefore lacking for disease genes. The CLN3 gene underlies classic juvenile CLN3 disease, also known as juvenile neuronal ceroid lipofuscinosis, a rare paediatric neurodegenerative disorder. The most common cause of this biallelic disorder is a 1-kb intragenic deletion that removes two internal coding exons (exons 7 and 8). Here, we report findings from the first long-read RNA sequencing targeting CLN3 in blood samples derived from control individuals and from patients clinically and genetically diagnosed with juvenile CLN3 disease. We find that CLN3 transcription is complex, with >80 different transcripts encoding >35 different open reading frames (ORF) of different lengths, and no dominantly expressed transcript. The 1-kb deletion has direct consequences on this. This is consistent across patients, with total loss of some transcripts including those encoding the canonical 438 amino acid protein and other significant smaller isoforms. The highest expressed disease transcripts include those lacking exons 7 and 8 and encoding a 181 amino acid protein isoform, and other novel isoforms that lack additional exons and encode longer ORFs. The different effects on transcription of other CLN3 disease-causing variants are revealed in single patients. Together, these findings confirm the complexity of transcription at the CLN3 locus, reveal the impact of the 1-kb deletion and other variants on isoform abundance, and highlight the importance of understanding the contribution of these isoforms to CLN3 function in health and disease. Moreover, they impact the future design and development of personalised therapeutics and the design and generation of disease models. Finally, they underline the importance of full annotation for disease genes.