Measuring transcriptomic diversity induced by genome SCRaMbLEing with nanopore direct RNA sequencing
"What if you could build a genome from scratch?" Aaron opened with this question, stating that writing entire genomes from scratch is now possible, and affordable, as the cost of DNA synthesis is rapidly decreasing. Increasingly ambitious projects have been initiated in recent years to synthesise genomes.
Aaron then introduced the model organism that he works with - yeast. After~300 million years of divergence, organisation of the genome between yeast species is highly conserved, with Aaron giving the example of the galactose (GAL) regulon; yet research data have revealed extensive transcript heterogeneity. For example, per gene there are over 26 major transcript isoforms with coding potential, of which at least 10 are typically needed to explain 80% of expression. Genome architecture is the key factor in this extensive transcriptome heterogeneity.
Aaron outlined the goal of the synthetic yeast genome project (Sc2.0), which is to synthesise all yeast chromosomes and combine them into a single strain. The project also aims to facilitate engineering of these synthetic genomes, by introducing LoxP sites that enable "on-demand" stochastic genome rearrangement using the Cre-recombinase. This technique is called Synthetic Chromosome Rearrangement and Modification By LoxP-mediated Evolution (SCRaMBLE), and results in heterogeneous populations of individual yeast strains with different genomic structures. Aaron works on such a synthetic yeast strain, and investigates how abrupt changes to genome organisation impact the transcriptome.
Aaron discussed his use of Direct RNA sequencing to successfully detect novel transcriptional events, functional isoform changes, and quantitative changes in gene expression, which result from abrupt changes to genome organisation. He stated that in his experience, short-read sequencing has limited ability to understand what happens to reorganised isoforms in the genome as lengths to do not span the entirety of rearranged LoxP sites. Using the Direct RNA sequencing Kit (SQK-RNA002), Aaron has analysed 67 SCRaMBLED yeast strains, using 106 MinION flow cells, obtaining a total of 120 million reads across all the strains, with average feature coverage of 30X. From these data, many novel junction types have been identified; interestingly, most of the individual junctions are unique, as 75% of junctions are present in less than 4 strains. Isoform changes are common in SCRaMBLEd synthetic chromosomes, and these changes are extensive on both the 5' and 3' ends of the SCRaMBLEd gene feature.
To conclude his talk, Aaron summarised that the aim of his work has been to try to understand if genome organisation is an engineer-able thing, and therefore ultimately whether we could build a predictive model from our understanding of the effects of known genomic alterations on the transcriptome. From the application of nanopore sequencing, he described how he has been able to successfully investigate the consequences of engineered genome rearrangement, and start to predict these effects at the transcriptome level.