Tonight, I watched the London Calling 2019 session by Irina Chelysheva from the University of Hamburg in Germany. The session’s title is “Small, modified, and highly structured: the challenge of tRNA sequencing.” Their lab is interested in sequencing tRNA because it is challenging. The major problems for sequencing, Chelysheva noted, include the short length of tRNAs (76-90 nt), structural stability, and modifications. The group designed a library preparation workflow starting with full-length RNA, primer annealing and ligation, and attachment of adapters. Notably, the team designed a specific adapter to the 3′ CCA tail of tRNAs. They began with in vitro transcribed standard tRNA molecules. The first nanopore sequencing attempt was for proline tRNA from E. coli. They obtained 100,000 reads and were unable to align using the Needleman-Wunsch algorithm… but this didn’t work! Chelysheva then used local alignment and optimized the library prep procedure, improving alignment. However, the number of reads for each tRNA did not correspond to the input proportion. Chelysheva used a variety of tRNAs and mixed them in known proportions. They identified the problem as missing sequences in both the 5′ and 3′ ends. The group then poly-A-tailed in vitro transcribed tRNAs. The mix of four tRNAs with poly-A tails increased the alignment score with longer alignments. There are still some issues with precision. Chelysheva tried sequencing total native tRNA. In E. coli, there are forty-seven different tRNAs encoded by 86 genes. However, upon sequencing they were unable to align. Numerous modifications to the tRNA have been made that are a problem for nanopore sequencing. The extensive modifications are challenging for signal base calling. Chelysheva is optimistic that tRNA nanopore sequencing will allow the exploration of novel tRNA structures and modifications. Chelysheva noted that they had only been working with nanopore sequencing for half a year!
