A hybrid transcriptome assembly reveals coding and non-coding landscapes in Drosophila hemocytes.

Abstract

Innate immune responses triggered by Drosophila larval hemocytes have been extensively characterized. However, the full extent of transcriptional and post-transcriptional regulation underlying these processes remains poorly understood. Here, we employed a hybrid sequencing strategy integrating Oxford Nanopore long-read and Illumina short-read sequencing to provide a more comprehensive transcriptome annotation. This enabled discovery of full-length transcripts with novel 5' and 3' boundaries and uncovered 349 previously unannotated long non-coding RNAs highly induced during late stages of wasp infestation. To ensure high confidence transcript models, we further eliminated potential intra-priming artifacts specific to long-read cDNA data. This high-confident full-length transcript models helped to reveal cell type-specific lncRNA markers in lamellocytes and crystal cells by single-cell analyses, which recapitulated hemocyte differentiation trajectories. Notably, RNAi-based depletion of two highly induced lncRNAs impaired lamellocyte formation under wasp infestation, highlighting their functional relevance. Collectively, our findings provide detailed insights into the Drosophila larval immune transcriptomes through the long-read sequencing and highlight the regulatory roles of non-coding RNAs in innate immunity.