The landscape, properties, and determinants of transcriptional activation of endogenous transposable elements in grapevine (Vitis vinifera L.) : A thesis submitted in partial fulfilment of the requirements for the Degree of Doctor of Philosophy at Lincoln University

Abstract

Transposable elements (TEs) are an intrinsic mutagen of eukaryotic genomes and have been proposed to be important in increasing genetic diversity in plants. It has been known that biotic and abiotic stress treatments induce TE transcription, the first stage in TE mobilisation. This research began with an investigation of TE transcription activity in grapevine embryogenic callus subjected to biotic stressors (Botrytis cinerea extracts and live Hanseniaspora uvarum cultures) to determine the location and regulation of autonomous TEs. Short-read RNA sequencing (RNAseq) has been commonly used to determine TE transcription patterns at a family level. This research sought to further these approaches by establishing an analysis pipeline to identify the expression of individual TE loci from Illumina RNAseq data. We efficiently identified that only 1.7%-2.5% of total annotated TE loci were transcribed in our system. This work identified a strong tendency for TE expression candidates to be found within introns of expressed genes. It was also discovered that these pairs of TEs and genes shared the same differential expression patterns in response to applied stressors. Our analysis pipeline was successfully validated using publically available RNAseq datasets from Arabidopsis, wild-type and epigenetic mutant (ibm2 and ddm1) lines, and Drosophila datasets of amyotrophic lateral sclerosis (ALS) models exhibiting a TE transcriptional storm. We successfully identified an Arabidopsis COPIA-93 locus previously proven to mobilise in ddm1 mutant and a subset of Drosophila TE loci that potentially contributed to full-length autonomous TE transcripts in the ALS models that have not been previously reported. Oxford Nanopore Technology (ONT) cDNA sequencing was deployed to determine whether autonomous TEs were being expressed as a precursor of mobilisation. Only low levels of full-length transcription of one Gypsy-V1 locus and three hAT-7 loci was detected in this data, suggesting rare intact transcription from autonomous TE loci despite stress treatments. This finding suggested that TE mobilisation might require inhibition of the epigenetic silencing system. We, therefore, treated embryogenic callus with the histone deacetylase inhibitors (HDACi), trichostatin A (TSA) or 4-phenylbutyric acid (4PBA), to alter the heterochromatic architecture of callus cells. Only the 4PBA treatment showed a noticeable shift in the transcriptional landscape of TE transcription, significantly increasing the proportion of intergenic TE loci in the expression candidate pool and resulting in significant up-regulation of 2,059 TE loci. ONT cDNA sequencing of these samples detected very low levels of intact sequencing reads from different yet a single Gypsy-V1 locus and six hAT-7 loci. Five genes participating in the RNA-dependent DNA methylation (RdDM) pathway (AGO2, AGO4, RDR1, RDR6, and NERD) were upregulated, suggesting that callus exposed to 4PBA responded by an enhancement of RdDM, maintaining effective control of TE transcription and therefore TE mobility. Overall, this thesis contributes to the understanding of the landscape, properties, and determinants of transcriptional activation of endogenous transposable elements, revealing the closely connected transcriptional relationship between TEs and co-localised genes. These findings shed light on the genetic and epigenetic impact of endogenous TE activation on genes in nature.