Posttranscriptional regulation with regard to cleavage factor I (CFI) associated alternative polyadenylation and methyltransferase A (MTA) assisted m6A modifications in Arabidopsis thaliana

Changing the expression of genes, posttranscriptional modifications and other regulatory pathways enable rapid adaption to both mild and extreme conditions in all domains of life. Recent research improved the understanding of posttranscriptional modifications, which are affecting the fate of a transcript. Alternative polyadenylation (APA) is associated with many regulatory aspects. It describes the addition of the poly(A) tail at alternative sites inside the transcript. This is achieved by the cleavage and polyadenylation machinery. Cleavage factor I (CFI) is a complex associated with APA and well-studied in humans, but not in Arabidopsis. N6-methyladenosine (m6A) is one of the most frequently observed RNA modifications, which is established by methyltransferase A (MTA) and other m6A writer proteins in Arabidopsis thaliana.<br /> This study revealed RNA binding sites of CFI subunit proteins GFP-CFI59 and CFI68-GFP transcriptome-wide in Arabidopsis thaliana using iCLIP2. It connected those and the binding sites of GFP-CFI25 to polyadenylation sites (PAS), altered in the knock out cfi25a and cfi59cfi68 mutants. Dominant binding behaviour at about 50 nt upstream of the distal PAS, which is primarily used in wild-type, was determined. In cfi mutants the PAS choice is predominantly shifted towards a proximal PAS. Thus, I conclude that the CFIs promote the usage of the distal PAS. This has to be confirmed in further experiments, for example with in vitro cleavage assays. A few more target transcripts are upregulated, than downregulated in cfi mutants. Thus, the usage of the proximal PAS, due to the loss of CFI, might affect the stability of the transcript. The UGUA binding motif, published for the human homologs, could be confirmed to be enriched for GFP-CFI25. Meanwhile GFP-CFI59 and CFI68-GFP tend to avoid the direct binding onto the motif, while their binding is enriched next to it. This might be due to a blocking of the motif by GFP-CFI25. Binding or avoiding of the motif and its relatives should be confirmed in vitro with electro mobility shift assays (EMSAs). Additionally, the presence of this motif seems to affect the stimulation of the CFIs onto PAS choice, as the 3’UTR of transcripts with altered PAS present a higher number of it. A higher affinity with oligonucleotides, containing multiple copies might be confirmed with EMSAs, too. As binding sites of GFP-CFI59 and CFI68-GFP were determined via two experiments each; further replicates might improve the identification of those. To investigate the impact of an altered PAS onto the transcript level, RNAseq should be done, as the PATseq data is biased towards all polyadenylated transcripts. This could reveal a more frequent presence of non-polyadenylated or differentially alternative spliced transcripts as well. The later might be interesting to analyse, because especially CFI68-GFP binding was proven to be enriched in the intronic sequences around splice sites, while GFP-CFI59 binding was only enriched for intronic sequences of 3’ splice sites. GFP-CFI25 binding sites are slightly enriched in exonic sequences around both kinds of splice sites. As translation might be altered with differential PAS choice, a proteomic approach might shed light onto consequences regarding the protein level. The mutants cfi25a and cfi59cfi68 showed a two-hour delay of the transcript level peaks of the circadian regulated genes CAB2, GRP7, TOC1 and LHY. Whereas in cfi59 the peaks were detected to occur one hour earlier than in the wild-type. This impact was not observed in cfi68. However, as the single mutants cfi59 and cfi68 were only analysed in one replicate and the mutants cfi25a and cfi59cfi68 in two replicates, which presented partly variable results; this has to be elucidated in additional replicates. The knockout mutant of cfi25a was confirmed and cfi59cfi68 was discovered to flower early. Flowering experiments with cfi68 revealed a milder early flowering phenotype, while cfi59 did not present one. Probably this is due to the altered PAS choice in the knockout mutants, caused by missing CFI binding, onto flowering associated transcripts, such as FKF1, GI and SOC1. <br /> Apart from that this study investigated the (pre-)mRNA binding sites of MTA in Arabidopsis thaliana with iCLIP2, transcriptome-wide and linked them to published m6A sites. Binding and m6A sites are both predominantly located in the 3’UTR. The analysis revealed an enrichment of binding sites in close proximity (-35 till +60 nt) to m6A marks, but also a high variability of m6A occurrence among already published data sets. Thus, I conclude the epi-transcriptomic methylome is highly dynamic and has to be studied under identical conditions as the binding behaviour of the proteins. Uncovering the binding sites of the other known methylation writer proteins MTB, FIP37, VIR and HAKAI, could support the understanding of the methylation machinery further. The binding behaviour of m6A erasers should be investigated to unravel this highly dynamic posttranscriptional modification. In vitro methylation assays might clarify the responsibility of one m6A writer protein for m6A establishment at a dedicated transcript site. Finally, a better understanding of the impact onto modified transcripts would complete the picture.<br /> My thesis supports the understanding of APA by CFI and m6A modification by MTA in Arabidopsis. This contributes to the uncovering of posttranscriptional modifications and their consequences for subsequent processes in the cell.

Level thesis:degree_level

thesis.doctoral

Grantor dc:publisher

Universität Bielefeld

Year

2024

Author dc:creator

• Kahmen, Sara

Repository record source_url

https://pub.uni-bielefeld.de/record/2990435

OAI identifier oai:identifier

oai:pub.uni-bielefeld.de:2990435