Lange nicht-codierende RNAs (lncRNAs) in der Reaktion auf DNA Schäden in Arabidopsis thaliana
Zusammenfassung der Projektergebnisse
In Arabidopsis thaliana seedlings that were exposed to DNA double strand break (DSB)- inducing ionizing radiation (X-rays) we discovered 86 novel up- or downregulated X-ray responsive transcripts that were annotated as long non-coding RNAs (lncRNAs). In this project we want to investigate if some of these transcripts play a role in the DNA-damage response (DDR), i.e. the recognition and/or repair of DNA damage. To date, in plants no lncRNAs involved in the DDR are known. XlincR is one of the most highly X-ray induced loci and annotated as a lncRNA in the Arabidopsis genome and lncRNA databases. It is also highly induced by other DNA damaging treatments like UV-C irradiation, zeocin and bleomycin, and its expression is dependent on SOG1, which is the plant master regulator of the DDR. Moreover, it is also induced by salt stress, which indirectly induces DNA damage, as it leads to a ROS accumulation that oxidises DNA. We discovered that XlincR contains a ~60 amino acids ORF and demonstrated that this peptide is expressed in plants. It carries a peroxisome targeting signal (PTS1) at the C-terminus, and in tobacco agroinfiltration experiments we observed that it is indeed located in peroxisomes. Consequently, XlincR was renamed Short Peroxisomal Protein Induced in DDR 1 (SPPiDDR1). Protein BLAST searches with SPPiDDR1 yielded several hits in most dicot genomes, suggesting that several SPPiDDR1 paralogs coexist within each plant genome. The SPPiDDR N- and C-termini including the PTS1 signal are highly conserved in all plant species, suggesting that SPPiDDR peptide function in plants is under positive selection pressure. The A. thaliana genome harbors three SPPiDDR paralogs, all on chromosome 4. The SPPiDDR1 and 2 promoters, but not SPPiDDR3, contain a SOG1 binding site and were shown to be direct targets for SOG1. SPPiDDR1 and 2 are both induced to high levels only one hour after UV-C irradiation, whereas SPPiDDR3 was not induced at all after 1 hour but strongly expressed after 24 hours. Under salt stress, sppiddr1 and sppiddr2 single mutants are less affected than wild type plants. The double mutant sppiddr1,2 is even less growth retarded than the single mutants, suggesting that SPPiDDR1 and 2 act partially redundantly. Under salt stress the sppiddr1,2 plants resemble the phenotype of salt tolerant sog1 mutant plants. Although it appears likely, it has to be proven more rigorously if this reflects a role in DDR and/or in oxidative stress response. Besides XlincR/SPPiDDRs we also study other DSB-induced lncRNAs. Most progressed is the work on LINDA and LNC4. LINDA (At3g00800) has one ChIP-seq confirmed SOG1 binding site in its promoter region and is SOG1-dependently induced only 30 to 60 minutes after treatment with UV-C or zeocin. For functional studies we created a deletion mutant by CRISPR/Cas12a. Preliminary phenotyping experiments indicate a slightly reduced sensitivity of the linda mutant compared to wild type seedlings against zeocin and UV-C and more extended cell death in the root meristem in comparison to wild type. To better understand the role of LINDA in the plant DDR, we performed RNA-seq analysis of linda deletion mutants after treatment with zeocin or UV-C. The RNA-seq data are currently analyzed. Another interesting candidate gene was initially described as a cluster of four lncRNAs and two hypothetical proteins. However, RACE PCR revealed that the whole cluster, covering a region of 5327 bp, is transcribed into one 1793 bp long lncRNA with 10 exons rather than six independent transcripts, which we termed LNC4. LNC4 induction by DNA double-strand breaks, peaking 6 hours after zeocin treatment, is strictly ATM and SOG1 dependent. Three lnc4 T-DNA mutants exhibit interesting, but different phenotypes: lnc4-1 has albino cotyledons and is unable to develop true leaves. lnc4-2 seedlings appear like wild type, but are more sensitive to zeoccin. For lnc4-3 no homozygotes can be isolated. Thus, we assume the mutation is embryo lethal. To further investigate the biological relevance of LNC4 we started to generate CRISPR/Cas12a deletion mutants, but this project could not be completed before the end of the funding period.
