CAPÍTULO 3: ANÁLISIS DE RESULTADOS
3. Todos y todas construyen la inclusión
In this dissertation the application of nocoRNAc to the genome of Strepto- myces coelicolor has been described. S. coelicolor is an important antibiotic pro- ducing model organism. To a large extent the mechanisms involved in the reg-
9.2. nocoRNAc uncovering putative ncRNA regulators in Streptomyces coelicolor ulation of the production of these secondary metabolites have remained unex- plored. Within the SysMO STREAM consortium S. coelicolor wild type and dif- ferent mutant strains were grown under various conditions and the transcriptome, proteome and metabolome of the organism was profiled in unprecedented de- tail [109, 164, 99, 13, 153, 3]. The transcriptomics data was used to validate the expression of predicted ncRNAs and to characterize them further [65].
Of 403 predicted ncRNA transcripts that were measured 317 showed expression under the tested conditions. It turned out that the expression of a high number of predicted cis-encoded antisense RNAs correlates with the expression of their protein-coding target in many cases. In addition, the expression of several predicted ncRNAs in intergenic regions has been confirmed. Some of them show a clear reaction to the nutrient limitation event and might be involved in the regulation of metabolic processes.
Furthermore, a systematic study of potential RNA-RNA interactions between predicted ncRNAs and mRNAs identified an element putatively involved in the reg- ulation of antibiotics production. For the ncRNA transcript, which was predicted upstream of the cold shock protein csp, an interaction with the mRNA of TetR, a global downregulator of antibiotics production, was predicted. In an earlier study by Mart´ınes-Costa et al. [100] the region upstream of csp has been introduced into S. coelicolor using high copy number plasmids, which led to an upregulation of an- tibiotics production. The microarray screen indicated that the ncRNA is transcribed and if the predicted interaction with the TetR mRNA takes place in vivo, this would explain why an overrepresentation of that locus in the cell induces a stronger an- tibiotics production. In this case the expression of TetR would be downregulated due to the interaction and this would silence its repressive influence on antibiotics production. In this context nocoRNAc’s ability to calculate z-scores and p-values for predicted interactions proved to be extremely useful, as the TetR mRNA would not be among the top-scoring candidates if only the free energy of the interaction would have been considered.
It has to be noted that the predicted interaction site between the csp-ncRNA and TetR is located in the middle of the TetR mRNA. Thus, the direct occlusion of the ribosome binding site cannot be the mechanism of regulation. However, the binding of the ncRNA might influence the secondary structure of the TetR mRNA in a way that represses translation or the degradation of the molecule is promoted. Furthermore, it has to be considered that the results presented in this thesis are purely based on in silico analyses, which are strengthened, however, by the study of Mart´ınes-Costa et al.. As a next step the transcription of the csp-ncRNA has to be verified using primer extension, for example. The insertion of a constitutive promoter could be used then to overexpress the ncRNA, possibly confirming the promoting effect on antibiotics production. To my knowledge this would be the first known trans-encoded ncRNA that is increasing antibiotics production in Strepto- myces, while one trans-encoded ncRNA [163] and one cis-asRNA [35] decreasing antibiotics production have already been described.
Non-coding RNAs in S. coelicolor have also been reported in earlier studies. P´anek et al. [114] identified 32 ncRNA of which 15 where also found in our study. In ad-
dition, Swiercz et al. [149] detected 9 ncRNAs of which we found 2. Later studies, however, made use of deep RNA sequencing techniques for the genome-wide exper- imental detection of ncRNAs in S. coelicolor. By this, Suess et al. [162] identified 63 ncRNA of which 29 are located antisense to a protein-coding gene. In a very re- cent study Moody et al. [104] identified ncRNAs in S. coelicolor, S. avermitilis and S. venezuelae confirming a high degree of antisense transcription in all three species. The authors could show that the expression of ncRNAs including asRNAs is highly species dependent even if the sequence is perfectly conserved in the three genomes. This leads to two possible conclusions. First, in silico analyses, which are based on genomic sequences, can only give an indication of where ncRNAs might be found in the target genome. They are not able to predict in which species or under which conditions the elements might be expressed. Second, RNA-seq analyses are also lim- ited to some degree as they are only able to detect ncRNAs that are expressed in the investigated organism under the investigated conditions. Given the specific ex- pression of many identified ncRNAs RNA-seq analyses alone are probably also not powerful enough to disclose the full ncRNA repertoire of an organism. In addition, a further characterization of experimentally identified elements is still necessary. Thus, the integration of experimental results and computational analyses promises to be fruitful, e.g., to assess the identified elements with respect to conservation of sequence and structure or with respect to their RNA-RNA interaction potential. Thus, nocoRNAc offers the integration with the results of the automated TSS pre- diction presented in section 6.1.2, which works on differential RNA-seq data [142] and which will be discussed in detail in section 9.4. By this, all integrated methods for ncRNA characterization can be applied to the identified elements. In addition, nocoRNAc can assist in the determination of the ncRNA transcript’s 3’ end by the utilization of predicted transcription terminator signals. While the differential RNA-seq technique allows for a precise localization of a transcript’s 5’ start, the exact localization of the 3’ end remains much more challenging.