ИСТИНА

Motivation and Aim: Mycoplasma gallisepticum is a cause agent of avian chronic respiratory disease. This disease has been responsible for major declines in house finch populations, damages commercial poultry industry over the world. Besides that, mycoplasmas are very interesting object for fundamental biology. These bacteria lack cell wall, their genome is reduced and contains few transcription factors. At the same time mycoplasmas demonstrate high adaptive potential. Recently, we demonstrated that transcription about half of coding sequences changed and antisense transcription was activated under heat stress [2]. However proteomic profile underwent few changes and demonstrated little correlation with transcriptional response. The causes of this phenomenon are not clear. The observed massive transcriptional response can potentially have negligible contribution to the stress adaptation [3]. Therefore, we used a combination of RNA-seq and ribosome profiling to dissect adaptive and non-adaptive responses. In addition we aimed to identify the determinants that guide RNAs towards effective translation under the stress. Methods and Algorithms: Mycoplasma gallisepticum S6 was cultured in liquid medium for exponential phase and then cells were exposed to sublethal heat stress at 46°C for 30 min. Isolation of ribosomes was performed as previously described [1]. Libraries of cDNA were sequenced in SOLiD platform. Read mapping was performed using Bowtie software. Quantitative analysis was performed using R packages as previously described [2]. Results: Under exponential growth, mRNA abundance in the ribosome-bound pool of mRNA showed a correlation of 0.87 with the total RNA pool. But under heat stress correlation drastically decreased. The profile of ribosome-bound mRNA in stress was more similar to the one under control conditions rather than to total mRNA profile. Using the multiple regression method, we built a model to predict mRNA abundance in the ribosome-bound pool of mRNA. Total mRNA abundance, RBS (ribosome binding site) strength, stability of secondary structure near RBS and GC content of the transcript were used as predictor variables. Our model explains 72 and 38% of total variance in exponential growth and heat stress, respectively. Total mRNA abundance was the best predictor for mRNA abundance in the ribosome-bound pool of mRNA under exponential growth, but under heat stress, it was RBS strength. Ribosome-bound mRNA demonstrated a phenomenon of high upregulation of lowabundant mRNAs in the stress. Previously using RNA-seq we demonstrated that numerous transcriptional terminators decrease efficiency under the stress. This effect results in the terminator read-throung and appending of downstream genes to the transcript. Using these data we found that 40% of low-abundant transcripts are upregulated in ribosome- Systems Biology and Bioinformatics • the Ninth International Young Scientists School • SBB-2017 31 bound mRNA due to this mechanism. Our previous study indicate that the heat stress induces large amount of noise to the transcription and the ribosome profiling generally confirms this observation. We propose a method to distinguish adaptive response from the noise-like. In our previous research we showed that the abundance of the majority of transcripts in mycoplasmas is lower than one copy per cell [2]. Using our previous data we calibrated absolute abundance of transcripts identified by RNA-seq to the copy number. We identified that under the stress a small amount of genes increase abundance to 1 copy per cell, while the rest of genes, highly induced in terms of fold-change are still expressed in the amount significantly less than one copy per cell. The latter should be considered as transcriptional noise. The genes involved in adaptive response code for chaperons, cell division proteins, immunoglobulin-binding proteins, and uncharacterized protein. Conclusion: Under heat stress, transcriptional response in M.gallisepticum is characterized as bimodal, not equal changes. The first mode is noise but the second is adaptive response. Ribosome binding to mRNA is every important and critical step for generation specific cell response. Numerous perturbations on the level of transcriptionare not transmitted by ribosomes to the level of translation. Availability: Deep-sequencing (translatomics) data were uploaded to the NCBI SRA database, accession number PRJNA301561. Acknowledgements: The work was funded by the Russian Science Foundation grant 14-24-00159, ”Systems research of minimal cell on a Mycoplasma gallisepticum model”. References 1. Fisunov G.Y., Evsyutina D.V., Arzamasov A.A., Butenko I.O., Govorun V.M., Profiling of Mycoplasma gallisepticum ribosomes. Acta Naturae. 2015. 2. Mazin P.V., Fisunov G.Y., Gorbachev A.Y., Kapitskaya K.Y., Altukhov I.A., Semashko T.A., Alexeev D.G., Govorun V.M., Transcriptome analysis reveals novel regulatory mechanisms in a genome-reduced bacterium. Nucleic Acids Res. 2014. 3. Fisunov G.Y., Evsyutina D.V., Garanina I.A., Arzamasov A.A., Butenko I.O., Altukhov I.A., Nikitina A.S., Govorun V.M., Ribosome profiling reveals an adaptation strategy of reduced bacterium to acute stress. Biochimie. 2016.