Transcriptome analysis of thermophilic methylotrophic Bacillus methanolicus MGA3 using RNA-sequencing provides detailed insights into its previously uncharted transcriptional landscape

Abstract

Background: Bacillus methanolicus MGA3 is a thermophilic, facultative ribulose monophosphate (RuMP) cycle methylotroph. Together with its ability to produce high yields of amino acids, the relevance of this microorganism as a promising candidate for biotechnological applications is evident. The B. methanolicus MGA3 genome consists of a 3,337,035 nucleotides (nt) circular chromosome, the 19,174 nt plasmid pBM19 and the 68,999 nt plasmid pBM69. 3,218 protein-coding regions were annotated on the chromosome, 22 on pBM19 and 82 on pBM69. In the present study, the RNA-seq approach was used to comprehensively investigate the transcriptome of B. methanolicus MGA3 in order to improve the genome annotation, identify novel transcripts, analyze conserved sequence motifs involved in gene expression and reveal operon structures. For this aim, two different cDNA library preparation methods were applied: one which allows characterization of the whole transcriptome and another which includes enrichment of primary transcript 5'-ends.

Results: Analysis of the primary transcriptome data enabled the detection of 2,167 putative transcription start sites (TSSs) which were categorized into 1,642 TSSs located in the upstream region (5'-UTR) of known protein-coding genes and 525 TSSs of novel antisense, intragenic, or intergenic transcripts. Firstly, 14 wrongly annotated translation start sites (TLSs) were corrected based on primary transcriptome data. Further investigation of the identified 5'-UTRs resulted in the detailed characterization of their length distribution and the detection of 75 hitherto unknown cis-regulatory RNA elements. Moreover, the exact TSSs positions were utilized to define conserved sequence motifs for translation start sites, ribosome binding sites and promoters in B. methanolicus MGA3. Based on the whole transcriptome data set, novel transcripts, operon structures and mRNA abundances were determined. The analysis of the operon structures revealed that almost half of the genes are transcribed monocistronically (940), whereas 1,164 genes are organized in 381 operons. Several of the genes related to methylotrophy had highly abundant transcripts.

Conclusion: The extensive insights into the transcriptional landscape of B. methanolicus MGA3, gained in this study, represent a valuable foundation for further comparative quantitative transcriptome analyses and possibly also for the development of molecular biology tools which at present are very limited for this organism.

Figure 1

Overview of the classification of putative TSSs within the B. methanolicus MGA3 genome sequence. (A) Schematic illustration of the different categories which were used for the classification of TSSs based on the respective genomic context. Putative TSSs are depicted as angled black arrows and are identified as described in section “Preparation of two different cDNA libraries for high-throughput sequencing” using the read starts obtained from RNA-seq data of enriched 5′-ends cDNA library. TSSs located in the upstream region and in coding direction of known CDSs (gray arrows) were classified as single TSSs or multiple TSSs. All TSSs overlapping in sense direction with known CDSs were categorized as novel intragenic TSSs. TSSs without annotated features downstream were classified as novel intergenic TSSs (black arrow), while TSSs antisense to annotated CDSs were classified as novel antisense TSSs (black arrow). (B) Process of TSSs analysis which includes the identification, filtering, manual verification and classification of putative TSSs. After manual inspection TSSs that belong to rRNA/tRNA and false-positive TSSs were removed from the automatically detected set, whereas the manually detected TSSs were added. The complete set of verified TSSs was divided into subsets depending on their genomic context.

Figure 2

Absolute number of identified transcription start sites in correlation to the length of their 5′-untranslated regions (5′-UTRs). The 1,642 TSSs located upstream or in coding direction of known CDSs were used to determine the length of the 5′-UTRs for each CDS. The 5′-UTR length was calculated as the distance between an identified TSS to the next TLS. The absolute number of TSSs is grouped in 5 bp intervals of 5′-UTR lengths (1–5, 6–10, 11–15 etc.), whereas the most distant right bar represents all 5′-UTRs longer than 500 bases.

Figure 3

Distribution of nucleotides within the ribosome binding sites and translation starts of B. methanolicus MGA3. The analysis of the nucleotide distribution in translation start sites and ribosome binding sites of B. methanolicus MGA3 was based on the TLS and upstream regions of genes for which a 5′-UTR was identified in the present study. The conserved TLSs and RBSs motifs were determined by using the motif-finding program Improbizer [23]. The conservation of a specific nucleotide at certain position is measured in bits and represented in the illustration by the size of the nucleotide. The depicted sequence logo was created with the software WebLogo [24].

Figure 4

Distribution of nucleotides within the −10 and −35 regions of B. methanolicus MGA3 promoter regions. The conserved sequences were determined by using the Improbizer motif-finding program [23]. For this analysis, the upstream regions of the 1,642 TSSs located in the 5′-UTR of annotated protein-coding genes were used. Conserved -10 motifs were detected in 1,619 sequences (98.6%), whereas 1,616 of the analyzed sequences contributed to identification of the -35 motif (98.4%). The conservation of a specific nucleotide at certain position is measured in bits and represented in the illustration by the size of the nucleotide. The hexamer of the core -10 region is underlined. The position values below the nucleotides are represented in relation to the positions of the identified TSSs, while the two spacers represent the mean distance between extended -10 region and TSS or -10 and -35 region, respectively. The depicted sequence logo was created with the software WebLogo [24].

Figure 5

Analysis of operon structures and comparison of the number of genes assigned to monocistronic transcripts, primary operons and suboperons, identified in B. methanolicus MGA3. The bars represent the different categories of transcripts. Within each category the number of genes is highlighted with a color code as depicted in the legend below.