Transcription Start Site Mapping and Small RNA Profiling of Leptospira biflexa serovar Patoc

doi:10.21203/rs.3.rs-2224410/v1

Download PDF

Research Article

Transcription Start Site Mapping and Small RNA Profiling of Leptospira biflexa serovar Patoc

https://doi.org/10.21203/rs.3.rs-2224410/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 21 Feb, 2023

Read the published version in World Journal of Microbiology and Biotechnology →

You are reading this latest preprint version

Leptospirosis is an emerging zoonotic disease caused by bacterial species of the genus Leptospira. However, the regulatory mechanisms and pathways underlying the adaptation of pathogenic and non-pathogenic Leptospira spp. in different environmental conditions remain elusive. Leptospira biflexa is a non-pathogenic species of Leptospira that lives exclusively in a natural environment. It is an ideal model not only for exploring molecular mechanisms underlying the environmental survival of Leptospira species but also for identifying virulence factors unique to Leptospira’s pathogenic species. In this study, we aim to establish the transcription start site (TSS) landscape and the small RNA (sRNA) profile of L. biflexa serovar Patoc grown to exponential and stationary phases via differential RNA-seq (dRNA-seq) and small RNA-seq (sRNA-seq) analyses, respectively. Our dRNA-seq analysis uncovered a total of 2,726 TSSs, which are also used to identify other elements, e.g., promoter and untranslated regions (UTRs). Besides, our sRNA-seq analysis revealed a total of 603 sRNA candidates, comprising 16 promoter-associated sRNAs, 184 5’UTR-derived sRNAs, 230 true intergenic sRNAs, 136 5’UTR-antisense sRNAs, and 130 open reading frame (ORF)-antisense sRNAs. In summary, these findings reflect the transcriptional complexity of L. biflexa serovar Patoc under different growth conditions and help to facilitate our understanding of regulatory networks in L. biflexa. Besides, the TSS and sRNA landscapes of L. biflexa can also be compared with its pathogenic counterparts, e.g., L. borgpetersenii and L. interrogans, to identify features contributing to their environmental survival and virulence.

Leptospira biflexa

leptospirosis

small RNA

transcription start site

transcriptome

Leptospirosis is a neglected zoonotic disease with a high morbidity and mortality rate across the globe (Picardeau et al. 2008). The main etiologic agents of leptospirosis are pathogenic species of the genus Leptospira, i.e., Leptospira borgpetersenii and Leptospira interrogans. Pathogenic Leptospira species have different modes of transmission, which are influenced by their environment, for instance, L. borgpetersenii usually spread via host-to-host transmission, whereas L. interrogans usually transmit through water (Picardeau et al. 2008). The strict host-to-host mode of transmission adopted by L. borgpetersenii may be attributed to genetic decay and loss of gene functions that hamper its survival outside mammalian hosts, therefore the strict dependence on the hosts for sustenance (Bulach et al. 2006).

One of the non-pathogenic, saprophytic species of the genus Leptospira is L. biflexa, which lives exclusively in soil or freshwater (Picardeau et al. 2008). Despite being a non-pathogenic bacterial species, L. biflexa is an ideal model for studying gene functions and features associated with the environmental survival of Leptospira species due to its fast growth rate and ease of genetic manipulation (Pětrošová and Picardeau 2014). Besides, L. biflexa may serve as a comparative reference that aids in identifying features that are unique to the pathogenic species L. borgpetersenii and L. interrogans, thereby providing insights into the evolution, environmental persistence, and virulence factors in pathogenic Leptospira species (Pětrošová and Picardeau 2014).

Pathogenic Leptospira species are exposed to hostile and ever-changing environments throughout their transmission cycle. Previous DNA microarray studies on L. interrogans exposed to various infection-associated stress conditions, e.g., iron depletion, temperature fluctuation, and osmotic stress, have revealed genes that facilitate the adaptation of pathogenic Leptospira species to diverse natural and host environments (Qin et al. 2006; Matsunaga et al. 2007; Lo et al. 2010; Xue et al. 2010). Compared to DNA microarray, another alternative known as the RNA-seq can provide a snapshot of the whole transcriptome while the former is only able to profile known transcripts using a probe-based hybridization method (Rao et al., 2019). Having the capacity to unravel the transcriptome architecture of bacteria, RNA-seq also provides information regarding gene expression levels, as well as reveals the co-regulated genes both in cis and trans. The rapid development of RNA-seq technologies over the past decade has witnessed the deposition of 10 RNA-seq datasets of Leptospira and two RNA-seq datasets of L. biflexa in the NCBI Bioproject database (Caimano et al. 2014; Iraola et al. 2016; Zhukova et al. 2017; Adhikarla et al. 2018; Gaultney et al. 2020; Zavala-Alvarado et al. 2020; Fule et al. 2021; Grassmann et al. 2021; Putz et al. 2021; Zavala-Alvarado et al. 2021). Despite this advancement, the transcriptional and post-transcriptional regulations of Leptospiral adaptation and virulence remain largely unknown, in particular, the transcriptional and post-transcriptional regulations of gene expression in Leptospira.

Driven by the poor apprehension of the Leptospiral transcriptional and post-transcriptional regulations, we present the transcription start sites (TSS) map and the identification of sRNAs, potential promoter motifs, and untranslated regions (UTRs) in L. biflexa. To the best of our knowledge, this is the first study reporting the TSS landscape of L. biflexa. It is anticipated that this study has the potential to improve the genome annotation in L. biflexa and subsequently augment our understanding of the complex gene regulatory networks in L. biflexa as compared to its pathogenic counterparts.

Bacterial growth conditions

Two hundred microliters of L. biflexa serovar Patoc (1X10⁹ lepto/ml) inoculum was inoculated into 10 ml of complete EMJH (1:50 dilution) in a 100 ml conical flask. This starter culture was grown at 30^°C with agitation at 60 rpm until reaching the mid-log phase (OD₄₂₀ of 0.4). Next, 200 microliters of the L. biflexa serovar Patoc starter culture prepared were inoculated into 200 ml of complete EMJH in a 1 L conical flask. The reading of OD₄₂₀ was taken using NanoPhotometer (Implen GmbH, Germany) at intervals of 6 hours for a total of 7 days to get a complete growth cycle. The growth curves were plotted using the triplicate OD₄₂₀ readings acquired for a period of 7 days.

RNA isolation

The total RNA extraction was performed using the TRI reagent (Molecular Research Center, USA). In brief, about 0.1 g of cell pellet was re-suspended in 1 mL of TRI reagent, and the cell lysate was thoroughly mixed with 0.2 mL of chloroform. The mixture was then incubated at room temperature for 10 minutes and centrifuged at 12,000 g and 4°C for 15 minutes. The upper aqueous layer was transferred to a new tube and mixed with 0.5 mL of isopropanol, followed by incubation at room temperature for 5 minutes and centrifugation at 12,000 g and 4°C for 10 minutes. The resulting supernatant was discarded and the pellet was washed with 1 mL of pre-chilled 80% ethanol via centrifugation at 12,000 g and 4°C for 8 minutes. The supernatant was then discarded and the pellet was air-dried before re-suspension in RNase-free water. The concentration and quality of the total RNA were determined with Bio-Photometer (Eppendorf, Germany). To access the RNA integrity, the total RNA was separated via denaturing urea-polyacrylamide gel electrophoresis (urea-PAGE: 7M urea, 8% polyacrylamide gel) and visualized under the Gel Doc XR + System (Bio-Rad Laboratories, Inc., USA).

Before constructing RNA-seq libraries, the total RNA samples were treated with DNase I to remove all traces of DNA. In brief, ~ 20 µg of the total RNA was added with 1X DNase I buffer, 10 U of DNase I, and 80 U of RiboLock RNase Inhibitor in a final volume of 50 µL, followed by incubation at 37°C for 30 minutes. Subsequently, the reaction mixture was added with 10 µL of 3 M sodium acetate, 40 µL of RNase-free water, and 10 µg of glycogen. The mixture was vortexed briefly and mixed thoroughly with 100 µL of phenol, followed by centrifugation at 12,000 g and room temperature for 5 minutes. The upper aqueous layer was transferred into a new tube and mixed thoroughly with 100 µL of chloroform, followed by centrifugation at 12,000 g and room temperature for 5 minutes. After that, the upper phase was transferred into a new tube and mixed thoroughly with 400 µL of pre-chilled absolute ethanol. After being incubated at -80°C for 30 minutes, the mixture was centrifuged at 12,000 g and 4°C for 20 minutes. The resulting supernatant was discarded and the pellet was washed with 1 mL of pre-chilled 80% ethanol via centrifugation at 12,000 g and 4°C for 8 minutes. Finally, the supernatant was then discarded and the air-dried pellet was re-suspended in 20 µL of RNase-free water.

Construction of cDNA libraries for the TSS mapping

For the TSS mapping, each total RNA sample was divided into two portions, the first portion was directly subjected to cDNA library (TEX-) construction, while the second portion was treated with Terminator™ 5’-phosphate-dependent exonuclease (TEX) to degrade 5’-monophosphorylated transcripts, leaving only the 5’-triphosphorylated RNAs (i.e., the primary transcripts) in the sample. In brief, 10 µL of the DNase I-treated RNA was added with 1X Terminator Reaction Buffer A, 2 U of TEX, and 60 U of RiboLock RNase Inhibitor to a final volume of 50 µL. Incubation was carried out at 30°C for 60 minutes and followed by phenol/chloroform purification as per the protocols described above. The resulting air-died pellet was re-suspended in 20 µL of RNase-free water. This together with the first, untreated portion, was subjected to the following treatments sequentially.

Tobacco acid pyrophosphatase (TAP) treatment

The RNA was incubated respectively with 1X TAP Reaction Buffer, 2.5 U of TAP, and 60 U of RiboLock RNase Inhibitor in a final volume of 50 µL. After being incubated at 37°C for 90 minutes, the mixture was purified with phenol/chloroform and the air-dried pellet was re-suspended in 20 µL of RNase-free water.

Ligation of 5’-adapter

The TAP-treated RNA was denatured at 90°C for 2 minutes, snap-cooled on ice for 2 minutes, and mixed with 100 picomoles of 5’-adapters (PE_DNAAdapt and PE_RNAAdapt), 1X T4 RNA ligase reaction buffer, 15 U of T4 RNA ligase, and 60 U of RiboLock RNase Inhibitor to a final volume of 30 µL, followed by incubation at room temperature for 60 minutes, 30°C for 30 minutes, and 37°C for 60 minutes. After that, the reaction mixture was added with 1X T4 RNA ligase reaction buffer, 10 U of T4 RNA ligase, and 40 U of RiboLock RNase Inhibitor to a final volume of 40 µL. The reaction was then incubated overnight at 4°C and purified with phenol/chloroform on the next day. Finally, the air-dried RNA pellet was re-suspended in 17.5 µL of RNase-free water.

Synthesis of cDNA: The 5’-adapter-ligated RNA was added with 100 picomoles of TRUmRN6 (with random hexamer), denatured at 90°C for 2 minutes, and snap-cooled on ice for 2 minutes, followed by the addition of 1X Transcriptor RT buffer, 20 U of Transcriptor Reverse Transcriptase, 60 U of RiboLock RNase Inhibitor, and 10 mM of dNTP to a final volume of 30 µL. The reaction mixture was sequentially incubated at 25°C for 10 minutes and 55°C for 40 minutes, followed by phenol/chloroform purification as per the above-described protocol. The air-dried pellet (first-strand cDNA) was re-suspended in 20 µL of RNase-free water and added with 1X FastStart PCR buffer containing 1.5 mM of MgCl₂, 2.5 U FastStart Taq DNA polymerase, 10 mM of dNTP, and 50 picomoles of primers complementary to the 5’- and 3’-adapters (1Mlx3 and 1Mlx5) to a final volume of 50 µL. The reaction mixture was subject to the following PCR amplification steps in a thermocycler (Bio-Rad Laboratories Inc., USA): denaturation at 94°C for 2 minutes, followed by 6 cycles of 94°C for 20 seconds, 62°C for 20 seconds, and 72°C for 25 seconds. The PCR products were subsequently purified using the High Pure PCR Cleanup Micro Kit (Roche, Germany) according to the manufacturer’s instructions and eluted in 15 µL of RNase-free water.

Barcoding: Prior to the barcoding of cDNA libraries, the double-stranded cDNA was purified to eliminate primer dimers (78 bp). Briefly, 15 µL of the purified cDNA samples were separated on a 2% low-melting agarose gel (1X TAE buffer) via electrophoresis at 100 V for 30 minutes. The desired gel fraction (100–200 bp) was excised for subsequent gel extraction using the E.Z.N.A. Gel Extraction Kit (Omega Bio-tek Inc., USA). After that, the eluted cDNA sample was added with 1X FastStart PCR buffer containing 1.5 mM of MgCl₂, 2.5 U FastStart Taq DNA polymerase, 10 mM of dNTP, and 50 picomoles of primers complementary to the 5’- and 3’-ends of cDNA library (TruseqUniv and IDX1-4, see Online Resource 1) to a final volume of 50 µL. The reaction mixture was subject to the following PCR amplification steps: denaturation at 94°C for 2 minutes, followed by 18 cycles at 94°C for 20 seconds, 63°C for 20 seconds, and 72°C for 30 seconds. The PCR products were then purified using the High Pure PCR Cleanup Micro Kit (Roche, Germany) and eluted in 15 µL of RNase-free water. Size selection and gel extraction were performed to eliminate adapter dimers (128 bp). A total of 15 µL of barcoded cDNA library were separated on a 2% low-melting agarose gel (1X TAE buffer) at 100 V for 30 minutes, and the desired fraction (150–250 bp) was excised out for gel extraction by using the E.Z.N.A. Gel Extraction Kit (Omega Bio-tek Inc., USA) and eluted in 30 µL of RNase-free water.

Amplification of cDNA libraries: To prepare a sufficient amount of starting materials for RNA-seq, each of the barcoded cDNA libraries was subjected to PCR amplification. This was achieved by the addition of 1X FastStart PCR buffer containing 1.5 mM of MgCl₂, 2.5 U FastStart Taq DNA polymerase, 10 mM of dNTP, and 50 picomoles of primers (IDXPCR and TruSeqUnivPCR) to a final volume of 50 µL. The PCR parameters were as follows: denaturation at 94°C for 2 minutes, followed by 15 cycles of 94°C for 20 seconds, 60°C for 20 seconds, and 72°C for 30 seconds with a final extension step at 72°C for 2 minutes. After that, the PCR products were purified using the QIAquick PCR Purification Kit (Qiagen, Germany) and eluted in 30 µL of RNase-free water.

Construction of cDNA libraries for sRNA-seq

For the sRNA-seq, the total RNA (100 µg) was subjected to electrophoresis on 8% denaturing urea-PAGE at 120 V for 60 minutes. Gel fractionation was carried out via excision of the bands that constitute RNA of less than 120 nts and overnight elution with 0.3 M sodium acetate (pH 5.2), followed by purification with an equal volume of chloroform and ethanol precipitation. The purified size-selected RNA samples were then subject to the TAP treatment as described above and the following treatments sequentially.

C-tailing and Ligation of 5’-adapter

The TAP-treated RNA was denatured for 2 minutes at 90°C, snap-cooled on ice for 2 minutes, and added with 20 mM of CTP, 1X tailing reaction buffer, 8 U of Poly(A) Polymerase, and 80 U of RiboLoRNase Inhibitor to a final volume of 40 µL. The reaction mixture was incubated at 37°C for 2 hours and purified with phenol/chloroform as described above on the next day. Finally, the air-dried RNA pellet was re-suspended in 22 µL of RNase-free water. After that, the C-tailed RNA was ligated with a 5’-adapter (PE_DNAAdapt and PE_RNAAdapt).

Synthesis of cDNA

200 picomoles of UningTruSeq8g were added to the C-tailed, 5’-adapter-ligated RNA, and the resulting mixture was denatured at 90°C for 2 minutes, snap-cooled on ice for 2 minutes, and added with 1X Transcriptor RT buffer, 20 U of Transcriptor Reverse Transcriptase, 60 U of RiboLock RNase Inhibitor, and 10 mM of dNTP to a final volume of 40 µL. The reaction mixture was sequentially incubated at 33°C for 10 minutes and 55°C for 45 minutes, followed by purification using Sephadex G-50 DNA purification Quick Spin Columns. The first-strand cDNA was eluted in 40 µL of RNase-free water and subject to the synthesis of the second strand of cDNA as per the above-described protocols.

Barcoding

Prior to the barcoding of cDNA libraries, the double-stranded cDNA was purified as described above to eliminate primer dimers (81 bp). After that, each cDNA library was barcoded with a unique identifier sequence (Online Resource 1) as described above and purified again to eliminate primer dimers (131 bp). To prepare a sufficient amount of starting materials for RNA-seq, the size-selected cDNA libraries were subject to PCR according to the above-described protocols.

RNA-seq

In summary, 4 cDNA libraries for TSS mapping and 2 size-selected cDNA libraries for sRNA-seq were outsourced for paired-end sequencing on the Illumina HiSeq 2000 platform (Beijing Genomics Institute, China) (Fig. 1). Raw sequence reads generated in this study have been submitted to the Sequence Read Archive (SRA) database and can be accessed using the BioProject accession: PRJNA884181.

Bioinformatics Analyses

The bioinformatics analysis workflow in this study is depicted in Online Resource 2. All bioinformatics tools and webservers are listed in Table S4 (Online Resource 1). All Shell scripts for bioinformatics analyses are provided in Online Resource 3.

Pre-processing of RNA-seq reads

All raw sequencing reads (FASTQ files) were interleaved with the SeqFu tool (Telatin, Fariselli, and Birolo 2021) using the script interleave.sh to merge paired-end read files into single, interleaved read file, which was then pre-processed to filter low-quality reads (quality score < 28) and remove adapter sequences with Cutadapt (Martin 2011) using the script trim.sh. The qualities of both pre- and post-processed reads were assessed with FastQC (Andrews 2010) and compiled with MultiQC (Ewels et al. 2016) using the scripts quality_report.sh and compile_report.sh, respectively.

Alignment to the reference genome: All trimmed paired-end reads were aligned to the reference genome of L. biflexa serovar Patoc strain (NCBI Accession: NC_010602, NC_010843, and NC_010844) with Bowtie2 (Langmead and Salzberg 2012) using the scripts index_genome.sh and mapping.sh. The resulting alignment (SAM) files were compressed into BAM files and indexed using the scripts convert2bam.sh and index_bam.sh, respectively. To generate coverage files from the alignment files for subsequent visualization, the BAM files were first converted to the bedgraph format, which was then converted to the bigwig format using the scripts generate_bedgraph.sh and bedgraph2bigwig.sh, respectively. After that, BAM files generated from TEX + and TEX- reads were subject to TSS mapping, while BAM files generated from sRNA-seq reads were subject to sRNA prediction as follows.

TSS mapping: TSS maps were extracted from BAM files generated from TEX + and TEX- reads using the TSSAR Client, which submitted them to the TSSAR Web Server (http://nibiru.tbi.univie.ac.at/TSSAR/server) for automated TSS annotation with a p-value threshold of 0.01, noise threshold of 10, and maximum merging range of 10 (Amman et al. 2014). Since the TSSAR Client can process BAM files with a single chromosome, the above BAM files were split by chromosome using the scripts splitbam_chr1.sh, splitbam_chr2.sh, and splitbam_p74.sh, respectively. The TSSAR Web Server generated a list of annotated TSSs, each of which is classified with respect to its genomic position relative to annotated genes as follows: primary TSS is located up to 250 nt upstream of an annotated gene, internal TSS is located within the open reading frame of an annotated gene, antisense TSS is located opposite to an annotated gene, and orphan TSS is located beyond the above-mentioned regions.

Predictions of other genomic features: The TSSs annotated in this study can be used to predict putative promoter regions and 5’UTRs. For the prediction of promoter regions, the genomic region within 50 nts upstream to each TSS was scanned for potential promoters (consensus sequence motif: “TANNNT”, “TTGACA”, “AAAAAARNR”, and “AWWWWWTTTTT”). Briefly, the entire genome sequences were scanned for promoter motifs with MEME FIMO (Bailey et al. 2009) using the scripts iupac2meme.sh and promoter_prediction.sh. The resulting list of potential promoter regions was screened for potential promoters located within 50 nts upstream to TSSs using the script promoter_TSS.sh. Meanwhile, 5’UTRs, which have not yet been made available in existing annotations, were predicted based on TSSs identified in this study using the script 5UTR.sh, which also simultaneously summarizes their length distribution and scans their sequences for any potential riboswitch and other highly conserved regulatory elements using the Infernal cmscan tool (Nawrocki and Eddy 2013) based on covariance models downloaded from the Rfam database (Nawrocki et al. 2015).

Identification and classification of novel sRNA candidates

The BAM files generated from sRNA-seq reads were subject to transcript assembly using the Rockhopper software, which is a graphical user interface (GUI) program for bacterial RNA-seq analysis, including the identification of novel transcripts, using the Bayesian approach (McClure et al. 2013). The predicted novel transcripts were extracted from the output file of Rockhopper software and converted to BED files for subsequent manipulations using the script sRNA_characterization.sh, which extracted the sequence of sRNA candidates in FASTA format and further categorized them according to their lengths and genomic locations relative to annotated genes into promoter-associated sRNAs, 5’UTR-derived sRNAs, 5’UTR-antisense sRNAs, ORF-antisense sRNAs, and possible individual intergenic sRNAs.

The total RNA samples extracted at mid-log and stationary growth stages of L. biflexa serovar Patoc were used to generate sRNA-seq libraries and dRNA-seq libraries, which were sequenced on the Illumina HiSeq 2000 platform.

Read pre-processing and alignments

The RNA-seq produced ~ 9.8–13.8 million raw read pairs per library, of which 3.5–9.7 million read pairs passed the pre-processing steps (Online Resource 1 and Online Resource 4). Trimmed reads are mapped to the reference genome (NC_010602, NC_010843, and NC_010844) to identify their locations for subsequent TSS annotation and sRNA identification. Here, the RNA-seq read pairs were aligned to the reference genomes. The resulting alignment rates range between 35.1% and 88.3% (Online Resource 1). We noticed that read pairs generated from TEX-treated samples not only have significantly low percentages of passing the pre-processing steps (28.1%-33.4%) but also yielded relatively low alignment rates (35.1%-52.1%). Furthermore, a close inspection of the trimming procedure revealed that the majority of reads generated from TEX-treated samples failed to pass the minimum length threshold of 20 nts, whereby 66.6% and 71.9% of read pairs generated from TEX-treated mid-log and stationary samples were too short and thus eliminated during trimming (Online Resource 4). However, there were still 1.3–1.8 million read pairs mapped to the reference genome for these TEX-treated samples.

TSS mapping revealed a total of 2,726 TSSs

In this study, the TSS mapping was performed based on the enrichment pattern of TEX + libraries relative to TEX- libraries using the TSSAR Client, which identified a total of 2,726 TSSs, of which 2,525 TSSs are located on chromosome I (NC_010602), 177 TSSs are located on chromosome II, and 24 TSSs are located on plasmid p74 (Online Resource 5). Among these 2,726 TSSs, we annotated 1,364 primary TSSs (pTSSs), 644 internal TSSs (iTSSs), 1,097 antisense TSSs (asTSSs), and 190 orphan TSSs (oTSSs), respectively (Fig. 2a). Notably, a number of them belong to more than one category (Fig. 2b and Online Resource 5).

On top of pTSSs, our dRNA-seq also uncovered a great number of TSSs of other classes, including iTSSs (644, 23.6%), which refer to TSSs located inside annotated genes in the same orientation, 1,097 asTSSs, which refer to TSSs located opposite to annotated genes, and 190 oTSSs, which refers to TSSs that are neither pTSSs, iTSSs, nor asTSSs (Sharma et al. 2010). These asTSSs and oTSSs represent the transcription of cis-encoded antisense and trans-encoded sRNAs in L. biflexa, respectively. However, only 28 asTSSs and 8 oTSSs are preceded by promoter motifs. An intersection with our sRNA candidates also revealed that only 17 asTSSs and three oTSSs potentially drive the transcription of antisense and intergenic sRNAs, respectively. In addition to the above-described TSS categories, we also identified 240 TSSs (~ 17.6% of pTSSs) that potentially give rise to leaderless transcripts, which refer to mRNAs that either entirely lack 5’UTRs or have extremely short (less than 6 nts) 5’UTRs. Furthermore, 158 out of those 350 leaderless transcripts have their TSSs located exactly at the start codons. In fact, TSSs associated with leaderless transcripts constitute exceptionally high densities across all chromosomes.

Other genomic features - Promoters and UTRs

The TSS annotation can provide additional information about promoters and UTRs. The 50-bp upstream regions of TSSs (which we considered to contain potential promoters where RNA polymerases bind and initiate transcription) were scanned for the presence of promoter motifs. A total of 123 promoters were identified for 130 TSSs, which comprise 93 pTSSs, 42 iTSSs, 28 asTSSs, and 8 oTSSs (Online Resource 5). About half of these TSSs (67 out of 130 TSSs) have promoter motifs centered within 30 bp of their upstream regions. In addition to promoter prediction, non-leaderless pTSSs also directly reveal the presence of 5’UTRs, which are non-coding regions extending from TSSs to start codons of mRNAs and potentially harbor cis-acting elements (e.g., riboswitches and RNA thermometers) that are implicated in the regulation of mRNA translation in response to changes in metabolite concentration or environmental cues. Here, we identified 1,124 5’UTRs for 901 genes with mean and median lengths of 89.45 nts and 77 nts, respectively. Long 5’UTRs are often associated with cis-encoded regulations, but unfortunately, the screening with covariance models downloaded from the Rfam database (Nawrocki et al. 2015) did not detect any matching to known riboswitch elements This is unexpected considering ~ 56% (627) of those 5’UTRs are long enough (> 60 nts) to carry cis-regulatory elements (Weinberg et al. 2007). However, we found potential 5’UTR-derived sRNA candidates in 16.4% of our 5’UTR data in subsequent analysis.

Novel sRNA candidates

The transcript assembly generated a total of 603 sRNA candidates, of which 546 candidates are located on chromosome I (NC_010602), 47 candidates are located on chromosome II (NC_010843), and 10 candidates are located on the plasmid p74 (NC_010844), which correspond well with the ratios of their lengths and number of genes (Online Resource 6). Among these candidates, there are only 33 sRNAs and 34 sRNAs with detectable promoters and TSSs within the upstream regions, respectively. We broadly divided these sRNA candidates into intergenic and antisense sRNAs with respect to the annotated genes.

Our RNA-seq analysis detected a total of 328 intergenic sRNAs, which were further subdivided into the following 3 categories based on their genomic locations: promoter-associated sRNAs, 5’UTR-derived sRNAs, and true intergenic sRNAs. Here, we identified 16 intergenic sRNAs that overlap the promoters of downstream genes in both sense and antisense orientations, potentially representing promoter-associated sRNAs that are usually transcriptional interference (TI) by-products (Online Resource 6). We also uncovered 184 5’UTR-derived sRNA candidates, which locate between pTSSs and start codons. After excluding the above two sub-categories of sRNAs, the remaining 230 intergenic sRNAs are true intergenic sRNA candidates, of which only 16 candidates and three candidates with promoters and TSSs, respectively.

In this study, we identified 275 cis-antisense sRNA candidates, of which there are 18 candidates with promoters and 26 candidates with TSSs (Online Resource 6). These antisense sRNA candidates were further categorized into the following two categories based on their location with respect to their opposite target genes: 5’UTR-antisense sRNAs and ORF-antisense sRNAs. Here, we detected 136 antisense sRNA candidates that overlap with 5’UTRs in antisense orientation, while the remaining antisense sRNA candidates (139) are transcribed in an antisense orientation to ORFs without overlapping 5’UTRs and start codons.

Bacterial transcripts contain both the primary transcripts (which have a 5’-triphosphate end) and processed transcripts (which contain 5’-monophosphate or 5’-hydroxyl groups). The differential RNA-seq approach aims to specifically identify the start sites of primary transcripts via the depletion of the processed transcripts. In this study, the RNA samples were divided into two parts. The first part was not subjected to any form of treatment to retain both the 5’-PPP-containing primary transcripts and 5’-P-containing processed transcripts. The second part was subjected to treatment with TEX, which specifically degrades the processed transcripts, enriching the primary transcripts (Sharma and Vogel 2014). Sequencing then reveals the characteristic enrichment patterns of the cDNA coverages at the 5’-end of the primary transcripts, facilitating the annotation of TSSs. Here, we report for the first time the TSS landscape of L. biflexa with a total of 2,726 TSSs, which can be further classified into pTSS, iTSS, asTSS, and oTSS. Among these TSSs, pTSSs are the most abundant class of TSSs, but precede only ~ 36% of annotated genes in L. biflexa serovar Patoc, probably due to limited culture conditions being covered in this study.

This study identified a myriad of iTSSs that represent the initiation of transcription from regions inside annotated genes of the same strand. It was previously believed that gene transcriptions are initiated upstream of the start codons, but recent studies have uncovered iTSSs in bacteria, such as E. coli (Thomason et al. 2015). In bacteria, iTSSs are postulated to be attributed to: (1) alternative transcription that yields transcript variants encoding for truncated proteins; (2) small, densely packed genomes of bacteria, in which the transcription of some genes, including 3’-UTR-derived sRNAs, is initiated from the coding regions of upstream genes (Miyakoshi, Chao, and Vogel 2015). Here, we identified 195 iTSSs that are also pTSSs of downstream genes, but only three sRNA candidates are preceded by iTSSs, indicating that the majority of iTSSs may represent the transcription of truncated transcripts. Despite relatively few pTSSs representing the transcription of annotated genes, the dRNA-seq analysis revealed abundant asTSSs and oTSSs, which represent the transcription of cis-encoded antisense and trans-encoded sRNAs in L. biflexa, respectively. In fact, asTSSs alone constitute ~ 40% of TSSs uncovered in this study. However, only 17 asTSSs and three oTSSs potentially drive the transcription of sRNA candidates uncovered in this study. The most plausible regulatory role of the antisense sRNAs is via their propensity to form a duplex with their corresponding strand, which then facilitates co-degradation or co-stabilization (Georg and Hess 2011; Cheah et al. 2018). However, the concomitant detection of both the sense and antisense RNAs is challenging, and this is aggravated by the relatively lower expression level of either one. Hence, the abundance of asTSSs and oTSSs lacking downstream detectable sRNA candidates might be attributed to transcriptional noise due to spurious promoters throughout the genome (Lloréns-Rico et al. 2016). This is also known as pervasive antisense transcription, which is the side product of gene expression that does not have definite functions (Raghavan, Sloan, and Ochman 2012; Martinez et al. 2022).

Leaderless transcripts refer to mRNAs that either entirely lack 5’UTRs or have extremely short (less than 6 nts) 5’UTRs. Here, our analysis found a fraction of pTSSs that give rise to leaderless transcripts, on which the initiation of translation relies on the direct binding of the 70S ribosomal subunit to the first nucleotide of the transcripts due to the lack of ribosome-binding site (RBS) or Shine-Dalgarno (SD) sequence (Shell et al. 2015). There are many factors that may influence the translational efficiency of leaderless transcripts, e.g., the presence of a phosphate group at the 5’-end, CA repeats near the start codon, the distance of the 5’-end to the start codon, and the stability of the RNA structure that surrounds the codon (Leiva and Katz 2022). The discovery of 158 out of those 350 leaderless transcripts that have their TSSs located exactly at start codons suggests their dependency on only 70S ribosomal subunit to initiate translation. Leaderless transcripts have been reported in a wide range of bacteria, but their actual functional implications remain unclear (Shell et al. 2015). Interestingly, it has been shown that the translation of some leaderless transcripts in E. coli is initiated by 70S ribosomes carrying 16S rRNA whose anti-SD sequence at the 3’-end has been removed by the stress-induced endoribonuclease MazF (Moll and Engelberg-Kulka 2012). Extending this to the context of L. biflexa, the translational activities of leaderless transcripts can also be triggered by certain conditions or stresses, and thus, are potentially functionally distinct from leadered transcripts. Under these adverse conditions, the translation of the leaderless transcripts can be initiated while that of the canonical leadered transcripts are somehow impeded.

Our sRNA-seq analysis yielded a total of 603 sRNA candidates, which can be further divided into intergenic and antisense sRNAs. Previous in silico screening for bacterial sRNAs often focused on intergenic regions (Livny and Waldor 2007). However, the current genome annotation often lacks features such as UTRs and promoters, potentially leading to erratic classification or even making the target prediction of sRNAs challenging. For example, using genome annotation to predict intergenic sRNAs can be challenging, as they might be derived from or antisense to UTRs of neighboring genes. However, our TSS map allows the prediction of 5’UTRs and promoters, which helps improve the classification and target prediction of sRNAs. TSS landscape (particularly pTSS) is useful for the prediction of the 5’UTRs of mRNAs, which potentially carry cis-acting elements (e.g., riboswitches and RNA thermometers). Previous studies showed that UTR-derived stable degradation products often enter small-size cDNA libraries, but some of them may be functional and serve as trans-antisense sRNAs, for instance, the riboswitches SreA and SreB in Listeria monocytogenes also act as trans-antisense sRNAs that regulate the expression of the virulence regulator PrfA (Loh et al. 2009).

Here, our RNA-seq analysis detected a total of 328 intergenic sRNAs, which were further subdivided into the following three categories based on TSSs and features predicted using TSSs: promoter-associated sRNAs, 5’UTR-derived sRNAs, and true intergenic sRNAs. Promoter-associated sRNAs are usually transcriptional interference (TI) by-products, most of which lack any functions (Palmer, Barry Egan, and Shearwin 2011). Here, we identified 16 intergenic sRNAs that overlap the promoters of downstream genes in both sense and antisense orientations (Online Resource 6). We also uncovered 184 5’UTR-derived sRNA candidates, which locate between pTSSs and start codons. Besides, we also identified a total of 230 true intergenic sRNA candidates, of which only 16 candidates and three candidates with promoters and TSSs, respectively.

On the other hand, cis-encoded antisense sRNAs are transcribed opposite to their target genes (Georg and Hess 2011; Cheah et al. 2018). With perfectly complementary base pairing, antisense sRNAs regulate their target mRNAs via various mechanisms, including transcriptional attenuation, modulation of translation, and alteration of mRNA stability (Georg and Hess 2011; Cheah et al. 2018). In this study, we identified 275 cis-antisense sRNA candidates, of which there are 18 candidates with promoters and 26 candidates with TSSs (Online Resource 6). These antisense sRNA candidates were further classified into the following two sub-categories based on their locations with respect to ORF regions or 5’UTRs of their opposite target genes: 5’UTR-antisense sRNAs and ORF-antisense sRNAs. The identification or prediction of UTRs using RNA-seq data enabled the filtering of such sRNAs. In bacteria, 5’UTR-antisense sRNAs usually regulate their target mRNAs by binding to regions associated with translation initiation, i.e., RBSs and start codons, or by altering the stability of their target mRNAs (Georg and Hess 2011; Cheah et al. 2018). For instance, the base-pairing of SymR to the 5’UTR of symE mRNA in E. coli not only masks its RBS from ribosomes but also decreases its stability (Kawano, Aravind, and Storz 2007).

Taken together, the novelty of our study lies in the successful genome-wide annotation of TSSs and sRNAs in L. biflexa serovar Patoc under different growth stages. The RNA-seq analysis uncovered a total of 2,726 TSSs and 603 sRNA candidates, which reflect the transcriptional complexity of L. biflexa serovar Patoc. In short, this study provides insights into the transcriptional annotation that potentially facilitates comparison with other pathogenic Leptospira species. However, these features require further experimental validation and characterization, which we propose to be carried out for future studies.

Funding

This project was supported by Fundamental Research Grant Scheme (FRGS) from the Ministry of Higher Education (MOHE) Malaysia (Grant No.: 203/CIPPT/6711628).

Competing Interest

The authors declare no potential conflict of interest relating to the article’s research, authorship, and publication.

Authors Contributions

Siti Aminah Ahmed and Thean-Hock Tang conceived and planned the study. Siti Aminah Ahmed performed the construction of cDNA libraries for RNA-seq. Hong-Leong Cheah designed the computational framework and carried out the bioinformatic analyses of RNA-seq data and the interpretation of results. Hong-Leong Cheah wrote the manuscript with input and critical feedback from all authors. Thean-Hock Tang supervised the project. Hong-Leong Cheah and Siti Aminah Ahmed contribute equally to the work.

Data Availability

The raw RNA-seq datasets generated in this study are available in the Sequence Read Archive (SRA) database and can be accessed using the BioProject accession: PRJNA884181.

Adhikarla H, Wunder EA, Mechaly AE et al (2018) Lvr, a signaling system that controls global gene regulation and virulence in pathogenic leptospira. Front Cell Infect Microbiol 8:45. https://doi.org/10.3389/fcimb.2018.00045
Amman F, Wolfinger MT, Lorenz R et al (2014) TSSAR: TSS annotation regime for dRNA-seq data. BMC Bioinform 15:89. https://doi.org/10.1186/1471-2105-15-89
Andrews S (2010) FastQC: A Quality Control Tool for High Throughput Sequence Data
Bailey TL, Boden M, Buske FA et al (2009) MEME SUITE: tools for motif discovery and searching. Nucleic Acids Res 37:W202–W208. https://doi.org/10.1093/nar/gkp335
Bulach DM, Zuerner RL, Wilson P et al (2006) Genome reduction in Leptospira borgpetersenii reflects limited transmission potential. Proc Natl Acad Sci 103:14560–14565. https://doi.org/10.1073/pnas.0603979103
Caimano MJ, Sivasankaran SK, Allard A et al (2014) A model system for studying the transcriptomic and physiological changes associated with mammalian host-adaptation by Leptospira interrogans serovar copenhageni. PLoS Pathog 10:e1004004. https://doi.org/10.1371/journal.ppat.1004004
Cheah H-L, Raabe CA, Lee L-P et al (2018) Bacterial regulatory RNAs: complexity, function, and putative drug targeting. Crit Rev Biochem Mol 53:335–355. https://doi.org/10.1080/10409238.2018.1473330
Ewels P, Magnusson M, Lundin S, Käller M (2016) MultiQC: Summarize analysis results for multiple tools and samples in a single report. Bioinform 32:3047–3048. https://doi.org/10.1093/bioinformatics/btw354
Fule L, Halifa R, Fontana C et al (2021) Role of the major determinant of polar flagellation FlhG in the endoflagella-containing spirochete Leptospira. Mol Microbiol 116:1392–1406. https://doi.org/10.1111/mmi.14831
Gaultney RA, Vincent AT, Lorioux C et al (2020) 4-Methylcytosine DNA modification is critical for global epigenetic regulation and virulence in the human pathogen Leptospira interrogans. Nucleic Acids Res 48:12102–12115. https://doi.org/10.1093/nar/gkaa966
Georg J, Hess WR (2011) Regulatory RNAs in cyanobacteria: developmental decisions, stress responses and a plethora of chromosomally encoded cis-antisense RNAs. Biol Chem 392:291–297. https://doi.org/10.1515/BC.2011.046
Grassmann AA, Zavala-Alvarado C, Bettin EB et al (2021) The FUR-like regulators PerRA and PerRB integrate a complex regulatory network that promotes mammalian host-adaptation and virulence of Leptospira interrogans. PLoS Pathog 17:e1009078. https://doi.org/10.1371/journal.ppat.1009078
Iraola G, Spangenberg L, Lopes Bastos B et al (2016) Transcriptome sequencing reveals wide expression reprogramming of basal and unknown genes in Leptospira biflexa biofilms. mSphere 1:e00042–e00016. https://doi.org/10.1128/mSphere.00042-16
Kawano M, Aravind L, Storz G (2007) An antisense RNA controls synthesis of an SOS-induced toxin evolved from an antitoxin. Mol Microbiol 64:738–754. https://doi.org/10.1111/j.1365-2958.2007.05688.x
Langmead B, Salzberg SL (2012) Fast gapped-read alignment with Bowtie 2. Nat Methods 9:357–359. https://doi.org/10.1038/nmeth.1923
Leiva LE, Katz A (2022) Regulation of leaderless mRNA translation in bacteria. Microorganisms 10:723. https://doi.org/10.3390/microorganisms10040723
Livny J, Waldor MK (2007) Identification of small RNAs in diverse bacterial species. Curr Opin Microbiol 10:96–101
Lloréns-Rico V, Cano J, Kamminga T et al (2016) Bacterial antisense RNAs are mainly the product of transcriptional noise. Sci Adv 2. https://doi.org/10.1126/sciadv.1501363
Lo M, Murray GL, Khoo CA et al (2010) Transcriptional response of Leptospira interrogans to iron limitation and characterization of a PerR homolog. Infect Immun 78:4850–4859. https://doi.org/10.1128/IAI.00435-10
Loh E, Dussurget O, Gripenland J et al (2009) A trans-acting riboswitch controls expression of the virulence regulator PrfA in Listeria monocytogenes. Cell 139:770–779. https://doi.org/10.1016/j.cell.2009.08.046
Martin M (2011) Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet j 17:10. https://doi.org/10.14806/ej.17.1.200
Martinez B, Bharati BK, Epshtein V, Nudler E (2022) Pervasive transcription-coupled DNA repair in E. coli. Nat Commun 13:1702. https://doi.org/10.1038/s41467-022-28871-y
Matsunaga J, Lo M, Bulach DM et al (2007) Response of Leptospira interrogans to physiologic osmolarity: relevance in signaling the environment-to-host transition. Infect Immun 75:2864–2874. https://doi.org/10.1128/IAI.01619-06
McClure R, Balasubramanian D, Sun Y et al (2013) Computational analysis of bacterial RNA-Seq data. Nucleic Acids Res 41:e140–e140. https://doi.org/10.1093/nar/gkt444
Miyakoshi M, Chao Y, Vogel J (2015) Regulatory small RNAs from the 3’ regions of bacterial mRNAs. Curr Opin Microbiol 24:132–139. https://doi.org/10.1016/j.mib.2015.01.013
Moll I, Engelberg-Kulka H (2012) Selective translation during stress in Escherichia coli. Trends Biochem Sci 37:493–498. https://doi.org/10.1016/j.tibs.2012.07.007
Nawrocki EP, Burge SW, Bateman A et al (2015) Rfam 12.0: Updates to the RNA families database. Nucleic Acids Res 43:D130–D137. https://doi.org/10.1093/nar/gku1063
Nawrocki EP, Eddy SR (2013) Infernal 1.1: 100-fold faster RNA homology searches. Bioinform 29:2933–2935. https://doi.org/10.1093/bioinformatics/btt509
Palmer AC, Barry Egan J, Shearwin KE (2011) Transcriptional interference by RNA polymerase pausing and dislodgement of transcription factors. Transcription 2:9–14. https://doi.org/10.4161/trns.2.1.13511
Pětrošová H, Picardeau M (2014) Screening of a Leptospira biflexa mutant library to identify genes involved in ethidium bromide tolerance. Appl Environ Microbiol 80:6091–6103. https://doi.org/10.1128/AEM.01619-14
Picardeau M, Bulach DM, Bouchier C et al (2008) Genome sequence of the saprophyte Leptospira biflexa provides insights into the evolution of Leptospira and the pathogenesis of leptospirosis. PLoS ONE 3:e1607. https://doi.org/10.1371/journal.pone.0001607
Putz EJ, Sivasankaran SK, Fernandes LGV et al (2021) Distinct transcriptional profiles of Leptospira borgpetersenii serovar Hardjo strains JB197 and HB203 cultured at different temperatures. PLoS Negl Trop Dis 15:e0009320. https://doi.org/10.1371/journal.pntd.0009320
Qin J-H, Sheng Y-Y, Zhang Z-M et al (2006) Genome-wide transcriptional analysis of temperature shift in L. interrogans serovar lai strain 56601. BMC Microbiol 6:51. https://doi.org/10.1186/1471-2180-6-51
Raghavan R, Sloan DB, Ochman H (2012) Antisense transcription is pervasive but rarely conserved in enteric bacteria. mBio 3:e00156–e00112. https://doi.org/10.1128/mBio.00156-12
Rao MS, Van Vleet TR, Ciurlionis R et al (2019) Comparison of RNA-seq and microarray gene expression platforms for the toxicogenomic evaluation of liver from short-term rat toxicity studies. Front Genet 9:636. https://doi.org/10.3389/fgene.2018.00636
Sharma CM, Vogel J (2014) Differential RNA-seq: the approach behind and the biological insight gained. Curr Opin Microbiol 19:97–105. https://doi.org/10.1016/j.mib.2014.06.010
Sharma CM, Hoffmann S, Darfeuille F et al (2010) The primary transcriptome of the major human pathogen Helicobacter pylori. Nature 464:250–255. https://doi.org/10.1038/nature08756
Shell SS, Wang J, Lapierre P et al (2015) Leaderless transcripts and small proteins are common features of the mycobacterial translational landscape. PLoS Genet 11:e1005641. https://doi.org/10.1371/journal.pgen.1005641
Telatin A, Fariselli P, Birolo G (2021) SeqFu: A suite of utilities for the robust and reproducible manipulation of sequence files. Bioeng 8:59. https://doi.org/10.3390/bioengineering8050059
Thomason MK, Bischler T, Eisenbart SK et al (2015) Global transcriptional start site mapping using differential RNA sequencing reveals novel antisense RNAs in Escherichia coli. J Bacteriol 197:18–28. https://doi.org/10.1128/JB.02096-14
Weinberg Z, Barrick JE, Yao Z et al (2007) Identification of 22 candidate structured RNAs in bacteria using the CMfinder comparative genomics pipeline. Nucleic Acids Res 35:4809–4819. https://doi.org/10.1093/nar/gkm487
Xue F, Dong H, Wu J et al (2010) Transcriptional responses of Leptospira interrogans to host innate immunity: significant changes in metabolism, oxygen tolerance, and outer membrane. PLoS Negl Trop Dis 4:e857. https://doi.org/10.1371/journal.pntd.0000857
Zavala-Alvarado C, Huete G, Vincent S et al (2021) The oxidative stress response of pathogenic Leptospira is controlled by two peroxide stress regulators which putatively cooperate in controlling virulence. PLoS Pathog 17:e1009087. https://doi.org/10.1371/journal.ppat.1009087
Zavala-Alvarado C, Sismeiro O, Legendre R et al (2020) The transcriptional response of pathogenic Leptospira to peroxide reveals new defenses against infection-related oxidative stress. PLoS Pathog 16:e1008904. https://doi.org/10.1371/journal.ppat.1008904
Zhukova A, Fernandes LG, Hugon P et al (2017) Genome-wide transcriptional start site mapping and sRNA identification in the pathogen Leptospira interrogans. Front Cell Infect Microbiol 7. https://doi.org/10.3389/fcimb.2017.00010

No competing interests reported.

Download PDF

Journal Publication

published 21 Feb, 2023

Read the published version in World Journal of Microbiology and Biotechnology →

Editorial decision: Major revision
16 Dec, 2022
Reviews received at journal
05 Dec, 2022
Reviewers agreed at journal
18 Nov, 2022
Reviewers invited by journal
02 Nov, 2022
Editor assigned by journal
02 Nov, 2022
Submission checks completed at journal
01 Nov, 2022
First submitted to journal
01 Nov, 2022

You are reading this latest preprint version

Transcription Start Site Mapping and Small RNA Profiling of Leptospira biflexa serovar Patoc

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Materials And Methods

Bacterial growth conditions

RNA isolation

Construction of cDNA libraries for the TSS mapping

Construction of cDNA libraries for sRNA-seq

RNA-seq

Bioinformatics Analyses

Results

Read pre-processing and alignments

TSS mapping revealed a total of 2,726 TSSs

Other genomic features - Promoters and UTRs

Novel sRNA candidates

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1