University of Groningen RNA regulation in Lactococcus lactis van der Meulen, Sjoerd Bouwe

(1)

RNA regulation in Lactococcus lactis

van der Meulen, Sjoerd Bouwe

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date: 2018

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

van der Meulen, S. B. (2018). RNA regulation in Lactococcus lactis. Rijksuniversiteit Groningen.

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

(2)

525070-L-bw-Meulen 525070-L-bw-Meulen 525070-L-bw-Meulen 525070-L-bw-Meulen Processed on: 24-10-2018 Processed on: 24-10-2018 Processed on: 24-10-2018

Processed on: 24-10-2018 PDF page: 113PDF page: 113PDF page: 113PDF page: 113

Target identification of the cold stress induced

sRNA CisR in

Lactococcus lactis

S.B. van der Meulen1,2_{and J. Kok}1,2

1_{Department of Molecular Genetics, Groningen Biomolecular Sciences and Biotechnology Institute,}

University of Groningen, Groningen, The Netherlands

2_{Top Institute Food and Nutrition (TIFN), Wageningen, The Netherlands}

(3)

114

ABSTRACT

Small non-coding RNAs (sRNAs) are post-transcriptional regulators of gene expression. They have been found in all bacterial and Archaean genomes and are involved in various cellular processes like metabolism, virulence and stress. Most identified sRNAs have not been functionally characterized yet. We have previously pinpointed 186 sRNAs that derive from intergenic regions in the lactic acid bacterium Lactococcus lactis. Ten of these were located

within the 3’-UTR of a gene. Here we have studied CisR, an 81-nt sRNA that is located in the 3’-UTR of the gene of unknown function, llmg_1129. CisR was previously shown to

be induced after cold stress. We have used a combination of in silico target prediction,

overexpression and MS2-affinity purification coupled to RNA sequencing, to perform target identification. Of most interest, CisR seems to down-regulate most of the tRNAs in L. lactis.

(4)

115

5

INTRODUCTION

Bacterial genomes encode a plethora of protein and RNA regulators that are employed to control gene expression at the transcriptional or post-transcriptional level. RNA-dependent gene expression control either operates in cis or in trans. Riboswitches and RNA

thermometers are examples of cis-acting RNA structures that respond to various ligands

and temperature, respectively, resulting in a change in the RNA structural conformation that switches expression of the downstream gene ON or OFF (1, 2). Small regulatory RNAs (sRNAs) are trans-acting RNAs of 50 to 350 nucleotides in size and are encoded in all bacterial

and Archaeal genomes. Most of these transcripts are non-coding; they act by antisense binding to their target mRNAs (3, 4). Some sRNAs also encode a (small) peptide, which has the potential to work in synergy with the non-coding RNA (5). For example, the sRNA SgrS protects Escherichia coli against high levels of glucose-phosphate by stimulating degradation

of the ptsG mRNA that encodes the major glucose transporter. Besides the ptsG mRNA, also

the genes encoding the mannose and secondary glucose transporter manXYZ are targeted

by SgrS, preventing manX from being translated (6). SgrS also encodes the small peptide

SgrT, which inhibits the transport activity of the major glucose permease PtsG by binding to the N-terminal domain of PtsG (7).

sRNAs can exert their effect on gene expression by blocking access of ribosomes to the Shine Delgarno sequence, causing a decrease in translation (8). On the other hand, the 5’-end of a transcript can base-pair in such a way that the RBS is blocked. Some sRNAs have been reported to bind to these mRNAs such as to unfold this self-inhibitory structure, after which translation can occur (9). Often, the sRNA-mRNA duplex attracts RNases that recognize double stranded RNA, such as RNase III and RNase E/Y, which results in their degradation (10). SgrS is the only example thus far of an sRNA that activates gene expression by interfering with RNase accessibility; it binds to the dicistronic pldB-yigL mRNA and blocks

the endonucleolytic turnover by RNase E. As a consequence, the yigL part of the transcript

is protected from degradation and is the only gene that can be translated (11).

Since next generation sequencing is being applied to study RNA transcripts (RNA-seq), either using total RNA or an enriched RNA pool e.g. obtained through co-purification with

a specific RNA-binding protein, the identification of novel sRNAs in bacteria has increased tremendously over the last decade or so. To characterize their biological functions is currently one of the main challenges in the field. The differential expression of an sRNA during specific (stress) conditions may provide some information on the process in which the sRNA might be involved. Notwithstanding this, target mRNA identification remains an essential approach in uncovering the regulatory role of any sRNA. In silico predictions that

are based on sRNA structural information and on the complementarity between sRNA and possible target mRNA can provide potential targets (12, 13), but they often lead to many false-positive hits. Most research on sRNAs, therefore, focuses first on the effect of sRNA

(5)

116

deletion or overproduction on the total transcriptome. The obtained results are subsequently validated by studying translational fusions with the identified target genes and/or by proteomics research. A common way to identify sRNA target transcripts employs a short period of overexpression (pulse-expression) of the sRNA, followed by total transcriptome analysis by DNA microarrays or RNA-seq (14). This can quickly reveal the potential mRNA targets of the sRNA, albeit that secondary effects may overwhelm the initial response if the timing of sRNA expression was not accurate. This might either be caused by stress evoked by the overexpressed sRNA itself or by one or more of the other affected genes, triggering a second round of induction or repression. Besides, sRNAs that act on translation of their target mRNAs might be overlooked in this transcriptomics strategy. More direct methods for target identification have recently been developed. MS2-Affinity Purification coupled with RNA-Seq (MAPS) uses the RNA binding protein MS2 fused to the maltose binding protein (MS2-MBP) in combination with an sRNA that is tagged to the aptamer (ms2) to which

MS2 binds. Using the untagged sRNA as a control, RNAs that base-pair with the ms2-sRNA

fusion can be fished out with the MS2-MBP chimera and sequenced (15). Global small non-coding RNA target Identification by Ligation and sequencing, or GRIL-seq, can identify the segments that base-pair between the complementary strands of sRNA and mRNA (16). Both techniques have been used successfully to identify new targets for sRNAs that already had been studied before.

Typical sRNAs are located within intergenic regions and have their own promoters. Recently, a new class of sRNAs has been discovered that overlap within the 3’-UTR of mRNAs. These sRNAs derive from processing of a parent mRNA or are transcribed from their own promoters (17). We have previously identified 10 sRNAs that originate from 3’-UTRs, from a total of 186 intergenic sRNAs identified in L. lactis by differential RNA-seq (dRNA-seq)

(18). One of these sRNAs, ArgX, overlaps with the 3’-UTR of argR. While ArgR regulates

arginine metabolism as a transcriptional regulator, ArgX was shown to down-regulate the arginine catabolic operon arc (van der Meulen et al., submitted). Another 3’-UTR-derived

sRNA in L. lactis, LLMGnc_082, was upregulated 7-fold after applying 5 min of cold stress

(19). LLMGnc_082 will be here referred to as Cold-induced small RNA, or CisR. In this study, we have used in silico target predictions and have overexpressed CisR in combination with

(6)

117

5

MATERIALS AND METHODS

Bacterial growth and microtiter plate assays. Table 1 presents the L. lactis strains used in

this study. E. coli was used as a host for construction of pCS166-based plasmids, designed

for genomic integration in L. lactis and for expression of the MS2-MBP fusion protein. E. coli was grown in LB medium with erythromycin (150 μg ml−1_{) and ampicillin (75 μg ml}−1₎

for selection. L. lactis was routinely grown as standing cultures at 30°C in CDMPC (20) or

M17 broth (Difco, Becton Dickinson, Le Pont de Claix, France) containing 0,5% (w/v) glucose (GM17), and on GM17 agar plates. For selection purposes, chloramphenicol (5 μg ml−1₎

and/or erythromycin (5 μg ml−1_{) were added. Microtiter plate assays were performed by}

loading 200 µl of a mixture of CDM medium and cell culture in a 96-wells microtiter plate. Culture optical density (at 600 nm; OD₆₀₀) and GFP fluorescence (excitation wavelength of 485 nm and emission wavelength of 535 nm) were recorded in a Tecan F200 (Tecan Group, Männedorf, Switzerland).

TABLE 1. L. lactis strains and plasmids used in this study

Strain or plasmid Relevant phenotype or genotype* Reference

Strains

MG1363 L. lactis subsp. cremoris, plasmid-free derivative of _NCDO712 (40)

NZ9000 MG1363, pepN::nisRK (41)

SVDM2016 NZ9000, CisR deletion mutant This work SVDM2017 NZ9000, RNaseIII gene deletion mutant (Δrnc) This work SVDM2018 Emr_,_{MG1363, P}

CisR-sfgfp fusion integrated in pSEUDO_10 This work

SVDM2019 Cmr_{, NZ9000 ΔCisR, pSVDM5009} _{This work}

SVDM2020 Cmr_{, NZ9000 ΔCisR}_{, Δrnc, pSVDM5009} _{This work}

SVDM2021 Cmr_{, NZ9000 ΔCisR, pSVDM5010} _{This work}

SVDM2022 Cmr_{, NZ9000 ΔCisR}_{, Δrnc, pSVDM5010} _{This work}

Plasmids

pNZ8048 Cmr_{, high copy number cloning vector} ₍₄₁₎

pCS1966 Emr_,_{oroP, integration vector} ₍₂₁₎

pSEUDO-GFP Em_{pseudo_10 locus of}r, vector for integration of sfgfp fusions in the

L. lactis (22)

pSVDM5003 Cmr_{, pGhost containing P}

nisA-recT

van der Meulen et al., submitted

pSVDM5008 Emr_{, P}

CisR-sfgfp fusion in pSEUDO_10-GFP This work

pSVDM5009 Cmr_{, pNZ8048 carrying P}

nisA-CisR This work

pSVDM5010 Cmr_{, pNZ8048 carrying P}

nisA-MS2-CisR This work * Cmr_{, chloramphenicol; Em}r_{, erythromycin.}

(7)

118

General DNA techniques and L. lactis strain construction. Purification of plasmid DNA

and PCR fragments was done, essentially as described by the manufacturer, with the NucleoSpin® Plasmid kit and NucleoSpin® Gel and PCR Clean-up kit (Machery-Nagel GmbH, Düren, Germany), respectively. For the isolation of plasmid DNA from L. lactis, cell lysis was

improved by a 45-min incubation in buffer A1 supplemented with 10 mg/ml lysozyme. The L. lactis ΔCisR and ΔRNaseIII strains were constructed by amplifying the flanking regions of the

respective genes with Fusion polymerase (Life Technologies, Bleiswijk, the Netherlands). In both instances, the PCR fragment containing the upstream region of the gene was digested with XbaI and PstI, while the fragment encompassing the downstream region of the gene was cut with the restriction enzymes PstI and XhoI. Both fragments were inserted into XbaI and XhoI-digested integration vector pCS1966 (21) using T4 ligase (all enzymes: Life Technologies). The proper constructs were obtained in E. coli, isolated and used to integrate

into the L. lactis chromosome via double cross-over recombination (22). The transcriptional

fusion between the P_CisR promoter and the gene for superfolder GFP (sfgfp; (23)) and the

pNZ8048 derivatives overexpressing CisR with and without ms2 tag were constructed using

restriction-and-ligation-independent cloning (24). The ms2 tag sequence was introduced

in one of the primers for PCR amplification. The vector backbone and the required inserts were separately amplified with polymerase pfuX7 (25), purified and subsequently treated with the USER enzyme mix (New England Biolabs, Ipswich, MA). The P_CisR-sfgfp fusion vector

pSVDM5008 was integrated in the transcriptionally silent pSEUDO_10 locus (22).

For transformation with plasmids and ligation mixtures, electroporation was applied for L. lactis cells using a Bio-Rad Gene Pulser (Bio-Rad Laboratories, Richmond, CA) at 2.5 kV, 25 uF

and 200 Ohm. E. coli cells were prepared chemically for heat-shock (45 sec at 42°C). RNA isolation and quality control. RNA was isolated for RNA sequencing and Northern

analysis as described before (18). In short, frozen cell pellets were thawed on ice, re-suspended in 400 μl TE-buffer (10 mM Tris, 1 mM EDTA; pH 8.0), after which 50 μl of 10% Sodium Dodecyl Sulphate (SDS), 500 μl phenol/chloroform and 0.5 g glass beads (75-150 μm, Fischer Scientific) were added. The cells were disrupted in a Biospec Mini-BeadBeater (Biospec Products, Bartlesville, OK) through 2 cycles of 45 sec, with a 1-min interval on ice. Chloroform extraction was used to recover nucleic acids, that were subsequently treated with DNase I and RiboLock RNase inhibitor (Fermentas/Thermo Scientific, Vilnius, Lithuania) in DNase I buffer for 30 min at 37°C. Total RNA was then retrieved using standard phenol/ chloroform extraction and sodium acetate/ethanol precipitation. Resulting RNA pellets were dissolved in elution buffer from the High Pure RNA Isolation Kit (Roche Diagnostics, Almere, the Netherlands) and stored at -80°C. RNA concentration was measured using a Nanodrop ND-1000 (Thermo Fischer Scientific), the RNA quality was assessed by checking the integrity of the 16S/23S rRNA and the presence of any DNA contamination on 1% agarose/1% bleach gels (26).

(8)

119

5

Northern hybridization. Total RNA (10 μg) was separated on 8% polyacrylamide gels in TAE

buffer (40mM Tris, 20mM acetic acid, and 1mM EDTA). As a denaturing agent, 1% bleach was used instead of the 7M urea normally employed (van der Meulen et al, submitted). The

separated RNAs were then transferred to a positively charged Zeta-Probe nylon membrane (Bio-Rad Laboratories BV), using semi-dry electro blotting. RNAs were covalently cross-linked to the nylon membranes at 1200 mJ in a UVC-508 Ultraviolet Crosslinker (Ultra-Lum Inc., Carson, CA). ssDNA oligonucleotides were labeled with 32_{P-γATP (PerkinElmer BV,}

Groningen, the Netherlands) using polynucleotide kinase (Fermentas/Thermo Scientific), according to instructions of the manufacturer. The membranes were incubated for hybridization overnight at 42°C in PerfectHyb Plus Hybridization buffer (Sigma-Aldrich Chemie Gmbh, Munich, Germany) with 9 µl of the labeled probe. Membranes were washed twice in 2x saline sodium citrate (SSC) buffer with 0.1% SDS, after which a Phosphor Screen was exposed to the labeled membrane overnight. A Cyclone Plus Phosphor Imager and OptiQuant software (PerkinElmer, Groningen, the Netherlands) was used for imaging.

RNA deep sequencing and data analysis. Total RNA samples from the CisR overexpression

study and enriched RNA samples from the MAPS experiment were sequenced at the Primbio Research Institute (Exton, PA). Ribo-Zero™ rRNA removal and library preparation using the AmpliSeq™ kit (ThermoFischer Scientific) was performed prior to sequencing of the cDNA libraries on an Ion Proton sequencer (ThermoFischer Scientific). Raw sequence reads of the CisR overexpression experiment were analyzed for quality, trimmed with a PHRED score >28 and aligned to the genome L. lactis NZ9000 using Bowtie 2 (27). RKPM values were

used as an input for the T-REx analysis pipeline (28) together with a text file describing the factors, contrasts and classes specifying genes that were induced (blue) and repressed (red) after cold shock. Sequence reads from the MAPS experiment were treated as above. T-REx was used to analyze the data. The RNA-seq data has been uploaded under GEO accession number GSE115974.

MS2-MBP fusion protein purification. The MS2-Maltose binding protein fusion (MS2-MBP)

was expressed in a 1-L E. coli culture at an OD600 of 0.5 by induction with 1mM IPTG. Cells

were incubated for another 3 hours at 37°C and harvested by centrifugation at 6000 rpm for 10 min. They were then lysed by sonicating 10 times 10 sec with 30 sec intervals, all on ice, using a Vibra Cell sonicator (Beun de Ronde, Abcoude, the Netherlands) at 50% intensity and re-suspended in buffer A (20 mM HEPES pH 7.9, 200 mM KCl, 1 mM EDTA). The MS2-MBP protein was loaded on a 5-ml amylose column (Sigma-Aldrich Chemie Gmbh), which was first washed with 40 ml buffer A and then with 10 ml buffer B (buffer A with 20 mM KCl) to lower the salt concentration. The protein was eluted with 20 ml buffer C (buffer B with 10 mM maltose). The protein peak fractions were pooled and concentrated to 1 ml using an Amicon-30 centrifugal filter (Millipore Merck, Darmstadt, Germany). The fusion

(9)

120

protein was further purified by NGC (Bio-Rad Laboratories BV) using a 1-ml HiTrap Heparin column (GE Healthcare Bio-Sciences AB, Uppsala, Sweden). The column was equilibrated with 10 ml buffer 1 (20 mM HEPES, pH 7.9, 1 mM EDTA, 20 mM KCl) before the protein concentrate was loaded, and then washed with 5 ml buffer 1. The MS2-MBP fusion protein was eluted by applying a gradient from 20 to 400 mM KCl over 10 ml (10 column volumes). The peak fractions were pooled and concentrated to ~500 µl using an Amicon-30 as before and glycerol was added to end concentration of 10%. The MS2-MBP protein samples were divided into 65 µl aliquots and stored at -80°C.

MS2-affinity purification coupled with RNA sequencing (MAPS)

The MS2-affinity purification assay was inspired by the experimental set-up described elsewhere (29). L. lactis strains were grown in duplo to an OD600 of 0.4, after which they were

induced for 10 min with 5 ng/ml nisin (Sigma-Aldrich Chemie Gmbh) to express ms2-CisR

and CisR. Per strain, the cells from 50 ml of culture were harvested by centrifugation for 1 min at 10.000 rpm, washed with 1 ml of buffer A (20 mM Tris-HCl at pH 8.0, 150 mM KCl, 1 mM MgCl2 and 1 mM DTT) and re-suspended in 2 ml buffer A. Cells were lysed on ice using a Vibra Cell sonicator at 80% intensity (5 cycles of 45 sec ON, 15 sec OFF). The lysate was cleared by centrifugation (14.000 rpm, 30 min, 4°C) and used for affinity chromatography at 4°C. The column was prepared by adding 75 μL of amylose resin (New England Biolabs) to a Bio-Spin disposable chromatography column (Bio-Rad Laboratories BV). The column was first washed with 3 ml of buffer A after which 100 pmol of the purified MS2-MBP fusion protein was loaded and immobilized onto the column. Subsequently, the amylose resin was washed with 2 ml of buffer A. The cleared lysate was loaded onto the MS2-MBP affinity column, which was then washed with 5 ml of buffer A. RNA was eluted from the column with 1 ml of buffer A that contained 15 mM maltose. Eluted RNA was isolated, purified and concentrated by phenol/chloroform extraction and sodium acetate/ethanol precipitation as described above. RNA pellets were dissolved in elution buffer from the High Pure RNA Isolation Kit (Roche Diagnostics) and the samples were stored at -80°C. RNA concentration was measured on a Quantus device using the Quantifluor RNA system according to the manufacturer (Promega Benelux, Leiden, the Netherlands).

RESULTS

CisR discovery, structure and expression

Previously, we have identified the putative small regulatory RNA LLMGnc_082 (CisR) by differential RNA-seq (18). Mfold (30) predicted the molecule to contain two hairpin structures at the 5’-end and a terminator structure (Figure 1A). CisR overlaps with the

3’-UTR of llmg_1129, which is part of the operon llmg_1127-29. The genes of this operon

(10)

121

5

protein. Downstream of CisR, an uncharacterized sRNA LLMGnc_083 is located (Figure 1B).

Both sRNAs have their own predicted promoter sequences (18).

We previously studied the transcriptome of L. lactis cells subjected to six industrially

relevant stress conditions and observed that CisR was upregulated 7-fold after a 5-min exposure to cold stress at 10°C (19), and 6.3-fold downregulated after 5 min of starvation in phosphate-buffered saline (PBS). In contrast, the genes from the operon llmg_1127-29

were up-regulated 3.4±1.1 after starvation, but not affected by any of the other stressors. To verify the existence and the size of the CisR RNA molecule, Northern analyses were performed on total RNA isolated from L. lactis cells at various time points during growth

and after 0 to 60 min of exposure to a temperature of 10°C (Figure 1C). The size of CisR (~81

nt) agrees with what was expected on the basis of the original dRNA-seq data (18) and the end of the predicted terminator structure (Figure 1A). The expression of CisR seems rather

stable during exponential growth but peaks at an OD₆₀₀ of 2.0. Two hours after the culture had reached an OD₆₀₀ of 2.0, and also after overnight growth, expression is undetectable. A culture with an OD₆₀₀ of 1.0 was centrifuged and the cells were re-suspended and aliquoted over six tubes with fresh pre-cooled (at 10°C) GM17 medium. The cultures were incubated further at 10°C and one each was harvested at the time points 0, 5, 10, 20 , 30 or 60 min. Based on the Northern analysis, it seems that the expression of CisR is elevated after 5 to 10 min of cold stress and that the expression remains high throughout the remaining 60 min at 10°C.

(11)

122

Figure 1. CisR structure, genomic location and expression. (A) Secondary structure of the CisR sRNA

as predicted by Mfold. It has a dG of -20.80 (30). (B) Schematic overview of the genomic location

of the CisR gene. Grey arrows indicate genes; white arrows represent sRNAs; black bent arrows are promoters; predicted terminators are shown as lollipops. (C) CisR (LLMGnc_082) expression under

various growth conditions (indicated in the top margin) was verified by Northern analysis, using an 8% acrylamide/1% bleach gel and a CisR-specific radioactively labeled probe. The 5S rRNA was used as a control. Nucleotide size marker is shown in the left margin.

(12)

123

5

The promoter of CisR is highly conserved and active at the end of the exponential growth phase

To search for potential homologs of CisR in other bacterial species, we performed an online GLASSgo search (http://rna.informatik.uni-freiburg.de). CisR homologs were present in all sequenced L. lactis strains. ClustalW (GenomeNet Database, Japan) was used to align CisR

and its upstream region from four L. lactis subsp. lactis and four subsp. cremoris strains

(Figure 2A). As expected, CisR sequences are most different between the two subspecies

of L. lactis. Interestingly, the region between the second putative hairpin structure and

the terminator sequence is invariant in all strains. This suggests that these nucleotides are potentially important for base pairing with the target mRNA(s), and could represent the so-called seed region or seed sequence (31). The most distant CisR homolog was found in the human isolate L. lactis subsp. lactis CV56; it contains 12 mutations compared to L. lactis

MG1363 CisR. These mutations do not lead to changes in the predicted structure but the first hairpin in L. lactis CV56 CisR is shorter and also its first nucleotide is predicted to not

base-pair. The latter could have an impact on CisR stability, as in Bacillus subtilis it was shown

that the RNA pyrophosphohydrolase RppH requires at least two unpaired nucleotides at the 5’-end of an RNA to effectively convert a 5’-terminal triphosphate into a monophosphate. This conversion subsequently triggers 5’-exonucleolytic degradation of the RNA by RNase J (32). The total number of mutations in the CisR homologs is relatively high compared to another sRNA characterized in L. lactis, ArgX. In the ArgX variants of ten L. lactis strains

only one mutation was observed (van der Meulen et al, submitted). The almost perfect -10 box (TAGAAT) of the promoter of CisR was conserved in all strains analyzed. A previously identified putative CodY binding box in the CisR promoter region, 38 bases away from the CisR transcription start site (18) is present in all eight L. lactis sequences. A 2.9-fold increase

of CisR was apparent in an L. lactis MG1363 codY deletion mutant using RNA-sequencing

(van der Meulen et al., submitted).

To study the promoter of CisR in more detail, a transcriptional fusion between P_CisR and a superfolder GFP gene (sfgfp) (23) was constructed and integrated in the transcriptionally

silent pseudo_10 locus of L. lactis (22). The activity of PCisR-sfgfp was examined under various

stress conditions. P_CisR was active at all conditions, but activity specifically peaks towards the end of the exponential growth phase (Figure 2B). Furthermore, more GFP-fluorescence per

cell was observed in the presence of 2% NaCl, while the growth was strongly inhibited. With 1% NaCl the strain growth was normal and P_CisR-sfgfp expression was similar to that of the

(13)

124

Figure 2. CisR nucleotide and promoter analysis. (A) Nucleotide sequence of CisR (in the black box)

and its upstream region including the promoter, aligned to corresponding sequences in the genomes of eight L. lactis strains. The black bent arrow indicates the transcription start of CisR, Conserved

nucleotides are shown in red, alternative nucleotides in blue or black. (B) Analysis of PCisR-sfgfp activity

in L. lactis SVDM2018 cells treated for 60 min with the following stressors: cold (10°C), heat (42°C)

and starvation (incubation in PBS). The cells were then centrifuged and re-suspended in fresh CDMPC medium. For salt stress, 2% NaCl was added to the medium during growth. Each data point represents the average of eight measurements in a Tecan plate reader.

Overexpression of CisR downregulates tRNAs

To gain insight into the potential mRNA target(s) of CisR, pulse expression of CisR was performed by inducing its gene, present on a high-copy number plasmid and downstream of the nisin-inducible promoter P_nisA, for 10 min with nisin. Table 2 presents an overview

of the results of CisR overexpression. The results of CisR overexpression on the L. lactis

(14)

125

5

expressed upon cold stress (Figure 3A). Vice versa, the differentially expressed genes after

CisR overexpression were plotted in the transcriptome data after 10 min of cold stress (Figure 3B). CisR itself was induced 60 times. Most notable was the downregulation of 52

tRNAs from the 63 that are described in L. lactis MG1363 (33). A decrease in tRNAs was

also observed after a 5-min exposure to various stress conditions (19). The fruACR operon

for fructose utilization was downregulated -8.7±1.6 fold after CisR overexpression; this operon was also observed to be downregulated upon cold, acid, osmotic and starvation stress. In particular, rRNA genes for the 16S small ribosomal subunit were upregulated after CisR overexpression while, in contrast, they were generally downregulated in our previous study investigating the effect of various stress conditions on L. lactis gene expression (19).

Besides the downregulation of the regulator gene fruR, the glutamine synthetase

repressor-encoding glnR gene was also negatively affected (-2-fold). Two other regulators, of unknown

function, were upregulated: llmg_0709 (2.6-fold) and llmg_2339 (2.3-fold).

Figure 3. Transcriptome analysis by RNA-seq of CisR overexpression or a 5-min exposure to cold stress. (A) Volcano plot of the effect on the transcriptome of 10 min of CisR overexpression in L. lactis

relative to that in an empty expression vector control strain. Blue and red dots represent genes that are activated or repressed, respectively, upon cold stress. (B) Effect of 5-min cold stress (10°C) from

our previous study in which CisR was upregulated 7-fold (19). Blue and red dots represent genes that are activated and repressed, respectively, after 10 min overexpression of CisR. Genes outside the grey areas have a p-value of ≥ 0.05 and a fold change of ≥ 2. Genes outside dashed lines: p-value of ≥ 0.01

(15)

126

Table 2. Effect of overexpression of CisR on L. lactis transcriptome. The CisR sRNA was induced for 10

min. Genes that were enriched ≥2 fold and with p-value ≤0.05 and a logCPM value >1 are presented.

The CisR gene and 26 other genes were activated while 75 genes were repressed.

Gene locus Gene alias (Putative) function Fold change p-value

LLMGnc_082 CisR sRNA 60.1 2.80E-71

LLMGnc_184 LLMGnc_184 sRNA 10.8 2.10E-05

as-llmg_1354 as-llmg_1354 antisense RNA 7.1 2.40E-02 llmg_rRNA_4a llmg_rRNA_4a 16S ribosomal RNA 6 1.60E-02 llmg_rRNA_7a llmg_rRNA_7a 16S ribosomal RNA 5.9 1.30E-02 llmg_rRNA_48a llmg_rRNA_48a 16S ribosomal RNA 5.7 2.30E-02 llmg_rRNA_6b llmg_rRNA_6b 16S ribosomal RNA 5.5 2.00E-02 llmg_rRNA_5a llmg_rRNA_5a 16S ribosomal RNA 5.4 2.20E-02

llmg_0743 amyY alpha-amylase 5.3 4.60E-02

llmg_rRNA_1 llmg_rRNA_1 16S ribosomal RNA 4.6 3.40E-02

llmg_0960 llmg_0960 beta-glucosidase 4.4 4.40E-02

llmg_0737 malG maltose ABC transporter permease _malG 4 1.20E-02 llmg_0436 llmg_0436 hypothetical protein 3.7 6.20E-04 llmg_1091 llmg_1091 hypothetical protein 3.6 4.20E-02

llmg_0319 pepN peptidase N 3.3 2.40E-08

llmg_1886 intTn5276 IntTn5276 protein 2.8 4.10E-02

llmg_2338 llmg_2338 hypothetical protein 2.7 1.20E-02 llmg_0709 llmg_0709 PadR-like family transcriptional _regulator 2.6 2.30E-02

llmg_0449 llmg_0449 hypothetical protein 2.4 3.60E-02 llmg_2339 llmg_2339 transcriptional regulator 2.3 2.70E-02 llmg_2459 llmg_2459 hypothetical protein 2.2 1.50E-02 llmg_1069 llmg_1069 DNA-binding protein 2.2 4.50E-02 llmg_2144 llmg_2144 hypothetical protein 2.1 1.60E-02 llmg_0560 hemK protoporphyrinogen oxidase 2.1 1.60E-02 llmg_1629 llmg_1629 methyltransferase 2.1 2.20E-02

llmg_tRNA_45 llmg_tRNA_45 tRNALys _-2 _2.90E-03

llmg_0281 llmg_0281 anaerobic ribonucleoside _{triphosphate reductase} -2 5.00E-03

llmg_2485 glnR glutamine synthetase regulator _protein -2 1.60E-02

(16)

127

5

llmg_2473 rpsR 30S ribosomal protein S18 -2.1 2.40E-02

llmg_tRNA_11 llmg_tRNA_11 tRNAThr _-2.1 _9.80E-04

llmg_0282 nrdG anaerobic ribonucleoside-triphosphate reductase activating protein

-2.1 7.20E-03

llmg_tRNA_54 llmg_tRNA_54 tRNAGln _-2.1 _1.90E-03

llmg_2269 cutC copper homeostasis protein CutC -2.1 8.50E-04

llmg_tRNA_57 llmg_tRNA_57 tRNATyr _-2.2 _2.00E-02

llmg_tRNA_17 llmg_tRNA_17 tRNAArg _-2.2 _1.60E-03

llmg_tRNA_56 llmg_tRNA_56 tRNATrp _-2.3 _4.80E-04

llmg_tRNA_20 llmg_tRNA_20 tRNAAsp _-2.3 _1.20E-03

llmg_tRNA_46 llmg_tRNA_46 tRNAAsp _-2.3 _1.20E-03

llmg_tRNA_61 llmg_tRNA_61 tRNAGlu _-2.3 _1.60E-03

llmg_1986 llmg_1986 hypothetical protein -2.3 1.60E-02 llmg_rRNA_48b llmg_rRNA_48b 5S ribosomal RNA -2.3 1.60E-04

llmg_tRNA_04 llmg_tRNA_04 tRNAPhe _-2.4 _4.20E-04

llmg_rRNA_6a llmg_rRNA_6a 5S ribosomal RNA -2.4 1.20E-04

as-ps118 as-ps118 antisense RNA -2.4 5.50E-03

llmg_tRNA_29 llmg_tRNA_29 tRNAPhe _-2.4 _3.20E-04

llmg_0711 tnpR DNA-invertase/resolvase -2.4 3.70E-02

llmg_tRNA_37 llmg_tRNA_37 tRNASer _-2.4 _1.50E-04

llmg_0443 llmg_0443 hypothetical protein -2.5 2.60E-05

llmg_tRNA_16 llmg_tRNA_16 tRNACys _-2.5 _2.20E-05

llmg_rRNA_2a llmg_rRNA_2a 5S ribosomal RNA -2.6 1.60E-05

llmg_tRNA_34 llmg_tRNA_34 tRNALys _-2.6 _2.60E-05

llmg_0532 def peptide deformylase -2.6 6.50E-05

llmg_2426a llmg_2426a 5S ribosomal RNA -2.7 2.50E-06

LLMGnc_015 LLMGnc_015 sRNA -2.8 5.70E-03

llmg_2466a llmg_2466a 5S ribosomal RNA -2.8 1.50E-06

llmg_tRNA_31 llmg_tRNA_31 tRNAIle _-2.8 _2.30E-05

llmg_tRNA_24 llmg_tRNA_24 tRNAGly _-2.9 _4.80E-07

llmg_tRNA_06 llmg_tRNA_06 tRNAIle _-3 _4.00E-06

llmg_tRNA_42 llmg_tRNA_42 tRNAGly _-3 _3.70E-07

llmg_2495 llmg_2495 hypothetical protein -3 8.60E-05

llmg_tRNA_30 llmg_tRNA_30 tRNAGly _-3 _1.30E-06

(17)

128

llmg_tRNA_06a llmg_tRNA_06a 5S ribosomal RNA -3.1 7.90E-07

llmg_tRNA_52 llmg_tRNA_52 tRNALeu _-3.1 _1.50E-04

llmg_tRNA_05 llmg_tRNA_05 tRNAGly _-3.2 _2.50E-07

llmg_tRNA_03 llmg_tRNA_03 tRNAMet _-3.3 _8.30E-07

llmg_tRNA_01 llmg_tRNA_01 tRNAGlu _-3.7 _1.00E-07

llmg_tRNA_59 llmg_tRNA_59 tRNAAla _-3.7 _1.70E-08

as-llmg_1269 as-llmg_1269 antisense RNA -3.9 2.40E-04

llmg_tRNA_18 llmg_tRNA_18 tRNAAla _-4 _2.00E-09

llmg_tRNA_47 llmg_tRNA_47 tRNAVal _-5 _2.10E-13

llmg_tRNA_58 llmg_tRNA_58 tRNAVal _-5.1 _1.90E-13

llmg_tRNA_19 llmg_tRNA_19 tRNAVal _-5.1 _6.70E-14

llmg_1568 fruA PTS 2C fructose-specific IIBC _components -6.5 7.30E-04 llmg_1569 fruC tagatose-6-phosphate kinase -9.2 8.80E-03

(18)

129

5

MAPS shows enrichment of tRNAs, busR and fruR

Besides sRNA overexpression to identify the potential targets of CisR, MAPS technology (MS2-affinity purification coupled with RNA sequencing) was employed for the same reason and to confirm transcript identities obtained by either approach. In particular, MAPS should provide more direct evidence for targets as sRNA overexpression can result in secondary transcriptomic effects. MAPS relies on the affinity of the RNA-binding MS2 protein for the

ms2 RNA aptamer. The ms2 aptamer was fused to the 5’-end of CisR and placed in a vector

in which the ms2-CisR gene fusion could be induced with nisin. CisR without ms2 aptamer

was used as a control. In previous MAPS studies in E. coli, an RNase E mutant strain was used

that is defective in assembling the degradosome (34). This results in less cleavage of sRNA-mRNA duplexes, enabling enrichment of potential sRNA-mRNA targets, especially those that are negatively regulated by the sRNA. L. lactis RNase Y acts as a functional homolog of RNase

E, but no studies have been performed to create degradosome defective L. lactis mutants.

An attempt to delete RNase Y failed, most probably because the gene is essential in L. lactis

(data not shown). Instead, we deleted the gene encoding RNase III, an enzyme that has been shown to play an important role in the digestion of double-strand RNAs in Gram-positive bacteria (35). Since it was unclear whether or not this approach would provide proper MAPS results, the experiments were performed twice in the L. lactis Δrnc strain,

lacking RNase III (Llmg_1753) and also two times in the strain with an intact rnc gene. All

strains used for MAPS additionally contained a deletion of the entire chromosomal CisR gene. Table 3 gives an overview of enriched RNAs after MAPS. MAPS reveals the enrichment

of three tRNAs, namely tRNAArg_{(11.2 fold), tRNA}Pro_{(6.3 fold) and tRNA}Ser_{(3.4 fold). The latter}

was also negatively regulated after CisR overexpression, while the other two were not. Two mRNAs that encode transcriptional regulators, busR and fruR, were enriched 3.8- and

2.4-fold, respectively. The fruR mRNA was strongly downregulated after CisR overexpression,

and also after cold, acid, osmotic and starvation stress. The BusR regulator represses glycine betaine transport by BusAB (36), of which the genes were activated upon osmotic stress (19).

(19)

130

Table 3. MAPS results. RNAs of genes that were enriched ≥2 fold with a p-value ≤0.05 and a logCPM

value >1 are indicated.

Gene locus Gene alias Gene function Ratio ms2-CisR/CisR p-value

llmg_tRNA_33 llmg_tRNA_33 tRNAArg _11.2 _4.60E-03

llmg_0852 ps357 phage protein, Abi-like 7 1.70E-10

llmg_tRNA_15 llmg_tRNA_15 tRNAPro _6.3 _1.80E-02

llmg_1997 pflA pyruvate-formate lyase _{activating enzyme} 4.5 4.10E-07 llmg_1047 busR transcriptional repressor BusR 3.8 1.80E-06

llmg_tRNA_37 llmg_tRNA_37 tRNASer _3.4 _3.10E-02

llmg_2298 llmg_2298 hypothetical protein 2.9 3.50E-02 llmg_0853 ps358 phage protein with unknown _function 2.8 3.10E-02 llmg_2015 llmg_2015 hypothetical protein 2.5 1.80E-02 llmg_1570 fruR fructose transport regulator _FruR 2.4 3.10E-02

DISCUSSION

Overexpression of CisR and the MAPS approach followed by RNA-seq were used to identify potential targets of CisR, an sRNA that is induced 7-fold after a 5-min exposure to cold stress. CisR is located in the 3’-UTR of llmg_1129, the last gene of a functionally uncharacterized

operon that specifies three hypothetical cell wall surface proteins. The only other 3’-UTR-derived sRNA from L. lactis that has been studied so far, ArgX, overlaps with the 3’-UTR of

the gene encoding the transcriptional regulator ArgR (van der Meulen et al., submitted). In

this specific case, both the sRNA ArgX and the transcription factor ArgR are involved in the regulation of the arc operon for the catabolism of arginine. In Salmonella typhimurium, the

sRNA CpxQ is cleaved from the 3’-UTR of the transcript encoding CpxP, a periplasmic protein that combats cell envelope stress. CpxQ sRNA complements the function of the protein CpxP by repressing the mRNAs of several envelope proteins (17). These two examples show that 3’-UTR-derived sRNAs can have a functional relationship with the gene immediately preceding it. One might expect this to be the case especially when the sRNA and the protein coding gene share a common promoter. It would be interesting to investigate whether a functional link exists between the llmg_1127-1129 operon and CisR.

A transcriptional fusion between the CisR promoter region and the superfolder GFP gene was used to measure promoter activity during or after various stress conditions for 12 hours of growth. Surprisingly, the promoter seemed to be active under all conditions tested, including the one without stress. This result would contradict the results of the Northern hybridization experiments, which showed an increase of CisR RNA upon the application of a

(20)

131

5

cold stress. We have previously also observed an increase of CisR RNA in the transcriptome of L. lactis cells exposed briefly (5 min) to a low temperature of 10°C. CisR was not induced

after any of the other stress conditions employed in that study (19). The activity of P_CisR could be explained by the use of a CDM medium in which the plate reader assay was conducted, that potentially triggers the promoter of CisR. Alternatively, the expression of CisR may not fully depend on the activity of its promoter, but may be subjected to post-transcriptional control e.g. by other regulatory RNAs or by protein chaperones.

A 10-min overexpression of CisR in combination with full-genome transcription analysis by RNA-seq revealed a majority of the tRNA genes of L. lactis being downregulated. Since

downregulation of tRNA gene expression was also observed under various stress conditions (19), and the fact that CisR was induced after 5 min of cold stress in the same study, it is tempting to speculate that CisR is responsible for the decrease in tRNA expression. tRNAs are regulated at several levels. They undergo extensive processing at both their 5’- and 3’-ends (37) and various nucleotides at different positions in the molecules are chemically modified (33). The status of amino acid charging of the tRNAs plays an important role in the regulation of the concomitant aminoacyl-tRNA synthetase genes through their interaction with T-box elements in the 5’-UTR regions (38). To our knowledge, there have so far been no reports of an sRNA that is able to regulate one or more tRNAs by base pairing interactions. The possibility that the sRNA CisR is involved in direct or indirect regulation of tRNAs in L. lactis is of high interest and deserves future research focus.

Ten potential targets of CisR were identified by MAPS. In comparison with a previous MAPS study performed by Lalaouna et al (29), we observed relatively low enrichment of RNAs

by co-purification with ms2-CisR bait. The highest ratio of 11.2 was observed for a tRNA

(tRNAArg_{). Lalaouna}_{et al (29) used a degradosome-defective RNase E mutant of E. coli. Since}

a functionally comparable RNase Y mutation could not be obtained in L. lactis, it is highly

likely that duplexes of CisR and its mRNA target(s) were degraded, which would explain the relatively low ratios of enrichment in our MAPS experiment. Lalaouna et al (27) made

a similar observation when the wildtype strain was used for MAPS: this severely reduced enrichment of targets under negative regulation of the sRNA (15). The possibility that the MS2 tag interfered with CisR function, should be examined in the future.

The sRNA target prediction program TargetRNA2 was used to generate a list of potential CisR targets. Besides, we used the genes that were differentially expressed upon CisR overexpression or those that were enriched in the MAPS experiment. No potential regions were identified in CisR that could base-pair with the genes from the overexpression and/or MAPS experiment.

The promoter region of CisR contains a putative CodY binding box (18) and the gene was previously shown to be upregulated in L. lactis ΔcodY (van der Meulen et al., submitted). L. lactis CodY is a global transcriptional regulator and acts as a repressor during nutrient

(21)

132

acids (39). CisR overexpression resulted in the down-regulation of 52 of 63 tRNAs, while three tRNAs (tRNAPro_{, tRNA}Arg_{and tRNA}Ser_{) co-purified with}_{ms2-tagged CisR. How could}

the protein CodY and the sRNA CisR be linked to each other? We hypothesize that CisR is activated upon stress, either with or without simultaneous nutrient deprivation, and then acts to decrease protein synthesis by (in)directly lowering the tRNA pool. Two other targets of CisR with a link to stress, namely busR and fruR, could play additional roles during

different stress responses (Figure 4).

Figure 4. Working model of CisR regulation. CisR is repressed by CodY and activated upon (cold)

stress. tRNAs are down-regulated by CisR. Two potential targets of CisR that have been previously linked to stress are included in this model. The fruR gene is downregulated upon CisR overexpression

and its transcript was enriched in MAPS. The busR transcript was enriched in MAPS, but how CisR

affects busR is unknown, as indicated by the dotted line.

ACKNOWLEDGEMENTS

We thank the Jörg Vogel group for supplying the E. coli strain with the plasmid containing

MS2-MBP for the MAPS experiments. We thank dr. Andrius Buivydas for help in purifying MS2-MBP. Furthermore, we thank the TIFN functional fermentation consortium for helpful discussions.

(22)

133

5

REFERENCES

1. Narberhaus F, Waldminghaus T, Chowdhury S. 2006. RNA thermometers. FEMS Microbiol Rev 30:3-16.

2. Mandal M, Breaker RR. 2004. Gene regulation by riboswitches. Nature Reviews Molecular Cell

Biology 5:451-463.

3. Waters LS, Storz G. 2009. Regulatory RNAs in bacteria. Cell 136:615-628.

4. Storz G, Vogel J, Wassarman KM. 2011. Regulation by small RNAs in bacteria: expanding frontiers.

Mol Cell 43:880-891.

5. Gimpel M, Brantl S. 2017. Dual-function small regulatory RNAs in bacteria. Mol Microbiol 103:387-397.

6. Rice JB, Vanderpool CK. 2011. The small RNA SgrS controls sugar–phosphate accumulation by

regulating multiple PTS genes. Nucleic Acids Res 39:3806-3819.

7. Lloyd CR, Park S, Fei J, Vanderpool CK. 2017. The Small Protein SgrT Controls Transport Activity of

the Glucose-Specific Phosphotransferase System. J Bacteriol 199:10.1128/JB.00869-16.

8. Gottesman S. 2005. Micros for microbes: non-coding regulatory RNAs in bacteria. TRENDS in

Genetics 21:399-404.

9. Prévost K, Salvail H, Desnoyers G, Jacques J, Phaneuf É, Massé E. 2007. The small RNA RyhB

activates the translation of shiA mRNA encoding a permease of shikimate, a compound involved in siderophore synthesis. Mol Microbiol 64:1260-1273.

10. Saramago M, Bárria C, dos Santos RF, Silva IJ, Pobre V, Domingues S, Andrade JM, Viegas SC, Arraiano CM. 2014. The role of RNases in the regulation of small RNAs. Curr Opin Microbiol 18:105-115.

11. Papenfort K, Sun Y, Miyakoshi M, Vanderpool CK, Vogel J. 2013. Small RNA-mediated activation

of sugar phosphatase mRNA regulates glucose homeostasis. Cell 153:426-437.

12. Wright PR, Richter AS, Papenfort K, Mann M, Vogel J, Hess WR, Backofen R, Georg J. 2013.

Comparative genomics boosts target prediction for bacterial small RNAs. Proc Natl Acad Sci U S A 110:E3487-96.

13. Kery MB, Feldman M, Livny J, Tjaden B. 2014. TargetRNA2: identifying targets of small regulatory

RNAs in bacteria. Nucleic Acids Res 42:W124-W129.

14. Vogel J, Wagner EGH. 2007. Target identification of small noncoding RNAs in bacteria. Curr Opin

Microbiol 10:262-270.

15. Lalaouna D, Prévost K, Eyraud A, Massé E. 2017. Identification of unknown RNA partners using

MAPS. Methods 117:28-34.

16. Han K, Tjaden B, Lory S. 2016. GRIL-seq provides a method for identifying direct targets of

bacterial small regulatory RNA by in vivo proximity ligation. Nature microbiology 2:16239. 17. Miyakoshi M, Chao Y, Vogel J. 2015. Regulatory small RNAs from the 3’ regions of bacterial

mRNAs. Curr Opin Microbiol 24:132-139.

18. van der Meulen SB, de Jong A, Kok J. 2016. Transcriptome landscape of Lactococcus lactis reveals

many novel RNAs including a small regulatory RNA involved in carbon uptake and metabolism. RNA Biol 13:353-366.

(23)

134

lactis to Environmental Stresses Reveals Differentially Expressed Small Regulatory RNAs and

tRNAs. Frontiers in Microbiology 8:1704.

20. Goel A, Santos F, Vos WM, Teusink B, Molenaar D. 2012. Standardized assay medium to measure

Lactococcus lactis enzyme activities while mimicking intracellular conditions. Appl Environ

Microbiol 78:134-143.

21. Solem C, Defoor E, Jensen PR, Martinussen J. 2008. Plasmid pCS1966, a new selection/

counterselection tool for lactic acid bacterium strain construction based on the oroP gene, encoding an orotate transporter from Lactococcus lactis. Appl Environ Microbiol 74:4772-4775. 22. Pinto JP, Zeyniyev A, Karsens H, Trip H, Lolkema JS, Kuipers OP, Kok J. 2011. pSEUDO, a genetic

integration standard for Lactococcus lactis. Appl Environ Microbiol 77:6687-6690.

23. Overkamp W, Beilharz K, Detert Oude Weme R, Solopova A, Karsens H, Kovacs A, Kok J, Kuipers OP, Veening JW. 2013. Benchmarking various green fluorescent protein variants in Bacillus

subtilis, Streptococcus pneumoniae, and Lactococcus lactis for live cell imaging. Appl Environ

Microbiol 79:6481-6490.

24. Geu-Flores F, Nour-Eldin HH, Nielsen MT, Halkier BA. 2007. USER fusion: a rapid and efficient

method for simultaneous fusion and cloning of multiple PCR products. Nucleic Acids Res 35:e55. 25. Nørholm MH. 2010. A mutant Pfu DNA polymerase designed for advanced uracil-excision DNA

engineering. BMC biotechnology 10:21.

26. Aranda PS, LaJoie DM, Jorcyk CL. 2012. Bleach gel: a simple agarose gel for analyzing RNA quality.

Electrophoresis 33:366-369.

27. Langmead B, Salzberg SL. 2012. Fast gapped-read alignment with Bowtie 2. Nature methods 9:357-359.

28. de Jong A, van der Meulen S, Kuipers OP, Kok J. 2015. T-REx: Transcriptome analysis webserver

for RNA-seq Expression data. BMC Genomics 16:663.

29. Lalaouna D, Carrier M, Semsey S, Brouard J, Wang J, Wade JT, Massé E. 2015. A 3′ external

transcribed spacer in a tRNA transcript acts as a sponge for small RNAs to prevent transcriptional noise. Mol Cell 58:393-405.

30. Zuker M. 2003. Mfold web server for nucleic acid folding and hybridization prediction. Nucleic

Acids Res 31:3406-3415.

31. Gorski SA, Vogel J, Doudna JA. 2017. RNA-based recognition and targeting: sowing the seeds of

specificity. Nat Rev Mol Cell Biol 18:215-228.

32. Hsieh PK, Richards J, Liu Q, Belasco JG. 2013. Specificity of RppH-dependent RNA degradation in

Bacillus subtilis. Proc Natl Acad Sci U S A 110:8864-8869.

33. Puri P, Wetzel C, Saffert P, Gaston KW, Russell SP, Cordero Varela JA, Vlies P, Zhang G, Limbach PA, Ignatova Z. 2014. Systematic identification of tRNAome and its dynamics in Lactococcus

lactis. Mol Microbiol 93:944-956.

34. Lopez PJ, Marchand I, Joyce SA, Dreyfus M. 1999. The C-terminal half of RNase E, which organizes

the Escherichia coli degradosome, participates in mRNA degradation but not rRNA processing in

vivo. Mol Microbiol 33:188-199.

35. Lasa I, Toledo-Arana A, Dobin A, Villanueva M, de los Mozos IR, Vergara-Irigaray M, Segura V, Fagegaltier D, Penades JR, Valle J, Solano C, Gingeras TR. 2011. Genome-wide antisense

transcription drives mRNA processing in bacteria. Proc Natl Acad Sci U S A 108:20172-20177. 36. Romeo Y, Obis D, Bouvier J, Guillot A, Fourçans A, Bouvier I, Gutierrez C, Mistou M. 2003.

(24)

135

5

Osmoregulation in Lactococcus lactis: BusR, a transcriptional repressor of the glycine betaine

uptake system BusA. Mol Microbiol 47:1135-1147.

37. Shepherd J, Ibba M. 2015. Bacterial transfer RNAs. FEMS Microbiol Rev 39:280-300.

38. Wels M, Groot Kormelink T, Kleerebezem M, Siezen RJ, Francke C. 2008. An in silico analysis

of T-box regulated genes and T-box evolution in prokaryotes, with emphasis on prediction of substrate specificity of transporters. BMC Genomics 9:330-2164-9-330.

39. Guédon E, Serror P, Ehrlich SD, Renault P, Delorme C. 2001. Pleiotropic transcriptional repressor

CodY senses the intracellular pool of branched-chain amino acids in Lactococcus lactis. Mol

Microbiol 40:1227-1239.

40. Gasson MJ. 1983. Plasmid complements of Streptococcus lactis NCDO 712 and other lactic

streptococci after protoplast-induced curing. J Bacteriol 154:1-9.

41. Kuipers OP, de Ruyter PG, Kleerebezem M, de Vos WM. 1998. Quorum sensing-controlled gene

(25)