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: 149PDF page: 149PDF page: 149PDF page: 149

CHAPTER 7

(3)

150

The work presented in this thesis has investigated the presence and function of small regulatory RNAs (sRNAs) in the lactic acid bacterium Lactococcus lactis. This Gram-positive bacterium is used world-wide in milk fermentations and has been very instrumental as a model organism for lactic acid bacteria (LAB). The industrial application of L. lactis, not only in the dairy industry but also in entirely novel biotechnological approaches, has greatly been improved by fundamental knowledge of how the cells are functioning. This has been gained from genetic information on the level of DNA and RNA, but also from proteome data. With this work we have provided more insight in the transcriptome landscape of L.

lactis and have uncovered several novel sRNAs and their regulons.

Introduction

The concept of regulation by RNAs has now been documented in all domains of life. For a long time however, in the central dogma in molecular biology RNA has been mainly considered as a messenger, an adaptor and a structural molecule and the potential regulatory roles of RNA molecules have been overlooked. We now know that sRNAs from bacteria are not only abundant in number, but also diverse in their origins, modes of action and the cellular processes in which they play a role (1). Although no experimental evidence was published on the presence, let alone function, of small regulatory RNAs in Lactococcus lactis

when this work was started, it came as no surprise that they existed (Chapter 2). The real challenge after discovering hundreds of candidate RNA regulators lies in the dilemma which RNA should be prioritized to studied first? A combination of determining sRNA expression levels under different conditions, promoter predictions, identifying possible transcription factor binding sites as well as terminator and secondary structures, looking for conservation among (related) species and in silico target prediction have influenced the choice of studying

a number of the putative sRNAs identified here. The sRNAs studied in this thesis function in different cellular processes, ranging from arginine and carbon metabolism to stress. Below, a summary is presented of the highlights of this study, difficulties in the field, possible applications and prospects for future research.

General remarks and findings

Using differential RNA sequencing (dRNA-seq), we were able to pinpoint transcription start sites (TSS) for described and as yet unknown genes of L. lactis MG1363 in Chapter 2. This so-called transcriptome landscape led to the identification of several hundred potential RNA regulators. These novel RNAs were classified into 186 trans-encoded small regulatory RNAs

(sRNAs) located intergenically and 60 cis-encoded antisense RNAs (asRNAs), transcribed

from the strand opposite to that on which a gene is located. Furthermore, 129 long 5’-UTRs (≥100 nt) were identified that potentially harbor cis-acting riboswitches. Based on

homology searches using the BSRD database (2), riboswitches for various tRNAs (T-boxes), flavin mononucleotide (FMN), fluoride, lysine, purine, thiamine pyrophosphate (TPP) and

(4)

151

7

prequeuosine 1 (preQ1) have been found in one or more leader sequences.

In total, the presence and size of fifteen sRNAs were validated by Northern hybridization. Information about the expression of the sRNAs was obtained using RNA derived from L. lactis

MG1363 cells taken from different points during growth or after exposure to various stress conditions. Since this approach is rather laborious, an experiment was set-up in Chapter 3 in which L. lactis cells were exposed to a 5-min period to cold (10°C), heat (42°C), acid (pH

4.5), osmotic (2.5% NaCl), oxidative (shaking) or starvation (PBS) stress. RNA was isolated from these stress-induced cells and analyzed by RNA-seq. Thus, we aimed to focus on the primary effects of these conditions on the transcriptome. The majority of genes that have been reported previously to be induced after a certain stress condition were also identified but, next to that, we provided expression data on all novel sRNAs and asRNAs, In addition, we could show that tRNAs are predominantly down-regulated after a short period of several stress conditions. One of the sRNAs, CisR, was shown to be clearly induced after cold stress, an observation that has led to a follow-up study aiming, among others, at identifying its targets (Chapter 5).

Four of the novel sRNAs were cloned in inducible gene expression plasmids and were overexpressed during 10 min, after which RNA-seq was performed to identify possible mRNA targets. In Chapter 2, we show that overexpression of LLMGnc_147 leads to a strong up-regulation of an operon involved in carbon uptake and metabolism. Although LLMGnc_147 was dominantly expressed when the cells were supplied with cellobiose, overexpression of the sRNA led to a better growth on galactose as the sole carbon source. Overexpression of the abundant and strongly conserved 6S RNA (LLMGnc_004) uncovered no clear targets but the expression of 6S was carbon source-dependent, suggesting that L. lactis 6S RNA is

regulated by the transcriptional repressor CcpA.

In Chapter 4 we show that the sRNA ArgX (LLMGnc_072) is transcribed from the 3’-UTR of the gene argR, encoding the transcriptional repressor of arginine metabolism in L. lactis

MG1363. ArgX is transcribed from its own promoter but shares a common terminator with

argR. A transcriptional fusion between the promoter of ArgX, PArgX, and the gene for green

fluorescent protein gfp identified arginine as an important inducer and CcpA as a repressor

of this promoter. The key finding is that the sRNA ArgX regulates the arc mRNA, encoding

the arginine deaminase pathway. In contrast to the protein regulator ArgR, which regulates

arc at the level of transcription, ArgX acts on arc mRNA stability, most probably through

binding at the ribosome binding site of arcC1, which would, in addition, lead to inhibition of

the translation of the latter gene.

In Chapter 5, the existence of the sRNA CisR (LLMGnc_082) was verified by Northern analysis and a promoter analysis was conducted. tRNAs were identified as potential targets of CisR using overexpression and MS2-affinity purification coupled with RNA sequencing (MAPS). A third method such as GRIL-seq (16) could be optionally used to further investigate the target spectrum of CisR. Further verification of already identified CisR targets such as busR and

(5)

152

fruR by fusing the mRNAs to LacZ or GFP, is recommended for future research. The analysis of large datasets

Ever since the “omics” era has arrived, molecular biologists have the luxury to generate and analyze large datasets derived from topics such as genomes, transcriptomes and metabolomes. These fields have provided a tremendous insight in how various processes such as general metabolism, cell division and adaptation function in bacterial cells. With every new discovery more questions arise and this is certainly the case for generating large datasets. Due to technological developments in high-throughput protein analysis and DNA sequencing, and the fact that the price for DNA sequencing decreases every year, more and more data is produced. Nowadays, the main challenge for researchers is not how to create data but rather how to mine and analyze it. In other words, how can one draw biological conclusions from large data sets? In this thesis, we have used RNA-seq for example to gain insight in cellular responses after a short period of various stress conditions (Chapter 3). During the analysis of the data set obtained during this study, we expanded the toolbox for RNA-seq data analysis after the point where RKPM values have been extracted from the dataset. All these separate analysis scripts eventually have led to a fully automated pipeline for data analysis. This Transcriptome analysis webserver for RNA-seq Expression data (T-REx) enables biologists performing statistical analysis and visualize the data, with a large output of tables, figures and matrices to choose from (Chapter 6). The whole analysis only requires a few text files as an input and a few minutes for the program to run. Some of the T-REx output can be further mined on an interactive webserver (http://genome2d.molgenrug.nl). To uncover the biological meaning in sets of genes that are differentially expressed in a dataset analyzed by for example T-REx, a Gene Set Enrichment Analysis tool was developed (GSEA-Pro; de Jong et al., submitted). GSEA-Pro allows for a quick examination of

high-throughput data for biological functions and is based on various classifications such as COG, KEGG, GO, InterPro and PFAM. Both the T-REx and the GSEA-Pro tool allow biologists to fast and reliably analyze their data sets, something that can be easily reproduced by other scientists who want to analyze the same data set.

Complexity of bacterial genomes

To unambiguously describe operons in bacteria is difficult. Differential RNA sequencing (dRNA-seq) has revolutionized the prediction of transcription start sites (TSSs) and therefore has allowed identifying novel RNA species as well as determining how operons are structured in bacterial genomes. Witnessing the outcome of these genome-wide RNA studies, it is not surprising that operon prediction has been such a challenging job. In L. lactis, as has also

been observed for example in Helicobacter pylori (3), many sub-operons exist because of

the presence of multiple promoters driving transcription until RNA polymerase encounters a terminator that is common to the (sub-) operons. Also, read-through of weak terminator

(6)

153

7

structures results in operons being longer than originally anticipated on the basis of mere DNA sequence gazing. Besides operon structures being more flexible than expected, (weak) promoters that are present in an open reading frame of a coding gene create the possibility that smaller versions of the encoded protein are being made. Some derive from mRNAs that contain a ribosomal binding site (RBS), while others are encoded by leaderless mRNAs (4). Since these mRNAs are transcribed from different promoters, their regulation might be different as well. Possibly, transcribing mRNAs with different lengths to produce various versions of the same protein is easier to evolve when employing two promoters then by integrating both functions under control of one promoter.

A perspective on the application of regulatory RNAs in the dairy industry

Other than proteins, sRNAs by themselves do not perform cellular functions such as e.g.,

cell wall synthesis or metabolic reactions. Some RNAs perform enzymatic reactions, such as RNase P (5), whereas riboswitches have the ability to bind a variety of metabolites (6). Regulatory sRNAs mainly interact with mRNAs and RNA binding proteins such as Hfq (7, 8) or ProQ (9). Thus, the presence or absence of certain sRNAs in L. lactis will not directly

influence functionalities that for example change the ability to metabolize a specific sugar or alter desired flavors or aromas in dairy products. Nevertheless, as shown in Chapter 2 for sRNA LLMGnc_147 (4), a regulatory small RNA can influence genes involved in cellular functions such as carbon uptake and metabolism. By controlling the expression of sRNAs, either by triggering the promoter when the inducer is known, or by inducible overexpression using genetic techniques, target genes can be regulated via interaction of the sRNA with the mRNAs. The latter genetic engineering method is currently not allowed and also not favored by the dairy industry and many dairy product consumers alike. As most sRNAs regulate more than one mRNA target it would be difficult to steer only one cellular process by controlling the expression of a single sRNA. In the past, antisense RNAs have been successfully designed to specifically target phage RNAs to increase phage resistance (10-12). Chapter 2 describes the discovery of 60 antisense RNAs, of which more than a third are encoded by (remnants of) pro-phage genomes (4). By studying these antisense RNAs in combination with recent mechanistic insights in the functioning of sRNAs, more effective antisense RNAs or sRNAs could be designed to defend against phage attack or to steer desirable processes in the cell. A first proof of principle was established In E. coli, which demonstrated the usefulness

of artificial sRNAs (13). In a further step, the CRISPR system was combined with asRNAs to enable the simultaneous reversible repression or de-repression of multiple target genes (14). From a multispecies perspective, such as is the case in complex starter cultures used in cheese making or yoghurt production, one could eliminate a certain strain by using antisense RNAs targeting essential genes. This could be done at a certain point during fermentation or cheese ripening to modify product properties. To increase the stability of antisense RNAs in the extracellular space (the fermentation broth) and their ability to enter a target

(7)

154

bacterium, Morpholino oligos (15) conjugated to cell-penetrating peptides (CPP) could be used. CPP-Morpholino’s have been employed to treat viral, bacterial and genetic diseases (16). Answers to a combination of ethical and economical questions should determine whether or not such RNA-based developments would be adopted in the dairy industry.

Future experiments and forecast

Despite the vigorous mining of the transcriptome landscape in L. lactis that was performed

in this study, it cannot be excluded that there are more sRNAs waiting to be discovered. Specifically those sRNAs that are processed from mRNAs were not identified here as in our TSS calling we focused on primary transcripts. Novel RNA-seq approaches such as Cappable-seq (17) can potentially improve the transcriptome landscape of L. lactis even further. Using

this method on RNA extracted from cells grown under a variety of conditions would most probably allow identifying additional sRNAs. The resolution can increase further with more precise enrichment strategies for the 5’- or 3’-RNA ends, in combination with a higher sequence depth, which would make TSS calling more reliable. As TSS calling algorithms are continuously improving, we expect to see a faster and more accurate annotation of existing and newly sequenced genomes. Novel methods such Grad-seq (9) can provide the transcriptional landscape with an extra dimension by also including the network of RNA-protein interactions.

We have discovered hundreds of novel RNA elements in L. lactis, of which 186 are denoted

as sRNAs. For a number of these sRNAs, we have identified targets and the cellular functions in which these sRNAs are likely to play a role: LLMGnc_147 in carbon metabolism, ArgX in arginine metabolism and CisR in (cold) stress. Future work on these sRNAs should include identifying more targets using approaches like GRIL-seq and, for some targets, more validation. A better understanding of the molecular mechanisms of the sRNAs ArgX and CisR would be of great interest. This might include detailed studies as to which nucleotides are exactly involved in the binding of the target mRNA, which RNases are involved in the cleavage of both target and sRNA and what the possible roles are of RNA chaperones, especially since the genome of L. lactis does not encode an Hfq and/or ProQ homologue

(see Chapter 1).

A completely unexplored possibility for the hundreds of sRNAs per bacterial genome is that some might have riboswitch-like capabilities of binding metabolites. Such an sRNA-riboswitch hybrid could undergo a structural change upon metabolite binding and thus influence the function of the sRNA as an additional switch. Depending on the presence or absence of the specific metabolite, the regulatory role of the sRNA would either be activated or inactivated.

Small Open Reading Frames (sORFs), resulting in small peptides of 10-50 amino acid residues (AA) in length, have long been overlooked (18). The fact that the existence and possible roles of small peptides have not been assessed is partly due to biochemical techniques not

(8)

155

7

being sensitive enough, but also because of practical reasons while otherwise the number of additional ORFs would increase dramatically, as seen during annotation of the yeast genome (19). In Chapter 2, we have assessed long mRNA leaders sequences, novel sRNAs and asRNAs for the presence of sORFs preceded by a minimal RBS and consisting of ≥ 20 codons. For none of these sORFs we have investigated whether or not they are actually translated or whether they could be functional. Characterization of small peptides in other work has shown them to have a great functional diversity; they play a role in cell division (20), spore formation (21), quorum sensing (22) or in the regulation of multidrug transporters (23). It is of high interest to study the function of sORFs in L. lactis in follow-up research.

Concluding remarks

In this thesis, a first genome-wide assessment of L. lactis transcripts was reported. Based on

differential RNA-seq, transcription start sites were detected for known coding genes, small ORFs and unknown putative RNA regulators. From a plethora of antisense RNAs, sRNAs from intergenic regions and riboswitches, a number of these entirely novel sRNAs have been studied in more detail. LLMGnc_147 was shown to induce/stabilize an operon involved in carbon uptake and metabolism, whereas ArgX regulates the arc operon in parallel with the

protein regulator ArgR. For the cold stress induced sRNA CisR, several potential targets have been identified, including tRNAs, a class of RNAs that so far have escaped the repertoire of possible sRNA-targets . We envision that this study is an excellent starting point for further characterization of a variety of RNAs in L. lactis.

(9)

156

REFERENCES

1. Wagner EGH, Romby P. 2015. Chapter Three-Small RNAs in Bacteria and Archaea: Who They Are,

What They Do, and How They Do It. Adv Genet 90:133-208.

2. Li L, Huang D, Cheung MK, Nong W, Huang Q, Kwan HS. 2013. BSRD: a repository for bacterial

small regulatory RNA. Nucleic Acids Res 41:D233-8.

3. Sharma CM, Hoffmann S, Darfeuille F, Reignier J, Findeiß S, Sittka A, Chabas S, Reiche K, Hackermüller J, Reinhardt R. 2010. The primary transcriptome of the major human pathogen

Helicobacter pylori. Nature 464:250-255.

4. 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.

5. Guerrier-Takada C, Gardiner K, Marsh T, Pace N, Altman S. 1983. The RNA moiety of ribonuclease

P is the catalytic subunit of the enzyme. Cell 35:849-857.

6. Breaker RR. 2012. Riboswitches and the RNA world. Cold Spring Harb Perspect Biol 4:10.1101/

cshperspect.a003566.

7. Vogel J, Luisi BF. 2011. Hfq and its constellation of RNA. Nature Reviews Microbiology 9:578-589. 8. Brennan RG, Link TM. 2007. Hfq structure, function and ligand binding. Curr Opin Microbiol

10:125-133.

9. Smirnov A, Förstner KU, Holmqvist E, Otto A, Günster R, Becher D, Reinhardt R, Vogel J. 2016.

Grad-seq guides the discovery of ProQ as a major small RNA-binding protein. Proceedings of the National Academy of Sciences 201609981.

10. Kim SG, Batt CA. 1991. Antisense mRNA-Mediated Bacteriophage Resistance in Lactococcus lactis

subsp. lactis. Appl Environ Microbiol 57:1109-1113.

11. Walker SA, Klaenhammer TR. 2000. An explosive antisense RNA strategy for inhibition of a

lactococcal bacteriophage. Appl Environ Microbiol 66:310-319.

12. McGrath S, Fitzgerald GF, van Sinderen D. 2001. Improvement and optimization of two engineered

phage resistance mechanisms in Lactococcus lactis. Appl Environ Microbiol 67:608-616. 13. Man S, Cheng R, Miao C, Gong Q, Gu Y, Lu X, Han F, Yu W. 2011. Artificial trans-encoded small

non-coding RNAs specifically silence the selected gene expression in bacteria. Nucleic Acids Res

39:e50.

14. Lee YJ, Hoynes-O’Connor A, Leong MC, Moon TS. 2016. Programmable control of bacterial gene

expression with the combined CRISPR and antisense RNA system. Nucleic Acids Res

44:2462-2473.

15. Heasman J. 2002. Morpholino oligos: making sense of antisense? Dev Biol 243:209-214. 16. M Moulton H. 2013. In vivo delivery of morpholino oligos by cell-penetrating peptides. Curr

Pharm Des 19:2963-2969.

17. Ettwiller L, Buswell J, Yigit E, Schildkraut I. 2015. A novel enrichment strategy reveals

unprecedented number of novel transcription start sites at single base resolution in a model prokaryote and the gut microbiome. bioRxiv 034785.

18. Storz G, Wolf YI, Ramamurthi KS. 2014. Small Proteins Can No Longer Be Ignored. Annu Rev

(10)

157

7

19. Basrai MA, Hieter P, Boeke JD. 1997. Small open reading frames: beautiful needles in the

haystack. Genome Res 7:768-771.

20. Karimova G, Davi M, Ladant D. 2012. The beta-lactam resistance protein Blr, a small membrane

polypeptide, is a component of the Escherichia coli cell division machinery. J Bacteriol

194:5576-5588.

21. Ebmeier SE, Tan IS, Clapham KR, Ramamurthi KS. 2012. Small proteins link coat and cortex

assembly during sporulation in Bacillus subtilis. Mol Microbiol 84:682-696.

22. Kleerebezem M, Quadri LE, Kuipers OP, De Vos WM. 1997. Quorum sensing by peptide

pheromones and two-component signal-transduction systems in Gram-positive bacteria. Mol Microbiol 24:895-904.

23. Hobbs EC, Yin X, Paul BJ, Astarita JL, Storz G. 2012. Conserved small protein associates with the

multidrug efflux pump AcrB and differentially affects antibiotic resistance. Proc Natl Acad Sci U S A 109:16696-16701.

(11)