University of Groningen Cell envelope related processes in Bacillus subtilis van den Esker, Mariëlle Henriëtte

(1)

Cell envelope related processes in Bacillus subtilis

van den Esker, Mariëlle Henriëtte

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 den Esker, M. H. (2018). Cell envelope related processes in Bacillus subtilis: Cell death, transport and cold shock. 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)

521545-L-bw-vdEsker 521545-L-bw-vdEsker 521545-L-bw-vdEsker 521545-L-bw-vdEsker Processed on: 25-7-2018 Processed on: 25-7-2018 Processed on: 25-7-2018

Processed on: 25-7-2018 PDF page: 43PDF page: 43PDF page: 43PDF page: 43

Mariëlle H. van den Esker, Maarten Mols, Jan Wintgens

and Oscar P. Kuipers

Are wall teichoic acids essential or not?

RNA sequencing reveals a large

secondary mutation in a

Bacillus

subtilis tagO mutant

(3)

44

Abstract

Wall teichoic acids (WTAs) are major constituents of the cell wall of Bacillus subtilis. These

anionic polymers have many functions, including autolysin regulation, but the molecular mechanisms behind them are not all understood. In this study we wanted to examine the influence of WTA abundance on lysis behavior and autolysin transcription. We created a conditional mutant in which the expression of tagO, the first gene of the WTA

synthetic pathway, could be depleted or repleted. Division and lysis were severely affected in this mutant. RNA sequencing indicated that expression of most major autolysins was repressed in the tagO depletion mutant. However, data mapping revealed that a big

genomic region of ~207 kb was deleted in our strain, pointing at a secondary site mutation. Although the behavior and morphology of our mutant resembled the previously published

tagO deletion strains, the occurrence of secondary mutations was never excluded in those

mutants. Therefore, our study raises the question whether it is possible to delete tagO

without causing concomitant other mutations, and whether tagO mutants in other studies

(4)

45

Introduction

Wall teichoic acids (WTAs) are major constituents of the Gram-positive cell wall; up to 60% of the cell wall mass can consist of this polymer 27,118_{. The molecular mechanisms}

behind the functioning of WTAs are largely elusive, but their abundance suggests that they contribute to many functions of the cell envelope, including maintaining its tensile strength, elasticity, shape and electrostatic steering 28_{. WTAs consist of a linkage unit}

attached to the peptidoglycan and a phosphate-rich, anionic backbone comprised of glycerol 3-phosphate in the model strain Bacillus subtilis 168 28_{. The charge of WTAs can}

be modified by D-alanylation, which adds a positive charge to the WTAs thereby reducing repulsion between the negatively charged WTAs 30_{. The WTA charge influences many}

cellular processes, including ion homeostasis and murein hydrolase activity 29,50,119,120_.

Murein hydrolases (or autolysins) are enzymes that cleave peptidoglycan bonds and enable remodeling of the cell wall during elongation and cell division. B. subtilis

has 35 proposed autolysins that are classified according to their function, i.e. which peptidoglycan bond they exactly hydrolyse and in which phase they are active (vegetative growth, sporulation) 121_{. Major autolysins that are active during vegetative growth are the}

Lyt proteins (LytC, LytD, LytE, LytF). These proteins are involved in cell wall turnover and separation, and their functions are partially redundant. Autolysin activity should be tightly regulated, since overactivity can be potentially lethal.

WTAs control autolysin activity in multiple ways, although not all details are yet understood. Firstly, WTAs influence the localization of certain autolysins 50,54,122_{. For}

example, the major autolysin LytF is present only at the division septum in B. subtilis, but

when WTAs are depleted, LytF is spread across the sidewall 54_{. Secondly, WTAs affect}

the activity of autolysins directly by creating local differences in the pH of the cell wall, especially via D-alanylation 29,53,123,124_{. Autolysins are active at various pH gradients, and the}

pH variations created by WTAs within the cell wall govern the activity of autolysins. A general accepted model suggests that the negatively charged backbones of WTAs bind the positively charged autolysins, and D-alanylation releases the binding, thereby inducing autolysin activity 28_{. Finally, studies have reported that the amount of WTAs present in}

the cell wall influences the transcription of the LytE and LytF murein hydrolases 122,125_.

Thus, WTAs play an important role in regulating autolysin activity and in other functions, and were long thought to be essential in B. subtilis13,25,28,126_{. However, a study}

of d’Elia et al. indicated that it is possible to remove tagO (from teichoic acid glycerol), the

first gene in the WTA synthetic pathway, resulting in complete absence of WTAs in the cell wall 33_{. This leads to an aberrant phenotype: cells lose their rod-shape and become}

(5)

46

circular, and aggregate together 33,34,126_{. Moreover, growth is severely restrained, and it}

has been proposed that the formation of small colonies is the reason that researchers previously thought it is impossible to delete tagO33_._{tag-genes encoding for proteins active}

later in the synthetic pathway cannot be deleted, unless tagO is deleted, possibly due to

the build-up of toxic intermediates 28,33_{. Although complementation of the mutant was not}

accomplished, d’Elia et al. speculate that the chance of suppressor mutations rescuing the tagO deletion mutant is low, because similar phenotypes were observed when genomic

DNA was transformed to the wild-type strain 33_.

It is known that WTAs play an important role in regulating autolysin activity, and that it is possible to (conditionally) delete genes involved in the WTA synthetic pathway, but the effect of WTA absence on lysis is poorly characterized in B. subtilis. Therefore,

in this study we evaluated the consequence of WTA depletion on morphology, growth and autolysis. Furthermore, we analyzed the impact of tagO deletion and repletion on

autolysin transcription using RNA sequencing. Remarkably, data mapping revealed that a big genomic region of ~207 kb was missing in our tagO mutant, which partially overlapped

with the SPβ-prophage region. This indicates that a large secondary mutation event did occur in our strain that was previously left unnoticed, raising again the debate of WTA essentiality.

Results and Discussion

Effect of wall teichoic acid deficiency on cell morphology and division

WTAs influence cell morphology and growth, and are essential for the regulation of autolysin activity, although complete mechanistic details are lacking for B. subtilis. To

gain more insight in the effect of WTA deletion on autolysin activity in B. subtilis 168, a

conditional knockout strain was constructed (∆tagO). The native tagO gene was replaced

by a neomycin cassette, and a second copy was placed back in the ectopic amyE locus

with the tight spank promoter in front of the gene enabling control of tagO expression via

IPTG induction.

The effect of tagO depletion on cell division and morphology was analyzed by time-

lapse microscopy. The wild-type and ∆tagO strain were grown on LB agarose slides for

10 hr using a membrane dye to visualize the cell membrane (Fig. 1A). The ∆tagO mutant

appeared to be much weaker compared to the wild-type; only a small part of the cells inoculated on a LB agarose slide grew out into a microcolony, whereas the major part lysed without dividing. For the wild-type, the fraction of growing cells was nearly 100% (data not shown). We quantified cell size and division time of cells in three growing

(6)

47

microcolonies of the wild-type and the ∆tagO mutant. Morphology of the ∆tagO mutant

was aberrant from the wildtype with round cells that lost their rod-shape, agreeing with previous studies 33,34,126_{. Membrane staining revealed that these circular cells consisted}

Figure 1. (A) Still frames from a time-lapse microscopy movie showing the outgrowth of a ∆tagO (upper frames) and wild-type (lower frames) microcolony. Different time points were selected due to variations in growth rates for both strains. The scale bar represents 5 µm. (B) Box plots showing the division time of ∆tagO and wild-type (WT). Dots represent the 5th and 95th percentile, and the whiskers show the 10th and 90th percentile of the data. (C) Box plots demonstrating the cell length of the ∆tagO mutant and the wild-type (WT). Dots represent the 5th and 95th percentile, and the whiskers show the 10th and 90th percentile of the data.

(7)

48

of upto four tiny daughter cells that had not yet separated (Fig. 1A; e.g. at 230 min). On LB, the wild-type chained, and due to fast growth (20-30 min division time; Fig. 1B) cells grew on top of each other after two hours. The division rate of the ∆tagO mutant was much

lower, with an average of 150-160 min per division (Fig. 1B). Daughter cells produced by the mutant were considerably smaller than cells of the wild-type (Fig. 1C).

Wild-type cells divided evenly by forming a septum in the middle of the cell at regular intervals, yielding two daughter cells of the same size. These cells first elongated followed by another round of division, whereas in the mutant no real elongation occurred. In B. subtilis, LytE and CwlO are the two main murein hydrolases involved in cell elongation

and shape determination; inactivation of these proteins blocks cell elongation 127,128_.

Remarkably, LytE transcription was previously shown to be induced in a tagO null mutant,

indicating that cells try to compensate for the lack of sidewall LytE activity by inducing its transcription 129_{. Hence, cell morphology, WTA abundance and murein hydrolase}

activity are tightly coupled.

The effect of WTA abundance on growth was further assessed by growing the wild-type and ∆tagO mutant in the presence and absence of IPTG in a 96-well plate.

WTA depletion significantly reduced growth in a 96 well plate, while IPTG addition partially restored this (Fig. 2). The cell shape of the ∆tagO mutant was reverted to

rod-shape in the presence of IPTG, but cells looked bigger and more irregular than the

Figure 2. Growth of the wild-type (circles) and ∆tagO mutant (triangles) in the presence (black) and absence (white) of 1 mM IPTG, with pictures on the right side showing the morphology of the different strains. Cultures were grown in 200 µl LB in a 96-well plate and the absorbance at 600 nm was quantified every 10 min. Scale bar is 5 µm.

(8)

49

wild-type (Fig. 2). The addition of 1 mM IPTG presumably resulted in overexpression of TagO and thereby, the overproduction of WTA linkage units. MraY (the first enzyme involved in the synthesis of Lipid I required for peptidoglycan synthesis) and TagO both use undecaprenyl-phospate (UDP) as precursor for peptidoglycan respectively WTA synthesis. Thus, when TagO is overexpressed, reduced availability of the UDP precursor for peptidoglycan synthesis could lead to this aberrant cell shape 25_{. Moreover, a longer}

lag-phase was observed for the ∆tagO mutant, as was described before 33_{, and the addition}

of IPTG did not alter this. Hence, TagO repletion or depletion led to severe deviations in growth, division and morphology, indicating that the cell wall and autolysin activity were affected.

Wall teichoic acid depletion affects autolysis in B. subtilis

It is already known that Staphylococcus aureus mutants lacking tagO are more prone to lysis

than wild-type cells 29,50_{, but the effect of WTA depletion on lysis in}_{B. subtilis has never}

been examined. Therefore, an autolysis assay was performed. The wild-type and ∆tagO

mutant were incubated in destH₂O and HEPES buffer and the decline in optical density was quantified. We observed a higher endogenous lysis rate of the wild-type in comparison to the ∆tagO mutant in both situations: in HEPES buffer, the wild-type density reduced to

20% of its initial turbidity after 2,5 hr, whereas the density of the ∆tagO culture decreased

to 50% (Fig. 3). In destH₂O, this difference was even more profound: the culture density of the wild-type decreased to 40% while ∆tagO declined to 80% (Fig. 3). Thus, in our study, tagO depleted cells were less prone to endogenous lysis compared to the wild-type.

This is contradictory to the lysis behavior of S. aureus, although in these assays the

detergent Triton-X-100 was used as lysis buffer 29,50_{. It seems plausible that ∆}_{tagO cells are}

more susceptible to lysis than the wild-type since the cell wall structure and the charges controlling autolysin activity are not interfered with in the wild-type, but previous studies have shown that B. subtilis and S. aureus cells also respond differently to a reduction in

D-alanylation. In S. aureus, lysis was reduced in a dlt deletion mutant, whereas lysis in B. subtilis enhanced when the dlt-operon was deleted 53,123,124_{. Furthermore, the structure of the}

B. subtilis cell wall is different from S. aureus: B. subtilis exhibits a gradient in the density of the

cell wall creating a periplasmic-like space and a gradient in D-alanylation that is regulated by the proton-motive force 17,30,61,130–132_{. Autolysin activity varies therefore in different regions}

of the cell wall. The cell wall of S. aureus does not comprise such a gradient in cell wall

density, and murein hydrolases are targeted directly to the division septum 18,133,134_{. This}

suggests that S. aureus and B. subtilis employ different strategies in the regulation of autolysin

activity, making it hard to compare study outcomes between both species.

(9)

50

RNA sequencing reveals the deletion of a large genomic region

WTAs regulate autolysin activity by creating local pH differences in the cell membrane, but studies revealed that LytE and LytF transcription is also influenced by the abundance of WTAs in the cell wall 122,125_{. As discussed above, the division and lysis behavior of}_B.

subtilis is strongly affected by WTAs, but the transcriptional response has so far not

been studied well. In order to create a more complete overview of the relation between autolysin transcription and WTA presence in the cell wall, we sequenced the RNA of the wild-type, and TagO-depleted and repleted cells. The wild-type and ∆tagO were grown in

LB with and without IPTG, and RNA was isolated in the exponential phase as described in the materials and methods section. IPTG addition did not change the transcription of any genes in the wild-type significantly, therefore, data of the wild-type with and without IPTG was compared to ∆tagO with and without IPTG. In total, the expression

of 1413 RNAs were changed significantly in the ∆tagO depletion mutant compared to the

wild-type, whereas 1246 RNAs were altered when tagO was repleted by IPTG addition

Figure 3. Autolysis assay of the ∆tagO mutant (white squares) and wild-type (black circles) in HEPES buffer and destH2O. The rate of autolysis was measured as decline in optical density at 600 nm.

(10)

51

(Table S1). Hence, about one third of all genes were affected in the ∆tagO mutant, and

supplementation with IPTG did not change this number considerably. Analysis of our data revealed that all gene classes were influenced. Most notable was the class of genes with a hypothetical function: 30% of all genes that were changed belonged to this class in both situations. Another major class affected were the membrane proteins, indicating that the membrane is severely adjusted in the ∆tagO mutant.

To study the differences in autolysin expression, a list was made of autolysins involved in cell elongation, separation and wall turnover during the exponential phase (Table 1). All these autolysins were significantly changed in expression with the major part being repressed (≥12.5x), and only lytE and cwlO induced in the ∆tagO mutant. These

results are in full correspondence with previous studies that found lytF to be repressed,

while lytE transcription was induced 122,125_{. This might also (partially) explain the reduced}

lysis behavior of the ∆tagO mutant: expression of most major autolysins is reduced in

this strain compared to the wild-type, suggesting that less autolysins are present in the cell wall. Hence, the deletion of WTAs has major impact on autolysin expression. The addition of IPTG did not alter the transcriptional response of autolytic genes, suggesting that the difference in ∆tagO transcription was not (solely) responsible for the changes in

autolysin gene expression (Table 1).

To create a general overview of the transcriptome data, the reads were mapped on the genome of B. subtilis 168 using the JBrowse genome server. Remarkably, when browsing

through the genome sequence, we found that a big region of ~207 kb in the genome of the ∆tagO mutant was not expressed in the ∆tagO mutant (Fig. 4A) . This region was

located around prophage SPβ (size of 132 kb), but did not completely overlap with it. The region encoded mainly genes with unknown function (Table S2). A volcano plot that was generated from the data comparing wild-type gene expression to ∆tagO expression

revealed that nearly all genes that were induced more than seven times were located in this unexpressed region (Fig. 4B).

Table 1. Autolysin gene expression in the wild-type vs ∆tagO and in the wild-type vs ∆tagO + 1 mM IPTG

Gene Fold change

WT vs ∆tagO

Fold change

WT vs ∆tagO + IPTG

Function

lytC 25.6 18.1 major autolysin, cell separation, wall turnover

lytD 17.9 12.5 major autolysin, cell separation

lytE -10.7 -12.8 major autolysin, cell elongation and separation

lytF 32.3 36.4 cell separation

cwlO -2.3 -2.5 cell wall synthesis, cell elongation

cwlS 10334 1733 cell separation

(11)

52

We investigated whether the genes in the unexpressed region were only not expressed, or that a big genomic deletion occurred in our mutant. We performed a PCR reaction using primers amplifying gene yorL, which is located in the unexpressed region,

and as a control, we used primers amplifying ysbA, a gene located outside this region. In

the ∆tagO mutant, yorL was not amplified whereas the control indicated that the genomic

DNA was pure (Fig. S1). In the wild-type, both genes were amplified. These results suggest that a big genomic region is deleted in the ∆tagO mutant, pointing at a secondary mutation

rescuing the ∆tagO phenotype. To further confirm this, the genomic DNA of the ∆tagO

strain will be sequenced in the near future.

The secondary mutation could also explain why IPTG addition did only partly complement the ∆tagO phenotype: the expression of genes located in the deleted region

was not restored. The changes in the transcriptome are caused by both the deletion of a large genomic region and tagO repletion and depletion, and differential gene expression

can therefore not be attributed to altered TagO amounts alone. The missing genomic region partly overlapped with the SPβ-prophage region. A previous study has shown that deletion of the SPβ prophage region alone does not have major consequences for cell viability or key physiological or developmental processes 135_{. Deletion of the other genes}

located outside the SPβ-prophage region (96 genes) as well as tagO deletion could therefore

be responsible for the phenotypic changes observed.

The question remains why specifically this region was deleted during tagO depletion:

Was it a coincidence or did the deletion of a gene in this region rescue the strain? WTAs contribute with their phosphate backbone to the negative charges in the cell wall. Therefore, the charge will be more positive when WTAs are depleted, unless this is compensated by e.g. increased lipoteichoic production, which is not the case according to our data. Furthermore, WTA deletion has many pleiotropic effects, including altered ion homeostasis, membrane permeability and they act as scaffolds for extracytoplasmic enzymes such as autolysins 136_{. In order to survive, the expression of genes involved in}

these processes will possibly be adjusted.

The deleted region consists of 270 coding genes. 85% are of unknown function, or with a function predicted based on homology (so-called y-genes). Therefore, it is hard

to speculate about their exact contribution and the potential rescue mechanism. Of the remaining 24%, some genes are involved in sporulation. Also csaA is deleted. The membrane

protein CsaA acts as a chaperone that promotes the export of a subset of preproteins in B. subtilis, and was hypothesized to act similar as SecB in E. coli137_{. It affects the transport of}

at least two proteins across the membrane, and could therefore influence the presence of extracellular proteins in the membrane. Moreover, the murein hydrolase CwlS is located

(12)

53

Figure 4 (A) The JBrowse Genome server shows that a big genomic region between 2,064,200 and 2,271,657 bp is not transcribed in the ∆tagO mutant. The yellow marked region shows the location of the SPβ-prophage region. (B) Volcano plot showing differentially expressed genes in the wild-type versus the ∆tagO mutant. The yellow dots represent genes located in the deleted region. In total, this region consists of 270 genes.

(13)

54

in the deleted region, explaining the tremendous ‘induction’ of this autolysin in the wild-type (Table 1). CwlS is involved in cell separation, and deletion of this gene thus reduces autolytic activity 138_{. CzrA is also deleted. This protein is involved in the regulation of}

resistance against toxic metal cations. WTAs are involved in metal cation homeostasis, thus this deletion might influence metal homeostasis 139_{. Additional experiments should}

focus on the replicability of this deletion before any definite conclusions can be drawn. After that, more focused search could reveal which genes rescue WTA depletion.

Most interesting is that our other results fully correspond to previous published data. The phenotype of our ∆tagO mutant is comparable to the phenotype of the mutant

published before, and also LytE and LytF transcription was changed in a similar way

33,34,125,129_{. It was reasoned before that the likelihood of suppressor mutations rescuing the}

tagO mutant was low, because transformation of genomic DNA from the mutant to the

wild-type resulted in a similar phenotype 33_{. However, this does not provide full evidence}

as other mutations could also be transformed into the wild-type, or new mutations could emerge during transformation. Furthermore, complementation of the mutant was never achieved. Therefore, it cannot be excluded that in these studies rescue mutants were used too, and additional experiments (e.g. sequencing of the full genome) should eliminate this possibility. Our results thus again questions WTA essentiality: it has been shown previously that tagO can be deleted leading to WTA absence in the cell wall, but if this

inevitably leads to concomitant secondary mutations rescuing the strain, it suggests that WTAs might after all be essential.

Conclusion

In the ∆tagO mutant generated in this study a large genomic region of ~207 kb was deleted,

which was not noticed until the entire transcriptome was sequenced. This mutant had a similar morphology as other tagO deletion mutants used in previous studies, and besides, lytE and lytF expression was altered in a comparable way. Although several studies have

shown that it possible to delete ∆tagO, the occurrence of rescue mutations in those

strains has never been eliminated. Our study therefore raises the question whether it is possible to delete tagO, and if mutants in other studies might also have undergone a

secondary mutation. To validate the results of previous studies, full genome sequencing

and complementation should be performed on those strains to investigate the presence of secondary suppressor mutations.

(14)

55

Experimental procedures

Bacterial strains, plasmids and media

Bacterial strains and plasmids that were used in this study are shown in Table 2. Strains were grown in Lysogeny Broth (LB-Lennox: 1% Tryptone, 0.5% Bacto-yeast extract, 0.5 % NaCl) at 37°C, 220. When necessary, antibiotics were added to the following concentrations: ampicillin (100 µg ml-1_{) for}_{E. coli, or kanamycin (km; 5 µg}

ml-1_{), spectinomycin (spec; 100 µg ml}-1_{) for}_{B. subtilis. For induction, 1 mM isopropyl}

β-D-1-thiogalactopyranoside (IPTG) was added to the medium, unless mentioned otherwise. Solid media were prepared by adding 1.5% (wt/vol) agar.

Molecular cloning

All primers used in this study are listed in Table 3. DNA purification, restriction, ligation and transformation to E. coli were carried out as described by Sambrook et al.

(1989) 110_{. PCRs were performed using chromosal DNA of}_{B. subtilis 168 as a template.}

Phusion polymerase, restriction enzymes and T4 ligase were purchased from Fermentas. All constructs were verified by PCR and DNA sequencing (Macrogen). B. subtilis

transformations were performed as previously described 109_.

Table 2. Bacterial strains and plasmids used in this study

Strain Relevant features Reference or source

E. coli

MC1061 F-, araD139, ∆(ara-leu)7696, ∆(lac)X74, galU,

galK, hsdR2, mcrA, mcrB1, rspL

Wertman et al., 1986 112

Bacillus subtilis

168 1A700 trpC2 Kunst et al., 1997 12

168 P_sp-tagO amyE::P_spank-tagO, SpR _{This study}

168 ∆tagO tagO:: neo, amyE::P_spank-tagO, SpR _{This study}

Plasmids

pUC21 bla, lacZ Vieira and Messing, 1991 140

pBEST501 bla, neo Itaya et al., 1989 141

pUC21::neo bla, neo This study

pDR110 bla amyE’ P_spankspec lacI ‘amyE Gift of D. Rudner

pDR110-tagO bla amyE’ P_spank– tagO spec lacI ‘amyE This study

(15)

56

Table 3. Oligonucleotides used in this study

Name Sequence (5’ → 3’) tagO-F+SalI+RBS GTAACGGTCGACTAGGAGGAGAGGAAATGCTTGACGAACGCATGATTCGC tagO-R+NheI+2stop GGCCGCGCTAGCCTATTAATTCCTTTTCACCAGCCG tagO-F-up GGCCCGCTCGAGAGGTGAACGCTGCTGTTTATG tagO-R-up-SalI GATAGGTCGACGAATCATGCGTTCGTCAAGC tagO-F-down-SphI GTATGGCATGCAAACTCCGGCTTATGTGC tagO-R-down GGATTGAGCTCGAGGATTACGCGACTAAAGG yorL-F GATGACCGACTTCAGTTGAG yorL-R GAGCACGAACCAATTATCCC ysbA-F CCGCTATACGGCATGTTATC ysbA-R CCTCCTGTTTCCCTTAATGG

Plasmid and strain construction

To enable overexpression of tagO, pDR110 was used. This plasmid carries the tight

IPTG-inducible P_spankpromoter, followed by restriction sites in which the gene of interest can be ligated. tagO was amplified by PCR with primer tagO-F+SalI+RBS and tagO-R+NheI+2stop.

PCR fragments were cleaved with SalI and NheI and ligated into the corresponding sites of pDR110. After regeneration of the plasmid in E. coli, the plasmid was isolated and

transformed to B. subtilis 168. Transformants were selected on LB plates containing spec.

To create the conditional knockout strain ∆tagO, tagO was replaced with a neomycin

cassette. This cassette was first cloned into pUC21 by restricting pBEST501 and pUC21 with KpnI and SalI, after which fragments were ligated resulting in pUC21::neo. tagO up

and down flanking regions of ~1 kb were amplified with primer tagO-F-up + tagO-R-up-SalI and tagO-F-down-SphI + tagO-R-down. The resulting PCR products and pUC21::neo were digested with SphI and SalI. After ligating the neomycin cassette of pUC21::neo to the up- and downstream flanking region, the mixture was concentrated in a SpeedVac concentrator and directly transformed to natural competent B. subtilis 168 P_sp-tagO cells.

B. subtilis 168 P_sp-tagO was pregrown with IPTG (200 µg ml-1_{) in order to induce}_tagO

expression, thereby enhancing transformation efficiency. Transformants were selected overnight on LB plates containing km and IPTG (spec was not included to promote growth of the ∆tagO mutant). Colonies were inoculated in LB medium with km, and

(16)

57

Growth curves

Cultures were inoculated from -80°C stocks and grown overnight in LB medium +/- IPTG with appropriate antibiotics at 37°C, 220 rpm. The next morning, cultures were diluted in prewarmed LB medium with different concentrations IPTG to an OD₆₀₀ of 0.1 (without antibiotics). Growth was followed by measuring the absorbance at 600 nm at different timepoints.

Growth was also measured in a plate reader. For this, cultures were inoculated from -80°C stocks on LB agar plates with antibiotics, and grown overnight at 37°C. The following day, a single colony was picked and grown overnight in LB containing antibiotics when required, at 37°C, 220 rpm. After overnight growth, cultures were diluted in a 96-well microtiter plate by adding 2 µl of the culture in 200 µl fresh medium. This plate was incubated in an Infinite 200 plate reader (Tecan) and i-control 1.10 software (Tecan) at 37°C, 4 mm orbital shaking. Measurements were taken every 10 minutes. Absorbance values were collected at 600 nm, Data of all samples were collected in duplicates, and processed in Microsoft Excel by correcting the OD₆₀₀for background absorbance of the medium.

Autolysis assay

Autolysis assays were performed by growing cells overnight in LB (with km and IPTG when appropriate) at 37°C, 220 rpm. The following morning, cells were diluted in fresh LB medium to an OD₆₀₀ of 0.1 and grown at 37°C, 220 rpm to mid-exponential phase (OD₆₀₀ of ~0.5-1). Cells were washed twice in ice-cold Phosphate-buffered Saline (PBS) by spinning for 1 min at 14,000 rpm, 4°C, and diluted to an OD₆₀₀of 1 in either 2 ml 0.1 M HEPES buffer (pH 7), or 2 ml destH₂O. The cultures were incubated as before, and the amount of lysis was calculated as relative decrease in absorbance at 600 nm over time. Time-lapse microscopy

To visualize cell morphology, common light microscopy was performed. Cells were inoculated from -80°C stocks and grown overnight at 37°C in LB medium containing km and IPTG. The following morning, cells were washed two times in PBS, and diluted in LB with or without IPTG. At different time points, samples were taken and –if the cell density was low- concentrated by centrifugation (14,000 rpm for 1 min). Samples were used for microscopy by spotting 1 µl on a slide prepared with 1% agarose in PBS, which was air-dried to solidify. When necessary, membranes were stained by adding 2 µl FM 1-43 dye (Invitrogen, USA; concentration 0.5 mg ml-1_{) to 500 ul PBS with agarose.}

(17)

58

Time-lapse microscopy was performed essentially as described previously 142_{, except}

that LB medium was used since the ∆tagO mutant did not grow on agarose slides prepared

with chemically defined medium. In short, overnight cultures were inoculated from -80°C stocks and grown at 37°C in LB medium +/- IPTG and km. The following morning, the cells were loaded on a slide containing LB and FM 1-43 dye, solidified with 1.5% high- resolution agarose (Sigma). The outgrowth of a single cell into a microcolony monolayer was achieved by growing the cells in an environmental chamber at 37°C for 16 h. Before capturing the first image, the cells were incubated for 1 h in the environmental chamber at 37°C to improve autofocusing.

Bacterial cells were visualized using an Olympus DeltaVision microscope (Applied Precision) with a 300W Xenon Light source, CoolSNAP HQ2 camera (Princeton Instruments) and a 100x phase contrast objective (Olympus PlanApo 1.40 NA). A GFP filterset (Chroma, excitation at 470/40 nm, emission 525/50 nm) was used to detect the membrane dye FM 1-43. Microscopy data were stored using softWoRx 3.6.0 (Applied Precision) and further processed and analyzed using ImageJ 143_.

RNA isolation and transcriptome experiments

B. subtilis 168 wild-type and ∆tagO were grown with and without 1 mM IPTG as described

above (growth curves). Cultures were harvested in exponential phase by centrifugation (1 min, 14,000 rpm) and the pellet was immediately frozen in liquid nitrogen and stored at -80°C until further extraction.

RNA was isolated by resuspending the cell pellet in 400 µl of TE (10 mM of Tris·HCl, 1 mM of EDTA, pH 8.0). Subsequently, 500 µl phenol/chloroform, 50 µl 10% SDS (Sodium dodecyl sulfate) and 0.5 g glass beads were added. The cells were lysed by pulsing two rounds of one minute in a Mini-BeadBeater (Biospec Products, Bartlesville, USA) at 4°C. In between, cells were cooled on ice. After centrifugation (14,000 rpm for 10 min), the supernatant was collected and 500 µl chloroform was added. After centrifugation as before, the supernatant was used to isolate RNA with the NorGen Total RNA purification kit (Cat #17200, Biotek Corp, Canada) according to manufacturer’s instructions. The optional DNase treatment step was included, and to avoid degradation of the RNA, 3 µl RiboLock (Fermentas, USA) was added to the enzyme mixture. The concentration of the extracted RNA was measured using NanoDrop ND-1000 (Thermo Scientific, USA) and quality was analyzed using bleach gel electrophoresis 144_.

After RNA isolation, RNA was sent to the PrimBio Research Institute (Exton, USA). There, quality was analyzed by determining the RNA Integrity Number and RNA was sequenced using an Ion ProtonTM_{Sequencer. The quality of the transcriptome data}

(18)

59

was checked and trimmed with a cutoff of > 30 nt, and the RNA sequence reads were mapped on the reference genome of B. subtilis 168. The JBrowse Genome Browser was

used to visualize the reads 145_{. The Reads Per Kilobase per Million reads (RPKM) were}

subsequently normalized and analyzed using the Transcriptome analysis webserver for RNA-seq expression data T-rex 146_{. To detect differentially expressed genes (DEGs), a}

fold change > 3 or < -3 was used, with an adjusted P-value of < 0.01. DEGs with extremely high fold changes due to very low expression (RPKM <10) were not investigated further.

Authors’ contributions

MHvdE designed and performed experiments, and wrote the chapter. MM and MHvdE created strain ∆tagO. JW helped planning and executing the experiments. OPK guided

the research, helped with interpretation of the results and revised the text.

Supplementary Information

Table S1. WTA_RNAseq_Final. A table with all DEGs can be downloaded at: http://www.molgenrug. nl/index.php/bacillus/marielle-van-den-esker

Table S2. DeletedRegion_Final. This table with an overview of the genes located in the deleted region can be downloaded at: http://www.molgenrug.nl/index.php/bacillus/marielle-van-den-esker

Figure S1. PCR performed on genomic DNA of the wild-type (WT) and ∆tagO mutant using primers amplifying yorL (left) and ysbA (right). Genomic DNA was amplified from cultures grown in the presence (+) and absence of 1 mM IPTG. As a reference, a 1 kb ladder (Fermentas) was used.

(19)