Development and optimization of an EGFP-based reporter for measuring the general stress response in Listeria monocytogenes

Abstract

A characteristic of the food-borne pathogen Listeria monocytogenes is its tolerance to the harsh conditions found both in minimally processed foods and the human gastrointestinal tract. This trait is partly under the control of the alternative sigma factor sigma B (σ^B). To study the mechanisms that trigger the activation of σ^B, and hence the development of stress tolerance, we have developed a fluorescent reporter fusion that allows the real-time activity of σ^B to be monitored. The reporter, designated P_lmo2230::egfp, fuses the strong σ^B-dependent promoter from the lmo2230 gene (which encodes a putative arsenate reductase) to a gene encoding enhanced green fluorescence protein (EGFP). The reporter was integrated into the genomes of the wild-type strain L. monocytogenes EGD-e as well as two mutant derivatives lacking either sigB or rsbV. The resulting strains were used to study σ^B activation in response to growth phase and hyperosmotic stress. The wild-type was strongly fluorescent in stationary phase or in cultures with added NaCl and this fluorescence was abolished in both the sigB and rsbV backgrounds, consistent with the σ^B-dependency of the lmo2230 promoter. During sudden osmotic upshock (addition of 0.5 M NaCl during growth) a real-time increase in fluorescence was observed microscopically, reaching maximal activation after 30 min. Flow cytometry was used to study the activation of σ^B at a population level by hyperosmotic stress during exponential growth. A strong and proportional increase in fluorescence was observed as the salt concentration increased from 0 to 0.9 M NaCl. Interestingly, there was considerable heterogeneity within the population and a significant proportion of cells failed to induce a high level of fluorescence, suggesting that σ^B activation occurs stochastically in response to hyperosmotic stress. Thus the P_lmo2230::egfp is a powerful tool that will allow the stress response to be better studied in this important human pathogen.

Introduction

The gram positive bacterium Listeria monocytogenes is a facultative intracellular pathogen that is found widely in the environment (reviewed by Barbuddhe, 2009).¹ As a foodborne pathogen L. monocytogenes is able to survive and multiply in food during processing and/or storage. Key virulence factors are controlled by the positive regulatory factor A (PrfA) and also by the alternative RNA polymerase sigma factor SigB (σ^B).^2-4 σ^B is involved in the initiation of transcription by recognizing σ^B promoter sequences and allowing the expression of a specific set of genes (σ^B regulon) when the appropriate conditions are encountered. A variety of stress and virulence-related phenotypes are associated with loss of the sigB gene, indicating that σ^B plays an important role both during infections and under stressful conditions (reviewed by O’Byrne, 2008).⁵ After the full genome sequence of L. monocytogenes became available⁶ a number of studies identified components of the σ^B regulon by looking at genes differentially expressed in L. monocytogenes wild-type and corresponding sigB mutants using proteomic approaches^7,8 or gene microarray technology.^9-12

The mechanism responsible for sensing stress that leads to the activation of σ^B has not been fully characterized in L. monocytogenes. However, the signal transduction system controlling σ^B activity in Bacillus subtilis, has been well studied (reviewed by Hecker, 2007).¹³ A high level of conservation between the sigB operons in B. subtilis and L. monocytogenes suggests a shared mechanism of post-translational regulation of σ^B activity. In B. subtilis this mechanism involves partner switching of an anti sigma factor called RsbW between σ^B and anti-anti sigma factor called RsbV; when RsbW is sequestered by RsbV, σ^B is free to participate in transcription. In order to test this model and to develop an understanding of how σ^B is controlled in L. monocytogenes, molecular tools for measuring the activity status of σ^B in this pathogen are required. Therefore in the present study we aimed to develop an effective reporter for measuring σ^B activity in L. monocytogenes, which would then allow important questions to be addressed about the conditions that trigger the activation of σ^B and how this activity is modulated.

A number of transcriptional reporter systems based on bioluminescence and enzymatic assays have been used to study the regulation of transcription and the investigation of promoter activity in L. monocytogenes.^14-16 However, the commonly used lacZ, gus or lux reporters are limited by the need for specific cofactors or exogenous substrates. In addition to these considerations we sought a system that could be detectable with multiple methods and not adversely affected by the measurement techniques or growth conditions. Fluorescent reporter systems containing green fluorescent protein (GFP), first described in Aequorea, jellyfish are extensively utilized in a wide range of other organisms including bacteria.¹⁷ As measurement of GFP activity in prokaryotes was reported to be affected reversibly by protein oxidation, the pH value of the medium and temperature,¹⁸ several chromophore variants of wild-type GFP (wtGFP) have been developed for use as a reporters under a wide range of conditions.¹⁹ For a reporter of σ^B activity the enhanced green fluorescent protein (EGFP) variant was selected, since it demonstrates 35-fold higher fluorescence and is more stable in terms of irreversible quenching and photo-bleaching, making the detection threshold of EGFP much lower than wtGFP.^20,21 EGFP has a single, strong red shifted excitation peak at 488 nm which corresponds to the line of FITC optics and line of argon ion lasers used in many flow cytometers. In addition, elimination of the UV excitation enables utilization of EGFP in living cells under less harmful conditions due to lower energies involved in visualization. EGFP folding and chromophore formation is also faster²² and its structure was reported to be stable over osmolarity and temperature ranges.²³

One of the key factors in considering a successful reporter of σ^B activity in L. monocytogenes is the choice of an effective σ^B-dependent promoter. In the pioneering in vitro studies of sporulation in B. subtilis Haldenwang and Losick (1979)²⁴ used the ctc gene in their transcription assays, which led to the discovery of one of the first bacterial alternative sigma factor, σ^B. Since then ctc has become the first-choice reporter gene for σ^B activity studies in B. subtilis, where it has been used to investigate a whole range of different stress stimuli.^25-27 The ctc gene, described as being positively regulated by σ^B in L. monocytogenes,¹¹ was also used in a transcriptomic approach to investigate blue and red light activation of σ^B in L. monocytogenes.²⁸ The study demonstrated that in a ΔsigB mutant there was still substantial ctc transcript levels, suggesting that in L. monocytogenes ctc transcription is not fully dependent on σ^B. The sigB gene itself is known to be autoregulated^3,29 and has previously been used in L. monocytogenes as a reporter for the role of σ^B in biofilm formation.³⁰ However, the proposed regulation of transcription of eight-gene sigB operon based on B. subtilis and a transcriptomic study of L. monocytogenes suggest that sigB might also be transcribed in σ^B-independent manner from the σ^A promoter located upstream from rsbR and yielding the transcript for the whole operon including rsbRSTU followed by rsbV, rsbW, sigB and rsbX²⁹ or within a 10 gene operon starting upstream from lmo0887.¹²

Our recently published work evaluated transcript levels of sigB and other σ^B-dependent genes as a reporter of σ^B activity.³¹ Out of four σ^B-dependent genes investigated (sigB, opuCA, lmo2085, lmo2230), expression of lmo2230, which encodes a putative arsenate reductase, was strongly σ^B-dependent and demonstrated the highest level of induction following osmotic stress. Selected studies on the σ^B regulon in various strains of L. monocytogenes over a range of different stress conditions^9-11,32-35 have confirmed the high induction and strong σ^B-dependence of lmo2230 expression. These studies also show that the σ^B-dependent activation of lmo2230 transcription occurs in different strains and in response to different stresses. The σ^B promoter sequence of lmo2230 was predicted with Hidden Markov model (HMM)-based searches and its transcriptional start site was confirmed by RACE-PCR.¹¹ lmo2230 (designated as arsC) has been already successfully used among other σ^B-regulated genes as a reporter of σ^B activity after exposure to light.²⁸ However, this valuable and quantitative study used expensive and time consuming RNA extractions followed by RT-PCR assays. Here we describe the construction of a reporter fusion utilizing the promoter of lmo2230 to monitor σ^B-dependent expression that will be quicker and more effective for testing over a wide range of conditions.

Materials and Methods

Bacterial strains, plasmids and growth conditions

The bacterial strains and plasmids used in this study are listed in Table 1. L. monocytogenes strains were grown in Brain Heart Infusion (BHI) broth or agar (LabM) at 37°C unless otherwise stated. Cells were grown under continuous shaking from a starting OD₆₀₀ = 0.05 to OD₆₀₀ = 0.6 for exponential phase and up to ~16 h for stationary phase experiments. The volume of the medium occupied no more than 10% of the flask volume to ensure sufficient aeration. Escherichia coli strains used as intermediate vector hosts were grown in Luria Bertani (LB) broth or agar (LabM) at 30°C. Antibiotics were incorporated into the media as follows: Chloramphenicol (Cml) 10 μg ml⁻¹ for L. monocytogenes; Ampicillin (Amp) 100 μg/ml for E. coli.³⁶

Table 1. Plasmids and strains used in this study.

	Source or reference
PLASMIDS
pKSV7 (pUC18 and pBD95 integrated shuttle vector for E. coli and L. monocytogenes carrying a temperature sensitive oriC from pE194)	36
pCR®-XL-TOPO®	Invitrogen™
pEC04 (pCR®-XL-TOPO® containing the ΔrsbV deletion cassette)	This study
pEC06 (pKSV7 containing ΔrsbV DNA deletion cassette)	This study
pBluescriptII-Plmo2230::egfp	Eurofins MWG Operon
pKSV7-Plmo2230::egfp	This study

STRAINS
	Source or reference	Collection number
E. coli DH5α/pKSV7	N. Freitag	COB082
One Shot® TOP10 Electrocomp™ E. coli	Invitrogen™	COB267
E. coli TOP10/pBluescriptII-Plmo2230::egfp	This study	COB515
E. coli TOP10/pKSV7-Plmo2230::egfp	This study	COB516
L. monocytogenes EGD-e	K. Boor	COB261
L. monocytogenes EGD-e ΔsigB	K. Boor	COB262
L. monocytogenes EGD-e/pEC06	This study	COB392
L. monocytogenes EGD-e::pEC06	This study	COB393
L. monocytogenes EGD-e ΔrsbV	This study	COB411
L. monocytogenes EGD-e/pKSV7	This study	COB517
L. monocytogenes EGD-e/pKSV7-Plmo2230::egfp	This study	COB476
L. monocytogenes EGD-e ΔsigB/pKSV7-Plmo2230::egfp	This study	COB477
L. monocytogenes EGD-e ΔrsbV/ pKSV7-Plmo2230::egfp	This study	COB498
L. monocytogenes EGD-e::pKSV7-Plmo2230::egfp	This study	COB491
L. monocytogenes EGD-e ΔsigB::pKSV7-Plmo2230::egfp	This study	COB495
L. monocytogenes EGD-e ΔrsbV::pKSV7-Plmo2230::egfp	This study	COB501

Construction of a L. monocytogenes EGD-e ΔrsbV mutant

L. monocytogenes EGD-e ΔrsbV mutant was constructed using the splicing by overlap extension (SOEing) PCR technique³⁷ followed by the construction of a shuttle vector containing the ΔrsbV deletion cassette. Primers A, B, C and D (COB385–388) were designed (based on EGD-e sequence published on http://genolist.pasteur.fr/ListiList) to amplify two fragments (AB and CD) of the gene to be deleted (Table 2). Primers A and D were designed to carry an EcoRI restriction endonuclease site at the 5′ end of each primer. Primer C was designed to have its 5′ region complementary to the 3′ region of primer B. The intermediate vector pCR-XL-TOPO® (Invitrogen) was used to generate pEC04 carrying ΔrsbV deletion cassette prepared by SOEing. The ΔrsbV deletion cassette was subcloned into the pKSV7 shuttle vector yielding pEC06. The integration of pEC06 into the L. monocytogenes chromosome was achieved by homologous recombination at 42°C (non permissive for pKSV7 replication) and finally the excision of the shuttle vector containing the wild type gene was obtained. Colonies were then screened for loss of the rsbV gene using PCR using primers COB470 and COB471 (Table 2) to confirm the elimination of rsbV. Primers COB330 and COB331 were used to verify that a fully intact neighbor rsbW gene existed in the ΔrsbV mutant and primers COB332 and COB333 provided similar finding for the sigB gene. Primers COB328 and COB329 were used to verify the rsbV deletion (312bp of 345bp).

Table 2. Primers used in this study.

Name	Sequence (5′ to 3′)	Anneal (°C)	Collection number
M13-F	CAGGAAACAGCTATGAC	50	COB162
M13-R	GTAAAACGACGGCCAG	50	COB163
rsbV-CACC-F	CACCATGAATATTAGTATAGAAATAAAAG	55	COB328
rsbV-his6-R	TCAATGATGATGATGATGATGTGCATTGTTGCCATTCATTTC	55	COB329
rsbW-CACC-F	CACCATGGCAACAATGCATGACAAAATTAC	55	COB330
rsbW-His6-R	TCAGTGGTGGTGGATGATGATGGGTTGAGATACTTTTGGC	55	COB331
sigB-CACC-F	CACCATGCCAAAAGTATCTCAACCTG	55	COB332
sigB-His6-R	TTAATGGTGATGGTGATGGTGCTCCACTTCCTCATTCTG	55	COB333
rsbV SOE A	GAATTCGAAGGCTTACCACTGGG	58	COB385
rsbV SOE B	CATCACTTCACCCCCATCTAATTTTT	58	COB386
rsbV SOE C	GGGGGTGAAGTGATGGTAGAGGGTGAAATGAATGGC	58	COB387
rsbV SOE D	GAATTCAGGTTGAGATACTTTTGGC	58	COB388
PBrsbV-His-F	CCGAATTCCTTTGAGTTGCATGATGACTTTACGC	60	COB470
PBrsbV-His-R	GCGAATTCTCAATGATGATGATGATGAGTGCATTGTTGCC	60	COB471
lmo2229-F	GCCTTGTCGCCATCTTTG	55	COB688
egfp-R	GGCCGTTTACATCTCCATC	55	COB689

Design and construction of P_lmo2230::egfp gene fusion strains

The reporter plasmid pKSV7-P_lmo2230::egfp (Fig. 1A) was designed to contain a fusion of the σ^B-dependent promoter region of lmo2230 (located upstream of the start codon based on the full genome sequence for L. monocytogenes EGD-e) with a codon-optimised (with GENEius software) egfp (accession AY192024.1), the gene coding for enhanced green fluorescent protein (EGFP).²⁰ First, egfp was synthesized together with 443 bp sequence of the lmo2230 promoter region upstream from its start codon¹¹ flanked by BamHI and EcoRI restriction sites (together this gene fusion construct was designated P_lmo2230::egfp). The gene fusion was synthesized by Eurofins MWG and inserted into pBluescriptII yielding pBluescriptII-P_lmo2230::egfp. The P_lmo2230::egfp fragment including the BamHI and EcoRI restriction sites was amplified by PCR with Ultra High-Fidelity DNA Polymerase (Agilent Technologies) and M13 primers (Table 2) using pBluescriptII-P_lmo2230::egfp as a template. After digestion with BamHI and EcoRI the P_lmo2230::egfp fragment was subcloned into pKSV7 using standard methodologies, generating plasmid pKSV7-P_lmo2230::egfp. This newly constructed plasmid was transformed into electrocompetent L. monocytogenes EGD-e wild type, ΔsigB and ΔrsbV derivative mutant strains as described earlier.³⁸ The transformants were selected on BHI agar plates containing 10 μg ml⁻¹ of chloramphenicol, kept for 48 h at 30°C.

Figure 1. Design of the reporter plasmid pKSV7-P_lmo2230::egfp containing a fusion of the σ^B–dependent promoter region of lmo2230 with egfp (A). Site-directed integration of the reporter fusion vector into the chromosome of L. monocytogenes EGD-e wild type and derivative integrants (B) occurring at homologous site of lmo2230 promoter region (443 bp) after incubation at 42°C (non-permissive for autonomous plasmid maintenance in L. monocytogenes) in the presence of chloramphenicol selection.

Chromosomal integration of the P_lmo2230::egfp reporter fusion

Homologous recombination was utilized to construct recombinant strains of L. monocytogenes EGD-e wild type, ΔsigB and ΔrsbV derivative mutants carrying single copies of the P_lmo2230::egfp reporter fusion integrated into the chromosome (Fig. 1B). The integration was achieved by growing the L. monocytogenes cells, electro-transformed with temperature sensitive pKSV7-P_lmo2230::egfp vector, on chloramphenicol plates, kept for 48 h at 42°C. At a non-permissive temperature the vector was unable to replicate autonomously and integrated into the P_lmo2230 homologous region on the host chromosome. Fast growing colonies were sub-streaked five times onto fresh BHI agar with chloramphenicol, each time for an additional 24 h growth at 42°C. Integration was confirmed using PCR with COB688 and COB689 primers, which bind to defined regions in both the pKSV7-P_lmo2230::egfp plasmid and EGD-e genomic DNA (Table 2).

Microscopic quantification of EGFP fluorescence

To visualize fluorescence due to σ^B-promoter-driven EGFP expression, cells were grown in BHI or BHI supplemented with 0.5 M NaCl under continuous shaking to OD₆₀₀ = 0.6. Then 1 ml of cells was immediately fixed in 1:1 volume of ice cold 1:1 (v/v) methanol/ethanol mixture for 10 min at -20°C maintaining fluorescence level similar to live cells when observed with microscopy. For quantitative fluorescence, 1 ml of fixed cells was diluted 1:10 by mixing with 9 ml of filter sterilised PBS and then cells were concentrated on 0.2 μm pore, 10 mm diameter, black polycarbonate membrane filters (Millipore). For preliminary analyses, 1 ml of bacterial cells was centrifuged, resuspended in 50 μl of sterile PBS and 5 μl was smeared on the slides for visualization. Phase-contrast and fluorescence microscopy was performed using a Nikon Eclipse E600 microscope with a CCD camera attached for digital photography. Expression of EGFP was visualized at a fixed exposure time of 2 sec in a dark room using the B-2A filter covering GFP wavelengths together with 1/8 Neutral Density filter (ND8) for minimizing EGFP photobleaching. Relative fluorescence intensities and detectable cell counts were reported after automated image processing of multiple fields with ImageJ 1.44 software³⁹ with appropriate manipulations as described by others.^40,41 For each experimental condition a minimum of three biological replicates were analyzed and ten randomly selected fields were captured.

Western blot analyses

Proteins were extracted using a sonication-based method³⁸ from 200 ml of stationary cultures grown for 16 h and the concentrations of protein extracts were determined by the RC DC Protein Assay Kit (BioRad). Extracts were normalized to 5 mg ml⁻¹ total protein concentrations and 10 µl of these samples were separated by SDS-PAGE and transferred to a nitrocellulose membrane. Western blot analyses were performed using chicken polyclonal anti-GFP antibody (Abcam) diluted 1:2,000 in 3% w/v skim milk and secondary antibody (HRP conjugated to goat anti-chicken IgY, Promega) diluted 1:20,000 in 3% w/v skim milk. Blots were viewed with a chemiluminescent substrate (SuperSignal® West Pico Chemiluminescent Kit, Pierce) captured in a dark room using a light sensitive film (Amersham Hyperfilm ECL, GE Healthcare).

Flow cytometry (FCM) analysis of the EGFP expressing cells

For FCM analysis of EGFP expression bacterial strains were grown from starting OD₆₀₀ = 0.05 to OD₆₀₀ = 0.6 in BHI or BHI supplemented with 0.3 M, 0.6 M, or 0.9 M NaCl. Cells were fixed in 1:1 volume of ice cold 1:1 (v/v) methanol/ethanol mixture for 10 min at -20°C. One ml of fixed cells was then harvested by centrifugation and resuspended in 0.5 ml of PBS. Quantification of single cell fluorescence was achieved by FCM analysis of 100 μl of the suspension with a BD Accuri C6 flow cytometer (Accuri Cytometers, Inc.) using the following instrument settings: 488 nm blue laser excitation, FL1 533/30 nm (e.g., FITC/GFP) emission channel and fast 66 µL/min flow rate from 96 round bottom plates (Sarstedt) and a minimum of 100,000 events for each sample recorded. The data collected were processed with BD CFlow® software to determine mean fluorescence values and plot side scatter and forward scatter values for two biological and three technical replicates.

Results

A P_lmo2230::egfp reporter fusion generates strongly σ^B-dependent fluorescence

A fusion construct consisting of 443 bp of the lmo2230 promoter region upstream from the egfp gene was synthesized and sub-cloned into pKSV7, generating a reporter plasmid designated pKSV7-P_lmo2230::egfp. This plasmid was then transformed into L. monocytogenes EGD-e (wild-type) as well as ΔsigB and ΔrsbV mutant derivatives of this strain. These transformants were investigated by fluorescent microscopy to determine whether EGFP was expressed in the three strains grown for 16 h in brain heart infusion (BHI) broth at 37°C. EGD-e wild-type cells harboring the pKSV7-P_lmo2230::egfp plasmid were brightly fluorescent, although considerable heterogeneity in the level of EGFP fluorescence could be observed in the population (Fig. 2). Analysis of randomly selected microscopic fields with phase contrast and fluorescence microscopy suggests that some portion of wild-type cells transformed with pKSV7-P_lmo2230::egfp were non-fluorescent or might express EGFP at a level not detectable with this method. In contrast to the wild-type, the ΔsigB mutant strain carrying pKSV7-P_lmo2230::egfp showed no fluorescence, suggesting that σ^B is necessary for activation of egfp expression from the lmo2230 promoter. Furthermore, the ΔrsbV strain harbouring this plasmid was also non-fluorescent, which is consistent with the proposed role of RsbV as an anti-anti sigma factor; in the absence of RsbV σ^B remains permanently sequestered by RsbW and is therefore unavailable to drive transcription from σ^B promoters.

Figure 2. Fluorescence of egfp-containing cells of L. monocytogenes EGD-e wild type, ΔsigB and ΔrsbV mutants after transformation with pKSV7-P_lmo2230::egfp. Phase-contrast and fluorescence microscopy of corresponding fields were performed after 1 ml of stationary phase culture was centrifuged, resuspended in 50 μl of sterile PBS and 5 μl was smeared on the slides for visualization. Images are representative of a minimum of ten randomly selected fields captured for three biological replicates for each strain.

To check if the lack of fluorescence of ΔsigB, and ΔrsbV strains bearing the reporter plasmid was due to a detection limit of fluorescent microscopy we looked at egfp expression at the protein level. Western blotting using an anti-GFP antibody was performed on stationary cell extracts from wild-type, ΔsigB, and ΔrsbV strains. EGFP was found to be expressed in wild-type only and neither full-length EGFP nor any indication of degraded protein bands were observed in the ΔsigB (Fig. 3A), and the ΔrsbV strains (Fig. 3B). These findings confirmed σ^B-dependent expression of EGFP and also indicated an absence of RsbV-independent activation of σ^B under the conditions investigated.

Figure 3. EGFP is expressed of L. monocytogenes EGD-e wild type but not in ΔsigB and ΔrsbV backgrounds. Western blot analyses were performed using protein extracts from stationary phase cultures (normalized to 5 mg ml⁻¹ total protein concentrations and 10 µl of these samples separated by SDS-PAGE) with chicken polyclonal anti-GFP antibody (Abcam) HRP-conjugated goat anti-chicken antibody (Promega) and chemiluminescent detection.

Quantification of lmo2230-promoter-driven expression with fluorescent microscopy

In order to eliminate any artifacts that might be associated with a multi-copy plasmid-based reporter fusion (e.g., variable copy number or effects of antibiotics used for maintenance) the reporter fusion vector (Fig. 1A) was integrated into the genomes of L. monocytogenes EGD-e wild type, ΔsigB and ΔrsbV strains by homologous recombination (Fig. 1B). Microscopic comparison of wild type cells carrying either a chromosomal copy or a plasmid copy of the reporter fusion did not show a distinguishable difference in fluorescence (data not shown). As expected the ΔsigB and ΔrsbV integrants were not fluorescent. Integrant strains bearing single chromosomal copies of egfp could be grown without the need for chloramphenicol selection and consequently were more stable than plasmid bearing transformants, which required selection for plasmid maintenance.

It was possible to quantify the relative fluorescence levels in cultures growing under different experimental conditions using fluorescent microscopy and image analysis (Fig. 4). This method was based on detection of fluorescent cells using algorithms for automatic edge detection of bacteria, background subtraction and counting of the objects (Fig. 4B) and subsequently determining the mean fluorescent intensity of the objects detected (Fig. 4C). To evaluate this approach we examined the wild-type integrant strain in the presence of salt stress (0.5 M NaCl) and in stationary phase, conditions that σ^B is known to be activated by, and under which EGFP structure and its fluorescence are stable.²³ As expected EGFP expression was increased by salt in exponential phase and a higher fluorescence was observed in stationary phase when compared with the exponential phase of growth (Fig. 4). The salt induction in exponential phase could be visually observed in randomly selected microscopic fields (Fig. 4A; n ≥ 30) and was further demonstrated by quantifying the number of fluorescent objects detected (Fig. 4B) and their relative fluorescence (Fig. 4C). However, the activation of σ^B by salt was not found to be statistically significant in stationary phase (Fig. 4C). Cells grown up to stationary phase of growth (~16 h) were more fluorescent than exponentially growing cells regardless of the salt addition. Thus the ability of the P_lmo2230::egfp fusion to act as a reporter of σ^B activity is not adversely affected by genome integration.

Figure 4. σ^B -dependent expression of EGFP is increased in L. monocytogenes EGD-e wild-type integrant in stationary phase and by osmotic stress. Cells of integrants bearing P_lmo2230::egfp fusion were grown from starting OD₆₀₀ = 0.05 up to either exponential (OD₆₀₀ = 0.6) or stationary phase (16 h) in BHI or BHI supplemented with 0.5 M NaCl. A similar number of cells (equivalent to 1 ml of OD₆₀₀ = 0.6 and diluted 1:2 by the fixing procedure) was diluted 1:10 in PBS and concentrated on 10 mm diameter, 0.2 μm pore polycarbonate membrane (Millipore) by filtration. For each experimental condition three biological replicates were analyzed and a minimum ten randomly selected fields of each membrane were captured with a CCD camera attached to Nikon Eclipse E600 with B-2A and ND8 filters used at a fixed exposure time of 2 sec with representative fields shown (A). Fluorescence levels of cells were quantified with automated image processing of microscopic fields by counting numbers of detectable particles (B) and their relative fluorescence intensities (C) with ImageJ 1.44 software with appropriate manipulations as described by others.^39,40

Activation of σ^B is rapid after osmotic upshock

To determine whether our reporter system is able to detect quick changes of σ^B activity reported by earlier studies,^31,32 the relative fluorescence was investigated after osmotic upshock. In contrast to steady-state growth experiments, cells were grown without NaCl supplementation up to OD₆₀₀ = 0.6 then osmotically shocked by addition of solid NaCl (0.5 M) and samples were taken at suitable intervals from vigorously shaken cultures for microscopic observation (Fig. 5A) and fluorescence intensity quantification (Fig. 5B). A statistically significant increase (p < 0.05, Student's t-test) of relative fluorescence between osmotically shocked and untreated culture was observed from 10 min after the upshock (Fig. 5B). The reporter approach demonstrated an induction of σ^B activity with the highest variability up to 20 min after salt addition and then stable fluorescence subsequent to 30 min after treatment. Together these results show that a chromosomally integrated P_lmo2230::egfp reporter fusion can reliably detect a rapid activation of σ^B following osmotic upshock.

Figure 5. σ^B -dependent EGFP expression is induced after osmotic upshock in L. monocytogenes EGD-e wild-type P_lmo2230::egfp integrant. Cells bearing P_lmo2230::egfp fusion grown in BHI from OD₆₀₀ = 0.05 up to OD₆₀₀ = 0.6 when 0.5 M NaCl was added. Similar number of cells (equivalent to 1 ml of OD₆₀₀ = 0.6 and diluted 1:2 by the fixing procedure) was taken at appropriate intervals from untreated and osmotically shocked cultures. Fixed cells were diluted 1:10 in PBS and concentrated on 10 mm diameter, 0.2 μm pore polycarbonate membrane (Millipore) by filtration. For each experimental condition three biological replicates were analyzed and a minimum of ten randomly selected fields were captured with a CCD camera attached to Nikon Eclipse E600 with B-2A and ND8 filters were used at a fixed exposure time of 2 s with representative fields shown (A). Fluorescence levels of cells were quantified with automated image processing of captured microscopic fields by counting relative fluorescence intensities (B) with ImageJ 1.44 software with appropriate manipulations as described by others.^39,40

Flow cytometry reveals heterogeneous activation of σ^B in the population

Digital image processing is cheap and straight forward but also time-consuming especially in terms of the investigation of multiple experimental conditions or for determining the proportion of the population that is fluorescent. For a more automated analysis of cells expressing EGFP flow cytometry (FCM) was utilized and evaluated in cultures with increasing concentrations of salt. FCM allows individual cells to be analyzed whereas other methods give average data for the whole population. When wild-type cells without a reporter fusion were analyzed by FCM the population was found to have a low level of intrinsic fluorescence (autofluorescence). This “background” fluorescence level was used to define the gate range used to detect EGFP expressing cells. When the wild-type P_lmo2230::egfp fusion strain was analyzed, following aerobic growth to exponential phase (OD₆₀₀ = 0.6) at 37°C, 27.8% of the particles detected had fluorescence levels that fell within the defined gate (Fig. 6). When a FCM analysis was performed on a culture of the ΔsigB mutant grown under the same conditions, no significant fluorescence was detected above the background level, indicating that σ^B was responsible for any EGFP expression detected in the wild type. When cultures were grown under conditions of hyperosmotic stress (NaCl concentrations of 0.3, 0.6 or 0.9 M) the wild-type reporter strain showed a clear increase in the proportion of cells that were expressing EGFP, while no EGFP expression was detected in the ΔsigB mutant (Fig. 6). The proportion of cells expressing EGFP more than doubled as the NaCl concentration increased from 0 to 0.9 M, reaching 62.2% of the population at 0.9 M NaCl. The data presented are in line with our previous transcriptomic reports of σ^B activity being proportional to the extent of NaCl stress.³¹

Figure 6. Heterogeneity of fluorescence within EGFP-expressing population and σ^B activation proportional to the extent of osmotic stress was revealed with FCM. Cells of parent wild-type strain (WT), wild-type-P_lmo2230::egfp (WT-egfp) and ΔsigB-P_lmo2230::egfp (Δσ^B-egfp) derivative strains were grown from a starting OD₆₀₀ = 0.05 up to OD₆₀₀ = 0.6 in BHI (no salt) or BHI supplemented with 0.3 M, 0.6 M or 0.9 M NaCl. FCM was performed using BD Accuri C6 flow cytometer on fixed cells (concentration equivalent to 1 ml of OD₆₀₀ = 0.6, fixed and resuspended in sterile PBS) for three technical replicates of two biological replicates for each strain and each experimental condition. Levels of intrinsic fluorescence (autofluorescence) were determined for wild-type cells without a reporter fusion (WT) and were used to define the gate range of EGFP expressing wild-type-egfp strain population. Mean fluorescence intensities (MFI) for egfp gate and particles located outside the gate (autofluorescence) were calculated by BD CFlow^® software.

The heterogeneity of fluorescence within the wild-type EGFP expressing population was significant; with ~38% of the population exposed to 0.9 M NaCl failing to induce fluorescence levels that fell within the predetermined gate (Fig. 6A). The gate was determined by defining the limit of autofluorescence of wild-type L. monocytogenes. However it was clear that there was an increase in the level of autofluorescence when cells were exposed to 0.9 M NaCl and that increase was σ^B-dependent (Fig. 6). This result indicates that there is some overlap between cellular autofluorescence and EGFP fluorescence and also shows that the range of σ^B activities within the population is considerable. The heterogeneity of σ^B activities within the population exposed to hyperosmotic stress raises new and interesting questions about how individual cells sense and respond to osmotic stress.

Discussion

In the present study, we developed a rapid EGFP-based reporter assay for measuring σ^B activity in L. monocytogenes. The system was carefully designed based on available literature to have the strong σ^B-dependent promoter of lmo2230, which ensured that cells were highly fluorescent following the activation of σ^B. The reporter approach showed increased σ^B activity in stationary phase and during hyperosmotic stress, which was applied either continuously or suddenly. Furthermore, an analysis by flow cytometry revealed that σ^B activation was heterogeneous in the population and proportional to the extent of stress. The σ^B-dependency of the reporter fusion was confirmed since strains lacking either the sigB or rsbV genes failed to induce EGFP expression under any condition tested. Together the data presented here demonstrate that the P_lmo2230::egfp reporter is a useful tool that can be used to investigate the conditions and mechanisms that trigger the activation of σ^B in L. monocytogenes.

Expression of lmo2230 was shown to be rapid and transient both after heat shock at 48°C, with a 10-fold induction observed after 3 min of heat stress³² and also 15 min after osmotic upshock when more than 160-fold higher levels of lmo2230 in wild-type were reported comparing to unstressed cells at time zero.³¹ Taken together these findings give an opportunity for monitoring activation of σ^B in real time by following lmo2230-promoter-driven expression after a sudden change of environmental conditions. However, there are also some limitations in terms of reporting transient activation of σ^B with the EGFP-based reporter system caused by the high stability of the fluorescent protein under most of the experimental conditions. The long half-life of EGFP (estimated at greater than 24 h)⁴² makes it impossible to observe a drop in σ^B activity after the removal of stress, while this decrease can be demonstrated by monitoring the transcript levels of lmo2230 and other σ^B-regulated genes under similar conditions³¹ and after heat shock.³² Future versions of this reporter could include less stable GFP variants (with ssrA RNA tags recognized by housekeeping proteases)⁴² in order to overcome this limitation. When the time scale for σ^B activation was compared between the RNA based approach described earlier³¹ and the P_lmo2230::egfp reporter (Fig. 5B), the maximal σ^B activation was evident approximately 15 min earlier when monitoring mRNA levels. It seems likely that this lag observed with microscopy is due to the time required for EGFP translation, protein folding and the chromophore maturation process.^19,43 Overall, despite limitations related to the stability and maturation of EGFP, this reporter system should prove very useful for applications that require knowledge about the induction of σ^B activity in response to stress.

Another unique advantage of the P_lmo2230::egfp reporter is that σ^B activation can be studied both at the single cell and the population levels, using microscopy and flow cytometry. Using these approaches heterogeneity of the population and stochastic cell-to-cell variability of gene expression in L. monocytogenes can be studied in response to strictly defined stimuli. The flow cytometry data presented here indicate that cells from the same vigorously shaken liquid culture exhibit significant differences in fluorescence intensities (Fig. 6). This result indicates that σ^B activation does not occur uniformly in the population and further suggests that not all cells may perceive osmotic stress in the same way. Phenotypic variations in clonal populations have been investigated in other bacteria and partly explained by fluctuations in the amount of cellular components together with the noise and the asynchrony in gene expression.⁴⁴ It will be interesting in future studies to determine whether cell-cell differences in σ^B activation influences the survival of individual cells in foods or within the gastrointestinal tract.⁴⁵

In an ΔrsbV background no EGFP expression was detected from the P_lmo2230::egfp reporter (Figs. 2 and 3B), consistent with the current model for σ^B activation, which proposes that RsbV acts as an anti-anti-sigma factor that facilitates σ^B activation when it interacts with the anti-sigma factor RsbW.¹³ Indeed there is genetic data from L. monocytogenes 10403S to support this model; rsbV and sigB mutants are found to have essentially the same phenotype, suggesting that RsbV is required for σ^B activation.⁴⁶ This highlights the potential use of the P_lmo2230::egfp reporter in conducting studies on the regulation of σ^B. A key question that remains to be answered concerns the nature of the stress sensing mechanism. In B. subtilis a high molecular weight (~2MDa) sensory organelle called a stressosome is involved in sensing although the mechanism has not yet been elucidated. Stressosomes are also likely to exist in L. monocytogenes, since homologs of all of the B. subtilis stressosome proteins are present in the genome, but their existence has not yet been established experimentally. By making targeted mutations in the predicted components of the stressosome it should be possible to use the P_lmo2230::egfp reporter to determine which components are required for stress sensing and σ^B activation.

In summary, the P_lmo2230::egfp fusion was developed and shown to be a reliable indicator of σ^B activity in L. monocytogenes, both during different phases of growth and under conditions of hyperosmotic stress. The tight σ^B-dependence of this promoter combined with the simplicity of detecting GFP fluorescence means that this reporter can be used for many applications in the study of the stress responses of this important food-borne pathogen.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.