kdpE and a putative RsbQ homologue contribute to growth of Listeria monocytogenes at high osmolarity and low temperature

Abstract

The kdp locus of Listeria monocytogenes encodes products with homology to structural proteins of a high-affinity potassium uptake system and to a two-component signal transduction system commonly involved in controlling gene expression. We have investigated the role of kdpE, encoding the transcriptional response regulator, as well as of the downstream gene, orfX, in adaptation of L. monocytogenes EGD to NaCl and low temperature. When grown in chemically defined medium the addition of NaCl to 2% decreased the growth rate of a mutant with an insertional inactivated kdpE, while mutants carrying in-frame deletions of either kdpE or orfX were unaffected by high osmolarity. Transcriptional analysis of kdpE and orfX revealed that their products are encoded by the same transcript. Thus, our data indicate that the absence of both KdpE and OrfX influences growth under osmotic pressure. Interestingly, OrfX contains a conserved domain of α/β-hydrolases and resembles RsbQ that in Bacillus subtilis participates in the activation cascade of the general stress sigma factor SigB. When shifted to low temperature the deletion mutant lacking orfX resumed growth slightly faster than the wild-type. This phenotype was shared by a mutant carrying an in-frame deletion of sigB supporting the notion that OrfX could be a RsbQ homologue.

Listeria monocytogenes, Osmotic stress, Low temperature, kdpE, sigB, rsbQ

1 Introduction

Since ancient times NaCl has been used to preserve food due to growth inhibition of microorganisms. However, some microorganisms including pathogenic bacteria have the ability to grow at high osmolarity. One such bacterium is the food-borne pathogen Listeria monocytogenes, which is able to grow in the presence of 10% NaCl in complex medium [10]. In addition, growth of L. monocytogenes has been observed at temperatures as low as 1°C [20] allowing it to survive and multiply at conditions normally found in chilled, lightly preserved food products. The ability of L. monocytogenes to survive both high salt concentrations and low temperature is attributed mainly to the accumulation of compatible solutes such as glycine betaine [9]. The accumulation is achieved by osmotically activated transport from the medium rather than by de novo synthesis [9]. Several different systems responsible for transport of glycine betaine and carnitine have been found in L. monocytogenes and in some cases expression has been suggested to be controlled by the alternative sigma factor σ^B (for review see [18]).

When bacteria adapt to osmotic stress an initial uptake of potassium prevents plasmolysis and leads to restoration of the turgor pressure [5]. In Escherichia coli the initial rapid uptake of potassium is provided by the high-affinity potassium uptake system encoded by the kdpABC operon. During change to high-osmolarity conditions, KdpE activates transcription of the kdpABC operon resulting in increased expression and thus uptake of potassium [15,19]. KdpE is the regulator of a two-component sensor–regulator pair that is phosphorylated by its cognate sensor histidine kinase, KdpD, during high osmolarity or when the potassium concentration becomes low. Thus, in E. coli, expression of the high-affinity potassium uptake system is dependent on KdpE. The signals determining the kinase activity of the sensor, KdpD, have been suggested to be the intracellular K⁺ concentration, cell turgor, rate of transmembrane K⁺ flux or membrane stretch (for review see [18]).

In L. monocytogenes LO28 the kdpE gene was identified as a putative two-component response regulator and an insertion mutant of kdpE showed reduced growth rate at high osmolarity, and impaired virulence in a mouse model [8]. In this report we have analysed the kdp locus of L. monocytogenes EGD. In-frame, non-polar deletion mutants were constructed in kdpE and in a downstream gene, orfX, and growth was followed at high osmolarity and at low temperature. Furthermore, a potential role of OrfX in the activation of the alternative sigma factor sigB was investigated by construction of an in-frame sigB deletion mutant. Our data suggest that OrfX might be a RsbQ homologue of L. monocytogenes.

2 Materials and methods

2.1 Bacterial strains, media, and transformation

L. monocytogenes strains were propagated in brain heart infusion media (BHI, Difco), while E. coli DH5α was grown in Luria–Bertani broth (Difco) [17]. Erythromycin was used at final concentrations of 5 µg ml⁻¹ for L. monocytogenes and 150 µg ml⁻¹ for E. coli. L. monocytogenes EGD and E. coli DH5α were transformed by electroporation [12,17]. A chemically defined minimal medium (Improved Minimal Medium, IMM) containing 1% glucose was used [13]. For low-potassium experiments IMM was modified by substitution of the buffer component KH₂PO₄ with Na₂PO₄ and addition of KCl to a final concentration of 0.048 mM potassium.

2.2 Construction of mutants

For construction of non-polar deletion mutants in the kdpE, orfX and sigB genes, L. monocytogenes EGD chromosomal DNA was used as template for amplification of DNA fragments containing either the 5′ end of the gene and upstream sequences or the 3′ end of the gene and downstream sequences. For the kdpE gene primers SOEA and SOEB2 (212 bp) and primers SOEC2 and SOED (195 bp) were used, for orfX primers PXA and PXB (381 bp) and primers PXC and PXD (392 bp), while primers SIA and SIB (359 bp) and primers SIC and SID (330 bp) were used for the sigB gene. In a second round of polymerase chain reaction (PCR) the kdpE, orfX or sigB fragments were joined by using the splicing by overlap extension PCR method [7], creating PCR fragments each containing an in-frame deletion of 630 bp, 753 bp and 738 bp in kdpE, orfX and sigB, respectively. Subsequently, the ΔkdpE and ΔorfX PCR fragments were digested with Hin dIII and Xba I and cloned into Hin dIII–Xba I-digested pAUL-A, while the ΔsigB fragment was digested with Eco RI and Xba I and cloned into Eco RI–Xba I-digested pAUL-A. After transformation of E. coli DH5α, plasmids pLB157 (ΔkdpE), pLB180 (ΔsigB) and pLB183 (ΔorfX) were isolated (Table 1). Due to the temperature-sensitive origin of replication of pAUL-A, chromosomal integration in kdpE, orfX, or sigB could be selected for by incubation of L. monocytogenes EGD containing either pLB153, pLB180, or pLB183 at 42°C in the presence of erythromycin. This was followed by growth at 30°C in the absence of erythromycin, thus allowing allelic exchange between the intact genes and the deleted genes to take place. Finally, PCR on erythromycin-sensitive colonies verified that the deletions were transferred to the chromosome of L. monocytogenes EGD in strains LB1168 (ΔkdpE), LB1171 (ΔsigB), and LB1174 (ΔorfX).

1

Bacterial strains and plasmids used in this study

Strain	Genotype	Reference
E. coli
DH5α	φ80lacZΔM15 D(lacZYA-argF) U169 recA1 endA1 hsdR17 supE44 thi-1 gyrA96 relA1	Laboratory strain
L. monocytogenes
EGD	Wild-type	W. Goebels
LB1168	EGD ΔorfX	This study
LB1171	EGD ΔsigB	This study
LB1174	EGD ΔkdpE	This study
LB1180	EGD kdpE::pAUL-A	This study
Plasmids
pAUL-A	Shuttle vector, erm, orits	[4]
pBK18–1	pAUL-A::PCR internal kdpE	[8]
pLB157	pAUL-A::SOEing PCR ΔkdpE	This study
pLB180	pAUL-A::SOEing PCR ΔsigB	This study
pLB183	pAUL-A::SOEing PCR ΔorfX	This study

An insertion mutant in kdpE was constructed by repeated streaking of L. monocytogenes EGD carrying pBK18–1 (a pAUL-A derivative that carry a 280-bp internal kdpE fragment) [8] on BHI plates containing erythromycin at 42°C. Chromosomal integration was verified in the resulting strain LB1180 by PCR.

2.3 Growth

Cultures were incubated with shaking and growth was followed by measurement of OD₆₀₀. For growth experiments at high osmolarity strains of L. monocytogenes were grown exponentially in BHI at 37°C to an OD₆₀₀ of 0.2, cells were harvested and the pellets were resuspended in either BHI or BHI supplemented with 7% NaCl. For experiments using minimal medium (IMM) strains of L. monocytogenes were grown exponentially in IMM containing 1% glucose at 37°C to an OD₆₀₀ of 0.2, cells were harvested and the pellets were resuspended either in standard IMM medium supplemented with 2% NaCl or in IMM with low potassium content supplemented with 2% NaCl. For growth experiments at low temperature strains of L. monocytogenes were grown exponentially in BHI to an OD₆₀₀ of 0.2, cells were harvested and the pellets were resuspended in BHI medium that was pre-cooled to either 3 or 5°C. Cultures were incubated at either 3 or 5°C and growth was followed by daily measurement of OD₆₀₀.

2.4 RNA techniques

For purification of RNA, L. monocytogenes EGD was grown in BHI to OD₆₀₀= 0.6. Total RNA was isolated using a hot acid phenol procedure [14]. Purified RNA was treated with RNase-free DNase I (Amersham) according to the manufacturer's procedure. For cDNA synthesis, 0.1 pmol PXB primer (Table 2) was allowed to anneal to 0.5 µg RNA in 1×AMW buffer (Finnzymes) and 1.1 pmol dNTP in a total volume of 10 µl. To initiate cDNA synthesis, 2 U of AMW reverse transcriptase (Finnzymes) was added and the reaction was allowed to proceed for 30 min at 42°C. 0.25 µl of the cDNA reaction was used as template in 25-µl PCR amplification reactions using forward primers SOEC2, PXA or PEI (Table 2), reverse primer PXB and Taq DNA polymerase (Promega) as described by the manufacturer. In control reactions, RNA or chromosomal DNA was used as template.

2

Primers used in this study

Primer	Sequence (5′ to 3′)
SOEA	5′-GGGGGTCTAGAGAATTCCGGAAAATCGTTT-3′
SOEB2	5′-CATTTTCCATCTCCTC-3′
SOEC2	5′-GAGGAGATGGAAAATGCCGGCAGAACCTGCATAC-3′
SOED	5′-GGGGGAAGCTTTAACAGGAGCAAAGCCATGAA-3′
PXA	5′-GGGGGTCTAGACGCCAAATGACCATATTATTCG-3′
PXB	5′-AACGTCTACGCTACG-3′
PXC	5′-CGTAGCGTAGACGTTGACTTCTTATTCCGCGT-3′
PXD	5′-GGGGGAAGCTTCCTAAGTGTTGGCTACAG-3′
SIA	5′-GGGGGTCTAGAAGCCGTAAGTGAAGCC-3′
SIB	5′-CGCCTCTTTATCAGG-3′
SIC	5′-CCTGATAAAGAGGCGGTGTCTAGAATCCAA-3′
SID	5′-GGGGGGAATTCTAATAGCTATCGCAGC-3′
PEI	5′-CGGCGAGCGATTACT-3′

Underlined sequence: Xba I restriction enzyme site.

Bold sequence: sequence complementary to SOEB2.

Underlined sequence: Hin dIII restriction enzyme site.

Bold sequence: sequence complementary to PXB.

Bold sequence: sequence complementary to SIB.

Underlined sequence: Eco RI restriction enzyme site.

3 Results

3.1 Effects of kdpE insertion and deletion mutations in L. monocytogenes EGD on growth at high salt concentrations

In a previous study of L. monocytogenes LO28, a two-component response regulator, KdpE, was implicated in salt tolerance and in virulence [8]. Since growth at high salt concentrations plays a major role in the survival of L. monocytogenes in food products we wished to extend these findings. In order to capitalise on the available genome sequence of L. monocytogenes EGD-e [6],kdpE of L. monocytogenes EGD was disrupted by insertion mutagenesis. To this end an internal fragment of 280 bp inserted into pAUL-A [8] was used to inactivate the kdpE gene of L. monocytogenes EGD. However, when the resulting mutant (LB1180) was challenged with 7% NaCl in BHI, growth was unaffected by the mutation (data not shown). Since KdpE in E. coli has a role in adaptation to high osmolarity during low potassium concentrations (for review see [18]) we investigated growth in IMM, where the potassium concentration can be manipulated. When using a standard concentration of potassium (48 mM) we observed a slightly reduced growth rate for the insertion mutant following addition of 2% NaCl (Fig. 1A). This effect was even more pronounced when the potassium concentration of the IMM medium was reduced to 0.048 mM (Fig. 1B) while in the absence of NaCl the parental strain and the insertion mutant grew identically (data not shown). Subsequently, we constructed a non-polar, in-frame deletion mutation in the kdpE gene. However, growth of the resulting mutant (LB1168) was identical to the parent strain after addition of 2% NaCl irrespective of the potassium concentration in the IMM medium (data not shown).

1

Growth at high osmolarity in chemically defined medium. Exponentially growing cells of L. monocytogenes strains EGD (wild-type) and LB1180 (kdpE insertion mutant) were shifted to either (A) standard IMM medium+2% NaCl or (B) IMM medium containing only 0.048 mM potassium+2% NaCl at t=0. OD₆₀₀ is represented as closed circles for L. monocytogenes EGD (wild-type), while OD₆₀₀ of L. monocytogenes LB1180 (kdpE insertion mutant) is represented as open circles.

3.2 The kdp operon includes orfX located downstream of kdpE

Since the kdpE deletion mutant did not display the same phenotype as the kdpE insertion mutant we analysed the genome sequence for genes located adjacent to kdpE whose expression might be affected by the insertion. Immediately upstream of kdpE we found genes encoding products with homology to KdpA, KdpB, KpdC and KdpD of other bacteria (Table 3). While the genetic organisation (kdpABCDE) of this region is similar to what is found in E. coli, L. monocytogenes encodes a sixth gene (orfX) located just downstream of kdpE. Inspection of the Listeria innocua genomic sequence revealed that it also encodes an OrfX homologue downstream of kdpE although it lacks the C-terminal 69 amino acids compared to OrfX of L. monocytogenes EGD. Analysis of the L. monocytogenes DNA sequence did not reveal any termination signals within the kdp locus suggesting that the five kdp genes together with orfX form a transcriptional unit. We addressed this notion by analysing the transcriptional units covering the kdp locus using RT-PCR analysis. The result presented in Fig. 2 shows that the same mRNA encodes both kdpE and orfX. Furthermore, kdpD and orfX are located on the same transcript (data not shown), suggesting that the three genes form an operon. In contrast, only a very faint RT-PCR product was obtained using primers in orfX and kdpC indicating that these genes may not be transcriptionally linked, while kdpABC were found to be encoded by the same transcript (data not shown).

3

Features of the kdp operon of L. monocytogenes EGD-e

Gene	Protein
	aa	kDa	pI	Homology	Identity	e value
kdpA	561	59.3	9.9	KdpA of Yersinia pestis	55% in 547 aa	e-124
				KdpA of E. coli	53% in 550 aa	e-136
kdpB	683	72.6	4.9	KdpB of Yersinia pestis	65% in 680 aa	0
				KdpB of E. coli	65% in 683 aa	0
kdpC	190	20.8	5.5	KdpC of Ralstonia solanacearum	48% in 193 aa	6e-36
				KdpC of E. coli	35% in 189 aa	4e-24
kdpD	896	100.2	6.1	KdpD of Clostridium acetobutylicum	45% in 897 aa	0
				KdpD of E. coli	36% in 900 aa	0
kdpE	231	26.2	5.1	KdpE of Clostridium acetobutylicum	53% in 232 aa	e-66
				KdpE of E. coli	42% in 222 aa	e-48
orfX	270	30.6	4.6	YugF of Bacillus subtilis	23% in 276 aa	1e-6

2

RT-PCR analysis of transcription of the kdpD, kdpE, and orfX genes. Lane 1: DNA marker (size of fragments: 100, 200, 300, 400, 500 (strong band), 600, 700, 800, 900, 1000, 1500 bp); lane 2: RT-PCR, RNA template, primers PXB and SOEC2; lane 3: PCR, RNA template, primers PXB and SOEC2; lane 4: PCR, DNA template, primers PXB and SOEC2; lane 5: RT-PCR, RNA template, primers PXB and PXA; lane 6: PCR, RNA template, primers PXB and PXA; lane 7: PCR, DNA template, primers PXB and PXA; lane 8: RT-PCR, RNA template, primers PXB and PEI; lane 9: PCR, RNA template, primers PXB and PEI; lane 10: PCR, DNA template, primers PXB and PEI; lane 11: DNA marker.

3.3 Effects of orfX deletion mutation in L. monocytogenes EGD on growth at high salt concentrations and low temperature

To investigate the role of orfX in L. monocytogenes adaptation to increased osmolarity we constructed an in-frame orfX deletion mutation. When we examined growth of the resulting mutant it was unaffected by addition of salt in both complex and minimal media suggesting that the encoded gene product is not involved per se in tolerance to salt (data not shown). In L. monocytogenes growth at high osmolarity is physiologically linked to growth at low temperature, since compatible solutes accumulate under both conditions to compensate for the exclusion of water [9]. Likewise, potassium uptake and control of this may also be important for growth at low temperature. Therefore we compared growth of the wild-type with the kdpE and orfX mutants after lowering the temperature from 37°C to 3°C in complex medium. We found that absence of kdpE and orfX did not affect the growth rate at 3°C (Fig. 3). However, the lag phase of the orfX mutant was slightly decreased compared to the wild-type (Fig. 3). This phenomenon was reproducible and was also observed at 5°C (data not shown).

3

Growth at low temperature in complex medium. Exponentially growing cells of L. monocytogenes strains EGD (wild-type), LB1168 (ΔkdpE), LB1171 (ΔsigB) and LB1174 (ΔorfX) were shifted from 37°C to 3°C. OD₆₀₀ is represented as closed circles for L. monocytogenes EGD (wild-type) (A–C), closed triangles for L. monocytogenes LB1168 (ΔkdpE) (A), open circles for L. monocytogenes LB1174 (ΔorfX) (B) and crosses for L. monocytogenes LB1171 (ΔsigB) (C).

3.4 orfX may encode a putative RsbQ homologue of L. monocytogenes

Since the orfX mutant had a distinct phenotype at low temperature we analysed the amino acid sequence of OrfX and found that the protein contains a conserved domain of the α/β superfamily of predicted hydrolases or acetyltransferases (COG059). While the primary sequence of this protein family is not conserved, they share a common three-dimensional core structure containing the catalytic domain [11]. Recently, it was found that an α/β-hydrolase of B. subtilis designated RsbQ takes part in the activation cascade of the alternative sigma factor σ^B of B. subtilis[3]. Through an interaction with RsbP, RsbQ is required for the energy stress signalling pathway [3], which in turn dephosphorylates the anti-anti-sigma RsbV. Thus, in B. subtilis RsbQ is a positive factor in the activation cascade of the alternative sigma factor σ^B during energy stress. Analysis of the L. monocytogenes genome sequence revealed six members of this family and of these only OrfX shows homology to the α/β-hydrolase YugF of B. subtilis (Table 3), which again is homologous to RsbQ of B. subtilis (26% identity in 257 amino acids). Hence, orfX may encode a putative RsbQ homologue of L. monocytogenes. To address this notion we constructed an in-frame deletion of the sigB gene in L. monocytogenes EGD. Since the only phenotype observed for the orfX mutant was the decreased lag phase at low temperature we investigated the sigB deletion mutant under similar conditions. Interestingly, we found that the sigB mutant also had a decreased lag phase following a shift in temperature from 37°C to 3°C (Fig. 3). Thus, orfX might encode a functional homologue of the B. subtilis RsbQ; however, further experimentation is required to determine the role of OrfX in L. monocytogenes stress tolerance.

4 Discussion

When L. monocytogenes is present in food products its ability to adapt to high salt concentrations and low temperature is important for the bacterium to survive and potentially multiply. Here, we have investigated the importance of the kdp locus in L. monocytogenes EGD for growth under these conditions. The kdp locus comprises genes encoding products with homology to structural proteins of a high-affinity potassium uptake system and a two-component sensor as well as a response regulator likely to be involved in transcriptional regulation. While the composition of this genetic locus resembles that found in other bacteria we observed that L. monocytogenes EGD-e in addition encodes an open reading frame designated orfX located just downstream of kdpE. When we analysed expression by RT-PCR we found that kdpD, kdpE and orfX are encoded by the same transcript suggesting that the function of orfX might be related to that of the kdp locus.

It was shown in a previous study that insertional inactivation of kdpE of L. monocytogenes LO28 affects salt tolerance [8]. However, in L. monocytogenes EGD deletion of either kdpE or orfX did not affect growth compared to wild-type cells when growing in the presence of various concentrations of NaCl and potassium. In contrast, insertional inactivation of kdpE of L. monocytogenes EGD resulted in a decreased growth rate when conditions were changed to 2% NaCl in minimal medium, and it was further reduced when the potassium concentration of the growth medium was lowered. The different phenotypes observed for a kdpE deletion mutant and a kdpE insertion mutant may be explained by a polar effect of the insertional mutation on expression of orfX thus creating a phenotypically kdpEorfX double mutant. Therefore, our results indicate that while that absence of kdpE or orfX, individually, has no effect on adaptation to high osmolarity, expression of both genes is required for optimal growth when cells are shifted to higher concentrations of NaCl; an effect that is dependent on the potassium concentration. Thus, the kdp locus influences the ability of potassium to serve as osmoprotectant for L. monocytogenes EGD.

Analysis of the L. monocytogenes EGD-e genome sequence showed that it also encodes a putative Trk-like potassium uptake system that in E. coli has a low affinity for potassium and is constitutively expressed (for review see [18]). The Trk transport system of L. monocytogenes is most likely responsible for transport of potassium under conditions where the potassium concentration is high. In addition to the Trk-like potassium uptake system we found that the L. monocytogenes genome encodes a putative ion channel subunit (LMO 2059) that may be involved in potassium uptake. The lack of phenotype of the kdpE deletion mutant during growth at high osmolarity supports the hypothesis that other potassium uptake systems are functional in L. monocytogenes EGD. In E. coli a phenotype of a kdp mutant is only observed when several of the other four functional potassium uptake systems are inactivated [16].

OrfX contains a conserved domain of the α/β superfamily of predicted hydrolases or acetyltransferases (COG059). In B. subtilis one member of this family, RsbQ, is involved in modulating the activity of the general stress sigma factor, Sigma B [3]. Among the six members of this protein family encoded by the L. monocytogenes genome OrfX is the most likely candidate for a L. monocytogenes RsbQ homologue. The catalytic activity of RsbQ appears to be required for activation of RsbP during energy stress in B. subtilis[3]. This modification may be necessary for RsbP to receive the energy stress signal and subsequently dephosphorylate the anti-anti-sigma RsbV, which releases σ^B from the anti-sigma, RsbW [3]. Thus, under conditions where RsbQ is responsible for the onset of the activation cascade of σ^B an rsbQ mutant may show the same phenotype as a sigB mutant. The sigB operon of L. monocytogenes encodes proteins highly homologous to RsbV and RsbW of B. subtilis[1,21]. Inspection of the L. monocytogenes genome revealed only two proteins (LMO 0799 and LMO 1699), which, like RsbP of B. subtilis, carry a PAS domain. The function of the PAS domain is currently unknown; however, RsbQ has been suggested to modify this domain [3]. Since LMO 0799 furthermore shows homology to protein kinases, LMO 0799 is the most likely L. monocytogenes RsbP candidate. Thus, it is likely that an activation cascade of σ^B similar to that found in B. subtilis is present in L. monocytogenes. To this end we constructed an in-frame sigB deletion mutant and found that when temperature was shifted from 37 to 3°C both the sigB and the orfX mutants showed a slightly decreased lag phase compared to wild-type cells. This sigB phenotype is in contrast to results published by Becker and co-workers [2] however; this may be due to strain, medium and temperature differences. Most notably, the same phenotype was found for both the orfX and the sigB deletion mutants indicating that lack of OrfX and Sigma B resulted in the same physiological response. Future work might clarify the role of OrfX in adaptation of L. monocytogenes to sub-optimal conditions and the potential involvement in the activation cascade of Sigma B.

Acknowledgements

We sincerely appreciate the expert technical assistance of T. Birk, L. Gertmann, and J. Pedersen. This work was supported by The Danish Research Council and the Centre for Advanced Food Studies.