|Home | About | Journals | Submit | Contact Us | Français|
The regulation of the compatible solute transport systems in Listeria monocytogenes by the stress-inducible sigma factor σB was investigated. Using wild-type strain 10403S and an otherwise isogenic strain carrying an in-frame deletion in sigB, we have examined the role of σB in regulating the ability of cells to utilize betaine and carnitine during growth under conditions of hyperosmotic stress. Cells lacking σB were defective for the utilization of carnitine but retained the ability to utilize betaine as an osmoprotectant. When compatible solute transport studies were performed, the initial rates of uptake of both betaine and carnitine were found to be reduced in the sigB mutant; carnitine transport was almost abolished, whereas betaine transport was reduced to approximately 50% of that of the parent strain. Analysis of the cytoplasmic pools of compatible solutes during balanced growth revealed that both carnitine and betaine steady-state pools were reduced in the sigB mutant. Transcriptional reporter fusions to the opuC (which encodes an ABC carnitine transporter) and betL (which encodes an a secondary betaine transporter) operons were generated by using a promoterless copy of the gus gene from Escherichia coli. Measurement of β-glucuronidase activities directed by opuC-gus and betL-gus revealed that transcription of opuC is largely σB dependent, consistent with the existence of a potential σB consensus promoter motif upstream from opuCA. The transcription of betL was found to be sigB independent. Reverse transcriptase PCR experiments confirmed these data and indicated that the transcription of all three known compatible solute uptake systems (opuC, betL, and gbu), as well as a gene that is predicted to encode a compatible solute transporter subunit (lmo1421) is induced in response to elevated osmolarity. The osmotic induction of opuCA and lmo1421 was found to be strongly σB dependent. Together these observations suggest that σB plays a major role in the regulation of carnitine utilization by L. monocytogenes but is not essential for betaine utilization by this pathogen.
The gram-positive food-borne pathogen Listeria monocytogenes is noted for its ability to grow over a wide range of environmental conditions. In particular, it can grow over a wide range of osmolarities (5), it can adapt effectively to acidic conditions (7, 24), and it can grow at temperatures as low as −0.1°C (5, 28). These properties, combined with the fact that the organism is almost ubiquitous in the environment, make the control of this pathogen problematic in certain foods.
Survival in such harsh environments requires the ability to respond rapidly to changes in the environment, and in bacteria these responses are frequently coordinated at the transcriptional level. Global changes in transcription are often coordinated by specific sigma factors whose levels and activities fluctuate in response to environmental cues. In several gram-negative genera, σs (encoded by rpoS) plays a central role in regulating the transcription of genes required for protection against environmental stresses such as high osmolarity, low pH, and oxidative stress (18, 21). In a number of gram-positive genera (e.g., Staphylococcus and Bacillus), the stress-inducible sigma factor σB plays a similar role in coordinating stress responses at the transcriptional level (14, 16). The gene encoding σB in Listeria monocytogenes (sigB), which was identified based on its homology to the sigB gene from Bacillus subtilis, plays a role in resistance to low-pH stress (30), cryotolerance (3), and oxidative stress resistance and survival of carbon starvation (8). It is also known to play an important role in osmotolerance (2). Reduced osmotolerance in a ΔsigB strain has been attributed to a defect in the ability to transport betaine, an important compatible solute (2). The transcription of the sigB gene itself was found to be strongly dependent on the osmolarity of the medium, with induced transcription detected under conditions of hyperosmotic stress (2).
Several groups have investigated the response of L. monocytogenes to hyperosmotic stress, and it is now clear that at least three distinct transport systems allow the organism to utilize osmoprotectants in the environment (22). There are two betaine transporters: a sodium-dependent secondary transporter, BetL (25), and a substrate binding protein-dependent ABC (ATP binding cassette) transporter, Gbu (19). A single gene, betL, encodes the BetL transporter (25), while Gbu is encoded by three genes, gbuA, gbuB, and gbuC (19). Carnitine is accumulated via another substrate binding protein-dependent ABC transporter, OpuC, encoded by the opuC operon, which consists of four genes, opuCA, opuCB, opuCC, and opuCD (9). The analysis of carnitine pools in a mutant strain lacking OpuC suggests that an alternative, low-affinity mechanism for carnitine uptake also exists (10). The identification of two open reading frames (lmo1421 and lmo1422) in close proximity to the opuC operon and with significant sequence similarity to opuC has led to speculation that this operon may encode the low-affinity carnitine uptake system (10).
Sequence analysis of the DNA sequences upstream from the translation initiation codons of betL and opuCA reveals the presence of putative σB promoter motifs at positions −33 and −58 from the start ATG, respectively (9, 25). No σB consensus promoter sequence motif is evident upstream from the gbuA gene (19). Additional evidence implicating σB in regulating the uptake systems for betaine and carnitine comes from a recent study demonstrating that σB plays a role in cryotolerance in L. monocytogenes (3). In that study the accumulation of the compatible solutes betaine and carnitine at 8°C was found to be impaired in a sigB mutant (2). It now seems likely that compatible solute accumulation plays an important role in allowing L. monocytogenes to grow at low temperatures (3, 19, 20), although the molecular basis for this cryoprotection is not yet clear.
In the present study we have sought to clarify the role of σB in regulating the accumulation of the two principle osmoprotectants of L. monocytogenes, betaine and carnitine. Using an in-frame sigB deletion mutant, we show that the ability to utilize carnitine as an osmoprotectant is σB dependent but that this is not the case for betaine. We show that the transport of carnitine is strongly dependent on the presence of a wild-type sigB gene. Betaine accumulation is also impaired in a sigB background, although significant betaine accumulation occurs independently of σB. Transcriptional reporter fusions to betL and opuCA and reverse transcriptase PCR (RT-PCR) analysis of transcript levels reveal an important role for σB in regulating the transcription of opuCA but indicate that σB is not essential for betL transcription during exponential growth.
The wild-type L. monocytogenes strain 10403S (serotype 1/2a) (4) and its ΔsigB derivative (30) were used throughout. Cells were cultured in defined medium (DM), which has an osmolarity of 260 mOsM (1), with 0.4% (wt/vol) glucose at 30°C and with vigorous shaking. For cloning experiments Escherichia coli OneShot (Invitrogen, Carlsbad, Calif.) was used as a host. Betaine and l-carnitine HCl were supplied by Sigma Chemicals.
Cells were grown overnight in 25 ml of DM (1) supplemented with limiting glucose (0.04%, wt/vol) in a 125-ml flask at 30°C with vigorous shaking. Under these conditions, growth is arrested at an optical density at 600 nm (OD600) of ca. 0.25 as a result of carbon starvation. We have found that the lag phase observed when glucose is added to these overnight cultures is appreciably less than if the cultures are allowed to grow into stationary phase in the presence of excess glucose. Glucose (0.4%, wt/vol) was added to this culture, and growth was continued to an OD600 of ~0.4. This culture was then used to inoculate 25 ml of DM (0.4% [wt/vol] glucose), DM supplemented with 1 mM carnitine or betaine, DM with 0.8 M NaCl, or DM with 0.8 M NaCl supplemented with 1 mM carnitine or betaine to an OD600 of 0.05. Cell growth was measured spectrophotometrically (Ultrospec 4050; LKB, Biochrom) by measuring the OD600 of 1-ml samples at regular intervals during growth. Specific growth rates (μ) were calculated from the doubling times (g) by using the relationship μ = ln2/g. g was determined directly from the growth curves.
Cells were grown overnight in DM supplemented with 0.04% (wt/vol) glucose. Glucose (0.4% [wt/vol]) was added to the overnight culture, and growth was continued to an OD600 of 0.4. This culture was then used to inoculate 50 ml of DM (with 0.4% [wt/vol] glucose) in a 250-ml flask to a starting OD600 of 0.05. Cells were grown at 30°C with shaking to mid-exponential phase (OD600 = 0.4). The uptake of carnitine and betaine was then measured as previously described (9), utilizing l-[3H]carnitine-HCl (Amersham Biosciences) and [14C]betaine (ICN Pharmaceuticals Ltd.). Chloramphenicol (50 μg ml−1) was included in the assay buffer in order to prevent protein synthesis during the course of the assay. The measured uptake rates therefore reflected the levels of functional transporter present in the cells immediately prior to initiating the assay (i.e., during exponential growth in DM). Assays were performed in the presence or absence of an osmotic stimulant (NaCl, 0.5 M). The initial uptake rates were determined from the uptake plot (solute concentration versus time) by using linear regression (Microsoft Excel) to determine the slope of the line over the first 4 min of uptake. Transport rates were linear over this time period. The mean rate of accumulation was calculated from two separate experiments on separate days, with two repeats within each experiment. The error represents the standard deviation from the mean uptake rate.
Cells were grown in DM as described for the transport assays above, except for the following changes. Cells were cultured in 25 ml of DM (with 0.4% [wt/vol] glucose) in 125-ml flasks in the presence of 200 μM carnitine or betaine and grown to an OD600 of 0.2. A 2.5-ml aliquot was transferred to a sterile 25-ml test tube containing 300 nCi of l-[3H]carnitine HCl or [14C]betaine. Steady-state carnitine or betaine pools were then measured as described previously (9).
The plasmid pNF580 (11) was used to construct L. monocytogenes strains containing chromosomal gus reporter fusions for either betL or opuCA, each in the wild-type strain 10403S and in an isogenic sigB null mutant (30). pNF580 is a derivative of the shuttle vector pKSV7, which carries the gus gene, encoding β-glucuronidase (11).
To construct a transcriptional gus fusion to betL, a 279-bp fragment bearing the predicted betL promoter region was amplified by PCR from L. monocytogenes 10403S with primers betL-F and betL-R (Table (Table1).1). For opuCA transcriptional fusions, a 312-bp fragment bearing the predicted opuCA promoter region was amplified by using primers opuCA-F1 and opuCA-R1 (Table (Table1).1). PCR was performed with Vent polymerase (New England Biolabs, Beverly, Mass.) and the following cycling conditions: 3 min at 94°C for one cycle; 30 s at 94°C, 30s at 58°C, and 30s at 72°C for 35 cycles; and a final extension for 5 min at 72°C.
PCR products were purified by using the QiaQuick PCR purification kit (Qiagen, Valencia, Calif.) and digested with XbaI and BamHI for cloning upstream of the promoterless gus gene in pNF580. The digested PCR products were ligated to vector pNF580, which had first been cut with XbaI and BamHI, generating plasmids pDS1 (containing a betL-gus fusion) and pDS2 (containing an opuCA-gus fusion). These plasmids were transformed into OneShot E. coli cells (Invitrogen), and restriction analysis was used to confirm the presence of the correct inserts. Each of these plasmids was then purified and electroporated into both L. monocytogenes 10403S and a nonpolar L. monocytogenes sigB null mutant, FSLA1-254 (30). Transformants were selected on brain heart infusion (BHI) (Difco, Sparks, Md.) agar plates containing chloramphenicol (10 μg ml−1). Isolates were serially passaged at 42°C in BHI broth containing chloramphenicol (10 μg ml−1) to force chromosomal integration of the fusion plasmids. Chromosomal integration of the plasmids was confirmed by sequence analysis of PCR products spanning the integration junctions. PCR was used to verify the presence or absence of the sigB deletion in each of the four strains.
β-Glucuronidase activities directed by the chromosomal betL-gus and opuCA-gus fusions were monitored by using a fluorometric assay measuring hydrolysis of 4-methylumbelliferyl β-d-glucuronide (4-MUG) (Sigma, St. Louis, Mo.) to the fluorescent molecule methylumbelliferone.
β-Glucuronidase activities were assayed for bacterial cells grown to an OD600 of 0.4 in DM. Specifically, an overnight culture of each fusion strain was diluted 1:50 into 10 ml of DM broth containing 0.04% glucose and incubated with shaking at 30°C. Once cultures reached an OD600 of 0.25, glucose was added to a final concentration of 0.4% (wt/vol), and incubation was continued with shaking at 30°C. When these cultures reached an OD600 of 0.4, they were diluted 1:200 into 100 ml of DM (with 0.4% [wt/vol] glucose) and incubated with shaking at 30°C to an OD600 of 0.4.
Bacterial cells were collected essentially as described by Youngman (31). Specifically, cells from 1-ml aliquots were collected by centrifugation at 17,000 × g for 7 min. The cell pellet was resuspended and washed in 1 ml of buffer AB without Triton X-100 (60 mM K2HPO4, 40 mM KH2PO4, 0.1 M NaCl [pH 7.0]), and the washed cell pellet was resuspended in 350 μl of buffer AB without Triton X-100. The OD600 was measured in a Packard (Meriden, Conn.) SpectraCount instrument with clear Costar (Corning, N.Y.) 96-well flat-bottom plates for 200-μl aliquots of these cell suspensions. To quantify β-glucuronidase activities, 100 μl of the cell suspensions was thoroughly mixed with 100 μl of buffer AB containing 0.2% Triton X-100 and incubated at room temperature for 60 min to lyse the cells. In white Microfluor U-bottom 96-well plates (Dynex, Franklin, Mass.), 50 μl of the lysed cells was mixed with 10 μl of 4-MUG (0.4 mg/ml in dimethyl sulfoxide). The fluorogenic reaction mixture was incubated at room temperature for at least 60 min. Fluorescence was measured in a Packard FluoroCount instrument with an excitation filter of 360 nm and an emission filter of 460 nm. Fluorescence units were converted to picomoles of methylumbelliferone by using a standard curve of known methylumbelliferone (Sigma) concentrations. β-Glucuronidase activity was expressed in activity units, which are defined as picomoles of 4-MUG hydrolyzed per milliliter of cells (OD600 = 1.0) per minute.
The RT-PCR method was adapted from that of Cotter et al. (6). Cells were grown to an OD600 of 0.4 as described in “Growth experiments” above, and then 1.5 ml of cells was removed, harvested, resuspended in 500 μl of ice-cold lysis buffer (20 mM sodium acetate [pH 5.5], 1 mM EDTA, 1% sodium dodecyl sulfate), and added to 500 μl of preheated (65°C) phenol-chloroform-isoamylalcohol (1:1:24) and 200 μl of glass beads. Samples were incubated at 65°C for 10 min with frequent vortexing and then centrifuged, and the aqueous phase was reextracted by repeating the hot-phenol step. The aqueous phase was added to 1 ml of 99% ethanol and precipitated at −20°C. After precipitation and washing in 70% ethanol, RNA was resuspended in 15 μl of resuspension buffer (10 mM MgCl2, 1 mM dithiothreitol, 10 mM Tris [pH 7], and 1 mM EDTA made up in diethyl pyrocarbonate-treated water containing 5 U of DNase I [RNase free] [Roche] and 5 U of RNase inhibitor 100 μl−1 [Promega]) and incubated at 37°C for 60 min. cDNA was synthesized from 8.5 μl of diluted RNA by using Expand reverse transcriptase with random primer p(dN)6 (both supplied by Roche). Aliquots (2 μl) of the resulting cDNA were subjected to 12, 18, 24, 30, and 36 cycles of PCR and run on agarose gels. Primers for the 16S rRNA (23) were used as controls. Non-reverse-transcribed RNA was used as template for PCRs to ensure complete removal of genomic DNA. Specific primers for opuCA, gbuA, betL, and lmo1421 (Table (Table1)1) were used in conjunction with cDNA generated from L. monocytogenes 10403S wild-type and ΔsigB strains grown in the presence or absence of 0.5 M NaCl. Relative band intensities were determined by densitometry with a Kodak imaging system and one-dimensional image analysis software. RNA and cDNA were prepared twice, from independent cultures, and the RT-PCRs for each set of primers were repeated at least twice from each cDNA preparation.
The presence of potential σB-dependent promoters upstream from the start codons of the betL and opuCA genes (9, 25) prompted us to investigate the ability of L. monocytogenes to utilize betaine and carnitine as osmoprotectants in wild-type and ΔsigB backgrounds. In batch cultures the wild-type and ΔsigB strains grew with identical specific growth rates and lag times when cultured in either BHI broth or BHI broth with added NaCl (0.5 M). The final OD600s of the cultures after 16 h were also found to be essentially identical (data not shown). When growth in DM in the presence or absence of added NaCl (0.8 M) was examined, the two strains had identical growth rates and both exhibited a 75% growth rate reduction in the presence of 0.8 M NaCl (Fig. (Fig.1).1). To determine the ability of the wild-type and ΔsigB strains to use the compatible solute betaine or carnitine as an osmoprotectant, these solutes were added (1 mM) to DM cultures containing 0.8 M NaCl. The presence of 1 mM carnitine stimulated the growth of the wild type, increasing its growth rate approximately 2.5-fold compared to that of the DM culture with 0.8 M NaCl and no added carnitine. In contrast, the growth of the strain lacking σB was found to be stimulated by less than 1.5-fold in presence of 1 mM carnitine and 0.8 M NaCl (Fig. (Fig.1A).1A). When the osmoprotective effect of betaine was studied, the growth rates of both strains were stimulated by the addition of 1 mM betaine in the presence of 0.8 M NaCl. However, the sigB mutant achieved a growth rate that was approximately 20% less than that observed for the wild type (Fig. (Fig.1B).1B). When these experiments were performed with a lower NaCl concentration (0.5 M), the results were qualitatively identical, although the extent of the salt-induced growth inhibition was lower, making it more difficult to quantify accurately the growth rate changes. The addition of betaine or carnitine to DM without added NaCl had no stimulatory effect on the growth of either strain (data not shown). These data suggest an important role for σB in allowing cells to use carnitine as an osmoprotectant and indicate that under these growth conditions, betaine utilization is largely σB independent.
In order to derive an osmoprotective effect from compatible solutes, bacterial cells must accumulate a high intracellular concentration of these compounds. The growth data described above suggested that mutants of L. monocytogenes lacking σB might be unable to accumulate carnitine efficiently. Therefore, the rates of carnitine uptake were assayed for wild-type and ΔsigB cells grown in DM to mid-exponential phase (OD600 = 0.4). Transport was assayed in the presence or absence of an osmotic stimulus (0.5 M NaCl) and in the presence of the protein synthesis inhibitor chloramphenicol (50 μg ml−1). Under these assay conditions, the transport rates give an indication of the levels of functional compatible solute transporter present in the cell at a particular point during growth, and this level is fixed during the assay by blocking further translation with chloramphenicol. The assay therefore allows the effects of a given genetic background (i.e., wild type or sigB) on the levels of functional transporter present in the cell to be determined at a given time during growth (OD600 = 0.4 in the experiments described here). Using this assay, the wild type accumulated carnitine at approximately 50 nmol min−1 mg of cell protein−1, and this rate was stimulated slightly (approximately 1.5-fold) by the inclusion of NaCl in the assay medium. In contrast, the accumulation of carnitine was almost completely abolished in the ΔsigB mutant (Fig. (Fig.2A2A).
The steady-state cytoplasmic pools of carnitine were also measured during exponential growth in DM. This measurement enables an estimation of the total intracellular pool of carnitine during balanced growth, using growth conditions that were identical to those used to grow cells prior to the transport assays. In wild-type cells the cytoplasmic pool of carnitine was approximately 650 nmol mg of cell protein−1, and this increased slightly when cells were grown in the presence of 0.5 NaCl. The steady-state pool of carnitine in the wild type was approximately 3.5-fold higher than the pool observed in cells lacking σB. These data suggest that σB plays a significant role in allowing L. monocytogenes to accumulate carnitine, although a residual pool of this compatible solute can be accumulated in the absence of σB.
The accumulation of betaine in these strains was also investigated. In the absence of added NaCl, the mutant and the wild type accumulated betaine at similar rates (approximately 2 nmol min−1 mg of cell protein−1). When osmotic stress was imposed by the inclusion of 0.5 M NaCl during the transport assay, both strains showed a large increase in the betaine accumulation rate, though the wild type accumulated betaine at about twice the rate of the ΔsigB mutant under these conditions (Fig. (Fig.3A).3A). The increase in uptake rates observed in the presence of 0.5 M NaCl is indicative of an osmotic stimulation of the activity of the existing betaine transport systems (no effects on expression will be seen during the course of this assay, since translation was inhibited by the presence of chloramphenicol). The differences in uptake rates between the wild type and the sigB mutant must therefore reflect an effect of the genotype on the expression of the betaine transport systems prior to the uptake assay (i.e., during growth in DM). When the steady-state pools of betaine were measured during growth, the mutant strain was found to accumulate a reduced pool of this compatible solute (Fig. (Fig.3B).3B). Together these data suggest that the accumulation of betaine is at least partially dependent on σB.
L. monocytogenes accumulates betaine via at least two independent systems: the Na+-dependent secondary transporter BetL and the ATP-dependent ABC transporter Gbu. By examining betaine accumulation in the absence of sodium, the Gbu transporter can be studied independently of BetL (13). The uptake of betaine was examined in wild-type and ΔsigB cells either without osmotic stimulation or in the presence of either 0.5 M NaCl or KCl. As before, wild-type cells were found to increase the rate of betaine accumulation in response to osmotic stimulation, but the uptake rate was reduced by approximately 50% in the presence of KCl compared with NaCl (Fig. (Fig.4).4). The increase in betaine transport observed in KCl-treated cells is likely to be due to osmotic stimulation of the Gbu transporter alone. In contrast to the results obtained with NaCl, the uptake rates in the presence of 0.5 M KCl revealed no difference in betaine transport between the wild type and the mutant (Fig. (Fig.4).4). This result suggests that σB does not regulate the expression of the Gbu transporter, at least under the growth conditions tested.
The betL and opuCA genes are both preceded by sequence motifs that strongly resemble the predicted σB consensus promoter motif (9, 25). Taken together with the growth and transport data described above, this observation suggested a possible role for σB in the transcriptional regulation of the opuC and betL operons. To directly test this hypothesis, gus transcriptional reporter fusions to either opuCA or betL were constructed (see Materials and Methods). These reporter fusions were recombined into the chromosomes of the ΔsigB mutant and its isogenic parent, generating strains FSL S1-049 (betL-gus), FSL S1-046 (betL-gus ΔsigB), FSL S1-063 (opuCA-gus), and FSL S1-059 (opuCA-gus ΔsigB). These strains were grown to mid-exponential phase (OD600 = 0.4) in DM, and the β-glucuronidase activity from each was measured. Control strains, which did not carry gus reporter fusions, had no detectable β-glucuronidase activity (data not shown). β-Glucuronidase directed from opuCA was found to be approximately sevenfold higher in the wild-type strain than in the ΔsigB mutant (Fig. (Fig.5).5). In contrast, the expression of betL was essentially identical in the presence and absence of σB. These data indicate that the transcription of opuC is under the control of σB, while betL appears to be transcribed in a σB-independent manner, at least under these growth conditions.
Several attempts were made to assay the transcription of betL and opuCA during growth at high osmolarity. These attempts were unsuccessful because the strains harboring the gus fusions failed to grow in DM in the presence of 0.5 M NaCl. The reason for this growth defect was not clear but is probably related to the presence of pNF580 in the chromosomes of these strains, since the parent strains grew well in DM with 0.5 M NaCl (data not shown).
In order to study the role of σB in regulating the transcription of the known compatible solute uptake systems in the presence of osmotic stress, RT-PCR was used as an alternative to the gus reporter system. Primers directed against betL, opuCA, gbuA, and lmo1421 were used in RT-PCR experiments to assess the transcriptional regulation of BetL, OpuC, Gbu, and the hypothetical OpuC-related transporter encoded by lmo1421 and lmo1422 (Fig. (Fig.6).6). All four transcripts were found to increase significantly in the wild type as a consequence of growth in high-osmolarity medium (DM with 0.5 M NaCl). The absence of σB had no detectable effect on the transcript levels of betL, gbuA, or lmo1421 under low-osmolarity growth conditions. However, under low-osmolarity conditions the levels of opuCA mRNA were reduced in the ΔsigB background to about 30% of the wild-type level, consistent with the reduced transcription detected with the opuCA-gus reporter fusion. Under conditions of high osmolarity, the osmotic induction of betL and gbuA transcription was retained in the ΔsigB strain. In contrast, the osmotic induction of opuCA and lmo1421 transcription was completely abolished in the absence of σB. These data demonstrate that σB plays an essential role in mediating the osmotic activation of opuCA and lmo1421 transcription but that it is largely dispensable for the osmotic stimulation of betL and gbuA transcription.
The ability of L. monocytogenes to adapt and grow under conditions of hyperosmotic stress depends on its capacity for accumulating osmoprotectants from the environment; strains lacking the appropriate transport systems are osmotically sensitive (9, 19, 25, 26). The results presented here show that the ability to utilize the compatible solute carnitine as an osmoprotectant is strongly dependent on the stress-inducible sigma factor σB. These data are consistent with an earlier report demonstrating an increased lag phase in a sigB::Km strain, compared to its parent, growing in a high-salt medium with added carnitine (2). Here we have shown that in the absence of σB the osmoprotective effect of carnitine is almost completely abolished (Fig. (Fig.1A),1A), and this is consistent with the finding that both the transport rate (Fig. (Fig.2A)2A) and cytoplasmic pool (Fig. (Fig.2B)2B) of carnitine are reduced dramatically in a ΔsigB background. Since the osmoprotective effect of compatible solutes is based on their accumulation to high levels in the cytoplasm (thereby reversing the water loss that is consequent upon an increase in the extracellular osmolarity), a strain that is unable to accumulate high intracellular levels of this compatible solute will, ipso facto, be unable to utilize it as an osmoprotectant.
The growth and transport data suggest strongly that σB plays a role in regulating the expression of one of the carnitine uptake systems in L. monocytogenes. Measurements of opuCA transcription by using an opuCA-gus fusion and by using RT-PCR to assess transcript levels confirm the role of σB in regulating the transcription of opuCA (Fig. (Fig.55 and and6).6). In the absence of σB the levels of opuCA transcript are reduced (most dramatically for cells growing in the presence of 0.5 M NaCl), and this is predicted to lead to a reduced level of the functional OpuC transporter in the cell, which would account for the transport and growth defects observed in a ΔsigB background. The presence of a putative σB promoter sequence upstream from opuCA (9) suggests that σB plays a direct role in regulating the transcription of opuCA in response to osmotic stress. Indeed, the transcription of sigB itself is up-regulated strongly in response to osmotic stress (2), suggesting that it is likely to play a central role in adjusting transcription in the cell to cope with high-osmolarity environments.
The transport data presented here indicate that the activity of the carnitine transporter in L. monocytogenes is not stimulated dramatically in the presence of NaCl (Fig. (Fig.2A).2A). We routinely observe less than a 1.5-fold increase in activity in the presence of 0.5 M NaCl when cells are grown in DM. In contrast, betaine transport rates are stimulated approximately sevenfold by the imposition of hyperosmotic stress during the transport assay, when either NaCl or KCl is used as the osmolyte (Fig. (Fig.4).4). When KCl is used as an osmolyte, the uptake of betaine is predominantly via the Gbu transporter (13). Therefore, these data suggest a fundamental difference between the ways in which the two related uptake systems (OpuC and Gbu) respond to hyperosmotic shock. Both transporters belong to the substrate binding protein-dependent subgroup of the ABC transporter superfamily. A more detailed biochemical analysis of these systems will be required in order to establish the molecular basis for this apparent difference in the osmotic regulation of these two compatible transport systems.
The data presented here indicate that σB plays an active role in controlling the expression of the compatible solute uptake systems even in cells growing in medium without added NaCl (DM). First, the initial uptake rate of carnitine is reduced in the ΔsigB background when cells are grown in DM (Fig. (Fig.2A).2A). Second, the steady-state pools of both carnitine and betaine are reduced in the mutant when it is growing in DM with added betaine or carnitine (Fig. (Fig.2B2B and and3B).3B). These data are supported by the observation that opuCA transcription is reduced in cells growing in DM in the absence of functional σB (Fig. (Fig.5).5). These data lead us to conclude that σB must be expressed and available to participate in transcription in wild-type cells grown in DM, even though this medium is not designed to deliberately induce a stress response. The osmolarity of this growth medium (in the absence of added NaCl) is approximately 260 mOsM, and this may be sufficient to stimulate the σB-dependent transcription of the compatible solute uptake systems. The cells are then well placed to accumulate compatible solutes when they become available in the environment.
The availability of the L. monocytogenes genome sequence (15) has allowed us perform in silico searches for candidate genes encoding compatible solute uptake systems. Two genes identified in this way are lmo1421 and lmo1422 (10). Although they have not yet been shown experimentally to play a role in compatible solute transport, they are related closely to known ABC compatible solute transporters from Listeria as well as other bacterial genera. They are located approximately 2.4 kb downstream from the opuC operon on the chromosome of L. monocytogenes EGD and are oriented in the opposite direction. The RT-PCR data presented here show that the transcription of lmo1421 is osmotically activated and that this activation is dependent on σB (Fig. (Fig.6).6). Interestingly, the DNA sequence upstream from the predicted start codon (position −85 with respect to the start ATG) of lmo1421 contains a sequence (GAATAT-n14-GGGTAA) with similarity to the σB consensus promoter motif, as established in B. subtilis (16). Although this result does not give us any further clues about the function of lmo1421, it does suggest that σB plays a direct role in regulating the transcription of lmo1421 in response to changes in the osmolarity of the environment. It is possible that the protein products of lmo1421 and lmo1422 are responsible for the low-affinity carnitine transport activity observed in a ΔopuCA mutant (10).
We also show here that betaine accumulation in L. monocytogenes is at least partially σB dependent. This confirms the earlier work of Becker et al. (2), who showed a role for σB in betaine accumulation in L. monocytogenes 10403S. The data we present in this paper indicate that it is the sodium-dependent betaine transport system that is defective in a strain lacking σB; no difference in betaine uptake is detected between the wild type and the ΔsigB mutant when potassium chloride is used to raise the osmolarity and sodium is absent from the transport assay buffer (Fig. (Fig.4).4). This result suggests that gbuA is not under σB control, at least under these growth and assay conditions, and this conclusion is consistent with the RT-PCR data, which show no significant change in gbuA transcript levels in the absence of σB (Fig. (Fig.6).6). The only known sodium-dependent betaine transporter in L. monocytogenes is BetL, encoded by the betL gene (12, 25). This suggested that betL transcription might be dependent on σB, particularly as the betL open reading frame is preceded by a sequence motif that is similar to known σB promoter motifs (25). However, we found no significant difference in the β-glucuronidase activity directed by a betL-gus reporter in the presence or absence of σB (Fig. (Fig.5)5) under growth conditions that did show a difference in betaine transport (Fig. (Fig.4),4), and neither did the RT-PCR analysis of betL transcription reveal any significant decline in transcript levels in the absence of σB (Fig. (Fig.6).6). These results suggest that σB is not directly involved in regulating the transcription of betL and are consistent with a report that no transcript could be detected starting from the σB consensus motif upstream from betL under any condition tested (3). Indeed, the predicted −35 region of the σB consensus promoter motif differs from the sequence upstream from betL at two positions; the σB consensus −35 region is GTTTAA, while the sequence upstream from betL is GTTTCC (22, 25). This difference may be sufficient to render the hexameric sequence unrecognizable by σB, although it has not yet been possible to establish the σB consensus promoter sequence in L. monocytogenes, since only one σB-dependent promoter has been determined experimentally in this pathogen (2). This leaves the question of how σB influences betaine uptake. Two possible explanations for these data remain. (i) An additional, as-yet unidentified, sodium-dependent betaine transporter is present in L. monocytogenes and is transcribed in a σB-dependent manner. A double mutant of L. monocytogenes LO28 lacking both known betaine transporters, BetL and Gbu, has recently been described, and no residual betaine transport activity was observed (29). This result suggests that BetL and Gbu are the only betaine uptake systems present in L. monocytogenes, although extrapolating this conclusion to the present study must be done cautiously, since strain-strain variations in compatible solute transport activities are known to exist (27). (ii) σB may indirectly regulate BetL expression (or activity) at the posttranscriptional level. The σB regulon in B. subtilis consists of at least 100 genes (17), and we have observed significant differences in the proteome of L. monocytogenes in the absence of σB (C. P. O'Byrne, unpublished data). Therefore, the ΔsigB mutation is likely to have pleiotropic effects on L. monocytogenes, which could include posttranscriptional effects on BetL expression or activity. For example, if sodium ion homeostasis was perturbed in the ΔsigB mutant, the activity of BetL could be altered, since its activity is dependent on the sodium ion gradient. Further experiments to determine which of these explanations accounts for the observed data are under way.
In summary, we have shown here that the stress-inducible sigma factor σB plays an essential role in allowing L. monocytogenes to utilize carnitine as an osmoprotectant and that it is clearly involved in regulating the osmotic induction of opuC and lmo1421 transcription. Although a role for σB in regulating betaine transport is demonstrated, this regulation appears not to involve a direct role for σB in modulating the transcription of betL and gbu, the only known betaine transporters in L. monocytogenes. Together these findings highlight the central role of σB in controlling the response to osmotic stress in this important food-borne pathogen.
We are grateful to Roy Sleator and Colin Hill for assistance with the RT-PCR experiments and for helpful discussions. We also thank Ian Booth for useful discussions. We thank Nancy Freitag for providing us with plasmid pNF580.
This work was supported in part by a Unilever BBSRC-CASE studentship. This research was supported in part by the Cornell University Agricultural Experiment Station federal formula funds, project no. NYC-143422, received from Cooperative State Research, Education, and Extension Service, U.S. Department of Agriculture.
Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the authors and do not necessarily reflect the view of the U.S. Department of Agriculture.