Understanding the molecular basis of acid tolerance in the food-borne pathogen Listeria monocytogenes is important as this property contributes to survival in the food-chain and enhances survival within infected hosts. The aim of this study was to identify genes contributing to acid tolerance in L. monocytogenes using transposon mutagenesis and subsequently to elucidate the physiological role of these genes in acid tolerance. One mutant harboring a Tn917 insertion in the thiT gene (formerly lmo1429), which encodes a thiamine (vitamin B1) uptake system, was found to be highly sensitive to acid. The acid-sensitive phenotype associated with loss of this gene was confirmed with an independently isolated mutant, from which the thiT gene was deleted (ΔthiT). Cells of both wild-type and ΔthiT mutant that were thiamine depleted were found to be significantly more acid sensitive than control cultures. Thiamine-depleted cultures failed to produce significant concentrations of acetoin, consistent with the known thiamine dependence of acetolactate synthase, an enzyme required for acetoin synthesis from pyruvate. As acetoin synthesis is a proton-consuming process, we suggest that the acid sensitivity observed in thiamine-depleted cultures may be owing to an inability to produce acetoin.
The gram positive bacterium Listeria monocytogenes is a saprophyte that is ubiquitous in the environment, but is also an intracellular pathogen well adapted to the life in the cytosol of eukaryotic host cells. It is the causative agent of food-borne listeriosis and is associated with a high mortality rate (Freitag et al., 2009). Listeria monocytogenes is highly adapted to survive under acidic environments encountered at a number of stages during the course of its infectious cycle, from food to the gastrointestinal tract and finally in the phagosome of macrophages (Cotter & Hill, 2003; Gray et al., 2006). Listeria monocytogenes’ tolerance to acidic stress is considered as a virulence factor (Werbrouck et al., 2009) and the acid survival strategies employed by the cells have been widely investigated (Davis et al., 1996; Dykes & Moorhead, 2000; Cotter & Hill, 2003; Ferreira et al., 2003). One of the most important acid-adaptive responses in L. monocytogenes is the phenomenon known as the acid tolerance response (ATR), which permits cells to survive lethal acid when first exposed to sublethal acid stress during the exponential phase growth (Davis et al., 1996; Ferreira et al., 2003; Skandamis et al., 2008; Chorianopoulos et al., 2011). Although the molecular basis for this response is not yet understood, cellular components that contribute to acid tolerance have been identified, including both the glutamate decarboxylase system (Cotter et al., 2001) and the arginine deiminase system (Ryan et al., 2009). The alternative sigma factor Sigma B has also been identified as an important regulator of acid tolerance (Wiedmann et al., 1998).
The initial aim of the present study was to identify genetic components that contribute to acid tolerance using transposon mutagenesis. One mutant with a Tn917 insertion in the thiT gene (lmo1429) proved to have a highly acid-sensitive phenotype. This gene was known to encode a thiamine uptake system (Schauer et al., 2009). Thus, the remainder of the study focused on establishing the role of ThiT in acid tolerance in L. monocytogenes and on determining if thiamine itself is required for an acid tolerant phenotype in this pathogen. ThiT is an integral membrane protein containing six transmembrane helices for thiamine recognition and binding (Erkens & Slotboom, 2010). It is predicted to act as the substrate binding S subunit of subclass II factors belonging to the energy coupling factor (Ecf) class of transporters. These also comprise A and T subunits that act as an energizing module during transport (Rodionov et al., 2009; Eitinger et al., 2011). A recent study has demonstrated the requirement for the EcfA and EcfT subunits for thiamine transport by ThiT in Lactococcus lactis (Erkens et al., 2011). In L. monocytogenes, these subunits are thought to be encoded by lmo2601, lmo2600, and lmo2599 (Schauer et al., 2009). The presence of a thi box in the 5′ untranslated region suggests that thiamine pyrophosphate (TPP) influences thiT transcription via a riboswitch mechanism (Winkler et al., 2002; Eudes et al., 2008). Thiamine is an essential co-factor in L. monocytogenes as not all the genes involved in thiamine biosynthesis are present in the genome. TPP, the biologically active form of thiamine, is used as a co-factor by several metabolic enzymes including those of central function. The involvement of ThiT in thiamine uptake was demonstrated by the binding of thiamine to the ThiT protein and by the resistance of ThiT-negative mutants to the toxic thiamine analog pyrithiamine (Schauer et al., 2009). This uptake system may also play a role during pathogenesis as mutants lacking the system are defective for intracellular growth (Schauer et al., 2009).
The results of the present study established that thiT is necessary for full acid tolerance in L. monocytogenes and demonstrated that thiamine-depleted cells are more acid sensitive than control cells. Cultures grown without thiamine were found to produce dramatically lower levels of acetoin, a metabolite derived from pyruvate, and we discuss the possibility that this deficiency might be responsible for the acid sensitivity of thiamine-starved cells.
Materials and methods
Bacterial strains, plasmids, and culture conditions
Listeria monocytogenes EGD (serotype 1/2a), wild-type, and an isogenic mutant derivative EGD ∆thiT (Schauer et al., 2009) were used throughout this study. A mutant derivative of EGD carrying a Tn917 insertion in the thiT gene was independently isolated as described below. Listeria monocytogenes strains were streaked on brain heart infusion (BHI; Lab M Ltd) agar plates and incubated at 37 °C for 24 h. Overnight cultures were obtained from a single colony inoculated into 5 mL of BHI broth in 20 mL universal tubes and incubated at 37 °C with shaking at 160 r.p.m. (New Brunswick Scientific Bio Gyrotory® Shaker). Listeria monocytogenes strains were cultivated in BHI broth or in a chemically defined medium (DM; Amezaga et al., 1995) with or without thiamine supplementation (1.0 mg L−1) at 37 °C, with shaking. When required, the pH of DM solution was reduced using 10 M HCl.
Tn917 library construction and screening
A mutant library consisting of 4800 individual Tn917 insertion mutants was generated in L. monocytogenes EGD by transposon mutagenesis, using the shuttle vector pLTV3 as a source of the Tn917 and following the method described by Camilli et al. (1990). Mutants were stored at −80 °C in 96-well microtitre plates until required. Mutants were first cultured at 30 °C for 16 h in BHI in 96-well microtitre plates using a stainless steel 96-well replica plater to inoculate the wells. Then, mutants were transferred to an acidified medium (BHI acidified to pH 3.0 with 10 M HCl) using the replica plater. After 1 h at pH 3.0, survivors were replica plated onto BHI agar and mutants showing poor survival (evidenced by reduced growth) after an overnight incubation at 30 °C were selected for further study. Southern blotting was used to confirm the presence of a single Tn917 insertion in genomes of isolates that were investigated further following the screen. The junction region between the Tn917 transposon and the EGD chromosome was amplified by inverse PCR (Ochman et al., 1988) and the resulting product sequenced to identify the disrupted gene. The DNA sequence was determined using a Perkin-Elmer Applied Biosystems 377 automated sequencer and analyzed using dnastar Inc. software.
The growth experiments were carried out in 250 mL conical flasks containing 25 mL of medium. One mL aliquots of overnight cultures of the wild-type and the ∆thiT mutant, grown in BHI at 37 °C, were centrifuged at 3420 g for 5 min and the bacterial pellets were washed twice with DM, either with or without thiamine before resuspending in the same volume of DM. The flasks were then inoculated with this suspension to an initial OD600 nm of 0.05 in 25 mL of fresh DM with or without thiamine and incubated at 37 °C with shaking. The OD600 nm was then measured at appropriate intervals throughout the growth.
Acid survival experiments
Cultures grown in 25 mL BHI in 250 mL flasks with shaking at 37 °C were assayed for acid tolerance by diluting 1 : 10 into BHI medium adjusted to pH 3.0. At suitable intervals, samples were removed, serially diluted, and 10 μL aliquots of each dilution were plated on BHI agar plates. Colonies were counted after 24 h at 37 °C and survival was calculated as a percentage of the cell count at time zero. For acid-adapted cultures, cells were first diluted 1 : 10 into BHI medium adjusted to pH 5.0, incubated for 1 h, and then further diluted 1 : 10 into BHI medium adjusted to pH 3.0. For experiments on thiamine-depleted cells, cultures were grown in DM either with or without thiamine supplementation (3 μM), and then after 12 h of growth, cells were diluted 1 : 20 into DM adjusted to pH 3.0 (which contained thiamine). Survival was determined by serial dilution and plating on BHI agar plates as described above.
Real-time RT-PCR measurement of transcript levels
Relative transcript levels of thiT in exponentially growing cells (OD600 nm = 0.6) at pH 5.5 or pH 5.0 compared to pH 7.0 were measured by real-time RT-PCR as previously described (Utratna et al., 2011).
Acetoin was determined by the modified Voges–Proskauer reaction of Westerfeld (1945), with slight modification. Stationary phase cultures of L. monocytogenes wild-type and mutant grown in both DM supplemented with thiamine and DM without thiamine were recovered and centrifuged at 14 500 g for 5 min. The supernatants were used to measure the acetoin production. To 1.0 mL of culture supernatant in DM, diluted appropriately to give a reading within the range of the calibration curve for acetoin, 0.2 mL of 0.5% (w/v) l-arginine monohydrochloride and 0.2 mL of 5% (w/v) α-naphthol in 2.5 N NaOH were added, in that order. The pink color that developed after 1-h incubation was measured by recording the absorbance at 530 nm using a UV-VIS spectrophotometer (Spectronic® 20 Genesys™). The concentration of acetoin was estimated from a linear calibration curve based on measurements of standard acetoin solutions (0.01–40 μg mL−1).
The thiT gene (lmo1429) is required for acid tolerance
To identify genetic determinants of acid tolerance in L. monocytogenes, a library of 4800 transposon (Tn917-lacZ) mutants was screened for mutants displaying an acid-sensitive phenotype at pH 3.0. One acid-sensitive mutant, initially designated ads12, was found to induce a poor adaptive ATR at pH 5.0 compared to the wild-type, indicated by a dramatically reduced ability to survive at pH 3.0 (Fig. 0001a). Inverse PCR was used to amplify the transposon-chromosome junction and subsequent sequence analysis identified the gene carrying the insertion as lmo1429. At the time this screen was conducted, no information was available on the function of lmo1429, although it was highly similar to the yuaJ gene from Bacillus subtilis, which also had no known function. Subsequently in an independent study, lmo1429 was shown to encode a thiamine uptake system and was renamed thiT (Schauer et al., 2009). To confirm the role of thiT in acid tolerance, suggested by the phenotype of the lmo1429::Tn917-lacZ transposon mutant, the ability of a ∆thiT deletion mutant to withstand an acid challenge at pH 3.0 was compared to the wild-type (EGD) using cells cultured in BHI, both before and after the induction of an ATR. After induction of an ATR (1 h at pH 5.0), the ∆thiT strain lost viability rapidly after exposure to pH 3.0 whereas the wild-type was essentially unaffected by this challenge pH (Fig. 0001b). For unadapted exponentially growing cultures, the presence of a thiT deletion also reduced the ability to survive at pH 3.0; after 90 min at pH 3.0, approximately 60-fold more wild-type survivors were counted than mutant survivors (Fig. 0001b), and no mutant survivors could be detected at later sampling times. These data indicate that the thiT gene contributes significantly to the acid tolerance of L. monocytogenes.
To investigate whether the thiT gene was itself induced under acidic conditions, real-time RT-PCR was used to measure the relative transcript levels in EGD cells growing at pH 5.5 or pH 5.0 versus an untreated control culture (pH 7.0). The results indicated that thiT was induced approximately 1.9-fold at pH 5.5 and 2.3-fold at pH 5.0 (P < 0.05; data not shown), conditions that are known trigger the induction of an ATR (Davis et al., 1996).
Thiamine limitation causes premature growth arrest of the ∆thiT mutant
As thiT is known to play a role in thiamine uptake in L. monocytogenes, the results described above suggested that thiamine limitation might be responsible for the acid-sensitive phenotype observed. To test this directly, the growth of a wild-type and ∆thiT mutant was measured in a chemically DM with and without thiamine supplementation (1 mg L−1). In this medium, the wild-type and ∆thiT mutant both grew with similar specific growth rates (0.45 and 0.46 h−1, respectively) when thiamine was present (Fig. 0002), suggesting that neither strain is limited for thiamine in this growth medium. In the absence of thiamine, the wild-type entered stationary phase 8 h after inoculation (OD600 nm = 0.83) while the ∆thiT mutant was growth arrested at 5 h (OD600 nm = 0.21) (Fig. 0002). In both cases, growth arrest was shown to be caused by thiamine starvation as the addition of thiamine to the medium after growth had arrested allowed the cells to resume normal growth (data not shown). These data suggested that the ∆thiT mutant had a lower intracellular pool of thiamine than the wild-type at the point of inoculation and therefore became thiamine limited after a fewer number of generations.
Thiamine-limited cells are sensitive to acid stress
To test whether the thiamine status of the cell influences acid tolerance, acid survival was determined for both wild-type and ∆thiT mutant after 12 h of growth in DM, either with or without thiamine supplementation. For both strains, survival was significantly reduced when cells were first starved for thiamine (Fig. 0003). No survivors were detected at pH 3.0 subsequent to the 75-min time point for thiamine-depleted wild-type cells whereas the thiamine-replete culture still had > 105 CFU mL−1 survivors at 150 min (Fig. 0003). Likewise, the mutant strain was dramatically more sensitive to acid when it was first starved for thiamine by culturing in a thiamine-free medium. The mutant was also significantly more sensitive than the wild-type when they were grown either in the presence or absence of thiamine, but the magnitude of the differences was smaller (Fig. 0003). Thus, the availability of thiamine in the cell has a significant influence on acid survival in L. monocytogenes.
Thiamine availability influences acetoin production
In L. monocytogenes, the biosynthesis of acetoin is known to be dependent on thiamine (Romick & Fleming, 1998), as acetolactate synthase, the enzyme that converts pyruvate to acetolactate (a precursor of acetoin), depends on thiamine as a co-factor (Romick & Fleming, 1998; Xiao & Xu, 2007). As acetoin production has been implicated in pH homeostasis in other bacteria (Tsau et al., 1992; Cañas & Owens, 1999), we investigated whether the availability of thiamine in the culture medium influenced acetoin accumulation in the wild-type and the ∆thiT mutant. Acetoin levels were measured in the culture supernatants at suitable intervals during growth in DM, either with or without thiamine supplementation. As expected, cultures grown in the presence of thiamine accumulated acetoin as the cultures entered stationary phase (approximately 8 h), consistent with the findings of an earlier study (Romick & Fleming, 1998). Cells grown in the absence of thiamine accumulated dramatically reduced levels of acetoin. There was approximately 12 times more acetoin in the wild-type culture after 12 h when thiamine was present than when it was absent (Fig. 0004). The ∆thiT mutant also produced significantly less acetoin than the wild-type when both strains were grown under thiamine-limiting conditions (P < 0.5 Student's t-test, n = 6). Taken together, these data highlight the involvement of thiamine in acetoin production and suggest the possibility that acetoin could play a role in acid tolerance in L. monocytogenes.
In this study, we have provided evidence that thiamine plays a critical role in the ATR of L. monocytogenes. Mutants that are defective for thiamine uptake displayed reduced acid tolerance both after acid adaptation and when growing exponentially without adaptation. The availability of thiamine in the growth medium also had a significant impact on the ability to tolerate a lethal acid challenge. Cultures starved for thiamine produced dramatically reduced acetoin levels, presumably owing to reduced activity of the thiamine-dependent enzyme acetolactate synthase, and this may contribute to the acid-sensitive phenotype displayed by these cultures.
When the growth of the wild-type was compared to the ∆thiT mutant in a chemically DM, they were found to grow at essentially identical rates when thiamine was present in the medium. This result was unexpected because an earlier study had found that the same mutant grows with a significant growth lag, although the growth rates were similar (Schauer et al., 2009). It seems likely that this difference in growth resulted from the different media or experimental procedures used in the two studies. However, both data sets suggest that L. monocytogenes might encode a rescue pathway or an alternative uptake system for thiamine that is capable of meeting the thiamine needs of the cell during growth in media containing thiamine. One possibility is that the putative EcfA and EcfT components of the ThiT transporter, thought to be encoded by the operon lmo2601, 2600, 2599 (Schauer et al., 2009), could associate with an alternative, as yet unidentified, S subunit.
The way in which thiamine contributes to acid tolerance in L. monocytogenes is not clear at present, but it seems likely that a thiamine-dependent enzyme reaction is required for protection against low pH. Several enzymes are known to be dependent on this co-factor, including pyruvate dehydrogenase, pyruvate oxidase, transketolase, 2-oxoglutarate decarboxylase, and acetolactate synthase (Schauer et al., 2009). 2-Oxoglutarate decarboxylase catalyzes the decarboxylation of α-ketoglutarate to succinyl semialdehyde, a metabolite that is also thought to be produced by a pathway involving the metabolism of γ-aminobutyrate (GABA). As GABA is known to be involved in acid tolerance in L. monocytogenes (Karatzas et al., 2010), it is possible to speculate that succinyl semialdehyde production could influence acid tolerance by modulating the metabolism of GABA. Further experiments will be required to address this possibility.
In this study we show that acetoin production is influenced by the thiamine status of the cells, a result that suggests reduced acetolactate synthase activity. This thiamine-dependent enzyme catalyzes the decarboxylation of pyruvate to acetolactate, a reaction that has been shown to play a critical role in pH homeostasis in Lactobacillus plantarum (Tsau et al., 1992) as well as in Leuconostoc mesenteroides (Cañas & Owens, 1999). This conversion consumes a cytoplasmic proton and a further proton is consumed when acetolactate is decarboxylated (by acetolactate decarboxylase) to form acetoin, thereby raising the intracellular pH. Indeed, the genes encoding both acetolactate synthase (alsS; lmo2006) and acetolactate decarboxylase (alsD; lmo1992) in L. monocytogenes are upregulated significantly in response to acid stress (Bowman et al., 2010). Furthermore, a recent study describing the response of L. monocytogenes to organic acids has shown that cells respond by increasing acetoin production (Stasiewicz et al., 2011). Together these data suggest that the role for thiamine in acid tolerance may well result from the requirement for acetoin production from pyruvate under conditions of acid stress, although further experiments will be required to test this model rigorously.
The authors are grateful to members of the Bacterial Stress Response Group at NUI Galway for helpful discussions and comments on the manuscript. We thank Prof Simon Foster for providing us with EGD (pLTV3). The work was supported by a Science Foundation Ireland SIRG award to K.A.K. (09/SIRG/B1570) and by an Irish Research Council for Science, Engineering and Technology EMBARK award to M.U.