Introduction

The foodborne pathogen Listeria monocytogenes emerged in the late 20th century and has been cause of many outbreaks of listeriosis with high case fatality rates. Listeria survival depends on membrane lipid homeostasis and on its ability to adjust lipid composition to acclimatize to different environments. The first step to understanding the role of the membrane in adaptation is to define in detail the composition of the membrane. Mastronicolis et al. (1996a) isolated phospholipids of L. monocytogenes and reported that four main phospholipids represented 88% of the total L. monocytogenes polar lipids.

Listeria monocytogenes cells exposed to low temperature (Annous et al. 1997; Mastronicolis et al. 2006), osmotic (Chihib et al. 2003), anaerobic stresses (Jydegaard-Axelsen et al. 2004) and low pH stress (Giotis et al. 2007) undergo alterations in membrane fatty acid (FA) profiles. As a component of cellular membranes, FA are essential from a qualitative and a quantitative point of view. In optimal conditions L. monocytogenes is characterised by >85% branched chain fatty acids (BCFA; principally anteiso C15:0 [ai-C15:0] and anteiso C17:0 [ai-C17:0]) (Mastronicolis et al. 2005). It appears that membrane modifications differ from those documented in other genera, where exposure to moderately acidic conditions results in increased levels of monounsaturated long-chain FA in the membrane; in Listeria, higher proportions of straight chain fatty acids (SCFA) are incorporated into the membrane (Ryan et al. 2008). The physiological response to mild acidic stress (acid tolerance response, ATR) has the potential to influence the survival of bacteria in acidic environments, sub-optimal to growth (Gahan et al. 1996). Acids inhibit essential molecular reactions by several mechanisms; one of these is by increasing H+ concentration, which results in a decrease in internal pH (pHi) (Brown and Booth 1991). This drop in pHi is a major cause of growth inhibition by weak acids (Beales 2004). The antimicrobial effect of weak acids is both pH-dependent and non-pH dependent, although this has been long established to be more active in an acid than in a neutral environment (Eklund 1983; Vasseur et al. 1999).

Inducing resistance to one stress also increases resistance of bacteria to other types of stress. L. monocytogenes cells exposed to sublethal acid stress have increased resistance to heat, ethanol and salt stress (Bayles 2004). Acid adaptation can increase L. monocytogenes resistance to nisin activity (van Schaik et al. 1999). The cross-resistance of acid adapted cells to other stresses has important implications, since foods commonly encounter sublethal acidic treatments during processing (Gandhi and Chikindas 2007). The acid response of cells involves changes in cell physiology. Modifications in fatty acid composition are extensive enough to alter membrane fluidity and permeability, in order to minimise energy expenditure and to optimise growth. Moreover, the surface physicochemical properties of cells that govern biofilm formation, such as hydrophobicity, van der Waals and Lewis acid–base characteristics and electrical properties, in part depend on membrane phospholipid composition and their modulation (Haest et al. 1972; Gianotti et al. 2008; Prokopovich and Perni 2009). L. monocytogenes has a known ability to attach to and survive on surfaces found in food-processing plants (Giovannacci et al. 2000).

In order to determine whether membrane FA adaptation is a specific survival mechanism for L. monocytogenes grown at pH 5.5 with hydrochloric, acetic or lactic acid, (conditions under which L. monocytogenes is acid-sensitive) we examined whether this phenomenon also occurs in the presence of benzoic acid at neutral pH. The aim of the study was to investigate changes in the L. monocytogenes phospho- or neutral lipids and their FA profiles, in response to environmental pH, under the influence of various acids in order to provide a better understanding of acid adaptation mechanism in this bacterium.

Materials and methods

Culture of the organism

An avirulent strain of L. monocytogenes, DP-L1044 (D. Portnoy, University of Pennsylvania), prepared by a transposon insertion (Camilli et al. 1991) in the parent strain (Lm10403S), was grown in Brain Heart Infusion broth (BHI, Difco Laboratories) at 30°C (24 h). A 10 ml aliquot of this was then inoculated into 1 l of BHI broth, which was then incubated at 30°C (Lmcontrol) until early stationary phase (OD600 = 0.8; 8 h). Four aliquots (10 ml) of the same stock were each inoculated into 1 l BHI that were, respectively, adjusted to pHinitial 5.5 with (i) HCl (LmHCl), (ii) l-lactic acid (Fluka) (LmLA), or (iii) acetic acid (Merck) (LmAA). Another 10 ml aliquot was used to inoculate 1 l BHI with the addition of 1.00 g benzoic acid (Merck, pa) (LmBA) pHinitial 7.3. All the above cultures were incubated at 30°C until early stationary phase.

Extraction of total lipids (TL) and isolation of polar (PL) and neutral (NL) lipid classes

Cells pelleted by centrifugation were washed twice in phosphate buffer (pH 7.0). TL were extracted (Folch et al. 1957) and separated into PL and NL lipids by solid-phase extraction (SPE) as previously described (Berger et al. 1992; Mastronicolis et al. 2005).

Methanolysis of lipids and gas chromatography (GC)

Each culture lipid sample (TL, PL and NL) was converted to methyl esters (FAME) as previously described (Vance and Sweely 1967; Mastronicolis et al. 2006). FAME were separated by using a Shimadzu gas chromatograph with a split/splitless injector, flame ionisation detector, capillary column DB-5ms (Phenyl Arylene polymer virtually equivalent to 5%-phenyl-methylpolysiloxane; 30 m by 0.251 mm, 0.25 μm film thickness, Cat No. 1225532, Serial No. US3270652 H, J&W Scientific). Carrier gas (He) flow was 44 ml/min. The injector and detector were maintained at 250 and 280°C, respectively. The oven temperature was ramped from 90 to 170°C (70°C/min) and from 170 to 210°C (5°C/min). The initial and final holding times were 1 and 2 min, respectively. Two standard mixtures were used: bacterial FAME (CP Matreya, Inc.) and Marine oil FAME (Restek Corp.).

Identification and quantification of membrane phospholipids by thin-layer chromatography (TLC) and phosphorus analysis

TL phosphorus analysis was performed as previously described (Long and Staples 1961). Chloroform/methanol/acetic acid/water (50/25/6/2, vol/vol/vol/vol) was used for one dimensional thin-layer chromatography (1D-TLC; silica gel G60 aluminum chromatoplate [Merck 1.05554], 0.25 mm thick, 10 × 10 cm). Spots were visualised using iodine, ninhydrin and phosphomolybdenum blue (Dittmer and Lester 1964). The standards cardiolipin (CL), phosphatidylglycerol (PG), phosphatidylinositol (PI), phosphatidylcholine (PC) and lyso-phosphatidylethanolamine (L-PE) were purchased from Sigma Chemical Co. Individual phospholipids (PhL) bands (cells LmBA or LmAA) were quantified after 1D-TLC fractionation of each PL sample (10 μg applied phosphorus): each iodine-visualised band (n = 4) was scraped into a digestion tube and the total phosphorus content was determined as previously published (Kariotoglou and Mastronicolis 1998).

Statistical analysis

Results were evaluated by analysis of variance (ANOVA) (Downie and Heath 1974; Zar 1999). T-test for unpaired observations was tested at a confidence level of 95%.

Results

Lipid fractionation and lipid class composition of acid adapted L. monocytogenes

Growth of L. monocytogenes in BHI medium to early stationary phase, in the presence of four different acids, was determined by measuring absorbance (OD) at 600 nm. The obtained OD600 values: LmHCl = 0.271 ± 0.013, 24 h (n = 5); LmLA = 0.083 ± 0.018, 168 h (n = 4); LmAA = 0.210 ± 0.031, 72 h (n = 12); and LmBA = 0.682 ± 0.013, 10 h (n = 9) provided information on the sensitivity of cells in each culture adapted to different conditions. Cells grown at neutral pH, in presence of 1 g/l benzoic acid, displayed little difference from normal cells Lmcontrol (OD = 0.816 ± 0.032, 10 h, n = 8).

TL were extracted from adapted and non-adapted cells of L. monocytogenes at early stationary phase, and fractionated into PL and NL classes. NL from each culture were quantified by weight and shown in Table 1. The growth of L. monocytogenes in the presence of hydrochloric, acetic, or lactic acid at pH 5.5, caused a substantial increase (31.9, 49.3, and 41.7%, respectively, P < 0.01) of NL percentage compared to Lmcontrol cells. Interestingly, the growth of L. monocytogenes in the presence of benzoic acid caused a slight decrease (6%, P < 0.05) in the NL percentage. The lipid phosphorus content of each TL sample is shown in Table 1.

Table 1 % neutral lipids (NL), the ratio of neutral to polar lipids (NL/PL) and % phosphorus content of each L. monocytogenes total lipid sample extracted from cells grown in optimum conditions (Lmcontrol) or grown in the presence either of benzoic acid at pH 7.3 (1 g/l) (LmBA) or at pH 5.5 [by hydrochloric (LmHCl), acetic acid (LmAA), or l-lactic acid (LmLA)]
Full size table

Listeria monocytogenes growth in the presence of three different acids at pH 5.5, or in presence of benzoic acid at neutral pH, was accompanied by a significant reduction (P < 0.01) in lipid phosphorus compared to Lmcontrol cells as follows: LmBA 33.9%, LmHCl 33.6%, LmLA 67.3% and LmAA 51.3%.

Effect of acids on L. monocytogenes TL, PL and NL fatty acid composition

TL FA profiles of discrete adapted L. monocytogenes cells were similar and dominated by four major FA (in order: ai-C15:0, ai-C17:0, C16:0 and C18:0), which totalled 69.5, 69.3, 60.0 and 69.4% for LmBC, LmHCl, LmAA and LmLA, respectively (Fig. 1). In contrast, ai-C15:0 and ai-17:0 alone represented 69.1% of the Lmcontrol TL FA, as observed previously (Mastronicolis et al. 2006). Upon acid stress, the proportion of saturated SCFA (C16:0, C18:0) in the L. monocytogenes TL increased at the expense of BCFA (ai-C15:0 and ai-C17:0). For each batch of acid adapted cells, the ratio of total BCFA to total saturated SCFA (∑BCFA/∑saturated SCFA) of their TL, PL and NL FA profiles compared to Lmcontrol is shown in Fig. 2. It is worth noting that the NL FA ratio in adapted cells shows a similar large (×8 fold) decrease in relation to the ratio of the Lmcontrol cells, while the PL FA ratio shows a smaller decrease (averaging ×1.7 fold). These findings appear to indicate that the mechanism for L. monocytogenes growth under the influence of acids, regardless of acid type, involves a strong increase of ∑ saturated SCFA at the expense of ∑BCFA as reflected in the FA composition of the TL due to NL FA. Total unsaturated FA (including C16:1 and C18:1) increased from 1.7% in the TL FA of Lmcontrol cells to 7.9, 8.3, 14.5 and 15.4% in LmBA, LmHCl, LmAA and LmLA cells, respectively. This increase is due mainly to an increase in NL FA.

Fig. 1
figure 1

Mean concentration (and standard deviation) of fatty acids of (i) total lipids, (ii) polar lipids and (iii) neutral lipids in Listeria monocytogenes before (Lmcontrol) and after acid stress exposure a at pH 5.5 by HCl (LmHCl), acetic acid (LmAA) or l-lactic acid (LmLA) and b at pH 7.3 by benzoic acid (LmBA). Individual fatty acids comprising 0.1–2.2% (C14:1, C17:0, C18:2 and C20:0), with the exception of iso C17:0 which amounted to 1.7–4.8%, were ignored

Full size image
Fig. 2
figure 2

Ratio of total branched chain fatty acids, BCFA, to total saturated straight chain fatty acids, SSCFA (∑BCFA/∑SSCFA) of total, polar and neutral lipid FA profiles of cells grown in the presence either of benzoic acid (1 g/l, pH 7.3) (LmBA) or at pH 5.5 with hydrochloric (LmHCl), or acetic acid (LmAA) or l-lactic acid (LmLA)

Full size image

Effect of benzoic or acetic acid in L. monocytogenes phospholipid composition

LmBA and LmAA PL samples, separated by 1D-TLC, contained four major phosphocomponents (CL, PG, PhAL, PI) similar to the pattern previously reported (Mastronicolis et al. 1996b, 2008). A new phosphocomponent (designated as A) was observed in both acid adapted cells, and co-chromatographed close to a phosphatidylcholine (PC) standard (Fig. 3). Table 2 shows the lipid phosphorus distribution of each lipid component.

Fig. 3
figure 3

Thin-layer chromatography separation of individual L. monocytogenes phospholipids derived from cells grown in presence of acetic LmAA or benzoic acid LmBA (6th and 7th zone, respectively, from the left). The diagram shows these standards from the left to right: CL cardiolipin, PG phosphatidylglycerol, PI phosphatidylinositol, L-PE lyso-phosphatidylethanolamine and PC phosphatidylcholine. The plate was developed using the solvent system chloroform/methanol/acetic acid/water [50/25/6/2, vol/vol/vol/vol]. Visualisation: after spraying with molybdenum blue reagent. L. monocytogenes cells contained, from top to bottom: CL, PG, PhAL (phosphoaminolipid), PI, and phosphocomponent A

Full size image
Table 2 Separationa and composition (% of lipid Pb) of L. monocytogenes phospholipids, grown in optimal conditions cultures, Lmcontrol and grown either in the presence of benzoic acid (1 g/l, pH 7.28), LmBA, or of acetic acid (pH 5.5) LmAA
Full size table

The quantitative composition of the main phospholipids from LmBA cells diverged strongly (P < 0.01) from that of Lmcontrol cells. When CL, PG and PI were combined to give the total anionic phospholipids, the value amounted to 53.1 ± 3.2% for LmBA and 75.7 ± 2.2% for Lmcontrol cells.

Unlike LmBA cells, the percentage of the phospholipid species from LmAA cells did not differ greatly from the Lmcontrol. Specifically the data indicated a slight but significant (P < 0.01) decrease of CL and a major decrease (P < 0.01) in PG content compared to Lmcontrol cells. For the remaining two varieties of phospholipid (PI, PhAL) there appeared to be no statistically significant change.

Discussion

This study is the first approach to undertake an observation of the involvement of phospho- or neutral lipids in the mechanism of L. monocytogenes cellular adaptation to exposure to acid stress at low or neutral pH, besides their FA profile alteration. Hydrochloric acid will be dissociated whereas acetic (pK a = 4.74) and lactic acid (pK a = 3.79) will be undissociated at pH 5.5. The latter form of both acids is membrane permeable and thus allows acetic and lactic acid to enter the microbial cell. Benzoic acid (pK a = 4.19) at pH 7.3 will be in its dissociated form (benzoic anion) and showed a lower antimicrobial effect on L. monocytogenes. The adaptation mechanism of L. monocytogenes in the presence of acidic stress at pH 5.5 (enhanced antimicrobial activity) caused: (i) a substantial increase of ∑saturated SCFA at the expense of ∑BCFA. This increase is due mainly to an increase in NL FA. (ii) A substantial increase of the percentage of NL, calculated from the increased mass value among TL. Such lipid changes might play a protective role when cells are subjected to (acidic) stress by decreasing the membrane fluidity. It is worth mentioning that the growth of L. monocytogenes in the presence of benzoic acid at neutral pH caused: (i) a similar alteration in the membrane FA as reported for low pH. (ii) No alteration in the percentage of NL contents, while instead, the relative quantities of phospholipids were dramatically affected by neutralisation of the negative head group charges. Such lipid changes also promote a decrease in membrane fluidity.

Effect of acids on lipid-phosphorus amount and NL content changes

Our data indicated that the acid adaptation response of L. monocytogenes caused a significant decrease in TL phosphorus of each culture (Table 1). In cells grown at low pH the decrease in phosphorus reflects a higher content of NL, while in cells grown in the presence of benzoic acid at neutral pH, this may reflect a higher content of non-phospholipid components (in the PL class), such as glyco- or aminolipids as indicated by ninhydrin on 1D-TLC. Moreover the lipid phosphorus content of PL from each culture after separation by SPE was also analysed and only in low pH adapted cells was the PL phosphorus content lower than that of Lmcontrol cells (unpublished data). These data reflect the NL accumulation in cells grown at low pH. A similar mechanism was found for cold (5°C) adapted cells in previous studies (Mastronicolis et al. 2006). Furthermore, other authors (Phan-Thanh and Mahouin 1999; Gandhi and Chikindas 2007) reported that the protein GroEL showed increased synthesis during the growth of L. monocytogenes at low temperatures and was also induced under acid stress. Reduction in the content of phospholipids has also been utilised by L. monocytogenes and other bacteria in response to different environmental stresses such as antibiotics (membrane disrupting peptide) in Pseudomonas aeruginosa (Conrad and Galano 1989) or bacteriocins in L. monocytogenes (Ming and Daeschel 1995) or ethambutol in Mycobacterium tuberculosis (Sareen and Khuller 1990).

Links between alterations in FA and acid treatment

Alterations in the FA profile may alter membrane permeability and fluidity, which in turn may contribute to acid tolerance. Our data show that, irrespective of pH, the acid adaptation response will lead to a similar alteration in FA composition, mainly in the NL class of adapted cultures. Changes in the FA profile of newly synthesized NL class were also observed in cold (5°C) adapted L. monocytogenes from our previous studies (Mastronicolis et al. 1998, 2006). The increased percentage in the proportion of C16:0 and C18:0 saturated SCFA (high-melting point FA) was maintained in all the acid adapted cultures. Such modifications are known to decrease membrane fluidity (Russell 1989). The higher proportion of total cis-monounsaturated fatty acids may reflect the induction of adaptive mechanisms in acid adapted L. monocytogenes cells to counteract reduced membrane fluidity, resulting from the increase of saturated SCFA and the reduction of BCFA.

Other authors have reported that exposure of bacteria to moderately acidic conditions results in the incorporation of higher SCFA proportion into their membrane (van Schaik et al. 1999; Fozo and Quivey 2004; Giotis et al. 2007). van Schaik et al. (1999) and Giotis et al. (2007) examined whole cell FA extracts of acid adapted L. monocytogenes and observed only subtle alterations in FA composition. An increase in saturated SCFA and decrease in BCFA was also observed in Aeromonas spp. (Chihib et al. 2005) and Bacillus subtilis (Lopez et al. 2006) subjected to high NaCl concentrations. An increase in C16:0 and C18:0 SCFA (Mazzotta and Montville 1997; Naghmouchi et al. 2007) and in the ratio of unsaturated to saturated acyl-chain species of phosphatidylglycerol (Vadyvaloo et al. 2002) was also observed for bacteriocin resistant L. monocytogenes.

The mechanism behind these adaptive responses seemed to be non-specific as similar changes were observed not only under acid stress but also under osmotic stress or in presence of various of bacteriocins.

The effect of benzoic or acetic acid on L. monocytogenes phospholipid composition

Until now there has been no previous reports of the type of phospholipid subspecies present in LmBA or LmAA cells. Interestingly the relative quantity of the four phospholipids CL, PG, PhAL and PI was dramatically affected by benzoic acid (Table 2; Fig. 3). A decrease was observed in the net negative charge of the lipid bilayer of cells. Related data were reported by other authors (Crandall and Montville 1998) for a decrease of total anionic (CL + PG) and an increase of the zwitterionic phosphatidylethanolamine in the resistance to nisin. Additionally neutralisation of the negative head group charges may induce a condensation of these phospholipids resulting in a more rigid membrane.

Unlike LmBA the relative quantity of four phospholipids of L. monocytogenes was not dramatically affected in response to acetic acid at low pH.

Other authors (Brown and Booth 1991) reported that an increase in synthesis of aminophospholipids was observed in Staphylococcus aureus during incubation at low pH. Such changes provide the membrane surface with a net positive charge, which may act as a barrier to protons. Furthermore, the synthesis of lysyl-PG has been implicated in resistance against cationic antimicrobial peptides (CAMPs) (Peschel et al. 2001). The synthesis of aminoacylated PG derivatives has been described in several microorganisms. Recently there has been fresh interest in these processes. Klein et al. (2009) found that approximately 0.5% alanyl-PG content in the membrane composition is sufficient to significantly increase the acidic resistance of P. aeruginosa. Toke (2005) reported that the effects of pH and CAMP resistance are linked. This means that ‘tightening’ the cell membrane against protons may also render the membrane structure more resistant to CAMPs, to the beta-lactam cefsulodin, but also to additional compounds like Cr3+ ions or sodium lactate (Klein et al. 2009). Additionally, Lopez et al. (2006) investigated the importance of anionic phospholipid content (CL and PG) in the osmotic adaptation and in the membrane structure of B. subtilis. Since changes in the lipid headgroup region of lipid bilayers could lead to modifications in the structure of transmembrane alpha-helices of proteins and as a result, changes in the packing of these helices, it is important to maintain the presence of charged lipids in the bilayer (Lopez et al. 2006).

It has become clear in recent years that bacteria in nature do not prevalently exist in free floating cells but as biofilms attached to surfaces (Costerton 1995; Kolter and Losick 1998). Moreover some microbiologically produced, surface active compounds such as lipids and polysaccharides may influence the interactions of bacterial cells with interfaces (Chamberlain 1992). The increased amount of long chain FA, the accumulation of neutral lipids and the changes in polar head groups of phospholipids observed in our work will affect the physicochemical surface properties of cells. Gianotti et al. (2008) reported dramatic differences between the FA composition of adhered cells of two L. monocytogenes strains (acid adaptive and non-adaptive). It is reasonable to presume that a major proportion of straight FAs is required for the interaction of phospholipids, or their specific loci, with the cell surface proteins that mediate the contact of the cell envelopes with abiotic surfaces (Gianotti et al. 2008). Moreover quorum-sensing allows bacterial cells to collaborate in a way that benefits an entire biofilm population (Baysse et al. 2005; Tarver 2009). Interestingly other authors (Baysse et al. 2005) have shown that the quorum-sensing system of P. aeruginosa can also be activated by nutritional stress, independently of cell density, and therefore may be part of a more general adaptive response to stressful environmental conditions.

Conclusions

Our data points to a role for an acid adaptation response of L. monocytogenes, irrespective of pH, in triggering a significant increase of straight chain FA. This increase is reflected mainly in the NL class FA of acid adapted cells. Additionally our results showed that only low pH enhances the antimicrobial activity of an acid and the mechanism behind this adaptation response includes NL accumulation in the cells. The nature of the newly-synthesised neutral lipids remained to be determined. Benzoic acid has a slight antimicrobial activity and the percentage of NL and PL remained unchanged, but the relative quantities of four phospholipid subspecies were dramatically affected, while these alterations were not observed in acetic acid adapted cells. Such changes in the lipid profile of the plasma membrane alter membrane permeability and fluidity, which may in turn contribute to tolerance, decreasing the ability of weak acid preservatives to pass through the membrane and thus conferring protection.

The above observations shed further light on how L. monocytogenes is able to adapt its physiology to allow growth in acidic environments. FA biosynthesis is co-ordinately regulated with NL synthesis and growth, as part of the bacterial response to a changing environment and offers unique sites for selective inhibition to combat microbial infection (chemotherapeutic agents). Understanding microbial stress response mechanisms will improve the effective use of preservative factors, help manipulate the survival and growth of foodborne microorganisms, and exert microbial control at these points.