Variations in the Degree of d-Alanylation of Teichoic Acids in Lactococcus lactis Alter Resistance to Cationic Antimicrobials but Have No Effect on Bacterial Surface Hydrophobicity and Charge

Abstract

An increase of the degree of d-alanylation of teichoic acids in Lactococcus lactis resulted in a significant increase of bacterial resistance toward the cationic antimicrobials nisin and lysozyme, whereas the absence of d-alanylation led to a decreased resistance toward the same compounds. In contrast, the same variations of the d-alanylation degree did not modify bacterial cell surface charge and hydrophobicity. Bacterial adhesion to polystyrene and glass surfaces was not modified either.

The gram-positive cell wall is formed by a thick peptidoglycan layer decorated with proteins, polysaccharides, and mainly polymers of alternating phosphate and alditol groups called teichoic acids (5). Teichoic acids are either covalently linked to the peptidoglycan (wall teichoic acids) or anchored to the membrane through a glycolipid (lipoteichoic acids [LTAs]) (20). In many gram-positive bacteria, the products of the dlt operon genes are involved in d-alanylation of teichoic acids (20, 24). Mutants with teichoic acids lacking d-alanine esters have been shown to exhibit a variety of phenotypic changes, especially altered resistance to cationic antimicrobials (13, 15, 23, 25) and modified adhesion and biofilm formation (1, 8, 10, 30).

Lactococcus lactis is widely used in dairy fermentations and also serves as a model organism for biological studies of lactic acid bacteria. In L. lactis, the characterization of dlt mutants has revealed the role of teichoic acid d-alanylation in UV sensitivity, autolysis, and protein secretion (7, 21, 27). The aim of this study was to examine the impact of d-alanylation of L. lactis teichoic acids on resistance to cationic antimicrobials and, in parallel, on the physicochemical properties of the bacterial surface and on bacterial adhesion to solid surfaces.

Overexpression of the dlt operon and d-alanine quantification of teichoic acids.

The genetic organization of the dlt operon in L. lactis MG1363 is depicted in Fig. 1. To overproduce the dlt genes in L. lactis, we used the nisin-controlled expression system (16). The dlt operon was placed under the control of the nisA promoter (6). For this, a 4,274-bp fragment encompassing the complete dlt operon and starting at the ATG codon of dltA was amplified by PCR using primers U-DLT-NcoI (CATGCCATGGGAAAATTATTAGACAAATTTTTG [NcoI site underlined]) and l-DLT-KpnI (GGGGTACCTTATTTTAAGACAGACTCTG [KpnI site underlined]), with MG1363 DNA as the template and with TripleMaster Taq polymerase (Eppendorf). The PCR product was cloned downstream of the nisA promoter into the pNZ8048 plasmid (obtained from NIZO, The Netherlands Institute for Dairy Research, Ede, The Netherlands) (14), yielding plasmid pNZdlt. The recombinant plasmid was used to transform L. lactis NZ9000, a derivative of MG1363 with nisRK genes integrated into its chromosome, by electroporation (Table 1). In this host, nisin can activate the nisA promoter via signal transduction mediated by a two-component regulatory system composed of histidine kinase NisK and response regulator NisR (16).

FIG. 1.

Genetic organization of the dlt operon in L. lactis subsp. cremoris MG1363 (NCBI database accession number NC_009004). Open reading frames corresponding to the four genes of the dlt operon are indicated with arrows. The dltA (1,499-bp), dltB (1,223-bp), dltC (239-bp), and dltD (1,286-bp) genes encode the d-alanine:d-alanyl carrier ligase DltA, the d-alanyl carrier protein DltC, and the membrane proteins DltB and DltD, respectively. P_DLT, putative promoter; stem-loop structure, putative transcriptional terminator.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Relevant properties or description	Reference or source
L. lactis strains
MG1363	Plasmid-free and prophage-cured derivative of L. lactis NCDO712	9
MG1363dltD	L. lactis MG1363 dltD mutant by ISS1 transposition	7
NZ9000	L. lactis MG1363 pepN::nisRK	16
NZ9000(pNZdlt)	L. lactis NZ9000 containing the pNZdlt plasmid; Cmr	This study
NZ9000(pNZ)	L. lactis NZ9000 containing the empty pNZ8048 plasmid; Cmr	This study
Plasmids
pNZ8048	Gene expression vector with nisin-inducible PnisA promoter; Cmr	16
pNZdlt	Plasmid pNZ8048 containing the dlt operon under nisin-inducible promoter; Cmr	This study

The amount of dlt transcript in strain NZ9000(pNZdlt) after induction with nisin (1 ng/ml) (Sigma) was measured by real time reverse transcription-PCR. RNA extraction and cDNA synthesis were performed as described previously (12). Primers (GTTCTCGGTTCGTCAGAAATGG and TAATGTCGTTGTTCCGGGTTG) located inside the dltD gene were designed using Primer Express software from Applied Biosystems. Real-time PCR was carried out using Sybr green PCR master mix (Applied Biosystems) as recommended by the supplier. Reactions were run in duplicates with a Mastercycler ep realplex instrument (Eppendorf). The cycle threshold was used to determine the relative dltD gene expression levels. Data were computed using the comparative critical threshold (2^−ΔΔCT) method. The results were normalized with the L. lactis tuf gene, encoding the elongation factor TU, as a control as described previously (26). The results indicated a 50-fold increase of the amount of dlt transcript in the NZ9000(pNZdlt) strain compared to the levels for the control strain NZ9000(pNZ) after nisin induction.

The amount of d-alanine esterified to teichoic acids was measured for the overexpressing strain NZ9000(pNZdlt) as well as for the dltD-negative mutant MG1363dltD, obtained previously by Duwat et al. (7). Bacteria were grown overnight at 30°C in M17-glucose broth supplemented with chloramphenicol (5 μg/ml) and nisin (1 ng/ml) for NZ9000(pNZdlt) and the control strain NZ9000(pNZ). d-Alanine was released from whole cells by alkaline hydrolysis, as reported previously for Streptococcus pneumoniae (13) and group A Streptococcus (15), and quantified by high-performance liquid chromatography after derivatization with Marfey's reagent (1-fluoro-2,4-dinitrophenyl-5-l-alanine amide; Sigma) (28). The results showed that a larger amount of d-alanine was released from the dlt overexpressing strain NZ9000(pNZdlt) than from the control strain NZ9000(pNZ) (1.5-fold-larger amount) (Fig. 2). On the other hand, almost no d-alanine was released from the dltD mutant. The lack of d-alanylation in dlt mutants of a number of gram-positive pathogens has also been reported (1, 8, 15, 25).

FIG. 2.

Amounts of d-alanine released from whole cells by alkaline hydrolysis for L. lactis mutant strain MG1363dltD, dlt-overexpressing strain NZ9000(pNZdlt) (filled bars), and their respective control strains, MG1363 and NZ9000(pNZ) (open bars). Data represent the mean values ± standard deviations of results from three independent experiments.

Sensitivity to cationic antimicrobials.

In the absence of d-alanine substitutions, teichoic acids are expected to bear higher negative charges, since the positively charged amino groups of d-alanyl esters partially counteract the negative charges of the backbone phosphate groups (20). In several bacterial species, dlt-deficient mutants have been found to be more sensitive to cationic antimicrobials (13, 15, 23, 25). In this study, the MICs for nisin and lysozyme (pI values of nisin and lysozyme, 10.5 and 11.35, respectively) were determined for the four strains, using the optical density modeling method which was described by Guillier et al. (11). The strains NZ9000(pNZdlt) and NZ9000(pNZ) were precultured twice in the presence of chloramphenicol. Then, bacterial growth was monitored in medium without chloramphenicol, using a Bioscreen C apparatus with microtiter plates, in the presence of 10 different concentrations of each antimicrobial. Afterwards, growth rate was estimated by fitting the data to the modified Gompertz model (2) and a dose response curve was obtained by using the Lambert-Pearson model (17), leading to MIC determination. The MIC of nisin for the dltD-negative mutant was reduced compared to that for MG1363, as previously reported by Kramer et al. (14), which confirms the involvement of teichoic acid d-alanylation in nisin resistance in L. lactis (Fig. 3A). Similarly, the MIC of lysozyme for the dlt-deficient mutant was greatly reduced compared to that for MG1363 (Fig. 3B). In contrast, the dlt-overexpressing strain NZ9000(pNZdlt) was more resistant to both nisin and lysozyme than the control strain NZ9000(pNZ) as well as the wild-type strain MG1363 (Fig. 3A and B).

FIG. 3.

MICs of nisin (A) and lysozyme (B) determined for L. lactis mutant strain MG1363dltD, dlt-overexpressing strain NZ9000(pNZdlt) (filled bars), and their respective control strains, MG1363 and NZ9000(pNZ) (open bars). Strains were grown at 30°C in M17-glucose broth without chloramphenicol in microtiter plates with different concentrations of antimicrobial. Each bar represents the mean values ± standard deviations. The data are from two independent experiments, with two wells inoculated in parallel for each experiment.

Surface physicochemical properties and adhesion to solid surfaces.

In several previous studies, d-alanine deprivation of teichoic acids was hypothesized to result in a more negative cell surface (10, 15, 22). Besides, LTA is recognized as a main determinant of cell surface hydrophobicity (19). However, in these studies, the bacterial surface properties were not investigated with physicochemical methods. In our study, first we examined the hydrophobic/hydrophilic cell surface properties of the dlt mutant strains and their Lewis acid-base characteristics with the microbial-adhesion-to-solvents method, described by Bellon-Fontaine et al. (3). Strains were compared for their affinities to two pairs of solvents: hexadecane-chloroform and decane-ethyl acetate. The results for the microbial-adhesion-to-solvents method did not reveal any significant differences between the dltD mutant, the overexpressing strain, and the wild-type or the control strain as regards their surface hydrophobicities and polarities (data not shown). Low affinity to apolar solvents (decane and hexadecane) (around 10%) revealed the hydrophilic character of the surface of MG1363 as well as the three other strains. Second, the electrical properties of the bacterial surface were assessed by electrophoretic mobility measurements, as previously described (4). Surprisingly, no significant differences at the overall net surface charge were observed between the dlt mutant and MG1363 or between the overexpressing strain and the control strain (Fig. 4). These data reveal that the variations of the degree of d-alanine substitution of L. lactis teichoic acids, at least in the range tested in this study, do not modify the global bacterial surface charge.

FIG. 4.

Electrophoretic mobility (E.M.) (10⁻⁸ m² s⁻¹ V⁻¹) curves for (A) the L. lactis MG1363dlt mutant strain and wild-type MG1363 and (B) dlt-overexpressing strain NZ9000(pNZdlt) and control strain NZ9000(pNZ). Bacteria were suspended in 1.5 mM NaCl at pH values from 2 to 7 after overnight growth at 30°C in M17-glucose broth (chloramphenicol and nisin were added when required). The typical standard deviation for the electrophoretic mobility mean was 0.3 × 10⁻⁸ m² s⁻¹ V⁻¹.

Furthermore, we examined whether the altered d-Ala content of teichoic acids could influence L. lactis interactions with solid surfaces. We tested the adhesion of bacteria to polystyrene microtiter plates by using the method described by van Merode et al. (29) and also to glass slides (25 by 25 mm) by quantifying the number of attached cells through acridine orange staining and epifluorescence microscopy, as previously described (18). No significant differences in adhesion to hydrophobic polystyrene and hydrophilic glass were observed between the strains (data not shown). This result correlates with the absence of differences in cell surface physicochemical properties between the different strains tested.

In sum, the present results show a correlation between the degree of d-alanylation of teichoic acids in L. lactis and the resistance to cationic antimicrobials, in agreement with the data obtained for several other gram-positive bacterial species (13, 15, 23, 25). However, no modification of the bacterial surface physicochemical properties was observed in relation to the teichoic acid d-alanylation degree in the range tested. In L. lactis MG1363, the ratio of d-Ala to glycerol-phosphate (GroP) on LTA was estimated to be 28.5% (27). Taking into account this value and the results of d-Ala quantification shown on Fig. 2, we can estimate that the d-Ala/GroP ratios in LTAs of the tested strains ranged from 4% to 45%. The electronegative and hydrophilic characters of the surfaces of MG1363 and its derivative strains suggest that wall teichoic acids and/or LTAs are exposed at the cell surface and responsible for the global negative charge. However, the fact that the d-alanylation level does not modify the surface global charge suggests that, in L. lactis, the d-alanyl substituents of teichoic acids are located inside the cell wall rather than exposed at the cell surface. Nisin and lysozyme are small molecules that can diffuse inside the cell wall to exert their inhibitory activity, in the case of nisin at the level of the cytoplasmic membrane through lipid II binding (31). The d-Ala substituents on teichoic acids located inside the cell wall could reduce the accumulation of these cationic antimicrobials inside the cell wall and in the vicinity of the cytoplasmic membrane and thus could increase the bacterial resistance to these molecules.