Staphylococcus aureus requires cardiolipin for survival under conditions of high salinity
© Tsai et al; licensee BioMed Central Ltd. 2011
Received: 21 June 2010
Accepted: 18 January 2011
Published: 18 January 2011
Skip to main content
© Tsai et al; licensee BioMed Central Ltd. 2011
Received: 21 June 2010
Accepted: 18 January 2011
Published: 18 January 2011
The ability of staphylococci to grow in a wide range of salt concentrations is well documented. In this study, we aimed to clarify the role of cardiolipin (CL) in the adaptation of Staphylococcus aureus to high salinity.
Using an improved extraction method, the analysis of phospholipid composition suggested that CL levels increased slightly toward stationary phase, but that this was not induced by high salinity. Deletion of the two CL synthase genes, SA1155 (cls1) and SA1891 (cls2), abolished CL synthesis. The cls2 gene encoded the dominant CL synthase. In a cls2 deletion mutant, Cls1 functioned under stress conditions, including high salinity. Using these mutants, CL was shown to be unnecessary for growth in either basal or high-salt conditions, but it was critical for prolonged survival in high-salt conditions and for generation of the L-form.
CL is not essential for S. aureus growth under conditions of high salinity, but is necessary for survival under prolonged high-salt stress and for the generation of L-form variants.
Staphylococcus aureus is an opportunistic pathogen that causes a wide range of diseases in both immunologically normal and compromised hosts. The natural habitat of S. aureus is the nasal cavity of warm-blooded animals. Over the past ~50 years, S. aureus has undergone genetic changes that have resulted in antibiotic-resistant strains [1, 2]. Importantly, the methicillin-resistant strains (MRSA) are now the most common cause of nosocomial S. aureus infections and are spreading throughout communities .
Staphylococcus aureus has a number of characteristics that allow it to survive host bactericidal factors and environmental stresses, including drastic changes in osmotic pressure [4–6]. Osmoprotectants such as choline, glycine betaine, and proline accumulate in cells in response to osmotic stress [7–11]. Multiple genes, including the branched-chain amino acid transporter gene brnQ  and the arsenic operon regulatory gene arsR , cooperatively participate in salt tolerance. In addition, a very large cell wall protein, Ebh, is involved in tolerance to transient hyperosmotic pressure .
In general strategy, the phospholipid composition of bacteria changes in response to growth phase or environmental stressors such as osmolality , pH [16, 17], temperature, and the presence of organic solvents [18, 19]. In the 1970s, the molecular mechanism of staphylococcal salt resistance was studied, focusing on a phospholipid, cardiolipin (CL) . CL possesses four acyl groups and carries two negative charges . In stationary phase, 30% of the S. aureus cell membrane is composed of CL . It has been reported that CL can stabilize liposomes during osmotic stress  and that it is required for the growth of Escherichia coli and Bacillus subtilis under high-salt conditions [24, 25]. However, the role of CL in the molecular mechanism of staphylococcal resistance to high salinity remains unknown.
In this study, we used an improved lipid extraction method to assess the phospholipid composition of S. aureus and performed molecular genetic analyses to evaluate the role of CL in the resistance of S. aureus to high salinity.
No difference in susceptibility to antibiotics affecting cell walls (vancomycin, teicoplanin, cefarotin, cefmetazole, and cefazoline), quinolones (ofloxacin, norfloxacin, ciprofloxacin, and nalidixic acid), arbekacin, or the antimicrobial peptides ASABF-α  and nisin was observed between the N315 and its cls mutants (data not shown). The MIC of nisin for both S. aureus N315 and its cls mutants was 80 μg ml-1.
Cardiolipin is known to play a role in the adaptive mechanisms of some bacteria to high salinity stress [15, 20, 37]. For example, a deficiency in CL decreases the growth rate in B. subtilis under conditions of 1.5 M (8.76%) NaCl . Additionally, salt-sensitive S. aureus mutants contain no or only a small amount of CL [38, 39]. Therefore, we were surprised to find that the growth of S. aureus under conditions of high salinity did not depend on CL (Figure 6). This may be attributable to the presence of other mechanisms, including species-specific systems such as variations in cell wall proteins , that give staphylococci the ability to cope with high-salt stress [11, 40]. However, this study is, to our knowledge, the first to demonstrate that CL is important for long-term fitness of S. aureus under conditions of high salinity. This is an important finding in understanding the NaCl resistance of S. aureus, which is itself important for commensal growth on skin and mucus membranes, survival on dry surfaces during indirect transmission, and persistence in foods with a high salt content .
Cardiolipin depletion did not increase the susceptibility of S. aureus to cell wall-targeted antibiotics, suggesting that CL alone is not responsible for bacterial survival against these challenges. We also examined the susceptibility of S. aureus and its cls mutants to cationic antimicrobial peptides; because cardiolipin is negatively charged, a decrease in total negative charge on the membrane surface may contribute to reduced cationic antimicrobial peptide resistance . However, CL depletion had no effect on susceptibility to the antimicrobial peptides ASABF-α and nisin. It is possible that the net negative charge is compensated for by other membrane components such as PG. In fact, the PG level was increased in the mutants that did not accumulate CL. The importance of positively charged lysyl-phosphatidylglycerol (LPG) (or MprF protein) in resistance to cationic antimicrobial peptides has been reported [43, 44], and the LPG level was not different between wild-type S. aureus and cls mutants. In addition to the probable effect of cell surface charge, we have previously reported that cell wall thickness is an important factor affecting resistance to the antimicrobial peptide ASABF-α . Furthermore, in the present study, ASABF-α-resistant strains had cell walls that interfered with CL extraction (Additional file 1, Figure S1). Cell wall thickness may also be related to resistance against other antimicrobial peptides in S. aureus [45, 46].
Our data indicate that lysostaphin treatment is critical for the efficient extraction of CL from S. aureus. Previous reports have suggested that CL is not readily extractable from B. subtilis and other Gram-positive bacteria without lysozyme treatment . This may be attributable to its large molecular mass (~1500 Da) relative to that of other phospholipids, owing to its four acyl residues. However, ~25-kDa globular hydrophilic molecules can pass freely through the ~2-nm holes in the peptidoglycan polymer that forms the cell wall of Gram-positive bacteria . Instead, the efficiency of CL extraction is likely reduced by its physical interactions with cell wall components; for example, when CL is bound to cell wall components, it will not efficiently enter the organic phase during extraction.
The membrane of the L-form variants of S. aureus is thought to express certain features that support cell growth and survival in the absence of a rigid cell wall. One study reported that a particular L-form strain had an increased CL level . Our data demonstrate that the cls2 gene is important for normal L-form generation. However, the cls1/cls2 double mutant still produced L-form cells, suggesting the existence of a CL-independent mechanism. Thus, multiple mechanisms may function in cooperation to generate L-form variants. The production of a number of factors such as carotenoids, catalase, coagulase, lipase, fibrinolysin, hemolysin, and enterotoxin changes upon L-form generation and reversion [49–52]. However, none of these represents a common L-form variant phenotype, suggesting that L-form generation is associated with a drastic phenotypic conversion. The increase in CL content may be important, but not essential, for membrane stabilization.
In this study, both cls1 and cls2 were shown to encode functional CL synthases. Under our experimental conditions, cls2 was the dominant CL synthase gene. The cls1 mutant did not differ from the parental strain in its growth rate, survival at stationary phase, total CL accumulation, or L-form generation. However, our data indicate that the synthesis of CL by Cls1 helps the cls2 mutant to survive prolonged incubation under high-salt conditions (Figure 5E), suggesting that Cls1 has a specific function under stress conditions if Cls2 is unavailable. Future studies should examine the functional characteristics of these two CL synthases, including possible differences in their subcellular localizations.
Improved lipid extraction and molecular genetic analyses showed that both cls1 and cls2 participate in CL accumulation. The cls2 gene appears to serve a housekeeping function, while cls1 is active under stress conditions. Staphylococcus aureus can grow under conditions of high salinity without CL, but CL is required to survive prolonged high salinity stress and to generate L-form variants. This CL-dependent survival helps to explain the success of S. aureus as a human pathogen and skin/mucus membrane commensal.
Bacterial strains, plasmids, and primers used in this study
Strain or plasmid
S. aureus strain
8325-4 derivative that accept foreign DNA
RN4220 Δ SA1155; Cmr
RN4220 Δ SA1891; Tetr
RN4220 Δ SA1155/ Δ SA1891; Cmr Tetr
pre-methicillin resistant strain
N315 Δ SA1155; Cmr
N315 Δ SA1891; Tetr
N315 Δ SA1155/ Δ SA1891; Cmr Tetr
NCTC 8325 cured of prophages, rsbU (-)
8325-4 Δ SA1155; Cmr
8325-4 Δ SA1891; Tetr
8325-4 Δ SA1155/ Δ SA1891; Cmr Tetr
8325-4 derivative, rsbU repaired, rsbU (+)
SH1000 Δ SA1155; Cmr
SH1000 Δ SA1891; Tetr
SH1000 Δ SA1155/ Δ SA1891; Cmr Tetr
isolate from healthy host, able to generate L-form
MT01 Δ SA1155; Cmr
MT01 Δ SA1891; Tetr
MT01 Δ SA1155/ Δ SA1891; Cmr Tetr
thermosensitive, pE194ts-based delivery vector
pMAD-based SA1155 targeting vector
pMAD-based SA1891 targeting vector
shuttle vector, tet (TetR)
shuttle vector, cm (CmR)
CATATT GTCGAC TAAGTGATGAAATACTG
G GAATTC CTGTTATAAAAAAAGGATCAAT
CGA GTCGAC GATAAAGTGGGATATTTT
C GAATTC CGGGGCAGGTTAGTGACATT
T GGATCC TGATATTGCTTACATACT
TGG GTCGAC AAAAAGTACAAATAGC
CAC AGATCT TATGGACTTTAGAAGTT
TTT AGATCT CAATATCATCCAAATTA
C GGATCC AATAGTCCGACGATAGCT
GAA GTCGAC GTCCTAATAGTAAGTA
T GAATTC ACAAAAGCACGTTATGCT
TGA AGATCT AACATCACAACGGCATA
Cells equivalent to 8 × 108 CFU were collected at exponential or stationary phase, washed in 2% NaCl, and resuspended in 200 μl of 2% NaCl. Lysostaphin (0.1 mg ml-1) was added, and the mixture was incubated at 37°C for 3 min. The lysed cell suspension was extracted with chloroform-methanol. Briefly, a five-fold volume of chloroform-methanol (2:1; v/v) was added, mixed vigorously for 2 min, and left at room temperature for 10 min. Following the addition of a three-fold volume of chloroform-2% NaCl (1:1; v/v) and centrifugation, the lower layer was recovered and concentrated under vacuum. The lipids were dissolved in chloroform-methanol (2:1; v/v), applied to silica thin-layer chromatography (TLC) plates (Silica gel 60; Merck, Darmstadt, Germany), and developed with chloroform-methanol-acetic acid (65:25:10; v/v/v). The TLC plates were sprayed with CuSO4 (100 mg ml-1) containing 8% phosphoric acid and heated at 180°C to visualize the phospholipids. A digital image was obtained using a scanner, and signal intensities were quantified by ImageJ software (version 1.43U, NIH).
To construct S. aureus mutants, we used pMAD, which is designed for efficient allelic replacement in Gram-positive bacteria , and incorporated the tetracycline or chloramphenicol resistance gene to make it suitable for use with erythromycin-resistant hosts such as S. aureus N315. The tet gene was amplified with primers Tet-F(Sal) and Tet-R(EcoRI) using pHY300PLK (Takara Shuzo Co. Ltd., Kyoto, Japan) as a template and ligated into the Eco RI- Sal I site of pMAD to generate pMADtet. The cat gene was amplified from pRIT5H  with primers CAT(Sal)F and CAT(Eco)R and ligated into the Eco RI- Sal I site of pMAD to generate pMADcat.
Target vectors were designed to replace the SA1155 (cls1) and SA1891 (cls2) genes with cat and tet, respectively. Two regions encompassing SA1155 were amplified with the primer pairs clsU1p and clsU2p (upstream region) and clsD1p and clsD2p (downstream region), restricted at the primer-attached sites, and sequentially ligated into the Bam HI- Sal I and Bgl II sites of pMADcat to generate the target plasmid pMADcat1155. Similarly, the upstream and downstream regions of SA1891 were amplified with the primer pairs 1891U1 and 1891U2, and 1891D1 and 1891D2, and then sequentially ligated into the Bam HI- Sal I and Eco RI- Bgl II sites of pMADtet to generate pMADtet1891. These target vectors were introduced into S. aureus RN4220 and N315 by electroporation. Each mutant was isolated as described previously . Briefly, β-galactosidase-positive colonies carrying the target vector were plated on TSB agar (TSA) containing antibiotic (12.5 μg ml-1 Cm or 5 μg ml-1 Tet) and 100 μg ml-1 X-gal, followed by incubation at 42°C overnight. Several resulting blue colonies were pooled and subjected to three cycles of growth in drug-free TSB at 30°C for 12 h and at 42°C for 12 h. Dilutions were plated on drug-free TSA plates containing 100 μg ml-1 X-gal. Homologous recombination in white colonies was detected by PCR and Southern blot analyses. The SA1155/SA1891 double mutants of RN4220 and N315, the SA1155 and SA1891 single mutants, and the SA1155/SA1891 double mutants of SH1000, 8325-4, and MT01 were obtained by phage transduction. The absence of the genes in each mutant was confirmed by Southern blot analysis and/or PCR.
Cells were grown overnight in 5 ml of drug-free Muller-Hinton (MH) broth at 37°C with shaking (180 rpm; BR-1; TAITEC). These cells were diluted with MH (×10-4) and plated onto MH agar. Antibiotic susceptibilities of the strains were compared using the disk diffusion method (BD BBL sensi-Disk; Becton, Dickinson and Co., Franklin Lakes, NJ). The susceptibilities to ASABF-α were measured as described previously . The minimum inhibitory concentration (MIC) of nisin (from Lactococcus lactis; Sigma, St. Louis, MO) was determined by microdilution with 104 cells per well and a 20-h incubation at 37°C.
Cells were cultured in BHI without antibiotics, and 100 μl of the overnight culture were spread onto BHI agar plates containing 5% NaCl, 5% sucrose, 10% heat-inactivated horse serum, and 100 μg ml-1 penicillin. The presence of serum selects for the stable L-form of S. aureus . The plates were incubated at 37°C, and colonies showing the L-form ('fried egg shape') were counted for 8 days post-inoculation .
We thank Dr. Michel Débarbouillé (Institut Pasteur, CNRS) for providing the pMAD vector.
This study was supported by The Sasakawa Scientific Research Grant from The Japan Science Society 21-421 to MT, JSPS Grant-in-Aid for Scientific Research (C) 17590405 to HH, The Ichiro Kanehara Foundation 09KI202 to KM, and The Salt Science Research Foundation 0826 to KM.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.