Alternative cardiolipin synthase Cls1 compensates for stalled Cls2 function in Staphylococcus aureus under conditions of acute acid stress


Correspondence: Kazuya Morikawa, Faculty of Medicine, University of Tsukuba, Tsukuba 305-8575, Japan. Tel./fax: +81 29 853 3928; e-mail:


Staphylococcus aureus possesses two distinct cardiolipin (CL) synthase genes, cls1 and cls2. It was previously shown that cls2 encodes a housekeeping-type CL synthase. However, the role of cls1 is elusive; a cls1 mutant was found to be equal to the wild type in terms of CL accumulation and stress tolerance. Here, we report that the physiological role of cls1 is to synthesize CL under conditions of acute low-pH stress. Below pH 2.6, the cls1 mutant (i.e. carrying Cls2 alone) could not produce CL, while the cls2 mutant (carrying Cls1) effectively accumulated CL. The cls1-dependent CL production was quick (within 5 min) and did not require de novo protein synthesis. Together with the results of phylogenetic analyses, our findings suggest that cls1 was generated through the duplication of cls2 after the divergence of the genus Staphylococcus and that the alternative CL synthase encoded by this gene confers improved survival in the face of acute acid stress.


Staphylococcus aureus is a Gram-positive bacterium that naturally inhabits the nasal cavity of warm-blooded animals. It has a number of characteristics that allow it to survive host bactericidal responses and stressors associated with the surface environment, including drastic changes in osmotic pressure (Clements & Foster, 1999; Garzoni & Kelley, 2009; Morikawa et al., 2010). It is also an opportunistic pathogen that causes a wide range of diseases in both immunologically normal and compromised hosts. Importantly, methicillin-resistant strains (MRSA) are now the most common cause of nosocomial S. aureus infections and are spreading throughout communities (Chambers & Deleo, 2009).

In general, the phospholipid composition of bacteria changes in response to the growth phase or environmental stressors such as osmolality (Romantsov et al., 2009), pH (Gould & Lennarz, 1970; Minnikin & Abdolrahimzadeh, 1974), temperature and the presence of organic solvents (Ramos et al., 2002; Bernal et al., 2007). The major phospholipid in logarithmic-phase staphylococcal cells is phosphatidylglycerol (PG). PG is converted to cardiolipin (CL) during cell growth, and it constitutes 30% of the cell membrane in stationary-phase cells (Short & White, 1971). CL, which possesses four acyl groups and carries two negative charges (Schlame, 2008), can stabilize liposomes against osmotic stress (Nagamachi et al., 1992). In 1970s, biochemical studies indicated that CL was induced under conditions of high salt. Recently, we reported that CL is dispensable for growth under high salinity, but is essential for long-term survival under high salt conditions, suggesting that membrane composition needs to be modulated to adapt to conditions of high salinity (Tsai et al., 2011).

In S. aureus, two CL synthase genes, cls1 and cls2, are responsible for CL synthesis (Koprivnjak et al., 2011; Tsai et al., 2011). A previous molecular genetic study indicated that cls2 encodes the major CL synthase that is responsible for CL accumulation under both normal and high salt conditions. In contrast, the absence of cls1 had no significant effect on CL accumulation under the experimental conditions employed (Tsai et al., 2011). In addition, the cls1 mutant exhibited no difference from the wild type (WT) in any of the tested phenotypes, including growth rate, salt resistance and L-form generation (Tsai et al., 2011). These results raised the question why S. aureus has cls1 in addition to the housekeeping gene cls2. Koprivnjak et al. (2011), and we found that CL synthesis by cls1 is responsive to stress: CL production in a cls2 mutant was induced during culture in high salt (15% and 25% NaCl), at a moderately low pH (pH 5.0), under anaerobic conditions (Tsai et al., 2011), and during phagocytosis by polymorphonuclear leucocytes (Koprivnjak et al., 2011). In the present study, we aimed to clarify the stress responsive role of cls1, and we explored the conditions under which cls1, but not cls2, is exclusively responsible for CL synthesis.

Materials and methods

Database search and phylogenic analysis

We used the FASTA search algorithm to examine the genomes of 30 bacteria whose genome projects have been completed. Cls homologues were downloaded from the KEGG database (Kanehisa et al., 2002). The amino acid sequences of the Cls homologues obtained from our FASTA search were aligned using the clustalx program (Jeanmougin et al., 1998). The alignment was used for phylogenic analysis with the protdist and neighbour programs of the phylip 3.6 package (Retief, 2000). The phylogenic tree was inferred by the neighbour-joining method (Saitou & Nei, 1987) and tested by 100 replications of bootstrap analysis, which was carried out using the seqboot and consense programs and visualized using the treeview program (Page, 1996).

Bacterial strains and culture conditions

The S. aureus strains used in this study were N315, Ncls1 (N315 ΔSA1155; Cmr), Ncls2 (N315 ΔSA1891; Tetr) and Ncls1/cls2 (N315 ΔSA1155/ΔSA1891; Cmr, Tetr) (Tsai et al., 2011). Luria–Bertani (LB) broth was used as the basic culture medium. Cells were precultured at 37 °C overnight with shaking (180 r.p.m.; BR-15: TAITEC, Tokyo, Japan). This culture (50 μL) was inoculated into 5 mL of LB at 37 °C with shaking (180 r.p.m.; BR-15). Logarithmic-phase cells were collected at an OD600 of 0.3. Cells from an overnight culture were harvested 15 h after inoculation from a glycerol stock. To inhibit transcription/translation, cells were treated with 100 μg mL−1 rifampicin and 100 μg mL−1 chloramphenicol for 60 min prior to harvesting.

Stress assays

Cells equivalent to 8 × 108 colony-forming units (CFU) were collected at the logarithmic or stationary phase, washed with PBS and suspended in high osmotic or acid solutions. The high osmotic solutions were 4 M NaCl, 4 M KCl and 20% raffinose. The acid solutions were 10 mM HCl (pH 2.0) and citrate-phosphate buffer (pH 2.6, 4.6 or 6.6) supplemented with 100 mM NaCl and 10 mM KCl. After incubation for 5, 15 or 60 min, the cells were washed with PBS and collected for subsequent viability testing (CFU counting) and thin-layer chromatography (TLC). For heat- or cold-shock treatment, cultures containing 8 × 108 CFU were directly shifted to the appropriate temperature.

Lipid extraction and TLC

Lipid extraction and TLC were carried out as described previously (Tsai et al., 2011). Cells equivalent to 8 × 108 CFU were washed with PBS and resuspended in 200 μL of 2% NaCl. Lysostaphin was added to the cell suspension (final concentration 0.1 mg mL−1) and incubated at 37 °C for 3 min. The lysed cell suspension was then subjected to chloroform–methanol extraction. Lipids were dissolved in chloroform–methanol (2 : 1; v/v), applied to silica 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 100 mg mL−1 CuSO4 containing 8% phosphoric acid and heated at 180 °C to detect phospholipids. A digital image was obtained by a scanner, and the signal intensities were quantified using Image J software (version 1.44p; NIH).


The origin of the two cls genes

The number of CL synthase genes varies among bacterial species (Supporting Information, Table S1). Staphylococcal cls1 (SA1155) and cls2 (SA1891) share higher levels of similarity with each other than with cls genes from other species. They were grouped with Bacillus subtilis cls (BSU36590) and Listeria monocytogenes lmo2503, but not with B. subtilis ywjE (BSU37190) and ywiE (BSU37240) or L. monocytogenes lmo0008 (Fig. S1). This indicates that the two staphylococcal cls genes were not acquired by horizontal gene transfer from different species. We found a single insertion/deletion (INDEL) site in the N-terminal region of Cls (Fig. S1). The INDEL in Cls2 is considered to be the ancestral type because it is shared with the Cls of other bacterial species. This suggests that cls1 was generated through the duplication of cls2, after the divergence of the genus Staphylococcus.

Acid stress rapidly increases the CL level in cls2 mutant cells

In our previous study, we failed to identify the conditions under which Cls1 plays a major role in CL synthesis. The previously tested conditions were high salinity, continuous culture at a low/high temperature (30 and 42 °C), mildly acidic conditions (pH 5.0) and anaerobiosis (Tsai et al., 2011). Here, we further explored the conditions under which Cls1 plays a dominant role in CL production, and we tested the effect of stressors that would physically alter the cell membrane. The tested conditions were a temperature shift (from 37 to 0, 4, 30, 42 and 48 °C over 15 min), antibiotic treatment (at the MIC of oxacillin, vancomycin and nisin for 15 min), high osmotic pressure and acid stress. Our results indicated that the temperature shift and antibiotics did not affect CL accumulation in the tested strains (data not shown).

Treatment with 4 M NaCl, 4 M KCl or 20% raffinose induced CL accumulation in the cls2 mutant (Ncls2), although the effect of 4 M KCl was relatively weak. This suggests that Cls1 can induce CL production in response to broad high osmolality stressors (Fig. 1). However, the CL level did not change significantly in WT and Ncls1 cells under conditions of high osmotic pressure. We found that a low pH (4.6, 2.6 and 2.0) induced CL accumulation in Ncls2 cells (Fig. 1) more efficiently than mildly acidic conditions (pH 5.0: Tsai et al., 2011). The low-pH response was faster (< 15 min) than the osmotic stress response (Fig. 1). Importantly, the CL level in Ncls1 did not increase after 15 min of exposure to a pH of 2.6 or 2.0, resulting in a statistically significant difference compared with S. aureus N315 cells. This suggests that Cls2 function is impaired by this type of low-pH treatment. Cells of both types from overnight (Fig. 1a and b) and logarithmic-phase (Fig. 1c and d) cultures exhibited a similar tendency.

Figure 1.

Phospholipid composition of Staphylococcus aureus N315 and the cls mutants under conditions of high osmotic pressure and low pH. Cells harvested from an overnight culture (a and b) or logarithmic-phase culture (c and d) were challenged with osmotic or low-pH stress for 15 min. (a and c) Phospholipid compositions were analysed by TLC. LPG, lysyl-phosphatidylglycerol. (b and d) The means and standard deviations of the relative CL intensities on the thin-layer chromatograms were calculated from three independent experiments. N315: diamonds, cls1 mutant: circles, cls2 mutant: squares. *Significant differences as determined by Student's t-test (< 0.05).

Survival under conditions of acute low-pH stress requires cls1 but not cls2

The cls1 mutant exhibited 100-fold increased susceptibility in the logarithmic phase upon a sudden change in pH from 7.4 to 2.6 (Fig. 2a, log phase). The cls1/cls2 double mutant was 10-fold more susceptible compared with the cls1 mutant, but the survival of the cls2 single mutant was equal to that of the WT. Namely, survival against acute acid stress depends largely on cls1 and does not rely on cls2 when cls1 is available. The importance of cls1 in acute acid stress was also observed in an overnight culture, but the difference was not statistically significant. Acute acid stress is the first condition under which cls1 has been found to be physiologically important for S. aureus survival: the cls1 mutant was equal to the WT in terms of long-term survival under conditions of high salinity and susceptibility to antibiotics and antimicrobial peptides (Tsai et al., 2010, 2011), as well as extended incubation at pH 4.6 (Fig. 2b).

Figure 2.

Cell survival under acidic conditions. (a) Cells recovered from logarithmic-phase or overnight cultures were exposed to a pH of 2.6 for 15 min. The means and standard deviations of the CFU values are shown (n ≥ 3). (b) Cells were suspended in buffer (pH 4.6), and the CFU were measured for 6 h (n ≥ 4). *Significant differences as determined by Student's t-test (< 0.05).

Stress responsive CL accumulation by cls1 cells does not require de novo protein synthesis

We noticed that the increase in CL at a low pH in cls1 was very fast – within 5 min (Fig. 3). The rapid accumulation of CL in the cls2 mutant was observed in nutrition-free solutions. To gain insight into this quick response, we tested whether it requires de novo protein synthesis. Cells were treated with an excess concentration of rifampicin and chloramphenicol to inhibit transcription and translation, respectively and then exposed to a low pH. Analysis by TLC showed that the increase in CL in Ncls2 was unaffected by treatment with these inhibitors (Fig. 3).

Figure 3.

Low-pH responsive CL synthesis by Cls1 (in Ncls2) does not require de novo protein synthesis. Cells were pretreated with both rifampicin and chloramphenicol then exposed to low pH for 5 or 15 min. Ncls1 and Ncls1/cls2, which are resistant to chloramphenicol, served as a reference; N315 and Ncls2 are susceptible to the drug. rif, rifampicin; cm, chloramphenicol.


In the present study, we first showed that Cls1 compensates for the stalled function of Cls2 under conditions of acute low-pH stress. This response did not require de novo Cls1 synthesis, suggesting that Cls1 is equipped with a backup system that can respond swiftly to such an emergency.

In the human body, low-pH conditions play a protective role against pathogens. In a fasting stomach, the pH is 1–1.5, which is a strong barrier against incoming bacteria. The acidic environment of the vagina (˜pH 4) is maintained by commensal Lactobacillus spp. (Dover et al., 2008). Also, the surface of the skin is enriched with various organic acids, including propionic acids, lactic acid and pyruvic acid produced by host cells and the cells of the microbiota (Holland, 1993). Quick drying of the skin concentrates these organic acids, leading to a sudden acid shock. In macrophages, engulfed bacteria are challenged by a series of bactericidal factors, including acidification in the phagosome lumen (pH 5). Staphylococcus aureus, as a commensal bacterium and opportunistic pathogen, is occasionally challenged by an acidic environment; however, it is capable of increasing its acid tolerance through its Cls1 backup system.

Membrane composition can significantly affect cell survival in response to acid exposure. In Streptococcus mutans, an increase in monounsaturated fatty acids is important for acid adaptation (Fozo & Quivey, 2004). Furthermore, the same group recently reported that CL is a reservoir for monounsaturated fatty acids, and they showed that a cls mutant of S. mutans was acid-sensitive (Macgilvray et al., 2012). Consistent with this, CL in S. aureus is also important for acid resistance (compare with wild-type cells vs. the Ncls1/cls2 double mutant in Fig. 2). An important difference is that in S. mutans CL synthesis depends on a single Cls, while S. aureus has a Cls1 backup system in addition to the housekeeping gene cls2.

The present study raises a number of questions regarding the Cls1 backup system, including how Cls2 is inactivated by a low pH and how Cls1 function is initiated. Future studies should focus on the subcellular localization of these proteins, the optimal pH for enzymatic activity and activity control through specific modifications. It is also important to address why other types of stress induce Cls1-dependent CL synthesis. In the present study, we tested the effect of ‘single’ stressors on Cls1 function; however, in a natural environment, multiple stressors assault S. aureus simultaneously (e.g. acid stress and oxidative stress in the phagosome). The present study does not limit the function of the Cls1 backup system to acute low-pH stress.


This study was supported in part by the Program to Disseminate Tenure Tracking System, MEXT, Japan (to RLO).

Authors' contribution

R.L.O. and K.K. contributed equally to the work.