APPLIED AND ENVIRONMENTAL MICROBIOLOGY,0099-2240/97/$04.0010

May 1997, p. 1889–1897 Vol. 63, No. 5

Copyright © 1997, American Society for Microbiology

Characterization of an NaCl-Sensitive Staphylococcus aureus Mutantand Rescue of the NaCl-Sensitive Phenotype by Glycine

Betaine but Not by Other Compatible SolutesURIWAN VIJARANAKUL, MATHEW J. NADAKAVUKAREN, DARRELL O. BAYLES,

BRIAN J. WILKINSON, AND RADHESHYAM K. JAYASWAL*

Microbiology Group, Department of Biological Sciences, Illinois StateUniversity, Normal, Illinois 61790-4120

Received 19 November 1996/Accepted 4 March 1997

To further study mechanisms of coping with osmotic stress-low water activity, mutants of Staphylococcusaureus with transposon Tn917-lacZ-induced NaCl sensitivity were selected for impaired ability to grow on soliddefined medium containing 2 M NaCl. Southern hybridization experiments showed that NaCl-sensitive mu-tants had a single copy of the transposon inserted into a DNA fragment of the same size in each mutant. TheseNaCl-sensitive mutants had an extremely long lag phase (60 to 70 h) in defined medium containing 2.5 M NaCl.The osmoprotectants glycine betaine and choline (which is oxidized to glycine betaine) dramatically shortenedthe lag phase, whereas L-proline and proline betaine, which are effective osmoprotectants for the wild type, wereineffective. Electron microscopic observations of the NaCl-sensitive mutant under NaCl stress conditionsrevealed large, pseudomulticellular cells similar to those observed previously in the wild type under the sameconditions. Glycine betaine, but not L-proline, corrected the morphological abnormalities. Studies of the uptakeof L-[14C]proline and [14C]glycine betaine upon osmotic upshock revealed that the mutant was not defectivein the uptake of either osmoprotectant. Comparison of pool K1, amino acid, and glycine betaine levels underNaCl stress conditions in the mutant and the wild type revealed no striking differences. Glycine betaineappears to have additional beneficial effects on NaCl-stressed cells beyond those of other osmoprotectants. TheNaCl stress protein responses of the wild type and the NaCl-sensitive mutant were characterized and comparedby labeling with L-[35S]methionine and two-dimensional gel electrophoresis. The synthesis of 10 proteinsincreased in the wild type in response to NaCl stress, whereas the synthesis of these 10 proteins plus 2 othersincreased in response to NaCl stress in the NaCl-sensitive mutant. Five proteins, three of which were NaClstress proteins, were produced in elevated amounts in the NaCl-sensitive mutant under unstressed conditionscompared to the wild type. The presence of glycine betaine during NaCl stress decreased the production ofthree NaCl stress proteins in the mutant versus one in the wild type.

It is believed that staphylococcal gastroenteritis is a leadingcause of food-borne illness (18, 20). The disease is classified asa food-borne intoxication. Staphylococcus aureus proliferatesin food and releases one or more heat-stable enterotoxins.Because staphylococcal food poisoning is not an officially re-portable disease (it is estimated that only 1 to 5% of cases arereported in the United States), it is difficult to get an exactestimate of its incidence.

One of the distinguishing characteristics of S. aureus is itsconsiderable tolerance of NaCl, and inclusion of high concen-trations of NaCl in media makes them selective for S. aureus(8). The significance of this in the context of food-borne dis-ease was elegantly summarized by Troller (42). S. aureus is theonly food-borne bacterial pathogen that is routinely capable ofgrowth at water activity (aw) levels below 0.90 (2.6 M NaCl)(high osmotic strength). The organism is a relatively poor com-petitor with other bacteria in food, but at lower aw, growth ofother bacteria is suppressed while that of S. aureus is not.Indeed, salt-cured ham (3.5% NaCl) caused 24% of the out-breaks of staphylococcal food poisoning reviewed by Holmbergand Blake (18). There is potentially practical value in under-

standing the mechanisms utilized by S. aureus to grow in high-osmolarity–low-aw environments.

When a bacterium is subjected to osmotic stress, such as bybeing inoculated into a high-osmotic-strength medium, it re-sponds by rapid accumulation of compatible solutes (11, 13).Accumulation of compatible solutes by S. aureus in response tohyperosmotic stress has been examined in some detail over thepast 5 years. Transport systems for glycine betaine (4, 32, 39)and proline (5, 41) are activated, choline transport is induced(23), and glycine betaine and proline accumulate in high con-centrations (14, 24, 29).

Bacteria also initiate a program of gene expression in re-sponse to osmotic and/or ionic stress by high NaCl concentra-tions, for example, which is manifested as a set of proteinsproduced in increased amounts in response to the stress (44).Limited attention has been paid to alterations in gene expres-sion occurring in response to NaCl stress in S. aureus. Qoron-fleh et al. (33, 34) investigated the S. aureus heat shock re-sponse by using L-[35S]methionine labeling and sodium dodecylsulfate-polyacrylamide gel electrophoresis and reported thatmost of the 12 or so heat shock proteins were also induced byhigh NaCl concentrations. By using stained one-dimensionalgels, Armstrong-Buisseret et al. (3) identified four proteinsproduced in increased amounts in NaCl-grown S. aureus. Thegene encoding a 25-kDa protein was cloned and sequencedand shown to encode a homolog of a subunit of enteric bac-terial alkyl hydroxyperoxide reductase, which is a part of the

* Corresponding author. Mailing address: Department of BiologicalSciences, Illinois State University, Normal, IL 61790-4120. Phone:(309) 438-5128. Fax: (309) 438-3722. E-mail: drjay@rs6000.cmp.ilstu.edu.

1889

on April 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from

oxidative stress response system in these bacteria. The stressprotein response in gram-positive bacteria has been muchmore intensively studied in Bacillus subtilis than in S. aureus.NaCl stress induces B. subtilis to produce a set of general stressproteins under the control of an alternative sigma factor, sB

(16).Microorganisms present in food may be stressed by a variety

of physical or chemical conditions used in food processing thatmay injure the organisms but not kill them (19, 38). On non-selective agar media, injured cells repair the damage and growbut are unable to grow on selective media, which impose ad-ditional stress. For example, in heat-treated S. aureus, injuredcells are defined as those able to grow on tryptic soy agar butnot tryptic soy agar containing 7.5% (wt/vol) NaCl (21). Thus,microbial injury can lead to failure to detect a food-bornepathogen or underestimation of its numbers. Damage to avariety of systems has been described in injured bacteria (19,22), including membrane permeability properties, ribosomes,chromosomes, and enzyme and transport activities. However,as pointed out by Farber and Peterkin (12), much more workis needed in the area of heat- and/or stress-induced injury.

In the present study, we created some NaCl-sensitive S. au-reus mutants to gain further insights into the osmoregulatorymechanisms of S. aureus and potential lesions in injured cells.

MATERIALS AND METHODS

Bacterial strains and culture conditions. NaCl-sensitive mutants were derivedfrom S. aureus ISP2018, a derivative of strain RN450 containing plasmid pTV32,obtained from Peter A. Pattee. S. aureus 8325 and RN450 (8325-4; no prophage)were used for comparison in physiological experiments. All strains were main-tained on tryptic soy agar (Difco) plates (for NaCl-sensitive mutants, the platescontained erythromycin at 20 mg ml21) at 4°C. Organisms were grown in trypticsoy broth or the defined medium described by Townsend and Wilkinson (41).Cells were grown with shaking (200 rpm) at 37°C in 300-ml nephelometer flaskscontaining 50 ml of medium. A 2% (vol/vol) inoculum from an overnight culturein defined medium was the standard inoculum. When required, erythromycin at20 mg ml21 and osmoprotectants at a final concentration of 1 mM were used.Growth was measured turbidimetrically by measurement of optical density at 580nm (OD580) with a Spectronic 20 (Bausch & Lomb, Rochester, N.Y.). CFU weredetermined by plating suitable dilutions on tryptic soy agar plates. Colonies werecounted after incubation at 37°C for 24 to 36 h. Growth experiments wereperformed four times, and the results presented are those of representativeexperiments.

Transposon mutagenesis and screening for NaCl-sensitive mutants. Transpo-son Tn917-lacZ is a 5.4-kb transposon encoding macrolide, lincosamide, andstreptogramin resistance that is present in plasmid pTV32, which has a chlor-amphenicol resistance marker outside the transposon and a temperature-sensi-tive replicon, so that transposition can be forced to occur at 43°C. To selecttransposition events, strain ISP2018 carrying plasmid pTV32 was grown on de-fined medium at 43°C in the presence of erythromycin. Erythromycin-resistantcolonies were scored for NaCl sensitivity by patching them onto defined-mediumplates with and without 2 M NaCl. Colonies unable to grow on plates containing2 M NaCl were designated NaCl-sensitive mutants. Four thousand erythromycin-resistant colonies obtained in four transposon-induced mutagenesis experimentswere scored for NaCl sensitivity. The results presented here were obtained withone of the NaCl-sensitive mutants, designated SS20. Similar results were ob-tained with the other NaCl-sensitive mutants examined.

Southern blot analysis. Chromosomal DNA was isolated from NaCl-sensitivemutants and strain RN450 as described earlier (26) and digested with EcoRI orHindIII. After electrophoresis on a 0.7% agarose gel, the fractionated DNAfragments were transferred to a nylon membrane and probed with Tn917 labeledwith [32P]dCTP.

Uptake of solutes upon hyperosmotic stress (upshock). The uptake of [14C]glycine betaine and L-[14C]proline in response to osmotic upshock in definedmedium was measured as described by Kaenjak et al. (23) and Townsend andWilkinson (41). Briefly, mid-exponential-phase cells (OD580, 0.4) were harvestedand resuspended in a small volume of defined medium. This was used to inoc-ulate defined medium containing a radiolabeled substrate with and without 2 MNaCl. Samples were taken at intervals for measurement of OD580 and uptake ofradioactivity by membrane filtration and scintillation counting.

Pool extraction and analysis. The solute pool fraction was obtained by ex-tracting cells with 5% (wt/vol) trichloroacetic acid as described by Graham andWilkinson (14). Potassium was determined by inductively coupled plasma atomicemission spectrometry (30). Amino acid analysis was carried out by the high-performance liquid chromatography method of Schuster (37). 13C nuclear mag-

netic resonance (NMR) spectroscopy was carried out as described by Amin et al.(1).

Labeling of proteins induced by NaCl stress. To label NaCl-induced proteins,defined medium was constituted with 68.3 mM D-glucose instead of glycerol todecrease the baseline expression of stress proteins (16, 33). Twenty-five millilitersof defined medium was inoculated with an overnight culture of S. aureus andgrown at 37°C to the mid-exponential phase (OD580, 0.4). The cells in 0.5 or 1 mlof medium were recovered by microcentrifugation and washed once with anequal volume of defined medium lacking methionine. Washed cells were resus-pended in methionine-free defined medium containing either no NaCl, 2 MNaCl, or 2 M NaCl and 1 mM glycine betaine. Each group of cells was trans-ferred to a 10-ml tube and allowed to equilibrate at 37°C for 5 min, after which5.5 3 106 Bq of Trans35S-label (L-[35S]methionine and [35S]cysteine; 4.4 3 1010

Bq mol21; ICN) ml21 was added. Cells stressed with 2 M NaCl with and without1 mM glycine betaine were labeled for 1 h at 37°C. Unstressed cells (no NaCl)were labeled for 5 min at 37°C. Labeled cells were recovered by microcentrifu-gation and resuspended in an equal volume of cold acetone for 15 min. Acetone-treated cells were recovered by microcentrifugation. Supernatant was decanted,and the pellet was allowed to air dry. S. aureus cells were lysed by incubation at37°C in 0.05 M Tris-HCl, pH 7.5, containing 0.145 M NaCl and 0.25 mg oflysostaphin ml21 until lysis became visibly apparent. Incorporation of L-[35S]me-thionine into trichloroacetic acid-insoluble protein was quantitated by themethod of Berg et al. (6).

Two-dimensional gel electrophoresis. We subjected protein samples to two-dimensional gel electrophoresis in accordance with the procedure of Hoch-strasser et al. (17) by using a Bio-Rad Protein II xi 2-D cell. First-dimensionisoelectric focusing gels were loaded with equal sample volumes which contained500,000 cpm of trichloroacetic acid-insoluble protein and 100 mg of unlabeledtotal protein from a similarly treated, unlabeled sample. Total protein wasdetermined by using the Bio-Rad DC Protein Assay reagents with bovine serumalbumin fraction V as the standard. Second-dimension slabs were 12.75% acryl-amide–1.25% N,N9-bis-methylene-acrylamide continuous gels without sodiumdodecyl sulfate. The 14C-labeled protein molecular mass standards lysozyme(14.3 kDa), carbonic anhydrase (30.0 kDa), ovalbumin (46.0 kDa), bovine serumalbumin (66.0 kDa), phosphorylase b (97.4 kDa), and myosin (220.0 kDa) (Am-ersham) were loaded in the reference lane of each second-dimension gel. Fol-lowing electrophoresis, the gels were fixed with 30% ethanol–10% acetic acid,dried onto filter paper, and exposed to Kodak Biomax MR film for 65 h at roomtemperature.

Electron microscopy. Samples were examined by transmission electron mi-croscopy as described by Vijaranakul et al. (43).

Molecular genetic procedures. DNA isolation, oligolabeling, and Southernblotting were performed as described by Novick (31) and Sambrook et al. (36).All enzymes were used in accordance with the manufacturer’s specifications.

RESULTS

Analysis of Tn917-lacZ insertion in NaCl-sensitive mutants.The chromosomal DNAs of strain RN450 and NaCl-sensitivemutants were isolated and analyzed by Southern blotting byusing Tn917 as the probe. The probe hybridized to a single17.7-kb EcoRI fragment and to both 10-kb and 8-kb HindIIIfragments in each of the mutant DNA isolates digested witheither EcoRI or HindIII (Fig. 1). The probe did not hybridizewith RN450 chromosomal DNA. This result indicated thatNaCl-sensitive mutants contain a single copy of Tn917 insertedin a DNA fragment of the same size in all of the mutants.

It has been reported that the insertion of Tn917 is random inthe bacterial genome. Whether nine NaCl-sensitive mutantsisolated in four independent experiments having the Tn917insertion within same-size fragments are due to a hot spot orinadvertent sibling selection in the experiment remains to bedetermined.

To prove that the NaCl-sensitive phenotype was due totransposon insertion, we mobilized the erythromycin markerfrom the mutants into another strain by using phage f11.Analysis of the erythromycin-resistant transductants revealedNaCl sensitivity, indicating that insertion of the transposon isresponsible for the salt-sensitive phenotype.

Growth and survival under NaCl stress conditions. In de-fined medium containing 2.5 M NaCl, the mutant showed avery long lag phase of 60 to 70 h compared to the 6 h of theparent strain (Fig. 2). When growth did resume, the rate andfinal culture turbidity achieved were similar to those of parentstrain RN450. In contrast, no difference in growth pattern was

1890 VIJARANAKUL ET AL. APPL. ENVIRON. MICROBIOL.

on April 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from

observed between these two strains under unstressed condi-tions (defined medium alone). Estimation of the viability oflag-phase cells by determining CFU indicated that there wasno wholesale loss of viability of the NaCl-sensitive mutantduring the long lag phase. The cells that eventually grew re-tained erythromycin resistance and did not appear to be rever-tants to NaCl insensitivity, suggesting that the reversion fre-quency is negligible. When SS20 cells growing exponentially indefined medium containing 2.5 M NaCl were inoculated intomedium containing 2.5 M NaCl, the culture grew without a lag.However, if the cells were first grown in defined medium alone,they again showed a long lag phase in medium containing 2.5M NaCl.

In complex medium (tryptic soy broth) containing 2.5 M

NaCl, the mutant showed a significantly slower growth rate,lower maximum OD580 values, and a slightly longer lag phasethan the parent strain grown under the same conditions (datanot shown).

Influence of osmoprotectants on growth of the NaCl-sensi-tive mutant in high-NaCl medium. Various osmoprotectantsare known to stimulate the growth of S. aureus in high-osmotic-strength media. It was of interest to study the effects of thesecompounds on the growth of the NaCl-sensitive mutant inhigh-NaCl medium. First, the influence of several osmopro-tectants on the growth of strain RN450 in defined mediumcontaining 2.5 M NaCl was determined. In accord with previ-ous findings (1, 43), glycine betaine, choline, and proline be-taine were highly effective osmoprotectants, with similar poten-cies, as evaluated by their effects on growth rate and maximumculture turbidity, whereas L-proline was a somewhat weakerosmoprotectant, and taurine had markedly lower activity (datanot shown). When this experiment was carried out with theNaCl-sensitive mutant, strikingly different results were found(Fig. 3). Glycine betaine reduced the lag phase from 60 to 12 h,and the exponential growth rate of the culture was similar tothat of strain RN450 in the presence of 2.5 M NaCl and glycinebetaine. Choline, which is oxidized to glycine betaine (23),stimulated the growth rate like glycine betaine; however, thecells exited the lag phase about 6 h later than with glycinebetaine, which is probably explained by the necessity to inducethe choline transport system (23). Proline betaine and L-pro-line had very little osmoprotectant activity for the NaCl-sen-sitive mutant, even less than taurine. These observationsdemonstrate a qualitative difference in the potencies of thedifferent osmoprotectants, with glycine betaine (or its precur-sor choline) being able to rescue the NaCl-sensitive mutant.

Electron microscopic observation of the NaCl-sensitive mu-tant under NaCl stress conditions. Previous studies (43) withstrain RN450 showed that growth in defined medium contain-ing high NaCl concentrations led to large, pseudomulticellularcells and that the morphological abnormality was corrected byinclusion of glycine betaine, but not by inclusion of L-proline,in the medium. Cells of SS20 were grown to the exponentialphase in defined medium with and without 2.5 M NaCl withand without 1 mM glycine betaine or 1 mM L-proline wereexamined by transmission electron microscopy. Under NaClstress conditions, the NaCl-sensitive mutant showed an even

FIG. 1. Southern analysis of EcoRI (lane 1)- and HindIII (lane 11)-digestedDNA from RN450 and EcoRI (lanes 2 to 10)- and HindIII (lanes 12 to 20)-digested DNAs from NaCl-sensitive mutants SS13, SS14, SS20, SS1.5, SS3.3,SS8.4, SS9.4, SS15.1, and SS19.3, respectively. SS13, SS14, and SS20 were frommutagenesis screening experiment 1: SS1.5 and SS3.3 were from experiment 2,SS8.4 and SS9.4 were from experiment 3, and SS15.1 and SS19.3 were fromexperiment 4. DNA was electrophoresed in 0.7% agarose and transferred to anylon membrane. The blot was probed with 32P-labeled pTV32. A single band ofEcoRI-digested DNA is approximately 17.7 kb (there is no EcoRI site withinTn917). The upper and lower bands of HindIII-digested DNA are approximately10 and 8 kb, respectively (this is due to the presence of an HindIII site withinTn917). The bars indicate the positions of 23- and 9.4-kb molecular size stan-dards (top and bottom, respectively). The arrows indicate the bonds hybridizingwith 32P-labeled probe.

FIG. 2. Growth of S. aureus RN450 and SS20, an NaCl-sensitive mutant, indefined media with and without 2.5 M NaCl. Symbols: E and h, RN450 andSS20, respectively, in defined medium; F and ■, RN450 and SS20, respectively,in defined medium containing 2.5 M NaCl.

FIG. 3. Influence of osmoprotectants on the growth of NaCl-sensitive mutantSS20 in defined medium (E) or defined medium containing 2.5 M NaCl plusglycine betaine (h), L-proline (Ç), no added substance (F), choline (■), prolinebetaine (å), or taurine (ç).

VOL. 63, 1997 ANALYSIS OF NaCl-SENSITIVE MUTANTS OF S. AUREUS 1891

on April 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from

1892 VIJARANAKUL ET AL. APPL. ENVIRON. MICROBIOL.

on April 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from

more extreme pseudomulticellular appearance, with sup-pressed separation of daughter cells (Fig. 4A), than previouslyobserved with strain RN450 (43). Addition of glycine betainereversed this effect (Fig. 4B), whereas addition of proline didnot (Fig. 4C).

Osmoregulation of the NaCl-sensitive mutant. (i) Uptake ofL-[14C]proline and [14C]glycine betaine upon hyperosmoticstress. Because accumulation of compatible solutes is a criticalaspect of bacterial adaptation to NaCl stress, various aspects ofosmoregulation were compared in the NaCl-sensitive mutantand the wild type. More specifically, the experiments describedabove raised the possibility that the mutant was defective inproline and proline betaine accumulation. Accordingly, the up-take of L-[14C]proline and [14C]glycine betaine in response toNaCl stress was compared in the two strains. Previous studieshave shown the accumulation of large amounts of L-[14C]pro-line and [14C]glycine betaine in NaCl-shocked S. aureus (39,41). Mid-exponential-phase cultures were subjected to hyper-osmotic stress with 2 M NaCl, and the uptake of both L-[14C]proline and [14C]glycine betaine was measured. The uptake ofL-[14C]proline by the wild type and SS20 in response to up-shock with 2 M NaCl is shown in Fig. 5. L-[14C]proline wastaken up rapidly and to high levels in response to 2 M NaClwith essentially no difference between the two strains. Similarfindings of no difference between the two strains were madewhen [14C]glycine betaine uptake was studied (data not shown).

(ii) Solute pool composition of NaCl-stressed cells. The sol-ute pools of mid-exponential-phase cultures of the wild type

FIG. 4. Thin-section electron micrographs of exponential-phase SS20 cultures grown in defined medium containing 2.5 M NaCl (A), defined medium containing2.5 M NaCl and 1 mM glycine betaine (B), or defined medium containing 2.5 M NaCl and 1 mM L-proline (C). Bars, 1 mm.

FIG. 5. L-Proline uptake by S. aureus RN450 and SS20 upon osmotic up-shock. Mid-exponential-phase cells were harvested by centrifugation and used toinoculate defined medium containing L-[14C]proline with and without 2 M NaCl.Cell-associated radioactivity was determined at intervals by membrane filtrationand scintillation counting. The values shown are means of three separate exper-iments. Symbols: E and F, SS20 in the absence and presence of 2 M NaCl,respectively; h and ■, RN450 in the absence and presence of 2 M NaCl,respectively.

VOL. 63, 1997 ANALYSIS OF NaCl-SENSITIVE MUTANTS OF S. AUREUS 1893

on April 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from

and the NaCl-sensitive mutant were analyzed for K1 andamino acid compositions and by 13C NMR spectroscopy. K1

levels were high in both the wild type and the mutant grown intryptic soy broth or defined medium and did not change ap-preciably in cells grown in the presence of 2.5 M NaCl. Prolinewas the major amino acid accumulated in both the wild typeand the mutant grown in tryptic soy broth containing 2.5 MNaCl. 13C NMR spectroscopy revealed a high concentration ofglycine betaine accumulated from the medium in both strainsgrown in tryptic soy broth in the presence of 2.5 M NaCl. Therewere no obvious differences in the pool amino acid composi-tions and 13C NMR spectra of the wild type and the mutantgrown in defined medium containing 2 M NaCl. Under theseconditions, glycine betaine did not appear to be present andthere was no strikingly dominant organic osmolyte in eithercase. Thus, there were no striking differences between the poolcompositions of the mutant and wild-type strains grown underNaCl stress conditions.

Protein synthesis in response to NaCl stress. S. aureus 8325and SS20 were stressed with 2 M NaCl, and changes in proteinsynthesis were measured by labeling with L-[35S]methionineand two-dimensional electrophoresis (Fig. 6 and 7 and Table1). Stressing of strain 8325 with 2 M NaCl resulted in increasedsynthesis of 10 proteins, numbered 1 to 10 (Fig. 6b), withapparent masses of 58.8, 37.5, 34.6, 33.8, 32.8, 27.2, 26.3, 18.0,15.1, and 14.5 kDa. These 10 proteins plus 2 more, numbered11 and 12, with apparent masses of 24.5 and 22.4 kDa, respec-tively, were produced when the cells were stressed with 2 MNaCl in the presence of glycine betaine (Fig. 6c). The presenceof glycine betaine during NaCl stress appeared to decrease theproduction of protein 2. When strain SS20 was NaCl stressed,the same 10 proteins (1 to 10) were produced in increasedamounts, plus two others, proteins 16 and 17, of 100 and 27.5kDa, respectively (Fig. 7b), which were not observed in strain8325. Proteins 11 and 12 were also produced in the mutantstressed with NaCl in the presence of glycine betaine (Fig. 7c),as in the parent strain. The production of protein 2 was de-creased by NaCl stress in the presence of glycine betaine com-pared to NaCl stress alone in SS20 (Fig. 7c), as it was in 8325.In SS20, but not strain 8325, the production of two additionalproteins, 5 and 9, was decreased by glycine betaine, with aparticularly large decrease in protein 9 production in SS20(Fig. 7c). Also, the production of one of the additional stressproteins in the mutant, spot 17, was decreased by the presenceof glycine betaine during NaCl stress. Proteins exhibiting de-creased production in SS20 in response to NaCl stress andglycine betaine had levels of synthesis comparable to those ofunstressed SS20. When the basal level of proteins in unstressedcultures of 8325 and SS20 were compared, five proteins, 1, 3, 5,13 (32.1 kDa), and 14 (24.4 kDa), were produced in elevatedamounts compared to strain 8325, whereas one protein, no. 15(19.3 kDa), was produced in decreased amounts (Fig. 6a and7a). Three of the proteins, 1, 3, and 5, were NaCl stress pro-teins, whereas the production of proteins 13, 14, and 15 was notincreased by NaCl (Fig. 6b and 7b and Table 1).

FIG. 6. Two-dimensional protein profile of S. aureus 8325 in response to 2 MNaCl stress. Each gel was loaded with protein corresponding to 500,000 cpm oftrichloroacetic acid-insoluble protein along with 100 mg of unlabeled protein.Symbols: E, proteins exhibiting increased synthesis in response to 2 M NaCl; h,proteins exhibiting increased synthesis in response to 2 M NaCl and 1 mM gly-cine betaine; ¿, proteins increased in SS20 by 2 M NaCl stress; Ç, proteins withaltered levels of synthesis in unstressed 8325 compared to unstressed SS20. a,unstressed control, 5-min labeling; b, stressed with 2 M NaCl, 60-min labeling; c,stressed with 2 M NaCl in the presence of 1 mM glycine betaine, 60-min labeling.

1894 VIJARANAKUL ET AL. APPL. ENVIRON. MICROBIOL.

on April 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from

Sensitivity of the NaCl-sensitive mutant to other stress con-ditions. Compared to RN450, strain SS20 showed increasedsensitivity to 2.5 M KCl, 2 M sucrose, 2 M sorbitol, pH 5, andelevated temperature (46°C), as judged by some combinationof the effects on lag-phase length, growth rate, or maximumculture turbidity attained (data not shown).

DISCUSSION

Several mutants with transposon Tn917-induced NaCl-sen-sitivity were isolated. The mutants showed a very long lagphase in defined medium containing 2.5 M NaCl. This lag wasshortened dramatically by glycine betaine and choline, which isoxidized to glycine betaine (23), but not by L-proline or prolinebetaine. These results raised the question of whether the mu-tant was defective in L-proline accumulation. Uptake measure-ments revealed that the mutant accumulated high levels ofL-[14C]proline and [14C]glycine betaine, similar to the parentstrain, upon hyperosmotic stress with 2 M NaCl. Pool analysesshowed levels of K1, proline, and glycine betaine which weresimilar in the mutant and wild-type strains grown in the pres-ence of a high salt concentration. These results suggest that adefect in compatible-solute accumulation is not responsible forthe NaCl-sensitive phenotype of the mutant.

One question that is raised is why glycine betaine restores

TABLE 1. Molecular masses of proteins involved inthe S. aureus NaCl stress response

Protein spota Avg mol mass(kDa) 6 SD

1 ........................................................................................ 58.8 6 0.62 ........................................................................................ 37.5 6 0.43 ........................................................................................ 34.6 6 0.34 ........................................................................................ 33.8 6 0.35 ........................................................................................ 32.8 6 0.56 ........................................................................................ 27.2 6 0.47 ........................................................................................ 24.3 6 0.38 ........................................................................................ 18.0 6 0.49 ........................................................................................ 15.1 6 0.310 ...................................................................................... 14.5 6 0.4

11 ...................................................................................... 24.0 6 0.512 ...................................................................................... 22.4 6 0.3

13 ...................................................................................... 32.1 6 0.414 ...................................................................................... 24.4 6 0.515 ...................................................................................... 19.3 6 0.4

16 ...................................................................................... 100.0 6 1.817 ...................................................................................... 27.5 6 0.4

a Spot numbers correspond to those in Fig. 7. Spots 1 to 10 represent NaClstress proteins in 8325 and SS20. Spots 11 and 12 represent glycine betaine-dependent stress proteins in 8325 and SS20. Spots 13 to 15 represent proteinsexpressed differently in unstressed SS20 versus 8325. Spots 16 and 17 representadditional NaCl stress proteins in SS20.

FIG. 7. Two-dimensional protein profile of S. aureus SS20 in response to 2 MNaCl stress. Each gel was loaded with protein corresponding to 500,000 cpm oftrichloroacetic acid-insoluble protein along with 100 mg of unlabeled protein.Symbols: E, proteins whose synthesis in the wild type was increased in responseto 2 M NaCl; h, proteins whose synthesis in the wild type was increased inresponse to 2 M NaCl and 1 mM glycine betaine; ¿, proteins whose synthesis wasincreased in the mutant in response to 2 M NaCl; Ç, proteins with alteredsynthesis in the unstressed mutant compared to the unstressed wild type. a,unstressed control, 5-min labeling; b, stressed with 2 M NaCl, 60-min labeling; c,stressed with 2 M NaCl in the presence of 1 mM glycine betaine, 60-min labeling.

VOL. 63, 1997 ANALYSIS OF NaCl-SENSITIVE MUTANTS OF S. AUREUS 1895

on April 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from

near-normal growth of the mutant under NaCl stress whereasL-proline does not. It does not appear to be a simple matter ofturgor restoration, because proline is accumulated in highamounts upon osmotic stress but nevertheless does not act asan osmoprotectant in the mutant. A broadly similar observa-tion has been made recently in an NaCl-resistant S. aureusstrain (25). In strain RN4220, proline was not an osmopro-tectant in cultures in defined medium containing 10% (wt/vol)NaCl at 41°C, even though it was accumulated, but was anosmoprotectant at 37°C. Glycine betaine and choline were osmo-protectants at both temperatures. In most bacteria, glycine be-taine is the most effective osmoprotectant (10, 11). The supe-rior activity of glycine betaine over other osmoprotectants ismore pronounced in our NaCl-sensitive S. aureus mutant thanin the parent strain. The mechanism of compatible-solute func-tion is not fully understood (13). Compatible solutes appear tobe preferentially excluded from the immediate surface of pro-teins (2). By minimizing entropy and reinforcing the hydropho-bic effect, compatible solutes provide a general stabilizing ef-fect, opposing unfolding and denaturation of proteins (13).They also combat the destabilizing effects of NaCl. Interest-ingly, large amounts of NaCl appear to accumulate in the cyto-plasm of S. aureus stressed with high NaCl concentrations (9,28). Cayley et al. (7) have shown that glycine betaine causes asignificant increase in the volume of cytoplasmic free water inosmotically stressed Escherichia coli. Perhaps whatever the de-fect is that makes the mutant more NaCl sensitive than the wildtype can only be counteracted by glycine betaine because of itssuperior protective and cytoplasmic water volume-enhancingeffects. In the wild type, L-proline and proline betaine havesufficient beneficial activities to promote its growth under NaClstress conditions. In Salmonella typhimurium, mutants thatwere not protected by glycine betaine could still use proline asan osmoprotectant in high-osmolarity media (15). Glycine be-taine has been shown to have a variety of beneficial physiolog-ical effects on osmotically stressed cells. The colony-formingability of E. coli was lost upon exposure to 0.8 M NaCl, al-though the organisms were not dead, but could be rescued byglycine betaine (35).

The inhibitory effects of osmotic stress on DNA replicationand cell division are reversed by glycine betaine in osmoticallystressed E. coli (27). In S. aureus, glycine betaine corrects theincreased cell size-pseudomulticellularity and shortened gly-cine interpeptide bridge of cells grown in defined medium con-taining a high NaCl concentration (43). The NaCl-sensitivemutant showed an exaggerated version of the morphologicalabnormalities caused by a high NaCl concentration, whichwere reversed by glycine betaine but not by L-proline. Possiblereasons for the abnormal cell morphology and the role of gly-cine betaine in their reversal have been discussed by Vijara-nakul et al. (43).

Protein production in response to NaCl stress has only beenstudied to a limited extent in S. aureus (3, 33). In the presentstudy, 10 NaCl stress proteins were identified. The simulta-neous presence of glycine betaine and NaCl stress revealed twoadditional proteins whose synthesis was apparently dependenton the presence of glycine betaine and down regulation of thesynthesis of one protein by glycine betaine. Previous studies ofS. aureus NaCl stress proteins have been conducted with cellsgrowing in complex media which would thus be expected tomodulate the NaCl stress response because of the presence ofglycine betaine in complex media.

Several striking observations can be made in comparing theprotein profiles of the NaCl-sensitive mutant with the control.First, the mutant produced several proteins at higher levelsthan the wild type under unstressed conditions and one protein

at significantly reduced levels. Two additional NaCl stress pro-teins were observed in SS20. Under NaCl stress conditions,glycine betaine had more modulating activity than in the par-ent, apparently turning down the synthesis of four of the NaClstress proteins. Overall, the impression is that the mutant hasa tendency to overproduce several stress proteins and thatglycine betaine may modulate this response.

Mutations in different bacteria conferring sensitivity to hy-perosmotic stress have been reviewed by Csonka (10). In somecases, the mutations appear to affect compatible-solute accu-mulation, whereas in others, other cellular processes appear tobe affected. Overall, this is not a well-understood area.

To determine the molecular nature of the observed NaClsensitivity, we recently cloned and sequenced the regions flank-ing the Tn917 insertion in the NaCl-sensitive S. aureus chro-mosome. Our preliminary sequence analysis revealed signifi-cant homology between the mutated gene and branched-chainamino acid carrier proteins of other bacteria (40). We are alsoinvestigating the mechanism by which an apparent mutation ina branched-chain amino acid carrier protein results in NaClsensitivity. A detailed analysis of the mutated gene may pro-vide important information regarding cellular targets affectedby physical or chemical stresses in injured bacteria and theensuing physiological consequences.

ACKNOWLEDGMENTS

We thank Ukti Amin Mani for help with the 13C NMR spectroscopyanalyses and Debra Spayer for help in screening mutants.

This work was supported in part by a student stipend (to U.V.) fromthe American Heart Association-Illinois Affiliate, award 93-37201-9560 from the National Research Initiative Competitive Grants Pro-gram of the U.S. Department of Agriculture, and grant 1 R15 A135778-01 from the National Institutes of Health.

REFERENCES

1. Amin, U. S., T. D. Lash, and B. J. Wilkinson. 1995. Proline betaine is a highlyeffective osmoprotectant for Staphylococcus aureus. Arch. Microbiol. 163:138–142.

2. Arakawa, T., and S. N. Timasheff. 1985. The stabilization of proteins byosmolytes. Biophys. J. 47:411–414.

3. Armstrong-Buisseret, L., M. B. Cole, and G. S. A. B. Stewart. 1995. Ahomologue to the Escherichia coli alkyl hydroperoxide reductase AhpC isinduced by osmotic upshock in Staphylococcus aureus. Microbiology 141:1655–1661.

4. Bae, J. H., S. H. Anderson, and K. J. Miller. 1993. Identification of ahigh-affinity glycine betaine transport system in Staphylococcus aureus. Appl.Environ. Microbiol. 59:2734–2736.

5. Bae, J. H., and K. J. Miller. 1992. Identification of two proline transportsystems in Staphylococcus aureus and their role in osmoregulation. Appl.Environ. Microbiol. 58:471–475.

6. Berg, G. R., W. E. Inniss, and J. J. Heikkila. 1987. Stress proteins andthermotolerance in psychrotrophic yeasts from aquatic environments. Can. J.Microbiol. 33:383–389.

7. Cayley, S., B. A. Lewis, and M. T. Record, Jr. 1992. Origins of the osmo-protective properties of betaine and proline in Escherichia coli K-12. J.Bacteriol. 174:1586–1595.

8. Chapman, G. H. 1945. The significance of sodium chloride in studies ofstaphylococci. J. Bacteriol. 50:201–203.

9. Christian, J. H. B., and J. A. Waltho. 1964. The composition of Staphylo-coccus aureus in relation to the water activity of the growth medium. J. Gen.Microbiol. 35:205–218.

10. Csonka, L. N. 1989. Physiological and genetic responses of bacteria to os-motic stress. Microbiol. Rev. 53:121–147.

11. Csonka, L. N., and A. D. Hanson. 1991. Prokaryotic osmoregulation: geneticsand physiology. Annu. Rev. Microbiol. 45:569–606.

12. Farber, J. M., and P. I. Peterkin. 1991. Listeria monocytogenes, a food-bornepathogen. Microbiol. Rev. 55:476–511.

13. Galinski, E. A. 1995. Osmoadaptation in bacteria. Adv. Microbiol. Physiol.37:273–328.

14. Graham, J. E., and B. J. Wilkinson. 1992. Staphylococcus aureus osmoreg-ulation: roles for choline, glycine betaine, proline, and taurine. J. Bacteriol.174:2711–2716.

15. Gutierrez, J. A., and L. N. Csonka. 1995. Isolation and characterization ofadenylate kinase (adk) mutations in Salmonella typhimurium which block the

1896 VIJARANAKUL ET AL. APPL. ENVIRON. MICROBIOL.

on April 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from

ability of glycine betaine to function as an osmoprotectant. J. Bacteriol.177:390–400.

16. Hecker, M., W. Schumann, and V. Volker. 1996. Heat shock and generalstress response in Bacillus subtilis. Mol. Microbiol. 19:417–428.

17. Hochstrasser, D. F., M. G. Herrington, A. C. Hochstrasser, M. J. Miller, andC. R. Merril. 1988. Methods for increasing the resolution of two-dimensionalprotein electrophoresis. Anal. Biochem. 173:424–455.

18. Holmberg, S. D., and P. A. Blake. 1984. Staphylococcal food poisoning in theUnited States. New facts and old misconceptions. JAMA 251:457–489.

19. Hurst, A. 1977. Bacterial injury: a review. Can. J. Microbiol. 23:935–944.20. Hurst, A., and D. L. Collins-Thompson. 1979. Food as a bacterial habitat.

Adv. Microb. Ecol. 3:79–134.21. Iandolo, J. J., and Z. J. Ordal. 1966. Repair of thermal injury of Staphylo-

coccus aureus. J. Bacteriol. 91:134–142.22. Iandolo, J. J. 1974. Repair of stress-induced macromolecular alterations in

Staphylococcus aureus. Ann. N. Y. Acad. Sci. 236:160–173.23. Kaenjak, A., J. E. Graham, and B. J. Wilkinson. 1993. Choline transport

activity in Staphylococcus aureus induced by osmotic stress and low phos-phate concentrations. J. Bacteriol. 175:2400–2406.

24. Kunin, C. M., and J. Rudy. 1991. Effect of NaCl-induced osmotic stress onintracellular concentrations of glycine betaine and potassium in Escherichiacoli, Enterococcus faecalis and staphylococci. J. Lab. Clin. Med. 118:217–221.

25. Louch, H. A., M. L. Eck, and K. J. Miller. 1997. Osmoadaptation by Staph-ylococcus aureus: analysis of several strains linked to food poisoning out-breaks. J. Food Prot. 60:139–143.

26. Mani, N., P. Tobin, and R. K. Jayaswal. 1993. Isolation and characterizationof autolysis-defective mutants of S. aureus created by Tn917-lacZ mutagen-esis. J. Bacteriol. 175:1493–1499.

27. Meury, J. 1988. Glycine betaine reverses the effects of osmotic stress onDNA replication and cellular division in Escherichia coli. Arch. Microbiol.149:232–239.

28. Michell, P., and J. Moyle. 1959. Permeability of the envelopes of Staphylo-coccus aureus to some salts, amino acids and non-electrolytes. J. Gen. Mi-crobiol. 20:434–441.

29. Miller, K. J., S. C. Zelt, and J. H. Bae. 1991. Glycine betaine and proline arethe principal compatible solutes of Staphylococcus aureus. Curr. Microbiol.23:131–137.

30. Munter, R. C., T. L. Halverson, and R. D. Anderson. 1984. Quality assurancefor plant tissue analysis by inductively coupled plasma-atomic emission spec-troscopy. Commun. Soil Sci. Plant Anal. 15:1285–1322.

31. Novick, R. P. 1991. Genetic systems in staphylococci. Methods Enzymol.204:587–636.

32. Pourkomailian, B., and I. R. Booth. 1994. Glycine betaine transport byStaphylococcus aureus: evidence for feedback regulation of the activity of thetwo transport systems. Microbiology 140:3131–3138.

33. Qoronfleh, M. W., U. N. Streips, and B. J. Wilkinson. 1990. Basic features ofthe staphylococcal heat shock response. Antonie Leeuwenhoek 58:79–86.

34. Qoronfleh, M. W., W. Weraarchakul, and B. J. Wilkinson. 1993. Antibodiesto a range of Staphylococcus aureus and Escherichia coli heat shock proteinsin sera from patients with S. aureus endocarditis. Infect. Immun. 61:1567–1570.

35. Roth, W. G., M. P. Leckie, and D. N. Dietzler. 1988. Restoration of colony-forming activity in osmotically stressed Escherichia coli by betaine. Appl.Environ. Microbiol. 54:3142–3146.

36. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: alaboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, ColdSpring Harbor, N.Y.

37. Schuster, R. 1988. Determination of amino acids in biological, pharmaceu-tical, plant and food samples by automated precolumn derivatization andhigh-performance liquid chromatography. J. Chromatogr. 431:271–284.

38. Smith, J. L., and D. L. Archer. 1988. Heat-induced injury in Listeria mono-cytogenes. J. Indust. Microbiol. 3:105–110.

39. Stimeling, K. W., J. E. Graham, A. Kaenjak, and B. J. Wilkinson. 1994.Evidence for feedback (trans) regulation of and two systems for glycinebetaine transport by Staphylococcus aureus. Microbiology 140:3139–3144.

40. Stucky, K., A. Hagting, J. R. Klein, H. Matern, B. Henrich, W. N. Konings,and R. Plapp. 1995. Cloning and characterization of brnQ, a gene encodinga low-affinity, branched-chain amino acid carrier in Lactobacillus delbruckiisubsp. lactis DSM7290. Mol. Gen. Genet. 249:682–690.

41. Townsend, D. E., and B. J. Wilkinson. 1992. Proline transport in Staphylo-coccus aureus: a high-affinity system and a low-affinity system involved inosmoregulation. J. Bacteriol. 174:2702–2710.

42. Troller, J. A. 1986. Water relations of foodborne bacterial pathogens—anupdated review. J. Food Prot. 49:656–670.

43. Vijaranakul, U., M. J. Nadakavukaren, B. L. M. de Jonge, B. J. Wilkinson,and R. K. Jayaswal. 1995. Increased cell size and shortened peptidoglycaninterpeptide bridge of NaCl-stressed Staphylococcus aureus and their rever-sal by glycine betaine. J. Bacteriol. 177:5116–5121.

44. Volker, U., H. Mach, R. Schmid, and M. Hecker. 1992. Stress proteins andcross protection by heat shock and salt stress in Bacillus subtilis. J. Gen.Microbiol. 138:2125–2136.

VOL. 63, 1997 ANALYSIS OF NaCl-SENSITIVE MUTANTS OF S. AUREUS 1897

on April 23, 2021 by guest

http://aem.asm

.org/D

ownloaded from