Altered Competitive Fitness, Antimicrobial Susceptibility, and Cellular Morphology in a Triclosan-Induced Small-Colony Variant of Staphylococcus aureus

Abstract

Staphylococcus aureus can produce small-colony variants (SCVs) that express various phenotypes. While their significance is unclear, SCV propagation may be influenced by relative fitness, antimicrobial susceptibility, and the underlying mechanism. We have investigated triclosan-induced generation of SCVs in six S. aureus strains, including methicillin-resistant S. aureus (MRSA). Parent strains (P0) were repeatedly passaged on concentration gradients of triclosan using a solid-state exposure system to generate P10. P10 was subsequently passaged without triclosan to generate X10. Susceptibility to triclosan and 7 antibiotics was assessed at all stages. For S. aureus ATCC 6538, SCVs were further characterized by determining microbicide susceptibility and competitive fitness. Cellular morphology was examined using electron microscopy, and protein expression was evaluated through proteomics. Triclosan susceptibility in all SCVs (which could be generated from 4/6 strains) was markedly decreased, while antibiotic susceptibility was significantly increased in the majority of cases. An SCV of S. aureus ATCC 6538 exhibited significantly increased susceptibility to all tested microbicides. Cross-wall formation was impaired in this bacterium, while expression of FabI, a target of triclosan, and IsaA, a lytic transglycosylase involved in cell division, was increased. The P10 SCV was 49% less fit than P0. In summary, triclosan exposure of S. aureus produced SCVs in 4/6 test bacteria, with decreased triclosan susceptibility but with generally increased antibiotic susceptibility. An SCV derived from S. aureus ATCC 6538 showed reduced competitive fitness, potentially due to impaired cell division. In this SCV, increased FabI expression could account for reduced triclosan susceptibility, while IsaA may be upregulated in response to cell division defects.

INTRODUCTION

Staphylococcus aureus small-colony variants (SCVs) are characterized by low growth rate and the formation of small nonpigmented colonies (1, 2). They are commonly, but not exclusively, related to antibiotic exposure (3) and have been shown to display diverse phenotypic characteristics, including reduced beta-hemolysis, coagulase, and DNase activities (4), enhanced intracellular survival (5), impaired biofilm formation (6), reduced virulence (6), and low intrinsic susceptibility to certain antibiotics, cationic microbicides, and antimicrobial peptides (7, 8). While all SCVs are not physiologically the same, certain SCVs have been reported to cause persistent skin, bone, and device-associated infections, and they have been isolated from patients undergoing prolonged antibiotic therapy (2, 9, 10). Due to their uncommon morphological features and pinpoint colony size, SCVs may be overlooked or misidentified in clinical microbiology laboratories, potentially confounding their identification.

The phenotypic variation observed in S. aureus SCVs is often attributed to auxotrophy for menadione, hemin, or thiamine due to mutations in their respective genes. This results in impaired synthesis of menaquinone and cytochromes, causing defects in the electron transport chain (1, 11, 12). A resulting reduction in transmembrane potential leads to a decrease in ATP production, which may cause impaired cell wall synthesis as well as a decrease in growth rate, resulting in a smaller bacterial colony size (12, 13). This metabolic change can also lead to alterations in pigmentation and exotoxin expression (5, 6, 14).

S. aureus SCVs recovered from clinical specimens taken from the bronchial secretions of cystic fibrosis (CF) patients have displayed auxotrophy for thymidine due to mutations in thymidylate synthase (thyA), an enzyme involved in thymidine synthesis through the production of dTMP (9, 15). Recent studies indicate that thyA mutants exhibit resistance to trimethoprim-sulfamethoxazole, a common treatment for CF (15). Trimethoprim-sulfamethoxazole inhibits tetrahydrofolic acid production, which is a cofactor for thymidylate synthase and is thus involved in thymidine synthesis (15). It is therefore apparent that exposure to certain antibiotics may provide a selective pressure for SCVs.

A decrease in transmembrane potential in SCV electron transport-defective mutants may result in reduced susceptibility to certain antibiotics and cationic microbicides due to a reduction in cell wall metabolism, lower growth rate, and impaired uptake of positively charged molecules to the bacterial cell (16–18). For example, clinical SCV isolates of S. aureus have previously shown reduced susceptibility to β-lactams and aminoglycosides (15, 16, 19). Furthermore, it has been suggested that SCVs may potentially gain a survival advantage within the host by their ability to persist within phagocytes, due in part to a decrease in alpha-toxin production, and are therefore shielded from host immune defenses as well as the actions of antibiotics (5, 20, 21). The clinical significance of these purported attributes, however, depends on whether the SCV can revert to full virulence following cessation of treatment, which in turn depends upon the stability of the responsible mutations and the relative fitness of the SCV compared to the parent strain (22, 23).

Induction of the SCV phenotype in S. aureus after subeffective exposure to triclosan has been previously reported (6, 24, 25). Triclosan is a bisphenol microbicide that is often incorporated into disinfectant washes, toothpastes, cosmetics, and household products for the purpose of antisepsis and disinfection (26–28). Triclosan exerts bacteriostatic activity through inhibition of FabI, an enoyl-ACP reductase, which participates in fatty acid synthesis (29–31). At higher concentrations, triclosan is bactericidal due to direct effects on the cytoplasmic membrane (32). While resistance to in-use concentrations of microbicides is rare, certain bacteria are reported to exhibit reduced susceptibility to triclosan after subinhibitory exposure in vitro (6, 25, 33). This may be due to point mutations in the fabI gene (34), overexpression of FabI or due to increased efflux pump activity leading to the removal of the compound from the cell (35, 36). In S. aureus, sublethal exposure to triclosan reportedly may induce the formation of triclosan-insusceptible SCVs that display alterations in metabolism, virulence, and reduced susceptibility to gentamicin (6, 24, 25).

We have previously described the generation of an SCV in S. aureus ATCC 6538 in response to repeated sublethal triclosan exposure (6) that displayed reduced susceptibility to triclosan, lower growth rate, impaired biofilm formation, and reduced pathogenicity compared to the parent strain. The current investigation evaluates the effect of triclosan on the induction of the SCV state in five other strains of S. aureus, as well as further characterizing the phenotypic changes in the previously generated SCV with respect to susceptibility to antibiotics and cationic microbicides, alterations in competitive fitness, cellular morphology, and protein expression.

MATERIALS AND METHODS

Chemical reagents and growth media.

Bacteriological growth media were purchased from Oxoid (Basingstoke, United Kingdom). Chemical reagents were purchased from Sigma-Aldrich (Dorset, United Kingdom) unless otherwise stated. Vantocil (a 20% [vol/vol] aqueous solution of polyhexamethylene biguanide [PHMB]) was obtained from Arch Chemicals, Inc. (Manchester, United Kingdom).

Bacterial strains and growth media.

Staphylococcus aureus strains ATCC 6538 and ATCC 43300 (methicillin-resistant S. aureus [MRSA]) were supplied by the American Type Culture Collection. Strains Newman, NCTC 6571, NCTC 13277 (MRSA), and NCTC 13142 (MRSA) were obtained from Public Health England (Salisbury, United Kingdom). Bacteria were grown on tryptone soy agar (TSA) or tryptone soy broth (TSB). Cultures were incubated aerobically at 37°C for 18 to 24 h unless otherwise stated. Bacteria were archived at −80°C prior to triclosan exposure (parent strain P0), after 10 passages across a triclosan gradient (P10 [SCV]), and after a further 10 passages in the absence of triclosan (X10).

Selection of isolates with reduced triclosan susceptibility.

Reproducible concentration gradients of triclosan were created on TSA by deposition of stock solutions of triclosan (100 μg/ml to 10 mg/ml) with a Wasp II spiral plater (Don Whitley, Shipley, United Kingdom) (6). Plates were dried for 1 h at room temperature prior to radial deposition of an overnight suspension of S. aureus and incubated for 4 days aerobically at 37°C. Growth observed at the highest triclosan concentration was removed and used to inoculate further gradient plates. This process was repeated for 10 passages. A further 10 passages were performed on triclosan-free TSA. Isolates (the parent P0 strain), those passaged 10 times on triclosan (strain P10), and those passaged a further 10 times on triclosan-free TSA (strain X10) were archived at −80°C for subsequent analyses.

MICs and MBCs.

MIC values were determined using the microdilution method as described previously (37). Briefly, overnight bacterial cultures were adjusted to an optical density at 60 nm (OD₆₀₀) of 0.8 and diluted 1:100 in TSB to produce a bacterial inoculum for susceptibility testing. Inocula were incubated with doubling dilutions of the relevant microbicide at 37°C for 24 h. The MIC was defined as the lowest concentration for which bacterial growth did not occur. Growth was defined as turbidity (496 nm) in comparison to an uninoculated well (negative control). Aliquots (10 μl) from wells exhibiting no growth were transferred to sterile TSA and incubated at 37°C. The minimum bactericidal concentration (MBC) was defined as the lowest concentration of microbicide at which no bacterial growth occurred after 4 days of incubation.

Disc diffusion tests.

Antibiotic susceptibilities were determined for ciprofloxacin (1 μg), cephalothin (30 μg), ampicillin (10 μg), kanamycin (10 μg), tetracycline (10 μg), gentamicin (10 μg), and trimethoprim-sulfamethoxazole (25 μg). Disc diffusion assays were performed according to the standardized British Society for Antimicrobial Chemotherapy (BSAC) disc diffusion method for antimicrobial susceptibility testing (38). Plates were incubated for 48 h at 37°C.

Protein extraction and IEF.

Bacterial cultures were grown in TSB at 37°C and 100 rpm for 18 h, diluted 1:100, and incubated at 37°C with shaking at 100 rpm to the mid-log phase (OD₆₀₀ of 0.4). Cultures were pelleted at 12,000 × g, washed in phosphate-buffered saline (PBS) (0.01 M phosphate buffer, 0.0027 M potassium chloride, 0.137 M sodium chloride [pH 7.4], 3 × 3 ml), and resuspended in PBS (1 ml). To extract protein, lysostaphin (50 μg/ml) was added, and the suspensions were incubated for 15 min on ice prior to sonication at an amplitude of 10 μ in 6× 30-s bursts. Protein was precipitated in a 1:1:8 solution of cell lysate with trichloroacetic acid (6.1 N) and acetone and incubated at −20°C for 1 h. Protein was pelleted by centrifugation at 16,000 × g, washed three times in acetone (1 ml), and dissolved in rehydration buffer consisting of 9 M urea, 2% CHAPS {3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate}, 1% dithiothreitol (DTT), 2% carrier ampholytes, 0.5% protease inhibitor, and 0.001% bromophenol blue (2 ml). The soluble protein concentration was quantified using the Bradford assay (Sigma, Poole, United Kingdom). Between 250 μg and 500 μg of protein per 200 μl total volume of buffer was loaded per 11-cm ReadyStrip immobilized protein gradient (IPG) strip at pH 5 to 8 (Bio-Rad, Hertfordshire, United Kingdom). Strips were rehydrated under active conditions overnight using a Protean isoelectric focusing (IEF) cell (Bio-Rad, Hertfordshire, United Kingdom). After rehydration, isoelectric focusing was conducted as follows: 250V for 15 min, linear voltage to 8,000 V, and 500 V until the run was completed.

Two-dimensional gel electrophoresis.

IPG strips were equilibrated using equilibration buffer 1 (6 M urea, 2% SDS, 50 mM Tris-HCl [pH 8.8], 2% glycerol, 1% DTT [5 ml]) followed by equilibration buffer 2 (6 M urea, 2% SDS, 50 mM Tris-HCl [pH 8.8], 2% glycerol, 2.5% iodoacetamide [5 ml]). Polyacrylamide casting gels (34 ml distilled water, 25 ml 1.5 M Tris-HCl [pH 8.8], 0.5 ml of 20% SDS, 40 ml of 30% bis-acrylamide) were polymerized by the addition of 10% ammonium persulfate (0.5 ml) and tetramethylethylenediamene (TEMED) (100 μl). Stacking gel solution (34 ml distilled water, 6.25 ml of 1 M Tris-HCl, 0.25 ml of 20% SDS, 8.5 ml of 30% bis-acrylamide, 0.25 ml ammonium persulfate, 50 μl TEMED) was poured above the set casting gel, and IPG strips were loaded above the stacking gel. Gels were run at 20 V for 1 to 2 h and then at 55 V for 15 to 18 h before being fixed for 8 h (500 ml ethanol, 400 ml water, 100 ml acetic acid) at room temperature and stained with a Coomassie blue stain (0.8 g Coomassie blue R350, 400 ml of 40% methanol, 400 ml 20% acetic acid) for 18 h at room temperature and 20 rpm. After destaining, (500 ml methanol, 400 ml water, 100 ml acetic acid), gel spots of interest were excised and proteins were identified using tandem mass spectrometry (MS-MS), performed at the Biomolecular Analysis Facility within The University of Manchester.

TEM.

Cultures (50 ml) were grown to an OD₆₀₀ of 0.3 in TSB, and bacterial cells were pelleted via centrifugation at 16,000 × g for 10 min. Cells were resuspended in 0.25% glutaraldehyde (1 ml) at 4°C, further fixed in 2% osmium tetroxide, and passed through an ethanol dehydration series. Cells were sectioned (80 nm), and transmission electron microscopy (TEM) was conducted using a FEI Polara 300-kV FEG transmission electron microscope (FEI, Hillsboro, OR) at The University of Manchester imaging suite.

Competitive fitness assay.

Competitive fitness was assessed using methods outlined previously (39). Overnight cultures of S. aureus P0 or P10 were diluted 1:10 and adjusted to an OD₆₀₀ of 1.5. Sterile TSB (250 ml) was inoculated in triplicate with P0 or P10, alone or in combination (final inoculum volume, 500 μl). Flasks were incubated at 37°C with shaking at 100 rpm for 24 h. At 0 h and 24 h, dilutions from each flask (10⁻² to 10⁻⁶) were plated onto TSA and TSA containing 1 μg ml⁻¹ triclosan (TSA_TCS) in triplicate and incubated at 37°C for 18 h. Bacterial viable counts were determined after 18 h of incubation, and relative fitness was assessed for bacteria grown independently and in combination, using the equation W = ln (RF/RI)/ln (SF/SI), where W refers to relative fitness, RI and SI refer to the numbers of SCV and susceptible cells at the start point, respectively, and RF and SF refer to the numbers of SCV and susceptible cells at the endpoint.

RESULTS

Altered triclosan and antibiotic susceptibility in triclosan-exposed S. aureus.

In addition to the SCV (replicate 1 [R1]) previously induced by the exposure of S. aureus ATCC 6538 to triclosan (6), SCVs were similarly formed by the replicate triclosan exposure of S. aureus ATCC 6538 (R2), as well as by strains Newman, ATCC 43300, and NCTC 13277. Colony morphology in S. aureus strains NCTC 6571 and NCTC 13142, however, remained unchanged after repeated triclosan exposure.

Triclosan susceptibility (MIC and MBC) significantly decreased in all P10 strains (SCV and non-SCV) compared to that in the respective parent strains (P < 0.01) (Table 1). After passage in the absence of triclosan (X10), MICs and MBCs frequently partially reverted but remained significantly higher than the preexposure values for all test strains (P < 0.01). When comparing the susceptibility of the P0 to P10 strains, for SCVs, increases in MIC ranged from 4- to 31-fold, while increases in MBC ranged from 3- to 16-fold. In comparison, for non-SCV strains MICs increased from 5- to 11-fold, while MBCs increased from 4- to 8-fold.

TABLE 1.

Triclosan susceptibility of Staphylococcus aureus before, during, and after repeated triclosan exposure^a

S. aureus strainb	MIC, μg/ml (SD)			MBC, μg/ml (SD)
	P0	P10	X10	P0	P10	X10
ATCC 6538 R1*	1	31	7	4	63	14 (4)
ATCC 6538 R2*	2 (1)	21 (8)	14 (4)	8 (3)	94 (34)	29 (6)
NCTC 6571	2	21 (8)	16	17 (6)	63	31
Newman*	4	16	16	31	125	63
ATCC 43300*	2	16	14 (3)	21 (6)	63	41 (17)
NCTC 13277*	4	18 (6)	14 (3)	16	63	57 (13)
NCTC 13142	4	18 (6)	16	16	125	47 (17)

^aShown are mean MICs and minimum bactericidal concentrations (MBCs) of S. aureus before triclosan exposure (P0), after sublethal triclosan exposure (P10), and after recovery in a triclosan-free environment (X10). Data show results from duplicate experiments with three technical replicates. Where data varied between replicates, standard deviations (SD) are given in the parentheses. Bold text indicates a statistically significant difference (P < 0.001) in MIC or MBC compared to that of the parent strain (P0).

^bAsterisks indicate strains that formed SCVs after triclosan exposure. R1 and R2 indicate replicates 1 and 2, respectively.

In terms of antibiotic susceptibility, our previously formed S. aureus ATCC 6538 R1 showed a significant increase in sensitivity to all test antibiotics, with the exception of ampicillin, compared to the parent strain (P < 0.05) (Table 2). Antibiotic susceptibilities partially reverted to preexposure values when the SCV was allowed to recover in the absence of triclosan. However, the susceptibilities of X10 remained significantly higher than those of the parent strain for cephalothin, gentamicin, kanamycin, and trimethoprim-sulfamethoxazole (P < 0.05). Replicate SCV strain ATCC 6538 R2 also exhibited a significant increase in susceptibility to cephalothin, gentamicin, and tetracycline, which remained elevated in the absence of triclosan for cephalothin and gentamicin (X10; P < 0.05). S. aureus ATCC 43300 (SCV) increased in susceptibility to all test antibiotics, with the exception of gentamicin, kanamycin, and ciprofloxacin; however, all increases in susceptibility fully or partially reverted back to preexposure levels once the bacteria were passaged without triclosan (X10; P < 0.05). S. aureus NCTC 13277 (SCV) was more susceptible to gentamicin and trimethoprim-sulfamethoxazole after triclosan exposure, while X10 strains showed no significant difference in susceptibility compared to the unexposed parent strain (P0). S. aureus Newman (SCV) exhibited increased susceptibility to ciprofloxacin, cephalothin, kanamycin, and gentamicin and a decrease in trimethoprim-sulfamethoxazole susceptibility; however, X10 strains only showed a significantly different susceptibility than P0 to trimethoprim-sulfamethoxazole and cephalothin. In non-SCV-forming strains, NCTC 6571 exhibited a significant increase in cephalothin susceptibility after repeated triclosan exposure, while NCTC 13142 showed a reduction in trimethoprim-sulfamethoxazole susceptibility, neither of which fully reverted to preexposure levels in the absence of triclosan (P < 0.05).

TABLE 2.

Antibiotic susceptibility in Staphylococcus aureus before, during, and after repeated triclosan exposure^a

S. aureus strainb	Zone of inhibition, mm (SD)
	TET			KAN			GEN			SXT			AMP			CIP			CEF
	P0	P10	X10	P0	P10	X10	P0	P10	X10	P0	P10	X10	P0	P10	X10	P0	P10	X10	P0	P10	X10
ATCC 6538 R1*	26 (1)	29 (1)	25 (1)	20 (1)	29 (1)	25 (1)	21 (1)	29 (1)	24 (1)	29 (1)	37 (1)	33 (1)	44 (1)	46 (3)	46 (2)	24 (2)	27 (2)	23 (1)	43 (1)	45 (1)	44 (1)
ATCC 6538 R2*	27 (1)	30 (1)	28 (2)	18 (1)	19 (1)	19 (1)	21 (1)	26 (1)	24 (1)	28 (1)	27 (1)	26 (2)	45 (3)	48 (4)	48 (1)	25 (2)	24 (2)	23 (3)	43 (1)	47 (1)	45
NCTC 6571	25 (3)	25 (4)	26	18 (2)	17 (2)	17 (1)	22 (3)	22 (1)	21 (1)	28 (3)	26 (2)	27 (1)	44 (4)	42 (2)	43 (1)	28 (3)	28 (2)	27 (1)	37 (1)	43 (1)	41 (1)
Newman*	28 (2)	29 (1)	26 (1)	14 (3)	20 (1)	16	25 (3)	29 (2)	23 (1)	20 (1)	18 (1)	22 (1)	18 (2)	19 (1)	18 (1)	26 (2)	30 (1)	23 (2)	33 (1)	37 (1)	35 (1)
ATCC 43300*	25 (1)	32 (1)	30 (1)	0	0	0	9 (1)	32 (1)	28 (1)	26 (3)	30	26 (1)	15 (2)	22 (1)	18 (1)	22 (1)	21 (1)	22 (1)	27 (1)	48 (1)	46 (1)
NCTC 13277*	31 (4)	30 (1)	28 (1)	0	0	0	22 (4)	29 (1)	23 (2)	27 (4)	29 (1)	25 (1)	11	11 (1)	10 (1)	0	0	0	0	0	0
NCTC 13142	27 (3)	28 (1)	28 (1)	15 (4)	15 (2)	16 (1)	23 (1)	22 (2)	23 (2)	29 (2)	25 (2)	25 (1)	12 (2)	13 (1)	13	20 (4)	23 (1)	23 (2)	28 (1)	28 (1)	28 (1)

^aShown are antibiotic disc diffusion zones of inhibition of S. aureus strains before triclosan exposure (P0), after sublethal triclosan exposure (P10), and after recovery in a triclosan-free environment (X10). TET, tetracycline; KAN, kanamycin; GEN, gentamicin; SXT, trimethoprim-sulfamethoxazole; AMP, ampicillin; CIP, ciprofloxacin; CEF, cephalothin Data show results from duplicate experiments, each with three technical replicates. Where data varied between replicates, standard deviations (SD) are given in parentheses. Bold text indicates a statistical difference (P < 0.05) in inhibition zone size compared to that of the parent strain (P0).

^bAsterisks indicate strains that formed SCVs after triclosan exposure (P10).

Two-dimensional (2D) gel electrophoresis of a parent and triclosan-exposed strain of S. aureus ATCC 6538 revealed differences in protein expression.

Proteins of interest were identified using tandem mass spectrometry (MS-MS) after electrospray ionization (Fig. 1A and B). Notably, upregulation of triclosan target enzyme FabI was observed in the SCV strain. There was an evident increase in peptide deformylase (Def) production after triclosan exposure, which is a participant in protein synthesis in bacteria. A possible increase in expression of transglycosylase IsaA, an autolysin involved in cell wall cleavage during cell replication, was also detected in the SCV strain.

FIG 1.

(A) Two-dimensional gels showing protein expression profiles in the P0 and P10 strains of S. aureus ATCC 6538. Proteins of interest were excised and identified using electrospray ionization MS-MS. Indicated proteins were identified as FabI (spot 1) IsaA (spot 2), and Def (spot 3). (B) Identities and functions of proteins selected from 2D gels of S. aureus ATCC 6538 that were upregulated in triclosan-induced SCVs. Accession numbers are from the SwissProt database.

A triclosan-adapted S. aureus ATCC 6538 SCV exhibits abnormal cell morphology.

The internal cellular morphologies of the S. aureus parent strain (P0), SCV (P10), and recovered X10 strain were visualized using TEM (Fig. 2). High-resolution micrographs revealed that the SCV exhibited a higher frequency of irregular-shaped or abnormally dividing cells due to asymmetrical septum formation. The mean diameters of the SCV cells were on average 32.8% and 28.3% greater than those of the P0 or X10 strain, respectively (P < 0.001). There was no significant difference between the diameters of P0 and X10 cells.

FIG 2.

(A) Cellular morphology of S. aureus ATCC 6538 parent strain P0 and the P10 and X10 strains visualized by TEM. (B) Mean cell diameters of the P0 (black), P10 (white), and X10 (gray) strains. The asterisk indicates significant difference in cell diameter of P10 compared to that of P0 or X10 (P < 0.001).

Altered cationic microbicide susceptibility in a triclosan-induced S. aureus ATCC 6538 SCV.

Compared to P0, MICs for polyhexamethylene biguanide (PHMB), chlorhexidine, and benzalkonium chloride (BAC) significantly decreased in the S. aureus SCV (P10) from 3.6 μg/ml to 1.8 μg/ml for benzalkonium chloride and chlorhexidine and from 15.6 μg/ml to 3.6 μg/ml for PHMB (P < 0.001) (Table 3). The MBCs of both biguanides were also significantly decreased in P10 from 93.8 μg/ml to 31.6 μg/ml for PHMB and from 15.6 μg/ml to 7.8 μg/ml for chlorhexidine. The MBC for BAC did not change between P0 and P10 (Table 3). After passage in the absence of triclosan (X10), the MIC of PHMB partially reverted to the preexposure level, whereas MICs of chlorhexidine and BAC did not revert. In terms of bactericidal activity, the MBCs of PHMB and chlorhexidine fully reverted to preexposure levels in the absence of any microbicide (X10).

TABLE 3.

Susceptibility of Staphylococcus aureus ATCC 6538 to cationic microbicides before, during, and after repeated triclosan exposure^a

Microbicide	MIC, μg/ml (SD)			MBC, μg/ml (SD)
	P0	P10	X10	P0	P10	X10
Benzalkonium chloride	3.6	1.8	1.8	15.6	15.6	15.6
Chlorhexidine	3.6	1.8	1.8	15.6	7.8	15.6
Polyhexamethylene biguanide	15.6	3.6	7.8	93.8 (34)	31.2	93.8 (34)

^aShown are mean MICs and minimum bactericidal concentrations (MBCs) of S. aureus ATCC 6538 before triclosan exposure (P0), after sublethal triclosan exposure (P10 [SCV]), and after recovery in a triclosan-free environment (X10). Data show results from duplicate experiments with three technical replicates. Where data varied between replicates, standard deviations (SD) are given in parentheses. Bold text indicates a statistically significant difference (P < 0.001) in MIC or MBC compared to that of the parent strain (P0).

Reduced competitive fitness of an S. aureus ATCC 6538 SCV compared to the parent strain.

The overall productivity (CFU per milliliter) of P10 after 24 h of growth was significantly lower than that of P0. This deficit in growth was substantially more pronounced when the strains were grown in competition (Fig. 3). The relative Darwinian fitness (W) levels of P0 and P10 were compared when grown separately and when in competition with each other (Fig. 3). By definition, a relative fitness of 1 indicates no fitness effect between strains, a value of below 1 implies impaired fitness, and a value above 1 indicates enhanced fitness (39). The relative fitness (W) of P10 to P0 during individual growth was 0.97, compared to 0.51 during competition. Therefore, in a noncompetitive environment P10 grew 3% slower than P0, whereas when in a competitive environment, P10 grew 49% slower than P0.

FIG 3.

Competitive fitness of P0 versus P10 (SCV) in S. aureus ATCC 6538. Black and gray bars show CFU per milliliter of P0 and P10, respectively, after 24 h of growth axenically or in binary culture. White bars indicate relative fitness (W) under axenic and binary growth. Data are means and standard deviations from four separate experiments with three technical replicates. Error bars show standard deviations. Asterisks indicate statistically significant differences (P < 0.001) compared to the parent strain (P0).

DISCUSSION

In the present investigation, the repeated exposure of S. aureus to triclosan selected for substantially reduced triclosan susceptibility in 6/6 test strains, while only 4/6 formed the SCV phenotype. In SCVs, antibiotic susceptibility significantly increased in 3/5 strains for tetracycline, 2/5 for kanamycin, 5/5 for gentamicin, 3/5 for trimethoprim-sulfamethoxazole, 1/5 for ampicillin, 2/5 for ciprofloxacin, and 4/5 for cephalothin. The only decrease in antibiotic susceptibility observed in an SCV was in S. aureus Newman for trimethoprim-sulfamethoxazole, which reverted in the absence of triclosan. In the two non-SCV-forming strains, only NCTC 6571 showed a significant increase in antibiotic susceptibility after triclosan exposure (to cephalothin), while non-SCV-forming strain NCTC 13142 showed a significant decrease in trimethoprim-sulfamethoxazole susceptibility. S. aureus strain ATCC 6538 SCV R1 exhibited the largest increase in both triclosan and antibiotic susceptibilities compared to the P0 strain and was therefore further evaluated for alterations in protein expression, competitive fitness, cationic microbicide susceptibility, and cellular morphology.

Proteomic analysis of the S. aureus ATCC 6538 SCV R1 (P10) and parent strain (P0) revealed changes in protein expression after repeated sublethal triclosan exposure, notably an upregulation of triclosan target enzyme FabI, which may explain previously observed decreases in triclosan susceptibility (6). An increase in the expression of peptide deformylase, Def, a metalloenzyme involved in protein synthesis, may indicate an overall elevation in protein synthesis in the SCV (40), possibly as part of a generalized stress response. An increase in IsaA expression was also observed in this SCV strain. A major role of this enzyme is the hydrolysis of bonds within peptidoglycan, thus allowing cell wall expansion and cell growth (41). TEM analysis of cell morphology revealed a high proportion of SCV cells with an abnormal shape and impaired septation, resulting in significantly larger cells than the parent (P0) and the partly recovered (X10) strains. It is therefore possible that the overexpression of IsaA may occur in response to this morphological defect, in an attempt to compensate for the lack of cell division observed in this SCV strain. Both thymidine and hemin auxotrophic SCVs have previously presented as enlarged cocci with multiple cross walls when viewed using scanning electron microscopy (13), which is consistent with the impaired cell septation observed in the current SCV. However, previous analysis of this SCV strain revealed no auxotrophy for thymidine or hemin (6). This defective cell division may help further account for the reduced growth rate and small colony size of the SCV compared to the parent (P0) strain.

The generation of S. aureus SCVs by exposure to various antimicrobials has been previously associated with decreased susceptibility to certain antibiotics, cationic microbicides, and recently, to human antimicrobial peptides (8, 42). This reduction in susceptibility is often attributed to defects in the electron transport chain, as well as reduced growth rates and hypermutability (11, 15, 18). In contrast, in the present study a triclosan-induced SCV in S. aureus strain ATCC 6538 (R1) exhibited increased susceptibility to 6/7 test antibiotics and to the cationic microbicides PHMB, chlorhexidine, and BAC. Interestingly, the only antibiotic to which this SCV's susceptibility did not increase significantly was ampicillin, a transpeptidase inhibitor that interferes with bacterial cell wall formation. TEM revealed impaired cross-wall formation in the SCV strain, and proteomic analysis suggested an increase in expression of IsaA a lytic transglycoylase involved in the hydrolysis of peptidoglycan and thus cell wall expansion, turnover, and cell growth (41). The overexpression of IsaA may represent an adaptation to this functional deficit. Such phenotypic compensation may reduce the effectiveness of ampicillin, potentially ameliorating susceptibility increases in this SCV. Impairments in cell wall synthesis leading to a possible increase in cell wall permeability may further explain why P10 was more susceptible to the majority of antibiotics, as well as to the membranotopic cationic microbicides.

When comparing the relative levels of fitness of the parent (P0) and SCV (P10) strains in S. aureus ATCC 6538 SCV (R1), P10 grew at a 3% lower rate than P0 when grown independently, but it was 49% slower when grown in competition, highlighting the competitive advantage of the P0 strain. The impaired ability of this SCV to undergo cell division may in part account for this reduced relative fitness. Previous investigations have demonstrated fitness costs associated with antimicrobial resistance (22, 43, 44). It has been theorized that the fitness of a bacterium is directly proportional to its rate of transmission and ability to compete with other strains within the host, and this may be inversely proportional to its rate of clearance from the host (45). Therefore, fitness, however measured, may be an important predictor of the clinical significance and potential for environmental persistence of a bacterium. For example, a bacterium acquiring a mutation that results in antimicrobial resistance but also in a fitness burden may not establish in its environment due to reduced competitive fitness. Alternatively, the bacterium could persist at a low level for a prolonged period. However, if an adapted bacterium cannot compete with its congeners or has a markedly reduced specific growth rate, then its pathogenic capability may be reduced (22). When grown in binary culture with the mother strain, the triclosan-induced SCV in S. aureus ATCC 6538 (R1) in the present investigation was outcompeted, indicating the functional implications of adaptation.

Conclusion.

Repeated exposure to triclosan may select for SCVs in S. aureus exhibiting reduced susceptibility to triclosan but significantly increased susceptibility to certain antibiotics. In an SCV generated from S. aureus ATCC 6538, reduced triclosan susceptibility may be partly attributed to the overexpression of target enzyme FabI. This SCV also exhibited impaired competitive fitness, which may be due to defective cell division and an associated reduction in planktonic growth. Additionally, this SCV strain showed increased susceptibility toward 6/7 test antibiotics and all tested cationic microbicides. Unlike some previous reports, the formation of the SCV phenotype in triclosan-exposed S. aureus ATCC 6538 appears not to be due to defects in hemin, menadione, or thymidine synthesis but possibly due to impairment in cell wall formation.