Upload
nguyennhu
View
217
Download
2
Embed Size (px)
Citation preview
1
Attenuated Virulence and Biofilm 1
Formation in Staphylococcus aureus 2
following Sublethal Exposure to Triclosan 3 4
Joe Latimer#, Sarah Forbes# and Andrew J. McBain* 5
School of Pharmacy and Pharmaceutical Sciences, 6 The University of Manchester, Manchester, M13 9PT, U.K. 7
8
9
Running title: Triclosan-adapted Staphylococcus aureus 10
Key Words: Small colony variants, Galleria mellonella infection model, triclosan 11 susceptibility 12 13 #These authors contributed equally to this work 14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
*For correspondence: Andrew McBain, School of Pharmacy, The University of Manchester, 30 Oxford Road, Manchester M13 9PT, UK. Tel: 00 44 161 275 2360; Fax: 00 44 161 275 2396; 31 Email: [email protected] 32
33
Copyright © 2012, American Society for Microbiology. All Rights Reserved.Antimicrob. Agents Chemother. doi:10.1128/AAC.05904-11 AAC Accepts, published online ahead of print on 19 March 2012
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
2
Sub-effective exposure of Staphylococcus aureus to the biocide triclosan can reportedly 34
induce a small colony variant (SCV) phenotype in Staphylococcus aureus. S. aureus SCVs 35
are characterised by slow growth rates, reduced pigmentation and lowered antimicrobial 36
susceptibility. Whilst they may exhibit enhanced intracellular survival, there are conflicting 37
reports regarding their pathogenicity. The current study reports the characteristics of a SCV-38
like strain of S. aureus, created by repeated passage on sub-lethal triclosan concentrations. S. 39
aureus ATCC 6538 (P0) was serially exposed ten times to concentration gradients of triclosan 40
to generate strain P10. This strain was then further passaged ten times on triclosan-free 41
medium (designated x10). The minimum inhibitory and bactericidal concentrations of 42
triclosan for P0, P10 and x10 were determined and growth rates measured in biofilm and 43
planktonic culture. Haemolysin, DNAse and coagulase activities were measured and virulence 44
determined using a Galleria mellonella pathogenicity model. Strain P10 exhibited decreased 45
susceptibility to triclosan and characteristics of a SCV phenotype, including considerably 46
reduced growth rate and the formation of pinpoint colonies. However, this strain also had 47
delayed coagulase production, impaired haemolysis (p<0.01), was defective in biofilm 48
formation and DNAase activity, and displayed significantly attenuated virulence. Colony size, 49
haemolysis, coagulase activity and virulence were only partially restored in strain x10, 50
whereas planktonic growth rate was fully restored. However, x10 was at least as defective in 51
biofilm formation and DNAse production as P10. These data suggest that although repeated 52
exposure to triclosan may result in a SCV-like phenotype, this is not necessarily associated 53
with increased virulence, and adapted bacteria may exhibit other functional deficiencies. 54
55
56
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
3
INTRODUCTION 57
Staphylococcus aureus is an important human pathogen that is responsible for a range of 58
hospital and community-acquired infections (10, 15). Successful treatment of such infections 59
is complicated not only by the emergence of meticillin-resistant strains (MRSA) (10) but also, 60
it has been claimed, by the spontaneous generation of small-colony variants (SCVs) (55). 61
62
SCVs are slow growing S. aureus sub-populations that may be recovered from patients with 63
persisting or relapsing infections (44), particularly those undergoing treatment with 64
aminoglycoside antibiotics (13), which exhibit decreased susceptibility to antimicrobials. 65
SCVs display a distinct phenotype, characterised by the formation of pinpoint colonies on 66
agar, low growth rate and reduced pigmentation. Such characteristics may lead to their 67
misidentification in the hospital laboratory (28), potentially complicating diagnosis (43). 68
69
The SCV phenotype has been commonly attributed to defects in the electron transport chain 70
due to mutations in hemB, ctaA or menD which inhibit haemin, haemin A and menadione 71
biosynthesis, respectively (14, 54). Some SCVs result from thymidine auxotrophy, due to 72
mutations in genes encoding thymidine synthesis or transport (9). Such SCVs can survive on 73
exogenous thymine, for instance in the airway secretions of CF patients, which are rich in 74
necrotic cells (9, 24, 25). Thymidine is also required for menadione biosynthesis, suggesting 75
that menadione auxotrophy may be a common factor in SCV formation in menadione and 76
thymidine mutants. However, SCVs isolated from infections can result from mechanisms 77
other than menadione and thymidine auxotrophy (47), suggesting diversity in SCV 78
mechanisms that has not been systematically investigated and that SCVs which emerge from 79
various sources are not necessarily functionally equivalent. 80
81
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
4
Reduced susceptibility to antibiotics in SCVs is likely to be multifactoral, depending on 82
whether the particular SCV is attributable to aminoglycoside-selected auxotrophy or was 83
generated by other means, since electron transport chain defects, result in reduced 84
electrochemical gradients which may impair the uptake of aminoglycosides and other cationic 85
molecules (4, 37). Additionally, lowered growth rate in SCVs could non-specifically reduce 86
antimicrobial susceptibility by non-specific mechanisms associated with reduced expression 87
of pharmacological targets. Alternatively, enhanced survival within the host cells could shield 88
the bacteria from inhibitory concentrations of antibiotics and biocides. The clinical 89
significance of this, however, depends on whether SCVs can revert to full virulence following 90
cessation of treatment (37) and on whether the expression of SCV phenotypes influences the 91
ability to form recalcitrant biofilms on surfaces. 92
93
The generation of the SCV phenotype in S. aureus has reportedly been induced in vitro by 94
exposure to triclosan (7, 49), a trichlorinated diphenyl ether biocide. Effective concentrations 95
of triclosan are highly bactericidal towards susceptible species where the molecule causes 96
membrane damage through direct interaction and by perturbing lipid biosynthesis (36, 46, 97
53). Triclosan is used as an antibacterial in a variety of consumer and clinical applications (5, 98
12, 16) including triclosan wash solutions, which have been used to reduce carriage of 99
methicillin-resistant S. aureus (MRSA) in hospitals (1, 26, 45, 56). The enoyl-acyl carrier 100
protein reductase FabI has been identified as a primary pharmacological target (31, 36) for 101
triclosan. Inhibition renders the cell unable to synthesise fatty acids and growth is therefore 102
inhibited and, depending on concentration and time, bacteria can be permanently inactivated 103
(21). Due to the demonstrable therapeutic efficacy, concern has been expressed that the use of 104
triclosan and other antiseptics should be limited to applications where a clear health benefit 105
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
5
can be demonstrated, due the possibility of reduced susceptibility to the molecule and 106
potentially also to chemically-unrelated compounds (18). 107
108
Reduced susceptibility to triclosan has been shown in a number of in vitro studies but not, to 109
date, in the environment (32, 38). In one study, SCVs generated by the in vitro exposure of 110
bacteria to sub-lethal concentrations of triclosan impregnated in silicone disks exhibited 111
reduced susceptibility to the molecule and altered colony morphology (7), Since triclosan-112
exposed S. aureus has the potential to display a SCV-like phenotype and some clinical SCV 113
isolates appear to be able to cause disease (49), it has been inferred that triclosan use might 114
generate SCVs which could potentially be of clinical significance (7). However, these in vitro 115
studies do not necessarily reflect the clinical impact of triclosan-selected SCVs (48). For 116
example, in a previous report, SCVs generated using triclosan-impregnated disks 117
demonstrated slow growth rate, reduced coagulase and DNAse activity and decreased 118
haemolysis (7), which is suggestive of decreased virulence. 119
120
There is a clear need to determine the ability of triclosan-generated SCVs to initiate infection 121
(48). To our knowledge, there are currently no published reports regarding the impact of 122
triclosan exposure on the pathogenic capability of resulting SCV-like strains. Additionally, 123
whilst antibiotic-selected SCVs may revert to wild-type phenotype in vivo following cessation 124
of treatment, this has not been reported for triclosan-selected SCVs. The present study 125
therefore investigates the growth and virulence of a SCV-like S. aureus strain created by 126
serial exposure to sub-inhibitory concentrations of triclosan. The potential reversion of this 127
strain following subsequent repeated passage on triclosan-free medium was also studied. 128
129
130
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
6
MATERIALS AND METHODS 131
Bacterial strains and growth media. Staphylococcus aureus ATCC 6538 was 132
acquired from the American Type Culture Collection and cultured on Mueller Hinton agar 133
(MHA), Colombia blood agar (CBA) or mannitol salt agar (MSA, Oxoid, UK) aerobically at 134
37˚C. For supplementation experiments, haemin (1, 10 or 100 μM, final concentrations), 135
menadione (0.4, 4 or 40 μM, final concentrations) or thymidine (0.62, 6.2 or 62 μM final 136
concentrations) were added to MHA. Cryogenic stocks of the bacteria were archived at -80˚C. 137
Selection of isolates with reduced susceptibility to triclosan. Reproducible 138
concentration gradients of triclosan were created on Mueller Hinton agar by depositing stock 139
solutions of triclosan (100 μg/ml or 1 mg/ml) with a Wasp II spiral plater (Don Whitley, 140
Shipley, UK) (30, 35). Plates were dried for 1 hour at room temperature prior to radial 141
deposition of an overnight suspension of S. aureus and incubated for 4 days aerobically at 142
37˚C. Growth observed at the highest triclosan concentration was aseptically removed and 143
used to inoculate further gradient plates. This process was repeated for ten passages. A further 144
ten passages were performed on triclosan-free MHA. Wild-type bacteria (P0), those passaged 145
ten times on triclosan (P10) and those passaged a further ten times on triclosan-free MHA 146
(x10) were archived at -80 for subsequent MIC and MBC determination. 147
Determination of bacterial minimum inhibitory concentrations (MIC). Overnight 148
cultures of S. aureus were diluted 1:100 in sterile Mueller Hinton broth (MHB) and aliquots 149
(120 μl) were delivered to wells of polystyrene 96-well plates (Corning Ltd, Corning, USA). 150
Stock solutions (1.16 mg/ml) of triclosan were prepared in 25% ethanol, filter sterilised and 151
stored at –80˚C. Doubling dilutions of triclosan were added to the diluted overnight cultures 152
(final concentrations 232 μg/ml to 0.23 μg/ml). Sterile and triclosan-free controls were also 153
included. Plates were incubated for 24 h aerobically at 37 ˚C with shaking at 100 rpm. MICs 154
were determined as the lowest concentration of triclosan showing no turbidity in comparison 155
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
7
to a sterile negative control. Three technical and two biological replicates were conducted. 156
Previous validation studies with staphylococci and other genera showed that the triclosan 157
solvent has no effect on the outcome of MIC or MBC determinations (30). 158
Determination of bacterial minimum bactericidal concentrations (MBC). 159
Aliquots (10 μl) taken from MIC plates (above) were spot-plated onto MHA in triplicate and 160
incubated aerobically at 37˚C. MBCs were determined as the lowest concentration of triclosan 161
at which no growth was observed after 4 days of incubation. 162
Planktonic growth rate measurement. Overnight suspensions of S. aureus P0, P10 163
and x10 were diluted 1:100 in MHB (30 ml) and incubated at 37 oC with shaking at 150 rpm. 164
Samples were regularly removed and diluted as appropriate. Optical density was measured at 165
600 nm using a Helios spectrophotometer (Pye Unicam Ltd., Cambridge, UK). 166
Determination of biofilm growth rates. Overnight suspensions of S. aureus P0, P10 167
and x10 were diluted 1 in 40 in MHB and aliquots (200μl) delivered to wells of polystyrene 168
96 well plates. Sterile controls were also included. Plates were incubated statically aerobically 169
or in an anaerobic workstation (Don Whitley, Shipley, UK) at 37 ˚C in humid containers to 170
minimise evaporation. At 6 time points up to 50 h, planktonic and biofilm growth was 171
measured. Planktonic suspensions were transferred into semi-micro cuvettes and cell density 172
was measured spectrophotometrically at 600 nm. The remaining biofilm was stained for 60 173
seconds with 1% (w/v) crystal violet (200 μl) which was then removed and wells were 174
washed three times with distilled water before allowing plates to air-dry. Biofilm-bound CV 175
was eluted in 100 % ethanol (200 μl) and absorbance was measured spectrophotometrically at 176
600 nm. Biofilm units’ (arbitrary units) were calculated by dividing biofilm absorbances by 177
their corresponding planktonic OD. 178
Determination of biofilm architecture and viability. Overnight suspensions of S. 179
aureus P0 and P10 were diluted 1 in 40 in MHB (40 ml) in a Duran bottle containing a 180
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
8
partially-submerged sterile glass microscope slide. Bottles were incubated at 37 oC statically 181
for 24, 48 or 72 h before gentle rinsing in PBS (50 ml). Biofilm viability and structure was 182
visualised by fluorescence microscopy. Briefly, a working solution of BacLight LIVE/DEAD 183
stain (Invitrogen Ltd, Paisley, UK) was prepared by adding 1 μl each of SYTO 9 (component 184
A) and propidium iodide (component B) to 98 μl distilled water. This solution (20 μl) was 185
applied directly to the biofilm and covered with a glass cover slip. Slides were incubated at 186
room temperature in the dark for 15 minutes, according to the standard BacLight staining 187
protocol. Biofilms were visualised with an Axioskop 2 fluorescence microscope with a 10 x 188
objective lens (Carl Zeiss Ltd, Rugby, UK) and images captured using a digital microscope 189
eyepiece (Cosmos Biomedical, Derbyshire, UK) and exported as JPEG files. Bacterial cells 190
incubated in the presence of both stains fluoresce either green (viable) or red (dead). The 191
excitation and emission maxima for these dyes are 480 and 500nm for SYTO 9 and 490 and 192
635nm for propidium iodide. The percentage viable biomass was determined by calculating 193
the proportion of red fluorescence as a percentage of total fluorescence. Biofilm three-194
dimensional structure was visualised with an LSM Confocor 2 confocal microscope with a 10 195
x objective lens (Carl Zeiss Ltd, Rugby, UK). Images were processed, surface-rendered and 196
quantified using Imaris (Bitplane AG, Zurich, Switzerland). 197
Determination of haemolysin activity. This method was adapted from Hathaway and 198
Marshall (19). Overnight suspensions of S. aureus P0, P10 and x10 were diluted 1:50 in MHB 199
(50 ml) in a conical flask and incubated with shaking at 37 oC to an optical density at 600 nm 200
of 0.3. Whole defibrinated horse blood (Oxoid, UK) was added to a final concentration of 5% 201
(v/v) before incubating for a further 3 h. Negative and positive controls (sterile MHB and 202
distilled water, respectively) were included. Aliquots (1 ml) were centrifuged at 16,000 x g 203
(1-14 microfuge, Sigma, UK) for 4 min at room temperature and the absorbance of the 204
supernatant was measured spectrophotometrically at 540 nm. Colony counts were performed 205
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
9
by plating appropriate dilutions on MSA. To avoid variation caused by differences in cell 206
number, specific haemolysis was calculated as ∆A540/cfu. 207
Determination of DNAse activity. This method was adapted from the Health 208
Protection Agency National Standard Thermonuclease Test Method (20). Overnight 209
suspensions of S. aureus P0, P10 and x10 were diluted 1 in 50 in MHB (50 ml) in a conical 210
flask and incubated with shaking at 37 oC to an optical density at 600 nm of 0.3. Aliquots 211
(1ml) were incubated at 100 oC for 15 min, centrifuged at 13,000 rpm for 5 min and 212
supernatants retained. Wells were cut into plates containing DNAse test agar (Oxoid, UK) and 213
filled with supernatant (40 μl). Plates were incubated overnight at 37 oC and then overlaid 214
with 1M HCl. Polymerised DNA makes the agar opaque, and zones of clearing surrounding 215
the wells indicate levels of DNase activity. Zone diameters were recorded and colony counts 216
were performed by plating appropriate dilutions of unheated cultures on MSA to account for 217
variations caused by differences in cell number. 218
Coagulase assay. Suspensions of S. aureus P0, P10 and x10 in Mueller Hinton Broth 219
(1 ml, OD600 = 0.4) were added to rabbit plasma with EDTA (Bactident coagulase, Merck, 220
Darmstadt, Germany, 3 ml), in triplicate, and incubated at 37 oC in a water bath. Tubes were 221
monitored for signs of coagulation over 3 h and scored on a five-point scale according to 222
manufacturers’ instructions. 223
Galleria mellonella pathogenesis assay. The pathogenesis model was adapted from 224
Peleg et al. (42). Final larval-stage G. mellonella (Live Foods Direct, Sheffield, UK) were 225
stored in the dark at 4 oC for less than 7 days before randomly assigning 16 to each treatment 226
group and incubating at 37 oC for 30 min. Overnight suspensions of S. aureus strains P0, P10 227
and x10 were washed twice in PBS and diluted appropriately in PBS to achieve an optical 228
density at 600 nm of 0.1 (5-8 x 105 cfu/ml, as confirmed by colony counts on MSA) Aliquots 229
of each suspension (5 μl) were injected into the hemocoel of each larva via the last left proleg 230
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
10
using a Hamilton syringe (2.5-4 x103 cfu per individual). Larvae were incubated in plastic 231
petri dishes at 37 oC and the number of surviving individuals was recorded daily (dead larvae 232
were unresponsive to touch and appeared black). An untreated group and a group injected 233
with sterile PBS were used as negative controls. The experiments were terminated when at 234
least 2 individuals in a control group had died. Three independent replicates were performed 235
and significance was calculated using the Log-Rank test. Data were plotted as survival curves 236
and representative data are presented. Dead individuals were homogenised in sterile PBS, 237
diluted appropriately and aliquots were spread on MSA to calculate bacterial load per 238
individual at death. 239
240
RESULTS 241
Triclosan exposure selects for isolates with reduced triclosan susceptibility. The 242
minimum inhibitory concentration (MIC) of triclosan to Staphylococcus aureus ATCC 6538 243
(P0) was 0.23 μg/ml. After 5 passages (P5), this increased to 9.67 μg/ml and after 10 244
exposures (P10), to 29 μg/ml (a 126-fold increase compared to the parent strain; Fig. 1.). 245
Colonies of P10 were markedly smaller than those of their parent strain (Fig. 2.). Further 246
passaging on triclosan-free medium resulted in a partial reversion of susceptibility and colony 247
size; following 5 passages on MHA, the MIC was 7.3 μg/ml, and after 10 passages, 1.1 μg/ml. 248
Minimum bactericidal concentration (MBC) data showed a similar trend (Fig. 1.). 249
Triclosan-adapted S. aureus grows more slowly than its parent strain in 250
planktonic and biofilm modes. The growth kinetics of triclosan-adapted S. aureus (P10), its 251
parent strain (P0) and the strain further passaged on triclosan-free medium (x10) were 252
compared in shaken broth cultures (planktonic growth). As shown in Figure 3, P10 underwent 253
an extended lag phase and entered stationary phase later than P0 and x10. The planktonic 254
growth kinetics of P0 and x10 were not significantly different. The exponential phase growth 255
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
11
rate of P10 (0.19/h) was less than half that of P0 (0.43/h) and X10 (0.42/h), Biofilm growth 256
was measured over 50 h using the well-characterised microplate assay (41). Formation of 257
biofilm by P10 and x10, relative to planktonic growth, was significantly lower than that of P0 258
(Fig. 4.). Strain x10 formed less biofilm mass than P10 for the first 15 h of growth, after 259
which levels were not significantly different. Strains P10 and x10 were also significantly 260
biofilm defective when grown anaerobically (data not shown). Biofilms formed on glass 261
slides by P0, P10 and x10 differed markedly with respect to aerial coverage (Fig. 5.). 262
Additionally, epifluorescence and confocal microscopy showed that P0 formed relatively 263
thick three-dimensional microcolony structures whereas strains P10 and x10 formed thinner, 264
less structurally complex biofilms (Fig. 6 A.). Analysis of representative images indicated that 265
the density of biomass in P0 biofilms was markedly higher than in those of P10 or x10 over 266
72 h (Fig. 6 B.). 267
The in vitro haemolytic activity of triclosan-adapted S. aureus is lower than that 268
of the parent strain. The ability of planktonic P0, P10 and x10 populations to lyse 269
erythrocytes was investigated. Higher absorbance readings are indicative of lysis and release 270
of haemoglobin into the supernatant. Haemolysis by P10 was, on average, 2% of the P0 value, 271
a statistically significant difference (Fig. 7. p<0.01). Although x10 showed a partial reversion 272
of haemolytic activity, it was still significantly lower than the parent strain (5% of the P0 273
value). 274
The in vitro DNAse activity of triclosan-adapted S. aureus is lower than that of 275
the parent strain. The levels of thermostable DNase produced by planktonic P0, P10 and 276
x10 populations was investigated with an agar diffusion test. Zones of clearing after addition 277
of heated culture supernatant indicate deoxyribonuclease activity. Activity (mean zone 278
diameter 9.47 ± 0.53 cm) was only observed surrounding wells filled with culture supernatant 279
from the P0 strain. No DNAse activity was observed surrounding wells filled with culture 280
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
12
supernatants from the P10 or x10 strains (Table 2.). Cell density in the P0 cultures (4.6 x 108 281
cfu/ml) was not significantly different from those in the P10 or x10 populations (p=0.2). 282
The in vitro coagulase activity of triclosan-adapted S. aureus is lower than that of 283
the parent strain. The levels of coagulase produced by planktonic P0, P10 and x10 284
populations were investigated using a standard tube coagulase test. Coagulase activity was 285
observed in all strains. Although strain P0 displayed a positive result after 30 minutes, 286
significant coagulation was delayed in strains P10 and x10, both showing a positive result 287
after 2 h (Table 1). 288
Supplementation with haemin, menadione or thymidine does not restore a non-289
SCV phenotype. To ascertain whether the SCV phenotype was caused by mutations affecting 290
biosynthesis of haemin, menadione or thymidine, MHA containing these molecules was 291
inoculated with P10. The concentrations used were based on the previous reports that 292
maximal restoration of normal phenotypes occurred with 1.0 µM haemin, 0.375 µM 293
menadione or 6.2 µM of thymidine (9, 56). Further plates were inoculated with P0 and x10 to 294
control any growth-inhibitory effects of medium supplementation. No supplementation 295
affected the colony morphologies of any strain (Table 2), suggesting that the phenotype of 296
P10 is not attributable to defects in haemin, menadione or thymidine biosynthesis. 297
Triclosan-adapted S. aureus is less virulent in an invertebrate model. The Galleria 298
mellonella pathogenesis assay was used to assess the relative virulence of the test strains. 299
Groups of larvae were injected with P0, P10 or x10 and incubated at 37oC. Dead individuals 300
were recorded daily and data were plotted as survival curves (Figure 8A). Over 6 days of 301
incubation, the P10 treatment group suffered the fewest deaths whereas survival in the group 302
inoculated with the parent strain P0 fell rapidly, levelling out at 5 individuals in the example 303
shown. In comparison, x10 displayed moderate virulence, suggesting a partial reversion in 304
phenotype. A Log Rank test including all replicates showed that all three treatment groups 305
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
13
were significantly different (n=48). Within 24 h of death, the bacterial load of each individual 306
was measured. Although numbers varied markedly between individuals (107 – 1011), there 307
was no significant difference between treatment groups (Figure 8B). Colony morphologies of 308
bacteria recovered from dead larvae were similar to the inoculum in all groups, suggesting 309
that the SCV-like strain did not revert to wild-type, or vice versa, during infection. Results 310
have been summarized in Table 2. 311
312
DISCUSSION 313
S. aureus is known to generate phenotypically-distinct sub-populations termed small colony 314
variants (SCVs), which have occasionally been isolated from patients with chronic and 315
relapsing infections. Clinically-occurring SCVs commonly result from haemin, menadione or 316
thymidine auxotrophy, but may also result from mutations in other genes (47). SCV-like 317
variants can arise in response to sub-lethal exposure to aminoglycoside antibiotics (13) and 318
the biocide triclosan. Thus, concerns have been expressed that exposure of S. aureus to 319
triclosan might lead to the emergence of S. aureus SCVs with reduced susceptibility to 320
antimicrobials (7). However, care must be taken when extrapolating clinical relevance from in 321
vitro data. The fact that triclosan is normally applied topically at concentrations considerably 322
higher than those used to generate SCVs should be taken into account (48), as should the fact 323
that although elevated (0.0029% w/v), the MICs for triclosan in the adapted strains described 324
in the current study remained orders of magnitude lower than the triclosan concentrations in 325
washes used to control MRSA in hospitals (0.3 – 1.0 % w/v) (11, 57). Little is known about 326
the physiology and pathogenicity of triclosan-induced SCVs and doubts have been raised 327
concerning their ability to initiate infection in healthy individuals (48). 328
329
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
14
There have been conflicting reports concerning the virulence of SCVs, whether of clinical 330
origin or created in the laboratory. For example, a hemB mutant was more virulent than its 331
parent strain in a murine septic arthritis model (23), but demonstrated no significant 332
difference in virulence in a rabbit endocarditis model (6). Another report suggested that hemB 333
and menD mutants and SCVs of clinical origin were reduced in virulence (50). Clinically-334
isolated SCVs may however be physiologically distinct from hem and men mutants generated 335
in the laboratory and in this respect, recent proteomic analyses revealed significant 336
differences in protein expression between a hemB mutant, a gentamycin-induced SCV and a 337
clinical SCV isolate (29). Markedly increased expression of glycolytic and pyruvate 338
dehydrogenase enzymes was detected in the laboratory mutant and gentamicin-induced 339
strains, compared to the clinical isolate. However, there has previously been little information 340
available to indicate whether strains selected by triclosan exposure vary in virulence 341
compared to the progenitor strains. 342
343
The current investigation has demonstrated that repeated exposure of S. aureus to triclosan 344
can result in a phenotype with marked similarities to SCVs, but that are attenuated in biofilm 345
formation capacity, haemolysis, DNAase, coagulase activities and pathogenesis in an animal 346
model (Table 2). This suggests that whilst repeated exposure to triclosan may result in an 347
SCV-like phenotype, this is not necessarily associated with increases in variables associated 348
with pathogencity. 349
350
Serial exposure of bacteria to concentration gradients of antimicrobials in agar is a highly 351
selective procedure that has been validated as an effective method for selecting isolates with 352
decreased susceptibility (35). The G. mellonella infection model has been previously used to 353
study bacterial pathogenesis (5, 22, 39), including antimicrobial insusceptibility in S. aureus 354
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
15
(42). Data generated with this model in the current study showed a clear distinction between 355
the virulence of strains P0, P10 and x10. Benefits of G. mellonella over other non-mammalian 356
systems such as C. Elegans, include the fact that individual animals may be injected with a 357
defined inoculum, individual animals can be maintained at human body temperature, and that 358
the system has been shown to broadly reflect virulence patterns seen in mammalian systems 359
(22). 360
361
With respect to specific virulence factors, DNAse activity was markedly reduced following 362
triclosan exposure. It has been suggested that staphylococcal DNAse interferes with the 363
antimicrobial activity of neutrophil-produced extracellular traps (NETs) through the break 364
down of the chromatin backbone. DNAse-mediated NET degradation has been demonstrated 365
for S. aureus in a murine model, in which mice infected with wild-type S. aureus exhibited a 366
significantly higher mortaility rate than those infected with the nuclease-deficient mutant (8). 367
Host extracellular DNA in G. mellonella reportedly plays a similar role in the immune 368
response as it does in humans (3), suggesting that a reduction in DNase activity might 369
partially account for the reduced virulence of the S. aureus SCV observed in our G.mellonella 370
assay. 371
372
Importantly, S. aureus became markedly defective in its ability to form biofilms following 373
triclosan adaptation. Strains P10 and x10 formed biofilm less readily than the parent strain on 374
polystyrene over 50 h incubations and biofilms grown on glass were less dense, exhibiting 375
simpler architecture than the parent strain (Figs. 4-6). To our knowledge, there are currently 376
no data in the literature concerning the ability of triclosan-selected SCVs to form biofilms, 377
and data on other SCVs otherwise generated have not been conclusive; for example, S. aureus 378
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
16
SCVs created by exposure to amikacin formed greater biofilm mass on polystyrene (52) but 379
not on glass (51) and a Staphylococcus epidermidis SCV formed by a mutation in hemB was 380
almost entirely biofilm-defective at 24 h but, contrary to findings presented in the current 381
study, reverted to levels approximating the wild-type at 48 h (2). However, since the authors 382
did not measure biofilm relative to planktonic growth, this may be a reflection of the 383
inherently slow growth of the SCV phenotype, rather than a specific measure of biofilm-384
forming capability. In the current study, biofilm formation on polystyrene was measured 385
quantitatively over 50 h and these data were then converted to relative biofilm-forming units 386
(absorbance of the biofilm-bound crystal violet divided by the corresponding planktonic OD) 387
and therefore reflect biofilm formation, irrespective of planktonic bacterial productivity. Since 388
initial attachment does not appear to be impaired and biofilm deficiency is not due solely to 389
reduced growth rate, strains P10 and x10 appear to be specifically deficient in biofilm 390
maturation. This is potentially a consequence of repression of the intercellular 391
polysaccharides or proteins PIA or Aap, respectively. The staphylococcal accessory regulator 392
gene sarA, which controls several virulence determinants, including biofilm formation, 393
haemolysins and DNAse, may also be implicated. Furthermore, recent analysis has shown 394
that an increase in DNAse production resulting from a mutation in the stress response factor 395
sigB inhibited biofilm growth by breaking down extracellular DNA (eDNA) (27). Although 396
this appears to be in contrast to data presented in the current investigation, there is much to 397
learn about the role of eDNA in such a multi-factoral process as biofilm formation. 398
Importantly, since S. aureus infections related to medical devices are strongly associated with 399
biofilm formation (17), the selection for biofilm-deficient strains may be clinically 400
advantageous since biofilms are markedly less susceptible to antimicrobial therapy than their 401
planktonic counterparts (34, 40), a phenomenon that is due to biofilm-specific physiology as 402
well as diffusion-resistance (33). Therefore, the inability to form biofilm could potentially 403
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
17
outweigh changes in susceptibility related to the SCV phenotype. Further, despite a raised 404
MIC to a single biocide, an inability to form biofilm may leave such SCVs vulnerable to a 405
wide range of antimicrobial and immunological challenges. 406
Although the phenotype of the triclosan-adapted strain (P10) was practically identical to 407
previously-described SCVs in terms of colony size and planktonic growth, initial 408
supplementation data indicated that P10 was not an electron transport chain-defective SCV. It 409
is however possible that mechanisms other than auxotrophy to haemin, menadione or 410
thymidine were responsible for the induction of a SCV-like response in this strain. 411
S. aureus generated from 10 further passages on triclosan-free medium (x10) showed partial 412
reversion in planktonic growth rate and, to a lesser extent, in colony diameter and triclosan 413
susceptibility. Only minor reversions in haemolytic activity and biofilm formation were 414
observed and like P10, no DNAse activity was detected. Strain x10 was significantly less 415
pathogenic than the parent strain in the G. mellonella model and exhibited a marked increase 416
in triclosan susceptibility. Furthermore, P10 did not revert to normal colony size in vivo 417
during the course of the infection model (data not shown). 418
In conclusion, we have shown that whilst repeated exposure to triclosan might select for S. 419
aureus strains with decreased susceptibility to triclosan, this may be associated with 420
deficiencies in growth and virulence. 421
422
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
18
REFERENCES 423
1. Al-Mukhtar, A., V. K. Wong, H. Z. Malik, M. Abu-Hilal, M. Denton, M. Wilcox, 424 J. P. Lodge, G. J. Toogood, and K. R. Prasad. 2009. A simple prophylaxis regimen 425 for MRSA: its impact on the incidence of infection in patients undergoing liver 426 resection. Ann R Coll Surg Engl 91:35-8. 427
2. Al Laham, N., H. Rohde, G. Sander, A. Fischer, M. Hussain, C. Heilmann, D. 428 Mack, R. Proctor, G. Peters, K. Becker, and C. von Eiff. 2007. Augmented 429 expression of polysaccharide intercellular adhesin in a defined Staphylococcus 430 epidermidis mutant with the small-colony-variant phenotype. J Bacteriol 189:4494-431 501. 432
3. Altincicek, B., S. Stotzel, M. Wygrecka, K. T. Preissner, and A. Vilcinskas. 2008. 433 Host-derived extracellular nucleic acids enhance innate immune responses, induce 434 coagulation, and prolong survival upon infection in insects. J Immunol 181:2705-12. 435
4. Aperis, G., B. B. Fuchs, C. A. Anderson, J. E. Warner, S. B. Calderwood, and E. 436 Mylonakis. 2007. Galleria mellonella as a model host to study infection by the 437 Francisella tularensis live vaccine strain. Microbes Infect 9:729-34. 438
5. Barbolt, T. A. 2002. Chemistry and safety of triclosan, and its use as an antimicrobial 439 coating on Coated VICRYL* Plus Antibacterial Suture (coated polyglactin 910 suture 440 with triclosan). Surg Infect (Larchmt) 3 Suppl 1:S45-53. 441
6. Bates, D. M., C. von Eiff, P. J. McNamara, G. Peters, M. R. Yeaman, A. S. Bayer, 442 and R. A. Proctor. 2003. Staphylococcus aureus menD and hemB mutants are as 443 infective as the parent strains, but the menadione biosynthetic mutant persists within 444 the kidney. J Infect Dis 187:1654-61. 445
7. Bayston, R., W. Ashraf, and T. Smith. 2007. Triclosan resistance in methicillin-446 resistant Staphylococcus aureus expressed as small colony variants: a novel mode of 447 evasion of susceptibility to antiseptics. J Antimicrob Chemother 59:848-53. 448
8. Berends, E. T., A. R. Horswill, N. M. Haste, M. Monestier, V. Nizet, and M. von 449 Kockritz-Blickwede. Nuclease expression by Staphylococcus aureus facilitates 450 escape from neutrophil extracellular traps. J Innate Immun 2:576-86. 451
9. Besier, S., A. Ludwig, K. Ohlsen, V. Brade, and T. A. Wichelhaus. 2007. 452 Molecular analysis of the thymidine-auxotrophic small colony variant phenotype of 453 Staphylococcus aureus. Int J Med Microbiol 297:217-25. 454
10. Boucher, H., L. G. Miller, and R. R. Razonable. 2010. Serious infections caused by 455 methicillin-resistant Staphylococcus aureus. Clin Infect Dis 51 Suppl 2:S183-97. 456
11. Brady, L. M., M. Thomson, M. A. Palmer, and J. L. Harkness. 1990. Successful 457 control of endemic MRSA in a cardiothoracic surgical unit. Med J Aust 152:240-5. 458
12. Braid, J. J., and M. C. Wale. 2002. The antibacterial activity of triclosan-459 impregnated storage boxes against Staphylococcus aureus, Escherichia coli, 460 Pseudomonas aeruginosa, Bacillus cereus and Shewanella putrefaciens in conditions 461 simulating domestic use. J Antimicrob Chemother 49:87-94. 462
13. Brouillette, E., A. Martinez, B. J. Boyll, N. E. Allen, and F. Malouin. 2004. 463 Persistence of a Staphylococcus aureus small-colony variant under antibiotic pressure 464 in vivo. FEMS Immunol Med Microbiol 41:35-41. 465
14. Clements, M. O., S. P. Watson, R. K. Poole, and S. J. Foster. 1999. CtaA of 466 Staphylococcus aureus is required for starvation survival, recovery, and cytochrome 467 biosynthesis. J Bacteriol 181:501-7. 468
15. Deleo, F. R., M. Otto, B. N. Kreiswirth, and H. F. Chambers. 2010. Community-469 associated meticillin-resistant Staphylococcus aureus. Lancet 375:1557-68. 470
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
19
16. DeSalva, S. J., B. M. Kong, and Y. J. Lin. 1989. Triclosan: a safety profile. Am J 471 Dent 2 Spec No:185-96. 472
17. Francolini, I., and G. Donelli. 2010. Prevention and control of biofilm-based 473 medical-device-related infections. FEMS Immunol Med Microbiol 59:227-38. 474
18. Gilbert, P., and A. J. McBain. 2003. Potential impact of increased use of biocides in 475 consumer products on prevalence of antibiotic resistance. Clin Microbiol Rev 16:189-476 208. 477
19. Hathaway, S. C., and R. B. Marshall. 1980. Haemolysis as a means of 478 distinguishing between Leptospira interrogans serovars balcanica and hardjo. J Med 479 Microbiol 13:477-81. 480
20. Health Protection Agency. 2010. Thermonuclease Test. National Standard Method 481 BSOP TP 34, UK. 482
21. Heath, R. J., and C. O. Rock. 1995. Enoyl-acyl carrier protein reductase (fabI) plays 483 a determinant role in completing cycles of fatty acid elongation in Escherichia coli. J 484 Biol Chem 270:26538-42. 485
22. Jander, G., L. G. Rahme, and F. M. Ausubel. 2000. Positive correlation between 486 virulence of Pseudomonas aeruginosa mutants in mice and insects. J Bacteriol 487 182:3843-5. 488
23. Jonsson, I. M., C. von Eiff, R. A. Proctor, G. Peters, C. Ryden, and A. Tarkowski. 489 2003. Virulence of a hemB mutant displaying the phenotype of a Staphylococcus 490 aureus small colony variant in a murine model of septic arthritis. Microb Pathog 491 34:73-9. 492
24. Kahl, B., M. Herrmann, A. S. Everding, H. G. Koch, K. Becker, E. Harms, R. A. 493 Proctor, and G. Peters. 1998. Persistent infection with small colony variant strains of 494 Staphylococcus aureus in patients with cystic fibrosis. J Infect Dis 177:1023-9. 495
25. Kahl, B. C., G. Belling, R. Reichelt, M. Herrmann, R. A. Proctor, and G. Peters. 496 2003. Thymidine-dependent small-colony variants of Staphylococcus aureus exhibit 497 gross morphological and ultrastructural changes consistent with impaired cell 498 separation. J Clin Microbiol 41:410-3. 499
26. Kampf, G., and A. Kramer. 2004. Epidemiologic background of hand hygiene and 500 evaluation of the most important agents for scrubs and rubs. Clin Microbiol Rev 501 17:863-93, table of contents. 502
27. Kiedrowski, M. R., J. S. Kavanaugh, C. L. Malone, J. M. Mootz, J. M. Voyich, 503 M. S. Smeltzer, K. W. Bayles, and A. R. Horswill. Nuclease modulates biofilm 504 formation in community-associated methicillin-resistant Staphylococcus aureus. PLoS 505 One 6:e26714. 506
28. Kipp, F., B. C. Kahl, K. Becker, E. J. Baron, R. A. Proctor, G. Peters, and C. von 507 Eiff. 2005. Evaluation of two chromogenic agar media for recovery and identification 508 of Staphylococcus aureus small-colony variants. J Clin Microbiol 43:1956-9. 509
29. Kriegeskorte, A., S. Konig, G. Sander, A. Pirkl, E. Mahabir, R. A. Proctor, C. 510 von Eiff, G. Peters, and K. Becker. 2011. Small colony variants of Staphylococcus 511 aureus reveal distinct protein profiles. Proteomics 11:2476-90. 512
30. Ledder, R. G., P. Gilbert, C. Willis, and A. J. McBain. 2006. Effects of chronic 513 triclosan exposure upon the antimicrobial susceptibility of 40 ex-situ environmental 514 and human isolates. J Appl Microbiol 100:1132-40. 515
31. Levy, C. W., A. Roujeinikova, S. Sedelnikova, P. J. Baker, A. R. Stuitje, A. R. 516 Slabas, D. W. Rice, and J. B. Rafferty. 1999. Molecular basis of triclosan activity. 517 Nature 398:383-4. 518
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
20
32. Levy, S. B. 2001. Antibacterial household products: cause for concern. Emerg Infect 519 Dis 7:512-5. 520
33. Lewis, K. 2007. Persister cells, dormancy and infectious disease. Nat Rev Microbiol 521 5:48-56. 522
34. Lewis, K. 2001. Riddle of biofilm resistance. Antimicrob Agents Chemother 45:999-523 1007. 524
35. McBain, A. J., R. G. Ledder, P. Sreenivasan, and P. Gilbert. 2004. Selection for 525 high-level resistance by chronic triclosan exposure is not universal. J Antimicrob 526 Chemother 53:772-7. 527
36. McMurry, L. M., M. Oethinger, and S. B. Levy. 1998. Triclosan targets lipid 528 synthesis. Nature 394:531-2. 529
37. Melter, O., and B. Radojevic. Small colony variants of Staphylococcus aureus--530 review. Folia Microbiol (Praha) 55:548-58. 531
38. Meyer, B., and B. Cookson. 2010. Does microbial resistance or adaptation to 532 biocides create a hazard in infection prevention and control? J Hosp Infect 76:200-5. 533
39. Mylonakis, E., R. Moreno, J. B. El Khoury, A. Idnurm, J. Heitman, S. B. 534 Calderwood, F. M. Ausubel, and A. Diener. 2005. Galleria mellonella as a model 535 system to study Cryptococcus neoformans pathogenesis. Infect Immun 73:3842-50. 536
40. Nickel, J. C., I. Ruseska, J. B. Wright, and J. W. Costerton. 1985. Tobramycin 537 resistance of Pseudomonas aeruginosa cells growing as a biofilm on urinary catheter 538 material. Antimicrob Agents Chemother 27:619-24. 539
41. O'Toole, G. A., and R. Kolter. 1998. Initiation of biofilm formation in Pseudomonas 540 fluorescens WCS365 proceeds via multiple, convergent signalling pathways: a genetic 541 analysis. Mol Microbiol 28:449-61. 542
42. Peleg, A. Y., D. Monga, S. Pillai, E. Mylonakis, R. C. Moellering, Jr., and G. M. 543 Eliopoulos. 2009. Reduced susceptibility to vancomycin influences pathogenicity in 544 Staphylococcus aureus infection. J Infect Dis 199:532-6. 545
43. Proctor, R. A., and G. Peters. 1998. Small colony variants in staphylococcal 546 infections: diagnostic and therapeutic implications. Clin Infect Dis 27:419-22. 547
44. Proctor, R. A., C. von Eiff, B. C. Kahl, K. Becker, P. McNamara, M. Herrmann, 548 and G. Peters. 2006. Small colony variants: a pathogenic form of bacteria that 549 facilitates persistent and recurrent infections. Nat Rev Microbiol 4:295-305. 550
45. Rashid, A., L. K. Solomon, H. G. Lewis, and K. Khan. 2006. Outbreak of epidemic 551 methicillin-resistant Staphylococcus aureus in a regional burns unit: management and 552 implications. Burns 32:452-7. 553
46. Russell, A. D. 2004. Whither triclosan? J Antimicrob Chemother 53:693-5. 554 47. Schaaff, F., G. Bierbaum, N. Baumert, P. Bartmann, and H. G. Sahl. 2003. 555
Mutations are involved in emergence of aminoglycoside-induced small colony 556 variants of Staphylococcus aureus. Int J Med Microbiol 293:427-35. 557
48. Seaman, P. F., D. Ochs, and M. J. Day. 2007. Comment on: Triclosan resistance in 558 methicillin-resistant Staphylococcus aureus expressed as small colony variants: a 559 novel mode of evasion of susceptibility to antiseptics. J Antimicrob Chemother 560 60:175-6; author reply 176-7. 561
49. Seifert, H., H. Wisplinghoff, P. Schnabel, and C. von Eiff. 2003. Small colony 562 variants of Staphylococcus aureus and pacemaker-related infection. Emerg Infect Dis 563 9:1316-8. 564
50. Sifri, C. D., A. Baresch-Bernal, S. B. Calderwood, and C. von Eiff. 2006. 565 Virulence of Staphylococcus aureus small colony variants in the Caenorhabditis 566 elegans infection model. Infect Immun 74:1091-6. 567
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
21
51. Singh, R., P. Ray, A. Das, and M. Sharma. 2010. Enhanced production of 568 exopolysaccharide matrix and biofilm by a menadione-auxotrophic Staphylococcus 569 aureus small-colony variant. J Med Microbiol 59:521-7. 570
52. Singh, R., P. Ray, A. Das, and M. Sharma. 2009. Role of persisters and small-571 colony variants in antibiotic resistance of planktonic and biofilm-associated 572 Staphylococcus aureus: an in vitro study. J Med Microbiol 58:1067-73. 573
53. Villalain, J., C. R. Mateo, F. J. Aranda, S. Shapiro, and V. Micol. 2001. 574 Membranotropic effects of the antibacterial agent Triclosan. Arch Biochem Biophys 575 390:128-36. 576
54. von Eiff, C., C. Heilmann, R. A. Proctor, C. Woltz, G. Peters, and F. Gotz. 1997. 577 A site-directed Staphylococcus aureus hemB mutant is a small-colony variant which 578 persists intracellularly. J Bacteriol 179:4706-12. 579
55. von Eiff, C., R. A. Proctor, and G. Peters. 2000. Staphylococcus aureus small 580 colony variants: formation and clinical impact. Int J Clin Pract Suppl:44-9. 581
56. Webster, J., J. L. Faoagali, and D. Cartwright. 1994. Elimination of methicillin-582 resistant Staphylococcus aureus from a neonatal intensive care unit after hand washing 583 with triclosan. J Paediatr Child Health 30:59-64. 584
57. Zafar, A. B., R. C. Butler, D. J. Reese, L. A. Gaydos, and P. A. Mennonna. 1995. 585 Use of 0.3% triclosan (Bacti-Stat) to eradicate an outbreak of methicillin-resistant 586 Staphylococcus aureus in a neonatal nursery. Am J Infect Control 23:200-8. 587
588 589
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
22
590 FIGURE LEGENDS 591 592 Fig. 1. Minimum inhibitory concentrations (MICs, white bars) and minimum bactericidal 593 concentrations (MBCs, black bars) of triclosan to S. aureus ATCC 6538 before exposure 594 (P0), after 5 or 10 passages in the presence of triclosan (P5 and P10, respectively) and after 5 595 or 10 subsequent passages on triclosan-free medium (X5 and x10 respectively). Error bars 596 show standard deviation (n=6). 597 598 Fig. 2. Colonies formed on Colombia blood agar by S. aureus parent strain (P0, A), triclosan-599 adapted strain (P10, B) and one passaged a further ten times on triclosan-free medium (x10, 600 C) after aerobic incubation at 37 oC for 48 h. Colonies formed by the triclosan-adapted strain 601 were markedly smaller and more heterogeneous than those formed by the parent strain. 602 603 Fig. 3. Planktonic growth of S. aureus parent strain (P0, triangles), triclosan-exposed strain 604 (P10, squares) and one passaged a further ten times on triclosan-free medium (x10, circles). 605 P10 exhibited an extended lag phase, slower growth and a delayed stationary phase when 606 compared to P0. Error bars show standard deviation (n=3). 607 608 Fig. 4. Biofilm growth of S. aureus parent strain (P0, triangles), triclosan-exposed strain (P10, 609 squares) and one passaged a further ten times on triclosan-free medium (x10, circles) on 610 polystyrene. Data are shown as ‘biofilm units’. A biofilm unit is defined as the absorbance of 611 the biofilm-bound crystal violet divided by the corresponding planktonic OD, and corrects the 612 data to give a representation of biofilm formation irrespective of planktonic mass. Biofilm 613 formation by P10 was markedly lower than P0. This difference was significant at all time 614 points from 6 h (p<0.001, n=12). 615 616 Fig. 5. Biofilm growth on glass of S. aureus parent strain (P0), triclosan-exposed strain (P10) 617 and one passaged a further ten times on triclosan-free medium (x10) as viewed by 618 epifluorescence microscopy. Green fluorescence represents viable cells and red fluorescence 619 represents non-viable cells. Representative fields of view at 24 h are shown. 620 621 Fig. 6. Biofilm growth on glass of S. aureus parent strain (P0), triclosan-exposed strain (P10) 622 and one passaged a further ten times on triclosan-free medium (x10) as viewed by confocal 623 microscopy. Specialist software was used to surface-render the images, allowing comparison 624 of relative density and structure (A), and measurement of biomass density (B). Representative 625 images and data shown (P0, black bars; P10, white bars; x10 striped bars). 626 627 Fig. 7. Haemolytic activity of S. aureus parent strain (P0, black bar), triclosan-exposed strain 628 (P10, white bar) and one passaged a further ten times on triclosan-free medium (x10, striped 629 bar). Whole blood (final concentration 5% v/v) was added to pre-stationary phase cultures and 630 free haemoglobin from lysed erythroctes was measured spectrophotometrically at 450 nm. 631 Data are expressed as mean percentages of the mean P0 value. Error bars show standard 632 deviation (n=3). 633 634 Fig. 8. Example survival curve showing virulence of S. aureus parent strain (P0, green fine-635 dash curve), triclosan-exposed strain (P10, red dot-dash curve) and one passaged a further ten 636 times on triclosan-free medium (x10, blue coarse-dash curve) in a G. mellonella survival 637 assay (A, n=16) and bacterial load at death (B, each point corresponds to the bacterial load in 638
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
23
an individual animal). P0 caused significantly higher death rates than both other strains, 639 whereas the virulence of x10 was not fully restored. Untreated and PBS-treated control data re 640 also shown. Although the bacterial load recovered from dead larva varied, there was no 641 significant difference between treatment groups. Example images of treatment groups are also 642 shown.643
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
24
Table 1. Coagulase activity of S. aureus parent strain (P0), triclosan-adapted strain (P10) and 644 one passaged a further ten times on triclosan-free medium (x10). Tubes were monitored for 645 signs of coagulation over 3 h and scored on a five-point scale according to manufacturers’ 646 instructions. 647
648
Strain
P0 P10 X10
0.5 h ++++ - -
1 h ++++ - +
2 h ++++ +++ +++
3 h ++++ ++++ ++++
Negative symbol: no coagulase detected; +: small separate clots (negative); ++: small joined 649 clots (negative); +++: extensively coagulated clots (positive); ++++: complete coagulation 650 (positive) (Bactident coagulase, Merck, Darmstadt, Germany).651
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
25
652 Table 2. Summary indicating presence (+) or absence (-) of tested characteristics in strains P0, 653 P10 and x10 654
Strain
P0 P10 X10
MIC to triclosan (μg/ml) 0.23 29 1.1
Slow planktonic growth - + -
Impaired biofilm growth - + +
Impaired DNAse activity - + +
Impaired haemolytic activity - + +
Delayed coagulase activity - + +
Impaired virulence - + +
Reversal by hemin supplementation - - -
Reversal by menadione supplementation - - -
Reversal by thymidine supplementation - - -
655
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
26
656
FIG. 1. Latimer et al. 657
658
0
10
20
30
40
50
60
70
PO P5 P10 X5 X10
Strain
Tri
clo
san
co
nce
ntr
atio
n (μ
g/m
l)
659
660
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
27
661
FIG. 2. Latimer et al. 662
663
664
665
A
B C
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
28
FIG. 3. Latimer et al. 666
667
668
669
670
0
1
2
3
4
5
6
7
0 5 10 15 20 25 30 35Time (h)
OD
63
0
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
29
FIG. 4. Latimer et al. 671
672
673
0
0.2
0.4
0.6
0.8
1
1.2
0 5 10 15 20 25 30 35 40 45
Time (h)
Bio
film
Uni
ts (
Bio
film
A 6
30/P
lank
toni
c O
D 6
30)
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
30
FIG. 5. Latimer et al. 674
675
676
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
31
FIG. 6. Latimer et al. 677
678
679
680
A
B
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
32
FIG. 7. Latimer et al. 681
0
10
20
30
40
50
60
70
80
90
100
Ha
em
oly
sis
(%
of
wild
-ty
pe
va
lue
)
682
on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from
33
FIG. 8. Latimer et al. 683
684
685
686
687
688
Untreated
PBS
P10
x10
P0
A
B on February 13, 2018 by guest
http://aac.asm.org/
Dow
nloaded from