α-Hemolysin Activity of Methicillin-Susceptible Staphylococcus aureus ...

α-Hemolysin Activity of Methicillin-Susceptible S. aureus Predicts Ventilator-1

Associated Pneumonia 2

3

Lukas Stulik1, Stefan Malafa1, Jana Hudcova2, Harald Rouha1, Bence Z. Henics1, 4

Donald E. Craven3, Agnes M. Sonnevend4, Eszter Nagy1 5

6

1 Arsanis Biosciences GmbH, Vienna, Austria, 2 Departments of Surgical Critical 7

Care and 3 Infectious Diseases, Lahey Hospital and Medical Center, Burlington, 8

MA,USA, 4 Department of Microbiology and Immunology, United Arab Emirates 9

University, College of Medicine and Health Sciences, Al-Ain, United Arab Emirates. 10

11

Corresponding author: 12

Eszter Nagy, M.D., Ph.D. 13

Arsanis, Inc, Arsanis Biosciences 14

Helmut-Qualtinger-Gasse 2, 1030 Vienna, Austria 15

P: +43-1-7990-117-10, F: +43-1-7990 117 88 16

E-mail: [email protected] 17

18

19

Author’s Contribution: 20

Conception and design of the study: ENA, LST 21

Experimental work: LST, SMA, BHE 22

Analysis and interpretation of data for the work: LST, ENA, SMA, JHU, HRO, ASO, 23

DCR 24

Preparation and review of the manuscript: ENA, LST, HRO, DCR, JHU, ASO, SMA 25

26

of 50 AJRCCM Articles in Press. Published on 10-October-2014 as 10.1164/rccm.201406-1012OC

Copyright © 2014 by the American Thoracic Society

Sources of support: 27

This work was supported by the General Program grant of the Austrian Research 28

Promotion Agency (grant No: FFG 841918), awarded to Arsanis Biosciences. 29

30

Running head: Highly hemolytic MSSA in ETA predicts VAP 31

32

Descriptor Number according to Subject Category List: 10.12, 10.07 33

34

Total word count: 3441 35

36

At a Glance Commentary: 37

Scientific Knowledge on the Subject: 38

Airway colonization by S. aureus is a precursor for the development of MRSA or 39

MSSA induced ventilator-associated tracheobronchitis (VAT) and/or pneumonia 40

(VAP). However, little is known about the pathogen-associated factors of S. aureus 41

that promote progression from colonization to pneumonia. Detection of S. aureus 42

isolates with a high propensity to cause VAP and identification of simple, sensitive 43

and specific biomarkers would greatly support prophylaxis or initiation of earlier 44

antibiotic therapy for VAT and VAP. 45

46

What This Study Adds to the Field: 47

This study represents a comprehensive characterization of S. aureus isolates from 48

endotracheal aspirates (ETA) obtained by serial sampling of ventilated patients. The 49

majority of VAP cases were caused by MSSA; in contrast, MRSA isolates were 50

mainly recovered from colonized patients. High alpha-hemolysin activity of MSSA, 51

but not MRSA, isolates was a marker for the progression to or the presence of VAP. 52



The sheep blood agar hemolysis test is a simple assay that can be performed in 53

routine microbiological laboratories to measure alpha-hemolysin activity and can 54

serve as predictor for VAP in ventilated patients colonized with MSSA isolates. 55

These findings may expedite initiation of earlier therapy to improve patient 56

outcomes, such as survival, decreased ventilator days and length of ICU stay. 57

58

Online Data Supplement: 59

This article has an online data supplement, which is accessible from this issue’s 60

table of content online at www.atsjournals.org 61

62

Key findings: Airway colonization by MSSA strains with high alpha-hemolysin 63

production is a valuable predictor for VAP. These data may help differential 64

diagnosis and patient management as well as confirming the important role of 65

alpha-hemolysin in S. aureus VAP, previously demonstrated only in animal models. 66

67



3

ABSTRACT 68

69

Rationale. Colonization of lower airways by Staphylococcus aureus is a risk factor 70

for the development of ventilator-associated tracheobronchitis (VAT) and 71

pneumonia (VAP). However, little is known about the virulence factors of methicillin-72

sensitive and -resistant S. aureus (MSSA and MRSA) that may influence host 73

colonization and progression to VAT and VAP. 74

Objectives. We evaluated MRSA and MSSA endotracheal aspirates (ETA) for 75

genotype and alpha-hemolysin activity in relation to the development of VAT and 76

VAP. 77

Methods. Serial S. aureus ETA isolates from ventilated patients were analyzed for 78

methicillin resistance, molecular-type by MLST- and spa-typing and alpha-hemolysin 79

activity by semi-quantitative analysis of hemolysis on sheep blood agar and 80

quantitative measurement of cytolysis of human lung epithelial cells. The virulence 81

of selected strains was assessed in mice by intranasal challenge. 82

Measurements and Main Results. We detected S. aureus from ETA samples in a 83

quarter of the 231 ventilated patients analyzed; one third of them developed VAP. 84

VAP patients (n=15) were mainly infected by MSSA strains (87%), while colonized 85

individuals (n=18) not progressing to disease mainly carried MRSA strains (68%). 86

MSSA isolates from colonized or VAT patients exhibited significantly lower alpha-87

hemolysin activity than those from VAP cases, however, no such relationship was 88

found with MRSA strains. Alpha-hemolysin activity of S. aureus isolates was 89

predictive for virulence in mouse pneumonia model. 90



4

Conclusions. MSSA strains with strong blood-agar hemolysis and high alpha-91

hemolysin activity are markers for VAP, but not VAT and, might be considered in 92

differential diagnosis and initiation of therapy. 93

94

Total word count of the Abstract: 244 95

96

Keywords: Staphylococcus aureus; biomarker; alpha-hemolysin; ventilator-97

associated pneumonia 98



5

INTRODUCTION 99

100

Mechanically ventilated patients are at high risk of bacterial colonization that 101

may progress to ventilator-associated respiratory infections, manifested as 102

tracheobronchitis (VAT) or pneumonia (VAP). Both conditions contribute to 103

prolonged ventilation and stay in the intensive care unit (ICU), increased healthcare 104

costs and VAP is associated with increased mortality (1-4). VAT has been proposed 105

as an intermediate condition between simple colonization of the upper airways and 106

VAP, although there are controversies concerning diagnostic criteria and true 107

distinction from VAP (5). Recently, a new approach to disease management has 108

been implemented using serial microbial analysis of endotracheal aspirates (ETAs), 109

which is performed to identify and quantify bacteria colonizing the lower airways 110

(4,6). Heavy colonization defined as many (4+) and moderate (3+) growth by semi-111

quantitative analysis of endotracheal aspirates (SQ-ETA) or quantitative ETA >105 112

colony forming units (CFU)/ml of a respiratory pathogen(s) is a risk factor for 113

progression to VAT and/or VAP (4,6). Detection of specific bacterial pathogens 114

helps to guide earlier, targeted antibiotic treatment and antibiotic stewardship efforts 115

(7). 116

Staphylococcus aureus, both MSSA and MRSA, is a frequent causative 117

pathogen for VAT and VAP (1,2). S. aureus virulence and pathogenesis have been 118

extensively studied in animals, but the role of virulence factors in human disease is 119

poorly understood (8,9). Identification of relevant virulence factors and biomarkers 120

would support more effective prevention strategies, and initiation of earlier therapy 121

of high-risk patients. One of the hallmarks of S. aureus pathogenesis is the 122

production of cytotoxins, of which the best characterized is alpha-hemolysin (Hla). 123

Hla is highly potent in lysing bronchial and alveolar epithelial cells, as well as 124



6

macrophages and lymphocytes, and is implicated in the induction of pro-125

inflammatory processes (10). Its dominant role in the pathogenesis of S. aureus 126

pneumonia has been demonstrated in several different animal models (10). 127

Recently published seroepidemiology studies suggested a correlation between 128

higher serum levels of anti-Hla antibodies and favorable clinical outcome in the case 129

of sepsis (11,12). Therefore, the relationship between pre-existing Hla-neutralizing 130

antibody levels, susceptibility to VAP, and disease progression deserves future 131

investigation. Hla has shown promise as a vaccine antigen and monoclonal antibody 132

target in animal models of S. aureus disease (13-16), and is currently being 133

evaluated in human trials of both active and passive immunization. 134

We compared methicillin resistance, genotypes and Hla-activity of S. aureus 135

isolates from serial ETA samples from ventilated patients, to assess the association 136

between these markers and progression to VAP. 137

Some of the results of these studies have been previously reported in the 138

form of an abstract (17). 139



7

METHODS 140

141

Patients and Samples 142

ETA samples and clinical data were collected from 231 ventilated patients 143

hospitalized between May and December 2010 in two medical and one surgical ICU 144

at the Lahey Hospital and Medical Center (Burlington, MA, US). A natural history 145

study based on data from 188 of these patients with ETA samples analyzed by 146

quantitative and semi-quantitative techniques was published earlier (4). SQ-ETA 147

data were obtained from all 231 patients (while Q-ETA for 188 patients) and was 148

used in the current study. Heavy colonization is defined based on semi-quantitative 149

microbiology analysis of ETA samples with many (4+) or moderate (3+) bacterial 150

growth. In the original study authors observed good correlation between Q-ETA 151

criteria ≥ 105 CFU/mL and SQ-ETA with at least moderate growth: 93% agreement 152

and a kappa value = 0.86 (4). 153

The characteristics of the additional 43 patients included in this study were 154

comparable to those 188 reported earlier (4). VAT diagnosis was based on heavy 155

colonization, plus at least two clinical criteria (fever, leukocytosis or purulent 156

sputum). VAP was diagnosed as for VAT plus a new and persistent infiltrate on 157

chest radiograph. 158

The research protocols were approved by the Lahey Clinic Institutional 159

Review Board. 160

161

Microbiological Analyses and Molecular Typing of S. aureus Isolates 162

To determine the MRSA/MSSA-status, strains were grown on selective agar 163

and subjected to multiplex PCR to detect mecA. S. aureus strains were genotyped 164



8

by MLST-, spa-, capsule-, SCCmec- and hla-typing according to standard methods 165

(18-22). More details are provided in the Online Data Supplement. 166

167

Blood Agar Hemolysis Test 168

S. aureus isolates were incubated for 16h at 37°C on Columbia agar 169

containing 5% sheep blood (COS, bioMérieux, Marcy-l'Etoile, France) and 170

hemolysis profiles were evaluated by visual inspection and semi-quantitative 171

assessment of cleared (complete hemolysis) zones around bacterial colonies. This 172

test was performed with three different clones of the same isolate and results 173

evaluated by three different individuals. 174

175

In Vitro Cytotoxicity Assay 176

Cytotoxicity assays were performed using a human alveolar epithelial cell line 177

(A549 ATCC® CCL-185™, LGC Standards, Teddington, UK) and bacterial culture 178

supernatants of bacteria grown overnight in three different culture media: TSB, CCY 179

and RPMI + casamino acids. More details are provided in the Online Data 180

Supplement. 181

182

Murine Pneumonia Model 183

Virulence studies were performed by intranasal challenge of anesthetized 184

BALB/cJRj mice (Janvier Labs, Saint-Berthevin Cedex, France) with three different 185

challenge doses of clinical isolates. Survival of animals was monitored for 7 days 186

after challenge. More details are provided in the Online Data Supplement. 187

188

189

190



9

Data-Analyses and Statistics 191

For spa-typing and MLST, published software-packages were applied. Spa-192

types were analyzed with the Ridom Staph TypeTM software (Ridom GmbH, 193

Würzburg, Germany). We acknowledge the use of the S. aureus MLST database 194

which is located at Imperial College London and is funded by the Wellcome Trust. 195

Five types of statistical methods were applied: Fisher exact two-tailed probability 196

test using the Prism 6 software (GraphPad, La Jolla, CA, USA) and the one-way 197

ANOVA, two-sample t-test, Kruskal-Wallis test and two-sample Wilcoxon rank-sum 198

test using the SPSS V19.0 software (IBM, Armonk, NY, US). The method used for 199

each dataset is described in the corresponding text and the figure legends. 200



10

RESULTS 201

202

Collection of S. aureus Isolates from ETA Samples of Ventilated Patients. 203

S. aureus was isolated from 56 of the 231 study patients (24%), and 45 of 204

them infected only with S. aureus were selected into this study (Figure 1). 17 205

patients developed VAT (38%), of which five progressed to VAP, giving a total of 15 206

patients who developed VAP (33%) (Figure 1). With the exception of six non-207

VAT/non-VAP patients (out of 18), study subjects had heavy colonization detected 208

by semi-quantitative techniques (corresponding to > 105 CFU/ml in a Q-ETA sample 209

when analyzed by quantitative growth) (Figure 1). 210

Patient characteristics, including disease scores, co-morbidity and crude 211

mortality were comparable between VAP, VAT and colonized groups. Patients 212

treated in the medical ICUs had higher APACHE II scores than those in the surgical 213

ICU. All patients received antibiotics during ventilation (Table 1). Significantly higher 214

proportion of S. aureus VAP cases occurred in the surgical ICU compared to non-215

VAP cases (S. aureus colonized or diagnosed with S. aureus VAT) (p=0.0393). 216

Patients diagnosed with VAP and/or VAT had significantly more ventilator days than 217

those who were colonized by S. aureus, but did not progress to VAT or VAP (Table 218

1). Duration of ventilation was comparable in the surgical vs. medical ICU patient 219

groups. 220

221

Distribution of MRSA and MSSA Strains in the Different Patient Groups. 222

MRSA and MSSA were found in 23 and 26 samples of 45 patients, 223

respectively (three patients carried both MRSA and MSSA) (Figure 2). 75% of 224

isolates from the surgical ICU were MSSA (12/16), while more MRSA than MSSA 225

were detected among the medical ICUs isolates (19/33) (p=0.0164, Fisher exact 226



11

two-tailed probability test). Nasal swabs - taken at admission to ICUs followed by 227

weekly sampling - were approximately 75% predictive for the presence of MRSA or 228

MSSA in subsequently collected ETA samples (Figure 2). 229

Distribution of MRSA and MSSA strains in the different patient groups was 230

significantly different. While 67% of colonized patients (12/18) carried MRSA, 87% 231

(13/15) of VAP patients were infected with MSSA (p=0.0062, two-sample t-test). 232

Only four ETA samples from VAP patients contained MRSA, two of these occurred 233

together with MSSA. MRSA and MSSA strains were equally represented in VAT 234

patients who did not progress to VAP (7 and 6 out of 12, respectively, Figure 2). 235

These data strongly suggest that MSSA strains were more likely to cause 236

VAP than MRSA strains in this study population. 237

238

Determination of S. aureus Genotypes. 239

We further characterized the isolates based on MLST and spa-typing. At least 240

one strain from each patient was included in the analysis. When multiple samples 241

were collected, the first and last available isolates were tested, and even more when 242

different sheep blood agar hemolysis patterns or both MSSA and MRSA were 243

detected in individual patients. MLST and spa-typing revealed that all 26 MSSA 244

strains belonged to a different clonal type, while the 23 MRSA strains were 245

represented by 5 sequence types. The most dominant clonal lineage was the 246

ST5-II-t002, a well-known hospital-associated (HA)-MRSA also represented by the 247

USA100 PFGE standard strains. This was detected in 37% of patients and was 248

responsible for 71% of MRSA infections. The ST5-II-t002 clone was isolated 249

throughout the seven months of the study. Three MRSA clonal types occurred in 250

two different patients without temporal association (Table E1 in the online data 251

supplement). 252



12

253

Sheep blood Agar Hemolysis Patterns 254

Upon culturing the S. aureus isolates on sheep blood agar, we detected a 255

broad range of hemolytic activities (Figure 3A). The majority of isolates displayed a 256

hemolysis pattern characteristic for alpha-hemolysin (Hla) causing beta-hemolysis 257

and a clear halo on sheep blood agar plates. A few isolates showed hemolysis with 258

a turbid halo, which is typical for beta-hemolysin (Hlb) production of S. aureus. The 259

presence of the functional hlb gene in these isolates was confirmed by PCR (data 260

not shown). 261

We detected comparable proportion of high and low hemolytic S. aureus 262

isolates among MRSA and MSSA (4+ or 3+: 12/23 and 12/26, respectively). 263

However, VAP patients significantly more likely carried highly hemolytic (3+ or 4+) 264

MSSA strains than non-VAP patients (10 out of 13 vs. 2 out of 11, respectively; 265

p=0.0123, Fisher exact two-tailed probability test). Such correlation between 266

hemolysis strength and VAP occurrence was not found with MRSA strains (Figure 267

3B). VAP patients who were infected with MSSA strains exhibiting moderate, low or 268

no blood agar hemolysis were taken off the ventilator earlier than those infected with 269

highly hemolytic strains (Figure 2). Isolates from serial samples of the individual 270

patients displayed similar blood agar hemolysis profile, except when more than one 271

molecular type was identified by MLST- or spa-typing (data not shown). 272

Based on these data, MSSA strains strongly hemolytic on sheep blood agar 273

are much more likely to cause VAP than weakly hemolytic isolates. 274

275

Alpha-Hemolysin Activity of Isolates 276

To correlate the hemolytic profiles on sheep blood agar with alpha-hemolysin 277

activity, we determined the strength of cytolysis of serially diluted overnight culture 278



13

supernatants from S. aureus isolates using a human lung alveolar epithelial cell line 279

(A549). The specificity of this assay for Hla was demonstrated by the lack of cell 280

lysis by a hla gene-deletion mutant S. aureus strain (TCH1516, USA300) and in the 281

presence of a Hla-neutralizing monoclonal antibody (Figure E1 in the online data 282

supplement). 283

Since it is known that the composition of growth media greatly influences 284

toxin expression, S. aureus isolates (two individually picked clones) were grown in 285

three different culture media: TSB, CCY or RPMI supplemented with casamino 286

acids. We found that blood agar hemolysis strength was overall in agreement with 287

cytolytic activity, and CCY culture supernatants showed the best resolution between 288

the different hemolysis category groups (Figure 4). Hla-activity of isolates that 289

expressed beta-hemolysin became quantifiable (masked by the turbid hemolysis 290

pattern on blood agar). 291

Similarly to blood agar hemolysis, a great variation in alpha-hemolysin activity 292

measured in the A549 viability assays was observed among the S. aureus isolates. 293

The serial samples of individual patients, often collected more than a week apart, 294

were remarkably similar to each other in their cytolytic pattern (Figure 5A). When 295

patients were infected with two different S. aureus clonal types, these were 296

recognizably different from each other (for example patients P22, P39 and P198). 297

The ST5-II-t002 isolates behaved differently when isolated from different patients 298

and displayed a broad range of cytotoxicity from no or low to high activity (Figure 299

5B). We also observed different Hla-activities with other clonal types when isolated 300

from different patients (examples shown in Figure 5C). 301

302

By comparing the cytotoxicity of MRSA vs. MSSA strains (the first available 303

isolate from each patient), we found a comparable distribution and median-value 304



14

(Figure 6), which is in good agreement with the semi-quantitative blood agar 305

hemolysis data. However, culture supernatants of MSSA strains obtained from VAP 306

patients were overall more cytotoxic than those from colonized or VAT patients 307

based on group median-values. This trend was consistent with all three growth 308

media tested (Figure 6, panels A, B and C), but did not reach statistical 309

significance. Such trend was not observed with MRSA strains. 310

311

Virulence Testing in Murine Pneumonia Model 312

We investigated whether MRSA and MSSA strains, both with high and low 313

hemolytic activity, were associated with different virulence profiles in murine 314

pneumonia. The three strains characterized with high Hla-activity were associated 315

with 100% lethality at the highest dose tested, whilst the majority of animals 316

survived at the same dose of isolates characterized with low or no detectable in vitro 317

Hla-activity (Table 2). Two lower bacterial challenge doses also differentiated well 318

between strains with low and high Hla-activity. 319

These results confirmed the major role of Hla in the murine S. aureus 320

pneumonia pathogenesis, however, the survival outcomes did not reflect MRSA or 321

MSSA background or disease outcome (colonized vs. VAP). 322



15

DISCUSSION 323

324

In this study, we analyzed a unique collection of S. aureus isolates obtained 325

by serial sampling of ventilated patients and correlated our data with the diagnosis 326

of patients. We found that one third of the patients heavily colonized with S. aureus 327

developed VAP and the vast majority of them carried MSSA strains. VAP and MSSA 328

strains were much more likely to be detected in surgical ICU, while VAT and 329

colonized patients and MRSA strains were more likely to be found in the medical 330

ICU. Higher prevalence of VAP in general and MSSA induced VAP in particular in 331

surgical/trauma vs. medical ICU patients have been also reported by others (23,24). 332

The significantly lower prevalence of MRSA vs. MSSA in VAP patients and the 333

inverse prevalence in colonized patients, suggests that MRSAs are associated with 334

lower pneumonia causing potential, supporting the notion that antibiotic resistance is 335

associated with reduced virulence (25). 336

While MSSA strains were genetically very diverse, 87% of MRSA isolates 337

were ST5-II-t002 and ST5-II-t242. These molecular types were reported to belong to 338

the USA100 PFGE type, the dominant HA-MRSA in North America until recently 339

(26-30). The isolates characterized in this study were collected in 2010 in North 340

America (Burlington, MA). Since then, USA300 CA-MRSA strains replaced a 341

significant portion of USA100 infections in US hospitals (31). 342

The most important finding of this study is the significantly higher blood agar 343

hemolytic activity of MSSA strains isolated from VAP patients compared to those 344

from VAT or colonized patients. Such correlation was not observed for MRSA 345

isolates. By using a quantitative cytolytic assay with human airway cells, specifically 346

measuring alpha-hemolysin activity, we detected a good correlation with the blood 347

agar hemolysis profile. There was a tendency for higher Hla-activity of MSSA 348



16

associated with VAP compared to VAT and asymptomatic colonization (without 349

reaching statistical significance with this low sample size). 350

Alpha-hemolysin is one of the best characterized virulence factors in S. 351

aureus pneumonia pathogenesis in animals (10). However, very little evidence 352

exists that Hla contributes to pneumonia in clinical settings. Gene prevalence 353

studies are not supportive to assess the role of alpha-hemolysin, as the hla gene is 354

present in all S. aureus isolates sequenced to date (> 300 genomes) and is also 355

present in all isolates of this study (based on PCR analysis, data not shown). We 356

are aware of only one study that correlated the extent of hemolysis on sheep blood 357

agar to disease severity in the case of peritonitis (32). 358

It is known from the literature that alpha-hemolysin expression is influenced 359

by the composition of the culture medium and growth conditions. The presence of 360

red blood cells is known to up-regulate virulence factor expression (33), therefore 361

culturing on sheep blood agar as done for assessing the hemolysis profiles in a 362

semi-quantitative manner is a highly relevant condition. It was rather surprising to 363

detect a very broad range of Hla-activities from isolates representing the same 364

clonal lineage but isolated from different patients. We could even detect different 365

responses of isolates in terms of Hla expression to the different culture media. 366

Based on analysis of serially collected samples, Hla-activity remained unchanged 367

during colonization and over the course of disease within individual patients. This 368

suggests that strains have a “personal history” and undergo changes while 369

colonizing individuals. Hla-expression is under complex regulation (10). Mutations in 370

global regulators of toxin expression, such as the agr-system and Sar are described 371

in MRSA strains (34-36). Genetic alterations affecting hla itself are also implicated in 372

altered activity in both directions (37,38). Further genetic studies are needed to 373



17

uncover the molecular basis of the high variability in Hla-expression of clonally 374

related S. aureus strains. 375

It is highly relevant for the ICU patient population - being extensively treated 376

with multiple anti-infectives - that several antibiotics, even at sub-inhibitory 377

concentrations, have been shown to increase toxin production of S. aureus strains 378

(39,40). In several studies, improved therapeutic efficacy of linezolid over 379

vancomycin was observed, however this was not associated with a more rapid 380

microbial clearance suggesting that mechanism(s) other than direct antibacterial 381

activity are involved (41,42). It was shown in a therapeutic rabbit pneumonia model 382

that the use of linezolid was associated with lower toxin production by S. aureus and 383

lower cytokine levels compared to treatment with vancomycin (43). 384

Upon testing S. aureus isolates in murine pneumonia model, we observed 385

good correlation between Hla-activity and virulence (assessed by lethality). 386

However, we did not detect higher virulence of MSSA VAP isolates compared to 387

MRSA colonizing strains, suggesting that murine models do not reflect all important 388

aspects of disease causing potential in humans. This is most likely due to the 389

dominant role of Hla in murine pneumonia model documented in numerous studies 390

(reviewed in 10). Since mice are not natural hosts for S. aureus, a high inoculum is 391

needed to induce lethal infection (approximately 108 CFU), leading to overestimation 392

of toxin mediated virulence mechanisms. Moreover, several virulence factors of S. 393

aureus have been shown to be species (human) specific (e.g. certain leukotoxins) 394

and mice do not have the major risk factors associated with VAP (heavy 395

colonization, underlying diseases and mechanical ventilation). Therefore, alternative 396

animal models, which are more representative for the human setting, such as the 397

pneumonia model in ventilated pigs using more physiological bacterial challenge 398

doses (approximately 106 CFU), should also be considered for assessing virulence 399



18

of S. aureus strains (44). In addition, well-designed clinical research studies 400

performed in different geographic regions are of utmost importance to understand 401

the difference between MSSA and MRSA pathogenesis and the host-pathogen 402

interactions. Changes in innate host immunity, such as complement activity and 403

cytokine expression are also reported to be associated with or predictive for VAT 404

and/or VAP that can have a contribution in this clinical setting (45,46). 405

In conclusion, our data suggest that airway colonization with MSSA strains 406

with high alpha-hemolysin activity is a predictor of progression to VAP. Measuring 407

Hla-activity in cell based assays would be too laborious and clinically not applicable. 408

Therefore, we propose to assess the hemolytic activity of MSSA isolates from ETA 409

samples of ventilated patients on blood agar plates. Although this test is semi-410

quantitative and therefore to a certain extent subjective, it is very simple and 411

performed routinely in hospital microbiology laboratories. Molecular diagnostic 412

approaches, such as quantification of alpha-hemolysin mRNA with RT-PCR or 413

measuring the amount of Hla with antibody-based methods are worth being 414

explored. Based on our data, the results of such analyses can support differential 415

diagnosis (VAP vs VAT vs colonization) and be useful for identifying patients for 416

earlier antibiotic therapy or prophylaxis with Hla-neutralizing monoclonal antibodies. 417

418



19

NOTES 419

420

Acknowledgements. We thank our team members Gabor Nagy for the critical 421

review of the manuscript, Marisa Caccamo, Zehra Visram, Karin Gross and Marian 422

Fürsatz for the technical help, Yuxiu Lei (Lahey Clinic Medical Center, Burlington, 423

MA) for support in clinical data collection, Valéria Szijártó (Arsanis), Robin Ruthazer 424

(Biostatistics Research Center, Institute for Clinical Research and Health Policy 425

Studies, Tufts Medical Center, Boston, MA) and Abderrahim Oulhaj (College of 426

Medicine and Health Sciences, Al-Ain, United Arab Emirates) for help with the 427

statistical analyses, Knut Ohlsen (University of Wuerzburg, Germany) for providing 428

the RN4220 S. aureus strain and Fuminori Kato (Hiroshima University, Japan) for 429

sharing the pKFT shuttle-vector. 430

431

Potential conflicts of interest. LST, SMA, HRO and ENA are employees and 432

shareholders in Arsanis, Inc (Delaware, US). All other authors report no conflict of 433

interest. 434



20

REFERENCES 435

436

1. Bassetti M, Taramasso L, Giacobbe DR, Pelosi P. Management of ventilator-437

associated pneumonia: epidemiology, diagnosis and antimicrobial therapy. 438

Expert Rev Anti Infect Ther 2012; 10:585-596. 439

440

2. Valencia M, Torres A. Ventilator-associated pneumonia. Curr Opin Crit Care 441

2009; 15:30-35. Review. 442

443

3. Kollef MH, Hamilton CW, Ernst FR. Economic impact of ventilator-associated 444

pneumonia in a large matched cohort. Infect Control Hosp Epidemiol 2012; 445

33:250-256. 446

447

4. Craven DE, Lei Y, Ruthazer R, Sarwar A, Hudcova J. Incidence and 448

outcomes of ventilator-associated tracheobronchitis and pneumonia. Am J 449

Med 2013; 126:542-549. 450

451

5. Nseir S, Ader F, Marquette CH. Nosocomial tracheobronchitis. Curr Opin 452

Infect Dis. 2009;222:148-153. 453

454

6. Grgurich PE, Hudcova J, Lei Y, Sarwar A, Craven DE. Diagnosis of 455

ventilator-associated pneumonia: controversies and working toward a gold 456

standard. Curr Opin Infect Dis 2013; 26:140-150. 457

458

7. Nseir S, Favory R, Jozefowicz E, Decamps F, Dewavrin F, Brunin G, Di 459

Pompeo C, Mathieu D, Durocher A. Antimicrobial treatment for ventilator-460



21

associated tracheobronchitis: a randomized, controlled, multicenter study. 461

Crit Care 2008; 12:R62. 462

463

8. Watkins RR, David MZ, Salata RA. Current concepts on the virulence 464

mechanisms of meticillin-resistant Staphylococcus aureus. J Med Microbiol 465

2012; 61:1179-1193. 466

467

9. Tong SY, Chen LF, Fowler VG Jr. Colonization, pathogenicity, host 468

susceptibility, and therapeutics for Staphylococcus aureus: what is the clinical 469

relevance? Semin Immunopathol 2012; 34:185-200. 470

471

10. Berube BJ, Bubeck Wardenburg J. Staphylococcus aureus α-toxin: nearly a 472

century of intrigue. Toxins (Basel) 2013; 5:1140-1166. 473

474

11. Adhikari RP, Ajao AO, Aman MJ, Karauzum H, Sarwar J, Lydecker AD, 475

Johnson JK, Nguyen C, Chen WH, Roghmann MC. Lower antibody levels to 476

Staphylococcus aureus exotoxins are associated with sepsis in hospitalized 477

adults with invasive S. aureus infections. J Infect Dis 2012; 206:915-923. 478

479

12. Fritz SA, Tiemann KM, Hogan PG, Epplin EK, Rodriguez M, Al-Zubeidi DN, 480

Bubeck Wardenburg J, Hunstad DA. A serologic correlate of protective 481

immunity against community-onset Staphylococcus aureus infection. Clin 482

Infect Dis 2013; 56:1554-1561. 483

484

13. Bubeck Wardenburg J, Schneewind O. Vaccine protection against 485

Staphylococcus aureus pneumonia. J Exp Med 2008; 205:287-294. 486



22

487

14. Foletti D, Strop P, Shaughnessy L, Hasa-Moreno A, Casas MG, Russel M, 488

Bee C, Wu S, Pham A, Zeng Z, Pons J, Rajpal A, Shelton D. Mechanism of 489

action and in vivo efficacy of a human-derived antibody against 490

Staphylococcus aureus α-hemolysin. J Mol Biol 2013; 425:1641-1654. 491

492

15. Hua L, Hilliard JJ, Shi Y, Tkaczyk C, Cheng LI, Yu X, Datta V, Ren S, Feng 493

H, Zinsou R, Keller A, O'Day T, Du Q, Cheng L, Damschroder M, Robbie G, 494

Suzich J, Stover CK, Sellman BR. Assessment of anti-alpha toxin mAb for 495

prevention and treatment of Staphylococcus aureus induced pneumonia. 496

Antimicrob Agents Chemother 2013; 58:1108-1117 497

498

16. Cheung GYC, Otto M. The potential use of toxin antibodies as a strategy for 499

controlling acute Staphylococcus aureus infections. Expert Opin Ther Targets 500

2012; 16:601–612. 501

502

17. Stulik L, Malafa S, Hudcova J, Rouha H, Henics BZ, Craven DE, Sonnevend 503

AM, Nagy E. Evaluation of S. aureus Isolates from Endotracheal Aspirates of 504

Ventilated Patients [abstract]. International Symposium on Staphylococci & 505

Staphylococcal Infections – ISSSI 2014; A233. 506

507

18. Aanensen DM, Spratt, BG. The multi-locus sequence typing network: 508

mlst.net. Nucleic Acids Res. 2005; 33 (Web Server issue):W728-33. 509

510

19. Harmsen D, Claus H, Witte W, Rothganger J, Claus H, Turnwald D, Vogel U. 511

Typing of methicillin-resistant Staphylococcus aureus in a university hospital 512



23

setting by using novel software for spa repeat determination and database 513

management. J Clin Microbiol. 2003; 41:5442-5448. 514

515

20. Louie L, Goodfellow J, Mathieu P, Glatt A, Louie M, Simor AE. Rapid 516

detection of methicillin-resistant staphylococci from blood culture bottles by 517

using a multiplex PCR assay. J Clin Microbiol. 2002; 40:2786-2790. 518

519

21. Oliveira DC, de Lencastre H. Multiplex PCR strategy for rapid identification of 520

structural types and variants of the mec element in methicillin-resistant 521

Staphylococcus aureus. Antimicrob Agents Chemother. 2002; 46:2155-2161. 522

523

22. Sau S, Bhasin N, Wann ER, Lee JC, Foster TJ, Lee CY. The Staphylococcus 524

aureus allelic genetic loci for serotype 5 and 8 capsule expression contain the 525

type-specific genes flanked by common genes. Microbiol. 1997; 143:2395-526

2405. 527

528

23. Dudeck MA, Weiner LM, Allen-Bridson K, Malpiedi PJ, Peterson KD, Pollock 529

DA, Sievert DM, Edwards JR. National Healthcare Safety Network (NHSN) 530

report, data summary for 2012, Device-associated module. Am J Infect 531

Control. 2013; 41:1148-1166. 532

533

24. Agbaht K, Lisboa T, Pobo A, Rodriguez A, Sandiumenge A, Diaz E, Rello J. 534

Management of ventilator-associated pneumonia in a multidisciplinary 535

intensive care unit: does trauma make a difference? Intensive Care Med 536

2007; 33:1387-1395. 537

538



24

25. Collins J, Rudkin J, Recker M, Pozzi C, O'Gara JP, Massey RC. Offsetting 539

virulence and antibiotic resistance costs by MRSA. ISME J. 2010; 4:577-584. 540

541

26. Tenover FC, Tickler IA, Goering RV, Kreiswirth BN, Mediavilla JR, Persing 542

DH; MRSA Consortium. Characterization of nasal and blood culture isolates 543

of methicillin-resistant Staphylococcus aureus from patients in United States 544

Hospitals. Antimicrob Agents Chemother 2012; 56:1324-1330. 545

546

27. Nichol KA, Adam HJ, Roscoe DL, Golding GR, Lagace-Wiens PR, Hoban DJ, 547

Zhanel GG. Canadian Antimicrobial Resistance Alliance. Changing 548

epidemiology of methicillin-resistant Staphylococcus aureus in Canada. J 549

Antimicrob Chemother 2013; 68:i47-55. 550

551

28. Hudson LO, Murphy CR, Spratt BG, Enright MC, Elkins K, Nguyen C, 552

Terpstra L, Gombosev A, Kim D, Hannah P, Mikhail L, Alexander R, Moore 553

DF, Huang SS. Diversity of methicillin-resistant Staphylococcus aureus 554

(MRSA) strains isolated from inpatients of 30 hospitals in Orange County, 555

California. PLoS One 2013; 8:e62117. 556

557

29. Nair N, Kourbatova E, Poole K, Huckabee CM, Murray P, Huskins WC, 558

Blumberg HM. Molecular epidemiology of methicillin-resistant 559

Staphylococcus aureus (MRSA) among patients admitted to adult intensive 560

care units: the STAR*ICU trial. Infect Control Hosp Epidemiol 2011; 32:1057-561

1063. 562

563



25

30. Roberts JC. Community-associated methicillin-resistant Staphylococcus 564

aureus epidemic clone USA100; more than a nosocomial pathogen. 565

Springerplus 2013; 2:133. 566

567

31. O'Hara FP, Amrine-Madsen H, Mera RM, Brown ML, Close NM, Suaya JA, 568

Acosta CJ. Molecular characterization of Staphylococcus aureus in the 569

United States 2004-2008 reveals the rapid expansion of USA300 among 570

inpatients and outpatients. Microb Drug Resist. 2012; 18:555-561 571

572

32. Barretti P, Montelli AC, Batalha JE, Caramori JC, Cunha MLRS. The role of 573

virulence factors in the outcome of staphylococcal peritonitis in CAPD 574

patients. BMC Infect Dis 2009; 9:212. 575

576

33. Malachowa N, Whitney AR, Kobayashi SD, Sturdevant DE, Kennedy AD, 577

Braughton KR, Shabb DW, Diep BA, Chambers HF, Otto M, DeLeo FR. 578

Global changes in Staphylococcus aureus gene expression in human blood. 579

PLoS One 2011; 6:e18617. 580

581

34. Traber KE, Lee E, Benson S, Corrigan R, Cantera M, Shopsin B, Novick RP. 582

agr function in clinical Staphylococcus aureus isolates. Microbiology 2008; 583

154:2265-2274. 584

585

35. Villaruz AE, Bubeck Wardenburg J, Khan BA, Whitney AR, Sturdevant DE, 586

Gardner DJ, DeLeo FR, Otto M. A point mutation in the agr locus rather than 587

expression of the Panton-Valentine leukocidin caused previously reported 588



26

phenotypes in Staphylococcus aureus pneumonia and gene regulation. J 589

Infect Dis. 2009; 200:724-734. 590

591

36. Zielinska AK, Beenken KE, Joo HS, Mrak LN, Griffin LM, Luong TT, Lee CY, 592

Otto M, Shaw LN, Smeltzer MS. Defining the strain-dependent impact of the 593

Staphylococcal accessory regulator (sarA) on the alpha-toxin phenotype of 594

Staphylococcus aureus. J Bacteriol 2011; 193:2948–2958. 595

596

37. DeLeo FR, Kennedy AD, Chen L, Bubeck Wardenburg J, Kobayashi SD, 597

Mathema B, Braughton KR, Whitney AR, Villaruz AE, Martens CA, Porcella 598

SF, McGavin MJ, Otto M, Musser JM, Kreiswirth BN. Molecular differentiation 599

of historic phage-type 80/81 and contemporary epidemic Staphylococcus 600

aureus. Proc Natl Acad Sci USA 2011; 108:18091-18096. 601

602

38. Liang X, Hall JW, Yang J, Yan M, Doll K, Bey R, Ji Y. Identification of single 603

nucleotide polymorphisms associated with hyperproduction of alpha-toxin in 604

Staphylococcus aureus. PLoS One 2011; 6:e18428. 605

606

39. Rudkin JK, Laabei M, Edwards AM, Joo HS, Otto M, Lennon KL, O'Gara JP, 607

Waterfield NR, Massey RC. Oxacillin alters the toxin expression profile of 608

community-associated methicillin-resistant Staphylococcus aureus. 609

Antimicrob Agents Chemother 2014; 58:1100-7. 610

611

40. Nielsen LN, Roggenbuck M, Haaber J, Ifrah D, Ingmer H. Diverse modulation 612

of spa transcription by cell wall active antibiotics in Staphylococcus aureus. 613

2012; 5:457. 614



27

615

41. Wunderink RG, Mendelson MH, Somero MS, Fabian TC, May AK, 616

Bhattacharyya H, Leeper KV Jr., Solomkin JS. Early microbiological response 617

to linezolid vs vancomycin in ventilator-associated pneumonia due to 618

methicillin-resistant Staphylococcus aureus. Chest 2008; 134:1200-1207. 619

620

42. Sicot N, Khanafer N, Meyssonnier V, Dumitrescu O, Tristan A, Bes M, Lina 621

G, Vandenesch F, Vanhems P, Etienne J, Gillet Y. Methicillin resistance is 622

not a predictor of severity in community-acquired Staphylococcus aureus 623

necrotizing pneumonia-results of a prospective observational study. Clin 624

Microbiol Infect. 2013; 19:E142-8. 625

626

43. Diep BA, Afasizheva A, Le HN, Kajikawa O, Matute-Bello G, Tkaczyk C, 627

Sellman B, Badiou C, Lina G, Chambers HF. Effects of linezolid on 628

suppressing in vivo production of staphylococcal toxins and improving 629

survival outcomes in a rabbit model of methicillin-resistant Staphylococcus 630

aureus necrotizing pneumonia. J Infect Dis 2013; 208:75-82. 631

632

44. Luna CM, Sibila O, Agusti C, Torres A. Animal models of ventilator-633

associated pneumonia. Eur Respir J 2009; 33:182-188. 634

635

45. Ramírez P, Ferrer M, Gimeno R, Tormo S, Valencia M, Piner R, Menendez 636

R, Torres A. Systemic inflammatory response and increased risk for 637

ventilator-associated pneumonia: a preliminary study. Crit Care Med 2009; 638

37:1691-1695. 639

640



28

46. Martin-Loeches I, Papiol E, Almansa R, López-Campos G, Bermejo-Martin 641

JF, Rello J. Intubated patients developing tracheobronchitis or pneumonia 642

have distinctive complement system gene expression signatures in the pre-643

infection period: a pilot study. Med Intensiva 2012; 36:257-263.644

645



29

Table 1. Baseline characteristics of the study patients 646

Baseline Variables #

VAP

(n=15)

VAT (n=12)

Colonized (n=18)

p value MICU (n=31)

SICU (n=14)

p value ‡

Age (years) 60.1 ± 12.7 62.8 ± 14.0 67.7 ± 13.5 0.2875 * 61.4 ± 2.7 69.4 ± 2.6 0.0751

Male (%) 46.7 58.3 66.7 0.5291 * 83.9 50.0 0.0169

BMI 26.8 ± 6.9 36.8 ± 17.9 30.9 ± 5.8 0.0805 * 32.1 ± 2.3 28.9 ± 2.0 0.3966

Charlson Comorbidity Index 0.9 ± 1.0 1.4 ± 1.0 2.1 ± 2.0 0.1039 * 1.6 ± 0.3 1.4 ± 0.5 0.7722

APACHE II 18.4 ± 4.1 18.5 ± 5.7 20.4 ± 6.8 0.5535 * 20.5 ± 1.1 16.7 ± 1.2 0.0377

GCS 13.3 ± 0.4 12.6 ± 2.9 12.9 ± 2.8 0.9401 * 12.0 ± 1.1 14.0 ± 0.3 0.1275

GCS-Intubated 7.2 ± 2.1 6.9 ± 1.6 7.4 ± 1.3 0.8463 * 7.3 ± 0.4 6.7 ± 0.7 0.4375

Mortality in Hospital (%) 20.0 41.7 22.2 0.4035 * 25.8 28.6 0.8503

Antibiotics during ICU stay (%) 100 100 100 n.s. 100.0 100.0 n.s.

SICU (%) 53.3 16.7 22.2 0.0393 ** n.a. n.a. n.a.

Emergency Surgery Trauma (%) 46.7 25.0 22.2 0.2914 * 9.7 78.6 <0.0001

Acute Renal Failure (%) 13.3 8.3 33.3 0.1885 * 25.8 7.1 0.1541

Chronic Organ Insufficiency (%) 6.7 8.3 22.2 0.3725 * 16.1 7.1 0.4232

Duration of Hospitalization (days) 26.2 ± 10.2 20.9 ± 9.7 16.2 ± 6.8 0.1989 / 0.0027 / 0.1426 ‡ 19.1 ± 1.6 24.6 ± 3.1 0.0800

Duration of ICU stay (days) 24.2 ± 9.7 15.3 ± 6.7 10.3 ± 5.2 0.0146 / <0.0001 / 0.0374 ‡ 13.6 ± 1.4 22.2 ± 3.0 0.0037

Ventilation (days) 14.0 ± 4.7 11.1 ± 3.8 6.8 ± 2.5 0.1089 / <0.0001 / 0.0013 ‡ 9.6 ± 0.9 11.9 ± 1.3 0.1491



30

APACHE = acute physiology and chronic health evaluation; BMI = body mass index; MICU = medical intensive care unit; SICU = surgical intensive care unit, 647

GCS= Glasgow coma scale; GCS-Intubated = GCS of patients intubated prior scoring. 648

# Mean ± SD shown for Age, BMI, Charlson Comorbidity Index, APACHE II, GCS, GCS-Intubated, Hospitalization-Days, ICU-Days and Ventilation-Days 649

categories 650

n.s = not significant; n.a. = not applicable. 651

* One-way ANOVA of mean of three groups; ‡ Two-sample t-test (VAP vs. VAT / VAP vs. Colonized / VAT vs. Colonized). ** Fisher exact two-tailed probability 652

test653



31

Table 2. Alpha-hemolysin mediated in vitro cytotoxicity predicts in vivo 654

virulence in mice. Four MRSA and two MSSA isolates selected based on their Hla-655

activity and sheep blood agar hemolysis profile were used for intranasal (i.n.) 656

challenge of mice at three different bacterial challenge doses (CFU – colony forming 657

units given per mouse). Survival is indicated as percentage of live vs. total number 658

of animals used (n=10 mice per isolate and challenge dose). 659

660

Isolate ID Hemolysis

Pattern

Percent Survival post i.n. Challenge (Day 7) Molecular Type 9x10

8

CFU 3x108

CFU 1x108

CFU

P125.1 (col) 1+ 100 100 100 ST5-II-t002

P159.1 (col)

-

30 100 100 ST5-II-t002

P28.1 (VAP) - 100 100 100 ST30-t018

Average Survival 76.7 100 100

P178.1 (col) 4+ 0 0 80 ST5-II-t002

P44.1 (col)

3+

0 0 0 ST5-II-t002

P151.1 (VAP) 4+ 0 50 100 ST72-t148

Average Survival 0 16.7 60

661



32

FIGURE LEGENDS 662

663

Figure 1. Study patient selection. The flow chart describes the selection of 664

patients into the study and their grouping in three patient groups. 665

666

Figure 2. Collection of serial samples from ventilated patients. Data are shown 667

in three groups according to the diagnosis of patients: VAP, VAT and colonized. 668

Length of ventilation is given as time-line (days) and time of VAT- and VAP-669

diagnosis is indicated by green and red diamonds, respectively. MSSA and MRSA 670

positive ETA samples are marked with closed and open circles, respectively. 671

Strength of Hla-hemolysis is expressed as -, and 1+ to 4+ (according to Fig. 3A) and 672

color-coded as light to dark red = weak to strong Hla-activity; blue: non-hemolytic; 673

green: Hlb-activity (where Hla-induced hemolysis cannot be evaluated). Nasal 674

swabs from patients tested negative for S. aureus carriage upon admission but 675

positive at later stage are shown in italics. Patients, not heavily colonized (SQ-ETA 676

≤2+) are indicated in grey font. 677

678

Figure 3. Hemolytic phenotypes of S. aureus isolates on sheep blood agar 679

plates. A: S. aureus isolates were plated on sheep blood agar plates and 680

categorized as strong to weak or no beta-hemolysis (from 4+ to -) (characteristic for 681

alpha-hemolysin activity), according to the diameter of cleared halo around the 682

colonies based on semi-quantitative evaluation. Turbid halos surrounding the 683

colonies were accounted for beta-hemolysin (Hlb) expression. B: Distribution of S. 684

aureus strains with different blood agar hemolysis pattern among the different 685

patient groups is indicated. The p-value is given for VAP vs. non-VAP patient groups 686

calculated with the Fisher exact two-tailed probability test. 687



33

688

Figure 4. Cytolysis of human alveolar epithelial cells by S. aureus culture 689

supernatants reflects blood agar hemolysis strength. A549 cells were incubated 690

with culture supernatants of S. aureus strains grown in three different culture media; 691

A: TSB, B: CCY and C: RPMI supplemented with casamino acids. Cytolytic indices 692

of S. aureus isolates were determined as described in the methods and are grouped 693

for MSSA and MRSA according to the hemolysis profile on sheep blood agar. Each 694

patient is represented by the first available isolate, or by two isolates in case two 695

different molecular types were identified. MSSA and MRSA isolates are represented 696

by closed or open circles, respectively. Horizontal lines indicate group median 697

values. 698

699

Figure 5. Cytotoxic activities of S. aureus isolates greatly differ, but are very 700

similar in serial samples from a given patient. A549 cells were incubated with 701

culture supernatants of S. aureus isolates grown in three different media as 702

indicated and cytotoxicity measured as described in the methods. Mean values +/- 703

SEM obtained with independent biological replicates of the same isolates are 704

shown. A: Serial samples from individual patients with the genotypes indicated; B: 705

ST5-II-t002 isolates; C: ST8-t334 isolates. 706

707

Figure 6. Cytotoxicity profiles of strains involved in colonization, VAT and 708

VAP. A549 cells were incubated with culture supernatants of S. aureus isolates 709

grown in A: TSB, B: CCY and C: RPMI supplemented with casamino acids. 710

Cytotoxicity indices were determined as described in the methods and are shown for 711

MSSA and MRSA isolates in different groups of patients as indicated. Each patient 712

is represented by the first available isolate, or by two isolates in case two different 713



34

molecular types were identified. MSSA and MRSA isolates are indicated by closed 714

or open circles, respectively; and multiple isolates from the same patient by 715

squares. Horizontal lines indicate group median values. 716



Figure 1

Ventilated Patients(n=231)

Excluded (n=177)No S. aureus recovered from ETA

Patients with S. aureuspositive ETA

(n=56)

Excluded (n=11)Co-infection with other pathogen (n=10)CAP by S. aureus at admission (n=1)

Patients with only S. aureus positive ETA

(n=45)

Colonized Patients(n=18)

VAP Patients(n=15)

Including progressed from VAT (n=5)

Low ColonizationSQ-ETA: ≤ ++

(n=6)

VAT Patients(n=12)

Excluding progressed to VAP (n=5)

Heavy ColonizationSQ-ETA: +++ or ++++

(n=12)



Figure 2

Key:

Diagnosis of VAP

MRSA isolated from ETA

Start of ventilation End of ventilation

MSSA isolated from ETA

Diagnosis of VAT

MSSA MRSA

MSSA, MRSA

MSSA

negative; MSSA

Nasal SwabVentilation and Strain-Isolation OverviewETA

3+

2+

3+

4+ negative; MSSA

7 80

Days of Ventilation

9 10 11 123 4 5 6 13 21 22141 2

3+

MSSA

MSSA, MRSA

MSSA

4+

3+

-

negative

1+ MSSA

MSSA

3+ MSSA

4+ 2+ negative; MSSA

MRSA

2+ MSSA

MSSA

15 16 17 18 19 20

P6

P92

3+3+

P100

P151

P248

P165

P142

P117

4+

P28

Patient ID

P39

P22

P53

P190

P210

P158

>

MSSA MRSAETA

1+ negative

1+ MRSA

0 1 2 21 223 4 5 6 7 8 15 16 179 10 11 12 13 14 18 19 20

Days of Ventilation

2+ negative

negative

3+ 4+ MSSA

1+ MSSA

1+ MSSA

3+ MRSA

3+

- MRSA

2+ MRSA

MRSA

- MSSA

P159

P249

P3

P85

Patient ID

P212

P198

P16

P75

P125

P233

P193

P156

Nasal SwabVentilation and Strain-Isolation Overview

MSSA MRSA

Days of Ventilation

15 16 17 18 19 209 10 11 12 13 140 1 2 3 4 5 6 7 8 21 22

P192 3+ MRSA

P134 3+ MRSA

P160 3+ negative

P46 1+ negative

P112 1+ MSSA

P97 1+ MRSA

P35 2+ MRSA

P58 3+ negative

P42 3+ MSSA

P140 2+ negative

P166 3+ MRSA

P5 3+ negative

P9 3+ MRSA

P4 1+ MSSA

P50 3+ negative

P45 1+ MSSA

P78 MSSA

P8 - MSSA

Patient ID Ventilation and Strain-Isolation OverviewETA

Nasal Swab

VA

P P

ati

en

tsV

AT

Pa

tie

nts

Co

lon

ize

d P

ati

en

ts



A

B

4+ 3+ 2+ 1+ - Hlb

4+ 3+ 2+ 1+ - Hlb

Per

cen

t o

f Is

ola

tes

Col

VAT

VAP

100

80

60

40

20

0

n=4 n=8 n=2 n=7 n=3 n=2

p=0.0123

MSSA

Figure 3

MRSAn.s.

4+ 3+ 2+ 1+ - Hlb

Per

cen

t o

f Is

ola

tes

Col

VAT

VAP

n=1 n=11 n=5 n=3 n=1 n=2

100

80

60

40

20

0



Cyt

oly

tic

Ind

ex

4+ 3+ 2+ 1+ - Hlb 4+ 3+ 2+ 1+ - Hlb

0

10

20

30

40

50

MSSA MRSA

Figure 4

AC

yto

lyti

c In

de

x

4+ 3+ 2+ 1+ - Hlb 4+ 3+ 2+ 1+ - Hlb

0

10

20

30

40

50

MSSA MRSA

Cyt

oly

tic

Ind

ex

4+ 3+ 2+ 1+ - Hlb 4+ 3+ 2+ 1+ - Hlb

0

5

10

15

MSSA MRSA

B

C



Figure 5

A

B

Cyto

lyti

c I

nd

ex

TS

B a

nd

CC

Y

Cyto

lytic

Ind

ex

RP

MI-C

AS

ST

81

-t1

27

ST

81

-t1

27

ST

5-t

06

7

ST

5-t

06

7

ST

5-t

06

7

ST

5-t

06

7

ST

30

-t7

26

ST

30

-t7

26

ST

30

-t7

26

ST

30

-t7

26

ST

81

-t1

30

69

ST

81

-t1

30

69

ST

5-I

I-t0

02

ST

5-I

I-t0

02

ST

5-I

I-t0

02

ST

5-I

I-t0

02

ST

45

-t0

40

ST

39

-t2

13

2

ST

39

-t2

13

2

ST

5-t

00

3

ST

5-t

00

3

ST

97

-t3

59

ST

97

-t3

59

ST

97

-t3

59

ST

8-I

V-t

00

8

0

10

20

30

40

0

5

10

15

P16 P158 P75 P6 P39 P198

Bacterial

Genotype

Patient-ID

Isolate N° 1 2 21 21 21 2143 43 43 5

P22

21 43 21 43

Cyto

lyti

c I

nd

ex

TS

B a

nd

CC

Y

Cyto

lytic

Ind

ex

RP

MI-C

AS

ST

5-I

I-t0

02

ST

5-I

I-t0

02

ST

5-I

I-t0

02

ST

5-I

I-t0

02

ST

5-I

I-t0

02

ST

5-I

I-t0

02

ST

5-I

I-t0

02

ST

5-I

I-t0

02

ST

5-I

I-t0

02

ST

5-I

I-t0

02

ST

5-I

I-t0

02

ST

5-I

I-t0

02

ST

5-I

I-t0

02

0

10

20

30

40

0

5

10

15

P156

1 2

P58

1 2

P193

1 2

P50

1 2

P125

1 2 3

P100

1

P159

Bacterial

Genotype

Patient-ID

Isolate N° 1

Cyto

lyti

c I

nd

ex

TS

B a

nd

CC

Y

Cyto

lytic

Ind

ex

RP

MI-C

AS

ST

8t3

34

ST

8t3

34

ST

8t3

34

ST

8t3

34

0

10

20

30

40

50

60

0

5

10

15

20

25

Bacterial

Genotype

Patient-ID

Isolate N°

P129

1 2

P190

1 2

C

Cyto

lyti

c I

nd

ex

TS

B a

nd

CC

Y

Cyto

lytic

Ind

ex

RP

MI-C

AS

ST

8t3

34

ST

8t3

34

ST

8t3

34

ST

8t3

34

0

10

20

30

40

50

60

0

5

10

15

20

25

Bacterial

Genotype

Patient-ID

Isolate N°

P129

1 2

P190

1 2

TSB CCY RPMI-CAS



A

Figure 6

B

C

Cyt

oly

tic

Ind

ex

MSSA MRSA MSSA MSSA MSSA MRSA MRSA MRSA

0

10

20

30

40

50

VAP VAT col VAP VAT coltotal

Cyt

oly

tic

Ind

ex


0

10

20

30

40

50


Cyt

oly

tic

Ind

ex


0

5

10

15




ONLINE DATA SUPPLEMENT 1

2

METHODS 3

4

Bacterial Strains and Culture Conditions 5

S. aureus was routinely cultured in tryptic soy broth (TSB, Sigma-Aldrich, 6

Steinheim, Germany) or on Columbia agar plates containing 5% sheep blood (COS, 7

bioMérieux, Marcy-l'Etoile, France). MSSA/MRSA-status was confirmed by selective-8

plating on chromID® SAID- and MRSA-plates (bioMérieux) incubated for 24 h at 9

37°C. Culture supernatants were collected by centrifugation (15 min, 4°C, 5000 rpm) 10

of S. aureus cultures grown for 16 h, at 37°C in TSB-medium, CCY-medium (3% 11

yeast extract (Fisher BioReagents, Vienna, Austria), 2 % bacto-casamino acids 12

(Fisher BioReagents), 2.3% sodium pyruvate (Fisher BioReagents), 0.63% Na2HPO4 13

(Fisher BioReagents), 0.041% KH2PO4 (Sigma-Aldrich), pH 6.7) or RPMI-medium 14

supplemented with bacto-casamino acids (RPMI-CAS; RPMI-1640 (InvitrogenTM, 15

LifeTechnologies, Paisley, UK), 1% bacto-casamino acids (Fisher BioReagents). 16

The isogenic S. aureus hla gene-deletion mutant was generated in the 17

sequenced, USA300 CA-MRSA strain TCH1516 (ATCC® BAA1717TM, LGC 18

Standards, Teddington, UK) with homologous recombination based on previously 19

published methods [E1] using the primer-pairs listed in Table E2 and the pKFT gene-20

deletion vector, a previously reported shuttle-vector for E. coli and S. aureus [E2]. 21

22

Preparation of Genomic DNA and Genotyping of S. aureus Strains 23

Total genomic DNA (gDNA), was isolated from pelleted over-night cultures of 24

S. aureus (20 ml TSB, 16 h, 37°C, 200 rpm) using the DNeasy Blood and Tissue kit 25

(QIAGEN, Hilden, Germany). 26



Multi-locus sequence typing (MLST; [E3]), spa-genotyping [E4], multiplexing 27

PCR (16S rRNA, nuc and mecA [E5]), SCCmec-typing [E6], as well as capsule-28

typing [E8] PCRs were performed as previously described. Hla genes were amplified 29

by a specific primer-pair. All oligonucleotides used in this study are listed in 30

Supplementary Table E2. PCRs were performed using purified, gDNA and the 31

Phusion High-Fidelity PCR Master Mix with HF Buffer (Thermo Scientific, Waltham, 32

MA, USA). 33

34

Evaluation of the Cytotoxicity of S. aureus Culture Supernatants Using a 35

Human Alveolar Epithelial Cell Line, A549 36

A549 cells (ATCC® CCL-185™, LGC Standards), were used to evaluate the 37

cytotoxicity of Hla on lung epithelial cells. Cells were grown as adherent cultures in 38

Ham´s F12-K medium (Gibco®, LifeTechnologies, Paisley, UK) supplemented with 1x 39

penicillin-streptomycin solution (10x stock, PAA Laboratories, Pasching, Austria) and 40

10% fetal calf serum (FCS, Sigma-Aldrich) in 175 cm2 tissue culture treated cell 41

culture flasks. For the cytotoxicity assays, 2x104 cells were seeded in cell culture 42

medium, into 96-well half-area plates 16 h prior to the experiment. Cells were 43

incubated with serial dilution of bacterial culture supernatants (50 µl) for 3 h, 37°C, 44

5% CO2) diluted in 50 µl fresh cell culture medium (F12-K medium+ 5% FCS). Cell 45

viability was measured (Cell Titer-Glo® Luminescent Cell Viability Assay Kit, 46

Promega, Madison, WI, USA) at 590/35 nm in a plate reader (Synergy HT, BioTek, 47

Bad Friedrichshall, Germany). Data are presented as cytolytic index that is the EC50-48

value indicating the fold-dilution of a given culture supernatant needed to achieve 49

50% of cell death and calculated from the dose-reponse curves using the Prism 6 50

(GraphPad, La Jolla, CA, USA) software-package. 51



For Hla-inhibition experiments, bacterial culture supernatants were pre-52

incubated with a a single-specific, monoclonal anti-Hla antibody (10µg/ml; Arsanis 53

Biosciences, Vienna, Austria) prior to incubation with cells. Percent cell viability was 54

calculated relative to mock-treated cells. Recombinant Hla and an isotype-matched 55

control mAb were used as positive and negative assay controls, respectively. 56

57

Animals and Murine Pneumonia Model 58

Female, six to eight weeks old BALB/cJRj mice were purchased from the SPF 59

breeding-facility of Janvier Labs (Saint-Berthevin Cedex, France) and were given 60

food and water ad libitum. The animals were housed and cared for in accordance to 61

Austrian Law for animal testing (BGBl Nr. 501/1989), which was reviewed and 62

approved by the Federal State Government of the City of Vienna (MA58, permit-63

number: M58/006304/2012/5). 64

The bacterial inocula were prepared by growing S. aureus to mid-logarithmic 65

phase (OD600 of 0.5) in TSB medium (37°C, 200 rpm). Bacterial cells were pelleted 66

and washed with sterile DPBS (pH 7.5; Gibco®) prior to storage as glycerol-stocks at 67

-80°C. The total number of CFU in each inoculum was quantified from three stocks 68

per strain by plating triplicates of serial dilutions. Prior to the challenge, thawed 69

bacterial pellets were re-suspended and washed once before setting the challenge-70

dose by dilution in sterile DPBS. For the pneumonia model, mice were anesthetized 71

with an intraperitoneal injection of 200 µl of 10% ketamine (Ketamidor®, Richter-72

Pharma, Wels, Austria) and 2% xylazine (Rompun®, Bayer, Vienna, Austria) in sterile 73

DPBS, followed by intranasal application of different challenge-doses of S. aureus in 74

a total volume of 40 µl. Animals were monitored daily for alterations in general health 75

for 7 days after the bacterial challenge. 76



References for the Online Data Supplement 77

78

E1. McNamara PJ. Genetic Manipulation of Staphylococcus aureus. In: Lindsay 79

JA, editor. Staphylococcus: Molecular Genetics, 1st ed., Norfolk, UK: Caister 80

Academic Press; 2008. p. 100-102. 81

82

E2. Kato K, Sugai M. A simple method of markerless gene deletion in 83

Staphylococcus aureus. J. Microbiol. Methods 2011; 87:76-81. 84

85

E3. Aanensen DM, Spratt, BG. The multi-locus sequence typing network: mlst.net. 86

Nucleic Acids Res. 2005; 33 (Web Server issue):W728-33. 87

88

E4. Harmsen D, Claus H, Witte W, Rothganger J, Claus H, Turnwald D, Vogel U. 89

Typing of methicillin-resistant Staphylococcus aureus in a university hospital setting 90

by using novel software for spa repeat determination and database management. J 91

Clin Microbiol. 2003; 41:5442-5448. 92

93

E5. Louie L, Goodfellow J, Mathieu P, Glatt A, Louie M, Simor AE. Rapid 94

detection of methicillin-resistant staphylococci from blood culture bottles by using a 95

multiplex PCR assay. J Clin Microbiol. 2002; 40:2786-2790. 96

97

E6. Oliveira DC, de Lencastre H. Multiplex PCR strategy for rapid identification of 98

structural types and variants of the mec element in methicillin-resistant 99

Staphylococcus aureus. Antimicrob Agents Chemother. 2002; 46:2155-2161. 100

101



E7. Sau S, Bhasin N, Wann ER, Lee JC, Foster TJ, Lee CY. The Staphylococcus 102

aureus allelic genetic loci for serotype 5 and 8 capsule expression contain the type-103

specific genes flanked by common genes. Microbiol. 1997; 143:2395-2405. 104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127



Supplemental Figure Legends 128

129

Figure E1. Human lung epithelial cell lysis by S. aureus culture supernatants is 130

mediated by alpha-hemolysin. A549 cells were incubated with culture supernatants 131

of S. aureus strains and cell viability was measured based on cellular ATP content. 132

Data are expressed as mean values ± SEM obtained with independent biological 133

replicates. A: TCH1516 (USA300) wild-type and isogenic ∆hla strains grown in three 134

different culture media; B: TSB culture supernatants of the TCH1516 wild-type strain 135

and four study isolates with high Hla-activity were used for intoxication in the 136

presence or absence of 66.7 nM of an Hla-neutralizing monoclonal antibody and a 137

corresponding isotype control mAb. Percent cell-viability was determined compared 138

to mock-treated cells. 139

140

Table E1. List of S. aureus genotypes identified in the study. 141

142

Table E2. Oligonucleotides used in the study. 143

144



297x420mm (300 x 300 DPI)



Table E1. List of S. aureus genotypes identified in the study.

Clonal Type Total Number Number of Patients MRSA-SCCmec / Capsule-Types MLST spa-type of Patients Colonized VAT VAP MSSA ST5 t002 16 9 5 2 MRSA-II 5 ST5 t003

1 1

MSSA 5

ST5 t045 2 1 1 MRSA-II 5 ST5 t067

1 1

MSSA 5

ST5 t242 2 1 1 MRSA-II 5 ST5 t306

1 1

MSSA 5

ST5 t548 1 1 MSSA 5 ST8 t008

2 1 1

MRSA-IV 5

ST8 t13069 1 1 MSSA 5 ST8 t334

1 1

MSSA 5

ST15 t084 1 1 MSSA n.t. ST15 t13068

1 1

MSSA n.t.

ST25t081 1 1 MSSA 5 ST30t012

1 1

MSSA 8

ST30 t018 1 1 MSSA 8 ST30 t726

1 1

MSSA 8

ST39 t2132 1 1 MSSA 8 ST45 t040

1 1

MSSA 8

ST45 t550 1 1 MSSA 8 ST72 t148

1 1

MSSA 5

ST81 t127 1 1 MSSA n.t. ST87 t216

1 1

MSSA 8

ST97 t359 1 1 MSSA 5 ST101 t056

1 1

MSSA 8

ST105 t002 1 1 MRSA-II 5 ST121 t645

1 1

MSSA 8

ST146 t002 1 1 MSSA 5 ST188t189

1 1

MSSA 8

ST199 t084 1 1 MSSA n.t. ST474 t127

1 1

MSSA 8

ST1970 t065 1 1 MSSA 8 total 49 18 13 18 23 / 26 different 31 10 9 17 5 / 26 MRSA / MSSA 12 / 6 6 / 7 4 / 14 MRSA - methicillin resistant S. aureus; MSSA - methicillin susceptible S. aureus; ST - sequence-type n.t. - non-typeable; VAT - ventilator-associated tracheobronchitis; VAP - ventilator-associated pneumonia



arc up TTGATTCACCAGCGCGTATTGTC

arc dn AGGTATCTGCTTCAATCAGCG

aro up ATCGGAAATCCTATTTCACATTC

aro dn GGTGTTGTATTAATAACGATATC

glp up CTAGGAACTGCAATCTTAATCC

glp dn TGGTAAAATCGCATGTCCAATTC

gmk up ATCGTTTTATCGGGACCATC

gmk dn TCATTAACTACAACGTAATCGTA

pta up GTTAAAATCGTATTACCTGAAGG

pta dn GACCCTTTTGTTGAAAAGCTTAA

tpi up TCGTTCATTCTGAACGTCGTGAA

tpi dn TTTGCACCTTCTAACAATTGTAC

yqi up CAGCATACAGGACACCTATTGGC

yqi dn CGTTGAGGAATCGATACTGGAAC

1095F (spa fwd) AGACGATCCTTCGGTGAGC

1517R (spa rev) GCTTTTGCAATGTCATTTACTG

cap5-1 GGTTTGCTGAAAAACCAGTC

cap5-2 CCTCATATGCTCCTACATTT

cap8-1 GCGCTACAAACATTAAGCAT

cap8-2 TTCTTAGCCTGCTGGCATC

16S-F AGAGTTTGATCATGGCTCAG

16S-R GGACTACCAGGGTATCTAAT

mecA-1 AAAATCGATGGTAAAGGTTGGC

mecA-2 AGTTCTGCAGTACCGGATTTGC

nuc-1 GCGATTGATGGTGATACGGTT

nuc-2 AGCCAAGCCTTGACGAACTAAAGC

CIF2 F2 TTCGAGTTGCTGATGAAGAAGG

CIF2 R2 ATTTACCACAAGGACTACCAGC

KDP F1 AATCATCTGCCATTGGTGATGC

KDP R1 CGAATGAAGTGAAAGAAAGTGG

MECI P2 ATCAAGACTTGCATTCAGGC

MECI P3 GCGGTTTCAATTCACTTGTC

DCS F2 CATCCTATGATAGCTTGGTC

DCS R1 CTAAATCATAGCCATGACCG

RIF4 F3 GTGATTGTTCGAGATATGTGG

RIF4 R9 CGCTTTATCTGTATCTATCGC

RIF5 F10 TTCTTAAGTACACGCTGAATCG

RIF5 R13 GTCACAGTAATTCCATCAATGC

IS431 P4 CAGGTCTCTTCAGATCTACG

pUB110 R1 GAGCCATAAACACCAATAGCC

pT181 R1 GAAGAATGGGGAAAGCTTCAC

MECA P4 TCCAGATTACAACTTCACCAGG

MECA P7 CCACTTCATATCTTGTAACG

LST_0001 CTGTCGCTAATGCCGCAGATTC

LST_0002 GAAGTCCAGTGCAATTGGTAGTC

ko-LST_0041 CCGCGGATCCCTATTAGATATTTCTATGTAATGGC

ko-LST_0042 CCCATCCACTAAACTTAAACA CATCATCCTTCTATTTTTTAAAACG

ko-LST_0043 TGTTTAAGTTTAGTGGATGGGGTAAATTATTTGTTCATGTACAAA

ko-LST_0044 TGGCACCCGGGCGTTAATTTCCCAATCATTAAAAACC

enzymatic restriction sites: underlined

fusion overlaps: italic

this study

tpi_

yqil

spa -typing E4

Multilocus

sequence-typing

(MLST)

arcc

E3

aroe

glpf

gmk_

pta_

gene-deletion hla this study

capsule-typing

type 5

E7

type 8

confirmation of

MSSA/MRSA

16S rRNA

E5mecA

nuc

SCCmec-typing E6

gene-typing hla

Table E2. Oligonucleotides used in the study.

Purpose Locus/Type Primer Nucleotide Sequence (5' - 3') Reference



α-Hemolysin Activity of Methicillin-Susceptible Staphylococcus aureus ...

Documents

Transcript of α-Hemolysin Activity of Methicillin-Susceptible Staphylococcus aureus ...