1

Immunomodulatory role of Clarithromycin in Acinetobacter baumannii infection 1

via Neutrophil Extracellular Traps formation. 2

Running Title: Clarithromycin in NET formation 3

Theocharis Konstantinidis1,*, Konstantinos Kambas1,*, Alexandros Mitsios1, Maria 4

Panopoulou2, Victoria Tsironidou1, Erminia Dellaporta3, Georgios Kouklakis3, 5

Athanasios Arampatzioglou1, Iliana Angelidou1, Ioannis Mitroulis4, Panagiotis 6

Skendros1,5, Konstantinos Ritis1,5,§ 7

*These authors contributed equally 8

1Laboratory of Molecular Hematology, Department of Medicine, Democritus 9

University of Thrace, Alexandroupolis, Greece. 10

2Laboratory of Microbiology, Democritus University of Thrace, University General 11

Hospital of Alexandroupolis, Alexandroupolis, Greece. 12

3Gastrointestinal Endoscopy Unit, Democritus University of Thrace, University 13

General Hospital of Alexandroupolis, Alexandroupolis, Greece. 14

4Department of Clinical Pathobiochemistry and Institute for Clinical Chemistry and 15

Laboratory Medicine, Faculty of Medicine, Technische Universität Dresden, Dresden, 16

Germany. 17

5First Department of Internal Medicine, Democritus University of Thrace, University 18

General Hospital of Alexandroupolis, Alexandroupolis, Greece. 19

20

21

AAC Accepted Manuscript Posted Online 7 December 2015Antimicrob. Agents Chemother. doi:10.1128/AAC.02063-15Copyright © 2015, American Society for Microbiology. All Rights Reserved.

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

2

§Correspondence to Konstantinos Ritis, MD, PhD, University Hospital of 22

Alexandroupolis, Dragana, Alexandroupolis, 68100, GREECE, Tel: +30 25513 23

51103, Fax: +30 25510 30378, E-mail address: kritis@med.duth.gr 24

Keywords: Clarithromycin, Neutrophil Extracellular Traps, Acinetobacter baumannii, 25

LL-37, immunomodulation. 26

27

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

3

Abstract 28

Macrolide antibiotics have been shown to act as immunomodulatory molecules in 29

various immune cells. However, their effect on neutrophils has not been extensively 30

investigated. In this study we investigated the role of macrolide antibiotics in the 31

generation of NETs. 32

By assessing ex vivo and in vivo NET formation we demonstrated that clarithromycin 33

is able to induce NET generation both in vitro and in vivo. Clarithromycin utilizes 34

autophagy in order to form NETs and these NETs are decorated with antimicrobial 35

peptide LL-37. Clarithromycin-induced NETs are able to inhibit Acinetobacter 36

baumannii growth and biofilm formation in a LL-37-dependent manner. Additionally, 37

LL-37 antimicrobial function depends on NET-scaffold integrity. 38

Collectively, these data expand the knowledge on the immunomodulatory role of 39

macrolide antibiotics via the generation of LL-37-bearing NETs, which demonstrate 40

LL-37-dependent antimicrobial activity and biofilm inhibition against A. baumannii. 41

42

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

4

Introduction 43

In the last two decades macrolides have been demonstrated as potential 44

immunomodulatory agents, due to their ability to induce the activity of various 45

immune cells and their regulatory role in chemokine and cytokine production (1,2). 46

The addition of macrolide antibiotics (azithromycin, clarithromycin or erythromycin) 47

to the standard treatment of severe community-acquired pneumonia (CAP) due to 48

macrolide-resistant microorganisms led to decreased mortality of patients, providing 49

evidence for their effect on immunomodulation (3). Moreover, in ventilator-50

associated pneumonia (VAP) patients with clarithromycin-resistant bacterial 51

infections, combined treatment with clarithromycin was positively correlated with the 52

resolution time of the infection (4). Although it was proposed that some macrolides 53

could induce degranulation and bacterial killing in polymorphonuclear neutrophils 54

(PMNs), the knowledge on their role on these cells is very limited (5,6). 55

Neutrophils are the most abundant circulating inflammatory cell and the first line of 56

defense against pathogens. They employ three major strategies to fight against 57

microbes: phagocytosis, degranulation and the release of neutrophil extracellular traps 58

(NETs) (7,8). The discovery of the mechanism of NETs has redefined our perception 59

of neutrophil role and functions. NETs are composed of chromatin which is decorated 60

with neutrophilic proteins (cytoplasmic, granular, nuclear). NETs do not only elicit 61

their antimicrobial effect through pathogen immobilization via entrapment (7), but 62

also have a direct microbicidal effect that depends on antimicrobial peptides (7,8), 63

histones (7,9) or DNA (10). On the other hand, it has been also reported that some 64

microorganisms escape NETs, such as Acinetobacter baumannii (11), or possess 65

inhibitory strategies against this mechanism (12,13). 66

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

5

One of the strategies that some microbes utilize to protect their colonies is the 67

formation of biofilms. Biofilms are constituted by exopolymeric extracellular matrix 68

comprised of lipids, proteins that frequently exhibit amyloid-like properties, exo DNA 69

and exopolysaccharides (14,15). Biofilms facilitate the attachment and growth of 70

microbes to abiotic surfaces and also render bacterial cultures less susceptible to 71

antibiotics (16). Recently, it has been shown that this mechanism is also found in a 72

multi-drug resistant nosocomial gram-negative bacterium, A. baumannii, and the 73

formation of biofilm contributed to the multi-resistance of this microorganism to 74

antibiotics (17). 75

Considering the above, we investigated the role of macrolide antibiotics in the 76

generation of NETs, how these NETs could affect pathogens that naturally are not 77

able to trigger the mechanism of NETosis and what is their effect on bacterial defense 78

mechanisms such as the formation of biofilm. 79

80

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

6

Materials and Methods 81

Patients and sample collection. 82

To examine the in vivo NET generation potential of clarithromycin, ten patients 83

suffering from Helicobacter pylori positive gastritis were recruited since these 84

patients are administered clarithromycin as standard treatment. Five patients with H. 85

pylori negative gastritis (omeprazole monotherapy) and 10 age-sex matched control 86

individuals were also recruited (details in Table 1). Additionally, to investigate the 87

ability of A. baumannii to induce in vivo NET generation, 4 septic patients with A. 88

baumannii bacteremia were enrolled in this study. 89

PMNs and sera were obtained from patients with H. pylori positive gastritis 12-24 90

hours before treatment (n =10) and during the 5th day of standard eradication 91

regimen. Moreover, PMNs and sera were obtained from the patients with H. pylori 92

negative gastritis before and during the 5th day of omeprazole treatment, as well as, 93

from the patients with A. baumannii bacteremia at various time-points of the disease 94

course. 95

The study protocol was in accordance with the Declaration of Helsinki and was 96

approved by the ethics review board of the University Hospital of Alexandroupolis. 97

Written informed consent was obtained from each individual. 98

Reagents 99

Antibiotics: Daptomycin (Novartis) was dissolved at a concentration of 50 mg/ml per 100

stock, Ciprofloxacin (Vianex AE) Stock 200mg/100ml or 2mg/ml, Amoxicillin 101

(Cooper pharmaceuticals) was dissolved at a concentration of 1g/vial per stock, 102

Azithromycin (Anfarm Hellas) was dissolved at a concentration of 100 mg/ml per 103

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

7

stock, Clarithromycin (Anfarm Hellas) was dissolved at a concentration of 50 mg/ml 104

per stock. 105

Antibodies 106

Primary antibodies: a mouse anti- myeloperoxidase (MPO) mAb, rabbit anti-107

neutrophil elastase (NE) mAb (Santa Cruz Biotechnology Inc) and mouse anti- 108

cathelicidin antimicrobial peptide LL-37 (LL-37) mAb (Santa Cruz Biotechnology), 109

anti LC3bAb and anti p62Ab, DAPI (Sigma-Aldrich) was used for DNA staining. 110

Polyclonal rabbit anti-mouse Alexa fluor 488 antibody and polyclonal goat anti-rabbit 111

Alexa fluor 647 (Invitrogen) were used as secondary antibodies. Anti-MPO mAb 112

(Upstate, Millipore) was used as a capture antibody for MPO-DNA complex ELISA. 113

MSU crystal preparation 114

Monosodium urate crystals (MSU) were prepared as previously described (18) in 115

pyrogen-free conditions. In particular, urate acid sodium salt (Sigma-Aldrich) was 116

dissolved in 1 M NaOH (25mg/ml) and boiled for 2 hours at 200°C prior to 117

crystallization. The solution was left to cool at room temperature and filtered through 118

a 0.2 μM filter. It was then incubated at room temperature for 7 days. The resulting 119

crystals were washed with ethanol and acetone and allowed to air dry under sterile 120

conditions. 121

PMN isolation, stimulation and inhibition studies 122

PMNs were isolated from heparinized venous blood and cultured in 5% CO2 at 37o C 123

in a total volume 500 μl of Roswell Park Memorial Institute (RPMI) medium (Gibco 124

BRL) in the presence of 2% heterologous healthy donor serum and different 125

stimulatory agents. For in vitro studies, PMNs from control individuals were treated 126

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

8

with antibiotics for 3 hours. Furthermore, PMNs were treated with other inflammatory 127

stimuli MSU crystals (non-infectious) or Phorbol 12-myristate 13-acetate (PMA) for 128

determination of LL-37 expression on NETs. 129

To study autophagy induction, neutrophils were treated with clarithromycin for 90 130

min and for 3h. To inhibit the autophagic machinery, neutrophils were pretreated with 131

Bafilomycin A1 (30 nM; Sigma-Aldrich) for 30 min before Clarithromycin 132

stimulation. 133

Immunofluorescence 134

NETs generation, preparation and visualization by immunofluorescence confocal 135

microscopy was performed as previously described (19). Samples were stained using 136

anti-MPO mAb, anti- LL-37 mAb or anti-neutrophil elastase (NE) polyclonal Ab. A 137

polyclonal rabbit anti-mouse Alexa fluor 488 antibody or a polyclonal goat anti-rabbit 138

Alexa fluor 647 (Invitrogen) were utilized as secondary antibodies. DAPI (Sigma-139

Aldrich) was used for DNA counterstaining. Visualization was performed in a 140

confocal microscope (Spinning Disk Andor Revolution Confocal System, Ireland) in 141

a PLAPON 606O/TIRFM-SP, NA 1.45 and UPLSAPO 100XO, NA 1.4 objectives 142

(Olympus). Percentage of NET releasing PMNs was calculated by counting 200 cells 143

in total. 144

To study autophagy induction samples were stained with anti-LC3b polyclonal 145

antibody, followed by a polyclonal goat anti-rabbit Alexa Fluor 488 antibody utilized 146

as secondary. DNA was counterstained using DAPI (Sigma-Aldrich). 147

NETs isolation 148

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

9

A number of 1.5x106 ex vivo or in vitro-stimulated neutrophils was cultured for 4 149

hours. After medium removal, cells were washed with RPMI. RPMI was added to 150

each well and NETs were collected on supernatant medium after vigorous agitation. 151

The medium was centrifuged at 20x g for 5 min and supernatant phase, containing 152

NETs, was collected and stored at -20°C until use. 153

MPO/DNA complex ELISA 154

To quantify NET release by ex vivo or in vitro-stimulated PMNs, MPO/DNA complex 155

was measured in NET structures isolated from 1.5x106 PMNs as previously described 156

(19) or in plasma from patients and control individuals. More specifically, capture 157

antibody, 5 μg/ml anti-MPO mAb (Upstate, Millipore) was coated onto 96-well plates 158

(dilution 1:500 in 50 μl) overnight at 4°C. After three washes (300 μl each), 20 μl of 159

samples was added to the wells with 80 μl incubation buffer containing a peroxidase-160

labeled anti-DNA mAb (Cell Death ELISAPLUS, Roche; dilution 1:25). The plate 161

was incubated for 2 hours, shaking at 300 rpm at room temperature. After three 162

washes (300 μl each), 100 μl peroxidase substrate (ABTS) was added. Absorbance at 163

405-nm wavelength was measured after 20 minutes incubation at room temperature in 164

the dark. NET release was represented as % increase compared to controls. 165

Western blot analysis 166

Western blot analysis for p62/SQSTM1 determination was performed in cells treated 167

with Clarithromycin for 3h, as previously described (20). Briefly, incubation of the 168

PVDF membranes was carried out at 4°C using an anti-p62 antibody (stock 169

concentration: 200μg/ml, 1/300 dilution). Membranes were probed with HRP-170

conjugated secondary antibody (stock concentration: 400μg/ml, 1/1000 dilution) for 171

45 min at room temperature. To verify equal loading, membranes were re-probed for 172

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

10

GAPDH. Moreover, LL37-specific mouse mAb (stock concentration: 200μg/ml, 173

1/500 dilution) was utilized for the measurement of LL-37 expressed in NETs, using 174

the same protocol. 175

Bacterial strains 176

A. baumannii strains, including reference strains (ATCC19606) and clinical isolated 177

strain were cultured as previously described (21). Bacteria were preserved in glycerol 178

broth at -80oC. Bacteria from an overnight culture on MacConkey agar were 179

suspended into saline to an optical density of 0.5 McFarland, corresponding to a 180

concentration of approximately 108 colony forming units (CFU)/ml. Bacterial strains 181

of A. baumannii were co-cultured with NETs isolated as described above ("NETs 182

isolation"), generated by clarithromycin or generic inducers such as MSU or PMA in 183

a 1/10 concentration. For LL-37 inhibition on NETs, NETs were incubated with anti-184

LL-37 antibody prior to their introduction to bacterial cultures. Mouse monoclonal 185

IgG1 antibody was used as control to anti-LL-37 and did not affect NET induced 186

bacterial killing. 187

Biofilm Assay 188

Biofilm assay was determined by microtitre plate method as described previously 189

(22). Briefly, each microbial strain was grown overnight in trypticase soy broth (TSB) 190

at 37oC. Next, the overnight growth was diluted in a ratio of 1:40 in TSB. Two 191

hundred microlitre of cell suspension was inoculated in sterile 96 well polystyrene 192

microtitre plates in the presence or absence of NETs in 1/10 concentration, isolated as 193

described above. After 24 h of incubation at 37oC, the wells were gently washed three 194

times with 200 microlitre of phosphate buffered saline (PBS) and stained with 0.5% 195

crystal violet for 15 min. The wells were rinsed again in 200 microlitre of 95% 196

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

11

ethanol to solubilize crystal violet. Each assay was performed in duplicate. Finally, 197

the optical density at 550 nm (OD 550) was determined using microplate reader. 198

OD550 values for each well were subtracted from those of the blank, which only 199

contained TSB without inoculums. 200

Statistical analysis 201

Statistical analyses were performed using one-way ANOVA (Scheffé test in uniform 202

n and LSD test in non-uniform n for post hoc comparisons). All statistical analyses 203

were performed with OriginPro 8. P values less than 0.05 were considered significant. 204

205

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

12

Results 206

Macrolide antibiotics induce generation of NETs. 207

Considering that macrolide antibiotics can act as immunomodulatory agents (1,2), we 208

sought to investigate whether this effect can be mediated through neutrophils and, 209

more specifically, through the generation of NETs. We initially studied the ability of 210

different antibiotic groups to generate NETs. In vitro stimulation of PMNs from 211

healthy individuals with clinically relevant concentrations of commonly used classes 212

of antibiotics have shown that macrolide antibiotics (azithromycin and 213

clarithromycin) induced NETs formation, whereas ampicillin (semi-synthetic 214

penicillin), penicilin, daptomycin and ciprofloxacin did not, as observed by 215

immunofluorescence (Fig. 1A-B) and MPO-DNA complex ELISA (Fig. 1C). 216

We next investigated the ability of clarithromycin to induce NETs in vivo. To assess 217

this, patients suffering from H. pylori positive gastritis were enrolled as study group, 218

since clarithromycin is used for standard eradication therapy and they do not 219

demonstrate clinical and laboratory evidence for systemic inflammation (23) that 220

could potentially influence in vivo NET formation. Thus, PMNs were isolated before 221

and during the 5th day of standard eradication therapy with clarithromycin, 222

amoxicillin and omeprazole. The collected PMNs during the 5th day of eradication 223

therapy demonstrated increased ex vivo NETs formation compared to PMNs before 224

clarithromycin initiation, PMNs from healthy individuals, or PMNs from patients 225

under sequential eradication treatment that comprises only amoxicillin plus 226

omeprazole for the first five days of eradication (details in Table 1), as determined by 227

immunofluorescence (Figure 2A-B) and MPO-DNA complex ELISA (Fig. 2C). 228

Patients with H. pylori negative gastritis that treated with omeprazole alone did not 229

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

13

demonstrate any ex vivo NET generation (Fig. 2A-C). These findings demonstrate the 230

ability of macrolides to induce NET formation. 231

Clarithromycin utilizes autophagy to induce NET generation 232

Several lines of evidence support the involvement of autophagy in neutrophil 233

functions, including NET release (18,24,25). Thus, we assessed endogenous LC3B 234

cellular distribution and p62/SQSTM1 degradation in PMNs from control subjects 235

treated with clarithromycin. Formation of LC3B puncta and increased p62/SQSTM1 236

degradation was observed in PMNs treated with clarithromycin compared to controls, 237

as determined by immunofluorescence and immunoblotting, respectively, suggesting 238

the induction of the autophagic machinery (Fig. 3 A-B). 239

In addition, PMNs pretreated with Bafilomycin A1 prior to stimulation with 240

clarithromycin demonstrated reduced NET generation compared to PMNs treated 241

with clarithromycin alone (Fig. 3C), further verifying the involvement of autophagy 242

in clarithromycin induction of NETs formation. These findings demonstrate that 243

clarithromycin induces NET formation in an autophagy-dependent manner. 244

NETs generated by clarithromycin are decorated with LL-37 245

Considering that LL-37 is a potent anti-microbial peptide (26,27) and that it is 246

released through NETs under certain inflammatory conditions (28,29), we examined 247

the presence of LL-37 on clarithromycin-induced NETs. NETs released by PMNs 248

from control subjects treated with clarithromycin were decorated with LL-37, as 249

determined by immunofluorescence and immunoblotting (Fig. 4A-B). 250

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

14

We then assessed LL-37 expression on NETs released from PMNs derived from 251

patients treated with clarithromycin. Indeed, PMNs derived from these patients 252

demonstrated presence of LL-37 on NETs (Fig. 4A). 253

Furthermore, to assess whether the presence of LL-37 is a generic finding of NETs we 254

investigated its presence on NETs using other alternative inducers of NETosis, such 255

as MSU and PMA. LL-37 was absent from NETs induced by MSU (Fig. 4A-B), as 256

assessed by immunofluorescence and immunoblotting of NET proteins. Similarly to 257

MSU, PMA is not able to induce LL-37-bearing NETs (Fig. 4A). 258

These findings suggest that clarithromycin has the ability to induce NETs containing 259

LL-37. 260

Clarithromycin-dependent LL-37 expression on NETs modulates the 261

antimicrobial activity against Acinetobacter baumannii 262

Since neutrophils during infection produce NETs as a defense mechanism against 263

pathogens (7,8), we investigated if clarithromycin-driven NETs conserve this ability. 264

To test the microbicidal activity of clarithromycin-driven NETs, our model was based 265

on A. baumannii, since this microorganism is not able to trigger the mechanism of 266

NET by itself in vitro (11) and ex vivo (Fig. S1) and it is also resistant to treatment 267

with clarithromycin (30). Clinical and ATCC 19606 strains of A. baumannii were 268

cultured in the presence or absence of clarithromycin-induced NETs. NETs were 269

obtained by either in vitro stimulation or ex vivo from H. pylori positive patients 270

treated with clarithromycin. We observed that both in vitro (Fig. 5A) and ex vivo (Fig. 271

5B) clarithromycin-induced NETs significantly reduced bacterial growth compared to 272

control cultures. This activity was attributed to the presence of LL-37, since LL-37 273

neutralization with anti-LL-37 antibody abolished the antimicrobial action of 274

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

15

clarithromycin-induced NETs (Fig. 5A-B). To further verify the specific activity of 275

clarithromycin-driven NETs, we used NETs obtained after stimulation from inducers 276

that are not able to generate LL-37-bearing NETs, such as MSU or PMA. These 277

generic NETs in contrast to clarithromycin-induced NETs indicated absence of LL-37 278

(Fig. 4A-B) and significantly reduced capacity to inhibit bacterial growth (Fig. 5A). 279

We further sought to investigate the effect of clarithromycin-induced NETs on the 280

generation of biofilm by A. baumannii. Interestingly, clarithromycin-induced NETs 281

significantly reduced the formation of biofilms in an LL-37-dependent manner, as 282

demonstrated by LL-37 inhibition (Fig. 5C). However, generic NETs induced by 283

MSU crystals or PMA did not affect biofilm generation by A. baumannii (Fig. 5C). 284

Treatment of NET structures with DNase attenuated the inhibitory effect on biofilms 285

(Fig. 5C). 286

Since NET scaffold affects the functionality of various NET-bound components 287

(7,19) we wanted to test whether LL-37 functionality is affected by NET integrity. 288

DNase treatment of clarithromycin-induced NETs abolished LL-37 antimicrobial 289

function on A. baumannii cultures (Fig. 5D). 290

These findings suggest an indirect antimicrobial role of clarithromycin against strains 291

of A. baumannii through NETs formation in a LL-37 dependent manner. 292

293

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

16

Discussion 294

The present study demonstrates for the first time the in vitro and in vivo ability of 295

macrolide antibiotics to induce NETs formation, expanding our understanding on 296

their immunomodulatory role. We show that clarithromycin-induced NETs are 297

decorated by the anti-microbial peptide LL-37, which is in its active form and it is 298

able to act against multi-drug resistant A. baumannii in vitro. 299

Clarithromycin-induced NETs exerted their antimicrobial function against both 300

clinical and ATCC 19606 strains of multi drug-resistant A. baumannii, a pathogen that 301

is not able to trigger the mechanism of NET by itself. However, specific NETs (such 302

as clarithromycin-driven) restrict A. baumannii in cultures. Additionally, 303

clarithromycin-induced NETs were able to inhibit A. baumannii biofilm formation. 304

This is an interesting finding and gives more information on the ongoing debate on 305

the crosstalk of biofilms and NETs, since neutrophils constitute indispensable biofilm 306

controllers but can also contribute to biofilm stability (31). These findings suggest an 307

indirect anti-microbial action of macrolide antibiotics, through the induction of NET 308

generation. As a side note, we also provide evidence that A. baumannii does not 309

induce NET generation in vivo, as observed in ex vivo isolated PMNs from patients 310

with A. baumannii bacteremia. 311

Since neutrophils have been shown to express LL-37 under certain inflammatory 312

conditions (28,29), we investigated the presence of this antimicrobial peptide on 313

clarithromycin-induced NETs. We demonstrated for the first time that clarithromycin 314

induces both in vitro and ex vivo NETs decorated with LL-37, while other generic 315

inducers of NETs, such as MSU or PMA, were not able to generate amounts of active 316

LL-37-bearing NETs. Recent studies described that PMA-induced NETs did not 317

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

17

express active LL-37. Although, traces of LL-37 have been detected by these authors 318

mainly observed on electron microscopy, the absence of LL-37 activity are in 319

accordance with our findings (32). Furthermore, considering the susceptibility of A. 320

baumannii to LL-37 (27), we further demonstrate that LL-37 mediated the 321

microbicidal activity of NETs and attenuated the formation of biofilms by this 322

pathogen. Additionally, digestion of chromatin scaffold with DNase abolished 323

completely the bactericidal activity of clarithromycin-induced NETs similarly to anti-324

LL37 inhibition. This indicates the importance of NET chromatin scaffold in LL-37 325

functionality, previously shown also in other NET-bound proteins (7,19). Moreover, 326

since autophagy plays an important role in NET generation we investigated whether 327

clarithromycin regulates this active mechanism to generate NETs. Indeed, 328

clarithromycin was involved in the in vitro upregulation of autophagy leading to 329

generation of NETs, as demonstrated by the end-stage autophagy inhibition with 330

Bafilomycin A1. These findings suggest that clarithromycin can act additionally as an 331

autophagy inducer in PMNs, thus expanding its role by regulating autophagy, a 332

necessary mechanism for NET formation. 333

A previous clinical trial has demonstrated that administration of clarithromycin in 334

patients with VAP accelerated the resolution of pneumonia and decreased the risk of 335

death from septic shock, suggesting a potential immunomodulatory effect of 336

clarithromycin on the host (4). It should be noted that most of the patients (54.5%) 337

were suffering from A. baumannii infection. Besides the limitations related to the 338

design of the study including the short period of clarithromycin administration (only 3 339

days), these results support our in vitro findings and encourage the investigation of the 340

in vivo synergistic role of macrolide-induced NETs in severe infections, such as those 341

caused by multidrug-resistant species of A. baumannii. Additionally, whether NETs 342

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

18

are involved in the adjunctive potential of clarithromycin, when administered in 343

combination with other antibiotics, is a query that needs to be investigated in future 344

studies. 345

In conclusion, this is the first report suggesting the formation of LL-37-bearing NETs 346

by clarithromycin as a part of its immunomodulatory role. Clarithromycin NET-347

inducing role is dependent on autophagy and the LL-37 present in NETs is important 348

for both microbicidal activity and biofilm inhibition. Of note, the functionality of LL-349

37 is highly dependent on NET chromatin scaffold integrity. Our study supports the 350

use of clarithromycin as synergic agent in clinical trials in patients with A. baumannii 351

infection in order to determine whether the observed in vitro microbicidal function 352

through NETs is repeated in vivo. 353

Funding information 354

This study was supported by the Scientific Committee of Democritus University of 355

Thrace Grant number - 80895. 356

Acknowledgements 357

Conflict of Interest - None declared 358

359

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

19

References 360

1. Kanoh S, Rubin BK. 2010. Mechanisms of action and clinical application of 361

macrolides as immunomodulatory medications. Clin Microbiol Rev 23:590–615. 362

2. Kovaleva A, Remmelts HH, Rijkers GT, Hoepelman AI, Biesma DH, Oosterheert 363

JJ. 2012. Immunomodulatory effects of macrolides during community-acquired 364

pneumonia: a literature review. J Antimicrob Chemother 67:530-540. 365

3. Restrepo MI, Mortensen EM, Waterer GW, Wunderink RG, Coalson JJ, Anzueto 366

A. 2009. Impact of macrolide therapy on mortality for patients with severe sepsis 367

due to pneumonia. Eur Resp J 33:1531–1539. 368

4. Giamarellos-Bourboulis EJ, Mylona V, Antonopoulou A, Tsangaris I, 369

Koutelidakis I, Marioli A, Raftogiannis M, Kopterides P, Lymberopoulou K, 370

Mouktaroudi M, Papageorgiou C, Papaziogas B, Georgopoulou AP, Tsaganos T, 371

Papadomichelakis E, Gogos C, Ladas M, Savva A, Pelekanou A, Baziaka F, 372

Koutoukas P, Kanni T, Spyridaki A, Maniatis N, Pelekanos N, Kotsaki A, Vaki I, 373

Douzinas EE, Koratzanis G, Armaganidis A. 2008. Effect of clarithromycin in 374

patients with sepsis and ventilator-associated pneumonia. Clin Infect Dis 46:1157-375

1164. 376

5. Schultz MJ, Speelman P, Hack CE, Buurman WA, van Deventer SJ, van Der Poll 377

T. 2000. Intravenous infusion of erythromycin inhibits CXC chemokine 378

production, but augments neutrophil degranulation in whole blood stimulated with 379

Streptococcus pneumoniae. J Antimicrob Chemother 46:235-240. 380

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

20

6. Iskandar I, Walters JD. 2011. Clarithromycin accumulation by phagocytes and its 381

effect on killing of Aggregatibacter actinomycetemcomitans. J Periodontol 382

82:497-504. 383

7. Brinkmann, V., Reichard U, Goosmann C, Fauler B, Uhlemann Y, Weiss DS, 384

Weinrauch Y, Zychlinsky A. 2004. Neutrophil extracellular traps kill bacteria. 385

Science 303:1532–1535. 386

8. Yipp BG, Petri B, Salina D, Jenne CN, Scott BN, Zbytnuik LD, Pittman K, 387

Asaduzzaman M, Wu K, Meijndert HC, Malawista SE, de Boisfleury Chevance 388

A, Zhang K, Conly J, Kubes P. 2012. Infection-induced NETosis is a dynamic 389

process involving neutrophil multitasking in vivo. Nat Med 18:1386-1393. 390

9. Wang Y, Chen Y, Xin L, Beverley SM, Carlsen ED, Popov V, Chang KP, Wang 391

M, Soong L. 2011. Differential microbicidal effects of human histone proteins 392

H2A and H2B on Leishmania promastigotes and amastigotes. Infect Immun. 393

79(:1124-1133. 394

10. Halverson TW, Wilton M, Poon KKH, Petri B, Lewenza S. 2015. DNA Is an 395

Antimicrobial Component of Neutrophil Extracellular Traps. PLoS Pathog 396

11(1):e1004593. 397

11. Kamoshida G, Kikuchi-Ueda T, Tansho-Nagakawa S, Nakano R, Nakano A, 398

Kikuchi H, Ubagai T, Ono Y. 2015. Acinetobacter baumannii escape from 399

neutrophil extracellular traps (NETs), J Infect Chemother 21:43-49. 400

12. Berends ET, Horswill AR, Haste NM, Monestier M, Nizet V, and von Kockritz-401

Blickwede M. 2010. Nuclease expression by Staphylococcus aureus facilitates 402

escape from neutrophil extracellular traps. J Innate Immun 2:576–586. 403

13. Lauth X, von Kockritz-Blickwede M, McNamara CW, Myskowski S, Zinkernagel 404

AS, Beall B, Ghosh P, Gallo RL, Nizet V. 2009. M1 protein allows Group A 405

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

21

streptococcal survival in phagocyte extracellular traps through cathelicidin 406

inhibition. J Innate Immun 1:202–214. 407

14. Hall-Stoodley L, Costerton JW, Stoodley P. 2004. Bacterial biofilms: from the 408

natural environment to infectious diseases. Nat Rev Microbiol 2:95–108. 409

15. Fuxman Bass JI, Russo DM, Gabelloni ML, Geffner JR, Giordano M, Catalano 410

M, Zorreguieta A, Trevani AS. 2010. Extracellular DNA: a major 411

proinflammatory component of Pseudomonas aeruginosa biofilms. J Immunol 412

184:6386-6395. 413

16. Donlan RM, Costerton JW. 2002. Biofilms: survival mechanisms of clinically 414

relevant microorganisms. Clin Microbiol Rev 15:167-193. 415

17. Espinal P, Martí S, Vila J. 2012. Effect of biofilm formation on the survival of 416

Acinetobacter baumannii on dry surfaces. J Hosp Infect 80:56-60. 417

18. Mitroulis I, Kambas K, Chrysanthopoulou A, Skendros P, Apostolidou E, 418

Kourtzelis I, Drosos GI, Boumpas DT, Ritis K. 2011. Neutrophil extracellular trap 419

formation is associated with IL-1β and autophagy-related signaling in gout. PLoS 420

One 6(12):e29318. 421

19. Stakos DA, Kambas K, Konstantinidis T, Mitroulis I, Apostolidou E, Arelaki S 422

Tsironidou V, Giatromanolaki A, Skendros P, Konstantinides S, Ritis K. 2015. 423

Expression of functional tissue factor by neutrophil extracellular traps in culprit 424

artery of acute myocardial infarction. Eur Heart J 36:1405-1414. 425

20. Chrysanthopoulou A, Mitroulis I, Kambas K, Skendros P, Kourtzelis I, Vradelis 426

S, Kolios G, Aslanidis S, Doumas M, Ritis K. 2011. Tissue factor-thrombin 427

signaling enhances the fibrotic activity of myofibroblasts in systemic sclerosis 428

through up-regulation of endothelin receptor A. Arthritis Rheum 63:3586-3597. 429

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

22

21. Bouvet PJ and Grimont PA. 1986. Taxonomy of the genus Acinetobacter with the 430

recognition of Acinetobacter baumannii sp. nov., Acinetobacter haemolyticus sp. 431

nov., Acinetobacter johnsonii sp. nov., and Acinetobacter junii sp. nov. and 432

emended descriptions of Acinetobacter calcoaceticus and Acinetobacter lwoffii. 433

Int J Syst Bacteriol 36:228-240. 434

22. O'Toole GA. 2011. Microtiter dish biofilm formation assay. J Vis Exp. 30:(47). 435

pii: 2437. doi: 10.3791/2437. 436

23. Suerbaum S, Michetti P. 2002. Helicobacter pylori infection. N Engl J Med 347: 437

1175-1186. 438

24. Mitroulis I, Kourtzelis I, Kambas K, Rafail S, Chrysanthopoulou A, Speletas M, 439

Ritis K. 2010. Regulation of the autophagic machinery in human neutrophils. Eur 440

J Immunol 40:1461-1472. 441

25. Remijsen Q, Vanden Berghe T, Wirawan E, Asselbergh B, Parthoens E, De Rycke 442

R, Noppen S, Delforge M, Willems J, Vandenabeele P. 2011. Neutrophil 443

extracellular trap cell death requires both autophagy and superoxide generation. 444

Cell Res 21:290-304. 445

26. Wnorowska U, Niemirowicz K, Myint M, Diamond SL, Wróblewska M, Savage 446

PB, Janmey PA, Bucki R. 2015. Bactericidal Activities of Cathelicidin LL-37 and 447

Select Cationic Lipids against the Hypervirulent Pseudomonas aeruginosa Strain 448

LESB58. Antimicrob Agents Chemother 59:3808-3815. 449

27. Feng X, Sambanthamoorthy K, Palys T, Paranavitana C. 2013. The human 450

antimicrobial peptide LL-37 and its fragments possess both antimicrobial and 451

antibiofilm activities against multidrug-resistant Acinetobacter baumannii. 452

Peptides 49:131-137. 453

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

23

28. Garcia-Romo GS, Caielli S, Vega B, Connolly J, Allantaz F, Xu Z, Punaro M, 454

Baisch J, Guiducci C, Coffman RL, Barrat FJ, Banchereau J, Pascual V. 2011. 455

Netting neutrophils are major inducers of type I IFN production in pediatric 456

systemic lupus erythematosus. Sci Transl Med 3:73ra20. 457

29. Lande R, Ganguly D, Facchinetti V, Frasca L, Conrad C, Gregorio J, Meller S, 458

Chamilos G, Sebasigari R, Riccieri V, Bassett R, Amuro H, Fukuhara S, Ito T, Liu 459

YJ, Gilliet M. 2011. Neutrophils activate plasmacytoid dendritic cells by releasing 460

self-DNA-peptide complexes in systemic lupus erythematosus. Sci Transl Med. 461

3:73ra19. 462

30. EUCAST Breakpoint tables for interpretation of MICs and zone diameters. 463

Version 3.0, January 2013. Available at 464

http://www.eucast.org/clinical_breakpoints/. Accessed 465

31. Hirschfeld J. 2014. Dynamic interactions of neutrophils and biofilms. J Oral 466

Microbiol. 6:26102. 467

32. Neumann A, Völlger L, Berends ET, Molhoek EM, Stapels DA, Midon M, Friães 468

A, Pingoud A, Rooijakkers SH, Gallo RL, Mörgelin M, Nizet V, Naim HY, von 469

Köckritz-Blickwede. M. 2014. Novel role of the antimicrobial peptide LL-37 in 470

the protection of neutrophil extracellular traps against degradation by bacterial 471

nucleases. J Innate Immun. 6(6):860-868. 472

473

474

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

24

Figure Legends 475

Figure 1. Macrolide antibiotics induce NET generation in vitro. A) NET 476

generation in control PMNs treated with relevant concentrations of antibiotics, as 477

assessed by immunofluorescence (confocal microscopy). Green: MPO, Red: NE, 478

Blue: DAPI/DNA. One representative out of six independent experiments is shown. 479

Original magnification 600x. Scale bar – 5μm. (B) Percentage of NET-releasing 480

PMNs and (C) MPO/DNA complex in isolated NET structures from control PMNs 481

treated with antibiotics. (B), (C). Data from six independent experiments presented as 482

mean±SD (* P < 0.05, n.s. non significant compared to control). 483

Figure 2. Clarithromycin induces NET generation in vivo in patients with H. 484

pylori gastritis. A) NET generation in ex vivo PMNs from patients with H. pylori 485

positive gastritis before and after treatment with clarithromycin, amoxicillin and 486

omeprazole and from control individuals, as assessed by immunofluorescence 487

(confocal microscopy). Green: MPO, Red: NE, Blue: DAPI/DNA. One representative 488

out of six independent experiments is shown. Original magnification 600x. Scale bar 489

– 5μm. (B) Percentage of NET-releasing PMNs and (C) MPO/DNA complex in 490

isolated NET structures from ex vivo PMNs isolated from patients with H. pylori 491

positive gastritis before and after standard eradication treatment with 492

clarithromycin/amoxicillin/omeprazole (n=6), patients with sequential eradication 493

(amoxicillin and omeprazole) (n=4), patients with H. pylori negative gastritis 494

(omeprazole monotherapy) (n=5) and control individuals (n=10). Data presented as 495

median ± 90% percentile (* P < 0.05, n.s. non significant compared to control). 496

Figure 3. Clarithromycin induced NET generation in PMNs by utilizing 497

autophagy. A) Autophagy induction assessed with LC3B staining as assessed by 498

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

25

confocal microscopy or (B) p62/SQSTM1 immunoblotting in control PMNs treated 499

with clarithromycin. Red: LC3B, Blue: DAPI/DNA. (C) NET generation in control 500

PMNs treated with clarithromycin in the presence or not of Bafilomycin A1 501

(autophagy inhibitor), as assessed by immunofluorescence (confocal microscopy). 502

Green: MPO, Red: NE, Blue: DAPI/DNA. (D) MPO/DNA complex in isolated NET 503

structures from control PMNs treated with clarithromycin in the presence or not of 504

autophagy inhibitor. (A), (B) & (C) One representative out of four independent 505

experiments is shown. Original magnification (A) 1000x, (C) 600x. Scale bar – 5μm. 506

(D) Data from four independent experiments presented as mean±SD (* P < 0.05, n.s. 507

non significant). 508

Figure 4. LL-37 is present on NETs induced by clarithromycin. A) Localization of 509

LL-37 on NETs in in vitro control PMNs stimulations with clarithromycin or ex vivo 510

patients treated with clarithromycin/amoxyciling/omeprazole, as observed by confocal 511

microscopy. MSU and PMA were used as generic NET inducers. Green: LL-37, Red: 512

NE, Blue: DAPI/DNA. Original magnification 600x. Scale bar – 5μm. (B) Expression 513

of LL-37 on NET proteins derived from in vitro stimulations of control PMNs with 514

clarithromycin or MSU, as assessed by immunoblotting. (A) & (B) One representative 515

out of four independent experiments is shown. 516

Figure 5. In vitro and ex vivo generated clarithromycin-induced NET act against 517

Acinetobacter baumannii in a LL-37 dependent manner. (A) Cultures of clinical 518

and ATCC 19606 strains of A. baumannii in the presence of clarithromycin-induced 519

NETs generated in vitro before and after pretreatment with anti-LL-37 antibody. MSU 520

and PMA were used as controls (LL-37 negative NETs). (B) Cultures of clinical and 521

ATCC 19606 strains of A. baumannii in the presence of NETs derived from H. pylori 522

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

26

positive patients treated with clarithromycin/amoxycilin/omeprazole, before and after 523

pretreatment with anti-LL-37 antibody. Data from six independent experiments 524

presented as mean±SD (* P < 0.05, n.s. non significant). (C) Determination of biofilm 525

generation in clinical and ATCC 19606 strains of A. baumannii in the presence of 526

clarithromycin-induced NETs generated in vitro before and after pretreatment with 527

anti-LL-37 antibody or DNase I, as assessed by crystal violet staining. MSU and 528

PMA were used as controls (LL-37 negative NETs). (D) Cultures of clinical and 529

ATCC 19606 strains of A. baumannii in the presence of clarithromycin-induced NETs 530

generated in vitro before and after pretreatment with DNase I (A), (B) & (C) Data 531

from six independent experiments presented as mean±SD (* P < 0.05, n.s. non 532

significant). (D) Data from four independent experiments presented as mean±SD (* P 533

< 0.05, n.s. non significant). 534

535

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

1

Table 1. Patient and control group characteristics 1

2

3

4

Standard eradication regimen: amoxicillin 1g bid, clarithromycin 500 mg bid, and omeprazole 20 mg bid, for 5

10 days 6

Sequential eradication regimen: amoxicillin 1gr bid plus omeprazole 20 mg bid for the first 5 days, followed by 7

clarithromycin 500 mg bid, metronidazole 500 mg bid and omeprazole 20 mg bid, for the next 5 days 8

HP: H. pylori, M: male, F: female 9

Hp positivegastritis

Hp negative gastritis Control

N=10 N=5 N=10 Gender, M/F 4/6 2/3 4/6 Age, mean±SD 42.9±14.3 44.2±6.4 35.4±5.7 Standard eradication regimen 6 - - Sequential eradication regimen 4 - - Omeprazole monotherapy - 5 -

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

on March 27, 2018 by guest

http://aac.asm.org/

Dow

nloaded from