TOWARDS IMPROVEMENT OF ANTIBIOTIC

THERAPY FOR TREATING BURKHOLDERIA

CEPACIA COMPLEX BIOFILM INFECTIONS IN

CYSTIC FIBROSIS PATIENTS

Anne-Sophie Messiaen

August 2013

Thesis submitted in the fulfillment of the requirements for the degree of

Doctor in Pharmaceutical Sciences, Microbiology.

Promotor: Prof. Dr. Tom Coenye

Co-promotor: Prof. Dr. Apr. Hans Nelis

Laboratory of Pharmaceutical Microbiology

The author and the promoter give the authorization to consult and to copy parts of

this thesis for personal use only. Any other use is limited by the laws of copyright,

especially the obligation to refer to the source whenever results from this thesis are

cited.

A-S. Messiaen – Towards improvement of antibiotic therapy for treating Burkholderia

cepacia complex biofilm infections in cystic fibrosis patients.

Ph.D. thesis, Faculty of Pharmaceutical Sciences, Ghent University, Ghent, Belgium.

Publicly defended in Ghent, October 4th, 2013.

EXAMINATION COMMITTEE

Prof. Dr. Kevin Braeckmans (Chairman)

Laboratory of General Biochemistry & Physical Pharmacy

Faculty of Pharmaceutical Sciences, Ghent University, Ghent, Belgium

Prof. Dr. Tom Coenye (Promotor)

Laboratory of Pharmaceutical Microbiology (LPM)

Faculty of Pharmaceutical Sciences, Ghent University, Ghent, Belgium

Prof. Dr. Hans Nelis (Co-promotor)

Laboratory of Pharmaceutical Microbiology (LPM)

Faculty of Pharmaceutical Sciences, Ghent University, Ghent, Belgium

Prof. Dr. Serge Van Calenbergh

Laboratory of Medicinal Chemistry

Faculty of Pharmaceutical Sciences, Ghent University, Ghent, Belgium

Prof. Dr. Niek Sanders

Laboratory of Gene Therapy,

Faculty of Veterinary Medicine, Ghent University, Ghent, Belgium

Prof. Dr. Francoise Van Bambeke

Cellular and Molecular Pharmacology

Department of Pharmaceutical Sciences, UCL-Bruxelles, Brussels, Belgium

Prof. Dr. Jo Demeester

Laboratory of General Biochemistry & Physical Pharmacy

Faculty of Pharmaceutical Sciences, Ghent University, Ghent, Belgium

Prof. Dr. Bruno De Geest

Laboratory of Pharmaceutical Technology

Faculty of Pharmaceutical Sciences, Ghent University, Ghent, Belgium

Na vier jaar werken is het eindelijk zover, mijn doctoraat ligt er en ik ben er trots op! Hier heb ik

echt naar uitgekeken. Ik stond er de voorbije 4 jaar echter nooit alleen voor en zou hiervoor

toch graag nog enkele mensen willen vereeuwigen in mijn dankwoord.

Professor Coenye, Tom, bedankt om me de kans te geven om in uw labo te doctoreren. Het

project waarop je me hebt aangesteld sprak me enorm aan. Wanneer je aan zoiets begint denk

je de wereld te kunnen verbeteren maar al gauw besef je dat je met kleine passen vooruitgang

boekt. Het liep niet steeds op wieltjes, vaak waren resultaten negatief, maar ik kon steeds op je

rekenen ook om me duidelijk te maken dat negatieve resultaten ook resultaten zijn. Je hulp bij

het opstellen van experimenten was verruimend en je feedback op mijn artikels en op mijn

doctoraatsthesis was razendsnel. Ik heb aan u een uitstekende promotor gehad!

Professor Nelis, ook u wil ik bedanken voor de kans om in uw labo te kunnen doctoreren. Ik ben

blij dat ik nog deel heb mogen uitmaken van uw laatste jaren op het labo. Uw oprechte

interesse in het verloop van de experimenten en uw opbeurende antwoorden bij tegenslagen

deden soms echt goed. Door uw lessen heb ik in eerste instantie de smaak voor microbiologie

te pakken gekregen. Uw uitgebreide kennis van het Engels zorgde er ook voor dat mijn artikels

telkens grammaticaal tip top in orde waren voor ze werden doorgestuurd naar de editor.

De sfeer op het labo was geweldig. Velen uit mijn nabije omgeving kunnen dit bevestigen, ik

ging amper met tegenzin gaan werken. Soms vraag ik me af of het mogelijk is om elders

opnieuw samen te kunnen werken met zo een gekke en toffe bende collega’s. Heleen, wij zijn

samen in het doctoreerbootje gestapt en stonden ook meteen samen in het labo. Hoewel we

elkaar tijdens onze studie farmacie nooit echt hebben leren kennen konden we het toch goed

met elkaar vinden. Je hulpvaardigheid en luisterend oor wist ik enorm te appreciëren. Steven,

Stevie, om jouw kun je moeilijk heen lopen op het labo. Iedereen heeft je onmiddellijk gezien

en als dat niet het geval is dan zorg jij er wel voor dat ze je hebben gehoord. Maar zo heb ik het

graag! Ik heb je gedurende die 4 jaar beter leren kennen en we hebben samen leuke momenten

gehad. We zijn gestart als collega’s maar ik heb er een goede vriend aan over gehouden. Davy,

jij bent net iets voor mij vertrokken op het labo en dat was toch voelbaar. Je werd onmiddellijk

gemist. Ik heb steeds het gevoel gehad dat ik altijd bij jouw terecht kon. Jij bent één van de

meest hulpvaardige personen die ik ken en ik ben dan ook heel blij dat ik jou heb leren kennen!

Gilles, jouw zie ik voornamelijk als een stille genieter. Altijd goedlachs, steeds klaar om een

woordje uitleg te geven en publiceren aan de lopende band. Inne, ons moeder, jij hebt een

groot hart. Je bezorgdheid en lieve woorden kwamen soms echt op het gepaste moment.

Petra, wie kan zeggen dat ze familie leren kennen op het werk? Ik wel! En met veel plezier.

Jouw hulp tijdens mijn laatste jaar kwam als een echt godsgeschenk. Nele, voor mij ben jij de

moraalridder van het labo. Je hebt steeds het beste voor met iedereen rondom je en je springt

duidelijk voor je medemens in de bres. Dit siert je enorm! Evelien, ons bakprinses, door jou

waren de pauzes vaak rijkelijk gevuld met zoetigheden en dat zal de weegschaal geweten

hebben. Je deelde graag en niet alleen je zoetigheden, als iemand een paar extra handen kon

gebruiken aan de bench stond jij altijd paraat. Ilse, met jouw heb ik vanaf het begin een goede

klik gehad. Jij was voor mij de eerste postdoc op het labo, heel toegankelijk en supervriendelijk.

Vele leuke babbels en grappige momenten hebben ervoor gezorgd dat wij elkaar waarschijnlijk

(hoop ik) nog wel terug zullen zien. Sarah, zolang jij op het labo bent valt de groep niet uiteen.

Daar ben ik zo goed als zeker van. Met bakken vol enthousiasme weet jij telkens weer de

geschikte labo activiteit te organiseren. Ik heb je leren kennen als enorm hulpvaardig, iemand

waarop je steevast kan rekenen! Freija en Annelien ik heb een jaartje met jullie kunnen

samenwerken. Lang genoeg om te beseffen dat ook jullie perfect in het labo passen.

Creativiteit en hilariteit daarmee vullen jullie de leutige bende collega’s perfect aan.

Niet alleen mijn collega’s maar ook mijn vrienden en familie zijn er steeds voor me geweest.

Alle berichtjes en lieve woorden wanneer het even iets minder ging en de continue interesse

naar de gang van zaken deden meer dan deugd. Jojo, jij bent vaak de ultieme remedie voor al

mijn zorgen geweest. Ik prijs mezelf enorm gelukkig met een vriendin als jij! Eva en Anne, de

“Gent crew”, wij moeten onze etentjes zeker in stand houden want dit waren voor mij keer op

keer topavonden. Ik wil ook graag al mijn vrienden bedanken voor jullie steun in de vorm van

feestjes, cafebezoekjes, nog meer cafebezoekjes, de zoete inval op onze terras, de zoete inval

elders, … Kortom om samen leuke momenten te beleven.

Mama, bedankt om dit alles mogelijk voor mij te maken. Jij bent zowaar de sterkste vrouw die

ik ken en wat ben ik blij dat ik je mijn mama kan noemen. Ik hoop dat ik vele eigenschappen van

jou heb meegekregen en later een even goede en trotse mama als jij kan worden.

Iemand die ook steeds aan mijn zijde heeft gestaan en die ik ook absoluut wil bedanken is mijn

keppe. Jij gaf me constant de ruimte die ik nodig had en stond dicht bij me wanneer ik het

moeilijk had. Ik ben blij dat jij er was toen ik mijn doctoraat begon en nog blijer dat je er nog

steeds staat nu het achter de rug is. Ik hoop dat wij er samen nog een lang en mooi vervolg aan

kunnen breien.

1

Table of Contents

List of abbreviations 3

PART 1 OVERVIEW OF THE LITERATURE 5

Chapter I: CYSTIC FIBROSIS 7

I.1 Introduction 7

I.2 Diagnosis 8

I.3 Clinical manifestation 9

I.4 Pulmonary therapy 17

Chapter II: BIOFILMS 25

II.1 Introduction 25

II.2 Biofilm formation 25

II.3 Resistance of biofilms 27

II.4 Biofilm formation in the CF lung 31

II.5 Intervention strategies related to CF biofilm infections 32

Chapter III: BURKHOLDERIA CEPACIA COMPLEX 45

III.1 General introduction 45

III.2 Taxonomy 46

III.3 Human pathogenesis 47

III.4 Genomics 49

III.5 Virulence factors 50

PART 2 OBJECTIVES 67

PART 3 EXPERIMENTAL WORK 71

Chapter IV: RESISTANCE OF THE BURKHOLDERIA CEPACIA COMPLEX AGAINST

FOSMIDOMYCIN AND FOSMIDOMYCIN DERIVATIVES 72

ABSTRACT 73

Introduction 74

Materials and methods 75

Results 78

Discussion 81

2

Chapter V: UPTAKE OF TOBRAMYCIN AND TOBRAMYCIN DERIVATIVES IN DIFFERENT

GRAM-NEGATIVE BACTERIA 84

ABSTRACT 85

Introduction 86

Materials and methods 89

Results and discussion 94

Conclusion 102

Supporting information 106

Chapter VI: INVESTIGATING THE ROLE OF MATRIX COMPONENTS IN PROTECTION OF

BURKHOLDERIA CEPACIA COMPLEX BIOFILMS AGAINST TOBRAMYCIN 112

ABSTRACT 113

Introduction 114

Materials and methods 116

Results 120

Discussion 124

Conclusion 127

Chapter VII: TRANSPORT OF NANOPARTICLES AND TOBRAMYCIN-LOADED LIPOSOMES IN

BURKHOLDERIA CEPACIA COMPLEX BIOFILMS 130

ABSTRACT 131

Introduction 132

Materials and methods 134

Results 139

Discussion 146

Conclusion 149

PART 4 GENERAL DISCUSSION 153

PART 5 SUMMARY-SAMENVATTING 165

Summary 167

Samenvatting 171

3

List of abbreviations

AMP: antimicrobial peptide

ASL: airway surface liquid

BAL: bronchoalveolar lavage fluid

Bcc: Burkholderia cepacia complex

BOC: di-tert-butyl dicarbonate

CAP: cationic antimicrobial peptide

CF: cystic fibrosis

CFTR: cystic fibrosis transmembrane conductance regulator

CGD: chronic granulomatous disease

CLSM: confocal laser scanning microscopy

DFO: deferoxamine

DIOS: distal intestinal obstruction

DMEDA: N,N-dimethylethylenediamine

DMF: dimethylformamide

DMSO: dimethylsulfoxide

DOPC: 1,2-dioleoyl-sn-glycero-3-phosphocholine

DPPC: 1,2-dipalmitoyl-sn-glycero-3-phosphocholine

DPPG: 1,2-dipalmitoyl-sn-glycero-3-phosphoglycerol

DSX: deferasirox

eDNA: extracellular DNA

EPS: extracellular polymeric substances

ET12: electrophoretic type 12

FDA: Food and Drug Administration

Fsr: fosmidomycin resistance gene

IHF: integrating host factor

IL-8: interleukin-8

IRT: immunoreactive trypsinogen

LPS: lipopolysaccharide

MBC: minimum bactericidal concentration

MH: Mueller Hinton

MI: meconium ileus

4

MIC: minimum inhibitory concentration

MRSA: methicillin-resistant Staphylococcus aureus

MSD: mean square displacement

NO: nitric oxide

NOR: nitric oxide reductase

OD: optical density

PBS: phosphate buffered saline

PIA: polysaccharide intercellular adhesin

PNAG: poly-β-1,6-N-acetylglucosamine

PS: physiological saline

QS: quorum sensing

rhDNase: recombinant human DNase

ROS: reactive oxygen species

SEM: scanning electron microscopy

SPT: single particle tracking

TFA: trifluoroacetic acid

THF: tetrahydrofuran

TNF: tumor necrosis factor

TLC: thin layer chromatography

5

PART 1

Overview of the literature

6

Anyone who has read the literature in cystic fibrosis and isn’t confused is confused (Efraim Racker)

7

Chapter I: CYSTIC FIBROSIS

I.1 Introduction

Cystic fibrosis (CF) is the most common genetically inherited disease in the Caucasian

population [1]. It is a recessive disorder which means that CF only occurs when an

individual inherits two malfunctioning copies (mutated alleles) of the “CF

transmembrane conductance regulator” gene (CFTR). When a detrimental mutation

occurs only on one allele, the individual is called a CF carrier. In populations of northern

European or northern American descent approximately one in 25 people are

asymptomatic carriers while one in 2500 newborns are suffering from CF [1].

The gene that causes CF, CFTR, was not detected until 1989 and is located on the long

arm of chromosome 7. CFTR encodes a cAMP-regulated-chloride channel belonging to

the ABC (ATP-binding cassette) transporter superfamily and is expressed on the apical

surface of epithelial cells in different organs [2].

The protein comprises 5 domains: two membrane-spanning domains (MSD1 and MSD2)

that form the chloride channel, two cytoplasmic nucleotide-binding domains (NBD1 and

NBD2) that hydrolyze ATP and a regulatory domain (Figure 1) [2].

Figure 1: Proposed model for CFTR structure

(http://www.genemedresearch.ox.ac.uk/cysticfibrosis/protein.html)

8

More than 1000 CF-associated mutations are already recorded but the most prominent

is a single deletion of phenylalanine at codon 508 in the NBD1, ΔF508

(http://www.genet.sickkids.on.ca/cftr/). This mutation gives rise to a protein that is

incapable of maturing properly and is tagged in the endoplasmic reticulum for

degradation. Mutations of the CFTR gene can be divided in 5 classes depending on the

effects they have on CFTR protein expression and function (Table 1). The genotype of

the mutation is in some cases linked to a specific phenotype which reflects the severity

of the disease. Some mutations have no or limited effect on CFTR function while others

are known to cause a severe form of CF disease. [3, 4]

Table 1: Classification of CFTR genotype mutations and the associated CFTR defect.

Mutation classification CFTR defect

Class I Absence of CFTR protein because of instable mRNA

Class II Translation of an abnormal protein that fails to mature properly

-> little plasma membrane expression of CFTR e.g. ΔF508

Class III Impaired activation and/or regulation of CFTR at the plasma membrane

Class IV Reduced chloride transport

Class V Reduced levels of normal CFTR protein

I.2 Diagnosis

The majority of CF patients has been diagnosed through family history and phenotypic

features consistent with the disease, including chronic sinopulmonary disease,

gastrointestinal and nutritional abnormalities, salt loss syndromes and genital

abnormalities in males. Today, many organizations, both in Europe and in the United

States, support the implementation of newborn screening for CF [5, 6] as early diagnosis

is associated with many benefits. Early therapeutic intervention can lead to reduced

disease severity resulting in a decreased burden of care throughout the patient his life

and reduced medical costs [7]. From the age of five days, newborns can be screened for

elevated levels of immunoreactive trypsinogen (IRT) in their blood which is indicative

for CF and is caused by pancreatic dysfunction. A positive screening has to be confirmed

by a second IRT test at the age of two weeks or by DNA screening for known CFTR gene

mutations. Both strategies have been reported to identify newborns at risk for CF with

9

approximately 90% to 95% sensitivity [6]. A positive screening result means the child is

at increased risk for developing CF. The result is communicated to the parents and

primary care provider and is followed by a sweat test with pilocarpine iontophoresis to

confirm the diagnosis [5].

I.3 Clinical manifestation

CFTR is expressed on the apical surface of specialized epithelial cells in sweatglands,

pancreas, gastrointestinal- and reproductive tract, submucosal glands and in the airway

epithelial cells [8]. Therefore, a malfunctioning chloride channel or other ion transport

defect which is caused by a mutated CFTR gene leads to complications in all these

organs, but especially involves the lungs and pancreas.

I.3.1 Sweat gland

In the sweat gland CFTR channels reabsorb chloride ions out of the sweat. In CF

patients, CFTR dysfunction leads to abnormalities in sodium chloride homeostasis

because of the reduced Cl- reabsorption [9]. This results in a high level of Cl- in the

sweat emerging on the skin. The main diagnostic test for CF, the pilocarpine

iontophoresis sweat test, is based on this clinical manifestation. In general, the sweat Cl-

concentration of a healthy individual is below 40 mmol/L. In CF patiensts Cl-

concentrations are usually ≥ 60 mmol/L and can be as high as 120 mmol/L [9].

I.3.2 Intestinal disease

Increased viscosity of intestinal fluids and a delayed intestinal transit time can lead to

meconium ileus (MI) when a baby with CF is born. MI is unique to CF and is

characterized by a complete obstruction of the intestine of the neonate. It occurs in 13

to 17% of all CF newborns and is associated with the CF genotype [10]. Homozygosity

for the ΔF508 mutation is correlated with the occurrence of MI [11].

Later on, in childhood, episodes of distal intestinal obstruction (DIOS) and constipation

can develop. Incidence of DIOS in childhood is relatively low (7-8%) but increases as

patients become older. In contrast, constipation is very common in both young and

older CF patients [10].

10

I.3.3 Liver disease

Hepatobiliary complications are the third leading cause of death in CF patients,

accounting for 2.5% mortality [12]. CFTR is expressed on the apical surface of

cholangiocytes (epithelial cells of the bile duct) where it controls the viscosity and

alkalinity of the bile fluid [13]. In CF patients an increased viscosity of the bile fluid

results in an obstruction of the intrahepatic bile ducts. An inflammation reaction is

initiated by elevated levels of toxic compounds in these obstructed ducts, Kupffer cells

become activated and produce collagen which leads to focal fibrosis, the first stage of

CF-associated liver disease [14]. The liver fibrosis causes an increased pressure in the

venae portae which can lead to splenomegaly and oesophagusvarices. This stage of liver

disease, where the overall liver function remains relatively unaffected, is still reversible

but without treatment it can slowly evolve in irreversible liver cirrhosis eventually

leading to liver failure [13, 15].

I.3.4 Reproductive tract disease

Men with CF are almost always infertile (over 98%) [16]. The human epididymis and

vas deferens are extremely vulnerable to the destructive effects of mutations in the

CFTR gene. Mucus plugging in the vas deferens during in utero development leads to

fibrosis and atrophy of the vas deferens resulting in congenital bilateral absence.

Although spermatogenesis proceeds normally, absence of the vas deferens causes an

obstructive azoospermia [17].

In women infertility is less common, up to 50% of CF women is able to become pregnant

[18]. Women usually have a normal anatomy but thick and tenacious cervical mucus

contributes to the reduced fertility by acting as a penetration barrier to sperm [17].

I.3.5 Pancreatic disease

CF is the primary cause of pancreatic exocrine dysfunction in childhood [19]. The

pancreas secretes digestive enzymes into the duodenum together with bicarbonate

(HCO3-)- rich fluid to neutralize gastric acid. When gastric acids enter the duodenum,

the release of secretin into the blood circulation is triggered. Secretin binds to its

receptor on pancreatic duct cells and stimulates the secretion of Cl- by CFTR through

cAMP production. Increased Cl- secretion is associated with increased Na+ secretion

11

followed by the movement of water into the lumen. Active CFTR also positively

regulates a Cl-/HCO3- exchange channel resulting in increased HCO3- levels in the

duodenum [20]. (Figure 2)

Figure 2: Regulation of pancreatic chloride and bicarbonate secretion. [20]

Reduced Cl- secretion in CF patients leads to a reduced HCO3- secretion by the epithelial

cells lining the ducts and a reduced water secretion. The resultant thick pancreatic

secretions progressively block the flow of digestive juices from the enzyme-secreting

cells (acini), leading to impaired digestion of fat, protein and starch. This causes

steatorrhoea, fat-soluble-vitamin deficiencies and malnutrition. A progressive

accumulation of digestive juices in the pancreatic duct will cause gradual degradation of

the enzyme-producing cells and subsequent pancreatic fibrosis [19]. Once a certain

degree of islet cells is no longer functional the patient will develop CF-related diabetes

mellitus because of insulin insufficiency and carbohydrate intolerance. [19, 21]

There is a clear correlation between genotype and pancreatic exocrine function.

Exocrine pancreatic insufficiency has been attributed to the presence of a severe

mutation in both CFTR alleles (Class I-III). Patients who carry a mild mutation usually

remain pancreatic sufficient [22].

12

I.3.6 Respiratory disease

Respiratory failure is the leading cause of morbidity and mortality in CF patients. The

lungs of CF patients appear normal at birth but become rapidly colonized by

Haemophilus influenzae, Staphylococcus aureus or both. Within a few years Pseudomonas

aeruginosa becomes the predominant lung pathogen [23].

I.3.6.1 Molecular basis of CF lung disease

There are several hypotheses to explain the molecular basis of this vulnerability of the

CF lung to chronic lung infections, the “low volume” hypothesis, the “high salt”

hypothesis, increased adherence to epithelial airway cells and reduced clearance of

bacteria after binding to CFTR receptors [24].

The “low volume” hypothesis is based on the assumption that the salt concentration in

airway surface liquid (ASL) in both CF patients and healthy individuals is the same.

Active CFTR channels in the lung epithelial cells do not only secrete chloride ions but

also inhibit the epithelial sodium channel, ENac, which reabsorbs sodium from the ASL

[24]. Therefore, the malfunctioning of CFTR in CF patients results in an increased

sodium absorption, through activation of ENac, accompanied with a Cl- absorption

through alternative pathways (either paracellularly or via other ion channels). This

increase in ion reabsorption results in the movement of water from the apical side of the

epithelial cells to the basolateral side to preserve the isotonic character of the ASL

(Figure 3). The resulting decrease in apical surface volume makes the cilia collaps,

impairing the normal ciliary movement of mucus. This defect in mucociliary clearing

would make the lungs of CF patients more susceptible to colonization with bacteria [25,

26].

13

Figure 3: Mechanism behind the low volume hypothesis. Ion transport in lung epithelial cells of a (A)

healthy individual compared to that in a (B) CF patient. [24]

The “high salt” hypothesis postulates that salt concentrations in the ASL of CF patients

are higher than in ASL of normal patients. Because of the ability of lung epithelial CFTR

to pump more Cl- ions into the cell compared to the amount of water, a hypotonic

environment is created outside the cell of healthy individuals. In CF patients this Cl-

reabsorption mechanism is disrupted. As a result, to preserve electroneutrality, less Na+

is absorbed, resulting in an increased ASL salt concentration (Figure 4) [25]. The lack of

an alternative pathway for Cl- absorption and the fact that CFTR does not inhibit the

ENaC channel are key features of this hypothesis [24]. Smith et al. showed that a salt-

sensitive antimicrobial peptide, of which the antimicrobial activity is inhibited by an

increased salt concentration, is present in the ASL. The increased sensibility of CF

patients for lung infections could thus be explained by this impaired innate host defense

mechanism [27].

Figure 4: Mechanism behind the high salt hypothesis. Ion transport in lung epithelial cells of a (A) healthy

individual compared to that in a (B) CF patient. According to this hypothesis there is no inhibition of ENaC

by CFTR nor an alternative Cl- absorption pathway.

14

Both hypotheses offer no explanation for the limited number of pathogens that colonize

the CF lung. Adherence of P. aeruginosa to respiratory epithelial cells is increased in CF

patients and during regeneration of the respiratory epithelium [28]. The basis for this

increased adherence could be related to the presence of a asialoganglioside-1 protein

(asialo GM-1) that is expressed on the apical surface of respiratory epithelial cells and

acts as a receptor for various pathogens including P. aeruginosa and S. aureus [28]. It

has been shown that asialo GM-1 is more abundant on the surface of cells expressing

mutated CFTR and on regenerating epithelial cells [29]. As inflammation frequently

occurs in the CF lung, areas of regenerating epithelium are likely to be present,

promoting P. aeruginosa adherence through the increased asialo GM-1 expression.

Some researchers have described the lack of binding of clinical P. aeruginosa isolates to

asialo GM-1 receptors, especially those isolates without pili or flagella including mucoid

strains refuting the relevance of asialo GM-1 in chronic P. aeruginosa infections [30, 31].

Another receptor for P. aeruginosa is CFTR itself. After binding to CFTR, P. aeruginosa is

internalized and cleared from the airways through shedding of the epithelial cells [32].

Reduced CFTR expression or expression of mutated CFTR impairs this clearance

mechanism and allows P. aeruginosa to persist in the airways. Moreover, the process of

P. aeruginosa binding to CFTR induces the activation of NF-κB and the production of

cytokines and antimicrobial peptides (AMPs) which are important factors in the innate

immunity towards infections [33]. As CFTR mediated internalization is specific for P.

aeruginosa this hypothesis is the first to provide a molecular basis for the high-level

association of P. aeruginosa and CF lung disease. [34]

I.3.6.2 Microbiology of CF lung disease

The upper respiratory tract of a healthy individual is colonized with a wide variety of

microorganisms, called commensals, including non-typeable H. influenzae and S. aureus.

In contrast, the lower airways remain sterile by means of various innate defense

mechanisms present in the lung. In CF patients some of these defense mechanisms are

impaired, making them more susceptible to colonization of the lower respiratory tract.

Only a limited number of different pathogens cause infections in CF patients and they

are frequently acquired in an age-dependent manner (Figure 5).

15

Figure 5: Isolation of respiratory germs from CF patients by age (adapted from the 2011 annual data

report published by the US CF Foundation).

Staphylococcus aureus

In infancy and early childhood, S. aureus is the most frequently detected in BAL

(bronchoalveolar lavage) samples [35]. The attention towards S. aureus has grown

during the last decade because of the increased isolation of methicillin-resistant S.

aureus (MRSA) from CF patients. Approximately 15% of CF patients are colonized with

MRSA and this is associated with poorer lung functions [36]. Although S. aureus has

been shown to persist in the CF lung over years, controversy about a prophylactic anti-

staphylococcal therapy remains. Some studies indicate that a reduction in early S.

aureus infections is associated with earlier or more frequent P. aeruginosa infections

[37-39]. Moreover, a prophylactic therapy in the presence of MRSA strains could

enhance the emergence of further resistance. [36]

Pseudomonas aeruginosa

The most significant pathogen colonizing the lungs of CF patients is P. aeruginosa.

Approximately 80% of the CF population gets infected with this pathogen [40]. Most P.

aeruginosa infections are acquired in the environment [41], outside the hospital, or

through cross-infection between siblings [42, 43]. In general, the overall risk of patient-

to-patient transmission with P. aeruginosa is considered to be low but interpatient

transmission in different CF care centers has been well documented [41, 44, 45].

Adequate hygienic measures in hospitals and CF centers and segregation of CF patients

according to microbiological status reduces the risk of hospital acquired cross-infections

16

[24]. Reducing the risk of acquisition is considered at least equally important as early

treatment of an established P. aeruginosa infection.

The average age for the first P. aeruginosa positive lower airway culture from CF

children is 23 months [46] and acquisition of this pathogen is associated with clinical

deterioration [47-49]. Early antibiotic treatment can prevent or delay chronic P.

aeruginosa infections which are usually preceded by a period of recurrent, intermittent

colonization with P. aeruginosa [50]. During intermittent P. aeruginosa infection

transition from a planktonic, non-mucoid phenotype to a sessile, mucoid phenotype

takes place. This mucoid phenotype, characterized by alginate production, is better

adapted to persist in the CF lung and grows as a sessile community, called a biofilm [51].

Microorganisms growing as a biofilm are protected from dehydration and phagocytosis

and are highly resistant towards antibiotics, providing a survival benefit in the CF lung

environment. The mucoid P. aeruginosa phenotype is not observed outside the host

environment and is associated with a more rapid decline in lung function [52, 53].

Burkholderia cepacia complex

The Burkholderia cepacia complex (Bcc) is a group of gram-negative β-proteobacteria

currently consisting of at least 18 closely related species. All Bcc species have been

isolated from CF patients except for Burkholderia ubonensis [54]. Burkholderia

cenocepacia and Burkholderia multivorans are the most frequently recovered Bcc species

from CF sputum samples followed by Burkholderia gladioli, not belonging to the Bcc

(Figure 6). Incidences of B. cenocepacia and B. multivorans infections change over time

and differ from place to place [55, 56]. Overall prevalence of Bcc infections in CF

patients varied between 3-5% during the last decade [40].

Figure 6: Distribution of Burkholderia species among U.S.

CF patients. “Indeterminate” refers to strains that

phylogenetically are members of Bcc but that can not be

placed into one of the 18 defined species. [57]

17

Although the majority of Bcc infections among CF patients are acquired from distinct

environmental sources, Burkholderia species have an enhanced capacity for interpatient

transmission compared to P. aeruginosa. Since the late 1980s, person-to-person spread

of Burkholderia species has been documented [58-60] followed by the identification of

epidemic strains, including the B. cenocepacia ET12 (electrophoretic type 12) strain

[61]. Nevertheless, stringent infection control policies have been shown to fully

eliminate the acquisition of epidemic strains.

The acquisition of Burkholderia generally occurs later in the course of CF airway disease

and most commonly results in chronic infections [62, 63].

Other pathogens

H. influenzae which is primarily known as a commensal of the upper respiratory tract is

isolated from the lungs of approximately 20% of CF children under the age of one year

[40].

Other pathogens like Stenotrophomonas maltophilia, Achromobacter xylosoxidans,

Aspergillus fumigatus, Mycobacteria and some viruses are also capable of infecting the

CF lung but usually in a later stage of CF disease. S. maltophilia is most likely acquired

from environmental sources, in- or outside the hospital [64, 65], and primarily causes

transient and recurrent infections [66]. As with Bcc, the true incidence of S. maltophilia

infection in CF patients is difficult to ascertain. The same is true for A. xylosoxidans,

which has long been recognized as a CF pathogen. Most A. xylosoxidans infections are

nosocomially acquired with outbreaks attributed to contaminated solutions [67, 68].

Although infections with A. xylosoxidans are frequently transient, chronic infections for

several years have been described [69, 70].

I.4 Pulmonary therapy

Pulmonary therapy in CF patients focuses on three main goals: the clearance of viscous

mucus out of the airways, the eradication of bacterial infections and the reduction of

pulmonary inflammation. Patients with end-stage pulmonary damage are considered

for lung transplantation.

18

I.4.1 Airway clearance

The clearance of airway secretions is a primary therapy for CF patients as the increasing

volumes of viscous mucus, that can not be cleared by the cilia, form the main source of

pulmonary complications. The accumulation of mucus is associated with the persistence

of pathogens trapped in this mucus, resulting in inflammation and destruction of lung

tissue. In addition, the thickened mucus obstructs the smaller airways leading to the

formation of hypoxic regions which induces P. aeruginosa biofilm formation [71].

I.4.1.1 Mechanical clearance

A variety of airway clearance therapies have been developed as primary therapy for CF

patients. Although no method has been proven superior to any other, there can be

differences in effectiveness between patients. Chest physiotherapy involves chest

percussion in different positions to mechanically induce the clearance of secretions from

all lobes of the lungs. Many clinicians believe in the long-term benefits of this therapy if

daily maintained. Nevertheless, the burden on the patient is not to be underestimated

and clearly influences patient compliance [72, 73]. To put less burden on the patient the

“High-frequency chest compression” vest can be used. This airway clearance device

applies high-frequency oscillations to the patients chest wall assisting in removing the

secretions [74].

Other techniques like “Active cycle of Breathing” and “Autogenic drainage” are breathing

techniques aiming to increase the expiratory airflow accompanied with the movement

of airway secretions out of the lungs [75, 76]. A PEP (positive expiratory pressure)

mask also helps in clearing retained mucus from the airways. By breathing out through

the mask, with an appropriate resistor in place, a positive pressure is build up in the

lungs. This pressure keeps the airways open en allows air to flow behind mucus plugs

helping them move upwards [77].

I.4.1.2 Mucolytics

Reducing the viscosity of the sputum with mucolytics also allows better airway

clearance. Both hypertonic saline and mannitol have been demonstrated to reduce

mucus viscosity [78]. Inhalation of hypertonic saline or mannitol, can correct the

underlying defect in CF as it induces the flow of water into the lumen of the airways.

The rehydration of the periciliary layer restores the function of the cilia allowing

19

improved clearance [79]. In addition, hypertonic saline may have an effect on airway

inflammation and bactericidal activity within ASL (Figure 7). Glucosaminoglycans

present in the ASL are well-known to bind AMPs and cytokines including interleukin-8

(IL-8) which is considered a key cytokine in the inflammatory respons. The interaction

of IL-8 with glucosaminoglycans can be disrupted by hypertonic saline. While at first

sight this seems to increase the chemotaxis of neutrophils by increasing the amount of

free IL-8, the increased susceptibility of IL-8 to proteolytic degradation reduces the

overall IL-8 burden [80]. By disrupting the electrostatic interactions between AMPs and

glucsoaminoglycans the bactericidal effect of ASL can also be improved [81]. Still, the

long-term effects of chronic hypertonic saline therapy have not been defined.

Figure 7: Different effects of hypertonic saline therapy in CF ASL [79].

Another widely administered treatment for this purpose is the once-daily inhalation of

recombinant human DNAse (rhDNAse), which was FDA (Food and Drug Administration)

approved in 1994 under the proprietary name of Pulmozyme [82]. rhDNAse decreases

the viscosity of CF mucus by cleaving polymeric extracellular DNA (eDNA), mainly

derived from damaged neutrophils as a result of progressive inflammation. It is shown

that rhDNAse improves lung function and decreases the risk of pulmonary

exacerbations over a relatively short period of time. A reduction in mortality and/or

morbidity after long-term Pulmozyme therapy still has to be proven [83, 84].

20

I.4.2 Anti-inflammatory therapy

Neutrophil-driven inflammatory reactions are characteristic for CF airway disease. If

untreated this results in localized dilatation of parts of the bronchial tree, caused by

destruction of muscle- and elastic lung tissue, and ultimately respiratory failure. Airway

inflammation has been observed in young patients, sometimes even without identifiable

infection [85, 86]. Some investigators believe that malfunctioning of CFTR directly

dysregulates the host inflammatory response leading to an imbalance between pro-

inflammatory and anti-inflammatory mediators and favouring excessive inflammation

[87].

The most promising anti-inflammatory agent at the moment is azithromycin which has

both an anti-inflammatory and antimicrobial effect [88]. Chronic therapy with

ibuprofen and/or oral corticosteroids are no first choice treatment for CF lung

inflammation because of their adverse effects. In some cases, high dose administration

of ibuprofen is associated with serious gastric bleedings [89] while chronic oral

prednisone significantly reduces growth [90].

I.4.3 Antibiotic therapy

Effective antibiotic therapy to treat CF lung disease has undoubtedly contributed to the

increasing live expectancy of CF patients and is directed at preventing, eradicating and

controlling bacterial lung infections.

I.4.3.1 Early antibiotic treatment

Early antibiotic treatment following the first isolation of bacteria from the lower

respiratory tract can delay the onset of a chronic colonization associated with

progressive lung damage [91]. Different studies described a total eradication of P.

aeruginosa following early antibiotic therapy [92, 93]. Prophylactic therapy in infants in

order to prevent acquisition of CF pathogens, like S. aureus, is under debate as several

studies question the benefit of such a therapeutic approach [36].

I.4.3.2 Maintenance therapy

Once chronic colonization with P. aeruginosa takes place, maintenance therapy is

needed in order to delay progressive lung damage and reduce the frequency of

21

pulmonary exacerbations. Chronic P. aeruginosa infections are generally treated with

inhaled colistin or tobramycin as first choice maintenance therapy [91, 94, 95].

Ciprofloxacin has a bactericidal activity against many CF pathogens and is orally active.

Nevertheless, the emergence of resistance in S. aureus and P. aeruginosa prevented its

use as first choice therapy [96, 97] and limited its use to treatment of P. aeruginosa

exacerbations [98].

Aztreonam lysine solution for inhalation has recently become an alternative for

managing chronic P. aeruginosa infections as it results in greater reduction in the

number of exacerbations and better pulmonary function compared to tobramycin

nebulizer solution when administered over a short period of time [99]. Another new

treatment option for CF pulmonary infections, once daily inhalation of levofloxacin, is

currently undergoing phase III trials and looks promising for treating P. aeruginosa

infections [100].

Although macrolides, like erythromycin and azithromycin, have little activity against

planktonic P. aeruginosa cells they are effectively used in the treatment of pulmonary P.

aeruginosa infections in CF patients. Sub-inhibitory concentrations of macrolides are

shown to inhibit alginate production, an important exopolysaccharide in the transition

of non-mucoid to mucoid P. aeruginosa strains [103]. Macrolides can thus prevent or

delay the transformation of early P. aeruginosa strains into a sessile, more resistant

phenotype which is better adapted to persist in the CF lung. It is even suggested that

apart from the antibiotic effect, macrolides exert an additional anti-inflammatory effect

which often leads to an overall improvement of the clinical status of the patient. Just like

for ciprofloxacin, the administration of macrolides is not recommended for children

younger than 6 years and long term impact of the treatment is yet to be investigated

[104].

A culture-driven approach in which the antibiotic therapy is stopped once sputum

cultures reappear sterile, is proven as effective as routine cycles of antibiotic therapy

every 3 months regardless of culture results [101], but because of the reduced time of

antibiotic treatment, a culture-driven approach is preferred. In the case of young

children, who are often unable to expectorate sputum, treatment decisions are

commonly based on oropharyngeal cultures which are poorly reflective of lower airway

colonization and thus not a good base for culture-driven therapy [102].

22

I.4.3.3 Exacerbation therapy

Once patients are chronically infected with CF pathogens they experience episodes of

pulmonary exacerbations. Repeated exacerbations have to be avoided because of their

negative effect on pulmonary function and life expectancy of the patient. Therapy for

these exacerbations include intravenous administration of antibiotics based on recent

cultures of airway secretions. A combination therapy of a β-lactam and an

aminoglycoside is often recommended for exacerbations caused by P. aeruginosa or B.

cenocepacia because of the provided synergy and the slow emergence of resistance

[103]. Therapy is initiated in the hospital to allow immediate follow up of the patients

clinical status and to monitor drug concentrations if necessary, especially for

aminoglycosides [98].

23

24

Biofilms: Slime and the City (Wendy Orent)

25

Chapter II: BIOFILMS

II.1 Introduction

In 1933 Henrici described the adherence of aquatic bacteria to a glass surface [104].

These bacteria presented on the glass in the form of microcolonies that increased in size,

indicating that these bacteria were actually growing. Henrici came to the conclusion

that most water bacteria are not free-swimming organisms but rather sessile organisms

that grow on submerged surfaces. Zobell continued in 1935 his research on fouling

based on the findings of Henrici. He also observed the adherence of water bacteria to

submerged surfaces and indicated them as “primary film formers” that favor the

attachment of subsequent fouling organisms by providing them with nutrients [105].

These findings together with the work of Bill Costerton (published in 1978) gave rise to

the realization that sessile communities, referred to as biofilms, dominate in different

microbial ecosystems [106]. Biofilm formation even lies at the basis of many, if not all,

persistent human infections. Antibiotic therapy towards biofilms often fails and

eradication by the host defense mechanism rarely takes place [107]. Many

investigations led to our current definition of a biofilms as a highly structured microbial

community enclosed in a self-produced matrix. However, biofilm cells differ from their

planktonic counterparts not only in a conformational point of view as also gene

expression in sessile cells differs profoundly from that in planktonic cells [108].

II.2 Biofilm formation

The application of confocal laser scanning microscopy (CLSM) for studying the

development of a biofilm dates from 1991. CLSM images show a sessile community of

bacterial cells enclosed in a matrix penetrated with open water channels. Through these

channels, biofilm cells have access to nutrients, including oxygen [109].

The structural development of a biofilm is often described as a five-stage process

(Figure 8).

26

Figure 8: Five stages of biofilm development. Each stage in the developmental process is associated with a

microscopic picture of a developing P. aeruginosa biofilm. All photomicrographs are shown at the same

scale (Modified from [110]).

The first step in biofilm formation is the reversible attachment of cells to the surface.

Attached bacterial cells can redistribute through surface motility, called twitching. Type

IV pili, which extend and contract, are responsible for this kind of movement. Through

twitching motility, attached cells move towards each other, aggregate and form

microcolonies [111]. Binary division of attached cells also contributes to this

microcolony formation. The second step in biofilm development is the production of

extracellular polymeric substances (EPS) providing irreversible attachment of bacteria

to the surface [112]. This self-produced matrix also provides structural support during

further development of the biofilm architecture (step 3 and 4). The matrix is composed

of polysaccharides, proteins and nucleic acids and is essential for acquiring structural

complexity in a biofilm. Less EPS production is associated with a flat, undifferentiated

biofilm while EPS overproduction stimulates the formation of mushroom- and tower-

shaped cell clusters [113, 114]. The viscoelastic character of the biofilm is also

determined by EPS. Through cross-linking of negatively charged polymers in the

extracellular matrix, with the aid of multivalent cations, biofilms possess a certain

degree of cohesive strength [115]. The strong well-organized unit makes it possible for

the biofilm to persist in detrimental environmental conditions. The last stage in biofilm

27

development is the dispersion of single cells. These cells can colonize new surfaces

starting the development of a new biofilm community.

II.3 Resistance of biofilms

Standard antimicrobial therapies often fail to eradicate a biofilm infection. Both biofilm

structure and the phenotype of microorganisms living in the biofilm are responsible for

this resistance to killing by cidal antimicrobial agents as well as by the innate human

defense system. Cells living in a biofilm can be 10 to 1000 times more resistant to

antimicrobial killing [116, 117]. Several resistance mechanisms have been suggested to

explain these great differences in susceptibility of planktonic cells and biofilm associated

cells [118].

II.3.1 Restricted penetration of the antimicrobial agent

It has been suggested that the biofilm matrix has the capacity to prevent the access of

antimicrobial agents to the biofilm cells. Although not all equally important, different

factors are involved in this penetration barrier, including sorption, electrostatic

interactions and enzymatic reactions [119].

II.3.1.1 Diffusion barrier

For a solute that does not react with biofilm matrix components the theoretically

calculated diffusion barrier can not account for the observed resistance of biofilms

towards antibiotics [119]. However, many matrix polymers are charged, allowing

electrostatic interactions with oppositely charged antimicrobials. These electrostatic

interactions (or even sorption to matrix components) can delay the penetration of

antimicrobials into the biofilm [120, 121]. In most cases, the penetration of small

antibiotic molecules is not entirely blocked, leading to a postponed antibiotic effect

rather than a lack of effect which does not explain the increased resistance of biofilms. If

electrostatic interactions are the primary cause of the delayed antibiotic penetration,

chronic administration during patient therapy should rapidly lead to saturation of these

binding sites resulting in increased diffusion rates in vivo. Still, biofilm infections located

in poorly accessible regions will be continuously exposed to lower antibiotic

concentrations. In these regions a delayed penetration of the antimicrobial agent

through the biofilm may play a role in biofilm tolerance [122].

28

II.3.1.2 Enzymatic reactions

When biofilm cells produce antibiotic modifying enzymes like β-lactamases, these

enzymes will be present in the biofilm matrix degrading the antimicrobial agent while it

penetrates into the biofilm. The combination of restricted penetration and degradation

has been shown to contribute to the resistance of β-lactamase producing biofilms [123].

β-lactamases can appear in the biofilm matrix either through membrane vesicles

containing the enzyme or through lysis of bacteria after exposure to antibiotics. For

imipenem and piperacillin it has been shown that they induce the production of β-

lactamase in P. aeruginosa biofilms, leading to their own degradation [124]. Similarly,

the production of catalase in a P. aeruginosa biofilm protects the cells from being killed

by hydrogen peroxide [125].

II.3.2 Altered growth rate/metabolic activity of biofilm cells

Because of the limited nutrient supply in a biofilm environment, biofilm cells grow

slower and present with lower metabolic activity than their planktonic counterparts.

Soren Molin’s research group indicated that microorganisms located in the center of

biofilm clusters are often even in a dormant phase, are characterized by a decreased

sensitivity to antimicrobials and can provide a nucleus for regrowth after antibiotic

treatment (Figure 9) [126].

Figure 9: Metabolic activity in a biofilm microcolony. Biofilm cells residing near the surface present with

greater metabolic activity than cells residing in the center of a microcolony. [110]

Many antibiotics are more effective in killing growing cells and some, like penicillin and

ampicillin, even totally lack killing activity against non-dividing cells [127]. It is

important to notice that bacterial cells growing in different regions of the biofilm will

experience different environmental stresses. Cells at the bulk water interface will grow

faster and have a higher metabolic activity compared to cells located deeper in biofilm

29

clusters. Therefore, the activity of many antimicrobials will be greater at the surface of

microcolonies than in the center [128]. In sputum of chronically infected CF patients a

stationary phase-subpopulation of P. aeruginosa cells seems to be present [129]. This

explains why monotherapy with antibiotics that are only active against dividing cells is

not efficient in eradicating these infections. In contrast, colistin is primarily active

against non-dividing biofilm cells. Metabolically active cells confer resistance to colistin

through the addition of aminoarabinose to the lipid A moiety of lipopolysaccharide

(LPS) molecules, resulting in decreased uptake of the antibiotic in the cell [130, 131]. In

addition, upregulation of efflux pumps in metabolically active cells also contributes to

this tolerance to colistin [131].

In general, an overall slow down of metabolic activity can protect bacterial cells from

environmental stresses. This can be regulated by the expression of RpoS, a sigma factor

known as a general regulator of stress responses [132]. By regulating the transcription

of other proteins, RpoS protects cells from environmental challenges and prepares them

for subsequent stresses. RpoS is transcribed in late-exponential growth phase in

planktonic cultures [133] as well as in environments with high cell density [134], like

biofilms, where it contributes to the increased resistance towards antimicrobial killing.

II.3.3 Increased frequency of mutation

Biofilm-growing bacteria have an increased mutation frequency compared to isogenic

planktonic bacteria. Downregulation of antioxidant enzymes in P. aeruginosa biofilms

increases the prevalence of DNA damage, enhancing the frequency of mutagenic events

[135]. Single mutations in specific genes can lead to e.g. altered antimicrobial targets,

reduced antimicrobial uptake and increased expression of efflux pumps. Because of

their close proximity, biofilm cells are also amenable to horizontal gene transfer by

which they can acquire genes, encoding antibiotic modifying enzymes or efflux pumps,

from other isolates [136, 137].

Because of the great incidence of mutation within a biofilm population some

investigators suggest the presence of hypermutable strains in biofilms. In CF patients

the isolation of hypermutable P. aeruginosa strains, associated with increased antibiotic

resistance, has been described [138]. The increased mutation rate of these isolates is

often due to alterations in genes of the DNA repair system. The increased amount of

oxidative stress present in a biofilm also influences the formation of hypermutable

30

strains. During antibiotic therapy resistant mutants will have better chances of survival

and will be selected for, making the problem of persistent infections even bigger.

II.3.4 Quorum sensing

Bacteria use signaling molecules to communicate with each other, a process referred to

as quorum sensing (QS) [139]. The accumulation of these signaling molecules enables

the bacteria to estimate the number of other bacteria in their environment. When the

signaling molecules reach a threshold concentration, bacteria can respond by inducing

transcription of specific genes. By this means bacteria can respond quickly to

environmental threats as a group rather than as autonomous individuals. The QS

molecules described for gram-positive bacteria are usually small peptides while gram-

negative bacteria communicate primarily through members of the class of N-acyl-L-

homoserine lactones [140]. It has been described that QS is responsible for the

tolerance to tobramycin and the innate inflammatory respons by polymorphonuclear

leucocytes in P. aeruginosa biofilm cells [140, 141].

II.3.5 Biofilm phenotype

The “biofilm phenotype”, described by Costerton, is characterized by a unique gene

expression pattern following attachment to a surface [142]. These gene expression

patterns are not observed in their planktonic counterparts what makes biofilm cells

morphologically and biochemically distinct from free-living cells. It is likely that

activation of certain genes during biofilm formation contributes to biofilm resistance.

Several biofilm resistance mechanisms are under the same regulatory control indicating

that multiple resistance mechanisms might be operating at the same time within the

biofilm population [107].

II.3.6 Persister cells

The majority of cells in a biofilm is often rapidly killed by bactericidal antibiotics that are

active against slow growing cells. A small fraction, referred to as persisters, is able to

survive the presence of such a bactericidal antibiotic [143]. The same observation has

been made in planktonic cultures (Figure 10A), although biofilms typically have a larger

fraction of persisters. The surviving fraction of a planktonic culture can normally be

cleared by a healthy immune system in contrast to the persistant fraction in a biofilm

31

which is protected by the polymeric matrix [144]. The immune system fails to clear

these biofilm persisters therefore providing a niche for regrowth once antibiotic

pressure is elevated. After regrowth of these persisters, a biofilm population

indistinguishable from the original biofilm is created (Figure 10B) [145]. This indicates

that persisters are not mutants. Until now the nature of persistence is unknown. It is

possible that the growth rate of persisters affects their resistance.

Figure 10: A) Biphasic killing pattern in response to antibiotics B) Biofim model of persister survival. Red

circles represent cells susceptible to antibiotic killing; blue circles respresent cells resistant to antibiotic

killing (persister cells). Once antibiotic concentration drops, persister cells regrow and form a new

biofilm population consisting of normal and persister cells. This clearly indicates that persister cells are

no mutants as they do not pass their phenotypic characteristics to their progeny. [145, 146]

II.4 Biofilm formation in the CF lung

The number of infectious diseases associated with biofilms is large. Especially oral

infections, chronic infections and colonization of medical devices are often related with

biofilm formation. The difficulty of eradicating these biofilm infections is of major

concern [107].

Biofilm infections can be confirmed by direct examination of infected tissue, often by

CLSM and scanning electron microscopy (SEM). If biofilms are present, microscopic

images show microorganisms living in clusters, surrounded by an extracellular matrix.

Such biofilm-like structures have been detected in different samples related to chronic

infections, including the sputum and autopsy specimens of CF lungs chronically infected

with P. aeruginosa (Figure 11) [147]. The presence of P. aeruginosa biofilms is thought

to be associated with the onset of chronic infections. Isolates colonizing the CF lung can

32

convert to a phenotype with better adherence properties and an increased production of

extracellular matrix material which coincides with permanent infection [148, 149].

Another sign that indicates biofilm phenotype is the absence of flagella, pili and motility

in many CF P. aeruginosa isolates from chronically infected patients [150]. It has been

described that when cells adhere to a surface and form a biofilm they adapt to their

environment by decreasing the expression of their motility machinery making them

more persistent and less invasive.

Figure 11: SEM images of P. aeruginosa PAO1 in artificial sputum medium (left image) and sputum

extracts (right image). In both conditions bacterial cells are encapsulated by polymeric substances

indicating biofilm formation in the CF lung environment. [147]

One of the problems with the formation of a biofilm during infections is that they

provide a niche for the persistence of bacteria at the place of infection. Through the

detachment of cells or cell aggregates from the biofilm, microbial cells can distribute and

colonize new uninfected surfaces. In addition, through the encapsulation in an

extracellular polymeric matrix, the opsonisation of bacteria mediating phagocytosis and

elimination of cells is impeded [151].

II.5 Intervention strategies related to CF biofilm infections

All components involved in biofilm formation can be considered as possible targets for

biofilm therapy. Components like pili that are important in the initial steps of biofilm

formation are possible targets for prophylactic therapy while exopolysaccharides that

are required for the maintenance of biofilm structure can be promising targets for

biofilm eradication.

33

II.5.1 Inhibition of microbial attachment

The protein lactoferrin is a component of the innate human defense system.

Independent from its antimicrobial activity it also prevents P. aeruginosa biofilm

formation, probably through the chelation of ferric iron [152]. Moreover, a destructive

effect on mature P. aeruginosa biofilms is noted.

Superti and colleagues described the binding of lactoferrin to B. cenocepacia cable pili,

thereby inhibiting bacterial attachment to mucin and epithelial cells [153]. These

results provide evidence for a role of this natural protein in preventing P. aeruginosa

and B. cenocepacia biofilm infections.

II.5.2 Matrix degradation

The biofilm matrix which is sometimes referred to as “the house of biofilm cells”, is

essential for biofilm stability [154]. The biofilm matrix is for over 90% comprised of

water next to EPS like lipids, polysaccharides, proteins and nucleic acids. The exact

composition can vary greatly among biofilms. Even different compositions between

strains of a single species have been described [155].

Degradation of the biofilm matrix as a remedy against biofilm infections is increasingly

being investigated. Matrix degrading components can either be used as monotherapy to

prevent biofilm formation or in combination with antibiotics to improve their

penetration into the biofilm.

II.5.2.1 Exopolysaccharides

In many bacteria exopolysaccharides are essential for biofilm formation and/or biofilm

maintenance. Exopolysaccharides serve a wide range of functions involving structural

stability, adherence and protection against environmental predators.

Alginate, Psl and Pel

Alginate which is abundantly present in the biofilm matrix of mucoid P. aeruginosa

strains is one of the best studied exopolysaccharides. It is a high molecular mass linear

copolymer, consisting of homopolymeric blocks of 1→4 linked β-D-mannuronate and

heteropolymeric sequences with random distribution of β-D-mannuronate and α-L-

guluronate residues (Figure 12A). P. aeruginosa alginate molecules are partially

34

acetylated at the O-2 and/or the O-3 positions on the mannuronate residues which is

important for cell-cell and cell-surface interactions [156]. Although alginate is not

essential for biofilm formation, it contributes to the development of a complex biofilm

architecture. Because of its polyanionic character, alginate can interact with positively

charged antibiotics, like aminoglycosides, thereby reducing the antibiotic effect and

protecting the biofilm community [157].

Non-mucoid P. aeruginosa biofilms contain little if any alginate. By contrast, two

carbohydrate-rich polymers, Pel and Psl, are the primary matrix polysaccharides. Psl

which is anchored to the cell surface, plays an important role in surface attachment and

cell aggregation by providing cell-cell and cell-surface interactions [158]. A recent study

indicated that this polysaccharide also plays a crucial role in mucoid biofilm

development [159]. In addition, Psl is able to protect P. aeruginosa from neutrophilic

phagocytosis by inhibiting efficient opsonization, providing a survival advantage for

non-mucoid P. aeruginosa strains in vivo [160].

A second important polysaccharide in a non-mucoid biofilm matrix is Pel which plays a

key role in maintaining cell-cell interactions in a developing biofilm. Apart from its

structural importance it also protects against aminoglycosides, probably through

sequestering or binding of the antibiotics [161].

poly-β-1,6-N-acetylglucosamine

Poly-β-1,6-N-acetylglucosamine (PNAG) is a positively charged linear homoglycan

composed of β-1,6-N-acetylglucosamine residues with approximately 20% deacetylated

residues (Figure 12B). It has primarily been described as a major component and

virulence factor in staphylococcal biofilms where it is referred to as “polysaccharide

intercellular adhesin” (PIA) [162]. Its production in other bacteria like Escherichia coli

[163], Actinobacillus pleuropneumoniae [164], Acinetobacter baumannii [165],

Aggregatibacter actinomycetemcomitans [166], Bordetella spp. [167], Yersinia pestis

[168] and some Bcc species [169] has also been confirmed. PNAG is attached to the

bacterial cell surface and provides protection against specific and non-specific human

defense mechanisms [162]. Degradation of PNAG by Dispersin B inhibits biofilm

formation and disperses established biofilms [169, 170]. This anti-biofilm effect has

been confirmed in some Burkholderia species although PNAG is only a minor component

of these biofilms [169].

35

Cepacian

The mucoid phenotype in B. cepacia complex strains is characterized by the production

of an exopolysaccharide, named cepacian (Figure 12C). Synthesis of cepacian occurs in

approximately 80 to 90% of Bcc strains isolated from CF patients where it contributes to

both colonization and biofilm persistence [171]. Cepacian is not essential for biofilm

formation but it contributes to the formation of thick and mature biofilms. In 2006,

Cescutti et. al. reported for the first time the isolation of a lyase for cepacian, produced

by a bacterium of the genus Bacillus [172]. Another study described the degradation of

cepacian after short exposure to NaClO implicating its role as a scavenger of reactive

oxygen species (ROS) produced by the innate immune system [173].

Figure 12: A) heteropolymeric region of alginate with α-L-guluronate (G) and β-D-mannuronate residues

(M) B) repeating unit of PNAG C) primary structure of cepacian repeating unit.

(http://www.intechopen.com)[172]

36

II.5.2.2 Extracellular DNA

eDNA has been detected in many different biofilms from various origins and species

composition. It is believed to be present in every biofilm as a residual material from

lysed cells. Some studies however indicate differences between genomic DNA and eDNA

indicating that an additional active secretion system should be involved [174, 175].

Whitchurch et al. were the first who reported on the functional relevance of eDNA in

biofilms [176]. They indicated that eDNA is essential for P. aeruginosa biofilm formation

and that young biofilms are amenable to dispersion by DNAse treatment, suggesting that

this could be useful to prevent biofilm-related infections. Apart from the structural

function of eDNA, Mulcahy et. al described a mechanism by which small amounts of

eDNA contribute to antibiotic resistance [177]. Through the chelation of divalent

cations that are essential for stabilizing the outer membrane, large amounts of eDNA can

exert an antibiotic effect in P. aeruginosa. In contrast, when eDNA is present in the

biofilm matrix at sub-inhibitory concentrations it induces antibiotic resistance. Through

the creation of a cation-limited environment the PhoPQ and PmrAB two-component

regulatory systems are activated, inducing the expression of a “cationic antimicrobial

peptide” (CAP) resistance operon. This increased expression of the CAP operon

initiates the activation of an LPS modification pathway resulting in the addition of 4-

amino arabinose molecules to the lipid A moiety of LPS. Through the addition of 4-

amino arabinose molecules, the overall negative charge of lipid A decreases, hampering

the initial binding and subsequent intracellular uptake of AMPs, aminoglycosides and

polymyxin in gram-negative bacteria. The abundance of eDNA in the sputum of CF

patients could also in part contribute to the resistance of P. aeruginosa biofilms to

positively charged antimicrobials.

The group of Bakaletz indicated that free- or eDNA is often stabilized by the presence of

an “integrating host factor” (IHF) [178]. IHF is a DNA binding protein that contributes to

the structural lattice formation of free DNA and is located at those points where DNA

strands are bending to form this grid-like structure. They showed that the presence of

such IHF molecules protect free DNA from being degraded by DNAse. Therefore, a

combination therapy of DNAse and anti-IHF serum could considerably improve

therapeutic outcomes in decreasing mucus viscosity and eradicating biofilm infections

in CF patients [179].

37

II.5.3 Other anti-biofilm strategies

Conventional antimicrobial therapy is usually ineffective in eradicating biofilm-

associated infections. There is an urgent need for the development of new antibiotics,

new delivery systems or effective combination therapies that will benefit patients

suffering from persistent biofilm infections.

II.5.3.1 Antimicrobial peptides

Over the last years natural AMPs, like colistin, have gained a lot of interest as promising

agents to treat biofilm infections [180]. Their bactericidal effect relies on their

capability of permeabilizing the outer membrane and/or forming pores in the

cytoplasmic membrane. The relatively low stability and a certain degree of non-specific

toxicity of AMPs have led to an increasing effort in the development of new AMPs and

peptidomimetics [181-183]. In general, AMPs have a broad antimicrobial spectrum,

there is a low incidence of resistance development and they remain active against slow

or even non-growing cells. As a large part of biofilm cells present with low metabolic

activity, AMPs could be a promising approach for eradicating biofilm-related infections,

either alone or in combination with antimicrobials active against fast-growing cells.

II.5.3.2 Fosmidomycin and fosmidomycin derivatives

Biofilm growth is associated with increased levels of mutations often resulting in

increased resistance and contributing to the survival of biofilms. This increases the

need for new treatment strategies even further. The non-mevalonate pathway is a

promising new target for antibiotic development [184]. Fosmidomycin and FR900098

are two well-known inhibitors of Dxr, the key enzyme in this non-mevalonate pathway

which is essential for isoprenoid synthesis. Apart from its traditional use against

Plasmodium falciparum to treat malaria, fosmidomycin is also active against some

bacteria, including E. coli and P. aeruginosa [185]. The major cause of fosmidomycin

resistance for treating bacterial infections is poor cell penetration [186]. Therefore, the

focus of several studies lies on the development of fosmidomycin derivatives that could

overcome this problem of poor uptake [187].

38

II.5.3.3 Quorum sensing inhibitors

The contribution of QS to the development of biofilm tolerance towards several

antimicrobials has been the subject of some debate. Several researchers noticed an

increased susceptibility of QS-deficient mutants of P. aeruginosa towards several

antimicrobials, including tobramycin, hydrogen peroxide, ciprofloxacin and ceftazidime

[140, 141, 188]. However, the QS-regulated genes involved in these processes still have

to be identified. It has also been shown that QS-deficient mutants were less virulent and

cleared faster compared to the wild type, indicating the involvement of QS in chronic P.

aeruginosa infections [189, 190]. The contribution of QS during chronic P. aeruginosa

lung infections was confirmed by the detection of QS signals in sputum of CF patients

[191, 192].

These observations led to the investigation of QS inhibitors as a treatment for various

biofilm related infections [193, 194]. Wu et. al. indicated that synthetic furanones, which

interfere with the P. aeruginosa QS system, attenuated the virulence of the pathogen

resulting in a reduced severity of the lung infection [195]. QS inhibitors to treat

Burkholderia biofilm infections have also been proposed as it was described that QS is

presumably involved in Burkholderia biofilm formation [196]. Several QS inhibitors

have already been shown to reduce biofilm formation of B. cenocepacia and B.

multivorans [197].

II.5.3.4 D-amino acids

In 2010, Kolodkin-Gal et. al. indicated that a mixture of D-amino acids, produced by

Bacillus subtilis, could disrupt existing biofilms as well as prevent biofilm establishment

of B. subtilis, P. aeruginosa and S. aureus [198]. Further research indicated that a

mixture of four different D-amino acids was responsible for this effect. D-leucine, D-

methionine, D-tryptophan and D-tyrosine were all active in preventing biofilm

formation at micromolar to millimolar concentrations, with D-tyrosine being the most

potent. In B. subtilis and S. aureus these amino acids replace D-alanine in the

peptidoglycan side chain where they dislodge anchored amyloid fibers [198, 199]. In

general these amyloid fibers connect cells present in a biofilm, explaining the dispersal

effect of these D-amino acids. Bacteria produce D-amino acids in stationary phase

planktonic cultures and mature biofilms, probably regulating biofilm disassembly when

39

nutrients become limiting. This implies that these amino-acids could complement

existing therapies for treatment or prevention of biofilm-associated infections.

II.5.3.5 Nitric oxide

The mucus layer lining CF airways represents an area with reduced oxygen

concentrations, stimulating robust P. aeruginosa biofilm formation [200]. As chronic

lung disease progresses, mucoid P. aeruginosa strains become the predominant

phenotype in the CF lung through the overproduction of alginate. P. aeruginosa

continues respiring under anaerobic conditions, using nitrate (NO3-) or nitrite (NO2-) as

terminal electron acceptors. Both NO3- and NO2- have been shown to be present at high

concentrations in CF ASL and sputum [201]. Nitrate reductase and nitrite reductase are

used to reduce NO3- to NO2- and NO2- to nitric oxide (NO), respectively (Figure 13).

Increased levels of NO have to be controlled through the production of NO reductase

(NOR) as overproduction of toxic NO would cause cell death [202]. As mucoid strains

present with very low NOR activity, increasing NO levels could be a possible novel

approach for treating mucoid P. aeruginosa biofilms. The increase of NO concentrations

through the addition of acidified nitrite (HNO2) derivatives under CF conditions,

resulted in a bactericidal effect on mucoid P. aeruginosa strains [203].

Figure 13: Anaerobic respiration in P. aeruginosa. NOR activity in mucoid strains is very low, resulting in

toxic NO concentrations when NO2- levels increase. (Modified from http://www.spring8.or.jp)

Moreover, general responses to low non-toxic levels of NO in P. aeruginosa include

upregulation of genes involved in motility and downregulation of virulence factors and

adhesins [204]. This implies that low levels of NO can increase the antibiotic sensitivity

of P. aeruginosa biofilms, suggesting a promising role for NO as adjunctive therapy for

treating CF chronic lung infections.

40

II.5.3.6 Iron depletion

Moreau-Marquis et. al. have observed that CF airway cells, expressing ∆F508 CFTR,

release more iron compared to airway cells from a healthy individual [205]. These

elevated iron concentrations have been shown to enhance P. aeruginosa biofilm

development in the CF lung. Reducing the availability of free iron through the addition

of iron chelators resulted in prevention of P. aeruginosa biofilm formation and/or

disruption of established biofilms. This effect was even more pronounced under

anaerobic conditions, which are frequently encountered in chronically infected CF lungs

[206]. In addition, clinical isolates seem to respond differently to various iron chelators,

indicating possible differences in effectiveness and mechanism of working between

various naturally and synthetic iron chelators. To date, two FDA approved iron

chelators, deferoxamine (DFO) and deferasirox (DSX), have been shown to potentiate

the bactericidal effect of tobramycin on P. aeruginosa biofilm formed on live CF airway

epithelial cells [207]. Moreover, this combination therapy prevented the formation of P.

aeruginosa biofilms under these conditions. Although the mechanism underlying this

improved bactericidal effect on P. aeruginosa biofilms is not yet fully understood, the

combination of tobramycin with DFO or DSX looks promising to treat CF P. aeruginosa

lung infections. As P. aeruginosa has an uptake mechanism for DFO-Fe, this could

explain why DFO is less effective compared to DSX in disrupting biofilms.

A second approach to disrupt the essential iron metabolism in bacteria is the use of the

transition metal gallium (also FDA approved) as many biologic systems are unable to

distinguish between iron and gallium [208]. Unlike Fe3+, Ga3+ can not be reduced, which

is an essential reaction in many iron-dependent bacterial processes [209]. Gallium is

taken up by P. aeruginosa cells and simultaneously reduces iron uptake. It has been

shown to inhibit P. aeruginosa growth, prevent biofilm formation and exert a

bactericidal effect on established biofilms. The activity of gallium is based in part on

repressing an Fe-responsive transcriptional regulator [208]. An additional advantage of

gallium is its activity against cells located in the center of biofilm clusters which are

usually the most resistant towards conventional antibiotics. Banin et. al. suggested a

novel approach for delivering gallium to P. aeruginosa cells [210]. They used a metallo-

complex, with DFO as siderophore (DFO-gallium), as P. aeruginosa is believed to have

two uptake systems for DFO [211]. This metallo-complex can function as a “Trojan

41

horse” improving the uptake of gallium into P. aeruginosa cells and resulting in

increased bacterial killing. DFO-gallium has been shown effective in killing P. aeruginosa

biofilm cells and this effect even became synergistic when DFO-gallium was combined

with gentamicin.

II.5.3.7 Improved drug delivery

In 1997 a tobramycin inhalation solution, called TOBI, has been FDA approved and is

now routinely used in CF patients with a P. aeruginosa lung infection [212]. Through

inhalation, the tobramycin concentration in the infected CF airways can easily be

increased compared to IV therapy resulting is less toxic effects of the aminoglycoside.

When taking TOBI in a 28 days on-off cycle a decreased need for IV anti-pseudomonal

therapy accompanied with a decreased rate of hospitalization has been demonstrated,

clearly improving patient comfort. This could be further improved by a dry powder

inhalation formulation which is at the moment being investigated for approval by the

FDA [213, 214]. The dry powder formulation can be administered by the use of a

handheld manually operated inhaler while TOBI has to be inhaled via a nebulizer and

compressor. Not only for the administration of tobramycin but also for other antibiotics

the use and development of inhalable formulations is investigated [215].

Inhalation therapy can be combined with nanoparticle engineering for the development

of inhalable antibiotic-loaded nanoparticles. Encapsulation of antibiotics in

nanoparticles prevents the interaction with sputum and biofilm components improving

antibiotic delivery near the biofilm cells. Antibiotics can be encapsulated in liposomes,

micelles or polymers that have to meet certain requirements. The carrier should be

smaller than 300 nm to avoid alveolar phagocytosis and should consist of biodegradable

and biocompatible components (Figure 14). The release of the antibiotic from its carrier

can be accomplished by an active process, e.g. an enzymatic reaction, as well as by

passive diffusion like is the case for Arikace [216]. Arikace is a liposomal encapsulated

amikacin for inhalation which is currently in clinical phase III trials. Amikacin has long

been recognized as an effective treatment for different gram-negative infections,

including P. aeruginosa infections. By encapsulation into liposomes the antibiotic can be

delivered closer to the bacteria improving both efficacy and safety of amikacin

treatment. By modification of the liposomes in which an antibiotic is encapsulated these

42

particles can even be targeted towards certain areas in the human body. Currently a lot

of research is focusing on such targeted and encapsulated drug design [217, 218].

Figure 14: The use of antibiotic-loaded nanoparticles to treat CF biofilm infections. The perfect carrier

must be able to penetrate deep into the mucus and biofilm matrix, deliver its encapsulated antibiotic near

the biofilm cells and evade phagocytosis by the innate immune system.

(http://www.ntu.edu.sg/home/kunnong/Research/default.html)

43

44

Burkholderia cepacia complex: opportunistic pathogens with important natural biology (Eshwar Mahenthiralingam)

45

Chapter III: BURKHOLDERIA CEPACIA COMPLEX

III.1 General introduction

The Bcc consists of at least 18 closely-related species belonging to the group of β-

proteobacteria [219]. Members of this complex are gram-negative, rod-shaped bacteria

that are widespread in natural and clinical environments where they exert both

beneficial and detrimental effects (Figure 15).

Burkholderia cepacia (formerly referred to as Pseudomonas cepacia) was first identified

by W. H. Burkholder in 1950 as the causative agent of soft onion rot [220]. Nevertheless,

several members of the Bcc live as plant commensals in close proximity to the

rhizosphere protecting them from plant disease by the synthesis of antimicrobial

metabolites and siderophores. The production of siderophores is implicated in

biocontrol through competition for iron, an essential component for bacterial growth.

Through the production of plant hormones, Bcc strains can even promote the growth of

certain crops like maize, rice and wheat. Another great advantage of environmental Bcc

strains is their unique metabolic capacity that allows them to be used as a

bioremediation agent. Bcc can degrade a wide variety of man-made pollutants like

phthalates, herbicides and chlorinated hydrocarbons [221].

In marked contrast with their ecologically beneficial effect, Bcc bacteria emerged as

highly problematic opportunistic human pathogens during the past 30 years in both CF

patients and immunocompromised patients [63, 222]. As Bcc bacteria are not carried as

commensals, infections are acquired either in hospital settings (from other patients or

from contaminated devices) or from the environment. The first reports of Bcc infection

in CF patients date from the late 1970s [223] and reported nosocomial outbreaks were

often due to contaminated devices, disinfectants and antiseptics [224]. The first

isolation of Bcc strains from the sputum of CF patients was described by Ederer in 1972

[225].

46

Figure 15: Role of Bcc bacteria in environmental and clinical environments. Most environmental effects

are ecologically beneficial while in clinical environments Bcc bacteria have emerged as highly problematic

opportunistic pathogens in CF and immunocompromised patients. [226]

III.2 Taxonomy

After extensive research on DNA-DNA homology, 16S rRNA sequences, cellular lipid and

fatty acid compositions, seven Pseudomonas species were in 1992 transferred to a new

genus Burkholderia [227]. In 1997, the group of Peter Vandamme investigated 89

isolates primarily isolated form CF patients and identified as B. cepacia. A marked

heterogeneity among these strains has led to the subdivision of “B. cepacia” strains in at

least five distinct genomic species, referred to as genomovars [228]. This group of five

genomovars, representing phenotypically similar but genotypically distinct species, was

collectively referred to as the Bcc. Over the years more diversity was revealed and more

genomovars were described. Currently, at least 18 species belonging to the Bcc are

described and each has received a formal binomial designation (Table 2) [229].

47

Table 2: Overview of the Bcc [229]

Species (former genomovar designation)

Natural environment Clinical environment

B. cepacia (I) Rhizosphere, soil, water CF, medical solutions B. multivorans (II) Rhizosphere, soil, water CF, CGD, non-CF B. cenocepacia (III) Rhizosphere, soil, water, plant, animal CF, non-CF B. stabilis (IV) Rhizospere CF, hospital equipment B. vietnamiensis (V) Rhizosphere, soil, water, plant, animal CF B. dolosa (VI) Maize rhizosphere, plant CF B. ambifaria (VII) Rhizosphere, soil CF B. anthina (VIII) Rhizosphere, soil CF B. pyrrocinia (IX) Rhizosphere, soil, water, plant CF, non-CF B. ubonensis Soil Nosocomial infection B. latens No environmental strain reported CF B. diffusa Soil, water CF, hospital equipment, non-CF B. arboris Rhizosphere, soil, water CF, non-CF B. seminalis Rice rhizosphere CF, nosocomial infection B. metallica No environmental strain reported CF B. contaminans Soil, water, animal CF, non-CF, hospital equipment B. lata B. pseudomultivorans

Soil, water, flower Rhizosphere

CF, non-CF CF, non-CF

CF: CF CGD: chronic granulomatous disease

III.3 Human pathogenesis

Bcc strains are encountered as opportunistic pathogens, which means that they do not

usually infect healthy persons but only those that are immunocompromised. Lungs of

CF patients are particularly susceptible to infections with Bcc strains. Isles et. al. were

the first to describe Bcc strains as emerging problematic pathogens in CF patients

during a Bcc epidemic in Toronto (Canada) in the early 1980s [63]. They observed an

accelerated and uncontrollable decline in overall clinical status in about 10% of infected

CF patients.

Patients suffering from chronic granulomatous disease (CGD) also have an increased

risk of developing Bcc infections. In patients with CGD a genetic defect results in the

inability of polymorphonuclear leukocytes to produce ROS [230]. Therefore, these

patients are not capable of clearing Bcc strains from their lungs resulting in sepsis

(invasive Bcc infections) and pneumonia, the second leading cause of death in CGD

patients [226].

In addition, Bcc strains have been isolated from immunocompetent individuals, like

patients with chronic otitis media and pharyngeal infections [231-234]. Because of their

tolerance to various disinfectants and antiseptics, Bcc bacteria can persist in hospital

environments, resulting in an increased number of cases of Bcc-related septicaemia;

48

even in non-CF patients. Among these hospitalized patients, hemodialysis, permanence

in intensive care units and use of central venous catheters, urinary catheters and

endotracheal tubes are recognized as risk factors contributing to Bcc acquisition [235].

During the last 10 years B. cenocepacia is also increasingly being recognized as an

emergent nosocomial pathogen among immunocompromised oncology patients [236].

III.3.1 Bcc in CF infection

The number of Bcc infections in CF patients increased dramatically during the 1980s.

Compared to P. aeruginosa, Bcc strains infect a relatively small fraction of CF patients.

Nevertheless, pulmonary colonization with Bcc strains is often correlated with a worse

prognosis, longer hospital stays and an increased risk of death [23, 237]. A great

variability in clinical outcome has been noted among Bcc infected CF patients, even

among patients infected with the same clonal strain. Some patients are asymptomatic

carriers while others experience a rapid decline in clinical status associated with

necrotizing pneumonia and septicemia (cepacia syndrome) [238]. Both strain and host

factors are thought to be responsible for these differences in disease pathogenesis. All

Bcc strains, except for B. ubonensis, have been isolated from the lungs of CF patients.

Still, the most prevalent isolated pathogens are B. cenocepacia and B. multivorans [239].

Strains belonging to several species of the Bcc complex like B. cenocepacia, B. cepacia, B.

multivorans and Burkholderia dolosa, have been shown to be highly transmissible among

CF patients with B. cenocepacia being the most notorious among them [60, 239]. B.

cenocepacia has been shown to replace B. multivorans during chronic CF infection [240]

and epidemic strains with international impact, like B. cenocepacia ET12, have been

described [61]. During the late 1980s and throughout much of the 1990s this highly

epidemic lineage spread among CF patients from Canada to the United Kingdom

resulting in high mortality rates mainly because of its ability to cause cepacia syndrome.

Although transmission of Bcc strains between siblings is most frequently reported,

transmission between hospitalized CF patients [60] and even between CF patients and

non-CF patients [241] has been noted. As transmission occurs by direct contact or by

inhalation of aerosol droplets, several infection control guidelines recommend to

segregate people with Bcc infections from each other and from non-Bcc infected CF

patients in clinical settings [242, 243]. They also advise to minimalize close contact

outside the hospital environment. Although such segregation policies have proven to be

49

efficient not everyone agrees with these guidelines because of the great impact on social

and psychological well-being of CF patients [244].

III.4 Genomics

The first genetic map of a Bcc strain, later identified as a B. multivorans strain, was

published in 1994 by Cheng and Lessie [245]. Apart from the overall large genome size

(∼ 7 Mb), it demonstrated a rather unusual genomic organisation. The genome of this B.

multivorans strain (ATCC 17616) was made up of three large replicons (3.4 Mb – 2.5 Mb

– 0.9 Mb) and a plasmid (170 kb). The large replicons were referred to as chromosomes

as they all encoded ribosomal RNA operons. One year later, Rodley et al. confirmed this

genomic arrangement for another Bcc strain, the B. cepacia type strain [246]. During the

following years different other Bcc genomes were sequenced and until now they all

possess the same genomic structure of three chromosomes in combination with one to

five plasmids [247]. In general, chromosomes are never lost although they could

undergo several rearrangements and all chromosomes encode essential genes, with

most housekeeping genes located on the largest chromosome [248].

The genome of B. cenocepacia J2315, a reference strain of the ET12 epidemic lineage,

was published in 2009 and gave more insight into the success of this pathogen in

colonizing and persisting in the CF lung [249]. This genome consists of three large

circular chromosomes and one small plasmid and has a total size of 8.056 Mb. The

presence of several types of insertion sequences, small transposable mobile DNA

fragments, is possibly involved in genomic plasticity and regulation of gene expression.

The genome also contains a large amount of genomic islands which are regions of

genomic DNA with unusual characteristics compared to the most of the genomic DNA

sequence, indicating that they may have been acquired during evolution. Genomic

islands are present in the genome of many gram-negative bacteria although normally to

a lesser extent [250]. Approximately 10 % of the total genome of B. cenocepacia J2315 is

comprised of genomic islands, occurring on all three chromosomal replicons [249]. The

B. cenocepacia pathogenicity island (cci) was the first to be defined in the entire Bcc and

its contribution to B. cenocepacia virulence has been described [251]. Interest in this

region was prompted because of the presence of a “B. cenocepacia epidemic strain

marker” which is associated with virulent B. cenocepacia strains capable of infecting CF

patients and transmission between patients. Apart from this marker, cci also contains

50

an amidase gene playing a role in persistence and two signaling genes and a porin gene

also involved in virulence. In B. cenocepacia J2315, 14 genomic islands have been

identified [249] contributing to the pathogens genomic plasticity and metabolic

diversity.

III.5 Virulence factors

Bcc strains are intrinsically resistant towards a broad range of antibiotics and are

perfectly adapted to colonize and persist in the CF lung. Putative virulence factors and

pathogenic mechanisms contributing to these traits are extensively being investigated.

III.5.1 Lipopolysaccharides

LPS are a general virulence factor in gram-negative bacteria and are located on the outer

leaflet of the outer membrane. LPS are composed of a lipid A moiety, a core

oligosaccharide and an O-antigen (Figure 16).

Figure 16: LPS from the outer membrane of gram-negative bacteria. The two glucosamine molecules

(oblongs) of the lipid A moiety contain phosphodiester-linked 4-amino-4-deoxyarabinoses (brown circles)

in Burkholderia. The inner core which is attached to lipid A is comprised of less 3-deoxy-D-manno-oct-2-

ulosonic acids (KDO; dark blue hexagon) compared to LPS molecules of other gram-negative organisms.

[226]

51

The LPS of Burkholderia species can be distinguished from LPS of other gram-negative

bacteria as it contains less phosphate and/or KDO residues on their core

oligosaccharide, making it less negatively charged. In addition, 4-amino-4-

deoxyarabinose molecules are bound to the phosphate residues on the lipid A moiety

also reducing the negative charge of this LPS component [252]. The addition of 4-

amino-4-deoxyarabinose molecules to lipid A phosphate residues is a well described

mechanism used by gram-negative bacteria, like P. aeruginosa, to confer resistance

against positively charged antimicrobials [253]. The uptake of such positively charged

antimicrobials, e.g. aminoglycosides, polymyxin and AMPs, is initiated by the binding to

lipid A phosphate. By binding to these molecules, divalent cations cross-bridging two

adjacent LPS molecules are displaced, destabilizing the outer membrane and improving

the antimicrobial’s uptake.

LPS from Bcc species is four to five times more endotoxic than P. aeruginosa LPS [254].

Therefore, most B. cenocepacia strains consistently induce cytokine production,

including tumor necrosis factor (TNF) and IL-8, leading to inflammation reactions. This

endotoxic effect is considered to be attributed to the lipid A moiety of LPS which is

biologically the most active component.

III.5.2 Biofilm formation

Bcc strains are believed to persistently colonize the CF lung as a biofilm. In vitro biofilm

formation of all Bcc strains and even the formation of mixed biofilms with P. aeruginosa

strains has been confirmed [255]. Biofilm formation is a complex process involving

many different determinants. The first requisite for biofilm formation is the initial

attachment to a surface which is influenced by the availability of iron and the presence

of cable pili and adhesins on the surface of Bcc. The overload of free iron in CF sputum

can induce both P. aeruginosa and B. cenocepacia non-motile growth forms leading to

aggregate formation and biofilm development [256]. Interaction between these

aggregated bacteria through the production of acyl-homoserine lactone signaling

molecules has been shown to be important in Bcc biofilm formation although not under

all growth conditions [196]. The production of cepacian by mucoid Bcc strains partly

contributes to the formation of thick, mature biofilms [171]. EPS production in Bcc also

inhibits neutrophil chemotaxis and allows to scavenge ROS protecting Bcc biofilm

bacteria from innate neutrophil-mediated host defenses [257]. In general, biofilm

52

formation is associated with increased resistance towards different antibiotics and

components of the innate human defense system contributing to the persistence of Bcc

infections in CF patients.

III.5.3 Quorum sensing

QS implies a cell-to-cell communication system that is cell density dependent and allows

bacteria to behave as a multicellular entity. The first QS system detected in Bcc was in a

B. cencocepacia K56-2 strain and was designated as the cepRI quorum-sensing system

[258]. Similar QS systems have currently been detected in all Bcc species through which

bacteria produce and respond to N-acyl-homoserine lacton molecules. The cepRI system

is involved in the regulation of many different Bcc virulence phenotypes including

biofilm formation, motility, toxin production and protease and lipase production [259].

In addition, it is also suggested to be important in communicating with P. aeruginosa in

mixed species biofilms. Certain strains of B. cenocepacia possess a second acyl-

homoserine lacton producing cell-to-cell communication system. This system is

designated as cciRI as is encoded on the B. cenocepacia pathogenicity island [251]. The

presence of this second QS system even further emphasizes the complexity of bacterial

cell communication.

III.5.4 Innate resistance to antibiotics

Bcc organisms are intrinsically resistant to a broad range of antibiotics. This is partly

due to the special structure of their LPS molecules which make the outer membrane

impermeable for certain antimicrobials. Antibiotics that do manage to penetrate the

bacterial membrane can be excluded again by the activity of different efflux pumps

[226]. Resistance towards β-lactam antibiotics can be explained by the induction of

chromosomally encoded β-lactames [260]. These enzymes can reside in the biofilm

matrix or in other in vivo environments like sputum, hydrolyzing the β-lactams while

approaching the bacteria.

53

References

1. O'Sullivan, B.P. and S.D. Freedman, CF. Lancet, 2009. 373(9678): p. 1891-904. 2. Gadsby, D.C., P. Vergani, and L. Csanady, The ABC protein turned chloride channel

whose failure causes CF. Nature, 2006. 440(7083): p. 477-83. 3. Du, K., M. Sharma, and G.L. Lukacs, The DeltaF508 CF mutation impairs domain-

domain interactions and arrests post-translational folding of CFTR. Nat Struct Mol Biol, 2005. 12(1): p. 17-25.

4. Koch, C., et al., European Epidemiologic Registry of CF (ERCF): comparison of major disease manifestations between patients with different classes of mutations. Pediatr Pulmonol, 2001. 31(1): p. 1-12.

5. Castellani, C., et al., European best practice guidelines for CF neonatal screening. J Cyst Fibros, 2009. 8(3): p. 153-73.

6. Farrell, P.M., et al., Guidelines for diagnosis of CF in newborns through older adults: CF Foundation consensus report. J Pediatr, 2008. 153(2): p. S4-S14.

7. Farrell, P.M., et al., Evidence on improved outcomes with early diagnosis of CF through neonatal screening: enough is enough! J Pediatr, 2005. 147(3 Suppl): p. S30-6.

8. Brennan, A.L. and D.M. Geddes, CF. Curr Opin Infect Dis, 2002. 15(2): p. 175-82. 9. Mishra, A., R. Greaves, and J. Massie, The Relevance of Sweat Testing for the

Diagnosis of CF in the Genomic Era. Clin Biochem Rev, 2005. 26(4): p. 135-53. 10. van der Doef, H.P., et al., Intestinal obstruction syndromes in CF: meconium ileus,

distal intestinal obstruction syndrome, and constipation. Curr Gastroenterol Rep, 2011. 13(3): p. 265-70.

11. Blackman, S.M., et al., Relative contribution of genetic and nongenetic modifiers to intestinal obstruction in CF. Gastroenterology, 2006. 131(4): p. 1030-9.

12. Moyer, K. and W. Balistreri, Hepatobiliary disease in patients with CF. Curr Opin Gastroenterol, 2009. 25(3): p. 272-8.

13. Herrmann, U., G. Dockter, and F. Lammert, CF-associated liver disease. Best Pract Res Clin Gastroenterol, 2010. 24(5): p. 585-92.

14. Westaby, D., cystic fibrosi: liver disease, in CF in the 21st Century, A. Bush, et al., Editors. 2006, Karger: Basel. p. 251-261.

15. Colombo, C., Liver disease in CF. Curr Opin Pulm Med, 2007. 13(6): p. 529-36. 16. Boyd, J.M., A. Mehta, and D.J. Murphy, Fertility and pregnancy outcomes in men

and women with CF in the United Kingdom. Hum Reprod, 2004. 19(10): p. 2238-43.

17. Ahmad, A., A. Ahmed, and P. Patrizio, CF and fertility. Curr Opin Obstet Gynecol, 2013.

18. Sueblinvong, V. and L.A. Whittaker, Fertility and pregnancy: common concerns of the aging CF population. Clin Chest Med, 2007. 28(2): p. 433-43.

19. Taylor, C.J. and N. Aswani, The pancreas in CF. Paediatr Respir Rev, 2002. 3(1): p. 77-81.

20. Naruse, S., M. Kitagawa, and H. Ishiguro, Molecular understanding of chronic pancreatitis: a perspective on the future. Mol Med Today, 1999. 5(11): p. 493-9.

21. Davis, P.B., M. Drumm, and M.W. Konstan, CF. Am J Respir Crit Care Med, 1996. 154(5): p. 1229-56.

54

22. Walkowiak, J., A. Lisowska, and M. Blaszczynski, The changing face of the exocrine pancreas in CF: pancreatic sufficiency, pancreatitis and genotype. Eur J Gastroenterol Hepatol, 2008. 20(3): p. 157-60.

23. Lyczak, J.B., C.L. Cannon, and G.B. Pier, Lung Infections Associated with CF. Clin Microbiol Rev, 2002. 15(2): p. 194-222.

24. Gibson, R.L., J.L. Burns, and B.W. Ramsey, Pathophysiology and management of pulmonary infections in CF. Am J Respir Crit Care Med, 2003. 168(8): p. 918-51.

25. Krouse, M.E., Is CF lung disease caused by abnormal ion composition or abnormal volume? J Gen Physiol, 2001. 118(2): p. 219-22.

26. Tarran, R., et al., The CF salt controversy: in vivo observations and therapeutic approaches. Mol Cell, 2001. 8(1): p. 149-58.

27. Smith, J.J., et al., CF airway epithelia fail to kill bacteria because of abnormal airway surface fluid. Cell, 1996. 85(2): p. 229-36.

28. Imundo, L., et al., CF epithelial cells have a receptor for pathogenic bacteria on their apical surface. Proc Natl Acad Sci U S A, 1995. 92(7): p. 3019-23.

29. de Bentzmann, S., et al., Asialo GM1 is a receptor for Pseudomonas aeruginosa adherence to regenerating respiratory epithelial cells. Infect Immun, 1996. 64(5): p. 1582-8.

30. Bryan, R., et al., Overproduction of the CFTR R domain leads to increased levels of asialoGM1 and increased Pseudomonas aeruginosa binding by epithelial cells. Am J Respir Cell Mol Biol, 1998. 19(2): p. 269-77.

31. Schroeder, T.H., T. Zaidi, and G.B. Pier, Lack of adherence of clinical isolates of Pseudomonas aeruginosa to asialo-GM(1) on epithelial cells. Infect Immun, 2001. 69(2): p. 719-29.

32. Pier, G.B., M. Grout, and T.S. Zaidi, CF transmembrane conductance regulator is an epithelial cell receptor for clearance of Pseudomonas aeruginosa from the lung. Proc Natl Acad Sci U S A, 1997. 94(22): p. 12088-93.

33. Pier, G.B., CFTR mutations and host susceptibility to Pseudomonas aeruginosa lung infection. Curr Opin Microbiol, 2002. 5(1): p. 81-6.

34. Schroeder, T.H., et al., Transgenic CF mice exhibit reduced early clearance of Pseudomonas aeruginosa from the respiratory tract. J Immunol, 2001. 166(12): p. 7410-8.

35. Armstrong, D.S., et al., Lower airway inflammation in infants and young children with CF. Am J Respir Crit Care Med, 1997. 156(4 Pt 1): p. 1197-204.

36. Goss, C.H. and M.S. Muhlebach, Review: Staphylococcus aureus and MRSA in CF. J Cyst Fibros, 2011. 10(5): p. 298-306.

37. Nolan, G., et al., Antibiotic prophylaxis in CF: inhaled cephaloridine as an adjunct to oral cloxacillin. J Pediatr, 1982. 101(4): p. 626-30.

38. Ratjen, F., et al., Effect of continuous antistaphylococcal therapy on the rate of P. aeruginosa acquisition in patients with CF. Pediatr Pulmonol, 2001. 31(1): p. 13-6.

39. Stutman, H.R., et al., Antibiotic prophylaxis in infants and young children with CF: a randomized controlled trial. J Pediatr, 2002. 140(3): p. 299-305.

40. CF, F., Annual data report to the center directors. 2011, CF Foundation patient registry: Bethesda, Maryland.

41. Scott, F.W. and T.L. Pitt, Identification and characterization of transmissible Pseudomonas aeruginosa strains in CF patients in England and Wales. J Med Microbiol, 2004. 53(Pt 7): p. 609-15.

55

42. Grothues, D., et al., Genome fingerprinting of Pseudomonas aeruginosa indicates colonization of CF siblings with closely related strains. J Clin Microbiol, 1988. 26(10): p. 1973-7.

43. Wolz, C., et al., Pseudomonas aeruginosa cross-colonization and persistence in patients with CF. Use of a DNA probe. Epidemiol Infect, 1989. 102(2): p. 205-14.

44. O'Carroll, M.R., et al., Clonal strains of Pseudomonas aeruginosa in paediatric and adult CF units. Eur Respir J, 2004. 24(1): p. 101-6.

45. Speert, D.P., et al., Epidemiology of Pseudomonas aeruginosa in CF in British Columbia, Canada. Am J Respir Crit Care Med, 2002. 166(7): p. 988-93.

46. Burns, J.L., et al., Longitudinal assessment of Pseudomonas aeruginosa in young children with CF. J Infect Dis, 2001. 183(3): p. 444-52.

47. Rosenfeld, M., et al., Early pulmonary infection, inflammation, and clinical outcomes in infants with CF. Pediatr Pulmonol, 2001. 32(5): p. 356-66.

48. Emerson, J., et al., Pseudomonas aeruginosa and other predictors of mortality and morbidity in young children with CF. Pediatr Pulmonol, 2002. 34(2): p. 91-100.

49. Nixon, G.M., et al., Clinical outcome after early Pseudomonas aeruginosa infection in CF. J Pediatr, 2001. 138(5): p. 699-704.

50. Valerius, N.H., C. Koch, and N. Hoiby, Prevention of chronic Pseudomonas aeruginosa colonisation in CF by early treatment. Lancet, 1991. 338(8769): p. 725-6.

51. Moreau-Marquis, S., B.A. Stanton, and G.A. O'Toole, Pseudomonas aeruginosa biofilm formation in the CF airway. Pulm Pharmacol Ther, 2008. 21(4): p. 595-9.

52. Govan, J.R. and V. Deretic, Microbial pathogenesis in CF: mucoid Pseudomonas aeruginosa and Burkholderia cepacia. Microbiol Rev, 1996. 60(3): p. 539-74.

53. Li, Z., et al., Longitudinal development of mucoid Pseudomonas aeruginosa infection and lung disease progression in children with CF. JAMA, 2005. 293(5): p. 581-8.

54. LiPuma, J.J., et al., Disproportionate distribution of Burkholderia cepacia complex species and transmissibility markers in CF. Am J Respir Crit Care Med, 2001. 164(1): p. 92-6.

55. Govan, J.R., A.R. Brown, and A.M. Jones, Evolving epidemiology of Pseudomonas aeruginosa and the Burkholderia cepacia complex in CF lung infection. Future Microbiol, 2007. 2(2): p. 153-64.

56. Reik, R., T. Spilker, and J.J. Lipuma, Distribution of Burkholderia cepacia complex species among isolates recovered from persons with or without CF. J Clin Microbiol, 2005. 43(6): p. 2926-8.

57. Lipuma, J.J., The changing microbial epidemiology in CF. Clin Microbiol Rev, 2010. 23(2): p. 299-323.

58. LiPuma, J.J., et al., Ribotype analysis of Pseudomonas cepacia from CF treatment centers. J Pediatr, 1988. 113(5): p. 859-62.

59. Govan, J.R., et al., Evidence for transmission of Pseudomonas cepacia by social contact in CF. Lancet, 1993. 342(8862): p. 15-9.

60. LiPuma, J.J., et al., Person-to-person transmission of Pseudomonas cepacia between patients with CF. Lancet, 1990. 336(8723): p. 1094-6.

61. Johnson, W.M., S.D. Tyler, and K.R. Rozee, Linkage analysis of geographic and clinical clusters in Pseudomonas cepacia infections by multilocus enzyme electrophoresis and ribotyping. J Clin Microbiol, 1994. 32(4): p. 924-30.

62. Thomassen, M.J., et al., Pseudomonas cepacia colonization among patients with CF. A new opportunist. Am Rev Respir Dis, 1985. 131(5): p. 791-6.

56

63. Isles, A., et al., Pseudomonas cepacia infection in CF: an emerging problem. J Pediatr, 1984. 104(2): p. 206-10.

64. Hauben, L., et al., Genomic diversity of the genus Stenotrophomonas. Int J Syst Bacteriol, 1999. 49 Pt 4: p. 1749-60.

65. Denton, M., et al., Stenotrophomonas maltophilia contamination of nebulizers used to deliver aerosolized therapy to inpatients with CF. J Hosp Infect, 2003. 55(3): p. 180-3.

66. Demko, C.A., R.C. Stern, and C.F. Doershuk, Stenotrophomonas maltophilia in CF: incidence and prevalence. Pediatr Pulmonol, 1998. 25(5): p. 304-8.

67. Vu-Thien, H., et al., Investigation of an outbreak of wound infections due to Alcaligenes xylosoxidans transmitted by chlorhexidine in a burns unit. Eur J Clin Microbiol Infect Dis, 1998. 17(10): p. 724-6.

68. Liu, L., et al., Ribosomal DNA-directed PCR for identification of Achromobacter (Alcaligenes) xylosoxidans recovered from sputum samples from CF patients. J Clin Microbiol, 2002. 40(4): p. 1210-3.

69. Peltroche-Llacshuanga, H., H. Kentrup, and G. Haase, Persistent airway colonization with Alcaligenes xylosoxidans in two brothers with CF. Eur J Clin Microbiol Infect Dis, 1998. 17(2): p. 132-4.

70. Ronne Hansen, C., et al., Chronic infection with Achromobacter xylosoxidans in CF patients; a retrospective case control study. J Cyst Fibros, 2006. 5(4): p. 245-51.

71. Ramsey, D.M. and D.J. Wozniak, Understanding the control of Pseudomonas aeruginosa alginate synthesis and the prospects for management of chronic infections in CF. Mol Microbiol, 2005. 56(2): p. 309-22.

72. Flume, P.A., et al., CF pulmonary guidelines: airway clearance therapies. Respir Care, 2009. 54(4): p. 522-37.

73. Pisi, G. and A. Chetta, Airway clearance therapy in CF patients. Acta Biomed, 2009. 80(2): p. 102-6.

74. Hansen, L.G. and W.J. Warwick, High-frequency chest compression system to aid in clearance of mucus from the lung. Biomed Instrum Technol, 1990. 24(4): p. 289-94.

75. Sutton, P.P., et al., Assessment of the forced expiration technique, postural drainage and directed coughing in chest physiotherapy. Eur J Respir Dis, 1983. 64(1): p. 62-8.

76. Schoni, M.H., Autogenic drainage: a modern approach to physiotherapy in CF. J R Soc Med, 1989. 82 Suppl 16: p. 32-7.

77. Andersen, J.B., J. Qvist, and T. Kann, Recruiting collapsed lung through collateral channels with positive end-expiratory pressure. Scand J Respir Dis, 1979. 60(5): p. 260-6.

78. Pettit, R.S. and C.E. Johnson, Airway-rehydrating agents for the treatment of CF: past, present, and future. Ann Pharmacother, 2011. 45(1): p. 49-59.

79. Reeves, E.P., et al., Hypertonic saline in treatment of pulmonary disease in CF. ScientificWorldJournal, 2012. 2012: p. 465230.

80. Reeves, E.P., et al., Nebulized hypertonic saline decreases IL-8 in sputum of patients with CF. Am J Respir Crit Care Med, 2011. 183(11): p. 1517-23.

81. Bergsson, G., et al., LL-37 complexation with glycosaminoglycans in CF lungs inhibits antimicrobial activity, which can be restored by hypertonic saline. J Immunol, 2009. 183(1): p. 543-51.

82. Anon, Summary of product characteristics: Pulmozyme. 1996, Roche Products Limited: Hertfordshire.

57

83. Christopher, F., et al., rhDNase therapy for the treatment of CF patients with mild to moderate lung disease. J Clin Pharm Ther, 1999. 24(6): p. 415-26.

84. Fuchs, H.J., et al., Effect of aerosolized recombinant human DNase on exacerbations of respiratory symptoms and on pulmonary function in patients with CF. The Pulmozyme Study Group. N Engl J Med, 1994. 331(10): p. 637-42.

85. Khan, T.Z., et al., Early pulmonary inflammation in infants with CF. Am J Respir Crit Care Med, 1995. 151(4): p. 1075-82.

86. Balough, K., et al., The relationship between infection and inflammation in the early stages of lung disease from CF. Pediatr Pulmonol, 1995. 20(2): p. 63-70.

87. Prescott, W.A., Jr. and C.E. Johnson, Antiinflammatory therapies for CF: past, present, and future. Pharmacotherapy, 2005. 25(4): p. 555-73.

88. Southern, K.W. and P.M. Barker, Azithromycin for CF. Eur Respir J, 2004. 24(5): p. 834-8.

89. Garcia Rodriguez, L.A. and H. Jick, Risk of upper gastrointestinal bleeding and perforation associated with individual non-steroidal anti-inflammatory drugs. Lancet, 1994. 343(8900): p. 769-72.

90. Eigen, H., et al., A multicenter study of alternate-day prednisone therapy in patients with CF. CF Foundation Prednisone Trial Group. J Pediatr, 1995. 126(4): p. 515-23.

91. Doring, G., et al., Antibiotic therapy against Pseudomonas aeruginosa in CF: a European consensus. Eur Respir J, 2000. 16(4): p. 749-67.

92. Taccetti, G., et al., Early antibiotic treatment for Pseudomonas aeruginosa eradication in patients with CF: a randomised multicentre study comparing two different protocols. Thorax, 2012. 67(10): p. 853-9.

93. Taccetti, G., et al., Early eradication therapy against Pseudomonas aeruginosa in CF patients. Eur Respir J, 2005. 26(3): p. 458-61.

94. Ramsey, B.W., et al., Intermittent administration of inhaled tobramycin in patients with CF. CF Inhaled Tobramycin Study Group. N Engl J Med, 1999. 340(1): p. 23-30.

95. Wiesemann, H.G., et al., Placebo-controlled, double-blind, randomized study of aerosolized tobramycin for early treatment of Pseudomonas aeruginosa colonization in CF. Pediatr Pulmonol, 1998. 25(2): p. 88-92.

96. Blumberg, H.M., et al., Rapid development of ciprofloxacin resistance in methicillin-susceptible and -resistant Staphylococcus aureus. J Infect Dis, 1991. 163(6): p. 1279-85.

97. Radberg, G., L.E. Nilsson, and S. Svensson, Development of quinolone-imipenem cross resistance in Pseudomonas aeruginosa during exposure to ciprofloxacin. Antimicrob Agents Chemother, 1990. 34(11): p. 2142-7.

98. Ramsey, B.W., Management of pulmonary disease in patients with CF. N Engl J Med, 1996. 335(3): p. 179-88.

99. Assael, B.M., et al., Inhaled aztreonam lysine vs. inhaled tobramycin in CF: A comparative efficacy trial. J Cyst Fibros, 2012.

100. Azoicai, D. and S.A. Antoniu, MP-376 (Aeroquin) for chronic Pseudomonas aeruginosa infections. Expert Opin Investig Drugs, 2013. 22(2): p. 267-76.

101. Ratjen, F., et al., Treatment of early Pseudomonas aeruginosa infection in patients with CF: the ELITE trial. Thorax, 2010. 65(4): p. 286-91.

102. Rosenfeld, M., et al., Diagnostic accuracy of oropharyngeal cultures in infants and young children with CF. Pediatr Pulmonol, 1999. 28(5): p. 321-8.

103. Turpin, S.V. and M.R. Knowles, Treatment of pulmonary disease in patients with CF, in CF, P.B. Davis, Editor. 1993, Marcel Dekker: New York. p. 277-344.

58

104. Henrici, A.T., Studies of Freshwater Bacteria: I. A Direct Microscopic Technique. J Bacteriol, 1933. 25(3): p. 277-87.

105. Zobell, C.E. and E.C. Allen, The Significance of Marine Bacteria in the Fouling of Submerged Surfaces. J Bacteriol, 1935. 29(3): p. 239-51.

106. Costerton, J.W., G.G. Geesey, and K.J. Cheng, How bacteria stick. Sci Am, 1978. 238(1): p. 86-95.

107. Costerton, J.W., P.S. Stewart, and E.P. Greenberg, Bacterial biofilms: a common cause of persistent infections. Science, 1999. 284(5418): p. 1318-22.

108. Sauer, K., et al., Pseudomonas aeruginosa displays multiple phenotypes during development as a biofilm. J Bacteriol, 2002. 184(4): p. 1140-54.

109. Lawrence, J.R., et al., Optical sectioning of microbial biofilms. J Bacteriol, 1991. 173(20): p. 6558-67.

110. Davies, D., Understanding biofilm resistance to antibacterial agents. Nat Rev Drug Discov, 2003. 2(2): p. 114-22.

111. O'Toole, G.A. and R. Kolter, Flagellar and twitching motility are necessary for Pseudomonas aeruginosa biofilm development. Mol Microbiol, 1998. 30(2): p. 295-304.

112. Danese, P.N., L.A. Pratt, and R. Kolter, Exopolysaccharide production is required for development of Escherichia coli K-12 biofilm architecture. J Bacteriol, 2000. 182(12): p. 3593-6.

113. Hentzer, M., et al., Alginate overproduction affects Pseudomonas aeruginosa biofilm structure and function. J Bacteriol, 2001. 183(18): p. 5395-401.

114. Watnick, P.I., et al., The absence of a flagellum leads to altered colony morphology, biofilm development and virulence in Vibrio cholerae O139. Mol Microbiol, 2001. 39(2): p. 223-35.

115. Flemming, H.C., J. Wingender, and C. Mayer, Physico-chemical properties of biofilms, in Biofilms: recent advances in their study and control, L.V. Evans, Editor. 2000, Harwood academic publishers: Amsterdam. p. 19-34.

116. Nickel, J.C., et al., Tobramycin resistance of Pseudomonas aeruginosa cells growing as a biofilm on urinary catheter material. Antimicrob Agents Chemother, 1985. 27(4): p. 619-24.

117. Gristina, A.G., et al., Adhesive colonization of biomaterials and antibiotic resistance. Biomaterials, 1987. 8(6): p. 423-6.

118. Mah, T.F. and G.A. O'Toole, Mechanisms of biofilm resistance to antimicrobial agents. Trends Microbiol, 2001. 9(1): p. 34-9.

119. Stewart, P.S., Theoretical aspects of antibiotic diffusion into microbial biofilms. Antimicrob Agents Chemother, 1996. 40(11): p. 2517-22.

120. Suci, P.A., et al., Investigation of ciprofloxacin penetration into Pseudomonas aeruginosa biofilms. Antimicrob Agents Chemother, 1994. 38(9): p. 2125-33.

121. Hoyle, B.D., J. Alcantara, and J.W. Costerton, Pseudomonas aeruginosa biofilm as a diffusion barrier to piperacillin. Antimicrob Agents Chemother, 1992. 36(9): p. 2054-6.

122. Bjarnsholt, T., et al., Pseudomonas aeruginosa biofilms in the respiratory tract of CF patients. Pediatr Pulmonol, 2009. 44(6): p. 547-58.

123. Anderl, J.N., M.J. Franklin, and P.S. Stewart, Role of antibiotic penetration limitation in Klebsiella pneumoniae biofilm resistance to ampicillin and ciprofloxacin. Antimicrob Agents Chemother, 2000. 44(7): p. 1818-24.

59

124. Giwercman, B., et al., Induction of beta-lactamase production in Pseudomonas aeruginosa biofilm. Antimicrobial Agents and Chemotherapy, 1991. 35(5): p. 1008-10.

125. Stewart, P.S., et al., Effect of catalase on hydrogen peroxide penetration into Pseudomonas aeruginosa biofilms. Appl Environ Microbiol, 2000. 66(2): p. 836-8.

126. Sternberg, C., et al., Distribution of bacterial growth activity in flow-chamber biofilms. Appl Environ Microbiol, 1999. 65(9): p. 4108-17.

127. Tuomanen, E., et al., The rate of killing of Escherichia coli by beta-lactam antibiotics is strictly proportional to the rate of bacterial growth. J Gen Microbiol, 1986. 132(5): p. 1297-304.

128. Duguid, I.G., et al., Effect of biofilm culture upon the susceptibility of Staphylococcus epidermidis to tobramycin. J Antimicrob Chemother, 1992. 30(6): p. 803-10.

129. Yang, L., et al., In situ growth rates and biofilm development of Pseudomonas aeruginosa populations in chronic lung infections. J Bacteriol, 2008. 190(8): p. 2767-76.

130. Haagensen, J.A., et al., Differentiation and distribution of colistin- and sodium dodecyl sulfate-tolerant cells in Pseudomonas aeruginosa biofilms. J Bacteriol, 2007. 189(1): p. 28-37.

131. Pamp, S.J., et al., Tolerance to the antimicrobial peptide colistin in Pseudomonas aeruginosa biofilms is linked to metabolically active cells, and depends on the pmr and mexAB-oprM genes. Mol Microbiol, 2008. 68(1): p. 223-40.

132. Dong, T. and H.E. Schellhorn, Role of RpoS in virulence of pathogens. Infect Immun, 2010. 78(3): p. 887-97.

133. Hengge-Aronis, R., Regulation of gene expression during entry into stationary phase, in Escherichia coli and Salmonella: Cellular and Molecular Biology, F.C. Neidhart and e. al., Editors. 1996, ASM Press. p. 1497-1512.

134. Liu, X., C. Ng, and T. Ferenci, Global adaptations resulting from high population densities in Escherichia coli cultures. J Bacteriol, 2000. 182(15): p. 4158-64.

135. Driffield, K., et al., Increased mutability of Pseudomonas aeruginosa in biofilms. J Antimicrob Chemother, 2008. 61(5): p. 1053-6.

136. Sorensen, S.J., et al., Studying plasmid horizontal transfer in situ: a critical review. Nat Rev Microbiol, 2005. 3(9): p. 700-10.

137. Molin, S. and T. Tolker-Nielsen, Gene transfer occurs with enhanced efficiency in biofilms and induces enhanced stabilisation of the biofilm structure. Curr Opin Biotechnol, 2003. 14(3): p. 255-61.

138. Mandsberg, L.F., et al., Antibiotic resistance in Pseudomonas aeruginosa strains with increased mutation frequency due to inactivation of the DNA oxidative repair system. Antimicrob Agents Chemother, 2009. 53(6): p. 2483-91.

139. Whitehead, N.A., et al., Quorum-sensing in Gram-negative bacteria. FEMS Microbiol Rev, 2001. 25(4): p. 365-404.

140. Bjarnsholt, T. and M. Givskov, The role of quorum sensing in the pathogenicity of the cunning aggressor Pseudomonas aeruginosa. Anal Bioanal Chem, 2007. 387(2): p. 409-14.

141. Bjarnsholt, T., et al., Pseudomonas aeruginosa tolerance to tobramycin, hydrogen peroxide and polymorphonuclear leukocytes is quorum-sensing dependent. Microbiology, 2005. 151(Pt 2): p. 373-83.

142. Espinosa-Urgel, M., A. Salido, and J.L. Ramos, Genetic analysis of functions involved in adhesion of Pseudomonas putida to seeds. J Bacteriol, 2000. 182(9): p. 2363-9.

60

143. Brooun, A., S. Liu, and K. Lewis, A Dose-Response Study of Antibiotic Resistance in Pseudomonas aeruginosa Biofilms. Antimicrob Agents Chemother, 2000. 44(3): p. 640-6.

144. Lewis, K., Riddle of Biofilm Resistance. Antimicrob Agents Chemother, 2001. 45(4): p. 999-1007.

145. Lewis, K., Persister cells and the riddle of biofilm survival. Biochemistry (Mosc), 2005. 70(2): p. 267-74.

146. Fauvart, M., V.N. De Groote, and J. Michiels, Role of persister cells in chronic infections: clinical relevance and perspectives on anti-persister therapies. J Med Microbiol, 2011. 60(Pt 6): p. 699-709.

147. Sriramulu, D.D., et al., Microcolony formation: a novel biofilm model of Pseudomonas aeruginosa for the CF lung. J Med Microbiol, 2005. 54(Pt 7): p. 667-76.

148. Pedersen, S.S., Lung infection with alginate-producing, mucoid Pseudomonas aeruginosa in CF. APMIS Suppl, 1992. 28: p. 1-79.

149. Parsek, M.R. and P.K. Singh, Bacterial biofilms: an emerging link to disease pathogenesis. Annu Rev Microbiol, 2003. 57: p. 677-701.

150. Mahenthiralingam, E., M.E. Campbell, and D.P. Speert, Nonmotility and phagocytic resistance of Pseudomonas aeruginosa isolates from chronically colonized patients with CF. Infect Immun, 1994. 62(2): p. 596-605.

151. Donlan, R.M. and J.W. Costerton, Biofilms: Survival Mechanisms of Clinically Relevant Microorganisms. Clin Microbiol Rev, 2002. 15(2): p. 167-93.

152. Kamiya, H., T. Ehara, and T. Matsumoto, Inhibitory effects of lactoferrin on biofilm formation in clinical isolates of Pseudomonas aeruginosa. J Infect Chemother, 2012. 18(1): p. 47-52.

153. Ammendolia, M.G., et al., Bovine lactoferrin interacts with cable pili of Burkholderia cenocepacia. Biometals, 2010. 23(3): p. 531-42.

154. Flemming, H.C., T.R. Neu, and D.J. Wozniak, The EPS matrix: the "house of biofilm cells". J Bacteriol, 2007. 189(22): p. 7945-7.

155. Branda, S.S., et al., Biofilms: the matrix revisited. Trends Microbiol, 2005. 13(1): p. 20-6.

156. Tielen, P., et al., Alginate acetylation influences initial surface colonization by mucoid Pseudomonas aeruginosa. Microbiol Res, 2005. 160(2): p. 165-76.

157. Allison, D.G. and M.J. Matthews, Effect of polysaccharide interactions on antibiotic susceptibility of Pseudomonas aeruginosa. J Appl Bacteriol, 1992. 73(6): p. 484-8.

158. Ma, L., et al., Assembly and development of the Pseudomonas aeruginosa biofilm matrix. PLoS Pathog, 2009. 5(3): p. e1000354.

159. Ma, L., et al., The roles of biofilm matrix polysaccharide Psl in mucoid Pseudomonas aeruginosa biofilms. FEMS Immunol Med Microbiol, 2012. 65(2): p. 377-80.

160. Mishra, M., et al., Pseudomonas aeruginosa Psl polysaccharide reduces neutrophil phagocytosis and the oxidative response by limiting complement-mediated opsonization. Cell Microbiol, 2012. 14(1): p. 95-106.

161. Colvin, K.M., et al., The pel polysaccharide can serve a structural and protective role in the biofilm matrix of Pseudomonas aeruginosa. PLoS Pathog, 2011. 7(1): p. e1001264.

162. Vuong, C., et al., Polysaccharide intercellular adhesin (PIA) protects Staphylococcus epidermidis against major components of the human innate immune system. Cell Microbiol, 2004. 6(3): p. 269-75.

61

163. Wang, X., J.F. Preston, 3rd, and T. Romeo, The pgaABCD locus of Escherichia coli promotes the synthesis of a polysaccharide adhesin required for biofilm formation. J Bacteriol, 2004. 186(9): p. 2724-34.

164. Izano, E.A., et al., Poly-N-acetylglucosamine mediates biofilm formation and antibiotic resistance in Actinobacillus pleuropneumoniae. Microb Pathog, 2007. 43(1): p. 1-9.

165. Choi, A.H., et al., The pgaABCD locus of Acinetobacter baumannii encodes the production of poly-beta-1-6-N-acetylglucosamine, which is critical for biofilm formation. J Bacteriol, 2009. 191(19): p. 5953-63.

166. Izano, E.A., et al., Poly-N-acetylglucosamine mediates biofilm formation and detergent resistance in Aggregatibacter actinomycetemcomitans. Microb Pathog, 2008. 44(1): p. 52-60.

167. Parise, G., et al., Role of a putative polysaccharide locus in Bordetella biofilm development. J Bacteriol, 2007. 189(3): p. 750-60.

168. Bobrov, A.G., et al., Insights into Yersinia pestis biofilm development: topology and co-interaction of Hms inner membrane proteins involved in exopolysaccharide production. Environ Microbiol, 2008. 10(6): p. 1419-32.

169. Yakandawala, N., et al., Characterization of the poly-beta-1,6-N-acetylglucosamine polysaccharide component of Burkholderia biofilms. Appl Environ Microbiol, 2011. 77(23): p. 8303-9.

170. Izano, E.A., et al., Differential roles of poly-N-acetylglucosamine surface polysaccharide and extracellular DNA in Staphylococcus aureus and Staphylococcus epidermidis biofilms. Appl Environ Microbiol, 2008. 74(2): p. 470-6.

171. Cunha, M.V., et al., Studies on the involvement of the exopolysaccharide produced by CF-associated isolates of the Burkholderia cepacia complex in biofilm formation and in persistence of respiratory infections. J Clin Microbiol, 2004. 42(7): p. 3052-8.

172. Cescutti, P., et al., First report of a lyase for cepacian, the polysaccharide produced by Burkholderia cepacia complex bacteria. Biochem Biophys Res Commun, 2006. 339(3): p. 821-6.

173. Cuzzi, B., et al., Investigation of bacterial resistance to the immune system response: cepacian depolymerisation by reactive oxygen species. Innate Immun, 2012. 18(4): p. 661-71.

174. Allesen-Holm, M., et al., A characterization of DNA release in Pseudomonas aeruginosa cultures and biofilms. Mol Microbiol, 2006. 59(4): p. 1114-28.

175. Yang, L., et al., Effects of iron on DNA release and biofilm development by Pseudomonas aeruginosa. Microbiology, 2007. 153(Pt 5): p. 1318-28.

176. Whitchurch, C.B., et al., Extracellular DNA required for bacterial biofilm formation. Science, 2002. 295(5559): p. 1487.

177. Mulcahy, H., L. Charron-Mazenod, and S. Lewenza, Extracellular DNA Chelates Cations and Induces Antibiotic Resistance in Pseudomonas aeruginosa Biofilms. PLoS Pathog, 2008. 4(11).

178. Goodman, S.D., et al., Biofilms can be dispersed by focusing the immune system on a common family of bacterial nucleoid-associated proteins. Mucosal Immunol, 2011. 4(6): p. 625-37.

179. Gustave, J.E., et al., Targeting bacterial integration host factor to disrupt biofilms associated with CF. J Cyst Fibros, 2012.

62

180. Batoni, G., et al., Use of antimicrobial peptides against microbial biofilms: advantages and limits. Curr Med Chem, 2011. 18(2): p. 256-79.

181. Flemming, K., et al., High in vitro antimicrobial activity of synthetic antimicrobial peptidomimetics against staphylococcal biofilms. J Antimicrob Chemother, 2009. 63(1): p. 136-45.

182. Hua, J., R.W. Scott, and G. Diamond, Activity of antimicrobial peptide mimetics in the oral cavity: II. Activity against periopathogenic biofilms and anti-inflammatory activity. Mol Oral Microbiol, 2010. 25(6): p. 426-32.

183. Qi, X., et al., Novel short antibacterial and antifungal peptides with low cytotoxicity: Efficacy and action mechanisms. Biochem Biophys Res Commun, 2010. 398(3): p. 594-600.

184. Davey, M.S., et al., A promising target for treatment of multidrug-resistant bacterial infections. Antimicrob Agents Chemother, 2011. 55(7): p. 3635-6.

185. Uh, E., et al., Antibacterial and antitubercular activity of fosmidomycin, FR900098, and their lipophilic analogs. Bioorg Med Chem Lett, 2011. 21(23): p. 6973-6.

186. Messiaen, A.S., et al., Resistance of the Burkholderia cepacia complex to fosmidomycin and fosmidomycin derivatives. Int J Antimicrob Agents, 2011. 38(3): p. 261-4.

187. Jackson, E.R. and C.S. Dowd, Inhibition of 1-deoxy-D-xylulose-5-phosphate reductoisomerase (Dxr): a review of the synthesis and biological evaluation of recent inhibitors. Curr Top Med Chem, 2012. 12(7): p. 706-28.

188. Davies, D.G., et al., The involvement of cell-to-cell signals in the development of a bacterial biofilm. Science, 1998. 280(5361): p. 295-8.

189. Pearson, J.P., et al., Pseudomonas aeruginosa cell-to-cell signaling is required for virulence in a model of acute pulmonary infection. Infect Immun, 2000. 68(7): p. 4331-4.

190. Wu, H., et al., Pseudomonas aeruginosa mutations in lasI and rhlI quorum sensing systems result in milder chronic lung infection. Microbiology, 2001. 147(Pt 5): p. 1105-13.

191. Erickson, D.L., et al., Pseudomonas aeruginosa quorum-sensing systems may control virulence factor expression in the lungs of patients with CF. Infect Immun, 2002. 70(4): p. 1783-90.

192. Middleton, B., et al., Direct detection of N-acylhomoserine lactones in CF sputum. FEMS Microbiol Lett, 2002. 207(1): p. 1-7.

193. Brackman, G., et al., Quorum sensing inhibitors increase the susceptibility of bacterial biofilms to antibiotics in vitro and in vivo. Antimicrob Agents Chemother, 2011. 55(6): p. 2655-61.

194. Christensen, L.D., et al., Synergistic antibacterial efficacy of early combination treatment with tobramycin and quorum-sensing inhibitors against Pseudomonas aeruginosa in an intraperitoneal foreign-body infection mouse model. J Antimicrob Chemother, 2012. 67(5): p. 1198-206.

195. Wu, H., et al., Synthetic furanones inhibit quorum-sensing and enhance bacterial clearance in Pseudomonas aeruginosa lung infection in mice. J Antimicrob Chemother, 2004. 53(6): p. 1054-61.

196. Conway, B.A., V. Venu, and D.P. Speert, Biofilm formation and acyl homoserine lactone production in the Burkholderia cepacia complex. J Bacteriol, 2002. 184(20): p. 5678-85.

63

197. Brackman, G., et al., Use of quorum sensing inhibitors to interfere with biofilm formation and development in Burkholderia multivorans and Burkholderia cenocepacia. Res Microbiol, 2009. 160(2): p. 144-51.

198. Kolodkin-Gal, I., et al., D-amino acids trigger biofilm disassembly. Science, 2010. 328(5978): p. 627-9.

199. Hochbaum, A.I., et al., Inhibitory effects of D-amino acids on Staphylococcus aureus biofilm development. J Bacteriol, 2011. 193(20): p. 5616-22.

200. Worlitzsch, D., et al., Effects of reduced mucus oxygen concentration in airway Pseudomonas infections of CF patients. J Clin Invest, 2002. 109(3): p. 317-25.

201. Jones, K.L., et al., Elevation of nitrotyrosine and nitrate concentrations in CF sputum. Pediatr Pulmonol, 2000. 30(2): p. 79-85.

202. Hassett, D.J., et al., Anaerobic metabolism and quorum sensing by Pseudomonas aeruginosa biofilms in chronically infected CF airways: rethinking antibiotic treatment strategies and drug targets. Adv Drug Deliv Rev, 2002. 54(11): p. 1425-43.

203. Yoon, S.S., et al., Anaerobic killing of mucoid Pseudomonas aeruginosa by acidified nitrite derivatives under CF airway conditions. J Clin Invest, 2006. 116(2): p. 436-46.

204. Barraud, N., et al., Nitric oxide signaling in Pseudomonas aeruginosa biofilms mediates phosphodiesterase activity, decreased cyclic di-GMP levels, and enhanced dispersal. J Bacteriol, 2009. 191(23): p. 7333-42.

205. Moreau-Marquis, S., et al., The DeltaF508-CFTR mutation results in increased biofilm formation by Pseudomonas aeruginosa by increasing iron availability. Am J Physiol Lung Cell Mol Physiol, 2008. 295(1): p. L25-37.

206. O'May, C.Y., et al., Iron-binding compounds impair Pseudomonas aeruginosa biofilm formation, especially under anaerobic conditions. J Med Microbiol, 2009. 58(Pt 6): p. 765-73.

207. Moreau-Marquis, S., G.A. O'Toole, and B.A. Stanton, Tobramycin and FDA-approved iron chelators eliminate Pseudomonas aeruginosa biofilms on CF cells. Am J Respir Cell Mol Biol, 2009. 41(3): p. 305-13.

208. Kaneko, Y., et al., The transition metal gallium disrupts Pseudomonas aeruginosa iron metabolism and has antimicrobial and antibiofilm activity. J Clin Invest, 2007. 117(4): p. 877-88.

209. Chitambar, C.R. and J. Narasimhan, Targeting iron-dependent DNA synthesis with gallium and transferrin-gallium. Pathobiology, 1991. 59(1): p. 3-10.

210. Banin, E., et al., The potential of desferrioxamine-gallium as an anti-Pseudomonas therapeutic agent. Proc Natl Acad Sci U S A, 2008. 105(43): p. 16761-6.

211. Banin, E., M.L. Vasil, and E.P. Greenberg, Iron and Pseudomonas aeruginosa biofilm formation. Proc Natl Acad Sci U S A, 2005. 102(31): p. 11076-81.

212. Cheer, S.M., J. Waugh, and S. Noble, Inhaled tobramycin (TOBI): a review of its use in the management of Pseudomonas aeruginosa infections in patients with CF. Drugs, 2003. 63(22): p. 2501-20.

213. Vandevanter, D.R. and D.E. Geller, Tobramycin administered by the TOBI((R)) Podhaler((R)) for persons with CF: a review. Med Devices (Auckl), 2011. 4: p. 179-88.

214. Konstan, M.W., et al., Safety, efficacy and convenience of tobramycin inhalation powder in CF patients: The EAGER trial. J Cyst Fibros, 2011. 10(1): p. 54-61.

215. Serisier, D.J., Inhaled antibiotics for lower respiratory tract infections: focus on ciprofloxacin. Drugs Today (Barc), 2012. 48(5): p. 339-51.

64

216. Okusanya, O.O., et al., Pharmacokinetic and pharmacodynamic evaluation of liposomal amikacin for inhalation in CF patients with chronic pseudomonal infection. Antimicrob Agents Chemother, 2009. 53(9): p. 3847-54.

217. Cheow, W.S., M.W. Chang, and K. Hadinoto, Antibacterial efficacy of inhalable antibiotic-encapsulated biodegradable polymeric nanoparticles against E. coli biofilm cells. J Biomed Nanotechnol, 2010. 6(4): p. 391-403.

218. Cheow, W.S. and K. Hadinoto, Green preparation of antibiotic nanoparticle complex as potential anti-biofilm therapeutics via self-assembly amphiphile-polyelectrolyte complexation with dextran sulfate. Colloids Surf B Biointerfaces, 2012. 92: p. 55-63.

219. Peeters, C., et al., Burkholderia pseudomultivorans sp. nov., a novel Burkholderia cepacia complex species from human respiratory samples and the rhizosphere. Syst Appl Microbiol, 2013.

220. Burkholder, W.H., Sour skin, a bacterial rot of onion bulbs. phytopathology, 1950. 40(1): p. 115-117.

221. Parke, J.L. and D. Gurian-Sherman, Diversity of the Burkholderia cepacia complex and implications for risk assessment of biological control strains. Annu Rev Phytopathol, 2001. 39: p. 225-58.

222. Goldmann, D.A. and J.D. Klinger, Pseudomonas cepacia: biology, mechanisms of virulence, epidemiology. J Pediatr, 1986. 108(5 Pt 2): p. 806-12.

223. Laraya-Cuasay, L.R., M. Lipstein, and N.N. Huang, Pseudomonas cepacia in the respiratory flora of patients with CF. Pediatr Pulmonol, 1977. 11: p. 502.

224. Shigeharu, O. and A. Kamiya, Microbial contamination of antiseptics and disinfectants. Am J Infect Control, 1996. 24(5): p. 389-395.

225. Ederer, G.M.a.M., J. M., Colonization and infection with Pseudomonas cepacia. J Infect Dis, 1972. 125(6).

226. Mahenthiralingam, E., T.A. Urban, and J.B. Goldberg, The multifarious, multireplicon Burkholderia cepacia complex. Nat Rev Microbiol, 2005. 3(2): p. 144-156.

227. Yabuuchi, E., et al., Proposal of Burkholderia gen. nov. and transfer of seven species of the genus Pseudomonas homology group II to the new genus, with the type species Burkholderia cepacia (Palleroni and Holmes 1981) comb. nov. Microbiol Immunol, 1992. 36(12): p. 1251-75.

228. Vandamme, P., et al., Occurrence of multiple genomovars of Burkholderia cepacia in CF patients and proposal of Burkholderia multivorans sp. nov. Int J Syst Bacteriol, 1997. 47(4): p. 1188-200.

229. Vial, L., et al., The various lifestyles of the Burkholderia cepacia complex species: a tribute to adaptation. Environ Microbiol, 2011. 13(1): p. 1-12.

230. Johnston, R.B., Jr., Clinical aspects of chronic granulomatous disease. Curr Opin Hematol, 2001. 8(1): p. 17-22.

231. Marioni, G., et al., Cervical necrotizing fasciitis: a novel clinical presentation of Burkholderia cepacia infection. J Infect, 2006. 53(5): p. e219-22.

232. Bayram, M., et al., Community-acquired Burkholderia cepacia pneumonia: a report of two immunocompetent patients. Tuberk Toraks, 2011. 59(4): p. 380-3.

233. Memish, Z.A., et al., Outbreak of Burkholderia cepacia bacteremia in immunocompetent children caused by contaminated nebulized sulbutamol in Saudi Arabia. Am J Infect Control, 2009. 37(5): p. 431-2.

65

234. Marioni, G., et al., Burkholderia cepacia complex nasal isolation in immunocompetent patients with sinonasal polyposis not associated with CF. Eur J Clin Microbiol Infect Dis, 2007. 26(1): p. 73-5.

235. Sousa, S.A., C.G. Ramos, and J.H. Leitao, Burkholderia cepacia Complex: Emerging Multihost Pathogens Equipped with a Wide Range of Virulence Factors and Determinants. Int J Microbiol, 2011. 2011.

236. Mann, T., et al., An outbreak of Burkholderia cenocepacia bacteremia in immunocompromised oncology patients. Infection, 2010. 38(3): p. 187-94.

237. Tablan, O.C., et al., Pseudomonas cepacia colonization in patients with CF: risk factors and clinical outcome. J Pediatr, 1985. 107(3): p. 382-7.

238. Corey, M. and V. Farewell, Determinants of mortality from CF in Canada, 1970-1989. Am J Epidemiol, 1996. 143(10): p. 1007-17.

239. Mahenthiralingam, E., A. Baldwin, and P. Vandamme, Burkholderia cepacia complex infection in patients with CF. J Med Microbiol, 2002. 51(7): p. 533-8.

240. Mahenthiralingam, E., et al., Infection with Burkholderia cepacia complex genomovars in patients with CF: virulent transmissible strains of genomovar III can replace Burkholderia multivorans. Clin Infect Dis, 2001. 33(9): p. 1469-75.

241. Holmes, A., et al., An epidemic of burkholderia cepacia transmitted between patients with and without CF. J Infect Dis, 1999. 179(5): p. 1197-205.

242. Lipuma, J.J., Preventing Burkholderia cepacia complex infection in CF: is there a middle ground? J Pediatr, 2002. 141(4): p. 467-9.

243. Saiman, L. and J. Siegel, Infection control in CF. Clin Microbiol Rev, 2004. 17(1): p. 57-71.

244. Duff, A.J., Psychological consequences of segregation resulting from chronic Burkholderia cepacia infection in adults with CF. Thorax, 2002. 57(9): p. 756-8.

245. Cheng, H.P. and T.G. Lessie, Multiple replicons constituting the genome of Pseudomonas cepacia 17616. J Bacteriol, 1994. 176(13): p. 4034-42.

246. Rodley, P.D., U. Romling, and B. Tummler, A physical genome map of the Burkholderia cepacia type strain. Mol Microbiol, 1995. 17(1): p. 57-67.

247. Mahenthiralingam, E. and P. Drevinek, Comparitive genomics of Burkholderia species, in Burkholderia Molecular Microbiology and Genomics, T. Coenye and P. Vandamme, Editors. 2007, horizon bioscience: United Kingdome. p. 300.

248. Agnoli, K., et al., Exposing the third chromosome of Burkholderia cepacia complex strains as a virulence plasmid. Mol Microbiol, 2012. 83(2): p. 362-78.

249. Holden, M.T., et al., The genome of Burkholderia cenocepacia J2315, an epidemic pathogen of CF patients. J Bacteriol, 2009. 191(1): p. 261-77.

250. Hacker, J. and J.B. Kaper, Pathogenicity islands and the evolution of microbes. Annu Rev Microbiol, 2000. 54: p. 641-79.

251. Baldwin, A., et al., The Burkholderia cepacia epidemic strain marker is part of a novel genomic island encoding both virulence and metabolism-associated genes in Burkholderia cenocepacia. Infect Immun, 2004. 72(3): p. 1537-47.

252. De Soyza, A., et al., Chemical and biological features of Burkholderia cepacia complex lipopolysaccharides. Innate Immun, 2008. 14(3): p. 127-44.

253. Hancock, R.E. and A. Bell, Antibiotic uptake into gram-negative bacteria. Eur J Clin Microbiol Infect Dis, 1988. 7(6): p. 713-20.

254. De Soyza, A., et al., Burkholderia cenocepacia lipopolysaccharide, lipid A, and proinflammatory activity. Am J Respir Crit Care Med, 2004. 170(1): p. 70-7.

255. Tomlin, K.L., O.P. Coll, and H. Ceri, Interspecies biofilms of Pseudomonas aeruginosa and Burkholderia cepacia. Can J Microbiol, 2001. 47(10): p. 949-54.

66

256. Berlutti, F., et al., Iron availability influences aggregation, biofilm, adhesion and invasion of Pseudomonas aeruginosa and Burkholderia cenocepacia. Int J Immunopathol Pharmacol, 2005. 18(4): p. 661-70.

257. Bylund, J., et al., Exopolysaccharides from Burkholderia cenocepacia inhibit neutrophil chemotaxis and scavenge reactive oxygen species. J Biol Chem, 2006. 281(5): p. 2526-32.

258. Lewenza, S., et al., Quorum sensing in Burkholderia cepacia: identification of the LuxRI homologs CepRI. J Bacteriol, 1999. 181(3): p. 748-56.

259. Venturi, V., et al., Quorum sensing in the Burkholderia cepacia complex. Res Microbiol, 2004. 155(4): p. 238-44.

260. Beckman, W. and T.G. Lessie, Response of Pseudomonas cepacia to beta-Lactam antibiotics: utilization of penicillin G as the carbon source. J Bacteriol, 1979. 140(3): p. 1126-8.

67

PART 2

Objectives

68

69

The primary pathogen isolated from the CF lung is P. aeruginosa. Infections with Bcc

bacteria are much less prevalent compared to P. aeruginosa infections but the clinical

outcome of Bcc infected CF patients is highly variable and often associated with a rapid

decline of the lung function. In addition, as Bcc bacteria are capable of person-to-person

transmission, isolation measures for Bcc infected patients have been frequently

implemented in many different CF care centres. Eradication of Bcc infections is difficult

because of the innate resistance of the Bcc to many antibiotics and the ability of biofilm

formation, providing additional protection against both antibiotics and host immune

components. The urge for new treatment modalities to deal with Bcc infections is high

as the development of new antibiotics to potentially overcome the problem of innate

resistance is on decline.

The aim of my PhD thesis was to investigate different strategies to treat Bcc infections.

A first aim was to investigate whether fosmidomycin and fosmidomycin derivatives

could be useful to treat Bcc infections. Secondly, we investigated also whether Bcc

bacteria use self-promoted uptake for tobramycin; this could pave the path for the

development of selective tobramycin derivatives. To overcome the problem of a

decreased bioavailability of free tobramycin because of interaction with biofilm matrix

components, the combination of tobramycin with matrix degrading components was

investigated in a third study. We hypothesised that these antibiotic-matrix interactions

can also be inhibited by the encapsulation of tobramycin in liposomal carriers. This has

the additional advantage of targeted drug delivery near the biofilm cells. Therefore, a

fourth objective was to investigate the potential of different liposomal tobramycin

formulations to treat Bcc biofilm infections.

70

71

PART 3

Experimental work

72

Chapter IV: RESISTANCE OF THE BURKHOLDERIA CEPACIA COMPLEX

AGAINST FOSMIDOMYCIN AND FOSMIDOMYCIN DERIVATIVES Redrafted from:

Messiaen AS, Verbrugghen T, Declerck C, Ortmann R, Schlitzer M, Nelis H, Van Calenbergh

S, Coenye T. (2011). Resistance of the burkholderia cepacia complex against fosmidomycin and

fosmidomycin derivatives. Int J Antimicrob Agents 38(3):261-264.

73

ABSTRACT

The Bcc is a group of at least 18 closely related opportunistic pathogens that are able to

infect the respiratory tract of CF patients. Bcc bacteria are intrinsic resistant to many

antibiotics and are therefore difficult to eradicate. Fosmidomycin could be a new

therapeutic agent to treat Bcc infections as it inhibits Dxr, a key enzyme in the non-

mevalonate pathway essential in Bcc bacteria for isoprenoid synthesis. In the present

study the antimicrobial activity of fosmidomycin and eight fosmidomycin derivatives

towards 40 Bcc strains was investigated. All Bcc strains were resistant, although the

addition of glucose-6-phosphate reduced MIC values of FR900098, the fosmidomycin

acetyl derivative, from 512 mg/L to 64 mg/L for B. multivorans and B. cepacia. This

enhanced activity was linked to an increased expression of the genes involved in

glycerol-3-phosphate transport, which seems to be the only route for fosmidomycin

import in Bcc bacteria. Further, the upregulation of a fosmidomycin resistance gene

(fsr), encoding an efflux pump, was observed during fosmidomycin and FR900098

treatment. These results strongly suggest that the observed resistance in Bcc bacteria is

due to an insufficient uptake accompanied by fosmidomycin and FR900098 efflux.

74

Introduction

In the early 1980s Bcc bacteria emerged as problematic pathogens in CF patients, with B.

cenocepacia and B. multivorans being the most frequently isolated species. Compared to

P. aeruginosa, Bcc infections only account for a small percentage of the respiratory

infections within the CF population, however, they are often associated with rapid

deterioration of the lung function and increased mortality [1]. Once infection with Bcc

bacteria is established, therapeutic options are limited because of the innate resistance

of these bacteria to many antibiotics. Different resistance mechanisms have already

been described, including increased efflux and decreased permeability of the outer

membrane [2]. Despite these high levels of resistance there has been little progress in

developing new therapeutic strategies to fight these infections.

The non-mevalonate pathway for isoprenoid synthesis was discovered in the early

1990s as an alternative to the classical mevalonate pathway. While most bacteria

exclusively synthesize isoprenoids, like coenzyme Q, through this pathway, mammalian

cells do not possess it [3]. Fosmidomycin, a natural phosphonic acid antibiotic, is a

specific inhibitor of a key enzyme of the non-mevalonate pathway, i.e. Dxr. Although it

has shown good safety and efficacy, e.g. for the treatment of uncomplicated Plasmodium

falciparum malaria [4], the rapid development of resistance during therapy is of great

concern. White et al. suggest that this resistance is due to a decreased uptake of the

compound [5]. The uptake of fosmidomycin is hypothesized to follow the same route as

its structural analog fosfomycin, either through hexose-phosphate or through glycerol-

3-phosphate transporters because mutations in genes encoding these transport systems

confer resistance to both fosmidomycin and fosfomycin [6]. Further, the B. cenocepacia

genome contains a fosmidomycin resistance gene (fsr), encoding an efflux pump, which

transports fosmidomycin out of the cell [7].

The aim of the present study was to evaluate the in vitro activity of fosmidomycin and

eight derivatives against the Bcc and to elucidate the mechanisms of resistance in this

group of organisms.

75

Materials and methods

Strains

The following 40 Bcc strains were used: B. cepacia LMG 18821 and LMG 1222T; B.

multivorans LMG 18822, LMG 18825, LMG 13010T and LMG 17588; B. cenocepacia LMG

16656T, LMG 18828, LMG 18829, LMG 18830, LMG 6986 and K56-2; B. stabilis LMG

14294T and LMG 14086; B. vietnamiensis LMG 18835 and LMG 10929T; B. dolosa LMG

18943T and LMG 18941; B. ambifaria LMG 19182T and LMG 19467; B. anthina LMG

20980T and LMG 20983; B. pyrrocinia LMG 14191T and LMG 21824; B. ubonensis LMG

20358T and LMG 24263; B. latens R-11768 and LMG 24264; B. diffusa LMG 24065T and

LMG 24266; B. arboris LMG 24066T and R-132; B. seminalis LMG 24067T and LMG

24272; B. metallica LMG 24068T and R-2712; B. lata LMG 6992 and R-9940; and B.

contaminans LMG 16227 and R-12710. All strains were obtained from the BCCM/LMG

Bacteria Collection (Ghent, Belgium) or were kindly provided by Dr P. Vandamme

(Ghent University, Belgium). We used two E. coli K-12 strains as positive controls as it

was previously shown that fosmidomycin is active against these isolates [8].

Antimicrobial agents

Fosmidomycin was obtained from Invitrogen (Eugene, OR). All fosmidomycin

derivatives were synthesized as previously described [9-13] (See Figure 1 for an

overview of the structures).

Figure 1: Overview of the

molecular structures of

fosmidomycin and the

fosmidomycin derivatives

used in the present study.

1: fosmidomycin;

2: FR900098; 3: TVI 064;

4: Schl 7150;

5: Schl 7498; 6: TVII 046;

7: TVII 007; 8: TVB 006;

9: TVI 118.

76

Determination of the MIC

MICs were determined according to the EUCAST standard broth microdilution protocol

[14]. A two-fold antibiotic dilution series ranging from 1 to 512 mg/L was tested. MICs

were also determined in the presence of glucose-6-phosphate and glucose-1-phosphate

as sole carbon sources. To this end, a chemically defined minimal salt medium

supplemented with 50 μM FeCl3.6H2O (designated as CDM) [15] and equimolar

concentrations of glucose-6-phosphate and glucose-1-phosphate instead of glucose

(designated as CDM-G) was used.

RNA extraction and cDNA synthesis

Cells were cultured in either CDM or CDM-G medium. Antibiotic concentrations,

corresponding to the previously determined MIC values were used during treatment. If

MIC values were > 512 mg/L, cultures were grown in medium containing 512 mg/L of

the antimicrobial agent. After 24 h of incubation, with constant agitation, cells were

collected and RNA was extracted using the Ambion RiboPure Bacteria Kit (Ambion,

Austin, TX). Manufacturer’s instructions were followed except for the DNAse treatment

that was prolonged to 1 h. After RNA extraction, RNA yields were assessed using the

RNA Quant-iT kit (Invitrogen). 15 μL RNA samples were used to synthesize cDNA using

the iScript cDNA Synthesis Kit.

Quantitative real time PCR

The expression of 7 genes (including 2 reference genes) was investigated by means of

quantitative RT-PCR (qPCR). Primers were developed using the website of NCBI

(http://www.ncbi.nlm.nih.gov) and their specificity was checked using BLAST (See

Table 1 for primer sequences). All qPCR experiments were performed on a Bio-Rad

CFX96 Real-Time System C1000 Thermal Cycler using the iQ SYBR Green Supermix kit

(Bio-Rad, Hercules, CA) according to the manufacturer’s instructions. To activate the

polymerase included in the iQ SYBR Green Supermix kit the reaction mixture was heated

at 95°C for 3 min, followed by 40 amplification cycles, consisting of 15 s at 95°C and 60 s

at 60°C. A melting curve analysis was included at the end of each run. Real-time PCR

data were normalized to the geometric mean of the expression of two reference genes

(BCAL0289 and BCAL1861), which were stably expressed in all conditions investigated

in the present study (data not shown).

77

Table 1: Forward (FW) and reverse (RV) primers used in real-time PCR experiments.

Gene Annotation Orientation Primer sequence (5’ to 3’) Concentration

BCAL2085 Dxr FW GCGCTCGAGGAAGGCGGTAT 600 nM

RW TCCGGCGCTCGAGAAATGCT 600 nM

BCAL1451 Efflux pump FW TGTTCCAGGTGGGCGGCAAT 300 nM

RW CCAGTGGCCGATCTGCGTGA 300 nM

BCAL0282 Glycerol-3P-transporter FW ATGCGAAGACGGGCCACCTC 200 nM

RW GCCCGCCTTCTTGAACGCATC 200 nM

BCAL0284 Glycerol-3P transporter membrane

protein

FW GGGCTTCGACCTGTTCTGCCA 200 nM

RW ACGCGACGTAGACGGGGAAC 200 nM

BCAL0285 Glycerol-3P transporter ATP binding

subunit

FW CTGAGCTTGAAGGGCGTCAGGA 200 nM

RW GGGCCGACCAGCACGATGAA 200 nM

BCAL0289 Glutamate synthase FW ATCATCCAGCAGGGTCTGAAGA 600 nM

RW GCCATTTCCTCGCGATAGAA 600 nM

BCAL1861 Acetoacetyl-CoA reductase FW GATCACCTGCTTCGTGACGTT 600 nM

RW GACGTCGTGTTCCGCAAGAT 600 nM

78

Results

Susceptibility testing in Mueller Hinton broth

For all Bcc strains tested, MICs of fosmidomycin were > 512 mg/L, whereas those for the

two E. coli K-12 strains, serving as positive controls, ranged from 0.5 to 2 mg/L, which is

in agreement with previously reported data [8].

The acetyl derivative, FR900098, was capable of inhibiting growth of 18 of the 40 strains

tested at concentrations ≤ 512 mg/L. B. multivorans LMG 18822 and LMG 18825, B.

cenocepacia LMG 18830 and LMG 6986, B. stabilis LMG 14294T, B. vietnamiensis LMG

18835 and LMG 10929T, B. dolosa LMG 18941, B. pyrrocinia LMG 21824, B. ubonensis

LMG 20358T and LMG 24263, B. seminalis LMG 24067T and B. metallica LMG 24068T and

R-2712 were inhibited at concentrations of 512 mg/L. Further, B. cepacia LMG 1222T, B.

multivorans LMG 13010T and B. lata LMG 6992 were inhibited at concentrations of 256

mg/L and B. anthina LMG 20983 at a concentration of 128 mg/L.

The seven other derivatives were tested against three Bcc type strains, B. cepacia LMG

1222T, B. multivorans LMG 13010T and B. cenocepacia LMG 16656T, but growth

inhibitory concentrations of these derivatives were never below 512 mg/L (Table 2).

Table 2: MICs of fosmidomycin and 8 fosmidomycin derivatives for 3 Bcc type strains (experiments

were performed in duplicate).

Fos: fosmidomycin

MIC (mg/L)

Strain Fos FR900098 TVII 007 TVII 046 TVI 064 TVI 118 TVB 006 Schl 7150 Schl 7498

B. cepacia

LMG 1222T > 512 256 512 > 512 512 > 512 > 512 512 > 512

B. multivorans

LMG 13010T > 512 256 512 > 512 512 512 > 512 512 > 512

B. cenocepacia

LMG 16656T > 512 > 512 > 512 > 512 512 > 512 > 512 512 > 512

79

Susceptibility testing in the presence of glucose-6-phosphate

First, the activity of fosmidomycin in the presence of glucose-6-phosphate was

investigated by adding 25 mg/L glucose-6-phosphate to Mueller Hinton (MH) broth.

However, MICs remained > 512 mg/L for all 40 strains tested (data not shown).

To investigate the activity of fosmidomycin in the presence of glucose-6-phosphate as

sole carbon source, CDM medium supplemented with Fe3+ [15] (designated CDM) and

CDM with glucose-6-phosphate replacing glucose at equimolar concentrations

(designated as CDM + G6P) was used. Although it should be noted that the reduced

growth in these media precluded the determination of the exact MIC value, an increased

sensitivity of B. multivorans LMG 13010T and B. cepacia LMG 1222T to fosmidomycin in

CDM + G6P compared to CDM was observed (Figure 2).

Figure 2: Activity of fosmidomycin in the presence of glucose-6-phosphate as sole carbon source.

Full line: activity of

fosmidomycin in CDM + G6P

medium.

Dotted line: activity of

fosmidomycin in CDM

medium;

∎ : B. multivorans LMG

13010T; ▲ : B. cepacia LMG

1222T.

Because of the higher activity of FR900098 the effect of glucose-6-phosphate was

further investigated using this acetyl derivative. MICs for B. cenocepacia LMG 6986, B.

multivorans LMG 13010T and B. cepacia LMG 1222T were determined in CDM with

glucose-6-phosphate and glucose-1-phosphate as sole carbon sources (designated as

CDM-G). Glucose-1-phosphate had no effect on MIC values but improved growth to

levels equivalent to the growth in MH broth (data not shown). MICs of FR900098 for all

three strains tested were 64 mg/L in CDM-G compared to 512 mg/L for B. cenocepacia

LMG 6986 and 256 mg/L for B. multivorans LMG 13010T and B. cepacia LMG 1222T in

MH.

80

Effect of glucose-6-phosphate on the expression of genes involved in glycerol-3-phosphate

transport during FR900098 treatment

The expression of three genes involved in glycerol-3-phosphate transport in B.

cenocepacia LMG 16656T (BCAL0282, BCAL0284 and BCAL0285) during treatment with

64 mg/L FR900098 was investigated. The expression levels for the three genes tested

were significantly higher (p < 0.05) in the presence of glucose-6-phosphate (CDM-G)

compared to controls (CDM) (Figure 3a). The expression of BCAL0283, which encodes a

putative ABC transporter permease involved in glycerol-3-phosphate transport was too

low to allow accurate quantification (data not shown).

Expression of BCAL2085 (dxr) and BCAL1451 (fsr)

The expression levels of dxr and fsr in the presence of fosmidomycin or FR900098 (64

mg/L) in CDM-G medium relative to untreated controls are shown in Figure 3b. The fsr

gene encoding the efflux pump was significantly upregulated during treatment

compared to untreated controls (p < 0.05), while the gene encoding the target enzyme

(Dxr) was stably expressed in both conditions.

Figure 3: Gene expression data (Experiments were performed in triplicate).

a) Transcriptional response of genes belonging to the glycerol-3-phosphate transport system to the

addition of glucose-6-phosphate in both untreated and FR900098 treated conditions (expressed as fold

change in CDM-G compared to CDM).

b) Expression of BCAL2085 (dxr) and BCAL1451 (fsr) in the presence of fosmidomycin and FR900098 (64

mg/L) (expressed as fold change in CDM-G with fosmidomycin or FR900098 compared to CDM-G).

Mean values of three replicates are displayed with error bars representing standard deviations. A more

than 1.5 fold change in expression is considered upregulated (*).

81

D iscussion

Bcc bacteria are known for their resistance against a wide range of antimicrobial agents.

The aim of the present study was to evaluate the potential of fosmidomycin for the

treatment of Bcc infections.

Despite the presence of dxr (BCAL2085) in the B. cenocepacia J2315 genome, a clear

resistance against fosmidomycin and its derivatives was observed for all Bcc strains

tested. This indicates that the presence of the non-mevalonate pathway, and more

specifically Dxr, does not guarantee susceptibility towards inhibitors of this pathway

and that the resistance may be due to altered transport. The glycerol-3-phosphate

transporter (GlpT) is reportedly the main importer of fosmidomycin in E. coli [6]. No

glpT homologs were found in the B. cenocepacia J2315 genome, nevertheless an

alternative system involved in glycerol-3-phosphate import is present. This transporter

belongs to the ABC transporter superfamily and consists of a periplasmic substrate

binding protein (ugpB; BCAL0282), two permease proteins (ugpA; BCAL0283 and ugpE;

BCAL0284) and an ATP binding protein (ugpC; BCAL0285). Because of the structural

similarity of fosmidomycin to glycerol-3-phosphate we suspect that fosmidomycin can

enter B. cenocepacia cells through this alternative glycerol-3-phosphate transport

system as well.

The activity of both fosmidomycin and FR900098 was enhanced in the presence of

glucose-6-phosphate when used either as a sole carbon source or in combination with

glucose-1-phosphate, suggesting the presence of an inducible sugar phosphate transport

system through which fosmidomycin can be imported into Bcc bacteria. The hexose

phosphate transporter (UhpT) is a main transport system for fosfomycin, a structural

analog of fosmidomycin, in E. coli cells [6, 16]. Based on the genes involved in the

hexose-phosphate transport system in E. coli, we searched for their homologs in the

genome of B. cenocepacia J2315 using BLAST. However, we did not identify uhp

homologs, suggesting that B. cenocepacia lacks a specific sugar-phosphate transport

system.

This indicates that fosmidomycin can only be imported into B. cenocepacia cells through

its glycerol-3-phosphate transport system. Therefore, we investigated the effect of

glucose-6-phosphate on the expression of ugpB, ugpC and ugpE. A significant

upregulation of these genes in CDM-G medium, compared with CDM medium, was

observed. As the only difference between these media is the carbon source, the

82

upregulation must be attributed to the presence of glucose-6-phosphate or glucose-1-

phosphate in CDM-G medium. As previous observations indicated that glucose-1-

phosphate had no influence on MIC values, the upregulation is likely caused by glucose-

6-phosphate.

A gene conferring resistance to fosmidomycin (fsr) was shown to be involved in

fosmidomycin efflux in E.coli [7]. As a fsr homolog is present in the genome of B.

cenocepacia J2315 and is significantly upregulated in treated samples this could in part

explain the observed resistance in Bcc bacteria.

In conclusion, our results suggest that the only transport system for fosmidomycin in

Bcc bacteria is a glycerol-3-phosphate transport system, which is induced in the

presence of glucose-6-phosphate. Consequently, the uptake of fosmidomycin can be

increased by the addition of glucose-6-phosphate. However, a fosmidomycin efflux

pump is capable of compensating for this increased fosmidomycin uptake, providing

resistance of Bcc bacteria against fosmidomycin and FR900098.

Acknowledgements

This research was supported by FWO-Vlaanderen and BOF (Ghent University).

T.V. is a Fellow of the Agency for Innovation by Science and Technology of Flanders

(IWT Vlaanderen)

83

References

1. Mahenthiralingam, E., T.A. Urban, and J.B. Goldberg, The multifarious, multireplicon Burkholderia cepacia complex. Nat Rev Microbiol, 2005. 3(2): p. 144-156.

2. Avgeri, S.G., et al., Therapeutic options for Burkholderia cepacia infections beyond co-trimoxazole: a systematic review of the clinical evidence. Int J Antimicrob Agents, 2009. 33(5): p. 394-404.

3. Eisenreich, W., et al., Biosynthesis of isoprenoids via the non-mevalonate pathway. Cell Mol Life Sci, 2004. 61(12): p. 1401-1426.

4. Na-Bangchang, K., et al., Pharmacokinetics and pharmacodynamics of fosmidomycin monotherapy and combination therapy with clindamycin in the treatment of multidrug resistant falciparum malaria. Malar J, 2007. 6.

5. White, R.J., et al., Targeting metalloenzymes: a strategy that works. Curr Opin Pharmacol, 2003. 3(5): p. 502-507.

6. Castaneda-Garcia, A., et al., The Glycerol-3-Phosphate Permease GlpT Is the Only Fosfomycin Transporter in Pseudomonas aeruginosa. J Bacteriol, 2009. 191(22): p. 6968-6974.

7. Fujisaki, S., et al., Cloning of a gene from Escherichia coli that confers resistance to fosmidomycin as a consequence of amplification. Gene, 1996. 175(1-2): p. 83-87.

8. Neu, H.C. and T. Kamimura, In vitro and in vivo anti-bacterial activity of FR-31564, a phosphonic acid anti-microbial agent. Antimicrob Agents Chemother, 1981. 19(6): p. 1013-1023.

9. Fokin, A.A., et al., Synthesis of the antimalarial drug FR900098 utilizing the nitroso-ene reaction. Organic Letters, 2007. 9(21): p. 4379-4382.

10. Haemers, T., et al., Synthesis of beta- and gamma-oxa isosteres of fosmidomycin and FR900098 as antimalarial candidates. Bioorg Med Chem, 2008. 16(6): p. 3361-3371.

11. Haemers, T., et al., Synthesis of alpha-substituted fosmidomycin analogues as highly potent Plasmodium falciparum growth inhibitors. Bioorg Med Chem, 2006. 16(7): p. 1888-1891.

12. Verbrugghen, T., et al., Synthesis and Evaluation of alpha-Halogenated Analogues of 3-(Acetylhydroxyamino)propylphosphonic Acid (FR900098) as Antimalarials. J Med Chem, 2010. 53(14): p. 5342-5346.

13. Ortmann, R., et al., Acyloxyalkyl ester Prodrugs of FR900098 with improved in vivo anti-malarial activity. Bioorg Med Chem Lett, 2003. 13(13): p. 2163-2166.

14. European Committee for Antimicrobial Susceptibility Testing of the European Society of Clinical, M., Determination of minimum inhibitory concentrations (MICs) of antibacterial agents by broth dilution.

15. Peeters, E., H.J. Nelis, and T. Coenye, Resistance of planktonic and biofilm-grown Burkholderia cepacia complex isolates to the transition metal gallium. J Antimicrob Chemother, 2008. 61(5): p. 1062-1065.

16. Takahata, S., et al., Molecular mechanisms of fosfomycin resistance in clinical isolates of Escherichia coli. Int J Antimicrob Agents, 2010. 35(4): p. 333-337.

84

Chapter V: UPTAKE OF TOBRAMYCIN AND TOBRAMYCIN

DERIVATIVES IN DIFFERENT GRAM-NEGATIVE BACTERIA

Redrafted from:

Anne-Sophie Messiaen*, Martijn Risseeuw*, Krisztina Fehér, José Martins, Hans

Nelis, Serge Van Calenbergh, Tom Coenye. (2013). Uptake of tobramycin and

tobramycin derivatives in different gram-negative bacteria. J Antibiot, Submitted.

*Both authors contributed equally to the manuscript

85

ABSTRACT

Chronic P. aeruginosa infections, the main cause of morbidity and mortality in CF

patients, are often treated with daily tobramycin inhalation therapy. The cellular uptake

of aminoglycosides, including tobramycin, can be divided in three phases: an initial

energy-independent electrostatic interaction at the cell surface and subsequently two

energy-dependent phases. Nevertheless, fundamental differences exist in

aminoglycoside uptake between different Gram-negative and Gram-positive bacteria. In

Gram-negative bacteria, aminoglycosides can cross the outer membrane either through

diffusion through hydrophilic channels formed by porin proteins or through a process

referred to as self-promoted uptake. In the present study, we investigated the mode of

uptake of tobramycin in different Gram-negative bacteria, including P. aeruginosa, E. coli

and Bcc organisms. To this end we determined the activity of tobramycin in the

presence and absence of Mg2+. In addition, we determined the activity of three

tobramycin derivatives, i.e. tobramycin-OH-BODIPY, tobramycin-N-Me2 and tobramycin-

OH-EDTA, on both planktonic and biofilm cells. Increased tobramycin MICs in the

presence of Mg2+ confirmed self-promoted uptake in P. aeruginosa. In contrast,

decreased MICs in the presence of Mg2+ for Bcc organisms suggested alternative uptake

mechanisms. Based on the MIC of tobramycin-OH-BODIPY for planktonic cells and the

high activity of tobramycin-OH-EDTA towards biofilms, we hypothesized that

tobramycin is at least in part taken up through self-promoted uptake in E. coli, as both

derivatives are too large to be taken up through porins. The observed selective activity

of tobramycin-N-Me2 against P. aeruginosa biofilms appears to be biofilm specific.

86

Introduction

CF is the most prevalent genetically inherited disease in the Caucasian population and is

caused by mutations in both copies of the CFTR gene [1]. CFTR encodes a chloride

channel which is expressed on the apical surface of different organs, including the lung

[2]. A malfunctioning chloride channel in lung epithelial cells results in the secretion of

thick viscous mucus which impairs mucociliary clearance and thereby sensitizes the CF

lung to chronic bacterial infections [3]. Within a few years after birth, P. aeruginosa

becomes the predominant pathogen isolated from the CF lung. Early treatment of these

infections is essential as they are the main cause of morbidity and mortality in CF

patients [4]. Chronic P. aeruginosa infections are often treated with tobramycin which

has been shown to improve lung function and reduce P. aeruginosa cell numbers in CF

sputum samples [5]. Aminoglycosides, including tobramycin, promote mRNA

mistranslation, resulting in the incorporation of incorrect amino acids into the

elongating peptide strain [6]. Mistranslated membrane proteins play a crucial role in

the bactericidal effect of aminoglycosides. These proteins are incorporated in the

bacterial cell membrane, thereby creating non-specific channels through which small

molecules, including aminoglycosides (500 to 600 daltons), can diffuse. The increased

access of the drug results in increased ribosome inhibition and ultimately leads to cell

death [7]. In addition, by the incorporation of mistranslated proteins into the bacterial

cell membrane, an envelope stress and redox responsive two-component system is

activated, resulting in the production of ROS involved in bacterial killing [8].

The uptake of aminoglycosides in bacterial cells is an active process which can be

divided in three phases. The first phase involves the initial binding of the positively

charged aminoglycoside to the bacterial cell membrane, which is energy-independent,

followed by two energy-dependent phases, EDPI and EDPII [9-11]. In Gram-negative

bacteria, the initial binding of the aminoglycoside is mainly directed towards negatively

charged LPS, located on the outer leaflet of the outer membrane. In Gram-positive

bacteria, aminoglycosides can bind to the polar heads of the phospholipids and to

teichoic acids [12]. After initial binding, the antibiotic is transported across the outer

membrane of Gram-negative bacteria by either non-specific hydrophilic diffusion

channels formed by membrane porin proteins or by a self-promoted transport

mechanism, like has been observed for P. aeruginosa [13]. During the EDPI phase,

aminoglycosides cross the cytoplasmic membrane in response to the membrane

87

potential of the bacterial cell [14, 15]. The mechanism by which this occurs is still

unclear. Nevertheless, it has been demonstrated that this process is energy-dependent

as it can be blocked by inhibitors of the electron transport chain [10, 11] or oxidative

phosphorylation [16]. EDPI is characterized by a slow rate of uptake of aminoglycosides

preceding the inhibition of protein synthesis and bacterial killing [14]. During the EDPII

phase, aminoglycosides are transported more rapidly across the cytoplasmic membrane,

also in an energy-dependent manner. Active protein synthesis by aminoglycoside

sensitive ribosomes is essential for the initiation of and continuation of the EDPII phase.

The increased permeability of the cytoplasmic membrane caused by the incorporation

of aminoglycoside-induced mistranslated proteins could explain the necessity for active

protein synthesis for this accelerated uptake phase [17].

Apart from these three major phases in aminoglycoside uptake, fundamental differences

exist in aminoglycoside uptake between different Gram-negative and Gram-positive

bacteria. In Gram-negative bacteria, divalent cations stabilize the outer membrane by

cross linking adjacent LPS [18]. Therefore, by adding divalent cations, the outer

membrane is increasingly stabilized and less permeable to aminoglycosides which

compete for the same binding sites on LPS molecules during the energy-independent

uptake phase [19]. Donovick et al. were the first to observe an effect of different

electrolytes on the activity of streptomycin, with the greatest interference caused by

magnesium and calcium [20]. Later, it was observed that Mg2+ and Ca2+ significantly

increased MIC values of gentamicin for P. aeruginosa strains while for other Gram-

negative bacteria, including E. coli, Klebsiella pneumoniae and Serratia marcescens, this

effect was less pronounced [21, 22]. This is in agreement with data from Gray and

Wilkinson who observed a greater bactericidal effect of the divalent cation chelator,

EDTA, on P. aeruginosa compared to E. coli, suggesting that LPS are more essential for

structural integrity of the outer membrane of P. aeruginosa than for E. coli [23]. It was

hypothesized that polycationic antibiotics cross the outer membrane of different Gram-

negative bacteria by replacing Mg2+ at their binding site on LPS, resulting in increased

outer membrane permeability [13, 24], a process referred to as self-promoted uptake.

In the present study we investigated the activity of tobramycin against several bacterial

species in the presence and absence of divalent cations in order to determine whether

self-promoted uptake is involved. Based on these data, we developed tobramycin

derivatives with possible selective activity. The activity of these tobramycin derivatives

88

was tested on P. aeruginosa, E. coli, S. aureus and E. faecalis planktonic and biofilm

cultures.

89

Materials and methods

Strains

Throughout the present study, different P. aeruginosa, E. coli, S. aureus and Bcc strains

were used (Table 1, see Results section) to study tobramycin uptake. MICs and

bactericidal activity of tobramycin derivatives were determined for 4 quality control

strains recommended by EUCAST for monitoring the microbroth dilution protocol for

tobramycin MIC determinations, i.e. P. aeruginosa ATCC 27853, E. coli ATCC 25922, S.

aureus ATCC 29213 and E. faecalis ATCC 29212. All strains were obtained from the

BCCM/LMG Bacteria Collection (Ghent University, Belgium) or were kindly provided by

Dr. P. Vandamme. Isolates from endotracheal tubes were previously isolated by our

research group [25]. The bacteria were stored in Microbank tubes at -80°C and were

transferred twice on MH agar before use in any experiment.

Synthesis of tobramycin derivatives

Three new tobramycin derivatives were synthesized and evaluated in this study (Fig. 1).

Figure 1: Chemical structures of the tobramycin derivatives used in the present study.

To allow derivatisation at position 6”, tobramycin was converted to the BOC-protected

6”-azide intermediate 1 using known procedures (Scheme 1)[26]. Briefly, the sequence

commenced with BOC-protection of all amine groups followed by regiospecific (6”-OH)

sulfonylation using triisopropylbenzenesulfonyl chloride. Displacement of this sulfonate

with sodium azide at elevated temperature afforded the azide intermediate 1.

Staudinger reduction of 1 to the corresponding amine and subsequent conjugation with

the activated succinimidyl ester of 4-azidobutyric acid [27] gave azidobutyramide 2. A

second azide reduction, PyBop mediated condensation of known C-carboxy-EDTA

90

(5)[28], followed by acid mediated deprotection yielded EDTA fuctionalized tobramycin

3 as the TFA salt in 72% yield.

Since initial experiments indicated that the BODIPY fluorophore proved incompatible

with the typical acidic conditions used to remove BOC groups, we pursued a strategy in

which these protecting groups could be removed prior to introduction of the BODIPY

fluorophore. It was estimated that the copper catalyzed 3+2 azide alkyne cycloaddition

(CuAAC), in which the 6”-azido group was used as ligation handle, would fit. Thus, TFA-

mediated removal of the BOC groups of 1, followed by CuAAC with alkyne 6 [29],

afforded, after purification by preparative TLC, the tobramycin-OH-BODIPY derivative 4.

Scheme 1: Synthesis of tobramycin-OH-EDTA (3) and tobramycin-OH-BODIPY (4). Reagents and

conditions: i) Me3P, THF, H2O. ii) 4-Azidobutyric acid N-hydroxysuccinimide ester, Et3N, DMF (82%

two steps). iii) 5, PyBOP, DiPEA, DMF. iv) TFA, CH2Cl2 (3: 72% two steps 4: quant.). v) 6, CuI, DMF,

H2O, Et3N (68 %).

91

The synthesis of tobramycin-N-Me2 commenced with the regiospecific Cbz protection of

tobramycin using carbonate 9 and concomitant BOC protection of the remaining four

amino groups (Scheme 2). Hydrogenation in the presence of formaldehyde conveniently

transformed the Cbz protected amine to the dimethylamino species in one step. Acidic

deprotection yielded 6-N,N-dimethyltobramycin (8) in 68% over two steps as the TFA

salt. Additional information concerning the synthesis of the tobramycin derivatives is

added as supporting information at the end of this chapter.

Scheme 2: Synthesis of tobramycin-N-Me2 (8). Reagents and conditions: i) 9, Et3N, DMSO, water. ii)

Boc2O, Et3N, THF/H2O, 45°C (40% two steps). iii) H2, Pd/C, H2CO, HCO2H, MeOH. iv) TFA, CH2Cl2

(68% two steps).

Determination of the MIC of tobramycin and tobramycin derivatives in the presence and

absence of Mg2+

MICs were determined in duplicate according to the EUCAST standard broth

microdilution protocol [30] using unsupplemented MH for standard MIC tests. MICs

were also determined in the presence of Mg2+ at concentrations without any effect on

bacterial growth. Mg2+ was added as magnesium acetate because acetate has no

antagonistic effect on the growth inhibiting action of aminoglycosides [20]. Differences

in MIC values in the presence of Mg2+ were considered biologically relevant if the

decrease or increase observed was more than 2-fold compared to MICs determined in

the absence of Mg2+. Concentrations of all tobramycin derivatives are shown as the

quantity of underivatized tobramycin per mL of antibiotic solution.

92

Determination of the MBC of tobramycin after EDTA pretreatment

Standardized bacterial suspensions (OD=0.05; λ= 595 nm) were prepared in MH broth

and were aerobically incubated in falcon tubes at 37°C with constant agitation. After

24h of incubation, tubes were vortexed until a homogenous suspension was obtained

and centrifuged at 5000 rpm at 20°C during 5 min. The supernatant was discarded and

10 mL of physiological saline (0.9 % NaCl) (PS) was added in which the pellets were

resuspended. Suspensions were centrifuged a second time, supernatant was discarded

and pellets were resuspended in 10 mL of PS. 4 mL of the bacterial suspension was

transferred to a control tube and 4 mL was transferred to a test tube. Both falcon tubes

were centrifuged and supernatants were discarded. Pellets were resuspended in 4 mL

of Tris-HCl buffer (0.1 M, pH 8). Subsequently, 4 mL of Tris-HCl buffer supplemented

with 0.8 mM EDTA was added to the test tube. The tube was vortexed and after exactly

1 min, 100 µL of the suspension was transferred to 10 mL of PS in order to stop the

EDTA chelating reaction. This EDTA concentration and treatment duration had no

growth inhibiting or bactericidal effect on its own (data not shown). The control tube

was treated with 4 mL of unsupplemented Tris-HCl buffer during 1 min before 100 µL

was transferred to 10 mL of PS. These suspensions (∼ 5 X 105 CFU/mL) were used to

inoculate the wells of a 96 well mictrotitre plate filled with tobramycin at concentrations

ranging from 2048 mg/L to 16 mg/L for Bcc strains and ranging from 128 mg/L to 1

mg/L for P. aeruginosa, S. aureus and E. coli strains. Control wells without antibiotic

were also included. After 24h of incubation at 37°C the content of the wells was

streaked on agar plates and these were incubated for another 24h at 37 °C. CFUs were

determined and compared to control conditions. The MBC was considered as the

concentration of the antibiotic required to kill at least 99.9% of the bacteria. Differences

in MBC values were considered biologically relevant if the increase or decrease

observed was more than 2-fold compared to MBCs determined without EDTA

pretreatment.

Bactericidal effect of tobramycin and tobramycin derivatives on biofilms

Biofilms were grown in a round-bottomed 96-well microtitre plate. Wells were

inoculated with 100 µL of a standardized bacterial suspension in MH broth (108

CFU/ml). After 4h of incubation at 37°C, the supernatant containing planktonic cells

was aspirated from the wells and adhered cells were carefully rinsed with PS.

93

Subsequently, 100 µL of fresh sterile MH was added to the wells and plates were

incubated for another 20h at 37°C. After 24h of biofilm formation, biofilms were rinsed

with PS and treated with tobramycin or with a tobramycin derivative at concentrations

4XMIC in PS. Plates were incubated for an additional 24h at 37°C, followed by

determination of cell numbers by plate counting. These experiments were performed in

12 wells per condition and the content of the wells was pooled before plate counting.

Experiments were repeated three times on three separate occasions (n=3). The

significance of differences between treatment groups was determined using the Mann-

Whitney non-parametric test, with p < 0.05 considered significant.

94

Results and discussion

MICs of tobramycin in the presence and absence of Mg2+

The self-promoted uptake mechanism is associated with a well-known antagonistic

effect between divalent cations and aminoglycosides in several Gram-negative

organisms as both molecules compete for the same binding site on LPS [31]. In the

present study, we investigated the effect of Mg2+ on the activity of tobramycin against 6

P. aeruginosa strains, 2 S. aureus strains, 8 E. coli strains and 26 Bcc strains (Table 1) in

order to determine if tobramycin uptake occurs through self-promoted uptake.

Table 1: Strains used in the present study with their corresponding tobramycin MICs in the

presence and absence of Mg2+.

Strain Isolation source

Tobramycin

MIC

Tobramycin MIC in

the presence of Mg2+

Pseudomonas aeruginosa ATCC 9027 Outer ear infection 1 8-16 *

P. aeruginosa ATCC 15442 Animal room water bottle 1 8 *

P. aeruginosa PAO1 Infected wound 1-2 4 *

P. aeruginosa JWJ Endotracheal tube 2 4

P. aeruginosa LMW 1 Endotracheal tube 1-2 4 *

P. aeruginosa MPM 1 Endotracheal tube 32 32-64

Staphylococcus aureus HYL MRSA+ Endotracheal tube 2 4

S. aureus DWR 1 Endotracheal tube 1 2

Escherichia coli K12 Human faeces 1 1

E. coli ATCC 8739 Faeces 4 2

E. coli DH5α Developed in the laboratory 1-2 0.5-1

E. coli BC 1 Endotracheal tube 2 2-4

E. coli ECE Endotracheal tube 2 2-4

E. coli HYL Endotracheal tube 2 2

E. coli VA Endotracheal tube 2-4 2-4

E. coli BC 2 Endotracheal tube 2 2

95

Burkholderia cepacia LMG 1222 Onion 16-32 8

B. multivorans LMG 18822 CF (CF) patient, Canada 128 32-64

B. multivorans LMG 18825 CF patient, UK 128 64

B. multivorans LMG 13010 CF patient, Belgium 32-64 32

B. multivorans LMG 17588 Soil, USA 64 16 *

B. cenocepacia LMG 16656 CF patient, sputum, UK 256 256

B. cenocepacia LMG 18828 CF patients, Canada 128 128

B. cenocepacia LMG 18829 CF patient, USA 128 32 *

B. stabilis LMG 14086 Respirator, UK 16 16

B. vietnamiensis LMG 10929 Rhizosphere soil, Vietnam 16 4-8

B. dolosa LMG 18943 CF patient, USA 64-128 64

B. dolosa LMG 18941 CF patient, USA 256 128

B. ambifaria LMG 19182 Pea rhizosphere 16 8

B. ambifaria LMG 19467 CF patients, Australia 128 64

B. anthina LMG 20980 Soil, USA 8-16 8

B. antina LMG 20983 CF patient, sputum, UK 16 4 *

B. pyrrocinia LMG 14191 Soil, Japan 32 8-16

B. ubonensis LMG 20358 Soil, Thailand 32-64 32

B. ubonensis LMG 24263 Nosocomial infection, Thailand 64 32

B. diffusa LMG 24065 CF patient, USA 64 64

B. diffusa LMG 24266 CF patient, Canada 32-64 32

B. seminalis 24067 CF patient, USA 64 64

B. seminalis LMG 24272 Nosocomial infection, Thailand 32-64 32

B. metallica LMG 24068 CF patient, USA 64 32

B. metallica R-2712 CF patient, Canada 64 64

B. lata LMG 6992 Soil, Trinidad & Tobago 64 64

All MICs were determined in duplicate; if two different values were obtained, both are represented in the

table.

Mg2+ was added at concentrations of 25 mM to P. aeruginosa, E. coli and S. aureus cultures and at

concentrations of 6.25 mM to Bcc cultures as 25 mM had a growth inhibiting effect on Bcc strains.

*: Different MIC compared to MIC determined in the absence of Mg2+.

96

Four P. aeruginosa strains showed an increased MIC in the presence of Mg2+ compared

with control MICs, confirming the hypothesis of self-promoted uptake of

aminoglycosides in P. aeruginosa [32]. MICs of tobramycin for the E. coli strains tested

were not altered by the presence of Mg2+. Medeiros et al. previously investigated the

effect of different salt concentrations on the in vitro susceptibility of some Gram-

negative bacteria to aminoglycosides [21]. Although they observed an antagonistic

effect of 30 mM MgCl2 on gentamicin and kanamycin activity for E. coli in nutrient broth,

this could not be confirmed in MH broth. In contrast, the antagonistic effect of Mg2+ on

gentamicin and kanamycin activity for P. aeruginosa was retained both in nutrient and

MH broth, suggesting that additional environmental factors play a role in the uptake of

aminoglycosides in E. coli, e.g ionic strength (conductivity in MH broth is five times

higher than in nutrient broth). The addition of Mg2+ had a growth inhibiting effect on

Bcc strains (data not shown), therefore, Mg2+ concentrations used were 4 times lower

(6.25 mM) compared to the concentrations used for MIC determinations for P.

aeruginosa, E. coli and S. aureus (25 mM). Still, even at these lower concentrations, an

overall trend of decreased MIC values in the presence of Mg2+ was observed. This effect

was biologically relevant for B. multivorans LMG 17588, B. cenocepacia LMG 18829 and

B. anthina LMG 20983. Although it has been demonstrated that polymyxin binds to

purified Burkholderia LPS molecules at the same binding sites of Mg2+, it was not capable

of binding to whole cells [33]. Probably, the negative charges on Burkholderia LPS are

partially shielded, inhibiting binding of polycationic antibiotics. Some Gram-negative

organisms, including Bcc, modify their LPS through the addition of 4-amino-4-deoxy-L-

arabinose molecules to the lipid A residue, neutralizing the negative charge of the

phosphate groups to which polycations bind [34]. It has been shown that Bcc lipid A

molecules contain at least one 4-amino-4-deoxy-L-arabinose molecule [35], resulting in

poor binding of polycationic antibiotics to Bcc outer membrane LPS. In accordance with

previous data, our results suggest that another mechanism besides self-promoted

tobramycin uptake exists in Bcc strains [33, 36] and that Mg2+ ions probably influence

other processes involved in tobramycin susceptibility.

Because of the lack of an outer membrane, Gram-positive microorganisms are not

capable of carrying out self-promoted uptake and as was expected, no influence of Mg2+

ions on the MIC of tobramycin for S. aureus was observed.

97

MBC of tobramycin after EDTA pretreatment

If tobramycin enters the bacterial cell through self-promoted uptake, its uptake should

increase after pretreating bacterial cells with divalent cation-chelating components,

including EDTA. It is essential to perform these EDTA assays on bacteria grown in

media with sufficiently high Mg2+ concentrations as the effect of EDTA is correlated with

the quantity of Mg2+ ions in the bacterial outer membrane.

Table 2: MBC of tobramycin after EDTA pretreatment.

Strain MBC MBC after EDTA pretreatment

P. aeruginosa ATCC 9027 2 0.5*

P. aeruginosa MPM 1 > 64 32*

S. aureus HYL MRSA+ 32 32

S. aureus DWR 1 4 4

E. coli K12 64 32

E. coli HYL 32 16

B. cenocepacia LMG 16656 256 512

B. multivorans LMG 13010 128 128

*: Different MBC compared to MBC determined without EDTA pretreatment.

We observed an increased susceptibility of P. aeruginosa to tobramycin when cells were

pretreated with EDTA, confirming the hypothesis of self-promoted uptake for

tobramycin in P. aeruginosa (Table 2). No difference in MBC values, determined with or

without EDTA pretreatment, was observed for E. coli (Table 2). This observation

confirms that self-promoted uptake is not the only route for tobramycin uptake in E. coli,

and that it is likely that E. coli also uses the hydrophilic porin-protein uptake pathway

for aminoglycosides [37, 38]. Hancock et al. noticed an outer membrane permeabilizing

effect of aminoglycosides in E. coli favouring the self-promoted uptake mechanism [39].

However, they also observed a superior affinity of aminoglycosides for LPS compared to

Mg2+, indicating that Mg2+ could only compete for the aminoglycoside binding site at

sufficiently high concentrations. This could explain the lack of biologically relevant

differences in MIC and MBC values in the presence and absence of divalent cations in the

98

present study. For the Bcc strains tested no difference in MBC values determined with

or without EDTA pretreatment was observed, confirming our hypothesis that Bcc strains

are not using the self-promoted uptake pathway for tobramycin. As was expected, S.

aureus strains also showed no difference in MBC values with or without EDTA

treatment.

MICs of tobramycin derivatives

Based on previous studies we chose to modify the 6’’ primary alcohol of tobramycin

with two different substituents. Tobramycin-OH-BODIPY was synthesized for another

study on biofilm penetration of this aminoglycoside using confocal scanning microscopy.

The results obtained with pretreatment of EDTA on the activity of tobramycin against P.

aeruginosa, encouraged us to synthesize a tobramycin conjugate with this cation

chelator, tobramycin-OH-EDTA.

Aminoglycoside modifying enzymes play an important role in bacterial resistance

against aminoglycosides. Aminoglycoside acetyltransferases are known to be one of the

most important determinants to infer resistance to aminoglycosides [40]. Recently, Yan

et al. have demonstrated that acetylation at the 6’-NH2 group is sufficient to increase the

MIC to a clinically ineffective level [41]. Possibly the loss of the positive charge at

physiological conditions is at the basis of this loss of activity. This led us to synthesize a

6’-dimethylamine derivative (tobramycin-N-Me2) that is insusceptible to the

abovementioned acylation. We tested the activity of these three tobramycin derivatives

(Table 3). Tobramycin-OH-BODIPY and tobramycin-N-Me2 were the most active

derivatives, with the lowest MIC values for P. aeruginosa. According to NCCLS

breakpoints for tobramycin intravenous therapy [42], E. coli ATCC 25922, S. aureus

ATCC 29213 and E. faecalis ATCC 29212 are resistant to all of these derivatives, while P.

aeruginosa ATCC 27853 is intermediately susceptible to tobramycin-OH-BODIPY and

tobramycin-N-Me2 and is resistant to tobramycin-OH-EDTA. Surprisingly, the

tobramycin-OH-EDTA derivative shows the highest MICs for all organisms tested while

we expected that by chelating divalent cations it would be more active against P.

aeruginosa compared to other organisms in which self-promoted uptake does not occur.

In P. aeruginosa, a mechanism conferring resistance against polycations is regulated by

the PhoPQ and PmrAB two-component systems [43, 44]. Under low Mg2+ conditions

these two-component systems are activated, resulting in the addition of 4-amino-deoxy-

99

L-arabinose to lipid A which reduces the outer membrane permeability to polycations

[45]. Through the Mg2+ chelating capacity, a Mg2+-limited environment can be created

which could explain the high MIC values of tobramycin-OH-EDTA for P. aeruginosa. If E.

coli would use the hydrophilic pathway for aminoglycoside uptake one would expect a

high MIC value for bulky tobramycin derivatives, including tobramycin-OH-BODIPY and

tobramycin-OH-EDTA, as these molecules are too big to be taken up through E. coli

porins. Porins on the outer membrane of E. coli cells exclude molecules bigger than 500

to 600 daltons, approaching the size of aminoglycosides [16]. The MIC value of

tobramycin-OH-BODIPY for E. coli is not that high (32 mg/L) as would be expected

based on uptake exclusively through porins. Therefore, we hypothesize that E. coli, at

least in part, uses self-promoted uptake for aminoglycosides, including tobramycin. It

has been shown that EDTA and polycations can permeabilize the outer membrane of P.

aeruginosa to lysozyme, a protein of 14000 daltons [13]. This suggests that self-

promoted uptake can facilitate the uptake of large molecules while the hydrophilic

uptake route through outer membrane porins is dependent on porin sizes. The

exclusion limits of these porins likely only permit the uptake of smaller molecules like

tobramycin-N-Me2.

Table 3: MICs of tobramycin and tobramycin derivatives (experiments were performed in

duplicate).

Strain Tobramycin Tobramycin +

25 mM Mg2+

Tobramycin-

BODIPY-OH

Tobramycin-

NMe2

Tobramycin-

EDTA-OH

P. aeruginosa

ATCC 27853 0.5 4 8 8 512

E. coli

ATCC 25922 1 2 32 32 512

S. aureus

ATCC 29213 1 1 > 32 16 256

E. faecalis

ATCC 29212 16 32 > 32 > 32 > 512

100

Bactericidal effect of tobramycin and tobramycin derivatives on biofilms

The activity of tobramycin-OH-BODIPY and tobramycin-N-Me2 was tested on P.

aeruginosa, E. coli, S. aureus and E. faecalis biofilms at concentrations of 64 mg/L. The

tobramycin-OH-EDTA derivative was only tested on P. aeruginosa and E. coli biofilms

(Figure 2).

Figure 2: Number of CFU in biofilms after treatment with tobramycin or tobramycin derivatives.

Cell numbers of untreated biofilms are added as a reference. (Data shown are averages ± SD)

*: Significantly different from untreated biofilms (p < 0.05) (n=3)

Tobramycin at concentrations of 4 mg/L did not result in a significant bactericidal effect

for any of the strains tested. The tobramycin-OH-BODIPY derivative showed a small, but

significant bactericidal effect on S. aureus ATCC 29213 biofilms. The tobramycin-N-Me2

demonstrated a selective bactericidal effect towards P. aeruginosa ATCC 27853 biofilms.

This selective activity is likely linked to the biofilm phenotype as the MIC value of this

derivative for P. aeruginosa did not differ from the MIC for S. aureus and differed only 4

fold from the MIC for E. coli. Finally, tobramycin-OH-EDTA had approximately the same

activity against P. aeruginosa biofilms as the tobramycin-N-Me2 derivative. Surprisingly,

101

this derivative showed high activity against E. coli biofilms. Although MIC values of

tobramycin-OH-EDTA for E. coli and P. aeruginosa were the same (512 µg/mL), this

antibiotic was capable of almost totally eradicating E. coli biofilms (∼ 7 log reduction in

CFUs compared to untreated controls) while resulting only in approximately 2 log

reductions for P. aeruginosa biofilms. A possible explanation for the lower activity of

tobramycin-OH-EDTA against P. aeruginosa biofilms could be that by chelating divalent

cations, more negatively charged groups are exposed in the biofilm matrix. These

negatively charged groups, e.g. on alginate or eDNA, can interact with positively charged

tobramycin residues, delaying the penetration of the antibiotic into the biofilm [46]. It is

well-known that P. aeruginosa biofilms have large amounts of eDNA in their biofilm

matrix, and that this eDNA is capable of binding Mg2+ [47]. Less is known about the

amount of eDNA in E. coli biofilms and further research would be needed to confirm our

hypothesis. The high activity of tobramycin-OH-EDTA against E. coli biofilms implies

that this molecule is taken up into the cell. As this molecule is too large to be taken up

by porins, this supports our hypothesis that E. coli also uses self-promoted uptake for

aminoglycosides.

102

Conclusion

In the present study, we investigated the mode of uptake of tobramycin in different

Gram-negative bacteria. The self-promoted uptake route for aminoglycosides in P.

aeruginosa was confirmed. Our data also indicated the lack of self-promoted uptake in

Bcc strains. For E. coli, we hypothesized that tobramycin is taken up, at least in part,

through self-promoted uptake. We identified a tobramycin derivative (tobramycin-N-

Me2) with selective activity against P. aeruginosa biofilms and one (tobramycin-OH-

EDTA) with high activity towards E. coli biofilms. The use of selective antibiotics could

have major advantages in the clinic. By selectively targeting some or only one

pathogenic species, the commensal flora of the patient will be less disturbed compared

to traditional antibiotic therapy. This will reduce the number and frequency of side

effects associated with antibiotic therapy, including gastro-intestinal complaints. In

addition, the selection for antibiotic-resistant organisms during therapy will diminish.

Acknowledgements

This research has been funded by the Interuniversity Attraction Poles Programme

initiated by the Belgian Science Policy Office and by a concerted action grant by Ghent

University.We thank Petra Rigole and Tine Vandermeulen for their excellent technical

assistance.

103

References

1. O'Sullivan, B.P. and S.D. Freedman, CF. Lancet, 2009. 373(9678): p. 1891-904. 2. Brennan, A.L. and D.M. Geddes, CF. Curr Opin Infect Dis, 2002. 15(2): p. 175-82. 3. Matsui, H., et al., Reduced three-dimensional motility in dehydrated airway mucus

prevents neutrophil capture and killing bacteria on airway epithelial surfaces. J Immunol, 2005. 175(2): p. 1090-9.

4. Lyczak, J.B., C.L. Cannon, and G.B. Pier, Lung Infections Associated with CF. Clin Microbiol Rev, 2002. 15(2): p. 194-222.

5. Ratjen, F., F. Brockhaus, and G. Angyalosi, Aminoglycoside therapy against Pseudomonas aeruginosa in CF: a review. J Cyst Fibros, 2009. 8(6): p. 361-9.

6. Davies, J., L. Gorini, and B.D. Davis, Misreading of RNA codewords induced by aminoglycoside antibiotics. Mol Pharmacol, 1965. 1(1): p. 93-106.

7. Davis, B.D., L.L. Chen, and P.C. Tai, Misread protein creates membrane channels: an essential step in the bactericidal action of aminoglycosides. Proc Natl Acad Sci U S A, 1986. 83(16): p. 6164-8.

8. Kohanski, M.A., D.J. Dwyer, and J.J. Collins, How antibiotics kill bacteria: from targets to networks. Nat Rev Microbiol, 2010. 8(6): p. 423-35.

9. Andry, K. and R.C. Bockrath, Dihydrostreptomycin accumulation in E. coli. Nature, 1974. 251(5475): p. 534-6.

10. Miller, M.H., et al., Gentamicin uptake in wild-type and aminoglycoside-resistant small-colony mutants of Staphylococcus aureus. Antimicrob Agents Chemother, 1980. 18(5): p. 722-9.

11. Bryan, L.E. and H.M. Van den Elzen, Streptomycin accumulation in susceptible and resistant strains of Escherichia coli and Pseudomonas aeruginosa. Antimicrob Agents Chemother, 1976. 9(6): p. 928-38.

12. Kusser, W., K. Zimmer, and F. Fiedler, Characteristics of the binding of aminoglycoside antibiotics to teichoic acids. A potential model system for interaction of aminoglycosides with polyanions. Eur J Biochem, 1985. 151(3): p. 601-5.

13. Hancock, R.E., V.J. Raffle, and T.I. Nicas, Involvement of the outer membrane in gentamicin and streptomycin uptake and killing in Pseudomonas aeruginosa. Antimicrob Agents Chemother, 1981. 19(5): p. 777-85.

14. Bryan, L.E. and H.M. Van Den Elzen, Effects of membrane-energy mutations and cations on streptomycin and gentamicin accumulation by bacteria: a model for entry of streptomycin and gentamicin in susceptible and resistant bacteria. Antimicrob Agents Chemother, 1977. 12(2): p. 163-77.

15. Damper, P.D. and W. Epstein, Role of the membrane potential in bacterial resistance to aminoglycoside antibiotics. Antimicrob Agents Chemother, 1981. 20(6): p. 803-8.

16. Taber, H.W., et al., Bacterial uptake of aminoglycoside antibiotics. Microbiol Rev, 1987. 51(4): p. 439-57.

17. Hurwitz, C., C.B. Braun, and C.L. Rosano, Role of ribosome recycling in uptake of dihydrostreptomycin by sensitive and resistant Escherichia coli. Biochim Biophys Acta, 1981. 652(1): p. 168-76.

18. Page, M.G., The role of the outer membrane of Gram-negative bacteria in antibiotic resistance: Ajax' shield or Achilles' heel? Handb Exp Pharmacol, 2012(211): p. 67-86.

104

19. Moore, R.A., N.C. Bates, and R.E. Hancock, Interaction of polycationic antibiotics with Pseudomonas aeruginosa lipopolysaccharide and lipid A studied by using dansyl-polymyxin. Antimicrob Agents Chemother, 1986. 29(3): p. 496-500.

20. Donovick, R., A.P. Bayan, and et al., The influence of certain substances on the activity of streptomycin; differential effects of various electrolytes on the action of streptomycin. J Bacteriol, 1948. 56(1): p. 125-37.

21. Medeiros, A.A., et al., Effect of salt concentration on the apparent in-vitro susceptibility of Pseudomonas and other gram-negative bacilli to gentamicin. J Infect Dis, 1971. 124 Suppl: p. S59-64.

22. Gilbert, D.N., et al., Effect of the concentrations of magnesium and calcium on the in-vitro susceptibility of Pseudomonas aeruginosa to gentamicin. J Infect Dis, 1971. 124 Suppl: p. S37-45.

23. Gray, G.W. and S.G. Wilkinson, The effect of ethylenediaminetetra-acetic acid on the cell walls of some gram-negative bacteria. J Gen Microbiol, 1965. 39(3): p. 385-99.

24. Nicas, T.I. and R.E. Hancock, Outer membrane protein H1 of Pseudomonas aeruginosa: involvement in adaptive and mutational resistance to ethylenediaminetetraacetate, polymyxin B, and gentamicin. J Bacteriol, 1980. 143(2): p. 872-8.

25. Vandecandelaere, I., et al., Assessment of microbial diversity in biofilms recovered from endotracheal tubes using culture dependent and independent approaches. PLoS One, 2012. 7(6): p. e38401.

26. Disney, M.D. and O.J. Barrett, An aminoglycoside microarray platform for directly monitoring and studying antibiotic resistance. Biochemistry, 2007. 46(40): p. 11223-30.

27. Tsitovich, P.B., et al., A chemoenzymatic route to diversify aminoglycosides enables a microarray-based method to probe acetyltransferase activity. Chembiochem, 2010. 11(12): p. 1656-60.

28. Chiu, F.C.K., et al., 9-aminoacridine-EDTA conjugates as hydroxy radical footprinting reagents with no intrinsic cutting specificity. Bioorg Med Chem Lett, 1995. 5(15): p. 1689-1694.

29. Verdoes, M., et al., Acetylene functionalized BODIPY dyes and their application in the synthesis of activity based proteasome probes. Bioorg Med Chem Lett, 2007. 17(22): p. 6169-71.

30. European Committee for Antimicrobial Susceptibility Testing of the European Society of Clinical, M., Determination of minimum inhibitory concentrations (MICs) of antibacterial agents by broth dilution.

31. Zimelis, V.M. and G.G. Jackson, Activity of aminoglycoside antibiotics aganst Pseudomonas aeruginosa: specificity and site of calcium and magnesium antagonism. J Infect Dis, 1973. 127(6): p. 663-9.

32. Hancock, R.E., Alterations in outer membrane permeability. Annu Rev Microbiol, 1984. 38: p. 237-64.

33. Moore, R.A. and R.E. Hancock, Involvement of outer membrane of Pseudomonas cepacia in aminoglycoside and polymyxin resistance. Antimicrob Agents Chemother, 1986. 30(6): p. 923-6.

34. Raetz, C.R., et al., Lipid A modification systems in gram-negative bacteria. Annu Rev Biochem, 2007. 76: p. 295-329.

35. De Soyza, A., et al., Chemical and biological features of Burkholderia cepacia complex lipopolysaccharides. Innate Immun, 2008. 14(3): p. 127-44.

105

36. Jassem, A.N., et al., In vitro susceptibility of Burkholderia vietnamiensis to aminoglycosides. Antimicrob Agents Chemother, 2011. 55(5): p. 2256-64.

37. Nakae, R. and T. Nakae, Diffusion of aminoglycoside antibiotics across the outer membrane of Escherichia coli. Antimicrob Agents Chemother, 1982. 22(4): p. 554-9.

38. Foulds, J. and T.J. Chai, New major outer membrane proteins found in an Escherichia coli tolF mutant resistant to bacteriophage TuIb. J Bacteriol, 1978. 133(3): p. 1478-83.

39. Hancock, R.E., et al., Interaction of aminoglycosides with the outer membranes and purified lipopolysaccharide and OmpF porin of Escherichia coli. Antimicrob Agents Chemother, 1991. 35(7): p. 1309-14.

40. Wright, G.D. and P. Ladak, Overexpression and characterization of the chromosomal aminoglycoside 6'-N-acetyltransferase from Enterococcus faecium. Antimicrob Agents Chemother, 1997. 41(5): p. 956-60.

41. Yan, X., et al., The use of aminoglycoside derivatives to study the mechanism of aminoglycoside 6'-N-acetyltransferase and the role of 6'-NH2 in antibacterial activity. Bioorg Med Chem, 2007. 15(8): p. 2944-51.

42. Performance standards for antimicrobial susceptibility testing; 15th informational supplement, in CLSI/NCCLS M100-S15. 2005, Clinical and Laboratory Standards Institute: Wayne.

43. Macfarlane, E.L., A. Kwasnicka, and R.E. Hancock, Role of Pseudomonas aeruginosa PhoP-phoQ in resistance to antimicrobial cationic peptides and aminoglycosides. Microbiology, 2000. 146 ( Pt 10): p. 2543-54.

44. Mulcahy, H., L. Charron-Mazenod, and S. Lewenza, Extracellular DNA Chelates Cations and Induces Antibiotic Resistance in Pseudomonas aeruginosa Biofilms. PLoS Pathog, 2008. 4(11).

45. Moskowitz, S.M., R.K. Ernst, and S.I. Miller, PmrAB, a two-component regulatory system of Pseudomonas aeruginosa that modulates resistance to cationic antimicrobial peptides and addition of aminoarabinose to lipid A. J Bacteriol, 2004. 186(2): p. 575-9.

46. Tseng, B.S., et al., The extracellular matrix protects Pseudomonas aeruginosa biofilms by limiting the penetration of tobramycin. Environ Microbiol, 2013.

47. Lewenza, S., et al., Quorum sensing in Burkholderia cepacia: identification of the LuxRI homologs CepRI. J Bacteriol, 1999. 181(3): p. 748-56.

106

Supporting information

General experimental organic synthesis

All reactions described were performed under an argon atmosphere and at ambient

temperature unless stated otherwise. Tobramycin was purchased from TCI Europe All

other reagents and solvents were purchased from Sigma, Aldrich, Acros Organics used

as received. Reactions were monitored by TLC analysis using TLC aluminum sheets

(Macherey-Nagel, Alugram Sil G/UV254) with detection by spraying with a solution of

(NH4)6Mo7O24•4H2O (25 g/L) and (NH4)4Ce(SO4)4•2H2O (10 g/L) in H2SO4 (10%)

followed by charring or Ninhydrin (2% in ethanol). Column chromatography was

performed on Biosolve 60Å silica gel (32-63 μm). High resolution mass spectra were

recorded with a Waters LCT Premier XE Mass spectrometer. 1H-and 13C-NMR spectra

were measured on a Varian Mercury-300BB (300/75 MHz), an Avance II Bruker

(700/176 MHz) spectrometer equipped with a 1H/13C/15N TXI-Z probe or an Avance III

Bruker spectrometer operating at a 1H frequency of 500 MHz, equipped with a BBI

probe. Chemical shifts are given in ppm (δ) relative to tetramethylsilane as an internal

standard (1H NMR) or CDCl3, CD3OD or SO(CD3)2 (13C NMR). Coupling constants are

given in Hz.

To a solution of 6”-azido-(penta-NHBoc)-tobramycin (1)[1] (0.50 mmol, 497 mg) in THF

(5 mL) is added Me3P (5.0 mL, 1.0M in THF, 10eq.). After stirring for 4h, the solution is

concentrated in vacuo and the residue is dissolved in water (5 mL). After 1h the solution

is taken to dryness, coevaporated thrice with (ClCH2)2 (2 mL) and taken up in DMF (5

mL). 4-Azidobutyric acid N-hydroxysuccinimide ester [2] (0.7 mmol, 159 mg) and

triethylamine (200 µl) are added and the solution is stirred overnight. The reaction is

taken to dryness and purified via silica gel chromatography (0�15% MeOH in CH2Cl2).

Product fractions are collected concentrated and taken up in CH2Cl2 and precipitated

107

with cold hexane yielding the product (2) (442 mg, 0.41 mmol, 82%) as an off white

amorphous solid. (ESI-MS for C47H84N9O19 [M+H] found, 1078.5933 ; calcd, 1078.5883)

DMSO-d6 1H

700/176

MHz 13C

1H

13C

Ring A Ring C 1 4.97 98.0 1 4.93 98.3 2 3.50 50.0 2 3.26 70.8 3 1.89 33.6 3 3.52 55.6 1.51 4 3.00 68.8 4 3.36 65.1 5 3.92 71.7 5 3.55 71.8 6 3.40 40.3 6 3.29 41.8 3.19 3.24 Linker Ring B 1 - 173.0 1 3.44 81.8 2 2.21 32.4 2 3.49 75.2 3 1.76 24.9 3 3.39 82.2 4 3.34 50.8 4 3.39 49.8 5 1.83 35.2 1.43 6 3.37 50.2

To a solution of 4-Azidobutyryl tobramycin (2) (400 mg, 0.37 mmol) in THF (10 mL) is

added Me3P (3.0 mL, 1.0 M, 3.0 mmol) and the mixture is stirred for 4h. Water (1 ml) is

added and stirring is continued for 1h. The mixture ins concentrated in vacuo and

coevaporated twice with (ClCH2)2 (3 mL). The residue is dissolved in DMF (20 mL). To

the solution are added EDTA-tetra-tert-butylester carboxylic acid [3] (280 mg, 0.5

mmol), PyBop (234 mg, 0.45 mmol) and triethylamine (300 µl). The mixture is stirred

overnight, concentrated in vacuo and the residue is purified on silica gel

chromatography (0�15% MeOH in CH2Cl2) to yeld the fully protected derivative (ESI-

MS for C74H132N9O28 [M+H] found, 1594,9191 ; calcd, 1594,9176). The material is taken

up in a mixture of CH2Cl2 (4 mL) and TFA (4mL) for 90 minutes and concentrated in

108

vacuo yielding the deprotected product (3) (444 mg, 0.27 mmol, 72%) as a beige

amorphous powder. (ESI-MS for C33H60N9O18 [M+H] found, 870.4066 ; calcd, 870.4056)

DMSO-d6 1H

700/176

MHz 13C

1H

13C

Ring A Ring C 1 5.81 91.9 1 4.96 99.8 2 3.41 47.5 2 3.70 68.1 3 1.97 29.9 3 3.10 54.3 2.03 4 3.31 66.1 4 3.40 64.3 5 3.78 70.7 5 3.74 69.7 6 3.79 54.2 6 3.24 39.7 Linked EDTA 2.92 1 - 173.3 Ring B 2 2.14 32.3 1 3.55 83.3 3 1.63 24.97 2 3.63 73.7 4 3.04 38.1 3 3.79 75.7 5 - 170.0 4 3.28 48.0 6 3.67 60.7 5 2.40 27.4 7 3.19 53.7 1.75 8 3.38 38.9 6 3.39 49.4 3.03 9 3.56 52.6 3.46

6”-azido-(penta-NHBoc)-tobramycin (149 mg, 0.15 mmol) was taken up in a mixture of

TFA and CH2Cl2 (5mL, 1:1 v/v) and stirred for 2h. The solution was taken to dryness and

dissolved in DMF (2.5 mL). Added are CuI (29 mg, 0.15 mmol, 1 eq), Et3N (50 µL) and

BODIPY alkyne 6 [4] (50 mg, 0.15 mmol, 1 eq). The reaction mixture was stirred for 4h

at room temperature and concentrated in vacuo. Purification with preparative TLC

(CH2Cl2/MeOH/NH4OH) afforded the desired fluorescent triazole conjugate (4) (102 mg,

0.10 mmol 68 %). ESI-MS for C37H59BF2N10O8 [M+H] found, 821.4665; calcd, 821.4651.

109

DMSO-d6 1H

700/176

MHz 13C

1H

13C

Ring A Ring C 1 4.80 100.36 1 4.94 99.53 2 2.75 49.26 2 3.06 72.08 3 1.42 37.12 3 2.87 55.07 1.82 4 2.80 70.71 4 3.17 65.97 5 4.21 69.69 5 3.44 73.27 6 4.39 50.35 6 2.57 42.15 4.49 2.86 BODIPY Ring B 1/12 2.38 13.91 1 2.98 86.08 2/4/9/11 - 140.7 2 3.29 74.06 5/13 2.37 15.82 3 2.92 89.7 3/10 6.21 121.43 4 2.54 49.65 15 1.63 30.54 5 0.99 38.39 16 1.81 29.18 1.77 17 2.68 24.49 6 2.56 51.44 14 2.95 27.43 19 7.86 122.78 6/8 - 130.6 7 - 152.9 18 - 146.8

Tobramycin (1.87 g, 4 mmol) is taken up in a mixture of DMSO (200 mL) and water (20

mL). To this solution are added: N-Benzyloxycarbonyloxy-5-norbornene-2,3-

dicarboximide [5] (1.253 g, 4.0 mmol) and triethylamine (600 µl). The mixture is stirred

overnight and concentrated in vacuo. The residue is purified by silica gel

chromatography (CH2CH2/MeOH/NH4OH, 5/4/1, v/v/v). The monoprotected

aminoglycoside (950 mg, 1.57 mmol) is dissolved in a mixture of THF (70 mL) and water

(70mL). Next di-tert-butylpyrocarbonate (1.75 g , 8.0 mmol) and triethylamine (500µL)

are added and the mixture is heated to 45°C for 16 h. The solution in concentrated

vigorously and the residue is precipitated from a mixture of dichloromethane and

hexanes to yield the protected tobramycin (7) (1.6 g, 1.57 mmol, 40% based on

tobramycin) as an off white amorphous powder. ESI-MS for C46H76N5O19 [M+H] found,

1002,5151; calcd, 1002,5129.

110

DMSO-d6 1H

700/176

MHz 13C

1H

13C

Ring A Ring C 1 4.92 98.0 1 4.92 98.3 2 3.49 50.0 2 3.28 70.7 3 1.88 33.7 3 3.52 56.2 1.51 4 3.28 67.5 4 3.36 64.9 5 3.76 73.6 5 3.56 71.8 6 3.55 60.6 6 3.43 42.2 3.26 Cbz Ring B C=O 157.1 1 3.41 82.3 CH2 5.04 65.7 2 3.48 75.2 Ph 7.32 128.2 3 3.38 82.2 7.37 128.8 4 3.38 49.7 7.37 128.2 5 1.85 35.3 1.38 6 3.36 50.0

Protected tobramycin 7 (501 mg, 0,5 mmol) is taken up in methanol (50 mL) and the

vessel is purged with nitrogen. To the solution are added formaldehyde (1 mL, 38% aq.),

formic acid (1 mL) and Palladium on charcoal (50 mg, 10% Pd). Hydrogen is bubbled

through the black suspension for 2 h. The mixture is filtered taken to dryness and the

residue is purified by silica gel chromatography (CH2Cl2/MeOH/Et3N 40:10:0.1 v/v/v)

to yield the protected dimethylamino species. (ESI-MS for C40H74N5O17 [M+H] found,

896.5109; calcd, 896.5080). The material is taken up in CH2Cl2 (5 mL) and TFA (5mL)

for 1h and taken to dryness. The residue is lyophilized (tBuOH/H2O) to yield the product

(8) (362 mg, 0.34 mmol, 68%) as a beige amorphous solid. ESI-MS for C20H42N5O9

[M+H] found, 496,2967; calcd, 496,2977.

111

D2O 1H

700/176

MHz 13C

1H

13C

Ring A Ring C 1 5.72 93.8 1 5.03 100.7 2 3.61 47.5 2 3.88 67.9 3 2.21 29.0 3 3.41 54.7 1.97 29.0 4 3.63 65.3 4 3.69 64.7 5 3.84 73.0 5 4.06 69.6 6 3.78 59.7 6 3.46 58.2 3.71 59.7 NMe2 Ring B Mea 2.90 43.9 1 3.71 83.7 Meb 2.93 45.1 2 3.80 74.0 3 3.92 77.4 4 3.49 48.3 5 2.48 27.7 1.88 27.7 6 3.53 49.5

References

1. Disney, M.D. and O.J. Barrett, An aminoglycoside microarray platform for directly

monitoring and studying antibiotic resistance. Biochemistry, 2007. 46(40): p. 11223-30.

2. Tsitovich, P.B., et al., A chemoenzymatic route to diversify aminoglycosides enables a microarray-based method to probe acetyltransferase activity. Chembiochem, 2010. 11(12): p. 1656-60.

3. Chiu, F.C.K., et al., 9-aminoacridine-EDTA conjugates as hydroxy radical footprinting reagents with no intrinsic cutting specificity. Bioorg Med Chem Lett, 1995. 5(15): p. 1689-1694.

4. Verdoes, M., et al., Acetylene functionalized BODIPY dyes and their application in the synthesis of activity based proteasome probes. Bioorg Med Chem Lett, 2007. 17(22): p. 6169-71.

5. Gao, F., et al., Synthesis and use of sulfonamide-, sulfoxide-, or sulfone-containing aminoglycoside-CoA bisubstrates as mechanistic probes for aminoglycoside N-6'-acetyltransferase. Bioorg Med Chem Lett, 2008. 18(20): p. 5518-22.

112

Chapter VI: INVESTIGATING THE ROLE OF MATRIX COMPONENTS IN

PROTECTION OF BURKHOLDERIA CEPACIA COMPLEX BIOFILMS

AGAINST TOBRAMYCIN

Redrafted from:

Anne-Sophie Messiaen, Hans Nelis, Tom Coenye. (2013). Investigating the role of

matrix components in protection of burkholderia cepacia complex biofilms against

tobramycin. J Cyst Fibros, 2013 Aug 8.

113

ABSTRACT

Bcc organisms produce a wide variety of potential virulence factors, including

exopolysaccharides, and exhibit intrinsic resistance towards many antibiotics. In the

present study we investigated the contribution of Bcc biofilm matrix components,

including eDNA, cepacian and PNAG, to tobramycin susceptibility. To this end, the in

vitro bactericidal activity of tobramycin in combination with rhDNase, NaClO and

dispersin B was tested against Bcc biofilms. Exopolysaccharide degradation by NaClO

pretreatment and specific PNAG degradation by dispersin B significantly increased the

bactericidal effect of tobramycin towards some of the Bcc biofilms tested, including

strains of B. cenocepacia, B. cepacia and B. metallica. The presence of rhDNase during

biofilm treatment and/or development had no influence on tobramycin activity. These

results suggest that exopolysaccharides play a role in tobramycin susceptibility of Bcc

biofilms and that matrix degrading combination therapy could improve treatment of Bcc

biofilm infections.

114

Introduction

Chronic P. aeruginosa lung infections are the primary cause of morbidity and mortality

in CF patients [1]. Although P. aeruginosa is the dominant pathogen isolated from the

chronically infected CF lung, infections with Bcc species often lead to a more rapid

decline in lung function, resulting in increased mortality [2]. Chronic Bcc infections are

very difficult to eradicate because of the capacity of Bcc species to cause invasive disease

and their high level of intrinsic antibiotic resistance. The production of different

exopolymeric substances by mucoid Bcc bacteria provides additional protection against

both antibiotics and host immune components [3]. Exopolysaccharides produced by

Burkholderia cenocepacia have been shown to inhibit neutrophil chemotaxis and to

reduce the levels of neutrophil derived ROS [4]. This reduction in ROS levels is caused

by the ROS-scavenging capacities of these exopolysaccharides [4]. Cuzzi et al. observed

a protective effect of cepacian, a Bcc specific exopolysaccharide, against hypochlorite

(ClO-), a highly bactericidal ROS produced by neutrophils. This protective effect was

also provided by scavenging, resulting in extensive deacetylation of cepacian followed

by degradation of the polysaccharide backbone [5]. Although cepacian is the most

common exopolysaccharide produced by Bcc species, the production of at least six other

polysaccharides, either synthesized alone or in combination with cepacian, has been

demonstrated [6]. B. vietnamiensis LMG 10929, used in this study, has been shown not

to produce cepacian as main exopolysaccharide but a polysaccharide containing a fucose

residue in its backbone. In addition, the production of PNAG, also known as PIA, an

important virulence factor and major constituent of staphylococcal biofilms, has been

documented in several Bcc biofilms [7]. Although it appears to be a minor component

compared to cepacian it seems essential for initiating biofilm formation as well as

maintaining biofilm integrity. Dispersin B, an enzyme capable of degrading PNAG, was

first isolated from the periodontal pathogen Actinobacillus actinomycetemcomitans [8].

Since its discovery in 2004, dispersin B is widely investigated as a biofilm removing

agent and as a possible coating agent to prevent biofilm formation primarily focusing on

Staphylococcus epidermidis biofilms [9-11]. Apart from exopolysaccharides, another

important constituent of the biofilm matrix is eDNA. Most research about the

contribution of eDNA to biofilm formation and resistance has focused on P. aeruginosa in

which eDNA appears to be essential for the formation and stabilization of a complex

biofilm structure [12]. Witchurch et al. first demonstrated a reduction in P. aeruginosa

115

biofilm formation in the presence of DNase. Subsequently, the contribution of eDNA to

biofilm formation of a variety of bacterial species, including S. epidermidis and S. aureus,

was investigated [13-15]. In P. aeruginosa and Salmonella enterica biofilms the

contribution of eDNA to aminoglycoside tolerance has already been shown. By binding

divalent cations, a cation-limited environment is created which in turn induces the

expression of pmr genes conferring resistance towards aminoglycosides [16, 17]. In

addition, Chiang et al. observed a shielding effect of eDNA in P. aeruginosa biofilms,

mediating a delayed penetration of aminoglycosides [18]. Still, little is known about the

role of eDNA in Burkholderia biofilms, although it seems to be essential for cellular

aggregation in Burkholderia thailandensis [19].

The aim of the present study was to investigate whether Bcc biofilm matrix components,

including eDNA, cepacian and PNAG play a role in the protection of biofilm cells against

tobramycin. To this end, the bactericidal effect of tobramycin in combination with

matrix degrading components (rhDNase, NaClO and dispersin B) was tested against

several in vitro grown Bcc biofilms and compared with the bactericidal effect of

tobramycin alone.

116

Materials and methods

Bcc Strains

Strains used (Table 1) were obtained from the BCCM/LMG Bacteria Collection (Ghent,

Belgium) or were kindly provided by Dr. P. Vandamme (Ghent University, Belgium). The

bacteria were stored in Microbank tubes (Prolab Diagnostics, Richmond Hill, ON,

Canada) at -80°C and transferred twice on MH (Lab M, Heywood, UK) agar plates before

use in any experiment. All cultivations were performed under aerobic conditions at

37°C.

Table 1: Bcc strains used in the present study.

Strain Isolation source Tobramycin MIC (µg/mL)

Burkholderia cenocepacia LMG 16656 CF patient 256

Burkholderia multivorans LMG 13010 CF patient 64

Burkholderia cepacia LMG 1222 Onion 32

Burkholderia anthina LMG 20980 Soil 16

Burkholderia vietnamiensis LMG 10929 Soil 16

Burkholderia metallica LMG 24068 CF patient 64

Reagents

A highly purified solution of rhDNase (1.0 mg/mL) (Pulmozyme, Genentech Inc., San

Francisco, CA, USA) which selectively degrades DNA, was obtained from the pharmacy of

the Ghent University hospital (Ghent, Belgium). It was stored between 2-8°C, protected

from light. A NaClO solution (4–4.99% available chlorine), was purchased from Sigma-

Aldrich (St. Louis, MO, USA) and stored between 2-8°C protected from light. Lyophilized

dispersin B was obtained from Kane Biotech Inc. (Winnipeg, MB, Canada) and was

dissolved in equal volumes of glycerol and 100 mM phosphate buffer pH 5.9,

supplemented with 200 mM NaCl. The dispersin B solution was stored at -20°C.

Biofilm formation

An inoculum was prepared by suspending bacteria from a pure culture on MH agar in

MH broth. The inoculum was standardized at ∼108 CFU/mL and the wells of a round-

bottomed 96-well microtitre plate (TPP, Trasadingen, Switzerland) were filled with 100

117

µL of this standardized bacterial suspension. Plates were incubated at 37°C for 4h

without shaking. After 4h, the supernatant, containing planktonic cells, was aspirated

from the wells. Adhered cells were carefully washed with 100 µL of PS and 100 µL of

fresh sterile MH broth was added. Plates were incubated for another 20h at 37°C

without shaking.

Quantification of eDNA in the biofilm matrix

Eppendorf protein LoBind microcentrifuge tubes (1.5 mL) (Eppendorf AG, Hamburg,

Germany) were inoculated with 600 µL of a standardized bacterial solution (∼108

CFU/mL) in cation-supplemented MH broth (0.015 % (w/v) CaCl2; 2.0 mM MgCl2) (csMH

broth), with or without the addition of rhDNase (50 µg/mL). Experiments were

performed in csMH broth as Ca2+ and Mg2+ are essential for DNase I activity [20]. Tubes

were incubated aerobically for 4h at 37°C without shaking. After 4h, the supernatant

was gently poured from the tubes to remove non-adhered planktonic cells. Tubes were

rinsed with PS followed by the addition of 600 µL of fresh sterile csMH broth, with or

without 50 µg/mL rhDNase. After 20h of additional incubation, a biofilm had formed on

the walls of the tubes (Figure 1) and these were gently rinsed with PS without

disturbing the adherent film. Subsequently, 300 µL of DNA rehydration solution

(Promega, Madison, WI, USA) was added and biofilms were dispersed in this solvent by

vortexing the tubes.

Figure 1: Determination of cell numbers in Burkholderia biofilms in the presence (grey bars) and

absence (white bars) of rhDNase when grown in centrifuge tubes. (Data shown are averages ± SD)

118

Biofilm cells were separated from the matrix by centrifugation at 5000g for 10 min at

4°C. The supernatant was aspirated and filtered through a 0.2 µm cellulose acetate filter

(Whatman GmbH, Dassel, Germany) and the amount of eDNA was quantified using the

Quantifluor dsDNA System kit (Promega, Madison, WI, USA) (The manufacturer claims

that the Quantifluor dye is an intercalating DNA dye which usually does not bind to DNA

fragments smaller than 5 bp). It can not entirely be excluded that a small amount of

eDNA is remaining on the filter after filtration of the supernatant. eDNA concentrations

were normalized to the number of biofilm cells, determined by plate counting after 24h

of biofilm growth.

Determination of tobramycin MICs

MICs were determined in duplo according to the EUCAST broth microdilution protocol

using unsupplemented MH broth. In brief, bacterial suspensions were standardized at

∼5 × 105 CFU/mL before inoculating a flat-bottomed 96-well plate filled with 100 µL of

tobramycin serial dilutions in MH. The range of tobramycin concentrations used was

from 2 to 1024 mg/L. Plates were incubated at 37°C for 20 h and the optical density was

determined at 590 nm using a multilabel microtitre plate reader (Envision, Perkin Elmer

LAS, Waltham, MA, USA). The MIC is the lowest concentration of antibiotic for which a

similar optical density was observed in the inoculated and blank wells.

Effect of rhDNase on tobramycin susceptibility of 24h-old Bcc biofilms

Biofilm formation was carried out in csMH broth with or without 50 µg/mL of rhDNase.

After 24h, the supernatant was discarded and biofilms were washed with PS.

Subsequently, 100 µL of tobramycin in cation adjusted PS (0.9% NaCl; 0.015% CaCl2)

with or without 50 µg/mL of rhDNase was added. A tobramycin concentration of 4XMIC

was used as previous research from our group indicated that these concentrations

resulted in a substantial bactericidal effect against Bcc biofilms [21]. After 24h of

additional incubation at 37°C, cell numbers were determined by plate counting.

Effect of NaClO on tobramycin susceptibility of 24h-old Bcc biofilms

Biofilm formation was carried out as described above. After 24h, the supernatant was

aspirated from the wells and biofilms were washed with PS. Before treatment with

tobramycin (4XMIC in PS), biofilms were exposed to a 15 min pretreatment with NaClO

(60 µM) (pH 7-8) in phosphate buffered saline (PBS; pH 7) or with non-supplemented

119

PBS. The NaClO concentration used was based on data published by Cuzzi et al. [5].

Subsequently, biofilms were incubated for another 24h, followed by the determination

of cell numbers by plate counting.

Effect of dispersin B on tobramycin susceptibility of 24h-old Bcc biofilms

After 24h of biofilm formation, biofilms were washed with PS and treated either with

phosphate buffer (100 mM, pH 5.9), with phosphate buffer supplemented with

tobramycin (4XMIC) (pH 7) or with phosphate buffer supplemented with tobramycin

(4XMIC) and dispersin B (final concentration of 100 µg/mL) (pH 7). Plates were

incubated for an additional 24h, followed by determination of cell numbers by plate

counting.

Statistical analyses

Microtitre plate biofilm assays were performed in 12 wells per condition and their

content was pooled before plate counting. All experiments were repeated three times

on three separate occasions (n=3). Data are represented as means ± SD of the mean.

The significance of differences between treatment groups was determined using the

Mann-Whitney non-parametric test, with p < 0.05 considered as significant.

120

Results

Quantification of eDNA in biofilm matrix

Biofilms were grown on the walls of eppendorf tubes for 24h in the presence or absence

of rhDNase (50 µg/mL) (Figure 1). Treatment with rhDNase alone did not lead to

dispersion of the biofilm, as no difference in cell numbers was observed when biofilms

were grown in the presence or absence of rhDNase. The eDNA content of the biofilm

differs considerably between the six Bcc strains investigated, with B. multivorans LMG

13010 showing the highest amounts (Figure 2). By adding rhDNase to the growth

medium, a significant reduction in the amount of eDNA was observed for B. multivorans

LMG 13010, B. anthina LMG 20980, B. cepacia LMG 1222 and B. cenocepacia LMG 16656

(Figure 2). No such difference was observed after rhDNase treatment for biofilms of B.

metallica LMG 24068 and B. vietnamiensis LMG 10929.

Figure 2: Quantification of eDNA in 24h-old Burkholderia biofilms in the presence and absence of

rhDNase. (Data shown are averages ± SD)

The amount of eDNA was normalized for the number of cells adhered to the tube walls. *: significantly different compared to biofilms grown in the absence of rhDNase (p < 0.05).

121

Figure 3: Average number of biofilm cells after 24h of growth and 24h of treatment in the presence (grey bars) and absence (white bars) of a matrix

degrading component. (Data shown are averages ± SD)

Medium-only biofilms were grown in MH (NaClO and dispersin B assay) or csMH (rhDNase assay) and treated with PS (rhDNase and NaClO assay) or phosphate

buffer (dispersin B assay). For the NaClO assay, medium-only biofilms were pretreated with PBS for 15 min before treatment with PS. To determine the

bactericidal activity of the matrix degrading components, rhDNase was added at concentrations of 50 µg/mL to csMH (this was used for growing biofilms) and to

cation-supplemented PS (0.9% NaCl; 0.015% CaCl2) (this was used for treating the biofilms), NaClO was added (final concentration of 60 µM in PBS buffer) for 15

min before treatment with PS and Dispersin B was added at concentrations of 100 µg/mL to phosphate buffer.

122

Effect of rhDNase on tobramycin susceptibility of 24h-old Bcc biofilms

rhDNase (50 µg/mL) alone had no effect on the viability or dispersion of Bcc biofilms

(Figure 3). The bactericidal activity of tobramycin was tested in three experimental

conditions. Firstly, tobramycin (4XMIC) was added to a biofilm formed in the presence

of 50 µg/mL rhDNase. Secondly, tobramycin (4XMIC) in combination with 50 µg/mL

rhDNase was added to a biofilm formed in the presence of 50 µg/mL rhDNase. Finally,

tobramycin (4XMIC) was added to a biofilm formed in the absence of rhDNase. No

additional bactericidal effect, caused by the presence of rhDNase during biofilm

treatment and/or during biofilm formation, was observed (Table 2).

Table 2: Average log reductions (±SD) in biofilm cell numbers after tobramycin (4XMIC) treatment

alone or in combination with rhDNase. (n=3)

Biofilm

formation

Biofilm

treatment

B. cenocepacia

LMG 16656

B. multivorans

LMG 13010

B. cepacia

LMG 1222

B. anthina

LMG 20980

B.

vietnamiensis

LMG 10929

B. metallica

LMG 24068

csMH Tobra 3.87 (±0.13) 3.01 (±0.13) 3.14 (±0.37) 4.00 (±0.33) 4.46 (±0.10) 3.55 (±0.18)

csMH +

rhDNase Tobra 3.75 (±0.19) 3.05 (±0.30) 2.89 (±0.39) 3.70 (±0.33) 4.50 (±0.11) 3.60 (±0.16)

csMH +

rhDNase

Tobra +

rhDNase 3.66 (±0.18) 2.78 (±0.08) 2.74 (±0.37) 3.44 (±0.24) 4.17 (±0.16) 3.52 (±0.18)

Tobramycin was added to biofilms formed in csMH broth (control), biofilms formed in csMH broth in the presence of rhDNase and biofilms formed and treated in the presence of rhDNase. No statistically significant difference between the control condition and the conditions in which rhDNase was added was observed.

Effect of NaClO pretreatment on tobramycin susceptibility of 24h-old Bcc biofilms

A short pretreatment of the biofilms with a 60 µM NaClO solution (pH 7-8) had no

bactericidal effect on its own for all Bcc biofilms tested (Figure 3). For some Bcc species,

tobramycin (4XMIC) added to NaClO pretreated biofilms resulted in a significantly

increased bactericidal effect compared to tobramycin added to PBS pretreated biofilms

(Table 3). The most pronounced effect of NaClO pretreatment was noticed for B.

metallica LMG 24068, for which an additional 1.42 log reduction in cell numbers was

observed compared to tobramycin treatment alone. No increased bactericidal effect of

tobramycin was observed in NaClO pretreated B. anthina LMG 20980 and B.

vietnamiensis LMG 10929 biofilms.

123

Table 3: Average log reductions (±SD) in biofilm cell numbers after tobramycin (4XMIC) treatment

and pretreatment with PBS or NaClO. (n=3)

Pretreatment

B. cenocepacia

LMG 16656

B. multivorans

LMG 13010

B. cepacia

LMG 1222

B. anthina

LMG 20980

B. vietnamiensis

LMG 10929

B. metallica

LMG 24068

PBS 3.92 (±0.19) 3.09 (±0.19) 4.01 (±0.13) 3.21 (±0.18) 4.24 (±0.17) 3.83 (±0.22)

NaClO (60 µM) 4.67* (±0.24) 4.25* (±0.16) 4.86* (±0.13) 3.38 (±0.30) 4.41 (±0.17) 5.25* (±0.12)

*: statistically significant difference compared to PBS pretreatment (p < 0.05).

Effect of dispersin B on tobramycin susceptibility of 24h-old Bcc biofilms

Dispersin B at concentrations of 100 µg/mL had no bactericidal effect on any Bcc biofilm

tested (Figure 3). The bactericidal effect of tobramycin dissolved in phosphate buffer

(100 mM, pH 5.9) (Table 4; first row) was significantly reduced compared to the

bactericidal effect of tobramycin dissolved in PS (Table 3 and Table 2; first row). Still,

for three Bcc isolates, B. cenocepacia LMG 16656, B. cepacia LMG 1222 and B. metallica

LMG 24068, dispersin B significantly increased the bactericidal effect of tobramycin.

Table 4: Average log reductions (±SD) in biofilm cell numbers after tobramycin (4XMIC) treatment

alone (controls) or in combination with dispersin B of 24h-old biofilms.

Treatment

B. cenocepacia

LMG 16656

B. multivorans

LMG 13010

B. cepacia

LMG 1222

B. anthina

LMG 20980

B. vietnamiensis

LMG 10929

B. metallica

LMG 24068

Tobramycin 0.21 (±0.18) 0.71 (±0.12) 2.73 (±0.20) 2.11 (±0.26) 1.22 (±0.20) 2.57 (±0.15)

Tobramycin +

dispersin B 0.46* (±0.16) 0.72 (±0.16) 3.03* (±0.27) 2.06 (±0.23) 1.38 (±0.18) 2.91* (±0.14)

*: statistically significant difference compared to controls (p < 0.05) (n=3).

124

D iscussion

In the present study we investigated the role of different Bcc biofilm matrix components

in the protection against the antimicrobial activity of tobramycin. Tobramycin, a

positively charged aminoglycoside antibiotic, is likely to bind to EPS or eDNA, which will

decrease the bioavailability of free tobramycin.

Pulmozyme (a highly purified solution of rhDNase) was approved in 1993 by the US FDA

for the treatment of CF lung disease. rhDNase degrades eDNA, which is present in large

amounts in the sputum of CF patients, resulting in reduced sputum viscosity and

improved cough clearance and mucociliary transport [22]. In addition, rhDNase may

have an impact on biofilm lung infections. By degrading eDNA present in the biofilm

matrix it could sensitize bacteria to antibiotic therapy. In the present study we

demonstrated the presence of eDNA in all Bcc biofilms investigated. The amount of

eDNA differed among Bcc strains with the highest amount of eDNA measured for B.

multivorans LMG 13010. These high levels of eDNA could possibly contribute to the

virulence of the bacteria as eDNA has previously been shown to induce cationic

antimicrobial resistance in P. aeruginosa and S. enterica [17]. rhDNase added at the start

of biofilm formation significantly reduced the amount of eDNA in some Bcc strains. For

two Bcc strains, B. metallica LMG 24068 and B. vietnamiensis LMG 10929 no decrease in

eDNA concentration was measured after rhDNase treatment. Increasing the enzyme

concentration could potentially increase the effect of rhDNase but we did not explore

this further, as the concentrations used (50 µg/mL) were already at least 5 times higher

than the concentrations measured in sputum of CF patients after inhalation of 2.5 mg

rhDNase I [23]. When added alone, rhDNase had no effect on the viability or dispersion

of Bcc biofilm cells, which is in contrast with the observations of Tetz and Tetz for

Escherichia coli and S. aureus biofilms [24]. The ability of DNase to disperse established

P. aeruginosa biofilms has also been described [12]. However, although DNase I was

capable of dispersing young biofilms, older biofilms were much less affected by DNase I,

suggesting that other matrix components became increasingly important in stabilizing P.

aeruginosa biofilms during growth. We subsequently investigated whether rhDNase

could increase tobramycin susceptibility of Bcc biofilm cells. No increased bactericidal

effect of tobramycin was observed if Bcc biofilms were formed in the presence of 50

µg/mL of rhDNase, nor if they were formed and treated in the presence of rhDNase,

which is in agreement with tests previously performed in CF sputum [25]. Possible

125

explanations are that, although eDNA is depolymerized by rhDNase, tobramycin can still

bind to small eDNA molecules or that eDNA is not sufficiently degraded by rhDNase.

Goodman et al. indicated that DNA binding (DNAB II) proteins contribute to the

stabilization and structural integrity of eDNA within the matrix of different bacterial

biofilms, including Bcc biofilms, which could protect eDNA from degradation by

rhDNase [26, 27]. Recently, it has been observed that antiserum directed against DNAB

II proteins was capable of reducing B. cenocepacia biofilm mass, sensitizing the biofilm

towards antimicrobials. In contrast, a rhDNase treatment increased the biofilm biomass

[27]. A possible explanation for this is that smaller eDNA fragments, present in the

biofilm matrix after rhDNase treatment, are increasingly stabilizing the Bcc biofilm

matrix by cross-linking with other EPS, including polysaccharides.

Cepacian is considered the major exopolysaccharide produced by isolates of the Bcc

[28]. A protective effect of B. cenocepacia exopolysaccharides, and more specifically

cepacian, towards ROS has been demonstrated [4]. By scavenging NaClO, the backbone

of cepacian was degraded and acetyl substituents were cleaved off [5]. The loss of these

acetyl substituents hampered the ability of cepacian chains to aggregate, reducing the

viscosity of the polysaccharide network around Bcc cells [29]. To date, apart from

PNAG, the production of seven different exopolysaccharides by Bcc species has been

demonstrated. The exact contribution of exopolysaccharides in biofilm formation and

protection against stress has yet to be determined [30]. Benincasa et al. already

demonstrated a considerable decrease of the antibacterial activity of antimicrobial

peptides in the presence of polysaccharides produced by members of the Bcc [31]. To

study the effect of cepacian on tobramycin bioactivity, we pretreated Bcc biofilms with

60 µM NaClO solution, in order to degrade cepacian, before adding tobramycin. When

combined with tobramycin, we observed a significantly improved bactericidal activity of

the latter against B. cenocepacia LMG 16656, B. multivorans LMG 13010, B. cepacia LMG

1222 and B. metallica LMG 24068 biofilms. On average, 30% less biofilm cells survived

after combination therapy compared to tobramycin alone. The reactivity of ROS is most

likely not limited to the degradation of cepacian but could also affect the integrity of

other exopolysaccharides. This is corroborated by our results which demonstrate a

significantly increased bactericidal effect of tobramycin after NaClO pretreatment for B.

cenocepacia LMG 16656, although this strain had previously been shown not to produce

cepacian [32]. The use of a cepacian specific degrading enzyme, like the cepacian lyase

126

isolated by Cescutti et al., would allow for better differentiation between the role of

cepacian and other exopolysaccharides in reducing the activity of tobramycin [33].

The production of PNAG has been described for B. multivorans, B. vietnamiensis, B.

ambifaria, B. cepacia and B. cenocepacia where it seemed to be important for biofilm

formation and maintenance [7]. In addition, a mutant incapable of producing PNAG

showed increased sensitivity to tobramycin, which suggests a role for PNAG in Bcc

antibiotic resistance [7]. In contrast to these findings of Yakandawala et al., we did not

observe a biofilm dispersing effect of dispersin B on Bcc biofilms. The relatively small

biofilm killing effect of tobramycin monotherapy in this experiment is likely attributed

to the buffer recommended for optimal dispersin B activity (100 mM phosphate buffer

supplemented with 200 mM NaCl). As it has previously been noticed that tobramycin

activity can be significantly reduced in high ionic strength solutions [34], this could

explain the decreased bactericidal effect of tobramycin. Nevertheless, the combination

of dispersin B (100 µg/mL) with tobramycin significantly increased the latter’s

bactericidal effect against B. cenocepacia LMG 16656, B. cepacia LMG 1222 and B.

metallica LMG 24068. The combination therapy resulted in a 10-45% decrease in

biofilm cell numbers compared to tobramycin treatment alone for these isolates.

127

Conclusion

We investigated the in vitro bactericidal activity of tobramycin in combination with

matrix degrading substances. rhDNase did not show any additional effect on antibiotic

killing which could be due to a biofilm stabilizing effect of small eDNA fragments present

in the biofilm matrix after rhDNase treatment. Exopolysaccharide degradation, by

NaClO pretreatment, significantly increased the bactericidal effect of tobramycin

towards some of the Bcc biofilms tested, as did selective PNAG degradation by dispersin

B.

Acknowledgements

This research has been funded by the Interuniversity Attraction Poles Programme

initiated by the Belgian Science Policy Office and by a concerted action grant by Ghent

University.

We thank Kane Biotech for providing dispersin B and Petra Rigole and Laure Pape for

their excellent technical assistance.

128

References

1. Murray, T.S., M. Egan, and B.I. Kazmierczak, Pseudomonas aeruginosa chronic colonization in cystic fibrosis patients. Curr Opin Pediatr, 2007. 19(1): p. 83-8.

2. Lyczak, J.B., C.L. Cannon, and G.B. Pier, Lung Infections Associated with Cystic Fibrosis. Clin Microbiol Rev, 2002. 15(2): p. 194-222.

3. Zlosnik, J.E., et al., Mucoid and nonmucoid Burkholderia cepacia complex bacteria in cystic fibrosis infections. Am J Respir Crit Care Med, 2011. 183(1): p. 67-72.

4. Bylund, J., et al., Exopolysaccharides from Burkholderia cenocepacia inhibit neutrophil chemotaxis and scavenge reactive oxygen species. J Biol Chem, 2006. 281(5): p. 2526-32.

5. Cuzzi, B., et al., Investigation of bacterial resistance to the immune system response: cepacian depolymerisation by reactive oxygen species. Innate Immun, 2012. 18(4): p. 661-71.

6. Cescutti, P., et al., Structure of a novel exopolysaccharide produced by Burkholderia vietnamiensis, a cystic fibrosis opportunistic pathogen. Carbohydr Polym, 2013. 94(1): p. 253-60.

7. Yakandawala, N., et al., Characterization of the poly-beta-1,6-N-acetylglucosamine polysaccharide component of Burkholderia biofilms. Appl Environ Microbiol, 2011. 77(23): p. 8303-9.

8. Kaplan, J.B., et al., Enzymatic detachment of Staphylococcus epidermidis biofilms. Antimicrob Agents Chemother, 2004. 48(7): p. 2633-6.

9. Pavlukhina, S.V., et al., Noneluting enzymatic antibiofilm coatings. ACS Appl Mater Interfaces, 2012. 4(9): p. 4708-16.

10. Brindle, E.R., D.A. Miller, and P.S. Stewart, Hydrodynamic deformation and removal of Staphylococcus epidermidis biofilms treated with urea, chlorhexidine, iron chloride, or DispersinB. Biotechnol Bioeng, 2011. 108(12): p. 2968-77.

11. Kaplan, J.B., Therapeutic potential of biofilm-dispersing enzymes. Int J Artif Organs, 2009. 32(9): p. 545-54.

12. Whitchurch, C.B., et al., Extracellular DNA required for bacterial biofilm formation. Science, 2002. 295(5559): p. 1487.

13. Izano, E.A., et al., Differential roles of poly-N-acetylglucosamine surface polysaccharide and extracellular DNA in Staphylococcus aureus and Staphylococcus epidermidis biofilms. Appl Environ Microbiol, 2008. 74(2): p. 470-6.

14. Steinberger, R.E. and P.A. Holden, Extracellular DNA in single- and multiple-species unsaturated biofilms. Appl Environ Microbiol, 2005. 71(9): p. 5404-10.

15. Kaplan, J.B., et al., Recombinant human DNase I decreases biofilm and increases antimicrobial susceptibility in staphylococci. J Antibiot (Tokyo), 2012. 65(2): p. 73-7.

16. Mulcahy, H., L. Charron-Mazenod, and S. Lewenza, Extracellular DNA Chelates Cations and Induces Antibiotic Resistance in Pseudomonas aeruginosa Biofilms. PLoS Pathog, 2008. 4(11).

17. Johnson, L., et al., Extracellular DNA-induced antimicrobial peptide resistance in Salmonella enterica serovar Typhimurium. BMC Microbiol, 2013. 13(1): p. 115.

18. Chiang, W.C., et al., Extracellular DNA Shields against Aminoglycosides in Pseudomonas aeruginosa Biofilms. Antimicrob Agents Chemother, 2013. 57(5): p. 2352-61.

129

19. Anderson, M.S., E.C. Garcia, and P.A. Cotter, The Burkholderia bcpAIOB Genes Define Unique Classes of Two-Partner Secretion and Contact Dependent Growth Inhibition Systems. PLoS Genet, 2012. 8(8).

20. Gueroult, M., et al., How cations can assist DNase I in DNA binding and hydrolysis. PLoS Comput Biol, 2010. 6(11): p. e1001000.

21. Peeters, E., H.J. Nelis, and T. Coenye, In vitro activity of ceftazidime, ciprofloxacin, meropenem, minocycline, tobramycin and trimethoprim/sulfamethoxazole against planktonic and sessile Burkholderia cepacia complex bacteria. J Antimicrob Chemother, 2009. 64(4): p. 801-9.

22. Wagener, J.S. and O. Kupfer, Dornase alfa (Pulmozyme). Curr Opin Pulm Med, 2012. 18(6): p. 609-14.

23. Sanders, N.N., et al., Role of magnesium in the failure of rhDNase therapy in patients with cystic fibrosis. Thorax, 2006. 61(11): p. 962-8.

24. Tetz, V.V. and G.V. Tetz, Effect of extracellular DNA destruction by DNase I on characteristics of forming biofilms. DNA Cell Biol, 2010. 29(8): p. 399-405.

25. Hunt, B.E., et al., Macromolecular mechanisms of sputum inhibition of tobramycin activity. Antimicrob Agents Chemother, 1995. 39(1): p. 34-9.

26. Goodman, S.D., et al., Biofilms can be dispersed by focusing the immune system on a common family of bacterial nucleoid-associated proteins. Mucosal Immunol, 2011. 4(6): p. 625-37.

27. Novotny, L.A., et al., Structural stability of Burkholderia cenocepacia biofilms in reliant on eDNA structure and presence of a bacterial nucleic acid binding protein. PLoS One, 2013. 8(6).

28. Cunha, M.V., et al., Studies on the involvement of the exopolysaccharide produced by cystic fibrosis-associated isolates of the Burkholderia cepacia complex in biofilm formation and in persistence of respiratory infections. J Clin Microbiol, 2004. 42(7): p. 3052-8.

29. Herasimenka, Y., et al., Macromolecular properties of cepacian in water and in dimethylsulfoxide. Carbohydr Res, 2008. 343(1): p. 81-9.

30. Ferreira, A.S., et al., Insights into the role of extracellular polysaccharides in Burkholderia adaptation to different environments. Front Cell Infect Microbiol, 2011. 1: p. 16.

31. Benincasa, M., et al., Activity of antimicrobial peptides in the presence of polysaccharides produced by pulmonary pathogens. J Pept Sci, 2009. 15(9): p. 595-600.

32. Ferreira, A.S., et al., Distribution of cepacian biosynthesis genes among environmental and clinical Burkholderia strains and role of cepacian exopolysaccharide in resistance to stress conditions. Appl Environ Microbiol, 2010. 76(2): p. 441-50.

33. Cescutti, P., et al., First report of a lyase for cepacian, the polysaccharide produced by Burkholderia cepacia complex bacteria. Biochem Biophys Res Commun, 2006. 339(3): p. 821-6.

34. Beggs, W.H. and F.A. Andrews, Role of ionic strength in salt antagonism of aminoglycoside action on Escherichia coli and Pseudomonas aeruginosa. J Infect Dis, 1976. 134(5): p. 500-4.

130

Chapter VII: TRANSPORT OF NANOPARTICLES AND TOBRAMYCIN-

LOADED LIPOSOMES IN BURKHOLDERIA CEPACIA COMPLEX

BIOFILMS

Redrafted from:

Anne-Sophie Messiaen, Katrien Forier, Hans Nelis, Kevin Braeckmans, Tom

Coenye (2013). Transport of nanoparticles and tobramycin-loaded liposomes in

Burkholderia cepacia complex biofilms. Plos one, Accepted.

131

ABSTRACT

Due to the intrinsic resistance of the Bcc to many antibiotics and the production of a

broad range of virulence factors, lung infections by these bacteria, primarily occurring in

CF patients, are very difficult to treat. In addition, the ability of Bcc organisms to form

biofilms contributes to their persistence after colonization of the CF lung. As Bcc

infections are associated with poor clinical outcome, there is an urgent need for new

effective therapies to treat these infections. In the present study, we investigated

whether the encapsulation of tobramycin in liposomes could increase its anti-biofilm

effect against Bcc. Single particle tracking (SPT) was used to study the transport of

positively and negatively charged nanospheres in Bcc biofilms as a model for the

transport of liposomes. Negatively charged nanospheres became immobilized in close

proximity of biofilm cell clusters, while positively charged nanopsheres interacted with

fiber-like structures, probably eDNA. Based on these data, encapsulation of tobramycin

in negatively charged liposomes appeared promising for targeted drug delivery.

However, the anti-biofilm effect of tobramycin encapsulated into neutral or anionic

liposomes did not increase compared to that of free tobramycin. Probably, the fusion of

the anionic liposomes with the negatively charged bacterial surface of Bcc was limited

by electrostatic repulsive forces at close proximity. The lack of a substantial anti-biofilm

effect of tobramycin encapsulated in neutral liposomes could be further investigated by

increasing the liposomal tobramycin concentration. However, this was hampered by the

low encapsulation efficiency of tobramycin in these liposomes.

132

Introduction

CF is the most prevalent hereditary disease in the Caucasian population and is caused by

mutations in both cftr alleles encoding a chloride channel [1]. The absence of a

functional chloride channel (CFTR channel) results in the secretion of thick, viscous

mucus in several organs, including the lungs and the gastrointestinal tract [2]. The thick

mucus in CF lungs impairs the mucociliary clearance, rendering them more susceptible

to infections. In CF patients, lung infections can not be cleared easily and rapidly evolve

to chronic infections, the main cause of morbidity and mortality [3]. The most

significant pathogen colonizing the lungs of CF patients is P. aeruginosa. Over time,

approximately 80% of the CF population becomes infected with this pathogen [4].

Compared to P. aeruginosa, Bcc infections only account for a small percentage of the

respiratory infections within the CF population. However, they are often associated

with rapid deterioration of the lung function and increased mortality [5]. The capacity

of Bcc species to cause invasive disease and their high level of intrinsic antibiotic

resistance make them particularly difficult to eradicate. Through the production of

different EPS additional protection is provided against antibiotics and host immune

components [6]. As there is still no optimal treatment regimen for Bcc infections in CF

patients [7], these pathogens continue to be highly problematic [8] and there is an

urgent need for effective antibiotic therapies. As for P. aeruginosa, the production of EPS

by Bcc strains is essential for the formation of thick and mature biofilms [9]. These EPS,

including polysaccharides and eDNA, can delay the penetration of an antibiotic through

the biofilm [10]. By the encapsulation of the antibiotic in a liposomal carrier,

interactions between the antibiotic and EPS could probably be avoided, resulting in an

improved anti-biofilm effect. In addition, liposome-encapsulated antibiotics would be

protected from degradation by antibiotic-inactivating enzymes (like β-lactamases)

which can accumulate in the biofilm matrix [11]. Another important factor contributing

to antibiotic resistance in several Gram-negative bacteria, including P. aeruginosa and

Bcc species, is their low outer membrane permeability [12-14]. The uptake of

antibiotics encapsulated in neutral liposomal carriers is not affected by this outer

membrane barrier as this occurs through direct fusion of the liposome with the bacterial

membrane [15]. It has been shown that subinhibitory concentrations of tobramycin

encapsulated in neutral liposomes has high bactericidal activity against planktonic

cultures of several bacterial species, including P. aeruginosa and B. cenocepacia [16-18],

133

probably due to an enhanced uptake of the antibiotic. Besides the improved transport

through the biofilm and a better uptake in the cell, liposomes could provide selective

delivery to the target site by incorporating specific ligands to the liposome surface.

Currently, several liposome based drugs have been approved for clinical use and various

others are in clinical trials [19]. Arikace, a neutral liposomal amikacin formulation,

currently undergoes phase III clinical trials for the treatment of lung infections [19, 20].

It is the first liposomal drug being investigated for aerosol delivery.

In the present study, we investigated whether the bactericidal effect of tobramycin

against Bcc biofilms could be increased by encapsulating the antibiotic in liposomes.

Therefore, we analyzed the behavior of positively and negatively charged nanoparticles

in Bcc biofilms by means of SPT as a model for liposomal transport. In contrast to

liposomes, which are composed of phospholipids, the nanoparticles used in the model

are made up of polystyrene and thus can not be degraded by enzymes or other

components present in the biofilm matrix. According to our observations, negatively

charged particles were directed towards cell clusters, while positively charged particles

became immobilized by interactions with fiber-like structures (likely eDNA) in the

biofilm matrix. Based on these results, tobramycin encapsulated into negatively charged

liposomes could be promising in terms of targeted drug delivery. We consequently

evaluated the activity of neutral and anionic liposomal tobramycin formulations against

Bcc biofilms.

134

Materials and methods

Strains

We used 5 Bcc strains in the present study: B. cenocepacia LMG 16656 and LMG 18829,

B. cepacia LMG 1222, B. multivorans LMG 18825 and B. dolosa LMG 18943. All Bcc

strains were obtained from the BCCM/LMG Bacteria Collection (Ghent, Belgium) or were

kindly provided by Dr. P. Vandamme (Ghent University, Belgium). The bacteria were

stored in Microbank tubes (Prolab Diagnostics, Richmond Hill, ON, Canada) at -80°C and

transferred twice on MH (Lab M, Heywood, UK) agar plates before use in any

experiment.

Lipids

1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) and cholesterol were obtained

from Sigma-Aldrich (St. Louis, MO, USA) and stored at -20°C. 1,2-dioleoyl-sn-glycero-3-

phosphocholine (DOPC) and 1,2-dipalmitoyl-sn-glycero-3-phosphoglycerol sodium salt

(DPPG), were obtained from Corden Pharma International (Plankstadt, Germany) and

stored at -20°C. Lipids were dissolved in chloroform and stock solutions were stored

between 2-8°C.

Nanoparticles

Yellow-green (λex: 505/ λem: 515) fluorescent carboxylate-modified polystyrene

FluoSpheres of 0.2 μm diameter were purchased from Invitrogen (Carlsbad, CA, USA).

Positively charged nanospheres were prepared from the carboxylate-modified

FluoSpheres by modification with N,N-dimethylethylenediamine (DMEDA) (Sigma-

Aldrich) as described previously [21]. The size and zeta-potential of the positively

charged nanospheres were measured in HEPES (pH 7.3, 20 mM) using the Zetaziser

Nano-ZS (Malvern, Worcestershire, UK).

Single Particle Tracking setup

All SPT measurements were performed on a custom-built laser wide-field fluorescence

microscope setup as described previously [22]. Briefly, two solid-state lasers, a 100

mW Calypso 491 nm (Cobolt, Solna, Sweden) and a 100 mW Spectra Physics Excelsior

642 nm (CA, USA), were used to excite the fluorophores in samples mounted on a

TE2000-E (Nikon BeLux, Brussels, Belgium) inverted microscope equipped with a Plan

Apo VC 100× 1.4 NA oil immersion objective lens (Nikon). A diffuser in the illumination

135

path ensured even illumination of the sample. An acousto-optical tunable filter (AOTF)

was used to modulate the intensity of the laser beams. The AOTF was synchronized to

the EMCCD camera (Cascade II:512; Roper Scientific, Tucson, AZ, USA)in order to limit

photobleaching during imaging. Videos of the moving nanoparticles were acquired with

the NIS Elements software package (Nikon).

Analysis of SPT data

Analysis of the videos was performed off-line using software developed by Braeckmans

et al. [23]. First, the particles were localized in all frames of the SPT movie. A selection

of “real” particles is made according to user-defined criteria, including size, contrast

relative to the local background and sphericity. Subsequently, individual trajectories of

the nanospheres are calculated using a nearest neighbor algorithm. The positions of

nanospheres that are closest to each other are connected in subsequent image frames,

taking into consideration the maximum distance a particle can reasonably move from

one frame to another. The trajectories are further analyzed based on mean square

displacement (MSD) analysis. The MSD is calculated for every available time lag (t) , i.e.

the multiples of time between two subsequent images in an SPT movie.

A weighted least squares fitting was performed of

MSD = Γtα + 4 σ2

to the MSD vs. t curves, yielding for each trajectory the parameters Gamma (Γ), alpha (α)

and sigma (σ). Gamma is the so-called transport coefficient, alpha the anomalous

exponent and sigma a parameter taking into account the limited precision with which a

particle can be localized. The parameter of interest is alpha, which contains information

on the mode of motion. For free diffusion, α equals 1 while α < 1 and α > 1 represent

sub- and super-diffusion, respectively. By analyzing the MSD versus t plots of many

trajectories, a distribution of α values is obtained. In addition, the apparent diffusion

coefficient, Da, was calculated corresponding to the first time lag according to:

Da = MSD/(4t)

The distribution of Da values was further processed using a maximum entropy method

(MEM) that improves the resolution and at the same time smoothes the curve by only

retaining the features that are statistically warranted by the data [22].

136

SPT measurements in biofilms

Bcc biofilms were cultured in uncoated 35 mm glass bottom culture dishes (MatTek,

Ashland, MA, USA). The dishes were inoculated with 2 mL of a standardized bacterial

suspension in MH broth (∼108 CFU/mL) in the presence or absence of 10 µg/mL

dornase alfa (Pulmozyme, Genentech, SF, USA) and were incubated at 37°C. After 4h, the

supernatant, containing non-adhered cells, was removed and adhered cells were

carefully washed with PS. Subsequently, 2 mL of fresh MH broth (with or without 10

µg/mL dornase alfa) was added and the dishes were incubated for another 20h at 37°C

to allow biofilm formation. After 24h of biofilm formation, biofilms were gently washed

with PS to remove nonadhering cells. Biofilm cells were stained by adding 1 mL of a

Syto 59 (Invitrogen) solution (5 mM in PS). After 15 min of incubation at room

temperature, protected from light, the biofilm was washed twice with PS. The

nanoparticle stock suspensions were sonicated for 10 min before dilution in PS (∼

0.002% solids). One mL of this particle suspension was added to the biofilms,

immediately before recording movies. Movies of 5 s with a temporal resolution of 39.2

ms (time between two subsequent images, time lag) and an image acquisition time of 3

ms were recorded. Dual color image acquisition allowed easy navigation within the

biofilm. Between 25 and 30 movies on different locations in the biofilm were recorded.

Each experiment was carried out in triplicate at room temperature.

Liposome preparation

Liposomes were prepared by a dehydration-rehydration method as described

previously [24]. In brief, for the preparation of neutral liposomes, 50 µmol of DPPC and

25 µmol of cholesterol were dissolved in 1 mL of chloroform in a round-bottomed flask.

For the anionic liposomes, 53 µmol of DOPC and 6.6 µmol of DPPG were dissolved in 1

mL of chloroform. The round-bottomed flask was connected to a rotary evaporator to

dry the lipid film under controlled vacuum at 50°C. The lipid film was rehydrated with 2

mL of a sucrose solution in distilled water (1:1, w/w, sucrose to lipid). Sucrose is

needed to stabilize the liposomes during freeze drying. Lipid suspensions were

vortexed to form multilamellar vesicles and then sonicated for 5 min in an ultrasonic

bath (Branson 3510), followed by two additional cycles of vortexing and sonication. The

resulting suspension was then mixed with 1 mL (40 mg/mL) of tobramycin in PS. The

mixture was frozen (-20°C) and immediately freeze dried. After freeze drying, the

137

powdered formulations were stored at 4°C until use. For rehydration, 200 µL of distilled

water at 50°C was added. The suspension was vortexed and incubated for 30 min at

50°C. These steps were repeated with 200 µL phosphate-buffered saline (PBS, pH 5.9).

After incubation at 50°C, 1.6 mL of PBS was added, the mixture was vortexed and

incubated for another 30 min at 50°C. Non-encapsulated tobramycin was removed

following three rounds of PBS wash (18300 x g, 15 min at 4°C) and the pellet was

resuspended in 2 mL PS. The size and zeta-potential of the liposomes were measured in

HEPES (pH 7.3, 20 mM) using the Zetaziser Nano-ZS (Malvern, Worcestershire, UK).

Quantification of the amount of encapsulated tobramycin

The concentration of liposome encapsulated tobramycin was measured by agar

diffusion, using a laboratory strain of Bacillus subtilis (ATCC 6633) as indicator

organism. B. subtilis ATCC 6633 spore suspensions were prepared as described in the

European Pharmacopoeia [25]. In brief, B. subtilis was grown at 35-37°C for 7 days on

Antibiotic medium 1 (AM1) supplemented with 0.001 g/L manganese sulphate. After at

least 7 days, the growth, which mainly consists of spores, was washed off using sterile

water. The obtained suspension was heated at 70°C for 30 min and diluted to give an

appropriate concentration of spores (107 – 108 per mL). The spore suspensions were

stored at 4°C until use. We used this suspension to prepare a 1% suspension of spores

(4 mL/ 400 mL) in warm (47.5 °C) AM 11 agar. The agar was poured into a sterile glass

plate and was left to solidify for 15 min at room temperature. To lyse the liposomes, 0.2

% Triton X-100 was added to the liposomal solutions. This level of Triton X-100 had no

effect on the performance of the assay (data not shown). Wells of 5 mm diameter were

punched in the agar and filled with 200 µL of standard tobramycin solutions or with the

sample. The plate was covered with a steel lid and incubated for 20 hours at 35°C after

4h of pre-diffusion at 4°C. We subsequently measured the inhibition zones and the

average of 4 measurements was used for data analysis. A standard curve with known

concentrations of free tobramycin (0.03125-1 µg/mL) was constructed and was utilized

to calculate the amount of encapsulated tobramycin that was released after Triton X-100

treatment.

138

Determination of the MIC of tobramycin

MICs were determined in duplicate according to the EUCAST broth microdilution

protocol using unsupplemented MH broth. In brief, bacterial suspensions were

standardized at ∼5 × 105 CFU/mL before inoculation of a flat-bottomed 96-well plate

filled with 100 µL of tobramycin serial dilutions in MH. The range of tobramycin

concentrations used was from 2 to 1024 mg/L. Plates were incubated at 37°C for 20 h

and the optical density was determined at 590 nm using a multilabel microtitre plate

reader (Envision, Perkin Elmer LAS, Waltham, MA, USA). The MIC is the lowest

concentration of antibiotic for which a similar optical density was observed in the

inoculated and blank wells.

Activity of liposome encapsulated tobramycin against Bcc biofilms

Biofilms were cultured in round-bottomed 96-well plates. An inoculum was prepared

by suspending bacteria from a pure culture on MH agar in MH broth. The inoculum was

standardized at ∼108 CFU/mL and the wells of the microtitre plate were filled with 100

µL of this suspension. Plates were incubated at 37°C for 4h without shaking. After 4h,

the supernatant, containing planktonic cells, was aspirated from the wells. Adhered

cells were carefully washed with 100 µL of PS and 100 µL of fresh sterile MH broth was

added. Plates were incubated for an additional 20h at 37°C. After 24h of biofilm

formation, biofilms were washed with 100 µL of PS before treatment with free

tobramycin (final concentrations of 140 µg/mL or 4XMIC of tobramycin in PS) or

liposome encapsulated tobramycin (final concentrations of 140 µg/mL or 4XMIC of

tobramycin in PS). The anionic liposomal stock solutions were diluted in PS to obtain

the concentrations used in the experiments. After 24h of treatment at 37°C, cell

numbers were determined by plate counting.

139

Results

Properties of nanoparticles and liposomes

The average sizes and average zeta-potentials of nanospheres and liposomes used in the

present study are shown in Table 1. For the liposomes, the concentrations of

encapsulated tobramycin (i.e. the concentration that is released after breaking up all

liposomes in the stocksolution with Triton X-100) are also shown in Table 1.

Table 1: Characteristics of the nanospheres and liposomes used in the present study (n=3).

Particle

Average

size (nm )

( ± SD)

Average

zeta-potential (mV)

( ± SD)

Total lipsomal tobramycin

concentration (µg/mL)

( ± SD)

Carboxylate-modified

nanosphere 224.0 (± 0.7) - 48.4 (± 0.7) /

DMEDA-modified

nanosphere 231.9 (± 1.2) 30.4 (± 0.6) /

DPPC/cholesterol

liposomes

(2/1, molar ratio)

426.3 (± 26.4) - 0.5 (± 0.1) 141 (± 35)

DOPC/DPPG

liposomes

(8/1, molar ratio)

228.5 ( ± 34.9) - 22.3 (± 0.5) 1128 ( ± 16)

Transport of negatively charged nanoparticles in Bcc biofilms

The mobility of carboxylate-modified polystyrene nanospheres (0.2 µm) added to Bcc

biofilms was studied by analyzing individual trajectories of the nanospheres in the

biofilm. The distribution of both α values and Da coefficients of the carboxylate-modified

nanospheres is shown in Figure 1. Average α and Da values for all conditions are shown

in Table 2. For all strains investigated, negatively charged particles display subdiffusion

( 0 < α < 1 ) when added to the biofilm. This effect was the most pronounced in B.

cenocepacia LMG 16656 and B. cenocepacia LMG 18829 biofilms and to a lesser extent in

B. multivorans LMG 18825 and B. cepacia LMG 1222 biofilms. This is also reflected in the

Da distributions where an immobilized fraction is seen in both B. cenocepacia LMG

16656 and B. cenocepacia LMG 18829 biofilms. Thus it is clear that the negatively

charged particles strongly interact with the biofilms. In addition, on average the

140

particles diffuse slowly in the biofilm as their apparent diffusion coefficients are much

lower than those in water (Dw = 1.96 µm2/s [26]). The highest diffusion rates were

observed in B. cepacia LMG 1222 biofilm (average Da = 0.42 µm2/s) while the lowest

were observed in B. multivorans LMG 18825 biofilm (average Da = 0.15 µm2/s). The

mobility of the negatively charged nanospheres in both B. cenocepacia LMG 16656 and

LMG 18829 biofilms is similar, yielding average diffusion coefficients of 0.26 and 0.28,

respectively.

Table 2: Average apparent diffusion coefficients, both in Bcc biofilms and in water [26], and

average α values of the nanospheres in Bcc biofilms.

Biofilm Condition Particle Av Da (µm2/s)

(± SD) Av α

Av Dw (µm2/s)

(±SD)

B. cenocepacia

LMG 16656

Control Carboxylate-modified 0.26 (± 0.29) 0.43 1.96 (± 0.02)

Control DMEDA-modified 0.29 (± 0.21) 0.39 1.88 (± 0.09)

rhDNase DMEDA-modified 0.37 (± 0.44) 0.44

B. cenocepacia

LMG 18829

Control Carboxylate-modified 0.28 (± 0.36) 0.48 1.96 (± 0.02)

Control DMEDA-modified 0.23 (± 0.30) 0.23 1.88 (± 0.09)

rhDNase DMEDA-modified 0.54 (± 0.30) 0.49

B. cepacia

LMG 1222

Control Carboxylate-modified 0.42 (± 0.25) 0.81 1.96 (± 0.02)

Control DMEDA-modified 0.27 (± 0.28) 0.62 1.88 (± 0.09)

rhDNase DMEDA-modified 0.45 (± 0.26) 0.97

B. multivorans

LMG 18825

Control Carboxylate-modified 0.15 (± 0.11) 0.77 1.96 (± 0.02)

Control DMEDA-modified 0.38 (± 0.20) 0.65 1.88 (± 0.09)

rhDNase DMEDA-modified 0.42 (± 0.30) 0.67

B. dolosa

LMG 18943

Control Carboxylate-modified 0.32 (± 0.36) 0.57 1.96 (± 0.02)

Control DMEDA-modified 0.25 (± 0.30) 0.35 1.88 (± 0.09)

rhDNase DMEDA-modified 0.19 (± 0.30) 0.53

141

Transport of positively charged nanoparticles in Bcc biofilms

The mobility of DMEDA-modified polystyrene nanospheres (0.2 µm) in different Bcc

biofilms was also studied (Figure 2). Positively charged nanospheres displayed

subdiffusion in all Bcc biofilms when grown in the absence of dornase alfa. Alfa values

for B. cepacia LMG 1222 show a bimodal distribution, with part of the particles

displaying free diffusion. Anomalous diffusion was most pronounced in B. cenocepacia

LMG 18829 (mean α = 0.23). In contrast, α values closest to 1 were obtained in B.

multivorans LMG 18825 and B. cepacia LMG 1222 biofilms, with mean α values of 0.65

and 0.62, respectively. This correlates with the highest average apparent diffusion

coefficient (average Da = 0.38) of positively charged nanospheres in B. multivorans LMG

18825 biofilms. Although the average diffusion coefficient in B. cepacia LMG 1222 is

much smaller, the distribution is clearly bimodal, with two groups of particles behaving

differently. The peak of the faster moving population represents particles with an

apparent diffusion coefficient of 0.41 µm2/s. Still, positively charged nanospheres

diffuse much more slowly in the Bcc biofilms than in water (Dw = 1.88 µm2/s [26]).

Together with the low α values of these particles in Bcc biofilms this clearly indicates a

hindered transport, which could be due to interaction with negatively charged matrix

components. It was previously observed that the positively charged nanospheres

interacted with fiber-like structures in Bcc biofilms, greatly impairing their mobility and

this was confirmed in the present study. As these fiber-like structures are likely eDNA,

the experiments with DMEDA-modified nanospheres were repeated in Bcc biofilms

grown in the presence of dornase alfa (10 µg/mL) (Figure 2). The most pronounced

effect of dornase alfa on the mobility of positively charged nanospheres was observed in

B. cepacia LMG 1222 and B. cenocepacia LMG 18829 biofilms for which average diffusion

coefficients approximately doubled, from 0.27 to 0.45 µm2/s and from 0.23 to 0.54

µm2/s, respectively. In addition the α distributions of the positively charged

nanospheres displayed a shift to higher values, although this shift was less pronounced

in B. cenocepacia LMG 16656 and B. multivorans LMG 18825 (Figure 2).

142

Figure 1: Mobility of 0.2 µm carboxylate-modified nanospheres in Bcc biofilms.

Distribution of α values, describing the mode of movement, and Da values, describing the rate of movement for A) B.

cenocepacia LMG 16656, B) B. cenocepacia LMG 18829, C) B. cepacia LMG 1222, D) B. multivorans LMG 18825 and E) B.

dolosa LMG 18943. Dotted curves show the average α and average Da value of the carboxylate-modified nanospheres in

water [26].

143

Figure 2: Mobility of 0.2 µm DMEDA- modified nanospheres in Bcc biofilms.

Distribution of α values, describing the mode of movement, and Da values, describing the rate of movement for A) B.

cenocepacia LMG 16656, B) B. cenocepacia LMG 18829, C) B. cepacia LMG 1222, D) B. multivorans LMG 18825 and E) B.

dolosa LMG 18943. Green, solid lines represent the distribution for DMEDA-modified particles added to control biofilms.

Red, dotted lines represent the distributions for DMEDA-modified particles added to rhDNase treated biofilms. Grey

dotted lines show the α and Da value distributions of the DMEDA-modified nanospheres in water [26].

144

Activity of liposome encapsulated tobramycin against Bcc biofilms

We tested the activity of both neutral and negatively charged tobramycin containing

liposomes (0.2 µm) against Bcc biofilms (Figure 3). When added at a final concentration

of 140 µg/mL (Figure 3a), free tobramycin showed modest activity against B.

cenocepacia LMG 18829, B. multivorans LMG 18825 and B. dolosa LMG 18943 and strong

activity against B. cepacia LMG 1222. This is in agreement with the MIC values

observed. The MIC for B. cepacia LMG 1222 is low (32 µg/mL) while these for B.

cenocepacia LMG 18829, B. multivorans LMG 18825 and B. dolosa LMG 18943 were only

slightly below (128 µg/mL) the tobramycin concentration used (140 µg/mL) in the

present study. Finally, the MIC of tobramycin for B. cenocepacia LMG 16656 (256

µg/mL) is considerably above the concentration used in this experiment, explaining the

lack of bactericidal activity (Figure 3A). Both neutral and negatively charged liposomes

containing 140 µg/mL of tobramycin (≈ 4XMIC) showed a bactericidal effect against B.

cepacia LMG 1222 biofilms but no increased bactericidal activity compared to free

tobramycin was observed. In B. multivorans LMG 18825 biofilms, neutral liposomes

containing 140 µg/mL tobramycin (≈ MIC) showed the same activity as free tobramycin.

Previous research from our group has indicated that tobramycin at concentrations of

4XMIC yielded a substantial bactericidal effect against Bcc biofilms [27]. Increasing the

liposomal tobramycin concentration to 4XMIC was only possible for the negatively

charged liposomes as the maximum achievable concentration of encapsulated

tobramycin in neutral liposomes was 140 µg/mL. However, at this increased

tobramycin concentration, negatively charged liposomes only showed bactericidal

activity against B. cepacia LMG 1222 (Figure 3B).

145

Figure 3: Activity of free and liposomal tobramycin against Bcc biofilms.

A) Bactericidal effect of free and liposomal tobramycin (anionic and neutral liposomes) at a final

concentration of 140 µg/mL against Bcc biofilms. B) Bactericidal effect of free and liposomal tobramycin

(anionic liposomes) at a final concentration of 4XMIC against Bcc biofilms.

146

D iscussion

In the present study we investigated the transport of negatively and positively charged

nanospheres in Bcc biofilms. The transport of nanospheres can be used as a model to

predict the transport of antibiotic-containing liposomes. Liposomes are made up of

phospholipids and can act as biodegradable delivery systems. Antibiotics encapsulated

in liposomes have lower toxicity and higher bioactivity and bioavailability [18, 28]. It

has already been demonstrated that liposomal antibiotic formulations show increased

bactericidal activity against biofilms compared to free antibiotics [29, 30]. However, a

first condition is that efficient transport of liposomes in the biofilm matrix should occur.

Therefore, as a first step in developing a drug-delivery system with activity against Bcc

biofilms, we studied the transport of model nanospheres in these biofilms. In a previous

study we have shown that PEGylated neutral nanospheres displayed normal

(unobstructed) diffusion in B. multivorans LMG 18825 biofilms and that these particles

diffuse in the biofilm at similar rate as in water [26]. This was in contrast to the

movement of carboxylate- and DMEDA-modified nanopsheres in a B. multivorans LMG

18825 biofilm which was found to be strongly impeded [26]. In the present work we

expanded the panel of Bcc strains to study the diffusion in biofilms of both carboxylate-

and DMEDA-modified nanospheres. The negatively charged nanospheres moved 5 to 13

times more slowly in all Bcc biofilms tested compared to in water. In all biofilms, a

fraction of relatively immobile negatively charged particles, characterized by Da values

close to zero, was observed. This immobilization is probably caused by an electrostatic

interaction between positively charged components of the biofilm and the negatively

charged nanospheres, which correlates with the shift towards α < 1 observed for these

particles, indicating anomalous subdiffusion. It was previously observed that positively

charged particles are immobilized on fiber-like structures, which were hypothesized to

be eDNA [26] (Figure 4). Here we tested this further by studying the transport of

DMEDA-modified nanospheres in biofilms grown in the presence of dornase alfa. In

general it was found that dornase alfa treatment reduced the extent of anomalous

diffusion and improved the diffusion rate, supporting the view that positively charged

particles can get trapped by binding to eDNA in the biofilm matrix. As DMEDA-modified

nanospheres interact with eDNA, the application of positively charged liposomes as

antibiotic carrier systems can only be considered in combination with e.g. dornase alfa.

However, as previous research has indicated that the use of dornase alfa could be

147

contraindicated in CF patients infected with B. cenocepacia [31], we did not test the

effect of tobramycin encapsulated in positively charged liposomes. Instead, the effect of

tobramycin encapsulated in negatively charged liposomes was compared to that of

tobramycin encapsulated in neutral liposomes. Neutral liposomal formulations were

tested as a reference as these appear, at first sight, less interesting to treat Bcc biofilm

infections based on the mobility of neutral nanospheres which showed no enrichment in

B. multivorans LMG 18825 biofilms. In contrast, enrichement of carboxylate-modified

nanospheres at sites close to the bacteria has been observed [26] (Figure 4). Therefore,

if anionic liposomes behave similarly in Bcc biofilms as anionic nanospheres,

tobramycin could potentially be targeted near the cell clusters when encapsulated in

anionic liposomes, resulting in an increased anti-biofilm effect.

Figure 4: Confocal images of nanoparticles is a B. multivorans LMG 18825 biofilm [26].

Confocal image of A) 0.1 µm carboxylate-modified nanospheres and B) 0.1 µm DMEDA-modified nanospheres in a B. multivorans LMG 18825 biofilm. The bacteria are shown in red (Syto 59) and the nanospheres in green. The insert in A) shows anionic nanospheres adhered to the bacteria.

At free tobramycin concentrations equal to 4XMIC, a substantial bactericidal effect

against all Bcc biofilms was observed. Unfortunately, this anti-biofilm effect could not

be increased by encapsulating tobramycin in either neutral or anionic liposomes. As the

surface of Bcc cells is negatively charged it could be that the fusion of anionic liposomes

with the bacterial cells is limited by repulsive forces at close proximity. The bactericidal

activity of tobramycin encapsulated in neutral liposomes at concentrations of 4XMIC

could only be tested against B. cepacia LMG 1222 as this strain showed the lowest MIC

(32 µg/mL) and not more than 140 µg/mL of tobramycin could be encapsulated in these

neutral liposomes. The neutral liposomal tobramycin formulation did not show an

increased bactericidal effect compared to that of free tobramycin. Although the

observed relatively unobstructed motion of neutral particles in B. multivorans LMG

18825 biofilms suggests no penetration difficulties for neutral liposomes in the biofilm,

148

we did not find a substantial anti-biofilm effect. Still, the effect of the neutral liposomal

tobramycin at a concentration of 140 µg/mL was not decreased compared to the effect

of free tobramycin in B. multivorans LMG 18825 biofilms, indicating that the

encapsulated tobramycin does enter the cell. Possible explanations for the lack of an

anti-biofilm effect of neutral formulations of liposomal tobramycin in B. cenocepacia

LMG 16656, B. cenocepacia LMG 18829 and B. dolosa LMG 18943 are that a hydrophilic

EPS layer in close proximity to the biofilm cells protects them from fusion with the

liposomal carrier system. Alternatively, it could be that the transport of nanospheres is

not a good model for the transport of liposomes in Bcc biofilms, especially for the

neutral liposomes which are twice the size of the nanospheres. Smaller sized neutral

liposomes with equal tobramycin encapsulation efficiency would allow to investigate

this further but the low encapsulation efficiency of tobramycin into neutral liposomes

[24] makes this impossible at present.

149

Conclusion

Both positively- and negatively charged nanospheres show slow subdiffusion in Bcc

biofilms. While positively charged nanospheres likely interact with eDNA, negatively

charged nanospheres probably bind to positively charged matrix components. As

previous research from a collaborating research group has indicated an enrichment of

negatively charged nanospheres close to biofilm clusters, anionic liposomes could

possibly serve as antibiotic delivery systems with high potential for treatment of Bcc

biofilm infections. However, anionic liposomal formulations with tobramycin

concentrations of 4XMIC only displayed bactericidal activity against B. cepacia LMG

1222 biofilms. In contrast, free tobramycin at this concentration showed high

bactericidal activity against all Bcc biofilms tested. The fusion of the negatively charged

liposome with the negatively charged bacterial membrane is likely limited by

electrostatic repulsive forces at close proximity. Although we observed unobstructed

diffusion of neutral nanospheres, the use of a neutral tobramycin formulation did not

result in an increased anti-biofilm effect compared to free tobramycin. Additional

research for the development of an optimal tobramycin carrier to treat Bcc biofilm

infections is needed.

Acknowledgements

This research has been funded by the Interuniversity Attraction Poles Programme

initiated by the Belgian Science Policy Office and by a concerted action grant by Ghent

University.

150

References

1. O'Sullivan, B.P. and S.D. Freedman, Cystic fibrosis. Lancet, 2009. 373(9678): p. 1891-904.

2. Brennan, A.L. and D.M. Geddes, Cystic fibrosis. Curr Opin Infect Dis, 2002. 15(2): p. 175-82.

3. Lyczak, J.B., C.L. Cannon, and G.B. Pier, Lung Infections Associated with Cystic Fibrosis. Clin Microbiol Rev, 2002. 15(2): p. 194-222.

4. Cystic Fibrosis, F., Annual data report to the center directors. 2011, Cystic Fibrosis Foundation patient registry: Bethesda, Maryland.

5. Mahenthiralingam, E., T.A. Urban, and J.B. Goldberg, The multifarious, multireplicon Burkholderia cepacia complex. Nat Rev Microbiol, 2005. 3(2): p. 144-156.

6. Zlosnik, J.E., et al., Mucoid and nonmucoid Burkholderia cepacia complex bacteria in cystic fibrosis infections. Am J Respir Crit Care Med, 2011. 183(1): p. 67-72.

7. Horsley, A. and A.M. Jones, Antibiotic treatment for Burkholderia cepacia complex in people with cystic fibrosis experiencing a pulmonary exacerbation. Cochrane Database Syst Rev, 2012. 10: p. CD009529.

8. Sousa, S.A., C.G. Ramos, and J.H. Leitao, Burkholderia cepacia Complex: Emerging Multihost Pathogens Equipped with a Wide Range of Virulence Factors and Determinants. Int J Microbiol, 2011. 2011.

9. Cunha, M.V., et al., Studies on the involvement of the exopolysaccharide produced by cystic fibrosis-associated isolates of the Burkholderia cepacia complex in biofilm formation and in persistence of respiratory infections. J Clin Microbiol, 2004. 42(7): p. 3052-8.

10. Tseng, B.S., et al., The extracellular matrix protects Pseudomonas aeruginosa biofilms by limiting the penetration of tobramycin. Environ Microbiol, 2013.

11. Giwercman, B., et al., Induction of beta-lactamase production in Pseudomonas aeruginosa biofilm. Antimicrob Agents Chemother, 1991. 35(5): p. 1008-10.

12. Nikaido, H., Outer membrane barrier as a mechanism of antimicrobial resistance. Antimicrob Agents Chemother, 1989. 33(11): p. 1831-6.

13. Nikaido, H., Prevention of drug access to bacterial targets: permeability barriers and active efflux. Science, 1994. 264(5157): p. 382-8.

14. Hancock, R.E., Resistance mechanisms in Pseudomonas aeruginosa and other nonfermentative gram-negative bacteria. Clin Infect Dis, 1998. 27 Suppl 1: p. S93-9.

15. Mugabe, C., et al., Mechanism of enhanced activity of liposome-entrapped aminoglycosides against resistant strains of Pseudomonas aeruginosa. Antimicrob Agents Chemother, 2006. 50(6): p. 2016-22.

16. Beaulac, C., S. Sachetelli, and J. Lagace, In-vitro bactericidal efficacy of sub-MIC concentrations of liposome-encapsulated antibiotic against gram-negative and gram-positive bacteria. J Antimicrob Chemother, 1998. 41(1): p. 35-41.

17. Mugabe, C., A.O. Azghani, and A. Omri, Liposome-mediated gentamicin delivery: development and activity against resistant strains of Pseudomonas aeruginosa isolated from cystic fibrosis patients. J Antimicrob Chemother, 2005. 55(2): p. 269-71.

18. Halwani, M., et al., Bactericidal efficacy of liposomal aminoglycosides against Burkholderia cenocepacia. J Antimicrob Chemother, 2007. 60(4): p. 760-9.

151

19. Chang, H.I. and M.K. Yeh, Clinical development of liposome-based drugs: formulation, characterization, and therapeutic efficacy. Int J Nanomedicine, 2012. 7: p. 49-60.

20. Li, Z., et al., Characterization of nebulized liposomal amikacin (Arikace) as a function of droplet size. J Aerosol Med Pulm Drug Deliv, 2008. 21(3): p. 245-54.

21. Symens, N., et al., Nuclear inclusion of nontargeted and chromatin-targeted polystyrene beads and plasmid DNA containing nanoparticles. Mol Pharm, 2011. 8(5): p. 1757-66.

22. Braeckmans, K., et al., Sizing nanomatter in biological fluids by fluorescence single particle tracking. Nano Lett, 2010. 10(11): p. 4435-42.

23. Braeckmans, K., et al., Single particle tracking, in Nanoscopy and Multidimensional Optical Fluorescence Microscopy, D. A, Editor. 2010, Chapman & Hall/CRC: Boca Raton, FL, USA. p. 1-17.

24. Mugabe, C., A.O. Azghani, and A. Omri, Preparation and characterization of dehydration-rehydration vesicles loaded with aminoglycoside and macrolide antibiotics. Int J Pharm, 2006. 307(2): p. 244-50.

25. Europe, C.o., European Pharmacopoeia. 5th ed. Vol. 5. 2005, France: Edqm. 26. Forier, K., et al., Transport of nanoparticles in cystic fibrosis sputum and bacterial

biofilms by single-particle tracking microscopy. Nanomedicine (Lond), 2012. 8(6): p. 935-49.

27. Peeters, E., H.J. Nelis, and T. Coenye, In vitro activity of ceftazidime, ciprofloxacin, meropenem, minocycline, tobramycin and trimethoprim/sulfamethoxazole against planktonic and sessile Burkholderia cepacia complex bacteria. J Antimicrob Chemother, 2009. 64(4): p. 801-9.

28. Omri, A., Z.E. Suntres, and P.N. Shek, Enhanced activity of liposomal polymyxin B against Pseudomonas aeruginosa in a rat model of lung infection. Biochem Pharmacol, 2002. 64(9): p. 1407-13.

29. Meers, P., et al., Biofilm penetration, triggered release and in vivo activity of inhaled liposomal amikacin in chronic Pseudomonas aeruginosa lung infections. J Antimicrob Chemother, 2008. 61(4): p. 859-68.

30. Seidler, M., S. Salvenmoser, and F.M. Muller, Liposomal amphotericin B eradicates Candida albicans biofilm in a continuous catheter flow model. FEMS Yeast Res, 2010. 10(4): p. 492-5.

31. Novotny, L.A., et al., Structural stability of Burkholderia cenocepacia biofilms in reliant on eDNA structure and presence of a bacterial nucleic acid binding protein. PLoS One, 2013. 8(6).

152

153

PART 4

General discussion

154

155

CF is the most common inherited disease in the European population, affecting

approximately 1 in 2500 live births in Belgium [1]. The disease is caused by mutations

in both cftr alleles, encoding a chloride channel with secretory function in different

organs, including lung, liver, gut, pancreas and sweat glands. The hallmark of CF and

major cause of morbidity and mortality in CF patients are chronic lung infections. Early

lung infections, often by 1 year of age, are mainly caused by H. influenzae, S. aureus or P.

aeruginosa [2]. These early infections can most of the time successfully be cleared,

either with antibiotic therapy or spontaneously. In later stages, P. aeruginosa is

considered the major pathogen recovered from the CF lung and transient infections

often progress to chronic P. aeruginosa infections which are difficult to treat. Although

infections with Bcc species are much less prevalent than P. aeruginosa infections in CF

patients, the clinical outcome of Bcc infections is highly variable and often associated

with a rapid decline in lung function [3]. Some Bcc infected CF patients even experience

“cepacia syndrome”, characterized by necrotizing pneumonia and septicaemia, resulting

in early dead [4]. As Bcc bacteria are intrinsically resistant to many antibiotics, produce

a wide variety of potential virulence factors and have been shown to be capable of

person-to-person transmission, these infections are highly feared among CF patients.

Much effort has already been put in control strategies to limit the spread of these

infections [5] because, once a patient is infected with Bcc, eradication is very difficult [6].

In addition, biofilm formation has been associated with the persistence of Bcc lung

infections as the biofilm mode of growth provides additional protection against

antibiotics and host immune components [7]. Understanding the basis of the increased

resistance associated with biofilm formation is the main goal of many researchers.

In vitro biofilm formation is a common trait of Bcc strains. During biofilm formation,

bacteria produce EPS providing protection against many environmental factors

resulting in increased resistance against both host immune factors and antibiotics.

Although Bcc strains are intrinsically resistant to tobramycin, tobramycin shows high

bactericidal activity against both stationary phase planktonic cultures and sessile cells at

concentrations 4 times higher than the MIC [8]. However, a decreased susceptibility to

tobramycin of sessile cells compared to planktonic cells has been observed [8]. This

decreased susceptibiltiy could be due to a number of biofilm specific factors. A possible

explanation, that was further investigated in chapter VI, is that tobramycin interacts

through the biofilm.

156

As the aminoglycoside antibiotic, tobramycin, is positively charged under physiological

conditions, it can easily bind to negatively charged components of the biofilm. In non-

mucoid P. aeruginosa biofilms, such ionic interactions have already been demonstrated

to limit the penetration of tobramycin in the biofilm, resulting in increased tolerance to

tobramycin [9]. eDNA, an important biofilm matrix component in many bacterial

species, has been shown to protect P. aeruginosa biofilm cells by acting as a shield,

binding aminoglycosides and AMPs [10, 11]. The presence of eDNA in Bcc biofilms was

confirmed, although the amount of eDNA significantly varied among different Bcc

strains. rhDNase, at high concentrations (50 µg/mL), resulted in ≤ 50 % reduction of

eDNA concentrations but, in contrast to what was previously observed in E. coli, S.

aureus and P. aeruginosa [12, 13], it had no effect on Bcc biofilm dispersion. A

combination of tobramycin and rhDNase did not display an increased bactericidal effect

compared to tobramycin alone. Possibly, the eDNA is not sufficiently degraded to

eliminate a potential shielding effect of the eDNA, like observed in P. aeruginosa

biofilms. As rhDNase concentrations used were already up to 5 times higher than

concentrations measured in CF sputum after conventional rhDNase inhalation therapy

[14], we did not investigate the effect of higher rhDNase concentrations. Novotny et al.

suggest targeting DNABII proteins, which have been shown to stabilize eDNA in Bcc

biofilms [15]. DNABII proteins bind to and bend eDNA, contributing to the lattice

structure of eDNA in different bacterial biofilms. A dispersal effect of antisera directed

against these DNABII proteins in B. cenocepacia biofilms has been observed [15]. After

dispersal, the resulting biofilm and planktonic cells displayed an increased susceptibility

to antibiotic therapy. In contrast, treatment of B. cenocepacia biofilms with rhDNase

resulted in more robust biofilms, suggesting a contra-indication of rhDNase therapy in

CF patients infected with B. cenocepacia [15].

In addition to eDNA, exopolysaccharides may also confer tolerance to aminoglycosides

in Bcc biofilms. Exopolysaccharides are known to play an important role in the

adaptation of Bcc to different stress conditions and the type of exopolysaccharides

produced as well as the amount is regulated in response to environmental signals [16].

Cepacian, which is considered the major exopolysaccharide produced by most isolates of

the Bcc, is negatively charged due to the presence of a carboxylate residue in the

polymer repeating unit, allowing electrostatic interaction with tobramycin. The

polymer also contains acetyl substituents and the number of acetyl groups per repeating

157

unit is strain dependent [16]. A protective effect of cepacian against ROS has been

described. ROS, including NaClO, are produced in vivo by neutrophils in order to kill and

degrade infecting microorganisms after phagocytosis. ClO- concentrations in

phagolysosomes have been estimated to be around 100 mM [17]. By scavenging NaClO,

the backbone of cepacian is degraded and acetyl substituents are cleaved of [17]. The

loss of these acetyl substituents hampers the ability of cepacian chains to form duplexes,

reducing the viscosity of the polysaccharide network around Bcc cells [18]. Based on

the effects of ROS on the structure of cepacian, we investigated whether the anti-biofilm

effect of tobramycin would increase after a short (15 min) NaClO (60 µM) pretreatment

of Bcc biofilms. This combination therapy displayed significantly increased bactericidal

activity compared to tobramycin monotherapy in several Bcc biofilms tested, including

B. cenocepacia LMG 16656. As this strain does not produce cepacian [19], it is likely that

the polymer degrading effect of ROS is not limited to cepacian but that ROS also affect

the structural integrity of other polysaccharides present in the biofilm matrix. In vitro

experiments indicated a substantial (3 log reduction) bactericidal effect of NaClO on

planktonic B. pyrrocinia cultures when exposed to a 60 µM NaClO solution for only 15

min [17]. However, the same NaClO treatment did not result in killing of bacteria when

tested on Bcc biofilms (chapter VI). At higher concentrations, starting from 100 µM

NaClO, a small bactericidal effect was observed. The higher concentrations needed to

kill biofilm cells indicate that the biofilm mode of growth offers protection against ROS,

probably by the production of exopolysaccharides. In addition, through the formation of

an exopolysaccharide network around the bacterial cells, bacterial surface antigens are

masked, avoiding recognition by the immune system and subsequent phagocytosis [16].

Therefore, the production of exopolysaccharides by Bcc species likely contributes to the

persistence of these microorganisms and efficient degradation of these polysaccharides

could increase the efficacy of the immune system and of antibiotics.

Electrostatic interactions between tobramycin and the matrix could also be avoided by

adding competitors for binding to matrix components, e.g. divalent cations, as has been

investigated by Tseng et al. [9]. Divalent cations will occupy binding sites to which

tobramycin is also capable of binding, thus improving tobramycin penetration through

the biofilm. However, it has to be taken into account that these divalent cations will also

compete with tobramycin for the same binding sites on LPS in Gram-negative bacteria

(chapter V). The initial step necessary for aminoglycoside activity in Gram-negative

158

bacteria is the binding to LPS followed by subsequent uptake of the antibiotic into the

cell. Divalent cations are well-known to stabilize the outer membrane of Gram-negative

bacteria by cross-linking two adjacent LPS molecules, decreasing outer membrane

permeability and acting as antagonists for aminoglycoside activity. Mg2+ and Ca2+ have

been shown to display the biggest antagonistic effect on gentamycin activity in P.

aeruginosa. In chapter V an antagonistic effect of 25 mM Mg2+ on tobramycin activity

against P. aeruginosa was observed. However, this antagonistic effect could not be

confirmed for Bcc strains. In contrast, for some Bcc strains, an increased susceptibility

to tobramycin in the presence of 6.25 mM Mg2+ was observed. This indicates that the

combination of tobramycin with divalent cations to treat Bcc biofilm infections could

potentially be beneficial. However, the antagonistic effect of these cations on

tobramycin activity against P. aeruginosa suggests a decreased activity of this

combination therapy for the treatment of P. aeruginosa biofilm infections compared to

tobramycin monotherapy. As P. aeruginosa is the major pathogen isolated from CF

patients, the potential of this combination therapy of tobramycin with divalent cations

to treat CF lung infections seems relatively low.

Another component present in Bcc biofilms, contributing to their resistance against

antibiotics and their persistence in the CF lung is PNAG, a capsule polysaccharide. As

PNAG is positively charged, direct electrostatic interaction with tobramycin does not

occur. However, a biofilm produced by a B. multivorans mutant, incapable of producing

PNAG, showed increased susceptibility to tobramycin compared to the wild type [20]. In

chapter VI we tested if a combination of tobramycin and dispersin B, a selective PNAG

degrading enzyme, displayed superior bactericidal activity against Bcc biofilms

compared to tobramycin alone. In some biofilms, the combination of tobramycin with

100 µg/mL dispersin B resulted in a 10-45% additional reduction in biofilm cell

numbers compared to tobramycin treatment alone. This increased effect, due to the

addition of 100 µg/mL dispersin B, which did not cause any dispersal or cell death on its

own, is likely associated with the degradation of a protective polymeric network in close

proximity to the biofilm cells. Positively charged PNAG can possibly form complexes

with polyanionic matrix components, e.g. eDNA and cepacian, increasing the density of

the polymeric network surrounding the bacterial cells, thereby limiting tobramycin

penetration. The lack of an increased anti-biofilm effect in some Bcc biofilms tested

could either be due to the lack of PNAG production or limited PNAG degradation.

159

Besides being a diffusion barrier for antibiotics, the dense polymeric network could also

block the access of dispersin B to PNAG, inhibiting degradation of the latter. In contrast

to PNAG degradation therapy, in which the polymer is seen as an obstacle to efficient

antibiotic therapy, the presence of PNAG in Bcc biofilms could also positively contribute

to the treatment of Bcc infection. PNAG is produced on the surface of various multidrug

resistant bacteria, including MRSA and members of the Bcc. Thereby, immunotherapy

based on PNAG as an antigen could provide a broad-spectrum therapy for either treating

patients infected with PNAG-producing bacteria or as prophylaxis in patients at risk of

developing multidrug resistant bacterial infections [21]. A fully human monoclonal

antibody directed to PNAG has already successfully completed phase I clinical trials [21].

The development of a liposomal tobramycin delivery system could also prevent

interactions between the antibiotic and matrix components, possible increasing the

penetration rate of tobramycin in the biofilm (chapter VII). The mobility of neutral,

anionic and cationic nanospheres in Bcc biofilms was used as a model to predict the

mobility of liposomal delivery systems in these biofilms. Previous research from the

Laboratory of General Biochemistry & Physical Pharmacy in Ghent observed free

diffusion of neutral nanospheres in B. multivorans LMG 18825 biofilms and obstructed

diffusion of both anionic an cationic nanospheres [22]. The subdiffusion of these anionic

and cationic nanospheres was confirmed in all Bcc biofilms tested (chapter VII).

Cationic particles were observed to interact with fiber-like structures, likely eDNA,

which suggests that the use of cationic liposomes as delivery systems for tobramycin

would benefit from co-administration of the liposomal tobramycin formulation with

rhDNase. As rhDNase therapy has been suggested to be contra-indicated in CF patients

infected with Bcc [15] this was not further investigated. As anionic particles were

immobilized in close proximity to biofilm cell cluster, the use of anionic liposomal

carriers for tobramycin looked promising in terms of “targeted” drug delivery.

Unfortunately, tobramycin encapsulated in anionic liposomes displayed less activity

towards Bcc biofilms than free tobramycin. Possibly, the fusion of the anionic liposome

with the negatively charged bacterial cell surface is inhibited by repulsive forces. The

passive diffusion of tobramycin out of the liposomes is expected to be slow as the lipid

membranes of the liposome present with low permeability for hydrophilic drugs.

Therefore, without triggered release of tobramycin, e.g. by heat or enzymatic reaction,

the use of anionic delivery systems for tobramycin to treat Bcc biofilm infections would

160

present with low tobramycin bioavailability. However, phospholipases, which are

capable of degrading anionic liposomes [23], have been detected in CF sputum [24].

This suggests that the lack of an in vitro bactericidal effect of tobramycin encapsulated in

anionic liposomes against Bcc biofilms, does not necessarily mean there could be no

activity in vivo. Currently, Arikace, a neutral liposomal formulation of amikacin, is being

evaluated in phase II clinical trials for the inhalation treatment of pulmonary infections

with Gram-negative bacteria [25]. Arikace presented with superior activity against P.

aeruginosa biofilms compared to free amikacin and the penetration of these liposomes

(250-300 nm) in both CF sputum and P. aeruginosa biofilms has been demonstrated

[26]. In B. multivorans LMG 18825 biofilms, the activity of both free and liposomal

encapsulated tobramycin (same lipid composition as Arikace) was similar. The

observed bactericidal effect indicates that the liposomes are capable of reaching the

biofilm cells and releasing their content. However, the activity of the neutral liposomal

tobramycin formulation against Bcc biofilms was observed to be strain specific. This

could indicate either differences in penetration of the liposome through different Bcc

biofilms or differences in release of the encapsulated tobramycin. It is well-known that

the composition of the biofilm matrix is strain-specific, resulting in strain-dependent

diffusion of the relatively large liposomal carriers (400 nm) among Bcc strains. In

addition, exopolysaccharides in close proximity to the cell surface, e.g. PNAG, could

prevent fusion between the liposomal carrier and the bacterial surface. Triggered

release of liposome contents, like has been suggested in P. aeruginosa biofilms, could

increase the bioavailability of the encapsulated tobramycin in Bcc biofilms. In P.

aeruginosa biofilms, the release of amikacin from the Arikace liposomal formulation is

believed to be mediated by rhamnolipids, glycolipids often referred to as bacterial

surfactants produced by P. aeruginosa [26]. Rhamnolipid production has also been

observed in various Burkholderia species, but not in the Bcc [27]. Additional research is

needed to develop a good drug delivery system for tobramycin to increase the

effectiveness of eradication of Bcc in CF patients, in particular focusing on higher

encapsulation efficiency of tobramycin in smaller liposomes and triggered drug release.

Today, antibiotic resistance among pathogenic microorganisms is increasing,

confronting clinicians with the inability to eradicate many life-threatening infections. As

is the case for Bcc infections, resistance can be both intrinsic and associated with the

capability of forming biofilms. Until today, no guidelines for optimal antibiotic

161

treatment regimen of CF patients suffering from chronic Bcc infections are available

[28]. The discovery of new targets or the improvement of existing antibiotic treatments,

for example by means of combination therapy, different administration routes or

different antibiotic formulations, could drastically improve the prognosis of CF patients

infected with Bcc.

162

References

1. Farrell, P.M., The prevalence of CF in the European Union. J Cyst Fibros, 2008. 7(5): p. 450-3.

2. CF, F., Annual data report to the center directors. 2011, CF Foundation patient registry: Bethesda, Maryland.

3. Leitao, J.H., et al., Pathogenicity, virulence factors, and strategies to fight against Burkholderia cepacia complex pathogens and related species. Appl Microbiol Biotechnol, 2010. 87(1): p. 31-40.

4. Mahenthiralingam, E., T.A. Urban, and J.B. Goldberg, The multifarious, multireplicon Burkholderia cepacia complex. Nat Rev Microbiol, 2005. 3(2): p. 144-156.

5. Saiman, L. and J. Siegel, Infection control in CF. Clin Microbiol Rev, 2004. 17(1): p. 57-71.

6. Sousa, S.A., C.G. Ramos, and J.H. Leitao, Burkholderia cepacia Complex: Emerging Multihost Pathogens Equipped with a Wide Range of Virulence Factors and Determinants. Int J Microbiol, 2011. 2011.

7. Cunha, M.V., et al., Studies on the involvement of the exopolysaccharide produced by CF-associated isolates of the Burkholderia cepacia complex in biofilm formation and in persistence of respiratory infections. J Clin Microbiol, 2004. 42(7): p. 3052-8.

8. Peeters, E., H.J. Nelis, and T. Coenye, In vitro activity of ceftazidime, ciprofloxacin, meropenem, minocycline, tobramycin and trimethoprim/sulfamethoxazole against planktonic and sessile Burkholderia cepacia complex bacteria. J Antimicrob Chemother, 2009. 64(4): p. 801-9.

9. Tseng, B.S., et al., The extracellular matrix protects Pseudomonas aeruginosa biofilms by limiting the penetration of tobramycin. Environ Microbiol, 2013.

10. Chiang, W.C., et al., Extracellular DNA Shields against Aminoglycosides in Pseudomonas aeruginosa Biofilms. Antimicrob Agents Chemother, 2013. 57(5): p. 2352-61.

11. Mulcahy, H., L. Charron-Mazenod, and S. Lewenza, Extracellular DNA Chelates Cations and Induces Antibiotic Resistance in Pseudomonas aeruginosa Biofilms. PLoS Pathog, 2008. 4(11).

12. Tetz, V.V. and G.V. Tetz, Effect of extracellular DNA destruction by DNase I on characteristics of forming biofilms. DNA Cell Biol, 2010. 29(8): p. 399-405.

13. Whitchurch, C.B., et al., Extracellular DNA required for bacterial biofilm formation. Science, 2002. 295(5559): p. 1487.

14. Sanders, N.N., et al., Role of magnesium in the failure of rhDNase therapy in patients with CF. Thorax, 2006. 61(11): p. 962-8.

15. Novotny, L.A., et al., Structural stability of Burkholderia cenocepacia biofilms in reliant on eDNA structure and presence of a bacterial nucleic acid binding protein. PLoS One, 2013. 8(6).

16. Ferreira, A.S., et al., Insights into the role of extracellular polysaccharides in Burkholderia adaptation to different environments. Front Cell Infect Microbiol, 2011. 1: p. 16.

17. Cuzzi, B., et al., Investigation of bacterial resistance to the immune system response: cepacian depolymerisation by reactive oxygen species. Innate Immun, 2012. 18(4): p. 661-71.

163

18. Herasimenka, Y., et al., Macromolecular properties of cepacian in water and in dimethylsulfoxide. Carbohydr Res, 2008. 343(1): p. 81-9.

19. Ferreira, A.S., et al., Distribution of Cepacian Biosynthesis Genes among Environmental and Clinical Burkholderia Strains and Role of Cepacian Exopolysaccharide in Resistance to Stress Conditions. Appl Environ Microbiol, 2010. 76(2): p. 441-50.

20. Yakandawala, N., et al., Characterization of the poly-beta-1,6-N-acetylglucosamine polysaccharide component of Burkholderia biofilms. Appl Environ Microbiol, 2011. 77(23): p. 8303-9.

21. Roux, D., G.B. Pier, and D. Skurnik, Magic bullets for the 21st century: the reemergence of immunotherapy for multi- and pan-resistant microbes. J Antimicrob Chemother, 2012. 67(12): p. 2785-7.

22. Forier, K., et al., Transport of nanoparticles in CF sputum and bacterial biofilms by single-particle tracking microscopy. Nanomedicine (Lond), 2012.

23. Davidsen, J., et al., Drug delivery by phospholipase A(2) degradable liposomes. Int J Pharm, 2001. 214(1-2): p. 67-9.

24. Sahu, S. and W.S. Lynn, Enzymes of phospholipid metabolism in airway secretions of patients with asthma, CF, and alveolar proteinosis. Inflammation, 1977. 2(2): p. 93-104.

25. Li, Z., et al., Characterization of nebulized liposomal amikacin (Arikace) as a function of droplet size. J Aerosol Med Pulm Drug Deliv, 2008. 21(3): p. 245-54.

26. Meers, P., et al., Biofilm penetration, triggered release and in vivo activity of inhaled liposomal amikacin in chronic Pseudomonas aeruginosa lung infections. J Antimicrob Chemother, 2008. 61(4): p. 859-68.

27. Abdel-Mawgoud, A.M., F. Lepine, and E. Deziel, Rhamnolipids: diversity of structures, microbial origins and roles. Appl Microbiol Biotechnol, 2010. 86(5): p. 1323-36.

28. Horsley, A. and A.M. Jones, Antibiotic treatment for Burkholderia cepacia complex in people with CF experiencing a pulmonary exacerbation. Cochrane Database Syst Rev, 2012. 10: p. CD009529.

164

165

PART 5

Summary-Samenvatting

166

167

Summary

The Bcc is a group of at least 18 closely related opportunistic pathogens that are able to

infect the respiratory tract of CF patients. Although infections with Bcc species are much

less prevalent than P. aeruginosa infections in CF patients, the clinical outcome of Bcc

infections is highly variable and often associated with a rapid decline in lung function.

Due to their intrinsic resistance to many antibiotics and the production of a broad range

of virulence factors, Bcc lung infections are very difficult to treat. In addition, biofilm

formation has been associated with the persistence of Bcc infections in the CF lung, as

the biofilm mode of growth provides additional protection against antibiotics and host

immune components. The aim of this thesis was to search for an antibiotic treatment

with superior Bcc anti-biofilm efficacy compared to tobramycin.

The first three chapters (Part 1) present an overview of the literature concerning CF,

biofilms and the Bcc; three topics related to the content of this thesis.

In chapter IV the potential of fosmidomycin to treat Bcc infections was investigated.

Fosmidomycin inhibits Dxr, a key enzyme in the non-mevalonate pathway, which is

essential for isoprenoid synthesis in Bcc. The antimicrobial activity of fosmidomycin

and eight fosmidomycin derivatives towards forty Bcc strains was investigated. All Bcc

strains were resistant, although the addition of glucose-6-phosphate reduced MIC values

of FR900098, the fosmidomycin acetyl derivative, from 512 mg/L to 64 mg/L for B.

multivorans and B. cepacia. This increased susceptibility in the presence of glucose-6-

phosphate was associated with an increased expression of the genes involved in

glycerol-3-phosphate transport, likely the only route for fosmidomycin import in the

Bcc. In addition, the upregulation of fsr, encoding an efflux pump, was observed during

treatment with fosmidomycin and FR900098. Therefore, the lack of activity of

fosmidomycin and fosmidomycin derivatives against Bcc bacteria is likely due to an

insufficient uptake accompanied by increased efflux of these antibiotics in Bcc cells.

Chapter V deals with the mode of uptake of tobramycin in different Gram-negative

bacteria, including E. coli, P. aeruginosa and Bcc bacteria. The cellular uptake of

aminoglycosides, including tobramycin, can be divided in three phases: an initial energy-

independent electrostatic interaction at the cell surface and subsequently two energy-

dependent phases. However, fundamental differences exist in aminoglycoside uptake

between different Gram-negative bacteria. Aminoglycosides can cross the outer

membrane either by diffusion through hydrophilic channels formed by porin proteins or

168

by a process referred to as self-promoted uptake. In order to determine whether self-

promoted uptake is involved, the activity of tobramycin in the presence and absence of

Mg2+ was investigated. Increased tobramycin MICs in the presence of Mg2+ confirmed

self-promoted uptake in P. aeruginosa. In contrast, decreased MICs in the presence of

Mg2+ for Bcc organisms suggested alternative uptake mechanisms. Tobramycin activity

against E. coli cells was unaltered in the presence of Mg2+. In order to further investigate

the mode of tobramycin uptake in these organisms, the activity of three tobramycin

derivatives, i.e. tobramycin-OH-BODIPY, tobramycin-N-Me2 and tobramycin-OH-EDTA,

on both E. coli and P. aeruginosa planktonic and biofilm cells was investigated. Based on

the MIC of tobramycin-OH-BODIPY for E. coli (32 µg/mL) and the high activity of

tobramycin-OH-EDTA towards E. coli biofilms, we hypothesized that tobramycin is at

least in part taken up through self-promoted uptake in E. coli, as both derivatives are too

large to be taken up through porins. However, the high MIC of tobramycin-OH-EDTA for

E. coli (512 µg/mL) could not be explained. The observed selective activity of

tobramycin-N-Me2 against P. aeruginosa ATCC 27853 biofilms could be promising for

treating P. aeruginosa lung infections in CF. By selectively targeting the pathogenic

species, the commensal flora of the patients will be less disturbed than by traditional

antibiotics. This will reduce the number and frequency of side effects as well as the

selection of antibiotic-resistant organisms associated with antibiotic therapy.

The contribution of Bcc biofilm matrix components, including eDNA, cepacian and PNAG,

to tobramycin susceptibility was investigated in chapter VI. To this end, the in vitro

bactericidal activity of tobramycin in combination with rhDNase, NaClO and dispersin B

was tested against different Bcc biofilms. Exopolysaccharide degradation by NaClO

pretreatment and specific PNAG degradation by dispersin B significantly increased the

bactericidal effect of tobramycin against some of the Bcc biofilms tested, including

strains of B. cenocepacia, B. cepacia and B. metallica. In contrast, the presence of

rhDNase during biofilm treatment and/or development had no influence on tobramycin

activity against Bcc biofilms. These results indicate that exopolysaccharides likely play a

role in tobramycin susceptibility of Bcc biofilms and that a combination of tobramycin

with exopolysaccharide degrading components could improve the eradication of Bcc

biofilm infections.

In chapter VII we investigated the potential of treating Bcc biofilms with liposomal

tobramycin formulations. Single particle tracking was used to study the transport of

169

positive and negative nanospheres in Bcc biofilms as a model for the transport of

liposomes. Negative nanospheres were immobilized close to biofilm cell clusters while

positive nanopsheres interacted with fiber-like structures, probably eDNA. Based on

these data, encapsulating tobramycin into negatively charged liposomes looked

promising in terms of targeted drug delivery. However, the anti-biofilm effect of

tobramycin did not increase compared to free tobramycin when it was encapsulated

into neutral or anionic liposomes. Likely, the fusion of the anionic liposomes with the

negatively charged bacterial surface of Bcc was limited by electrostatic repulsive forces

at close proximity. The lack of a substantial anti-biofilm effect of tobramycin

encapsulated into neutral liposomes could be further investigated by increasing the

liposomal tobramycin concentration. However, this was hampered by the low

encapsulation efficiency of tobramycin in these liposomes.

170

171

Samenvatting

Het Bcc is een groep van minstens 18 nauw verwante species die luchtweginfecties bij

mucoviscidose patiënten kunnen veroorzaken. Hoewel de prevalentie van Bcc infecties

laag is in vergelijking met P. aeruginosa infecties en de symptomen variabel zijn,

resulteren Bcc infecties vaak in een snelle achteruitgang van de longfunctie. Bcc

infecties zijn moeilijk te behandelen omwille van hun intrinsieke resistentie tegen vele

antibiotica en door de productie van een brede waaier aan virulentie factoren. De

mogelijkheid om een biofilm te vormen, die extra bescherming biedt tegen antibiotica en

componenten van het immuunsysteem van de patiënt, draagt bij tot de persistentie van

Bcc infecties in mucoviscidosepatiënten. Het doel van dit doctoraatsproefschrift was om

op zoek te gaan naar een verbeterde antibioticatherapie om Bcc biofilm infecties in

mucoviscidosepatiënten te behandelen.

De eerste drie hoofdstukken (deel 1) geven een overzicht van de literatuur over

mucoviscidose, biofilms en het Bcc; drie onderwerpen met betrekking tot de inhoud van

deze doctoraatsthesis.

In hoofdstuk IV wordt het potentieel van fosmidomycine om Bcc infecties te

behandelen nagegaan. Fosmidomycine inhibeert Dxr, een sleutelenzym in de non-

mevalonaat pathway welke essentieel is voor de productie van isoprenoïden in Bcc. De

anti-microbiële activiteit van fosmidomycine en acht fosmidomycine derivaten tegen

veertig Bcc stammen werd onderzocht. Alle Bcc stammen vertoonden resistentie,

hoewel het toevoegen van glucose-6-fosfaat de MIC waarden van FR900098, een

fosmidomycine acetyl derivaat, deed dalen van 512 mg/L tot 64 mg/L voor B.

multivorans en B. cepacia. Deze verhoogde gevoeligheid in aanwezigheid van glucose-6-

fosfaat ging gepaard met een verhoogde expressie van genen die betrokken zijn in het

transport van glycerol-3-fosfaat. Het glycerol-3-fosfaat transport systeem lijkt de enige

weg voor fosmidomycine opname in Bcc cellen te zijn. Tijdens behandeling met

fosmidomycine en FR900098 werd ook een opregulatie van fsr, dat codeert voor een

effluxpomp, waargenomen. De resistentie tegen fosmidomycine en fosmidomycine

derivaten in Bcc organismen is dus waarschijnlijk te wijten aan onvoldoende opname

gecombineerd met een verhoogde efflux van deze antibiotica.

In hoofdstuk V wordt de opname van tobramycine in verschillende Gram-negatieve

micro-organismen besproken, waaronder E. coli, P. aeruginosa and Bcc species. De

opname van aminoglycosiden in Gram-negatieve micro-organismen kan onderverdeeld

172

worden in 3 fasen: een initiële elektrostatische interactie ter hoogte van de

celmembraan die energie-onafhankelijk is, gevolgd door twee energie-afhankelijke

fasen. Toch bestaan nog fundamentele verschillen tussen Gram-negatieve bacteriën

onderling. Aminoglycosiden kunnen de buitenste membraan doorkruisen via diffusie

doorheen hydrofiele kanalen, gevormd door porines, of via een “self-promoted uptake”

mechanisme. Het eventueel gebruik van het “self-promoted uptake” mechanisme wordt

onderzocht door de activiteit van tobramycine in aan- en afwezigheid van Mg2+ te

bepalen. De verhoogde MIC waarden voor P. aeruginosa in de aanwezigheid van Mg2+

bevestigen het gebruik van “self-promoted uptake”. Dit in tegenstelling tot de lagere

MIC waarden voor Bcc species in de aanwezigheid van Mg2+, die een andere manier van

opname suggereren. De activiteit van tobramycin tegen E. coli bleef ongewijzigd in aan-

of afwezigheid van Mg2+. Om de opname van tobramycine verder te bestuderen werd de

activiteit van drie tobramycine derivaten, (tobramycine-OH-BODIPY, tobramycine-N-

Me2 en tobramycine-OH-EDTA) tegen E. coli en P. aeruginosa planktonische en biofilm

cellen onderzocht. Gebaseerd op de MIC waarde van tobramycine-OH-BODIPY voor E.

coli (32 µg/mL) en de hoge activiteit van tobramycine-OH-EDTA tegen E. coli biofilms,

werd verondersteld dat tobramycine op zijn minst gedeeltelijk wordt opgenomen via

“self-promoted uptake” in E. coli. Deze hypothese werd voorop gesteld omdat deze

derivaten te groot zijn om te worden opgenomen via porines en ze toch activiteit

vertonen. De hoge MIC waarde van tobramycine-OH-EDTA voor E. coli (512 µg/mL) kon

niet worden verklaard. Er werd ook een selectieve activiteit van tobramycine-N-Me2

tegen P. aeruginosa ATCC 27853 biofilms waargenomen. Een selectief antibioticum zal

de commensale flora van de patiënt na antibiotica behandeling minder verstoren dan

een traditioneel antibioticum. Hierdoor kan het aantal en de frequentie van

nevenwerkingen geassocieerd met antibioticatherapie en ook de selectie van

antibiotica-resistentie micro-organismen teruggedrongen worden.

De invloed van matrix componenten, waaronder extracellulair DNA, cepacian en PNAG,

aanwezig in de biofilm van Bcc bacteriën, werd onderzocht in hoofdstuk VI. De in vitro

activiteit van tobramycine in combinatie met rhDNase, NaClO en dispersin B werd getest

tegen verschillende Bcc biofilms. De degradatie van exopolysacchariden door een korte

NaClO voorbehandeling en de specifieke degradatie van PNAG door dispersin B

resulteerde in een significant verhoogde activiteit van tobramycine in meerdere Bcc

biofilms, waaronder deze gevormd door B. cenocepacia, B. cepacia and B. metallica

173

stammen. De aanwezigheid van rhDNase tijdens biofilm vorming en tijdens

tobramycine behandeling van een 24u oude biofilm had geen effect op de activiteit van

tobramycine in Bcc biofilms. Hieruit werd besloten dat exopolysacchariden

hoogstwaarschijnlijk een rol spelen in de gevoeligheid van Bcc biofilms voor

tobramycine en dat een combinatie van tobramycine met componenten die in staat zijn

om exopolysacchariden af te breken de eradicatie van deze biofilms mogelijks kan

verbeteren.

In hoofdstuk VII wordt het potentieel van liposomaal tobramycine om Bcc biofilms te

behandelen besproken. Er werd gebruik gemaakt van single particle tracking om het

transport van positieve en negatieve nanopartikels in Bcc biofilms te bestuderen als

model voor het transport van liposomen. Negatieve partikels werden geïmobiliseerd

rondom cel clusters terwijl positieve partikels interactie vertoonden met draadachtige

structuren, waarschijnlijk extracellulair DNA. De encapsulatie van tobramycine in

negatief geladen liposomen leek dus veelbelovend om tobramycine af te leveren dicht bij

de biofilm cellen. Helaas werd geen verhoogde anti-biofilm activiteit waargenomen van

tobramycine geëncapsuleerd in negatief geladen liposomen, noch in neutrale liposomen.

Mogelijks word de fusie van het negatief geladen liposoom met het negatief geladen

buitenste membraan van Bcc cellen verhinderd door elektrostatische afstoting op korte

afstand. Het gebrek aan een substantieel anti-biofilm effect van tobramycine

geëncapsuleerd in neutrale liposomen kan verder onderzocht worden door de

concentratie aan geëncapusleerd tobramycine te verhogen. Dit wordt echter

belemmerd door de lage encapsulatie efficiëntie van tobramycine in deze liposomen.

174

Curriculum Vitae

General information

Full name: Anne-Sophie Messiaen

Date of birth: March, 3th,1986

Place of birth: Kortrijk

Educational background

2007-2009: Ghent University, Master of drug development.

Thesis: “Isolatie van een niet beschreven campylobacter taxon bij paarden.”

2009-2013: Ghent University, Ph.D. in Pharmaceutical Sciences (Laboratory of Pharmaceutical

Microbiology).

Thesis: “Towards improvement of antibiotic therapy for treating Burkholderia cepacia complex

biofilm infections in cystic fibrosis patients.”

List of publications

• Forier K, Messiaen AS, Coenye T, Nelis H, Celen S, Van Calenbergh S, De Smedt S,

Demeester J, Braeckmans K. (2010). Transport and delivery of antimicrobial agents in

Burkholderia biofilms. J Control Release. 2010 Nov 20;148(1).

• Messiaen AS, Verbrugghen T, Declerck C, Ortmann R, Schlitzer M, Nelis H, Van

Calenbergh S, Coenye T. (2011). Resistance of the Burkholderia cepacia complex to

fosmidomycin and fosmidomycin derivatives. Int J Antimicrob Agents. 2011 Sep;38(3).

• Forier K, Messiaen AS, Raemdonck K, Deschout H, Rejman J, De Baets F, Nelis H, De

Smedt SC, Demeester J, Coenye T, Braeckmans K. (2013). Transport of nanoparticles in

cystic fibrosis sputum and bacterial biofilms by single-particle tracking microscopy.

Nanomedicine (Lond). 2013 Jun;8(6).

• Messiaen AS, Nelis H, Coenye T. (2013). Investigating the role of matrix components in

protection of Burkholderia cepacia complex biofilms against tobramycin. J Cyst Fibros.

2013 Aug 8.

• Messiaen AS, Forier K, Nelis H, Braeckmans K, Coenye T. (2013). Transport of

nanoparticles and tobramycin-loaded liposomes in Burkholderia cepacia complex

biofilms. Plos One, accepted sept 2013.

Poster presentations

• Messiaen AS, Declerck C, Verbrugghen T, Nelis H, Van Calenbergh S, Coenye T. (2010).

Biofilm resistance of the Burkholderia cepacia complex against fosmidomycin and

fosmidomycin derivatives. Biofilms IV. September 1-3, 2010, Winchester, United

Kingdom.

• Messiaen AS, Celen S, Forier K, Risseeuw M, Braeckmans K, Nelis H, Van Calenbergh S,

Coenye T. (2011). Development and evaluation of tobramycin derivatives with selective

activity against Pseudomonas aeruginosa biofilms. Eurobiofilms. July 6-8, Copenhagen,

Denmark.

• Messiaen AS, Forier K, Nesis H, Demeester J, De Smeth S, Braeckmans K, Coenye T.

(2012). Singel particle tracking in Burkholderia cepacia complex biofilms as a tool to

study transport of drug-loaded liposomes. 6th

ASM conference on biofilms. September

29 – October 4, 2012, Miami, USA.

• Messiaen AS, Nelis H, Coenye T. (2013). Matrix components provide protection against

tobramycin in Burkholderia cepacia complex biofilms. Eurobiofims. September 9-12,

2013, Ghent, Belgium.

Oral presentation

• Messiaen AS, Forier K, Nelis H, Demeester J, De Smeth S, Braeckmans K, Coenye T.

(2012). Single particle tracking in Burkholderia cepacia complex biofilms as a tool to

study transport of drug-loaded liposomes. BSM annual symposium, November 30, 2012,

Brussels, Belgium.

Scientific activities

Supervision of students:

• Charlotte Declerck (2009). De susceptibiliteit van het Burkholderia cepacia complex voor

fosmidomycine en fosmidomycine analogen.

• Tine Vandermeulen (2010). Tobramycine opname bij bacteriële pathogenen betrokken

in luchtweginfecties bij mucoviscidose.

• Laure Pape (2011). In vitro evaluatie van de combinatie van antibiotica en pulmozyme

voor de behandeling van biofilm gerelateerde luchtweginfecties bij mucoviscidose

patiënten.

Assistance in practical courses of microbiology (2009-2012)