EXOPOLYSACCHARIDES OF THE

PSEUDOMONAS AERUGINOSA BIOFILM MATRIX

A Thesis

Presented to

The Honors Tutorial College

Ohio University

In Partial Fulfillment of the Requirements for Graduation from the Honors Tutorial College with the degree of

Bachelor of Science in Biological Sciences

by

Elizabeth Mathias

April 2014

This thesis has been approved by

The Honors Tutorial College and the Department of Biological Sciences

Dr. Donald Holzschu Associate Professor, Biological Sciences Thesis Adviser

Dr. Soichi Tanda Honors Tutorial College, DOS Biological Sciences

Jeremy Webster Dean, Honors Tutorial College

1

Table of Contents

Abstract 2

Introduction 4

Pseudomonas aeruginosa Significance 4

Biofilms 12

The Matrix Exopolysaccharides 22

Alginate 22

Psl and Pel 35

Experimental Overview and Research Questions 49

Materials and Methods 52

Results 67

Discussion 82

Conclusion 88

Acknowledgements 89

References 90

2

Abstract Pseudomonas aeruginosa is a common bacterium in the environment. In humans, it is an opportunistic pathogen that is a prevalent cause of hospital-acquired infections. P. aeruginosa is also a frequent cause of pulmonary infections in patients with cystic fibrosis. The majority of patients with cystic fibrosis develop chronic infections of P. aeruginosa that cannot be eradicated. P. aeruginosa is the primary cause of pneumonia in patients with cystic fibrosis. Chronic infection of the cystic fibrosis lung with P. aeruginosa is associated with a major decline in lung function and eventually leads to death in most cases. P. aeruginosa infection of the lung is particularly difficult to treat because these bacteria typically grow in biofilms.

Biofilms are complex bacterial communities in which the bacteria are encased in a thick matrix composed mostly of secreted exopolysaccharides. Bacteria within biofilms are protected from environmental stresses as well as from the host immune system and antimicrobials. P. aeruginosa produces three exopolysaccharides that contributes to the biofilm matrix: alginate, Psl, and Pel. Alginate confers additional protection to antimicrobials and the immune system while Psl and Pel contribute to aggregation and adherence.

The goal of this research was to use a reporter gene under the control of key promoters from the exopolysaccharides’ biosynthetic operons to elucidate the regulation of exopolysaccharide production over the course of biofilm development.

EGFP was placed under the control of the AlgD, PelA, and PslA reporters. Because the PelA and PslA reporters are not well defined, three variations of sequence length upstream of the transcriptional start site were used. These constructs were integrated 3

into P. aeruginosa strains of different morphotypes in order to examine whether promoter activity differs between strains and between biosynthetic operons. Biofilms were then microscopically imaged. However, significant changes in the expression of the reporter were not yet detected.

4

Introduction

1. Pseudomonas aeruginosa Significance

Pseudomonas aeruginosa is a Gram-negative, aerobic gammaproteobacterium that is commonly found in soil and water (Iglewski 1996). It is one of the most well- known of the nearly 200 species in the genus Pseudomonas. Rod-shaped and measuring 0.5-0.8 µm by 1.5-3.0 µm, these bacteria typically have one polar flagellum for motility as well as pili. P. aeruginosa is a pathogen of both plants and humans. Not usually found in the normal flora of humans, it is opportunistic and typically affects damaged tissue or individuals who are immunocompromised. P. aeruginosa was first identified in the mid-19th century on surgical wound dressings due to blue and green discoloration on the bandages caused by pigments that these bacteria produce. Today, it is notorious as an antibiotic resistant pathogen. The frequency of the isolation of multidrug resistant P. aeruginosa has increased 10% in almost 20 years (Lister et al.

2009). With a fatality rate of 50%, P. aeruginosa causes 15% of Gram-negative bacteremia as well as 10% of hospital-acquired infections (Iglewski 1996; Aloush

2006). It is the fourth leading cause of nosocomial infections and the second most frequent cause of nosocomial pneumonia (Lister et al. 2009). P. aeruginosa is also a leading cause of pneumonia and chronic infection of the lung in individuals with cystic fibrosis (CF). In 80-95% of patients with CF, respiratory failure ultimately resulting in death is caused by chronic bacterial infection (Lyczak et al. 2002).

Because of the incidence and clinical significance of P. aeruginosa infection, this pathogen is being widely studied. 5

P. aeruginosa is a notable pathogen not only because of the number and severity of P. aeruginosa infections, but also because of the wide range of tissues in which it causes infection. It is a significant cause of wound and urinary tract infections as well as pneumonia, otitis externa, keratitis, and folliculitis (Gellatly and Hancock

2013). Additionally, P. aeruginosa is a significant cause of hospital-acquired infection because its resistance to antimicrobials and its ability to survive in low- nutrient environments make it difficult to eradicate. The ability to infect many tissues and survive in both natural and hospital environments is due to the great adaptability of P. aeruginosa. P. aeruginosa is extremely metabolically versatile; it is able to catabolize a vast number of hydrocarbons, giving it the ability to survive in even the most nutrient poor settings, including surfaces in medical facilities and water lines.

The P. aeruginosa genome contributes to its adaptability and metabolic versatility. P. aeruginosa has one of the largest known bacterial genomes, containing ‘core’ regions of conserved genes across different strains as well as regions of genomic plasticity.

Within the genome, 10% of genes encode regulatory proteins (Gellatly and Hancock

2013). Many of these proteins are part of two-component regulatory systems, which are signal transduction pathways that facilitate rapid adaptation to environmental changes. Several regulatory systems control quorum sensing, biofilm formation, and virulence factors. This adaptability allows P. aeruginosa to survive in many environments, to be a pathogen of both plants and humans, and to colonize and cause infection in a wide range of tissues in humans.

P. aeruginosa is a significant cause of chronic lung infection in patients with 6

CF. CF is a homozygous recessive disease that is caused by defects in the cystic fibrosis transmembrane conductance regulator (CFTR), which is a cAMP-dependent chloride channel in the apical membrane of epithelial cells (Goldberg and Pier 2000).

Over 900 different homozygotic mutations in the CFTR gene can cause CF, and the effects can be seen in a wide range of tissues including pancreatic, gastrointestinal, and, most noticeably, pulmonary. Patients with CF are often diagnosed due to elevated salt levels in their sweat, and CF is associated with low body mass due to an inability to properly digest and absorb nutrients in the gut. In the lungs, CF is characterized by thickened airway surface liquid and a dehydrated mucus layer (Fig. 1). These secretions can harbor microbes and inhibit the ability of the immune response to eradicate them (Gellatly and Hancock 2013). Mucociliary clearance, an important mechanism for clearing microbes and other foreign substances in the lung, is inhibited because the mucus is more strongly adhered to the epithelial surface, thereby retarding the movement of beating cilia. Additionally, the high ionic strength of the mucus and lung secretions may inhibit the activity of antimicrobial peptides (Lyczak et al. 2002).

Damage to the lung tissue associated with P. aeruginosa infection is largely caused by chronic inflammation (Lyczak et al. 2002). While P. aeruginosa from chronic lung infections typically downregulate virulence factors and lose inflammatory features, P. aeruginosa infection of the lung is marked by recruitment of neutrophils; these release reactive oxygen species and other molecules that, during the course of a chronic infection, lead to long-term inflammation and significant tissue damage. Proteases, exotoxin A, and the pigment pyocyanin secreted by P. aeruginosa 7

also contribute to this damage. P. aeruginosa infections of the CF lung that are untreated often lead to the early death of the patient; with extensive treatment, life expectancy is increased to 35 years (Gaspar et al. 2013).

Fig. 1: Differences in the secretions of normal and CF-affected lungs (Lyczak et al. 2002). The dehydrated mucus layer of the CF lung results in increased bacterial adherence and decreased mucociliary clearance, causing chronic infection.

Because of its prevalence, the development of P. aeruginosa infection in the

CF lung has been well studied. It is thought that the lung is typically first exposed to

P. aeruginosa from the environment within three years of birth; the lung is also exposed to other common pathogens including Staphylococcus aureus and

Haemophilus influenza (Fig. 2; Lyczak et al. 2002). Pathogens are typically inhaled and then settle into the secretions in the obstructed airways of the lung (Lyczak et al.

2002; Gellatly and Hancock 2013). Serological studies have indicated that P. aeruginosa is present in CF patients prior to their fourth year of age in 97.5% of cases 8

(Burns et al. 2001). However, these bacteria are not often recovered from lung secretions, indicating that P. aeruginosa is being eradicated by the immune system.

Through these early years of infection, P. aeruginosa is not the predominant bacterium in the CF lung; S. aureus is typically more prevalent (Lyczak et al. 2002).

The pathology of S. aureus in CF lung infection is not clear. Patients are often treated when these bacteria are found in clinical isolates, despite the fact that treatment and clearance of S. aureus has not been shown to have a positive effect on the patient’s prognosis. This indicates that S. aureus itself may not be detrimental. However, a study has shown that patients treated for three years continuously with prophylactic antistaphylococcal antibiotics contracted P. aeruginosa infections sooner than those who were not treated or were treated intermittently, especially between ages 0-6

(Lyczak et al. 2002). Furthermore, CF patients over the age of 18 whose sputum contained S. aureus but not P. aeruginosa had increased survival. These studies indicate that S. aureus may actually retard or prevent colonization by P. aeruginosa, perhaps due to competition.

By about 18 years of age, P. aeruginosa is usually the predominant bacterium within the lung, and 80% of CF patients by that age will have acquired lung infections of P. aeruginosa that they will never get rid of and that will be a significant contributing factor to their morbidity and mortality (Lyczak et al. 2002; Gaspar et al.

2013). Chronic P. aeruginosa infection is marked by decline in lung function due to tissue damage from chronic inflammation (Lyczak et al. 2002). Additionally, the build-up of dead cells and aggregates of cells, both host and bacterial, in the airways 9

and alveoli of the lung decrease respiratory function. The effects of chronic infection contribute to death in the majority of cases.

Fig. 2: Incidence of infection with common pathogens of the lung in patients with CF (Gaspar et al. 2013). The occurrence of P. aeruginosa infection increases with the age of the patient, and P. aeruginosa becomes the predominant pathogen in patients over 20 years old.

The CF lung has been shown to be hypersusceptible to P. aeruginosa colonization and infection. Homozygous CFTR mutant mice were unable to clear P. aeruginosa from the lung after exposure via drinking water, in contrast to wild-type or heterozygous mice (Coleman et al. 2003). Using a variety of strains of P. aeruginosa, an average of 86% of mutant homozygotes, in contrast to only 26% of wild- type, were infected after 29 weeks following exposure. Transgenic mice overexpressing

CFTR had a rate of colonization of only 9%. This indicates that the absence of functional CFTR increases the susceptibility of the lung to P. aeruginosa infection and 10

inhibits clearance of the infection.

There are several theories as to why P. aeruginosa is the predominant bacterial pathogen in CF lung infections, though none have been firmly proven. Research has indicated that CF cells have a higher number of asialo-GM1 gangliosides and fucosylated oligosaccharides on the apical surface when compared to normal cells

(Goldberg and Pier 2000). These molecules may serve as receptors for the type IV pili and flagella of some strains of P. aeruginosa. The adherence of wild-type P. aeruginosa was reduced in vitro when the CF cells were treated with a polyclonal rabbit antibody to asialo-GM1, indicating binding to these gangliosides. Furthermore, the sequence of a P. aeruginosa gene shows homology to bacterial neuraminidases and is believed to encode a sialidase. Bacterial neuraminidases cleave a residue off of ganglioside-GM1, resulting in asialo-GM1. The introduction of neuraminidase inhibitors was shown to reduce P. aeruginosa adherence to CF cells. These data may indicate that P. aeruginosa may bind asialo-GM1 and thus may have increased adherence to CF cells due to increased numbers of asialo-GM1 on the epithelial cell surface. Other studies have indicated that asialo-GM1-mediated binding of P. aeruginosa to CF cells may cause increased cytotoxicity (Comolli and Waite 1999).

Canine kidney cells that incorporated asialo-GM1 gangliosides exogenous into their membrane were exposed to P. aeruginosa in vitro and had an eightfold increase in bacterial adherence and fivefold increase in cytotoxicity. These effects were dependent on the presence of P. aeruginosa type IV pili. Cytotoxicity was shown to be caused by the bacterial exotoxin ExoU. The binding of P. aeruginosa to asialo-GM1 has also 11

been shown to increase recruitment of neutrophils by stimulating the release of interleukin-8 (Goldberg and Pier 1999). While the evidence that asialo-GM1 in CF cells increases P. aeruginosa adherence and cytotoxicity is compelling, it may not fully explain why P. aeruginosa infections predominate in the CF lung. Though the increases in adherence and cytotoxicity are significant, they are still small.

Furthermore, these effects have not been observed in vivo.

Research has also shown that CFTR may serve as a receptor that mediates the internalization of P. aeruginosa into epithelial cells via the bacterial lipopolysaccharide (LPS) outer core (Pier and Grout 1996). Internalization of bacteria into epithelial cells, which then undergo desquamation, is believed to be an important host response for clearing infection. Mutant CFTR inhibits this internalization and therefore bacterial clearing. Histological analysis of CF lung tissue from patients showed decreased internalization of P. aeruginosa. Furthermore, using a mouse model of pneumonia, it was shown that LPS-binding to CFTR significantly increased internalization and that inhibition of this binding increased the extracellular bacterial load (Pier and Grout 1997). CFTR does not modulate internalization of other common pathogens, so this theory is popular because it provides an explanation for the overwhelming predominance of P. aeruginosa over other bacteria in the CF lung.

P. aeruginosa may escape the host immune response by utilizing syndecans

(Goldberg and Pier 2000). Syndecan-1, the most abundant syndecan on epithelial surfaces, is a transmembrane protein and receptor. In response to tissue damage, these molecules shed their extracellular domains, or ectodomains. The ectodomain can also 12

be shed due to the activity of P. aeruginosa elastase in vivo, encoded by the lasA gene.

In a mouse model, shed ectodomains were shown to enhance P. aeruginosa virulence by increasing colonization and tissue damage (Park and Pier 2001). While the mechanism for this is unclear, it has been suggested that the shed ectodomains may inhibit the activity of antimicrobial peptides and surfactant proteins, which are parts of the innate immune system. Additionally, shed ectodomains may inhibit cytokines involved in phagocyte recruitment and inhibit the activity of neutrophil elastase, further impairing the host defense to P. aeruginosa infection. While none of these theories completely explain why the CF lung is hypersusceptible to infection by P. aeruginosa compared to other microbes, it is possible that the reason that P. aeruginosa infection predominates in the CF lung is a combination of several of these factors. The interaction of many factors and characteristics of both the host and pathogen may contribute to the prominence of P. aeruginosa infection.

P. aeruginosa shows significant ability for adaptation to the CF lung environment. The CF lung is dynamic and full of microenvironments including nutrient, oxygen, and ion gradients with a variety of changing selective pressures

(Behrends et al. 2012). With its many regulatory systems and genomic plasticity, P. aeruginosa is a competent colonizer of this environment. Although infection of the lung is often clonal, P. aeruginosa undergoes dramatic adaptive radiation and niche specialization, leading to a wide range of phenotypes (Hogart and Heesemann 2010).

Several of these phenotypes are due to the biofilm matrix exopolysaccharides and will be discussed later. Others involve metabolic adaptations. P. aeruginosa has been 13

shown to adapt metabolite uptake to both metabolic cost and the availability of nutrients, and these adaptations increase P. aeruginosa metabolic efficiency (Behrends et al. 2012). Additionally, the thick, dehydrated secretions coating the epithelial surface in the CF lung results in a largely hypoxic environment that promotes regulatory changes and adaptation in P. aeruginosa. In place of oxygen, P. aeruginosa

- - can use NO3 or NO2 as electron acceptors (Worlitzsch et al. 2002). CF isolates from chronic infection often are nonmotile due to loss of flagella and pili (Hogardt and

Heesemann 2010). It is suggested that the loss of these features decreases the efficiency of phagocytosis by the host immune response. Additionally, virulence factors such as pyoverdin, elastase, the type III secretion system, and exotoxins are downregulated, possibly to reduce energy costs and evade the immune response.

These adaptations provide advantages for P. aeruginosa within the CF lung.

2. Biofilms

P. aeruginosa is also of interest to researchers because it often grows in biofilms.

A biofilm is a structured community of bacteria encased in a thick, protective and structural matrix. The biofilm matrix is composed of extracellular polymeric substance

(EPS), which largely consists of secreted exopolysaccharides but also contains nucleic acids, proteins, and other molecules (Mann and Wozniak 2012). Within a biofilm, only 10-20% of the volume consists of bacteria (Høiby et al. 2001). Evidence of prokaryotic biofilms dates back over three billion years ago in the fossil record, but biofilms were first recognized by microbiologist Antonie Van Leeuwenhoek in the

17th century on the surface of his teeth (Donlan and Costerton 2002; Hall-Stoodley et 14

al. 2004). Biofilms likely developed as a protective environment for prokaryotes, and this facilitated the development of cell-cell interactions and signaling pathways.

Biofilms are very prevalent both in the environment and in human infections. It is estimated that over 99.9% of all bacteria in the environment are sessile and within a biofilm, as opposed to planktonic, or free-floating (Donlan and Costerton 2002).

Additionally, biofilms have been estimated to be involved in more than 60% of recognized types of bacterial infection (Fux et al. 2005). The prevalence of biofilms is likely due to the protective environment that they provide. They protect the cells from environmental stresses such as desiccation, UV light, acidity and temperature fluctuations (Hall-Stoodley et al. 2004). Additionally, because the bacteria are housed in a compact, contained structure, scavenged nutrients such as iron are concentrated within the biofilm and dispersed among the bacteria. This provides these bacterial communities with a more stable environment in which to grow.

Though initially believed to be simple clumps of bacterial cells, biofilm structures are actually quite complex; composed of microcolonies, they contain channels for wastes, water, and nutrient distribution, leading to a variety of microenvironments with different oxygen, nutrient, and waste exposures (Hall-

Stoodley et al. 2004). They contain areas of cells with different metabolic activity and spatial and temporal heterogeneity. Additionally, biofilms are often polymicrobial.

Though the morphology of biofilms varies depending on environmental conditions, the thickness varies typically between 13-60 µm (Høiby et al. 2001). The structure of a biofilm depends upon the genetics of the organism as well as the conditions in which it 15

grows (Stapper et al. 2004). A biofilm typically grows horizontally during the initial colonization of a surface. If the substratum has limited nutrients or carbon sources, horizontal growth may continue. After initial colonization, biofilms grow vertically.

Mature P. aeruginosa biofilms are typically mushroom-shaped (Fig. 3). Vertical growth can also occur if horizontal space is limited. The physical and environmental conditions, as well as variations in gene expression, lead to a great variety in biofilm morphology.

Fig. 3: A confocal micrograph of a wild-type P. aeruginosa biofilm after three days of growth (Mann and Wozniak 2012). This biofilm displays typical characteristics of a mature morphology including a mushroom-shaped three- dimensional structure and a flat monolayer.

Biofilm formation and development is a complex process that is usually categorized into five stages (Fig. 4). Though the general stages are evident in the biofilm development of many species of bacteria, P. aeruginosa is commonly used as a model organism for the study of bacterial biofilms. First, mobile, planktonic cells loosely associate with a surface, either biological (e.g. heart valve, alveolar surface) or nonbiological (e.g. catheter, implant; Stoodley et al. 2002). This contact is initially reversible and the cells are not committed to the phenotypic changes of sessile cells 16

within the biofilm. These cells are still capable of motility via twitching or gliding.

Soon after association of P. aeruginosa to a surface, gene expression changes within cells and results in the increased production of 25% of detectable proteins (Sauer et al.

2002). Within 15 minutes, upregulation of a key protein in exopolysaccharide production, AlgC, occurs (Kuchma and O’Toole, 2000). The second stage is irreversible attachment, which is mediated by type IV pili and secreted exopolysaccharides (Stoodley et al. 2002). The cells become nonmotile (Sauer et al.

2002). Additionally, the P. aeruginiosa Las quorum sensing system is activated.

Quorum sensing is a bacterial method of cell-cell communication that relies on population density to trigger regulatory pathways via chemical signals. The Las system is important for proper formation of the biofilm; Las mutants had only 20% of the thickness and lacked the differentiation of wild-type biofilms (Stoodley et al.

2002). Irreversible attachment results in differential expression, either up- or downregulation, in 39% of the detectable proteome in P. aeruginosa (Sauer et al.

2002). During the third stage of development, the biofilm structure grows and begins to mature, forming water channels and unique niches. Another quorum sensing system, Rhl, is induced (Sauer et al. 2002). The Rhl system induces expression of

RpoS, an important sigma factor for induction of a stress response that also affects exopolysaccharide production (Stoodley et al. 2002). The significance of RpoS to biofilm growth will be further discussed below. During the fourth stage, the biofilm is fully mature. This is sometimes defined as when the biofilm reaches an average thickness of 100 µm. When compared to planktonic cells, there is over six-fold 17

difference in expression in up to 50% of detectable proteins (Stoodley et al. 2002).

During the final stage of biofilm development, a cavity begins to form in the center of the biofilm due to cell death and autolysis, initiating the process of seeding or swarming dispersal and forming a hollow mound (Ma et al. 2009). Autolysis is controlled by cidAB and lrgAB, encoding a holin and antagonistic anti-holin, respectively. The cells within the center of the microcolony die (called liquification), possibly releasing enzymes to facilitate clearance of the area as well as nutrients. This cavity fills with planktonic, motile cells which are eventually released (Hall-Stoodley et al. 2004). The size of the cleared area as well as the initiation of dispersal may be affected by nutrient availability and other environmental conditions (Purevdorj-Gage et al. 2005). Additionally, the quorum sensing Las and Rhl systems are required for seeding dispersal and the formation of the hollow mound. The motile cells that are released may have the persister phenotype (Hall-Stoodley et al. 2004). This phenotype has not been fully characterized, but these cells do not undergo apoptosis and are especially resistant to environmental stresses. Cells in the dispersion stage downregulate 35% of the proteome; because the switch from planktonic to sessile is marked by massive gene upregulation, this downregulation suggests that these cells may be reverting back to the planktonic phenotype (Sauer et al. 2002).

18

Fig. 4: Representation of the five stages of biofilm development (Ma et al. 2009). During stage 1, bacteria adhere reversibly to a surface. During stage two, those attachments become irreversible. During stage three, the bacteria aggregate and continue to secrete exopolysaccharides, and the biofilm grows. At stage four, the biofilm is a mature and complex structure. In the final stage, a cavity in the biofilm forms and fills with planktonic cells, which are then released.

Biofilms are common in a wide range of infections including native valve endocarditis, otitis media, chronic bacterial prostatitis, periodontitis, and, of course,

CF (Donlan and Costerton 2002). They are also commonly found on abiotic surfaces that pose potential health risks, such as prosthetic heart valves, urinary catheters, central venous catheters, contact lenses, and intrauterine devices. Though the presence of biofilms deep in the lung is difficult to detect, evidence suggests that P. aeruginosa does typically grow in biofilms during chronic infection of the CF lung. Transmission electron microscopy identified matrix-contained P. aeruginosa in the sputum of CF patients (Fig. 5; Singh et al. 2000). Additionally, quorum sensing signals similar to those of in vitro-grown biofilms were found in sputum, indicating that biofilms are 19

present in the CF lung.

Fig. 5: Image of P. aeruginosa in biofilm matrix, using transmission electron microscopy, isolated from CF sputum (Singh et al. 2000).

Biofilms have been shown to contribute greatly to the antimicrobial resistance of P. aeruginosa. For example, P. aeruginosa within biofilms have been shown to have greatly increased resistance to the antibiotics tobramycin and ciprofloxacin, and this effect was reversed when cells from the biofilm return to the planktonic lifestyle

(Walters et al. 2003). A main contributor to this resistance is the biofilm matrix, which presents a physical barrier through which antimicrobials must cross (Donlan and

Costerton 2002). The biofilm matrix has been shown to decrease antimicrobial penetrance. Tobramycin has been shown to penetrate 18-times more slowly through a

P. aeruginosa biofilm than through a media control (Walters et al. 2003). In a polymicrobial biofilm of P. aeruginosa and Klebsiella pneumoniae, another common cause of nosocomial infections and pneumonia in CF patients, the common 20

disinfectant chlorine was not able to even reach 20% of the volume of the biofilm

(Mah and O’Toole 2001). In a 2% suspension, a key exopolysaccharide component of the P. aeruginosa matrix called alginate inhibited diffusion of the antibiotics gentamycin and tobramycin; when the suspension was treated with alginate lyase, the inhibition was reversed, indicating the antimicrobial properties of matrix (Donlan and

Costerton 2002). Differences in antibiotic penetrance depend on the composition of the biofilm, i.e. thick or thin, as well as the identity of the antimicrobial. For instance, aminoglycosidic antibiotics such as tobramycin penetrate more poorly than fluoroquinolones such as ciprofloxacin, though ciprofloxacin has been shown to have retarded penetrance through biofilms as well (Suci et al. 1994; Walters et al. 2003).

This is likely due to chemical interactions between the antimicrobial and the matrix.

For example, alginate-predominant matrices are negatively charged and retard the passage of cationic antibiotics (Fux et al. 2005). Furthermore, it is possible that antimicrobials can pass through the fluid channels within the biofilm without penetrating the interiors of most microcolonies (Walters et al. 2003). All of these factors likely contribute to the antibiotic resistance of biofilms.

Additionally, cells within biofilms are also more antimicrobial resistant because they have a decreased rate of growth (Mah and O’Toole 2001). The inhibition of growth is likely due to nutrient limitation resulting from the high population density within biofilms. Studies have shown that biofilms with the slowest growing cells are more resistant to antibiotics (Donlan and Costerton 2002). Furthermore, the heterogeneity of biofilms affects the growth rate of cells and therefore antibiotic 21

resistance. For example, cells within hypoxic areas, which are less metabolically active and replicate more slowly, are more resistant to antibiotics (Walters et al. 2003).

Cells in more oxygen rich areas, such as the border between the biofilm and the external media, have a greater response to antimicrobials, while cells more towards the interior are more resistant (Mah and O’Toole 2001; Walters et al. 2003). Decreased growth rate increases the antibiotic resistance of cells protected by a biofilms.

Furthermore, cell growth may be limited by the general stress response, which induces physiological changes and is simply initiated by involvement in a biofilm

(Mah and O’Toole 2001). RpoS is an alternative sigma factor found in many human pathogens including P. aeruginosa, enterohemorrhagic Escherichia coli, Vibrio cholerae, and Salmonella enterica (Dong and Schellhorn 2010). It is considered to be the “master regulator” of the general stress response and is essential for virulence in many pathogens including those mentioned above. While its effects vary depending on the organism, in P. aeruginosa RpoS modulates the stress response to osmotic and heat shock, carbon starvation, oxidative stress, and low pH (Schuster et al. 2004).

RpoS also upregulates the production of virulence factors such as exotoxin A, alginate, and LasA (Dong and Schellhorn 2010). RpoS responds to environmental stimuli (such as those resulting from involvement in a biofilm) to quickly induce changes that may help the bacterium adapt to a new environment; these changes can include slower growth as well as differential expression of adhesion factors, virulence factors, and metabolic enzymes. High levels of mRNA encoding RpoS in the sputum of CF patients with established, chronic P. aeruginosa infections have been 22

demonstrated using RT-PCR (Foley et al. 1999). The general stress response and slowed growth induced by RpoS protect cells within the biofilm not only from antimicrobials but also oxidative stresses from the immune response.

3. The Matrix Exopolysaccharides

Biofilms can greatly vary in morphology, which affects their resistance to antibiotics, the host immune response, and environmental stresses. One factor of this variety is the composition of the biofilm matrix. While the matrix contains a variety of molecules including proteins and DNA, as much as 90% of the organic carbon in a P. aeruginosa biofilm is found in exopolysaccharides secreted by cells in the biofilm

(Yang et al. 2011). P. aeruginosa produces three matrix exopolysaccharides: alginate,

Psl and Pel. Each has different properties that determine its role in biofilm development and bacterial survival.

Alginate

Alginate is a high molecular weight, branching exopolysaccharide that is composed of β-1, 4-linked L-guluronic and D-mannuronic acids in an irregular, nonrepetitive pattern (Mann and Wozniak 2012). The β-1, 4 bond adds rigidity to the structure of the polysaccharide. D-mannuronic acid is typically O-acetylated at C2 and/or C3. Alginate is synthesized by the gene products of the 12 gene alg operon (Fig.

6; Franklin et al. 2011). Some of these proteins form a complex that extends from the inner membrane, through the periplasmic space, to the outer membrane. The precursor

GDP-mannuronate is synthesized from fructose-6-phosphate, a metabolite from the 23

central carbon metabolism, by AlgA, AlgC, and AlgD. AlgD is a GDP-mannose dehydrogenase; the reaction that this enzyme catalyzes is irreversible and thus its product is committed to the alginate pathway. Additionally, it is the rate-limiting reaction. GDP-mannuronate is incorporated into a homopolymer by Alg8 and Alg44, which are transmembrane proteins in the inner membrane. AlgG, AlgL, AlgK, AlgX, and AlgE complex to form a scaffold in the inner membrane to transport the polymer through the periplasm (Rehman et al. 2013). There, AlgI, AlgJ, and AlgF contribute to

O-acetylation of the polymer. Additionally, AlgG epimerizes D-mannuronic acid residues into L-guluronic acid (Franklin et al. 2011). AlgK and AlgX likely contribute to polymerization, perhaps by stabilizing Alg8 and Alg44 (Rehman et al. 2013).

Alginate is secreted through the outer membrane by the β-barrel pore protein AlgE.

Alginate is not bound to the cell surface.

Fig. 6: Alginate biosynthetic operon and proposed model of the biosynthetic machinery (Franklin et al. 2011). The 12 gene operon encodes all but one protein necessary for alginate synthesis, and many of these proteins form a complex that spans from the inner to outer membrane. 24

Alginate is the most well studied of the exopolysaccharides because of the prevalence of mucoid isolates from chronic infection of the CF lung. Mucoid strains are characterized by the gross overproduction of alginate, giving colonies a “mucus”- like appearance (Fig. 7). The high prevalence of mucoid P. aeruginosa in CF chronic infection is shocking; while only 1% of isolates from other sites of infection are mucoid, 85% of isolates from the CF lung of chronically infected patients are mucoid

(Stapper et al. 2004). However, mucoid P. aeruginosa is not typically responsible for initial colonization of the CF lung; nonmucoid P. aeruginosa is usually acquired at a very young age and then undergoes a phenotypic conversion resulting in mucoid strains that are typically detectable around 11 years of age, though it can occur as soon as a year after infection (Li et al. 2005). Longitudinal studies have shown that the incidence of the mucoid morphotype one to three years after initial infection increases from 10% to 40% (Behrends et al. 2013). By age 16, studies have shown that in 92% of CF patients with P. aeruginosa, mucoid strains can be isolated (Li et al. 2005).

Mucoid P. aeruginosa is associated with chronic infection and significant decline in lung function. Studies have shown that acquisition of mucoid P. aeruginosa corresponds with a sharp decrease in the forced expiratory volume and forced vital capacity of the lungs, which are standard measures of lung function, as well as increased coughing and wheezing (Parad et al. 1999; Li et al. 2005). The clinical significance of mucoidy has made alginate the subject of much research. 25

Fig. 7: Wild-type (A) and mucoid (B) P. aeruginosa colony morphology (Ramsey and Wozniak 2005). The distinctive morphology of the mucoid colonies is the result of the overproduction of alginate.

Because mucoidy is so prevalent during the later stages of infection of the CF lung, it likely confers selective advantages within that environment. Several studies have shown that mucoid P. aeruginosa has increased resistance to the host immune response. Alginate itself is resistant to depolymerization by free radicals produced by macrophages and neutrophils, as well as degradation during phagocytosis (Simpson et al. 1993). Additionally, antibodies produced in response to alginate do not promote opsonic killing of P. aeruginosa (Pier et al. 2004). Mucoid P. aeruginosa within a biofilm showed increased resistance to opsonic activity in vitro when exposed to antibodies from the sera of chronically infected patients; this effect was reduced when biofilms were treated with alginate lyase to degrade alginate (Meluleni et al. 1995).

Evidence suggests that this may be because the unstable deposition of C3 complement 26

proteins in mucoid strain biofilms prevents effective binding to complement receptors on phagocytes. Additionally, the O-acetylation of alginate is important to the resistance of mucoid strains to opsonization and phagocytosis. Acetylated mucoid strains have exhibited nearly 100% cell survival in the presence of complement proteins in vitro; in contrast, acetylation-deficient strains had a maximum survival of

37% (Pier et al. 2001). Acetate groups have been shown to reduce activation of the complement system by reducing binding of opsonins to the cell surface. While many of these studies give compelling evidence for the benefits of alginate production, there is conflicting evidence about the in vivo significance of alginate. In a mouse model, alginate was shown to increase resistance of P. aeruginosa to the respiratory defense system (Bragonzi et al. 2005). Other studies have shown that in a mouse model, mucoid strains do not have increased persistence when compared to non-mucoid strains (Bragonzi et al. 2009). However, other researchers have produced conflicting evidence that over time, mucoid strains have a higher bacterial load in the murine lung, suggesting increased persistence (Song et al. 2003). So while the properties of alginate may confer a selective advantage for mucoid strains in the CF lung environment, this has not been clearly demonstrated in vivo.

Alginate may protect bacteria within biofilms from more than just the host immune system. Based on the similarity of its chemical properties to polysaccharides such as hyaluronic acid, it is believed to have a significant capacity for water retention, perhaps up to 1 kg of water per gram (Mann and Wozniak 2012). This may allow alginate to protect biofilms against desiccation, which may be advantageous 27

both in the environment but also the dehydrated mucus found in the CF lung. The induction of alginate production has been demonstrated under desiccation conditions in other Pseudomonas species (Hay et al. 2014). There is also evidence that alginate increases bacterial resistance to antimicrobials in addition to the resistance contributed by the biofilm structure itself. Some studies have indicated that bacteria within mucoid biofilms are significantly more antibiotic resistant than nonmucoid due to the presence of alginate, even up to 1000-fold more resistant (Hentzer et al. 2001). However, the absence of alginate from a nonmucoid biofilm does not seem to affect antibiotic resistance (Wozniak et al. 2003). This suggests that resistance stems from the massive amount of alginate present in mucoid biofilms, not merely the presence of alginate.

Alginate is considered a capsular polysaccharide, which by definition means that it has a largely protective rather than structural function (Mann and Wozniak

2012). Because of the prevalence of both biofilms and the mucoid morphotype in chronic infection of the CF lung, alginate was long considered to be essential to biofilm formation. However, more recent data has suggested that is not the case.

Research has shown that alginate-deficient (Alg-) strains in vitro form biofilms comparable to both wild-type and mucoid strains (Stapper et al. 2004). Alg- biofilms develop at a rate comparable to isogenic wild-type biofilms (Wozniak et al. 2003).

There is conflicting evidence whether or not alginate is even expressed by non-mucoid strains during biofilm development. Studies have indicated that alginate-binding antibodies are active (though weakly) against nonmucoid strains during infection of a mouse model, indicating the presence of alginate (Pier et al. 2004; Bragonzi et al. 28

2005). Other studies have failed to show expression of alginate biosynthetic genes in non-mucoid biofilms in vitro (Wozniak et al. 2003). However, it has also been shown using antibodies that nonmucoid strains isolated from the sputum of CF patients produce alginate under anaerobic conditions (Bragonzi et al. 2005). While the role of alginate in biofilm development is unclear, it does not seem to be essential for formation.

The presence of alginate does seem to affect biofilm structure, however. When compared to isogenic wild-type and Alg- strains, a mucoid strain biofilm was shown to be more compact; the Alg- strain was more filamentous and diffuse than either mucoid or wild-type (Stapper et al. 2004). Furthermore, the mucoid biofilm was more heterogeneous and had more complex architecture. Mucoid strains have been shown to form more projected, three dimensional structures than nonmucoid strains, which tend to form monolayers (Fig. 8; Nivens et al. 2001). When biofilms of isogenic strains were compared, the mucoid biofilm had more distinct microcolonies and water channels (Hentzer et al. 2001). Additional studies have shown that alginate may be involved in retaining live cells in the “cap” of the mature, mushroom-shaped biofilm; alginate-deficient strains showed few live cells in this area while alginate-producing and mucoid strains had high densities of live cells (Ghafoor et al 2011). Furthermore, research has indicated the O-acetylation of alginate affects biofilm structure (Nivens et al. 2001). Acetate-deficient mucoid strains had decreased attachment and inhibited biofilm growth. The clinical significance of these differences is unclear. 29

Fig. 8: Scanning confocal micrograph images of wild-type and mucoid biofilms over several days of growth (Hentzer et al. 2001). Differences in the morphology are clear; the wild-type biofilm is a monolayer, while the mucoid biofilm is more projected and heterogeneous.

The regulation of alginate production is complex. Transcription of the alg biosynthetic operon is predominantly regulated by the alternate sigma factor AlgU

(also known as AlgT and σ22), which binds a promoter upstream of algD (Hay et al.

2014). AlgU is transcribed from a gene within the same operon as MucA, MucB,

MucC, and MucD. While the role of MucC is unclear, MucA, MucB and MucD are all important to the regulation of AlgU. MucA and MucB are key anti-sigma factors of

AlgU that modulate its activity (Fig. 9A). MucA binds to AlgU and sequesters it just inside the cytoplasmic inner membrane. MucB binds MucA and prevents proteolysis.

Mutations inactivating MucA lead to deregulation of AlgU and thus the alginate 30

biosynthetic operon, resulting in the overproduction of alginate and the mucoid morphotype.

AlgU is activated by a cascade of regulated intramembrane proteolysis (RIP;

Fig. 9B; Hay et al. 2014). First, AlgW, a protease, degrades the periplasmic domain of

MucA. Then, MucA is further degraded by MucP and the protease complex clpXP inside the inner membrane, leading to the release of AlgU (Damron and Goldberg

2012). RIP can be initiated by outer membrane proteins including MucE that activate

AlgW in response to envelope stress (Hay et al. 2014). These outer membrane proteins can be degraded by MucD, preventing activation of AlgW and thus the RIP cascade. MucD also interacts with AlgX and inhibits alginate production; deletions of mucD result in mucoidy (Rehman et al. 2013). Additionally, LPS that has been destabilized due to envelope stress causes the release of MucB from MucA, allowing for MucA to be proteolysed (Hay et al. 2014). AlgU also regulates key virulence factors such as the type III secretion system, exotoxin A, and type IV pili, which are downregulated in mucA mutants (Jones et al. 2010). Though much research has been directed towards this system in vitro, in remains unclear in vivo what environmental signals initiate this cascade.

31

) )

-

sigma sigma

-

. A) A) .

)

4

9: Regulation of of Regulation 9:

.

inhibited when AlgU is is AlgU when inhibited when occurs alginate of Fig via production alginate cascade RIP the and AlgU 201 al. et (Hay Expressionis operon the of by anti sequestered MucB. MucA and B factors when the Expressionoccurs AlgU causes cascade to RIP from its released anti be become and factor sigma Overproduction C) active. mutationsprevent in MucA allowing MucB, to binding be to constitutively AlgU active.

promoter (Hay et al. 2014). al. et (Hay promoter

algD

iptionalregulators the in

Fig. 10: DNA binding sites DNA 10: binding transcr of Fig.

32

Though AlgU is the predominant sigma factor regulator of the alginate biosynthetic pathway, it is not the only one. The housekeeping factor RpoD (σ70) competes with AlgU for the algD promoter but does not drive alginate production

(Yin et al. 2013). RpoD is normally sequestered by anti-σ70 factors including AlgQ during alginate production, allowing AlgU to bind to the promoter; induction of anti-

σ70 factors results in mucoidy. Another alternate sigma factor, RpoN (σ54), also regulates alginate production. RpoN has been shown to drive expression of the alginate biosynthetic operon in the presence of the muc23 mutation, a common mutation in the mucABCD operon that confers mucoidy (Boucher et al. 2000). It also drives alginate production in nitrogen-replete conditions, in which AlgU-driven production is usually inhibited by an unknown mechanism. RpoN can also inhibit

AlgU-dependent transcription, possibly by binding and blocking AlgU-polymerase holoenzyme from binding DNA.

A variety of DNA-binding proteins also regulate the expression of the alginate biosynthetic operon and thus alginate production (Fig. 10). AmrZ is a protein that is required for expression (Hay et al. 2014). In addition to promoting alginate production, it inhibits motility and is believed to be a positive regulator of biofilm development and chronic infection. AlgQ is another protein that promotes alginate production, possibly by contributing to the sequestration of RpoD. AlgP is a histone- like protein and positive regulator, as is IHB. CysB is also a positive regulator of transcription. Vfr is a cAMP-binding protein that is believed to promote transcription and may be involved in catabolite repression. 33

Alginate production is also regulated by two two-component regulatory systems (TCS) in response to environmental signals that are not defined. The KinB-

AlgB system is required for P. aeruginosa virulence (Chand et al. 2012). KinB regulates binding of AlgB, a positive regulator to the algD promoter. KinB can act as a phosphatase that dephosphorylates AlgB; phosphorylated AlgB is associated with alginate production, and although it is unclear how it is phosphorylated it has been suggested that it is also phosphorylated by KinB. The KinB-AlgB system is also a negative regulator of algU transcription, and KinB mutants are mucoid. FimS-AlgR is another TCS that regulates alginate production. AlgR is a DNA-binding protein that binds to two far-upstream regions and one proximal site of the algD promoter in order to promote transcription (Mohr et al. 1992). This activity is responsive to nitrogen- limiting conditions. The functional interaction between FimS and AlgR is unclear; while AlgR mutants have decreased production of alginate, FimS mutants produce more alginate.

There is post-transcriptional and post-translational regulation of the alginate biosynthetic operon as well. Three families of sRNAs (RsmZ, RsmY and RsmX) bind and inhibit the activity of RsmA, which binds and prevents translation of mRNAs

(Hay et al. 2014). Target mRNAs may include those of algD and alg8. Additionally, an antisense transcript of mucD, whose product inhibits alginate production, has been identified, though its regulation is unknown. Furthermore, bis-(3’-5’)-cyclic dimeric guanosine monophosphate (c-di-GMP) is an important molecule for post-translational regulation of alginate biosynthetic machinery. c-di-GMP binding to the PilZ domain 34

of Alg44 is necessary for alginate production, and mutations in this domain prevent binding and inhibit alginate production (Franklin et al. 2011). Activation of Alg8 by

Alg 44 may require a conformational change in Alg44 induced by c-di-GMP binding

(Rehman et al. 2013). An inner membrane protein, MucR, has been identified that may modulate Alg44 activity by producing and degrading c-di-GMP near Alg44, possibly in response to O2 or NO (Hay et al. 2014). Finally, AlgC is an important checkpoint in the exopolysaccharide production pathway. AlgC expression has been shown to be induced soon after attachment to a surface, initiating biofilm formation (Kuchma and

O’Toole 2000). The syntheses of alginate, Psl, and Pel draw from biosynthetically related precursors, mannose-1-phosphate and glucose-1-phosphate, which are produced by AlgC from mannose-6-phosphate and glucose-6-phosphate, respectively

(Ma et al. 2012). Therefore, the biosynthetic pathways of each exopolysaccharide compete for these precursors; overproduction in alginate results in a decrease in Psl and vice versa.

The large number of regulatory proteins and systems that affect alginate production allow for modulation in response to a large number of intracellular, intercellular, and environmental cues. Unfortunately, these cues are poorly understood, particularly regarding their clinical significance. This is partially due to the difficulty of studying P. aeruginosa in vivo, especially within a biofilm. Many of the cues are related to virulence and the establishment of chronic infection and thus in vitro models are often unable to adequately mimic pertinent environmental conditions. Disruption of this regulatory network at numerous points results in the mucoid morphotype. The 35

mutations that are commonly found inducing mucoidy in clinical isolates show patterns, however. Mucoid isolates typically contain at least two mutations in the chromosome at varying locations that contribute to mucoidy (Ramsey and Wozniak

2005). Alginate-producing nonmucoid strains inoculated into a murine lung were shown to revert to a non-alginate producing phenotype when cultured seven days later, indicating that alginate production is due to changes in gene regulation rather than new, stable mutations such as those that cause mucoidy (Bragonzi et al. 2005). Stable mutations causing mucoidy become more prevalent in the later stages of chronic infection. In a study of clinical isolates from CF lungs, mucA was one of the most commonly mutated genes, with a mutation in 63% of strains (Feliziani et al. 2010).

Most of these mutations consisted of small deletions. 37.5% of nonmucoid strains still had a mutation in the mucA gene, suggesting reversion from mucoidy (Feliziani et al.

2010). Many clinical isolates have the common mutations muc-2 or muc-22, which inactivate MucA and cause mucoidy (Fig. 9C; Ramsey and Wozniak 2005). Though much is understood about the direct modulation of alginate production, further research is still needed to understand the initiating cues and the significance of the mucoid morphotype in infection.

Psl and Pel

Psl (named for polysaccharide synthesis locus) is an exopolysaccharide composed of a repeating pentamer of D-mannose, L-rhamnose, and D-glucose subunits (Mann and Wozniak 2012). It is synthesized by the products of a 15 gene operon, though only 11 of these genes are required for a Psl-dependent biofilm (Fig. 36

11). PslB is the only gene within the operon that contributes to the production of Psl precursors. AlgC, which also contributes to precursor production in alginate synthesis, is also necessary for Psl precursors. PslF, PslH, PslI, and PslC contribute to the formation of a repeating polymer from the precursors. PslA, PslE, PslJ, PslK, and PslL are transmembrane proteins in the inner membrane that also contribute to polymerization. PslA likely is the site for assembly of the polymer onto an isoprenoid lipid intermediate for transport across the inner membrane, while PslE may affect the length of the polysaccharide chain. The specific functions of the other three proteins

(PslJ, PslK, PslL) are still unknown. PslD and PslG are proteins involved in exporting

Psl through the outer membrane, though this mechanism is still unknown as well.

Once exported, Psl can exist in two forms. In its smaller form, it is soluble and can be isolated from the supernatant of cell culture (Ma et al. 2009). The larger form is anchored to the cell itself. On the cell surface, the anchored Psl is arranged in a helical pattern around the cell. This likely reflects the localization of its biosynthetic machinery. 37

Fig. 11: The Psl biosynthetic operon and proposed model of the biosynthetic machinery (Franklin et al. 2011).

Psl was identified after researchers had realized that alginate did not serve to provide structural support for biofilm development and therefore hypothesized that other exopolysaccharides must be produced for this role (Mann and Wozniak 2012).

Unlike alginate, Psl is very important for biofilm formation. Psl-deficient mutants were shown have a decrease in the ability to attach to a surface and thus initiate biofilm formation (Jackson et al. 2004; Overhage et al. 2005). Psl- mutants were not only unable to bind to abiotic surfaces such as glass, but also mucin-coated surfaces

(mimicking epithelial surface within the lung) and epithelial cells in tissue culture (Ma et al. 2006). They also had a decreased ability for cell-cell interaction, resulting in an inability to form mature biofilms. Overproduction of Psl resulted in cell hyperaggregation and increased adherence to a surface. This ability to stick to surfaces 38

as well as other cells is not just important to the initiation of biofilm formation, but also maintenance of the biofilm structure as the biofilm matures. When production of

Psl under conditional control was stopped midway through biofilm development, the biofilms were less thick due to erosion of the existing biofilm structure (Ma et al.

2006). Overproduction of Psl caused more three dimensional, projected structures than wild-type strains. These studies demonstrate Psl’s role in biofilm development and in adhesion and aggregation.

Psl localization changes in the biofilm as it matures. In a younger biofilm, Psl was shown to be present throughout the mushroom structure (Ma et al. 2009).

However, as a biofilm matures, Psl is localized only on the periphery, not in the center of the “cap”. It is hypothesized that this is likely because of the degradation of existing

Psl as well as the cessation of Psl production; this reduced production is speculated to be caused by a lack of nutrients and perhaps post-transcriptional regulation as well, rather than the inhibition of transcription, though this is still unclear. The peripheral

Psl may allow the biofilm to continue to recruit planktonic cells. Other studies have indicated that Psl is essential for the association of motile and sessile cells (Yang et al.

2011). The Psl cavity in the biofilm appears to be related to seeding dispersal (Ma et al. 2009). Clearance of Psl from the center of the biofilm occurs in conjunction with a rise in the amount of extracellular DNA (eDNA), which indicates an increase in cell death. The localization of eDNA was also shown to be affected by Psl, though the significance of this is unclear (Yang et al. 2011). It is therefore hypothesized that the clearance of Psl within the cavity is part of the process of clearing the biofilm cavity 39

and cell liquification in preparation for seeding dispersal. Psl is important for biofilm formation even in mucoid strains (Yang et al. 2012). Psl deficiency in mucoid strains greatly reduces biofilm formation. Psl- mucoid strains were also shown to be outcompeted when cultured with a Psl+ mucoid strain, which evidences the importance of Psl in the biofilm formation of different morphotypes.

Because of the relatively recent discovery of Psl, little is known about its clinical significance in infection of the CF lung. It has been shown that Psl production is increased over the course of infection (Huse et al. 2013). One study showed that Psl production is elevated in 72% of isolates from chronic infection of the CF lung when compared to earlier isolates (Huse et al. 2013). Recent studies have shown that Psl may play a protective role as well as a structural one. Psl-deficient mutants had decreased antibiotic resistance, likely because Psl facilitates the close interaction of microcolonies, preventing the diffusion of antibiotics (Yang et al. 2011). It has also been shown that Psl-producing strains modify the activity of neutrophils and inhibit their release of reactive oxygen species in a mechanism that is unclear but that is possibly mediated by the small, unattached form of Psl (Mishra et al. 2012). Psl- producing mucoid strains also have shown increased resistance to phagocytosis when compared to mucoid Psl- strains (Mishra et al. 2012; Yang et al. 2012). Only a third as many cells were phagocytized in Psl-producing strains as compared to Psl-deficient strains (Mishra et al. 2012). This is likely because fewer C3 opsonins bound in Psl+ strains, preventing opsonophagocytosis. Disruption of Psl with cellulase reversed this effect. Deposition of complement proteins C5 and C7, which are part of the membrane 40

attack complex that mediates cell killing, was also impaired. Additionally, Psl- producing strains had increased survival to intracellular killing by neutrophils. These advantages may increase the resistance of biofilm-involved cells to clearance by the host immune system.

Like Psl, Pel was identified relatively recently as a component of the P. aeruginosa biofilm matrix. Of the three main exopolysaccharides that contribute to the

P. aeruginosa biofilm matrix, the least is known about Pel (Mann and Wozniak 2012).

Its structure is still unknown, and little is known about its biosynthesis. Most of the functions assigned to proteins encoded in this operon are based on homology and sequence gazing. The pel operon consists of only seven genes (Fig. 12). Pel precursors do not appear to be produced by gene products from the pel operon, suggesting that its precursors are produced by genes products from a different biosynthetic pathway. PelF is believed to involved in polymerization, while PelD, PelE, and PelG, located in the inner membrane, may be involved in polymer transport. PelB may serve as a scaffold for the polymer in the periplasm, and PelA and PelC may be involved in polymer modification. Of the three biosynthetic exopolysaccharide pathways, the pel operon is the least conserved across several Pseudomonas species. 41

Fig. 12: The Pel biosynthetic operon and proposed model of the biosynthetic machinery (Franklin et al. 2011).

The significance of Pel within a biofilm is still not clear because it seems functionally redundant to Psl. Pel mutants were initially identified by their inability to form an air-liquid (A-L) biofilm, also known as a pellicle, in a Psl-deficient strain

(Mann and Wozniak 2012). Pel may contribute to cell adhesion to a surface. While

Pel deficiency has been shown to have no impact on attachment in a Psl-deficient background, overproduction of Pel results in increased adherence both with and without Psl (Colvin et al. 2011). Pel seems to be more important to cell aggregation as evidenced by its necessity for forming pellicles in liquid culture. Additionally, in a

Psl-deficient strain, Pel overexpression was shown to double the number of cells involved in biofilms. However, this increase was not seen in a Psl-producing strain.

Like Psl, Pel contributes to the formation of the microcolonies that eventually develop into the more complex, mushroom-like structure of mature biofilms (Yang et al. 42

2011). In a Psl-deficient biofilm, Pel is essential for the formation of microcolonies and the development of three dimensional structure (Colvin et al. 2011). However, Pel deficiency does not have an effect on these characteristics in Psl-producing strains. Pel does seem to have a unique role in the association of motile cells to the sessile cells involved in biofilms. While Psl is also necessary for this interaction, Pel-deficient mutants were unable to form this association even in the presence of Psl, though the reason for this is unknown (Yang et al. 2011).

These studies demonstrate Pel’s properties as an aggregative exopolysaccharide. However, they also suggest that Pel plays a largely secondary role in Psl-dependent biofilms. It is possible that because cell aggregation and adherence are essential to biofilm formation, Pel serves as a “back-up” for strains unable to produce Psl. Indeed, Pel production is only maximal in strains that do not produce Psl

(Ma et al. 2012). As mentioned before, AlgC produces the precursors necessary for exopolysaccharide production. Because Pel and Psl production require common precursors, increased production of Psl results in a decrease in Pel production (Ma et al. 2012). Additionally, some Psl-producing strains have exhibited decreased expression of PelC in comparison to Psl-deficient strains, showing that the inverse regulation of Psl and Pel is not merely post-translational (Colvin et al. 2012).

Interestingly, PelA levels do not vary based on Psl status, so the significance of this is unclear. In support of the idea that Psl and Pel redundancy serves as an adaptive advantage, it has been shown that Psl mutant strains can acquire mutations that result in the increased production Pel, enabling the strains to resume biofilm formation 43

(Colvin et al. 2012). Additionally, it is possible that these exopolysaccharides may be utilized in different environments, as has been demonstrated in different biofilm- forming organisms producing multiple exopolysaccharides including E. coli and

Pseudomonas putida as well as Vibrio and Salmonella species.

Because of their relatively recent discovery, mechanisms for the regulation of the production of Psl and Pel are not fully elucidated. Psl and Pel production is highly dependent on intracellular c-di-GMP level (Hickman et al. 2005). Increased intracellular c-di-GMP increases exopolysaccharide production. Modulation of intracellular c-di-GMP levels has been shown to be, at least in part, the result of the

Wsp chemosensory system. WspR is a diguanylate cyclase that produces c-di-GMP from GTP (Hickman et al. 2005). It is encoded in the wsp gene cluster, which also contains genes involved in forming the Wsp complex. This complex is anchored in the outer membrane by WspA, a receptor responding to growth conditions (O’Connor et al. 2012). WspF, a methylesterase, and WspC, a methyltransferase, regulate the methylation status of WspA, which affects WspR activity. Loss-of-function mutations in wspF result in the hypermethylation of WspA and the constitutive activation of

WspR. This causes increased c-di-GMP levels and therefore the overproduction of Pel and Psl (Hickman et al. 2005; Starkey et al. 2009). Interestingly, the Wsp chemosensory system is also involved in chemotaxis; increased intracellular c-di-GMP leads to decreased motility, and thus exopolysaccharide production is inversely regulated with motility (O’Connor et al. 2012). Unlike other chemotaxis sensory systems, Wsp complexes are not only located at the polar ends of the cell but rather 44

laterally along the surface. The inverse control of motility and exopolysaccharide synthesis by this system as well as this unique localization and response to growth and surface conditions may indicate that this system is important to biofilm formation.

Recent work suggests that Psl itself may stimulate the production of c-di-GMP

(Irie et al. 2012). Overproduction of Psl using a conditional mutant as well as addition of exogenous, extracellular Psl resulted in increased intracellular c-di-GMP levels.

Psl-dependent production of c-di-GMP was shown to be mediated by the diguanylate cyclases SadC and SiaD, but not WspR; this activity does not seem to be the result of transcriptional upregulation of the diguanylate cyclases but perhaps direct binding of

Psl to the proteins. It was also shown that biofilms containing Psl stimulated planktonic cells to produce elevated levels of c-di-GMP. This may be a mechanism by which biofilms are perpetuated and recruit planktonic cells.

One known mechanism of the exopolysaccharide response to c-di-GMP involves FleQ. FleQ is a transcriptional regulator of both the psl and pel operons

(Hickman and Hardwood 2008). FleQ mutants upregulate the expression of these operons, suggesting that FleQ acts as a repressor. Though the mechanism by which it regulates the psl operon is unclear, it has been shown to bind to the pel promoter at two locations and repress transcription (Baraquet et al. 2012). Repression is increased in the presence of FleN, which is believed to bind FleQ and increase its DNA-binding affinity (Hickman and Hardwood 2008). Additionally, FleN has been to shown to cause DNA bending when bound to FleQ in both sites (Baraquet et al. 2012). FleQ activity is modulated by c-di-GMP levels. Data indicate that c-di-GMP binds FleQ, 45

causing a conformational change and de-repression of exopolysaccharide production.

C-di-GMP binding may even cause FleQ to become an activator of the pel operon, though these data are inconclusive and the mechanism for this action is still unknown.

Nevertheless, FleQ is an important regulator of exopolysaccharide synthesis.

C-di-GMP also plays a role in the control of Pel and possibly Psl production in response to quorum sensing signals. Las quorum sensing system mutants have decreased expression of the pel operon and decreased production of Pel (Sakuragi and

Kolter 2007). A recent study demonstrated that the presence of Las quorum sensing signals resulted in an increase in the expression of TbpA, a tyrosine phosphatase

(Ueda and Wood 2009). Inactivation of TbpA results in an increase in intracellular c- di-GMP, increased expression of the pel operon, and increased exopolysaccharide production through a mechanism that is still unknown.

Psl production is controlled transcriptionally by several mechanisms. Studies suggest that transcription is driven by the alternate sigma factor RpoS, which, as discussed above, is a regulator of the general stress response (Irie et al. 2010). Another transcriptional regulator, AmrZ, also affects production of Psl (Jones et al. 2013). As described above, it upregulates alginate synthesis. It has also been shown to repress expression of the psl operon by binding to the pslA promoter, possibly blocking polymerase binding. Because AmrZ is part of the AlgU regulon, Psl production is indirectly regulated by AlgU. AlgU mutants produce less Psl (Bazire et al. 2010).

Additionally, it has been shown that mucA mutants, unable to encode this anti-sigma factor, have 30% decreased pslA promoter activity (Ghafoor et al. 2011). This is likely 46

because AlgU is not sequestered by nonfunctional MucA and can therefore repress this promoter.

Post-transcriptionally, like alginate, Psl and Pel production is affected by

RsmA. RsmA has been shown to bind to the 5’ untranslated region (UTR) of the pslA mRNA transcript (Irie et al. 2010). It prevents ribosomal binding, possibly by inducing a double-stranded stem loop in the mRNA transcript at the Shine-Delgarno sequence.

RsmA repression of Pel and Psl production has been shown to be modulated by the activity of rsmZ sRNA, which in turn is controlled by the GacA-GacS TCS in response to environmental signals (Ryder et al. 2007). Pel production is also affected post-translationally by c-di-GMP. C-di-GMP binds PelD, and mutations that prevent c-di-GMP binding result in the loss of Pel production (Franklin et al. 2011).

As in the case of alginate, little is known about the environmental cues that signal this complicated system of regulation to produce these exopolysaccharides.

Like alginate, there is also an extreme Pel and Psl overproduction phenotype associated with infection of the CF lung. Overproduction of Pel and Psl result in a colony morphology called rugose small-colony variants (RSCV; Starkey et al. 2009).

RSCV strains are autoaggregative and hyperadherent, resulting in distinctively small, wrinkly colonies when grown on solid media and clumping of the cells in liquid media

(Fig. 13). RSCV strains, though not as prevalent as mucoid strains, are typically isolated from CF lungs during chronic infection with P. aeruginosa. In the lab, RSCV strains can be formed by creating mutations in wspF, resulting in increased levels of c- di-GMP and thus the overproduction of Pel and Psl. RSCV strains with wspF 47

mutations can be isolated from the CF lung as well and are estimated to account for as much of 68% of RSCV isolates (Starkey et al. 2009). Though mutations can occur in other loci, RSCVs nearly always have increased levels of intracellular c-di-GMP.

Metabolically, RSCVs also differ from wild-type strains in that they are unable to efficiently utilize amino acids as a carbon source and are unable to utilize cysteine as a sulfur source. The mechanism for this and the significance of this is unknown. RSCVs were shown to be able to develop into biofilms several hours faster than a Psl- strain producing a wild-type level of Pel, and these biofilms were more projected and three- dimensional (Drenkard and Ausubel 2002).

Fig. 13: Colony morphology of wild-type P. aeruginosa and an RSCV strain (Irie et al. 2010). The RSCV morphology is due to the overproduction of Pel and Psl.

The RSCV morphotype may confer increased antibiotic resistance to P. aeruginosa cells. One study found that 30% of variants that developed antibiotic resistance to kanamycin in vitro had the RSCV morphotype (Drenkard and Ausubel

2002). Additionally, biofilms formed by RSCV strains were more antibiotic resistant than non-RSCV biofilms. In vivo, there is a correlation between the incidence of

RSCV isolates and antibiotic treatment. For example, samples taken from one patient undergoing intravenous antibiotic treatment at the time revealed a 29% increase in 48

RSCV isolates over two days (Drenkard and Ausubel 2002). Increased resistance to antibiotics may be linked to the overproduction of Pel specifically. While studies have shown that the absence of Pel does not increase antibiotic sensitivity in comparison to a wild-type strain, overproduction of Pel is linked with slightly increased resistance to aminoglycoside antibiotics especially during stationary phase (Colvin et al. 2011).

This increased production may not be specific to Pel, but rather may be the result of a more compact, impenetrable biofilm. Increased resistance to antibiotics may explain why RSCV isolates are associated with chronic infection of the CF lung. The RSCV morphotype may also increase cell resistance to the host immune system. RSCVs were shown to induce a reduced chemokine response in vitro (Starkey et al. 2009).

Furthermore, mucoid Psl/Pel double mutants had a reduced bacterial load during infection of the murine lung, suggesting that these polysaccharides contribute to resistance from clearance (Yang et al. 2012). These studies suggest that RSCVs may have selective advantages during infection of the CF lung. It is unclear whether these advantages are specifically due to increased amounts of Pel and Psl or the affects that these exopolysaccharides have on biofilm morphology.

49

Experimental Overview and Research Questions

P. aeruginosa is an organism that is important to study for many reasons. P. aeruginosa has the ability to adapt to an astonishingly wide range of environments, from soil and water to the human lung, skin, ear and eye. It is a noted and significant component of biofilms, which are gaining recognition for their importance in infection and disease. Additionally, the incidence of P. aeruginosa antibiotic resistance is steadily climbing, and it is quickly gaining notoriety as an opportunistic pathogen

(Lyczak et al. 2002). These properties of P. aeruginosa are due to a vast number of complex regulatory systems that respond to a wide range of environmental cues, many of which are poorly understood or likely not even recognized. Unfortunately, this gap in knowledge makes it especially difficult to study biofilms, which are extremely responsive to the environment in which they are grown. This has led to a great deal of inconsistency in the study of biofilms. Biofilms are grown and analyzed at different times and in different media and conditions; conclusions are drawn using non-isogenic lab and clinical strains, many of which are not fully characterized; biofilms are studied at discrete time points without consideration to their stage in biofilm development. As research and understanding of biofilms increases, studies pay more attention to these factors, but there are still basic gaps in the understanding of biofilm growth.

Despite the attention that matrix exopolysaccharides have received in recent years as important components of biofilms, there is still insufficient information about their roles in biofilm development. Little is known about how exopolysaccharide production changes over the course of biofilm development. Similarly, there is little 50

information about spatial and temporal differential expression of the biosynthetic operons of these exopolysaccharides. Because of the differences in the properties of the exopolysaccharides, one can imagine that their production may change over the course of biofilm development. Information about the production of exopolysaccharides during biofilm development would further elucidate their role and significance to the stages of biofilm development and would therefore increase understanding about biofilm initiation, growth, and maturation.

Therefore, the goal of this thesis was to characterize the production of P. aeruginosa biofilm matrix exopolysaccharides over the course of biofilm development. This was intended to be performed under the same conditions in several strains to allow comparison of the temporal regulation of exopolysaccharide production. The strains that were compared include isogenic lab strains of differing morphotypes as well as strains of morphotypes comparable to the lab strains that were either isolated from in vivo or in vitro biofilms. The approach used was to put key promoters in the exopolysaccharide biosynthetic operons in control of a reporter gene.

These promoter-reporter constructs were then integrated into the bacterial chromosome of P. aeruginosa strains. The expression of the reporter gene monitored over the course of biofilm development would provide a measure of promoter activity and thus the temporal regulation of exopolysaccharide biosynthetic machinery. This thesis was designed to answer the following questions:

1. Does the regulation of the production of P. aeruginosa biofilm matrix

exopolysaccharides change as the biofilm develops? 51

2. Does exopolysaccharide regulation differ between strains of different

morphotypes (e.g. mucoid and RSCV), especially with regard to

uncharacterized isolates?

52

Materials and Methods

Fig. 14: Schematic overview of strain construction including promoter isolation, vector construction, and integration into the P. aeruginosa chromosome (Hoang et al. 2000). 53

Overall Construction Strategy

In order to monitor the regulation of exopolysaccharide biosynthetic operons in

P. aeruginosa, promoter-reporter strains were created in several steps (Fig. 14). First,

AlgD, PelA, and PslA promoter sequences were amplified by PCR. Because the PelA and PslA promoters are not well characterized, three variations of sequence length upstream of the transcriptional start site of these operons were amplified. These amplimers were inserted individually upstream of EGFP (Enhanced Green Fluorescent

Protein) in the vector pEGFP. EGFP was chosen as the fluorescent protein because it has a different emission peak than pyoverdin, a fluorescent siderophore that P. aeruginosa naturally produces. Then, the promoter-EGFP constructs were inserted into the vector miniCTX-1, which does not replicate in P. aeruginosa and integrates into the bacterial chromosome via a phage attachment site. The miniCTX-1 vectors with promoter-reporter constructs were transferred into P. aeruginosa via conjugation.

Then, the pFLP2 vector was also conjugated into these strains in order to excise the unnecessary miniCTX-1 backbone.

Strains

All P. aeruginosa strains used in this research were obtained from the lab of

Dr. Dan Wozniak in the Ohio State University Center for Microbial Interface Biology.

Five strains of P. aeruginosa were used in this research (Table 1a). PAO1 is a wild- type strain of P. aeruginosa. It is capable of producing alginate, Psl, and Pel, but its biofilms are typically Psl-dependent (Wozniak et al. 2003). PDO300 is isogenic to

PAO1 but contains a mutation in mucA, rendering it mucoid. PAO1ΔwspF is an 54

RSCV isogenic variant of PAO1. These three strains are all well characterized lab strains. They were chosen because while isogenic except for the noted mutations, they also represent the extreme phenotypes that are associated with chronic P. aeruginosa infection of the CF lung. FRD1 is a mucoid strain that was isolated from the sputum of a CF patient (Ohman and Chakrabarty 1981). MJK8 is an RSCV that was isolated from a biofilm grown in a drip-flow reactor and was one of the original RSCVs to be isolated and recognized (Kirisits et al. 2005). The FRD1 and MJK8 strains are not fully genetically characterized, and thus the regulation of their exopolysaccharide production may differ from the mucoid and RSCV lab strains, PDO300 and

PAO1ΔwspF, respectively. The E. coli strains TB1 and SM10 were used in vector construction and conjugation with P. aeruginosa, respectively (Table 1b).

Table 1a: P. aeruginosa strains.

Strain Morphotype PAO1 wild-type PDO300 mucoid (isogenic to PAO1) FRD1 mucoid (CF clinical isolate) PAO1ΔwspF RSCV (isogenic to PAO1) MJK8 RSCV (biofilm-derived)

Table 1b: E. coli strains

Strain Genotype (New England Biolabs) TB1 F- ara Δ(lac-proAB) [Φ80dlac Δ(lacZ)M15] rpsL(StrR) thi hsdR SM10 F’ thi-1 thr-1 leuB26 tonA21 lacY1 supE44 recA integrated RP4-2 Tcr::Mu aphA+

Promoter Identification and Isolation

The AlgD, PslA, and PelA promoters were chosen for this research because 55

they regulate transcription of their respective operons (Ramsey and Wozniak 2005).

The AlgD promoter has been well characterized (Mohr et al. 1990; Hay et al. 2014).

Mohr et al. (1990) used a reporter gene under the control of sequences spanning varying lengths upstream of the transcriptional start site in order to determine the sequence eliciting maximal expression. Based on this study, a DNA segment representing the AlgD promoter was used in this research (Fig. 15). Additionally, it inspired the use of similar methods for isolation of the PslA and PelA promoters. PslA and PelA promoter regions have been identified previously (Overhage et al. 2005;

Baraquet et al. 2012). However, because the regulation of these operons has not been fully elucidated, the possibility of far upstream sites such as those found in the AlgD promoter still exists. Therefore, differing DNA segments containing the known promoters of PelA and PslA, the transcription start site, and varying lengths of upstream sequence were used (Table 2).

Table 2: Isolated promoter sequence lengths (see also Figs. 15-17).

Promoter Name Promoter length (bp) AlgD 787 PslA1 206 PslA2 646 PslA3 976 PelA1 319 PelA2 588 PelA3 985

The PCR primers used to isolate and amplify the promoter sequences were designed from the sequence of PAO1 (Table 3). Purified PAO1 DNA was used as the template for all promoter sequence amplifications (Sambrook and Russell 2006). 56

BamHI and HindIII restriction sites were designed into the ends of the primers for vector construction. The promoter sequences were amplified using Q5 High-Fidelity

Polymerase (New England Biolabs) and purified using QIAQuick PCR Purification

Kit (Qiagen; Fig 15-17).

Table 3: PCR primer sequences.

Name Sequence AlgD-L ATTAAGCTTTCGACACCGAGTTCAAGGA HindIII AlgD-R ATTGGATCCTCAAGTTGCTCTGCCCATA BamHI PelA1-L ATTAAGCTTACGGCGTCTACCGAACCGA HindIII PelA2-L ATTAAGCTTTCCGTCACTACGCGACAGGT HindIII PelA3-L ATTAAGCTTACGCTACCATCGGTG HindIII PelA-R ATTGGATCCTCCGGTGGCAACGTCGAAT BamHI PslA1-L ATTAAGCTTTGGCGCCAGAAATACGTCA HindIII PslA2-L ATTAAGCTTAACGTCTCGCAGAAGGT HindIII PslA3-L ATTAAGCTTATGATGTACCAGGAATCTT HindIII PslA-R ATTGGATCCTTGTTTGCTCTGCCGATCA BamHI MiniCTX-R TAATACGACTCACTATAGGGCGAA MiniCTX-L ATCCACCGGCGCGCAATTAAC AttB-R AGTTCGCCTGGTGGAACAACTCG AttB-L CGAGTGGTTTAAGGCAACGGTCTTGA Junction1-R ACATCGTAGGTCTGCTGAATCA

Junction2-L ACGCGAGTGAAAATGGACAGTA

57

1 cggcgcggtgaagcgcaaggcgctgcaagcgcgggtcatcgacaccgagttcaaggacct AlgD-L -> 61 gaagaacggccagtacaagatcatcagcttctacgccaagaaggctcgcggcatgatggc 121 tcgctacgtgatccgcgagcgcctgcgcgacccggccgggctgaaggacttcaatgccca 181 cggctattacttcagcgccgagcaatccggcccggatcagctggtattcctcagggacgc 241 cccgcaagactgatctccccggcccgccgtcctggcgggccgctcctctttcggcacgcc 301 gacgcctcctggcgctaccgttcgtccctccgacacccctgctgcgtcgcctttctcccc 361 ggaaaagcccttgtggcgaataggcctactcaaccgttcgtctgcaagtcatgcggaact 421 gcatcacattttttcacgcccagcccacagacttttccccaattcaaggcggaaatgcca 481 tctccggcgtaatccattggccattaccagcctcccgccattacatgcaaattacgattg 541 caaagtgcatgggtcgaagattaaggaatccttaaggtttgcttaaggcggtaaaagcgg 601 cttctgtttcattccggtgacgcggaactttcagccgccatgcattctgcaactagtggc 661 cattggcaggcatttaacggaaaggccatcaagttggtatcaagtgatatcaaacggata 721 tttccaaatatttcgcgagcgggacaaacggccggaacttccctcgcagagaaaacatcc 781 tatcaccgcgatgcctatcgatagttatgggcagagcaacttgaaaccgtctcgaataat +1 <- AlgD-R

Fig. 15: AlgD promoter sequence. AlgD-R and AlgD-L are underlined and the transcriptional start site is in bold.

1 gcgctggggcgcgccggtccaggtgccgatggcgatgatgtaccaggaatcttccttcaa PslA3-L -> 61 gcatgacgccctgccgccgcgctactacttcctcggtttcatcccctggggccgggtcag 121 ctccgcctacggctacgcccaggccaaggacgagacctgggccgactacaagcgcgaggc 181 cggcggctggggcgccagccgcgacgacttcgccgacgccctggacttcatgggctggta 241 catccagaagagccagcgggtcaacggcgtctccaagtgggacgcctacggccagtacct 301 gaactaccacgagggctggggcggctaccgcaaccgcagctacgacgccaagccatggct 361 gaagaacgtctcgcagaaggtccagtcccgcgcctcgctgttcggcgcccagtaccgcag PslA2-L -> 421 ttgccagcaggagctgtcgcgcggcggctggttctggtagagccgtccgcgcacgccgcg 481 ggcccgaacctcttccgccttcgacgagggcaaacgtccgccaagccccccaggtcggac 541 cggcacgctcgaacggccaacctgcgtcatggaaccgcaggcgcatcctgcccagccagc 601 ccgcagcgccagtggacaggcgggtcgtcgaaccggcggctggccacctggccgagcggc 661 ctgccctcacctttcgccccgcttcgcttcccaggccagagcgctcgcggattggcggcg 721 tcagatttcctcgtctactgtttggataaaagtttggcgccagaaatacgtcaataaatt PslA1-L -> 781 gactaaaaaaacttacccagactacggatatttccctgggaatgctaagatagctatcac 841 aaagccactatcgacgaatgaacctattcgacgggaaaatgactaaaccgcgtggcaaat 901 gaaaaatagtcactaaattgacgcttcaccgccttgctcttccctatccactcaatggac +1 961 tgcccgtgatcggcagagcaaacaacatgcattcgaagtcggtagatagc <- PslA-R

Fig. 16: PslA promoter sequence. PslA3-L, PslA2-L, PslA1-L and PslA-R are underlined and the transcriptional start site is in bold.

58

1 gacgctaccatcggtggcgctcgccggtgtcgataccttgcgccacgcagcgctcgcacg PelA3-L -> 61 ccgttacggcaccctgctgcgctggcgtggggcgcacttgacgggccacgcggccgcgca 121 tgacgagctgctccaggctgtgttggcggtcacactcgccgtagcctccagcgcgacctt 181 ggtggcgtggcggcacatcgaatacggccagtgggcgatctggtcgtccgccagcgtggt 241 taccggcaatgccgccagcgcgcacctgaagctgcgcgatcgcgggcttggcgcgctgct 301 cggcgtaccgctcggcctcgcggtaggggccgctctaccgcatgttctattggtccacac 361 cttggcgactcttgcctcggtcctcaccctcgtcggattccgtcactacgcgacaggttt PelA2-L -> 421 tgccgcgcgttgcgcgtgcatcgcctgcgcctcctgggcgattcagcagtcgcccctggt 481 cgcctcgctgcgggtcgccgaagtactggcgggtggggcgataggcgtcgcctcggtgtg 541 gctggtgcggttcctcgcacgcaactgaaggcggtgcgaccgcacgcgcgtgcgccaaat 601 gcccggcgcctttcgttggccggaaaagacaacctgtttcacggatttaataaatattta 661 gttattagcgaacggcgtctaccgaaccgatgggtctagacgaagtgacttccgggggct PelA1-L -> 721 tccggagggccgtcgactttcctgatgaagtgcgtgaaacctgggctttccactttgcca 781 cagcggtgtctcgggaagcgcggagttgacctgcaaagcgtcacgggcaggcaaaaggac 841 accatggcgtcatgtgtgcgcctaaaaattgacagtttccgcttaaaaatttagcaatta 901 gcatatttagtcattagattgacgttaatcgcctgcggctgctttgcccgtgcccgcact +1 961 gccagggattcgacgttgccaccggaaaccgcatccatgcttccgacctgccgcttcggt <- PelA-R 1021 cggtatcgaccctgccgcgtaggctgggcatgcggttcagcaagaaagga

Fig. 17: PelA promoter sequence. PelA3-L, PelA2-L, PelA1-L and PelA-R are underlined and the transcriptional start site is in bold.

Vector Construction

The plasmid pEGFP (BD Biosciences Clontech) was grown in E. coli strain

TB1 and purified using QIAprep Spin Miniprep Kit (Qiagen). This plasmid was the

source of the reporter gene, EGFP (Fig. 18). It was digested with the restriction

enzymes BamHI (New England Biolabs) and HindIII (New England Biolabs) in the

multicloning site (MCS) for insertion of the promoter sequences. The amplified

promoter sequences were also digested with BamHI and HindIII. The digested vector

was ligated with each promoter sequence using DNA ligase (New England Biolabs)

overnight at 14 °C. The ligation mix was then transformed into E. coli strain TB1.

Because the pEGFP vector carries an ampicillin resistance gene, successful 59

transformants were selected using LANS (tryptone 10 g/L, yeast extract 5 g/L, agar 15 g/L) with 100 mg/L ampicillin. Individual transformants were then grown overnight in

LBNS (tryptone 10 g/L, yeast extract 5 g/L) with 100 mg/L ampicillin, and the plasmids were purified using QIAprep Spin Miniprep Kit (Qiagen). Successful insertion of the promoter sequence was confirmed by PCR amplification and restriction enzyme digests with BamHI, HindIII, and PstI (New England Biolabs).

Fig. 18: Map of vector pEGFP (BD Biosciences Clontech). Promoter sequences were inserted into the 5’ MCS in order to regulate the expression of the downstream gene encoding EGFP.

The promoter-reporter gene constructs were inserted into mini-CTX1, which was also obtained from the Wozniak Lab at the Ohio State University (Fig. 19). This vector does not replicate in P. aeruginosa, but integrates directly into the chromosome at a phage attachment site, AttB, via the AttP site on the vector (Hoang et al. 2000).

Additionally, the vector contains FRT sites that flank sequences outside of the MCS.

In the presence of the pFLP2 plasmid, which encodes Flp recombinase, genes that are 60

unnecessary once integrated, such as the integrase and antibiotic resistance, are recombined out of the integration site via Flp and the FRT sites (Hoang et al. 1998).

This approach was chosen for several reasons. The regulation of genes on plasmids is often not indicative of regulation of chromosomal genes due to effects such as supercoiling (Hoang et al. 2000). Thus, integration into the chromosome would allow the reporter gene to be expressed more like a single copy chromosomal gene, giving a more accurate picture of the gene expression from native promoter activity. This vector also contains transcriptional terminators (Ω) that flank the MCS so that the regulation of the reporter gene is isolated from the transcription of surrounding genes and runaway transcriptional complexes. Additionally, the site of integration for this system is precise; therefore, vector integration does not interrupt any gene and cause an unwanted mutation. Finally, it has been shown that integrants are stably maintained. 61

A) B)

C)

Fig. 19: (A) Map of vectors mini-CTX1 and (B) pFLP2; C) diagram of vector integration and excision via Flp recombinase (Hoang et al. 1998, 2000).

The promoter-reporter constructs were cut out of pEGFP using restriction enzymes HindIII, EcoRI, and NotI. HindIII and EcoRI were used for all constructs except for the AlgD construct, for which HindIII and NotI (New England Biolabs) 62

were used. Similarly, the mini-CTX1 vector was cut with the same enzymes and ligated with the promoter-reporter constructs. The ligation mixtures were then transformed into TB1 and selected on LANS containing 15 mg/L tetracycline.

Individual transformants were then picked and grown overnight in LBNS with 15 mg/L tetracycline. Plasmids were purified from transformants and those with insertions were identified using PCR and confirmed by restriction digest with EcoRI,

HindIII, and NotI. The PCR was done using the primers mini-CTX1-R and mini-

CTX1-L, which flank the MCS within mini-CTX1 and thus amplify insertions.

Finally, the promoter-reporter constructs in mini-CTX1 plasmids were sequenced by the Ohio University Genomics Facility to confirm the correct orientation and identity of the inserts.

Integration of mini-CTX1 in P. aeruginosa and excision with Flp recombinase

Because the efficiency of transforming or electroporating the mini-CTX1 plasmid into P. aeruginosa is extremely low, the vectors were introduced into P. aeruginosa from E. coli by conjugation. The protocol for biparental mating was received from the Wozniak lab through personal communication but is based on work by Hoang et al. (2000). The mini-CTX1-construct vectors were transformed into E. coli strain SM10 and transformants were selected using LANS with 15 mg/L tetracycline. Separately, SM10 with the mini-CTX1-construct vectors and P. aeruginosa strains were plated on LANS at about 300 colonies per plate. The plates were flooded with 2 mL LBNS and the colonies were scraped off and collected. Next,

1 μL, 3 μL, and 6 μL spots of the collected SM10/mini-CTX1-construct strains were 63

plated on LANS and allowed to dry. On top of those spots, 6 μL, 3 μL, and 1 μL spots of P. aeruginosa were plated, respectively, and the plates were allowed to dry. After incubation overnight at 37 °C, the SM10/ P. aeruginosa mix from the spots was plated using an inoculating loop on LANS with 100 mg/L tetracycline and 25 mg/L irgasan.

P. aeruginosa is resistant to this concentration of irgasan, while SM10 is not, so the irgasan selects for P. aeruginosa. Tetracycline selects for the presence of the vector.

Because mini-CTX1 cannot replicate in P. aeruginosa, the vector should be integrated into the chromosome. Selected colonies of P. aeruginosa were plated again on LANS with 100 mg/L tetracycline and 25 mg/L irgasan to confirm their antibiotic resistant phenotypes. The presence of the promoter-reporter sequence was confirmed in strains by PCR using primers mini-CTX-R and mini-CTX-L (Fig. 26a). Additionally, integration at the phage site was confirmed by amplifying both chromosome-vector junctions using primer pairs with one primer internal and one external to the vector

(AttB-L and Junction1-L; AttB-R and Junction2-R; Fig. 27a).

The pFLP2 vector was mated into P. aeruginosa strains via biparental mating as described above. Conjugants were selected using LANS with 300 mg/L carbenicillin and 25 mg/L irgasan. Because the vector contains sacB, which confers sensitivity to sucrose, the presence of the pFLP2 vector was then selected against using LANS with 5% sucrose and incubation at 30 °C overnight. Antibiotic sensitivity, indicating the absence of pFLP2 and the vector backbone containing tetracycline resistance, was then tested using LANS with 100 mg/L tetracycline and

300 mg/L carbencillin. 64

Microtiter Plate Biofilm Assay

The microtiter plate biofilm assay, also called the rapid attachment assay, is a commonly used method of growing biofilms (Merrit et al. 2011). As the alternate name suggests, it is typically used for growth over short amounts of time, from 30 minutes to 48 hours. Benefits of this system include that only common laboratory items are needed and many biofilms can be grown and monitored simultaneously using a plate reader.

The basic set-up of the microtiter plate biofilm assay was the same in all experiments (Merrit et al. 2011). P. aeruginosa strains were grown overnight in LBNS to stationary phase and then diluted 1:100 in fresh LBNS. Then, 100 μL of the diluted culture was deposited in wells of a polystyrene 96-well microtiter plate. Cells adhere to the polystyrene well, which serves as the surface for biofilm growth. The plate was covered and incubated, unshaken, at room temperature for a variable amount of time.

In order to continue biofilm growth for a longer period of time, the liquid media was refreshed so that nutrient limitation did not hinder biofilm growth. The biofilm assay was first done using PAO1 to determine if the liquid media could be replaced without disturbing biofilm growth. The plate was set up as described above.

Every 8, 12, or 16 hours, about 85 μL of media was removed from the wells and replaced with an equal amount of fresh LBNS. After 24 hours, all media was removed from all wells. The plate was then washed by submerging it in deionized water and shaking it. Then, it was turned upside down and blotted on paper towel remove excess water. After that, crystal violet staining was used to measured biofilm growth (Merrit 65

et al. 2011). Crystal violet is a dye that stains cells as well as biofilm matrix components. A 125 μL aliquot of 0.1% crystal violet in water was added to each well including wells that contained no cells as controls. This was left to incubate at room temperature for 10 minutes. Then, the plate was washed again, blotted, and left to dry.

Once completely dry, 200 μL of 30% acetic acid in water was added to each well and incubated for 15 minutes at room temperature to solubilize the biofilms and crystal violet. After mixing by pipetting, 125 μL of the solubilized biofilm was transferred to a new plate. Then, the optical density at 600 nm was measured using a plate reader.

This measurement serves as an indicator of the amount of biomass involved in a biofilm.

CellTiter-Glo Luminescent Cell Viability assay (Promega) was used to generate a measure of viable cells in a well over time. This assay uses ATP levels to determine the number of viable cells in culture. This information is necessary as a baseline to determine if differences in the expression of the reporter gene, controlled by the selected promoters, are due to changes in cell number or differences in regulation. A microtiter plate was set up as described above. PAO1, PDO300, FRD1,

PAO1ΔwspF, and MJK8 in LBNS were all used. Every 12 hours, the media was refreshed as described above. The reagent was added to the cells at various time points over 48 hours. After shaking on an orbital shaker for 2 minutes and incubating at room temperature for 10 minutes, the luminescence was measured using the Synergy H1

Hybrid Multi-mode Microplate Reader (Biotek).

PAO1, PDO300, FRD1, PAO1ΔwspF, MJK8, and PAO1-PelA1-EGFP were 66

grown in a microtiter plate biofilms assay set up as described above. Every 12 hours,

10 μL of LBNS was added to each. At 0.5, 3, 4, 8, 12, 24, 36, 48, 72, and 96 hours, the biofilm biomass from each strain was measured using crystal violet staining as described above. These data show the amount of biofilm biomass per well over time.

Microscopy

Fluorescent microscopy was used to directly observe fluorescence resulting from expression of the reporter gene EGFP. Biofilms were grown in a microtiter plate as described above for 24 hours, with the media refreshed at 12 hours. The media was removed, then the biofilm at the bottom of the microtiter wells was scraped with a pipet tip and washed in 10 μL of water on a slide. The slide was then viewed on a fluorescent microscope (Nikon Eclipse E400) and imaged using the Lumenera M14 camera. The fluorescence in the biofilm images was quantified in the program Image J

(NIH).

Statistical Analysis

Statistical analysis of biofilm growth and mean fluorescence was performed in

SPSS Statistic Version 21 (IBM). Wilcoxon rank sum analysis, a non-parametric statistical test, was used because of the small sample size and because the data were not normally distributed (Wilcoxon et al. 1970).

67

Results

Strain Construction

The chosen promoter sequences were amplified from PAO1 DNA using PCR.

The amplimers were observed by gel electrophoresis and were consistent with the

predicted sequence length (Fig. 20).

(A) (B) (C) (D) (E)

Lanes: 1 2 1 2 1 2 3 4 1 2 1 2 Fig. 20: Gel electrophoresis images of promoter amplimers. Lane 1 in each image is 1 kb ladder. A) Lane 2: PelA1; B) lane 2: PelA2; C) lane 2: PelA3, lane 3: PslA2, lane 4: PslA3; D) lane 2: PslA1; E) lane 2: AlgD.

The pEGFP vector (3.4 kb) was extracted from E. coli strain TB1 and digested

with BamHI and HindIII to confirm its length and identity (Fig. 21).

Lanes: 1 2 Fig. 21: Gel electrophoresis image of 1 kb ladder (lane 1) and pEGFP vector cut with BamHI and HindIII (lane 2). 68

The promoter segments were then ligated into pEGFP and the reactions were transformed into E. coli TB1. After selection on ampicillin, plasmids were extracted and the PCR primers used to first amplify the promoters were used to confirm the presence of promoter segments cloned into pEGFP (Fig. 22).

(B)

(A)

Lanes: 1 2 3 4 5 6 7 8 1 2 3 4 5 Fig. 22: Example gel electrophoresis images of pEGFP vectors screened for successful insertion of the promoters by PCR. Lane 1 in each image is 1 kb ladder. A) lane 4: AlgD, lane 6: PelA1; lanes 2-3, 5, and 7-8 are negative samples. B) lanes 2-3: PelA2; lanes 4-5 are negative samples.

The pEGFP vectors that were shown by PCR amplification to contain the appropriate promoter segments were then digested with restriction enzymes to confirm the presence of the promoter sequence (Fig. 23). BamHI and HindIII were used to cut on either side of the insert in the MCS. Additionally, PstI cuts between BamHI and

HindIII in the MCS, so in the pEGFP-promoter vector, this site is gone. This enzyme was used to confirm that the segment between the BamHI and HindIII sites was removed by showing that the vector was not digested by it. No vectors with promoter inserts were cut by PstI. When digested with both BamHI and HindIII, the inserts were cut out of the vector and had the predicted size (Fig. 23). 69

Lanes: 1 2 3 4 5 6 7 8 9 10 11 12 Fig. 23: Example gel electrophoresis image of promoter-pEGFP construct digested to confirm insertion. Lane 1: l kb ladder, lane 2: undigested pEGFP with no insert; 3: undigested pEGFP-AlgD; 4: pEGFP-AlgD cut with BamHI; 5: pEGFP-AlgD cut with HindIII; 6: pEGFP-AlgD undigested by PstI; 7: pEGFP-Psl1 cut with BamHI and HindIII; 8: undigested pEGFP-Psl1; 9: pEGFP-Psl1 cut with BamHI; 10: pEGFP-Psl1 cut with HindIII; 11: pEGFP-Psl1 undigested by PstI; 12: pEGFP-Psl1 cut with BamHI and HindIII. The white arrows indicate the bands that represent promoter segments cut out of the vector by BamHI and HindIII.

The promoter-EGFP segments of these constructs were excised using restriction enzymes (Fig. 24). The Pel and Psl constructs were excised using EcoRI and HindIII. The AlgD construct was excised by NotI and HindIII.

Lanes: 1 2 Fig. 24: Example electrophoresis image of PelA2-EGFP excised from pEGFP. Lane 1 is 1 kb ladder. Lane 2 is pEGFP-PelA2 digested with EcoRI and HindIII. PelA2 is about 600 bp and EGFP is about 700 bp, so the total construct size is about 1300 bp and is indicated by the arrow. 70

The cut constructs were ligated with cut miniCTX-1 and transformed into TB1.

Successful insertions into miniCTX-1 were confirmed by PCR amplification of the promoter (Fig. 25). Vectors with successful amplification were then sequenced at the

Ohio University Genomics Facility. Each displayed the correct sequence and orientation of the promoter-reporter in the vector.

Lanes: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Fig. 25: Example electrophoresis image of promoters amplified from several extracted miniCTX-1 vectors to check for insertion. Lane 1: 1 kb ladder; lanes 3-4: PelA2; lane 13: PelA1; lane 16: PslA1. Lanes 2, 5-12, and 14-15 contain negative samples.

The promoter-miniCTX-1 constructs were then introduced into P. aeruginosa strains by conjugation as described above. The presence of the plasmid in transformants was confirmed by PCR with a set of primers (miniCTX-L and miniCTX-R) internal to the insert (Fig. 26). Because a common set of primers was used for all seven constructs, the resulting amplimers differ in size due to the different lengths of the promoter sequences but show the predicted size. 71

A)

B)

Lanes: 1 2 3 4 5 6 7 8 9 10 11 12 13 Fig. 26: Example gel electrophoresis image of PAO1 colonies screened for the integration of miniCTX-1 constructs. A) Schematic of where the miniCTX-L and -R primers anneal and the region they amplify. B) Lane 1: 1 kb ladder, lanes 2-3:PelA1, 4-5: PelA2, 7: PelA3, 8: AlgD, 10-11: PslA1, 12-13: PslA3. Lanes 6 and 9 are negative samples.

PCR was also used to confirm integration of the plasmid at the phage attachment site into the P. aeruginosa chromosome (Fig. 27). Primers flanking the attB site were used as the external primers and were designed based on Hoang et al.

(2000). These primers were paired with primers specific to the miniCTX-1 vector in order to amplify the junctions of chromosome and integrated vector. Together, the amplification of the specific promoter sequences and the amplification of the junctions confirm the presence of the integrated vector as well as the identity of the construct. 72

A)

B)

Lanes: 1 2 3 4 5 6 7 8 9 10 11 12 13 Fig. 27: Example electrophoresis gel of PAO1 colonies screened for integration of the vector. A) Schematic of primer annealing sites and amplification of the vector-chromosome junctions. B) PCR amplification of junction 1 (288 bp) and junction 2 (339 bp), alternating: lane 1: 1 kb ladder; 2- 3: PelA2; 4-5: PelA3; 6-7: PslA2; 8-9: PslA3; 10-11: AlgD; 11-12: PelA1. Lanes 2-11 are amplified from PAO1 and lanes 12-13 are amplified from PAO1ΔwspF.

Excision of the miniCTX-1 backbone using Flp recombinase was attempted three times as described above. The first time, colonies sensitive to carbenecillin and tetracycline were obtained. This indicates that the neither the miniCTX-1 vector nor pFLP2 vector were not present. However, excision could not be confirmed by PCR amplification of the junction (data not shown). The last two attempts did not result in colonies with sensitivity to tetracycline, indicating that the gene responsible for tetracycline resistance was not excised and thus that recombination with Flp was not successful (data not shown). A list of all constructed strains is shown in Table 4.

73

Table 4: Constructed Strains

Strains PDO300-PslA1-EGFP FRD1-AlgD-EGFP PAO1-PslA1-EGFP PDO300-PslA3-EGFP PAO1ΔwspF-PslA1-EGFP PAO1-PslA2-EGFP PDO300-PelA1-EGFP PAO1ΔwspF-PelA1-EGFP PAO1-PslA3-EGFP PDO300-PelA2-EGFP PAO1ΔwspF-PelA3-EGFP PAO1-PelA1-EGFP PDO300-PelA3-EGFP PAO1ΔwspF-AlgD-EGFP PAO1-PelA2-EGFP PDO300-AlgD-EGFP MJK8-PelA1-EGFP PAO1-PelA3-EGFP FRD1-PelA1-EGFP MJK8-PelA3-EGFP PAO1-AlgD-EGFP FRD1-PelA3-EGFP MJK8-AlgD-EGFP

Developing a Method for Biofilm Growth

In order to determine if refreshing the media of static biofilms could decrease the effects of nutrient limitation, biofilms were grown for 24 hours. 85 μL of LBNS was removed and replaced at different time points in order to determine the effect on biofilm growth. Media was replaced every eight hours (twice), every 12 hours (once), and every 16 hours (once) with the experiment ending at 24 hours. Biofilm growth was measured using crystal violet staining (Fig. 28). Biofilm growth was greatest when the media was refreshed after 12 hours.

0.4

0.3

0.2

0.1 Absorbance nm) (600

0 Every 8 hrs 12 hrs 16 hrs None Time Point of Media Refreshing

Fig. 28: The effect of media refreshing (85 µL removed and replaced) on biofilm growth.

74

Following this pilot experiment, biofilms were grown for 48 hours with 85 µL of media removed and replaced every 12 hours. The goal of this was to determine the effect of replacing the media on biofilm growth over a longer period of time. The amount of total viable cells was measured using CellTiter-Glo Luminescent Cell

Viability assay (Fig. 29). This method led to visible disturbances (e.g. holes) in the biofilm, probably due to the removal of parts of the biofilm along with the media.

Additionally, the resulting growth curves showed alternating increases and decreases that also suggest that biofilm was being removed and disturbed. This method also measures cell viability of all cells in the well including planktonic, not only those in biofilms. Therefore, it is not necessarily an accurate measure of biofilm growth.

1.E+05

1.E+04

1.E+03 PAO1 PDO300 1.E+02 FRD1

PAO1ΔwspF Luminescence Luminescence (RLU) 1.E+01 MJK8

1.E+00 0 12 24 36 48 Time (hr)

Fig. 29: Cell viability in microtiter plate assay with the media removed and replaced every 12 hours.

75

Because removing media appeared to disturb the biofilm, 10 μL of media was added to each well every 12 hours instead. The biofilms were monitored over 96 hours and biofilm growth was measured by crystal violet staining. This was done with

PAO1, PDO300, FRD1, PAO1ΔwspF, and MJK8 as well as PAO1-PelA1-EGFP in order to determine if the inserted plasmid affects biofilm growth. The addition of media every 12 hours as compared to no additional media resulted in less disturbance in the growth curve, suggesting more stable biofilm formation (Fig. 30). The alternating increases and decreases in Fig. 30b may be because nutrient limitation and waste accumulation cause biofilm dissociation, resulting in less stable biofilm. There was no significant difference in biofilm growth as measured by crystal violet staining between PAO1 with and without the promoter-reporter construct at any of the measured time points (Table 5). This indicates that the integrated vector does not affect the ability of the strain to form a biofilm. Analysis was done with Wilcoxon rank sum tests. The Z-score is a measure of standard deviations away from the mean, and the p value indicates statistical significance under 0.05.

76

(A) 1

PAO1 PDO300 0.1 FRD1 PAO1ΔwspF

Absorbance nm) (600 MJK8 PAO1-PelA1

0.01 0 24 48 72 96 Time (hr)

1

(B)

PAO1 PDO300 0.1 FRD1 PAO1ΔwspF

Absorbance nm) (600 MJK8 PAO1-PelA1

0.01 0 24 48 72 96 Time (hr)

Fig. 30: Biofilm biomass measured by crystal violet staining. A) 10 μL of media was added every 12 hours. B) No media was added.

77

Table 5: Wilcoxon rank sum analysis of PAO1 and PAO1-PelA1-EGFP

Time (hr) Z-score p value 0.5 -0.272 0.785 3 -1.633 0.102 4 -1.604 0.109 8 -1.826 0.068 12 -1.826 0.068 24 0 1.000 36 -1.461 0.144 48 -1.461 0.144 72 -1.860 0.068 96 -0.535 0.068 .

Microscopy and Fluorescence Analysis

Biofilms were grown for 24 hours as described above with 10 μL LBNS added at 12 hours. They were then viewed and imaged as described above (Fig. 31)

Fig. 31: Images of PAO1 biofilm with visible light (A) and UV light (B) at 1000x magnification.

The mean fluorescence of these images was quantified in ImageJ and compared between strains and constructs. Because not all strains were constructed in time, not all were imaged. Of the PAO1 strains with promoter-reporter constructs, 78

none had significantly different fluorescence than PAO1 with no construct (Fig. 32;

Table 6). This may be due to a small sample size and a large amount of variance.

50.00

40.00

30.00

20.00

10.00

0.00 MeanFluorescence (arbitrary units)

Fig. 32: Mean fluorescence of PAO1 and PAO1 promoter-reporter constructs.

Table 6: Wilcoxon rank sum analysis of PAO1 and PAO1 with promoter-reporter constructs.

In comparison with PAO1 Z-score p value PAO1-PslA1-EGFP -0.524 0.600 PAO1-PslA2-EGFP -1.363 0.173 PAO1-PslA3-EGFP -1.153 0.249 PAO1-PelA1-EGFP -0.943 0.345 PAO1-PelA2-EGFP -0.314 0.753 PAO1-PelA3-EGFP -1.572 0.116 PAO1-AlgD-EGFP -0.734 0.463

The mean fluorescence of PDO300 and PDO300 constructs was also compared

(Fig. 33). Only the mean fluorescence of PDO300-PelA1-EGFP was significantly different from PDO300, but it had decreased fluorescence so this was likely not due to 79

the construct, which should result in increased fluorescence (Table 7).

40

30 *

20

10

MeanFluorescence (Arbitrary Units) 0 No PslA1 PslA3 PelA1 PelA2 PelA3 AlgD Construct

Fig. 33: Mean fluorescence of PDO300 and PDO300 promoter-reporter constructs.

Table 7: Wilcoxon rank sum analysis of PDO300 and PDO300 promoter-reporter constructs.

In comparison with Z-score p value PDO300 PDO300-PslA1-EGFP -0.734 0.463 PDO300-PslA3-EGFP -1.363 0.173 PDO300-PelA1-EGFP -1.992 0.046 PDO300-PelA2-EGFP -0.105 0.917 PDO300-PelA3-EGFP -0.314 0.753 PDO300-AlgD-EGFP -0.524 0.600

The mean fluorescence of PAO1ΔwspF-PelA3-EGFP, FRD1-AlgD-EGFP, and

MJK8-PelA3-EGFP was also compared to their respective wild-type strains (Fig. 34).

None of these constructs resulted in significantly different mean fluorescence (Table

8). 80

80

60

40

20

0 MeanFluorescence (arbitrary units)

Fig. 34: Mean fluorescence of wild-type strains and strains with promoter- reporter constructs.

Table 8: Wilcoxon rank sum analysis of mean fluorescence of wild-type strains and strains with promoter- reporter constructs.

Strain comparison Z-score p value (Mean fluorescence) FRD1 and -0.730 0.465 FRD1-AlgD-EGFP PAO1ΔwspF and -1.095 0.273 PAO1ΔwspF-PelA3-EGFP MJK8 and -0.730 0.465 MJK8-PelA3-EGFP

The corrected mean fluorescence (mean fluorescence of wild-type strain subtracted from the mean fluorescence of the construct strain) of the RSCV strains

(PAO1ΔwspF and MJK8) with the PelA3-EGFP construct was compared to the corrected mean fluorescence of PAO1-PelA3-EGFP. Additionally, the corrected mean fluorescence of the mucoid strains (PDO300 and FRD1) with AlgD-EGFP constructs was compared to that of PAO1-AlgD-EGFP. None of these comparisons resulted in a 81

significant difference (Table 9).

Table 9: Wilcoxon rank sum analysis of corrected mean fluorescence

Strain comparison Z-score p value (corrected mean fluorescence) PAO1-AlgD-EGFP and -1.363 0.173 PDO300-AlgD-EGFP

PAO1-AlgD-EGFP and -0.730 0.465 FRD1-AlgD-EGFP PDO300-AlgD-EGFP -1.095 0.273 and FRD1-AlgD-EGFP PAO1-PelA3-EGFP and 0 1.0 PAO1ΔwspF-PelA3-EGFP PAO1-PelA3-EGFP and -1.461 0.144 MJK8-PelA3-EGFP PAO1ΔwspF-PelA3-EGFP -1.095 0.273 and MJK8-PelA3-EGFP

82

Discussion

Strain Construction

All promoter-reporter constructs (AlgD-/ PslA1-/ PslA2-/ PslA3-/ PelA1-/

PelA2-/ PelA3-EGFP) were successfully inserted into the miniCTX-1 vector.

However, not all of the vectors with the constructs were successfully integrated into all P. aeruginosa strains. After the biparental mating, there were sometimes no antibiotic resistant colonies, indicating that the vector was not integrated into the P. aeruginosa chromosome. This could be for two reasons: either conjugation did not occur and the vector was not transferred or the vector was transferred but did not integrate into the chromosome. Hoang et al. (2000) only tested this method of miniCTX-1 conjugation and integration in PAO1. It is possible that other strains do not undergo conjugation as readily as PAO1. For example, positive colonies were especially rare in the RSCV strains; this may be because these cells autoaggregate, perhaps preventing conjugation. The idea that other strains undergo conjugation less readily is supported by the fact that all seven constructs were successfully integrated into PAO1 after only a couple of attempts, but several attempts with the same protocol resulted in the integration of only a couple of constructs in the other strains. This indicates that conjugation may be less efficient with strains other than PAO1.

Additionally, Hoang et al. (2000) tested the miniCTX-1 integration efficiency and stability with inserts up to 600 bp of sequence homologous to the chromosome in the vector. The largest constructs in this research (PelA3 and PslA3) had nearly 1000 bp of homologous sequence. Because this is a longer homologous sequence, there is a 83

greater chance of unwanted homologous recombination. The vector with the homologous promoter segment may recombine with the endogenous promoter, preventing integration at the phage attachment site. This may explain colonies with tetracycline resistance but from which the chromosome-vector junction could not be amplified.

Sometimes following biparental mating, antibiotic resistant colonies grew on the selective media but the colony morphology indicated that the bacteria had reverted from the desired morphotype (e.g. mucoid or RSCV) back to wild-type. Reversion is well documented in strains grown in vitro, especially in mucoid strains (DeVries and

Ohman 1994; personal communication, Wozniak Lab). FRD1 was especially prone to reversion. It is likely that successful integrants of the correct morphoype would have been isolated with further attempts.

Excision of the miniCTX-1 backbone by Flp recombination was not achieved after three attempts. In the first attempt, colonies sensitive to carbenicillin and tetracycline were obtained, indicating that it is likely that the recombination event was successful. However, PCR amplification of the expected junctions of the promoter- reporter construct and the chromosome could not confirm this. At about the same time, other PCR reactions were failing using primers obtained at the same time, indicating that this may have been due to faulty primers. Unfortunately, these putative recombinants were lost. After the next two attempts, no colonies sensitive to both antibiotics were found. In the literature, Hoang et al. (1998) gave little indication of possible complications in this procedure. It is possible that contamination interfered 84

with the biparental mating and conjugation of pFLP2 into P. aeruginosa. Additionally, this system was only tested by Hoang et al. (1998) in PAO1. Different strains may behave differently. Due to lack of time, Flp recombination was not pursued further.

However, comparison of the biofilm growth curves of PAO1 and PAO1-PelA1-EGFP revealed that there was no significant difference in the growth of these strains, and therefore it is unlikely that the miniCTX-1 vector (including the tetracycline resistance gene and integrase gene that would have been excised by Flp recombinase) has a significant effect on biofilm growth.

Developing a Method for Biofilm Growth

One of the major difficulties of this research was that growing biofilms under static conditions is not as ideal as growing biofilms in continuous-flow systems. These systems give biofilms more exposure to nutrients and oxygen as well as clear away metabolic byproducts. The accumulation of waste products and limitation of nutrients has been shown to affect biofilm development (Dunne 2002). These conditions cause biofilms to dissociate, stunting biofilm growth and shifting cells towards the planktonic lifestyle instead. Therefore, growing biofilms under static conditions often limits the time period in which they are viable and useful for study (Merritt et al.

2011). However, static biofilms are easy to grow in that many can be grown at once, as in a 96-well microtiter plate, and they require less specialized equipment.

From this research, it was found that adding small amounts of media (10 µL every 12 hours) had the best effect on biofilm growth. However, there was still a decreased in biomass after only 24 hours for most strains. This is likely because 85

adding media without any kind of mixing can only have a limited affect; metabolic wastes still accumulate. However, removing media and replacing it with fresh media visually disturbed the biofilm, causing gaps and holes. Further work is required to determine if perhaps adding more media or removing and replacing less would have a more positive affect on biofilm growth.

Microscopy and Fluorescence Analysis

Originally, it was intended that as biofilms in a microtiter were developing, the fluorescence from these biofilm would be measured in a fluorescent plate reader. This would allow many biofilms to be monitored at once, and the same biofilms could be monitored over several days without disturbance. Additionally, filters in the plate reader would have allowed the detection of only the emission wavelength of the reporter, EGFP. This would have been beneficial because P. aeruginosa naturally produces pyoverdin, a fluorescent pigment (Iglewski 1996). The fluorescence of pyoverdin causes the background fluorescence measured in the wild-type strains as shown in figures 32-34.

The plate reader must be more sensitive in order to measure biofilm fluorescence in this way. The fluorescent signal from an IPTG-induced E. coli culture with pEGFP, in which EGFP was under the control of the lacZ promoter, was not detected using a plate reader (data not shown). In E. coli, pEGFP has a high copy number and when these cells were visualized microscopically, they fluoresced brightly

(BD Biosciences Clontech). This indicates that the plate reader is not very sensitive.

The promoter-reporter construct used in this research was integrated into the P. 86

aeruginosa chromosome, so there is a single copy per cell. Therefore, even peak expression would likely be too weak to detect using this method.

Instead, microscopic analysis of the biofilms was used in an effort to detect any differential fluorescence and determine if these constructs successfully expressed

EGFP in P. aeruginosa. Biofilms from the microtiter plate well were picked and placed in water on a slide. The problem with this method is a lack of control. It does not allow the thickness of the biofilm to be measured and this cannot be controlled.

Therefore, any differences in fluorescence could be due to differences in cell number in the observed area. There was a lot of variance between the mean fluorescence of samples of the same strains, which is probably due to differences in biofilm thickness.

Additionally, the media used in this study, LBNS, has some autofluorescence. While the biofilms were suspended in water on the slide, some of this media could have been trapped within the biofilm, increasing variability between samples. This variability and the low sample number contributed to the fact that none of the constructs resulted in a significant difference in fluorescence between construct-containing strains and wild-type strains, as well as between different strains containing the same construct. It is therefore inconclusive whether the constructs result in the expression of EGFP.

The next step in this research would be to confirm that the constructed strains successfully express the reporter, EGFP. One possible way of doing this would be to grow biofilms as described before for 24 hours, collect the biofilms, and use Western-

Blot analysis with antibodies to EGFP to detect the presence of this protein. This may be sensitive enough to detect low levels of EGFP. Additionally, a colony count of 87

some of the collected biofilm mixture could be performed in order to account for the number of cells. Furthermore, quantitative PCR could be used to quantify the transcripts from each biosynthetic operon. This method would provide a sensitive measure of promoter activity.

88

Conclusion

In conclusion, P. aeruginosa is a dangerous opportunistic pathogen that can cause a wide range of tissue infections. The antibiotic resistance of these bacteria makes infection very difficult to treat. This is especially true of P. aeruginosa infection in the lungs of CF patients. P. aeruginosa has a very high incidence of infection in those with CF, and most infections are ineradicable. This is largely due to the ability of these bacteria to form biofilms, which serve to protect bacteria from antibiotics and the host immune system. Therefore, in order to identify potential therapeutic targets and strategies, it is important to understand biofilm development and the components of the protective biofilm matrix.

The goal of this research was to better understand biofilm development and the regulation of the production of biofilm matrix exopolysaccharides. One promoter from each of the three exopolysaccharide biosynthetic operons (alg, pel, and psl) was isolated and placed in control of a reported gene, EGFP. This was integrated into the bacterial chromosome of several P. aeruginosa strains of different morphotypes, two of which are associated with chronic infection of the CF lung. After working on a method for static biofilm growth, biofilms of strains with and without a construct were imaged via microscopy and the mean fluorescence was analyzed. There was no significant difference in the fluorescence of strains with and without a construct, so it is inconclusive as to whether the reporter is successfully expressed.

89

Acknowledgements

First of all, I would like to thank Dr. Holzschu, my thesis advisor, for his advice, time, and help on this project. I really appreciate that he was willing to invest so much effort in a project outside of his own area of expertise, and this work would not have been possible without his advice. Thanks also to Dr. Tanda, who provided guidance for this project and encouraged me to continue through all of the difficulties.

Dr. Tanda has always provided great direction and support. This would also not have been possible without his help funding this project as well. I also want to thank Dr.

Colvin for allowing me to use his plate reader and helping me use it as well as Dr.

Duerr for providing advice on measuring fluorescence.

This project was also funded by a grant from the Provost’s Undergraduate

Research Fund. Thank you to Dr. Kittle for helping me to write that grant proposal and for suggesting the idea of studying P. aeruginosa in the first place.

I also owe a huge thanks to my friends and family. I am very thankful that I had Dan Garrett to help me through this project. He encouraged me to persevere and was there to support me whenever I needed it, which was often. My family was also there to support me and provide much needed encouragement. I think that I only got through this challenging year because of the personal support I received from these people. 90

References

Aloush V, Navon-Venezia S, Seigman-Igra Y, Cabili S, Carmeli Y. 2006. Multidrug- resistant Pseudomonas aeruginosa: Risk factors and clinical impact. Antimicrob Agents Chemother 50(1):43-8.

Baraquet C, Murakami K, Parsek MR, Harwood CS. 2012. The FleQ protein from Pseudomonas aeruginosa functions as both a repressor and an activator to control gene expression from the pel operon promoter in response to c-di-GMP. Nucleic Acids Res 40(15):7207-18.

Bazire A, Shioya K, Soum-Soutera E, Bouffartigues E, Ryder C, Guentas- Dombrowsky L, Hemery G, Linossier I, Chevalier S, Wozniak DJ, et al. 2010. The sigma factor AlgU plays a key role in formation of robust biofilms by nonmucoid Pseudomonas aeruginosa. J Bacteriol 192(12):3001-10.

Behrends V, Ryall B, Zlosnik JE, Speert DP, Bundy JG, Williams HD. 2013. Metabolic adaptations of Pseudomonas aeruginosa during cystic fibrosis chronic lung infections. Environ Microbiol 15(2):398-408.

Boucher J, Schurr M, Deretic V. 2000. Dual regulation of mucoidy in Pseudomonas aeruginosa and sigma factor antagonism. Mol Microbiol 36(2):341-51.

Bragonzi A, Paroni M, Nonis A, Cramer N, Montanari S, Rejman , Di Serio C, D ring , T mmler B. 2009. Pseudomonas aeruginosa microevolution during cystic fibrosis lung infection establishes clones with adapted virulence. American Journal of Respiratory and Critical Care Medicine 180(2):138-45.

Bragonzi A, Worlitzsch D, Pier GB, Timpert P, Ulrich M, Hentzer M, Andersen JB, Givskov M, Conese M, Doring G. 2005. Nonmucoid Pseudomonas aeruginosa expresses alginate in the lungs of patients with cystic fibrosis and in a mouse model. J Infect Dis 192(3):410-9. 91

Chand NS, Clatworthy AE, Hung DT. 2012. The two-component sensor KinB acts as a phosphatase to regulate Pseudomonas aeruginosa virulence. J Bacteriol 194(23):6537-47.

Coleman FT, Mueschenborn S, Meluleni G, Ray C, Carey VJ, Vargas SO, Cannon CL, Ausubel FM, Pier GB. 2003. Hypersusceptibility of cystic fibrosis mice to chronic Pseudomonas aeruginosa oropharyngeal colonization and lung infection. Proceedings of the National Academy of Sciences 100(4):1949-54.

Colvin KM, Gordon VD, Murakami K, Borlee BR, Wozniak DJ, Wong GC, Parsek MR. 2011. The Pel polysaccharide can serve a structural and protective role in the biofilm matrix of Pseudomonas aeruginosa. PLoS Pathogens 7(1):e1001264.

Colvin KM, Irie Y, Tart CS, Urbano R, Whitney JC, Ryder C, Howell PL, Wozniak DJ, Parsek MR. 2012. The Pel and Psl polysaccharides provide Pseudomonas aeruginosa structural redundancy within the biofilm matrix. Environ Microbiol 14(8):1913-28.

Comolli JC, Waite LL, Mostov KE, Engel JN. 1999. Pili binding to asialo-GM1 on epithelial cells can mediate cytotoxicity or bacterial internalization by Pseudomonas aeruginosa. Infect Immun 67(7):3207-14.

Damron FH, Goldberg JB. 2012. Proteolytic regulation of alginate overproduction in Pseudomonas aeruginosa. Mol Microbiol 84(4):595-607.

DeVries CA, Ohman DE. 1994. Mucoid-to-nonmucoid conversion in alginate- producing Pseudomonas aeruginosa often results from spontaneous mutations in algT, encoding a putative alternate sigma factor, and shows evidence for autoregulation. J Bacteriol 176(21):6677-87.

Dong T, Schellhorn HE. 2010. Role of RpoS in virulence of pathogens. Infect Immun 78(3):887-97. 92

Donlan RM, Costerton JW. 2002. Biofilms: Survival mechanisms of clinically relevant microorganisms. Clin Microbiol Rev 15(2):167-93.

Drenkard E, Ausubel FM. 2002. Pseudomonas biofilm formation and antibiotic resistance are linked to phenotypic variation. Nature 416(6882):740-3.

Dunne WM. 2002. Bacterial adhesion: Seen any good biofilms lately? Clin Microbiol Rev 15(2):155-66.

Feliziani S, Luján AM, Moyano AJ, Sola C, Bocco JL, Montanaro P, Canigia LF, Argaraña CE, Smania AM. 2010. Mucoidy, quorum sensing, mismatch repair and antibiotic resistance in Pseudomonas aeruginosa from cystic fibrosis chronic airways infections. PloS One 5(9):e12669.

Franklin MJ, Nivens DE, Weadge JT, Howell PL. 2011. Biosynthesis of the Pseudomonas aeruginosa extracellular polysaccharides, alginate, Pel, and Psl. Front Microbiol 2:167.

Fux CA, Costerton JW, Stewart PS, Stoodley P. 2005. Survival strategies of infectious biofilms. Trends Microbiol 13(1):34-40.

Gaspar MC, Couet W, Olivier J-, Pais AACC, Sousa JJS. 2013. Pseudomonas aeruginosa infection in cystic fibrosis lung disease and new perspectives of treatment: A review. European Journal of Clinical Microbiology & Infectious Diseases 32(10):1231-52.

Gellatly SL, Hancock RE. 2013. Pseudomonas aeruginosa: New insights into pathogenesis and host defenses. Pathog Dis 67(3):159-73.

Ghafoor A, Hay ID, Rehm BHA. 2011. Role of exopolysaccharides in Pseudomonas aeruginosa biofilm formation and architecture. Applied and Environmental Microbiology 77(15):5238-46. 93

Goldberg J, Pier G. 2000. The role of the CFTR in susceptibility to Pseudomonas aeruginosa infections in cystic fibrosis. Trends Microbiol 8(11):514-20.

Hall-Stoodley L, Costerton JW, Stoodley P. 2004. Bacterial biofilms: From the natural environment to infectious diseases. Nat Rev Microbiol 2(2):95-108.

Hay ID, Wang Y, Moradali M, Rehman ZU, Rehm BHA. 2014. Genetics and regulation of bacterial alginate production. Environ Microbiol 16:1-15

Hentzer M, Teitzel GM, Balzer GJ, Heydorn A, Molin S, Givskov M, Parsek MR. 2001. Alginate overproduction affects Pseudomonas aeruginosa biofilm structure and function. J Bacteriol 183(18):5395-401.

Hickman JW, Harwood CS. 2008. Identification of FleQ from Pseudomonas aeruginosa as ac‐ di‐ GMP‐ responsive transcription factor. Mol Microbiol 69(2):376-89.

Hickman JW, Tifrea DF, Harwood CS. 2005. A chemosensory system that regulates biofilm formation through modulation of cyclic diguanylate levels. Proc Natl Acad Sci U S A 102(40):14422-7.

Hoang TT, Karkhoff-Schweizer RR, Kutchma AJ, Schweizer HP. 1998. A broad-host- range flp-FRT recombination system for site-specific excision of chromosomally- located DNA sequences: Application for isolation of unmarked Pseudomonas aeruginosa mutants. Gene 212(1):77-86.

Hoang TT, Kutchma AJ, Becher A, Schweizer HP. 2000. Integration-proficient plasmids for Pseudomonas aeruginosa: site-specific integration and use for engineering of reporter and expression strains. Plasmid 43(1):59-72.

Hogardt M, Heesemann J. 2010. Adaptation of Pseudomonas aeruginosa during persistence in the cystic fibrosis lung. Int J Med Microbiol 300(8):557-62. 94

Hoiby N, Krogh Johansen H, Moser C, Song Z, Ciofu O, Kharazmi A. 2001. Pseudomonas aeruginosa and the in vitro and in vivo biofilm mode of growth. Microbes Infect 3(1):23-35.

Huse HK, Kwon T, Zlosnik JE, Speert DP, Marcotte EM, Whiteley M. 2013. Pseudomonas aeruginosa enhances production of a non-alginate exopolysaccharide during long-term colonization of the cystic fibrosis lung. PloS One 8(12):e82621.

Iglewski BH. 1996. Chapter 27: Pseudomonas. In: Medical microbiology. Baron S, editor. 4th ed. Galveston, TX: University of Texas Medical Branch at Galveston.

Irie Y, Borlee BR, O'Connor JR, Hill PJ, Harwood CS, Wozniak DJ, Parsek MR. 2012. Self-produced exopolysaccharide is a signal that stimulates biofilm formation in Pseudomonas aeruginosa. Proc Natl Acad Sci U S A 109(50):20632-6.

Irie Y, Starkey M, Edwards AN, Wozniak DJ, Romeo T, Parsek MR. 2010. Pseudomonas aeruginosa biofilm matrix polysaccharide Psl is regulated transcriptionally by RpoS and post‐ transcriptionally by RsmA. Mol Microbiol 78(1):158-72.

Jackson KD, Starkey M, Kremer S, Parsek MR, Wozniak DJ. 2004. Identification of Psl, a locus encoding a potential exopolysaccharide that is essential for Pseudomonas aeruginosa PAO1 biofilm formation. J Bacteriol 186(14):4466-75.

Jones AK, Fulcher NB, Balzer GJ, Urbanowski ML, Pritchett CL, Schurr MJ, Yahr TL, Wolfgang MC. 2010. Activation of the Pseudomonas aeruginosa AlgU regulon through mucA mutation inhibits cyclic AMP/Vfr signaling. J Bacteriol 192(21):5709-17. 95

Jones CJ, Ryder CR, Mann EE, Wozniak DJ. 2013. AmrZ modulates Pseudomonas aeruginosa biofilm architecture by directly repressing transcription of the Psl operon. J Bacteriol 195(8):1637-44.

Kirisits MJ, Prost L, Starkey M, Parsek MR. 2005. Characterization of colony morphology variants isolated from Pseudomonas aeruginosa biofilms. Appl Environ Microbiol 71(8):4809-21.

Kuchma SL, O'Toole GA. 2000. Surface-induced and biofilm-induced changes in gene expression. Curr Opin Biotechnol 11(5):429-33.

Li Z, Kosorok MR, Farrell PM, Laxova A, West SE, Green CG, Collins J, Rock MJ, Splaingard ML. 2005. Longitudinal development of mucoid Pseudomonas aeruginosa infection and lung disease progression in children with cystic fibrosis. Jama 293(5):581-8.

Lyczak JB, Cannon CL, Pier GB. 2002. Lung infections associated with cystic fibrosis. Clin Microbiol Rev 15(2):194-222.

Ma L, Conover M, Lu H, Parsek MR, Bayles K, Wozniak DJ. 2009. Assembly and development of the Pseudomonas aeruginosa biofilm matrix. PLoS Pathogens 5(3):e1000354.

Ma L, Jackson KD, Landry RM, Parsek MR, Wozniak DJ. 2006. Analysis of Pseudomonas aeruginosa conditional Psl variants reveals roles for the Psl polysaccharide in adhesion and maintaining biofilm structure postattachment. J Bacteriol 188(23):8213-21.

Ma L, Wang J, Wang S, Anderson EM, Lam JS, Parsek MR, Wozniak DJ. 2012. Synthesis of multiple Pseudomonas aeruginosa biofilm matrix exopolysaccharides is post‐ transcriptionally regulated. Environ Microbiol 14(8):1995-2005. 96

Mah TC and O'Toole GA. 2001. Mechanisms of biofilm resistance to antimicrobial agents. Trends Microbiol 9(1):34-9.

Mann EE and Wozniak DJ. 2012. Pseudomonas biofilm matrix composition and niche biology. FEMS Microbiol Rev 36(4):893-916.

Meluleni GJ, Grout M, Evans DJ, Pier GB. 1995. Mucoid Pseudomonas aeruginosa growing in a biofilm in vitro are killed by opsonic antibodies to the mucoid exopolysaccharide capsule but not by antibodies produced during chronic lung infection in cystic fibrosis patients. J Immunol 155(4):2029-38.

Merritt JH, Kadouri DE, O'Toole GA. 2005. Growing and analyzing static biofilms. Curr Protoc Microbiol Chapter 1:Unit 1B.1.

Mishra M, Byrd MS, Sergeant S, Azad AK, Parsek MR, McPhail L, Schlesinger LS, Wozniak DJ. 2012. Pseudomonas aeruginosa Psl polysaccharide reduces neutrophil phagocytosis and the oxidative response by limiting complement‐ mediated opsonization. Cell Microbiol 14(1):95-106.

Mohr C, Martin D, Konyecsni W, Govan J, Lory S, Deretic V. 1990. Role of the far- upstream sites of the algD promoter and the algR and rpoN genes in environmental modulation of mucoidy in Pseudomonas aeruginosa. J Bacteriol 172(11):6576-80.

Mohr CD, Leveau JH, Krieg DP, Hibler NS, Deretic V. 1992. AlgR-binding sites within the algD promoter make up a set of inverted repeats separated by a large intervening segment of DNA. J Bacteriol 174(20):6624-33.

Nivens DE, Ohman DE, Williams J, Franklin MJ. 2001. Role of alginate and its O acetylation in formation of Pseudomonas aeruginosa microcolonies and biofilms. J Bacteriol 183(3):1047-57. 97

O'Connor JR, Kuwada NJ, Huangyutitham V, Wiggins PA, Harwood CS. 2012. Surface sensing and lateral subcellular localization of WspA, the receptor in a chemosensory‐ like system leading to c‐ di‐ GMP production. Mol Microbiol 86(3):720-9.

Overhage J, Schemionek M, Webb JS, Rehm BH. 2005. Expression of the psl operon in Pseudomonas aeruginosa PAO1 biofilms: PslA performs an essential function in biofilm formation. Appl Environ Microbiol 71(8):4407-13.

Parad RB, Gerard CJ, Zurakowski D, Nichols DP, Pier GB. 1999. Pulmonary outcome in cystic fibrosis is influenced primarily by mucoid Pseudomonas aeruginosa infection and immune status and only modestly by genotype. Infect Immun 67(9):4744-50.

Park PW, Pier GB, Hinkes MT, Bernfield M. 2001. Exploitation of syndecan-1 shedding by Pseudomonas aeruginosa enhances virulence. Nature 411(6833):98- 102.

Pier GB, Boyer D, Preston M, Coleman FT, Llosa N, Mueschenborn-Koglin S, Theilacker C, Goldenberg H, Uchin J, Priebe GP, et al. 2004. Human monoclonal antibodies to Pseudomonas aeruginosa alginate that protect against infection by both mucoid and nonmucoid strains. J Immunol 173(9):5671-8.

Pier GB, Coleman F, Grout M, Franklin M, Ohman DE. 2001. Role of alginate O acetylation in resistance of mucoid Pseudomonas aeruginosa to opsonic phagocytosis. Infect Immun 69(3):1895-901.

Pier GB, Grout M, Zaidi TS. 1997. Cystic fibrosis transmembrane conductance regulator is an epithelial cell receptor for clearance of Pseudomonas aeruginosa from the lung. Proc Natl Acad Sci U S A 94(22):12088-93. 98

Pier GB, Grout M, Zaidi TS, Olsen JC, Johnson LG, Yankaskas JR, Goldberg JB. 1996. Role of mutant CFTR in hypersusceptibility of cystic fibrosis patients to lung infections. Science 271(5245):64-7.

Purevdorj-Gage B, Costerton WJ, Stoodley P. 2005. Phenotypic differentiation and seeding dispersal in non-mucoid and mucoid Pseudomonas aeruginosa biofilms. Microbiology (13500872) 151(5):1569-76.

Ramsey DM and Wozniak DJ. 2005. Understanding the control of Pseudomonas aeruginosa alginate synthesis and the prospects for management of chronic infections in cystic fibrosis. Mol Microbiol 56(2):309-22.

Rehman ZU, Wang Y, Moradali MF, Hay ID, Rehm BH. 2013. Insights into the assembly of the alginate biosynthesis machinery in Pseudomonas aeruginosa. Appl Environ Microbiol 79(10):3264-72.

Ryder C, Byrd M, Wozniak DJ. 2007. Role of polysaccharides in Pseudomonas aeruginosa biofilm development. Curr Opin Microbiol 10(6):644-8.

Sakuragi Y, Kolter R. 2007. Quorum-sensing regulation of the biofilm matrix genes (pel) of Pseudomonas aeruginosa. J Bacteriol 189(14):5383-6.

Sambrook J, Russell DW. 2006. Purification of nucleic acids by extraction with phenol:chloroform. CSH Protoc 2006(1):10.1101/pdb.prot4455.

Sauer K, Camper AK, Ehrlich GD, Costerton JW, Davies DG. 2002. Pseudomonas aeruginosa displays multiple phenotypes during development as a biofilm. J Bacteriol 184(4):1140-54.

Schuster M, Hawkins AC, Harwood CS, Greenberg EP. 2004. The Pseudomonas aeruginosa RpoS regulon and its relationship to quorum sensing. Mol Microbiol 51(4):973-85. 99

Simpson JA, Smith SE, Dean RT. 1993. Alginate may accumulate in cystic fibrosis lung because the enzymatic and free radical capacities of phagocytic cells are inadequate for its degradation. Biochem Mol Biol Int 30(6):1021-34.

Singh PK, Schaefer AL, Parsek MR, Moninger TO, Welsh MJ, Greenberg EP. 2000. Quorum-sensing signals indicate that cystic fibrosis lungs are infected with bacterial biofilms. Nature 407(6805):762-4.

Song Z, Wu H, Ciofu O, Kong KF, Hoiby N, Rygaard J, Kharazmi A, Mathee K. 2003. Pseudomonas aeruginosa alginate is refractory to Th1 immune response and impedes host immune clearance in a mouse model of acute lung infection. J Med Microbiol 52(9):731-40.

Stapper AP, Narasimhan G, Ohman DE, Barakat J, Hentzer M, Molin S, Kharazmi A, Hoiby N, Mathee K. 2004. Alginate production affects Pseudomonas aeruginosa biofilm development and architecture, but is not essential for biofilm formation. J Med Microbiol 53(7):679-90.

Starkey M, Hickman JH, Ma L, Zhang N, De Long S, Hinz A, Palacios S, Manoil C, Kirisits MJ, Starner TD. 2009. Pseudomonas aeruginosa rugose small-colony variants have adaptations that likely promote persistence in the cystic fibrosis lung. J Bacteriol 191(11):3492-503.

Stoodley P, Sauer K, Davies DG, Costerton JW. 2002. Biofilms as complex differentiated communities. Annu Rev Microbiol 56:187-209.

Suci PA, Mittelman MW, Yu FP, Geesey GG. 1994. Investigation of ciprofloxacin penetration into Pseudomonas aeruginosa biofilms. Antimicrob Agents Chemother 38(9):2125-33. 100

Ueda A, Wood TK. 2009. Connecting quorum sensing, c-di-GMP, Pel polysaccharide, and biofilm formation in Pseudomonas aeruginosa through tyrosine phosphatase TpbA (PA3885). PLoS Pathogens 5(6):e1000483.

Walters MC, Roe F, Bugnicourt A, Franklin MJ, Stewart PS. 2003. Contributions of antibiotic penetration, oxygen limitation, and low metabolic activity to tolerance of Pseudomonas aeruginosa biofilms to ciprofloxacin and tobramycin. Antimicrob Agents Chemother 47(1):317-23.

Wilcoxon F, Katti S, Wilcox RA. 1970. Critical values and probability levels for the Wilcoxon rank sum test and the Wilcoxon signed rank test. Selected Tables in Mathematical Statistics 1:171-259.

Worlitzsch D, Tarran R, Ulrich M, Schwab U, Cekici A, Meyer KC, Birrer P, Bellon G, Berger J, Weiss T, et al. 2002. Effects of reduced mucus oxygen concentration in airway Pseudomonas infections of cystic fibrosis patients. J Clin Invest 109(3):317-25.

Wozniak DJ, Wyckoff TJ, Starkey M, Keyser R, Azadi P, O'Toole GA, Parsek MR. 2003. Alginate is not a significant component of the extracellular polysaccharide matrix of PA14 and PAO1 Pseudomonas aeruginosa biofilms. Proc Natl Acad Sci U S A 100(13):7907-12.

Yang L, Hu Y, Liu Y, Zhang J, Ulstrup J, Molin S. 2011. Distinct roles of extracellular polymeric substances in Pseudomonas aeruginosa biofilm development. Environ Microbiol 13(7):1705-17.

Yang L, Hengzhuang W, Wu H, Damkiær S, Jochumsen N, Song Z, Givskov M, Høiby N, Molin S. 2012. Polysaccharides serve as scaffold of biofilms formed by mucoid Pseudomonas aeruginosa. FEMS Immunology & Medical Microbiology 65(2):366-76. 101

Yin Y, Withers TR, Wang X, Hongwei DY. 2013. Evidence for sigma factor competition in the regulation of alginate production by Pseudomonas aeruginosa. PloS One 8(8):e72329.