EFFECTS OF NONTHERMAL PLASMA ON PROKARYOTIC AND EUKARYOTIC CELLS

A dissertation submitted to Kent State University in partial fulfillment of the requirements for the degree of Doctor of Philosophy

by James R. Ferrell May, 2013

Dissertation written by James R. Ferrell B.S., North Carolina Central University, May 2008 Ph.D., Kent State University, May 2013

Approved by ___________________________________ Christopher J. Woolverton, Advisor ___________________________________ Judith A. Fulton, Member ___________________________________ Derek S. Damron, Member ___________________________________ Gail C. Fraizer, Member

Accepted by ___________________________________ Laura Leff, Interim Chair, Department of Biological Sciences ___________________________________ Raymond Craig, Dean, College of Arts and Sciences

i

TABLE OF CONTENTS Page LIST OF FIGURES ..............................................................................vi LIST OF TABLES ...............................................................................xii ACKNOWLEDGEMENTS ...................................................................xiii CHAPTER I. General Introduction ............................................................1 Chronic Wounds ..........................................................................1 Wound healing ............................................................................5 Virulence factors ........................................................................10 Bacterial biofilms .......................................................................17 Nonthermal plasmas ....................................................................22 General overview and approach .....................................................27 Dissertation rationale and hypotheses .............................................28 References .................................................................................32

CHAPTER II. Physical Characteristics, Composition, and Germicidal Characteristics of a Novel “Spark-Based” Nonthermal Plasma Generator Discharge ...............................38 Abstract ....................................................................................38 Introduction ...............................................................................39 Methods ....................................................................................43 Physical description of the GPP ...........................................43 Plasma spectra characteristics ..............................................45 Plasma composition measurements .......................................45 ii

pH measurement ...............................................................48 Temperature measurement ..................................................49 Nutrient media variations ...................................................49 Bacterial cultures ..............................................................50 Depth experiments ............................................................50 Diameter experiments ........................................................51 Colony counts and viability..................................................52 Ultraviolet radiation filtration ..............................................52 Reactive oxygen and nitrogen species filtration........................53 Results ......................................................................................53 Discussion .................................................................................69 References .................................................................................78

CHAPTER III. Bacterial Biofilm Physiological and Structural Responses to Nonthermal Plasma ..........................................................................................................81 Abstract ....................................................................................81 Introduction ...............................................................................82 Methods ....................................................................................84 Bacterial biofilm strains and growth conditions........................84 Biofilm reactor construction ................................................85 Biofilm reactor units ..........................................................87 Biofilm staining and confocal imaging ..................................88 Z-Stack analysis ...............................................................88 iii

Statistical analysis .............................................................91 Results ......................................................................................92 Discussion ...............................................................................103 References ...............................................................................114

CHAPTER IV. Physiological Alterations in Virulence Factor Producing Pathogens Following Nonthermal Plasma Exposure ...........................................................117 Abstract ..................................................................................117 Introduction .............................................................................118 Methods ..................................................................................121 Strains and media ............................................................121 Optical density.................................................................121 Logarithmic plate counts....................................................122 Vital staining ..................................................................122 PBP-2a PCR....................................................................123 PBP-2a Purification..........................................................124 PBP-2a Quantification.......................................................125 Cell Membrane Damage Assays .........................................125 MRSA Screen Latex Agglutination .....................................126 Results ....................................................................................127 Discussion ...............................................................................139 References ...............................................................................154

iv

CHAPTER V. Stimulation of Wound Healing in Applied Human Skin In vitro Models ..................................................................................................................156 Abstract ..................................................................................156 Introduction .............................................................................157 Methods ..................................................................................161 Media and gel preparation..................................................161 Human tissue and cell culture ............................................162 Cell proliferation assay .....................................................163 Cell migration assay.........................................................164 Punch biopsy assay ..........................................................164 Histology and image analysis ............................................165 Immunohistochemistry .....................................................165 Results ....................................................................................167 Discussion ...............................................................................175 References ...............................................................................188

CHAPTER VI. General Conclusions .................................................................191 References ...............................................................................207

v

LIST OF FIGURES Page CHAPTER I. General Introduction Figure 1. Comparison of the early wound healing stages between acute and chronic wounds ..........................................................................................................4 Figure 2. General overview of the wound healing stages: progression through the inflammatory, proliferation, and wound remodeling phases ......................................7 Figure 3. Wound healing delays in chronic wounds due to bacterial infection and failed immune response ............................................................................................13 Figure 4. Two common mechanisms that Methicillin-resistant S. aureus utilize to combat penicillin-derived antibiotics. (A) Beta-lactamase activity and (B) Utility of penicillinbinding protein 2’ ...........................................................................................16 Figure 5. Formation, development, and progression of physiological events in bacterial biofilms ........................................................................................................18 Figure 6. Quorum sensing is a significant factor that regulates biofilm formation and sustainment ...................................................................................................20 CHAPTER II. Physical Characteristics, Composition, and Germicidal Characteristics of a Novel “Spark-Based” Nonthermal Plasma Generator Discharge

vi

Figure 7. Schematic of the pulsed nonthermal plasma discharge generator..................44 Figure 8. Illustration of spectrophotometry and radiometry setup and analyses ...........46 Figure 9. Spectrophotometric nonthermal plasma discharge profile generated from room temperature ambient air gas mixture ...................................................................47 Figure 10. Detection of molecular oxygen from nonthermal plasma discharge in: (A) Candle soot streaks and (B) Kapton H polyimide wafer disks ..................................56 Figure 11. Detection of nitric oxide (A) and ozone (B) from a sealed chamber exposed to continuous nonthermal plasma discharge .............................................................57 Figure 12. pH of different media suspensions exposed to nonthermal plasma ..............59 Figure 13. Temperature of localized plasma exposure zone exposed to continuous nonthermal plasma discharge ............................................................................60 Figure 14. Nonthermal plasma discharge penetration through uniform, inoculated agar matrices in both (A) P. aeruginosa and (B) S. aureus models ..................................62 Figure 15. Diameters of germicidal zones from a fixed exposure focal point in Grampositive and Gram-negative bacteria ...................................................................64 Figure 16. Survivability of S. aureus cultures exposed to nonthermal plasma at different levels of filtration ...........................................................................................67 Figure 17. Survivability of S. aureus cultures exposed to nonthermal plasma within different starting aqueous media ........................................................................68 CHAPTER III. Bacterial Biofilm Physiological and Structural Responses to Nonthermal Plasma vii

Figure 18. Schematic of the biofilm continuous culture setup ..................................86 Figure 19. LOWESS statistical curves in P. aeruginosa biofilm structural characteristics ....................................................................................................................

95

Figure 20. LOWESS statistical curves in S. aureus biofilm structural characteristics ....97 Figure 21. S. aureus biofilms exposed to nonthermal plasma. (A) No exposure at 100x TM, (B) 1 minute at 100x TM, (C) 2 minutes at 100x TM, (D) No exposure at 400x TM, (E) 1 minute at 400x TM, and (F) 2 minutes at 400x TM ......................................100 Figure 22. COMSTAT-generated binary images of Z-stacks in S. aureus biofilms. (A) No exposure D1 biofilm, (B) 1 minute exposure D1 biofilm, (C) 2 minutes exposure D1 biofilm, (D) No exposure D4 biofilm, (E) 1 minute exposure D4 biofilm, and (F) 2 minute exposure D4 biofilm ............................................................................101 Figure 23. ImageJ-generated 3-dimension composite images of S. aureus biofilm stacks. (A) No exposure - Z dimension, (B) No exposure - all dimensions, (C) 1 minute - Z dimension, (D) 1 minute - all dimensions, (E) 2 minutes - Z direction, and (F) 2 minutes all dimensions ..............................................................................................102 CHAPTER IV. Physiological Alterations in Virulence Factor Producing Pathogens Following Nonthermal Plasma Exposure Figure 24. Methicillin-resistant S. aureus negative control OD values. (A) No nonthermal plasma and (B) No methicillin ........................................................................129 Figure 25. Methicillin-resistant S. aureus OD values following simultaneous methicillin and NTPD exposure ......................................................................................130 viii

Figure 26. Scaled percentages indicative of increased methicillin sensitivity in MRSA cultures exposed to nonthermal plasma and re-introduced to methicillin...................132 Figure 27. Logarithmic plate counts of surviving MRSA cultures following methicillin and/or NTPD ................................................................................................133 Figure 28. Propidium Iodide & Cyto-9TM viability dual stain in MRSA cultures. (A) No methicillin or NTPD, (B) Methicillin only, (C) 0.5 min NTPD only, (D) 1 minute NTPD and methicillin, (E) 2 minutes NTPD and methicillin, and (F) 0.5 minute NTPD and methicillin ...................................................................................................135 Figure 29. PCR gel that indicates a band of DNA that corresponds the size of the mecA gene ...........................................................................................................137 Figure 30. Early SDS-PAGE gel confirming slight presence of a protein band with a size indicative of PBP-2’ ......................................................................................138 Figure 31. Formation of membrane-derived carbonyls in MRSA cultures exposed to NTPD .........................................................................................................143 Figure 32. Fluorescence intensity of membrane-derived aldehyde fluorescent probes in MRSA cultures exposed to NTPD ....................................................................146 Figure 33. PBP-2’ agglutination and viability in MRSA cultures exposed to sub-lethal NTPD intervals. (A) Agglutination measurement immediately after NTPD, (B) Agglutination measurement 24 hours after NTPD (recovery), and (C) Logarithmic survivability of cultures exposed to NTPD .........................................................150 CHAPTER V. Stimulation of Wound Healing in Applied Human Skin In vitro Models ix

Figure 34. Keratinocyte proliferation over three weeks post nonthermal plasma exposure ..................................................................................................................

168

Figure 35. Fibroblast migration into damaged cell monolayers post nonthermal plasma exposure .....................................................................................................169 Figure 36. In vitro skin bunch biopsy model histology slices at the wound edge. (A) No plasma exposure, (B) 3 minutes plasma exposure, and (C), 6 minutes plasma exposure ..................................................................................................................

171

Figure 37. Average counts of keratinocytes found within the nascent epithelium .......172 Figure 38. Average thickness of nascent epithelium in the skin punch biopsy model after nonthermal plasma exposure ...........................................................................173 Figure 39. Immunohistochemistry slide montage detecting the relative intensity of integrin beta-1 after nonthermal plasma exposure ................................................179 Figure 40. Immunohistochemistry slide montage detecting the relative intensity of cytokeratin 10 after nonthermal plasma exposure ................................................180 Figure 41. Calculated fluorescence intensity of integrin beta-1 ..............................183 Figure 42. Calculated fluorescence intensity of cytokeratin 10 expression in immunohistochemistry specimens exposed to nonthermal plasma ..........................184 Figure 43. Comparison of immunohistochemical fluorescence between integrin beta-1 and cytokeratin 10 in an approximate region of epithelium of specimens prepared using skin punch biopsy and plasma exposure ............................................................187

CHAPTER VI. General Conclusions x

Figure 44. Compilation of effects caused in prokaryotic and eukaryotic cells over the course of a fixed nonthermal plasma exposure interval .........................................205 Figure 45. Changes in bacterial physiology in planktonic culture and biofilm architecture caused by sub-lethal NTPD exposure intervals ....................................................206

xi

LIST OF TABLES CHAPTER III. Bacterial Biofilm Physiological and Structural Responses to Nonthermal Plasma Table 1. Changes in measured gray scale intensity of textural and structural parameters of Pseudomonas aerguinosa biofilms after varying exposures to nonthermal plasma discharge over time ................................................................................................110-111

Table 2. Changes in measured gray scale intensity of textural and structural parameters of Staphylococcus aureus biofilms after varying exposures to nonthermal plasma discharge over time ......................................................................................................112-113

CHAPTER IV. Physiological Alterations in Virulence Factor Producing Pathogens Following Nonthermal Plasma Exposure Table 3. Complete Optical density values of MRSA cultures exposed to both NTPD and methicillin [4 ug/mL] .....................................................................................151 Table 4. Complete methicillin re-sensitivity percentages compared to germicidal reductions ....................................................................................................152 Table 5. Logarithmic survivability of MRSA plated cultures exposed to NTPD or methicillin at varied levels of filtration ..............................................................153

xii

ACKNOWLEDGEMENTS

I would like to acknowledge the Ohio Third Frontier Innovation Fund, Akron General Medical Center Foundation, Kent State University Graduate Student Senate, and 5iTech as important funding sources for the work in this dissertation. I would like to acknowledge the members of my doctoral dissertation committee, Drs. Chris Woolverton, Judy Fulton, Derek Damron, Gail Fraizer, and Oleg Lavrentovich for their invaluable input leading to the completion of my dissertation work. I would like to additionally recognize Drs. Chris Woolverton and Judy Fulton as significant influences on the direction of my professional and academic career has been shaped into. I also acknowledge my labmates in the Woolverton Lab (past and present), Dr. Steve Fiester, James Redfearn, and Mike Shilling for their valued support, scientific debate, and general camaraderie over the years. I also would like to acknowledge other friends I’ve learned from and leaned upon during the ups and downs that accompany graduate school: Drs. Kurtis Eisermann, Jessie Guinn, Jr., Stephan Woods, Shandilya Ramdas, and Allison Brager. I would also like to take a moment to acknowledge Dr. Jessie Francl for her countless hours of love and support through my dissertation work and process. I would also like to recognize and thank my mother, Jeannie Ferrell, for her years of sacrifice and encouragement. Additionally, her instilment of the values of hard work, discipline, and integrity has shaped me personally and professionally.

xiii

CHAPTER I

General Introduction Chronic wounds The skin serves as the outer boundary and primary line of defense against infection, fluid loss, and external environmental insults. Long term disruption of skin integrity due to chronic injury or infection can lead to subsequent morbidity or mortality. A critical objective during wound repair is re-epithelialization of the wound in a timely manner (Lanza et al. 2007). In 1992, it was estimated that over thirty-five million cases of significant skin injury required medical treatment, and twenty percent of these wounds became chronic (Ruckley 1997). Chronic wounds are a major health-care problem because they can lead to amputation, disability, and decreased quality of life in patients. The incidence of chronic wounds is currently 0.78% of the general population (Crovetti et al. 2004). In 2006, the cost associated with chronic wounds in the United States exceeded ten billion dollars annually, half of the total cost of all skin disorders, including burns (Nosenko et al. 2009). Chronic wounds are found most commonly in people who are over the age of 60 (Mustoe 2004). Because a significant number of the current U.S. population was born during the 1940s to 1960s, the cases of chronic wounds is expected to rise as this population cohort ages. Furthermore, the increasing incidence and 1

2

prevalence of obesity and diabetes also contribute to increases in chronic ulcerations moving into the future. Acute wounds are caused by sudden, external damage to intact skin. Examples include surgical wounds, bites, cuts, abrasions, and severe traumatic wounds such as gunshots or crush injuries (Davis et al. 1992). Chronic wounds contrast acute wounds and describe wounds that fail to progress through the normal phases of wound healing in a timely sequence (Figure 1). Chronic wound persistence is thought to be caused in part by an elevated shift in destructive processes at the expense of proliferative processes (Mustoe 2004; Kirketerp-Moller et al. 2008). Chronic wounds can be classified into three main categories: venous ulcers, diabetic ulcers, or pressure ulcers. Of these, venous ulcers account for approximately 70% of chronic wounds are primarily associated with the elderly (Snyder 2005). Imbalances in venous blood flow to the extremities are the most common cause of this specific sub-type of chronic ulcer. Additional complications caused by ischemia and reperfusion further exacerbate these wounds. Diabetic ulcers are of growing concern due to increasing rates of diabetes mellitus in the U.S. population. Diabetics have a higher risk of amputation compared to the general population due to decreased perception of pain and neuropathy disorders (Armstrong et al. 1998). Diabetics often present with delayed immune response and damage to small blood vessels complicating adequate blood oxygenation at the extremities (Pozzilli et al. 1994). These patients often fail to detect non-healing wounds and suffer re-injury or microbial infection as a result. Pressure ulcers typically results as a consequence from movement limitation of body parts that are commonly subjected to elevated pressure. Pressure ulcers are caused by ischemia that occurs when pressure

3

placed on the tissue is greater than the pressure in capillaries, restricting blood flow into the area. Current treatment options for chronic wounds vary in their approaches but seek to address the sources of chronic wound complications, including ischemia, bacterial load, and protease imbalances. Topical antibiotic creams or gels are often used to reduce microbial viability and keep the wound area moist. Disinfectants are not effective options because they can damage tissues and delay wound contraction (Thomas et al. 2009). The active ingredient is often imparied by organic matter in wounds like blood and exudate, eliminating their utility in open wounds. Topical antibiotics are ineffective clearing chronic wounds complicated by bacterial biofilms or antibiotic-resistant microorganisms. Severe wounds are often surgically debrided to remove as much devitalized tissue as possible. Debridement and drainage of wound fluid are an especially important part of the treatment for diabetic ulcers. Elevated wound exudate and necrotic tissue can serve as a positive medium for bacterial growth and increases likelihood of infection (Robson 1997). However, it is often impossible to completely remove all of the bacterial biomass or devitalized tissue. This raises increases the likelihood of re-infection at a later time. Patients with venous ulcers may undergo arterial revascularization procedures to correct vein dysfunction and re-establish blood oxygenation at wound sites. Patients that are not candidates for this surgery can instead increase localized tissue oxygenation with Hyperbaric Oxygen Therapy (HBOT). This therapy can correct hypoxia, kill bacteria, increase the rate of growth factor production, fibroblast growth, and angiogenesis by saturating localized environments with balanced concentrations of

4

Figure 1. Comparison of the early wound healing stages between acute and chronic wounds. (Reprinted with permission)

5

oxygen (Wattel et al. 1990). Topical or localized oxygen application requires precise balance because extremely high oxygen levels can lead to production of destructive reactive oxygen species (ROS) byproducts. Additional wound healing therapies include constant re-introduction of growth factors and proteins involved in wound healing administered topically or skin-grafting. However, these options are time-consuming and expensive with limited guaranteed results.This necessitates a mechanism that can balance and optimize oxygen and its derivatives in wound care.

Wound healing Wound healing is a highly dynamic process that can be compromised by endogenous pathological morbidities such as poor oxygenation of peripheral extremities, metabolic diseases such as diabetes mellitus, or lifestyle choices such as tobacco usage, and exogenous factors such as microbial infection or extensive burns (Bowler et al. 2001). The general process of wound healing is divided into three main phases: inflammatory, proliferation, and maturation (Bowler 2002) (Figure 2). These phases a The inflammatory phase of wound healing begins after hemostasis and initial fibrin clot formation at the wound site. Vasoconstriction also occurs during this time to reduce blood loss following injury. Secreted chemokines attract and stimulate polymorphonuclear neutrophils (PMNs) to the injury site. Inflammatory proteins then trigger a vasodilation response that elevates the localized porosity of capillaries, allowing immune cells easier access to the wound site. During the first day after injury, PMNs work to kill bacteria and clear damaged tissues or extracellular materials. Macrophages

6

then gradually replace PMNs as the predominant cell type found at the local wound site (Goldman 2004). The macrophage's main functions during wound healing are phagocytosis of bacteria and damaged tissue and debridement of damaged tissue through secreted proteases. Macrophages act as a bridge that connects the inflammatory and initial proliferative processes. Macrophages secrete a number of growth factors and other cytokines three to four days after injury that attract cells involved in the proliferation stage of healing to the wound site. Macrophages also establish the pre-cursors for angiogenesis and keratinocyte activation responsible for wound re-epithelialization (Goldman 2004). Tissue engineered skin equivalent models suggest that the presence of macrophages can delay wound contraction and that macrophage dispersal from the wound site is essential for wound healing progression (Newton et al. 2004). Inflammation is a normal response associated with wound healing and is necessary during early wound healing as immune cells remove bacteria, devitalized tissue, and other foreign debris from the wounded area. However, chronic inflammation can damage or kill host cells and tissues and leading to delays in wound closure (Bjarnsholt et al. 2008). Persistent and elevated inflammatory cell activity is considered to be an indicator of pathogenesis in chronic wounds (Yager et al. 1999; Mustoe 2004; Homey et al. 2006). Furthermore, diminished blood flow, hypoxia, and ischemia exacerbate inflammatory imbalances. In general, microbial infection extends the inflammatory and debridement phases of wound healing (Mustoe 2004; Homey et al. 2006). This can serve as positive feedback for already hyperactive inflammatory processes, further delaying wound closure. Necrotic tissue with localized hypoxia creates

7

Figure 2. General overview of the wound healing stages: progression through the inflammatory, proliferation, and wound remodeling phases

8

favorable environments for anaerobic bacteria growth (Roy et al. 2007). Wound infection and other co-morbidities associated with chronic wounds can lead to even further deficient host immune response (Bowler 2002; Clark et al. 2007; Bjarnsholt et al. 2008). Management of inflammatory processes and regulation of antioxidant pathways help mitigate endogenous damage in local wound environments. The macrophage plays an important role overseeing this balance and progression into subsequent stages. Fibroblasts migrate into the wound site and proliferate once macrophages as macrophages disperse. Fibroblasts bind to cross-linked fibrin fibers laid down during late inflammatory phase and migrate through the wound site. The origins of these fibroblasts are from the adjacent uninjured cutaneous tissue or derived from blood-borne, circulating adult stem cells (Broughton et al. 2006). Angiogenesis occurs concurrently and lays the foundation for later processes that occur during the proliferative phase. Fibroblasts secrete growth factors in order to encourage local cell proliferation, migration, and development of early provisional ECM components that leads to granulation tissue formation (Eckes et al. 2000). Granulation tissue consists of nascent blood vessels formed during angiogenesis, fibroblasts, inflammatory cells, endothelial cells, and a provisional extracellular matrix (ECM) environment. Granulation tissue is a hydrated matrix composed of fibronectin, collagen, glycosaminoglycans, proteoglycans, fibronectin and hyaluronan.(Bowler 2002). Granulation tissue differs incomposition from the ECM in normal tissue because the majority of its components originate from fibroblasts (Broughton et al. 2006). This temporary matrix facilitates cell migration but does not provide significant protection against re-injury of the healing wound. Prior to collagen

9

deposition, the fibrin clot and developing granulation tissue supports and strengthens the overall wound environment. Granulation tissue will ultimately be replaced in later stages by a collagen-based ECM that resembles extracellular matrices found in non-injured tissue. Collagen deposition is important because collagen increases the strength and stability within the overall wound environment as the wound completely heals. The formation of granulation tissue in an open wound allows re-epithelialization to take place. This process is defined by migration of epithelial cells across the surface of new tissue in order to establish a new barrier between the healing wound and the external environment. Keratinocytes found at wound edges and around dermal appendages such as hair follicles, sweat glands and sebaceous glands are the sources of the cells responsible for re-epithelialization (Martin 1997). Keratinocyte migration across the wound site is initiated by lack of contact inhibition and stimulated by signal molecules such as nitric oxide (Hunt et al. 2000). Fibrin, collagen, and fibronectin within the ECM further signals cells to divide and migrate. Like fibroblasts, migrating keratinocytes attach to fibrin and fibronectin scaffolding materials laid down during inflammation in order to crawl across the surface. As keratinocytes migrate, new epithelial cells must form at the wound edges to replace migrating cells and provide additional cells for the advancing sheet (Hunt et al. 2000). Keratinocytes will continue to migrate across the wound bed until cells from either side meet in the middle, at which point contact inhibition triggers the end of migration. As keratinocyte migration ceases, re-formation of a new basement membrane occurs.

10

Contraction is a key phase of wound healing and fits into wound maturation and remodeling phases. During this phase, some sub-groups of fibroblasts gradually differentiate into myofibroblasts. Myofibroblasts are similar physiologically to smooth muscle cells and possess significantly more actin than their traditional fibroblast counterparts. The actin within the myofibroblasts contracts as the cells migrate across collagen tracks, pulling the wound edges together. Fibroblasts support the contraction process by laying down supplemental collagen to reinforce the wound environment during wound contraction (Stadelmann et al. 1998). As the provisional matrix is replaced with a long-term, stable collagen-based matrix, proliferation and migration ends once cessation signals are received by myofibroblasts (Stadelmann et al. 1998). The final stage of wound healing is the wound maturation phase and is characterized by collagen synthesis and degradation cycles. Collagen fibers are rearranged, cross-linked, and aligned along tension lines, signaling remodeling and stabilization of the wound environment. As wound maturation progresses, the tensile strength of the wound increases over time. The repaired tissue will regain 50% of its tensile strength when compared to undamaged tissue after three months of healing. This new tissue can ultimately gain back up to 80% of its tensile strength (Broughton et al. 2006).

Virulence factors Virulence factors are a collection of biochemical molecules or enzymes used by pathogenic bacteria to achieve any or all of the following: colonization of a host,

11

maintenance of a host niche once established, evasion of a host’s immune response, inhibition of a host’s immune response, or direct acquisition of nutrients from the host (Drake et al. 1988; Lowy 1998). Virulence factors can range from structural or motility components such as membrane-bound proteins or flagella to secreted cytolytic toxins or immunosuppressant superantigens. Pathogenic bacteria do not constitutively produce virulence factors. They often require precise environmental and metabolic cues, typically from within a specialized niche within a host (DeBell 1979). Many pathogenic bacteria couple the detection of specific nutrients with the regulation of genes that encode virulence proteins needed for interaction with target host cells. For example, Streptococcus pyogenes expresses several virulence genes in an ordered temporal pattern that is heavily influenced by nutrient availability, such as the iron-inducing production of streptolysin exotoxin. (Smoot et al. 2001). Likewise, Clostridium difficile produces its toxin A and B in response to intrinsic biotin deficiency and detection of biotin from target cells (Dupuy et al. 1998). This ordering is critically important in bacterial pathogenesis since many global transcription regulators involved in this process have been identified, and mutation of these regulators result in large variations in virulence (Kreikemeyer et al. 2003). Pathogenic bacteria are able to utilize metabolic pathways and up-regulate needed proteins in order to adapt to changes in their micro-environment as colonization and subsequent exploitation of a niche occurs. Wound microbiology is highly complex and detailed investigations have shown that wounds are polymicrobial, composed of both aerobic and anaerobic bacteria (BROOK et al. 1981; Rotstein et al. 1985; Bowler et al. 1999). Staphylococcus aureus

12

and Pseudomonas aeruginosa are aerobes commonly associated with skin infections, including chronic wounds and extensive burns (Schaberg et al. 1991; Kirketerp-Moller et al. 2008). Both are frequently part of human flora found in the nasal cavities and on the surface of the skin. However, if the skin is ruptured, opportunistic infection can occur. Wound infection is defined as the presence of replicating organisms with a total number of microorganisms greater than 106 per gram of tissue or mL of fluid within a wound with subsequent host injury (Bowler and Davies 1999). Microbial infection delays and limits the early stages of wound closure (Figure 3). There are varying theories regarding how bacteria delay wound healing. One school of thought suggests that the microbial load is too great for the host immune system to clear efficiently (Bendy Jr et al. 1964; LEVINE et al. 1976; Raahave et al. 1986; Noble 1999). Another theory postulates that virulence factors produced by polycolonized wounds lock the wound healing response in early phases characterized by limited germicidal action and localized host tissue damage causedby inefficient, chronic inflammation (Tarnuzzer et al. 1996; Yager and Nwomeh 1999; Ovington 2003). Virulence factor content in wounds can be extremely high and can stimulate extensive superoxide radical production from neutrophils (Bianca et al. 2002). Proteinase imbalance is also thought to be an important determinant in the pathogenesis of chronic wounds. Bacteria mediate the secretion of enzymes designed to degrade the extracellular matrices of cells and damage the structural layering of cells in order to penetrate deeper into the wound bed (Tarnuzzer and Schultz 1996). Endotoxins released by cells killed by the host’s immune response can further damage the wound site long after the offending

13

Figure 3. Wound healing delays in chronic wounds due to bacterial infection and failed immune response

14

pathogen has been eliminated. Endotoxins in the wound environment stimulate the PMNmediated release of pro-inflammatory molecules which in turn induce endogenous matrix metalloproteases (MMP) secretion. Unbalanced MMP levels further contribute to the local destruction of growth factors, receptors, and healing tissue components (C. Konturek 2001; Power et al. 2001; Ovington 2002). The two most common bacteria associated with wound infections are S. aureus and P. aeruginosa and each possess individualized virulence factors used to adapt and colonize a novel wound environment (Drake and Montie 1988; Lowy 1998). Coagulase is an enzyme used by S. aureus that cleaves fibrinogen to form fibrin and is used to help the bacteria generate a sticky fibrin clot that facilitates colonization. Protein A is a virulence factor found on the cell membrane of S. aureus that is able to bind immunglobulins or complements together, impairing their native functions (Rogers 1954; Kronvall et al. 1971). Staphylococcal extracellular adherence proteins have also been directly implicated in cessation of effective immunological responses that are necessary during early wound healing. These proteins permit S. aureus binding to host extracellular matrix components such as fibronectin, fibrinogen, vitronectin, and collagen (Flock et al. 1987; Boden et al. 1989; Paulsson et al. 1992; Flock 1999). Exotoxin A produced by P. aeruginosa is commonly found in clinical human wound isolates (Lyczak et al. 2000) and inhibits protein synthesis by inhibiting elongation factor 2 via ADP-ribosylation. P. aeruginosa is also flagellated and is capable of motility and evasion of host immune defenses during the inflammatory processes associated with early wound healing (Young 1980; Lyczak et al. 2000). Because the wound environment is highly variable and

15

experiences rapid immunological response and radical turnover as the wound site attempts to heal, virulence factors are necessary for the immediate survival and prolonged colonization of offending bacteria. Virulence factors have also been linked to facilitation of antibiotic resistance (Figure 4). Microorganisms can acquire or develop mechanisms to adapt to antibiotics including: alteration of the antibiotic’s target, enzymatic modification of the antibiotic, efflux pumps, or heightened impermeability at the cell membrane (Levy 1998; Chambers et al. 2009; KONG et al. 2010). Hiramatsu et al. detected an additional 30 to 50 kb of additional chromosomal DNA found in methicillin-resistance Staphylococcus aureus (MRSA) isolates not found in methicillin-susceptible Staphylococcus aureus (MSSA) isolates, labeled the mec class of genes (Hiramatsu et al. 1996). Methicillin resistance in MRSA is frequently due to production of a low-affinity penicillin binding protein named PBP2 α, which conveys β-lactamase resistance (Chambers 1997). However, methicillin resistance in S. aureus is not only linked to the PBP2 α protein coded by the mecA gene class. Methicillin resistance can also be characterized by the overproduction of βlactamase from the blaZ gene instead of solely pencillin-binding proteins activity. Overproduction of β-lactamase describes low-level or borderline resistance. This resistance phenotype is associated with lower MIC values and is additional dynamic that must be considered when evaluating experimental or clinical S. aureus strains for resistance patterns or mechanisms. The methicillin resistance phenotype exists on a continuum with intermediate or heterogeneous stages that depend on the antibiotic and the culture environment (Matthews et al. 1984; Hartman et al. 1986).

16

Figure 4. Two common mechanisms that Methicillin-resistant S. aureus utilize to combat penicillin-derived antibiotics. (A) Beta-lactamase activity and (B) Utility of penicillinbinding protein 2’

17

Bacterial biofilms Bacterial biofilms consist of growing bacterial colonies that display altered physiological states compared to their planktonic forms. The bacteria in these communities form into biofilms in response to a variety of environmental conditions including nutrient starvation, changes in temperature, or introduction of a novel surface substrate (Costerton et al. 1999). Biofilm formation follows a series of ordered events that begins with early, reversible colonization of a surface that gradually leads to irreversible attachment and subsequent recruitment of other microorganisms in the nearby microenvironment (Figure 5). As the biofilm grows in size and complexity, an exopolysaccharide (EPS) matrix is produced and is used to encapsulate the bacteria, protecting it from damage and other external threats (Parsek et al. 2003). This external layer also provides additional attachment sites leading to formation of secondary subbiofilms characterized by subsequent species’ colonization events. Once a biofilm reaches a maximum size or uses up a local environment’s resources, the peripheral members within the biofilm cede and migrate away from the overall biofilm structure. These bacteria gradually revert back to their free-floating forms and can lead towards secondary colonization events and formation of entirely new biofilms at distant sites. The bacteria within the biofilm shift their metabolism as they interact with each other and additional bacterial species within a growing biofilm. Mature biofilms tend to have a “colonial” center composed of bacteria that initially colonized the surface and is

18

Figure 5. Formation, development, and progression of physiological events in bacterial biofilms

19

surrounded by growing micro-colonies that branch out to and around the periphery of the community. It is not uncommon to observe differences in growth or metabolic events even in different portions of the same biofilm (Lewis 2005). Regardless of the relative complexity of the microbial community, biofilm formation is dependent on cellular interactions, particularly quorum signaling. Quorum sensing is a system of stimulus and response proportional to population density. Quorum signaling molecules released are used to identify the number of total individuals within the overall biofilm population (Figure 6). Quorum sensing is directly involved with EPS production, changes in motility, and alterations in gene expression which collectively lead to shifts in metabolic uptake, cell swarming, and aggregation events. P. aeruginosa biofilms are a standard model used to understand how formation and development are mediated by quorum sensing. Mutants that are incapable of quorum signaling are not able to form biofilms when compared to wild-type organisms (Shih et al. 2002; Zhu et al. 2003). Biofilms are fairly ubiquitous and can be found growing in a wide variety of environments. Evidence of biofilm formation has been found in hydrothermal environments dating back to the early fossil record (Westall et al. 2001). Biofilm formation can be considered an extension of bacterial life analogous to sporulation. However, even though biofilm formation is widespread, the physical characteristics, metabolic processes, and interaction with other organisms vary from one biofilm community to another. Whereas some bioremediation utilities have been explored using biofilms, pathogenic bacteria that form biofilms tend to not only create new disease, but also often exacerbate existing morbidities (Singh et al. 2006; James et al. 2008).

20

Figure 6. Quorum sensing is a significant factor that regulates biofilm formation and sustainment

21

Furthermore, biofilm formation is particularly problematic in medical settings where patients require extended hospitalization following acute injury, extensive burn injuries, or invasive surgical procedures. Antibiotic resistance and subsequent nosocomial infection rates continue to grow and remain one of the top public health challenges that health-care providers encounter (Schaberg et al. 1991; Vincent et al. 1995; Rosenthal et al. 2010; Yanagi et al. 2010; Raad et al. 2011). The presence of biofilms in hospitals contributes towards these rates for a variety of reasons. The exopolysaccharide matrix secreted by maturing biofilms serves to physically shield embedded bacteria by blocking or denaturing the antibiotics before they can reach targeted cells sequestered deep within the biofilm. In addition, because the metabolism and physiology of bacteria within biofilms differ from their planktonic counterparts, antibiotic therapies with specific mechanisms of action are often unable to efficient affect their molecular targets (Stewart et al. 2001; Stewart 2002). Any time antibiotic therapies are not completed properly, the potential for evolution of resistance can occur. This is especially problematic if antibiotic-resistant cells disassociate from the biofilm and colonize a secondary location. Colonization events and biofilm formation on abiotic surfaces such as joint replacement constructs or catheters is an additional problem that can lead to inadvertent secondary infections that may require additional surgical intervention (Saint et al. 2003). Collectively, biofilm formation in medical settings leads to extended hospitalization, increased cost, morbidities, mortalities, diminished quality of life, and contributes to the growing threat of antibiotic resistance (Maple et al. 1989; Costerton et al. 1999; Neuhauser et al. 2003; James et al. 2008).

22

Microbial infection has been associated with deleterious effects on wound healing for decades. Control of the bacterial burden must be considered as treatment and management regimes are designed. Biofilm-related diseases are persistent infections that develop transiently and are rarely completely resolved by host immunological response (Costerton et al. 1999; Parsek and Singh 2003). Biofilm formation has been shown to correlate directly with delays in wound healing (James et al. 2008; Schierle et al. 2009). Biofilms have been identified in association with a variety of chronic wound subtypes including diabetic wounds, venous stasis ulcers, and pressure sores (Wolcott et al. 2008). Biofilms in chronic wounds often resemble slime layers equipped with extracellular matrix materials that closely resemble EPS. Debridement of bacterial biofilms and devitalized tissue in chronic wounds has limited success due to the cryptic, penetrative, and hardy nature of bacterial biofilms. Failure to remove significant portions of the biofilm can often lead to re-colonization and re-infection events. This in turn stimulates imbalanced inflammatory processes in chronic wound patients, resulting in a feedback loop that ultimately provides little to no long term positive wound healing progress.

Nonthermal plasmas Plasma science has been of particular interest to researchers spanning several disciplines since the 1990s. This continuously evolving field allows investigations into a diverse series of applications including industrial sterilization, pollution control, polymer science, food safety, and biomedicine (Birmingham et al. 2000; Lerouge et al. 2001; Kim

23

2004; Vleugels et al. 2005; Gaunt et al. 2006; Fridman et al. 2008; Kong et al. 2009). The potential for interdisciplinary collaborative efforts that focus on plasma science is vast. Likewise, there is a growing need to further develop additional applications or mechanisms of action within plasma science. Wound healing is one such applied example within medical arenas, as one would have to assess plasma effects on the damaged microenvironment (devitalized tissue, debris, foreign and host bacteria or other prokaryotes) and on the eukaryotic cells (keratinocytes, fibroblasts, PMNs, etc.) that function together during a wound healing response. Two categories of plasma, thermal and nonthermal plasma, are categorized based on their creation. Thermal plasmas are obtained at high pressures (usually greater than 105 Pa) and need a significant amount of power (up to 50 MW) to be observed continuously. In addition, these plasmas generate a significant amount of thermal energy (5 to 24 K) (Moreau et al. 2008). Nonthermal plasmas produced by various discharge categories possess certain common properties. These plasmas possess very low energy cost because the majority of the electrical energy is used to produce energetic electrons, instead of heating the surrounding gas. This implies that the electron temperature should be higher than the gas temperature for this to occur. The elevated temperature and excitatory nature of the charged electrons and plasma molecules are responsible for initializing the chain reactions associated with observed phenomenology. For many types of discharges, the gas kinetic temperature is room temperature. This low gas temperature can be achieved by limiting the time the gas spends in the electrification process (Napartovich 2001).

24

Plasma is generated when sufficient energy is supplied to a neutral gas to initiate charge production. Plasmas are composed of particles that interact in continuous chain reactions with other particles which lead to further downstream effects. These particles include photons, electrons, ions, free radicals, and other excited molecules. Electric fields are used to ionize starting gaseous environments and create plasma from gaseous environments (Napartovich 2001). Electric fields are successful for plasma generation because the charge densities are able to excite the tiny masses of electrons. It is theorized that the excitatory action of charged electrons and ions lead to what is called an “electron avalanche” or massive chain reaction of downstream molecules. This concept is defined by the Townsend mechanism and is an important concept within the nature of individual plasmas (Yokoyama et al. 1990; Massines et al. 2005). The electrification of particles in a localized environment leading to the amplification and extension of surrounding particles is a significant concept of plasma activity. The very nature of the localized environment can also vary depending on the type of plasma that is needed. Plasmas can be created using noble gasses, ambient air, or aqueous vapors (Ogata et al. 2000). The starting medium that is used to generate the plasma can convey different properties inside produced plasma discharge profiles. For example, using water vapor can dramatically increase the amount of hydroxides that make up the plasma’s composition (Lukes et al. 2004). The ability to “customize” a plasma specific sub-type of plasma in order to perform a specific task is something that merits additional investigation. The specificity of plasmas are dependent upon gas compositions, pressure, air flow, frequency of the applied voltage, construction of the

25

cathode/anode configuration, and sustainment of the electric field (Conrads et al. 2000; Fridman et al. 2008). The most common method used for creating plasma involves the electrical breakdown of a neutral gas in the presence of an electrical field. There are several general classes of known plasma generators based on: direct current, alternating current, or pulseproducing periodic discharge (Moisan et al. 2001; Napartovich 2001; Akishev et al. 2002; Kim 2004). The class of plasma generators that rely on alternating current can be further divided into sub-categories: high-frequency discharge and dielectric barrier discharge. Plasma generators based on direct current (constant voltage) function by introducing gas through a discharge gap constructed from dielectric materials. A strong electric field then forms between a conic anode and a sharp cathode, also called a needle. Ionization occurs inside the anode which causes a “jet” of plasma to form at the end of the cathode. Plasma generators that use high-frequency discharge are constructed using a sharp metal-alloy pin inserted coaxially into a thermoplastic tube. Ionization of the gaseous ambient environment occurs, thereby allowing the charged plasma molecules to flow through the needle. Plasma created using this type of generator is usually higher in temperature (100°C) compared to other types (Schutze et al. 1998). Plasma generated using high-frequency discharge is composed of a continuous stream with no discernible discharge pattern or periodicity (Moreau 2007). Dielectric barrier discharges (DBDs) are created between flat, parallel metal plates covered with thin layers of dielectric material. The plates are driven by a high frequency electric

26

current that creates an electric field used to ionize gasses. The dielectric material enables proper plasma discharge function by limiting the charge as it is transported through the system and distributed uniformly over the electrode area (Kogelschatz et al. 1999; Kogelschatz 2003). DBD plasma utility can be limited in certain applications because these plasma generators require flat, dry surfaces (Kong et al. 2009). While this may be advantageous for industrial sterilization, its utility in certain biomedical arenas may be impractical or ineffective. Plasma generators that utilize pulsed periodic discharge produce an uncontrolled random discharge, sometimes referred to as a “spark” discharge (Chang 2001; Fridman et al. 2005). These devices contain a nozzle within an installed rod electrode that is sharpened at one end and connects to the source of a gaseous medium. An output electrode extends beyond the nozzle configuration. The electric field that is generated occurs between the electrode in the needle and the electrode in the head nozzle creating a discharge gap. As the spark is produced, the plasma materials are expelled out through the gap. Since there is no direct electric field outside of the discharge gap, this plasma discharge subtype is safe to use on biological specimens. Plasma generated using this type of device is lower in temperature and larger in overall volume compared to other types of plasma (Sugiarto et al. 2001). This current investigation utilizes a variant of the pulse-producing periodic discharge capable of controlled, uniform nonthermal plasma discharges.

27

General overview and approach Plasma science and the associated technologies it has produced over the last twenty years continue to grow and reach further into different scientific fields of study. The field of plasma science and its development is predominantly steeped in physics and engineering. Plasma has been successfully associated with industrial processes such as decontamination of noxious components found in the environment and mass sterilization of large non-biotic surfaces and constructs (Kogelschatz et al. 1999; Kogelschatz 2003; Lukes et al. 2004; Massines et al. 2005). There have been many observed phenomenon reported in the literature including the germicidal capabilities of a wide variety of plasma generators. Nonthermal plasma has been shown to be effective as an antimicrobial agent capable of sterilizing Gram-positive and Gram-negative bacteria, fungi, viruses, and associated spores and biofilms (Vleugels et al. 2005; Fridman et al. 2008; Kolb et al. 2008). Plasma studies have also demonstrated beneficial biological effects in human cells such as stimulation of apoptosis in cancer cells or up-regulation of signals needed for blood coagulation (Fridman et al. 2006; Fridman et al. 2007). Recent investigations discuss the validity and necessity of future development of nonthermal plasma discharges customized for chronic wounds (Nosenko et al. 2009). While this is advantageous for the overall future of plasma science as a whole, a multi-disciplinary approach is needed to answer remaining questions and mechanisms involving biological responses to nonthermal plasma in an applied manner such as wound healing complicated with microbial interaction and infection.

28

Unanswered questions involving prokaryotic cell response to nonthermal plasma discharge include (1) how is the viability and structural integrity of bacterial biofilms affected by nonthermal plasma discharge and (2) does nonthermal plasma discharge alter expression or potency of virulence factors produced by pathogenic bacteria prior to germicide? Questions also exist from the eukaryotic cells’ point of view and include (3) does the composition of nonthermal plasma discharge possess reactive species sufficient to stimulate wound healing responses such as proliferation and migration at wound edges? Additionally, (4) what is an optimal nonthermal plasma exposure range needed to observe bactericidal or bacteriostatic effects in prokaryotic cells and simultaneous stimulation of proliferative and migratory effects in eukaryotic cells without significant eukaryotic cell injury. The final question asks that (5) should such a proposed interaction exist, what are the mechanisms involved?

Dissertation rationale and hypotheses It is the overarching aim of the following studies to determine how nonthermal plasma is able to disrupt biofilm formation and three-dimensional structure, alter virulence factor production in pathogenic bacteria, and stimulate wound healing responses including, but not limited to keratinocyte proliferation and migration in in vitro constructs. The proposed study will evaluate these concepts using the following three hypotheses:

29

Hypothesis 1: Nonthermal plasma application to S. aureus or P. aeruginosa biofilms will reduce bacterial viability and alter cellular physiology, characterized by fluorescence and confocal microscopy. Hypothesis 2: Pathogenic bacteria treated with sub-germicidal doses of nonthermal plasma will show reduced virulence factor expression but not at the expense of viability. Hypothesis 3: Whole pieces of human tissue exposed to nonthermal plasma will show increased repair and maturation as determined by microscopy and immunohistochemical detection of protein markers. The first hypothesis is based on observed research done initially in pathogenic bacteria sterilization experiments mediated by nonthermal plasma exposure. The composition and physical parameters of the produced nonthermal plasma discharge was characterized and found to penetrate and radiate fixed values from a central exposure focus. This exposure zone was able to penetrate through agar plates inoculated with bacteria. In these experiments, agar served as a hydrated biopolymer material comparable to early stage biofilms. This observation provided one of the core foundations of this proposed study. This foundational physical experimentation can be found in its entirety in Chapter II. The second hypothesis is designed to test the physiological responses of both Gram-positive (S. aureus) and Gram-negative (P. aeruginosa) bacteria following nonthermal plasma leading up to cell death. Extensive findings are spread throughout the

30

literature documenting nonthermal plasma sterilization. The plasma discharge studied here was also confirmed to have significant germicidal capabilities. However, paucities exist in the literature identifying concrete cellular activity and events leading up to cell death mediated by nonthermal plasma usage. The production of virulence factors are a hallmark of healthy cellular function and maximization of a host niche in pathogens. Evaluation of the totality and functionality of isolated and purified virulence factors following nonthermal plasma stress at sub-germicidal levels provides a more complete understanding of how the discharge affects pathogenic cells up to death. The third hypothesis is based on observed research done initially using nonthermal plasma in order to stimulate eukaryotic cells in culture and observe their migration and proliferation. Early experimentation was designed to evaluate in vitro cell culture assays. Initial studies indicated stimulation of both migration and proliferation in two significant cell types involved in wound healing: keratinocytes and fibroblasts. An applied model was constructed to simulate wounds using injured whole tissue suspended in a collagen-based extracellular matrix following nonthermal plasma exposure. The initial proliferation and migration results laid the groundwork for this specific aim in the present study. The subsequent chapters address these hypotheses and conclude with a final synopsis that aims to determine two additional core objectives. A preliminary dose curve that balances germicidal effects in prokaryotes and proliferative effects in eukaryotes will

31

be provided. Mechanisms to explain these observed phenomena will also be postulated based on the totality of the findings.

32

References Akishev, Y., M. Grushin, et al. 2002. Novel AC and DC non-thermal plasma sources for cold surface treatment of polymer films and fabrics at atmospheric pressure. Plasmas and Polymers 7(3): 261-289. Armstrong, D. G., L. A. Lavery, et al. 1998. Validation of a diabetic wound classification system: the contribution of depth, infection, and ischemia to risk of amputation. Diabetes care 21(5): 855-859. Bendy Jr, R., P. Nuccio, et al. 1964. RELATIONSHIP OF QUANTITATIVE WOUND BACTERIAL COUNTS TO HEALING OF DECUBITI: EFFECT OF TOPICAL GENTAMICIN. Antimicrobial agents and chemotherapy 10: 147. Bianca, R., N. Wayman, et al. 2002. Superoxide dismutase mimetic with catalase activity, EUK-134, attenuates the multiple organ injury and dysfunction caused by endotoxin in the rat. Medical science monitor: international medical journal of experimental and clinical research 8(1): BR1. Birmingham, J. G. and D. J. Hammerstrom 2000. Bacterial decontamination using ambient pressure nonthermal discharges. Plasma Science, IEEE Transactions on 28(1): 51-55. Bjarnsholt, T., K. Kirketerp Møller, et al. 2008. Why chronic wounds will not heal: a novel hypothesis. Wound Repair and Regeneration 16(1): 2-10. Boden, M. K. and J. Flock 1989. Fibrinogen-binding protein/clumping factor from Staphylococcus aureus. Infection and Immunity 57(8): 2358. Bowler, P., B. Duerden, et al. 2001. Wound microbiology and associated approaches to wound management. Clinical Microbiology Reviews 14(2): 244. Bowler, P. G. 2002. Wound pathophysiology, infection and therapeutic options. Annals of medicine 34(6): 419-427. Bowler, P. G. and B. J. Davies 1999. The microbiology of acute and chronic wounds. Wounds 11: 72-78. BROOK, I. and J. G. RANDOLPH 1981. Aerobic and anaerobic bacterial flora of burns in children. The Journal of Trauma 21(4): 313. Broughton, G. I. I., J. E. Janis, et al. 2006. The basic science of wound healing. Plastic and reconstructive surgery 117(7S): 12S. C. Konturek, T. B., SJ Konturek, S. Kwiecien, A. Dembinski, EG Hahn, P. 2001. Influence of bacterial lipopolysaccharide on healing of chronic experimental ulcer in rat. Scandinavian journal of gastroenterology 36(12): 1239-1247. Chambers, H. F. 1997. Methicillin resistance in staphylococci: molecular and biochemical basis and clinical implications. Clinical Microbiology Reviews 10(4): 781-791. Chambers, H. F. and F. R. DeLeo 2009. Waves of resistance: Staphylococcus aureus in the antibiotic era. Nature Reviews Microbiology 7(9): 629-641. Chang, J. S. 2001. Recent development of plasma pollution control technology: a critical review. Science and Technology of Advanced Materials 2(3-4): 571-576. Clark, R. A. F., K. Ghosh, et al. 2007. Tissue engineering for cutaneous wounds. Journal of Investigative Dermatology 127(5): 1018-1029.

33

Conrads, H. and M. Schmidt 2000. Plasma generation and plasma sources. Plasma Sources Science and Technology 9: 441. Costerton, J. W., P. S. Stewart, et al. 1999. Bacterial biofilms: a common cause of persistent infections. Science 284(5418): 1318. Crovetti, G., G. Martinelli, et al. 2004. Platelet gel for healing cutaneous chronic wounds. Transfusion and Apheresis Science 30(2): 145-151. Davis, M., P. Dunkley, et al. 1992. Cause: types of wound and ulcers. The wound programme. Centre for Medical Education, Dundee, United Kingdom: 109–132. DeBell, R. M. 1979. Production of exotoxin A by Pseudomonas aeruginosa in a chemically defined medium. Infection and Immunity 24(1): 132. Drake, D. and T. C. Montie 1988. Flagella, motility and invasive virulence of Pseudomonas aeruginosa. Microbiology 134(1): 43. Dupuy, B. and A. L. Sonenshein 1998. Regulated transcription of Clostridium difficile toxin genes. Molecular microbiology 27(1): 107-120. Eckes, B., P. Zigrino, et al. 2000. Fibroblast-matrix interactions in wound healing and fibrosis. Matrix Biology 19(4): 325-332. Flock, J., G. Fröman, et al. 1987. Cloning and expression of the gene for a fibronectinbinding protein from Staphylococcus aureus. The EMBO Journal 6(8): 2351. Flock, J. I. 1999. Extracellular-matrix-binding proteins as targets for the prevention of Staphylococcus aureus infections. Molecular medicine today 5(12): 532-537. Fridman, A., A. Chirokov, et al. 2005. Non-thermal atmospheric pressure discharges. Journal of Physics D: Applied Physics 38: R1. Fridman, G., G. Friedman, et al. 2008. Applied plasma medicine. Plasma Processes and Polymers 5(6): 503-533. Fridman, G., M. Peddinghaus, et al. 2006. Blood coagulation and living tissue sterilization by floating-electrode dielectric barrier discharge in air. Plasma chemistry and plasma processing 26(4): 425-442. Fridman, G., A. Shereshevsky, et al. 2007. Floating electrode dielectric barrier discharge plasma in air promoting apoptotic behavior in melanoma skin cancer cell lines. Plasma chemistry and plasma processing 27(2): 163-176. Gaunt, L. F., C. B. Beggs, et al. 2006. Bactericidal action of the reactive species produced by gas-discharge nonthermal plasma at atmospheric pressure: A review. Plasma Science, IEEE Transactions on 34(4): 1257-1269. Goldman, R. 2004. Growth factors and chronic wound healing: past, present, and future. Advances in skin & wound care 17(1): 24. Hartman, B. J. and A. Tomasz 1986. Expression of methicillin resistance in heterogeneous strains of Staphylococcus aureus. Antimicrobial agents and chemotherapy 29(1): 85-92. Hiramatsu, K., N. Kondo, et al. 1996. Genetic basis for molecular epidemiology of MRSA. Journal of Infection and Chemotherapy 2(3): 117-129. Homey, B., M. Steinhoff, et al. 2006. Cytokines and chemokines orchestrate atopic skin inflammation. The Journal of allergy and clinical immunology 118(1): 178. Hunt, T. K., H. Hopf, et al. 2000. Physiology of wound healing. Advances in skin & wound care 13(2 Suppl): 6.

34

James, G. A., E. Swogger, et al. 2008. Biofilms in chronic wounds. Wound Repair and Regeneration 16(1): 37-44. Kim, H. H. 2004. Nonthermal Plasma Processing for Air Pollution Control: A Historical Review, Current Issues, and Future Prospects. Plasma Processes and Polymers 1(2): 91-110. Kirketerp-Moller, K., P. O. Jensen, et al. 2008. Distribution, organization, and ecology of bacteria in chronic wounds. Journal of clinical microbiology 46(8): 2717. Kogelschatz, U. 2003. Dielectric-barrier discharges: Their history, discharge physics, and industrial applications. Plasma chemistry and plasma processing 23(1): 1-46. Kogelschatz, U., B. Eliasson, et al. 1999. From ozone generators to flat television screens: history and future potential of dielectric-barrier discharges. Pure and applied chemistry 71(10): 1819-1828. Kolb, J. F., A. A. H. Mohamed, et al. 2008. Cold atmospheric pressure air plasma jet for medical applications. Applied Physics Letters 92: 241501. KONG, K. O. K. F. A. I., L. Schneper, et al. 2010. Beta‐lactam antibiotics: from antibiosis to resistance and bacteriology. Apmis 118(1): 1-36. Kong, M. G., G. Kroesen, et al. 2009. Plasma medicine: an introductory review. New Journal of Physics 11: 115012. Kreikemeyer, B., K. S. McIver, et al. 2003. Virulence factor regulation and regulatory networks in Streptococcus pyogenes and their impact on pathogen-host interactions. TRENDS in Microbiology 11(5): 224-232. Kronvall, G., J. H. Dossett, et al. 1971. Occurrence of protein A in staphylococcal strains: quantitative aspects and correlation to antigenic and bacteriophage types. Infection and Immunity 3(1): 10. Lanza, R. P., R. S. Langer, et al. 2007. Principles of tissue engineering, Academic Press. Lerouge, S., M. R. Wertheimer, et al. 2001. Plasma sterilization: A review of parameters, mechanisms, and limitations. Plasmas and Polymers 6(3): 175-188. LEVINE, N. S., R. B. LINDBERG, et al. 1976. The quantitative swab culture and smear: a quick, simple method for determining the number of viable aerobic bacteria on open wounds. The Journal of Trauma 16(2): 89. Levy, S. B. 1998. The challenge of antibiotic resistance. Scientific American 278(3): 3239. Lewis, K. 2005. Persister cells and the riddle of biofilm survival. Biochemistry (Moscow) 70(2): 267-274. Lowy, F. D. 1998. Staphylococcus aureus infections. New England Journal of Medicine 339(8): 520. Lukes, P., A. T. Appleton, et al. 2004. Hydrogen peroxide and ozone formation in hybrid gas-liquid electrical discharge reactors. Industry Applications, IEEE Transactions on 40(1): 60-67. Lyczak, J. B., C. L. Cannon, et al. 2000. Establishment of Pseudomonas aeruginosa infection: lessons from a versatile opportunist*. Microbes and infection 2(9): 1051-1060. Maple, P., J. Hamilton-Miller, et al. 1989. World-wide antibiotic resistance in methicillin-resistant Staphylococcus aureus. The Lancet 333(8637): 537-540.

35

Martin, P. 1997. Wound healing--aiming for perfect skin regeneration. Science 276(5309): 75. Massines, F., N. Gherardi, et al. 2005. Glow and Townsend dielectric barrier discharge in various atmosphere. Plasma physics and controlled fusion 47: B577. Matthews, P. and P. Stewart 1984. Resistance heterogeneity in methicillin-resistant Staphylococcus aureus. FEMS microbiology letters 22(3): 161-166. Moisan, M., J. Barbeau, et al. 2001. Low-temperature sterilization using gas plasmas: a review of the experiments and an analysis of the inactivation mechanisms. International journal of Pharmaceutics 226(1-2): 1-21. Moreau, E. 2007. Airflow control by non-thermal plasma actuators. Journal of Physics D: Applied Physics 40: 605. Moreau, M., N. Orange, et al. 2008. Non-thermal plasma technologies: New tools for biodecontamination. Biotechnology advances 26(6): 610-617. Mustoe, T. 2004. Understanding chronic wounds: a unifying hypothesis on their pathogenesis and implications for therapy. The American Journal of Surgery 187(5): S65-S70. Napartovich, A. 2001. Overview of atmospheric pressure discharges producing nonthermal plasma. Plasmas and Polymers 6(1): 1-14. Neuhauser, M. M., R. A. Weinstein, et al. 2003. Antibiotic resistance among gramnegative bacilli in US intensive care units. JAMA: the journal of the American Medical Association 289(7): 885. Newton, P., J. Watson, et al. 2004. Macrophages restrain contraction of an in vitro wound healing model. Inflammation 28(4): 207-214. Noble, W. 1999. Ecology and host resistance in relation to skin disease. Fitzpatrick's dermatology in general medicine 1: 184-191. Nosenko, T., T. Shimizu, et al. 2009. Designing plasmas for chronic wound disinfection. New Journal of Physics 11: 115013. Ogata, A., N. Shintani, et al. 2000. Effect of water vapor on benzene decomposition using a nonthermal-discharge plasma reactor. Plasma chemistry and plasma processing 20(4): 453-467. Ovington, L. G. 2002. Overview of matrix metalloprotease modulation and growth factor protection in wound healing. Part 1. Ostomy/wound management 48(6 Suppl): 3. Ovington, L. G. 2003. Bacterial toxins and wound healing. Ostomy Wound Management 49(7; SUPP): 8-12. Parsek, M. R. and P. K. Singh 2003. Bacterial biofilms: an emerging link to disease pathogenesis. Annual Reviews in Microbiology 57(1): 677-701. Paulsson, M., O. D. Liang, et al. 1992. Vitronectin-binding surface proteins of Staphylococcus aureus. Zentralblatt für Bakteriologie: international journal of medical microbiology 277(1): 54. Power, C., J. Wang, et al. 2001. Bacterial wall products induce downregulation of vascular endothelial growth factor receptors on endothelial cells via a CD14dependent mechanism: implications for surgical wound healing. Journal of Surgical Research 101(2): 138-145.

36

Pozzilli, P. and R. Leslie 1994. Infections and diabetes: mechanisms and prospects for prevention. Diabetic medicine 11(10): 935-941. Raad, I. I., J. A. Mohamed, et al. 2011. The prevention of biofilm colonization by multidrug-resistant pathogens that cause ventilator-associated pneumonia with antimicrobial-coated endotracheal tubes. Biomaterials 32(11): 2689-2694. Raahave, D., A. Friis-Moller, et al. 1986. The infective dose of aerobic and anaerobic bacteria in postoperative wound sepsis. Archives of Surgery 121(8): 924. Robson, M. C. 1997. WOUND INFECTION:: A Failure of Wound Healing Caused by an Imbalance of Bacteria. Surgical Clinics of North America 77(3): 637-650. Rogers, H. J. 1954. The rate of formation of hyaluronidase, coagulase and total extracellular protein by strains of Staphylococcus aureus. Microbiology 10(2): 209. Rosenthal, V. D., D. G. Maki, et al. 2010. International nosocomial infection control consortium (INICC) report, data summary for 2003-2008, issued June 2009. American journal of infection control 38(2): 95-104. e102. Rotstein, O. D., T. L. Pruett, et al. 1985. Mechanisms of microbial synergy in polymicrobial surgical infections. Review of Infectious Diseases 7(2): 151. Roy, S., D. Patel, et al. 2007. Transcriptome-wide analysis of blood vessels laser captured from human skin and chronic wound-edge tissue. Proceedings of the National Academy of Sciences 104(36): 14472. Ruckley, C. V. 1997. Socioeconomic impact of chronic venous insufficiency and leg ulcers. Angiology 48(1): 67. Saint, S. and C. E. Chenoweth 2003. Biofilms and catheter-associated urinary tract infections. Infectious disease clinics of North America 17(2): 411. Schaberg, D. R., D. H. Culver, et al. 1991. Major trends in the microbial etiology of nosocomial infection. The American journal of medicine 91(3): S72-S75. Schierle, C. F., M. De la Garza, et al. 2009. Staphylococcal biofilms impair wound healing by delaying reepithelialization in a murine cutaneous wound model. Wound Repair and Regeneration 17(3): 354-359. Schutze, A., J. Y. Jeong, et al. 1998. The atmospheric-pressure plasma jet: A review and comparison to other plasma sources. Plasma Science, IEEE Transactions on 26(6): 1685-1694. Shih, P. C. and C. T. Huang 2002. Effects of quorum-sensing deficiency on Pseudomonas aeruginosa biofilm formation and antibiotic resistance. Journal of Antimicrobial Chemotherapy 49(2): 309. Singh, R., D. Paul, et al. 2006. Biofilms: implications in bioremediation. TRENDS in Microbiology 14(9): 389-397. Smoot, L. M., J. C. Smoot, et al. 2001. Global differential gene expression in response to growth temperature alteration in group A Streptococcus. Proceedings of the National Academy of Sciences 98(18): 10416. Snyder, R. J. 2005. Treatment of nonhealing ulcers with allografts. Clinics in dermatology 23(4): 388-395. Stadelmann, W. K., A. G. Digenis, et al. 1998. Physiology and healing dynamics of chronic cutaneous wounds. The American Journal of Surgery 176(2): 26S-38S.

37

Stewart, P. S. 2002. Mechanisms of antibiotic resistance in bacterial biofilms. International journal of medical microbiology 292(2): 107-113. Stewart, P. S. and J. William Costerton 2001. Antibiotic resistance of bacteria in biofilms. The Lancet 358(9276): 135-138. Sugiarto, A. T. and M. Sato 2001. Pulsed plasma processing of organic compounds in aqueous solution. Thin solid films 386(2): 295-299. Tarnuzzer, R. W. and G. S. Schultz 1996. Biochemical analysis of acute and chronic wound environments. Wound Repair and Regeneration 4(3): 321-325. Thomas, G. W., L. T. Rael, et al. 2009. Mechanisms of delayed wound healing by commonly used antiseptics. The Journal of Trauma 66(1): 82. Vincent, J. L., D. J. Bihari, et al. 1995. The prevalence of nosocomial infection in intensive care units in Europe. JAMA: the journal of the American Medical Association 274(8): 639. Vleugels, M., G. Shama, et al. 2005. Atmospheric plasma inactivation of biofilm-forming bacteria for food safety control. Plasma Science, IEEE Transactions on 33(2): 824-828. Wattel, F., D. Mathieu, et al. 1990. Hyperbaric oxygen therapy in chronic vascular wound management. Angiology 41(1): 59. Westall, F., M. J. de Wit, et al. 2001. Early Archean fossil bacteria and biofilms in hydrothermally-influenced sediments from the Barberton greenstone belt, South Africa. Precambrian Research 106(1-2): 93-116. Wolcott, R. D., D. D. Rhoads, et al. 2008. Biofilms and chronic wound inflammation. Journal of wound care 17(8): 333. Yager, D. R. and B. C. Nwomeh 1999. The proteolytic environment of chronic wounds. Wound Repair and Regeneration 7(6): 433-441. Yanagi, H. and R. J. Sherertz 2010. Infections of Intravascular Catheters and Vascular Devices. Staphylococci in Human Disease: 363-377. Yokoyama, T., M. Kogoma, et al. 1990. The mechanism of the stabilisation of glow plasma at atmospheric pressure. Journal of Physics D: Applied Physics 23: 1125. Young, L. S. 1980. The role of exotoxins in the pathogenesis of Pseudomonas aeruginosa infections. The Journal of Infectious Diseases 142(4): 626-630. Zhu, J. and J. J. Mekalanos 2003. Quorum sensing-dependent biofilms enhance colonization in Vibrio cholerae. Developmental cell 5(4): 647-65

CHAPTER II

Physical Characteristics, Composition, and Germicidal Characteristics of a Novel “Spark-Based” Nonthermal Plasma Generator Discharge

Abstract The application and utility of nonthermal plasma discharges depend on the endogenous gas mixtures used to supply a given device construct. The composition of produced plasma discharges depends on the starting gaseous medium. However, nonthermal plasma discharge also varies considerably based on exogenous factors. The physical characteristics of a novel nonthermal plasma discharge generator were assessed. The plasma generator used here was found to have high levels of reactive oxygen species, particularly nitric oxide, in its discharge spectrum. Ultraviolet radiation was also detected and is considered an additional key constituent associated with this discharge profile. Temperature and pH changes mediated by nonthermal plasma activity were limited. The discharge in this study was found to penetrate through complex media and radiate from a central coronal focus point. The discharge was also found to be highly germicidal and was capable of sterilizing both liquid and plated cultures. Germicidal efficacy was dependent on the liquid environment the cells were exposed in. High nutrient media was

38

39

found to mitigate observed germicidal response to nonthermal plasma; this is a key exogenous factor that affects plasma-mediated activities.

Introduction Plasma medicine is a rapidly growing and multi-disciplinary field of science with far-reaching potential that could lead to future medical treatment options including wound healing therapies (Fridman et al. 2008). The need for alternative methods to combat recalcitrant microorganisms and their byproducts continues to rise since antibiotic resistance and nosocomial infection rates remain important public health issues (Schaberg et al. 1991). Nonthermal plasma-associated technology’s extension into medicine is secondary to further development, optimization, and characterization of new biotechnology constructs. Antibiotic resistance in bacteria has been an ongoing problem for decades. Antibiotic misuse that leads to acquired resistance is a serious public health concern. Nosocomial infections are tremendously important in hospitals and other extended care facilities. Zhan and Miller reported excess mortality of 4% for infections stemming from medical care and 23% of mortality due to post-operative septicemia (Zhan et al. 2003). Antibiotic-resistant bacteria cause the majority of these nosocomial infection-related deaths. In 1989, the national costs of antimicrobial resistance for the United States was estimated to be between $100 million and $300 million annually (Phelps 1989). Acquired resistance to any biocidal agent may occur through genetic mutation, plasmid-mediated transfer, adaptation of specialized efflux pump systems, drug receptor inactivation, and

40

utilization of altered physiological states such as biofilms. Biofilms are difficult to treat with systemic antibiotics because they represent a distinct and separate growth phase in bacteria and differ from the traditional phases of growth that planktonic bacteria undergo. Due to the aforementioned difficulties associated with the use of standard therapies for the removal of antibiotic-resistant bacteria or biofilm-forming pathogens, alternative germicidal therapies could serve a tremendous role in medical settings. Nonthermal plasma has been successfully utilized to kill a variety of microorganisms and the associated germicidal effect is not linked to only one plasma generator subtype. (Laroussi 1996; Birmingham et al. 2000; Laroussi et al. 2003; Laroussi et al. 2004; Gaunt et al. 2006; Burts et al. 2009) However, there are several important aspects when evaluating the germicidal efficacy of nonthermal plasma within individual environments. The localized micro-environments exposed to nonthermal plasma discharge must also be taken under consideration. Bacteria grown in different media containing varied formulations of nutrient macromolecules, such as carbohydrates and sugars, may be shielded from nonthermal plasma discharge due to high densities of the makeup of the nutrient media itself (Laroussi 2002). Likewise, the pH (Locke et al. 2006) and temperature (Ogata et al. 2000) are susceptible to alteration following nonthermal plasma discharge due to the activity of reactive secondary messenger molecules and ions. (Liu et al. 2004) In this investigation, macromolecule concentrations, pH changes, and temperature fluxuations in aqueous bacterial microenvironments were evaluated concurrently with microbial viability after nonthermal plasma discharge. Collectively, this information provides a more comprehensive

41

understanding of the physical properties and capabilities of this specific plasma generator subtype. The specificity of plasmas depend upon gas compositions, pressure, air flow, frequency of the applied voltage, construction of the cathode and anode configuration, and sustainment of the electric field. (Conrads et al. 2000; Fridman et al. 2008). Nonthermal plasmas can be produced by various discharge methods (Moisan et al. 2001; Napartovich 2001; Akishev et al. 2002; Kim 2004). These plasmas possess very low energy cost which make them optimal tools for biological and biochemical experimentation in both in vitro and in vivo models. The excitatory nature of the charged electrons and plasma molecules are responsible for initializing biochemical chain reactions. For many types of discharges, the gas kinetic temperature is equivalent to room temperature values. Low gas temperature is achieved by limiting the time the gas spends in the electrification process (Napartovich 2001). The composition of plasmas generated with different starting mediums can be radically different due to electron amplification and subsequent creation of downstream molecules and free radicals. (Ogata et al. 2000). Some plasma generators rely on noble gasses such as argon or xenon whereas others function on ambient air made up primarily of nitrogen and oxygen. Gliding arc discharges supplied with water vapor as its initial gaseous medium display higher levels of hydroxyl radicals (Burlica et al. 2006). Changes in levels of oxygen content within in the starting plasma gaseous medium have been shown to affect the amount of produced nitrogen oxide. Zhao et al. described a critical

42

oxygen concentration needed to produce nitrogen dioxide gas instead of nitric oxide.(Zhao et al. 2005) Other investigations have used specific ratios of nitrogen and oxygen mixtures in their starting gaseous environment which led to the intentional production of nitric oxide. (Namihira et al. 2002) This concept was expanded upon and results in to the proposal of a nitric oxide generator designed for therapeutic purposes (Sakai et al. 2008). Determination of the physical parameters and composition of any generated plasma discharge is essential given the variety of ways nonthermal plasma can be generated. Here, we report the usage of a novel pulse-based nonthermal plasma generator that possesses significant germicidal capabilities. Characterization of the predominant plasma constituents is necessary in order to postulate mechanisms associated with observed results with scientific accuracy. The composition of the nonthermal plasma discharge from a novel pulse-based plasma generator was investigated in this study. Tests were designed to detect and quantify the following: atomic oxygen, ultraviolet radiation, ozone, visible light radiation, x-ray radiation, reactive oxygen species, and vaporized nozzle materials. In addition, the following physical operating parameters and device characteristics were evaluated: plasma discharge rate, pH fluctuation, temperature of the plasma discharge following extended usage, and efficacy on cellular activity in different liquid environments. The observed germicidal efficacy was also evaluated before and after the removal of several major plasma components within the nonthermal plasma discharge profile. Several authors have evaluated germicidal activities stemming from nonthermal plasma

43

and presented potential mechanisms (Lerouge et al. 2001; Laroussi et al. 2003; Moreau et al. 2008). However, the results generated from this novel generator aim to provide a new perspective and further elucidate the overall mechanism of plasma-microbe and plasmaenvironment interactions.

Methods

Physical Description of the Plasma Generator (GPP) The GPP generator is composed of three main components: a control module, plasma wand, and plasma nozzle. The gaseous medium is introduced into the control module through an air flow inlet (~1/8 of an inch in diameter). The gas is then passed to a connecting wand (Figure 7) composed of additional tubing. Electrical voltage is supplied to generate an electrical field within the device that ionizes the gaseous mixture. After the charged plasma molecules are transported downstream to the plasma nozzle, they pass between a discharge gap composed of two electrodes within the nozzle. The voltage creates a discharge between the electrodes which ionizes the gas in this localized environment. As discharge events continue inside of the discharge gap, the generated plasma is expelled through the orifice of the nozzle.

44

Figure 7. Schematic of the pulsed nonthermal plasma discharge generator

45

Plasma Spectra Characteristics To determine the makeup of ultraviolet (UV) radiation found in the plasma discharge, a TKA-PKM UV-radiometer (TKA, St. Petersburg, Russia) was utilized (Figure 8). The sensor was fixed 1 cm crosswise and lengthwise direction with respect to the discharge corona. Experimental data were collected continuously over 1 minute at a frequency discharge of 0.25 Hz. An AvaSpec-2048FT spectrophotometer (Avantes, Broomfield, CO) was used to measure spectral information from the nonthermal plasma discharge arc (Figure 9). The plasma discharge frequency was set to 0.25 Hz, and the spectrophotometer was placed 1 cm from the corona of the plasma discharge. An optical resolution of 0.3 nm was selected during subsequent and continuous analysis.

Plasma Composition Measurements Kapton H polyimide slices (5 mm diameter, 0.127 mm thickness; Dupont, Fayetteville, NC) were used to measure atomic oxygen content. The plasma discharge was fixed at a constant distance of 5 mm above the surface of the polyimide slices and fired continuously at exposure times that ranged from 15 seconds to 2 minutes. Images and measurements of induced voids were collected for mass analysis. Non-purified, room temperature air was introduced into the control module of the plasma generator. After electrification, the ionized molecules were discharged through the tip of the plasma wand into a sealed container. A nitric oxide detector (RAE Systems, San Jose, CA) was placed in close proximity (2 mm) to the pulsed plasma. In order to

46

UV radiometer

Focusing lens Plasma discharge

Optical fiber

Spectrometer

Plasma generator

Plasma power source

Figure 8. Illustration of spectrophotometry and radiometry setup and analyses

47

1

intensity, relative units.

0.8

0.6

0.4

0.2

0 250

350

450

550

650

750 wavelength, nm

Figure 9. Spectrophotometric nonthermal plasma discharge profile generated from room temperature ambient air gas mixture

48

reduce any electrical interference, a grounded copper Faraday cage was constructed and surrounded the monitor. Results were collected over ten minutes of continuous plasma discharge. Ozone within the plasma discharge was measured by confining the plasma in a hermetically sealed chamber. An ozone detector (EcoSensors, Newark, CA) was used to obtain readings over time by positioning the sensor approximately 5 inches from the plasma nozzle fixed inside the sealed chamber. To determine if the initiation voltage of the plasma was sufficient to produce soft X-rays, dental film (Kodak, Rochester, NY) covered with a perforated molybdenum sheet was used to develop X-rays through contrasting images. The film was exposed to the plasma for durations ranging from 2 minutes to 10 minutes at a fixed distance of 2 mm from the surface of the film.

pH Measurement An Accumet model pH meter (Fisher Scientific, Pittsburgh, PA) equipped with a Symphony probe (VWR, Radnor, PA) was used to measure changes in pH in prepared bacterial suspensions following plasma discharge. Prepared S. aureus suspensions were transferred to Costar 24-well plates (0.8 cm depth and 1.75 cm diameter; Corning Inc., Corning, NY) at a total depth of 5 mm. The tip of the plasma discharge was fixed 3 mm above the surface of the bacterial suspension. The plasma generator was operated for up

49

to ten minutes for each sample. Each sample was analyzed in real time using the pH probe during and after nonthermal plasma discharge. Temperature Measurement A RMS multimeter (Fluke, Everett, WA) was used to measure temperature increases in prepared bacterial suspensions in real time during plasma discharge. Prepared S. aureus suspensions were transferred to Costar 24-well plates (0.8 cm depth and 1.75 cm diameter; Corning Inc., Corning, NY) at a total depth of 5 mm. The probe adjusted to room temperature (23 °C) prior to experimentation. The tip of the plasma discharge was fixed 3 mm above the surface of the bacterial suspension. The plasma generator was operated for 10 minutes continuously with temperature values recorded at 30 second intervals.

Nutrient Media Variation (Macromolecular Concentrations) Overnight S. aureus cultures were obtained and re-suspended in: sterile distilled water, 1M phosphate buffered saline (PBS; pH 7.4), tryptic soy broth (TSB), Dulbecco’s Modified Eagle Medium (DMEM; pH 7.4; Invitrogen, Carlsbad, CA) enriched with 5% fetal bovine serum (Cleveland Clinic, Cleveland, OH), or Brain Heart Infusion Broth (BHIB) to provide a starting bacterial concentration of 3.0 x107 CFU/mL. Each culture was exposed to nonthermal plasma for lengths of time ranging from 0.5 minutes to 10 minutes at a fixed distance of 3 mm above the surface of the liquid. All cultures were then plated on nutrient agar plates and incubated overnight at 37oC. Plate counts were

50

collected the following day enabling survivability calculations based on cell culture densities. Bacterial Cultures The following bacterial strains were used in these studies: S. aureus ATCC 12598, P. aeruginosa ATCC 29260, Klebsiella pneumoniae ATCC 8045, Escherichia coli ATCC 25297, B. subtilis ATCC 465, and B. atrophaeus ATCC 6455. One hundred microliters of freezer stock culture for each bacterium was diluted with 14.9 mL of Brain Heart Infusion Broth (6 g Brain, Heart infusion from solids, 6 g Peptic digest of Animal Tissue, 5 g Sodium Chloride, 3 g Dextrose, 14.5 g Pancreatic digest of Gelatin, and 2.5 g Disodium Phosphate per liter of deionized water, pH 7.4) obtained from BD Biosciences, Sparks, MD. Following overnight incubation at 37°C, 1 mL aliquots for each bacterial species were taken from each suspension, placed into UV-transparent quartz cuvettes, and measured using a Genesys 10uV Thermospectronic spectrophotometer (Thermo Scientific, Pittsburgh, PA) at an optical density of 600 nm (OD600). The reading was in tandem with an established (average of five experiments) standard curve for the microorganism of interest which provided a starting concentration for each organism. For most experiments, the starting concentration of each respective bacterium was 3.0 x 106 CFU/mL unless otherwise noted.

Depth Experiments

51

Bacterial suspensions were prepared and then added into fresh, sterile Tryptic Soy Agar (15 g Pancreatic Digest of Casein, 5 g Digest of Soybean, 5 g Sodium Chloride, and 15 g Agar per liter of deionized water, pH 7.4, 56°C) obtained from BD Biosciences, Franklin Lakes, NJ. The ratio of inoculated bacterial culture to agar per mL was 1:50 unless otherwise noted. Prepared, inoculated agar was mixed thoroughly and poured into plates in 20 mL aliquots. This provided a depth of agar of: 0.5 cm. Other amounts of the inoculated agar were poured as indicated to yield different depths. The nonthermal plasma device was fixed 3 mm from the surface of each plate. Treatment with the cold plasma was marked on the bottom of each plate, indicating an individual exposure focus. Exposure doses ranged from 15 seconds to 3 minutes. Following exposure, all cultures were incubated for approximately 24 hours at 37°C and then photographed from the top and bottom of each plate using a BioMic plate reader (Giles Scientific, Santa Barbara, CA). A sterile scalpel was used to core around the perimeter of the treatment zones in the agar. These pieces were placed on a transparent glass surface and then analyzed for turbidity or cloudiness through the agar cores.

Diameter Experiments Overnight bacterial cultures (prepared as described above) were spread across the entire surface of TSA plates in 150 uL aliquots. Individual exposure zones were marked on the bottom of the plates to indicate where portions of the plate had been exposed to nonthermal plasma. Following exposure, the plates were placed in an incubator for 24

52

hours at 37°C. The plates were then removed and imaged. Following image acquisition, zones of inhibition were observed and then measured.

Colony Counts and Viability Sterile, deionized water was used to dilute bacterial stock cultures to experimental starting concentrations. One mL aliquots of a prepared bacterial suspension were placed in the wells of 24-well plastic plates with depths and a diameter of 0.8 cm and 1.75 cm respectively (Corning Inc., Corning, NY). The probe of the nonthermal plasma device was fixed 3 mm from the direct surface of the air-liquid interface. Exposure ranged from 15 seconds to 5 minutes. After plasma exposure, each individual sample well was then diluted sequentially at concentrations ranging from 10-4 to10-8 of exposed cultures. One hundred uL aliquots from each dilution group was then plated evenly across freshly prepared TSA plates. All plates incubated for 24 hours at 37°C. Original cell density (OCD) values were calculated for each sample using counted.

Reactive Oxygen and Nitrogen Species Filtration Magnesium fluoride windows with a 0.2 mm thickness (Edmund Optics, Barrington, NJ) were selected to transmit UV radiation and simultaneously block reactive oxygen and nitrogenous species. MgFl2 windows were placed on top of Costar 24-well plates (0.8 cm depth and 1.75 cm diameter; Corning Inc., Corning, NY). The plasma discharge probe was fixed 2 mm above the window, resulting in a total exposure gap of 5

53

mm from the surface of the liquid. Bacterial suspensions were then exposed to the nonthermal plasma at l intervals that ranged from 30 seconds to 4 minutes. Following exposure, resulting cultures were serially diluted and plated on nutrient agar. Bacterial colonies were counted after 18 hour incubations at 37oC.

Ultraviolet Radiation Filtration Nonthermal plasma filtration was repeated using sterile, “Breathe Easy film,” a 1 mm-thick non-toxic, polyurethane material (USA Scientific, Ocala, FL) painted black to absorb UV-radiation from the plasma discharge. The material was designed to remain porous, allowing oxygen and its reactive derivatives penetration through to cultures. S. aureus cultures were sealed with 24-well Costar plates using the film material. The plasma discharge probe was fixed 3 mm above the surface yielding a total discharge gap of 5 mm. Bacterial suspensions were then exposed to the nonthermal plasma at intervals that ranged from 30 seconds to 4 minutes. Following exposure, resulting cultures were serially diluted and plated on nutrient agar. Bacterial colonies were counted after 18 hour incubations at 37oC. Viability was determined using original cell density calculations of remaining survivors that grew on collected plates.

Results The physical dimensions and construction of the device are described as the following. The air flow rate through the nozzle was measured to be 10.6 cm3 per sec.

54

The input voltage required to generate the plasma was 220 volts, which corresponded to a plasma pulse rate of approximately 4 pulses per second. The diameter of the plasma discharge ranged from 3 to 5 mm from the central exposure point after continuous discharge ranged from 1 minute to 5 minutes. The plasma discharge extended over 1 cm out from the orifice of the nozzle as determined visually and confirmed utilizing photography and spectroscopy. The interaction between the positive nickel-chromium alloy center electrode and negative outer electrode formed a warm cathode spot on the brass outer electrode at continually varying places. These spots produce vaporized atoms and colloidal electrode particles that are deposited on surfaces downstream of the plasma discharge arc. Manipulation or optimization of the electrode materials could be a beneficial niche to exploit further. When air is used as generation matter, the discharge spectra are characterized by intense lines of oxygen and associated ionized derivatives. A peak at 257 nm corresponded to O+ ions and had a relative concentration of 0.38 units. Another peak at 281 nm corresponded to a high concentration of OH- with a relative concentration of 0.65 units. The last significant peak below 300 nm corresponded to O2+ and displayed a relative concentration of 0.77 units. The system of bands in the spectral range from 310 to 437 nm can be assigned to molecular oxygen. Low intensity levels of ozone were detected in the 310-340 nm spectral range. The light emission spectroscopy data displayed lines around 460 and 700 nm and can be identified as ionized copper-derived electrode materials (Figure 9).

55

The relative power densities of the three types of UV radiation were also determined using air as the plasma generation matter. Power densities of UV-A (long wave) and UV-B (medium wave) each displayed 180 mW/cm2 while UV-C displayed 330 mW/cm2. There is significantly more short wave UV radiation compared with the other two subtypes. Oxygen-mediated mass reduction in Kapton H polyimide slices was used to measure atomic oxygen content of the plasma discharge (Cooper et al. 2008). Erosion rings measured approximately 8 mm in diameter after 10 minutes of continuous plasma exposure. In addition, thr mass lost corresponded to an atomic flux of 1.15 x 107 atoms per cm2 per second (Figure 10). The level of detected nitric oxide (NO) was 18 parts per million (ppm) after one minute of continuous plasma discharge. Doubling the exposure time to two minutes accumulated 40 ppm. Four minutes of plasma discharge increased the NO concentration to 114 ppm, and eight minutes generated approximately 208 ppm. A steady value of 220 ppm NO was achieved following ten minutes of plasma exposure (Figure 11A). One minute of continuous discharge generated 0.03 ppm of ozone. Doubling the discharge exposure time resulted in a further increase to 0.07 ppm. At additional two minute intervals, the ozone concentration increased to 0.175 at four minutes, 0.265 at six minutes, 0.367 at eight minutes, and 0.44 ppm after ten minutes of continuous plasma discharge (Figure 11B). The pH of low nutrient aqueous media such as deionized water and 1x phosphate buffered saline (PBS) decreased whereas higher nutrient media sources increased

56

Figure 10. Detection of molecular oxygen from nonthermal plasma discharge in: (A) Candle soot streaks and (B) Kapton H polyimide wafer disks

57

Figure 11. Detection of nitric oxide (A) and ozone (B) from a sealed chamber exposed to continuous nonthermal plasma discharge

58

slightly. The pH of deionized water decreased from 7.04 to 6.83 after one minute of plasma exposure, and decreased further to values of 6.489, 5.88, and 5.18 at exposure intervals of two, five, and ten minutes respectively. The pH of brain heart infusion broth and Dulbecco’s Modified Eagle Medium rose from 7.38 and 7.73 to 7.43 and 7.9 following one minute of exposure. The pHs of these liquids increased further to 7.81 and 8.19 after ten minutes of continuous nonthermal plasma exposure respectively (Figure 12). A localized area exposed to one minute of continuous nonthermal plasma discharge increased in temperature by 1.7°C. Two minutes increased the temperature of the localized area by 2.9°C. At subsequent intervals of four, six, and eight minutes, the temperature rose 3.9°C, 4.6°C, and 5.83°C respectively. Ten minutes of continuous plasma generation increased the local environment’s temperature from 23°C to approximately 31°C (Figure 13). Thirty seconds of nonthermal plasma exposure penetrated through 0.133 inches of P. aeruginosa-embedded agar, equivalent to over 50% penetration of the total depth. When the exposure was increased to 1 minute, penetration through the first group increased to 0.186 inches, or approximately 74% total penetration. Two minutes of exposure increased the total penetration to 0.221 inches, or 88%. Four minutes of continuous exposure was sufficient to penetrate completely through 2.5 inches of inoculated agar. Nonthermal plasma exposure at 30 seconds, 1 minute, 2 minutes, and 4 minutes in the 0.5 inch agar model group generated penetration depths of 0.245 inches, 0.348 inches, 0.406, and 0.5 inches respectively. Likewise, in the 1.5 inches group,

59

pH of Suspension

9

DI Water 1x PBS TSB BHIB DMEM

8 7 6 5 4 0

1

2

3

4

5

6

7

8

9

10

Plasma Exposure (Minutes)

Figure 12. pH of different media suspensions exposed to nonthermal plasma

60

Temperature (°C)

32.5 30.0 27.5 25.0 22.5 20.0 0.0

2.5

5.0

7.5

10.0

12.5

Exposure Time (Minutes)

Figure 13. Temperature of localized plasma exposure zone exposed to continuous nonthermal plasma discharge

61

exposure doses of 30 seconds, 1 minute, 2 minutes, and 4 minutes generated penetration depths of 0.621 inches, 0.902 inches, 1.124 inches, and 1.5 inches (Figure 14). Similar results were observed in the embedded S. aureus agar plates. In the 0.25 agar model group, 30 seconds of plasma exposure generated penetration of 0.139 inches, or 55.5% of the total depth. When the dose was increased to 1 minute, 0.191 inches or 76% of the total depth was penetrated. Two minutes increased penetration to 0.222 inches, equivalent to approximately 89% of the total depth. In the 0.5 inch agar model experimental group, nonthermal plasma exposure of 30 seconds, 1 minute, 2 minutes, and 4 minutes generated penetration depths of 0.251 inches, 0.351 inches, 0.424 inches, and 0.5 inches respectively. In the 1.5 inch experimental agar model group, nonthermal plasma exposure at 30 seconds, 1 minute, 2 minutes, and 4 minutes generated penetration depths of 0.732 inches, 0.971 inches, 1.235 inches, and 1.5 inches respectively. Complete penetration was observed with 4 minutes of continuous exposure. A positive correlation between plasma exposure and total agar penetration was observed in all depth groups in both bacterial species. Overall nonthermal plasma-mediated penetration was reduced as the depth and overall amount of inoculated agar increased. However, this effect was not deemed significantly detrimental to penetration or lethality of the plasma discharge in most samples. Nonthermal plasma discharge-induced damage voids in S. aureus, P. aeruginosa, S. pyogenes, K. pneumonia, and E. coli, were measured after exposure. Fifteen seconds of exposure generated average void sizes of 0.642 centimeters, 0.736 centimeters, 0.788

62

Figure 14. Nonthermal plasma discharge penetration through uniform, inoculated agar matrices in both (A) P. aeruginosa and (B) S. aureus models

63

centimeters, 0.669 centimeters, and 0.531 centimeters in S. aureus, P. aeruginosa, S. pyogenes, K. pneumonia, and E. coli cultures respectively. Thirty seconds of exposure increased the average void sizes to 1.691 centimeters, 2.054 centimeters, 2.144 centimeters, 1.498 centimeters, 1.498 centimeters, and 1.242 centimeters. One minute of plasma exposure generated zones with diameters ranging in size from 2.471 centimeters observed in S. aureus plates to 2.218 centimeters in E. coli plates. As exposure increased to two minutes, inhibition zones also increased with measured values of 3.475 centimeters, 3.792 centimeters, 3.681 centimeters, 3.331 centimeters, and 3.055 centimeters in S. aureus, P. aeruginosa, S. pyogenes, K. pneumonia, and E. coli cultures respectively. Some of the measured zones surpassed 4 centimeters in total diameter. Although damage void sizes varied amongst the different bacterial species, a direct positive correlation between increased plasma exposure and increased void size diameter was observed (Figure 15). Fifteen seconds of continuous nonthermal plasma exposure reduced the logarithmic count of viable P. aeruginosa cultures by 0.24 logs. Log reduction increased to 1.24 once the exposure length was increased to 30 seconds of continuous exposure. A 1 minute exposure increased log reduction of viable bacteria by an additional 2.45 logs. Two minutes of exposure, reduced viable culture by 6.33 logs. Complete sterilization of cultures with an average starting logarithmic concentration of 7.42 occurred after 4 minutes of continuous plasma exposure in P. aeruginosa cultures. Similar log reductions in viability were noted in the S. aureus cultures. Fifteen seconds of plasma exposure reduced S. aureus cultures by 0.26 logs. When the exposure interval was increased to 30

64

Figure 15. Diameters of germicidal zones from a fixed exposure focal point in Grampositive and Gram-negative bacteria

65

seconds and 1 minute, the log reduction of viable organisms increased to 0.796 and 2.26 logs respectively. Nonthermal plasma exposure intervals of 2 minutes and 4 minutes generated log reductions of 6.08 logs and 7.56 logs respectively. Four minutes of exposure completely sterilized all S. aureus cultures (Figure 16). Unfiltered nonthermal plasma discharge reduced S. aureus cultures by 0.55 logs, or a 7.4% overall reduction, after 30 seconds of exposure. As plasma exposure increased to 1 and 2 minutes, bacterial logarithmic values decreased by 1.37 and 2.45 logs, equivalent to 18.3% and 32.8% overall reduction compared to negative control values. Four minutes of nonthermal plasma discharge was found to sterilize all cultures. UVfiltration reduced logarithmic survivability by only 5.7%, 7.8%, 15.4%, and 27.5% after 30 seconds, 1 minute, 2 minutes, and 4 minutes of plasma exposure. Reactive oxygen species (ROS) filtration reduced overall viability by only 1.62%, 4.41%, 5.9%, and 9.1% after 30 seconds, 1 minute, 2 minutes, and 4 minutes of plasma exposure. Plasma discharges that were selectively filtered for both ROS and UV radiation displayed almost no germicidal capability when compared to other experimental groups. Logarithmic reduction percentages equal to 0.61%, 1.02%, 1.5%, and 2.6% were observed after 30 seconds, 1 minute, 2 minutes, and 4 minutes of plasma exposure (Figure 16). Surviving S. aureus culture log scale values after 2 minutes of continuous nonthermal plasma discharge were normalized against negative control, non-exposed culture values: 98% (DMEM), 97% (BHIB), 96% (TSB), 81% (PBS), and 76% (water). As plasma exposure gradually increased to 5 minutes, logarithmic values changed to: 97% (DMEM), 95% (BHIB), and 94% (TSB).The PBS and water groups were completely

66

sterilized at 5 minutes. At 10 minutes of exposure, the surviving percentages decreased further in the heavier nutrient groups to: 94% (DMEM), 91% (BHIB), and 88% (TSB) (Figure 17).

Surviving Log10 S. aureus (CFU/mL)

67

8

ROS + UV Filtered ROS Filtered UV Filtered Unfiltered

7

6

5 2 1 0 0

1

2

3

4

5

Plasma Exposure (Minutes)

Figure 16. Survivability of S. aureus cultures exposed to nonthermal plasma at different levels of filtration

Surviving Log10 S. aureus ( CFU/mL)

68

10

BHIB DMEM TSB 1x PBS DI Water

8 6 4 2 0 0.0

0.5

1.0

2.0

5.0

10.0

Plasma Exposure (Minutes)

Figure 17. Survivability of S. aureus cultures exposed to nonthermal plasma within different starting aqueous media

69

Discussion A reasonably high atomic oxygen flux was observed during characterization and measurement of the GPP plasma’s main constituents. Atomic oxygen has been found to be an important component associated with plasma sterilization.(Moreira et al. 2004; CHOI et al. 2006; Hayashi et al. 2006). Its presence in the plasma discharge of the studied device is encouraging for future sterilization applications. Spectrophotometric analyses indicated the plasma discharge associated with this generator produces elevated amounts of O+, OH-, and O2+. Nitric oxide (NO), another main component found within this plasma discharge, is known to possess antimicrobial properties and also act as an important signaling molecule in a variety of systems including wound healing (Nathan et al. 1991). The amount of NO observed from a localized area from the plasma discharge is significantly elevated and merits additional investigation and application. NO is a highly reactive, volatile molecule. High concentrations of this reactive species are the result of elevated nitrogen species present in the starting gaseous medium fed into the plasma electrode. The production of nitric oxide is due to reactive ionization events caused during gaseous electrification prior to expulsion from the plasma generator. The significance of the concentration of NO observed in this study is apparent when compared to the OSHA Time Weighted Average of 25 parts per million (ppm) associated with eight hours of transient exposure (Phillips et al. 1999). All three types of ultraviolet radiation were detected and measured using radiometry. Although UV-A and UV-B have been linked to ultraviolet-radiation-induced damage in sporulating bacteria (Slieman et al. 2000), these wavelengths were detected at

70

lesser intensity than short wave UV-C radiation during spectrophotometric analyses of studied plasma discharges. High energy UV-C is likely more directly associated with germicide and DNA damage observed in exposed bacterial cultures (Kielbassa et al. 1997; Friedberg 2003). While the UV components detected within the plasma discharge spectral profile are independently antimicrobial, they may function collectively in tandem. Ozone was detected spectrophotometrically in the discharge associated with this plasma discharge. However, the measured concentration of ozone following extensive continuous nonthermal plasma discharge (>10 minutes) is less than half of the ozone threshold limit value concentration of 1 ppm per eight hours established by the American Conference of Governmental Industrial Hygienists of exposure and the Occupational Safety and Health Administration (Bergamini et al. 2004). Therefore, the ozone produced by the studied plasma is not sufficiently concentrated enough to have significant impact on any germicidal events upon biological experimentation. Nonthermal plasma-mediated temperature increases were not found to be significant. An 8°C increase from the starting room temperature value was observed after ten minutes of continuous discharge. Since S. aureus and P. aeruginosa are mesophilic bacteria capable of tolerating temperatures up to 44°C, plasma-mediated temperature increases to 31°C will have no effect on their bacterial physiology. The plasma discharge slightly altered the pH of various high nutrient media. Increases in the high nutrient media pH values were less than 1 log difference and not deemed significant. Likewise, the pH decrease associated with low nutrient media, deionized water and 1x PBS, while

71

greater than the change in the higher nutrient media, were not deemed significant. The changes in pH are not outside the threshold for normal pH tolerance for S. aureus which is typically found between pH 4 and 9 (Sutherland et al. 1994; Booth 1999). The heightened change in aqueous environments devoid of nutrient sources such as proteins and sugars may support the notion that the reactive species and ultraviolet radiation within nonthermal plasma discharges react with macromolecules in micro-environments alongside the intended, targets cells. Extensive testing with the penetrative properties of this plasma was performed in part to determine the effectiveness of using nonthermal plasma as a tool used to treat bacterial infections complicated by cryptic planktonic bacteria or biofilms. Agar was chosen to serve as a semi-solid matrix to embed bacteria uniformly throughout a fixed area. Since agar is also a polysaccharide polymer, it also provides a three-dimension barrier for the microorganisms embedded within. The agar medium served as a simple way to assess nonthermal plasma penetrative efficacy through biological polymers. Since the surrounding medium plays a role in how the plasma is able to interact with target cells, penetrative efficacy must be considered. Three different volumes of agar were chosen to evaluate plasma response to varied depth and total volume of biopolymerbacteria models. Although the plasma discharge was able to penetrate through all three agar model groups, the total penetration was reduced as the depth of starting agar concentrations increased. The penetrative nature of the plasma discharge is encouraging and allows for more advanced model development to occur. The limits of plasma

72

penetration provided additional capability information with respect to this plasma generator subtype. Agar penetration studies show that thirty seconds of nonthermal plasma exposure is able to penetrate through 0.133 inches of P. aeruginosa-embedded agar and 0.139 inches of S. aureus-embedded agar. Increased exposure lengths directly correspond to greater penetration through bacterial embedded agar hydrogel models. Four minutes of continuous plasma exposure penetrates completely through the largest volume of agar while simultaneously sterilizing uniform columns through these inoculated matrices. The dispersal of reactive oxygen species, charged ions, molecules, and secondary plasmaassociated molecules was observed using bacteria spread evenly on plates. As little as 15 seconds of exposure generated inhibition zones where bacteria had been killed and etched away from the surface. Fifteen seconds of exposure generated average void sizes of 0.642 centimeters, 0.736 centimeters, 0.788 centimeters, 0.669 centimeters, and 0.531 centimeters in each of the respective bacterial species plated. When the exposure increased to 2 minutes, the zone sizes increased to values of 3.475 centimeters, 3.792 centimeters, 3.681 centimeters, 3.331 centimeters, and 3.055 centimeters in each of the respective bacterial species. The other main physical parameter of experimental interest was the diameter of plasma-induced exposure zones. These voids served as indicators of the distance that plasma discharges radiate from a central fixed point. Void zones were observed in a variety of pathogenic bacteria, including P. aeruginosa and S. aureus (Figure 9). In

73

many of the experiments, a central void was accompanied by a surrounding corona that displayed a fringe germicidal effect. These coronas served as the boundary limits for all void measurements. Evidence of a biphasic, coronal effect suggests that the plasma has a direct, immediate sterilization effect on the localized area where the discharge is fixed, and also a secondary lesser germicidal effect at the periphery of its displayed range. There is a direct positive correlation between larger void zones and elevated nonthermal plasma usage. This expansion is due to an increase in the germicidal effect at the periphery of the growing zones due to continued release of germicidal free radicals and other plasma-associated ions. Because the chief components of the plasma discharge disperse rapidly into the atmosphere or interact with its local environment, cessation of the plasma discharge after shorter exposure lengths could explain reduced size of killing zone voids. Log reduction of viable bacteria following plasma exposure is a core confirmation of previous plasma germicidal capability in other plasma subtypes. It is also a fundamental observation within this study as a whole. Ultraviolet radiation, atomic oxygen, and reactive oxygen species all possess antimicrobial properties (Bandyopadhyay et al. 1999; Friedberg 2003; Bergamini et al. 2004). The nonthermal plasma discharge profile was found to possess each of these components, suggesting significant germicidal potential in this generator. Averages of germicidal values were obtained following multiple independent experiments and these results reveal increased nonthermal plasma exposure results in a direct negative correlation in planktonic bacteria assessed in vitro.

74

As little as fifteen seconds of nonthermal plasma discharge was sufficient to reduce bacterial viability in both Gram-positive and Gram-negative bacteria. Exposure up to two minutes of continuous nonthermal plasma exposure reduced bacterial viability by over six logs in cultures with starting bacterial concentrations of 3.0 x107 CFU/mL. Complete sterilization was observed between three and four minutes of plasma exposure. Significant sporicidal results were also collected from experiments using continuous plasma exposure. The length of time needed to reduce spore germination was much greater than the planktonic bacteria. Five minutes began to reduce viability and it was not until after ten minutes of exposure until more dramatic results were achieved. Complete sterilization of spores occurred around fifteen minutes of nonthermal plasma exposure (Data not shown). Collectively, these data provide a more complete profile for not only the physical parameters of the plasma discharge, but also the potential germicidal capabilities of this novel plasma generator. Reactive oxygen and nitrogen species have been implicated in the attack and destabilization of membranes in bacteria (Bandyopadhyay et al. 1999; Liu et al. 2004). A membrane attack event can be viewed as a short term germicidal effect because the induced damage would be a highly dynamic and amplifying sequence of events. The UV radiation profile of this nonthermal plasma is predominantly UV-C. Short wave ultraviolet radiation has been linked to DNA damage for decades. It is certainly probable that the UV radiation in the nonthermal plasma discharge is able to penetrate into the bacteria, especially as integrity of cellular membranes is weakened over time and then lost entirely. The accumulation of DNA damage ultimately results in genomic instability

75

which ultimately leads to necrotic lysis of the cell. Genomic damage in bacterial cells can be thought of as a long term killing event that contrasts the immediate membranedamaging attack on bacterial cell membranes. The data suggest that when reactive oxygen and nitrogen species are removed from the plasma discharge profiles, the time required to damage bacterial cells increases dramatically. This delay increases further in cultures exposed to ROS-filtered nonthermal plasma. Removal of the reactive species from the nonthermal plasma discharge profile eliminates the rapid membrane destabilizing mechanism of action associated with high ROS/RNS-associated membrane catastrophe. The drop off between ROS/RNS-filtered plasma suggests that a significant portion of the germicidal effect likely occurs very rapidly and could support this proposed “short-term” mechanism. The UV component within the nonthermal plasma discharge is still significant even though it independently takes longer to impact significant germicidal capability. Ultraviolet radiation is able to damage the DNA within the bacterial cell. The penetrative nature associated with UV radiation could be advantageous against higher starting concentrations of bacteria since the ROS, RNS, and other plasma ions could dissipate and react rapidly on the initial surface of an exposed area. UV-mediated germicide is a longer term effect that is effective at aiding removal of high density populations of cells. Bacteria possess a class of enzymes called superoxide dismutases to eliminate reactive oxygen species and their derivatives from localized environments. Should ROS attack bacterial membranes, bacteria can use these enzymes to neutralize offending ROS. However, if over time, the UV radiation damages the enzymes that are used to code for

76

this class of enzymes, the bacteria ultimately would become more susceptible to ROS damage. A two-pronged effect linking the two mechanisms together could be postulated in order to explain the synergistic effect of both constituents and their activity within nonthermal plasma discharge. Staphylococcal cultures exposed to the plasma in a variety of liquid media ranging from the highly fortified, nutrient-rich Brain Heart Infusion Broth (BHIB) to nutrient-devoid sterile distilled water indicated that a local environment does play a role in plasma penetration and its ability to affect targeted cells. Because each medium contains different types and ratios of proteins, sugars and vitamins, this study simulated different environmental conditions with respect to bacterial growth macromolecular content. The survivability of S. aureus observed in high nutrient media exposed to unfiltered plasma and low-nutrient media exposed to plasma with reactive species and charged molecules is similar. It remains possible that environments composed of higher amounts of macromolecules inhibit the reactivity of plasma molecules. Reactive species and free radicals are unstable and will react readily with a wide variety of compounds both in vitro and in vivo (Thorpe et al. 2003). Environments with elevated sugar or protein content may shield cells in localized areas from reactive species as the macromolecules could react with plasma-associated molecules and ions before reaching intended target cells. If the reactive and charged molecules are unable to reach the target cells, then some other component must be responsible for the reduction of bacteria. The reduction of viability in liquid cultures when both ROS and UV radiation was removed suggests that there are other components within the plasma discharge that

77

contributes to germicidal efficacy. A possible target could be the ionized colloidal nozzle materials produced during extended, continuous nonthermal plasma discharges. The ionized materials were detected spectrophotometrically and traces were visibly seen on glass slides and in sample solutions. These data suggest that the environment in which cells are placed has a role in the effectiveness of nonthermal plasma exposure. This further supports that the observed germicide in the S. aureus cultures are the result of more than one chief effector and possibly lesser secondary elements.

78

References Akishev, Y., M. Grushin, et al. 2002. Novel AC and DC non-thermal plasma sources for cold surface treatment of polymer films and fabrics at atmospheric pressure. Plasmas and Polymers 7(3): 261-289. Bandyopadhyay, U., D. Das, et al. 1999. Reactive oxygen species: Oxidative damage and pathogenesis. Curr Sci 77(5): 658-666. Bergamini, C. M., S. Gambetti, et al. 2004. Oxygen, reactive oxygen species and tissue damage. Current pharmaceutical design 10(14): 1611-1626. Birmingham, J. G. and D. J. Hammerstrom 2000. Bacterial decontamination using ambient pressure nonthermal discharges. Plasma Science, IEEE Transactions on 28(1): 51-55. Booth, I. R. (1999). The regulation of intracellular pH in bacteria, Wiley Online Library. Borriello, S. P. and P. Honour 1981. Simplified procedure for the routine isolation of Clostridium difficile from faeces. Journal of Clinical Pathology 34(10): 1124. Burlica, R., M. J. Kirkpatrick, et al. 2006. Formation of reactive species in gliding arc discharges with liquid water. Journal of electrostatics 64(1): 35-43. Burts, M. L., I. Alexeff, et al. 2009. Use of atmospheric non-thermal plasma as a disinfectant for objects contaminated with methicillin-resistant< i> Staphylococcus aureus. American journal of infection control 37(9): 729-733. CHOI, J. H., I. HAN, et al. 2006. Analysis of sterilization effect by pulsed dielectric barrier discharge. Journal of electrostatics 64(1): 17-22. Conrads, H. and M. Schmidt 2000. Plasma generation and plasma sources. Plasma Sources Science and Technology 9: 441. Cooper, R., H. P. Upadhyaya, et al. 2008. Protection of polymer from atomic-oxygen erosion using Al2O3 atomic layer deposition coatings. Thin solid films 516(12): 4036-4039. Fridman, G., G. Friedman, et al. 2008. Applied plasma medicine. Plasma Processes and Polymers 5(6): 503-533. Friedberg, E. C. 2003. DNA damage and repair. NATURE-LONDON-: 436-439. Gaunt, L. F., C. B. Beggs, et al. 2006. Bactericidal action of the reactive species produced by gas-discharge nonthermal plasma at atmospheric pressure: A review. Plasma Science, IEEE Transactions on 34(4): 1257-1269. Hayashi, N., W. Guan, et al. 2006. Sterilization of medical equipment using radicals produced by oxygen/water vapor RF plasma. Japanese journal of applied physics 45: 8358. Helfinstine, S. L., C. Vargas-Aburto, et al. 2005. Inactivation of Bacillus endospores in envelopes by electron beam irradiation. Applied and Environmental Microbiology 71(11): 7029. Kielbassa, C., L. Roza, et al. 1997. Wavelength dependence of oxidative DNA damage induced by UV and visible light. Carcinogenesis 18(4): 811.

79

Kim, H. H. 2004. Nonthermal Plasma Processing for Air Pollution Control: A Historical Review, Current Issues, and Future Prospects. Plasma Processes and Polymers 1(2): 91-110. Laroussi, M. 1996. Sterilization of contaminated matter with an atmospheric pressure plasma. Plasma Science, IEEE Transactions on 24(3): 1188-1191. Laroussi, M. 2002. Nonthermal decontamination of biological media by atmosphericpressure plasmas: Review, analysis, and prospects. Plasma Science, IEEE Transactions on 30(4): 1409-1415. Laroussi, M. and F. Leipold 2004. Evaluation of the roles of reactive species, heat, and UV radiation in the inactivation of bacterial cells by air plasmas at atmospheric pressure. International Journal of Mass Spectrometry 233(1-3): 81-86. Laroussi, M., D. A. Mendis, et al. 2003. Plasma interaction with microbes. New Journal of Physics 5: 41. Leighton, T. and R. H. Doi 1971. The stability of messenger ribonucleic acid during sporulation in Bacillus subtilis. Journal of Biological Chemistry 246(10): 3189. Lerouge, S., M. R. Wertheimer, et al. 2001. Plasma sterilization: A review of parameters, mechanisms, and limitations. Plasmas and Polymers 6(3): 175-188. Liu, C., N. Cui, et al. 2004. Effects of DBD plasma operating parameters on the polymer surface modification. Surface and Coatings Technology 185(2-3): 311-320. Locke, B., M. Sato, et al. 2006. Electrohydraulic discharge and nonthermal plasma for water treatment. Industrial & engineering chemistry research 45(3): 882-905. Moisan, M., J. Barbeau, et al. 2001. Low-temperature sterilization using gas plasmas: a review of the experiments and an analysis of the inactivation mechanisms. International journal of Pharmaceutics 226(1-2): 1-21. Moreau, M., N. Orange, et al. 2008. Non-thermal plasma technologies: New tools for biodecontamination. Biotechnology advances 26(6): 610-617. Moreira, A. J., R. D. Mansano, et al. 2004. Sterilization by oxygen plasma. Applied surface science 235(1): 151-155. Namihira, T., S. Katsuki, et al. 2002. Production of nitric oxide using a pulsed arc discharge. Plasma Science, IEEE Transactions on 30(5): 1993-1998. Napartovich, A. 2001. Overview of atmospheric pressure discharges producing nonthermal plasma. Plasmas and Polymers 6(1): 1-14. Nathan, C. F. and J. B. Hibbs 1991. Role of nitric oxide synthesis in macrophage antimicrobial activity. Current Opinion in Immunology 3(1): 65-70. Ogata, A., N. Shintani, et al. 2000. Effect of water vapor on benzene decomposition using a nonthermal-discharge plasma reactor. Plasma chemistry and plasma processing 20(4): 453-467. Phelps, C. E. 1989. Bug/drug resistance: sometimes less is more. Medical Care: 194-203. Phillips, M. L., T. A. Hall, et al. 1999. Assessment of medical personnel exposure to nitrogen oxides during inhaled nitric oxide treatment of neonatal and pediatric patients. Pediatrics 104(5): 1095. Rutledge, S. K., B. A. Banks, et al. 2000. Atomic oxygen treatment as a method of recovering smoke-damaged paintings. Journal of the American Institute for Conservation 39(1): 65-74.

80

Sakai, S., M. Matsuda, et al. (2008). Nitric Oxide Generator Based on Pulsed Arc Discharge. Schaberg, D. R., D. H. Culver, et al. 1991. Major trends in the microbial etiology of nosocomial infection. The American journal of medicine 91(3): S72-S75. Slieman, T. A. and W. L. Nicholson 2000. Artificial and solar UV radiation induces strand breaks and cyclobutane pyrimidine dimers in Bacillus subtilis spore DNA. Applied and Environmental Microbiology 66(1): 199-205. Sutherland, J., A. Bayliss, et al. 1994. Predictive modelling of growth of< i> Staphylococcus aureus: the effects of temperature, pH and sodium chloride. International journal of food microbiology 21(3): 217-236. Thorpe, S. and J. Baynes 2003. Maillard reaction products in tissue proteins: new products and new perspectives. Amino Acids 25(3): 275-281. Zhan, C. and M. R. Miller 2003. Excess length of stay, charges, and mortality attributable to medical injuries during hospitalization. JAMA: the journal of the American Medical Association 290(14): 1868. Zhao, G. B., S. Garikipati, et al. 2005. Effect of oxygen on nonthermal plasma reactions of nitrogen oxides in nitrogen. AIChE journal 51(6): 1800-1812.

CHAPTER III

Bacterial Biofilm Physiological and Structural Responses to Nonthermal Plasma

Abstract Bacterial biofilms were constructed in vitro from strains of P. aeruginosa and S. aureus using a modified, sequential bioreactor system. The structure and stability of bacterial biofilms were evaluated over time following exposure to nonthermal plasma discharge. Imaging and mathematical software were used to determine the textural and structural changes as biofilms grew over the course of seven days. Statistical modeling was also performed to assess nonthermal plasma’s ability to affect biofilm development at different periods of time. Several key parameters were ultimately found to be significantly affected by nonthermal plasma discharge whereas others were not affected at all. Changes in the 3D structure of biofilms after nonthermal plasma exposure was not limited to one period of development. The mechanism for this phenomenon is not completely understood but is likely to be a two-prong, synergistic effect due to the composition of the reactive gas species and other plasma-associated molecules previously identified in the nonthermal plasma discharge.

81

82

Introduction Studies with Pseudomonas aeruginosa biofilm formation indicate there is an ordered sequence that occurs as the biofilm develops and matures. The first two stages are generalized by a loose surface association followed gradually by irreversible adhesion on that surface. The third and fourth stages involve growth, division, and further aggregation of bacterial cells as they form into micro-communities. The final stage of biofilm formation is characterized by dispersal of peripheral members that have reverted back to their pre-biofilm, “planktonic-like” physiology (Sauer et al. 2002; Stoodley et al. 2002). The three-dimensional structure of the biofilm is important to consider when evaluating its formation and development over time. The structure can vary depends on local nutrient availability and plays a critical role in how the local micro-colonies develop and interact with other portions of the biofilm (Klausen et al. 2003). Analysis of key components of the three-dimensional biofilm architecture can provide further insight into how both endogenous and exogenous factors can destabilize the overall biofilm structure. Maintenance of biofilm structure as it develops significantly affects the overall community viability and tolerance of external factors. Laboratory investigation of biofilm growth and development requires a consistent, reproducible method in order to study the overall structure of the biofilm. Several methods have been proposed ranging from usage of individual flow cells to specifically constructed continuous culture chambers (de Beer et al. 1994; Ceri et al. 1999; Singh et al. 2002). The method chosen for this investigation is a modification of a continuous

83

culture construct utilized by Jackson et al. (Jackson et al. 2001). Consistent, repeatable results are not easily obtained due to intrinsic variability within individual biofilm aggregates. Biofilms constructed in vitro require replicate images taken from a large number of regions of each individual biofilm. Repetition is needed to avoid making incomplete assessments of limited areas of the overall total biofilm. Confocal microscopy affords users the opportunity to obtain stacks of images collected in three-dimensions, making it a valuable tool for investigations into biofilm architecture (Palmer 1999; Palmer et al. 2006). An established computer program used to integrate, measure, and analyze image information collected from confocal microscopy is paramount to investigations into biofilm structure. Lewandowski et al. developed a series of MATLAB-compatible programs that enable the collection of textural and volumetric parameters. It has been further developed since its creation in the late 1990’s and been successfully used in other investigations into biofilm formation and structure (Lewandowski et al. 1999; Heydorn et al. 2000; Hentzer et al. 2001; Davey et al. 2003; Beyenal et al. 2004; Labbate et al. 2004; Wood et al. 2006; Hansen et al. 2007; Lewandowski et al. 2007). In this study, the stability of biofilm architecture was investigated by utilizing nonthermal plasma as a disruptive stressor on the biofilm. Nonthermal plasma is a partially ionized atmospheric gas composed of reactive ions, UV photons, free radicals, and other charged molecules that possesses sufficient energy to initiate chemical reactions. Plasma is generated when enough energy is supplied to a neutral gas to cause charge production. We hypothesized that nonthermal plasma will alter the overall

84

structure of bacterial biofilms. This could directly impact how local environments survive within the biofilm short term immediately after exposure, and how plasma-mediated disruption of synchronized behavior associated with maintenance or stability of the biofilm could occur over time long term. Damaged biofilms may become further susceptible to secondary, follow-up therapies such as topical antibiotic application. Nonthermal plasma could become an important accessory or compliment application within the realm of biofilm control or removal.

Methods

Bacterial Biofilm Strain and Growth Conditions. Staphylococcus aureus (ATCC 12598; American Type Culture Collection, Manassas, VA) was grown in Brain Heart Infusion Broth (6 g brain, heart infusion from solids, 6 g peptic digest of animal tissue, 5 g sodium chloride, 3 g dextrose, 14.5 g pancreatic digest of gelatin, and 2.5 g disodium phosphate per liter, pH 7.4) obtained from BD Biosciences, Sparks, MD. Pseudomonas aerguinosa (ATCC 29260; American Type Culture Collection, Manassas, VA) was grown in Tryptic Soy Broth (17 g enzymatic digest of casein, 3 g enzymatic digest of soybean meal, 5 g sodium chloride, 2.5 g dipotassium phosphate, 2.5 g dextrose, pH 7.4) obtained from BD Biosciences, Sparks, MD. The TSB media was also enriched with 1% glycerol (v/v) and 1/20 volume of 1 M monosodium glutamate. The bacterial cultures were placed in an incubator for 18 hours at 37°C. Bacteria were adjusted to 3.0 x106 CFU/mL using sterile distilled water immediately prior to use. The prepared

85

suspensions were then aseptically inoculated into a larger volume of nutrient media at a total starting concentration of 2% within biosafety cabinets to prevent contamination.

Biofilm Reactor Construction. Biofilms were grown on 40 mm x 24 mm glass coverslips (Fisher Scientific, Pittsburgh, PA) fixed to the bottom of modified, autoclavable plastic boxes (VWR, Radnor, PA) and a water-resistant, non-toxic sealant was applied to prevent coverslip floatation or dislocation from the bottom of the reactors. These mini-“reactors” had the following dimensions: 11.5 cm length, 8 cm width, and 2.5 cm depth (discounting the lid) with a total working volume of 260 mL. Five mm holes were drilled into the two opposing faces of the reactors and were sealed with autoclavable inflow and outflow tubing lines (Tygon ® PharMed ® autoclavable, U.S. Plastic Corp., Lima, OH). The tubing had an inner dimension of 1/8”, an outer dimension of 1/4", and a wall dimension of 1/16”. Manostat ® peristaltic pumps (Cole-Palmer, Court Vernon Hills, IL) were used to maintain nutrient flow and recycle rates. The two lines connected from the reactor to two 1 mL glass serological pipettes (that had the cotton plugs removed) within a drilled rubber stopper fitting a 1 L filtering flask (VWR, Radnor, PA). Prior to use, all components of this setup were autoclaved and sterilized. A 20% Clorox ® bleach solution was cycled through the sealed system for two hours and flushed with sterile, deionized water overnight. The entire system can be seen based on a schematic illustration in Figure 18.

86

Figure 18. Schematic of the biofilm continuous culture setup

87

Biofilm Reactor Units. The mini-reactors were placed inside of a small Labline® incubator (Barnstead, Melrose Park, IL) set at 37°C over the course of the experiments. The flask containing the culture and media was placed atop a Cimarec® stirring hot plate (Barnstead, Melrose Park, IL) also set at 37°C. The cultures were kept oxygenated by using the stirring feature of this particular machine. The feed flow rate into the reactors was approximately 1.2 mL/min which corresponded to 60 RPMs/min. Nutrient cycling through the inflow tubing was continuous in the setup of these mini-reactors. The outflow tubing was not connected to a peristaltic pump and allowed to drain slowly using strategic positioning of the drilled outflow opening and placement of the flask directly beneath the outflow tubing exit. This permitted increased bacterial interaction and retention upon the glass coverslip surface substrates. Fresh media was added during the course of each weeklong experiment. The glass coverslips were removed from each reactor per the design of the experiment and replaced with additional sterilized coverslips. The system was then re-sealed and re-sterilized prior to a new experiment. Nonthermal plasma exposure followed removal of experimental glass coverslips from each bioreactor in a biosafety cabinet to prevent contamination. Nonthermal plasma exposure was fixed 5 mm from the surface of each coverslip in accordance with the design of each individual experiment. Coverslips coated with biofilms were exposed to nonthermal plasma in single intervals from 30 seconds to 2 minutes. Each time point during each experiment contained negative control biofilms collected concurrently that were not exposed to nonthermal plasma. Biofilm samples were immediately stained, imaged, and analyzed after nonthermal plasma exposure.

88

Bacterial Biofilm Staining and Confocal Imaging. Biofilms were stained with 1mL of 1 mg/mL 4’, 6-diamidino-2-phenylindole (DAPI, Invitrogen; Carlsbad, CA) after nonthermal plasma exposure. The stain was gently rinsed and removed from the coverslips prior to biofilm imaging. To quantify biofilm structure, stacks of biofilm images were collected and used to calculate a variety of parameters that would collectively assess the overall structure of each biofilm. An Olympus BX61 spinning disk confocal microscope (Olympus, Center Valley, PA) equipped with a Hamamatsu C10600 camera (Hamamatsu, Bridgewater, NJ) were configured to image fluorescently-labeled bacteria within biofilms. Image stacks were collected from 6-10 random locations around the region of biofilm that had been exposed to the nonthermal plasma discharge. The stacks of images were collected by setting an upper and lower limit. The distance between each cross section was then optimized to 0.18um. Images were collected from various biofilm locations Stack collection occurred daily during biofilm development (7 days), and was captured using Slidebook imaging software (Version 4.2.0.12, Olympus, Center Valley, PA). The image stacks obtained were in a 16-bit gray-scale TIFF format and were later converted to 8-bit and resized to 640 x 480 pixels using ImageJ freeware (NIH) for analysis by COMSTAT (provided by Arne Heydorn) (Heydorn et al. 2000). Z-Stack Analysis. Heydorn et al. (Heydorn et al. 2000) developed a software package called COMSTAT to calculate a wide variety of variables that directly measure the texture and structure of biofilms. COMSTAT was configured as an extension in MATLAB (MathWorks, Natick, MA) to quantify biofilm structural data from collected image stacks. A minimum of six stacks were collected from every biofilm sample,

89

collected in triplicate, for each exposure group (0 seconds, 30 seconds, 1 minute, and 2 minutes) every day during the course of each experiment. This process occurred independently for each independent experiment. Each individual image stack was merged into a binary composite image immediately prior to COMSTAT analysis. Lewandowski et al. (Lewandowski and Beyenal 2007) published a textbook that describes all of the parameters in COMSTAT used in this investigation. These parameters include, but are not limited to: textural entropy, energy, homogeneity, run lengths, nutrient diffusion distances, porosity, thicknesses, roughness, biovolume, and surface area calculation. Each term has been defined and is listed here: 1. Textural Entropy: Textural entropy is a measurement of the randomness of the pixels in a gray scale image. Higher textural entropy values correspond to greater overall biofilm heterogeneous characteristics. 2. Energy: Energy measures regularity in pixel patterning. Lesser energy values indicate repeated patterns of pixels clusters. 3. Homogeneity: Homogeneity measures the similarities of spatially close images.A higher value indicates a lesser overall differences between the overall biofilm structures. 4. Run Directions (X, Y, Z): Run lengths measure consecutive biomass pixels in each dimension’s respective direction. Higher values associated with this parameter would indicate homogeneous, compact biofilms.

90

5. Diffusion Distance: The diffusion distance is defined as the minimum linear distance in all three dimensions from one cluster of pixels to the nearest void pixel in a composite image. When diffusion distances are high, a substrate has to diffuse through many more clusters in a given area. 6. Porosity: Areal porosity is the ratio of void area to total area of a composite image. High values of porosity indicate a greater void to biomass ratio in stacks of biofilm images. 7. Thickness: Thickness is the highest point above each pixel in the bottom layer containing biomass. Mean biofilm thickness provides a measure of the spatial size of the biofilm. Simply, this parameter defines total biomass in the Z direction. 8. Roughness: The roughness parameter is the amount of variability associated with an increase or decrease of thickness in the Z direction. High roughness values can be associated with elevated variation along the Z direction of biomass. 9. Biomass: The relative amount of biomass in a composite image stack is all of the pixel information in three dimensions that falls above a certain minimum threshold set by the program user. 10. Biovolume: Biovolume is defined as the number of biomass pixels in all images of a stack multiplied by the pixel size in three dimensions divided by the substratum area. It represents the overall volume of the biofilm and estimates the biomass of the biofilm. Biovolume differs from volume because biovolume includes pixel intensity values associated with biomass, whereas volume encompasses all total volume.

91

Because these parameters describe structure of biofilms, the morphology must be calculated in three dimensions using variations of pixel intensities of three-dimensional image sets. The COMSTAT software was obtained from the authors and installed as an extension in the computer program MATLAB. Collections of image stacks for each biofilm sample were placed into directories and then analyzed using each of the individual programs of interest. Composite binary images of the combined stacks were created using the program for which all subsequent analyses performed. These composites were created using mathematical analyses of each slice within the presented stacks and their associated 3D voxel information. All control and treatment groups had a minimum of twelve stacks per variable condition and all data was analyzed in triplicate per day per variable. All textural and structural parameters were measured each day over a period of seven days. Emphasis was placed on the results of the structural analysis at days one, four, and seven.

Statistical Analysis. Each of the three stacks per parameter taken each day for each sample were averaged together to produce a dataset that contains, for each sample, a single measure of each structural characteristic of biofilm for each of the seven days data was collected. Jennrich and Schluchter's (Jennrich et al. 1986) covariance pattern modeling approach was used to simultaneously test fixed effects hypothesis regarding the effects of nonthermal plasma on parameters of nonthermal plasma-treated biofilms at initial application, over time and also directly modeling the error variance-covariance

92

structure of the repeated measures taken over the week, as these measures are likely to co-vary within each sample. While a number of variance-covariance structures were considered, an unstructured form that allowed all parameters in the variance-covariance matrix to be to be estimated fit the data best. Least squares means and F tests of fixed effects were calculated for each structural parameter to determine any significant changes to the structural parameters across different exposures to nonthermal plasma and time. To control for a potentially inflated false discovery rate (FDR) due to the large number of statistical tests, p-values are adjusted by using an adaptive step-up Bonferroni method developed by Hochberg and Benjamini (Benjamini et al. 2009). The covariance pattern models were calculated using PROC MIXED in SAS (Littell 2006). Last, to better understand how exposure to nonthermal plasma changes structural parameters over time, LOWESS (locally weighted scatterplot smoothing) curves were created (Cleveland et al. 1988). These plots were created using the R statistical program (Team 2010).

RESULTS In this study, textural and structural parameters of bacterial biofilms were evaluated using confocal microscopy and image analysis following single exposure events of nonthermal plasma discharge. Textural parameters quantify the gray scale intensity variations in biofilm images, where the gray scale values vary from 0 to 255. All textural and structural parameters were measured each day over a period of seven days. Emphasis was placed on the results of the structural analysis at days one, four, and seven. Collectively, this allowed evaluation of nonthermal plasma’s disruptiveness in both

93

Gram-positive and Gram-negative bacterial biofilms during three different periods of development. Experimental design dictated that multiple stacks were collected from each individual sample, including control (no exposure) and all exposure groups. A minimum of six stacks were collected from each biofilm replicate. Triplicate biofilms were collected per experimental group in each day of every experiment. A total of five independent week-long experiments were conducted using the following experimental groups: no exposure (negative control), 30 seconds, 1 minute, and 2 minutes. Biofilm repeatability was measured in two ways. First, the standard deviation of every COMSTAT biofilm structural characteristic was evaluated in all experimental groups (exposed biofilms and negative control, non-exposed biofilms) every day over the course of seven days for each independent experiment. For any given experiment, a minimum of three replicate biofilms were analyzed for every nonthermal plasma exposure group (zero to 2 minutes). Each experiment generated an average mean with accompanying standard deviation. LOWESS curves illustrate changes in gray scale intensity found along the Y axis over the seven days found along the X axis as seen in Figures 19 and 20. The different nonthermal plasma discharge exposure groups are plotted using four different lines. Differences in the starting point of the lines on day one reflect the effects of the initial exposure to nonthermal plasma on the biofilm's structural parameter. Each line’s

94

trajectory from that point reflects the effects of nonthermal plasma exposure on the growth of the biofilm parameter. Generally, four patterns can be observed with the use of these graphs: the first reflects no change in biofilm structure or growth over time displaying lines that overlay each other and remain level from day one to day seven. The second pattern demonstrates overlying lines indicative of no initial effect of nonthermal plasma exposure on a given parameter. However, biofilm growth and development can be seen in the sloping of the lines from day one to day seven. This pattern shows little to no response to nonthermal plasma exposure as the lines with increasing levels of exposure follow the control, which reflects normal biofilm development. The third pattern has different starting points for each of the lines at day one with each line remaining parallel to each other across time. This pattern demonstrates different levels of nonthermal plasma exposure effectiveness on a biofilm parameter without change across time. The fourth pattern also contains different starting points for each of the lines at day one, but the lines spread out over time, indicating both an impact of different exposures to nonthermal plasma and a sustained change in the growth of the biofilm parameter over time. The LOWESS curves for each structural parameter across the two bacteria with Pseudomonas aerguinosa in Figure 19 and Staphylococcus aureus in Figure 20 appear remarkably similar. Statistical tests that provide empirical evidence of the effect of different exposures to nonthermal plasma and over time are provided in Tables 1 and 2. Each parameter in both tables has been assigned a p-value from an F test that determines difference in least squares means across nonthermal plasma exposures per given day.

95

Figure 19. LOWESS statistical curves in P. aeruginosa biofilm structural characteristics

96

These values are presented in the last row of each respective parameter. In the final column for each parameter, a second p-value of an F test is presented that establishes possible difference in least squares means over time within each nonthermal plasma exposure level. he statistical analysis and classfication of the textural and structural parameters associated with P. aeruginosa biofilms were integrated into Table 1.The energy parameter falls into the first LOWESS pattern showing no statistically significant changes following increased nonthermal plasma exposure over time. Seven other parameters, average and maximum diffusion distance, textural entropy, homogeneity and run lengths X, Y and Z, all fall into the second pattern and demonstrate growth in the biofilm as evidenced by statistically significant change over time but no display no significant response to nonthermal plasma exposure. Porosity demonstrates differences in least-squares across different exposures to nonthermal plasma, in all days except day one, compelling us to place this parameter into the third pattern suggesting negative impact of exposure to different levels of nonthermal plasma without change over time. Finally, four other parameters, average biofilm thickness, biovolume, roughness and volume, fell into the fourth pattern with statistically significant changes across different nonthermal plasma exposures and across time. Of these four variables, three demonstrated an interaction between nonthermal plasma exposure and time effectively indicating that increased time of exposure to nonthermal plasma altered growth in a given biofilm parameter over time. For three parameters [average biofilm thickness (F18,16 = 11.66, p < 0.01), biovolume (F18,16 = 6.29, p < 0.01) and volume (F18,16 = 21.74, p < 0.01)] increased exposure to nonthermal plasma reduced structural biofilm growth,

97

Figure 20. LOWESS statistical curves in S. aureus biofilm structural characteristics

98

while the fourth parameter, roughness, which measures texture or structural instability of the biofilm, increased after longer exposure to nonthermal plasma (F18,16 = 13.98, p < 0.01). S. aureus biofilm parameters were also analyzed and characterized (Table 2). Energy, textural entropy and surface area are parameters that showed no statistically significant changes towards increased nonthermal plasma exposure over time, fitting into the first LOWESS pattern. Surface area was a measure of internal consistency throughout the experiment; its inclusion in this group was supportive of consistent methodology. Collection and measurement of the surface area parameter was an internal control showing that the given area analyzed in individual experiments as well as between all experiments was held consistent at each collected time point The second pattern contains six parameters, average and maximum diffusion distance, homogeneity and run lengths X, Y and Z characterized by biofilm development over time while and indifference towards increasing nonthermal plasma exposure intervals. Porosity and roughness fell into the third pattern characterized by statistically significant changes in least squares means with increased exposure to nonthermal plasma with little change over time. Lastly, the average biofilm thickness (F18,16 = 9.44, p < 0.01) and biovolume (F18,16 = 6.29, p < 0.01) suggested interaction between nonthermal plasma exposure and time, indicating increased nonthermal plasma exposure reduced structural biofilm growth. Biovolume showed statistically significant changes across different nonthermal plasma exposures and across time.

99

The parameters most significantly affected by nonthermal plasma exposure were roughness, porosity, thickness, and biomass. The loss of biomass suggests that the bacteria and exopolysaccharide materials within the biofilm and their corresponding pixel intensity was lost. This phenomenon was observed by Moisan and described as bacterial biofilm “etching” (Moisan et al. 2001). Visible loss of biomass associated with an intact biofilm structure was observed in this study and supports this observation (Figure 21). Likewise, there is an overall reduction in the total biomass per given surface areas following nonthermal plasma exposure. Figure 22 is a representative composite image of a given surface area that presents increases in randomness or dispersal of biomass from centralized, concentrated areas. This supports the concept of lost biomass after nonthermal plasma exposure. Both grouped effects displayed a positive correlation with increased length of exposure up to the maximum exposure length of this study. The 3D structure was also constructed using ImageJ stack-reader software and shows significant changes in overall structure, particularly in the Z-direction (Figure 23). Pixel intensity was measured using colorimetric scales.These imaging data support the COMSTATgenerated composite images and conclusions drawn from the LOWESS graphs.

100

Figure 21. S. aureus biofilms exposed to nonthermal plasma. (A) No exposure at 100x TM, (B) 1 minute at 100x TM, (C) 2 minutes at 100x TM, (D) No exposure at 400x TM, (E) 1 minute at 400x TM, and (F) 2 minutes at 400x TM

101

Figure 22. COMSTAT-generated binary images of Z-stacks in S. aureus biofilms. (A) No exposure D1 biofilm, (B) 1 minute exposure D1 biofilm, (C) 2 minutes exposure D1 biofilm, (D) No exposure D4 biofilm, (E) 1 minute exposure D4 biofilm, and (F) 2 minute exposure D4 biofilm

102

Figure 23. ImageJ-generated 3-dimension composite images of S. aureus biofilm stacks. (A) No exposure - Z dimension, (B) No exposure - all dimensions, (C) 1 minute - Z dimension, (D) 1 minute - all dimensions, (E) 2 minutes - Z direction, and (F) 2 minutes all dimensions

103

DISCUSSION The plasma generator used in this study possesses several properties that may impart additional advantages for sterilization of applied bacterial constructs such as biofilms. Plasma sterilization depends on the combination of the main components that make up the discharge: energized UV photons, reactive oxygen species, and charged electrons and ions. The relative importance of any one main plasma constituent can vary depending on the cell being exposed, the sample preparation, or the type of initial plasma gaseous medium. These variables can dramatically alter the produced plasma as well as impact how the plasma interacts with target cells in a specific environment or interface. Laroussi observed complete destruction of a population of P. aeruginosa with 10 minutes using a type of nonthermal plasma with very low UV output (Laroussi 1996). However, Baier reported that UV emitted by a radio-frequency discharge is not sufficiently able to act as the sole destructive force behind microorganism destruction (Baier et al. 1992). Thus, synergy between components within the plasma spectra is paramount towards understanding the mechanism(s) behind plasma sterilization. One of the most commonly studied plasmas is air plasma at room temperature (as in this study). Air plasma often contains high concentrations of reactive oxygen and nitrogen- based species (Laroussi et al. 2004). Atomic oxygen and reactive oxygen species are known to disrupt the integrity of bacterial cell membranes through downstream electrostatic processes resulting in cell membrane destabilization (Bandyopadhyay et al. 1999). Cells lose their ability to maintain proper intracellular pH if their cell membranes become increasingly permeable through free radical damage

104

(Benstaali et al. 1998; Moisan et al. 2001). It is believed that the UV radiation associated with this specific type of plasma device is able to penetrate through biofilm samples (Philip et al. 2002). Reactive species and other free radicals react with the local environment before they can penetrate through a deep, complex biofilm matrix. This suggests that there is some other component responsible for penetrating further through the biofilm. Given that the plasma discharge from this plasma generator can penetrate through 1 centimeter through biological matrices, it is probable that the UV radiation in the nonthermal plasma discharge is able to penetrate into bacterial biofilms. The UV photons could damage the DNA after penetrating through compromised matrices and damaged cells. UV-mediated damage in bacterial cells can be thought of as a longer term event contrasting the immediate disruptive ROS/RNS-mediated effects on bacterial cell membranes. Laboratory investigation of biofilms as they grow and develop requires a consistent, reproducible method that utilizes extensive repetition in order to study as much of the overall biofilm structure as possible. The method chosen for this investigation is a modification of a continuous culture construct utilized by Jackson et al. (Jackson et al. 2001). Confocal microscopy affords users the opportunity to obtain stacks of images collected in three-dimensions, making it a valuable tool for investigations into biofilm architecture (Palmer 1999; Palmer et al. 2006). Lewandowski et al. developed a series of MATLAB-compatible programs that enable the collection of textural and volumetric parameters. An established computer program used to integrate, measure, and analyze image information collected from confocal microscopy is paramount to

105

investigations into biofilm structure COMSTAT has been further developed and used with success in other studies investigating biofilm formation or structure (Lewandowski et al. 1999; Heydorn et al. 2000; Hentzer et al. 2001; Davey et al. 2003; Beyenal et al. 2004; Labbate et al. 2004; Wood et al. 2006; Hansen et al. 2007; Lewandowski and Beyenal 2007). Our use of these independent methods has allowed us to identify specific aspects of biofilm structure affected by air-generated nonthermal plasma. Textural entropy, energy, and homogeneity are all variables to measure the texture of a biofilm (Lewandowski and Beyenal 2007). Thus biofilm textural values are more of a unique “fingerprint” identifying specific microbial populations including environmental conditions in which they developed. Importantly, changes in textures resulting from environmental conditions produced by nonthermal plasma can serve as indicators of biofilm damage. Run lengths in the X, Y, and Z directional planes, average and maximum diffusion distances, porosity, average and maximum biofilm thickness, roughness, biovolume, and surface area are all variables used to measure the structure of biofilms (Lewandowski and Beyenal 2007). Structural, or volumetric, parameters express the relative morphology of biofilms and were calculated using pixels that represent the biomass of collected images. All structural parameters were calculated using composite binary images created by image thresholding. The statistical analysis of the collected biofilm 3D stack images show significant decreases in the thickness and biovolume values in both P. aeruginosa and S. aureus biofilms after nonthermal plasma exposure. The thickness parameter is an indicator of

106

growth in the Z-direction. Significant thickness loss affects the overall structure of the biofilm as developmental progression occurs. Dramatic increases in the roughness and porosity values were observed in both species of the bacterial biofilm constructs. Changes in these values suggest that uniform, compact areas of biomass seen in controls become scattered or diffuse throughout a given surface area following plasma exposure. This is further supported by large losses of biomass in the biofilm as seen in Tables 1 and 2. These data also suggest that the stage of biofilm development does not significantly limit the biocidal effectiveness of nonthermal plasma exposure in bacterial biofilms. Biofilm structural damage was affected by nonthermal plasma discharge as evidenced in resulting changes in biomass, thickness, roughness, and porosity values. This was a critical observation made due to the presumed differences in tolerance to nonthermal plasma as biofilms developed and matured. The other structural parameters investigated in this study, diffusion distances, or run distances in the X, Y, and Z directions, showed little significant response to nonthermal plasma. Much like the textural parameters studied, they seemed to be more of a unique function of the individual biofilms themselves. For several parameters, observed trends seemed to indicate effectiveness in specific parameters, but the internal variability served to cloud such observations. This may be resolved with additional experiments. As such, while the statistical power was sufficient for this study, it is limited by the unstructured variance-covariance matrix needed to model errors that co-vary within each sample and cannot precisely quantify precise incremental increases in textural and

107

structural parameters after exposure to nonthermal plasma and how those changes interact with each other over time. Previous studies have been concerned with the overall phenotype and viability of biofilms following exposure to various types of nonthermal plasma (Yokoyama et al. 1990; Birmingham et al. 2000; Montie et al. 2000; Moisan et al. 2002; Laroussi et al. 2003; Akitsu et al. 2005; Fridman et al. 2008; Joaquin et al. 2009; Lee et al. 2009). Atomic force microscopy of C. violaceum biofilms exposed to nonthermal demonstrated that cells become diffuse and scattered as the structure is lost (Joaquin et al. 2009). In our studies, P. aeruginosa and S. aureus biofilms exposed to nonthermal plasma (up to two minutes) and stained with 5-Cyano-2,3-Ditolyl Tetrazolium Chloride (CTC) showed marked differences in metabolic activity (data not shown). CTC has been used previously as an indicator of redox activity (Gruden et al. 2003). Biofilm samples not exposed to nonthermal plasma displayed a significantly higher CTC to CTF (a fluorescent metabolic byproduct) transition compared to biofilms exposed to nonthermal plasma at doses for as little as thirty seconds. Traces of cell damage were observed using vital dye staining. Composite MATLAB and 3D imaging indicated marked difference in structure between biofilms exposed and not exposed to doses of nonthermal plasma (Figures 21 and 22). The phenotype of the individual cells following plasma exposure displayed a marked reduction in uniformity and appears much more rough and acellular. The major components found within the plasma discharge generated by the device used in our studies contained atomic oxygen, reactive oxygen species, reactive nitrogen

108

species, and in particular, nitric oxide, low level ultraviolet radiation, and other trace ionized electrode materials (Personal communication from Dr. Bruce Banks, NASA Glenn, Cleveland, OH). This plasma subtype fits into the general class of air plasmas described previously. However, the manner in which the reactive species are created differs from other air plasmas. The overall concentration and makeup of the plasmaassociated molecules involved also differ. The impact of these components on bacterial biofilms suggests that reactive and charged species also affect the integrity of the polysaccharide matrix secreted by the bacteria in the biofilm aggregates. Nonthermal plasma exposure increased the amount of damage to the overall biofilm structure as evidence by porosity holes or loss of biomass and accompanying biofilm extracellular materials. Increases in surface free radicals and charged ions may explain why the biofilm structure breaks down over time after nonthermal plasma exposure. The importance of the actual biofilm structure and subsequent meaning for the viability of the remaining members following destabilization is also worth noting. Extensive damage in the local environment of a treatment zone of a biofilm could explain the structural data acquired in Tables 1 and 2. Should local areas within the biofilm environment sustain sufficient damage, it could cause surrounding structures to collapse, scattering clusters of biomass to other regions of the immediate plasma exposure zones. The instability of the structure of the surviving biomass left behind determined by biofilm porosity, roughness, and thickness would support this proposed observation. However, biofilm destabilization is important beyond immediate viability of the encapsulated cells. A longer term effect can

109

be associated with damaged or dying cells negatively impacting nearby viable cells, impacting future expolysaccharide production. Nonthermal plasma appears to penetrate bacterial biofilms in various phases of development due to destabilizing effects observed at all three developmental points in this study, thus it may destabilize biofilm structures of various maturity stages. While a specific mechanism is not completely defined, germicidal and destabilizing effects is likely a combination of short and long term effects acting in tandem. Nonetheless, we have demonstrated substantial biofilm disruption by 30 seconds of nonthermal plasma produced from ambient air at room temperature. The makeup and configuration of this specific plasma generator affords the user anti-biofilm capabilities and could serve as a possible tool in a variety of applications. The biofilms evaluated in this study were composed of pathogenic strains of S. aureus and P. aeruginosa. These biofilms were grown and designed in order to mimic bacterial biofilms found within slow-healing, chronic skin wounds. These organisms can also be found growing in biofilms on joint replacement constructs or catheters used in patient care. The plasma-mediated destabilization of biofilm structure likely weakens the overall integrity of the entire aggregate. This could allow introduction of secondary therapies or interventions such as systemic or topical antibiotics in order to further damage the integrity of the biofilm structure while simultaneously killing bacterial members within the biofilm itself with the goal of minimal re-colonization events.

110

Table 1: Changes in measured gray scale intensity of textural and structural parameters of Pseudomonas aerguinosa biofilms after varying exposures to nonthermal plasma discharge over time Textural and structural parameters Average diffusion distance

Average biofilm thickness

Biovolume

Energy

Textural entropy

Homogeneity

Maximum diffusion distance

Nonthermal plasma exposure 0 (control) 30 sec. 60 sec. 120 sec. p-value 0 (control) 30 sec. 60 sec. 120 sec. p-value 0 (control) 30 sec. 60 sec. 120 sec. p-value 0 (control) 30 sec. 60 sec. 120 sec. p-value 0 (control) 30 sec. 60 sec. 120 sec. p-value 0 (control) 30 sec. 60 sec. 120 sec. p-value 0 (control) 30 sec. 60 sec. 120 sec. p-value

Least squares means over time (days) 1 8.9 11.5 16.6 16.1 0.63 34.2 28.6 27.6 23.7 0.03 2.3 1.9 1.6 1.4