Citation
Use of Bacteriophages to Decontaminate Naturally and Experimentally Infected Oysters of Vibrio vulnificus

Material Information

Title:
Use of Bacteriophages to Decontaminate Naturally and Experimentally Infected Oysters of Vibrio vulnificus
Creator:
MARTIN, JULIO LZARO
Copyright Date:
2008

Subjects

Subjects / Keywords:
Bacteria ( jstor )
Bacteriophage typing ( jstor )
Bacteriophages ( jstor )
Cocktails ( jstor )
Control groups ( jstor )
Infections ( jstor )
Killing ( jstor )
Oysters ( jstor )
Plasmids ( jstor )
Sea water ( jstor )
Gulf of Mexico ( local )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Julio Lzaro Martin. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
12/31/2006

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

USE OF BACTERIOPHAGES TO DE CONTAMINATE NATURALLY AND EXPERIMENTALLY INFECTED OYSTERS OF Vibrio vulnificus By JULIO LAZARO MARTN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Julio Lazaro Martn

PAGE 3

This document is dedicated to the five most important people in my life: Julio, Juana, Jacqueline, Tommy, and Tammy Martn.

PAGE 4

iv ACKNOWLEDGMENTS I sincerely thank my advisor, Dr. Paul Gu lig, for all of his gu idance throughout my Master of Science program. I would also like to thank Dr. Donna Duckworth and Dr. Anita Wright for their insight and help th roughout the program. I also truly appreciate the technical support provided by Eric, Gopal, Qiu, and Patrick. Furthermore, I would like to thank my loving wife, Tammy L. Martn, for being so supportive and understanding throughout my studies, and my parents, Julio and Juana Mar tn, and sister, Jacqueline Delgado, for their continuous support.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES...........................................................................................................viii LIST OF FIGURES...........................................................................................................ix ABSTRACT....................................................................................................................... xi CHAPTER 1 INTRODUCTION........................................................................................................1 Bacteriophages..............................................................................................................1 The Discovery of Bacteriophages.........................................................................1 Phage Biology.......................................................................................................1 Early Phage Therapy Attempts..............................................................................3 Renewed Interest in Phage Therapy......................................................................6 Vibrio vulnificus ..........................................................................................................16 Vibrio vulnificus Pathogenesis and the Oyster Industry......................................16 Phage-Treatment of Experimentally V. vulnificus Infection of Mice..................19 Specific Aims..............................................................................................................21 Specific Aim 1: Isolation and Char acterization of Bacteriophages.....................21 Specific Aim 2: Establishment of an Experimental Oyster Infection Model and Treatment of the Experimentally Infected Oysters with Phages to Reduce Numbers of V. vulnificus .....................................................................21 Specific Aim 3: Phage Treatment of Naturally Infected Oysters........................21 2 MATERIALS AND METHODS...............................................................................22 Bacterial Strains, Media, and Growth Methods..........................................................22 Bacteriophage Strains, Isol ation, Amplification, Purifi cation, and Quantification....23 Bacteriophage Strains..........................................................................................23 Bacteriophage Isolation.......................................................................................24 Bacteriophage Amplification...............................................................................25 Broth phage amplification method...............................................................26 Plate phage amplification method................................................................26 Purification of Phage...........................................................................................27 Quantification of Phage.......................................................................................27

PAGE 6

vi Drop titers.....................................................................................................27 Full plate titers..............................................................................................28 Phage Typing..............................................................................................................28 Soft Agar Overlay Phage Typing Method...........................................................28 Microtiter Phage Typing Method........................................................................28 Phage Treatment of Experime ntally Infected Oysters................................................29 Experimental Infection of Oysters.......................................................................29 Phage Treatment..................................................................................................30 Harvest of Oysters...............................................................................................30 Phage Treatment of Naturally Infected Oysters.........................................................30 Phage Treatment..................................................................................................30 Harvest of Oysters...............................................................................................30 Bacterial Growth Assays............................................................................................32 pGTR902 Marker Plasmid..................................................................................32 Experimental infection of oysters................................................................33 Harvest of oysters.........................................................................................33 Ampicillin Treat ment Assay................................................................................33 Experimental infection of oysters................................................................34 Harvest of oysters.........................................................................................34 Statistical Analysis......................................................................................................34 3 RESULTS...................................................................................................................35 Rationale for Study.....................................................................................................35 Specific Aim 1: Isolation and Char acterization of Bacteriophages............................35 Soft agar Overlay Phage Typing Method............................................................36 Effect of Sea Salts on the In fectivity of Certain Phages.....................................43 Development of a Microtiter Phage Typing Method..........................................45 Phage Isolation and Transmission El ectron Micrographs of New Phages..........66 Phage isolation.............................................................................................66 Transmission electron micrographs of isolated phages................................68 Specific Aim 2: Establishment of an E xperimental Oyster Infection Model and Treatment of Experimentally Infected Oysters with Phages to Reduce Numbers of V. vulnificus .......................................................................................................71 Establishment of an Experiment al Oyster Infection Model................................71 Bacteriophage Treatment of Expe rimentally Infected Oysters...........................78 Treatment of Experimentally Infected Oysters with Phages 3a and CK-2.........80 Growth of V. vulnificus in Oysters......................................................................85 Marker plasmid pGTR902...........................................................................85 Ampicillin treat ment assays.........................................................................89 Attempts at Stimulation of V. vulnificus Growth Coupled with Phage Treatment.........................................................................................................94 Iron supplementation....................................................................................95 Elevated water temperature..........................................................................95 Specific Aim 3: Phage Treatment of Naturally Infected Oysters...............................98 Quantification Methods Eval uated for Enumeration of V. vulnificus Naturally Occurring in Oysters............................................................................................99

PAGE 7

vii 4 DISCUSSION...........................................................................................................108 The Potential use of V. vulnificus -Specific Phages in Treating V. vulnificus Infections..............................................................................................................108 Isolation and Characteriza tion of Bacteriophages....................................................108 Soft Agar Overlay Phage Typing Method.........................................................108 Microtiter Phage Typing Method......................................................................113 Growth of V. vulnificus in Oysters...........................................................................127 Marker Plasmid pGTR902 Assay......................................................................128 Ampicillin Enrichment Assays..........................................................................129 Attempts at Stimulation of V. vulnificus Growth Coupled with Phage Treatment.......................................................................................................132 Phage Treatment of Naturally Infected Market Oysters...........................................133 LIST OF REFERENCES.................................................................................................143 BIOGRAPHICAL SKETCH...........................................................................................150

PAGE 8

viii LIST OF TABLES Table page 3-1 Level of infectivity of each strain of V. vulnificus to each strain of V. vulnificus bacteriophage...........................................................................................................39 3-2 Host range for each strain of V. vulnificus -bacteriophage.......................................43 3-3 Host range of newly isolated phages on V. vulnificus ..............................................68 3-4 Minimal inhibitory c oncentration of Amp for V. vulnificus FLA077......................90

PAGE 9

ix LIST OF FIGURES Figure page 3-1 Effect of Mg+2 and Ca+2 on lysi s of V. vulnificus LL728 by phage CB-1............45 3-2 Effect of LB-N and LB-SW on the in fectivity of phages to V. vulnificus ..............48 3-3 Effect of using SW as diluent on the growth of V. vulnificus in either LB-N or LB-SW.....................................................................................................................51 3-4 Comparison of LB-SW, LB-IO, a nd LB-RS on growth of FLA042 and infectivity of phage CK-2.........................................................................................54 3-5 Comparison of V. vulnificus M06-24/O growth at various initial concentrations...55 3-6 Effect of different concentrations of V. vulnificus M06-24/O on phage infection...56 3-7 Final microtite r infections of V. vulnificus stains with 14 diffe rent phage strains at RT.........................................................................................................................5 9 3-8 Microtiter inf ection assays of V. vulnificus strains 2400112 and LL728 with 14 different phage strains at 37C.................................................................................65 3-9 Transmission electron microgr aphs of isolated phages............................................70 3-10 Effect of rifampicin and algae on ex perimental infection of oysters with V. vulnificus FLA042....................................................................................................74 3-11 Effect of the duration of infection time on the level of V. vulnificus infection in oysters.......................................................................................................................7 5 3-12 Effect of ultraviolet light filters on experimental infection of oysters with V. vulnificus ..............................................................................................................77 3-13 Analysis of the number of naturally o ccurring phages present in market oysters and water that are infectious to V. vulnificus M06-24/O and MLT403...................79 3-14 Retention of phage CK-2 in oysters.........................................................................80 3-15 Treatment of V. vulnificus FLA042-infected oyste rs with phage CK-2..................81

PAGE 10

x 3-16 Effect of CK-2 and 3a treatment of oysters infected for 1 h with V. vulnificus FLA042: before and after homogenization..............................................................82 3-17 Effect of CK-2 and 3a treatment of oysters infected for 6 h or 24 h with V. vulnificus FLA042....................................................................................................88 3-18 Segregation of the marker plasmid pGTR902 in V. vulnificus in experimentally infected oysters.........................................................................................................89 3-19 Effect of ampicillin treatment on V. vulnificus FLA042 in experimentally infected oysters.........................................................................................................92 3-20 Effect of ampicillin enrichment compar ed to treatment with phages CK-2 and 3a on V. vulnificus FLA042 in experimentally infected oysters...................................93 3-21 Effect of ampicillin enrichment on V. vulnificus FLA077 in experimentally infected oysters.........................................................................................................94 3-22 Effect of supplementation of seawater with FeCl3 on V. vulnificus FLA042 growth in experimentally infected oysters and on the efficacy of phage treatment.96 3-23 Effect of water temperature on the efficacy of phage treatment of V. vulnificus FLA042-infected oysters..........................................................................................97 3-24 Plating efficiencies of LB-N, VVM , and TCBS on marker oysters and on V. vulnificus FLA042 experimentally infected oysters...............................................101 3-25 Effect of treatment of market oysters with a cock tail of phages............................104 3-26 Effect of treatment of market oysters with a cock tail of phages............................105 3-27 Effect of treatment of market oysters with a cock tail of phages............................106

PAGE 11

xi Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science USE OF BACTERIOPHAGES TO DE CONTAMINATE NATURALLY AND EXPERIMENTALLY INFECTED OYSTERS OF Vibrio vulnificus By Julio Lazaro Martn December 2005 Chair: Paul A. Gulig Major Department: Molecular Genetics and Microbiology Vibrio vulnificus is a bacterium commonly foun d in shellfish and surrounding waters. V. vulnificus is an opportunistic pathogen capab le of causing serious illness in humans, usually through the consumption of raw oysters; it is the leading cause of reported fatalities associated with seafood consumption. In gestion of raw oysters can lead to fulminant primary septicemia, especia lly in individuals with pre-existing health conditions, such as liver disease, hemochroma tosis, or a compromised immune system. Additionally, the contact of wounds with raw oysters, oyster fluids, or seawater contaminated with V. vulnificus may also lead to wound infe ctions in otherwise healthy individuals, which can lead to severe necrosis that may require surgical debridement or amputation, or even sepsis and death. Bacteri ophages are viruses that specifically infect bacteria. Using an iron dext ran-treated mouse model of V. vulnificus infection, our laboratory determined that administration of V. vulnificus phages significantly inhibited infection of skin tissues, systemic infection of liver, and mortality of infected mice.

PAGE 12

xii This study was undertaken to examine the potential of V. vulnificus -specific phages to decontaminate oysters of V. vulnificus so as to render them safe for human consumption. Phages in our collection and ne wly isolated phages were characterized by plaque morphology, host range, and level of in fectivity using soft agar overlay and microtiter phage typing methods. Seventy-five percent of V. vulnificus strains examined were sensitive to at least one phage. An experimental oyster infection mode l was developed for the examination of phage treatment to decontaminate oysters infected with V. vulnificus . Pretreatment of oysters with rifampicin prior to in fection with rifampicin-resistant V. vulnificus FLA042 resulted in levels of infection as high as 106 CFU/g tissue. The treatment of FLA042infected oysters with phages CK-2 and 3a each at 1 x 108 PFU/mL resulted in decreases of contamination in the range of 10 to 100-fold compared to a control gro up. With the use of a marker plasmid and ampicillin enrichment assays, we showed that V. vulnificus grew very little in experimentally treate d oysters, possibly explaining the limited success of treatment of experimentally infected oysters with phages. The potential of decontaminating market oys ters with a cocktail of phages was also examined with the use of a selective me dium, VVM, and a DNA-hybridization probe for enumeration of V. vulnificus in oysters. This treatment resulted in a 10-fold decrease in the level of contamination, demonstrating the potential of dec ontaminating market oysters with V. vulnificus -specific phages. However, we were unable to improve upon the success of phage treatment to levels that we feel would be commercially acceptable. Modifications to the phage treatment method may circumvent the issue of the lack of V. vulnificus growth.

PAGE 13

1 CHAPTER 1 INTRODUCTION Bacteriophages The Discovery of Bacteriophages In 1915, Frederick Twort, a British bacterio logist, observed “degenerative changes” on some staphylococcal colonies that were extracted from calf lymph (1). He also observed that these “degenerative changes” were transmittable to a new culture of bacteria using the filtrate of the original culture. A lthough his observations were of significant value, Twort did not pursue his findings further, presumably due to financial constraints. Two years later, Felix d’Herelle, a French-Canadian self-made microbiologist working at the Pasteur Institu te in Paris, described how an “invisible microbe” from the filtrates of stool samples of dysentery patients lysed a broth culture and caused clear plaques on a bacterial lawn of dysentery bacillus (2). Regarding these observations, he later wrote: "In a flash I ha d understood: what caused my clear spots was in fact...a virus parasitic on bacteria." He correctly recognized that the plaques were caused by a bacterial virus that infects and lyses the bacteria. He later called these bacterial viruses bacteriophages, meaning “b acteria eaters.” The word “phage” is a shortened version of the word “bacteriophage.” Phage Biology Bacteriophages are on e of the most abundant biol ogical entities found on Earth (3). These viruses, which are specifically infecti ous to bacteria, are found all over the world, even in such extreme places as in hot springs . It has been estimate d that a milliliter of

PAGE 14

2 coastal seawater contains 107 phage-like particles (4). All bacteriophages contain nucleic acid, either DNA or RNA, encapsula ted in a protein coat called a capsid. The nucleic acid can be single or double stra nded. Morphologically, bacteriophages are composed of one of three forms: an icosah edral head containing a tail, an icosahedral head containing no tail, or a filamentous form . Ninety-six percent of all characterized bacteriophages are of the icosahedral head, containing dsDNA, and tail form (5). Due to the presence of a tail, these phages taxonomically fall under the Caudovirales order. Depending on the size of the tail, bacteriopha ges are further differe ntiated into three families: Siphoviridae , Myoviridae , or Podoviridae . The differences of tail morphology among the families are as follows: Siphoviridae consist of a long, non-contractile tail; Myoviridae consist of a contractile tail; and Podoviridae consist of a short, noncontractile tail. Bacteriophages are further characterized into two distinct life cy cles: virulent or temperate. In the virulent (lytic) life cycl e, a lytic phage binds to a specific bacterial surface receptor. For efficient binding of a phage to a receptor, a phage may require a cluster of the specific receptor or the pr esence of divalent cations, such as Mg+2 and Ca+2. After attachment of the phage to the receptor, lysins on the tail of the phage degrade the peptidoglycan matrix of the host permitting the pe netration of the tail into the cell. With the penetration of the tail tip into the host, the phage genome passe s through the tail of the phage and into the host. The phage genome encodes enzymes that override the metabolic machinery of the host. The seized metabolic machinery of the host is then used for the mass production of structural components for the eventual assembly of progeny phages. After numerous proge ny phages are made, phage genome-encoded

PAGE 15

3 lysins and holins lyse the bacterial cell. Holins assemble pores in the inner membrane of the cell wall allowing lysins to reach the peptidoglycan for degradation. In the temperate (lysogenic) life cycle, a temperate phage infects bacteria in the same manner as described with lytic phage s. Upon infection, however, a temperate phage has the ability to proceed through the pr ocesses of a lytic cycle, as described for lytic phages above, or proceed through a lysoge nic cycle. In the lysogenic cycle, the phage genome does not express enzymes that override the metabolic machinery of the host. Instead, the phage genome is either inte grated into the DNA of the host or exists as a plasmid, and the infection becomes latent. Th e phages in this state of latent infection are called prophages. A repressor protein bloc ks the expression of the lytic genes of the phage genome. However, the repressor may be inactivated and the prophage activated at any time by environmental factors, causing re version to a lytic lif e cycle and production of progeny phages. Early Phage Therapy Attempts Upon d’Herelle’s observations regarding pha ge biology in his la boratory, he sought to test the effectiveness of phage as a treatm ent for infectious diseases in animals. He demonstrated the prophylactic properties of phage against the infecti ous disease avian typhosis, a gastrointestinal di sease caused by the bacterium Salmonella gallinarum (6). Upon orally treati ng chickens with S. gallinarum -specific phage prior to their introduction to a chicken pen that housed chickens infected with avian typhosis, the pretreated chickens were less likely to experience S. gallinarum infection. In another phage therapy experiment using animals, d’He relle was able to protect water buffaloes from experimental infection of bovine hemo rrhagic septicemia, caused by the bacterium Pasteurella multocida , by pretreatment with P. multocida -specific phage (6) .

PAGE 16

4 After these and other encouraging results with phage therapy in treating infectious diseases in animals, d’Herelle attempted to tr eat humans afflicted with pathogens such as Vibrio cholerae and Yersinia pestis with either prophylactic or therapeutic phages, respectively (7,8). Prior to attempting phage treatments on patients, d’Herelle administered himself, co-workers, and even some of his own family members with phage suspensions to investigate their safety (6 ). Although the method d’Herelle used for verification of safety of phage suspen sions is not an accepted method by today’s standards, during his time this method was wi dely accepted. After he concluded that the filtrates were safe for human use, he sought to treat humans with therapeutic phages. One of d’Herelle’s most widely report ed cases of phage therapy on humans occurred while d’Herelle was stationed at th e League of Nations quarantine station in Alexandria, Egypt. This partic ular case was reported in th e French medical periodical, La Presse Mdicale , in which d’Herelle treated four people infected with the bacterium responsible for bubonic plague, Y. pestis , with Y. pestis -specific phage (9). He injected Y. pestis phage suspension directly into the buboes, the term given to the lymph nodes that are inflamed and infected with Y. pestis . He specifically inject ed the phage suspension into inguinal and axillary lymph nodes, and th e four patients eventually recovered from this dreaded disease. In another example of the potential benefits of phage therapy, V. cholerae -specific phages were used as a prophylax is in the Indian region of Naogaon resulting in a significant decr ease of deaths associated wi th cholera compared to the Habiganj region of the country, whic h refused to participate in the V. cholerae phages study (10,11).

PAGE 17

5 During the 1920s and 1930s, companie s in the United States produced commercially available phage products (12). Eli Lilly & Co. produced a commercial phage product called “staphylo jel” and othe r phage products that supposedly treated Streptococcus spp. and colon bacilli infections. Mo reover, E. R. Squibb & Sons had a phage product against Staphylococcus spp ., and a division of Abbo tt Laboratories had a phage product for Staphylococcus spp . and colon bacilli. Howe ver, the effectiveness of these commercially available phage products was in doubt. In the 1930s, d’Herelle and colleagues analyzed many commercial phage pro ducts and found that most were inactive. To confirm this finding, Max Delbrck analyz ed a commercial phage product that was advertised to consist of polyvalent phages (10). Analysis of the product, however, revealed that it only contained one viable phage type. Further inves tigation revealed that this particular company was amplifying the polyvalent phages in a single infection process, resulting in the selection of one phage, T7 in this case. Eventually many factors amassed in contri buting to the demise of phage therapy, including the lack of unders tanding of phage biology. Wh ile d’Herelle believed that bacteriophages were bacterial viruses, many in the scientific co mmunity at the time believed that phages were enzymes. In view of that fact, scientists were characterizing phage solutions as strong or weak prepar ations depending on the “strength” a phage solution would have to lyse a culture of b acteria. Moreover, phage products were also treated with chemicals, such as phenol and me rthiolate, for preserva tion. The addition of such chemicals was common practice at the time as a method to preserve vaccines and pharmaceutical products; however, these chemicals may have degraded or inactivated many of the phage preparations. Furthermor e, restriction and m odification processes

PAGE 18

6 possessed by bacteria were not complete ly understood until the 1950s, which was years after the first studies of phage therapy. Moreover, the role of the host immune system during phage treatment was mostly overlooke d or ignored. Anti bodies against phage after extended phage treatment can render phage therapy ineffective in such cases (13). After the discovery of antibiotics, scientific enthusiasm for phage therapy trials ceased in the United States and in most of Western Eu rope. Nevertheless, pha ge therapy continued in the Soviet Union and in some Eastern Eur opean countries, especially in Poland and in the Republic of Georgia (13). Renewed Interest in Phage Therapy During the 1980s in the United Kingdom, Smith and colleagues reexamined the efficacy of phage therapy in animal infections (14-17). In one particular experiment, they selected phages that required the Escherichia coli ( E. coli ) K1 capsular antigen as a receptor for phage attachment. Mice and fa rm animals were pre-treated with these selected phages and then expe rimentally infected with E. coli O18:K1:H7 ColV+. The E. coli -specific phages were effective in protecting the animals from the E. coli . Even though phage-resistant E. coli isolates were recovered from these animals, the resistant E. coli were less virulent due to the loss of the K1 capsule, an antigen necessary for pathogenesis. Furthermore, compared to multip le doses of certain antibiotics, including tetracycline, ampicillin, and chloramphenicol, a single dose of phage performed better in treating the infection, presumably due to th e fact that phages am plify exponentially as long as bacterial growth pers ists, whereas antibio tics begin to degrade at the point of administration. Following the reports by Smith and co lleagues of their phage studies in experimentally infected animals, a number of other research proj ects reinforced the

PAGE 19

7 potential of phage therapy as an antimicrobi al strategy against pathogens. Below is a summary of several reported experiments that have demonstrated encouraging results in treating various bacterial in fections in patients or c ontamination of foods with bacteriophages. Since Smith and colleagues’ experiments with E. coli and E. coli bacteriophages, others have also reported positive results in reducing or eliminating E. coli infections using E. coli phages. In one study, mice were orally infected with E. coli and then treated with four broad host range E. coli phages that survived passage through the gastrointestinal tract of mice when given in drinking water (18). The four administered bacteriophages lysed the E. coli . Furthermore, the normal E. coli gut flora was minimally lysed by the four administered E. coli phages, suggesting that the normal flora E. coli is somehow protected from infection and lysis by the four bacteriopha ges. The protection from phage lysis observed for the normal E. coli gut flora may demonstrate physical differences, such as the level of biofilm formation and/or colonization, between the normal E. coli gut flora and the E. coli that was orally administ ered. Nevertheless, this report suggests a possible use of phages for E. coli infections. In another study, the efficacy of the an tibiotic enrofloxacin was compared to bacteriophages for treating E. coli infections in broiler chickens (19). A mortality rate of 68% was observed when birds were experimentally infected with 104 CFU of E. coli that were injected directly into the thoracic ai r sac of birds. When enrofloxacin or 109 PFU each of two different bacteriophages were in tramuscularly injected immediately after bacterial infection, the mortal ity rate decreased to 3% and 15%, respectively. If both enrofloxacin and bacteriophages were admi nistered simultaneously, no mortality was

PAGE 20

8 observed. These observations suggest that ba cteriophage alone can considerably reduce the mortality rate of these infections, alt hough not as significantly as enrofloxacin, and that there is a possible synerg istic effect between the mode of actions of the antibiotic enrofloxacin and the two bacteriophages. This study further reinforced the potential of phages as a therapeutic against E. coli infections. Enterococcus faecium is a bacterium commonly f ound in the United States in hospitals and nursing ho mes. Additionally, E. faecium can cause fatal bacteremia and endocarditis, especially in immunocompromised individua ls. Merril and colleagues studied the efficacy of bacteriophage therapy against vancomycin-resistant E. faecium (VRE) (20). A VRE strain was injected intr aperitoneally (i.p) at a concentration of 109 CFU into mice, which caused 100% morbidity within 48 h. When 3 x 108 PFU of VREspecific phage was administered 45 min after the infection, 100% of the infected mice recovered from the infection. If the VRE pha ge was administered upon the first signs of illness, 50% of the infected mice were re scued. In total, the above experiments demonstrated a potential for VRE phage usage in E. faecium infections. Staphylococcus aureus , a bacterium that causes pyoge nic inflammatory diseases, is of great health concern due to its abil ity to cause toxic-shock syndrome and many iatrogenic infections. Furtheri ng the public health concern rega rding this bacterium is the fact that methicillin-resistant S. aureus (MRSA) strains are becoming more prevalent. In an attempt to study the potential of phage th erapy in a mouse model, mice were injected i.p. with 8 x 108 CFU of S. aureus , including methicillin-resistant strains. These doses caused bacteremia and death in greater than 80% of mice within 24 hr and in 100% of mice within 7 days (21). When the mice were injected i.p. with S. aureus -specific phage

PAGE 21

9 between 0 and 24 hr after bact erial infection, the mice were rescued from morbidity. In this particular phage experiment, the phage were effective at multiplicity of infection (MOI) between 0.1 and 200. Soothill also studied th e protective efficacy of S. aureus phage against experimental S. aureus wound infections in rabbits (2 2). Rabbits were injected subcutaneously (s.c.) with 8 x 107 CFU of S. aureus , followed immediately with s.c. injection at the site of the infection with 2 x 109 PFU of S. aureus phage. Of the eight phage-treated rabbits, only one rabbit devel oped an abscess after four days, whereas all eight rabbits in the control group developed abscesses. Furthermore, the number of bacteria was significantly lower in the pha ge-treated group compared to the control group. Rabbits injected with 8 x 107 CFU of S. aureus and then treated with 6 x 107, 6 x 106, or 6 x 105 PFU of phage developed abscesses af ter 4 days with the exception of one of the rabbits that received phage at 6 x 107 PFU. However, the size of the abscesses and the number of bacteria found in the absce sses proportionally decrea sed with increasing concentration of phage. Lastl y, rabbits infected with 5 x 107 CFU of S. aureus , and then treated with 3 x 109 PFU of S. aureus phage administered at 6, 12, or 24 h from the time of infection developed similar sized abscesse s after 4 days and equa l number of bacteria at the site of infection. Th ese experiments suggest that S. aureus phage possesses a prophylaxic property against S. aureus infections, at least in experimentally S. aureus infected rabbits. Soothill and colleague s, however, acknowledged that the protection achieved in these experiments was not as robust as phage experi ments reported with gram-negative bacteria, suggesting a possible obstacle for phage th erapy against grampositive bacterial infections. However, the experiments reported by Merril and

PAGE 22

10 colleagues detailed above (20) and the experi ments reported by Soothill and colleagues demonstrate a potential for phage treatments against S. aureus . Enterococcus faecalis is a facultative anaerobic bacterium that can cause endodontic infections in humans. E. faecalis infections, which are known to be resistant to several antibiotics, are occasionally persiste nt even after antibiotics and other measures are taken to eliminate the bacteria from the r oot canals and dentinal tubules of patients. In an ex vivo experiment studying the effectiveness of phage treatment against E. faecalis infections, the root canals a nd the dentinal tubules of huma n teeth were inoculated with E. faecalis and incubated for varying time periods (23). After the in fection period, they were treated with E. faecalis -specific bacteriophage at MO Is of 0.1, 1, and 10 for 3 h to 72 h. At all of the MOIs, the bacteriophage was able to inhibit the gr owth of the bacteria and/or reduce the number of bacteria. The result of this in vitro experiment implies a possible use of phage as an antimicrobial strategy against end odontic infections of E. faecalis . Clostridium difficile is an enteric bacterium that causes pseudomembranous colitis and death in humans. C. difficile infections commonly emerge as a secondary infection after individuals are treated with antibiotics for a different bacterial infection. Hamsters that were intragastrically injected with 103 CFU of C. difficile would die within 96 h (24). However, the hamsters survived the e xperimental infecti on if treated with C. difficile bacteriophage immediately after the infection with C. difficile , suggesting a possible role of phages against C. difficile infections. Two possible obstacles encountered with pha ge therapy are the release of bacterial endotoxins after phage treatment with detrimen tal effects in patients and the inability of

PAGE 23

11 phages to infect and lyse intr acellular bacterial infections. However, studies may provide solutions to these possible impediments. To address the issue of released endotoxin, a novel approach was undertaken in treating Pseudomonas aeruginosa infections with a non-replicating and non-lytic bacteriophage (25). P. aeruginosa is an opportunistic pathogen known to be resistant to several antibiotics. The ba cterium is one of the leading causes of acquired infections in hospitals and is the leading cause of mortality in people afflicted with cystic fibrosis (26). A P. aeruginosa filamentous phage was genetically engineered to become non-replicating a nd non-lytic. An export protein gene was exchanged with a restricti on endonuclease-methylase cassette gene. Mice that were experimentally infected with three times the minimum lethal dose of P. aeruginosa normally die within 24 h. However, if e ither the genetically engineered or the nonengineered phage was administ ered at an MOI of at leas t 1000, the mice were rescued from the P. aeruginosa infection. When mice were experimentally infected with five times the minimum lethal dose of P. aeruginosa , the survival rate was significantly higher with the non-replicating and non-lyti c phage compared to the lytic phage, 70% compared to 20%, respectively. The higher surv ival rate with the genetically engineered phage suggested that the higher survival rate was due to the decrease in th e release of bacterial endotoxins and, thus, inflammation. This study suggests that the usage of nonreplicating and non-lytic phages for tr eating certain infections, such as P. aeruginosa infections, may be more advantageous than the usage of lytic phages. A novel approach was undertaken to treat an intracellular bacter ial infection with phage (27). An intracellular bacterial inf ection is not usually treatable with phage because of the inability of phage to infiltrate into host cells. Mycobacterium tuberculosis ,

PAGE 24

12 an acid-fast bacterium that infects macropha ges, kills millions of people every year around the world. Although M. tuberculosis infections are treatable with antibiotics, the slow growth of the bacteria makes tr eatment difficult. The treatment of M. tuberculosis usually requires an uncommonl y extended time of antibiotic treatment compared to other bacterial infections. Of furt her concern to public health, an tibiotic-resistant strains of M. tuberculosis are emerging at an alarming rate. A bacterium with similar characteristics to M. tuberculosis , Mycobacterium avium is also a threat to public health. Like M. tuberculosis , M. avium is a slow growing bacteriu m that infects macrophages. M. avium infections occur in individuals afflic ted with acquired immune deficiency syndrome (AIDS), although increasing numbers of individuals not afflicted with AIDS have been reported to be infected with M. avium . Bermudez and colleagues infected a mouse peritoneal macrophage cell line, RAW 264.7, monolayer with M. tuberculosis or M. avium at an MOI of either 1 or 10 (27). The M. tuberculosis or M. avium infected macrophage monolayer was then treated with Mycobacterium smegmatis , a non-virulent mycobacterium, infected with TM4, a lytic phage that produces non-stable lysogens. M. smegmatis was used as a vehicle to transfer TM 4 into the intracellular environment of the macrophage monolayer where M. tuberculosis or M. avium were residing. M. smegmatis containing 7.8 x 107 PFU of TM4 was added for 30 min to the 24 h M. avium -infected macrophage monolayer. After two days, the bacteriophage significantly inhibited the growth of intracellular M. avium . After four days, the bacteriophage significantly redu ced the number of intracellular M. avium by 10-fold. To a 48 h M. avium -infected macrophage monol ayer, treatment with M. smegmatis infected with TM4 for 30 min resulted in a 100-fold decrease in the numb er of intracellular

PAGE 25

13 M. avium . Moreover, M. smegmatis infected with 6.7 x 107 PFU of TM4 was added for 30 min to a 24 h M. tuberculosis -infected macrophage monolayer. Approximate 10-fold and 100-fold decreases in the number of intracellular M. tuberculosis were observed after 2 and 4 days, respectively. This particular study suggests a possible method for phage treatment of in tracellular bacterial inf ection, in particular M. tuberculosis and M. avium , by use of non-virulent strains of bacteria as vehicles to transfer lytic phage into intracellular environments where phage usually would not reside. Although beyond the scope of this research proj ect, it must be not ed that numerous studies have also been repor ted on isolated phage-encode d lytic enzymes in treating bacterial infections in animal models. For example, Fischetti and colleagues successfully treated infections caused by bacteria such as Streptococcus pneumoniae and Bacillus anthracis in animal models with phageencoded lytic enzymes (28-36). Numerous studies have also focused on using bacteriophages as a method for biocontrol in food products. As mentioned ab ove, a couple of experiments studied the potential of phage on E. coli -infected animals (14-19). The effectiveness of phage treatment on E. coli -contaminated steak meat was also studied (37). Steak meat was infected with 2 x 102 CFU of E. col , and the contaminated meat was then treated with a phage cocktail of three phages each at 2 x 108 PFU/ml, corresponding to an MOI of 106. The bacteriophage cocktail completely eliminat ed bacteria in seven of the nine cases. Such observation would suggest a possible role of phages in reducing or eliminating the number of E. coli in contaminated steak meat. Fresh cut fruits and vegetables are more li kely to be contaminated with pathogenic bacteria than whole fruits and vegetables due to the fact th at the peel and rind act as a

PAGE 26

14 protective coat against bacter ial invasion and contaminati on. Leverentz and colleagues studied the usefulness of bact eriophage treatment as a method of biocontrol on freshly cut fruits that were artificially contaminated with Salmonella enteritidis (38). S. enteritidis infecting honeydew melon slices an d apple slices survived at temperatures as low as 5C. Furthermore, at temperatures of 10C and 20C the bacteria increased by 2 logs and 5 logs, respectively. When the sl ices were treated with bacter iophage, the concentration of bacteria was reduced on honeydew melon slices by 3.5 logs at 5C and by 2 logs at 20C. Moreover, the results obtained with bacteriopha ge treatment were better than the results obtained with chemical sanitizers. However, the phage treatments did not work on apple slices. The lack of phage kill ing on apple slices may be due to the acidic pH environment of apple slices. Leverentz and colleagues also studied the usefulness of bacteriophage as a method for biocontrol on freshly cut fruits artificially contaminated with Listeria monocytogenes (39). After the experimental infection of apple and honeydew melon slices with L. monocytogenes , the fruit slices were treated with L. monocytogenes phage. Compared to an untreated group, the phage-treated group re duced the number of bacteria by 2 to 4.6 logs in honeydew melon slices and by 0.4 logs in apples. Once more, the lack of a more pronounced decrease of bacterial levels in appl e slices may have been due to the acidic pH environment of apple slices. Leverent z and colleagues took these experiments one step further. They studied the efficacy of phage treatment accompanied with nisin, a bacteriocin, and nisin alone as a method for biocontrol of L. monocytogenes on experimentally infected fruit slices. When the fruit slices were treated with phage and nisin, the number of bacteria in honeydew melon slices was reduced by 5.7 logs and in

PAGE 27

15 apple slices by 2.3 logs. Moreover, nisi n alone reduced contamination in honeydew melon slices by 2.3 logs and in apple slices by 2 logs. These experiments by Leverentz and colleagues showed the usefulness of b acteriophages as a met hod for biocontrol of S. enteritidis and L. monocytogenes in experimentally infected fruit slices. In addition, the synergistic approach of using both phages and bacterio cins as a method of biocontrol may be useful. Campylobacter jejuni is a bacterium that causes gast rointestinal dis ease in humans. Poultry is a major reservoir of Campylobacter , which is easily spread in a broiler house. In an experiment to assess the efficacy of phage in artificia l infections with Campylobacter , chicken skins were experi mentally infected with Campylobacter at 104 CFU and 106 CFU (40). After 30 min of incubation, the contaminated skins were treated with phage at different MOIs ranging from 0.001 to 100,000 and stored at a temperature of either 4C or -20C. Although no signifi cant decrease was obser ved with the lowest phage treatment of 103 PFU, a significant decrease was observed at the highest phage treatment of 107 PFU. Furthermore, a more signifi cant decrease was observed on phagetreated chicken skins stored at -20C compared to phage-treated chicken skins stored at 4C, probably due to the harsher environmen t confronting the bacteria at the lower temperature. These experiments suggest a pos sible use of phage to lower the number of Campylobacter bacteria on contaminated chicken skin. Of interest to our project, Nakai et al . of Japan studied the efficacy of phage therapy on infectious diseas es in aquaculture (41). Lactococcus garvieae is an opportunistic pathogen of yellowt ail fish that has consider ably damaged the yellowtail aquaculture industry in Japan since its first outbreak in 1974. Nakai et al. demonstrated

PAGE 28

16 that L. garvieae -specific phage pre-treatment can pr otect and rescue yellowtail against experimental infection with L. garvieae . Another pathogen that Nakai et al. studied was Pseudomonas plecoglossicida that causes hemorrhagic asci tes in ayu, a popular fish for sport fishing in Japan (42). P. plecoglossicida phage-impregnated feed protected ayu against experimental infection with P. plecoglossicida . These two studies demonstrated a possible role of phages in treati ng aquatic infec tious diseases. Vibrio vulnificus Vibrio vulnificus Pathogenesis and the Oyster Industry The first known isolatio n and description of Vibrio vulnificus were reported in 1964 by the United States Centers for Disease Cont rol (CDC) (43-45). Ho wever, the isolate was incorrectly identified as Vibrio parahaemolyticus . A distinguishing characteristic of V. vulnificus from other Vibrio species is the ability of V. vulnificus to ferment lactose. Due to this distinction, V. vulnificus was recognized as a distinct member of the Vibrio genus in the mid to late 1970s. V. vulnificus is a gram-negative, curved rod-sh aped bacterium. Containing a single polar flagellum, this motile bacterium is commonly found in estuarine and marine environments. V. vulnificus is an opportunistic pathogen capable of causing serious illness in humans (46,47). V. vulnificus is commonly found in aquatic environments with tropical to subtropical temp eratures. Furthermore, V. vulnificus is a halophilic organism that preferentially grows in a quatic environments with sali nities ranging from 15 parts per thousand (ppt) to 25 ppt (48,49). Nevertheless, the bacterium has been isolated in harsher temperatures and salinity. V. vulnificus has been isolated at temperatures as low as 9C and as high as 31C. V. vulnificus has also been isolated at sa linity levels as high as 34 ppt; however, salinity levels above 25 ppt are de leterious to the growth of the bacterium.

PAGE 29

17 In the U.S., V. vulnificus is commonly found in coastal waters of the Gulf of Mexico coast, although the bacterium is also frequen tly found in estuaries on the west coast, New England (50,51), and in th e Chesapeake Bay (52). In the U.S., V. vulnificus is the leading cause of reported fatalities associated with seafood consumption, particularly oysters ( 53-55). Most human cases of infection associated with V. vulnificus occur in the Gulf Coast region, mainly Florida and Louisiana (56,57). Almost all cases of infection occur betw een the months of April and September (57) with warmer water temper atures, which increase the levels of V. vulnificus in the environment, especially in oyst ers. Ingestion of raw oysters can result in fulminant primary septicemia, especially in individuals with pre-existing health conditions (46,58). These pre-existing conditions include liver disease, such as cirrhosis due to hepatitis or alcoholism, and compromised immune systems, as occur in individuals with AIDS or people that are undergoing ch emotherapy. Another pre-existing condition frequently observed in humans infected with V. vulnificus is hemochromatosis, a condition which results in high saturation of iron binding proteins (59). In all these cases, V. vulnificus causes primary septicemia, symptoms of which include fever and shock, have a mortality rate of approximately 60% , even with antibiotic treatment (60,61). V. vulnificus can also cause wound infections in otherwise healthy individuals (62), with a mortality rate of approximately 25% (45,47,63). These wound infections typically occur when a wound comes in contact with s eawater or raw seafood contaminated with V. vulnificus . Wound infections caused by V. vulnificus usually include symptoms of pain, edema, and erythema. Moreover, wound inf ections may lead to severe necrosis that

PAGE 30

18 may require surgical debridement or amputati on. Furthermore, these infections can lead to septicemia. Due to the health hazards associated w ith the consumption or handling of raw oysters, the general public has become increasingly more apprehensive about consuming oysters. With consumer uneasiness about th e safety of oysters, the oyster industry has suffered economically. In fact in 2003, the state of California ba nned the importation of raw, untreated oysters harvested from the Gulf Coast due to safety concerns (64). This California ban has caused much discontent in the Gulf Coast, especially in the oyster industry due to the economic consequences of such a decision. The economic loss due to the California ban is estimated to be about 20 million dollars a year. As a result, the state of Louisiana is contemplating filing a lawsu it against California fo r what the government of Louisiana and many oystermen consider an unfair ban (65). The California ban of untreated oysters harvested from areas known to harbor V. vulnificus , mostly Louisiana and Florida, is an example emphasizing the devastating economic cost that naturally contaminated oysters have on the oyster industr y, an industry that has been valued at over 95 million dollars (66). A few post-harvest treatments (PHT) have been developed and employed to lower the level of V. vulnificus in oysters (67). One method involves treating oysters with hydrostatic-high pressure as high as 45,000 psi to kill V. vulnificus . In fact, this procedure eliminates othe r potentially harmful bacter ia, including the pathogen V. parahaemolyticus . This particular technique also opens the oyster allowing the quick removal of oyster meat, thus, reducing th e number of employees needed for shucking oysters. Developed by the AmeriPure Oyster Companies, cool past eurization is a second

PAGE 31

19 PHT utilized to decrease the level of V. vulnificus . The oysters are placed in seawater that is mildly heated to a temperature that kills V. vulnificus , and then the seawater is rapidly cooled. Another PHT utilized for reduction of V. vulnificus in oysters is the Individual Quick Freezing (IQF) method. The IQF method freezes oysters for a small time period effectively killi ng the bacteria. Although all of these methods reduce or eliminate V. vulnificus in oysters, the methods also ha ve some disadvantages. All of these methods kill the oyster. Another problem is that oyster enthusiasts complain that the texture and/or taste are changed by such PHT. Another setback encountered with such treatments is the increased economic co st for treating oysters, which is usually passed on to the consumer. Furthermore, PHT methods must d ecrease the level of V. vulnificus in oysters to non-detectable levels (<30 MPN (Most Probable Number)/g) for the product to be considered safe fo r consumption. The FDA approved method for enumeration of V. vulnificus in oysters is the procedure of Tamplin, et al., which is described in Chapter 9 of the FDA Bacteriological Analytical Manual , 7th Edition. Phage-Treatment of Experimentally V. vulnificus Infection of Mice In our laboratory, the po tential of bacteriophages as a therapy against V. vulnificus was studied in a mouse model (68). F our clinical isolates, M06-24/O, VV1009, 2400112, and NSV-5829, and four environmental isolates, MLT403, MLT365, MLT367, and 99-796DP-E7, of V. vulnificus were used. Mice were inje cted i.p. with iron dextran, resulting in an overload of iron levels. After a period of at least 30 min, mice were subsequently injected subcutaneous ly (s.c.) with a lethal dose of V. vulnificus . V. vulnificus -phages, CK-2, 153A-5, and 153A-7, susp ended in phosphate-buffered saline containing gelatin (BSG) were administered intr avenously (i.v.) at varying concentrations and at varying time periods after the inf ection. To observe the efficacy of phage

PAGE 32

20 treatment, five parameters were observed: r ectal temperature, skin lesion score, number of bacteria in the skin lesion, number of bacter ia in the liver, and survival rates. High levels of bacteria in skin lesions represent a localized infection, and high levels of bacteria in the liver represent a systemic in fection. Phage administ ration at a high titer, 108 PFU, significantly lowered the mortality rate of mice compared to a control group. Moreover, the phage treatment significantly lowered the freque ncy of localized V. vulnificus infection, measured by skin lesion scor e and the number of bacteria in the skin lesion, and systemic infection, measured by temperature and the number of bacteria in the liver. Interestingly, a sea salt effect was observed with one of the tested phages, phage 153A-7, on protecting mice against V. vulnificus infections. Since phage 153A-7 requires seawater for lysis of host bacteria, no protection was offered by this phage in the V. vulnificus infection mouse model. Overall, the phages were more effective if administered immediately after the infecti on. Postponing phage administration for more than 3 hr after infection made the phage treatment unsuccessful for the control or elimination of the infection. This study illustrated the potential of phage therapy in local and systemic infections with V. vulnificus . With these positive results, the research project described in this thesis was undertaken to examine the feasibility of treating experimentally and naturally contaminated market oysters with V. vulnificus -specific bacteriophages with the goal of decreasing the number of V. vulnificus infections associated with the consumption of oysters at the source of such infections – the oyster.

PAGE 33

21 Specific Aims Specific Aim 1: Isolation and Charac terization of Bacteriophages. Bacteriophages for V. vulnificus were isolated for greater range of killing of the V. vulnificus strains in our collection and for V. vulnificus strains that were insensitive or only slightly sensitive to the phages. All phages, both newl y isolated and those in our collection were characterized by host range, plaque morphology, and level of infectivity using a soft agar overlay phage typing method and a microtiter phage typing method. Plaque morphology was studied using the so ft agar overlay phage typing assays. Transmission electron microgra phs were obtained of certain phages for categorization by morphology. Specific Aim 2: Establishment of an Ex perimental Oyster Infection Model and Treatment of the Experimentally Inf ected Oysters with Phages to Reduce Numbers of V. vulnificus . An experimental oyster infection model was established for infection of oysters with a substantial level of V. vulnificus contamination. With the establishment of an oyster infection model, the efficacy of pha ge treatment was studied to examine the potential of phage treatments as a met hod to decontaminate infected oysters. Specific Aim 3: Phage Treatment of Naturally Infected Oysters. Naturally infected oysters were treate d with a pool of phages specific for V. vulnificus . The level of V. vulnificus was enumerated by either a selective medium for V. vulnificus or a DNA-hybridization probe specific to V. vulnificus to examine the potential of util izing a pool of V. vulnificus phages to decontaminate naturally infected oysters.

PAGE 34

22 CHAPTER 2 MATERIALS AND METHODS Bacterial Strains, Media, and Growth Methods V. vulnificus clinical strains MO624/O, LL728, 2400112, and VV1009 and environmental isolates MLT365, MLT367, and M LT403 were the primary strain utilized throughout these studies. In addition, we utilized a coll ection of 25 clinical and 25 environmental isolates of V. vulnificus provided by Dr. Angelo DePaola of the Food and Drug Administration (F.D.A.). V. vulnificus FLA042, a spontaneous rifampicin (Rif)resistant mutant of MLT403, and FLA077, a spontaneous Rif-resistant mutant of MO6-24/O, were utilized for the ex perimental oyster infections. V. vulnificus FLA077 containing the marker plasmid pGTR902 was us ed for the segregation assays described below. V. vulnificus was grown in Luria-Bertani brot h containing 0.85% (wt/vol) NaCl (LB-N) or on LB-N plates contai ning 1.5% (wt/vol) agar at 37C. V. vulnificus was occasionally grown in Luria-Bertani broth co ntaining seawater adju sted to a total salt concentration of 0.85% (wt/vol ) (LB-SW) or on LB-SW plat es containing 1.5% (wt/vol) agar at 37C. The salinity of seawater, or iginating from the University of Florida Whitney Laboratory for Marine Bioscience or purchased from the Sigma-Aldrich company, was determined with the use of a re fractometer. Luria-Bertani broth containing sea salts at a final concentration of 20 ppt from the commercially available products Red Sea Salt (Red Sea; Houston,Texas) (LB-RS) a nd Instant Ocean (Aquarium Systems, Inc.; Mentor, Ohio) (LB-IO) were also used. LB-SW or LB-SW pl ates were commonly

PAGE 35

23 required for assays involving phage. Soft agar of both LB -N and LB-SW was prepared containing 0.75% (wt/vol) agar. Soft agar was kept in a dry bath at 42C until immediately before use. V. vulnificus -selective medium (VVM) plates were prepared as described by Cerda-Cuellar et al. (69,70), containing 15 g of D-(+)-cellobiose, 10 g of NaCl, 4 g of yeast extract, 4 g of MgCl2H2O, 4 g of KCl, 40 mg of cresol red, 40 mg of bromothymol blue, 105 U/L of polymyxin B, 105 U/L of colistin methanesulfonate, and 15 g of agar per L of distilled deionized water (ddH2O), with the exception of the adjustment of the pH of the me dium to 8.5 prior to the addition of agar and boiling. All V. vulnificus strains were stored in LB-N containing 35% ( vol/vol) glycerol at 70C. A static overnight starter culture wa s prepared by inoculating 10 mL of LB-N and incubated at room temperature (RT) overnight . A fresh culture was prepared by diluting the static overnight starter culture 1:20 in to LB-N or LB-SW and shaking at 37C. Between 1 h and 1.5 h later, the shaking cu lture normally reached a cell density of approximately 2 x 108 CFU/mL corresponding to late loga rithmic phase of growth. The shaking culture density was determined by optical density read ing at 600 nm (OD600). The optical density reading of 0.39 was previo usly determined to be equivalent to approximately 1.0 x 108 CFU/mL and was used as a conversion factor between OD600 readings and CFU/mL values. Sterile phosphate-buffered sali ne containing 0.1% (wt/vol) gelatin (BSG) was used for dilution of bacteria. Bacteriophage Strains, Isolation, Amplific ation, Purification, and Quantification Bacteriophage Strains Of the 22 bacteriophages in our collec tion, 12 were obtained from Dr. Angelo DePaola of the F.D.A.: 152A-2, 152A-8, 152A-9 152A-10, 153A-5, 153A-7, 153A-8, 154A-8, 154A-9, 108A-9, 110A-7, and 7-8a ( 71,72). The remaining phages were

PAGE 36

24 isolated by our laboratory. Phage CK-2 wa s isolated by Dr. Donna Duckworth from estuarine mud sediment from Cedar Key, FL. Phage CB1 was isolated from mud sediment from Cedar Bay, FL, and phage EJc was isolated from oysters from Cedar Key, FL with the help of Eric Wilkening. Phag es 1a, 2a, 3a, 4a, 4b, AOIA-D, CKIA-B, and CKIF-G were isolated from oysters from Cedar Key, FL, or Apalachicola Bay, FL. Bacteriophage Isolation The phage enrichment and isolation pro cedures were conducted on both estuarine mud sediment and oysters. Oysters harves ted from Cedar Key, FL or Apalachicola Bay, FL were purchased from Northwest Seafood, Gainesville, FL. When using oysters for the phage enrichment and isolation procedur es, oysters were first scrubbed and washed under running deionized water. They were then shucked, and the oyster meat was removed for further use. Seawater at a salin ity of 20 ppt was added to the oyster tissue at 1 mL/g of oyster tissue, and the mixt ure was homogenized using a Stomacher 80 (Tekmar; Cincinnati, Ohio). For all experi ments using seawater, seawater was always used at 20 ppt. Fifty milliliters of either the homogenate or sediment mud contents was mixed with 50 mL of LB-N or LB-SW. The mixture was then inoculated with 1 mL of static overnight starter culture(s) consisting of V. vulnificus strain(s) of interest. The inoculated mixture was sh aken overnight at 37C. After shaking overnight, the mixture wa s centrifuged at 13,776 x g for 10 min at 4C to remove bacterial debris, oyster homoge nate, and mud sediment. The supernatant was then filtered through a 0.2 m filter. The amplified phages in the filtrate were further amplified by mixing 1 mL of the filtrate with 1 mL of static overn ight starter culture of V. vulnificus strain(s) of interest. This mi xture of phage and bacteria was supplemented with 8 mL of LB-N or LB-SW and shaken overnight at 37C. If more than

PAGE 37

25 one V. vulnificus strain was used for the phage enrichment procedure, the filtrate was amplified separately in each of the V. vulnificus strains. After the culture was shaken overnight, it was centrifuged at 13,776 x g for 10 min at 4C, and the resulting supernatant was filtered through a 0.2 m filter. Approximately 20 l of the filtrate was streaked on LB-SW plates, and 4 mL of LB -SW soft agar inoculated with 1 x 107 CFU/mL of log phase V. vulnificus was poured onto LB-SW plates from the least concentrated area towards the most concentrat ed area of the streaked filtrate. The LBSW plate(s) was incubated overnight at 37C. The next day, the clearest and most isolat ed plaques were picked using a Pasteur pipet. The agar plug was placed in 100 l of BSG containing 10 l of chloroform and was stored overnight at 4C. The mixtur e was then centrifuged at 13,776 x g for 10 min at 4C to remove bacterial debris and agar . The supernatant was used for a plaquepurification step. The supe rnatant was streaked on a LB -SW plate, and soft agar containing 1 x 107 CFU/mL of log phase V. vulnificus was poured over the LB-SW plate, as described above. The plate was incubated overnight at 37C, and the clearest and most isolated plaque was again picked using a Past eur pipet. The agar plug was placed in 100 l of BSG containing 10 l of chloroform and stored overnight at a temperature of 4C. The next day, the mixture was centrifuged at 13,776 x g for 10 min at 4C to remove bacterial debris and agar. The resulting supernatant was stored at 4C for the amplification procedure. Bacteriophage Amplification Both broth and plate methods were employe d for phage amplification. Almost all phages in our collection amplified efficien tly using the broth amplification method, although certain phages amplified more effici ently using the plate method. Thus, the

PAGE 38

26 broth amplification technique , which necessitates less time for phage amplification, was utilized as the prefer ential method unless, as detailed a bove, phages required the plate amplification techniqu e for more efficient amplification. Broth phage amplification method For the broth phage amplification method, 1 L of LB-N or LB-SW was inoculated with 5 mL of static over night starter culture of V. vulnificus . The culture was shaken at 37C until the culture reached an optical density corresponding to 2 x 107 CFU/mL. The culture was then infected w ith bacteriophage an MOI of 0.02 and shaken at 37C until a change in the culture was observed from tu rbid to clear, corresponding to the phageinduced lysis of the bacteria. One milliliter of chloroform was then added to the culture, which was shaken at 37C for an additional 15 min to lyse any remaining bacteria. The culture was centrifuged at 13,776 x g for 10 min at 4C to re move bacterial debris, and the supernatant was stored at 4C for the purification procedure. Plate phage amplification method For the plate phage amplification met hod, 10 mL of LB-N or LB-SW broth was inoculated with a static ove rnight starter culture of V. vulnificus at a dilution of 1:20. The culture was shaken at 37C until the culture reached a density of 2 x 108 CFU/mL, determined by OD600, and then 4 x 107 CFU was combined with a volume of phage equivalent to an MOI of 0.5, and the tube wa s vortexed. After a 10 min incubation period at RT, 4 mL of LB-SW soft agar was co mbined with the phage -bacteria mixture, vortexed, and poured onto a LB-SW plate. The soft agar overlay LB-SW plate was incubated overnight at 37C. The soft ag ar was removed using a sterile spatula, suspended in either 5 mL of BSG or seawater containing 20 L of chloroform, and stored at 4C. After at least 4 h, the mixture wa s centrifuged at 13,776 x g for 10 min at 4C to

PAGE 39

27 remove the soft agar and bacterial debris. The supernatant was stored at 4C for the purification procedure. Purification of Phage For purification of phage, 0.2 mL of 20% (wt/vol) polyethylene glycol (PEG) 8000, 2.5 M NaCl was added per mL of phage solution, vortexed, and stored overnight at 4C. The phage mixture was centrifuged at 13,776 x g for 10 min at 4C, and the resulting supernatant was discarded. To remove the remaining PEG in the phage suspension the mixture was centrifuged once more for approxi mately 1 min at 13,776 x g at 4C, and the remaining supernatant was removed using a Pa steur pipet. The pellet was suspended in seawater and filtered through a 0.2 m filte r. The filtrate was stored at 4C. Quantification of Phage For new phage solutions in which phage titers were unknown, the drop titer method was initially utilized to establish an approxi mate titer for each phage . The full plate titer method was utilized to establish a more accurate phage titer for each phage in our collection. Sterile seawater at 20 ppt was used for dilution of phage for all quantification assays. Drop titers A culture of V. vulnificus was grown to 2 x 108 CFU/mL as described above, and 4 x 107 CFU was inoculated into 4 mL of LB-N or LB-SW soft agar . The mixture was vortexed and poured onto a LB-SW plate. Afte rwards, 10 l of serially diluted phage stock was dropped onto the soft agar overlay, and the plates were incubated overnight at 37C. Individual plaques from the drops we re counted, and the approximate titer was calculated.

PAGE 40

28 Full plate titers A culture of V. vulnificus was grown to 2 x 108 CFU/mL, and 4 x 107 CFU was infected with 100 l of serially diluted pha ge and vortexed. After 10 min incubation at RT, 4 mL of LB-N or LB-SW soft agar wa s added and vortexed. The resulting mixture was then poured over a LB-SW plate, and th e plate was incubated overnight at 37C. The next day, the plaques were counted, and titer was calculated. Phage Typing The phage typing studies were undertaken to characterize host range and plaque morphology of each phage. Two methods were utilized for the phage typing studies: the soft agar overlay phage typing method and the microtiter phage typing method, developed as part of these thesis studies. Soft Agar Overlay Phage Typing Method A culture of V. vulnificus was grown to 2 x 108 CFU/mL as described above. After the bacterial culture reached the appropriate density, 4 mL of LB-SW soft agar was inoculated with 4 x 107 CFU and poured onto a LB-SW plate. After pouring the soft agar onto the LB-SW plate, 10 l containing 1 x 106 PFU of phage was dropped on top of the soft agar overlay. The plate was incubated ove rnight at 37C. The next day, the plates were analyzed for plaques, and results were scored as follows: 0no effect, 1faintly turbid confluent plaque, 2<10 small plaques, 3>10 small plaques, 4turbid confluent plaque, 5clear conflu ent plaque. Microtiter Phage Typing Method A culture of V. vulnificus was grown to 2 x 108 CFU/mL as described above. Each well of a 96 well microtiter plate was inoc ulated with the bacteria at varying concentrations from 1 x 106 CFU/mL to 1 x 108 CFU/mL. The wells contained phages at

PAGE 41

29 an MOI of either 1 or 10. The plate wa s read at 630 nm using a Bio-Tek ELx800 plate reader and KC Junior software (Bio-Tek Instru ments, Inc.) at time intervals ranging from every 15 min to every 1 h for a total tim e period of 4 to 24 h. Depending on the experiment, plates were incubated at RT or 37C between readings. Phage Treatment of Experimentally Infected Oysters Experimental Infection of Oysters Oysters were cleaned with running deionize d water and placed in either a small (9.5 L) or large (14.2 L) Nalgene polypropylene au toclavable pan that was aerated with a Maxi-Jet 400 aquarium pump (Marineland; Moorpark, California). The Nalgene polypropylene pans contained eith er 3 L or 6 L of autoclaved seawater for the small or large pans, respectively. After overnight ac climation, for some experiments the oysters were treated with Rif overnight by the addition of concentrated Rif to the seawater with a final concentration of 50 g/mL to kill the natural bacterial flora. After overnight treatment with Rif, the seawater was inoculat ed with log phase V. vulnificus FLA042 or FLA077 at a final concentration of 1 x 106 CFU/mL in the presence of Rif. Before addition of the bact eria into the seawater, the log phase bacterial culture, which was grown as detailed above , was centrifuged at 13,776 x g at RT for 10 min. The resulting pellet was suspended in seaw ater prior to the addi tion of the bacteria to the seawater to prevent the addition of nutrients from the bacterial culture into the oyster culture. Depending on the experiment, oys ters were incubated with bacteria from 0.5 h to 24 h. The V. vulnificus -infected seawater was samp led at varying time points, serially diluted, and plated on LB-N plates containing Rif.

PAGE 42

30 Phage Treatment After the infection period, the infected oys ters were transferred to another Nalgene polypropylene pan containing ster ile seawater. Depending on the number of oysters used for each experiment, a large pan containing 6 L of sterile seawater or a small pan containing 3 L of sterile seawat er was utilized for the experiment. Bacteriophage(s) was then added to the seawater at varying final concentrations ranging from 5 x 107 to 5 x 109 PFU/ml for various periods of time. Harvest of Oysters After the phage treatment period, the shel ls of the oysters were rinsed with deionized water, and the oysters were shucked. The oyster m eat was rinsed with sterile seawater and removed. Seawater was added to the oyster tissue at 1 mL/g of oyster tissue, and the mixture was homogenized, as detailed above. The homogenate was serially diluted and plated on LB-N plates containing Rif. Phage Treatment of Naturally Infected Oysters Phage Treatment Freshly purchased oysters were cleaned under running deionized water and placed in either a small or large Nalgene polypropyl ene pan containing 3 L or 6 L of sterile seawater, respectively, and aerated with an aquarium pump. The oysters were then immediately treated with a pool of phages fo r various periods of time ranging from 12 h to 48 h. Harvest of Oysters Viability of oysters was first confirme d by taping the she ll with a pipet and observing closing. The shell of oysters was ri nsed with deionized water, and the oysters were shucked. The oyster meat was rinsed wi th sterile seawater and removed. Seawater

PAGE 43

31 was added at 1 mL/g of tissue, and the ti ssue was homogenized us ing a stomacher. The homogenate was serially diluted and pl ated on selective medium VVM (69,70) to enumerate V. vulnificus . For some experiments, the oyster homoge nate was plated on LB-N plates for analysis with the V. vulnificus alkaline phosphatase-labeled oligonucleotide probe (VVAP), which was kindly provided by Dr . Anita Wright, for enumeration of V. vulnificus (73). The VVAP probe consisted of a segment of the cytolysin gene that was covalently linked with E. coli alkaline phosphatase. LB-N plates with colonies from the oyster experiment were lifted onto What man 541 filters. The filters were saturated with lysis solution (0.5 M NaOH and 1.5 M NaCl ) for approximately 5 min to lyse the bacterial cells. The filters were dried, and colonies were fixed onto the filter by microwaving at high until the filters appeared dry. The dried filters were treated with 20 mL of 2 M ammonium acetate neutralization solution per filter for 5 min at RT with shaking (80 to 125 RPM). The filters were then rinsed twice with 20 mL of 1X standard saline citrate buffer (SSC) per filter for 20 sec with shaking followed by incubation with 20 mL of 1X SSC containing 10 l of proteinase K (20 mg/mL) per filter in a water bath at 42C for 30 min. The filters were rinsed th ree times with 20 mL of 1X SSC per filter for 10 min with shaking. For every 10 filte rs, 30 mL of hybridization buffer: 0.5% (wt/vol) of bovine serum albumin, 1% (wt/vol ) of sodium dodecyl sulfate (SDS), and 0.5% (wt/vol) polyvininylpyrrolidone in 5X SSC buffer was added, and the filters incubated for 30 min at 56C with shaking (80 to 125 RPM). The filters were then treated with 2 mL of hybridization buffer containing 5l of VVAP per filter, and incubated at 56C with shaking (80 to 125 RPM) for 1 h. The filters were shaken with 20

PAGE 44

32 mL per filter of 1X SSC/1X SDS pre-warmed at 56C for 10 min at RT. The filters were then rinsed five times with 20 mL per filter of 1X SSC for 5 min at RT with shaking. To develop the probe, 20 mL of nitro blue tetrazolium/ 5-bromo4-chloro-3-indolyl phosphate solution, mixed 1 NBT/BCIP tablet per 10 mL of ddH2O, was added per 10 filters. The filters were incubated at RT w ith shaking. After the development of VVAP, the filters were rinsed three times with 20 mL per filter of ddH2O for 10 min at RT. The filters were dried in a dark room overn ight, and purple-brownish colonies were considered positive for V. vulnificus . Bacterial Growth Assays pGTR902 Marker Plasmid V. vulnificus FLA077 containing the marker plas mid pGTR902 ((Starks, A.M., et al., submitted to Mol. Microbiol.) was util ized to study the state of growth of V. vulnificus in experimentally infected oysters. The pGTR902 marker plasmid contains the pBADara C regulatory system. The pBAD-araC system st rictly controls the transcription of the pir gene, which produces the protein that is necessary fo r the replication of plasmids such as pGTR902 containing the R6K origin of replication. In the presence of the inducer arabinose, the AraC protein binds to an activati on region upstream of the pBAD promoter inducing the transcription of pir , and, thus, replication of pGTR902. However, in the absence of arabinose, the transcription of pir is not induced, resulting in nonreplication of pGTR902. Duri ng bacterial replication under non-permissive conditions, the lack of replication of pGTR902 results in the segrega tion of the plasmid from the bacterial population as plasmid-less cells are generated with each division. Consequently, examining the proportion of the recovered bacterial population containing pGTR902 (the growth proportion) after infect ion of oysters rev eals the number of

PAGE 45

33 replications of V. vulnificus during the infection. Convers ely, the level of death of V. vulnificus during an experimental oyster infect ion was calculated as the ratio of bacteria containing pGTR902 at a given time poi nt to the number of bacteria containing pGTR902 at the beginning of the experiment (killing proportion). Once a bacterial cell containing the non-replicating plasmid is kill ed, the plasmid cannot be restored to the population. Hence, a drop in the absolute number of marker plasmid-containing cells with time reveals the rate of bacterial death. Both growth and killing proportions are expressed as log values. Experimental infection of oysters V. vulnificus FLA077(pGTR902) was grown to log phase as described above with the exception that the LB-N broth contained 1% (wt/vol) arabinose and 300 g/mL of kanamycin (Kan). The log phase cultur e of FLA077(pGTR902) was centrifuged at 13,776 x g at 4C for 10 min, and the pellet wa s suspended in sterile seawater. The oysters were infected at a final concentration of 1 x 106 CFU/mL for 7 h. After the 7 h infection with FLA077(pGTR902), the infect ed oysters were transferred to pans containing sterile seawater. Harvest of oysters After varying periods of time in the sterile seawater, the oysters were harvested in the same matter as described above for expe rimental infection of oysters. Homogenates were serially diluted and plated on LB-N a nd LB-N containing 1% (wt/vol) arabinose and 300 g/mL of Kan. Ampicillin Treatment Assay The ampicillin treatment assay was used as an alternative method to examine the growth of V. vulnificus in the artificially infected oyster model. Ampicillin inhibits

PAGE 46

34 enzymes responsible for the synthesis of peptidoglycan and causes death of only replicating bacteria. Hence, the sensitivity or resistance of vibrio s infecting oysters to ampicillin can indicate if the vibrios we re replicating in the oysters. Experimental infection of oysters Oysters were infected as described above with V. vulnificus FLA042 or FLA077 at a final concentration of 1 x 106 CFU/mL in the presence of Rif. After a 7 h infection period, the infected oysters were transferre d to another Nalgene pan containing sterile seawater. The oysters were then treated overnight with ampic illin at concentrations of 10 g/mL, 33 g/mL, or 100 g/mL. Harvest of oysters After the overnight treatment with ampicillin, the oyste rs were harvested in the same manner as described above. The homoge nate was serially diluted and plated on LB-N plates containing Rif. Statistical Analysis For oyster experiments, log CFU/g tissu e was calculated. Mean and standard deviation were calculated for each group. A tw o-tailed student t-test was utilized for statistical analysis between groups. A p valu e of 0.05 or less was considered statistically significant.

PAGE 47

35 CHAPTER 3 RESULTS Rationale for Study The reevaluation of phage therapy, star ting with the studies reported by William Smith and colleagues in the 1980’s and later by other researchers (61), revealed the potential of bacteriophages as an antimicrobial strate gy in numerous real-life circumstances ranging from bacterial infections in patients to biocontrol of pathogens for foods. Furthermore, at a time in medicine in which bacterial pathogens are increasingly developing antibiotic resi stance, phage therapy may afford an alternative to antibiotics for treatment of infectious diseases. The successful results obtained by our laboratory utilizing V. vulnificus -specific phages in treating local and systemic disease caused by V. vulnificus in a mouse model (55) prompted the studies described in this thesis to investigate the potential of V. vulnificus bacteriophages to reduce or eliminate the number of V. vulnificus in both naturally and experimentally contaminat ed oysters. Bacteriophage-mediated decontamination of V. vulnificus in oysters prior to human consumption would result in prevention of the disease by blocking the most common pathway of the bacteria for infection in humans. This would, in turn, ha ve a positive effect on the shellfish industry. Specific Aim 1: Isolation and Charac terization of Bacteriophages. The phage typing studies were undertaken to characterize hos t range, level of infectivity, and plaque morphology of each pha ge in our collection. Two methods were utilized for the phage typing studies: the sta ndard soft agar overlay and the microtiter

PAGE 48

36 method, which was developed as part of these thesis studies. These methods helped us determine which phages to utilize in our V. vulnificus bacteriophage treatment experiments with experimentally and naturally V. vulnificus -infected oysters. Moreover, using the results of the phage typing experime nts, attempts were undertaken to isolate phages infective to strains of V. vulnificus that were insensitive or only slightly sensitive to the initial phages in our collection. Transmission electron micrographs (TEM) were obtained of most of the newl y isolated phages. Soft agar Overlay Phage Typing Method Using the standard soft agar overlay pha ge typing method, we examined the host range and the level of infectivity for each phage in our collection to each V. vulnificus isolate in our collection. The standard soft agar overlay phage typing experiments were performed as described in the Materials and Methods section. Briefly, 10 L of phage solution containing 1 x 106 PFU was dropped onto a soft ag ar overlay containing 1 x 107 CFU/mL of a V. vulnificus strain grown to log phase. After overnight incubation, plaques, if any, were visually examined and scored accordingly to the following scheme: 0 no effect, 1 faintly turbid confluent plaq ue, 2 <10 clear small plaques, 3 >10 clear small plaques, 4 turbid confluent pl aque, 5 clear confluent plaque. A score of 5 was assigned to a clear conf luent plaque caused by the phage infecting and lysing every cell of the bacteria. A scor e of 4 was assigned to a turbid confluent plaque, indicating that the phage could infect the strain with high efficiency, but that resistant bacterial cells either were pr esent in the initial population (i.e., the V. vulnificus strain was a mixed population) or arose afte r infection with phage (i.e., the phage was temperate and generated lysogens). Although it is possible that the turbid confluent plaques were caused by temperate phages, we obt ained results with full plate titer assays

PAGE 49

37 that suggested otherwise. For example, when full plate titer assa ys of each phage were performed, none of the plaques had a “bulls-e ye” appearance, a turb id center surrounded by a clear ring, which is usually associated with temperate phages. The issue regarding the turbid confluent plaques was not pursued any further. The scores of 3 and 2 corresponded to the number of clear plaques in the spot of the phage drop. These are somewhat confusing situations because 106 PFU were dropped onto the bacteria, yet only a countable number of plaques formed. This result suggests that the phage infected the particular V. vulnificus strain with very low efficiency. The inability of a phage to produce a clear confluent plaque on a particular V. vulnificus strain may indicate limiting factors by either the phage or bacterium. Limiting factors su ch as bacterial restriction modification systems, inefficient adsorption, or defective lysis of the bacterium by the phage are a few examples that may explain the results obse rved with scores between 2 and 3. A slightly turbid confluent plaque wa s given the lowest possible score of 1 due to the inability of confirming the confluent plaque as an actual turbid confluent plaque or merely an artifact caused by the addition of th e phage solution onto that spot of the soft agar containing bacteria. In Table 3-1a-c, the highest score of two soft agar overlay phage typing experiments is presented for each strain of V. vulnificus that we obtained from Angelo DePaola with each strain of bacteriophage in our collection, and the number of phages that affected each strain of V. vulnificus is presented as well. To guide choices for which phages would be suitable for use in treati ng oysters, we focused on a score of 5. However, to be able to more sensitively compare the host ranges of the phages, a lower score of 4 was also considered. If a very stringent score of 5 was used for scoring

PAGE 50

38 sensitivity, 36 of the 50 strains of V. vulnificus (72%) were sensitive to at least one phage in our collection. Us ing a score of 5, V. vulnificus FLA104, FLA107, FLA108, FLA110, FLA111, FLA112, FLA113, FLA118, FLA 119, FLA121, FLA126, FLA127, FLA128, and FLA138 were insensitive to all of the bacteriophages in our collection. If any confluent plaque, a score of 4 or 5, was used for consideration of sensitivity, 39 of the 50 V. vulnificus strains (78%) were sensitive to at least one phage, allowing the addition of V. vulnificus FLA111, FLA113, and FLA128 as sensitive to at least one of the phages in our collection. In Table 3-1d, the highest score of two so ft agar overlay phage typing experiments is presented for the seven V. vulnificus strains commonly used in our laboratory, MO6-24/O, 2400112, VV1009, LL728, MLT365, MLT367, and MLT403, with each strain of bacteriophage in our collection. Th e soft agar overlay phage typing experiments revealed that these seven strains of V. vulnificus developed complete lysis (score of 5) to at least three of the bacter iophages in our collection. If these seven strains of V. vulnificus are included with the 50 FDA strains, 75% (43 of the 57 strains) of the V. vulnificus strains had a score of 5 and 81% (46 of the 57 strains) had a score of at least 4. The host range of each phage in our collec tion was also determined using the soft agar overlay phage typing method using a score of 5 to define sensitivity (Table 3-2). Phage CK-2 had the narrowest host ra nge, infective to only one of the 57 V. vulnificus strains (2%), MLT403. Phages 154A-9, 7-8A , and CB-1 also had narrow host ranges, being infective to only four of the 57 strains each (7%). On the ot her end of the spectrum of host range using a cutoff scor e of 5, phage 3a had the broa dest host range, infecting 35

PAGE 51

39 of the 57 V. vulnificus strains (61%), and bacteri ophages 153A-7 and 108A-9 were infective to 33 out of the 57 st rains (58%). However, using a score of 4 as a cutoff, the host ranges increased for these and other phages; phage 3a continued to have the broadest host range, infecting 39 of the 57 (68%) V. vulnificus strains; phage 153A-7 infected 37 (65%); phages 153A-8, 154A-8, and 108A-9 inf ected 36 (63%) strain s (data not shown) . Table 3-1. Level of infec tivity of each strain of V. vulnificus to each strain of V. vulnificus -bacteriophage. Two soft agar overl ay assays were performed on 57 strains of V. vulnificus using 19 phage strains. Scoring scheme was as follows: 0 no effect, 1 faintly turbid confluent plaque, 2 <10 clear small plaques, 3 >10 clear small plaques, 4 turbid confluen t plaque, 5 clear confluent plaque. The highest score of the two soft agar overlay phage typing experiments is presented. Grey shad ed boxes represent a difference greater than 2 between the shown score and the score not shown. The bottom row labeled “total” indicates the number of phage strains that scored a 5 on a particular V. vulnificus strain. A Strains of V. vulnificus FLA 101 FLA 102 FLA 103 FLA 104 FLA 105 FLA 106 FLA 107 FLA 108 FLA 109 FLA 110 FLA 111 FLA 112 FLA 113 FLA 114 FLA 115 FLA 116 FLA 117 152A-2 0 3 0 0 3 0 0 0 0 0 0 0 0 0 0 0 4 152A-8 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 5 152A-9 0 3 0 0 2 0 0 0 0 0 0 0 0 0 0 0 5 152A-10 0 2 0 0 5 0 0 0 0 0 0 0 0 0 0 0 5 153A-5 0 1 4 0 5 4 1 0 0 0 0 0 4 0 0 0 4 153A-7 5 5 4 1 4 5 0 0 5 0 4 0 0 5 5 5 5 153A-8 5 5 5 1 4 4 0 0 4 0 4 0 0 5 5 5 5 154A-8 4 5 4 0 4 4 0 0 4 0 4 0 0 5 5 5 0 154A-9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 5 108A-9 5 5 5 0 4 4 0 0 5 0 4 0 0 5 5 5 5 110A-7 4 4 0 0 0 4 0 0 4 0 0 0 0 0 0 3 5 7-8a 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 CK-2 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 CB-1 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 EJc 0 5 0 0 0 0 0 0 0 0 0 0 0 0 0 5 5 1a 0 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 3a 5 0 5 0 5 5 0 0 5 0 4 0 0 5 5 0 5 4a 4 2 4 0 5 0 0 0 4 0 0 0 0 0 0 0 5 Strains of V. vulnificus-bacteriophage 4b 0 2 0 0 2 0 0 0 0 0 0 0 0 0 0 0 5 Total 4 5 3 0 4 2 0 0 3 0 0 0 0 5 5 5 14

PAGE 52

40 Table 3-1. Continued B Strains of V. vulnificus FLA 118 FLA 119 FLA 120 FLA 121 FLA 122 FLA 123 FLA 124 FLA 125 FLA 126 FLA 127 FLA 128 FLA 129 FLA 130 FLA 131 FLA 132 FLA 133 FLA 134 152A-2 0 0 0 0 0 0 0 5 0 0 0 4 0 0 0 0 0 152A-8 0 0 0 0 0 3 4 5 0 0 0 5 0 1 0 5 0 152A-9 0 0 0 0 0 3 0 5 0 0 0 4 0 1 0 5 0 152A-10 0 0 0 0 0 3 0 5 0 0 0 3 0 2 0 3 0 153A-5 0 0 0 0 0 5 5 5 0 0 0 5 4 1 0 1 5 153A-7 0 0 0 0 0 5 5 5 0 0 1 5 5 5 1 5 5 153A-8 0 0 4 0 1 0 5 5 0 0 0 5 5 5 0 5 5 154A-8 0 0 4 0 1 5 5 5 0 0 0 5 5 5 1 5 5 154A-9 0 0 4 0 0 0 0 5 0 0 0 0 0 2 0 2 0 108A-9 0 0 0 0 1 5 5 5 0 0 1 5 5 5 1 5 5 110A-7 0 0 4 0 0 3 5 5 0 0 0 4 5 2 0 5 0 7-8a 0 0 0 0 0 0 0 5 0 0 0 0 0 0 0 2 0 CK-2 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 2 0 CB-1 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 EJc 0 0 0 0 0 0 0 5 0 0 0 5 0 5 0 0 0 1a 0 0 0 0 4 3 0 5 0 0 0 0 2 1 0 0 0 3a 0 0 5 0 5 5 4 5 0 0 4 5 5 1 5 5 5 4a 0 0 4 0 4 5 0 5 0 0 0 4 4 0 1 5 5 Strains of V. vulnificus-bacteriophage 4b 0 0 0 0 0 3 0 5 0 0 0 3 5 0 0 0 0 Total 0 0 1 0 1 6 6 17 0 0 0 8 7 5 1 9 7

PAGE 53

41 Table 3-1. Continued C Strains of V. vulnificus FLA 135 FLA 136 FLA 137 FLA 138 FLA 139 FLA 140 FLA 141 FLA 142 FLA 143 FLA 144 FLA 145 FLA 146 FLA 147 FLA 148 FLA 149 FLA 150 152A-2 1 3 1 0 0 0 0 4 0 0 0 0 5 0 4 4 152A-8 1 5 1 0 5 0 4 5 0 4 1 0 5 0 5 5 152A-9 1 3 0 0 3 0 4 4 0 4 1 0 4 0 4 4 152A-10 2 2 0 0 2 0 0 3 0 2 0 0 5 0 3 3 153A-5 5 5 1 0 5 1 0 0 5 0 5 5 4 4 5 4 153A-7 5 5 1 0 5 5 5 5 5 5 5 0 5 4 5 5 153A-8 5 5 1 0 5 5 5 5 5 5 5 0 5 4 5 5 154A-8 1 5 0 0 5 5 5 4 5 5 5 0 5 4 5 5 154A-9 1 2 0 0 2 0 0 2 0 0 0 0 5 0 3 3 108A-9 1 5 3 0 5 5 5 5 5 5 5 0 5 5 5 5 110A-7 1 1 0 0 3 0 5 4 0 5 0 0 5 0 5 3 7-8a 1 2 0 0 0 0 0 0 0 0 0 0 5 0 4 0 CK-2 0 2 0 0 2 0 0 0 0 0 0 0 3 0 0 0 CB-1 0 0 0 0 0 0 0 2 0 0 0 5 0 0 2 0 EJc 0 0 0 0 0 0 0 2 0 0 0 0 1 0 3 5 1a 1 2 0 0 0 0 0 0 0 2 0 0 5 0 5 2 3a 1 5 5 3 5 5 5 5 5 5 5 5 5 1 5 4 4a 1 5 5 0 5 5 5 5 4 5 0 4 5 0 5 4 Strains of V. vulnificus-bacteriophage 4b 1 3 2 0 0 0 0 5 0 3 0 0 5 0 3 2 Total 3 8 2 0 8 6 7 7 6 7 6 3 14 1 10 6

PAGE 54

42 Table 3-1. Continued D Strains of V. vulnificus M06-24/O 2400112 VV1009 LL728 MLT365 MLT367 MLT403 152A-2 5 3 5 5 0 0 0 152A-8 5 1 5 0 0 0 0 152A-9 5 3 5 3 0 0 0 152A-10 5 2 5 0 0 0 0 153A-5 5 0 4 5 1 1 0 153A-7 5 5 5 2 5 5 0 153A-8 5 3 5 3 5 5 0 154A-8 5 5 5 0 5 5 0 154A-9 5 0 4 0 0 0 0 108A-9 5 5 5 0 5 5 0 110A-7 5 2 5 0 1 1 2 7-8a 5 2 5 0 0 0 2 CK-2 2 2 2 0 0 0 5 CB-1 5 2 2 5 0 0 3 EJc 5 5 5 3 0 0 0 1a 5 3 5 2 0 0 0 3a 5 1 5 5 5 5 5 4a 5 0 5 4 4 4 5 Strains of V. vulnificus-bacteriophage 4b 5 2 5 2 0 0 0 Total 18 4 15 4 5 5 3

PAGE 55

43 Table 3-2. Host range for each strain of V. vulnificus -bacteriophage. The soft agar overlay method was performed on 57 V. vulnificus strains using 19 phage stains. The number of V. vulnificus strains that exhibited a score of 5 with a particular phage strain is shown. Strains of V. vulnificus bacteriophage Number of strains of V. vulnificus sensitive to strain of V. vulnificus bacteriophage 152A-2 5 152A-8 12 152A-9 5 152A-10 6 153A-5 15 153A-7 33 153A-8 30 154A-8 27 154A-9 4 108A-9 33 110A-7 11 7-8a 4 CK-2 1 CB-1 4 EJc 10 1a 6 3a 35 4a 18 4b 7 Effect of Sea Salts on the Infectivity of Certain Phages While working with V. vulnificus bacteriophages, it came to our attention that the use of sea salts in LB-SW and of seawater as a diluent caused a considerable effect on the infectivity properties of certain phages. The use of sea salts signifi cantly increased the number of plaques during quantification assa ys of certain phages. The presence of divalent cations Mg+2 and Ca+2 has been previously reported to be important for efficient attachment and, thus, infection of certai n phages to bacteria (62). Moreover, Mg+2 and Ca+2 are present in seawater, the natural envi ronment of the isolated phages in our collection.

PAGE 56

44 Phage CB-1 was one of those phages for whic h sea salts had an enhancing effect on the infection and lysis of V. vulnificus LL728. To examine if Mg+2 and Ca+2 in sea salts explained the sea salt effect, we used LB-N, LB-N supplemented with 0.05 M MgCl2, LB-N supplemented with 0.038 M MgCl2 and 0.15 M CaCl2, and LB-SW soft agar (Fig. 3-1). A stock of phage CB-1 was used to infect V. vulnificus LL728 using the different soft agars poured onto LB-N plates, and the e fficiency of forming plaques was measured. The use of LB-SW soft agar enabled more pla ques to form than did LB-N soft agar (11.5 0.01 log PFU/mL and 9.5 0.15 log PFU/mL, respectively, (p=2.1 x 10-5). The supplementation of LB-N soft agar with 0.05 M MgCl2 (10.7 0.15 log PFU/mL) also enabled more plaques to form than w ith LB-N soft agar alone (p=5.7 x 10-4). Furthermore, the supplementation of LB-N soft agar with 0.038 M MgCl2 and 0.15 M CaCl2 (11.4 0.03 log PFU/mL) resulted in a nu mber of plaques comparable to the number of plaques on LB-SW soft ag ar. These results suggest that Mg+2 and Ca+2 in seawater are required by certain pha ges for efficient infection of V. vulnificus . The positive effect of using sea salts was also detected when phages were diluted with seawater instead of BSG and added into LB-N soft agar. No differences were observed on plaq uing efficiency when media were supplemented with seawater purchased from Sigma, obtained from the Whitney Laboratory, or reconstituted from dehydrated sea salts (Red Se a Salts). However, the use of Instant Ocean sea salts in the medium or as diluent for phages did not cause the same increase in infectivity of certa in phages and on the growth of V. vulnificus (data not shown). The fact that Instant Ocean sea salt s is synthetically formulated to mimic the chemical composition of natural seawater ma y explain its relative ineffectiveness for

PAGE 57

45 plaquing assays and on the growth of V. vulnificus . Therefore, seawater from the Whitney Laboratory was used for all subsequent experiments involving V. vulnificus specific bacteriophages. 0 1 2 3 4 5 6 7 8 9 10 11 12 LB-NLB-N + 0.05 M MgCl2LB-N + 0.15 M CaCl2 + 0.038 M MgCl2 LB-SWlog PFU/mL* * * # Figure 3-1. Effect of Mg+2 and Ca+2 on lysis of V. vulnificus LL728 by phage CB-1. LB-N, LB-N supplemented with 0.05 M MgCl2, LB-N supplemented with 0.038 M MgCl2 and 0.15 M CaCl2, and LB-S W soft agars were inoculated with LL728, which was infected with phage CB-1 and poured onto a LB-N plate. Triplicates of each were incuba ted overnight at 37C, and plaques were enumerated the next day. Bars represent mean standard deviation. * significant difference compared to LB -N (LBN + 0.5 M MgCl2, p=2.1 x 10-5; LBN + 0.15 M CaCl2 + 0.038 M MgCl2, p=5.7 x 10-4; LB-SW, p=2.7 x 10-5). # significantly different compared to LBN + 0.15 M CaCl2 + 0.038 M MgCl2 (p=1.6 x 10-3) or LB-SW (p=8.4 x 10-4). Development of a Microtiter Phage Typing Method A microtiter phage typing method was develo ped as an alternative technique to the soft agar overlay phage typing method to quantify the infectivity of phages to V. vulnificus . The microtiter method utilized the pl atform of a micro titer plate containing 96 wells, each with a bacterial culture infected with a phage. The optical density of the culture in each well was read using a micr otiter plate reader. One benefit of the

PAGE 58

46 microplate method was that multiple strains of V. vulnificus could be examined in various combinations with different phages. Such an experiment would necessitate numerous Petri dishes when using the soft agar ove rlay phage typing method. The small scale procedure of a microplate also required signi ficantly less volumes of media, phages, and bacterial cultures. Moreover, the sensitivity of a microplate reader permitted the determination of infectivity in considerab ly less time compared to the overnight incubation period of time required for the soft agar overlay phage typing method. Prior to using the microtiter phage typing method with the seven commonly used V. vulnificus strains, optimization of parameters such as bacterial concentration and MOI was done for the most effective visualization of lysis. Since seawater had an effect on infectivit y of certain bacteriophages, a microtiter experiment was performed to study the effect of LB-N and LB-SW on the infectivity of phages 152A-2, 153A-7, 153A-8, and 154A-8 on V. vulnificus strains M06-24/O, 2400112, LL728, and VV1009 (Figs. 2a-d). V. vulnificus was grown to exponential phase in the appropriate medium, diluted to 1 x 106 CFU/mL, and infected with phages at an MOI of 10. Each well was then supplemen ted with LB-N or LB-SW to a final volume of 200 l. The optical densities of the resul ting mixtures were measured every 1 h for 4 h. The uninfected control groups in e ither LB-SW or LB-N of any of the V. vulnificus strains grew to significantly high er densities every 1 h for the first 3 h, initially starting at an OD of approximately 0.04. While the uni nfected control group in LB-SW continued growing to significantly highe r densities at 4 h reaching ODs of approximately 0.50, the uninfected control group in LB-N began to level off at ODs of approximately 0.37. The phage-infected groups that were negative for pha ge infection followed similar patterns as

PAGE 59

47 their corresponding uninfected control group. However, the phage-infected groups that were positive for phage infection exhibited a decrease in density compared to their corresponding control group usually within the first couple of hours. That density was usually sustained for the rest of the expe riment, although occasionally further decreases in density were observed fo r some of the phages. Significant lysis of M06-24/O in LB-SW was observed with bacteriophages 152A2, 153A-7, 153A-8, and 154A-8 (range of OD630 0.04 to 0.26, p=0.002 to 0.011) compared to the uninfected control group (OD630 0.50) at 4 h. However, lysis of M0624/O in LB-N with some of the same bact eriophages was not as robust; although by 4 h some of the phages had begun to cause a decr ease in bacterial OD. The observation that LB-SW positively affected the ability of phage s to infect and lyse M06-24/O was also observed with the other V. vulnificus strains. Particularly, the ability of phage 152A-2 to infect and lyse M06-24/O was greater w ith LB-SW compared with LB-N. The remarkable effect of LB-SW on phage 152A-2 infection was further confirmed using V. vulnificus LL728 (Fig. 3-2c) and VV1009 (Fig. 3-2d). LB-SW also caused significantl y higher growth of some V. vulnificus strains in the microtiter experiments. At 4 h, strain M 06-24/0 grown in LB-SW had a significantly higher yield compared to the LB-N (OD630 0.50 0.03 vs. 0.37 0.001, p=0.024). Furthermore, between 3 and 4 h the bacteria ceased to grow in LB-N, while in LB-SW the bacteria were still growing to higher dens ities. The growth-stimulatory effect of LBSW on V. vulnificus was observed with other strains, with an average 24% increase in yield compared to LB-N.

PAGE 60

48 Figure 3-2. Effect of LB-N and LB-SW on the infectivity of phages to V. vulnificus. Using the microtiter method, the effect of LB-N and LB-SW on the infectivity of phages 152A-2, 153A-7, 153A-8, and 154A-8 was examined on V. vulnificus strains: a) M06-24/O, b) 2400112, c) LL728, and d) VV1009. The control group was not inf ected with phage. Duplicates were performed for each bacteria-phage combination. Bars represent mean standard deviation. * significant difference compared to control group (p= 4.2 x 10-4 to 0.034); # significant difference between the corresponding control groups in LB-N and LB-SW (p=0.024 to 0.031). 0.0 0.1 0.2 0.3 0.4 0.5 0.6 Control152A-2153A-7153A-8154A-8Control152A-2153A-7153A-8154A-8 MO6 (LBN)MO6 (LBSW) OD 600 nm 0 h r 1 h r 2 h r 3 h r 4 h r * * * * # **A

PAGE 61

49 0.0 0.1 0.2 0.3 0.4 0.5 0.6 Control152A-2153A-7153A-8154A-8Control152A-2153A-7153A-8154A-8 2400112 (LBN)2400112 (LBSW) OD 630 nm 0 hr 1 hr 2 hr 3 hr 4 hr 0.0 0.1 0.2 0.3 0.4 0.5 0.6 Control152A-2153A-7153A-8154A-8Control152A-2153A-7153A-8154A-8 LL728 (LBN)LL728 (LBSW) OD 630 nm 0 hr 1 hr 2 hr 3 hr 4 hr# * Figure 3-2. Continued B C

PAGE 62

50 0.0 0.1 0.2 0.3 0.4 0.5 Control152A-2153A-7153A-8154A-8Control152A-2153A-7153A-8154A-8 VV1009 (LBN)VV1009 (LBSW) OD 630 nm 0 hr 1 hr 2 hr 3 hr 4 hr* * * Figure 3-2. Continued The growth-stimulating effect of sea salts on the growth of V. vulnificus led us to study the effect of using SW as a diluent compared to BSG, which contains no sea salts, on the growth of V. vulnificus in either a LB-N or LB-S W medium using the microtiter phage typing method. Cultures of V. vulnificus M06-24/O, MLT403, and 2400112 were grown to exponential phase, dilu ted in either SW or BSG, a nd inoculated in either LB-N or LB-SW to a final concentration of either 1 x 105 CFU/mL or 1 x 106 CFU/mL. The bacterial cultures were read every 1 h for 6 h (Figs. 3a-c). As seen in the previous experiment (Figs. 2a-d), V. vulnificus consistently grew to higher densities in LB-SW than in LB-N, and the bacteria ceased to grow in LB-N, which was also seen in the previous experiment. Utilizing SW instead of BSG as diluent also caused an increase in densities of V. vulnificus in LB-N. Thus, SW was a mo re effective diluent for the microtiter experiments compared with BSG. D

PAGE 63

51 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1E5 CFU/mL (BSG) 1E6 CFU/mL (BSG) 1E5 CFU/mL (SW) 1E6 CFU/mL (SW) 1E5 CFU/mL (BSG) 1E6 CFU/mL (BSG) 1E5 CFU/mL (SW) 1E6 CFU/mL (SW) M06-24/O (LBN)M06-24/O (LBSW) OD 630 nm 0 h 1 h 2 h 3 h 4 h 5 h 6 h* * * * # # # Figure 3-3. Effect of using SW as diluent on the growth of V. vulnificus in either LB-N or LB-SW. V. vulnificus cultures were diluted to a final concentration of either 1 x 105 CFU/mL or 1 x 106 CFU/mL with diluent SW or BSG and supplemented with either LB-N or LB-S W. Triplicates were performed for each bacteria-phage combination. Bars represent mean standard deviation. * significant difference compared to the corresponding BSG diluted group (p=4.4 x 10-6 to 0.030); # significant difference compared to the corresponding LB-N group (p=5.3 x 10-6 to 0.047). A

PAGE 64

52 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1E5 CFU/mL (BSG) 1E6 CFU/mL (BSG) 1E5 CFU/mL (SW) 1E6 CFU/mL (SW) 1E5 CFU/mL (BSG) 1E6 CFU/mL (BSG) 1E5 CFU/mL (SW) 1E6 CFU/mL (SW) MLT 403 (LBN)MLT 403 (LBSW) OD 630 nm 0 h 1 h 2 h 3 h 4 h 5 h 6 h* * * * # # 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 1E5 CFU/mL (BSG) 1E6 CFU/mL (BSG) 1E5 CFU/mL (SW) 1E6 CFU/mL (SW) 1E5 CFU/mL (BSG) 1E6 CFU/mL (BSG) 1E5 CFU/mL (SW) 1E6 CFU/mL (SW) 2400112 (LBN)2400112 (LBSW) OD (630 nm) 0 h 1 h 2 h 3 h 4 h 5 h 6 h* * * *# # # # Figure 3-3. Continued B C

PAGE 65

53 As detailed for the soft agar overlay phage typing assays, an inhibitory effect was observed on phage infectivity and V. vulnificus growth using Instant Ocean sea salts compared to other sources of seawater. To ex amine if this effect also occurred using the microtiter phage typing method, the growth of V. vulnificus FLA042 and infectivity of phage CK-2 was studied using LB-SW, LB -IO, and LB-RS (Fig. 3-4). FLA042 was grown to exponential phas e, diluted to 2 x 106 CFU/mL, and infected with phage CK-2 at MOIs of 0.1, 1, 10, and 100. The resulting mixtures were measured at 0.5 h, 1 h, 2 h, 3 h, 4 h, and 6 h. No significant differences were observed for changes in OD for CK-2 infectivity related to the usage of the different sea salts. However, th e bacterial density of the control group in LB-IO was significantly lower compared to the control groups in LB-SW and LB-RS (at 4 h, LB-SW: OD630 0.47 0.01; LB-IO: OD630 0.44 0.01, p=0.006). These results indicated that the us e of LB-SW or LB-RS was optimal for the microtiter experiments. Therefore, LB-SW was used for the rest of the microtiter experiments because, of the three sea salt fo rmulations, it most likely reproduces the environmental conditions encountered by V. vulnificus and its phages in nature. We next determined the appropriate concentration of V. vulnificus M06-24/O for effective lysis of bacteria by phages in the microtiter phage typing method. We aimed for a concentration of V. vulnificus that exhibited exponen tial growth throughout the experiment and a density high enough to be detected with the microtiter plate reader at the earliest time point of th e experiment. To make the as say more rapid, we looked for the earliest detection of lysis. Starting with concentrations of exponential phase grown M06-24/O of 1 x 105, 1 x 106, 3.2 x 106, and 1 x 107 CFU/mL, growth was measured every 1 h for 4 h (Fig. 3-5). At 1 x 107 CFU/mL M06-24/O was detectable and grew

PAGE 66

54 mostly for the first 2 h, so phage infectivity a nd lysis of bacteria should also have been detectable over this time period. Thus, 1 x 107 CFU/mL of M06-24/O was chosen as the most probable efficient concentration for the microtiter phage infection experiments. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0123456 Time (h)OD 630 nm LBSW LBSW (100) LBSW (10) LBSW (1) LBSW (0.1) LBIO LBIO (100) LBIO (10) LBIO (1) LBIO (0.1) LBRS LBRS (100) LBRS (10) LBRS (1) LBRS (0.1)# * Figure 3-4. Comparison of LB-SW, LB-IO , and LB-RS on growth of FLA042 and infectivity of phage CK-2. Us ing the microtit er method, 2 x 106 CFU/mL of FLA042 was added to each well suppleme nted with either LB-SW, LB-IO, or LB-RS and infected with CK-2 at MOIs of 0.1, 1, 10, and 100. Duplicates were performed for each bacteria-phage combination. OD630 of the wells was read at 0.5 h, 1 h, 2 h, 3 h, 4 h, and 6 h. * significant difference compared to uninfected LB-SW (p=0.006); # sign ificant difference compared to uninfected LB-SW or LB-RS (LB-SW, p=0.001; LB-RS, p=0.040). Next, M06-24/O at concentrations of 1 x 106, 5 x 106, and 1 x 107 CFU/mL was infected with phages 7-8a, 153A-7, 153A-8, and 1a , and the plates were read at 1 h, 2 h, 3 h, 4 h, and 6 h (Fig. 3-6). As expected, a tr end of bacterial lysis caused by the phages among the different groups was observed, regard less of the initial c oncentration used for the assay. Every phage significantly lysed M 06-24/O with similar patterns of time of

PAGE 67

55 lysis and level of lysis. However, a larg er separation between the OD reading of the control group and the corres ponding phage groups was observe d at the highest bacterial concentration used for this assay, 1 x 107 CFU/mL. The larger separation between the treated and untreated groups at the 1 x 107 CFU/mL allowed for optimal visualization of phage lysis. Therefore, 1 x 107 CFU/mL was considered the optimal bacterial concentration to observe phage -induced bacterial lysis in th e shortest period of time. 0.0 0.1 0.2 0.3 0.4 0.5 01234 time (h)OD 630 nm 1E5 CFU/ml 1E6 CFU/ml 3.2E6 CFU/ml 1E7 CFU/ml Figure 3-5. Comparison of V. vulnificus M06-24/O growth at various initial concentrations. In a microtiter assa y, M06-24/O was added to wells at 1 x 105, 1 x 106, 3.2 x 106, and 1 x 107 CFU/mL. Growth was measured every 1 h for 4 h.

PAGE 68

56 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0123456 time (h)OD 630 nm M06 (1E6 CFU/ml) Control M06 (1E6 CFU/ml) 7-8a M06 (1E6 CFU/ml) 153A-7 M06 (1E6 CFU/ml) 153A-8 M06 (1E6 CFU/ml) 1a M06 (5E6 CFU/ml) Control M06 (5E6 CFU/ml) 7-8a M06 (5E6 CFU/ml) 153A-7 M06 (5E6 CFU/ml) 153A-8 M06 (5E6 CFU/ml) 1a M06 (1E7 CFU/ml) Control M06 (1E7 CFU/ml) 7-8a M06 (1E7 CFU/ml) 153A-7 M06 (1E7 CFU/ml) 153A-8 M06 (1E7 CFU/ml) 1a* #@ @ Figure 3-6. Effect of diff erent concentrations of V. vulnificus M06-24/O on phage infection. In a microtiter assay, initial concentrations of 1 x 106, 5 x 106, and 1 x 107 CFU/mL of M06-24/O were inf ected with phages 7-8a, 153A-7, 153A-8, and 1a at an MOI of 10. Duplicates were performed for each bacteria-phage combination. Plates we re read 1 h, 2 h, 3 h, 4 h, and 6 h. * significant differences between M06 (1E7 CFU/mL) compared to M06 (1E7 CFU/mL) infected with phage 153A-8 at 2 h, p=0.019. No significant differences between M06 (5E6 CFU/mL) or M06 (1E6 CFU/mL) to corresponding group infected with 153A-8 at 2 h (p=0.267 or 0.312, respectively). # significant di fference between c ontrol group and corresponding phage infected group (p=3.2 x 10-5 to 0.018). @ significant difference between control group and corresponding phage infected group (p=1.4 x 10-4 to 0.006). Final microtiter infections were perfor med for each of the seven commonly used laboratory strains of V. vulnificus (Figs. 7a-g). All V. vulnificus strains were grown to exponential phase in LB-SW, diluted in SW, and inoculated into wells containing LB-SW with a final concentration of 1 x 107 CFU/mL. The bacteria were infected with phages at an MOI of 10, and the plates were read ev ery 1 h for 12 h. In conjunction with the microtiter phage typing assay, a soft agar ove rlay phage typing experiment was also

PAGE 69

57 performed at the same time for comparison between both methods. For the soft agar overlay phage typing experiment, the same sc oring scheme as above was followed: 0 no effect, 1 faintly turbid conf luent plaque, 2 <10 clear small plaques, 3 >10 clear small plaques, 4 turbid confluent plaq ue, and 5 clear confluent plaque. All phages significantly lysed V. vulnificus M06-24/O and VV1009 by 4 h and 2 h, respectively, with the exception of phage CK-2 (Figs. 7a and 7b). These results agreed with those obtained using the soft agar overl ay phage typing experiment performed at the same time. In the microtiter experiments, phage CK-2 decreased the growth of both strains of V. vulnificus , although the effect was not as sign ificant as with the other phages (M06, p=0.023 to 0.259; VV1009, p=0.017 to 0.368). In contrast, phage CK-2 was scored as a 0 on these two V. vulnificus strains in the soft agar overlay assays. This small effect by phage CK-2 may suggest an incomp lete infection by CK-2 to both M06-24/O and VV1009 by a defective replicating and/or lysing mechanism, which can be detected with the microtiter method but not with the soft agar overlay method. For V. vulnificus MLT365 between 3 h and 4 h, phages 153A-5, 108A-9, and 3a (at 4h: range of OD630 0.39 to 0.48, p=4.7 x 10-4 to 0.001) caused the lowest bacterial densities that were signifi cantly different compared to the control group (at 4h Control: OD630 0.65 0.01) (Fig. 3-7c). Phages 108A-9 a nd 3a continued to significantly cause the lowest bacterial densities throughout the experiment. However, phage 153A-5 ceased to cause any significant decrease in bact erial density of MLT365, having OD values return to the range of the rest of the ne gative phages. Thus, phages 108A-9 and 3a but not phage 153A-5 were considered to have caused significant lysis. Phages 153A-5, 108A-9, and 3a caused turbid confluent plaque s in the soft agar overlay method. These

PAGE 70

58 results suggested that certain phages that caus e a turbid confluent plaque can be detected using the microtiter phage typing method while other phages are more likely not to be detected using the microtiter phage typing method. For V. vulnificus MLT367, phages 153A-5, 108A-9, 3a caused the greatest inhibition of bacterial growth compared to the control group between 3 h and 7 h (at 3 h, Control: OD630 0.55 0.02; 153A-5: OD630 0.13 0.001, p=0.; 108A-9: OD630 0.10 0.00, p=0.001; 3a: OD630 0.28 0.07, p=0.036) (Fig. 3-7d). In the soft agar overlay phage typing method, phages 153A-5, 108A-9, a nd 3a caused turbid plaques on MLT367. These results suggested that phages that cau sed a turbid confluent plaque on MLT367 in the soft agar overlay phage typing method may also be detected as infectious using the microtiter phage typing method. However, ot her phages that did not cause any plaques inhibited bacterial growth in th e microtiter assay, suggesting th at some level of infectivity of the phages to MLT367 was not detectable by th e soft agar assay. In fact, most of the phages caused significantly lower bacterial de nsities compared to the control group at certain time points. The signifi cance of this phenomenon is not known. For V. vulnificus MLT403, between 5 h and 6 h, phages CK-2, 3a, and 108A-9 (at 6 h: range of OD630 0.20 to 0.47, p=0.001 to 0.014) caused significantly lower cell densities compared to the control group (at 6 h Control: OD630 1.06) (Fig. 3-7e). Between 8 h and 9 h, as phages CK-2, 3a, and 108A-9 conti nued to inhibit growth, phages 110A-7 and 153A-7 began to inhibit bacterial growth compar ed to the control and the rest of the noninhibitory phages (at 9 h: range of p=0.001 to 0.014 ). Phages CK-2, 3a, 108A-9, 110A7, and 153A-7 caused clear plaques in the soft agar overlay phage typing method. Phage 152A-9, which caused a turbid confluent plaque using the soft agar overlay method, did

PAGE 71

59 not cause significant killing in the microtiter method, suggesting again that certain phages that cause a turbid confluent plaques in the soft agar overlay method may not be detectable as infective using the microtiter phage typing method. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0123456789101112 time (h)OD 630 nm Control 152A-2 (5) 152A-8 (5) 152A-9 (5) 152A-10 (5) 153A-5 (5) 153A-7 (5) 153A-8 (5) 154A-8 (5) 154A-9 (5) 108A-9 (5) 110A-7 (5) 7-8a (5) CK-2 (0) 3a (5)* * * * * * * * * * Figure 3-7. Final micr otiter infections of V. vulnificus stains with 14 different phage strains at RT. A microtiter assay was performed at RT for the seven commonly used V. vulnificus stains in our laborator y: a) M06-24/O, b) VV1009, c) MLT365, d) MLT367, e) MLT403, f) 2400112, and g) LL728. V. vulnificus strains were added at 1 x 107 CFU/mL per well, infected with phage at an MOI of 10, and supplemented with LB-SW. Duplicates were performed for each bacteria-phage combination. Plates were read every 1 h for 12 h. The number in parenthesis corresponds to a soft agar overlay assay with the following scoring scheme: 0 no effect, 1 faintly turbid confluent plaque, 2 <10 clear small plaques, 3 >10 clear small plaques, 4 turbid confluent plaque, 5 clear confluent plaque. All infected group( s) below the * for that particular time point are significantly different compared to the uninfected control group (p=0.001 to 0.041). A

PAGE 72

60 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0123456789101112 time (h)OD 630 nm Control 152A-2 (5) 152A-8 (5) 152A-9 (5) 152A-10 (5) 153A-5 (5) 153A-7 (5) 153A-8 (5) 154A-8 (5) 154A-9 (5) 108A-9 (5) 110A-7 (5) 7-8a (5) CK-2 (0) 3a (5)* * * * * * * * * * * 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0123456789101112 time (h)OD 630 nm Control 152A-2 (0) 152A-8 (0) 152A-9 (0) 152A-10 (0) 153A-5 (4) 153A-7 (0) 153A-8 (0) 154A-8 (0) 154A-9 (0) 108A-9 (4) 110A-7 (0) 7-8a (0) CK-2 (0) 3a (4)* * * * * * * * * Figure 3-7. Continued B C

PAGE 73

61 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0123456789101112 time (h)OD 630 nm Control 152A-2 (0) 152A-8 (0) 152A-9 (0) 152A-10 (0) 153A-5 (4) 153A-7 (0) 153A-8 (0) 154A-8 (0) 154A-9 (0) 108A-9 (4) 110A-7 (0) 7-8a (0) CK-2 (0) 3a (4)* * * * * * * * 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0123456789101112 time (h)OD 630 nm Control 152A-2 (0) 152A-8 (0) 152A-9 (4) 152A-10 (0) 153A-5 (0) 153A-7 (5) 153A-8 (0) 154A-8 (0) 154A-9 (0) 108A-9 (5) 110A-7 (5) 7-8a (0) CK-2 (5) 3a (5)* * * * * * * * * Figure 3-7. Continued D E

PAGE 74

62 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0123456789101112 time (h)OD 630 nm Control 152A-2 (0) 152A-8 (0) 152A-9 (5) 152A-10 (0) 153A-5 (0) 153A-7 (5) 153A-8 (5) 154A-8 (5) 154A-9 (0) 108A-9 (5) 110A-7 (5) 7-8a (0) CK-2 (0) 3a (5)* * 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 0123456789101112 time (h)OD 630 nm Control 152A-2 (0) 152A-8 (0) 152A-9 (5) 152A-10 (0) 153A-5 (5) 153A-7 (2) 153A-8 (5) 154A-8 (3) 154A-9 (0) 108A-9 (3) 110A-7 (3) 7-8a (0) CK-2 (0) 3a (5)* * * * * * * * * * * Figure 3-7. Continued F G

PAGE 75

63 The results of the microtiter method of the above phages overall agreed with the results of the soft agar overlay method. Howeve r, the results of the microtiter assays of V. vulnificus 2400112 and LL728, as detailed below, did not equate to the results of the soft agar overlay assays. For example, for V. vulnificus 2400112 at 10 h, phages 110A-7 and 152A-9, both of which caused clear conflu ent plaques in the soft agar overlay method, caused significantly lower cell densit ies compared to the control group and the other phage treatment groups (Fig. 3-7f). However, phages 108A-9, 152A-9, 153A-7, 154A-8, and 153A-8, which caused complete cl ear lysis using the soft agar overlay method, yielded ODs within the same range as that of phages that failed to produce plaques in the soft agar assay. For V. vulnificus LL728, only phages 3a and 153A-5 caused significantly lower cell densities starting at 2 h co mpared to the control group (Fig. 3-7g). These two phages caused complete ly sis with scores of 5 using the soft agar overlay method. Other phages that caused vari ous plaque scores of 2, 3, or 5 using the soft agar overlay method did not significantly inhibit bacterial growth in this microtiter experiment. For example phage 153A-8, whic h caused a score of 5, did not inhibit any growth using the microtiter method. Some of the discrepancies between results obtained using the microtiter and soft agar methods could have been due to inc ubation temperatures. The microtiter method involved RT infections, while the soft agar method was performed at 37C. Because of discrepancies in results obtained between the two methods for V. vulnificus 2400112 and LL728, the effect of an increasing the temper ature to 37C on microt iter experiments was studied. The same microtit er procedure, as detailed above, was followed with the exception of incubating the microtiter plates at 37C instead of RT. For 2400112 at

PAGE 76

64 37C, all phages that caused clear confluent plaques in the soft agar overlay method also caused significant lysis using the microtiter method starting between 3 h and 4 h (Fig. 38a). Furthermore, phages 152A-2, 153A -5 and 154A-9 significantly lysed 2400112, although they were scored a 0 in the soft agar overlay phage typing method. These results suggested that a temperature of 37 C is required by certain phages to infect V. vulnificus in the microtiter experiments. More over, these results showed that certain phages that do not cause plaques in the soft agar overlay experiment may actually cause lysis using the microtiter method at 37C, such as phages 152A-2, 153A-5 and 154A-9. For V. vulnificus LL728 at 37C, all phages that caused sc ores of 2, 3 or 5 in the soft agar overlay method also caused significant inhibiti on of bacterial growth in the microtiter method (Fig. 3-8b). Of note, phage 154A-9 significantly inhibited VV1009, although it was scored as a 0 in the soft agar method. Th ese results indicate that temperature is an important factor when using the micro titer method with certain strains of V. vulnificus , such as 2400112 and VV1009, and possibly even the other V. vulnificus strains in our collection. V. vulnificus grows more robustly at 37C than at RT, possibly allowing more efficient infection and lysis by phages. Furt hermore, using the microtiter method at 37C revealed infectivity of phag es for certain strains of V. vulnificus not consistently observed using the soft agar overlay method alone. The different environments encountered by both phage and V. vulnificus between the soft agar method, a semi-solid state, and the microtiter method, a liquid state, may explain why some phages that do not cause any significant lysis in the soft agar assays cause significant lysis in the microtiter assays. The liquid state may provide a greater probability for phage and bacteria to interact with each other. Although, the other V. vulnificus strains were not examined using the

PAGE 77

65 microtiter method at 37C, this assay could prove useful in more completely analyzing their interactions with phages. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0123456789101112 time (h)OD 630 nm Control 152A-2 (0) 152A-8 (0) 152A-9 (5) 152A-10 (0) 153A-5 (0) 153A-7 (5) 153A-8 (5) 154A-8 (5) 154A-9 (0) 108A-9 (5) 110A-7 (5) 7-8a (0) CK-2 (0) 3a (5)* * * * * * * * Figure 3-8. Microtiter infection assays of V. vulnificus strains 2400112 and LL728 with 14 different phage strains at 37C. A microtiter assay was performed at 37C for: a) 2400112 and b) LL728. V. vulnificus strains were added at 1 x 107 CFU/mL per well, infected with phage at an MOI of 10, and supplemented with LB-SW. Duplicates were performed for each bacteria-phase combination. The number in parenthesis corre sponds to a soft agar overlay assay with the following scoring scheme: 0 no effect, 1 faintly turbid confluent plaque, 2 <10 clear small plaques, 3 >10 clear small plaques, 4 turbid confluent plaque, 5 clear confluent pla que. All infected gr oup(s) below the * for that particular time point are sign ificantly different compared to the uninfected control group (p=0.001 to 0.047). A

PAGE 78

66 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0123456789101112 time (h)OD 630 nm Control 152A-2 (0) 152A-8 (0) 152A-9 (5) 152A-10 (0) 153A-5 (5) 153A-7 (2) 153A-8 (5) 154A-8 (3) 154A-9 (0) 108A-9 (3) 110A-7 (3) 7-8a (0) CK-2 (0) 3a (5)* * * * * * Figure 3-8. Continued Phage Isolation and Transmission El ectron Micrographs of New Phages Phage isolation Due to the inability of phages in our collecti on to fully lyse (soft agar score of 5) all V. vulnificus strains in our collection, isolation of new phages was undertaken to increase the percentage of V. vulnificus strains in our collection comp letely lysed by our pool of phages. The isolation of pha ges that have broad host ra nges and those that infect V. vulnificus strains that were insensitive or only slightly sensitive to the phages in our collection was attempted to enhance the probabi lity of successful treatment of naturally V. vulnificus -infected oysters with a cocktail of phages. To isolate phages, phage enrichment and is olation assays were performed using our seven commonly used V. vulnificus strains. Oyster homogenate or sedimentary mud was enriched with LB-SW and inoculated with V. vulnificus MO6-24/O, LL728, 2400112, B

PAGE 79

67 VV1009, MLT365, MLT367, and MLT403. After overnight incubatio n at 37C, the mixture was processed to isolat e individual phages that were in fective to any of the seven V. vulnificus strains, as described in the Materials and Methods. Phages CB1, EJc, 1a, 2a, 3a, 4a, and 4b were all isolated from oyste rs that originated fro m Cedar Key, FL. As above, these phages were isolated prior to ou r performing the soft agar overlay assays and, thus, were included in those assays. Due to the broad host range of phage 3a, this phage was also examined using the microtiter assays. Phage enrichment and isolation expe riments were also undertaken for V. vulnificus strains that were insensitive or only sli ghtly sensitive to any of the phages in our collection. Oyster homogenate was enrich ed with LB-SW and inoculated with V. vulnificus FLA104, FLA107, FLA108, FLA110, FLA112, FLA119, FLA121, FLA127, FLA 128, FLA 132, FLA137, and FLA1 38. After overnight incubation at 37C, the mixture was then processed to isolate phages infective for any of the V. vulnificus strains used in this experiment. Phages AO1A-D, AO2A-B, and CKIF-G were isolated on V. vulnificus FLA112, FLA104, FLA137, respectively. The nomenclature used for these phages indicates the source of the oysters. “AO” indicates that the oysters originated from Apalachic ola Bay, FL, and “CK” indicates that the oysters originated from Cedar Key, FL. Sin ce these phages were isolated after the soft agar overlay phage typing assays were perf ormed, additional soft agar overlay assays were performed on these newly isolated phage s. The phages were tested on all of the V. vulnificus strains used for the enrichment and is olation experiment (Table 3-3). Of note, phage AO2a-b caused only turbid pla ques on FLA104, indica ting that the phage isolate may perhaps be a temperate phage. A ttempts to isolate lytic isolates of phage

PAGE 80

68 AO2a-b were unsuccessful. Altogether, the phage enrichment and isolation assays proved to be effective in sele cting phages with very broad host ranges, such as phage 3a, and phages with capacity to lyse with a score of 5 V. vulnificus strains in our collection that were otherwise insensitive to the previ ously characterized phages in our collection, such as phage CKIF-G. Table 3-3. Host range of newly isolated phages on V. vulnificus . Using the soft agar overlay method, the newly isolated phages AO1a-d, AO2a-b, and CKIF-G were examined on V. vulnificus strains that were used for the enrichment and isolation assays of these phages. The scoring scheme is as follows: 0 no effect, 1 faintly turbid confluent pla que, 2 <10 clear small plaques, 3 >10 clear small plaques, 4 turbid conflu ent plaque, 5 clear confluent plaque. AO1a-d AO2a-b CKIF-G FLA104 0 4 0 FLA107 0 0 0 FLA108 0 0 0 FLA110 0 0 0 FLA112 5 0 0 FLA119 0 0 0 FLA121 0 0 0 FLA127 0 0 0 FLA128 0 0 0 FLA132 0 0 0 FLA137 0 0 5 FLA138 0 0 0 Transmission electron micrographs of isolated phages The morphologic characterization of is olated phages was determined by using transmission electron microscopy. Transmi ssion electron micrographs were obtained for phages EJc, AO1a-d, AO2a-b, and CKIf-g by the Electron Microscopy Core Laboratory at the University of Florida. All of these bacteriopha ges were of the Caudovirales order, consisting of tailed bacteriophages (Fig. 39). Furthermore, phages EJc, AO2a-b, and CKIf-g were of the Siphoviridae family, containing a long, no n-contractile tail. Phage AO1a-d was of the Myoviridae family, containing a contrac tile tail. All of the phages

PAGE 81

69 isolated were morphologically similar to phage s reported by DePaola et al.(56) However, none of these isolated phages belonged to the Podoviridae family, containing a short, non-contractile tail, as some phage isolates reported by DePaola. Nevertheless, phages 110A-7 and 7-8 in our collection belong to the Podoviridae family. In total, the use of the soft agar ove rlay phage typing method proved to be an effective tool to evaluate all the phages in our collection for such characteristics as host range and the level of infectivity to each V. vulnificus strain in our collection. Additionally, a microtiter phage typing method was developed and demonstrated to complement the results of the soft agar assays. In fact, the microtiter assays may also show phage reactivity on certain strains of V. vulnificus that would have not otherwise been observed using the soft agar overla y assays alone. Furthermore, the phage enrichment and isolation that were performe d to isolate new phages with characteristics such as a broad host range or infectious to a V. vulnificus strain in our collection that was insensitive to any of the phages in our coll ection, proved successful in isolating phages such as phage 3a, which had the broadest host range of any of the phages in our collection. Experiments examining the potenti al of phages in decontaminating naturally and experimentally infected oys ters were now possible, enable d by the data attained with both of the phage typing methods and with the isolation of useful phages.

PAGE 82

70 Figure 3-9. Transmission elec tron micrographs of isolated phages. Transmission electron micrographs of phages a) AO 1a-d, b) AO2a-b, c) CKIF-G, and d) Ejc. Phage AO1a-d is of the Myoviridae family, containing a contractile tail. Phages AO2a-b, CKIF-G, and EJc are of the Siphoviridae family, containing a contractile tail. Th e magnification was 151,000x.

PAGE 83

71 Specific Aim 2: Establishment of an Ex perimental Oyster Infection Model and Treatment of Experimentally Infected Oy sters with Phages to Reduce Numbers of V. vulnificus . The overall goal of the studies described in this thesis was to develop methods for reducing levels of contaminating V. vulnificus in oysters by using V. vulnificus -specific bacteriophages. Karen Cerveny, who obtained her master’s degree from this laboratory, attempted to develop a procedure to experimentally infect oysters with V. vulnificus so that more detailed and controlled phage treatment experiments could be performed. However, her results demonstrated the difficu lty of getting oysters to take up and stably maintain relevant levels of V. vulnificus . Therefore, the studies described below were undertaken to develop an experimental oyster in fection model to cause an increase in the levels of V. vulnificus contamination. After the esta blishment of an oyster infection model, the efficacy of phage treatment of ar tificially infected oysters was studied to evaluate the potential of phage treat ment as a method to decontaminate V. vulnificus infected oysters. Establishment of an Experime ntal Oyster Infection Model A problem confronted with experime ntal infection of oysters with V. vulnificus was the inability of oysters to stably maintain substantial levels of V. vulnificus . For example, when oysters that were acclimated overnight in fresh, sterile seawater were infected with 1 x 106 CFU/mL of V. vulnificus FLA042 for 6 h, the level of contamination in oysters was 3.7 0.24 log CFU/g tissue (Fig. 3-10). Although V. vulnificus levels in experimentally infected oysters were approximately 103 CFU/g tissue, this was too low of a number for the phage treatment experiments b ecause at this level of infection we would have a difficult time observing significant drops in contamination due to phage treatment. We hypothesized that the normal bacterial flora of the oyste rs was preventing the uptake

PAGE 84

72 of exogenously provided V. vulnificus in our experiments. Therefore, in an attempt to increase the level of V. vulnificus infection in oysters, oysters were pretreated with 50 g/mL of Rif overnight during the acclimati on period to eliminate their bacterial flora prior to infection with V. vulnificus . This resulted in minimal to nondetectable levels of bacterial flora of oysters when plated on LB-N plates (data not show n). Infection of Riftreated oysters with 1 x 106 CFU/mL of V. vulnificus FLA042 for 6 h significantly increased the level of V. vulnificus to 6.0 0.59 log CFU/g-tissue (p=3.6 x 10-4) compared with non–Rif-treated oysters (Fig. 3-10 ). This experimenent was repeated with similar results. Besides Rif pretreatment, the addition of al gae coupled with the infection of oysters with V. vulnificus FLA042 was examined for any increase in the level of V. vulnificus contamination in experimentally infected oys ters. Oysters eat planktonic algae for food, and it had been suggested that oysters accumulate V. vulnificus as a result of their adherence to algae. We examined the eff ect of providing algae on uptake of vibrios in two experimental infection experiments. Firs t, the effect of addi ng algae on the level of V. vulnificus contamination was studied by infecting acclimated oysters with 1 x 106 CFU/mL of FLA042 simultaneously with the ad dition of algae to oysters for 6 h. The commercially available algae (Shellfish Diet 1800; Reed Mariculture, Inc.; Campbell, California), specifically formulated for feed ing oysters, was added per manufacturer’s recommendations. The mixture of algae a nd FLA042 was incubated for 10 min prior to its addition to the oysters to give time for th e bacteria to adhere to the algae. The infection of FLA042 simultaneou sly with algae resulted in infection levels of 4.2 0.31 log CFU/g tissue compared to 3.7 0.24 log CFU/g tissue without algae (p=0.084). In

PAGE 85

73 the second experiment, oysters were pretr eated with Rif during the acclimation period and then infected with 1 x 106 CFU/mL of FLA042 simultaneously with algae for 6 h, as for the first experiment. With the combined pr etreatment of oysters with Rif and addition of algae, the level of FLA042 contamination was significantly increased to 5.7 0.19 log CFU/g tissue (p= 6.60 x 10-5) compared to oysters infected with FLA042 without a Rif pretreatment or addition of algae. This le vel of infection was similar to the level of infection with Rif pre-treatment. Due to the significant increase in the level of V. vulnificus infection of oysters with Rif pretreat ment, oysters were pretreated with Rif in all of the following experiments. Another parameter studied in our expe rimental infection of oysters with V. vulnificus was the effect of the duration of time of infection on the level of infection in oysters. Rif-pretreated oysters were infected with 1 x 105 CFU/mL of FLA042, and samples of water and oysters were taken at 0.5 h, 1 h, 1.5 h, 2.5 h, and 6 h (Fig. 3-11). The level of infection of oys ters with FLA042 reached 2.6 0.33 log CFU/g-tissue by 1 h and a significantly higher level of 5.0 0.75 log CFU/g-tissue by 6 h (p=0.013). Although the levels of FLA042 in oysters increased by 1000-fold throughout the experiment, the levels of FLA042 in the water only increased by 10-fold to 1.6 x 106 CFU/mL. Interestingly, the levels of cont amination in both water and oyster tissues reached equilibrium at approximately 105 to 106 CFU/mL (data not shown). Therefore, experimental infection of oysters for at le ast 6 h should result in maximum levels of contamination in oysters. For subsequent phage treatment of experimental oyster infections, Rif-pretreated oys ters were infected with V. vulnificus for 6 h to 7 h for optimal levels of contamination in oysters. However, we also examined a shorter period

PAGE 86

74 of infection of 1 h to examine decontamina tion with phage treatment with a lower level of contamination in oysters and briefer expos ure of vibrios to th e oyster environment. 0 1 2 3 4 5 6 7 Rif+ RifAlgaeAlgae/Riflog CFU/ g-tissue* * Figure 3-10. Effect of rifampicin and alg ae on experimental infec tion of oysters with V. vulnificus FLA042. Oysters were infected with 1 x 106 CFU/mL of V. vulnificus FLA042 for 6 h. Control (n=4) oy sters that were not treated with rifampicin or algae, + Rif (n=4) oyste rs treated with rifampicin overnight, Algae (n=3) oysters that were fed alg ae, and Algae/Rif (n=3) oysters that were fed algae and treated with rifa mpicin overnight. * significantly different compared to – Rif (+ Rif, p=3.6 x 10-4; Algae/Rif, p=6.6 x 10-5). Algae group was not significantly di fferent compared to – Rif (p=0.084).

PAGE 87

75 0 1 2 3 4 5 6 7 0123456 time (h)log CFU/ ml or g-tissue oyster Water Figure 3-11. Effect of the duration of infection time on the level of V. vulnificus infection in oysters. Rifampicin-pretreated oysters were infected with 1 x 105 CFU/mL of V. vulnificus FLA042. Water and oysters were sampled at 0.5 h (n=3), 1h (n=4), 1.5 h (n=5), 2.5 h (n=10), and 6 h (n =3). Data are reported as mean standard deviation. We hypothesized that UV treatment of the wa ter of experimentally infected oysters during a rinse cycle would elimin ate or decrease the level of V. vulnificus in the water, whereas the V. vulnificus infecting the oysters would be protected. The use of DNA damaging ultraviolet light (UV) is widely us ed in aquariums to decrease the level of pathogenic bacteria in the water. The UV f ilter used in aquariums is an elongated UV lamp inserted into a plastic tube contai ning openings on both ends of the tube where water enters and exits the tube. When water en ters the tube, UV inflicts lethal damage to the bacteria, killing them. Furthermore, the usage of a UV filter has also been reported for decontamination of oysters during a UV depuration process (63). The use of a UV

PAGE 88

76 filter in such an application would prevent a superficial infection of oysters due to the brief exposure to the contamination of the wate r bathing the oysters. Oysters were infected with 1 x 106 CFU/mL of FLA042 for 1 h and then rinsed with UV light treatment for 24 h (Fig. 3-12). No decrease in the level of FL A042 contamination in oysters was seen with the use of UV filters for 24 h when compared to the level of FLA042 contamination in oysters that were not UV treated. However, the level of FLA042 contamination in the water was decr eased by about 1.7 logs with UV filtration. These results suggested that most of the V. vulnificus in the oysters was stable and not merely due to superficial contamination with the water. However, these results were inconsistent with other UV f iltration experiments conducted in our laboratory. At times, the use of UV filters did not lower the level of V. vulnificus contamination in the water housing the oysters. Moreover, UV filt ers sometimes lowered the level of V. vulnificus contamination in oysters compared to oysters not treated with UV filtration. A possible reason for the lack of consistent killing of V. vulnificus in the water could have been that the UV filters were unable to perform properly at all times. One considered problem was the residual level of Rif in the water, whic h often colored the wate r light red and could have interfered with the UV filters. As we accumulated more experience with experimental infections, we felt confident that the levels of infection of oysters were not simple surface contamination with the water in which the oysters were sitting. Therefore, UV filters were deemed unnecessary for the subsequent experimentally infected oyster experiments.

PAGE 89

77 0 1 2 3 4 5 6 No UV treatmentUV treatmentlog CFU/g-tissue or log CFU/mL Water Oyster Figure 3-12. Effect of ultraviolet light filte rs on experimental infection of oysters with V. vulnificus . Oysters were experimentally infected with 1 x 106 CFU/mL of V. vulnificus FLA042 for 1 h and then transferred to fresh seawater. The water housing the oysters was either trea ted with UV light (n=12) or not (n=9) for 24 h. Bars represent mean sta ndard. The UV treated and nontreated oyster groups were not signi ficantly different (p=0.119). By examining parameters such as the effic acy of Rif treatment pr ior to infection of oysters, the beneficial addition of algae duri ng experimental infections of oysters, the duration of infection of oysters, and the e fficacy of UV filtration, we were able to develop an experimental infection of oys ter model that increased the level of V. vulnificus in oysters to levels to as high as 106 CFU/g tissue. The use of Rif-pretreatment without algae feeding was considered the most efficien t and reliable method to infect oysters with a Rif-resistant V. vulnificus strain; thus, the use of Rif alone was the method used for the rest of the experiments. With the development of this oyste r infection model, testing the efficacy of phage treatment of experiment ally infected oysters was now possible.

PAGE 90

78 Bacteriophage Treatment of Exp erimentally Infected Oysters After the parameters for experimental infection of oysters with V. vulnificus were established, phage treatment parameters were examined using experimentally infected oysters. However, prior to phage treatment e xperiments, the level of naturally occurring phages infective to the two strains of V. vulnificus that were to be used for the experimentally infected oysters, M06-24/O and MLT403, were analyzed in both oysters and in the water used for housing the oysters. After oysters were acc limated in seawater overnight, tissue was homogenize d, centrifuged to remove debr is, and filtered through a 0.2 m filter to remove bacterial cells. The f iltrate was serially diluted, mixed with 4 mL of LB-SW soft agar containing 1 x 107 CFU/mL of either M06-24/O or MLT403, vortexed, and poured on a LB-SW plate. Natu rally occurring phages infective for M0624/O (3.5 0.43 log PFU/g-tissue) and MLT 403 (2.0 0.30 log PFU/g-tissue) were isolated from the oysters (Fig. 3-13). Howe ver, no naturally occurr ing phages infective to either M06-24/O or MLT403 were recove red from the water. The isolation of naturally occurring phages in oysters as high as 3.5 logs PFU/g tissue may represent a possible interference with our phage treatment of experime ntally infected oysters. Furthermore, the inability of recovering any ly tic phages from the wate r may indicate that the phages are stably retained by the oysters. We next wanted to know how well phages would be taken up by and retained in oysters. The assumption was that phages would at least be in equilibrium with the water pumped by the oysters; however, we were unsur e if phages would be st ably maintained in oyster tissues after the water had been cha nged. Phage CK-2 incubation and retention was examined in oysters (Fig. 3-14). Oyster s were acclimated overnight in seawater and incubated with 1 x 107 PFU/mL of phage CK-2. Unlike the previous experiment, no

PAGE 91

79 naturally occurring phages inf ective for MLT403 were detected in the water or oysters before treatment with CK-2. Oysters that were treated with CK-2 had 7.3 0.13 log PFU/g tissue and 7.7 log PFU/mL in the water. When CK-2-treated oysters were rinsed with fresh seawater overnight , the level of CK-2 in tissu es, 6.7 0.12 log PFU/g-tissue, was similar to that from the pre-rinsed group. However, the level of CK-2 in the water was decreased to 5.2 log PFU/mL. The differe nce between the level of CK-2 in oyster tissue and water in the rinsed group suggested that phage CK-2 was stably retained by the oysters. 0 1 2 3 4 5 MO6MLT403 V. vulnificus strainlog PFU/g-tissue or PFU/ml Water Oyster Figure 3-13. Analysis of the number of natu rally occurring phages present in market oysters and water that are infectious to V. vulnificus M06-24/O and MLT403. After an overnight acclimation period, oysters (n=4) and th e water housing the oysters were analyzed using the soft agar overlay method to determine the number of lytic phages present in th e water and oysters for M06-24/O and MLT403.

PAGE 92

80 0 1 2 3 4 5 6 7 8 9 ControlCK-2CK-2 with rinselog PFU/g-tissue or PFU/ml Water Oyster* Figure 3-14. Retention of phage CK-2 in oysters. Oysters (n=9) were incubated overnight with (n=2) or without phage CK-2 at 1 x 107 PFU/mL. Oysters that were incubated overnight with phage CK -2 were either sampled immediately after the incubation period (CK-2 group)(n=3) or sampled after an overnight rinse period in fresh seawater (CK-2 w ith rinse group)(n=2). Bars represent mean standard deviation. * signi ficantly different compared to CK-2 (p=0.014). Treatment of Experimentally Infected Oysters with Phages 3a and CK-2 Phages 3a and CK-2 were studied both separately and combined to evaluate their potential to decontaminate oysters that were experimentally infected with V. vulnificus FLA042. Phages 3a and CK-2 were chosen for the phage treatment experiments due to certain characteristics of these phages that were considered beneficial. For example, we were able to obtain very hi gh titer lysates (CK-2) or th e phages exhibited broad host ranges (3a). Oysters were infected with V. vulnificus FLA042 for either 1 h or 6 h at 1 x 106 CFU/mL, as described above, and were treated overnight with either phage 3a (with 1 h infection only) or CK-2 at 1 x 108 PFU/mL. Oyster tissues were then harvested,

PAGE 93

81 homogenized, serially diluted, and plated on LB-N-Rif plates. Treatment of oysters experimentally infected with FLA042 with phages 3a or CK-2 al one did not cause any significant decrease in the level of contam ination (Fig. 3-15)(data for phage 3a not shown). 0.00 1.00 2.00 3.00 4.00 5.00 1 h7 hlog CFU/g-tissue Control Phage CK-2 treated Figure 3-15. Treatment of V. vulnificus FLA042-infected oyster s with phage CK-2. Oysters that were infected with 1 x 106 CFU/mL of V. vulnificus FLA042 for either 1 h (n=6) or 7 h (n=9) were then treated with 1 x 108 PFU/mL of phage CK-2 overnight. Bars represent mean standard deviati on. No significant differences between control and pha ge treated groups (1 h, p=0.134; 7 h, p=0.944). Since treatment of experimentally infected oysters with either phages 3a or CK-2 separately did not cause any significant decrease in the le vel of FLA042 contamination, treatment of experimentally infected oysters with a combination of phages 3a and CK-2 was studied. We hypothesized that the combination of both phages could increase the level of decontamination in oys ters. Oysters that were e xperimentally infected with

PAGE 94

82 FLA042 for 1 h were treated ove rnight with a combination of phages 3a and CK-2 (1 x 108 PFU/mL each). Oyster tissues were harv ested, homogenized, serially diluted, and plated on LB-N-Rif. With a 1 h infection, the combina tion of phages CK-2 and 3a caused a significant drop of 99% in leve ls of FLA042 from 5.7 0.43 log CFU/g-tissue (untreated control group) to 3.5 0.38 log CFU/g-tissue (C K-2 + 3a treated group)(p=2.9 x 10-4) (Fig. 3-16). 0 1 2 3 4 5 6 7 ControlCK-2 + 3aCK-2 + 3a after homogenizationlog CFU/g-tissue* Figure 3-16. Effect of CK-2 and 3a treat ment of oysters infected for 1 h with V. vulnificus FLA042: before and after homogeni zation. Oysters were infected with 1 x 106 CFU/mL of FLA042 for 1 h. The control group (n=20, combination of three experiments) was only infected with FLA042. The CK2 + 3a group (n=19, combination of thr ee experiments) was treated overnight with CK-2 and 3a at 1 x 108 PFU/mL each. The CK-2 + 3a after homogenization group (n=4, a single experiment) was a control group in which the oyster tissue was spiked with a phage solution containing CK-2 and 3a at 1 x 108 PFU/ml each prior to the homogenization of the oyster tissue. Bars represent mean standard de viation. * signi ficantly different compared to the control group (p=5.9 x 10-11). CK-2 + 3a after homogenization group was not significantl y different compared to the control group (p=0.801).

PAGE 95

83 In the 3a and CK-2 combined phage tr eatment of oysters experiment, it was possible that the phages did not kill the vibrios in the oyster tissues during treatment, but instead killed the bacteria after the tissues had been homogenized. This manner of phage efficacy would not be relevant to treatment of market oysters. Therefore, another group of infected oysters was examined for th e effect of adding phages 3a and CK-2 immediately prior to homogeni zation of oyster tissues, as op posed to incubating the live oysters with phage overnight. The same c oncentration of phages CK-2 and 3a (1 x 108 PFU/mL each) as used for the phage treatments of live oysters was added to oyster tissue instead of seawater, and the mixture was ho mogenized, serially diluted, and plated on LB-N–Rif. This brief post harvest phage in cubation of oyster tissue with phages 3a and CK-2 did not cause a significantly lower recovery of FLA042 (p=0.801)(Fig. 3-16). Therefore, the killing of FLA042 by phages CK -2 and 3a during treatment of live oysters had occurred during the phage treatment peri od, rather than after homogenization of the oyster tissue. This experiment was repeated again, and similar results were obtained. We considered the possib ility that the residual V. vulnificus in oyster tissues after phage treatment consisted of spontaneous phage-re sistant mutants of the infecting strains. Therefore, 5 colonies of FLA042 on the LB-N-R if plates per oyster were tested for phage sensitivity, and all of the colonies examined had retained their sensitivity to both phages CK-2 and 3a. Therefore, the inability of phages 3a and CK-2 to completely decontaminate FLA042 experimentally infected oysters was not likely due to spontaneous phage-resistant mutants. The effect of the duration of infection with FLA042 in experimentally infected oysters on the effectiveness of phage treatment was studied with phages 3a and CK-2.

PAGE 96

84 The same procedure for experimental infect ion of oysters and phage treatment with phages 3a and CK-2 was followed, as above, wi th the exception of in fecting the oysters for either 6 h or 24 h. When oysters were in fected with FLA042 for 6 h and then treated overnight with phages CK-2 and 3a, the leve l of contamination was lowered from 6.3 0.59 log CFU/g tissue (Control group) to 5.4 0.60 (CK-2 + 3a treated group)(p=0.008)(Fig. 3-17). When oysters were infected wi th FLA042 for 24 h prior to phage treatment with CK-2 and 3a, the leve l of contamination of FLA042 was lowered from 5.4 0.72 log CFU/g-tissue (Control) to 4.3 0.69 log CFU/g-tissue (CK-2 + 3a treated) (p=0.036) (Fig. 3-17). The most effective phage treatment with phages 3a and CK-2 was observed on experimentally infect ed oysters infected with FLA042 for 1h. This was unfortunate because treatment of na turally contaminated oysters would involve V. vulnificus that had resided in the oysters for ex tended periods of time, and they might be more resistant to phage treatment. The reasons behind the inability of phage s CK-2 and 3a to fully clear FLA042infected oysters of the bact eria were considered for im proving the methods. Possible mechanisms of failure of phage treatment c onsidered included: ge notypically resistant mutants (as examined above), a V. vulnificus strain not expressing phage receptors in the oyster environment and becoming phenotypica lly resistant, inability of phages to encounter V. vulnificus that had invaded into tissu es, and lack of growth of V. vulnificus in stably infected oysters. As detailed a bove, because colonies recovered after phage treatment of experimentally infected oysters retained sensitivity to both phages 3a and CK-2, we do not believe that spontaneous pha ge-resistant mutants were the cause for the incomplete elimination of V. vulnificus during phage treatment. We also do not believe

PAGE 97

85 V. vulnificus FLA042 is down regulating the expressi on of phage receptors in the oyster environment because some experiments demonstrated the ability of phages CK-2 and 3a to kill up to 99% of the initial bacter ial population. Had the bacteria shut down expression of the phage receptors, we would not have expected to see such efficient killing; however, we cannot rule out the possibi lity that a small population of bacteria in a specific environment had changed releva nt gene expression. Another possible explanation for the lack of complete decontamin ation in oysters was th e lack of growth of V. vulnificus , since phage infection and lysis require the active gr owth of bacteria. The CFU of V. vulnificus in experimentally infected oysters stabilized at approximately 6 h of infection at 105 to 106 CFU/g tissue, suggesting that the bact eria were in stationary phase. The more efficient decontamination obser ved with phage treatment of oysters experimentally infected for only 1 h, as opposed to 6 h, could have been due to the fact that fewer of the bacteria had reached stationary phase before phage treatment at 1 h. Growth of V. vulnificus in Oysters Since we had evidence that V. vulnificus might be replicating to a very limited extent in experimentally inf ected oysters and this could ha ve affected phage treatment, we used two complementary methods, a marker plasmid and ampicillin (Amp) enrichment, to examine the state of growth of V. vulnificus during experimental infections with FLA077 and FLA042. Marker plasmid pGTR902 The state of growth of V. vulnificus in experimentally infected oysters was initially studied with V. vulnificus FLA077 containing the marker plasmid pGTR902. The pGTR902 marker plasmid, which was constructed by Angela Starks (Starks, A.M., et al., submitted to Mol. Microbiol.) contains a pBADara C regulatory system that strictly

PAGE 98

86 controls the transcription of the pir gene, which produces a prot ein that is necessary for the replication pGTR902. In the presence of the inducer arabinos e, the AraC protein binds to an activation region upstream of the pBAD promoter , from which transcription of pir is induced, and replication of the plasmid is enabled. However, in the absence of arabinose, pir is not expressed resulting in non -replication of the plasmid. During bacterial replication under nonpermissive conditions, lack of replication of pGTR902 results in the segregation of the plasmid from the bacter ial population because of the accumulation of plasmid-free cells with ea ch division. Because the plasmid cannot replicate in oysters, monitoring the number of plasmid-containing bacteria at different times reveals how much bacterial death has occurred. Consequently, growth and killing of V. vulnificus can be analyzed using the marker plasmid with an oyster infection experiment. The growth proportion of V. vulnificus was calculated as the ratio of the number of bacteria containi ng pGTR902 to the total number of bacteria in oysters. The killing proportion was calculated as the ra tio of the number of bacteria containing pGTR902 at various time points during the e xperiment to the nu mber of bacteria containing pGTR902 at the beginning of the experiment. Oysters were experimentally infected with V. vulnificus FLA077(pGTR902) for 7 h in the same manner as detailed above with th e exception that the bacteria were grown to exponential phase in the presence of 1% (wt/vol) arabinose and 300 g/mL of Kan to maintain carriage of the plasmid. The oyste rs were harvested, homogenized, serially diluted, and plated on LB-N, LB-N–Rif, and LB -N–Kan-Ara. Details of the media are in the Materials and Methods. Since both the LB-N and the LB -N-Rif plates had comparable CFU throughout these experiments, only LB-N-Rif plates were used for

PAGE 99

87 analysis of the state of growth of FLA077. The growth proportion of FLA077 at various time points was calculated by the ratio of the number of bacteria containing the marker plasmid pGTR902, enumerated on LB-N-Kan-Ara pl ates, to the total number of bacteria in oysters, enumerated on LB-N -Rif plates, and is expressed as a log value. The death proportion of FLA077 in an experi mental oyster infection was also calculated by the ratio of the number of bacteria containing the ma rker plasmid pGTR902 at various time points during the experiment to the number of bact eria containing the marker plasmid pGTR902 in the beginning of the experiment, and is expressed as a log value. After a 7 h infection of oysters with FLA077(pGTR902), the oysters contained 4.9 0.85 log CFU/g tissue and a growth proportion of -1.4 (Fig. 3-18). This demonstrated that some initial replication had occurred, as previously observed with FLA042 between 1 h and 6 h. However, the growth proporti ons at 4 h, 16 h, and 40 h of incubation in fresh, sterile seawater were relatively consta nt (ranging from log 2.4 to log -2.7). The lack of change of the growth proportion from 4 to 40 h indicated a lack of growth of FLA077. If growth was occu rring during the later inc ubation periods, the growth proportions would have decreased due to the segregation of the marker plasmid in the absence of arabinose. There was a killing proportion drop of -0.6 l ogs at 4 h indicating some death of the bacteria had occurred betw een the end of the 7 h infection and the 4 h rinse. However, the killing proportions at 4 h, 16 h, and 40 h of incubation in fresh, sterile seawater were relativel y constant (ranging from log 0.6 to log -0.9), indicating a lack of death of the vibrio s over this time period. The results observed with the pGTR902 marker plasmid demonstrated th at very little, if any, growth of V. vulnificus occurred during experimental infection with V. vulnificus. This may be an important

PAGE 100

88 factor for the lack of complete decontamina tion of experimentally infected oysters with treatment of phages CK-2 and 3a. This experi ment was repeated, and similar results were obtained. 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 6 h infection24 h infectionlog CFU/g-tissue Control CK-2 + 3a* * Figure 3-17. Effect of CK-2 and 3a treatment of oysters infected for 6 h or 24 h with V. vulnificus FLA042. Oysters were infected with 1 x 106 CFU/mL of FLA042 for 6 h or 24 h. The CK-2 + 3a groups of either the 6 h (n=8) or 24 h (n=5) infection periods were treated overn ight with CK-2 and 3a at 1 x 108 PFU/mL each. Bars represent mean standard deviation. * significantly different compared to the control group (6 h, p=0.008; 24 h, p=0.036).

PAGE 101

89 0 1 2 3 4 5 6 7 After 7 h of infectionAfter 4 h of UV rinseAfter 16 h of UV rinseAfter 40 h of UV rinselog CFU/g-tissue Rif Kan 300 Growth proportion Killing proportion -0.9-0.7 -0.6 -1.4-2.4-2.7 -2.7 Figure 3-18. Segregation of the marker plasmid pGTR902 in V. vulnificus in experimentally infected oysters. Oysters were inf ected with 1 x 106 CFU/mL of FLA077(pGTR902) for 7 h. After the 7 h infection peri od, the infected oysters were either immediately sample d (After 7 h of infection group)(n=5) or placed in fresh seawater for rinsin g periods of 4 h, 16 h, or 40 h (n=5 each group). Oysters were harvested, homoge nized, and plated on both LB-N–Rif and LB-N–Kan-Ara. The growth propor tions and killing proportions are presented for each group. Bars represent mean standard deviation. No significant differences between after 7 h of infection (Rif) and 4 h, 16 h or 40 h of UV rinse (Rif) (p=0.114 to 0.832). No significant difference between after 4 h of UV rinse (Rif) and after 16 h of UV rinse (Rif) (p=0.094). No significant difference between after 7 h of infection (Kan) an d 4 h of UV rinse (Kan) (p=0.062). Significant differen ce for growth proportion after 7 h of infection (p=0.007). Significant differe nce between growth proportions after 7 h of infection and after 4 h of UV rinse (p=0.01). Ampicillin treatment assays The Amp treatment assays were undertaken to further study the state of growth of V. vulnificus in oysters. The mode of action fo r Amp is the inhibition of enzymes responsible for the synthesis of peptidogly can, which causes osmotic lysis of bacterial cells. If the bacteria are not replicating, peptidoglycan is not synthesized, and Amp will

PAGE 102

90 not lyse the cells. Phenotypi c resistance of genetically se nsitive bacteria to Amp or similar antibiotics can therefore serve as a marker for lack of replication. Therefore, if vibrios in the experimentally infected oysters were not killed by treatment with Amp, this could indicate the lack of growth of V. vulnificus . Amp is also useful for these experiments because it has good pene tration of tissues and cells. Prior to the Amp treatment experiments, th e effective concentration of Amp had to be determined for the oyster experimental infection assays. Th e minimal inhibitory concentration of Amp was determined for strain FLA077 by making two-fold serial dilutions starting from 100 g/mL of Amp in LB-N. The Amp broths were inoculated with a final concentration of 1 x 107 CFU/mL of FLA077 that was grown to exponential phase and incubated overnight at 37C. Th e minimal inhibitory concentration of Amp for FLA077 was 6.3 g/mL (Table 3-4). Table 3-4. Minimal inhibitory concentration of Amp for V. vulnificus FLA077. Twofold serial dilutions of Amp starting at 100 g/mL in LB-N. Strain FLA077 was grown to exponential phase and i noculated into each corresponding broth to a final concentration of 1 x 107 CFU/mL. The broths were incubated overnight at 37C and then visually inspected for turbidity. Ampicillin concentration (g/mL) Growth level of FLA077 25.0 Clear 12.5 Clear 6.3 Clear 3.1 Turbid 1.6 Turbid 0.0 Turbid Amp concentrations of 10, 33, and 100 g/mL were then tested on oysters experimentally infected with FLA042 to d ecide which concentration of Amp to use on the following Amp treatment experiments. Oy sters that were treated with Rif overnight were infected with FLA042 for 7 h and then rinsed overnight with SW containing 10, 33, or 100 g/mL of Amp. The oysters were ha rvested, homogenized, serially diluted, and

PAGE 103

91 plated on LB-N-Rif plates. The treatmen t of oysters with Amp, regardless of the concentration, significantly decreased the le vel of contamination by 2.4 to 2.5 (range of p=0.004 to 0.005). No significant differen ces were observed among the different concentrations of Amp used for the overnigh t rinse (Fig. 3-19). These results showed that any of the Amp concentration could work for the Amp enrichment experiments; the higher concentration of Amp (100 g/mL) was used for the rest of the Amp enrichment assays in oysters. Although th e intention of this particular experiment was to determine the appropriate concentration of Amp for the Amp enrichment assays detailed below, this experiment demonstrated that a subset of th e population of V. vulnificus FLA042 in the oysters was unable to be killed with Amp, approximately 103 CFU/g tissue. This observation suggested that a por tion of the bacteria was not growing adequately in the oysters for Amp killing. We then compared the abilities of treat ment with Amp at 100 g/mL, treatment with phages 3a and CK-2 (1 x 108 PFU/mL each), or a combination of these treatments to clear oysters experimentally infected for 6 h with FLA042. When oysters were treated with Amp, the level of FLA042 was lowered by 3.1 logs (99.9%) (p= 0.001)(Fig. 3-20). As above, this result suggested that a subset of the bacterial p opulation, approximately 103 CFU/g tissue, was not growing adequately to enable ampicillin killing of the bacteria. When oysters were treated with phages CK-2 and 3a, the level of contamination was lowered by 1.2 logs (97%) (p=0.026). These resu lts were comparable to the results of the earlier phage treatments of 3a and CK-2 on oysters experimentally infected with FLA042 for 6 h. When oysters were treated with bot h Amp and phages CK-2 and 3a, the level of FLA042 contaminating was lowere d by 3.8 logs (p=0.001). CFU of V. vulnificus in

PAGE 104

92 oysters treated with Amp, CK-2, and 3a were not significantly different from that in oysters treated with only Amp (p=0.26). This result suggests that bacteria killed by phages CK-2 and 3a could also be killed by Amp. However, Amp alone killed more bacteria than did phages CK-2 and 3a (p= 0.001). The greater efficiency in killing by Amp may be due to its ability to penetrate oyster tissues compared to phages CK-2 and 3a or in their differences in m ode of killing the bacteria. 0 1 2 3 4 5 6 7 8 ControlAmp 10 g/mlAmo 33 g/mlAmp 100 g/mllog CFU/g-tissue or log CFU/mL Water Oyster** * Figure 3-19. Effect of ampicillin treatment on V. vulnificus FLA042 in experimentally infected oysters. Oysters were infect ed with FLA042 for 7 h. The oysters were then transferred to fresh seaw ater containing 10, 33, 100 g/mL of Amp, or no Amp (Control group)(n=3 each group) . Bars represent mean standard deviation. * significantl y different compared to the control group (p=0.004 to 0.005).

PAGE 105

93 0 1 2 3 4 5 6 7 8 ControlAmp 100 g/mlCK-2 + 3a (1E8 PFU/ml)Amp 100 g/ml + CK-2 + 3a (1E8 PFU/ml)log CFU/g-tissue or log CFU/mL Water Oyster * * * # Figure 3-20. Effect of ampic illin enrichment compared to treatment with phages CK-2 and 3a on V. vulnificus FLA042 in experimentally infected oysters. Oysters were infected with FLA042 for 7 h. The oysters were then transferred to fresh seawater containing 100 g/mL Amp, 1 x 108 PFU/mL each of phages CK-2 and 3a, combination of both treatments, or no treatment (Control group)(n=4 each group). Bars represent mean standard deviation. * significantly different compared to the contro l group (Amp 100 g/mL, p=8.6 x 10-4; CK-2 + 3a, p=0.026; Amp 100 g/mL + CK-2 + 3a, p=6.6 x 10-4). # significantly different compared to the Amp 100 g /mL and Amp 100 g/mL + CK-2 + 3a (Amp 100 g/mL, p=8.7 x 10-4; Amp 100 g/mL + CK-2 + 3a, p=9.5 x 10-4). Since the marker plasmid assays were performed with V. vulnificus FLA077 and the Amp enrichment assays were performed with FLA042, the Amp enrichment assay was also performed with FLA077 to relate the two methods using one strain. Oysters were infected for 7 h with FLA077, as desc ribed above, and treated overnight with Amp (100 g/mL). As observed using FLA042, levels of FLA077 in oysters were significantly lowered with Amp treat ment by 2.6 logs (99%) (p=1.0 x 10-8)(Fig. 3-21). The inability of Amp treatment to comple tely eliminate FLA077 from oysters was

PAGE 106

94 consistent with the hypothesis that a portion of the bacterial populati on in the oysters was not growing sufficiently to enable killing by either Amp or phages. 0 1 2 3 4 5 6 ControlAMP 100 g/mllog CFU/g-tissue or log CFU/mL Water Oyster* Figure 3-21. Effect of ampicillin enrichment on V. vulnificus FLA077 in experimentally infected oysters. Oysters were infect ed with FLA077 for 7 h. The oysters were then transferred to fresh seawat er containing 100 g/mL of Amp or no Amp (Control group)(n=6 each group). Bars represent mean standard deviation. * significantl y different compared to the control group (p=1.0 x 10-8). Attempts at Stimulation of V. vulnificus Growth Coupled with Phage Treatment We considered that if the residual population of V. vulnificus not killed by phages in experimentally infected oysters was due to a lack of growth, we might be able to render this resistant populati on of bacteria susceptible by stimulating their growth. Therefore, attempts were made to stimula te the growth of FLA042 in oysters with subsequent treatment with phage s CK-2 and 3a. Two strategies were utilized to increase the growth of FLA042: addition of FeCl3 to the water and increase in water temperature.

PAGE 107

95 Iron supplementation To study the effect of supplementation of water with FeCl3 on V. vulnificus growth in oysters and efficiency of phage treatment, oysters were experimentally infected with FLA042 for 7 h, as described above, with the exception that 5 M of FeCl3 was added to the water during the infection pe riod. Supplementation with FeCl3 did not significantly increase the growth of FLA042 in oyste rs (p=0.389)(Fig. 3-22). Moreover, supplementation with Fe+3 did not significantly affect the effectiveness of phage treatment. In fact, the phage treatme nts in the oysters supplemented with Fe+3 resulted in nonsignificant values compared to the control group (p=0.280). Therefore, supplementation of water with 5 M FeCl did not stimulate growth of V. vulnificus or increase killing of V. vulnificus by phage. These results suggested that Fe+3 is not limiting the growth of V. vulnificus during oyster infections. This experiment was repeated, and similar results were obtained. Elevated water temperature To study the effect of an increase in water temperature on growth of V. vulnificus in experimentally infected oysters, water heater s were used to increase the temperature of the seawater to 32.5C from the usual RT of approximately 23C. The oysters were experimentally infected, as described above, and water te mperature was increased to 32.5C simultaneously with the tr eatment with phages 3a and CK-2. The increased water temperature significantly increa sed the level of FLA042 in non-phage-treated oysters by 0.9 logs CFU/g-tissue (p=5.60 x 10-4)(Fig. 3-23). The level of FLA042 in the phagetreated oysters at RT was significan tly decreased by 1.1 logs (p=8.7 x 10-5) while the phage-treated oysters at 32.5C was significantly decrea sed by1.8 logs (p=1.2 x 10-6). However, the final levels of contamination in oysters in both the RT and 32.5C phage-

PAGE 108

96 treated groups were not significantly differe nt from each other (p=0.521). These findings suggested that the phag e treatment killed the “extra” bact eria apparently stimulated by the increased temperature, resulting in the same level of residual contamination in either phage-treated group. These results suggested th at phages 3a and CK-2 were only able to kill the bacteria that were in a state of growth, since the final levels of bacterial contamination were the same for either the oyste rs that were at RT or at 32.5C. This experiment was repeated, and similar results were obtained 0 1 2 3 4 5 6 7 ControlCK-2 + 3aControl + FeCl3CK-2 + 3a + FeCl3log CFU/g-tissue Figure 3-22. Effect of supplem entation of seawater with FeCl3 on V. vulnificus FLA042 growth in experimentally infected oysters and on the efficacy of phage treatment. Oysters were infected with FLA042 for 7 h. The oysters were then transferred to fresh seawater with or without 5 M of FeCl3 and treated and with or without 1 x 108 PFU/mL each of phages CK-2 and 3a (n=3 each group). Bars represent mean standard deviation. No significant difference between control and control + FeCl3 (p=0.389). No signi ficant difference between control and phage-treated gr oups (Control vs. CK-2 + 3a, p=0.096; Control + FeCl3 vs. CK-2 + 3a + FeCl3, p=0.280).

PAGE 109

97 0 1 2 3 4 5 6 7 8 Control (RT)Control (32.5C)CK-2+3a (RT)CK-2+3a (32.5C)log CFU/g-tissue* # * Figure 3-23. Effect of water temperatur e on the efficacy of phage treatment of V. vulnificus FLA042-infected oysters. Oysters were infected with FLA042 for 7 h. The oysters were then incubated overnig ht in fresh seawater at either RT or 32.5C and treated with or without 1 x 108 PFU/mL each of phages CK-2 and 3a (n=10 each group). Bars represen t mean standard deviation. * significantly different compared to its corresponding control group (Control (RT) vs. CK-2 + 3a (RT), p=8.7 x 10-5; Control (32.5C) vs. CK-2 + 3a (32.5C), p=1.2 x 10-6). # significantly different compared to control (RT)(p=5.6 x 10-4). Altogether, we developed an oyster infection model for V. vulnificus enabling us to stably infect oysters with high levels of V. vulnificus . The establishment of this oyster infection model further enabled us to ex amine the potential of decontaminating experimentally infected oysters with phage s. Although phage 3a or CK-2 alone were unable to significantly lower the level of c ontamination in oysters, their combination resulted in significant decreases in the level of contaminati on in experimentally infected oysters ranging from a low of 0.9 logs to a high of 2.2, depending on the duration of infection of the oysters with V. vulnificus . With the use of pGTR902 marker plasmid and

PAGE 110

98 an Amp enrichment assay, we were able to de monstrate that in experimentally infected oysters a subset of the population of V. vulnificus was not adequately growing to be phage-sensitive. Specific Aim 3: Phage Treatment of Naturally Infected Oysters. With the limited success of phage treatmen t of experimentally infected oysters, attempts were undertaken at decontaminating naturally infected oysters with a pool of V. vulnificus -specific phages. Because the V. vulnificus strain composition of naturally contaminated oysters would be unknown, we hy pothesized that we would need a cocktail of phages for effective treatment. With the characterization of host range, morphology, and level of infectivity of the phages in our collection by using the soft agar and the microtiter phage typing methods, suitable phages were selected for use. In the experimentally infected oyster assays, detailed in Specific Aim 2, the level of contamination of a V. vulnificus spontaneous Rif-resistant stra in in oysters was measured with the use of LB-N -Rif plates. However, such an approach for enumeration of V. vulnificus contamination in naturally infected oysters would be unacceptable because the natural V. vulnificus strains would be sensitive to Rif and using plain LB-N plates would enable the growth of the other bacterial contaminat es of the oysters. Other technical approaches that are wide ly used for the enumeration of V. vulnificus contamination in naturally infected oysters were used for these experiments. This involved a selective medium, VVM, for V. vulnificus (58,59) and a DNA-hybridization probe specific to the V. vulnificus hemolysin/cytolysin gene (60).

PAGE 111

99 Quantification Methods Evaluated for Enumeration of V. vulnificus Naturally Occurring in Oysters We tested two different selective media wh ich have been widely used to enumerate the number of V. vulnificus in naturally infected oysters, TCBS and VVM, to evaluate their level of sensitivity in detecting V. vulnificus . TCBS (thiosulphate-citrate-bile saltssucrose) contains sucrose as the main car bon source and bile salts for selection of V. vulnificus . V. vulnificus grows as green colonies on TCBS. On the other hand, the selective medium VVM contains 105 U/L of the antibiotics polymyxin B and colistin methanesulfonate for selection of V. vulnificus , which is insensitive to these antibiotics. VVM contains D-(+)-cellobiose as the primary carbon source, thus V. vulnificus grows as yellow colonies on this medium . Because some selective media often yield lo w plating efficiencies for the bacteria of interest, we examined the pl ating efficiencies of different V. vulnificus strains on VVM and TCBS. Log phase cultures of V. vulnificus M06-24/O and MLT 403 were plated on each selective medium and on nonselec tive LB-N plates. The number of V. vulnificus colonies on VVM plates was comparable to that on LB-N plates. However, the number of the colonies on TCBS plates was consider ably lower compared to that on either VVM or LB-N plates. In fact, at th e dilutions that were utilized for this experiment no colonies were observed on TCBS, while approximately 107 CFU/mL were observed on both VVM and LB-N (data not shown). Because VVM exhibited no detrimental effect on plating efficiency, as opposed to TCBS, VVM plates we re chosen for use for the enumeration of V. vulnificus in oysters. To test the selectivity of VVM and TCBS, V. cholerae and E. coli were plated on each selective medium. V. cholerae grew as green colonies on VVM and yellow colonies on TCBS. E. coli did not grow on either medium, supporting the

PAGE 112

100 selective nature of VVM or TCBS for vibrios. However, as reported by Cerda-Cuellar et al. nonV. vulnificus bacteria can grow either yellow on VVM {3649,3648} or green on TCBS, suggesting that due care most be taken in interpreting our studi es that all yellow colonies on VVM or green col onies on TCBS are positive for V. vulnificus . Having established the selectivity and sens itivity of VVM and TCBS for vibrios, we examined the usefulness of TCBS and VVM to quantitatively identify V. vulnificus in both naturally contaminated oysters and oysters experimentally infected with V. vulnificus FLA042. For the naturally infected oys ters, LB-N plates yielded nearly 10fold higher numbers of colonies compared with yellow colonies on VVM plates (Fig. 324). This result was expected because V. vulnificus comprises only a fraction of the total bacterial flora of oysters, and all of the cu lturable bacteria would grow on LB-N while only a subset would grow on VVM, by design. However, a significant difference of 2.4 logs was observed between yellow CFU on VVM plates compared to green CFU on TCBS plates (p=6.3 x 10-4). This result suggested that either VVM was not selective enough and permitted non-vibrios to grow, or that TCBS was not enabling efficient plating of V. vulnificus from oysters. Based on the results above with V. vulnificus grown in culture, we favored the former hypothesis. In experimentally infected oysters, similar levels of FLA042 were enumerated on LB-N and VVM plates; however, a significantly lower number of FLA042 was enumerated us ing TCBS plates (p=0.004). Therefore, VVM plates were used to enumerate V. vulnificus in naturally contaminated oysters treated with a pool of phages, taking into ac count that not all ye llow colonies on a VVM plate are V. vulnificus .

PAGE 113

101 0 1 2 3 4 5 6 7 Naturally infected oystersFLA042 infected oysterslog CFU/g-tissue LB-N VVM TCBS* * Figure 3-24. Plating efficien cies of LB-N, VVM, and TC BS on marker oysters and on V. vulnificus FLA042 experimentally infected oysters. Oysters were either sampled immediately after arrival into the laboratory or sampled after an experimental infection with FLA042. The oysters were harvested, homogenized, serially diluted and pl ated on LB-N, VVM, and TCBS plates (n=4 each group). Bars represent mean standard deviation. * significantly different compared to LB-N or VVM (N aturally infected oysters: LB-N, p=4.9 x 10-4; VVM, p=6.3 x 10-4)(FLA042 infected oysters: L-BN, p=0.004; VVM, p=0.003). After the determination that VVM was suitable for enumeration of V. vulnificus in naturally contaminated oysters, phage treatm ent of naturally infected oysters with a cocktail of phages was performed. Phages with broad host ranges were selected using the data collected with the soft agar overlay a nd microtiter phage typing assays from Aim 1. Phages 153A-5, 153A-8, 154A-8, 3a and 4a were selected for treating naturally infected oysters because they exhibited broad host range s using the soft agar overlay phage typing method.

PAGE 114

102 Naturally infected market oysters were placed in sterile seawater. Immediately afterwards, half of the oysters were treated overnight with the cocktail of phages at 1 x 108 PFU/mL for each phage. Oysters were ha rvested, homogenized, serially diluted, and plated on VVM plates. All flat , yellow colonies typical of V. vulnificus on VVM plates were considered as V. vulnificus . In the phage cocktail-t reated group, the level of V. vulnificus was significantly decreased by 1.5 logs compared to the control group (p=0.052)(Fig. 3-25). This marginally signifi cant decrease in the phage-cocktail treated oysters demonstrated the potential of phages to lower the level of V. vulnificus contamination in market oysters. Furthermor e, the decrease of 1.5 logs observed in the phage-cocktail treated oyster s was comparable to the approximate 1 log decrease observed in the experimentally infected oysters that were infected with FLA042 for either 6 h or 24 h and treated with phages CK -2 and 3a. In fact, the level of V. vulnificus contamination could be artif actually high and the decontamin ation artifactually low in market oysters due to the po ssibility of counting some of the yellow colonies as V. vulnificus . To confirm and extend the first phage-coc ktail treatment of naturally infected oysters, three additional phages th at had broad host ranges were added to the in itial pool. Additionally, levels of V. vulnificus contamination of some of the market oysters were analyzed before the oysters were placed into water to evaluate changes in the levels of contamination caused by the incubation of oys ters in seawater during the experiment. Another set of market oysters was incubated overnight in ster ile seawater (control group), and a third group of oysters was placed into seawater and treated overnight with the expanded cocktail of phages at a lower concentration of 1 x 107 PFU/mL per phage.

PAGE 115

103 Oysters were harvested, homogenized, serially diluted, and plated on VVM plates. The incubation of oysters overnight in seawat er significantly increased the level of V. vulnificus contamination in oysters by 1.1 logs compared to the oysters that were harvested without being placed into seawater (p=0.002)(Fig. 326). This result showed that V. vulnificus had replicated in the oysters duri ng the overnight incubation. Treatment with the expanded pool of phages caused a 1.4 log decrease in V. vulnificus compared to the control group (p=3.7 x 10-4)(Fig. 3-26). This decrease in V. vulnificus contamination in the oysters was comparable to that obs erved in the first phage-cocktail treatment experiment detailed above. However, the number of V. vulnificus in the phage-treated group was not significantly different from that of the oysters harvested without ever being placed into seawater. This result s uggested that phage treatment only killed the population of V. vulnificus that grew as a result of the oys ters being placed in seawater during the overnight phage cockta il treatment period. This resu lt was also consistent with the hypothesis that a sign ificant portion of the V. vulnificus population is not growing in the oysters, as was hypothesized during treatment of experimentally infected oysters with phages CK-2 and 3a, the pGTR902 marker pl asmid segregation experiments, and the ampicillin experiments. Because VVM plates are not absolutely selective for V. vulnificus , we also used the VVAP DNA hybridization probe to examine the effectiveness of phage cocktail treatment in naturally V. vulnificus -contaminated oysters. Market oysters were placed in fresh, sterile seawater. Half of the oysters were immediately trea ted overnight with a cocktail of phages containing al l the phages in our collection with the exception of phage CB-1, at 1 x 107 PFU/mL for each phage for either 24 h or 48 h. The 48 h phage cocktail

PAGE 116

104 treatment period was included to examine if a longer phage treatment period would improve decontamination of naturally occurring V. vulnificus in oysters. No oysters were sampled before being placed into seawater. 0 1 2 3 4 5 6 7 ControlPhage treatedlog CFU/ g-tissue Water Oysters* Figure 3-25. Effect of treatme nt of market oysters with a cocktail of phages. Market oysters (n=8) were placed in fresh seawater and inc ubated overnight. During the overnight incubation pe riod, half of the oysters were treated with a cocktail of phages (phages 153A-5, 153A -8, 154A-8, 3a and 4a) each at 1 x 108 PFU/mL. Oyster were harvested, homogenized, serially diluted, and plated on VVM. Bars represent mean standard deviation. * significantly different compared to control group (p=0.032).

PAGE 117

105 0 1 2 3 4 5 6 Shucked same dayControlPhage treatedlog CFU/ g-tissue Water Oysters* # Figure 3-26. Effect of treatme nt of market oysters with a cocktail of phages. Market oysters were placed in fresh seawater and incubated overnight. During the overnight incubation period, half of the oys ters were treated with a cocktail of phages (phages 108A-9, 110A-7, 153A-5, 153A-7, 153A-8, 154A-8, 3a, and 4a) each at 1 x 107 PFU/mL (n=5 each group). Some market oysters were also sampled immediately after their arrival into the laboratory (Shucked same day group)(n=4). Oyster were harvested, homogenized, serially diluted, and plated on VVM. Bars represent mean standard deviation. * significantly different compared to control group (p=3.7 x 10-4). # significantly different compared to shucked same day group (p=1.7 x 10-3). No significant difference between shucked same da y and phage treated group (p=0.310). After 24 h of phage treatment, V. vulnificus contamination was significantly lowered by 1.1 logs compared to the contro l group (p=0.009)(Fig. 327). No further decontamination of V. vulnificus was observed with the longe r phage treatment of 48 h (p=0.012). The > 1 log decrease in V. vulnificus contamination from this phage cocktail treatment of naturally inf ected oysters using the VVAP hybridization probe confirmed the results observed in the experiment s using VVM plates for enumeration of V. vulnificus . Moreover, these results suggested th at a 24 h phage cocktail treatment is

PAGE 118

106 sufficient for phage treatment of naturally V. vulnificus -infected oysters, since no further killing was observed with a 48 h phage cocktail treatment. 0 1 2 3 4 5 2448 period of treatment (h)log CFU/g-tissue Control Phage cocktail* * Figure 3-27. Effect of treatme nt of market oysters with a cocktail of phages. Market oysters were placed in fres h seawater and incubated for either 24 h or 24 h. During the incubation periods, half of the oysters were treated with a cocktail of all the phages in our collection, wi th the exception of phage CB-1, each at 1 x 107 PFU/mL. Oyster were harvested, homogenized, serially diluted, and plated on LB-N. Using the LB-N plat es, colonies were enumerated and verified with the use of the VVAP DNA-hybridization probe (n=5 each group). Bars represent mean standard deviation. * significantly different compared to control group (24 h, p=0.009; 48 h, p=0.012). Although the use of VVM plates for enumeration of V. vulnificus was found to be very efficient in terms of time and effort, re sults were obtained a da y after the plating of the oyster homogenate onto VVM plates, there was considerab le concern that the yellow colonies on the VVM plates were truly V. vulnificus or other bacterial species that cause the same yellow colonies. Thus, the us e of the VVAP DNA-hybrid ization technique,

PAGE 119

107 although more time consuming compared to th e VVM plates, was considered the optimal method for enumeration of V. vulnificus in market oysters due to the probes specificity for V. vulnificus In summary, we applied the previous methods and results to attempt to decontaminate market oysters of natural V. vulnificus contamination. This included use of phage typing methods to choose pools of phages for treatment and culture and DNA hybridization methods to identify V. vulnificus from oysters. The best decontamination results yielded only approximately 90% decreases in V. vulnificus levels in oysters. Although promising, this level of efficiency is not likely to be acceptable to the seafood industry. Based on results obtained with the ex perimental infection model, we feel it is likely that there is a population of V. vulnificus in the oysters that is recalcitrant to killing by phages added to the water – either because they are not growing or because they are physically inaccessible to the phages.

PAGE 120

108 CHAPTER 4 DISCUSSION The Potential use of V. vulnificus -Specific Phages in Treating V. vulnificus Infections Since the reports in the 1980’s by Smith a nd colleagues reevaluating the potential of phage therapy (8-11), numerous other stud ies from other laboratories have reported successful treatment of bacterial infections wi th bacteriophages (reviewed in (61)). With these positive reports on phage therapy, our laboratory under took a project to assess the potential of phage thera py using a mouse model for V. vulnificus infection (55). V. vulnificus -specific phages were examined for use as a treatment of local and systemic infections by V. vulnificus . In doing so, our laboratory demonstrated that phages are capable of treating such infections when admi nistered at an appropr iate time after the infection. With these positive findings, the studie s detailed in this thesis were undertaken to evaluate the potential of using V. vulnificus phages to decontam inate oysters either experimentally or naturally infected with V. vulnificus . Isolation and Characterization of Bacteriophages Soft Agar Overlay Phage Typing Method The specificity of phages for infecting host bacterial stra ins has been widely used as a technique, i.e., phage t yping, to differentiate strains of bacteria. Besides bacterial identification purposes, the conventional soft agar overlay phage typing method has been utilized to characterize such phage propert ies as host range, plaque morphology, and level of infectivity. With the asse ssment of the characteristics of each phage, a more accurate determination of which phage or phages have pot ential for use as therapy is possible. For

PAGE 121

109 example, phages that exhibit broad host range s would be more useful because they can infect and lyse many strains of a bacterial species, increasi ng the probability of successful elimination of a bacterial inf ection. Moreover, phages with the most potential for phage therapy applications exhibit complete lysis on a soft agar overlay assay, because this demonstrates the ability of a phage to in fect and lyse every cell of the bacterial population. Phages with such efficient infect ion and lysis are more likely to completely and efficiently eliminate a bacterial infection in an individual. For these reasons, the soft agar overlay phage typing method was used w ith all of the phages in our collection to determine properties such as host range and level of infectivity for later consideration in the V. vulnificus decontamination experiments. A scoring scheme was developed for the soft agar assays to document the level of infectivity of each of the phages to 57 different strains of V. vulnificus . A clear confluent plaque was designated as 5, which we cons idered optimal for a phage therapeutic application. Phage-bacterial stra in combinations that resulted in a turbid confluent plaque were scored as a 4, while combinations that resulted in only a few clear plaques were scored as 2 or 3, depending on how many plaq ues were developed. A turbid confluent plaque suggested that the phage could lyse a large proportion of the bacteria, causing a confluent plaque, but that resistant bacteria l cells either were present in the initial population or arose after infecti on with phage. If the initial population contained a subset of bacteria that were resistan t to phage infection and lysis, then a turbid plaque could form due to the growth of the resistant bact eria. The resistance could be explained in these situations by either a mi xed population of two different V. vulnificus strains, or a subset of the population of a single strain of V. vulnificus being resistant to the phages.

PAGE 122

110 For the mixed population of two different V. vulnificus strains scenario, one strain of the bacteria can be completely susceptible to phage infection and lysis whereas a second strain of bacteria in the same culture could be resistant to the pha ge, allowing these cells to continue growing. Alternatively, a subs et of the population of a single strain of V. vulnificus could contain the ability to suspend the expression of a specific receptor that is necessary for certain phages to infect the bacteria, afford ing these cells resistance to phage infection, although this is probably unl ikely. Another explana tion of the confluent turbid plaque formations was that the phage may be a temperate phage, resulting in the generation of lysogens. However, inspecti on of individual plaques during the full-plate titer assays of each phage neve r revealed plaques with a “bu lls-eye” appearance, a turbid center with a clear ring, which is typical of temperate phages. Furthermore, no phage consistently caused turbid confluent plaques. All phages that caused a confluent turbid plaque to at least one V. vulnificus strain also caused a clear confluent plaque on at least four V. vulnificus strains. Although these findings are not direct proof that the phages in our collection were all virulent phages, they suggest that the more likely explanation for the turbid confluent plaques was that a subset of th e initial bacter ial population was resistant to the phage, either because the initially homogenous population had accumulated resistant mutants or that the bacter ial strain was actually a mixed culture of at least two strains of V. vulnificus . To determine if a bacteria l strain is actually a mixed culture, the strain can be str eaked on a LB-N plate for isola tion of colonies. Different colony morphologies would indicat e that the culture contains more that one strain of bacteria. The individual coloni es also should not exhibit a tu rbid confluent plaque with the soft agar method because they should repr esent a pure bacterial strain that should

PAGE 123

111 either be sensitive or resistant to the phage. Furthermore, examination of total cellular proteins by SDS-PAGE could also be used to determine if the isolated colonies are of the same strain. With the exception of V. vulnificus FLA111, all V. vulnificus strains that produced a turbid confluent plaque were also observed to produce other plaque formations, ranging from a few clear plaques to clear conflu ent plaques. These findings suggest that the more likely cause of the tu rbid confluent plaque s by these strains of V. vulnificus was the existence of a homogenous strain of V. vulnificus containing cells that were resistant to the phage. On the other hand, V. vulnificus FLA111 developed confluent turbid plaques with al l of the phages that were infectious to this bacterial strain, suggesting that this strain of V. vulnificus may be a mixed culture of two different strains of V. vulnificus . In other phage-bacterial strain combina tions, individual clear plaques, as opposed to confluent spots, were observed at the s pot where phage was dropped onto the soft agar overlay, in spite of the fact that 106 PFU were dropped onto the bacteria. This result suggested that the phage infected the particular V. vulnificus strain with very low efficiency. Limiting factors su ch as bacterial restriction m odification systems, inefficient adsorption, or defective mechanis m for lysis of the bacteria are possible explanations for these results. The isolation and amplification of such a clear plaque could result in the isolation of a phage that may now produce a clear confluent plaque at 106 PFU on that particular V. vulnificus strain. However, this was not examined. Seventy-five percent of our V. vulnificus strains scored a 5 for at least one of our phages. However, if a score of 4 was also considered, 3 additional V. vulnificus strains were sensitive to at least one of our phages, increasing the percentage of bacterial strains

PAGE 124

112 to 81% that were infected by at least one of the phages. The ability of our phages to infect and lyse up to 81% of the V. vulnificus strains indicated that experiments testing the potential of decontaminating oysters of V. vulnificus infections were possible using a cocktail of these phages. For those experi ments, a cocktail of phages with broad host ranges, detailed below, would hopefully enab le the killing of a large proportion of naturally occurring V. vulnificus strains contaminating oysters. The host range for each of our phages was studied using the soft agar overlay method for consideration of which phages to use for the phage treatment experiments on V. vulnificus -infected oysters, especially in natu rally infected oysters. For example, phage 3a, isolated from oyster homogenate, ha d a remarkably broad host range of 61% of the V. vulnificus strains in our collection using a score of 5 or 68% if a score of 4 is used. Phages 153A-7, 153A-8, 108A-9, and 154A-8 also exhibited broad host ranges, infecting 58% to 68% V. vulnificus strains in our collection, depending on whether the cutoff score was 4 or 5. Since phages 3a, 153A-7, 153A8, 108A-9, and 154A-8 demonstrated broad host ranges, they were include d in the cocktail used in the phage treatment experiments of naturally V. vulnificus -infected oysters, as deta iled in Specific Aim 3. On the other end of the spectrum of host range, phage CK-2 had the narrowest host range, infecting only V. vulnificus MLT403 with a score of 5. Phages 154A-9, 7-8A, and CB-1 also had narrow host ranges, infective to only 4 of the 57 strain s each (7%). Phage CK-2 was used for phage treatment of oysters experimentally infected with V. vulnificus FLA042. Although phage CK-2 had a narrow ho st range, it had certa in characteristics that were considered beneficial for the usag e of the phage for tes ting the potential of decontaminating V. vulnificus -infected oysters with phages. For example, phage CK-2

PAGE 125

113 created the clearest and larg est plaques when infecting V. vulnificus MLT403 or its Rifr derivative, FLA042. Additionally, we could obtai n very high titer lysa tes in the range of 1012 PFU/mL to produce large numbers of phage for our use. With the use of the soft agar method, we determined that the phages in our collection can infect up to 81% of the V. vulnificus strains with a score of at least 4. Thus, we felt that this level of coverage against V. vulnificus infections was sufficient for phage treatment of naturally infected oysters, detailed in Specific aim 3. Nevertheless, with the data obtained in the soft agar assays , phage enrichment and isolation assays were performed against V. vulnificus strains that were either insensitive or slightly insensitive to the phages in our collection, resulting in the isol ation of phages infective to some of these strains. Microtiter Phage Typing Method Although the soft agar overlay phage typi ng method was useful in characterizing the phages in our collection, a microtiter me thod was developed to minimize certain disadvantageous aspects of the soft agar ove rlay method. The soft agar overlay method required large amounts of materials fo r the testing of the 57 strains of V. vulnificus and an overnight incubation period to see results of the interaction between phage and bacterial strains. The soft agar overlay method also required human inspect ion and interpretation of the qualitative aspects of plaque formations (clear vs. turbid) of each phage; hence, the method was susceptible to some subjectivity. The appropriate medium, d iluent, concentration of V. vulnificus , and concentration of phage were studied for optimization of the microtiter assays. Since a sea salt effect was observed on the infectivity of certain phages, LB-N and LB-SW were studied in the microtiter method. In LB-SW V. vulnificus grew to an average 24% higher density in a

PAGE 126

114 shorter period of time compared to LB-N. The use of LB-SW also increased the ability of certain phages to lyse certain bacterial st rains. For example, a remarkable effect of LB-SW on phage 152A-2 infection and lysis of V. vulnificus MO6-24/O, LL728, and VV1009 was shown using the microtiter method (Figs. 2a, 2c, and 2d). Phage 152A-2 significantly lysed V. vulnificus MO6-24/O, LL728, and VV1009 in LB-SW, whereas in LB-N no significant lysis was observed with phage 152A-2. The use of SW instead of BSG as diluent also was beneficial for the growth of V. vulnificus in the microtiter assays (Figs. 3a-c). Higher bacterial densiti es were consistently observed with V. vulnificus M06-24/O, MLT403, and 2400112 when diluted in SW instead of BSG before their addition into LB-N. These resu lts suggested that the use of SW in the medium and/or as diluent provided a more natural environment for V. vulnificus and its phages in which to interact. Furthermore, we hypothesized that the sea salts contai ned important cations, such as Mg+2 and Ca+2, that are important for the adsorpti on of certain phages to bacteria. This sea salts effect was further demons trated on the infection and lysis of V. vulnificus LL728 by phage CB-1 (Fig. 3-1) The most su itable concentration of bacteria was also studied for the optimization of the assay; 1 x 107 CFU/mL for V. vulnificus was the optimal concentration to enable a ra pid determination of phage lysis. After the optimal microtiter parameters were established, the final microtiter experiments were performed on the seven V. vulnificus strains commonly used in our laboratory. The data compiled with this assay were then compared to the results of the soft agar overlay assays performed simultaneously, and there was general agreement between the methods. For a pha ge-bacterial strain combinati on that had no effect using the soft agar overlay method, the OD readings in a microtiter assa y usually increased

PAGE 127

115 with time and eventually plateaued similarl y to the no phage control. For a phagebacterial strain combination that had any e ffect using the soft ag ar overlay method, the OD readings in a microtiter assay usually de creased within the firs t couple of hours and plateaued at that level, alt hough some levels decreased even more or others gradually increased over time. The mi crotiter experiments of M 06-24/O and VV1009 completely agreed with the soft agar overlay phage typi ng assays. Phages that caused a significant decrease in bacterial densities also caused a clear conf luent plaque in the soft agar overlay assays. Moreover, the significan t lysis of M06-24/O and VV1009 in the microtiter assays was seen between 2 h and 4 h, which was a significantly shorter period of time for results compared to an overnigh t incubation period for the soft agar overlay experiments (Figs. 7a and 7b). Interestingl y, phage CK-2 caused a small but detectable decrease in OD of M06-24/O and VV1009 in the microtiter experiments, although no plaques were observed in the soft agar ove rlay phage typing assays. Therefore, the microtiter assays might be more sensitive than the soft agar method. The minute effect of phage CK-2 on these V. vulnificus strains may correspond to an incomplete infection by the phage. One possible explanation for this incomplete infection is the expression of a receptor needed by phage CK-2 for attachment and infection of the bacteria at very small numbers by only a subset of the bacterial population. Thus, the small probability of phage CK-2 to attach and infect the bacteria is reflective of the small but yet significant lysis observed on the microtiter a ssay, while no effect is observe d in the soft agar overlay assays. Another possible explanation for the incomplete infection of M06-24/O and VV1009 by CK-2 is that the phage has a defect ive replicating and/or lysing process for these strains of bacteria.

PAGE 128

116 As seen with M06-24/O and VV1009, phages that caused a clear, confluent plaque on a V. vulnificus strain using the soft agar ove rlay phage typing assays always significantly lysed the strain us ing the microtiter method. In contrast, phages that caused turbid confluent plaques on the soft ag ar overlay method did not always cause measurable lysis using the microtiter method. For example, for V. vulnificus MLT403 all phages that caused a clear confluent pla que (153A-7, 108A-9, 110A-7, CK-2, and 3a) caused significant lysis in the microtiter ex periment (Fig. 3-7e). However, phage 152A9, which caused a confluent turb id plaque on MLT403, did not cause significant lysis in the microtiter assay. A similar effect was observed with phage 153A-5 on MLT365 (Fig. 3-7c). However, phages 108A-9 and 3a, which caused a confluent turbid plaque in the soft agar overlay assays on MLT365, caused si gnificant lysis in th e microtiter assay. These results suggest that only some phages th at cause turbid conflu ent plaques using the soft agar overlay phage typing method may be detected using the microtiter method. As stated above, we considered a turbid conflu ent plaque on a soft agar overlay plate to represent a situation in which a number of bacterial cells was resistant to phage infection and, thus, the production of a turbid confluent plaque. A possible explanation for the occasional detection of lysis by phages that cau sed confluent turbid plaques is different levels of lysis. Some of the phages ma y cause enough lysis of the phage-sensitive bacterial cells to detect a d ecrease in bacterial density. However, some phages may not be able to cause enough lysis of the phage-sens itive bacterial cells to detect a decrease in bacterial density by the reader . Although the difference in turbidity of the confluent plaques cannot be detected by eye using a so ft agar overlay assay, such a difference in turbidity might be detect ed by the microtiter method.

PAGE 129

117 The microtiter method also revealed differe nces between phages with regard to the time required for lysis of different V. vulnificus strains. For example, phages 108A-9, CK-2, and 3a began to significantly lyse MLT403 between 4 h and 5 h. In contrast, phages 153A-7 and 110A-7 began to signifi cantly lyse MLT403 between 7 h and 8 h (Fig. 3-7e). These findings may have a si gnificant impact on the important task of choosing the appropriate phage or phages for use as a therapy. As stated before, V. vulnificus may possibly reside in oy ster tissues that phages are unable to infiltrate. Thus, the inability by phages to penetrate oys ter tissues may result in limited success of phage treatment of either naturally or expe rimentally infected oysters. However, an enhanced treatment of V. vulnificus -infected oysters may be achieved by treatment of infected oysters with a phage that lyses V. vulnificus at a slower rate. The vibrios infected in the accessible areas might invade into the phage-inaccessi ble areas and deliver the phage. The use of such a strategy may re sult in phages indirec tly “transported” to areas where phages would not otherwise be pr esent. The phage progeny can now infect and lyse V. vulnificus in these areas, resulting in an increase in the level of decontamination of V. vulnificus in oysters. Although not a part of this thesis, an enhanced treatment of V. vulnificus -infected mice, as detailed in the Introduction section, may have been achieved by treatment of infected mice with a phage that lyses V. vulnificus at a slower rate. An infected, ye t unlysed bacterium may shuttle a phage into the intercellular fluid were infection of V. vulnificus is reportedly localized, but where phages would not otherwise be present due to their inability to pass through the vascular endothelium. After the lysis of this phage-infected bacterium, progeny phages would now have the ability of infecting other bacteria that are localized in the

PAGE 130

118 intercellular fluid. A similar approach was reported by Broxmeyer and colleagues, detailed in the Introduction (20). By using a non-virulent mycobacterium, M. smegmatis , infected with a lytic phage that produced non-stable lysogens, TM4, they were able to treat a M. tuberculosis or M. avium infected macrophage monolayer. M. smegmatis was used as a vehicle to transfer phage TM4 into the intracellular environment of the macrophage monolayer where M. tuberculosis or M. avium were residing for treatment of these infections. Although infectivity of phages generally co rresponded between the microtiter and soft agar overlay methods, results obtained for V. vulnificus 2400112 and LL728 did not fully agree between the methods (Figs. 7f-g). Infection of V. vulnificus 2400112 with phages 108A-9, 152A-9, 153A-7, 154A-8, and 153A-8 and V. vulnificus LL728 with phage 153A-8 caused clear conf luent plaques on the soft ag ar overlay assays; however, these phages did not significantly lower the ODs in the microtiter assays. Furthermore, phages 153A-7, 154A-8, 108A-9, and 110A-7 that yielded scores of 2 or 3 on the soft agar overlay assays caused no significant ly sis of LL728 in the microtiter assays. However, when the microtiter assays for 2400112 and LL728 were repeated at an elevated temperature of 37C instead of the usual RT, the results completely corresponded between the methods, even for the phages that scored a 2 or 3 on the soft agar overlay method (Figs. 8a-b). Thes e results suggest that these strains of V. vulnificus require an elevated temperature to express th e receptors for these pha ges to infect and for replication. Furthermore, the detecti on of phages 153A-7, 154A-8, 108A-9, and 110A-7 on LL728 using the microtiter assays at 37C, even though they only scores of 2 or 3 in the soft agar overlay assays, s uggests that at this elevated te mperature the detection of all

PAGE 131

119 different types of plaques may be possi ble. The positive results observed with V. vulnificus 2400112 and LL728 at 37C also suggest s that the performance of the microtiter assays could be enhanced for the other strains of V. vulnificus at 37C instead of at RT. V. vulnificus preferentially grows at 37C comp ared to RT. At this preferential temperature, V. vulnificus may increase the ability of the pha ges to infect and replicate. Additionally, other phages significantly lysed stains 2400112 and LL728 in the microtiter method at 37C, even though they di d not cause lysis in the soft agar overlay assays, for example, phage 154A-9 (Figs. 8a and 8b). Possible reasons for the differences in the detection of infection of phages ma y be due to environmental differences the bacteria and phage encounter between both me thods. In the microtiter assays a liquid environment is encountered, while in the soft agar overlay assays a semi-solid environment is encountered by both phages a nd bacteria. Moreover, as detailed above, the difference in temperature in the microtiter assays may also have an impact on phage infection and lysis. Emphasizing this poi nt, although 153A-5 did not cause any plaques in the soft agar overlay phage typing assa ys, it significantly lysed strain 2400112 in the microtiter assay at 37C by 8 h (Fig. 3-8a) but not at RT (Fig. 3-7f). Interestingly, our laboratory reported sporadic prot ection of mice in fected with V. vulnificus 2400112 by phage 153A-5, but concluded that these re sults were of “questionable biological significance” (55). The temper ature condition in which this phage can infect and kill V. vulnificus 2400112 may have been pres ent in several of the V. vulnificus 2400112infected mice but not in all, thus a sporadic protection by the phage. For example, some of the infected mice may have had a more rapi dly decreasing temperature due to the onset

PAGE 132

120 of V. vulnificus -induced sepsis, resulting in a decrease in the ability of phage 153A-5 to infect and lyse V. vulnificus 2400112. The development of the microtiter method demonstrated the potential of this assay to examine the level of infectivity of phages on V. vulnificus strains in a small scale format. The ability of testing different bacterial strains of V. vulnificus in various combinations with different phages in a 96well microtiter plate is highly efficient and productive. Such an extensive experiment using the soft agar overlay method would require significantly more materials. Furt hermore, the sensitivity of detecting minute differences in bacterial densities by the microtiter plate reader also allows for rapid determination of phage sensitivity of V. vulnificus strains, usually detected in as little as 2 h. Furthermore, this method decreased the leve l of subjectivity that is encountered with the soft agar overlay methodeither it lyse d or not. The examination of phages in the microtiter assays at RT and 37C may indi cate temperature requirements for certain phages to infect and lyse certain V. vulnificus strains and different time requirements for lysis of bacteria. Lastly, th ese two properties, which are not easily examinable using the soft agar overlay method, may result in a more comprehensive understanding of the nature of the phages in our collection for later consideration in their usage in a therapeutic application. After the examination of the phages in our collection using both phage typing methods and the isolation of additional phages that could be useful for phage treatment applications, studies to examine the potential of phages to decontaminate V. vulnificus infected oysters were now possible. The establishment of an experimental oyster infection model was deemed an advantageous, controlled method to examine the

PAGE 133

121 potential of phages to eliminate V. vulnificus in oysters. However, attempts at stably infecting oysters with substantial levels of V. vulnificus similar to that observed in naturally infected oysters, as high as 6 log CFU/g tissue, were unsuccessful. Infecting oysters with V. vulnificus resulted in unacceptably low levels of V. vulnificus , usually between 3 and 4 log CFU/g tissue (Fig. 3-10) . This level of contamination was not considered to be high enough for phage treatment experiments, because the minimal level of detection of V. vulnificus in oyster tissues using LB-N plates was 1.3 log CFU/g tissue. We would therefore have difficu lty observing significant drops of V. vulnificus contamination in oysters trea ted with bacteriophages. Thus, we examined different parameters to achieve a higher le vel of infection in oysters. To increase the level of FLA042 contamina tion in oysters, oysters were pretreated with Rif prior to infection with FLA042. We hypothesized that the normal bacterial flora of the oysters was preventing the col onization of exogenously administered V. vulnificus in our oyster experiments. We further hypothesi zed that the pretreatment of oysters with Rif would eliminate the normal bacterial fl ora enabling more effective infection of oysters by the inoculated V. vulnificus strain. In fact, Rif-treat ed oysters did not yield any CFU on LB-N plates; apparently the Rif treatment was extrem ely effective in eliminating normal bacteria flora. The remarkable el imination of the normal flora, which also includes naturally contaminating V. vulnificus , with the use of Rif or any other antibiotic may be considered a potential method for dec ontaminating oysters. However, the U.S. government has a zero tolerance policy on th e usage of antibiotics for decontamination purposes on consumable products (GAO repor t number GAO-04-246). The Rif-treated oysters infected with FLA042 for 6 h had a significant increase in the level of

PAGE 134

122 contamination with FLA042 of 6.0 0.59 log CF U/g tissue (Fig. 3-10). The effect of longer duration of infection with V. vulnificus was also studied to examine if an infection longer that 6 h would result in an increase in the level of contamination in oysters. Infection of Rif-treated oysters with V. vulnificus FLA042 for longer than 6 h did not result in higher infection levels of FLA042 in oysters compared to Rif-treated oysters infected with FLA042 for 6 h. It has also been suggested that in the natural environm ent oysters acquire V. vulnificus infections during filter-feeding on algae. As opposed to filtering the vibrios directly, it is believed that the vibrios are att ached to the algae that the oysters filter. Thus, the effect of feeding oysters alg ae simultaneously with the infection of V. vulnificus was also studied for possible increases in the le vel of contamination in oysters. Infection of oysters with V. vulnificus FLA042 for 6 h coupled with feeding oysters algae did not result in any significantly higher infection levels of FLA042 in oysters compared to oysters that were infected for 6 h in the ab sence of algae (Fig. 3-10). Furthermore, the combination of Rif-pretreatment of oysters and feeding oysters algae simultaneously with the infection of V. vulnificus did not cause any higher levels of FLA042 infection compared to oysters that were only Rif-treated before infection with FLA042 (Fig. 3-10). Thus, we concluded that the treatment of oysters with Rif and infection with V. vulnificus for 6 to 7 h was the best method for inf ecting oysters to yield a high level of contamination. After the successful establishment of an e xperimental oyster in fection model, the examination of the potential of using bacteriophages to d econtaminate experimentally infected oysters was now possible. However, prior to any such experiments, naturally

PAGE 135

123 occurring phages in oysters and in the water housing the oys ters were enumerated using the soft agar overlay method. High levels of infectious phages naturally occurring in oysters and in the water could indicate a possible obstacle during phage treatment experiments involving experi mentally or naturally infected oysters. On V. vulnificus M06-24/O, which was exceptionally sensitiv e to many phages in our collection, 3.5 0.43 log PFU/g tissue were isolated from oyster homogenate (Fig. 3-13). On V. vulnificus MLT403, which was insensitive to ma ny of the phages in our collection, 2.0 0.30 log PFU/g tissue were isolated from oyster homogenate. DePaola and colleagues reported that the number of phages infecti ous to M06-24/O in oysters ranged from approximately 104 to 105 PFU/g tissue throughout the year, using a direct phage enumeration procedure similar to the soft agar overlay method we used for our experiments (56,57). Their fi nding, thus, corresponded with our findings; the slightly smaller number of phages enumerated on V. vulnificus M06-24/O by us compared to DePaola and colleagues may be related to an overnight incubation period of the oysters in seawater in our experiment, pr obably resulting in the “sheddi ng” of a few of the phages. No phages were enumerated in the water housing the oysters, most likely due to the levels being below the detectable level using the soft agar overlay method. Our inability to enumerate any phages in the water was sim ilar to the results of DePaola and colleagues in which they were unable to detect any pha ges in the seawater using the direct phage enumeration procedure. However, the water samples we tested were different to those experiments reported by DePaola et al.; we enumerated the number of phages in the water housing oysters after an incubation peri od, whereas DePaola et al. enumerated the number of phages from natural seawater. Ne vertheless, the inability of enumeration of

PAGE 136

124 phages lytic to V. vulnificus in waters surrounding oysters , either in a natural or experimental setting, suggests that the phages are mostly reta ined in oysters as opposed to freely flowing in the water. Having as high as 104 to 105 PFU/g tissue of naturally occurring infectious phages in oysters as repo rted by DePaola could result in interference of these phages with the attemp ts of treating either naturall y or experimentally infected oysters with phages. For example, some of the naturally occurring phages may be temperate phages that are infecting V. vulnificus , rendering them imm une to infection by the administered phages in our experime nts that have similar phage types by superinfection immunity. However, this ma y not be very plausible. Although we found high levels of infectious phages for M06-24/ O, we found smaller le vels of infectious phages for MLT403 in oysters. Since most of our experimental oyster infection experiments involved Rif resistant MLT403, FLA042, we felt that naturally occurring phages at the range of 102 PFU/g tissue should not significan tly interfere w ith our phage treatment experiments. We next examined the ability of phages to be retained by oysters using phage CK-2 by comparing the PFU in oyster tissues and th e water housing the oysters before and after a rinse. If similar concentra tions of phage CK-2 were presen t in oysters and in the water, the results would suggest that the phages are not retained by oysters, instead the phages are freely moving between the oysters and the wa ter. Prior to the commencement of this experiment, a portion of the oysters was samp led to enumerate the number of lytic phages infectious to MLT403. Unlike the previous experiment, no naturally occurring phages were found in the oysters or in the wate r housing the oysters. Oysters that were administered 1 x 107 PFU/mL of phage CK-2 had equal levels of phage CK-2 in the

PAGE 137

125 oysters and in the water, 7.3 0.13 log PFU /g tissue vs. 7.7 log PFU/mL, respectively (Fig. 3-14). However, after an overnight rins e in fresh seawater, the level of phage CK-2 in oysters was mostly retained compared to a significant decrease in phage CK-2 in the water, 6.7 0.12 log PFU/g tissue vs. 5.2 log PFU/mL, respectively. Altogether, this finding agreed with our earlier observation on naturally occurring phages and the experiments reported by DePa ola et al. (56,57) that oysters do retain phages infectious to V. vulnificus . As reported by DePaola et al., phages infectious to V. vulnificus are found throughout different oyster tissues and fluids. The lowest concentrations of phages were found in the hemolymph and the mantle fluid; however, the lowest concentrations of V. vulnificus were also found in these fluids. The relationship between phages and different oyste r tissues may block th e ability of oysters to eliminate the phages, resulting in the rete ntion of the phages. Furthermore, the fact that phages infectious to V. vulnificus are found throughout different oyster tissues and fluids may suggest possible elimination of V. vulnificus contamination with the administration of th erapeutic phages. The treatment of oysters experimentally infected with V. vulnificus FLA042 with either phage 3a or CK-2 separately at 1 x 108 PFU/mL did not result in any significant decrease in the level of V. vulnificus contamination in oyste rs (Fig. 3-15). We hypothesized that the combination of both phages 3a and CK-2 at 1 x 108 PFU/mL each could result in more killing of FLA042 in infected oysters. Since phages 3a and CK-2 most likely attach to different receptors, we hypothesized that the infection of cells by both phages 3a and CK-2 simultaneously may result in more efficient infection and killing, resulting in po ssible significant decreas es in the level of V. vulnificus

PAGE 138

126 contamination in oysters. Furthermore, th e use of two different phages would also increase the probability of killing the cells that may become resistant to one of the phages. Depending on the length of infec tion of oysters with FLA042, a significant decrease in the level of V. vulnificus contamination in oyste rs resulted with the combinational administration of both phages CK-2 and 3a. The highest decrease in contamination, as high as 2.2 logs (>99%), was observed with oysters that were infected with FLA042 for 1 h (Fig. 3-16). Smaller d ecreases, approximately 1 log (90%), were observed with oysters that were experimentally infected for 6 h or 24 h prior to treatment with phages CK-2 and 3a (Fig. 3-17). The difference in decontamination could indicate that the V. vulnificus were in a more susceptible physio logical phase at 1 h of infection compared to 6 h or 24 h. As stated above, by 6 h the level of infection of oysters with V. vulnificus seems to reach equilibrium. This l onger period of infection may allow the bacteria to reach equilibrium and attain a phys iological state that makes the bacteria more resistant to phage infection and lysis by phage s CK-2 and 3a. We hypothesized that this physiological state may correspond to a growth tr ansition of the bacteria into stationary phase. The transition of V. vulnificus growth to stationary phase might impede the ability of phages CK-2 and 3a to infect and lyse these bacteria. However, another plausible explanation is the ability of V. vulnificus to invade oyster tissues that phages are unable to penetrate, resulting in the protec tion of the bacteria from phage infection and lysis. This may explain the observed greater protection of the bacteria with a 6 h infection compared to a 1 h infection. The longer period of inf ection of oysters prior to phage treatment may have allowed the bacteria more ample time to invade oyster tissues that phages cannot infiltrate. If any of these e xplanations are true, then trea ting naturally infected oysters

PAGE 139

127 may be difficult, if not impossible, because oysters in the wild are infected with V. vulnificus for a considerable time. Another po ssible explanation for the inability of phages CK-2 and 3a to completely eliminate V. vulnificus FLA042 contamination in oysters may have been to the occurrence of spontaneous phage-resist ant bacterial cells. However, after testing several V. vulnificus colonies recovered from phage-treated oysters for susceptibility to phages CK-2 and 3a, all of the colonies were still susceptible to phage infection and lysis. One potential artifact of the decontamina tion of experimentally infected oysters treated with phages CK-2 and 3a was that the phages infected and lysed the bacteria only after the homogenizatio n of the oyster tissue. For exam ple, during the incubation period with phages CK-2 and 3a, V. vulnificus may have been residing in tissues that phages were unable to access; however, upon homoge nization of the oyster tissue, the phages would now be able to infect a nd lyse the bacteria. To test this possibility, tissues of oysters that were infected for 1 h with V. vulnificus were spiked with an equal concentration of phages CK-2 and 3a that th e phage-treated oysters in the experiments detailed above would have encountered dur ing phage treatment. This phage spikedoyster tissue was then homogenized and plated for enumeration of V. vulnificus . Treating V. vulnificus -contaminated oyster tissue with phage s immediately before homogenization did not result in any decontamination (Fi g. 3-16). Therefore, phages CK-2 and 3a decontaminated the oysters of V. vulnificus FLA042 during the treatment period of live oysters and not after the tissues were homogenized. Growth of V. vulnificus in Oysters As detailed above, one possibility for the lack of complete elimination of V. vulnificus was that the bacteria had slowed or stopped their growt h, rendering them

PAGE 140

128 phenotypically phage-resistant. Virulent phages depend on the metabolism of the host cell for replication, eventually involving the lysis of the bacterium to release phage progeny. The fact that phage treatment of oys ters infected for 1 h was more effective than treatment of oysters that were infect ed for 6 h or 24 h was consistent with the possibility that the bacteria had reached stationary phase. Furthermore, the leveling off of the level of V. vulnificus in experimentally infected oys ters may also be a direct, but not conclusive, evidence of the cessation of growth of V. vulnificus . Thus, two complementary methods were utilized to examine the level of growth of V. vulnificus in experimentally infected oysters: a marker plasmid and an Amp enrichment method. Marker Plasmid pGTR902 Assay A Rif-resistant M06-24/O strain of V. vulnificus , FLA077, containing the pGTR902 marker plasmid, which was constructed by Ange la Starks, was utilized to examine the level of growth of V. vulnificus . In the presence of arabinose, the marker plasmid is replicated by the bacteria. However, in the absence of arabinose, the marker plasmid is not replicated by the bacteria and segregates from the population of the bacteria during bacterial growth. The growth proportion was ca lculated as the ratio of the number of bacteria containing the marker plasmid to the total number of bacteria. Oysters that were infected for 7 h with FLA077(pGTR902) were analyzed for the number of bacteria containing or not containing the marker plasmid immediatel y after the 7 h infection period and after 4 h, 16 h, or 40 h incubation pe riod in fresh seawater (Fig. 3-18). The growth proportion after the 7 h infection was log -1.4. Howe ver, the growth proportions among the rinse periods were relatively consta nt at log -2.4 to l og -2.7. The relatively constant growth proportions indi cated a lack of growth of V. vulnificus FLA077(pGTR902) even after a 40 h rinse. Th e death proportion was calculated as the

PAGE 141

129 ratio of the number of bacter ia containing the marker plasmid at the various times to the number of bacteria containi ng the marker plasmid before rinsing. The death proportions of FLA077(pGTR902) were also relatively c onstant among the different rinse cycles ranging from log -0.6 to log -0.9. The rela tively constant death proportions among the sampling times indicated a lack of death of the bacterial cells. The relatively constant growth proportions and death proportions suggested that V. vulnificus was in stationary phase during the experimental oyster infection experiments. Ampicillin Enrichment Assays Further evidence of the lack of growth of V. vulnificus was obtained with the Amp enrichment experiments using both V. vulnificus FLA042 and FLA077. The antibiotic Amp works by the inhibition of enzymes re quired by bacteria for the synthesis of peptidoglycan. The inhibition of such enzyme s results in osmotic bacterial cell lysis. However, if bacteria are not replicating, the synthesis of peptidoglycan does not occur, and no osmotic bacterial lysis occurs. The mode of action of Amp can be used as a marker for bacterial replication. In our studies, if V. vulnificus in oysters were not killed after Amp enrichment, then this would indi cate a lack of growth of the bacteria. Oysters that were infected with V. vulnificus FLA042 for either 6 h or 7 h were treated overnight with 100 g/mL of Amp, resu lting in a decrease of CFU between 2.4 to 3.1 logs, (Figs. 19 and 20). Although a si gnificant portion (as high as 99.9%) of the population of V. vulnificus infecting oysters was susceptible to Amp, a level of V. vulnificus of approximately 103 CFU/g tissue was still not susceptible to the Amp treatments. The insensitivity to Amp by a s ubset of the population suggested a lack of growth by these bacteria. In comparison, phages CK-2 and 3a at 1 x 108 PFU/mL each decreased the level of V. vulnificus by 1.2 logs (97%), resulting in a level of V. vulnificus

PAGE 142

130 of approximately 105 CFU/g tissue not susceptible to phage treatments (Fig. 3-20). The approximate log drop in this experiment was consistent with our earlier results of phage treatment of oysters that were infected for 6 h. Another group of infected oysters was treated overnight with both Amp and phages CK2 and 3a. With the combinational treatment, the level of V. vulnificus infection decreased by 3.8 logs (Fig. 3-20). The decrease between the Amp only treated oysters and the combinational treatment of both Amp and phages CK-2 and 3a was not significan tly different. These results suggest that phages CK-2 and 3a killed the same bacteria that were killed by Amp treatment alone, although Amp alone killed more bacteria in total than phages CK-2 and 3a alone. The significantly better killing with Amp alone than with phages CK-2 and 3a alone were hypothesized to be differences between the phages and Amp, such as their mode of killing and/or physical properties. The more effective killing observed with Amp treatment, which readily diffuses through ti ssues and fluids of animals, may indicate a more efficient penetration of Amp through oyster tissues than phages CK-2 and 3a to kill the bacteria. This phenomenon may easily be explained by the larger physical size of phages CK-2 and 3a compared to the antibio tic Amp. Another possibility for the greater killing observed with Amp may represent a mi nute difference in the level of stationary phase the bacteria are exhibiti ng that may be detected with Amp but not with phages CK2 and 3a. Phages CK-2 and 3a may be unable to kill any bacteria that are not fully metabolically active, including the expressi on of necessary receptors for infection, whereas Amp can kill bacteria that are onl y slightly metabolica lly active. Another explanation may be that lethal concentra tions of Amp accumulated in non-actively growing bacteria in oysters during the Amp treatments. After the oyster tissues were

PAGE 143

131 homogenized and plated onto a rich medium , the bacteria transitioned to exponential growth resulting in the death of these bacteria by the mode of action of Amp. Since the marker plasmid assays involved V. vulnificus FLA077 and the above enrichment assays involved V. vulnificus FLA042, an Amp enrichment assay was performed with V. vulnificus FLA077 to bridge both V. vulnificus growth assays together. Oysters were infected for 7 h with V. vulnificus FLA077 and then treated overnight with 100 g/mL of Amp. The treatment of oysters with Amp significantly decreased the level of contamination in oysters by 2.6 logs (99%) (Fig. 3-21). This level of decrease was similar to the decrease observed with Amp treatment of V. vulnificus FLA042 infected oysters described above. Moreover, as seen with V. vulnificus FLA042, V. vulnificus FLA077 was not completely killed by Amp treatment. A level of approximately 102 CFU/g tissue contamination remained in the oys ters, again suggesting that a subset of the bacteria is not growing sufficien tly to be killed by Amp. At first analysis, it appeared that th e marker plasmid and Amp assays yielded discordant results for the replica tion or lack of replication of V. vulnificus in infected oysters. However, the marker plasmid a ssay showed significant growth up to 11 hours after initial infection. The l ack of growth occurred after that 11 hours of infection. The Amp assay showing 99.9% killing of supposedly replicating vibrios was performed after 7 hours of infection, within the period of growth observed with the marker plasmid assay. Therefore, there may not, in fact, be a discrepa ncy. If the Amp assay was performed after 11 hours of infection, we would hypothesize that the level of killi ng would significantly drop. The higher level of killing by Amp relative to that observed with phage can

PAGE 144

132 possibly be accounted for by the relative sensitivities of these agents to growth or statis of the target bacteria. Attempts at Stimulation of V. vulnificus Growth Coupled with Phage Treatment Since the marker plasmid and the Amp enrich ment assays indicated that a lack of growth of V. vulnificus may be the hindrance to complete elimination of V. vulnificus in oysters, attempts were undertaken to stimulate V. vulnificus growth in oysters to increase the level of killing by phages CK-2 and 3a in our experimental oyster infection model. V. vulnificus infections in our mouse model required the administration of iron prior to infection for a proper infection to ensue. We hypothesized that supplementation of the water with iron during th e phage treatment of oysters may stimulate the growth of V. vulnificus . However, the addition of iron at 5 M did not significan tly increase the level of growth of V. vulnificus in oysters (Fig. 3-22). Thus , we concluded that in the oyster environment iron is not a lim iting factor for the growth of V. vulnificus . Another method to attempt to increase the growth of V. vulnificus was an increase in water temperature during the phage treatmen t period from RT to 32.5C. The increase in water temperature increased the level of growth of V. vulnificus in oysters by 0.9 logs (Fig. 3-23). Moreover, the treatment of oysters with phages CK-2 and 3a at the elevated water temperature resulted in a larger decrease of V. vulnificus contamination compared to those at RT (1.8 logs vs. 1.1 logs)(Fig. 3-23). However, both phage-treated oyster groups (RT or 32.5C) were decontaminated to comparable levels of V. vulnificus , 4.5 0.57 log CFU/g tissue vs. 4.6 0.46 log CFU/g tissue, respectively. These results suggest that the larger decrease of bacteria in the oysters treated with phages CK-2 and 3a at 32.5C was caused by the phages killing the b acteria that were stimulated to grow by the increase in water temperature. These resu lts further suggested our hypothesis that the

PAGE 145

133 incomplete elimination of V. vulnificus contamination in experi mentally-infected oysters by phages 3a and CK-2 may be the lack of growth of the bacteria. After establishing an oys ter infection model for V. vulnificus that resulted in high levels of contamination comparable to leve ls seen in naturally infected oysters, the treatment of experimentally infected oyste rs with phages CK-2 and 3a resulted in significant decontamination. However, we felt that the levels of decontamination were not sufficient to be commercially acceptabl e. We hypothesized that the incomplete elimination of V. vulnificus with phage treatment in oysters was in part a result of the lack of growth of the bacteria. We were able to demonstrate with the use of a marker plasmid and an Amp enrichment assay that a subset of the population of V. vulnificus in oysters was not growing adequately, if at all. These observations suggest a possible reason for the inability of phage 3a and CK-2 to completely eliminate V. vulnificus in oysters. However, we have no data rega rding the state of growth of V. vulnificus in naturally infected oysters. Phage Treatment of Naturally Infected Market Oysters The main goal of this thesis project was to evaluate the potential of phage treatment for decontaminating naturally infected market oysters of V. vulnificus . The significant, albeit incomplete, decrease of V. vulnificus in experimentally infected oysters with phages CK-2 and 3a suggested that similar results could also be obtained with phage treatment of naturally infected oysters. Th is could eventually lead to a commercially viable strategy to render oysters safer for human consumption. Currently, as mentioned in the Introduction, a few post-ha rvest treatment (PHT) methods are utilized in the oyster industry to decrease the level of V. vulnificus in oysters. The use of hydrostatic-high pressure as high as 45,000 psi, mild elevati on of seawater temperature housing oysters,

PAGE 146

134 and the temporary quick freeze method have be en shown to successfully reduce the level of V. vulnificus contamination in harvested oysters . However, the use of such PHT methods can kill the oysters and lead to an a dditional significant expense that is usually passed on to consumers. Furthermore, oyster en thusiasts complain that the texture and/or taste of oysters are altere d by such PHT methods. If treatment of oysters with phages were satisfactory for the decontamination of oysters, it would have advantageous features compared to other PHT methods presently utilized in the oyster industry. For example, phages are relatively inexpensive to produce in high concentrations, which would likely lead to a lesser e xpense passed on to consumers. Due to the specificity of bacter iophages, bacteriophages that infect and kill only V. vulnificus would not alter the natural flora as seen with some of the other PHT methods. Moreover, bacteriophages are not known to have any adverse effects in oysters; hence, oysters remain alive during phage treatments, unlike other PHT methods. In contrast to the use of LB-N-Rif plates to enumerate Rif-resistant V. vulnificus strains in the experimentally infected oys ters, other enumeration methods had to be utilized for enumeration of V. vulnificus contamination in market oysters. The first quantification method evaluated was the use of selective media for enumeration of V. vulnificus . Both TCBS and VVM, or derivatives of VVM such as cellobiose-colistin (CC) and cellobiose-polymyxin Bcolistin (CPC), have been widely used to quantify the number of V. vulnificus in raw oysters (58,59). Since some selective media have low plating efficiencies for the bacteria of inte rest, the plating efficiencies of both VVM and TCBS media were compared and that of LB-N using V. vulnificus M06-24/O and MLT403. VVM had a higher plating efficacy co mpared to TCBS for both broth cultures

PAGE 147

135 of V. vulnificus and with either naturally infected or experimentally FLA042-infected oysters (Fig. 3-24). VVM had a 2.4 log highe r efficiency compared with TCBS for naturally infected or the experimentally in fected oysters. Furthermore, the plating efficacies of both VVM and LB-N were identical. VVM was selective against E. coli and was differential for V. cholerae because of the color differe nce produced by the bacteria. Therefore, VVM was used for several of th e phage cocktail treatment studies on market oysters. However, yellow colonies that were not V. vulnificus were observed on VVM plates during enumeration of V. vulnificus contamination in naturally infected oysters. Thus, careful consideration was taken when in terpreting data from the studies of phage treatment of naturally infected oysters using the selective medium VVM. A yellow colony that exhibited different colony mor phology to those typically associated with V. vulnificus was assumed not to be V. vulnificus , and, thus, was not taken into account during the enumeration process. Furthermore, we recognized that no t all yellow colonies that exhibited colony morphology similar to V. vulnificus were in fact that species of bacteria. However, we concl uded that similar proportions of V. vulnificus and nonV. vulnificus yellow colony-forming bacteria should be present in each individual oyster; thus, any significant difference observed be tween the phage-treated and untreated oyster groups most likely represented a decrease in the number of V. vulnificus . Erroneously counting nonV. vulnificus yellow colonies as V. vulnificus would have the effect of artifactually decreasing the calculated level of decontamination levels with phage. The selection of phages for our experime nts was made by the comparison of the host range and level of infectivity of each pha ge compiled during the use of the soft agar overlay and the microtiter phage typing me thods. Five phages (153A-5, 153A-8, 154A-8,

PAGE 148

136 3a, and 4a) were selected for the first expe riment due to their broad host ranges with a level of infectivity of 5. Phage cocktail treatment of naturally infected oysters with these phages each at 1 x 108 PFU/mL resulted in a signif icant decrease of 1.5 logs of V. vulnificus contamination in oysters (Fig. 3-25). This level of decrease was similar to the level of decontamination observed in expe rimentally infected oysters treated with phages CK-2 and 3a. In an attempt to increase the level of decontamination of V. vulnificus in oysters, three more phages exhibiting broad host range s were added to the cocktail of phages for the second experiment. Phagecocktail treatment of naturall y infected oysters with the larger pool of phages (phages 108A-9, 110A-7, 153A-5, 153A-7, 153A-8, 154A-8, 3a and 4a) but at a lower con centration of each at 1 x 107 PFU/mL resulted in a similar significant decrease of 1.4 logs of V. vulnificus contamination in oysters (Fig. 3-26). Furthermore, in the second experiment we de monstrated that the cocktail of phages was only able to decrease the level of contam ination of these oysters to levels of contamination enumerated from oysters that were sampled immediately upon their arrival into the laboratory (Fig. 326). The nonphage-treated oyste rs experienced a rise in V. vulnificus CFU during overnight incubation in s eawater. The fact that the phage cocktail treatments were able to bring the level of contamination in oysters to the same level of contamination they were initially at prior to any incubati on period in seawater suggested that the phages were only able to kill the bacteria that grew ov ernight in the oysters incubated in seawater. These findi ngs reflect the results observed with the experimentally infected oysters at RT or 32.5C in which a higher concentration of V. vulnificus was observed with the elevated temper ature; however, the treatment with

PAGE 149

137 phages CK-2 and 3a resulted in decreasing th e levels of contamination back down to where they were before the incubation at elevat ed temperature. It appeared that only the “extra” bacteria that were stimulated to grow by the elevated water temperature were the “extra” bacteria that were killed by phages CK -2 and 3a. Thus, the incomplete killing of V. vulnificus with phage-cocktail treatment may be due to the lack of growth of the bacteria, as demonstrated with experimentally infected oysters treat ed with phages CK-2 and 3a at RT or 32.5C. Because VVM plates were not absolutely selective and differential for V. vulnificus , a DNA-hybridization probe, VVAP, (60) wa s used to evaluate the level of decontamination of V. vulnificus with a cocktail of phages on naturally infected oysters. Oysters were treated with a cocktail of a ll of the phages in ou r collection, at 1 x 107 PFU/mL each, with the exception of CB-1, whic h was not included. Similar to the two previous phage-cocktail treatment experiment s, the phage-cocktail treatment significantly decreased the level of contamination of V. vulnificus in oysters by 1.1 logs (Fig. 3-27). The use of either the VVM plates or the VVAP DNA-hybridization technique for enumeration of V. vulnificus contamination in oysters demonstrated a decrease in V. vulnificus contamination of approximately 1 log. Although the use of VVM plates for enumeration of V. vulnificus was less laborious compared to the use of the VVAP probe, the specificity of the VVAP probe for V. vulnificus significantly decreased the probability of an error in counting bact erial colonies that were nonV. vulnificus during such experiments. Thus, future experi ments involving the enumeration of V. vulnificus contamination in market oysters should be enumerated with the use of the VVAP DNAhybridization probe. If needed, the VVM plat es can be used for estimation or as a

PAGE 150

138 secondary method for the determination of the level of V. vulnificus contamination in market oysters. Together, these results de monstrated a phage-induced decontamination of naturally infected oysters ranging from 1.1 logs to 1.5 logs. However, we do not consider the remaining levels of V. vulnificus contamination in oysters treated with a cocktail of phages, approximately 103 CFU/g tissue, to be safe for human consumption, or at least an acceptable commercial application. In conclusion, the treatment of experiment ally infected oysters with phages 3a and CK-2 and naturally infected oysters with a cocktail of phages result ed in a significant decrease in the level of V. vulnificus decontamination. However, the number of V. vulnificus that remained in the oysters after phage treatment, approximately 103 CFU/g tissue for phage-treated market oysters, was still an unacceptable level for human consumption. The acceptable level of V. vulnificus contamination in oysters set by the FDA for human consumption is less than 3 MP N/g tissue. However, this method is not equivalent to the quantitative plating method used in our experiments. Using both the marker plasmid pGTR902 and the Amp enrich ment assays, we demonstrated that a possible reason for the inability of complete elimination of the bacteria by phages was a lack of growth of the bacteria in the oysters. For phage infection a nd lysis of a bacterial cell to occur, bacteriophages necessitate the metabolic machinery of the host for replication. If an infected bacterium is in stationary phase, then the metabolic machinery is not readily available for phage replication and eventual lysis of the bacterium. This hypothesis was further supported with the result s of the experimentally FLA042-infected oysters treated with phages CK-2 and 3a at either RT or 32.5C. Phages CK-2 and 3a seem to have killed the bacteria that were stimulated to replicate by the increase in water

PAGE 151

139 temperature. This killing of the stimulat ed bacteria by phages CK-2 and 3a at the elevated temperature resulted in similar final levels of V. vulnificus as in the phagetreated oysters at RT, suggesting that the state of growth of th e bacteria is a factor for the incomplete killing of V. vulnificus in experimentally inf ected oysters. A similar phenomenon was observed with the naturally in fected oysters treated with a cocktail of phages. The treatment of naturally infected oysters with a pool of phages resulted in a significant decrease of contamination compar ed with control oyste rs not treated with phage but experiencing the same incubati on conditions; however, the final level of contamination of the phage-treated oysters was similar to the level of contamination seen in oysters that were never inc ubated overnight in seawater. This result suggested that the phages were only able to infect and lyse the bacteria that we re growing during the overnight incubation period in seawater. All in all, the findings in the experimentally and naturally infected oyster experiments consis tently suggested that a subset of the V. vulnificus population in oysters ar e in stationary phase. Our findings suggesting a lack of growth of V. vulnificus in oysters are in agreement with a report by Kolter et al. suggest ing that bacteria in a marine environment are generally in stationary phase due to the lack of nutrients in the environment (64). Marine bacteria are believed to go through wh at is known as a “feast or famine lifestyle” in which the bacteria are usually in stati onary phase with short periods of exponential growth when nutrients come available for ut ilization by the bacteria . During stationary phase, bacteria are observed to induce differe nt physiochemical properties in comparison to properties observed during exponential grow th. During stationary phase, bacterial cells decrease in size with the condensation of the cytoplasm. The nucleoid condenses as

PAGE 152

140 well, causing a decrease in gene expressi on. While the periplasmic volume increases during stationary phase, the cell wall decrea ses in fluidity and permeability. These physiochemical changes are known to induce significant resistance to environmental stresses such as osmotic or heat shoc k, including cell inf ection and lysis by bacteriophages. Thus, the physiological state of bacteria is important for phage infection and replication. However, some phage genomes, such as phage T4, are known to produce stable infective cente rs in bacterial cells in stationary phase, termed a “hibernation” period (62). When the bacteria l cells exit stationary phase to exponential phase, due to the availability of nutrients, the phages be gin the replication process eventually causing lysis of the bacteria. Furthermore, some P. aeruginosaspecific phages infect and lyse P. aeruginosa in stationary phase even after the bacteria were starved for five years (65). E. coli -phage T7M infects and lyses E. coli that is in stationary phase. The phages in our collec tion were not examined for their ability to infect and/or lyse V. vulnificus in stationary phase. However, such a study could elucidate the usability of th ese phages during phage treatments of either naturally or experimentally infected oysters, and coul d even explain the lack of complete decontamination in oysters with phage treatmen ts. Furthermore, gram-negative bacteria in stationary phase are also believed to be metabolically active, only at a significantly lower rate, expressing only vital genes for the survivability of the bacteria. This last finding may explain the greater decontamina tion levels obtained with Amp treatment compared to phage CK-2 and 3a treatment in our Amp enrichment experiments (Fig. 320).

PAGE 153

141 Attempts to stimulate growth of V. vulnificus in experimentally infected oysters during phage treatment with phages CK-2 and 3a were unsuccessful at stimulating all the bacteria to grow and, thus, increasing the proba bility of infection a nd lysis by the phages. Another supplementation to stimulate growth of V. vulnificus in oysters could be MgCl2 and KCl, which are present in the VVM sele ctive medium. These compounds have been reported to be stimulating growth factors for V. vulnificus (58,59). The addition of the divalent cation Mg+2 could also increase the abilit y of certain phages to infect V. vulnificus , as described with phage CB-1. Seawater obtained from the Whitney Laboratory was diluted appr oximately half with ddH2O, halving the concentration of these cations . This could have decreased the growth rate of V. vulnificus in oysters and also decreased the ability of our pha ges to optimally infect and lysis V. vulnificus in oysters. Another possible approach to increase the level of decont amination in oysters that was not examined during this project is an in crease in the concentr ation of phages used during the treatment process. In general, we treated contaminated oysters with concentrations of phages ranging from 1 x 107 PFU/mL to 1 x 108 PFU/mL. However, higher concentrations of pha ges in the range of 1 x 109 PFU/mL to 1 x 1010 PFU/mL may result in bacterial lysis from without. Lysi s from without results when a threshold of phage adsorptions occur to a single bacteria l cell, thus, overwhelming the bacterial cell causing non-proliferative lysis of the bact erium (62). Such a strategy may afford a technique for killing the bacteria that are in stationary phase, since the replication process of phages is not followed during this type of infection. However, the addition of such

PAGE 154

142 large quantities of phages may be an adde d expense burden if applied in a commercial setting. Altogether, the results presented in this th esis demonstrate that phage treatment of V. vulnificus contamination in both naturally and e xperimentally infected oysters results in decreases of contamination ranging from 1 log to 2 logs, depending on the particular experiment. However, the level of contamin ation after phage treatment was still too high for consideration as a commercial applica tion. We demonstrated that a possible explanation for the incomplete elimination of V. vulnificus contamination in oysters may be a lack of growth of V. vulnificus , resulting in the inability of the phages to infect and kill the bacteria. Other potential strategies in conjunction with phage treatment were presented above to clarify the potential lack of growth of V. vulnificus and possibly render V. vulnificus susceptible to phage infection and lysis, resulting in further decreases in contamination. However, due to a possible lack of growth of V. vulnificus in oysters, it may prove impossible to improve upon the level of decontamination of V. vulnificus in oysters with the use of therapeutic phages.

PAGE 155

143 LIST OF REFERENCES 1. Twort, F. W. 1915. An investigation on the nature of ultra-microscopic viruses. The Lancet 186 :1241-1243. 2. d'Herelle, F. 1917. Sur un microbe invisible antagoniste de bacilles dysenteriques. C. R. Acad. Sci. (Paris) 165 :373-375. 3. Chibani-Chennoufi, S., Bruttin, A., Dillmann, M. L., and Brussow, H. 2004. Phage-host interaction: an ecological perspective. J. Bacteriol. 186 :3677-3686. 4. Bergh, O., Borsheim, K. Y., Bratba k, G., and Heldal, M. 1989. High abundance of viruses found in aquatic environments. Nature 340 :467-468. 5. Ackermann, H. W. 2001. Frequency of morphological phage descriptions in the year 2000. Archives of Virology 146 :843-857. 6. d'Herelle,F. 1926. The bacteriophage and its behavior . Williams and Wilkins. Baltimore. 7. d'Herelle F, Malone RH, and Lahi ri M 1930. Studies on asiatic cholera. Indian Med. Res. Mem. 14 :1-161. 8. Summers, W. C. 1993. C holera and plague in India the Bacteriophage Inquiry of 1927-1936. Journal of the History of Medicine and Allied Sciences 48 :275301. 9. d'Herelle, F., and LeLouet, G. 1921. Sur la vaccination antibarbonique par virus attnu. C. R. Soc. Biol. (Paris) 85: 1011-1013. 10. Summers, W. C. 2001. Bacteriophage therapy. Annual Review of Microbiology 55 :437-451. 11. Summers,W.C. 2001. Felix d'Herelle and the origins of molecular biology . Yale University Press. New Haven and London. 12. Straub, M. E., and Applebaum, M. 1933. Studies on commercial bacteriophage products. JAMA 100 :110-113. 13. Kutter, E., and Sulakvelidze, A. 2004. Bacteriophages: Biology and applications. CRC Press. Boca Raton, FL.

PAGE 156

144 14. Smith, H. W., and Huggins, M. B. 1983. Effectiveness of phages in treating experimental Escherichia coli diarrhoea in calves, piglets and lambs. J. Gen. Microbiol. 129 :2659-2675. 15. Smith, H. W., and Huggins, M. B. 1982. Successful treatment of experimental Escherichia coli infections in mice using phage : its general superiority over antibiotics. J. Gen. Microbiol. 128 :307-318. 16. Smith, H. W., Huggins, M. B., and Sh aw, K. M. 1987. Factors influencing the survival and multiplication of bacteriophages in calves and in their environment. J. Gen. Microbiol. 133 :1127-1135. 17. Smith, H. W., Huggins, M. B., and Shaw, K. M. 1987. The control of experimental Escherichia coli diarrhoea in calves by means of bacteriophages. J. Gen. Microbiol. 133 :1111-1126. 18. Chibani-Chennoufi, S., Sidoti, J., Bruttin, A., Kutter, E., Sarker, S., and Brussow, H. 2004. In vitro and in vivo bacteriolytic activities of Escherichia coli phages: implications for phage therapy. Antimicrob. Agents Chemother. 48 :2558-2569. 19. Huff, W. E., Huff, G. R., Rath, N. C., Balog, J. M., and Donoghue, A. M. 2004. Therapeutic efficacy of bacteriophage and Baytril (enrofloxacin) individually and in combination to treat co libacillosis in broilers. Poult. Sci. 83 :1944-1947. 20. Biswas, B., Adhya, S., Washart, P., Paul, B., Trostel, A. N., Powell, B., Carlton, R., and Merril, C. R. 2002. Bacteriophage therapy rescues mice bacteremic from a clinical isolate of vancomycin-resistant Enterococcus faecium . Infect. Immun. 70 :204-210. 21. Matsuzaki, S., Yasuda, M., Nishikawa, H., Kuroda, M., Ujihara, T., Shuin, T., Shen, Y., Jin, Z., Fujimoto, S., Nasimuzzam an, M. D., Wakiguchi, H., Sugihara, S., Sugiura, T., Koda, S., Muraoka, A., and Imai, S. 2003. Experimental protection of mice against lethal Staphylococcus aureus infection by novel bacteriophage phi MR11. J. Infect. Dis. 187 :613-624. 22. Soothill, J. S. 1992. Treatment of experimental infections of mice with bacteriophages. J. Med. Microbiol. 37 :258-261. 23. Paisano, A. F., Spira, B., Cai, S., and Bombana, A. C. 2004. In vitro antimicrobial effect of bacteriopha ges on human dentin infected with Enterococcus faecalis ATCC 29212. Oral Microbiol. Immunol. 19 :327-330. 24. Ramesh, V., Fralick, J. A., and Rolfe, R. D. 1999. Prevention of Clostridium difficile -induced ileocecitis with bacteriophage. Anaerobe 5 :69-78. 25. Hagens, S., Habel, A., von Ahsen, U., von Gabain, A., and Blasi, U. 2004. Therapy of experimental pseudomona s infections with a nonreplicating genetically modified phage. Antimicrob. Agents Chemother. 48 :3817-3822.

PAGE 157

145 26. Elkin, S., and Geddes, D. 2003. Pseudo monal infection in cystic fibrosis: the battle continues. Expert. Rev. Anti. Infect. Ther. 1 :609-618. 27. Broxmeyer, L., Sosnowska, D., Milt ner, E., Chacon, O., Wagner, D., McGarvey, J., Barletta, R. G., and Bermudez, L. E. 2002. Killing of Mycobacterium avium and Mycobacterium tuberculosis by a mycobacteriophage delivered by a nonvirulent mycobacterium: a model for pha ge therapy of intr acellular bacterial pathogens. J. Infect. Dis. 186 :1155-1160. 28. Nelson, D., Loomis, L., and Fischetti, V. A. 2001. Prevention and elimination of upper respiratory colonization of mi ce by group A streptoc occi by using a bacteriophage lytic enzyme. Proc. Natl. Acad. Sci. U. S. A 98 :4107-4112. 29. Cheng, Q., Nelson, D., Zhu, S., and Fi schetti, V. A. 2005. Removal of group B streptococci colonizing the vagina and oropharynx of mice with a bacteriophage lytic enzyme. Antimicrob. Agents Chemother. 49 :111-117. 30. Yoong, P., Schuch, R., Nelson, D., and Fisc hetti, V. A. 2004. Identification of a broadly active phage lytic enzyme with le thal activity against antibiotic-resistant Enterococcus faecalis and Enterococcus faecium . J. Bacteriol. 186 :4808-4812. 31. Loeffler, J. M., Djurkovic, S., and Fisc hetti, V. A. 2003. Phage lytic enzyme Cpl1 as a novel antimicrobial for pneumococcal bacteremia. Infect. Immun. 71 :61996204. 32. Fischetti, V. A. 2003. Novel method to control pathogenic bacteria on human mucous membranes. Ann. N. Y. Acad. Sci. 987 :207-214. 33. Loeffler, J. M., and Fischetti, V. A. 2003. Synergistic lethal effect of a combination of phage lytic enzymes w ith different activi ties on penicillinsensitive and -resistant Strept ococcus pneumoniae strains. Antimicrob. Agents Chemother. 47 :375-377. 34. Schuch, R., Nelson, D., and Fischetti, V. A. 2002. A bacteriolytic agent that detects and kills Bacillus anthracis. Nature 418 :884-889. 35. Loeffler, J. M., Nelson, D., and Fi schetti, V. A. 2001. Rapid killing of Streptococcus pneumoniae with a ba cteriophage cell wall hydrolase. Science 294 :2170-2172. 36. Fischetti, V. A. 2001. Phage antibacterials make a comeback. Nat. Biotechnol. 19 :734-735. 37. O'Flynn, G., Ross, R. P., Fitzgerald, G. F., and Coffey, A. 2004. Evaluation of a cocktail of three bacteriophages for bi ocontrol of Escherichia coli O157:H7. Appl. Environ. Microbiol. 70 :3417-3424.

PAGE 158

146 38. Leverentz, B., Conway, W. S., Alavidze, Z., Janisiewicz, W. J., Fuchs, Y., Camp, M. J., Chighladze, E., and Sulakvelidze, A. 2001. Examination of bacteriophage as a biocontrol method for salmonella on fresh-cut fruit: a model study. J. Food Prot. 64 :1116-1121. 39. Leverentz, B., Conway, W. S., Camp, M. J., Janisiewicz, W. J., Abuladze, T., Yang, M., Saftner, R., and Sulakvelidze, A. 2003. Biocontrol of Listeria monocytogenes on fresh-cut produce by trea tment with lytic bacteriophages and a bacteriocin. Appl. Environ. Microbiol. 69 :4519-4526. 40. Atterbury, R. J., Connerton, P. L., Dodd, C. E., Rees, C. E., and Connerton, I. F. 2003. Application of host-specific bacter iophages to the surface of chicken skin leads to a reduction in recovery of Campylobacter jejuni. Appl. Environ. Microbiol. 69 :6302-6306. 41. Nakai, T., Sugimoto, R., Park, K. H ., Matsuoka, S., Mori, K., Nishioka, T., and Maruyama, K. 1999. Protective effects of bacteriophage on experimental Lactococcus garviea e infection in yellowtail. Dis. Aquat. Organ. 37 :33-41. 42. Park, S. C., Shimamura, I., Fukunaga , M., Mori, K. I., and Nakai, T. 2000. Isolation of bacteriophages specific to a fish pathogen, Pseudomonas plecoglossicid a, as a candidate for disease control. Appl. Environ. Microbiol. 66 :1416-1422. 43. Hollis, D. G., Weaver, R. E., Baker, C. N., and Thornsberr y, C. 1976. Halophilic Vibrio species isolated from blood cultures. J. Clin. Microbiol. 3 :425-431. 44. Morris, J. G., Jr., and Black, R. E. 1985. Cholera and other vibrioses in the United States. N. Engl. J. Med. 312 :343-350. 45. Blake, P. A., Merson, M. H., Weaver, R. E., Hollis, D. G., and Heublein, P. C. 1979. Disease caused by a marine Vi brio. Clinical characteristics and epidemiology. N. Engl. J. Med. 300 :1-5. 46. Vollberg, C. M., and Herrera, J. L. 1997. Vibrio vulnificus infection: an important cause of septicemia in patients with cirrhosis. South. Med. J. 90 :1040-1042. 47. Strom, M. S., and Paranjpye, R. N. 2000. Epidemiology and pathogenesis of Vibrio vulnificus . Microbes Infect. 2 :177-188. 48. Kaspar, C. W., and Tamplin, M. L. 1993. Effects of temperature and salinity on the survival of Vibrio vulnificus in seawater and shellfish. Appl. Environ. Microbiol. 59 :2425-2429. 49. Motes, M. L., DePaola, A., Cook, D. W., Veazey, J. E., Hunsucker, J. C., Garthright, W. E., Blodgett, R. J., and Chirtel, S. J. 1998. Influence of water temperature and salinity on Vibrio vulnificus in Northern Gulf and Atlantic Coast oysters ( Crassostrea virginica ). Appl. Environ. Microbiol. 64 :1459-1465.

PAGE 159

147 50. Kaysner, C. A., Abeyta, C., Jr., Wekell, M. M., DePaola, A., Jr., Stott, R. F., and Leitch, J. M. 1987. Virulent strains of Vibrio vulnificus isolated from estuaries of the United States West Coast. Appl. Environ. Microbiol. 53 :1349-1351. 51. O'Neill, K. R., Jones, S. H., and Grimes, D. J. 1992. Seasonal incidence of Vibrio vulnificus in the Great Bay estuar y of New Hampshire and Maine. Appl. Environ. Microbiol. 58 :3257-3262. 52. Wright, A. C., Hill, R. T., Johnson, J. A., Roghman, M. C., Colwell, R. R., and Morris, J. G. J. 1996. Distribution of Vibrio vulnificus in the Chesapeake Bay. Appl. Environ. Microbiol. 62 :717-724. 53. Hlady, W. G., Mullen, R. C., and Hopkin, R. S. 1993. Vibrio vulnificus from raw oysters. Leading cause of reported deaths from foodborne illness in Florida. J. Fla. Med. Assoc. 80 :536-538. 54. Hlady, W. G., and Klontz, K. C. 1996. The epidemiology of Vibrio infections in Florida, 1981-1993. J. Infect. Dis. 173 :1176-1183. 55. Food and Nutrition Boar d-Institute of Medicine 1991. Seafood Safety/Committee on Evaluation of the Safety of Fishery Products . National Academy Press. Washington, D.C. 56. Klontz, K. C., Lieb, S., Schreiber, M., Janowski, H. T., Baldy, L. M., and Gunn, R. A. 1988. Syndromes of Vibrio vulnificus infections. Clinical and epidemiologic features in Florida cases, 1981-1987. Ann. Intern. Med. 109 :318323. 57. Shapiro, R. L., Altekruse, S., Hu twagner, L., Bishop, R., Hammond, R., Wilson, S., Ray, B., Thompson, S., Tauxe, R. V., a nd Griffin, P. M. 1998. The role of Gulf Coast oysters harvested in warmer months in Vibrio vulnificus infections in the United States, 1988-1996. Vibrio Working Group. J. Infect. Dis. 178 :752759. 58. Kumamoto, K. S., and Vukich, D. J. 1998. Clinical infections of Vibrio vulnificus : a case report and review of the literature. J. Emerg. Med. 16 :61-66. 59. Muench, K. H. 1989. Hemochromatosis and infection: alcohol and iron, oysters and sepsis. Am. J. Med. 87 :40N-43N. 60. Hlady, W. G., Mullen, R. C., and Hopkin, R. S. 1993. Vibrio vulnificus infections associated with raw oyster consumption-Florida, 1981-1992. Arch. Dermatol. 129 :957-958. 61. Mascola, L., Tormey, M., Dassey, D., K ilman, L., Harvey, S., Medina, A., Tilzer, A., and Waterman, S. 1996. Vibrio vulnificus infections associated with eating raw oysters-Los Angeles, 1996. Morb. Mortal. Wkly. Rep. 45 :621-624.

PAGE 160

148 62. Halow, K. D., Harner, R. C., and Font enelle, L. J. 1996. Primary skin infections secondary to Vibrio vulnificus : the role of operative intervention. J. Am. Coll. Surg. 183 :329-334. 63. Linkous, D. A., and Oliver, J. D. 1999. Pathogenesis of Vibrio vulnificus . FEMS Microbiol. Lett. 174 :207-214. 64. Holly, M. 2003. California's ban on Gu lf Coast oysters costs Louisiana up to $20M so far in 2003. New Orleans CityBusiness . 65. Burdeau, C. 2005. Oyster industry ra cked by proposal to sue California. The Sacramento Union. 66. Keithly W.R., Jr., and DIOP, H. 2001. The Demand for Eastern Oysters, Crassostrea virginica, from the Gulf of Mexico in the Presence of Vibrio vulnificus. Marine Fisheries Review 63 :47-53. 67. Muth, M. K., Karns, S. A., Anderson, D. W., and Murray, B. C. 2002. Effects of post-harvest treatment requirements on the markets for oysters. Agricultural and Resource Economics Review . 68. Cerveny, K. E., DePaola, A., Duckwo rth, D. H., and Gulig, P. A. 2002. Phage therapy of local and systemic disease caused by Vibrio vulnificus in iron dextrantreated mice. Infect. Immun. 70 :6251-6262. 69. Cerda-Cuellar, M., Jofre, J., and Blan ch, A. R. 2000. A selective medium and a specific probe for detection of Vibrio vulnificus. Appl. Environ. Microbiol. 66 :855-859. 70. Cerda-Cuellar, M., Permin, L., Larsen, J. L., and Blanch, A. R. 2001. Comparison of selective media for the detection of Vibrio vulnificus in environmental samples. J. Appl. Microbiol. 91 :322-327. 71. DePaola, A., Motes, M. L., Chan, A. M., and Suttle, C. A. 1998. Phages infecting Vibrio vulnificus are abundant and diverse in oysters ( Crassostrea virginica ) collected from the Gulf of Mexico. Appl. Environ. Microbiol. 64 :346-351. 72. DePaola, A., McLeroy, S., and McManus, G. 1997. Distribution of Vibrio vulnificus phage in oyster tissues and other estuarine habitats. Appl. Environ. Microbiol. 63 :2464-2467. 73. Wright, A. C., Miceli, G. A., Landry, W. L., Christy, J. B., Watkins, W. D., and Morris, J. G., Jr. 1993. Rapid identification of Vibrio vulnificus on nonselective media with an alkaline phosphatase -labeled oligonucleotide probe. Appl. Environ. Microbiol. 59 :541-546. 74. Duckworth, D. H., and Gulig, P. A. 2001. Bacteriophages: Potential treatment for bacterial infections. BioDrugs 16 :57-62.

PAGE 161

149 75. Tamplin, M. L., and Capers, G. M. 1992. Persistence of Vibrio vulnificus in tissues of Gulf Coast oysters, Crassostrea virginica , exposed to seawater disinfected with UV light. Appl. Environ. Microbiol. 58 :1506-1510. 76. Kolter, R., Siegele, D. A., and Torm o, A. 1993. The stationary phase of the bacterial life cycle. Annu. Rev. Microbiol. 47 :855-874. 77. Schrader, H. S., Schrader, J. O., Walker , J. J., Wolf, T. A., Nickerson, K. W., and Kokjohn, T. A. 1997. Bacteriophage in fection and multiplication occur in Pseudomonas aeruginosa starved for 5 years. Can. J. Microbiol. 43 :1157-1163.

PAGE 162

150 BIOGRAPHICAL SKETCH Julio L. Martn was born in Havana, Cuba , immigrated to the United States in 1980 and resided in Miami. He received his B achelor of Science degree in 1999 from the University of Miami with a major in bi ology and minor in chemistry with honors.